UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Cabinet responsibility, the separation-of-powers and the makers and breakers of cabinets in Japanese… Steven, Robert P. G. 1973

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Notice for Google Chrome users:
If you are having trouble viewing or searching the PDF with Google Chrome, please download it here instead.

Item Metadata


831-UBC_1974_A1 S84_6.pdf [ 16.23MB ]
JSON: 831-1.0093520.json
JSON-LD: 831-1.0093520-ld.json
RDF/XML (Pretty): 831-1.0093520-rdf.xml
RDF/JSON: 831-1.0093520-rdf.json
Turtle: 831-1.0093520-turtle.txt
N-Triples: 831-1.0093520-rdf-ntriples.txt
Original Record: 831-1.0093520-source.json
Full Text

Full Text

c I • CABINET RESPONSIBILITY, THE SEPARATION-OF-POWERS AND THE MAKERS AND BREAKERS OF CABINETS IN JAPANESE POLITICS, 1890-1940 by ROBERT P.G. STEVEN B.A. (HONS.), OXFORD UNIVERSITY, 1970 A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in the Department of P o l i t i c a l Science We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA December, 1973 In presenting th is thesis in pa r t i a l fu l f i lment of the requirements for an advanced degree at the Un ivers i ty of B r i t i s h Columbia, I agree that the L ibrary shal l make it f ree ly ava i l ab le for reference and study. I fur ther agree that permission for extensive copying of th is thes is for scho la r ly purposes may be granted by the Head of my Department or by his representat ives . It is understood that copying or pub l i ca t ion of th is thes is for f inanc ia l gain sha l l not be allowed without my wri t ten permission. Department of P o l i t i c a l Science The Univers i ty of B r i t i s h Columbia Vancouver 8, Canada Date 21 January 197^. i i ABSTRACT According to parliamentary theory, an executive that i s made and unmade by the Lower House of the legislature alone i s responsible to that House. But an executive whose existence i s not solely dependent on the legislature i s not responsible to the legislature. In such systems, usually the main branches of government have specific functions, possess limited rights of veto over one another, and have independent existences. They are known as separation-of-powers systems. The purpose of the thesis i s to discover whether the prewar Japan-ese polity approximated more closely to a parliamentary system or a separation-of-powers system. Its method i s to identify a l l the p o l i t i c a l institutions which made and unmade the executive in 1890-1940. When institutions are not easily identifiable, for example, when a cabinet resigned because of public rioting, the influences responsible for Cabi-net changes are translated into politico-institutional forces. Because there was always a struggle over the selection of Prime Ministers and then over Cabinet seats, the selection of Prime Ministers is examined separately from the formation of cabinets. A classification of the reasons for Cabinet composition and i t s rise and f a l l i s used to determine whether institutional relationships are better understood in terms of parliamentary or separation-of-powers theory. The results of the investigation reveal that: i ) Each of the prewar p o l i t i c a l institutions had a separate identifiable i i i function and tried to have the executive pursue the policies i t desired i n matters related to i t s function. i i ) Each institution possessed a limited veto power over each of the others and used this power to ensure that the Cabinet included repre-sentatives from i t . The Cabinet regularly consisted of representatives from most institutions: the two Houses of the Diet, the Army, the Wavy, and the C i v i l Service. i i i ) Each institution had an existence independent of each of the others, and only the Cabinet never had an independent power base. Usually at least three institutions had to support a new Prime Minister before he could assume office, and usually two had to conspire to force his resignation. Because only rarely could any single institution on i t s own raise or pull down an entire ministry, the existence of the Cabinet was separate from each individual institution and the Cabinet was not responsible to any. Separation-of-powers theory alone emphasises the lack of the executive's t o t a l dependence on the legislature, or on any other institution for that matter. The need for at least three institutions to raise and two to p u l l down a ministry indicates that the Cabinet never had a completely inde-pendent existence. Not having i t s own separate power base, i t was the joint creation of other institutions. Though i t s existence was separate from each individual institution, i t s rise and f a l l was not independent of combinations of other institutions. The prewar Japanese polity, iv however, bore only a slight similarity to a parliamentary system, in which the executive i s entirely dependent on the Lower House of the legislature. Because only very rarely could the Lower House of the legislature on i t s own p u l l down an entire ministry, only occasionally were parliamentary type forces present, and the polity functioned regularly as a separation-of-powers system. TABLE OF CONTENTS LIST OF TABLES v i INTRODUCTION 1 CHAPTER 1. CABINET RESPONSIBILITY IN THE MEIJI CONSTITUTION 23 CHAPTER 2. AN EMERGING SEPARATION-OF-POWERS, 1890-1900 48 CHAPTER 3. A FULL-FLEDGED SEPARATION-OF-POWERS, 1900-1918 84 CHAPTER h. PARTY PREMIERS AND THE PERSISTENCE OF THE lU3 SEPARATION-OF-POWERS, 1918-1932 CHAPTER 5- LEARNING TO LIVE WITH THE SEPARATION-OF-POWERS: 191 NATIONAL UNITY CABINETS, 1932-1940 CONCLUSION 251 FOOT-NOTES 260 BIBLIOGRAPHY 286 APPENDICES 297 GLOSSARY 363 VI LIST OF TABLES Table 1. Proportion of Cabinet Members from the two main han, 80 1890-1918. Table 2. Major and Minor Roles in Cabinet-Making and -Breaking, 112 1900-1918. Table 3' Major and Minor Roles in Cabinet-Making and -Breaking, 177 1918-1932. Table 4. Major and Minor Roles in Cabinet-Making and -Breaking, 228 1932-19^0. Table 5. Institutional background of Lower House Members, 246 1928-1940. 1 INTRODUCTION I t has long been recognised t h a t r e l a t i o n s h i p s among prewar Japan-ese p o l i t i c a l i n s t i t u t i o n s were extremely complex, but both p r a c t i o n e r s of and commentators on p o l i t i c s under the M e i j i C o n s t i t u t i o n have, on the whole, been content t o leave i t at t h a t . The d r a f t e r s of the Con-s t i t u t i o n had only a vague id e a of how they hoped t o coordinate the a c t i v i t i e s of the Emperor, the P r i v y C o u n c i l , the Cabinet, the C i v i l S e r v i c e , the Army, the Navy, and the two Houses of the D i e t . Making the Emperor t h e o r e t i c a l l y sovereign, but at the same time denying him the r i g h t t o i n t e r f e r e i n p o l i t i c s , i n d i c a t e s t h a t a c t u a l r e l a t i o n s h i p s among i n s t i t u t i o n s was something t o which they gave very l i t t l e thought. Commentators on prewar Japanese p o l i t i c s have e i t h e r been too preoccu-p i e d w i t h r e l a t i n g the causes of the Second World War t o the M e i j i C o n s t i t u t i o n a l order, or too keen t o regard p o l i t i c s i n Japan as unique and c o o r d i n a t i o n as the r e s u l t of c u l t u r a l phenomena l i k e p ersonal net-works, t o give the problem serious c o n s i d e r a t i o n . No one has yet attemp-t e d t o provide a comprehensive h i s t o r i c a l a n a l y s i s of the a c t u a l ways i n which the v a r i o u s i n s t i t u t i o n s were r e l a t e d t o one another, or t o look at these r e l a t i o n s h i p s i n the l i g h t of c o n s t i t u t i o n a l theory. The prac-t i t i o n e r s of p o l i t i c s , w i t h a few e x c e p t i o n s , ! have been r e l u c t a n t t o t h e o r i s e , and too i n v o l v e d i n day t o day matters t o be conscious of the p o l i t i c o - i n s t i t u t i o n a l f o r c e s c o n s t r a i n i n g them, w i t h the r e s u l t t h a t t h e i r a c t i o n s have remained more r e v e a l i n g than t h e i r words. 2 Recent scholarship, however, particularly outside Japan, has become increasingly aware of the problem, and fewwwriters f a i l to make some reference to i t . Mayer-Oakes, in an introduction to his trans-lat ion of part of the Saionji-Harada Memoirs, writes :^ The Meiji Constitution . . . had left to the definition of practice the precise relationships to obtain among the Cabinet, Privy Council, and the agencies of the Supreme Command, . . . but each could assert a wide autonomy i f not independence of the others. In the unity provided by the collective counsel of the genro, the fundamentally schismatic character of Japanese govern-mental institutions was disguised, and the system as a whole was given apparent and temporary coordination. In a work on Minobe Tatsukichi, Mi l le r writes;3 The cabinet in Japan . . . had always been the victim of confl ict-ing pressures from the genro, the mili tary command, the imperial household offices, the privy council, the house of peers, and the p o l i t i c a l parties within the house of representatives, each of which was capable of obstructing policy. A Japanese scholar, Ito Takashi, writes:^ Each of the governmental organs began to insis t on i t s own power - the foreign ministry on i t s control over diplomacy, the ministry of justice on i t s control of the legal system, the army on i t s prerogative of supreme command, and the Privy Council and the House of Peers on their unique positions - so much so that p o l i t i -cal agreement became extremely d i f f i cu l t to maintain. While many have come to recognise the d i f f i cu l ty , only Vere Redman, who wrote in 1932, attempted to look at i t in the l ight of constitutional theory and to say something about what was actually happening. He began by asking the question, to whom i s the Cabinet responsible? His answer was :5 Ministers are obviously responsibletto whoever can appoint or dis-miss them; and in Japan, nominally of course, this appointing and dismissing power i s held by the Emperor. But according to the 3 theory t h a t the Emperor can do no wrong, some person or body must assume r e s p o n s i b i l i t y f o r e i t h e r of these I m p e r i a l a c t s . . . . ( M ) i n i s t e r s are r e s p o n s i b l e t o j u s t those persons or bodies on whose r e s p o n s i b i l i t y the Emperor appoints or dismisses them. Having provided a f a i r l y d e t a i l e d a n a l y s i s of whose advice was being f o l l o w e d on the appointment and d i s m i s s a l of m i n i s t e r s , Redman conclu-ded t h a t : "at present the Japanese executive i s not e n t i r e l y respon-s i b l e t o any s i n g l e organ i n the s t a t e : i t i s p a r t l y r e s p o n s i b l e t o a number of d i f f e r e n t organs." But Redman was unable t o go beyond des-c r i b i n g Japanese government as " a r b i t r a r y a c t i o n tempered by a number of i l l - d e f i n e d checks plus i n the l a s t a n a l y s i s , popular clamour. Although he was c o r r e c t i n a s c r i b i n g the power of making and breaking cabinets t o a v a r i e t y of i n s t i t u t i o n s , he f a i l e d t o r e a l i s e t h a t the " i l l - d e f i n e d " r e l a t i o n s h i p among them could be b e t t e r understood i n the l i g h t of a d i f f e r e n t c o n s t i t u t i o n a l theory from the one he was a p p l y i n g , namely, the theory of the separation-of-powers. The understanding of separation-of-powers theory has made great advances since Redman's time, and a l o g i c a l a n a l y s i s of i t s main theor-e t i c a l p resuppositions i s e s s e n t i a l before one can show how i t could have helped solve Redman's problem. I n a recent book which b r i n g s together most of the l i t e r a t u r e of the s u b j e c t , M.J.C. V i l e says t h a t the theory o f l t h e separation-of-powers i n i t s purest and most extreme form i m p l i e s t h a t : i ) "the government ( i s ) d i v i d e d i n t o three branches or departments, the l e g i s l a t u r e , the e x e c u t i v e , and the j u d i c i a r y , " i i ) "to each of these three branches there is a corresponding identi-fiable function of government," i i i ) no one is "allowed to be at the same time a member of more than one branch." The avoidance of overlapping membership is supposed to prevent overlapping functions.7 This extreme form of the doctrine, Vile points out, has never been implemented in practice, and even separation-of-powers theorists like Montesquieu and Locke never expounded i t . In practice as well as theory, modifications have been introduced, mainly in order to provide greater guarantees against the encroachment by one branch on the func-tions of another. The notion of checks-and-balances, which implies that one branch can check the actions of another, has become insepar-ably associated with the notion of a separation-of-powers. The term, "separation-of-powers," therefore refers to both what Vile regards as the pure form of the doctrine, and to the notion of checks-and-balances. This accounts for the somewhat paradoxical state of affairs in the writings of thinkers on the subject and in the American Constitution, that although each branch should be confined to the exercise of i t s particular function alone, "each branch was given the power to exercise a degree of direct (Vile's i t a l i c s ) control over the others by author-izing i t to play a part . . . in the exercise of the others' functions." For Montesquieu as well as the writers of the American Constitution, an institution's right to veto the actions of any other institution was as important as i t s exclusive right to exercise i t s own particular function 5 p a r a d o x i c a l as t h i s may seem. I t appears c o n t r a d i c t o r y t h a t one branch should be able t o defend i t s e x c l u s i v e r i g h t t o e x e r c i s e i t s f u n c t i o n by being allowed t o encroach on another branch's r i g h t t o i t s p a r t i c u l a r f u n c t i o n . This i s not n e c e s s a r i l y a c o n t r a d i c t i o n , V i l e argues, because the power t o i n t e r f e r e was l i m i t e d : "the b a s i c i d e a of a d i v i s i o n of func-t i o n s remained, modified by the view t h a t each o f the branches could e x e r c i s e some (author's i t a l i c s ) a u t h o r i t y i n the f i e l d of a l l three functions. " 9 But V i l e does not f u l l y recognise t h a t the theory of checks-and-balances was not merely intended t o guarantee the r i g h t of each branch t o the sole e x e r c i s e of i t s f u n c t i o n . A major preoccupation of Locke, Montesquieu, and the American Founding Fathers was t o prevent any s i n g l e branch from becoming too powerful. The d i v i s i o n of powers among separate branches of government, each of which was t o have a l i m i t e d power over each of the othe r s , was supposed t o prevent t y r a n n i c a l government. The p o s s i b i l i t y of any s i n g l e branch becoming t y r a n n i c a l had to be "checked" and "balanced". Another weakness i n V i l e ' s treatment of the t h e o r e t i c a l problems i n v o l v e d i n the separation-of-powers concept concerns h i s t h i r d charac-t e r i s t i c . The avoidance of over l a p p i n g membership among branches does not seem t o be a s u f f i c i e n t guarantee of each branch's independence. The w r i t e r s of the American C o n s t i t u t i o n wanted separate branches t o have separate e x i s t e n c e s . One branch should not be able t o dismiss the 6 members of another, and i n the case of the President and the two Houses of Congress, there was t o be no opportunity f o r one t o c o n t r o l appoint-ments t o another. The p o s i t i o n of the Supreme Court was ambiguous. I t s members were appointed by the P r e s i d e n t , they r e q u i r e d the r a t i f i c a t i o n o f the Senate, but they could not o r d i n a r i l y be dismissed by any other branch. A t h i r d problem area i s the number of separate branches and func-t i o n s . V i l e i s c o r r e c t i n n o t i n g t h a t the number of powers and f u n c t i o n s hase always been confined t o t h r e e , both by t h e o r i s t s and governments. But there does not seem t o be anything sacrosanct about the number t h r e e , or about the three corresponding f u n c t i o n s . I n theory as w e l l as p r a c t i c e , i t seems p o s s i b l e t o have a d d i t i o n a l branches f o r more s p e c i a l i s e d t a s k s without doing damage t o the separation-of-powers concept. A r e v i s e d l i s t of e s s e n t i a l c h a r a c t e r i s t i c s of a separation-of-powers system, a l i s t which takes account of V i l e ' s omissions, i s t h e r e -fore : i ) The government i s d i v i d e d i n t o separate branches, ranging from the e x e c u t i v e , l e g i s l a t u r e , and j u d i c i a r y alone, t o a l a r g e r number which may i n c l u d e , f o r example, s p e c i f i c branches f o r r a t i f y i n g f o r e i g n t r e a t i e s and f o r the conduct of m i l i t a r y a f f a i r s , i i ) Each branch has a corresponding i d e n t i f i a b l e f u n c t i o n , and greater or l e s s e r veto power over each of the o t h e r s , i n order t o prevent the others from usurping i t s f u n c t i o n too g r e a t l y and t o prevent 7 each from becoming too powerful, i i i ) Each branch is occupied by different persons, whose appointment and dismissal i s outside the control of any other branch, again the degree of separation being subject to variation. One factor that i s l i k e l y to determine the degree of separation i s the extent to which a branch has a separate power base. The American President and two Houses of Congress are dependent on different groups of voters and, once elected, cannot affect one another's existence. A second factor i s the extent to which a branch possesses a veto power over another branch. The greater the veto power, the less the latter branch's a b i l i t y to retain complete control over i t s composition. This factor may often be overshadowed by others. For example, i n spite of the mutual veto the American President and two Houses of Congress possess, the influence of the f i r s t factor predominates. The f i n a l factor i s whether an institution possesses both the powers to appoint and dismiss members of another in s t i t u -tion, or only one of these powers. The members of the American Supreme Court cannot, except by impeachment, be dismissed by other branches, although they owe their appointment to other branches. But had they not possessed security of office, their involvement as an independent' force in institutional struggles over policy-making i s l i k e l y to have been much less than i t was. It may have been even greater had they also possessed the power to make their own appointments, assuming of course that they retained their right of jud i c i a l review. 8 Only when the occupants of one branch are e n t i r e l y subject t o appointment and d i s m i s s a l by another branch does the i d e a of a separation-of-powers cease t o apply. Between t h i s extreme and the opposite one, i n which the existence of a branch i s e n t i r e l y inde-pendent of the c o n t r o l of any other, there are many p o s s i b l e degrees of s e p a r a t i o n . The importance of the three r e v i s e d c h a r a c t e r i s t i c s becomes c l e a r -e r when the separation-of-powers type of system i s c o n t r a s t e d w i t h the parliamentary type. The f i r s t person t o juxtapose the two was Walter Bagehot, who r e f e r r e d t o the B r i t i s h parliamentary system as a " f u s i o n " of powers. From Bagehot's w r i t i n g s , i t i s p o s s i b l e t o l i s t three c h a r a c t e r i s t i c s of a parliamentary system t h a t correspond t o the three of a separation-of-powers system: i ) and i i ) Although i t may be p o s s i b l e t o d i s t i n g u i s h more than one branch of government, i t i s not p o s s i b l e t o i d e n t i f y c o r r e s -ponding e x c l u s i v e f u n c t i o n s , i i i ) Because the power t o appoint and dismiss the executive i s the e x c l u s i v e preserve of the l e g i s l a t u r e , and because the members of the executive are at the same time a l s o members of the l e g i s l a t u r e , the executive cannot be described as separate from the l e g i s l a t u r e . The e l e c t o r a t e i s the only power base i n v o l v e d , and both are u l t i m a t e l y dependent on i t . In modern times, the e l e c t o r a t e chooses the one i n the very act of choosing the other, but t h i s a d d i t i o n t o what Bagehot s a i d does not a l t e r h i s c e n t r a l p o i n t . 9 Because Bagehot wrote his classic, The English Constitution, just before the introduction of manhood suffrage in 1867 and the develop-ment of mass parties in the country, i t has been argued that the develop-ment of greater discipline among parliamentary parties was also the result of this change. Maurice Duverger, one of the most noted modern writers on p o l i t i c a l parties, argues that party discipline and the party system are the major determinants of relations between the executive and the legis-lature. He says that the party system determines how a constitution w i l l function; the constitution does not determine how the party system w i l l funct ion .^ But Duverger fa i l s to take sufficient account of the absence of party discipline in the House of Commons before Bri ta in had achieved a real parliamentary system. One of Duverger's most cogent c r i t i c s , Leon Epstein, denies the argument that disciplined parties can prevent a separation-of-powers, while undisciplined ones can cause i t , regardless of the constitutional arrangement. Epstein correctly pointsoo.ut In the . . . West European situation, parliaments, though or iginal ly sharing power with monarchs, gradually made executive government (the ministries) responsible to elected representatives. By reducing the monarchs to figureheads or by substituting weak presidents, there was a drif t away from the separation of powers. . . . This often occurred, as i t did in Br i ta in , before the demo-cratic age of modern parties. When the parties did develop--indeed, as (author's i t a l i c s ) they developed—they could foresee f u l l executive and legislat ive control by obtaining a parliamentary majority. The constitutional structures of Bri ta in and America played a crucial part in the development of disciplined parties in Bri ta in and parties 10 with a much looser organization and discipline in America. Epstein's contention i s : ( l ) t i s safe to argue that Western democracies have maintained the general constitutional types with which they began. And, in each nation, the constitutional type existed i n some form before the parties. The parties, in other words, grew in the condition-ing environment of a given government structure. They had to adapt organizationally and electorally to i t . What they could become was fixed, within certain l imi t s , by the constitution as well as by the social conditions of the nation. . . . The American separation of executive and legislat ive powers give parties two electoral targets in the national arena. . . . Moreover, the executive power can be retained without continuous majority support in the legislat ive branch. Consequently, there i s absent i n the American system the strongly compelling force for legislat ive party unity that exists in a parliamentary system. He concludes:^ (T)he argument must rest heavily on the fact that there i s nothing about the United States, except the separation of powers, to distinguish i t s circumstances significantly from those of the nations with cohesive parties. . . . (T)he absence of parliamentary government i s decisive. With i t cohesive parties have regularly developed, except in France. Duverger's failure to hold culture constant (as far as this i s poss-ible) and vary the constitutional arrangement, prevented him from being able to determine the influence which the constitutional arrangement can have on party disc ipl ine . Epstein, who did use this technique, there-fore provides a much more convincing argument, and the conclusion that constitutional structure exerts a greater influence on party cohesion than vice versa i s not easy to refute. In recent years, another cr i t ic ism has been made of Bagehot's analysis, but although this cr i t ic ism is to the point, i t does not affect Bagehot's central argument. In an excellent h is tor ica l analysis of the 11 Br i t i sh Constitution, A.H. Birch argues that in modern times party discipline has made the legislature as dependent on the executive as vice versa, and that both are ultimately dependent on the electorate.!^" It must be granted that members of the House of Commons depend almost entirely for their reelection on their party leaders. The performance of the Cabinet and Shadow Cabinet, which consist of the leaders of the two main parties, i s the major factor determining the composition of the House. But this does not affect Bagehot's central point. Wo inst i tut ion besides the legislature, whether or not i t organises i t s e l f into disciplined hierarchical groupings under the control of their leaders, can make and unmake the executive. While the existence of the executive may depend entirely through a mass electorate on the majority party, i f one remembers that this majority party i s the majority party of the House of Commons, Bagehot's point remains. It must also be remem-bered that the major characteristic of the parties in Br i ta in , their d i s c i -pline, through which the executive i s "fused" with the legislature, i s a characteristic peculiar to parliamentary systems. Bagehot's central point i s that the extent to which one branch of government i s dependent for i t s very existence on another alone, indicates the degree of "fusion" or separation-of-powers. Because this point is frequently lost sight of, many criticisms of Bagehot are wide of the mark. The poss ibi l i ty that members of the legislature may be more dependent on the party leaders in the Cabinet and Shadow Cabinet than vice versa, and the ultimate dependence of both branches on the electorate, does not 12 affect the "fusion" of both branches in Br i ta in . Bagehot correctly-described the distinguishing characteristic of the parliamentary system as the to ta l dependence of the executive on the legislature, and that of the separation-of-powers system as the lack of such depend-ence of the executive on the legislature. The way to ascertain to which type a particular poli ty most closely approximates is to find out to whom the executive i s responsible, or who has the power to make and break i t . I f the legislature alone poss-esses this power, the poli ty i s of the parliamentary type. I f no other branch possesses i t , or i f a variety of branches do and the executive is not entirely dependent on any one of them, the executive i s separate from each and the poli ty i s of the separation-of-powers type. The consequence of either type which most interested Bagehot was the relationship between the executive and legislature, a relationship which determined whether the executive would be strong and efficient or weak and sluggish. In contemporary Br i ta in , where the House of Commons alone could make and break cabinets, Bagehot discovered a harmonious relationship between the executive and legislat ive branches, and a strong cabinet. He wrote, "the efficient secret of the English Consti-tution may be described as the close union, the nearly complete fusion, of the executive and legislat ive powers," a union effected by means of the Cabinet, which i s "the buckle which fastens the legislat ive part of the State to the executive part of the State." Because the Cabinet was "a board of control chosen by the legislature out of persons whom 13 i t trusts and knows, to rule the nation," conflict between the executive and legislature was eliminated.^5 Th e legislature had an ins t i tu t ional incentive to support the executive which i t , and only i t , had appointed and could dismiss, while the executive had an ins t i tu t ional incentive to follow the advice of the legislature which alone had created i t . The separation-of-powers type of system, on the other hand, provided no incentives for in ter- ins t i tu t ional harmony, but because each ins t i tu -t ion was occupied by different people, not entirely dependent on one another for their existence, in ter- inst i tu t ional deadlock was inherent. The American President and two Houses of Congress were chosen by d i f f -erent electorates, partly at different times, and were responsible to these electorates, not to one another. Not ppossessing ins t i tu t ional incentives for mutual support, both the executive and legislature were weakened by the continual need to win mutual consent, even over the most minor issues. The great vice of this type of system, according to Bagehot, was the ever-present conflict and deadlock that existed between the executive and the legislature, making i t very d i f f i cu l t for the business of government to be carried out at a l l . He wrote After saying that the division of the legislature and the executive in Presidential governments weakens the legislat ive power, i t may seem a contradiction to say that i t also weakens the executive power. But i t i s not a contradiction. The division weakens the whole aggregate force of Government. . . . The Ameri-can Government cal ls i t s e l f a Government of the supreme people; but at a quick c r i s i s , the time when the sovereign power i s most needed, you cannot find the supreme people. You have got a Congress elected for one fixed period, going out perhaps by fixed instalments, which cannot be accelerated or retarded--you have a Ik President chosen for a fixed period, and immovable during that period. . . . There i s no elastic element, everything i s r i g i d , specified, dated." To summarise Bagehot's classification of systems and their conse-quences, parliamentary systems have strong executives, because of a concentration of power in the Cabinet-chosen-by-the-Commons, whereas separation-of-powers systems have weak ones, because of a fragmentation of power among as many institutions as possess a veto over the executive and on which the executive is not solely dependent for i t s existence. In the former, the Cabinet is responsible to the Commons, because only the Commons can make and unmake i t , whereas in the latter, the executive i s either not responsible to any other institution at a l l , or is con-currently responsible to a number of institutions, on each of which i t is partly but never entirely dependent for i t s existence. The former i s characterised by the legislature's possession of the entire power to raise and pull down a cabinet, whereas the distinguishing characteristic of the latter' i s the legislature's lack of this entire power. One of Bagehot's major purposes i n comparing the B r i t i s h and American systems of government was to show why the B r i t i s h executive was strong and efficient, and the American executive comparatively weak and sluggish. He concluded that the institutional incentive for the British executive and legislature to cooperate was the main reason for the efficiency of that country's government. The lack of this incentive was the reason for the continual deadlock between the American executive and legislature and the comparative weakness of American government. 15 The hundred years since Bagehot wrote have made i t possible to dis-cover additional typica l consequences of either ins t i tu t ional arrangement. While each of them may in any given case be as much, i f not more so, due to cultural and economic forces, the purpose here i s to demonstrate that pol i t ico- ins t i tu t iona l forces can also exert an influence which makes their occurrence l i k e l y . In each case, the poss ibi l i ty that socio-economic conditions may reinforce or contradict pol i t ico- ins t i tu t ional forces must be born in mind. The purpose of this study i s not to assess the relative importance of social , economic, and pol i t ico- ins t i tu t ional forces in the p o l i t i c a l development of Japan, but to reassert the importance of the la t ter , which in recent years have been almost entirely ignored in scholarly writing on p o l i t i c a l development. In order to redress the balance of emphasis, this study, while not minimising the importance of economics and culture, takes them for granted and concentrates solely on demonstrating the influence of p o l i t i c a l inst i tut ions, an influence which may have rein-forced or worked against other influences. But no attempt has been made to weigh these influences against one another, because a consideration of a l l the factors for the period of more than half a century fa l l s beyond the scope of this study. It i s assumed, therefore, that the reader has a basic knowledge of Japanese social , economic, and p o l i t i c a l history, and is able to assess the relative importance of different forces. The f i r s t typica l consequence of a particular ins t i tu t ional arrange-ment, one that existed even before Bagehot's time, i s the likelihood of 16 high levels of corruption in separation-of-powers systems, and the comparative unlikelihood of the same in parliamentary ones. The reason i s that in the absence of ins t i tu t ional incentives for harmony between the executive and legislature, the flow of money from one to the other becomes the most easily effective deus ex machina for resolving inter-ins t i tu t ional deadlock. Corruption in eighteenth-century Br i ta in , when ins t i tu t ional forces were of the separation-of-powers type, was as endemic as in America, and the name of Robert Walpole became synonomous with corruption i t s e l f . It is not being claimed that a separation-of-powers system i s either a necessary or sufficient condition for high levels of corruption, but that separation-of-powers systems do not contain ins t i tu t ional forces that mitigate corruption, while parliamentary systems do. The actual causes of corruption in particular cases may involve a variety of socio-economic factors, and their importance i s not disputed. A second typical consequence of separation-of-powers systems, again other things being equal, i s the greater importance personality pol i t ics i s l i ke ly to have over policy p o l i t i c s . A legislature which is not the sole maker and breaker of the executive does not possess the ultimate sanction to compel the executive to do i t s bidding, and an executive, which i s not entirely dependent on the legislature, has no overriding incentive to l i s ten to that legislature. Moreover, the more institutions that share the power to make and unmake the executive or veto i t s actions, the less i t s need or ab i l i t y to pursue the policy of any one of them. 17 When there i s an Upper House, a Lower House, a Privy Council, an Army, a Navy, a C i v i l Service, and a group of Old Cronies, each of which can veto policy and unmake cabinets, the Cabinet i s hardly l i k e l y to pursue exclusively the policy of any single one. They i n turn, rea-l i z i n g that i t is unrealistic to expect th is , w i l l t ry instead to enlist sympathetic individuals to represent them in the Cabinet. Po l i t i cs becomes a problem of trying to get the most s k i l f u l and sym-pathetic personality to do his best in the competition among the various inst i tut ions, a world in which promises to implement policy can never be made. Prime Ministers or Presidents are chosen for their personal quali t ies, usually an ab i l i t y to deal with as many of the competing institutions as possible, the c lass ica l example once again being Robert Walpole. Related consequences are, f i r s t l y , that leaders capture parties, rather than that parties raise leaders. I f a Prime Minister or Presi-dent uis always chosen for his personal ab i l i t y to get the most out of a competitive in ter- ins t i tu t ional struggle, parties are on the keen look out to have such individuals assume their leadership. This i s the best way to get a party Premier or President. And because no one can become a Premier or President without among other things the support of a party, otherwise qualified individuals are keen to capture a party. Secondly, factionalism, frequently along personality l ines , and a lack of discipline in a l l institutions i s l i k e l y to develop. The reason i s that i f the executive's existence i s entirely dependent on the w i l l of a majority of only one inst i tut ion, that majority has a greater incentive 1 8 for discipline and unity than do the members of institutions which are not in sole possession of the power to make and break the executive. The probability that parties have less discipline in separation-of-powers systems than in parliamentary systems also exists for groups in other inst i tut ions. I f no single branch of government possesses the entire power to appoint and dismiss the executive, none receives pol i t ico-ins t i tu t ional incentives to increase i t s discipl ine . Although factionalism may often have basic causes unrelated to ins t i tu t ional forces, separation-of-powers systems do l i t t l e to mitigate that factionalism. It i s true that parliamentary systems provide only the Lower House of the legislature with incentives for discipl ine . But factionalism in powerless institutions i s less serious than in ones that can affect the executive's existence or block i t s pol ic ies . In separation-of-powers systems ins t i tu t ional faction-alism can be a great impediment to the smooth functioning of government. Had Redman been able to benefit from recent studies of separation-of-powers and parliamentary theory, as well as recent comparative studies of p o l i t i c a l institutions and p o l i t i c a l parties, his conclusions may have been quite different. He may have realised that separation-of-powers theory i s more conducive to an understanding of prewar Japanese ins t i tu -t ional development than parliamentary theory. The argument of this thesis i s that not merely did the Meij i Constitution create in Japan an ins t i tu t ional arrangement that approxi-mated very closely to the separation-of-powers type, but that in 19 practice institutional forces were of the kind that exist in America. Although the popular election of the American President on the one hand, and the more diverse makers and breakers of prewar Japanese Cabinets on the other, do indicate an important institutional d i f f -erence between the two countries, i t is not one which requires a reclassification of the Japanese system. Both America., and prewar Japan possessed a separation-of-powers institutional framework, and were susceptible to the same kinds of problems that occur readily in this type of system. In so far as the separation-of-powers concept conveys an inten-tion to prevent any single branch of government from becoming too powerful, the idea of checks-and-balances requires the major emphasis. But i n so far as i t also conveys an intention to have specific branches for specific tasks, the notion of a separation-of-powers i t s e l f should be emphasised. The two main ideas are inseparably bound up in the single concept, but one or the other may receive the greater emphasis in particular cases. In Japan, the idea of separate branches for separate tasks seems to have received greater attention, since branches besides the usual three were, granted highly specialised functions. But the desire to prevent any single branch from assuming too much power was also consciously incorporated in the Constitution. Because the idea of checks-and-balances received less emphasis in theory as well as practice, the term "separation-of-powers" rather than "checks-and-balances" is particularly applicable to Japan. But i t must be remembered 20 that part of what i s meant by "separation-of-powers" i s that checks-and-balances are involved. The ideal of col lect ively responsible cabinets has been cherished by many Japanese champions of democracy, but few have seen i t purely i n terms of a relationship between the executive and the legislature. Usually, responsible government has come to mean simply government that can be called to account by the public for whatever i t does. But because this type of government also exists in America, the central concern of Bagehot has been lost sight of. A number of Japanese scholars have made studies of the l ives of particular cabinets. Some have even looked specifical ly at the Cabinet and i t s occupants over a considerable period of time., and attempted to analyse reasons for Cabinet changes.^ But no one has examined the extent to which collective responsibility in Bagehot's sense existed in prewar Japan, or to relate the facts about the actual makers and breakers of cabinets in prewar Japan to theories of constitutionalism. This study draws conclusions about the type of ins t i tu t ional arrangement under which pol i t i cs in prewar Japan existed. The sole empirical evidence used i s a classif icat ion of the apparent reasons for the rise and f a l l of successive ministries, as well as of the individual qualifications of cabinet ministers. Relying on Bagehot's insight, i t i s assumed throughout that the Cabinet i s responsible to i t s makers and breakers as well as whoever i s able to effect i t s 21 composition, and that from this information types of institutional forces can be inferred. While the sources usually provide a l l the details of why min-is t r i e s came and went, they do not always indicate which institutions played the major parts in effecting the changes. In cases when the institutional role i s unclear, attempts have been made to classify the various pressures in institutional terms, because apart from the institutional context in which governments acted, governments could ignore these pressures with impunity. For example, public dissatis-faction with the Portsmouth Treaty in 1905 could not be ignored by the Katsura government, which had"made a deal with the Lower House to hand over the premiership to the leader of the majority party in return for a promise not to take up the public cause against the government's foreign policy. I have made similar assumptions about reasons for the appointment of particular ministers, about which the sources are also frequently silent. Because cabinets regularly consisted of one or more represent-atives of most institutions able to make and break i t , excepting the Privy Council and the Genro, I have assumed that the major reason for the appointment of ministers was membership i n these institutions. There is less justification for the view that those who made the actual appointments to the Cabinet were always f u l l y conscious of the reasons for their actions. Impersonal forces frequently constrain actions in 22 ways unnoticed by those directly involved. I assume the most important among them is ins t i tu t ional a f f i l i a t i on . Appendices I I to V indicate the ins t i tu t ional background of a l l ministers in I89O-I9U0, which I have assumed i s the major reason for their appoint-ment. When ministers had only one ins t i tu t ional qualif ication, the conclusion that they were appointed primarily because of i t i s hard to res is t . When they had more than one qualif ication, i t i s often impossible".to single out any one as the most important, and judgement becomes more d i f f i c u l t . In many cases such ministers seem to have been valued precisely because of their relations with a variety of inst i tut ions, i n the hope that they would be able to serve as bridges between them. The attempt to cover a considerable period of time and to classify rather than to establish reasons for ministerial changes has required that sources be l imited. I feel that a selected sample of the most reliable sources w i l l minimise the poss ibi l i ty of errors in the rea-sons for particular cabinet changes. But because I have not attempted to verify these reasons when my sources are in agreement, the possib-i l i t y of errors i n particular cases remains. More detailed studies on the rise and f a l l of individual ministries can be carried out la ter . My reference to secondary sources reflects my concern to infer a certain type of poli ty from facts that are not generally disputed. The emphasis is on the interpretation rather than the ver i f icat ion of reasons for cabinet changes. 23 CHAPTER 1 CABINET RESPONSIBILITY IN THE MEIJI CONSTITUTION  Background For almost the entire period after the expulsion from the inner c i rc le of the Meij i p o l i t i c a l leadership of Itagaki Taisuke, u n t i l the government proclamation i n l 8 8 l that a national assembly would be est-ablished in I 8 9 O , there was a great deal of discussion in the country as well as among the government leaders about the kind of constitutional government Japan should adopt. Itagaki took up the cause in the country and organised a number of p o l i t i c a l groups which agitated for the est-ablishment of a representative assembly. This agitation became what i s known as "The Movement for Constitutional Government" (Jiyu minken undo), and at one time or another embraced members of the samurai, landowner and peasant classes. It culminated i n the formation i n l 8 8 l of the-Jiyuto (Liberal Party), the f i r s t national p o l i t i c a l party i n Japan. Partly in response to this Movement, but mainly to prove to the Western Powers which held ext ra- ter r i tor ia l rights in Japan that Japan was a "c iv i l i zed" country, the leaders of the government also showed an interest in constitutional government, particularly Kido Takayoshi. After the death of the main f i r s t generation leaders of the Meij i Restoration, younger men, who had not played such important parts in the Restoration, became the most inf luent ia l government o f f i c i a l s , mainly Ito Hirobumi and Okuma Shigenobu. Both were strong advocates of constitutional government 24 even though they had their differences. In l88l these differences led to Okuma's expulsion from the government because of his desire to intro-duce immediately a system modelled on the Br i t i sh parliamentary one. Ito was more cautious. and advocated a ten year delay in order to work out the details of the system that would be the most suitable for Japan. His preference was for something similar to the Prussian Constitution. Okuma1s purge in l88l marked a new phase in the struggle over national assemblies. The government began to draft the Constitution, but not before Ito had spent some time in Europe, mainly Germany, study-ing European constitutionalism. The task was carried out in absolute secrecy, and no one knew exactly what the contents of the Constitution would be u n t i l i t was formally promulgated in I89O. 5kuma took up his cause in the country, much as Itagaki had done in the l870 's, and set up the second national party, the Kaishinto (Progressive Party). Its ideolo-gica l inspiration came from Br i t i sh thinkers l i ke . M i l l , Locke, and Spencer, while that of the Jiyuto came mainly from Rousseau. The socio-economic base of the Kaishinto was the urban capital is t class, while that of the Jiyuto was the rural classes. The Jiyuto was the more extreme of the two parties, and during the years 1883-1884, i t was the instigator of many acts of violence against the government. Itagaki was in Europe at the time, and returned to find his party detested by the world of officialdom. In 1884, under pressure from the more conservative landowner-wing of the party, which had not 25 always gone along with the more extreme peasant-wing, the Jiyuto was formally dissolved. The Kaishinto was dissolved in the same year, mainly because the urban capitalists were afraid of antagonising the government. The lack of a legitimate ins t i tu t ional base from which to carry out their operations i s of the greatest importance in the develop-ment of p o l i t i c a l parties in Japan. Both parties were revived by the promulgation of the Constitution and the f i r s t general election in I 8 9 O , which gave them legitimacy , but at the same time put constraints on their tactics and ambitions. The present chapter i s not intended to demonstrate the influence of a l l the parties and leaders on the f ina l Constitution as i t emerged in I89O. It i s limited to a discussion of the central problem with which this study i s concerned, and analyses the extent to which the Meij i Constitution was of the separation-of-powers type. Most of what follows i s therefore of a highly theoretical nature. Cabinet Responsibility and Constitutional Theory The intentions of the Meiji leaders on the question of Cabinet responsibility are more d i f f i cu l t to ascertain than on most other matters. The Constitution has l i t t l e to say on the central problem: "The respec-tive Ministers of State shall give their advice to the Emperor and be responsible for it."-'- By "Cabinet responsibility" the framers seem to have meant merely that the Cabinet take the blame or credit for what-ever was done in the Emperor's name. The only way to maintain the 26 i nv io lab i l i t y of an Emperor who could do no wrong was to avoid having him do anything p o l i t i c a l l y controversial. Ito Hirobumi made this clear in his Commentaries on the Constitution: "They w i l l not be able to release themselves from responsibility by pleading an Order of the S o v e r e i g n . T h e task of governing was therefore to be a min-i s t e r i a l rather than an Imperial responsibil i ty. But the intentions of the Meiji leaders are far from clear on the pertinent question. To whom was this responsibility to be born? Legally, the Cabinet was supposed to be responsible to the Emperor alone. According to Ito:3 The power of deciding upon his [the Minister of State's\ res-ponsibi l i ty belongs to the Sovereign of the State: He alone can dismiss a Minister, who has appointed him. . . . It i s only a legitimate consequence that the power of deciding as to the res-ponsibil i ty of Ministers, i s withheld from the Diet. Because in law the Emperor alone could appoint and dismiss Ministers, in law they were responsible to him alone. There i s reason to believe, however, that the cdifference between legal theory and practical pol i t ics did not entirely elude the Meij i leaders. The legal theory of responsi-b i l i t y to the Emperor is qualified in I to 's Commentaries by the require-ment that :^ (l)n making an appointment the susceptibil i t ies of the public mind must also be taken into consideration. This may be regarded as an indirect method of controlling the responsibili t ies of Minis-ters. . . . Ministers are directly responsible to the Emperor and indirectly so to the people. The Meij i leaders did not appreciate a l l the d i f f icu l t ies involved 27 i n upholding legal theories that bear l i t t l e relation to real p o l i t i c s . I f they real ly intended the wishes of the public as expressed in the Diet to be a consideration i n the appointment and dismissal of ministers, so that the Emperor's actions would be constrained by the wishes of the Diet, why did they not modify the legal theory accordingly? Although they were not adhering to a t radi t ion that demanded an active p o l i t i c a l role for the Emperor—for centuries he had been p o l i t i c a l l y passives-there was something about the t radi t ional way of l i f e that did fac i l i ta te making him the nation's central p o l i t i c a l figure. The closely knit nature of Japanese society gave to Japanese l i f e an essential unity that prevented the separation of ethics, re l ig ion, and pol i t ics into auton-omous and competing walks of l i f e , as well as the separation of ethical from religious or p o l i t i c a l authority. Because the nation's l i f e was one, i t was not possible to compartmentalise i t , or fragment the authority to which i t subscribed. As head of the family of the nation, the Emperor was regarded as the single symbol of the nation's authority i n a l l walks of l i f e . He was at the same time High Priest , Custodian of Morals, and P o l i t i c a l Leader. Loyalty to him meant loyalty to the nation and sub-mission to i t s value judgements.^ When the country was p o l i t i c a l l y divided during the feudal period, his authority in political/matters was not fu l ly recognised. Quite naturally, the great movement towards po l i t -i c a l unity in the nineteenth-century was accompanied by a growing recog-ni t ion of the Emperor's p o l i t i c a l authority. 28 To recount this history of the Emperor's position in Japanese society s t i l l does not explain why the Meiji leaders gave the Emperor such extensive p o l i t i c a l prerogatives. The answer probably l i e s in the general confusion of contemporary constitutional theory on which the drafters of the Meiji Constitution relied so heavily. I f they fa i l e d to understand the discrepancy between the legal theory of their consti-tution and the way i t was li k e l y to work in practice, this i s largely because the actual working of constitutional government in contemporary Europe was so imperfectly understood by European theorists themselves. The dominant notion was s t i l l the Montesquieuan concept of separation-of-powers, which required that executives and legislatures not only have separate and clearly defined functions, but have separate exist-ences. That this could in practice often result in deadlock between the executive and legislature was regarded as desirable by Americans, unavoidable by Frenchmen, capable of legal solution by Germans, and unimportant by Britishers. The British could afford to ignore the problem only because they did not really have i t . Their theory of a balanced constitution was in an important respect different from the Montesquieuan theory of a separation-of-powers.^ It made harmony between the executive and l e g i -slature the essence of good government, and recognised that'this was the result, not of the separation of powers, but of their inter-dependence through the selection of the executive by the legislature. 29 In Germany, however, where the legislature could not dismiss the executive, the problem of continual deadlock was as acute as in the United States. Unlike the Americans, however, the Germans disliked the resulting paralysis of government, even though no contemporary German theorist recognised the inadequacy of the proposed solution. While the executive was given legal sovereignty, this was insufficient to prevent the Diet from coming into continual conflict with i t . Making the executive legally responsible to an even higher authority, the Emperor, to whom i t could appeal for support against the Diet, merely altered the balance of power in the executive's favour. It also made of the Emperor a kind of deus ex machina to resolve deadlock between the two main branches of government, and therefore made him highly controversial. The Meiji leaders gave to the Japanese Emperor p o l i t i c a l preroga-tives that marked a sharp departure from tradition, because they believed this to be the only way to prevent the paralysis of government. They did not envisage him serving as a mere figurehead like the British monarch, because this would have meant either complete cabinet irresponsibility, or responsibility to the Diet. Nor did they intend him to be free to act contrary to the advice they gave him, or to take the side of the Diet in cases of conflict with the executive. On the contrary, they wanted the Emperor to be active in support of the policies they deemed to be in the national interest. Like the British Tories, they believed that they 30 were somehow less partisan than any one else. They therefore never appreciated, u n t i l after the opening of the f i r s t Diet, that there was an inherent contradiction in upholding the i n v i o l a b i l i t y of the Emperor and wanting him to support the executive in i t s conflicts with the Diet. The Emperor could not at the same time interfere i n pol i t i c s and remain uncontroversial. Wo one in Germany or Japan seemed to realise that the only kind of executive that would in practice be able to gain the support of the Diet was one which the Diet alone elected. This "efficient secret" to the harmonious functioning of exec-utives and legislatures was f u l l y understood only by a few British theorists. Actually, Walter Bagehot alone regarded the interdependence of executives and legislatures as the sole solution to the problem of conflict between them. The Br i t i s h theorists whose writings were read by the Meiji leaders, especially John Stuart M i l l and Sir Alpheus Todd, supported the selection of cabinets by the Commons for other reasons, ones which did not make the British system particularly attractive : for the ease with which a cabinet could be dismissed by the Commons.® The most probable reason for the Japanese rejection of British constitutionalism wastthat after 1867 i t was imperfectly understood by British theorists. During the l880's, the works of M i l l , Bagehot, and Todd were s t i l l regarded as the most authoritative, even though a l l of them were written before the great constitutional change of 1867. 31 Even Bagehot did not foresee that what would really make the Bri t i s h Cabinet stronger than ever was the support of a disciplined majority in the House, something which came only after Bagehot's time. In fact, i t was really only during the years 1832-67 that the House alone, as an independent body, made and unmade governments.9 Before this, cabinets had been kept in office by monarchs, and their a b i l i t y to purchase the support of a sufficient number of factions in Parliament. But from 1832-67, parliamentary groups had l i t t l e i f any cohesion, and members often changed sides over the most unimportant issues. Only during these years did the Commons have a corporate existence of i t s own and independently make and unmake cabinets. In no other period did the writings of M i l l , Bagehot, and Todd bear a l l that much relation to reality. After 1867, cabinets were kept in office by disciplined maj-ority parties, whose fortunes depended not on the ab i l i t y of great parliamentary orators to win over individual members, but on the results of elections. Electorates therefore became the makers and breakers of governments, even though they only exercised the limited choice between leaders of the two great parties. The efficient secret of British constitutionalism has since been that the leaders of the party that wins a majority of seats i n the Commons are the very people who occupy the Cabinet positions, thereby ensuring harmony between the executive and legislature. Not u n t i l 1885, however, when Henry Maine expressed his horror at the way the Commons 32 was coming under the control of two disciplined phalanxes, was this even noticed by constitutional theorists. They failed to see that what Bagehot had regarded as the greatest virtue of Br i t i s h govern-ment, i t s strength, was even more present now that the House was less f i c k l e . Without disciplined parties in the Commons, cabinets could on occasion be quite weak, and f a l l with every change of mood in the House, as happened quite a few times between I832-I867. The Meiji leaders were aware that majority parties had become the foci of p o l i t i c a l responsibility i n England. But because they were convinced of their own impartiality, they regarded this kind of government as sectional. Unfortunately, there was no British theorist to point out how strong and stable i t was, and how i t avoided inter-institutional deadlock. No one pointed out that i t i s far easier to win the support of a majority of voters in an election every four years, than of a majority of Diet members every time a budget or piece of legislation i s presented. Winning an election guarantees loyal support from the legislature u n t i l the next election, whereas winning the support of the Diet for one budget guarantees nothing for future budgets. It i s hardly surprising that the Meiji leaders, who were f u l l y aware of the need for strong stable government, especially during the years of rapid social and economic change, found l i t t l e to attract them i n a system that was supposed to provide for the ever present possibility of governments being made and unmade at the slightest change of mood in the Commons. The imperfect understanding of British 33 government "by British theorists i s possibly of greater importance than any other factor in explaining why the Meiji leaders rejected British constitutionalism. They turned to German theory, because the Germans were the only contemporary Europeans to stress the weaknesses of the prevailing separation-of-powers theory, and offer what seemed to be a practical alternative.!° The Meiji leaders believed that the theory of cabinet responsib-i l i t y to the Emperor avoided the problems of Montesquieuan constitue tionalism. They had no less an authority on European constitutional-ism then Hermann Roesler assure them that i t would prevent deadlock between the executive and the legislature Consequently the sovereign powers are not in any way divided by the constitution between the Emperor and the people. The whole government power remains concentrated in the person of the Emperor. . . . That power must be united and undivided because division of powers produces discord and dissolution. Divided powers can turn out as tyrannical as undivided ones; and in the case of disharmony one power must assume the pre-dominance over others, even unlawfully against the constitu-tion; otherwise the course of government would be stopped. A remedy against tyranny cannot be found in the division of powers. Ito agreed: "Because the imperial sovereignty is the cornerstone of our constitution, our system is not based on the European ideas of separation of powers or on the principle in force in some European countries of joint rule of the king and'the people."-^ Both Roesler and Ito believed that i t was possible to give the Diet the power to withhold i t s consent to laws and increased budgets, and at the same time avoid Montesquieuan constitutionalism. While they exp l i c i t l y deny that the Japanese Constitution embraces a separation-of-powers, what they describe i s in fact a system of checks-and-balances. Roesler says that the sovereign's power i s restricted in five ways l ) Every law, the annual budget and other important financial measures require the consent of the Diet; 2) every act of the executive power of the sovereign requires the advice and sig-nature of a minister of state; 3) the judicature shal l be exer-cised by independent courts of law according to law only; h) the respective domains of the legis la t ive , executive and jud ic ia l powers are to be constitutionally fixed as much as possible; 5) in a l l government affairs, the Diet can receive petitions, make addresses to the Emperor or representations to the govern-ment, and put questions to or demand explanations from the same. The Meiji Constitution did not merely enable the Diet to obstruct the acts of the Emperor and his ministers. It gave the two Houses equal powers, making them equally capable of creating deadlock. In phrases reminiscent of Montesquieu, Roesler approved of the Upper House because: i t forms a great check to hasty, one-sided, passionate and oppres-ive legislat ion . . . and secures more stable and harmonious rela-tions between the Crown and removes the frequent occasions for the exercise of the sovereign prerogatives in cases of confl ict , as of the right of veto, prorogation, dissolution, etc. I f he could say that the "Upper House should be able to hold the balance i n the deliberations of the national body,"-^ i t i s strange that he could s t i l l believe that the Constitution did not embrace a separation-of-powers . The legal position of each branch of government made i t highly l i k e l y that the Meij i Constitution would function differently from the way i t s framers intended. Either responsibility to the Emperor would have to be 35 sacrificed in practice, in which case i t was l i k e l y that a full-fledged separation-of-powers system would develop over time, or the Emperor would become highly controversial. I f a l l branches were in practice responsible to him alone, his ruling in favour of some and against others in cases of inter-^institutional conflict would make him a highly "po l i t i ca l " figure. Because of the emphasis the Meij i leaders placed on the Emperor's i nv io l ab i l i t y , the second alternative was unlikely. In the following section, the way the Constitution was l i k e l y to function i n practice i s analysed in the l ight of separation-of-powers and parliamentary theory. The discussion i s of a logica l and theoret-i c a l nature- but i s based on the legal provisions of the Meij i Consti-tution. Although largely a hypothetical argument, i t takes as i t s premises the legal ins t i tu t ional structure set up in I 8 9 O . How the Constitution was Likely to Work in Practice A central point in this study i s that channels of responsibility are impossible to trace in separation-of=powers systems, and that the theory of Cabinet responsibility i s not an aid to understanding how they function. Sometimes, even Roesler seems to have realised that the Meij i Constitution incorporated a par t ia l separation-of-powers, which could prevent the system from functioning as intended. The Privy Council, he s a i d : ^ has only consultative functions and no executive powers whatever; i t s opinion may be asked for or not . . . and i t may be accepted or rejected as the government may think f i t . Otherwise the mini-sters of state would be deprived of the l iberty of their advice and would be reduced to mere executive agents of the Privy Council. 36 But Roesler failed to realise that although in theory a l l i n s t i -tutions were responsible to the Emperor, none would be in practice, because the Emperor was not to get involved in p o l i t i c a l controversies. The Cabinet, which was supposed to be his major source of assistance and advice,- would constantly be confronted by a number of r i v a l i n s t i -tutions, each claiming to be the legitimate adviser on issues related to i t s constitutional prerogatives: the Upper House, the Lower House, the Privy Council, and the Armed Services. Their a b i l i t y to veto cer-tain policies and support others would enable them to insist that their •advice be heeded. The two Houses of the Diet could veto legislation and increased budgets; the Privy Council, which was to be the interpreter of the Constitution, could veto any policy but mainly Imperial Ordinances ahdttreaties with foreign c o u n t r i e s , a n d the Armed Services could -i Q exercise a veto on foreign p o l i c y . 0 Rather than be overruled by the others, however, the Cabinet was supposed to be the organ through which they attained access to the Throne to tendend their advice. That the Armed Services also had direct access to the Throne was not unique. Both Houses of the Diet could address'the Emperor directly, and the Privy Council and the Meiji leaders themselves, who later became known as the Genro, or "elder statesmen," could also circumvent the Cabinet. Nevertheless, the Cabinet was supposed to be a unique adviser. The reason for i t s establishment i n 1885 was precisely to bring unity to the government. ^9 3 7 Because the Cabinet was to be the major executive organ of govern-ment, other inst i tut ions, which were given specific functions of their own, were l i k e l y to make the Cabinet the organ through which they t r ied to guarantee their sole right to exercise these functions. For example, the Privy Council would try to have the Cabinet conclude only those treaties with which i t was i n agreement, the Diet would try to ensure the compilation of legislat ion and budgets of which i t approved, and the Armed Services would try to win Cabinet approval of the wars they wanted waged. The most obvious way for any inst i tut ion to gain Cabinet support for i t s pol ic ies , was for i t to veto those of unsympathetic cabinets in the hope that the lat ter would either change their minds, or, because of the deadlock resulting from the veto, be forced out of office and replaced by more favourably-disposed cabinets. In so far as an ins t i tu t ion succeeded i n raising or removing an entire ministry, parliamentary theory would describe the Cabinet as responsible to i t . But should a number of institutions employ the same tactic simultaneously and the exercise of vetoes and counter-vetoes lead to a rapid succession of Cabinet changes, the situation would become highly confused and would not be c la r i f ied by the notion of "concurrent responsibility" to a variety of inst i tut ions, except in a negative sense. I f such a rapid succession of Cabinet changes resulting from the vetoes of the 'different branches did occur in Japan, a theory would be 38 required that takes account of each branch's use of i t s veto power to have the Cabinet pursue i t s line of policy in those matters which are i t s constitutional preserve. The theory of a separation-of-powers embraces both the idea of separate branches for separate tasks ,and the idea that each branch possesses a veto power to ensure the acceptance of i t s advice. But because the theory of a separation-of-powers also presupposes that each institution, including the executive, has an independent existence, i t does not apply to the case in which the Cabinet can be dismissed as a result of any institution's veto power, as seemed possible in Japan. One should say that a Cabinet is "negatively" responsible to any institution that possesses the entire power to break i t and that i f each of a number of institutions can on i t s own dismiss an entire ministry, the Cabinet is best described as "concurrently res-ponsible," in a negative sense, to a l l these institutions. Cabinets which have no independent existence are also l i k e l y to have more than one institution responsible for their appointment. But whereas the uncompromising use by only one other institution of i t s veto power could possibly force the resignation of an entire ministry, i t seems unlikely that any single institution could use i t s veto to raise an entire ministry. Other institutions, which also wanted cabinets of their choice, could simply exercise their vetoes. The only solution would be compromise: antagonistic institutions would have to agree to more neutral ones domina-ting the Cabinet or to a f a i r division of influence. 39 A Cabinet that i s the compromise crea t ion of a number of Cabinet -makers, as was l i k e l y under the M e i j i C o n s t i t u t i o n , i s not w e l l descr ibed by the term "concurrently r e s p o n s i b l e , " which i s appropriate only f o r a Cabinet that can be dismissed s o l e l y by any one of a number of i n s t i t u t i o n s . The term "separation-of-powers" i s more a p p l i c a b l e , because as f a r as each i n d i v i d u a l i n s t i t u t i o n i s concerned, the power to r a i s e an e n t i r e cabinet independently o f others does not e x i s t . I f no s ing le i n s t i t u t i o n can on i t s own r a i s e an e n t i r e Cabinet , the existence of the Cabinet i s best regarded as separate from i t . The power to appoint i n d i v i d u a l m i n i s t e r s , however, as opposed to the power to appoint an e n t i r e m i n i s t r y , could be one Cabinet-makers shared. The compromise most l i k e l y to gain general acceptance would be each i n s t i t u t i o n ' s c o n t r o l over appointments to those Cabinet seats that were most c l o s e l y concerned with the exercise o f i t s c o n s t i t u t i o n a l l y determined f u n c t i o n , although t h i s would never be easy, because i n p r a c t i c e the task o f government i s d i f f i c u l t to fragment i n t o s e l f - c o n t a i n e d func t ions . Nevertheless , the Japanese Cabinet could be expected to con-s i s t of representat ives of the veto-possess ing i n s t i t u t i o n s , because, l i k e i n e ighteenth-century B r i t a i n , t h i s was the most obvious compromise. Such a compromise would not , however, enable the Japanese executive to escape the t y p i c a l problems executives i n separation-of-powers systems encounter. Because each of the veto-possess ing i n s t i t u t i o n s , except the Cabinet , had i n independent ex i s tence , they had no i n s t i t u t i o n a l ho incentive to cooperate. Each would tend to regard i t s e l f responsible to i t s own constituency alone: the Upper House to the aristocracy, the Lower House to the propertied class, the Armed Services to themselves and their recruiting ground, the peasantry, and the Privy Council to the old cronies who had sat in previous governments. The secret to the British solution of conflict among institutions was that a l l the p o l i t i -cally important ones were dependent on the same power base, a single electorate. After 1867, cabinets and parliaments rose and f e l l together, because the existence of both was ultimately dependent on the verdict of the electorate. This institutional incentive for harmony was espec-i a l l y important when the acts of one institution were distasteful to another, because even then cooperation was more or less guaranteed. The Japanese constitution, however, only provided legal fictions for the resolution of disagreement between institutions. If no single branch would in practice be responsible to the Emperor, the Emperor would not be able to smooth over the conflicts that arose among them, and deadlock would be l i k e l y . And deadlock was l i k e l y to be even more complex than in the United States, because the Meiji Constitution gave each "bf a greater number of separate power bases i t s own legal foothold. One of the f i r s t Japanese parliamentarians to realise that the main error the drafters of the constitution made was excessive/reliance on European constitutional theory was Uyehara Etsujiro. In 1910, he wrote This indicates a complete separation between the executive and the legislature. . . . It seems that the constitutional framers of Japan were s t i l l under the delusion of the old theory that the goodness of a Constitution consists in the entire separation of the executive and legislative branches of government. It i s ironic that the Meiji leaders ended up with the very kind of constitution the weaknesses of which they f u l l y understood and tried hard to avoid. While this did have a great deal to do with the state of contemporary p o l i t i c a l theory, i t was also partly due to a situation which confronts a l l revolutionary leaders. As long as they f e l t that the revolution which they had initiated would be threatened i f they were replaced by a government nominated by the newly created Diet, there could be no ultimate solution to conflict between the government and the Diet. In the early years, the f o l l y of handing over control of the executive to the opportunistic leaders of undisciplined parties, which represented the propertied class alone, was clear to a l l who sought to serve the national interest. But although It5 foresaw that party cabinets might become normal practice once responsible national parties developed, no one foresaw that deadlock between the government and the Diet would persist u n t i l pure party cabinets replaced ones which were not completely dependent on the Diet. One should remember that the English Parliament had existed for about six centuries before party cabinets became anything like normal practice. Almost throughout this period the king and his loyal servants had to gain the support of Members of Parliament in order to prevent k2 government from grinding to a halt. Gradually, as groups of loyal members came to band together to aid the king's measures through Parliament, the practice of choosing Cabinet ministers from among these groups became more common. Party did not have a connotation of opposi-tion to government, but of loyal support for government, so that i t f i n a l l y became possible to have loyal party cabinets. It was entirely unrealistic for Okuma and Itagaki to advocate immediate party cabinets in Japan, long before parties that could govern had emerged. The adoption of party cabinets in Japan was more complex than in Britain, because British parties had their origin i n the House of Commons. They were groups of representatives who came together i n order to support a certain kind of government. In Japan, however, parties had their origin in the country, where they agitated for support in order to oppose the government in the Diet. Immediate party cabinets were impractical, because i t takes a long time for country-based opposition parties to become parliamentary-based governing ones. Collective or Individual Responsibility? Finally, one must consider the intentions of the Meiji leaders on whether cabinets were to be collectively or individually responsible. American historians tend to place a great deal of emphasis on the absence from the Constitution of an explicit statement that the Cabinet ministers were to be collectively responsible.^1 i t i s concluded that the intention was for each minister to be individually subject to appointment and dismissal by the Emperor. This is unlikely. Since the Meiji leaders intended the Cabinet rather than the Emperor actually to govern the country, they must have realised that i t would have to act as a unit. There i s ample evidence that they did. It was precisely because the work of the government had not been well coordinated that the Cabinet was f i r s t set up in 1885. The change was made because, as Ito s a i d : ^ Ministers had no direct o f f i c i a l relations with the Emperor and were under no responsibility for the great affairs of State. . . .By the said reorganisation, the Ministers of State were made each separately to bear his share of responsibility to the Emperor directly. Over them was placed the Minister President of State. The object of this change was, on the one hand, to give weight to the functions of Ministers of State and to press upon them a higher sense of their responsibility, and, on the other, to maintain the unity of the Cabinet and to avoid a l l complications therein. The statement usually cited to support the theory of individual responsibility i s ambiguous There i s no joint responsibility among them in regard to such matters fmatters within their respective competency!. For, the Minister President and the other Ministers of State, being alike personally appointed by the Emperor-, the proceedings of each of them are, in every respect, controlled by the w i l l of the Emperor, and the Minister President himself has no power of control over the posts occupied by other Ministers, while the latter ought not to be dependent upon the former. In some coun-tries the Cabinet i s regarded as constituting a corporate body, the Ministers are not held to take part i n the conduct of the government each one in an individual capacity, but joint res-ponsibility i s the rule. The e v i l of such a system i s , that the power of party combination w i l l ultimately overrule the supreme power of the Sovereign. Such a state of things can never be approved of according to our Constitution. But with regard to important internal and external matters of State, the whole Government i s concerned, and no single Department can, therefore, be exclusively charged with the conduct of them. . . . A l l the Ministers of State shall take united counsel, and kk none of them is allowed to leave his share of the business a burden upon his colleagues. In such matters, i t would of course be proper for the Cabinet to assume joint responsibility. The intention seems to have been that ministers are individually responsible for the internal matters of their own departments, but collectively so for more important internal and external matters that affect the nation as a whole. This is exactly what i s supposed to be the case in Britain, although the Meiji leaders did not know i t . The possibility that any single institution could pull down an entire ministry may have made a kind of collective responsibility l i k e l y : concurrent collective responsibility in a negative sense. The u n l i k e l i -hood that any single one could raise an entire ministry and the likelihood that each would gain control of as many Cabinet seats as i t could made the collective rise of cabinets improbable. Although a l l ministers would be appointed at the same time, they would not a l l be appointed because of the influence of the same institution. Cabinets would rise neither as a homogeneous group, nor as a disparate collection of individuals, but as representatives of the institutions which effected their appointment and they would be neither individually nor collectively responsible. Conclusion While i t was not the intention of the-Meiji leaders to give their country weak and sluggish government or to institutionalise the separation-of-powers ideology, they were unable to avoid doing so, because the sine  qua non of constitutional government according to almost a l l European theorists was that executives and legislatures should not only have separate and clearly defined functions, hut separate existences. The Meiji Constitution did envisage separate functions and existences for each of i t s institutions, and the Meiji leaders also expected the Cabinet to have a separate existence. The Diet was granted the power to reject increased budgets and legislation, and because the two Houses were made equally capable of doing so, the executive could be in a doubly d i f f i c u l t position vis-a-vis the Diet. But the actual separation-of-powers in Japan was far greater than in America. Not merely was the Privy Council given powers similar to those of the American Supreme Court, but each of the Armed Services were given powers of veto in their respective domains. But unlike in America, while the judiciary was granted independence, i t was not given the power of j u d i c i a l review. Its lack of veto power prevented i t from assuming the role of yet another of the separate institutions. The occupants of the Cabinet were to be trusted imperial advisers, i n i t i a l l y the Meiji leaders themselves, and the function of this institution, though not mentioned in the Constitution, was to assist the Emperor in his executive capacity. The belief seems to have been that deadlock between the various institutions could be avoided by making the Emperor sovereign in a l l matters, and having each institution, including the Cabinet, accountable to him alone. But because the Meiji leaders wanted to maintain the kG i n v i o l a b i l i t y of the Emperor, they did not want him personally invol-ved i n p o l i t i c s , but wanted him to act only on advice. Whose advice? Here the intentions of the drafters of the Constitution are unclear, but circularity seems to be involved. On budgets and laws the Emperor would act on the advice of the two Houses of the Diet (provi-ded they could agree with each other); on the interpretation of the Constitution and the ratif i c a t i o n of foreign treaties he would act on the advice of the Privy Council; on foreign policy he would be confronted by Army and Navy advisers, as well as the Cabinet and the Foreign Office; and on administration i n general he would be advised by the Cabinet. But i f the Emperor was to avoid taking sides in cases of p o l i t i c a l conflict and act only on advice, he would be unable to exercise his theoretical sovereignty to resolve inter-institutional deadlock, and the separation-of-powers would- remain. If the Cabinet became the Emperor's major adviser and other institutions worked through i t , the power to make and break the Cabinet as well as affect i t s composition would indicate to whom the Japanese executive was i n practice responsible. It would also t e l l us whether parliamentary theory or separation-of-powers theory promotes the best understanding of prewar Japanese p o l i t i c s . To know one must examine practice. If the Emperor appointed and dismissed cabinets solely because of actions and wishes expressed by the Lower House of the Diet, then in practice the Cabinet was responsible to the House alone and there was no separation-of-powers. If, on each occasion, he considered the actions and wishes of a number of institutions, then the Cabinet was in practice responsible to no single one of them and there was a de  facto separation-of-powers. If he appointed and dismissed entire ministries solely because any one institution used i t s veto power, then in practice there was "concurrent collective responsibility." h8 CHAPTER 2 AN EMERGING SEPARATION-OF-POWERS, I89O-I9OO By the end of the f i r s t decade of constitutional development i t had become clear that relationships among the institutions established in I89O were typical of separation-of-powers systems. The Cabinet regularly consisted of a f a i r l y even spread of representatives of the new institutions, although most ministers s t i l l had contacts with more than one of them. It was not yet possible to link very many ministers with the influence of particular institutions, because only the Lower House of the Diet was really prepared to use i t s veto power. A l l cabinets were appointed by the Meiji leaders themselves, but their freedom to make these appointments gradually became constrained by the need to retain the support of institutions that could, and in the case of the House, did exercise their veto power. Because the GenrS found i t d i f f i c u l t to appoint ministries of their choice, the Cabinet was not quite responsible to them in the positive sense. And because the Lower House of the Diet sometimes needed help to unseat ministries, the Cabinet was not quite responsible to i t in the negative sense. Background By the early l890's, the national mood of infatuation with things Western that began soon after the Meiji Restoration in 1868 had subsided, and a reaction had set i n . The Meiji leaders had become disenchanted with Liberalism, i f indeed they ever embraced i t , and began to devote a l l 49 their efforts to economic development. The promulgation of the Constitution was intended both to demonstrate to the Western Powers that Japan was a "c i v i l i z e d " country and to prevent a re-occurence of the samurai and peasant rebellions that had rocked the state in the two preceeding decades. The latter consideration was perhaps less important, partly because the Army had attained sufficient strength to make the suppression of rebellion a much easier task than before. The leaders hoped to consolidate p o l i t i c a l authority, expand the economy, and use caution i n their dealings with foreign powers. The opposition to the government, which in the l870's and early 1880's had included large sectors of the peasant and samurai classes, had subsided. From 1884 u n t i l the promulgation of the Constitution i n I89O there was a period of comparative calm. The parties, which had been active agitators i n the country u n t i l the government promised in l88l to establish a representative assembly by the end of the decade, decided to wait for the promulgation of the Constitution before reorganising the opposition. Although they were somewhat disappointed by the role the Constitution gave them, at least they had a legal way of expressing themselves, and would try to exploit i t to the f u l l . They rapidly reorganised, but largely without their more radical peasant members, whom the Constitution excluded from the franchise. The two major party groupings, the Liberals and the Progressives, were loosely organised round their central figures, Itagaki Taisuke and 50 Okuma Shigenobu, both leading members of the Meiji oligarchy who had been excluded from the inner core. The Liberals were largely the party of the landowners, the Progressives of the financiers and urban capital-ists as well as of the intelligensia. Although the l890's have usually been regarded as years of intense struggle between the landowner and capitalist classes on the one hand, and the feudal oligarchy which was in charge of the state, on the other, the institutional context in which this struggle took place has not always received sufficient emphasis. George Akita's recent study, Foundations of Constitutional Government in Modern Japan, 1868-1900, has done a great deal to redress the balance. Acquaintance with this important work w i l l be assumed in this chapter, which concentrates almost entirely on the reasons for the Cabinet's composition as well as i t s rise and f a l l . I intend to provide a better theoretical understanding of early Meiji institutional development, whereas Akita emphasises the details of .what actually happened. The most important theoretical point to make about the years I89O-1900 i s that because the Genro were "almost" the sole creators of the Cabinet, the Japanese executive was "almost" responsible to the GenrS in the positive sense. And because the Lower House of the Diet was "almost" the sole breaker of the Cabinet, the executive was "almost" responsible to i t in the negative sense. But because neither the Genro nor the House can be regarded as the sole influence behind Cabinet-making 51 and-breaking, separation-of-powers theory explains what actually happened more accurately than parliamentary theory. Cabinet-Making As was expected, the Meiji leaders did not sacrifice their i n s i s t -ence on the i n v i o l a b i l i t y of the Emperor. They did not allow him to make p o l i t i c a l decisions and to become controversial. The Emperor appointed his ministers only after he had consulted the Meiji leaders themselves. The Genro, rather than the Emperor, became the controversial "institution" responsible for the appointment of Ministers. While in theory ministers remained solely Imperial appointments, in practice the Emperor only appointed men of whom the Genro approved. Responsibility to the Emperor was a f i c t i o n , because, as a contemporary remarked:^ (T)here i s not a single instance on record of the Emperor Mutsuhito taking any State matter into his own hands, independ-ently of the Ministers of State. . . . But in practice i t is generally understood that the outgoing Minister President advises the Emperor as to his successor, or else the Privy Council or an informal meeting of the so-called 'elder-states-men' decide who shall take the responsibility of a new adminis-tration, and advise the Emperor accordingly. The Emperor never appointed a Prime Minister u n t i l the Genro had held a formal Genro kaigi (Conference of elder statesmen), and agreed' on a single nominee. Yamagata was unanimously, although rather inform-ally , chosen in 1889, while Matsukata's selection i n I89I resulted from a formal Genro kaigi (Conference of elder statesmen), the f i r s t to be 52 publicly recognisable. Those who attended these Conferences were usually Ito Hirobumi, Yamagata Aritomo, Inoue Kaoru, Saigo Tsugumichi, Kuroda Kiyotaka, Yamada Akimasa, Matsukata Masayoshi, and Oyama Iwao.^ A l l had occupied high positions in the government for years before 1890. A l l the Prime Ministers of the 1890's were chosen by a Genro kaigi, and only in the case of the second Matsukata, second Ito, and f i r s t Okuma ministries did any institution other than the Genro wield an important influence. In each case this institution was the Lower House of the Diet. The one of their number the Genro considered most able to lead the government in 1892 was the one among them most favourably disposed towards compromise with the Lower House. His predecessor's use of election interference to change the composition of the House had back-fired, and the majority in the House was as hostile as ever. There was general agreement among the Genro that the person mainly responsible for the drafting of the Constitution should have an opportunity to try his way of "making constitutional government work." The Lower House must therefore be seen as an important force behind Ito's appointment as Prime Minister in 1892.3 But i t was far from being the major force: "The Genro made a united effort in organising the Cabinet," in an attempt to prevent the House from taking undue advantage of the new Prime Minister. In spite of their differences, both Ito and Yamagata served in the same government, so that a united Genro could be in the strongest position 53 vis-a-vis the House.4" The success of this ministry i n passing i t s budgets by means of an agreement with the Jiyuto, the sole price of which had been to reward this party's leader, Itagaki Taisuke, with a seat in the Cabinet, completely vindicated Ito's strategy. Even before the f a l l of his govern-ment, an agreement between Matsukata Masayoshi and the leader of the other major party in the House, Okuma Shigenobu, had been formed. Matsukata's appointment i n 1896 was not simply the result of his previous prime ministerial experience and favour with the Genro, but of the l i k e l i -hood that his agreement with a major party in the Lower House would make his government as great a success as i t s predecessor. The House's influence on Matsukata's selection i s underlined by his consultation of Okuma over the composition of the Cabinet even before i t had been formed. Ito had only admitted Itagaki into the Cabinet after a f u l l three and a half years of office. Okuma was in the important position of Foreign Minister from the beginning.5 Of Ito's appointment following the f a l l of Matsukata's second mini-stry, a contemporary British observer wrote:^ Resistence to the parties combined resulted in a deadlock, alliance with either of them threatened to strand the ship of state owing to the incompetence of the crew. . . . In 1898, Marquis Ito made an attempt to win the country back to non-party government and efficiency by forming an independent Ministry in defiance of the Liberal Jiyuto"? demands. But the result was a dismal fail u r e : the ministry lasted only five months. 54 The f i r s t Okuma government was mainly a Lower House creation. The two main parties had formally united to form the Kenseito (Consti-tutional Party) and had gained a massive victory in the election of 1898. It was impossible for It5, or any non-party Prime Minister, to keep "the ship of state" afloat. Although the appointment of Okuma, the Kenseito leader, was made at a formal Genro kaigi, and Yamagata was particularly hostile to the idea, i t was f e l t by a l l the Genro that there was no alternative. None of them was prepared to take on the premiership in the face of a united opposition in the Lower House.7 Although the Okuma government was largely the creation of the Lower House, i t did not result entirely from their influence. As Hackett correctly pointed out!: "In reality the Cabinet represented three distinct forces: one, the old Jiyuto members; another the old Shimpoto members, and the third, the service ministers." 0 Both Service Ministers were guard-ians of Genro interests: the one was Genro Saigo Tsugumichi, the other Yamagata's chief protege, Katsura Tar5. Oyama Iwao and Kuroda Kiyotaka had insisted that Saigo and Katsura remain in the Cabinet.9 The second Yamagata ministry came to power without the influence of any institution besides the Genro, although the Genro were not united behind him. Ito was in China at the time, and the other Genro held their meeting as soon as they could;, because they knew that he would not agree to Yamagata's appointment, but would have to accept a f a i t accompli.^ Because the choice of a new Prime Minister was always the subject of the greatest controversy at the time of each Cabinet change, the 55 symbols in Appendix I, which refer to the major influences behind the rise of ministries, actually only indicate the influences determining the selection of Prime Ministers. Because Prime Ministers never had accompletely free hand in forming their cabinets owing to the need to include institutional representatives, the Appendix is not strictly-accurate. But i t is more revealing to regard the influences respon-s i b l e for the selection of Prime Ministers as the ones raising whole ministries. Not to do so would mean that almost a l l institutions would have to be l i s t e d each time, and the real institutional struggles that took place over the appointment of Prime Ministers would be concealed. The tussel between the Lower House and the Genro over Cabinet-making in the 1890's would not receive sufficient emphasis. Because the Genro's freedom of appointment was on a number of occasions constrained by the need to nominate men who could win the support of the Lower House, none of the cabinets formed on these occasions can be regarded as s t r i c t l y responsible to the Genro. Neither the Genro1 nor the House exercised the entire power of appointment, and the Cabinet was s t r i c t l y speaking separate from both the GenrS and the House. The cabinets raised by both institutions i n the 1890's were: the second Ito, the second Matsukata, and the f i r s t Okuma ministries. But the remaining cabinets were Genro creations, and parliamentary theory correct-ly describes them as responsible to the Genro i n the positive sense. Cabinet-breaking was a different matter. Although the Emperor remained theoretically responsible for dismissing ministers, and although 56 the Genro also advised him on the use of this prerogative, the i n a b i l i t y of a Prime Minister to continue in office in the face of any in s t i t u -tion's uncompromising use of i t s veto power, made that institution the sole breaker of the Cabinet. But the f a l l of a few governments in the 1890's was due to the simultaneous use of their veto power by more than one institution. Because in almost every such case the Lower House was the sole necessary influence behind the government's demise, parliament-ary theory describes Cabinet-breaking more accurately than separation-of-powers theory. Cabinet-Breaking Although contemporary observers did not say in so many words that the Lower House of the Diet was the major institutional force respon-sible for the f a l l of icabinets in the 1890's, the fact remains that each and every one of them resigned because of i t s i n a b i l i t y to get what i t wanted from the House. This was usually passage of the budget in an unamended form. The government's need for expanded budgets each year to further economic development made i t s right to enact the previous year's budget should the House f a i l to give i t s consent, worth very l i t t l e . Throughout the l890's, the passage of the budget remained the most hotly contested issue between the Cabinet and the Lower House. The next most important issue was foreign policy, the House calling for a more vigorous foreign programme, the Cabinet, under the influence of the Genro, favouring moderation. In I89I the f i r s t Yamagata cabinet resigned because the House passed 57 an amended budget, which meant the "demolition of the government's financial p l a n . " ^ Yamagata had tr i e d to cudgel the House into compli-ance by issuing a new regulation which forbade the members of one party to communicate with members of another. But when this method failed, he resorted to corruption and was ultimately able to purchase s u f f i -cient support for a compromise.-^ He was so disgusted with the parties and the need to compromise with them, that he decided to step down because of the "extreme d i f f i c u l t y in making constitutional government function smoothly."^3 Yamagata's successor, having unsuccessfully tried by means of wholesale election interference to obtain a Diet which would pass i t s budgets, f i n a l l y resigned because of Cabinet disunity over the consequences of this interference. The Lower House had passed a reso-lut ion calling on the government to take responsibility for the election interference and resign.^ 4' Although Matsukata defied the House for a while, the members of his cabinet could not agree on the policy to be adopted towards the House. F i r s t , Shinagawa Yajiro, the Home Minister responsible for the election interference, was made to resign, because he was the main recipient of the House's c r i t i c i s m . ^ Then Mutsu Munemitsu, who had been a member of the House, resigned to dissociate himself from the government's policy.-^ They were replaced by men with close relations with the leaders of the two main parties: Soejima Taneomi and Kono Togama, indicating, as Akita points out, that "the government was predisposed to revert to compromise."17 58 When Soejima worked out a compromise with the Jiyuto to pass a supplementary budget, the other members of the Cabinet forced him out of office. Matsukata then indicated his intention to resign, but was persuaded by his colleagues not to. K5n6 Togama, Okuma's old friend, was made Home Minister to demonstrate to the House the government's good faith. But the Cabinet was f i n a l l y forced to resign when those ministers who opposed Matsukata's conciliatory policy towards the House, Takashima Tomonosuke and Kabayama Sukenori, refused to attend Cabinet meetings. Although the Army and Navy stepped in and indicated their refusal to find replacements for Takashima and Kabayama, who were the two Service Ministers, should Matsukata not go through with his intention to step down, the institutional pressures primarily responsible for Matsukata's predicament came from the Lower House. The policy of election inter-ference was the direct cause of the Cabinet disunity that ultimately brought Matsukata to the point of despair. The action of the Army and Navy only added the f i n a l touch to a hopeless situation.-'-® Uyehara also regarded the f a l l of the f i r s t Matsukata cabinet as the result of pressures from the Lower House': "Public confidence, owing to the inter-ference in the election, was completely shaken, and i t [the Cabinet] was forced to resign about two months after the end of the session."^ Okuma's biographer wrote: "The Opposition had succeeded in throwing out two ministries in two years--a good record for the fighting strength of p a r t i e s . " 2 0 59 The reasons behind the f a l l of the second Ito cabinet were equally-complex, but the major institutional force involved was once again the Lower House. In the f i f t h session of the Diet (November l893)> "the government was severely c r i t i c i s e d by the House, under the leadership of the Progressives, for i t s weak foreign policy. Ito, fearing that this might jeopardise negotiations that were being conducted with Britain, decided to dissolve the House. The new House then passed a vote of no-confidence in the government for dissolving without proper - Pi cause, and Ito dissolved the House a second time. J-Although Ito was having considerable .'difficulty in retaining the House's confidence over foreign policy, he was able to pass a l l his financial legislation because of a working agreement with the Jiyuto, which also refrained from taking part in the onslaught on the govern-ment 's foreign policy.^2 The maintenance of the agreement with the Jiyut5 ultimately required Ito to reward that party by giving i t s leader, Itagaki Taisuke, the important seat of Home Minister. When vacancies at the Foreign and Finance Ministries occurred, Ito found himself i n d i f f i c u l t i e s . The Genro Inoue Kaoru and Yamagata Aritomo recommended that Okuma Shigenobu and Matsukata Masayoshi be brought in as Foreign and Finance Ministers, and Ito made no objection. But Itagaki flatly-refused to accept the entrance into the Cabinet of the leader of the party which had spearheaded the attack on the government's foreign policy i n the House, and threatened to resign. Matsukata, who had been 6o in close consultation with Okuma, refused to accept the position of 23 Finance Minister unless Okuma became Foreign Minister. Akita concludes: "Ito, unable to f i l l the positions of foreign minister and finance minister without incurring the loss of Jiyuto support, resigned."^" The influence of the Lower House in the demise-.of the remaining governments in the 1890's, except the f i r s t Okuma one, was more direct and requires less detailed analysis. In each case, the resignation followed closely on some action or other taken by the House, and the causal link between them is easier to discern. The second Matsukata cabinet, which i t w i l l be remembered came to power as a result of an agreement between Matsukata and the Shimpoto, f e l l with the rupture of that agreement. Matsukata had promised to further certain Shimpoto aims, for example, to extend freedom of speech and of the press, but because the party f e l t he was not sufficiently responsive to their wishes, dissatisfaction among party members appointed to high C i v i l Service offices increased. When Matsukata dismissed the people concerned, Okuma resigned, and the House introduced a motion of no-confidence in the government. Matsukata dissolved the House, and three days later resigned. 2^ ItS's third ministry, which attempted to make constitutional govern-ment work without reliance on a major party in the House, was one of the most short-lived i n the entire prewar period. Financial problems resulting from high levels of government expenditure needed in part to support the Sino-Japanese war of 1895 made i t necessary to find new 61 sources of revenue. Ito turned to the landowner class for assistance. But this class was the one most strongly represented in the Jiyuto, which was in the forefront of the House's overwhelming rejection of Ito's Land Tax B i l l . To the House this B i l l was a matter of shikatsu ( l i f e and death). Ito replied hy dissolving the House, and within only twelve days the two major parties had combined to form a new party whose major pur-pose was to secure party control of the Cabinet. It called i t s e l f the Kenseito (Constitutional Party). Ito then requested permission from the other Genro to form a new "loyal" p o l i t i c a l party, but when they refused, P6 he resigned. The f a l l of the government that came to power because of the union between the two major parties was primarily due toithe break up of that union. The immediate cause was a "Republican Speech" made by Ozaki. Yukio, the Education Minister, who alluded to the hypothetical possibi-l i t y that Japan might become a republic in millenia to come. Army Minister Katsura Taro, one of the guardians of Genro interests in the Cabinet, the House of Peers, and the Privy Council, seized on the issue in an attempt to pull down the government. Ultimately Okuma was forced to sacrifice Ozaki, but this led to a struggle between the two factions (the old Jiyuto and old Shimpoto) of the Kenseito over Ozaki's successor. Okuma was adamant about appointing a member of his old party, and the old Jiyuto decided to revert to their former status as an independent party and retained the name of Kenseito for themselves. 62 Okuma's own faction, which now called i t s e l f the Kenseihonto (Orthodox Constitutional Party), had more seats in the House than any other party, and Okuma tried to reconstitute his Cabinet. He intended to use a general resignation to r i d his government of the opposition party members, but the GenrS stepped in- and appointed Yamagata before he could come up with a new l i s t of ministers. The action of the Genr5 was crucial, and was a necessary condition of the government's f a l l . But so was the break up of the Kenseito, which gave the Genro their chance, and the influence of the House must be placed above that of the Genro. Okuma no longer commanded the strong position vis-a-vis the House that had brought him to power, and this fact enabled the Genro to act as f i n a l executioners of an already condemned ministry.^ Yamagata's second cabinet, because of i t s working agreement with the Kenseito (old Jiyuto) and use of "gold p i l l s " to win votes in the Diet, was at long last able to pass the Land Tax B i l l . But the members of this party, once they saw that Yamagata had no intention to reward them with seats in his cabinet, terminated the agreement. Yamagata responded by tendering his resignation, although he agreed to remain in office u n t i l the suppression of the Boxer Rebellion in China. Hackett concluded: "Weary of governing, tired of the constant discontent expressed by the parties, and irked by the necessity of placating the opposition to get measures passed in the Diet, he withdrew for the last • time from service i n the Cabinet. 6 3 The Lower House's a b i l i t y to thwart the w i l l of any government by denying i t supply gave to the House the power to unmake cabinets. Although this was not enough to satisfy the parties in the House completely, i t was much more than the Meiji leaders had envisaged at the time of drafting the Constitution. A British observer appre-ciated the problem better than most: (T)he Opposition, i f i t could not have i t s way, could at least prevent the government from following i t s own course, with the result that a l l progress was stayed, and a veritable deadlock ensued. . . . The clan oligarchy . . . had yielded the right  of consent (author]s i t a l i c s ) to the Representatives of the people and withheld substantive powers over the Adminstration. The old dualism of the Stuart period or of the reign of George III was repeated in Japan, in spite of the plain teachings of English history. Unable to raise ministries of their choice, members of the Lower House united to bring down those raised largely by the Genro, in the only way they could: through the exercise of their veto over increased budgets. Unfortunately for themselves, Dietmen assumed a consistently antago-ni s t i c attitude towards the government, and did not know how to govern when given the chance, as.they demonstrated in the case of the f i r s t Okuma ministry. This habit of opposition to governments rather than cooperation with them persists to this very day among certain opposition members. It was succinctly described by Uyehara:2^ Most politicians . . . in Japan do not come down to the House intending to support the government and to pass certain govern-ment measures, like the majority in the House of Commons in England, except when the leading party or parties have been included for some reason or other to co-operate with the govern-ment; but they attend in order to ply the government with questions, 6k to ferret out faults . . . and so to shake public confidence. This is one of the most effective methods of demonstrating the power of the representative body of the people that can be adopted under the present Constitution. Using methods like these, the Lower House managed between Cabinet changes to effect a number of ministerial changes, although not a l l of their influence in this respect resulted from the employment of negative tactics. The replacement of''Shinagawa in March 1892 was, however, a classic case. He received, .more than any other Cabinet minister, the House's most vigorous criticism of the government's policy of election interference i n 1892.30 Mutsu Munemitsu's resignation at the same time was an expression of his dissociation from that policy. The Cabinet could not include a man who was sympathetic to the House and at the same time expect to attack that institution with impunity.31 Yamagata's decision to leave Ito's second cabinet resulted from his dislike of the compromises with the parties, without which government would have ground to a halt.3^ Goto Sh<5jir6" resigned from the same cabinet when charges of corruption, originally directed at and resulting in the impeachment of Hoshi Toru, spread to include him.33 Nomura Yasushi was edged out of his seat of Home Minister to make way for Itagaki, this being the only way Ito could reward the Jiyuto adequately for i t s cooperation. Finally, Agriculture and Commerce Minister Enomoto Takeaki was made to resign from the second Matsukata cabinet, after he was heavily c r i t i c i s e d in the House as the Minister responsible for the Kodoku jiken, an incident i n which many people died of copper pollution, caused by tailings from a mine which poisoned irrigation water.3^ 65 Although the House was not yet able to have whomsoever i t wanted appointed to the Cabinet, i t was able to remove from office anyone who incurred i t s wrath, as these and many other examples indicate. In addition to the House's a b i l i t y to remove undesirable Cabinet ministers, i t could remove entire ministries. The House's entire responsibility for the dismissal of a number of cabinets made these cabinets negatively responsible to i t . A l l governments except the f i r s t Matsukata and f i r s t Okuma ones were dismissed solely because of the House's influence. The f a l l of the two exceptions also resulted from the predominant pressure of the House, but in both cases another institution was the f i n a l executioner. The Army prevented the Matsu-kata government and the GenrS prevented the Okuma government from saving the situation. But because the Army and the Genro only added the f i n a l touch, the Cabinet in the 1890's should be seen as negatively responsi-ble to the Lower House of the Diet. The difference between Cabinet-making and--breaking also affected the development of collective responsibility. Ministers could be appointed individually but particularly towards the end of the decade cabinets usually resigned en bloc, and ministers were dismissed as members of teams. From Individual Responsibility to Institutional Representation During the l880's, the Meiji leadership had been able to control government appointments without any legitimate interference from the opposition. Ministries usually consisted of members of the inner core, 66 and when changes took place, usually only changes of portfolios were involved, and these were not too frequent. It is not surprising that during the f i r s t years under the Constitution, the practice of ministers coming and going only one or two at a time was continued. In the language of parliamentary theory, individual responsibility of ministers was the rule, although i t was not yet possible to decide to whom they were responsible. For example, the f i r s t Yamagata cabinet was, more than any other, simply a collection of individuals whose fortunes were only to a slight degree inter-dependent. Only two of i t s members never had seats in the preceding Kuroda cabinet, and when i t was succeeded by the f i r s t Matsukata cabinet, only the occupant of the Prime Minister|s Office changed.35 It is of course true that the number of ministerial changes per ministry cannot alone indicate the extent to which ministers come and go as individuals, because men who are more able to prolong the l i f e of the team may continually be brought in for this reason. Nevertheless, the unusually large number of changes that took place during the lives of the f i r s t Matsukata and second It5 ministries, largely for reasons unrelated to prolonging the l i f e of the whole,3^ confirms that ministers were not yet dependent on one another for their existence, but on them-selves as individuals. The second It5 cabinet saw more than a score of ministerial changes, and only the appointment of Itagaki Taisuke, the leader of the Jiyuto had anything to do with keeping the team in office for a while longer.37 The number of ministerial changes per cabinet 67 declined sharply after the f a l l of this government. Seven of the original ten members of the second Matsukata cabinet remained in office throughout i t s l i f e . Wine of them did i f one considers that his govern-ment f e l l soon after the resignation of Okuma Shigenobu, the leader of the ShimpotS, and Hachis.uk.a-' .• Mochiaki, which ended the government's alliance with that party. A l l but two of the original members of the third Ito ministry, a l l of those of the f i r s t Okuma one, which f e l l only two weeks after the resignation of Ozaki Yukio, and the entire second Yamagata ministry, saw their f u l l term of office. Although the f i r s t two Prime Ministers of the period merely assumed the leadership of cabinets formed by their predecessors,,the fate of the f i r s t Matsukata cabinet provided the f i r s t instance of what would become standard practice by the end of the decade. When the Prime Minister resigned, a l l his colleagues followed suit, and only one of them accepted a seat in the next cabinet. It took a few years for a fixed pattern to develop. Three members of the second It5 cabinet served in i t s successor, and so did two members of the f i r s t Okuma cabi-net. But when the third It5 cabinet resigned, for the f i r s t time, the two ministers who remained in office were those of the Army and Navy. This refusal of the Service Ministers to move with the times marked the beginning of a new trend: the Service Ministers became f i r s t and foremost, not members of the team, but representatives of the Army and Navy, who regarded themselves responsible to the Army and Navy alone. Both of Okuma's Service Ministers as well as Yamagata's Navy Minister i n his 68 second cabinet were the sole members of their ministries to serve under their successors. Separation-of-powers institutional forces had begun to affect the composition of the Cabinet,and prevent the development of i t s members into a tightly-knit team. If some members were being appointed primarily because of their ties with the Armed Services, others owed their appointment to ties with yet other institutions. A Cabinet con-sisting of representatives of the various institutions could not be expected to hold together in times of inter-institutional conflict. Nor could i t s members regard their existence as mutually dependent, because the existence of each ultimately depended on the institution he represented. If complete collective responsibility was therefore unlikely to develop, individual responsibility could never become the rule either, because ministers were appointed not merely for their individual qualities, but because of their membership in one or other of the veto-possessing institutions, whose cooperation the Cabinet required. After the f i r s t decade of constitutional government, cabinets continuedtorise and f a l l more or less as teams. The number of reshuffles was never large, and rarely indicated more than do reshuffles in Britain, that when the Prime Minister feels that accepting collective responsi-b i l i t y for the actions of a colleague w i l l endanger the l i f e of the whole, i t is better to replace the offending minister. 6 9 The tendency for changes in Service Ministers to coincide only occasionally with Cabinet changes continued throughout the prewar period and revealed that these ministers never came to think of them-selves primarily as members of a governing team. Although non-service ministers also remained primarily representatives of one or other of the separate institutions, few of them remained in office after the f a l l of a cabinet. When one did, like the representative of the Kenkyukai, the largest faction i n the Upper House, Oki Enkichi, who remained in office after the f a l l of the Takahashi cabinet i n 1922, i t was quite exceptional. Two contradictory forces were responsible for t h i s . The one, because ministers were appointed primarily as representatives of one of the separate institutions, worked against the development of solid-arity among the Cabinet members and of complete collective responsibility. The other, because the business of government requires coordination and unity among government programmes, tended to cement Cabinet members into teams, which came to have corporate identities of their own. The end product of these two antagonistic forces was not so much that min-isters came and went individually, as in the case of the Service Ministers and the occasional non-service minister, but that whole cabinets had only very brief existences. Ministers were appointed as representatives of institutions, and then came to acquire corporate identities. When inter-institutional conflict became severe, the only solution was to resign 70 en bloc, because their newly acquired corporate identities prevented each from simply reverting to his role of representative of one of the institutions. Because this pattern showed l i t t l e variation under the Meiji Constitution, no more w i l l be said about the relationship between collective responsibility on the one hand, and the timing of a minister's appointment and'the length of his term of office, on the other. The Qualifications of Ministers and the Sharing of Cabinet-Making Power That these new forces had come to exert the influences described is revealed in the composition of cabinets. Even i f the hi s t o r i c a l evidence does not always indicate specifically that a member was appoint-ed because he possessed a particular qualification, the assumption w i l l be made that, even though he may not have been appointed primarily for this reason, an unintended consequenceoof his membership in an i n s t i -tution i s that he w i l l to some extent regard himself under constraints from i t and be partly responsible to i t . A classification of the quali-fications of Cabinet members can therefore be used as evidence for the extent to which new institutional forces were operative. This reveals an incipient separation-of-powers. Appendices I and II show that in the l890's the Genro and their proteges a l l but monopolised Cabinet seats. Only the f i r s t Okuma cabinet never included a majority of them, but' this cabinet was not typical of the period. The only change that seemed to take place was that whereas in the early years representatives of the Genro were mainly 71 GenrS themselves, in later years these representatives tended to he mainly proteges of the Genro. For example, the second Ito cabinet, popularly known as a genkun naikaku (cabinet of veteran statesmen), at one time or another included a l l the Genro except Yamada Akimasa, whereas Ito's third cabinet included only three actual Genro, but four of Ito's proteges and two of Yamagata's. The same was true of the two Yamagata ministries, the f i r s t including five Genro and two of Yamagata's proteges, the second only three GenrS but four of Yamagata's proteges. The extent of Genro Cabinet-making power is also revealed by the fact that Okuma was the only non-Genr5 to serve as Prime Minister in this period. The steady increase in the number of Cabinet members who had contacts in the Lower House or who were actually party members shows that this institution was rapidly coming to acquire a share of Cabinet-making power. Akita concludes that "within a decade no government could be formed without at least their covert cooperation."3® The f i r s t Yama-gata cabinet, for example, included Iwamura Michitoshi, the brother of Hayashi Yuz5, a prominent Jiyuto leader, and Goto Shojiro, a founder of the earlier Jiyuto of the l880's and leader of the Daido Danketsu, an organization of the late l880's composed of ex-party members. It had urged the government to resist the Western powers and to increase freedom of speech and assembly. Goto entered the Kuroda cabinet to placate the party politicians whom the government had to face in the f i r s t Diet. 72 Just before the opening of the Diet, Iwamura was replaced by Mutsu Munemitsu, who had closer relations with more party members and was subsequently elected to the House. The next cabinet not only retained the services of Goto and Mutsu, but soon added Soejima Taneomi, who was close to both Okuma and Itagaki, and Kono Togama, an old friend of Okuma. One should not exaggerate the influence of the House in effecting these appointments, a l l of whom were only acquaintances of party members, not party members themselves, but, had the government not been having d i f f i c u l t y in obtaining the House's cooperation, few of them would have been considered suitable for Cabinet office. The second Ito cabinet, however, not only included men with con-tacts in the House such as Goto, Mutsu, K5no, and Saionji Kimmochi, but specifically brought in the leader of the Jiyuto, Itagaki, in order to reward that party for the cooperation i t had given the government.39 Then, even before the formation of the second Matsukata cabinet, Okuma was approached and consulted over the cabinet's composition, in order to cement an agreement for his party to cooperate with the new ministry. Finally, the f i r s t Okuma ministry represented the climax of the House's efforts to gain control over Cabinet seats, and a l l i t s c i v i l i a n members came from the majority party. In reaction to this disproportionate share of seats for the House--i f one remembers that the House was only one of the various i n s t i t u t i o n s — the second Yamagata cabinet included only Sone Arasuke, who had taken a 73 leading part in setting up the Kokumin Ky5kai (a pro-government party i n the House) in 1 8 9 2 , and had been elected to the House once. Yamagata had been able to negotiate an agreement with the Kenseit5 without having to pay the price of including any of i t s members.^0 Nevertheless, the gradual acceptance of the practice, usually indicated by the presence of major party leaders in the Cabinet, that government without the support of the House was impossible, confirms the development of the House into one of the separate veto-possessing institutions. Other institutions were also i n the process of acquiring a share of Cabinet-making power, particularly the Armed Services, although the dominant contenders i n the 1 8 9 0 ' s remained the GenrS and the Lower House of the Diet. The Armed Services had always possessed a good deal of influence, because, in the early years at least, i t was not easy to distinguish their leading members from the Genro, a l l of whom, except Ito Hirobumi and Inoue Kaoru, who were members of the C i v i l Service, were high ranking military officers. Nevertheless, once younger men, who were frequently at odds with the Genro, obtained Cabinet positions, i t became possible to distinguish the GenrS and the various Services as separate institutions with which the Cabinet would have to contend. For example, non-Genro military men in the f i r s t Yamagata cabinet were Enomoto Takeaki and Kabayama Sukenori, both of whom, together with Takashima Tomonosuke, served in the next cabinet. A l l the other cabinets of this period, except the third It5 and f i r s t Okuma ones, contained one or two non-Genro military men, indicating the development of a gradual i f almost imperceptible separation between the GenrS and the Armed Services. This i s confirmed by developments over the passage of time. In the early years non-Genro military men did not serve exclusively as Service Ministers. For example, Enomoto Takeaki and Takashima Tomonosuke were at one time Education and Colonies Minister respectively. By the end of the decade, however, non-Genr5 military men could only be found in the portfolios of Army and Navy Minister, and only Army men could be found in the former, and Navy-men i n the latter. The very large number of ministers whose qualifications included service in the upper levels of the C i v i l Service indicates that this institution also came to have an identity of i t s own as one which would take part in the struggles over Cabinet seats. Both the members of the f i r s t Yamagata cabinet who did not s i t in i t s predecessor had been Vice Ministers in the departments they were appointed to head, and Yoshikawa Kensei, although a Yamagata protege, had been Vice Minister in another department. A l l the men who became Cabinet ministers for the f i r s t time during the f i r s t Matsukata ministry, except Takashima, had been ministers in the pre-cabinet period. A l l were members of the Privy Council, the implications of which are spelt out below. Of those whose f i r s t Cabinet experience occurred during ItS's second ministry, Watanabe Kunitake had been Vice Minister of his department, and Ito's and Yamagata's proteges were a l l c i v i l servants of long 75 standing. The same pattern i s revealed hy the newcomers in the next cabinet. Nishi Tokujiro, Hamao Arata, and Kiyoura Keigo, came to head the departments in which they had held high positions, while Hachisuka Mochiaki and Yamada Nobumichi were also in government service. While these qualifications were crucial, that Kiyoura was also a Yamagata protege and a Peer, and Hamao and Hachisuka were also Peers, was not unimportant. Although the f i r s t Okuma cabinet also included men with experience in the C i v i l Service, for example, Okuma himself, i t s c i v i -lian members were there primarily because of their party membership. The second Yamagata cabinet included no one, except the Navy Minister, who had not served i n some previous cabinet. The fact that many of these C i v i l Service Cabinet members were also Genro proteges parallels the lack of complete separation between the Genro and the Armed Services. But many of them were not Genro proteges and had been Vice Ministers: i t seems that a career in the C i v i l Service would on i t s own become a sufficient qualification for Cabinet membership. It must be emphasised that because there was so l i t t l e inter-institutional conflict during the l8 Q0's other than that between the House and the Genro, the changes that were suggesting the acquisition of separate purposes and existences by different institutions were almost imperceptible. In the next period these changes were to become more obvious, but they had their origins in the l890's and were revealed by a gradually more even distribution of Cabinet seats among each of the 76 institutions established in 1890. The increasing tendency for ministers to be appointed largely because of their ties with certain institutions could be expected to lead these institutions to impose gradual constraints on their representatives. The latter would be required to look after their institution's interests. Inter-institutional conflict became acute not only because the Genro lost control of other institutions, but because of the growing tendency for Cabinet members to belong to only one institution and to be o f f i c i a l l y regarded as i t s representative, as was the case with Genro-House conflict in the l890's. Because the Cabinet members of. this decade who were members of the Privy Council frequently had some other qualifications as well, i t i s not easy to t e l l which particular one won them their seats. It seems that the great preference shown for people with previous Cabinet experi-ence was legitimised by making these people Privy Councillors, and then alleging that they were chosen as members of the Privy Council. Many were chosed perhaps not simply because of past experience, but because off the original qualification that won them a seat in the Cabinet in the f i r s t place, usually a career in the C i v i l Service. Although many Privy Councillors resigned from the Council when they entered the Cabinet, past membership is not distinguished from f u l l membership in this period, in order to illustrate the kinds of qualifications that were becoming necessary for Cabinet office. Thel'first Yamagata cabinet contained only one Privy Councillor, Aoki Shuzo, but Aoki was also one of Yamagata's proteges. 77 As noted above, a l l the men whose f i r s t Cabinet experience was in the ensuing Matsukata government, except Takashima Tomonosuke, were Privy Councillors with ministerial experience before 1885. But only Oki Kyonin, who had been the Council's President, Tanaka Fujimaro, and Sano Tsunetami can possibly have been appointed for this reason. The second Ito ministry contained five Privy Councillors, a l l of whom had previous Cabinet experience, but i n no case did their Privy Council membership make much difference. Of the Privy Councillors appointed in subsequent reshuffles, Inoue Kowashi and Saionji Kimmochi probably owed a bit to the Council. A l l the original members except Matsukata of his second cabinet who had served in previous cabinets had at one time or another been members of the Privy Council, but a l l would probably have been appointed anyway. Nishi Tokujiro was the only other member of the Council to enter the Cabinet in reshuffles, and also the only one who was probably appointed partly because of this membership. It5's-third ministry also had considerable overlapping membership with the Privy Council. Ito himself, Saigo Tsugumichi, Nishi Tokujiro, and Saionji Kimmochi had a l l been Privy Councillors, but only the latter two seemed to have owed anything to the Privy Council. The Okuma govern-ment included only the Premier and Saigo, but both would have served anyway. The same applies to the Councillors in the second Yamagata cabinet: the Premier, Saigo, Kabayama Sukenori, and Aoki Shuzo. 7 8 It seems therefore that the Privy Council also had some share of independent Cabinet-making power in this period, because although many Privy Councillors were Genro or GenrS proteges, or even possessed other qualifications, not a few were Privy Councillors in the f i r s t instance and owed their Cabinet positions to this more than to any-thing else. Examples are: Oki Kyonin, Tanaka Fujimaro, Sano Tsunetami, Rishi TokujirS, and Saionji Kimmochi. Moreover, many of the Privy Councillors who never served in the Cabinet could hardly be described as lackeys of the GenrS. One should not be surprised that the process by which each institution freed i t s e l f from Genro control and acquired a separate existence and purpose was far from complete in ten years. Even in hi 1910 Uyehara observed that: (U)p to now there has been no serious conflict between the Cabinet and the Privy Council, as both have been and s t i l l are occupied by men of the same mode of thinking . . . and both are responsible to the Emperor and not to the Diet. . . . When the Cabinet Ministers in the course of time become responsible to the Diet, the friendly relations now existing between the Cabinet and the Privy Council may not continue. The f i n a l institution to acquire a share of Cabinet-making power was the House of Peers, although Uyehara's remark on the Privy Council i s equally applicable to the Peers in this period. While many members of this institution served in cabinets, very few did so primarily or even partly because of this membership. The f i r s t Yamagata cabinet included three members of the Upper House, but for none of them could 79 this have made much of a difference. The same applies to the three Peers in the next cabinet. But of the three in the second Ito ministry, Saionji Kimmochi probably found his Peerage as useful as his other qualifications. The second Matsukata cabinet included five Peers, of whom Hamao Arata and Hachisuka Mochiaki were appointed partly, possibly even mainly, because of their Upper House membership. Of the seven Peers i n the third Ito cabinet, Saionji, Kaneko Kentaro, Suematsu Kencho, and Ito Miyoji may have owed something to this fact, although a l l were Ito proteges. For Toyama Masakazu, membership of the Upper House was probably more crucial. The Okuma cabinet included only Saigo Tsugumichi, who would have served anyway, while none of the Peers in the second Yamagata ministry could have been appointed because of their Peerage. In a l l , only Hachisuka, Hamao, and Toyama provide evidence for any real Upper House influence on the composition of the Cabinet, or for the acquisition of an independent identity by the Peers. The Peers' Cabinet-making and -breaking power would become more evident as conflict with the Lower House increased, a development which was inevitable, because, UP as Uyehara pointed out: (T)he House of Peers and the House of Representatives have an entirely different composition and represent different communities and interests. Therefore conflict is more lik e l y than harmony; and conflict must end in deadlock, or the supremacy of one party over the other. 8o The influence usually cited as the one to which the Cabinet was in practice responsible, the two han (clans) of Satsuma and Choshu, has not been mentioned yet. Because a l l the Premiers of the decade, except Okuma, were from these han, Quigley wrote, "responsibility to the Emperor has, u n t i l very recently, meant essentially responsibility to such prominent clan statesmen as Yamagata, Ito, Katsura, and others."^3 It i s true that the majority of Cabinet members in the l890's were from the two main han, but this majority was not as great as is sometimes made out. Table 1, which is compiled from information in Appendices II and III, reveals a steady decrease in the number of ministers from the main han. Table 1. Proportion of Cabinet Members from the two main han, I89O-I918' Ministry No. of Sat-Cho. Ministers Total No. of Yamagata I 7 13 Matsukata I 8 17 Ito II 11 20 Matsukata II 7 13 Ito 'III 7 12 Okuma I 2 10 Yamagata II 7 9 Ito IV 3 13 Katsura I 6 ik Saionji I 3 12 Katsura II 3 9 Saionji II 2 11 Katsura III k 10 Yamamoto I 3 12 Okuma II 2 15 Terauchi 2 11 81 Moreover, i t is very d i f f i c u l t to say that anyone was appointed primarily because of his Sat-Ch5 background. People from these han seem to have been appointed mainly because they were able to acquire the other requisite qualifications sooner then people from other han, an advantage they gradually lost. So while the Choshu clique may have proved to be dominant in the Army and the Satsuma clique in the Navy, this was not necessarily because the major forces creating conflict were feudal, but partly because the dominant han were i n i t i a l l y more able to capture control of the new institutions, which were really to set the Japanese ruling class at loggerheads with i t s e l f . For the same reason, the major conflicts among the GenrS were not between Satsuma and Choshu men, but, because It5 was sympathetic to the Lower House, between Ito and Yamagata, the two leading Choshu  Genro. This reflected the institutional conflict between the Genro as a whole and the Lower House of the Diet. Although the forces d i v i -ding the ruling class were as much politico-institutional as feudal, i t is true that the rival r y f e l t between men from Satsuma and Choshu did on occasion make a difference. For example, Matsukata, who was not unsympathetic to the Lower House, would have been wiser to support Ito against Yamagata, rather than allow his Satsuma lineage prevent him from allying with a man from Choshu. The fact remains, however, that cabinets can hardly be said to have been responsible to these two han. Very few Sat-Ch5 men who were not Genro, proteges of Genro, or military 82 men became Cabinet ministers. The only two who did were Nishi Tokujiro from Satsuma and Sone Arasuke from Choshu. While the association with the main han was an important qualification for both, i t was by no means alone. Although each of the newly created institutions had begun to exert an independent influence more important then feudal ti e s , the major conflicts of the l890's were between the Genro and the Lower House, because the former managed almost to monopolise Cabinet-making power, the latter Cabinet-breaking power. Only men who had the confidence of the GenrS were made ministers by Genro Premiers in the f i r s t place, andtthose who could not subsequently win the confidence of the House were compelled to resign. The kinds of qualifications possessed by ministers, membership in one or more of the various institutions, did indicate, however, that the Genro never had a completely free hand i n making their appointments. Their choices were constrained by the need to find representatives of the institutions without whose consent no cabinet could govern. Because these institutions, except the Lower House, were largely under GenrS control i n the 1890's, the GenrS's discretion was much greater than i t would be once they lost this control. Conclusion After only one decade of constitutional development, there were clear indications that separation-of-powers forces were becoming more important than parliamentary ones. The Lower House of the Diet, over 83 which the Genro had no control, was i n constant deadlock with the Cabinet, which in turn was far from sole dependence on the House. The Genro were never really in possession of the entire power to raise ministries, not merely because they frequently had to appoint Prime Ministers who could come to terms with the House, but because there were signs of the need for Prime Ministers to include in their cabinets a broad spectrum of institutional representatives. The power to appoint cabinets was becoming something which no single institution possessed, and the Cabinet was becoming separate from each. Although the power of breaking cabinets was largely in the hands of the Lower House in the 1890's and cabinets were mainly responsible to the House in the negative sense, there were also indications that no single institution could on i t s own always pull down an entire ministry. Even negatively, the Cabinet was not s t r i c t l y speaking responsible to the House. Deadlock in the 1890's was not 'as serious as predicted: only the Genro and the Lower House of the Diet were in serious conflict with each other. But this was because the Genro had almost complete control of institutions besides the House. Once the Genro lost their a b i l i t y to dominate even these institutions, deadlock between the Cabinet and a l l institutions became the rule. A full-fledged separation-of-powers developed and imparted to prewar Japanese politics one of i t s most distinctive characteristics. 84, CHAPTER 3 A FULL-FLEDGED SEPARATION-OF-POWERS, 1900-1918 The most pertinent characteristics of Japanese institutional dev-elopment during the years 1900-1918 were the development of a Genro-Lower House compromise and the gradual decline of Genro power within that compromise. Both developments reflected an increasing separation-of-powers. Cabinet-making and -breaking power were both consistently wielded by more than one institution, and the Cabinet was more f u l l y separate from each than ever before. The increasing separation-of-powers must be held at least in part responsible for the prevalence during these years of phenomena like high levels of corruption and personality p o l i t i c s . In the l890's the Genr5 had been the Constitu-tion's main deus ex machina and had mitigated inter-institutional conflict through their proteges, who worked from within the various institutions for a coherent set of policies which they, the Genro, had determined. Directly, or indirectly, they had been able to control the Upper House, the Privy Council, the Army, the Navy, and the C i v i l Service, and were therefore prevented from completely controlling the Cabinet only by the Lower House. But once they began to lose control of their proteges, or once the latter were no longer able to dominate the particular institutions to which they belonged, institutional autonomy increased. By 1918, the GenrS had lost control of the Navy and the C i v i l Service and partial control of the House of Peers and the Army, although they retained their grip on the Privy Council. 85 Examination of the part played "by each institution in the appointment and dismissal of ministries reveals the development of an increasing separation-of-powers. Also revealed are the typical consequences that occur readily in an institutional arrangement in which the executive i s not solely dependent on any single institution, but on a number of separate institutions occupied by people with different power bases. Background The Meiji leaders' three dominant concerns of the l890's—to con-solidate p o l i t i c a l authority, promote economic development, and abolish extra-territorial rights held by foreigners—ceased to command such overriding importance i n the f i r s t two decades of the twentieth-century. Their p o l i t i c a l authority had been institutionalised: the Genr5 were generally accepted as the men with the most experience to act as the Emperor's closest advisers on matters such as the appointment and dismissal of ministers, even though the Lower House f e l t their views should be given greater consideration by the Genro in carrying out this task. Moreover, although the Genro had come to realise that compromises with the Lower House were necessary, none of these compromises threatened the existence of the new state or the main paths which i t had chosen to follow. The industrial revolution had reached an irreversible stage, although Japan was s t i l l heavily dependent on foreign imports for her manufactures and heavy machines. No real industrial proletariat emerged u n t i l after the First World War, which gave the Japanese economy the 86 boost i t had been waiting for. Revision of extra-territorial rights had been making progress, particularly after Japan's defeat of China in 1895 brought her recognition as a Great Power. The conclusion of the Anglo-Japanese alliance in 1902, the victory over Russia in 1905, and the successful but limited Japanese role in the First World War a l l represented the culmination of Japan's drive for recognition as an equal among the world Powers. Although there remained some resentment over the Versailles Treaty, foreign affairs was far from occupying a central position in the years 19OO-I918, either i n the minds of the Meiji leaders or the public at large. The absence of any single overriding national problem i s perhaps the most important characteristic of this period. The Genr5 merely sought to keep the country on i t s present course, while the parties representing the capitalist and wealthy landed classes sought solely to increase their share of influence by gradualist legitimate means. Consensus and compromise was the national mood, and social and economic conditions were sufficiently constant to make Japan in the years 1900-1918 resemble in many respects Britain i n the f i r s t half of the eighteenth-century. The two major parties confined themselves to competing with each other as well as with other branches of government for the control of the Cabinet. Moderation being the order of the day, the parties refrained from identifying themselves too openly with the occasional mass agitation that occurred over foreign policy and the price of rice. 87 The only time the parties sought mass support was when the Genro shattered the consensus and broke the unspoken agreement to compromise with them in making the Constitution work. More than in any other period, the operation of constitutional government was the most important issue over which the influential sectors of Japanese society were prepared to take a strong stand. More than in any other period therefore, socio-economic conditions i n the country can be safely abstracted from politico-institutional problems without a loss of understanding. This does not mean, however, that socio-economic conditions were unimportant, but that the study of constitutional history can be carried out without reference to them more easily during the years 1900-1918 than at other times. Recent scholarship in America has contributed a great deal to the understanding of the p o l i t i c a l and institutional development of late Meiji and early Taisho Japan. Two studies in particular, by Tetsuo Najita, Hara Kei in the Politics of Compromise, 1905-1915> and Peter Duus, Party Rivalry and P o l i t i c a l Change in Taisho Japan, are indispensable reading, and acquaintance with them i s essential for the reader to under-stand the emphasis of this chapter. Like Akita, Duus and Najita provide a wealth of information about what actually happened but also f a i l to put this information into an appropriate theoretical framework. The purpose of the present chapter i s to show how a classification of the reasons for Cabinet composition, as well as i t s rise and f a l l , can make the i n s t i -tutional history of the period more comprehensible in terms of general theory. 88 The rise and f a l l of cabinets during the entire period resulted, with one or two exceptions, from the influence of a variety of i n s t i -tutions. The Cabinet consisted of representatives of a l l institutions except the Privy Council, and the sources are much more explicit than before in linking the influence of particular institutions with the appointment of particular ministers. Neither positively nor negatively was the Cabinet responsible to any single branch of government, because there was none on which the Cabinet's existence solely depended. The Cabinet was more independent of any single institution's control than ever before, and separation-of-powers theory alone sheds light on institutional relationships. Formation of a cabinet always resulted from a compromise, and at least two institutions had to conspire to cause a cabinet f a l l . Frequently the members of these institutions consciously cooperated. The major inter-institutional agreement was between the Genro and the Lower House of the Diet. Every government, with the exception of the f i r s t Yamamoto government, owed i t s appointment to this agreement in some form, and many governments owed their demise to i t as well. The following section describes the various forms the agreement took. The Genro-House Compromise in Cabinet-Making and -Breaking, 1900-1912 As in the 1890's, the Genro and the Lower House were the most influential institutions in the years 1900-1912. But now i t was no longer simply a matter of GenrS dominance in the making of cabinets and 89 Lower House dominance in their unmaking. It became more and more common for compromises to be struck between the GenrS and the House, and for them to share the responsibility for both the rise and f a l l of cabinets. By the end of the period, Hara Kei, the President of the Seiyukai since 19lU and i t s master strategist since 1903, was able to say: "Yamagata and I are the makers of the c a b i n e t A contemporary western scholar, who was keenly aware of institutional forces, spoke of a "cabinet in part responsible to the House of Representatives." Unable to control the Lower House, the Genro had learnt to surrender to i t a part of their influence over the Cabinet, while the House, unable to wish away the existence of the Genro, had to be satisfied with a Cabinet that was only partly "responsible" to i t . This modus vivendi at which the House and the GenrS arrived i s revealed in a variety of ways throughout the period. The f i r s t was when one of the Genro became a party leader. The second involved an agreement between the Genro and the majority party to take turns in form-ing ministries. The third and f i n a l compromise was an unspoken agreement to hand over the premiership to a third party, who was either neutral, or was on good terms with both the GenrS and the Lower House. Just before the f a l l of the second Yamagata government, GenrS Ito Hirobumi f i n a l l y decided that the only way to make constitutional govern-ment work was to have a government party i n the House that commanded a majority. He decided to defy Yamagata and to organise such a party. The old Jiyuto accepted his overtures, because they saw in a GenrS as their 90 President a way to obtain regular seats in the Cabinet. The result was the formation of the Seiyukai (The Association of P o l i t i c a l Friends) in 1900, with Ito as President and the old JiyutS providing the main body of members. Although Ito was not nearly as acute as Yamagata in drawing to himself powerful adherents from a number of institutions, he did bring a few influential bureaucrats into the party. Examples are: Suematsu KenchS, Watanabe Kunitake, Kaneko Kentaro, and Saionji Kimmochi. The organisation of the Seiyukai made Ito's appointment as Prime Minister in 1900 quite natural, even though Ito f e l t that his new party was not yet ready to assume office and that Yamagata wanted this lack of preparedness to discredit i t . The failure of the Kenseito to hold together in 1898 was probably uppermost in the minds of both Genro. As in 1898, however, there was really no alternative: after a decade of intense conflict between the House and the Genro, Ito, as a Genr5 and the Pres-ident of the largest party in the House, was the only person seriously -3 considered. The forces behind the rise of Ito's government contained the seeds of i t s f a l l . The Seiyukai represented two antagonistic groups of people, and a struggle between them was the f i r s t sign of the trouble to come. Ito had wanted GenrS Inoue Kaoru as his Finance Minister, but because his party objected, he gave the position to his associate, Watanabe Kunitake, who joined the party. This angered the old Jiyuto members, particularly Hoshi T5ru, who opposed a l l Watanabe's financial policies from the 91 beginning. Although the budget passed the Lower House, i t was rejected by the Peers, who were angry because Ito had organised a party and included Hoshi, whose reputation for corrupt dealings was well known. Hoshi was forced to resign, but Ito issued an Imperial Rescript to bring the Peers round on the budget. Ito then resigned, taking respon-s i b i l i t y for involving the Emperor in the whole a f f a i r . The Genro held a Conference and decided that because Ito s t i l l had a majority in the Lower House his resignation should not be accepted. Six weeks later Ito resigned again, allegedly because of Cabinet dis-unity. Hara Kei, Kaneko Kentaro, Hayashi Yuzo, and Matsuda Masahisa were united i n opposition to Watanabe's financial plans: Watanabe had wanted to cease the practice of selling bonds to pay for nationalised industries and to turn the industries over to private hands. The Cab-inet agreed to postpone the idea, but Watanabe wanted his policy reflected in the next budget. The other ministers opposed, and Watanabe privately expressed to Ito his intention to resign. Ito, without consulting his colleagues, tendered his own resignation. It i s d i f f i c u l t to decide whether the controversy with the Peers or Cabinet disunity was the main reason behind ItS's f i n a l resignation. It i s true that he did not have the p o l i t i c a l s k i l l to deal very well with intra-party squabbles, but i t would have been an easy matter to replace Watanabe, who was out on a limb from the very beginning. The main institutional pressures behind Ito's decision seem to have come, not from the Lower House, but from the Upper House, which had not merely 92 rejected Watanabe's budget and deepened the gap between him and the other Cabinet members, but had been demonstrating i t s disapproval of the use of the Imperial Rescript by passing a l l government measures without debate. It5 could not have remained in office for long i n the face of an Upper House which so openly demonstrated i t s lack of confidence in him.^ Ito's d i f f i c u l t y in keeping his party united foreshadowed-the collapse of the f i r s t form of the Genro-House compromise. He found himself unable to mitigate conflict between the House and the government of his successor, Katsura Taro, and abandoned his role as party leader. The details of how this happened belong to the following section. Yamagata, with the concurrence of the other Genro, chose Katsura Tar5 as Ito's successor. A number of Genro kaigi were held before the Genro came to this decision. Their f i r s t choice was Genro Inoue Kaoru, but Inoue could not form the Cabinet he wanted: Shibusawa E l i c h i , an important leader of the finance world, refused the offer of Finance Minister, and Kato Komei, Yamamoto Gombei, and Katsura, a l l of whom... had served under Ito, refused to remain in office. Katsura i n i t i a l l y refused the premiership as well, and recommended It5,whomhe knew would not accept. He did this to s o l i c i t Seiyukai concurrence in his appointment, because Ito was forced to come out openly in his favour. Although Ito reluctantly supported Yamagata's protege, Katsura's selection resulted almost entirely from Genro influence.5 The form the Genro-House compromise took under Katsura's government only emerged after the collapse of the f i r s t form of the compromise. In the 17th Diet (December 1902), Katsura's Increased Taxation B i l l was 93 rejected by both the Seiyukai and the Kenseihonto (Orthodox Constitutional Party), and Katsura dissolved the House. It5 was then asked to negotiate a compromise with his party: Katsura agreed to drop the idea of increased taxes i n return for an agreement to expand the Navy with money intended for railway construction.^ It5 agreed to the proposal, but his party did not want to sacrifice railway construction, which was one of the few ways members of the House could reward their constituents. Many le f t the Seiyukai under Ozaki Yukio and Hayashi Yuzo in dissatisfaction with Ito's dictatorial methods, and the party's strength in the House declined from 193 to 128.7 Because It5 did feel constrained to get as much as he could for his party, Katsura resigned in protest in July 1903• Katsura f e l t that his government could not survive with Ito in opposition, and Q he called on It5 to abandon his role of party leader. Yamagata then stepped in and engineered Ito's appointment as Pres-ident of the Privy Council. He presented the Emperor with a document advising this appointment because :'9 The present situation does not require a change in the cabinet but a strong united nation to solve our problems with Russia. Ito, however, as President of the Seiyukai, frequently obstructs the cabinet's action. Three factors were mainly responsible for a rapprochment between Katsura and the Seiyukai. The f i r s t was the dwindling strength of the party in the House. Party members like Hara Kei began to feel that cooperation with the government could bring concrete rewards i n terms of legislation and Cabinet seats and could help revive party fortunes. 9h This tendency for parties in times of low party morale and dwindling representation in the House to compromise with the other institutions was revealed again and again throughout the prewar period. The second factor was the outbreak of the Russo-Japanese war in 190k-, which led the House for patriotic reasons to support the government so long as the nation was threatened. During the Sino-Japanese war of 1895 "the House had also suspended a l l onslaughts on the government. But in both cases, the Prime Minister was obliged to reward the House, particularly i t s major party, for this cooperation, which he knew would not be forthcoming once h o s t i l i t i e s ceased. The third factor that promoted a rapprochement between Katsura and the Seiyukai was therefore Katsura's recognition that in the long run i t was impossible to govern without the concurrence of the House. None of his predecessors had been able to do so, and he must have realised that some modus vivendi had to be found. In the spring of 1905, the Seiyukai's main strategist since Ito's departure from the party in 1903, Hara Kei, came to an agreement with Katsura: Katsura would hand over the premiership to the new Seiyukai President, Saionji Kimmochi, in return for Lower House support for the peace terms the government was concluding with Russia, which were not expected to be well received by the public. When riots leading to martial law flared up in early September over the Portsmith Treaty, the Seiyukai remained passive, while members of the other parties took up the public 95 cause in the next session of the Diet. Hara was in a strong position to hold Katsura to his promise to step down and, as Najita t e l l s us, to "prevent interference in cabinet affairs by the Genro, the Privy Council, and the House of Peers. The f i r s t Katsura government resigned by mutual agreement: both the House and the Prime Minister consented to i t s replacement by a government under the Seiyukai President. Although the Genr5 never held a kaigi to select Saionji i n 1906 and Katsura himself recommended his successor directly to the Emperor, Yamagata and the other Genro raised no objections, on condition that the Saionji government pursue the policies of i t s predecessor. Because Katsura did not inform Yamagata of his intentions u n t i l i t was too late, the influence of the House in the dismissal of the one government and the appointment of the next must be placed above a l l others. Nevertheless, the GenrS 's perhaps reluctant acceptance of the need to compromise requires that they also be seen as a force responsibleffor the rise of the f i r s t Saionji cabinet. Because the compromise between Katsura and Hara Kei persisted u n t i l 1912 in the form of an agreement for Katsura and Saionji to take turns in forming a ministry, less detailed documentation of the events which determined the timing of Cabinet changes u n t i l 1912 is required. As long as the consensus on policy.continued, the Seiyukai supported Katsura governments in the belief that this would guarantee GenrS approval of Saionji governments. In each case the change took place because of growing impatience by the party to the agreement whose turn 96 to govern was coming up. Throughout this period no Genro kaigi was held, indicating the GenrS's submission to the best bargain their representative Katsura could strike with the Seiyukai, although Hara Kei remained in constant touch with Yamagata and assured him that the Genro were receiving stable government in return. From 1906 u n t i l just before the collapse of the compromise in 1912, each Cabinet change revealed an almost identical pattern: a growing impatience by the inferior force behind the government stepped up the pressure for i t to become the major force behind that government's successor. After two and a half years of Seiyukai-led government in 1908, the Genro began to get restless. Matsukata and Inoue, their two finan-c i a l experts, expressed dissatisfaction with the government's budgeting and called for higher taxes and a postponement of the "pork barrel" policy of railway construction. The Finance Minister, Sakatani Yoshiro, who was a protege of Inoue, gave in to the Genro's demands, and the Communications Minister, Yamagata IsabuirS (the GenrS's nephew), resigned in protest because of the cuts made in the appropriations for his depart-ment. The split in the Cabinet led to a general resignation, but when the Emperor asked ItS what to do, he was advised to accept only the resignations of Sakatani and Yamagata.^ Although the immediate c r i s i s was solved, i t became more and more d i f f i c u l t for the government to weather the general dissatisfaction of the Genro. Yamagata Aritomo was dissatisfied with the CabinetJs foreign policy, as well as i t s handling 97 of the incipient socialist movement, and i n vain tried to persuade the Army Minister, Terauchi Masatake, to resign and bring down the government. The f i n a l destruction of the Saionji ministry came when the Upper House, largely under the influence of the Yamagata-controlled Kenkyukai, rejected Hara Kei's plan to abolish the gun, or " d i s t r i c t s . " The councils of these gun were controlled by Yamagata through his appoint-ment of the gun chiefs, and they formed his main power base in the country. Najita describes the Peer's action as "a breakdown of the compromise relationship between Hara and Katsura," because Katsura failed l P to work in the Peers for the b i l l ' s passage. ^ To regard the Upper House's rejection of the gun B i l l as a "break-down of the compromise" is not s t r i c t l y accurate. The Seiyukai had overstepped the limits of the agreement by attacking Yamagata's power base and by remaining i n office longer than the GenrS regarded appro-priate. It i s better to regard the f a l l of the f i r s t Saionji cabinet as resulting from the terms of the agreement. While the GenrS and the Peers were the major forces occasioning the f a l l of the government, Saionji's resignation was not contrary to what, had been agreed on in 1905* and the Lower House must be assigned a minor role. Saionji explained to his colleagues that regardless of the party's victory in the election of May 1908, i t was time to hand over the premiership to Katsura.^ Again the GenrS did not meet to appoint a successor. Saionji nomi-nated Katsura directly to the Emperor, who consulted It5. ItS supported 98 Katsura's nomination and the other GenrS concurred. Najita i s not quite correct i n saying that the compromise was restored because Katsura needed support for his budget. The compromise had not really broken down, as indicated by the President of the Seiyukai's nomin-ik ation of Katsura as his successor. Katsura's resignation in August 1911 was not preceded by any particular conflict with the House, but by a growing dissatisfaction among party members that he should give way to Saionji. McLaren wrote: "By 1911 the inevitable tendency in Japanese p o l i t i c a l parties towards obstruction had developed to such an extent that Katsura did not think i t worth his while to continue any longer in office." He described the f a l l ofKKatsura's government as " v o l u n t a r y . T h e government had lost the support of the finance world and of the lower classes, and Katsura stepped down before the House was really prepared to take up their cause. The Kotoku jiken, an incident involving the execution of a group of anarchists, had contributed to the government's unpopularity in the country at large, and in January 1911 Katsura resigned because of i t s repercussions, even though his resignation was not accepted.^ When Katsura f i n a l l y resigned in August, he recommended Saionji as his successor directly to the Emperor, saying that he would return to office when Saionji resigned.-^ Katsura conspicuously by-passed the Genro, who were neither consulted over the appointment of the Prime Minister nor the composition of the Cabinet. But Hara did take pains to e l i c i t the support of Yamagata. His efforts in this regard, as well 99 as the Genro's failure to veto the fait accompli, indicate that once again the Cabinet was a joint Genro-House creation, this time with the Genro as l ft the junior partner. The Cabinet changes that took place under the Hara-Katsura agree-ment of 1905 were made with the joint concurrence of the Lower House and the Genro. Neither inst i tut ion was either the sole maker or breaker of any of the governments concerned. In the rise andr'fall of the fourth It5 cabinet and the rise of the f i r s t Katsura cabinet, the House and Genro were also c r i t i c a l l y involved, even though in these cases there was a different sort of agreement. Unt i l 1912 therefore, i f the role of the Upper House in unseating It5 i n 1901 and Saionji i n 1908 i s set aside, the Cabinet's appointment and dismissal resulted solely from the combined forces of the Genro and the Lower House. Neither inst i tut ion could on i t s own either raise or pu l l down the Cabinet, which i n these years was therefore separate from each inst i tut ion individually, although i t was dependent on both col lect ively . But after 1913, the dependence of the Cabinet on new combinations of inst i tut ions, which included hitherto passive ones, made the Cabinet even more separate from any single one of them, and the description of the period as one of a f u l l -fledged separation-of-powers became appropriate. Cabinet-Making and -Breaking, 1912-1918 The agreement to alternate the premiership between representatives of the Genro and the House collapsed when the consensus over policy broke down. This happened when the leading GenrS, Yamagata Aritomo, took the 100 side of the Army i n a dispute between t h i s i n s t i t u t i o n and the Lower House over the plan t o create two new army d i v i s i o n s at a time of across-the-board retrenchment. The importance of t h i s i n c i d e n t l i e s p a r t l y i n i t s r e v e l a t i o n o f the growing independence of the Army from the GenrS, because the plan was i n i t i a t e d by the Army and among the Genro only Yamagata supported i t . But because he, the most i n f l u e n t i a l of the Genro, d i d support i t , the House came i n t o c o n f l i c t w i t h the Genro as a whole as w e l l as w i t h the Army. Yamagata encouraged the Army M i n i s t e r t o r e s i g n i n f u l l knowledge of the agreement by e l i g i b l e o f f i c e r s not t o nominate a successor. I n the e l e c t i o n of May 1912 the S e i y u k a i won an absolute m a j o r i t y pledged t o economic retrenchment, and soon afterwards the government ordered a 10-15$ cut i n each department's expenditure. The Army i n s i s t e d t t h a t the savings r e s u l t i n g from i t s 1.95-million-yen cut be devoted t o the c r e a t i o n of two new army d i v i s i o n s , and the Navy i n s i s t e d on a 9-million-yen expansion programme. While S a i o n j i agreed t o the Navy's demand, he re f u s e d t o sanction t h a t of the Army. ^"9 The r e s u l t i n g c r i s i s l e d t o a regrouping of i n s t i t u t i o n a l a l l i a n c e s . The f i n a n c i a l experts of the GenrS, Inoue and Matsukata, were ambivalent but tended t o support the government's p o l i c y of across-the-board c u t s , and together w i t h General Oyama Iwao, favoured postponement of the Army's p l a n . Katsura, the mediator between the Genro and the House up t o now, was not i n a p o s i t i o n t o perform h i s former r o l e , because he had l o s t favour w i t h Yamagata. He had been " r e t i r e d " t o the palace as I m p e r i a l 101 Household Minister. Yamagata took the side of the Army from the outset, and soon Tanaka Giichi, Yamagata's protege and the Chief of the Army General Staff, won Inoue over to their side. Inoue led the "bankers' opposition to the Navy's plan, and had some success in his attempt to win the support of the Finance Minister, Yamamoto Tatsuo. The result was a clash between the Navy Minister, Saito Makoto, and the Finance Minister. Admiral Yamamoto Gombei acted as the "string puller" behind Saito and ensured that the Navy's position would not be sacrificed. The Genro Oyama, Inoue, and Matsukata held a meeting and t r i e d to 20 persuade the Army to compromise, but to no avail. On December 2 1912 the c r i s i s broke. After consulting Yamagata, Army Minister Uehara Yusaku resigned and a l l eligible Army officers agreed not to serve as a successor. On December 5 the second Saionji cabinet resigned. This time the compromise was shattered because Saionji was not prepared to nominate Katsura as his successor. This was the f i r s t government since the establishment of the Constitution that f e l l without any assistance from the Lower House. Because Yama-gata s t i l l had sufficient influence in the Army to have succeeded in dissuading i t from i t s strong course of action, i f he had so chosen, the f a l l of the government resulted f i r s t l y from the Army's use of i t s veto power, and secondly from the support Yamagata gave the Army.^ The rise of the third Katsura government reveals the attempt to restore the Genro-House compromise: for nearly two weeks the Genro, who had not met to select a Prime Minister since 1901 and whose number 102 now included Katsura, met almost solidly in an attempt to find someone who could break the impasse. Admiral Yamamoto Gombei, who had opposed the Army, was nominated but refused. Saionji would not remain in office, and the Genro Inoue, Oyama and Matsukata, who had also shown sympathy for the previous government, further refused to try to win over a hostile House. Finally, Yamagata decided that only Katsura would be able to restore some sort of compromise, and the GenrS nominated him. The rise of this the third Katsura government was therefore almost entirely due to an assertion of GenrS authority over the appointment of the Prime Minister. Katsura was the Genro's choice and the House had given l i t t l e indication that Katsura would be acceptable, although the Genro must have believed that he would be. The GenrS are assigned the major role in Appendix I, and the House and the Army, whom the Genro must have also considered, are each assigned a minor role. Katsura's appointment was badly received by the public, particularly by the Seiyukai. Hara was s t i l l prepared to bargain with Katsura, but this time at a higher price. Before long,- in spite of and partly because of Katsura's success in bringing together a new party grouping i n the House, he was forced to resign. The Seiyukai was so provoked by the formation of a new party that could threaten i t s hegemony in the House, that a l l willingness to deal with Katsura ceased. It was no longer possible for a Genro, as Katsura now was, to use the expedient which had brought ItS brief success a decade before. He not merely lost the 103 confidence of the Genro, but provoked the majority party into the strongest stand i t took on any issue throughout the period. The Movement for Constitutional Government was the response of most members of both major parties to Katsura's appointment. Its purpose was to restore predominant party influence over the Cabinet; i t s activities involved nationwide agitation. Because the details of the events during the next month or so are complex and not relevant to this study, only the main institutional forces that came together to force Katsura out of office need be discussed here. From the outset, the Navy was hostile towards the Katsura government. Katsura had used an Imperial Rescript to force Saito Makoto into service as Navy Minister, and Admiral Yamamoto Gombei, in his intrigues to topple the government, well represented the position of the Navy.23 Katsura's response to the Movement for Con-stitutional- Government and to the rumours that the Seiyukai intended to present a vote of no-confidence, was to sever relations with Yamagata and to organise a new party in the House. He was not aware that this course of action would translate the rumours into reality. Hara Kei was actually s t i l l prepared to come to terms with him. A brief digression w i l l be made to describe the origins of his party. In December 1912 Katsura organised the Doshikai (The Association of Friends) by bringing together a number of his and Yamagata's proteges as well as many young bureaucrats on the one hand, and the Chu5 Kurabu 10k (The Centre Club) and the "reform" faction of the KokumintS (The People's Party), on the other. The KokumintS had inherited the position of the main opposition party when the KenseihontS reorganised in 1910. The Kokuminto's reform faction, under Oishi Masami, like the Jiyuto in 1900, believed that a GenrS as i t s President was the best way to build up a party that could r i v a l the Seiyukai in the House. The KokumintS split into two almost equal parts, and kh of i t s Dietmen joined Katsura's DSshikai. The others, who had been among the main instigators of the Movement for Constitutional Government, remained in the party under the leadership of Inukai Tsuyoshi. The 30 members of the ChuS Kurabu, under Oura Kanetake and Adachi Kenzo, had been a party "loyal" to the Genro-dominated governments. In a l l , Katsura held the allegiance of 120 ok Dietmen. In anticipation of a dissolution and a Katsura-managed election, the Seiyukai prepared i t s e l f for a confrontation with the government in the Diet. When the session opened, Katsura prorogued the Lower House for five days, and as soon as the House reassembled a motion of no-confidence was presented. It c r i t i c i s e d Katsura for troubling the Sovereign for Imperial Rescripts and for suspending the session. Katsura responded by issuing an Imperial Rescript which commanded the House to withdraw the motion. Then the Navy stepped i n and took up the cause of the Seiyukai. 2^ Najita argues that Yamamoto Gombei used the incident to restore Satsuma influence over the government, which for years had been dominated 105 by men from Choshu. But to see Yamamoto acting merely on behalf of Satsuma interests i s to forget that a struggle between the Army and the Navy had been taking place over very concrete matters like the size of their respective share of national expenditure. Yamamoto supported the Seiyukai and gave that party the courage i t needed to defy the Imperial Rescript. He also advised Katsura to resign because he hoped, i t seems, that a Seiyukai-dominated government would be more generous to the Navy than a Doshikai-dominated one. The f a l l of Katsura's government f i n a l l y took place amid scenes of public rioting outside the Diet building. The House and the Navy, in that order, were the institutional forces that unseated Katsura, although his loss of favour with the Genro prevented them from coming to his rescue and requires that the Genro be assigned a minor role.^7 The selection of Katsura's successor was made at a formal GenrS  kaigi, attended by Yamagata, Oyama, Katsura, and Saionji Kimmochi, who retired from the Seiyukai and was admitted into the ranks of the Genro. Under the circumstances, the only kind of compromise that could satisfy both the Genro and the Lower House, was the summoning to the premiership of someone who represented a third force, and whom both sides could claim was as sympathetic to i t as to i t s r i v a l . This was the third form the GenrS-House compromise took in the years 1900-1918 and the one which characterised the latter part of the period. In these cases the GenrS kaigi was as much a barometer which assessed the 106 institutional balance of power as an assertion of the influence of the Genro themselves, who represented one of the institutions that affected the overall balance. The Genro not merely tried to find the person most able to make the Constitution work, but the person representing a force most l i k e l y to serve their own institutional interests. The Genro were not the only institution to play others off against one another. Hara Kei, according to Najita, used third parties "because they could be exploited to help the Seiyukai deal with the House of Peers, the military services, and the Privy Council. This was princi-pQ pally why Hara chose to work with Yamamoto." Yamamoto Gombei was a f a i r l y obvious choice in 1913* After con-sulting Hara, Saionji nominated him at the GenrS kaigi. Yamagata, who really wanted to nominate Terauchi Masatake, one of his proteges in the Army, made no objection to Yamamoto, probably because in spite of the House's overwhelming victory over Katsura, this choice prevented the appointment of a party Prime Minister, something Yamagata wanted at a l l costs to avoid. The House, the Navy, and the GenrS were therefore the institutional forces behind Yamamoto's selection. The order of their importance is more d i f f i c u l t to determine, but i t seems that the Genro got the worst of the compromise, because both the other institutions had been responsible for Katsura's fall. ^ 9 The f a l l of the Yamamoto government resulted almost entirely from the use by the Upper House of i t s veto power. Although the Genro were also partly responsible, the Peers were the major force, The details of 107 what happened are discussed below in the section on the House of Peers. It is sufficient to note here that because of Yamamoto's success in acting as an acceptable third force between the Lower House on the one hand, and the GenrS and the Army, on the other, the GenrS were on the keen look out for a successor in as similar a middle position as he. In one respect the GenrS f e l t the situation in February 1913 was different from the•one in March 1914: because the Seiyukai had got the better bargain in Yamamoto's selection, someone who could act as a counter force to the Seiyukai without losing the support of the House was needed, but this person was not to be a party man. Wow that there was a second potential majority party i n the House, the Genro's task was easier than i t might have been. They found in Okuma Shigenobu, who had resigned from the KenseihontS in 1907 > the ideal person. At a formal GenrS kaigi in March 1914, the Genro decided to r e c a l l Okuma from his retirement from active p o l i t i c s . Although he never attended any GenrS kaigi, he was really a GenrS himself, and could also command the sympathies of the champions of party cabinets. He represented the ideal third force. Because of his past association with the men who were now i n the Doshikai, a government based on the DSshikai could be used to put an end to the Seiyukai's drive for complete control of the Cabinet. When Okuma dissolved the House, the GenrS at long last witnessed the end of the Seiyukai's majority in the House.30 The election of 1915 raises some interesting theoretical points. One of the general characteristics of separation-of-powers systems is that the greater the number of institutions among which power is 108 fragmented, the greater the opportunity for those near the executive to play off against one another those with less access to the exec-utive. The American President's a b i l i t y to use the two Houses of Congress against each other is fa c i l i t a t e d by the control of the power to i n i t i a t e legislation and compile the budget. For the same reason, even though GenrS control of particular institutions diminished during the years 1900-1918, the GenrS were able to use the increasing separation-of-powers to play other i n s t i -tutions off against one another. This a b i l i t y increased after the development in the Lower House of a second party capable of mustering a majority. As long as there was only one such party, a government which had i t s support need not fear that the defection of a few members would mean a loss of the House's confidence. The f a l l of the Yamamoto government and the rise of i t s successor under Okuma clearly illustrated the new leverage the GenrS possessed. They could destroy a government with greater ease and less public outcry than ever before, simply by collaborating with the opposition party. One carefully managed election could then convert the new government's minority in the House into a majority. The increased manoeuverability of the GenrS had i t s counterpart in the decline of the House's influence. Before the appearance of the DSshikai, the Seiyukai could threaten t o t a l obstruction of the govern-ment in order to win influence over the Cabinet. Now such threats were less persuasive, because the GenrS could with v i r t u a l impunity replace 109 a Navy-Seiyukai government with one that was hostile to the Seiyukai and sympathetic to the GenrS themselves. The new situation did not, however, mean that the Genro could reign supreme. After the DSshikai's massive victory in the election of 1915> partly because of Okuma's popularity in the country, but largely because of DSshikai control of the Home Ministry and i t s a b i l i t y to "manage" the election, the Genro found themselves in almost the same position as before. Instead of a hostile House under the direction of the Seiyukai, they had to face a hostile House under the direction of the DSshikai. Nevertheless, the GenrS could repeat the tactic of collaborating with the opposition party to unseat a government of which they disapproved. They could then appoint a premier who was neutral between the two parties and rely on their mutual ho s t i l i t y to ensure the new Prime Minister's a b i l i t y to win at least the neutrality of one of them. Events over the next two years reveal that the GenrS used their new manoeuverability in this very way. Soon after the 1915 election the GenrS began to work for Okuma's f a l l . The policy of the Foreign Minister in particular, who failed to keep Yamagata informed and to show him important diplomatic documents, was something the GenrS could not tolerate. Yamagata ultimately insisted that KatS KSmei resign.3^ A l l that prevented the GenrS from the. immediate replacement of Okuma himself was the lack of a suitable successor. Okuma realised that he could not remain in office for much longer and began preparations to have KatS Komei succeed him. The-Genro-s- a b i l i t y to 110 prevent the appointment of the Doshikai leader resulted alone from the existence in the Lower House of an opposition party which could be used against the D5shikai. The institutional forces behind the rise of the Terauchi government were almost identical to those behind the rise of the Okuma government. The Genro played off one party against the other and could remove any party-supported government they liked. Although the Peers did the dirty work for them in 191k, they found that Okuma's departure in 19 l6 brought them l i t t l e public criticism, because the Seiyukai was ready to defend 32 Okuma's dismissal. Terauchi Masatake was chosen at a formal GenrS kaigi, which involved as great an assertion of GenrS power as in 1912 when Katsura was chosen for the third time. But now that there were two parties in the House, either of which could win an election i f supported by the government, a person independent of both parties could be appointed without the public h o s t i l i t y that had met Katsura.-^-" Although Terauchi had some i n i t i a l d i f f i c u l t i e s with the Diet, he proclaimed his cabinet a ehusei faheri naikaku or "a s t r i c t l y impartial cabinet," and found that he could win Seiyukai support in the House in return for government support in the next election. While the GenrS could use the Seiyukai to keep their protege in power, the Seiyukai could use Terauchi to transform them from the minority party to the majority one, and at last bring about the appointment of the f i r s t commoner party Prime Minister following Terauchi's resignation. To. I l l describe the rise of the Hara cabinet would be to anticipate what belongs to the next chapter. The f a l l of the Terauchi cabinet, however, reveals that while Cabinet-making may have been far from the sole pre-serve of the House in 1900-1918, Cabinet-breaking was something the House could on occasion accomplish on i t s own.3^ The sources are almost unanimous in ascribing the f a l l of the Terauchi cabinet to the rice riots of late 1918, which broke out after a 130$ rise in the price of rice in I91U-I918. But only Peter Duus mentions the institutional forces into which the riots were transformed: the Seiyukai refrained from identifying i t s e l f with the rioters and supported the government in the Diet.35 But Duus f a i l s to draw the appropriate conclusion. The Kenseikai,which was a reorganised version of the DSshikai formed in September 19l6,.actively took up the public cause, and i t is tempting to regard the mob and the Kenseikai as the breakerSTof the government, and the Seiyukai's caution as the reason for the rise of the Hara cabinet. Such an interpretation, however, would be incorrect. By refraining from taking up the rioters' cause, the Seiyukai took over the entire power to remove Terauchi, because i t could hold over him the threat to obstruct his policy and bring him down whenever i t chose. Had the Seiyukai joined the Kenseikai, not merely would i t have made Yamagata less l i k e l y to agree on Hara as a successor, but both parties would have been jointly responsible for the cabinet's demise. Instead, the Seiyukai cleverly translated the mob agitation not into the influence of the House as a whole, but into i t s own influence. 112 The period 1900-1918 began and ended with the Seiyukai as the predominant influence on the Cabinet's existence, although the rise of the party which challenged i t s hegemony in the House did not make i t the sole expression of the House's predominant influence. Appendix I reveals that the House, through one of i t s parties or other, played at least a minor role in the rise of every single govern-ment of the period. It also helped unseat every government except the second Saionji and f i r s t Yamamoto governments. The rise of five out of nine ministries and the f a l l of five also resulted from a major part played by the House. The House was therefore in possess-ion of the greatest Cabinet-making and -breaking power of a l l i n s t i -tutions . The second most influential institution was the Genro,who also exerted at least a minor influence in the rise of a l l cabinets, and also a major role in the rise of five or them. But the GenrS were only largely responsible for the f a l l of three ministries, although they were partly responsible for the f a l l of five. Table 2 indicates the breakdown of Appendix I into the major and minor roles played by each institution in Cabinet-making and -breaking in 1900-1918. Table 2. Major and Minor Roles i n Cabinet-Making and -Breaking, 1900-1918. Cabinet-Making Cabinet-Breaking House Genro Peers Army Wavy House Genro Peers Army Navy Major 5 5 0 0 1 5 3 3 1 1 Minor 4 4 0 1 0 2 5 0 0 0 113 The GenrS and the House were clearly the most important institutions of the period. It was inevitable that the functioning of constitutional government, which the national consensus on policy among the classes occupying these two institutions made easier, would force a compromise of some sort on how the Constitution was to work. These compromises were not merely revealed in the three main ways described, but In the composition of cabinets throughout the period. Cabinet Composition and the Genro-House Compromise The major qualifications of Cabinet members in this period reflect quite accurately the different compromises between the House and the Genro. The Fourth Ito cabinet included three veterans of the old Jiyuto and four personal followers of Ito, revealing his attempt to build bridges between the House and the GenrS by means of overlapping membership. 3°" Of the sixteen different people who served in the f i r s t Katsura ministry, only six were not proteges of Yamagata, and not one had any relationship with the House. Yet the cabinet survived longer than any other under the Meiji Constitution, because a working agreement with the Seiyukai obtained i t s support in exchange for control over the next cabinet. For the same reason, Katsura's second ministry, which included only two men who were not Genro proteges and no one from the House, lasted for three years. But the f i r s t Saionji cabinet failed to exploit the agreement to the f u l l , giving two non-Service seats to Yamagata men, and only three besides the premiership to the Seiyukai. Saionji's second ministry made better use of the compromise, and included no one from the Ilk Yamagata clique, three Seiyukai members besides himself, and one pro-Seiyukai financier. Without some compromise or other with the House, the third Katsura government, which consisted almost entirely of Katsura's personal followers.so that one writer described i t as a Katsura-batsu naikaku (a Katsura clique cabinet), was bound to die e a r l y . B u t when the compromise came, i t was of the wrong kind and too late. The sudden influx of Katsura's followers into the DSshikai was the kind of compromise ItS had unsuccessfully attempted a decade earlier. The f i r s t Yamamoto cabinet, born of GenrS submission to a Navy-Seiyukai alliance, included two Admirals, seven Seiyukai members, and one Seiyukai sympathiser. The second Okuma cabinet, which was the result of a GenrS-Doshikai- offensive against the Seiyukai, included two of Yamagata's followers, a couple of Katsura's, and seven members of the DSshikai. Finally, Terauchi was able to exploit to the f u l l the h o s t i l i t y • between the two main parties and could obtain Seiyukai support without i n i t i a l l y including a single member of i t in his cabinet. The party was prepared to support the government in exchange for support i n the next election. In each case, the cabinet's composition reflected the institutional compromise. The predominance of Lower House and GenrS representatives in cabi-nets throughout the period parallels their predominance in Cabinet-making and -breaking. Although Appendix I shows that the t o t a l number of Peers and C i v i l Servants or ex-Civil Servants exceeded the total 115 number of Lower House party members, Appendix III shows that many of these Peers and bureaucrats were also Lower House party members. If one considers that party members who were Peers or ex-officials could be expected to support the Lower House in times of inter-institutional conflict and that the f i r s t two Katsura cabinets included no party members, the predominance of the Lower House in Cabinet composition in the period as a whole becomes clearer. It also becomes clearer i f one recognises that the number of ministers who were solely Peers or solely bureaucrats f e l l far short of the to t a l number of party members. The period was one of compromise between the two main institutions, a compromise revealed not merely in their Cabinet-making and -breaking power, but in their a b i l i t y to have representatives included in the Cabinet. During the years 1900-1918, the Genro also gradually lost control of institutions which they had dominated in the l890's. Genro•inability to', ensure inter-institutional harmony owing to their at least partial loss of control of a l l institutions, except the Privy Council, became the second main characteristic of institutional development in the years I9OO-I918. The Decline of Genro Power and the Rise of Institutional Autonomy The following section deals with each institution's growing separation from the Genro. I assume that the a b i l i t y to act independently of the Genro in Cabinet-making and -breaking as well as the a b i l i t y to have 116 representatives who were not GenrS proteges included in cabinets indicates autonomous influence. I indicate when representatives of the C i v i l Service and the Upper House were not associated with any other institution as well, and can be regarded purely as C i v i l Service and Upper House representatives. This procedure w i l l indicate not merely the independence of institutions from the Genro, but their autonomy and a b i l i t y to act as completely independent forces. I discuss f u l l y each institution's Cabinet-making and -breaking power as well as i t s power over Cabinet composition before moving on to the next institution. The f i r s t institution to show5signs of a break away from the iron grip of the Genro was the House of Peers. Although the development of greater institutional autonomy took place mainly after 1912, the Peers showed a willingness to thwart Genro purposes much earlier than this. Their f i r s t assertion of autonomous influence took place when they opposed the Seiyukai's budget in 1901. Because the leaders of the opposition in the Peers included Yamagata supporters, who were, on Ito's request, instructed by their patron to pass the budget, some writers have concluded that the Peers' failure to do so was proof that Yamagata must have only half-heartedly urged them to cooperate with the govern-ment.-3 This interpretation does not accord with a number of facts. F i r s t l y , six Upper House factions, not merely those controlled by Yamagata, opposed the government. Secondly, of the fifteen Peers who sat on the special Cabinet committee that tried to work out a compromise, few were Yamagata supporters. Thirdly, the Genro held a number of meetings and 117 agreed that the Peers would have to be brought to their senses. A l l of them were reported to have been enraged by the Peers' intransigence.39 It i s therefore more accurate to say that in this instance the GenrS had lost control of the Peers. And although the f a l l of the ItS government f i n a l l y came after intense conflict between the staunch party members of the Cabinet and Watanabe Kunitake, a personal follower of ItS, i t would have been unable to continue in office for long in any case, be-cause i t had lost the confidence of the Peers by using an Imperial Rescript against them to pass the budget. The Cabinet f e l l not merely because of the contradiction i n having one man and his personal followers represent both sides of an antagonistic power relationship, but because a third force added intolerable complications.to an already delicate balance of forces. As Fahs wrote, "A Japanese cabinet may be forced out in so many different ways that the result can seldom be attributed to any one organ. The f a l l of the f i r s t Saionji cabinet, which has already been ascribed largely to the Upper House, and the f a l l of the f i r s t Yamamoto cabinet were more clear-cut cases of the Peers' cabinet-breaking power. Only the f a l l of the Yamamoto cabinet w i l l be described. When i t became known in 191k that high-ranking naval officers had accepted up to 5$ commission from German and British firms for the orders they placed, the Peers demanded Yamamoto's resignation. Although the attack was spearheaded by Yamagata protege Hirata Tosuke, • 118 the i n i t i a t i v e in the move to have the cabinet resign came from the Peers as a whole. Fahs described this as "the only clear-cut case of the downfall of a Cabinet as a result of adverse action by the House of Peers. It may be incorrect to speak of actual loss of Genro control over the Upper House before the death of Yamagata. But because the Upper House had opposed the Genr5 to the bitter end on one occasion, and initiated strong action against two cabinets which s t i l l had the confidence of the GenrS on another, i t i s not too much to say that the Upper House was gradually coming to acquire an institutional independ-ence and interest of i t s own. Once Yamagata l e f t the stage, this earlier tendency to ini t i a t e autonomous action increased. It should not be considered a sudden change coinciding with Yamagata's departure. The large number of Peers, many of whom were GenrS proteges, who sat i n the cabinets which were on the whole Genro creations, reveals the close ties between the Upper>:House and the GenrS. Few of the Peers in this category were at the same time members of Lower House parties, indicating that the GenrS seem to have abandoned the attempt to prevent deadlock between the two Houses by means of cabinet ministers who were members of both. Although cabinets that were largely Lower House crea-tions did include Peers who were also members of Lower House parties, this practice became less common. Lower House-dominated cabinets also came to recognise that the avoidance of deadlock with the Upper House would require at the very least the inclusion of Peers who were not too 119 closely identified with i t , or with the GenrS for that matter, hut represented the Upper House alone. The extent to which the Upper House was acquiring an institutional interest of i t s own and was becoming independent of the Genr5 therefore, i s indicated by the growing tendency for Peers who had ties with neither the GenrS nor the House to be included in the Cabinet. Until 1922, when Yamagata died, even though the Upper House was prepared to act independently of the Genro, i t was f a i r l y closely a l l i e d to them. After 1913> however, the alliance was much looser, and the influence of the DSshikai began to make i t s e l f f e l t on Cabinet composition. The f i r s t Katsura cabinet included eight Peers, five of whom were Genro proteges, and none of whom had any ties with the Lower House. But five of the six Peers i n Katsura's third ministry, which was born of an alliance between the DSshikai and dissidents of the Yamagata clique, were also members of the DSshikai, and a l l of them were person-a l l y tied to Katsura, some of them temporarily ex-members of the Yamagata clique. The DSshikai, more than the Seiyukai, had to learn that over-lapping membership was no longer able to preserve inter-institutional harmony, a lesson i t did not seem to have learnt by the time Okuma's second cabinet was appointed,, the result of another Genro-DSshikai alliance. Three of this cabinet's four Peers were also members of the DSshikai, and three were GenrS proteges. Of the five Peers in the 120 Terauchi cabinet, the last to be raised largely by the Genro in this period, three were Genro proteges, and only one was a member of the Seiyukai, revealing Terauchi's apparent recognition that the expedient of overlapping membership could not preserve the peace between the two Houses. The cabinets which were mainly Lower House creations show a slightly different tendency, fewer of their representatives of the Upper House being also Lower House party members. The fourth Ito cabinet's four Peers were a l l Seiyukai members, i t i s true, but because of that government's d i f f i c u l t i e s with the Upper House, three of the four Peers in the f i r s t Saionji cabinet had no ties with the Lower House, although the single Peer, besides Saionji, in his next cabinet, was regarded as sympathetic to the Seiyukai, and a l l three Peers in the f i r s t Yamamoto ministry were Seiyukai members. Although most Peers who served between 1900-1918 had ties with either the GenrS or the Lower House, there were a few who represented solely the Upper House, some of whom were included specifically in order to obtain the cooperation of that institution. Katsura's f i r s t ministry included three unaligned Peers, and the Terauchi ministry one, while both the unaligned Peers in the f i r s t Saionji cabinet were specifi-cally brought in in an attempt to woo the Upper House-. So although the House of Peers was not yet f u l l y recognised as an institution which could act. independently and cause deadlock, a beginning had been made. In the next period, however, no cabinet failed to include Peers 121 who were primarily i f not solely representatives of the Upper House, partly because i t was only in the next period that the Peers really came into their own, but partly because those involved did not at once understand the forces under which they were operating. The Peers, having been given an institutional base by the Meiji Constitution, were bound sooner or later to use their influence in their own interests, particularly after the death of those Genro like Yamagata, whose i n f l u -ence was more personal because of their prestige as creators of the Meiji state. Saionji, a later Genro, did not have the prestige of Yamagata and became a mere arbiter in the inter-institutional Struggle. Because the Peers had a permanent institutional base from which they could exercise enormous influence, their actual influence would increase as they discovered they had interests the GenrS did not share. After the f i r s t decade of constitutional development, the practice of giving as of right both the Army and the Navy a seat in the Cabinet was firmly established. When i t was legalised in 1900 by Yamagata, the potential of either institution to veto Cabinet composition and policy was abundantly clear, but as long as the GenrS could control them, they had l i t t l e chance to work at cross purposes with each other or with other institutions. When Fleet Admiral and GenrS, SaigS Tsugumichi, died in 1902, however, there was no longer any compelling reason for the Navy to neglect i t s own institutional interests in cases of conflict with other institutions. By 1918 the Navy had begun to pursue i t s interests with vigour. The actual break took place in late 1912, when the Saionji govern-ment, which had pledged i t s e l f to cut each department's estimates by 10-15$, 122 refused to allow the Army the use of the savings i t had made to acquire two new divisions, but agreed instead to the Navy's 9-million-yen expansion programme. While Hirota Naoe, one of the few contemporary observers to examine the problem of Cabinet changes, does recognise the conflict between Navy Minister Saito Makoto and the Army as well as Yamamoto Gombei's role as "wirepuller" behind the scenes, he insists on seeing i t in terms of Satsuma-Ch5shu rivalry rather than institutional rivalry.^3 Having gained what i t wanted from the Saionji cabinet, the Navy refused to cooperate in forming a cabinet under Katsura, who in order to placate the Army only f e l t able to offer the Navy the compromise of a 6-million-yen expansion programme.^ So Katsura had to obtain a Navy Minister by means of an Imperial Rescript, which ordered Saito "to contribute towards the betterment of the Navy's role in p o l i t i c s . It i s hardly surprising that the Navy actively plotted the down-f a l l of the government. Yamamoto Gombei, at the height of the public uproar against Katsura, told the Prime Minister in so many words to resign, and took seriously Katsura's suggestion that he be the next Prime Min i s t e r . ^ Yamamoto then informed the Seiyukai that he was prepared to succeed Katsura, and the party, encouraged by the Navy's determination to bring down the government, resolved, in defiance of an Imperial Order, to proceed with i t s no-confidence motion in the government. Najita sees in Yamamoto's move a motive which is related to his Satsuma origins, rather than his role as defender of the Navy in a constitutional arrangement in which those who did not fend for 123 themselves went under. The culmination of the Navy's drive for independence from the Genro came with the establishment of the Yamamoto cabinet, which was formed at a time of nationwide anti-Genro sentiment and activity. It would have been impossible to find the Navy keeping such company ten years earlier when the Genr5 had sufficient power to:.coordinate policy and see to each institution's interests. Few American scholars, whose own experience with politics under a separation-of-powers constitution has possibly made them less sensitive to the different strategies necessary for survival in separation-of-powers and parliamentary systems, appreciate the institutional forces which brought the Navy to this point. One who came f a i r l y close to doing so wrote Each of the major ."functional groupings in Japan did in fact have an institutional locus of power which generally enabled i t to exercise this kind of veto power. . . . Once the genro • lost their grip on the government the system was l e f t without a dominant coordinating e l i t e . Thereafter coordination was not a matter of centralised policy planning, but one of defining jurisdictions and negotiating differences among the- major institutional contenders. The Navy had become one of these contenders. Diminishing Genro control of the Army was not as apparent in this period. There were no cases of bitter Army resistence to the GenrS, nor even of minor conflicts between the two, although there was one occasion when the Army had the whole-hearted support of Yamagata alone among the Genro. The main reason for this seems to be that the Army's strategy for gaining i t s ends was through Genro influence, a strategy 12k which seemed obvious in view of the large number of Genro with close ties to the Army, The Navy on the other hand, always poorly represented among the GenrS and not at a l l since 1902, was forced much earlier to find an alternative to reliance on the GenrS. The Satsuma-ChSshu rivalry hypothesis i s unable to shed any light on this problem, because Oyama and Matsukata could have been used more effectively by the Navy to find Satsuma a l l i e s . The consequence of the Army's choice to al l y i t s e l f with the Genro was, however, a loss of influence in this period, which parallelled the GenrS's loss of influence. It would be some time before the Army would learn to use i t s power independently, although i t s bitter defeat in 1913 together with that of the GenrS marked the beginning of i t s d r i f t away from the alliance with the GenrS towards the acquisition of the more Machiavellian s k i l l s required for survival i n a separation-of-powers system. The Army learnt two lessons in 1913' F i r s t l y , i t discovered that refusal to provide an Army Minister on terms disagreeable to i t could bring down any cabinet i t liked, provided, this tactic was not used too frequently. That the Army insisted to the bitter end on obtaining two new divisions in spite of i n i t i a l opposition from Inoue and Katsura, the urging of postponement by Oyama, an ambivalent attitude by Matsukata, and the support only of Yamagata, indicated a remarkable change from the practice of passive obedience to GenrS instructions to assertion of i t s own interests and then the attempt to win a l l i e s . Even though there 125 was no real conflict with the GenrS, the Army discovered that Yama-gata was to no small degree dependent on i t s institutional legitimacy, and that i t could have independent Cabinet-making and -breaking power If i t so chose. Secondly, the Army must have realised that i t lost i n 1913 precisely because i t had a l l i e d i t s e l f to an unpopular institution whose power was rapidly declining. The slogans shouted by the mobs indicted clan government and the Genro, not the Army. Indeed, the Navy's victory and the means by which i t was obtained could not have been lost on the Army leaders. So although Yamagata seemed to retain control of the Army, that institution's apparently sudden assertion of i t s autonomy in the 1930's should be seen i n the light of the lessons i t began to learn in 1913• Because the c i v i l bureaucracy had no way to pull down a cabinet or to force i t s representatives into cabinets, the loss of GenrS control over i t was less dramatic. In the early years, the GenrS had exercised control over appointments to i t s higher levels, from among whose number they had selected many Cabinet members. P o l i t i c a l control over the C i v i l Service had therefore been maintained by means of the power of patronage. But during the 1890's impersonal c i v i l service examinations removed from GenrS patronage f i r s t the lower- and middle-level positions, those of hannin and sSnin rank, and then the top-level positions which had been subject only to the laws of patronage, the 150 or so o f f i c i a l s of chokunin rank. Although examination screening was intended to remove from party control those top positions whose occupants 126 did most to draft legislation and make policy, i t s effect was also to reduce GenrS influence over the bureaucracy and promote the development of a separate bureaucratic interest. By 1920, 82$ of a l l bureau chiefs had passed the Higher C i v i l Service Examinations, and by 1923 three-quarters of the Vice Ministers had. But the removal in 1913-lU of a few of the top positions, those of Vice Minister and a few others, from examination screening, did make i t a bi t easier for a while to control the bureaucracy by means of patronage.^9 Hara Kei, the acute Seiyukai Home Minister in a number of cabinets, used this limited appointive power more successfully than anyone else, and brought about a certain degree of "party-ization of the bureaucracy."-^ To under-stand this "party-ization'V.a theoretical perspective is required. In separation-of-powers systems, because of party control of high bureaucratic positions, lower members of the service who seek promotion have an incentive to identify with one or other of the parties controlling appointments. Before the American C i v i l Service was brought under competi-tive examination, the "party-ization" of that country's bureaucracy was enormous. But even today a change in Administrations is accompanied by an extensive change of bureaucrats in comparison with the almost imper-ceptible shuffle of o f f i c i a l s that accompanies a new government in Britain.51 From the parties' point of view, there is a greater need in a separation-of-powers system for a new government that wants to ensure bureaucratic compliance to replace large number of o f f i c i a l s than there is in a parliamentary system. 127 The reason is that in a separation-of-powers system, in which the government is weakened by i t s need to placate a variety of i n s t i -tutions each wanting something different and each possessing a negative veto power, the task of introducing new positive policies and of carry-ing out the business of government f a l l s on the bureaucracy, which tends to benefit more than any other institution from the general competition because of i t s proximity to the executive. The only way for any of the other competing institutions to exert influence over such a bureaucracy i s to have persons sympathetic to i t appointed to key posts.52 In prewar Japan, although most key C i v i l Service positions were under the Higher C i v i l Service Examinations, some were not, and these were the ones ambitious bureaucrats needed to occupy i f they were to become members of the Cabinet, over which the parties were exercising considerable control by the TaishS era. By this time therefore, many more party men than Genro proteges were receiving high office in the C i v i l Service, and many more high o f f i c i a l s were becoming party members than were joining the Yamagata clique. With the formation of the Doshikai in 1913 ambitious bureaucrats no longer chose between the Seiyukai and the Genro as vehicles of upward mobility, but between the Seiyukai and the Doshikai. Najita wrote, "It [[opening the vice ministership to party patronage] meant that their positions no longer depended on the patronage of the Yamagata faction, but on the parties. Within three months, seven o f f i c i a l s of vice minister rank joined the 128 Seiyukai, and this resulted in further pressure on o f f i c i a l s at lower levels, particularly bureau chiefs, to join the party. The way these forces were reflected in the composition of cabinets was in the change from ministries including a large number of ex-officials who had advanced under GenrS patronage and a few who had advanced under party patronage to the opposite situation. Five out of the six ex-officials, excluding the Prime Minister, i n the fourth ItS cabinet, had beenJ.dependent on GenrS patronage, and the sixth, Hara Kei, owed a great deal to Mutsu Munemitsu. The f i r s t Katsura cabinet contained ten ex-officials, half of whom were Yamagata proteges. Of the nine ex-officials who served in Saionji's f i r s t ministry, five had advanced under the patron-age of some or other GenrS, three were more or less neutral between the GenrS and the parties, and one was a member of the Seiyukai. A l l except one of the ex-officials in the second Katsura ministry had risen in the world on Yamagata's coattails. The new trend began with the formation of Saionji's second govern-ment, which, apart from the premier, contained three ex-officials who were independent of the GenrS and the parties, and one who was a member of the Seiyukai. After the formation of the DSshikai, pressure on bureaucrats to identify with a party became stronger, and five of the seven ex-officials in Katsura's third ministry joined his party. The Yamamoto cabinet contained four ex-officials, three of whom were Seiyukai members and the fourth had become friendly to the Seiyukai. Reflecting the Genro-DSshikai alliance, the Okuma cabinet reversed the trend slightly, 129 and included in addition to Okuma, one e x - o f f i c i a l who was purely a Yamagata protege, and four who were members of the Doshikai, three of whom were also Genro proteges. Finally, of the eight ex-officials who served under Terauchi, one was a Seiyukai member, one a Seiyukai supporter, one an ex-member of the Seiyukai, and one an ex-member of the DSshikai. Only three were simply high o f f i c i a l s , and none were Genro proteges. Even though i n the Okuma and Terauchi ministries there was a temporary reassertion of GenrS influence on the composition of cabinets, i t remains certain that the GenrS by 1913 had become unable to promote to the highest levels of the bureaucracy large numbers of proteges. The parties had supplanted the GenrS in control over the bureaucracy. But party influence over the C i v i l Service can never be extensive in a separation-of-powers system, and in Japan the Justice and Foreign Ministries remained beyond the control of the parties, preventing the party-ization of the o f f i c i a l s of these ministries. Throughout the prewar period, these o f f i c i a l s remained loyal to their ministry alone. The same i s partly true even of the o f f i c i a l s i n ministries over which the parties did exercise control. The autonomy and esprit de corps of the C i v i l Service was manifested less in the number of Cabinet seats o f f i c i a l s occupied or in the number of o f f i c i a l s who joined parties than in each ministry's day-to-day influence on policy. Unlike the other institutions, the C i v i l Service possesses an 130 influence which i s not accurately indicated by i t s a b i l i t y to affect the composition of the Cabinet. This is because i t is more able to i n i t i a t e policy than other institutions, which possess only a negative veto power. Moreover, in Japan the C i v i l Service has been powerful because cabinets have been too weak under the system of a separation-of-powers to appropriate for themselves the important task of governing the country. In France, where cabinets, although for different reasons, have also been weak u n t i l very recently, the bureaucracy has seemed equally invincible. Weak cabinets tend to be more responsible to the bureaucracy than vice versa. The Privy Council had no power as an independent force in this period, because Yamagata was i t s President from 1905-1922, except for five months in 1909 when Ito took the job. The Council was so completely under Yamagata's control that there was no need for a cabinet to deal separately with i t , or to include representatives from i t . Most of the Privy Councillors who served in cabinets during this period did so primarily for reasons other than their membership of the Council. The sole exception was Saionji who, as President of the Privy Council, became temporary Prime Minister after Ito's resignation in May 1901. What was s t i l l true of the Privy Council i n 1918 had in earlier years also been true of a l l institutions besides the Lower House: domin-ation by the GenrS, who in fact gradually became synonomous with Yamagata. No other GenrS took so much trouble toccultivate the s k i l f u l use of p o l i t i c a l influence. Oyama and SaigS had been the least p o l i t i c a l l y 1 3 1 involved, while Matsukata and Inoue had been influential in the financial rather than the p o l i t i c a l world.55 When ItS's power base became the Seiyukai, which soon sought to act independently of the Genro, he suddenly found himself without a legitimate institutional base from which to operate. Yamagata seemed to have understood the actual workings of the Constitution much better than Ito, and as Najita wrote, had his followers "firmly entrenched in the major bodies of the Meiji con-stitutional order—the House of Peers, the Privy Council, the Imperial Household, the bureaucracy, the army."-'0' While in the early years i t would have been more accurate to speak of "the emergence of the Yamagata faction as the dominant group in the Meiji p o l i t i c a l struc-ture, " 5 7 Yamagata came to lose control of one after another of the various institutions and in 1 9 1 8 ended up in a position not unlike that of Ito. Institutional forces gradually overshadowed the remnants of feudal ones, as well as the prestige of the leaders of the Restor-ation, and with the formation of the Doshikai, caused an irreparable split along instututional lines in the Yamagata faction. As Duus vrote:? 8 The strength of the oligarchic generation lay in their enormous prestige and extensive connections as makers of the Meiji state, but they were powerful as individuals, not asasm institution. The very personal nature of their power meant they could not bequeath i t to their proteges. The decline of GenrS power and the growth of a vi r t u a l l y complete separation-of-powers had typical consequences, a l l centring on the d i f f i c u l t y of forming a cabinet that would be able to avoid consti-tutional crises and deadlock. Otsu wrote, "The decline of hambatsu power i s quite evident i f one considers that the d i f f i c u l t y of organ-ising a cabinet has increased with every cabinet change."59 To prevent deadlock, cabinets resorted to a variety of expedients, the most common of which was to create a sense of national emergency and purpose. According to Richard Neustadt, American Presidents are able to get more cooperation from other institutions on matters of foreign policy because of the greater ease in obtaining a general agreement that there i s a c r i s i s which necessitates compromises. He refers to this as " c r i s i s c o n s e n s u s . I n Fourth Republican France, the tech-nique of creating crises as means of obtaining compromises was developed into a fine art.0'"'" In Japan in 191.6, Yamagata called for national unity to deal with the crises he saw in domestic and international p o l i t i c s . Duus wrote, "Under the facade of national unity (kyokoku i t c h i ) , the military services, the House of Peers, and the p o l i t i c a l parties might be induced to bury their differences and cooperate to support the cabinet.' Inter-institutional deadlock, a characteristic of separation-of-powers systems, had consequences in Japan that cannot be ascribed entirely to socio-economic forces. The cultures and economies of Japan and America were too different for the occurrence of similar phenomena not to be at least partly due to the existence of similar institutional forces. 133 Typical Consequences of the Separation-of-Powers Wo attempt is made in the present section to assess the role of socio-economic forces in bringing about the occurrence of phenomena like corruption and personality p o l i t i c s . These forces may have been their major cause in particular cases as well as a general cause through-out the period. I simply intend to demonstrate that p o l i t i c o - i n s t i -tutional forces were at least also partly, possibly even predominantly, responsible for the great emphasis politicians placed on personality and their free use of corruption, as well as for the gradual growth of public apathy towards constitutional government. The prevalences of these phenomena in Japan in 1900-1918 in particular i s noted by most scholars, but almost a l l of them either ascribe their occurrence to social forces, or regard them as normal in democratic societies. I intend to show that they are at least in part the result, not necessarily of democratic p o l i t i c s , but of poli t i c s in a separation-of-powers system. The f i r s t typical consequence of a separation-of-powers system i s the growth of personality as opposed to policy p o l i t i c s . Because each of the institutions can exercise a negative veto over policy, i t is most unlikely that any particular institution w i l l be able to have i t s policy preferences agreed on by the rest. The promise to implement certain policies i s not something any institution w i l l make in i t s attempts to win public support. Uyehara saw this very clearly 13h No p o l i t i c a l party i n Japan, therefore, has a specific concrete p o l i t i c a l programme. It is useless for i t to draw up one, for i t could not put i t into practice, even i f i t had a majority in the House. In the same way the Ministry never puts i t s own p o l i t i c a l programme definitely before the public. For i t is also uncertain whether the Ministry can carry out i t s measures to the extent i t desires, as i t largely depends upon i t s opp-ortunity of controlling the majority in the House. In the absence of a policy programme on the basis of which to seek support, personality becomes the central way candidates in elections distinguish themselves from one another. It is not fortuitous that personality plays such a large role i n Congressional and Presidential elections in the United States, although the law requiring the local residence of candidates in Congressional elections does introduce an additional factor. In Britain, someone almost entirely unknown to a local community can be elected purely on the basis of a promise to support a government of a certain kind. Members of the House of Commons owe their election to their party allegiance, not to their personal qualities. The separation-of-powers makes i t necessary for anyone seeking election to have the electors well acquainted with him and to convince them that he has the personality and s k i l l to do as much as possible for them in the actual inter-institutional struggles over policy. Personality p o l i t i c s in Japan i s not necessarily a feudal remnant that has stubbornly refused to disappear, but partly the result of new forces which were most pronounced during the years 1900-1918. One of the best examples of the influence of personality was the victory of the Doshikai in 1915 because of i t s association with an enormously popular Okuma. Yamagata could not have been as influential in a parliamentary system as he was in the Meiji separation-of-powers system. Equally important were the thousands of electoral candidates whose personal appeal was considerably enhanced through their a b i l i t y to dispense favours immediately in the form of cash. Corruption is the most characteristic consequence of a separation-of-powers constitution,aand the most widely used of a l l dei ex machina for resolving institutional deadlock. The flow of money from one institution to another, as well as from candidates in elections to electors, whether of the Upper, House or Lower House, had become quite common already in the l890's. On more than one occasion governments had given Dietmen money to gain support for budgets, as well as can-didates in elections whom i t hoped would be "loyal." The practice was so widespread and notorious that i t requires l i t t l e documentation, but i t does need more adequate explanation than has hitherto been offered. Part of the problem is that electoral corruption i s regarded by western scholars, most of whom are American, and who in recent years have not paid much attention to institutional forces, as quite normal. Duus refers to i t as "the increasing cost of p o l i t i c s , " and.in his frequent attempts to compare Japan to Anglo-Saxon countries, never distinguishes between Britain and the United States. The quantitative difference in the "cost of p o l i t i c s " between Britain on the one hand, and Japan and the United States on the other, i s so great as to be a 136 qualitative difference. Duus himself shows that whereas i t cost a typical Dietman Y2,500 to win a seat in I 8 9 O , i t cost Yl6,000 in 1924, and he admits that "much of the increased election expenditure went into the direct buying of votes," a practice which was wide-spread in Britain only in the eighteenth-century. In 1908 there were 2,338 reported cases of bribery in Japan, in 1924 there were 6k 13,986/ Not a l l contemporaries were unaware of the institutional forces promoting corruption. Uyehara compared i t to the "methods of Walpole," the master of resolving deadlock between the King and Parliament in eighteenth-century Britain by means of money flows, and the "American p o l i t i c a l 'Bosses,'" and asked rhetorically, "For, how else can a Ministry independent of the House of Representatives get a parlia-mentary majority to pass i t s necessary l e g i s l a t i o n ? M c L a r e n was another comtemporary to understand the consequences of a separation-of-powers constitution. He wrote, "Frequent elections meant financial ruin to members of the Lower House, and an unyielding opposition to the 66 Government destroyed a l l hope of profit," something which can never be said of elections in a parliamentary system. Money rather than programmes win elections under a separation-of-powers, and one of the few ways to get sufficient money is to make the government pay a high price for the House's support. One of the consequences on the body p o l i t i c of this kind of 137 politics i s a high level of public apathy. Again Duus i s the most accurate in describing the phenomenon but unable to identify i t s causes. He notes that in the 1890's most newspapers had been organs of the freedom and c i v i l rights movement, but that by 19lk the large dailies like the Asahi and journals like Chuo Koron had become independent and anti-establishment. "At i t s mildest, the new climate of public opinion was marked by apathy toward the game of party p o l i t i c s . To many c r i t i c s , i t seemed a mere mechanical struggle for power, devoid of any change of p o l i c y . " ^ But each of the typical com-plaints mentioned by Duus i s related to the separation-of-powers constitution and has close parallels in the United States but not Britain (1) The parties lacked fixed programs; instead of pursuing fundamental principles, they were guided almost exclusively by considerations of expediency in the struggle for power. (2) The parties compromised away the advantages of the people in their dealings with the bureaucrats, the military, the genro, and other nonparty elements. . . . (3) The parties were far more sensitive to the demands of special interests than they were to the needs of the people as a whole; party members worked to secure benefits for their local constit-uents or their businessmen backers, not to promote the national welfare, (k) The overriding concern of Diet members with winning office, building local jiban, and raising election funds resulted in endemic corruption i n the parties. Duus concludes that "such practices were necessary to party survival," but when he discovers them in the United States as well, he says they "are probably endemic in any representative system."^9 Public apathy, personality p o l i t i c s , and high levels of corruption 138 in prewar Japan must have resulted at least partly from politico-institutional forces. The logic of institutional relationships makes their occurrence more l i k e l y in separation-of-powers systems than parliamentary systems. Because in prewar Japan the existence of the Japanese executive was much more precarious than i t s American counter-part, coordination among Japanese institutions was much more d i f f i c u l t to achieve than in America. The result was even stronger institutional pressures in favour of personality p o l i t i c s , corruption, and public apathy than exist in America. By I918 the Japanese Constitution was without a central coordinating institution. Once the Genro lost at least partial control of a l l i n s t i -tutions besides the Privy Council, the inadequacy of the Emperor as an institution created to knit the Constitution together had been l a i d bare. A concluding word w i l l show why the Emperor system failed. Conclusion: The Emperor System and the Triumph of the Separation-of-Powers The Emperor was intended, and on occasion did, function as the most instantly effective of the Constitution's del ex machina for resolving inter-institutional conflict. Once the ideology that His commands required absolute obedience had become an accepted tradition, i t became possible for any institution which had his confidence to get an Imperial Rescript to bring opposing institutions into line. In the I89O's the Genro used Imperial Rescripts to bring the Lower House to cooperate with them; in 1901 Ito used an Imperial Rescript to gain Upper 139 House support for his budget, and in 1913 Katsura used one against the Navy. As long as they were not used too frequently, Imperial Commands instantly resolved deadlocks. So although Katsura under-estimated the degree of ho s t i l i t y which the Navy and Seiyukai bore him in 1913, i t i s none the less surprising that the party agreed in the end to disobey an Imperial Command. But because the con-sequence of a single disobeyed command was, as Najitappoints out, that "never again would a prime minister use the p o l i t i c a l prerog-atives of the Emperor against a p o l i t i c a l party in the Lower House,"7° i t became extremely important to emphasise the absolute binding nature on a l l subjects of such a command. Otherwise the Emperor would cease to be a deus ex machina, and either become another contender in the struggle like eighteenth-century British monarchs, or end up like modern British monarchs, merely performing ceremonial functions. In a sense therefore, the Emperor became as powerful as leftwing Japanese scholars allege he became, not merely because of the constant stream of government propaganda emphasising the absolute binding nature of his w i l l , but because of the effects on the popular mind of the many examples of instant obedience to his commands. A l l this, as well as the deference the government received by being accorded the privilege of "assisting" the Emperor, was a tremendously powerful weapon against uncooperative institutions. Essentially i a deus ex machina to resolve inter-institutional conflict, the Emperor system and the propaganda on which i t s effectiveness depended, worked against iko a l l kinds of conflict. From another point of view, however, the Emperor lacked sufficient power to remain effective as a deus ex machina. To he permanently employed in this role, the Emperor would become controversial, and would gradually be deprived of the mystique that gave his commands their special binding force. His effectiveness depended on a sparing use of Imperial Rescripts. This is why there was always more propa-ganda emphasising the absolute binding nature of the sovereign imperial w i l l than actual use of Imperial Rescripts against uncooperative institutions. There was another reason for the continuous emphasis by representa-tives of a l l institutions, particularly those without a popular base, on the sovereignty of the imperial w i l l and oncthe .extent of the royal prerogative. If each of the separate institutions wanted to maintain the legitimacy of i t s own sovereignty i n i t s special area, i t s most effective action was to insist on i t s own monopoly of the right to advise a theoretically a l l powerful but actually powerless Emperor on matters in that area. The Army, for example, particularly in the 1930's, insisted both on the sovereignty of the Emperor and on i t s sole right to advise him on foreign policy. In this way the ideology of the Emperor system servecL. to perpetuate the separation-of-powers system, paradoxical as i t may seem. In a passage which reveals how close he came to seeing that the model according to which prewar Japanese pol i t i c s should be analysed i s the separation-of-powers one, Duus wrote: The emperor was not too powerful in law, he was too weak in practice. In the absence of a central unifying force within the state, the power to make and carry out decisions was divided not in Montesquieuean fashion, but among an autonomous military high command, an independent professional bureaucracy, the.' leaders of the House of Peers, the Privy Council, and the parties in the House of Representatives. While the oligarchs lived, they managed to hold together this welter of competing elements by a kind of 'government by crony.' But once they had passed from the scene, the problems created by the fragmentation of power within the state became acute. The p o l i t i c a l leaders who emerged in the Taisho period were divided not simply by policy differences, bit by different bases of power and in some cases by radically opposite philosophies of government. The result was a continual struggle for control of the cabinet. Duus's conclusion can be improved upon by applying separation-of-powers theory to the reasons for Cabinet composition as well as i t s rise and f a l l . Although Appendix I indicates that the GenrS and the Lower House were the main makers and breakers of the Cabinet throughout the years 1900-1918, i t also indicates that only the f a l l of the Terauchi government resulted solely from the influence of only one institution, and that more than two institutions brought about the rise and f a l l of quite a few cabinets. The Cabinet was in no sense responsible to any institution, because i t s rise and f a l l regularly depended on more than one institution. The Cabinet's existence was therefore separate from each institution: as far as each was concerned, the Cabinet had an independent existence. The Cabinet's independent existence was always more visible in Cabinet-making than Cabinet-breaking. The cooperation of more i n s t i -tutions was necessary to raise a cabinet than to pull one down, and lk-2 Cabinet composition reveals even influence of a l l institutions except the Privy Council. The development of a full-fledged separation-of-powers following the break away of a l l institutions except the Privy Council from the control of the GenrS, who also became merely one of a number of con-tenders, characterised the institutional development of the period. It was unlikely that this development would suddenly be reversed and give way to a parliamentary system overnight. Later developments were to demonstrate that the separation-of-powers was as much a character-i s t i c of the Constitution i n the 1 9 2 0 ' s as i t had been in the previous decades, in spite of the dominant body of opinion-'to the contrary. ±k3 CHAPTER h PARTY PREMIERS AND THE PERSISTENCE OF THE SEPARATION-OF-POWERS, 1918-1932 Most contemporary as well as present-day observers of Japanese poli t i c s see in the years 1918-1932 a f a i r l y rapid development towards institutions like those of Britain, mainly because of the large number of party premiers and Cabinet ministers who served during these years. The presence of party premiers and Cabinet members, however, did not result in an increasing degree of party government, or a greater con-centration of power in the Cabinet. Cabinet-making and -'breaking power remained in the hands of a number of competitive semi-autonomous i n s t i -tutions, none of which showed any real sign of surrendering this autonomy. Indeed, one institution, the Privy Council, which had hitherto been under the firm grip of the Genro, came into i t s own in this period and joined the ranks of the active Cabinet-makers and -breakers. The agreement of the Genro to nominate mainly party Prime Ministers, who in turn chose mainly party Cabinet members, did not imply an agreement to a de facto amendment of the Constitution through working to undermine the power of the other institutions. Not infrequently, the GenrS took the side of one of the other institutions against the party with a majority in the Lower House. The reasons for' the failure of Japanese p o l i t i c a l institutions to evolve out of the separation-of-powers l i e in the timing of the occurrence ikk of certain crucial p o l i t i c a l forces, not merely of industrialisation and capitalism, as argued by Robert Scalapino. Of the greatest importance was the order in which p o l i t i c a l phenomena like the adoption of universal suffrage, the establishment of parliament, the establish-ment and abolition of the separation-of-powers, and the development of a unitary state came to exist. The evolution towards strong responsi-ble cabinets in Britain was very much the result of a specific order in which these phenomena developed in that country, and the lack of the same in Japan was due to their development in an entirely different order. Background The social and economic conditions which had been largely constant during the previous period underwent a sudden upheaval in the 1920's, largely due to the effects i n Japan of the First World War. The new situation in the country contributed greatly to the most important characteristic of Japanese institutional development in 19i8-1932: party control of the premiership and of most Cabinet seats. As far as the parties themselves were concerned, the change was not a l l that great. Their dominant influence on Cabinet composition was something they had come to regard as normal, and in the years 1918-1932 they behaved very much the same as before. But while i t was business as usual for the parties, growing dissatisfaction among the peasantry and urban working classes developed. Even after the adoption lh-5 of manhood suffrage i n 1925, the two main parties remained represent-ative of the wealthy classes alone and continued the struggle against the ri s i n g proletariat. The members of other institutions were on the whole content to allow the bourgeoisie in the Lower House to manage the social problems of the period, although they were not content to tolerate party control over foreign policy. Their interference in the conduct of foreign policy increased as foreign affairs came to occupy the centre of the stage in the late 1920's and early 1930's. The world depression which hit Japan in 1930-1931 made other institutions also less willing to leave the management of the nation's finances and economy in party hands, and interference i n these matters characterised the depression years as well. The economic changes wrought by the First World War were phenomenal. In the I87O's manufactures, which comprised about 30$ of commodity production, consisted mainly of food products and textiles, and the modernization of these two industries through imported raw materials and machines was the typical characteristic of economic change u n t i l the outbreak of the First World War. On the whole, this f i r s t phase of industrialisation was supported mainly by an expansion of the traditional agricultural sector, an expansion which began to. decline around 1905. The proportion of investments in the modern sector increased gradually and this sector became more able to support i t s e l f without the assist-ance of agriculture: tea and raw s i l k exports gave way to the outputs 146 of modern technology like s i l k and cotton fabrics. Until the War the import of manufactures greatly exceeded their export, but after the War the reverse was true. The boom of 191U-1919 brought with i t unprecedented prosperity for the business class, but the resulting inflation exacerbated the inequalities in the country between the industrial and agricultural classes on the one hand, and the capitalist and urban working classes, on the other. The f i r s t expression of lower class discontent was the rice riots of 1918, a discontent which became even more widespread following the slump that began in 1919 and lasted well into the 1920's, only to be deepened by the World Depression. Japanese capitalists responded to the post-war slump by rationalising industry and forming oligopolies, which became known as the zaibatsu (financial cliques), and each of the two major parties a l l i e d themselves closely with one of these oligopolies: the Seiyukai with the House of Mitsui, the Kenseikai with the interests of Mitsubishi. The lower classes now directed their attacks more point-edly against the two main parties, although the peasantry lagged behind the industrial working class i n detaching i t s e l f from the capitalist parties. The new working class, largely a creation of: the war boom, began to organise i t s e l f and prepare for a confrontation. The to t a l number of factory workers increased by 60$ from just under 1 million in I91U to just over 1.5 million in 1919 • The percentage of workers in textiles dropped from 6Cffo of the t o t a l to 55$, while the 147 percentage in the machines and chemicals industries rose from 17$ to 27$. The proportion of the population employed in the agricultural sector, however, remained above 50$ throughout the 1920's.2 Even during the war boom, the real wages of factory workers remained almost constant, and during these years there was a great upsurge in working class organisations: the number of new unions and labour disputes was greater than ever, particularly i n the manufacturing sector. Although the changes in social and economic conditions took place largely during the latter part of the previous period, their impact on the body p o l i t i c took place i n the 1920's. The rice riots of 1918 form a convenient break not only in institutional history, but also i n social history. They ushered in a period of party dominance over the Cabinet along with mass dissatisfaction with a l l parties and the classes they represented. In addition to a realignment of social forces and a partial realign-ment of p o l i t i c a l forces, the 1920's witnessed an awakening assurance about the future of "democracy" in Japan. Although this upsurge in interest in l i b e r a l ideas was mainly confined to intellectuals and journalists like Yoshino;. SakuzS and Baba Tsunego, i t was b r i e f l y able to hold the imagination of the lower classes. The agitation over uni-versal suffrage in the early 1920's owed a great deal to the ideas expressed in daily newspapers like the Asahi and semi-intellectual monthly magazines like Chu5 Koron. Although the lower classes soon lost their fervour over universal franchise, the new liberals remained an important influence i n bringing about i t s introduction i n 1925. Their cry was less for party control over the Cabinet than for the democrat-isation of the parties themselves. Their criticisms of the parties were echoed by other institutions and became an important factor in the parties' i n a b i l i t y to achieve complete control of the Cabinet. The alienation of the lower classes from the parties also assisted non-party institutions i n their efforts to prevent the development of party government. Through their greater emphasis on the right to strike and to organise rather than the right to vote, the lower classes contradicted the parties' claims to represent the nation. Even though the parties did not command''the affections of the media and the lower classes, i t i s not surprising that they exercised the dominant institutional influence in the 1920's. The major national problems were financial, industrial, and social, problems with which the capitalist class represented by the parties came into closest contact. Party influence on Cabinet composition and i t s rise and f a l l was l i k e l y to exceed that of the other institutions, which represented classes antagonistic to the lower classes but not yet antagonistic to the capitalists. The purpose of the present chapter i s to ill u s t r a t e that the Constitution functioned, as before, according to the rules of a separation-of-powers system. Less detail than before i s devoted to the reasons lk-9 for Cabinet changes, the major emphasis being on providing evidence that no fundamental change had taken place. There i s more emphasis on shoving that the Constitution had not become a parliamentary type than on showing that i t was of a separation-of-powers type. My evidence is therefore more anecdotal, and often only examples of separation-of-powers practices are presented. I feel i t would be tedious to demonstrate in the detail of the previous chapter that l i t t l e had changed. I also assume a greater knowledge of po l i t i c s in the l Q 2 0's, partly because the study's emphasis i s on the interpretation of events rather than their document-ation, but partly because Japanese p o l i t i c a l history in the 1920's is better known even by non-Japanese specialists than i t i s in other periods. It i s well known that in 1918-1932 the Japanese Cabinet, with a few exceptions, was under party prime ministers who appointed mainly party cabinets. What has not been recognised is that in i t s so-called heyday, party government never really existed. Compromises with other institutions were as apparent in Cabinet composition and i t s rise and f a l l as ever before. Party Premiers could not govern according to pre-determined party programmes but had to act according to their judgement of what was l i k e l y to provoke any of the other institutions into the kind of opposition that would result in their own demise. The choice that faced party prime ministers was the same as that faced by any previous occupant of the office: compromise with the other institutions, or deadlock, which 150 latter event meant p o l i t i c a l suicide. And as in Fourth Republican France, where the general mood favouring a compromise was always stronger after the c r i s i s of a cabinet change, the institutions of prewar Japan limped along through a series of successive crises and compromises. The long established modus vivendi between the Genro and the Lower House remained intact throughout this period, except b r i e f l y in 1922-2U, when the GenrS refused to nominate party men to the premiership. But the Genro-House agreement never showed signs of developing into the kind of alliance that grew up between the Monarchy and Commons in nine-teenth-century Britain, where the Monarch came to choose only majority party cabinets. On a number of occasions, the Genro acted against the majority party in the House, and allowed cabinets which enjoyed the latter's confidence to f a l l when other institutions flexed their muscles. The cardinal principle of parliamentary government, that majority parties and only majority parties form the Cabinet, was violated by the GenrS again and again. The GenrS-House Compromise in Cabinet-Making and -Breaking, I918-I92U The present section describes solely the various forms the Genro-House compromise took in 1918-1924. The roles of other institutions, which are discussed more f u l l y in the next section, are mentioned only in so far as they affected Genro-House relations in Cabinet-making and -breaking. I refer the reader to Appendix I for a summary account of the major and minor roles of Cabinet-makers and -breakers in these years 151 and proceed directly to explain the roles assigned. The great change wrought in 1918 with Hara Kei's appointment to the premiership was Genro recognition that party premiers could become the normal practice. Previously, the Genro had only reluctantly agreed to party men in this office and eagerly awaited an opportunity to replace party premiers with non-party men. Although Hara's appointment was not made with the intention of having a party man succeed him, after 192U party leaders occupied the office of Prime Minister u n t i l the end of the period, very much because Yamagata found Hara to have been an ideal choice .. As in the years 1900-1918, the Genro formally met to nominate Terauchi's successor. Following the deaths of Inoue Kaoru and Oyama Iwao in 19l6, their number now only included Yamagata, Matsukata, and Saionji. Although Yamagata wanted to appoint either Saionji or his protege, Hirata T5suke, Hara Kei was prepared to have his party take part in the rice riots should Terauchi f a i l to resign and agree to support Hara as his successor. Yamagata therefore eventually agreed on Hara for two main reasons. F i r s t l y , he was f u l l y aware of the pressure in the country for a party man to succeed Terauchi following the rice riots, and could think of no alternative to Hara as the man most l i k e l y to restore order. Secondly, after working so closely with Hara for the past fifteen years, Yamagata had come to respect Hara as an i n d i v i -dual who agreed with him on a l l matters except that of regular party prime ministers. While the Hara cabinet was a predominantly Lower House 152 creation, Yamagata's willingness to tolerate the Seiyukai President in the office he wanted to reserve for non-party men means that the GenrS also played a part, although a limited one, in Hara's appoint-3 ment. The f a l l of the Hara cabinet following the assassination of the Prime Minister need not detain us for long. Wo attempt has been made to identify institutional pressures that would have made Hara's resigna-tion l i k e l y . Labour relations and the agitation over universal suffrage were regarded by the Seiyukai President as security problems, and he had no need to fear the withdrawal of Yamagata's support. The Upper House was no source of trouble, and the working agreement Hara established with i t s Yamagata-controlled largest faction, the Kenkyukai, was in no danger of being abrogated by an aristocracy apprehensive about the possible consequences of universal suffrage.^ The appointment of Takahashi resulted from the identical line-up of forces that brought his predecessor to power, although the absence of public rioting in late 1921 made the avoidance of a party premier an easier matter than in 1918. There was no question of appointing the leader of the opposition party, which had taken part in the suffrage agitation. Because Saionji f e l t i t would set a bad precedent to turn out a party-dominated government because the party leader had been assassinated, he got Matsukata and Yamagata, who died soon afterwards, to support another Seiyukai man. But although both the House and the 153 GenrS were once again the sole institutions responsible for the ultimate choice, their influence was more equal this time. Takahashi only became the Seiyukai President after he became Prime Minister. Saionji was hard-pressed to find someone in the Seiyukai whom he f e l t , as Matsumoto Gokichi, Yamagata's and subsequently Saionji's private secretary, said "had the capacity (chikara) to become Prime Minister." He settled on Takahashi for two reasons. F i r s t l y , according to Matsumoto, Saionji regarded i t as a point in Takahashi's favour that "he was not a real member of the Seiyukai."5 Takahashi had been an indifferent party member, and belonged to the bureaucratic faction in the Upper House, the Chawakai (The Tea Group). Secondly, as Finance Minister i n the Hara cabinet, Takahashi had occupied the highest Cabinet, as opposed to party, position in the Seiyukai. The role of the GenrS was therefore more crucial in 1921 than in 19l8> even though the institutional line- up was the same and the GenrS-House compromise took the same form. The f a l l of the Takahashi cabinet, usually attributed simply to internal Cabinet disunity, had a great deal to do with the activities of the dominant groups in the Upper House as well as the prejudices of the GenrS. The dissident cabinet members, Nakahashi TokugorS and Motoda Hajime, were the ones whose projects had been rejected by the Upper House under the leadership of the Chawakai, a group with which the Prime Minister continued to have cordial relations. But the Kenkyukai, the largest and most i n f l u -ential group, suddenly indicated i t s readiness to change sides and support 15 k the two ministers against the Premier. Takahashi's only way to purge the dissidents, something British prime ministers are always doing, was through a general resignation, after which he hoped to be renominated and to construct a new cabinet without them. That the party endorsed the Prime Minister's actions and expelled the party dissidents under-lines the c r i t i c a l nature of the choice the Genro made in not renominating the man who had really only resigned in order to restore party discipline. The Genro and the Kenkyukai were more responsible for the cabinet's collapse than was Seiyukai disunity, which in fact only became serious after i t was clear that the Genr5's personal antipathy towards Takahashi 7 was preventing the rise of another Seiyukai cabinet. The Genro pleaded party disarray for their failure to nominate a party premier during the years 1922-1924, but this party disarray was more a consequence than a cause of the parties' distance from power. The GenrS did not see that the more a cabinet's existence depends solely on party unity, the greater w i l l that unity be, and vice versa. Wot having the need to keep a government in power as a unifying force, the Seiyukai and SeiyuhontS (Orthodox Seiyukai), the two parties into which the Seiyukai divided a year after Takahashi's resignation, themselves tended to become more factionalised. But the failure to nominate the leader of the Kenseikai during these years, was more due to the Genro 's dislike of i t s leader, KatS KSmei, than their fear of nominating the leader of a party without a majority in the House. Minority party leaders had been 155 nominated before and had always been able to create comfortable major-i t i e s in the subsequent elections they managed. By choosing KatS TomosaburS, Yamamoto Gombei, and Kiyoura Keigo, the GenrS boosted the legitimacy of the institutions on which their cabinets were primarily based, reinforcing the already well established separation-of-powers. The actual rupture of the GenrS-House compromise took place after the f a l l of the KatS TomosaburS government, and was only restored in 1924 when KatS KSmei was at last appointed to the office the GenrS had denied him for so long. The following section deals with the breakdown and restoration of the compromise, beginning with KatS TomosaburS's selection. The rise of the Kato Tomosaburo cabinet involved a complex line up of almost a l l institutions, whose order of importance is d i f f i c u l t to determine. The Seiyukai leaders were not considered e l i g i b l e , but some-one acceptable to some party or other had to be found. Because Saionji did not want to nominate the Kenseikai leader, Kato KSmei, he "became i l l , " and l e f t the negotiations up to Matsukata. Matsukata consulted the President of the Privy Council, Kiyoura Keigo, who f i r s t suggested Kato KSmei and then Admiral Kato Tomosaburo, who had overcome the opposition of the Naval General Staff to the cuts in naval expenditure negotiated by himself at the Washington Conference in 1922 and had achieved great popularity in the country. Matsukata then consulted the guardian of the Navy's interests for the past two 156 decades, Yamamoto Gombei, an opponent of the Washington agreement. The Navy Admirals held a meeting and advised Kato Tomosaburo not to become Prime Minister. When Matsukata saw that KatS Tomosaburo was reluctant to accept, he began to make plans for a Kato Komei cabinet. As soon as Tokonami Takejiro, one of the leaders of the anti-Takahashi faction of the Seiyukai, heard of this, he advised KatS TomosaburS to accept the nomination, apparently pledging Seiyukai support. Kato Tomosaburo then visited Takahashi, and although the Seiyukai did not agree to Kato]s request that Tokonami enter the government, the party gave him i t s support. Yamamoto Gombei then also concurred. It seems that the Navy's i n i t i a l reluctance was i t s fear that the cabinet's failure might harm the Navy i t s e l f . The President of the Privy Council also consented to KatS's nomination. Finally, Mizuno Rentaro, an influential member of the KSyu Kurabu (Friendship Club) in the Upper House and a member of the Seiyukai since 1913, strongly urged Kato to accept the nomination. Matsukata had been prepared to have Kato Komei serve, but when the inter-institutional struggle resolved i t s e l f in favour of KatS Tomosaburo, he went along with the general consensus. But the GenrS, because they would not have a Seiyukai premier, wielded at least a minor influence in the rise of the cabinet. The same applies to the President of the Privy Q Council, whom Shinobu describes as the "midwife" of the cabinet. The Navy, the Seiyukai, andtthe Upper House, on the other hand, played major 157 parts. The Seiyukai effectively prevented the rise of a Kenseikai cabinet, while the Wavy agreed to act as the Seiyukai's buffer against the Kenseikai. Through Mizuno Rentaro's pledge of support, his a b i l i t y to win over the Kenkyukai, and his almost complete freedom in selecting Kato's ministers, the Upper House exerted i t s influence. Matsumoto described the new cabinet as "actually a Mizuno Rentaro cabinet."9 The resignation of the Kato cabinet resulted from the illness of the Prime Minister, who died of stomach cancer soon afterwards. Shinobu says the government's resignation was "a sudden change in the p o l i t i c a l situation.""'"0 But the Genro-House compromise ended with Yamamoto's selection. The institutions raising the second Yamamoto government were, as i s to be expected under the circumstances, almost the same as those raising i t s predecessor, although the parts played by each were different. Because the Genro nominated Yamamoto directly to the Emperor, without f i r s t consulting Yamamoto, i t is not easy to assign weights to each i n s t i -tution. The degree of post facto support each gave and the tussel over Cabinet seats are the only ways to resolve the problem. The Genro's role was more .crucial this time: Saionji assumed the entire responsibility of appointing Yamamoto- and, according to Matsumoto, fe l t that "the leaders of the Kenseikai and Seiyukai were not equal to the task. Neither major party, both of whose leaders Yamamoto approach-ed with offers of Cabinet positions only to meet with refusal, played any role at a l l . Only Inukai Tsuyoshi, the leader of the smaller Kakushin 158 Kurabu (Renovationist Club), formed in March 1922 out of Kenseikai dissidents and the old Kokuminto (People's Party), entered the cabinet and assisted Yamamoto in the Lower House. The Kenkyukai, the largest Upper House faction, had wanted to extend the previous government and have Okano Keijiro take over the premiership. Kenkyukai refusal to support Yamamoto therefore requires that no role be assigned to the Second Chamber, even though Peers from other factions entered the cabinet. Because Yamamoto tried to construct a kyokoku i t c h i (national unity) cabinet and went out of his way to s o l i c i t the support of powerful bureaucrats like Goto Shimpei, Den Kenjiro, and Hiranuma Kiichiro, a l l of whom entered the cabinet and became i t s " p i l l a r s , " the C i v i l Service wielded an important influence in the cabinet's appointment. The struggle between Goto and Den for influence in the bureaucracy underlines the importance of Yamamoto's inclusion of both men. So does Saionji's encouragement of Yamamoto to work with Den. Because Ito Miyoji, one of the original drafters of the. Constitution arid a powerful member of the Privy Council, strongly urged Hiranuma and Okano Keijiro to enter the cabinet, and because both men were made Privy Councillors soon afterwards and became the Council's Vice Presidents—Okano in 1925 and Hiranuma in 1926^ -that institution's influence was at least a minor one. Finally, the Navy i t s e l f , through i t s most influential member's agreement to keep this welter of institutional representatives together, also played a large part in the making of the Yamamoto cabinet.^ 159 The distance between the Genro and the Lower House had hardly been greater since Katsura's appointment in 1913, although the House's actions were far from as dramatic. Following the great earthquake in Tokyo just after the new Prime Minister's appointment in 1924, both Houses of the Diet, the only two institutions not included in the alliance behind Yamamoto, began to work for his f a l l . Home Minister Got5 Shimpei tried to use the reconstruction of the city as a way to implement his policy to widen and increase the number of roads. The Seiyukai regarded this as unnecessary and did not want to let Goto take the credit for i t . The Lower House cut the budget by 20$ and the Peers passed the amended budget. Although Goto f i n a l l y backed down, his colleague and r i v a l , Den KenjirS, did not accept the Lower House's refusal to sanction his plan to lend money to f i r e insurance companies and resigned. Before the Prime Minister had decided what to do about Den's resignation, there was an attempt on the l i f e of the Prince Regent. Yamamoto tendered a general resignation of the cabinet, ostensibly taking responsibility for the incident. The Genro told him there was no need for the cabinet to resign, and Yamamoto consulted his colleagues, who replied that the government could not continue in office. The govern-ment's relations with the two Houses of the Diet had made i t s continuance in office impossible.^-3 Even though the Lower House had acted as a major force in unseating Yamamoto, the Genro did not yet restore the compromise. Their distrust i6o of Kato Komei and Takahashi exceeded their fear to provoke the parties into united opposition. Two other factors prompted Saionji, with Matsukata's approval, to nominate Kiyoura Keigo. F i r s t l y , Saionji did not want the forthcoming election to he "managed" hy either of the two parties and believed the President of the Privy Council to be the most neutral force between them. Secondly, Saionji chose Kiyoura because Kiyoura recognised that constitutional government required some working agreement with the Lower House. Saionji hoped Kiyoura would conclude such an agreement with the Seiyukai. The f i r s t step towards restoring the compromise had been taken. Developments did not proceed quite according to plan. After the Genro by-passed Takahashi for the third time, the anti-Takahashi faction of the party seceded and set up the Seiyuhonto (Orthodox Seiyukai), and Kiyoura could only come to terms with the smaller Seiyuhonto. The Lower House therefore played only a minor role in the rise of the Kiyoura cabinet. The Privy Council, of which Kiyoura was the President, and the Upper House, from which he selected almost a l l his ministers, both played major parts."'"4' The Seiyukai, Kenseikai, and Kakushin Kurabu responded to the new appointment by organising another Movement for Constitutional Govern-ment. Unlike the similar response in 1913, the parties in 1924 were unable to excite the public imagination, although the three concerned managed to win more seats in the election than the pro-government Seiyuhonto: their combined strength was 286 against the Seiyuhonto's 114. l 6 i Kiyoura resigned soon after the victorious parties translated their electoral alliance into an agreement on a minimal government programme.^ Shinohu describes the GenrS kaigi, the last attended by Matsukata, whose death soon afterwards l e f t Saionji as the sole GenrS u n t i l I 9 U 0 , as "a mere formality.""*"^ Past experience, however, indicates that the GenrS did have an alternative: they could have nominated the SeiyuhontS leader or someone more acceptable to that party than Kiyoura. If the Seiyukai had not immediately deserted the three-party-alliance and reunited with the SeiyuhontS, a SeiyuhontS managed election would in a l l probability have induced a merger. Although the predominant pressure raising KatS Komei came from the House, the GenrS played at least a minor r o l e . ^ From the f a l l of the Kato Tomosaburo cabinet t i l l the rise of the KatS Komei cabinet the leaders of the two main parties worked hard to restore some kind of compromise with the GenrS. The latter had learnt to manipulate the Constitution with considerable s k i l l and had kept government functioning without party premiers and with only three party cabinet ministers since Takahashi's resignation. They had constantly shifted the main institutional alliances behind successive governments, alliances which acted as counter-weights to the influence of the Lower House. Because Saionji did not believe party premiers to be the abnor-mality his predecessors did, he was not unwilling to nominate party members i f this would make his job to ensure the smooth working of 162 constitutional government an easier one. From 1924 u n t i l 1932 Saionji found party premiers the best way to f u l f i l his duty. The GenrS-House Compromise in Cabinet-Making and' -Breaking, 1924-1932 In the present section I simply intend to show that the selection of party premiers was a compromise and not a de facto amendment, even a temporary one, to the Constitution. Although a l l premiers in these years were party Presidents, Saionji only nominated majority party leaders when a premier with a majority died. On a l l other occasions minority party leaders were nominated. More pertinently, a l l prime ministers in 1924-32 who did not die in office were unseated by an institution other than the Lower House, without any objection from the GenrS. That the discretion of the Genro was enormous, and actually increased with the greater separation-of-powers, must be underlined. As long as only one or two institutions could make and break cabinets, the types of persons the Genro were free to nominate were few. But with the proliferation of semi-autonomous institutions, a greater variety of candidates could be considered, and the Genro could make more real as opposed to circum-scribed choices. I devote less detail than before to the events surrounding the rise and f a l l of-cabinets. Because only party leaders became prime ministers and the leader of the opposition party was always appointed after a cabinet f a l l , except when premiers died in office, I discuss only the f a l l of cabinets. Once the Genro sanctioned the f a l l of any government, 163 the terms of the unspoken agreement r e q u i r e d the appointment of the o p p o s i t i o n party l e a d e r . The f a l l of the Kato Komei,VHamaguchi, and Inukai cabinets f o l l o w i n g the deaths of the prime m i n i s t e r s i s not discussed i n t h i s s e c t i o n . The i n s t i t u t i o n a l f o r c e s that may have brought them t o r u i n had they not d i e d are discussed i n the s e c t i o n d e a l i n g w i t h i n s t i t u t i o n a l autonomy. I r e f e r the reader t o Appendices I and IV f o r a summary account of Cabinet-makers and -breakers and a chronology of prime m i n i s t e r s and t h e i r m i n i s t r i e s i n 1918-1932. The Genro, S a i o n j i , i s assigned a major r o l e i n s e l e c t i n g minor-i t y party l e a d e r s , a minor r o l e i n s e l e c t i n g m a j o r i t y party l e a d e r s . The p a r t i e s concerned i n both types of cases are assigned major r o l e s i n r a i s i n g the c a b i n e t s . I n the breaking of these party c a b i n e t s , the House i s assigned a minor r o l e , except i n the f a l l of Wakatsuki i n 1927 and the f a l l o f governments because of the death of t h e i r premiers. The Genro only sanctioned the f a l l o f cabinets r e s u l t i n g from other than n a t u r a l causes i n 1924-1932 because he intended t o nominate the opposi-t i o n p a r t y l e a d e r . When premiers d i e d and the same party provided successors, the House i s assigned no r o l e at a l l . I n the case o f Wakatsuki's f a l l i n 1927, "the House a c t u a l l y played a major r o l e . The r o l e s o f other i n s t i t u t i o n s which a f f e c t e d the appointment of a l l the party premiers i n 1924-32 and the p o s s i b l e f a l l of cabinets whose premiers d i e d are discussed i n the next s e c t i o n on i n s t i t u t i o n a l autonomy. The present s e c t i o n merely ^.demonstrates that a l l premiers other than those 164 who died in office were removed by institutions other than the Lower House. The section therefore begins with Wakatsuki's f a l l in 1927-The unseating of the f i r s t Wakatsuki cabinet is a partial exception to the generalisation that non-party institutions were the main causes of party cabinet f a l l . Although Wakatsuki's resignation i s usually attrib-uted solely to the Privy Council's veto of the cabinet 1s emergency r e l i e f -j Q measures for the Bank of Taiwan, the Prime Minister did not even consider doing battle with the Council. The working agreement in the House between the Kenseikai and Seiyuhonto on which his majority depended had threatened to break up three months previously. It only remained temporarily intact because Wakatsuki promised to resign at a suitable opportunity after the end of the session. Wakatsuki seemed to have regarded the Privy Council's obstruction as such an opportunity.^ Nevertheless, i f the GenrS was really trying to work against the separation-of-powers, he could have refused to sanction a resignation on the grounds of opposition from the Privy Council, and suggested waiting u n t i l the mood in the House dictated a cabinet change. The f a l l of the Tanaka cabinet was a clearer-cut case of Genro failure to side with the majority party in the House, come what may. The great merit of the British Cabinet system i s that when governments are unpopular, even among their own supporters, they are kept in power by institutional forces and given the opportunity to take strong though unpopular action, which they hope w i l l be vindicated before the next 165 quinquennial election. It i s precisely in';times of c r i s i s that extra incentives to support cabinets are so valuable, incentives which allow British cabinets to take, i f not a long-term view of policies, at least a five-yearly one. Prewar Japanese cabinets, even during the years of so-called parliamentary government, could operate only in the short-run. Because the Army's role in unseating Tanaka is discussed more f u l l y in the next section, only the support the Genr5 gavetthe Army needs mentioning here. The Army had strongly objected to the Prime Minister's attempt to lay on i t a l l the blame for the assassination of Chang Tso-lin, warlord of Manchuria. While Tanaka had previously reported to the Emperor that the Japanese Army was responsible for the incident, the Army forced him into the compromising position of reading to the Emperor .an o f f i c i a l statement which denied any Japanese complicity. The Emperor was furious because he was made to hear two contradictory versions of what had happen-ed and refused to receive Tanaka in audience except to receive his resig-nation. By f a i l i n g to side with the majority party prime minister, Saionji, as the major imperial adviser who could have smoothed matters over i f he PO had chosen, must assume responsibility for the Emperor's actions. The great virtue of parliamentarism is that a government need not fear for i t s existence, however unpopular or compromising i t s actions, as long as i t s own majority in the House remains loyal. The same verdict applies to the f a l l of the second Wakatsuki cabi-net. When the cabinet, backed by i t s majority, consistently opposed Army 166 demands for a stronger Manchurian policy, the Army began to work for a coalition "patriotic" cabinet under Adachi Kenzo, the MinseitS Home Minister who had been echoing the calls for a coalition patriotic cabi-net, and Inukai Tsuyoshi, the new President of the Seiyukai. A brief digression on party mergers w i l l explain party alignment i n 1930. Inukai's Kakushin Kurabu (Renovationist Club) merged with the Seiyukai in May 1925, and Inukai became Tanaka's successor as Seiyukai President. The Minseito (People's P o l i t i c a l Party) was the result of a merger between the SeiyuhontS and the Kenseikai following Tanaka's appointment in 1927-The institutional alliance which unseated Wakatsuki in 1931 was similar to the one which caused Takahashi's f a l l . But in 1931 the Army rather than the Peers encouraged internal Cabinet disunity. Wakatsuki also tried to purge his dissident member, Adachi, and was also abandoned by the Genro, who soon came to favour a new government under Inukai. Again, by partly siding with a non-party institution against a government backed by a majority in the House, the GenrS confirmed the legitimacy of the offending institution. He,^could have re-nominated Wakatsuki, who could then have formed a new cabinet without Adachi. While the Genro-House compromise did not develop into a permanent alliance, i t was not abandoned either. After 1924, even though the Genro never prevented non-party institutions from toppling majority-party-supported governments, he consistently handed the premiership over to 167 the leader of the minority opposition party, rather than to a member of the institution which played a major part in unseating the government. Saionji applied this principle when he nominated Tanaka, Wakatsuki for the second time, and Inukai. It was quite incorrect, however, to describe PP their nominations as "the normal course of constitutional government," because in modern Britain, as long as there is a majority party, a minority party is never asked to form a government. More normal perhaps was the nomination of men, as successors of prime ministers who died while in office, from the same party or i n s t i -tution as the dead premier: Takahashi, Yamamoto, and Wakatsuki on both occasions. But even here, there was an occasional element of the abnormal. Takahashi only became President of the Seiyukai after he was made Prime Minister by the Genro and did not assume office as leader of his party. The alternative to compromise, deadlock and demise, applied to the Cabinet's relations with institutions besides the Genro and the Lower House. But because i t was v i r t u a l l y impossible to please everyone at the same time, i t was inevitable that other institutions would sooner or later find themselves the cause of a Cabinet c r i s i s . In 1918-1932, the Upper House, the Army, the Navy, and the Privy Council a l l used their veto power vigorously and between them were pre-dominantly responsible for the f a l l of almost every government. The f a l l of two of the three exceptions also resulted from a major influence wielded by one of these institutions or another: only the Lower House 168 was a more important force than the Peers in unseating Yamamoto, and a greater obstacle than the Privy Council in the way of Wakatsuki's f i r s t government. After Yamagata's death i n 1922, there was no force to restrain these institutions other than their own understanding of how to use their autonomous power most effectively. The next section discusses the independent power of the major Cabinet-breakers of the period other than the Lower House and the Genro. The a b i l i t y of each to influence Cabinet composition is des-cribed alongside i t s a b i l i t y to unmake cabinets. No complete analysis of Cabinet composition is provided, for which the reader is referred to Appendix IV. Typical examples suffice to illustrate that the analysis of Cabinet composition in the previous chapter largely applies to the years 1918-1932. The main difference is that party members occupied more Cabinet seats than ever before. Apart from this, the number of representatives each institution had was more or less the same as before. But the sources are more explicit in linking the appointment of ministers with the independent power of these institutions. Institutional Autonomy, 1918-1932 Next to the Genro and the House, the Peers were the most effect-ive makers and breakers of cabinets in the 1920's. No ministry could assume office without coming to some agreement with them, the usual evidence of such agreement being the inclusion of at least one leading member of the Upper House's largest faction, the Kenkyukai. For example, 169 the Hara and Takahashi cabinets retained the support of the Peers through the inclusion of Oki Enkichi, a prominent Kenkyukai leader. 23 Even Kato Komei, when the three parties, on whose joint efforts in 1924 he had risen to power, were clamouring for a share of scarce cabinet seats, ok regarded i t as quite natural to give one to a Peer, Okada Ryohei. A few months after he assumed office in 1926, Wakatsuki gave a seat to a member of the Kenkyukai, because, as Colegrove observed, " i t is necessary, 25 of course, for the cabinet to come to terms with these parties." According to Tiedemann's study of the Hamaguchi cabinet, the Prime Minister and his successor after his death included Watanabe Chifuyu, "in an attempt to strengthen the cabinet's position i n the House of Peers by securing the support of the Kenkyukai, a group of which he was a leader." 2 6 Whenever there was not a party premier, the major group from which cabinet members were sought was the Kenkyukai. The Kato Tomosaburo cab-inet was described as a "Kenkyukai-centred' Kato-Upper House cabinet," 2^ and included four members of the Kenkyukai and three of the Koyu Club. The Yamamoto cabinet, which was based on the C i v i l Service, the Navy, as well as the Peers, included four Peers, but none of them were from the Kenkyukai, which refused to participate i n the cabinet because Saionji rejected i t s plan to extend the Kato ministry by only changing premiers. Predictably, the Peers were a major institution responsible for Yamamoto's f a l l . In his dissertation on the Upper House, Fahs concludes that "the 170 power of the Kenkyukai culminated in the establishment of the Kiyoura Cabinet, which . . . was based almost entirely on the support of the PR Kenkyukai and the K5yu Club. The Peers also did their f a i r share of cabinet-breaking. It was they who drove the wedge between the two factions i n the Takahashi cab-inet, forcing the Premier into a general resignation as the sole way to restore party unity. They were almost as responsible as the Lower House for the f a l l of the Yamamoto cabinet, which resigned because of the 20$ reduction by both Houses of the budget designed to restore the city of Tokyo after the great earthquake. The resignation of the minister res-ponsible for the plan to lend money to f i r e insurance companies, Den Kenjiro, who was "one of the p i l l a r s of the cabinet," was quickly followed by a general resignation. 2^ The reason Yamamoto himself gave, that his government was responsible for the attempted assassination of the Prince Regent, served the same purpose as the typical plea of i l l -health. When a similar attempted assassination occurred one month after the accession of the Inukai cabinet, a token general resignation was pre-sented, but there was no question of i t s acceptance.3® Although the Upper House cannot be credited with the f a l l of any other ministry i n these years, Kato Komei so greatly trimmed his programme of reform in anticipation of what i t would tolerate, that had he not died in office, the Peers would undoubtedly have been mainly responsible for his cabinet's f a l l , whenever i t occurred.3-L Redman concluded that the 171 Upper House's "indirect influence is unquestioned, and i t can therefore be said that the cabinet has a certain responsibility to (it)."32 The continued autonomy and strength of the two Armed Services was evident whenever foreign affairs came to the fore. Although the Army was not quite willing or able, during these years of public satisfaction with party premiers, to raise i t s own cabinet, a typical compromise was found. A leading Army General, Tanaka Giichi, became President of a p o l i t i c a l party, and was•supported for Prime Minister. This must have helped t i p the balance against Den KenjirS, whom, in spite of "the normal course of constitutional government," the Genro seriously con-sidered as Wakatsuki's successor i n 1927• That Tanaka was a party man as well as an Army General made him better qualified than Den in a system where the most important requirement of a Prime Minister is to do a balancing-act among a number of competitive institutions.33 The Army's power to unmake cabinets was more v i s i b l e . It was the major force behind the f a l l of the Tanaka and second Wakatsuki cab-inets, as well as a leading member of the anti-party-premier coalition, which, after the assassination of Inukai, f i n a l l y forced the Genro to disregard the precedent of nominating a man from the same party or institution as the dead prime minister. The Army's refusal to discipline the officers responsible for Chang Tso-lin's assassination forced the Prime Minister to contradict his previous imperial report. By not coming to Tanaka's rescue, the 172 Genro only sanctioned the Army's Cabinet-breaking power. The Army was also the major force that came between the members of the second Wakat-suki cabinet and caused i t s f a l l , although responsibility was placed on Cabinet disunity. Finally, although the Army could not be implicated o f f i c i a l l y in the May 15 Incident, in which Inukai was assassinated, i t s sympathies with the motives of the revolutionaries were well enough known for the GenrS to set aside precedent and nominate a non-party successor.-3 In addition to the more .dramatic examples of Army independence, the Army's uninterrupted occupation of the War Minister's seat in the Cabinet revealed i t s potential as a cause of Cabinet crises. Throughout the 1920's, the Army remained a major force with which a l l governments had to come to terms. The Navy's uninterrupted possession of the Navy Minister portfolio reflects that institution's possession of avveto power, which i t used vigorously. As long as great power p o l i t i c s , which i s always of great concern to navies, did not occupy the centre of the stage, the Navy was prepared to take a back seat. But with the conclusion of the popular Washington Agreement on naval construction in 1922, the Navy formed the backbone of the next government. Yamamoto succeeded KatS TomosaburS partly in his own right as a v i r t u a l GenrS, but mainly because the death of KatS was no reason to terminate Navy-led government. When great power politics loomed to the fore in the late 1920's 173 and keen inter-imperialist rivalry brought another agreement to limit the Japanese Navy i n 1930, that institution emerged from i t s apparent slumber: .a major section of the Navy revolted. The f a l l of the cab-inet was only prevented by the support the agreement received from both parties in the House, the Genro, the Army, and a section of the Navy i t s e l f . Only the Privy Council supported the opposition forces within the Navy. Had Hamaguchi not been assassinated, the Navy would have been the major institution responsible for the government's f a l l . The most dramatic display of Navy Cabinet-breaking power was that institution's implication in the May 15 Incident. Senior Navy officers were deeply involved, and Saionji was almost as apprehensive about Navy opposition to a party successor of Inukai as he was about Army opposition.35 In the years 1918-1932, the Navy, like the Army and the Upper House,decisively influenced the institutional balance of power. It almost equalled their Cabinet-breaking power, but surpassed their Cabinet-making power. No cabinet existed independent of the Navy. After 1922, the Privy Council, hitherto inconspicuous as a maker and breaker of cabinets, joined the throng of institutions able to veto government policy and topple cabinets. The usual explanation is that, while in earlier years the Cabinet and Council had been com-posed of similar kinds of men, clan bureaucrats, this was no longer the c a s e . B u t such an explanation does not shed light on the reasons 174 for the sudden and persistent opposition to cabinets that began in 1922, even ones led by Sat-cho men. It i s more"probable that the death in 1922 of Yamagata, who for eighteen years had been President of the Privy Council, f i n a l l y freed this institution from Genro control. No longer could cabinets assume that Privy Council support came automatically with GenrS support. The Privy Council was instrumental in raising the KatS TomosaburS and Yamamoto cabinets, and assumed major importance in raising the Kiyoura cabinet. Kato KSmei watered down his party's reform programme and caused himself considerable domestic d i f f i c u l t y , not merely out of deference to the Upper House but to the Privy Council. Both institutions were coming between him and his party, and i f he had fallen because of Cabinet disunity, 37 the Peers and the Privy Council would have been the main causes. 1 The Privy Council was the force that f i n a l l y compelled Wakatsuki to keep his promise and resign after the close of the 1927 Diet session. One should not assume that the coalition in the House would have broken up in any case. The SeiyuhontS was divided at the time, and after Tanaka's appoint-ment actually merged with Wakatsuki's Kenseikai.-^0 The next two governments also had largely the Privy Council to thank for the major problems they encountered. Tanaka's d i f f i c u l t y with the Privy Council over the Kellogg-Briand Pact almost equalled his d i f f i c u l t y with the Army over foreign policy i n Manchuria. The Council caused a great s t i r by objecting to the phrase, "in the names of their respective 175 peoples." Even the Minseito came to the defence of imperial sover-eignty. According to the Asahi daily, "there was no question that this was one of the major causes of the collapse of the Tanaka cabinet."39 This may be an over-statement, but the affair•contributed to Saionji's loss of confidence in Tanaka. The f i n a l display of Privy Council power was i t s five-month battle against the Hamaguchi cabinet over the London Naval Agreement.^0 Although the Council and the Navy f i n a l l y came round, the kind of atmos-phere that inspired Hamaguchi's assassin resulted partly from the feeling in the country that Hamaguchi may have gone too far. The only remarkable thing about the Privy Council's veto power was that there was l i t t l e evidence of i t in the composition of ministries. Kiyoura Keigo was the only Privy Councillor to serve in a cabinet through-out the period. It i s possible that the Council's legitimacy as a separate institution, entitled to a Cabinet seat as of right, was more d i f f i c u l t to justify than that of any other institution. The role of the C i v i l Service as a maker and breaker of cabinets is d i f f i c u l t to determine. Spaulding correctly points out that "the true measure of bureaucratic power is not how many ex-bureaucrats sat in the Cabinet, but how much voice bureaucratscco.llectively possessed in deciding policy and shaping l e g i s l a t i o n . " ^ 1 But a measure of influence on policy i s hard to find. While counting heads may not be an inaccurate indicator, i t i s the only practical one readily available. One must therefore 176 note not merely the large number of e x - c i v i l servants who sat in cabinets as members of one or other House of the Diet, but the increasing tendency for the positions of Foreign and Justice Minister to go to career members kp of these ministries. Even in the previous period, no party man, unless he was also Prime Minister, served as Foreign Minister. After 19l8, this proved to be even more the case: Tanaka was the sole party Prime Minister to serve as his own Foreign Minister. Inukai was his own Foreign Minister for only one month. A l l other occupants of the portfolio were career men of whom the Foreign Office approved. In 1918, Hara was able to reject Yamagata's choice of Foreign Minister.in favour of his own, Uchida Yasuya, only because Makino Shinken, a leading Foreign Office o f f i c i a l , supported Uchida.^3 T n 1923, the Foreign Office refused to accept Kiyoura's nominee, Fujimura YoshirS, and got i t s own, Matsui Keishiro, kk accepted. While many party men served as Justice Minister before 1918, only the Kato Komei, f i r s t Wakatsuki, and Inukai governments had party men i n this position i n 1918-1932. Hara was his own Justice Minister only u n t i l he could find a career man willing to accept the.job, the most influential of them, Hiranuma Kiichiro, having refused to serve under Hara.^"5 Because the seats of Foreign and Justice Minister gradually became the exclusive preserve of the C i v i l Service, cabinets were also under pressure to come to terms with bureaucratic interests. Even o f f i c i a l s 177 in other ministries could occasionally dislodge a man they disliked, as in 1931 when the Railway Ministry o f f i c i a l s forced Minister Egi Tasuku . kg to resign by passing a public resolution of no-confidence in him. The seats of Army, Navy, Foreign, and Justice Minister were typically held by non-party men throughout the period of so-called parliamentary government, and at least one seat went to the House of Peers. The number le f t for party control was never sufficient to make party government more than an aspiration. Cabinets in 1918-1932 were not merely made and unmade by a variety of institutional forces, but consisted of federations of institutional representatives. Appendices I and TV show the precise distribution of Cabinet seats in the period. Table 3 indicates the breakdown of Appendix I into the major and minor roles each institution, except the C i v i l Service which only played a part in raising the Yamamoto cabinet, played in the making of 11 cabinets, and the breaking of 9* The f a l l of the Kato Tomosaburo and Hara cabinets is not included. Table 3- Major and Minor Roles in Cabinet-Making and -Breaking, 1918-1932 Cabinet-Making Cabinet-Breaking House Genro Peers Army Navy Council House GenrS Peers Army Navy Council Major 9 6 2 1 2 0 3 5 3 3 2 4 Minor 2 5 0 1 0 2 2 2 0 0 0 0 As in 1900-1918, the Lower House and the Genro were the major makers and breakers of cabinets in 1918-1932. The House surpassed the GenrS 178 in raising ministries, but not i n pulling them down. The influences wielded by other institutions indicate, not merely as great a separation-of-powers as in the previous period, but a more even distribution of influence among the different branches. The existence of the Cabinet was also more separate from any single institution's control than ever before: according to Appendix 1, in 1900-1918 usually only two in s t i t u -tions at a time made and broke cabinets, but in 1918-1932, the average was almost three. Separation-of-powers theory alone emphasises this separation of the cabinet's existence from each of the other in s t i t u -tions . Typical manifestations of separation-of-powers forces were not hard to find during these years, although i t may be more d i f f i c u l t to attribute their occurrence to politico-institutional forces. In the following section no claim i s made that corruption and personality p o l i -t i c s resulted solely from institutional forces, although the timing of elections did seem to owe moretto these forces than to any others. Typical Consequences of the Separation-of-Powers The high degree of electoral and other forms of corruption in the 1920's was notorious, but because this phenomenon has already been commented on, i t is merely noted here that the separation-of-powers in the 1920's did nothing to mitigate corruption. Two other typical consequences of separation-of-powers forces w i l l be discussed more f u l l y : personality politics as i t affects the path to party leadership and 179 the timing of elections. The present section claims only that in 1918-1932 these phenomena, which s t i l l existed i n Japan during the 1970's, owed their occurrence at least partly to p o l i t i c o - i n s t i t u -tional forces. An interesting way in which the tendency towards personality politics manifests i t s e l f i s in the typical path to party leadership under the separation-of-powers. While in modern Britain, leaders rise within their party, owing everything to their party, in America, men who are potential presidents for other reasons, find i t necessary to capture a party. Often, like Mayor Lindsay of New York, they can even switch parties. As in post 1962 France and eighteenth-century Britain, the separation-of-powers also made i t necessary for men like Lecanuet or Palmerston to capture a party, the latter also being to switch parties. But no leader can switch parties in modern Britain with impunity, as Joseph Chamberlain found out i n I885, because the only qualification required by a prime minister is that he is a party leader. A highly popular Winston Churchill or Harold Wilson can lose to less popular leaders, simply because their parties have lost favour with the electorate. British government i s by parties, not personal-i t i e s . But Richard Nixon can win the Presidency in America, in spite of his party's unpopularity, because in that country, at least in Presidential elections, personality i s more important than party. An examination of the way Japanese prime ministers rose to the 180 position of party president reveals a close parallel with the American pattern. Although Hara was a man who, alone among Seiyukai presidents during this or any other period, owed his position more to his achieve-ments within the party than to anything else, even he had also been a member of the C i v i l Service, and according to Duus, was nominated as Prime Minister "in spite of (author's i t a l i c s ) the fact that he was a party leader," because of other qualities regarded by Yamagata as more important .^"7 Takahashi was even made Prime-Minister before he became party leader, so had l i t t l e d i f f i c u l t y in capturing the Seiyukai presi-dency. Japanese parties were on the keen lookout for leaders who had the potential to become premiers, and remind one of the frantic search by the American Democratic Party in 1972 to find a potential winner. More than any other Seiyukai President, Tanaka Giichi was a man with a number of attributes that qualified him for the premiership, who subsequently captured the party. His successor, Inukai, who had also achieved distinction outside the party, was an example of a man who had even switched parties. Hara Kei was the only Seiyukai Pres-ident who owed his position as party leader to his labours within the Seiyukai. MinseitS leaders were slightly less typical of this pattern, because a l l had risen to positions of prominence within their parties. Even so, a l l were also ex-bureaucrats, and only Hamaguchi was not a Peer. Their qualifications were not therefore limited to their party 181 membership. Theotiming of elections throughout the years 1918-1932 also con-formed to that of separation-of-powers systems, a fact which illustrates better than most that there was no fundamental change in institutional forces. When power is concentrated in a Cabinet whose existence depends solely on a majority in the House, as in modern Britain, electorates choose executives in the very act of choosing legislatures, and cabinet changes coincide with elections, rather than precede them. It i s not simply that executives and legislatures are chosen simultaneously at election time, but that the act of casting a single vote accomplishes both. In America, one does not choose a President in the very act of choosing a Representative, and many people vote for candidates from different parties in the two functionally different elections. Moreover, a new executive is not created every time a new Congress is chosen. In eighteenth-century Britain and prewar Japan, particularly after 1900, cabinet changes preceded elections, because the Cabinet's existence was not dependent on the composition of the legislature alone, While govern-ments needed loyal majorities to remain in office, these were not on their own sufficient. So even though cabinets usually rose for reasons other than the possession of a parliamentary majority, they soon after-wards held carefully controlled elections to provide them with such a majority. A l l the cabinets with party premiers during'this period that came 182 to power without a majority in the House were able to win one in the elections they held. The most remarkable reversals of party fortune took place in the elections of 1930 and 1932, managed by Hamaguchi, or rather his notoriously s k i l f u l election manager, Adachi KenzS, and Inukai respect-ively. When Hamaguchi came to power, his party had 2l6 seats to the Seiyukai's 217, but after the election i t was 273 to l^k. Inukai reversed the situation to give his party 301 to the Minseito's lk6. It is hardly l i k e l y that the electorate was "uninfluenced" in changing i t s mind so radically in such a short time. To understand why, a theoretical perspective is needed. Under separation-of-powers institutions, the party in power has an opportunity to influence elections in a way not possible in parliamentary systems. It can hardly be overemphasised that in eighteenth-century Britain and prewar Japan, no party-supported government ever lost an election, while in America, at least in the twentieth-century when national politics assumed much greater importance, few incumbent Pres-idents who sought re-election failed to secure i t . The power of the incumbent to enhance his personal appeal and to influence the electorate is greater in a separation-of-powers system than in a parliamentary system. There are two main reasons for".this. F i r s t l y , with the greater emphasis on personal a b i l i t y to get the most out of the inter-institutional struggle, incumbents have greater access to the media. Because everything they say or do is news, they have a greater opportunity to project the 183 type of personal image required. Secondly, because incumbents have control over the administration,, they can concretely demonstrate their a b i l i t y to manipulate the system by delivering actual favours, whereas the lack of concreteness about the opposition's promises usually results in these being regarded with greater scepticism. There is l i t t l e doubt that institutional forces in Japan in the 1920's were similar to those of previous periods, giving to the Japan-ese polity characteristics which made i t resemble in certain ways the American and eighteenth-century British po l i t i e s . To demonstrate why this was so, however, and why there was l i t t l e evolution away from" the separation-of-powers at a time when many people believed such a process to be taking place, i s less easy. A comparative analysis of the timing of a number of p o l i t i c a l forces in Britain and Japan, while suffering from the usual limitations of higher level' explanations which gloss over many important details, nevertheless comes closer to providing an understanding of what Scalapino sought .to explain in his' study of prewar Japanese parties, which he subtitled, The Failure of the .First  Attempt. Why the Separation-of-Powers Persisted The present section abstracts politico-institutional forces from the social, economic, and intellectual history of Britain and Japan, a procedure which is not meant to minimise the vastly different histories of these two island countries. For the purposes of the argument alone, l8k conditions other than politico-institutional ones, are assumed to he similar, even though such an assumption does not accord with the facts. The section also glosses over many crucial events and developments in the p o l i t i c a l histories of the two countries, because the length of the period covered requires that the comparison be limited to only the major relevant development s. Nine types of politico-institutional developments in each country are identified and the order in which they occurred i s compared. The completely different order indicates the different stages of ins t i t u -tional development Britain and Japan had achieved by the early 1930's. In England the order i n which the p o l i t i c a l forces occurred was as follows. F i r s t l y , already by 1297 Parliament had attained a legitimacy that compelled even the most autocratic Kings to consult i t or face armed rebellion. Secondly, i t was after the right of Parliament to be consulted had become unquestioned that England gradually developed into a unitary state with a strong.central government, and c i v i l war came to an end. This took place between the years 1066-1688. Those pleading the rights of Parliament could not be accused of unpatriotically fomenting c i v i l war. On the contrary, recognising the power of the parliamentary tradition, Kings like Henry VIII used Parliament as the major vehicle to promote national unity. Thirdly, the arr i v a l of separation-of-powers institutional forces took place well after the legitimacy of Parliament had become an established 185 fact. After 1688, when each of the separate institutions had i t s own legitimate function and was occupied by different people and inter-institutional deadlock became' the order of the day, much to the chagrin of their occupants, governmental deadlock tended to be attributed, not to an unwieldy Parliament, but to a Crown that failed to act in accord-ance with that Parliament. In the fourth place, the development of parties in Parliament was the earliest form p o l i t i c a l parties took, parties in the country dev-eloping much later. The Crown soon discovered that inter-institutional deadlock could be avoided simply byrrelying on these parties to get i t s business through Parliament. The parties in turn began to regard them-selves, not as channels of opposition to transcendental government, but as the vehicles of His Majesty's government i t s e l f . The result was gradually increasing party discipline and stable government, a develop-ment which began soon after the accession of G-eorge III in YJlk. It was much easier for the idea of parties as organs of government to gain currency before parties became frighteningly identified with the masses. Fifthly, the natural result was steady evolution away from the separation-of-powers into party cabinets, whose existence depended solely on majorities in the Commons. Evolution was smooth simply because Kings naturally preferred the greater efficiency of government unencumbered by constant inter-institutional deadlock. By the late eighteenth-century, they had given up trying to govern contrary to the w i l l of Parliament, 186 and by 1832, the Lords were forced to moderate their claims to equality with the Commons, even though u n t i l 1911 they remained a thorn i n the flesh of many a government. Had mass parties with programmes of social reform preceded this development, i t would not have taken place with such ease, as the ruling classes would not have allowed•institutions which-gave them a veto over proposals for radical change to become obsolete. Sixthly, when as a result of industrialisation under capitalism, movements for social reform sprang up, the power to introduce or resist such reform was concentrated in a cabinet responsible to a majority in the Commons alone. The Constitution, not particularly favouring the ruling-class or working class, could therefore be used with equal efficiency by either class. It was too late for a reactionary class to prevent social reform by retaining a foothold in one veto-possessing institution or another. Seventhly, therefore, when vi r t u a l manhood suffrage came soon after in 1868, movements for social reform were easily channelled into parlia-mentary activity. The result was that, eighthly, mass parties in the country became extensions of the already existing parliamentary parties. The road to party government would have been much more d i f f i c u l t had mass parties in the country existed before parliamentary parties. Finally, the development of a large C i v i l Service took place well after the supreme authority of the Cabinet, dependent, through the 187 Commons, on a mass electorate alone, had become an established trad-i t i o n . From the beginning, the Cabinet was able to exercise p o l i t i c a l control over the C i v i l Service, which never had the opportunity to become a veto-possessing institution. What gave i t the opportunity to do so in France and Japan was the: chronic weakness of Cabinet authority. In Japan the order of events was almost exactly the opposite, mainly because they a l l occurred in a mere fifty-year period spanning the last quarter of the nineteenth-century and the f i r s t quarter of the twentieth-century. F i r s t l y , a large and powerful bureaucracy, which existed already in Tokugawa Japan and acquired greater influence with every show of weakness by existing p o l i t i c a l authority, was the sole organ of govern-ment for more than thirty years before the establishment of the Diet. Wo wonder i t remained not merely one among many veto-possessing in s t i t u -tions, but one of the most powerful ones. Secondly, the maintenance of a unitary state under strong central- • ised government was the overriding concern just before, rather than well after, the establishment of the Diet. Those who championed the rights of this institution were quite naturally accused of fomenting c i v i l war, with the result that the legitimacy of the Diet was hard to establish. Thirdly, parties in the country, with some degree of grass-roots support, had developed not merely before parties in the Diet, but before the Diet i t s e l f , already in the early l88o's. Having arisen as channels 188 of opposition to transcendental government, they were unlikely to act or be treated like organs of government themselves for some time. . Fourthly, because separation-of-powers institutional forces came to exist at the same time as the Diet in I89O, the Diet had no special legitimacy vis-a-vis other institutions, and could not shift blame onto them for inter-institutional deadlock. On the contrary, the older i n s t i -tutions, the C i v i l Service, the Army, the Wavy, the Privy Council, and the Genro, were a l l more able to blame the newly created Diet for the i l l effects of inter-institutional deadlock. Fifthly, because the establishment of the Diet was so late in the sequence of events, i t took a long time for this institution to achieve a legitimacy even comparable to that of others. Sixthly, when parties with some degree of discipline did develop in the 1890's, because of their origins they tended to aim more at effective opposition to transcendental government than at taking over for themselves the task of government. It i s hardly surprising that they were not treated as capable of responsibility. Seventhly, when, as a result of industrialisation under capitalism, agitation for social reform became widespread, particularly after the First World War, power was s t i l l fragmented among a number of competitive institutions, some of which the ruling class, in order to have legitimate ways to veto social reform, prevented from obsolescence. So the Constitution did not evolve out of one which favoured the class preferring the status quo, 189 with the result that reformers turned to anti-constitutional means. Eighthly, when vi r t u a l manhood suffrage f i n a l l y came in 1925, i t was unable to give the House the popular boost to make i t the sole institution able to make and unmake cabinets. Manhood suffrage, not holding out the possibility of controlling an a l l powerful Cabinet, roused l i t t l e interest, and the House made no progress in i t s drive for sovereignty over other institutions. Finally, the abolition of the separation-of-powers did not take place naturally and came last of a l l , because natural evolution required a sequence of events that was not possible in Japan, given the need to telescope into f i f t y odd years what had taken centuries i n Britain. But while the problem of the separation-of-powers was solved by natural evolution in Britain, many other p o l i t i c a l developments in that country took place only after a great deal of bloodshed, which Japan was spared because of her telescopic development. Robert Scalapino's explanation of the "failure of the f i r s t attempt" in terms of the timing of industrialisation i s quite unconvincing, largely because of his imperfect understanding of what is concretely involved in running a successful democracy. By equating democracy with liberalism, Scalapino makes the running of a successful democracy synonomous with subscription to a particular p o l i t i c a l ideology. The equation makes his explanation purely a cultural one and sheds no light on the problem of the kinds of institutions required for democratic government. It makes 190 democracy just a matter of free-floating attitudes or "values", rather than of concrete procedures to he followed in a specific institutional framework. To look at the problem of those who genuinely wanted a more democratic Japan, and s t i l l do, in terms of the i n a b i l i t y of Japanese capitalism to spread more widely the creed of possessive individualism, which even in Britain was widely subscribed to only in the nineteenth-century, is completely to misunderstand their d i f f i c u l t y . The problem was largely politico-institutional, and requires a po l i t i c o - i n s t i t u -tional analysis. The 1930's are one of the most misunderstood periods in modern Japanese History. Very few western scholars have attempted to write a p o l i t i c a l history of these complex years, but those who have tr i e d emphasise foreign affairs, the absence of liberalism, and the power of the Army to the exclusion of almost a l l else. Developments during these years, however, show not that the Army was in complete control of the government, but that i t was only one of the more influential institutions in an inter-institutional struggle that had been accepted as the de facto "normal course of constitutional government." By the 1930's, the members of the veto-possessing institutions had learnt to live with the separation-of-powers.' 191 CHAPTER 5 LEARNING TO LIVE WITH THE SEPARATION-OF-POWERS: NATIONAL-UNITY CABINETS  1932-19^0 By the 1930's, the leaders of Meiji p o l i t i c a l institutions had come to recognise that i t was f u t i l e to strive for anything more than sover-eignty in the exercise of those functions which they regarded as peculi-arly their own and for control of the entire Cabinet. Having learnt to live with the separation-of-powers, the leaders of each institution merely tr i e d to ensure that they were not deprived of their due representation in the Cabinet and that no other institution encroached on matters f a l l i n g within their own prerogatives. The kyokoku i t c h i (national unity) cabi-nets of the 1930's were the result, not merely of a domestic nationalist reaction to foreign imperialism, but of politico-institutional forces as old as the Constitution i t s e l f . Only cabinets that represented each of the veto-possessing institutions could hope to retain their support, and the coalition Cabinet became the f i n a l attempt to make the separation-of-powers work. While in the days of "transcendental" and "party" cabinets the major deus ex machina had been the dominant institution, the Genr5 and the parties, the absence of a single dominant institution i n the 1930's brought new ways to promote inter-institutional harmony: committees of representatives from institutions in need of reconciliation, cross-institutional factional alliances, and a multi-institutional mass movement. 192 • Background'1' The misunderstanding of many aspects of Japanese history in the 1930's i s not least due to the complexity of this history. Economic and social changes as well as new relations with foreign powers and Japanese colonies were inter-related in a number df ways that cause headaches to bourgeois and Marxist scholars alike. Orthodox inter-pretations of the 1930's have to gloss over certain contradictions to resolve others, and unorthodox interpretations frequently rest on weak theoretical grounds and raise their own contradictions. No attempt w i l l be made to unravel the complexity of the period or to weigh the advantages and disadvantages of the various attempts to provide a general theory of what Japanese scholars regard as the "dark valley.'" Only a brief sketch of the -.different economic, social, and p o l i t i c a l developments of the period w i l l indicate that the politico-institutional analysis which follows has been largely abstracted from a very complex set of events. Economic changes in the 1930's were greater than in any other period. The relative importances of light and heavy industry were reversed. In 1929-1937 "the decline in the textile industry's share of t o t a l manufact-uring output and the rise in the share of machines and chemicals was greater than ever. In 1929, 50$ of factory workers were in textiles and 26$ in metals, machines, and vehicles, whereas in 1937, the figures were 35$ and 42$. By 1937 Japan could produce most of the machines required by her industries. The relative decline i n the growth of employment in 193 the textile industry resulted largely from technical improvements intro-duced under the impact of the depressed 1920's. The rationalisation of industry in the 1920's had greatly increased i t s efficiency and reduced the number of workers reguired for a given output in textiles more than needs in heavy industry could absorb. But the greater emphasis on heavy industry in the 1930 's was not solely a response to the deflation of the previous decade. The troubles in Man-churia that increased steadily after 1928 led the Armed Services to demand greater military budgets, and Inukai's Finance Minister, Takahashi, reversed the deflationary policies of his predecessor. Takahashi took Japan off the gold standard and, u n t i l his assassination in 193&, remained almost continually at the Finance Ministry. He used a cheap monetary policy to launch a massive programme of military expenditure:, in 1931 the share of Gross Rational Expenditure devoted to the Army and Ravy was 31$> but by 1936 the figure was h-lfo. Takahashi's budgeting was more than a response to demands from the Armed Services. It was an anticipation of Keynesian solutions to unemployment and depression, and Japan emerged from the slump soon after 1931* The years of Takahashi reflation were ones of rapid economic growth and higher employment. Because the economic growth of the 1930's was induced by an expan-sion of heavy industries, structural unemployment was bound to appear sooner or later. Already in 1936 bottlenecks in the form of a scarcity of skilled workers began to occur in the heavy industries, but a mass of 19 k unskilled workers were clamouring for scarce jobs. A huge migration from the countryside to urban areas and from agricultural trades to industrial trades took place: in 1930 agriculture and forestry absorbed 50$ of the occupied population, but in I9U0 the figure was only Kk-io of a larger t o t a l . The large pool of unskilled workers depressed the wage rate, which even in the depressed 1920's had not fallen a l l that much. An agricultural depression that began in the late 1920's and from which the country emerged only in 1935 led to a f a l l in the cost of l i v i n g , which alone prevented the wage rate in the 1930's from f a l l i n g more than i t did. The decline i n the price of raw s i l k and rice, which accelerated the rate of migration to the c i t i e s , prevented the decline of real wages in 1931-1936 from keeping pace with the decline in money wages. The hardest hit by f a l l i n g real wages were tenant farmers and workers in light industry. After Takahashi's assassination in 1936", when he decided to end his expansionary policy, the outbreak of h o s t i l i t i e s in China in 1937 pre-vented any further interference with the growth of heavy industry. Through-out the 1930's, declining levels of consumption and rising levels of military expenditure characterised^the Japanese economy. The Japanese working class and tenant farmers fought an uphill battle to prevent further - erosion of their standard of li v i n g , although they were more united in this endeavour than they had been in the 1920's. In 1925 the Ninon Rodo SodSmei (japan General Federation of Labour), 195 founded in 1921 and put under severe strains by government clampdowns and the destruction of many unions following the failure of the syndi-calist strikes in 1922, split into Communist and Social Democratic organisations. By the end of the 1920's, mass arrests in 1928 and 1929 had driven the Communists underground, and working class organisations of the 1930's remained social democratic, even though the legal leftwing organisations included many Communists. In 1932 the Shakai Taishuto (Social Masses Party) was formed. It was not the f i r s t socialist party in Japan, but i t was the f i r s t one to bring the legal l e f t and social democratic working class organisations into a single party to contest general elections. Its main support came from the organisations into which the old Ninon Rodo Sodomei split in 1925 and which tended to merge together again in the late 1930's. The Shakai Taishuto's electoral appeal was narrow. White-collar workers, small farmers, and the petty bourgeoisie remained loyal to the bourgeois parties, and only the workers and tenant farmers who were organ-ised gave their support to the new party. But organised workers and tenant farmers only comprised some 4-5$ of the t o t a l electorate in the 1930's, and the classes they represented formed only 30$ of the electorate. The Socialists, therefore aimed less at winning elections than at using elect-ions for propaganda purposes, and only in 1937 <lid they receive 10$ of the t o t a l vote, which came almost entirely from the large c i t i e s . The Shakai Taishuto was dominated by urban interests, because even 196 though the agricultural population was s t i l l enormous, farmers unions were d i f f i c u l t to organise. Only .33 million tenants were organised at the high point of farmer unionisation i n 1927' The impact of the working class on the body p o l i t i c in the 1930's was minimal. The right to strike and organise was not widely recognised, and the proletarian party's performance in elections was not much more than an ir r i t a n t to the bourgeois parties. The peasantry had a greater influence on major developments in the 1930 's, but because i t was poorly organised, i t s main institutional manifestation was through the Army, which recruited heavily from the peasantry. The Army was one of the major influences in shaping events in the 1930's. The late 1920's witnessed increased interference in Cabinet affairs by the Army and the Navy, because foreign affairs once again occupied the centre of the stage. In September 1931 the Army presented the Cabinet with a f a i t accompli occupation of Mukden, and later forced Japanese annexation of Manchuria. Public support for the Army's actions came not merely with the League of Nations' condemnation of the fatherland, but with the greater markets and population outlet the new colony provided, particularly for landless peasants. When trouble broke out in Shanghai in early 1932 and the Army came to the rescue of the Navy, international criticism of Japan only added to the Army's already growing prestige. The national mood of this island people was one of intense patriotism in the face of the growing inter-197 Imperialist rivalry in the Far East. The outbreak of the China war in 1937 prevented any likelihood that foreign affairs would give way to economic problems as the major national concern. Although the public on the whole was prepared to suffer and fre-quently support the Army's growing control over foreign policy, in the early 1930's, at least, public opinion was shocked by the Army's attempts to exert a greater influence on domestic affairs. From the assassination of Inukai in 1932 u n t i l the attempted assassination of Okada in 1936" the Army refused to live with the separation-of-powers. But after 1936, when the Army once again worked within the Constitution, public opinion was behind i t . The Army leaders themselves were not united on the proper degree of control they should exercise on foreign and domestic policy. The Kodoha (Imperial Way Faction), was the more influential group in the Army u n t i l at least 193^. It was also the more radical, because i t included the young officers whose ties with the hard-pressed peasantry were close: in 1920-1927, 30$ of those entering the cadet corps had agrarian and petty-bourseois origins. The older officers had a more aristocratic background, and had for many years worked for Army interests through compromise with other institutions. They advocated a less radical policy i n Manchuria and less radical methods to further that policy. The so-called "Showa Restorationists" in the KSdoha consisted mainly of young officers, who aimed to "restore" to the Emperor the sole right to control national 198 affairs, because they resented the practice of compromising with other institutions, particularly the Lower House. Their abortive coup d'etat on February 26 1936 spelt the end of Kodoha." supremacy in the Army, which thereafter lived with the separation-of-powers. The February 26 Incident marked the end of the peasantry's impact, albeit indirect, on the body p o l i t i c . The main opposition to the Kodoha.-, within the Army, the Toseiha (Control Faction), capitalised on public sympathy for their attempt to restore Army discipline and increased the overall influence of the Army vis-a-vis other institutions, particularly in foreign a f f a i r s . Members of other institutions regarded the Toseiha leaders as the only people who could restore discipline i n the Army and who were prepared to operate within the Constitution. After the outbreak of the China war in 1937, the Army brooked l i t t l e interference from other branches of government, except the Wavy and the Foreign Ministry, in the exercise of the function which was constitutionally i t s own. Predominant Army influence in war time was quite consistent with constitutional practice: only the Wavy and the Foreign Office could legitimately challenge Army supremacy in the decade when the major nat-ional concern was inter-imperialist rivalry. Problems arose in the Army's relations with institutions other than the Wavy and Foreign Office only because the desire to expand the Army's share of national expenditure required the approval of the institution constitutionally empowered to 199 reject increased expenditures. The two Lower House bourgeois parties remained a l l i e d to the zaibatsu (financial cliques) in the 1930's. Because the zaibatsu found that shifting resources into heavy industries preserved their profits, they offered l i t t l e resistance to the new economic policies. Until 1936" the expansion of heavy industry was financed by cheap money, and after 1936 only a part of the military appropriations came from increased taxes. Although there was never complete agreement between the Army and the zaibatsu parties on economic policy, the fundamental conflicts between the Lower House and the Army were not over the budget, but foreign policy. The maintenance of profits depended largely on good relations with Britain and America, and the parties advocated caution in dealing with Japan's major trading partners. The Army, however, tended to throw caution to the wind and wanted to use Japan's military potential in the way the actual users of that potential deemed f i t . The Lower House was compliant over the budget and even agreed to Army proposals for greater economic controls and planning. But i t constantly c r i t i c i s e d the Army for moving too close to Germany and for antagonising Britain and America. The Army's anti-capitalism of the early 1930's gave way to an anti-Anglo-Saxonism in the late 1930's, and the Lower House's anti-militarism of the early years gave way to an anti-Germanism in later years. Although the distance separating the two institutions most closely concerned with the two major national problems of the decade, foreign and 200 economic policy, was not unbridgeable, i t was sufficient to cause continual deadlock. Neither the Army nor the Lower House tolerated a prime minister representing the other, and the only governments that had any measure of success in the 1930's were under premiers from more neutral institutions. Examination of Cabinet composition and i t s making and breaking reveals that the Array and the Lower House were the two dominant i n s t i -tutions in the 1930's. Because most prime ministers came from branches other than these two specifically in order to keep the peace between them, both were major influences on the rise of a l l cabinets under third-force premiers. The Lower House was a major force causing the f a l l of a l l cabinets of the period, and often the Army conspired with i t . Because the role of third-force institutions was always c r i t i c a l in Cabinet-making, the raising of cabinets was as separate as ever from the control of any single institution. Cabinet-breaking power, however, was less fragmented than i n the 1920's, because usually only two institutions were needed to topple a government, and sometimes only one sufficed. The following section discusses only Cabinet composition and Cabinet-making, a procedure which indicates that more than ever before the Cabinet was a federation of institutional representatives under the Prime Minister the Genro regarded as the most able to keep together the often antagonistic forces they represented. After the f a l l of the Inukai government, the GenrS no longer believed that party premiers 201 were the best qualified for the task. The struggles over domestic reform in the 1920's when the parties were the best placed co-ord-inators of major national policies gave way to the foreign and economic policies of the 1930's, when the Array ref.used to serve in cabinets under party leaders. By their breaking of the only two Army-led governments in the 1930' s, the parties demonstrated their unwillingness to tolerate Army premiers. The Genro always chose a premier from among those men most able to retain the allegiance of a l l the veto-possessing in s t i t u -tions. By the 1930's, the need to compromise and live with the separation-of-powers was more generally accepted than ever before. The Genro and the parties gradually came to live with a smaller share of influence than they had wielded u n t i l then. The Genro-House compromise of the last three decades gave way to a more general compromise, which now brought institutions other than the Genro and the House more prominently into Cabinet-making. The General Compromise in Cabinet-Making, 1932-19^0 In a classic on British government, Sir Ivor Jennings wrote: Cabinet government i s government by committee, but i t differs from ordinary committee government because, through the party system, an attempt i s made to achieve uniformity of opinion. Usually a committee . . . contains members whose opinions differ widely. . . . The Cabinet i s not such;-;a committee. Normally, i t i s chosen from among the members of one party, who accept party policy as a matter of course. In one of the few studies on the Japanese Cabinet, Yamazaki Tansho 3 wrote : 202 After the May 15 Incident, when party cabinets came to an end, and the so-called national unity (kyokoku itc h i ) or mixed cabi-nets came into being, the Cabinet was formed by gathering rep-resentatives from a l l fields, with the result that i t became impossible to form a cabinet of people with the same p o l i t i c a l opinion. In order to maintain the support of the sole institution able to make and break i t , the British Cabinet is composed entirely of members of that institution, and agreement on policy among ministers is the rule. In prewar Japan, however, in order to maintain the support of a l l institutions able to make and break i t , the Cabinet was composed of rep-resentatives from a l l these institutions, regardless of their conflict with one another, and agreement on policy was rare. In the 1930's, this was as true, i f not more so, and the representation of institutions as balanced, as in any other period, because by then the rules of the separation-of-powers system had gained general recognition and accept-ance . Appendices I and '_"V show that throughout the decade, each of the institutions was consistently well represented, excepting the Privy Council, and the Genro, both of which gradually ceased to be independ-ent forces in Cabinet-making and -breaking and became mere barometers to assess the relative strength of each institution i n order to have the Emperor appoint the cabinet most likely to succeed. The Genro became less inclined to sanction any kind of lop-sided cabinet whatsoever. When a cabinet failed to include a proper balance of representatives, the offend-ed institution, usually the Lower House of the Diet, saw to i t that the 203 cabinet was short-lived. The Hayashi cabinet, which Ogata Taketora described as "one of the shortest-lived and most incapable Cabinets ever known in Japan,was unseated by. the House, because Hayashi fa i l e d to include party representatives and to come to terms with the House. Behind this need for a balanced apportionment of seats lay the Cabinet's attempt to win the cooperation of i t s potential breakers and to retain that of i t s makers. The Cabinet was not merely a c o l l -ection of people from different walks of l i f e who happened to be chosen for high office, i t was an attempt to coordinate conflicting politico-institutional forces by bringing their representatives together. According to Spaulding, c i v i l servants "thought of the executive branch as a confederation of autonomous ministries," and the Cabinet "as a summit c o n f e r e n c e . I n 1937 a contemporary observer wrote :^  Japan w i l l continue, for some time yet to come, to have Cabinets organised on the basis of compromise among the bureaucrats, the military and the p o l i t i c a l parties. It may even be said that this type of Cabinet has already become a well-established institution in p olitics a l a japonaise. Saionji himself referred to the Cabinet as a "patchwork operation" and a "sort of federation."^ Because the task of forming a successful and balanced "federation" f e l l mainly on the Prime Minister, the person chosen for this office had to be someone who would not provoke any institution. He had to be an uncontroversial, flexible, and frequently mediocre person. The Prime Ministers of the 1930's were of this type. 2C4 The present section does not go into the details of the events surrounding the rise of cabinets i n the 1 9 3 0 ' s . Only the main in s t i t u -tional forces are identified, and Cabinet composition is mentioned only when i t better illustrates the institutional line-up than does the mere selection of a premier. I refer the reader to Appendices I and V for the summary accounts of Cabinet composition and i t s makers in 1 9 3 2 - 1 9 4 0 . Because the role of the GenrS is discussed in a later section on the functions of institutions, the reasons for assigning Saionji an influence up to and including the appointment of Hayashi are not discussed in this section. The nature and Influence of the Genro's successor, the Jushin (Senior Retainers), are also discussed in the section on the functions of institutions. The present section, beginning with SaitS Makoto's appointment in 1 9 3 2 , identifies a l l the other makers of cabinets in 1 9 3 2 - 1 9 4 0 . In almost every case, the Army and the Lower House wielded a major influence, and other institutions acted as third-forces to keep the peace between the Army and the House. After Tnukai 1s assassination in 1 9 3 2 , the Armed Services, particularly the Army, were strongly opposed to a party successor, and even public opinion was hesitant about the appointment of another party premier. In the Lower House, only the mainstream faction of the Seiyukai under Suzuki KisaburS advocated a single-party cabinet. The Minseit5 and the anti-Suzuki faction of the Seiyukai preferred a coalition government. Saito was chosen as the man most l i k e l y to restore some kind of truce 205 between the parties and the Army. In 1936 Iwabuchi Tatsuo wrote : w [Saionji] saw in Viscount Saito the ideal type of man to organise a ministry that would strive primarily to appease and tranquilise the nation. He thought that this retired admiral of modesty. . . . was especially suited for the task of calming the people and easing the strained relations among the military, the parties and other factions. Before he was appointed, Saito had to pass a screening test. Saionji •consulted representatives of a l l institutions: acting Prime Minister Takahashi of the Seiyukai and President of the MinseitS, Wakatsuki; President of the Privy Council, Kuratomi Yusaburo; President of the House of Peers, Konoe Fumimaro; and Fleet Admiral Togo Heihachiro, Field Marshal Uehara Yusaku, and the two Service Ministers. In the end Saionji decided on Saito, because, as a representative of the Wavy, Saito could appease extremists in the Army without over-provoking the parties i n the Lower House. In his dissertation on Palace Of f i c i a l s in the 1930's, Titus i s not quite accurate in arguing that Saito was merely a "generalist" or "negotiator" and not "an institutional spokes-man. Saionji regarded him as both of these, and herein lay his suit-a b i l i t y as Prime Minister. Saito was meant to bring in the Wavy as a third force between the Army and the Lower House. After the Army's unabashed show of sympathy for the assassins of the Lower House premier, Inukai, an act condemned by the Wavy even though senior officers were implicated, Saito did well to win the i n i t i a l support of both parties in the House. The Saito cabinet was the creation of the Wavy, the House, and the Army.10 Saito's successor, another retired Admiral, was chosen for reasons 206 similar to the ones that led to his own appointment. According to Iwabuchi, "he [[Saionji} wanted another Premier as much like Viscount Saito as possible, and he could not have chosen better than he did to assure perpetuation of the harmless, moderate policy of the outgoing cabinet.""'-"'- Miller points out that both cabinets "consisted of bureau-cratic, military, and party men, reflecting the effort to produce harmony with the house of representatives, the house of peers, the privy council, and the military."-^ There were only two differences in the forces raising the SaitS and Okada cabinets: in 193^ i t was more important that a Navy man serve than i t was i n 1932, and in 193^ only the MinseitS supported the government. The Genro f e l t that the Okada government's major task was to participate successfully in the London Naval Conference in 1935, whereas the mission of the Saito ministry had been to bring domestic tranquility. J In March 1936, after the February 26 Incident, Iwabuchi observed that "again he ^Saionji] wanted peace and moderation, and again he found a Premier who was relatively popular in a l l circles and without bit t e r enemies. The Army had become ti r e d of Navy-led ministries and increas-ingly determined to brook no interference in i t s control of troops in Manchuria. It also wanted to force onto the Lower House some of i t s ideas on domestic economic reform. Ideal third-forces were obviously the Foreign Ministry and the House of Peers, and Saionji's f i r s t recommend-ation was the President of the House of Peers, Prince Konoe, who according to Harada Kumao, Saionji's private secretary, "was on friendly terms with 2 0 7 the military, the p o l i t i c a l parties, and the House of Peers. "1^> When Konoe declined the offer, Saionji nominated an influential member of the Foreign Office, Hirota Koki, who had been Okada's Foreign Minister. Hirota's a b i l i t y to act as a mediator between the Army and the House i s revealed by his willingness to make concessions to both sides. He gave way to Army objections to the entry into the cabinet of Wakajima Chikuhei, Shimoura Hiroshi, and Yoshida Shigeru, to Ohara Naoshi's remaining on as Justice Minister, and to Kawasaki Takukichi of the Minseito becoming Home Minister. But he would not accede to the demand that each of the major parties have only one representative instead of two assheshad promised. The next prime minister did not measure up to the requirements of the office, although the person ultimately appointed was only the Genro 's third choice. Saionji's f i r s t choice, because Hirota's f a l l resulted from a head-on col l i s i o n between the Lower House and the Army, was, as Baba Tsunego wrote, "a military man with an understanding of such [parliamentary p o l i t i c s . Unfortunately, although the parties were strongly i n favour of Ugaki Kazushige, the Army was completely hostile. When Saionji's second choice, President of the Privy Council, Hiranuma Kiichiro, refused to serve, Saionji f e l t the sole alternative was the man for whom Ishihara Kanji of the Army General Staff had been conducting a vigorous campaign, partly i n conjunction with certain party men who were also advocating the formation of a new-pro-military p o l i t i c a l party under 208 Prince Konoe. Because their number included Maeda Yonezo and Nakajima Chikuhei of the Seiyukai and Nagai Ryutaro of the Minseito, Saionji must have believed that Hayashi could assume the necessary middle-of--i Q the-road position between the Army and the Lower House. The actual balance of forces behind the Hayashi cabinet, however, is revealed in Ishihara's dominant influence on Hayashi's l i s t of prospective ministers and Hayashi's ina b i l i t y to have i t accepted by other institutions. Ishihara wanted Itagaki Seishiro as Army Minister, Suetsugu Nobumasa as Navy Minister, Ikeda Seihin as Finance Minister, and Tsuda Shingo as Commerce and Industry Minister. They a l l appeared on Hayashi's l i s t . But the Army vetoed Itagaki and chose Nakamura Kotaro, Suetsugu was unpopular in the Navy and was rejected in favour of Yonai Mitsumasa, and Ikeda and Tsuda refused to serve because they were reluctant to implement Ishihara's policy of economic reform. Finally, when Hayashi offered seats to Nakajima (Seiyukai), Nagai (Minseito), and Yamazaki Tatsunosuke (Showakai), only the latter was prepared to accept the condition of f i r s t resigning from his small party. The only positive institutional support behind the Hayashi cabinet was the Toseiha (Control Faction) of the Army, and possibly the Kenkyukai in the Upper House. Although the Foreign, Justice, and Home Ministries were headed by career men, who possibly brought Hayashi some support from the C i v i l Service, no fewer than four of the remaining seats were occupied by men who headed more than one ministry, indicating the Prime 209 Minister's i n a b i l i t y to f i l l certain positions with the usual balance of representatives. Wot surprisingly, the Hayashi cabinet lasted only four months. It was a chSzen naikaku (transcendental cabinet) rather than a kyokoku i t c h i cabinet, a type that had become quite unacceptable by the 1930's, and was soon unseated.^ The f a l l of two ministries had resulted from clashes between the Lower House and the Army, and not merely was a new third- force needed, but one which could also command the positive allegiance of the two belligerents. Otherwise, the paralysis of government would continue. Royama Masamichi wrote : "The structure of the f i r s t Konoe cabinet reflected this domestic situation clearly in the personnel of i t s Ministers gathered mainly from the administrative bureaucrats and the members of the House of Peers, whose institutional positions were deemed to be comparatively neutral i n the conflicts between the Diet and the M i l i t a r y . " 2 0 Of the nine o f f i c i a l s in the cabinet, four were purely C i v i l Service spokesmen, while three of the eight Peers represented the Upper House alone: two third-forces came between the House and the Army. But Konoe also won the positive support of both sides. By allowing rumours to circulate that he intended to form a new party, he e l i c i t e d the support of those Dietmen, who, like many in 1900 and 1913; believed that a prestigeous President was the only way to regain party control of the Cabinet. After the Seiyukai defeat i n the 1936" election, cooperation with Konoe became the Maeda-Wakajima group's major tactic to revive Seiyukai fortunes. In 210 his dissertation on Prince Konoe, Berger writes: "If party men wished to demonstrate that they could gain the confidence of the other elites and win the support of the ordinary citizenry, i t would clearly be to their advantage to have the respected and popular Prince Konoe leading them." The two party representatives in the cabinet, Nakajima and Nagai, were the leaders in their respective parties of the new party movement, even though Konoe minimised the importance of their appoint-ment by saying that they were serving as individuals rather than as representatives. But the leaders of their parties agreed that they pp entered the cabinet as representatives. Prince Konoe also had many friends in the Army. He had a long association with the Kod5ha (imperial Way Faction), and refused the premiership in 1936 because he was unwilling to preside over the liquidation of his friends after the February 26 Incident. Because he was supported by the Army and the Lower House and had few p o l i t i c a l enemies even in other institutions, Prince Konoe was superbly qualified to head the only practicable type of cabinet in a separation-of-powers system, a kyokoku i t c h i cabinet. 2^ The most striking thing about the advent of the Hiranuma ministry was the attempt to make i t r seem that no real change had taken place. When President of the Privy Council, Hiranuma, became Premier, Yoshioka Bunroku wrote : "In order to make known that the Cabinet change involved no question of policy, Prince Konoe was installed as President of the 211 Privy Council and concurrently Minister without portfolio in the new government, and the majority of the members of the preceding ministry joined the new cabinet."2^" In 193^ there had been a similar attempt to perpetuate a reasonably successful ministry. But Hiranuma differed from Konoe more than Okada did from Saito, and although his cabinet, which included four pure bureaucratic spokesmen, came as a neutral third force between the Army and the House, he found i t more d i f f i c u l t to win the allegiance of both sides, particularly because he was associated with the Kodoha of the Army and the Kokuhonsha (The National Foundation Society), a rightwing organisation of mainly bureaucrats and military men established in 1924.^5 Hiranuma therefore took special pains to publicise that he recog-nised and accepted the system as i t was. He had learnt to live with the separation-of-powers, and, according to Baba Tsunego, said as much ' P6 to newsmen: He would respect the Diet, for Japan has constitutional government, and therefore could not disregard the p o l i t i c a l parties, which he said came into being spontaneously wherever there is a legislature. His remarks were interpreted as recognition of the parties and assurance that he had no intention of crushing them. To demonstrate his good faith, he departed from the precedent set by Hayashi and Konoe, who insisted that party men join as individuals: he approached the leaders of the parties to request o f f i c i a l party represent-atives. The smooth transfer of power from Konoe to Hiranuma also demon-strated that the new government was meant to be an extension of the old. 212 In December 1938, consultations among only Konoe, Hiranuma, and the Privy Seal made i t almost a foregone conclusion that when Konoe chose to resign, Hiranuma would succeed him.2^ The next cabinet, under General Abe, lasted only two weeks longer than the abortive Hayashi cabinet, for similar reasons. Abe had been associated with the Ugaki school of thought in the Army and was placed on the reserve l i s t in 1936, because, as he said, the senior generals should be held responsible for the lack of discipline that lead to the February 26 Incident. Although the pro-German school of thought soon gained prominence, they lost face with the/conclusion of the Russo-German Non-;Aggression Pact in August 1939* Once again Iwabuchi under-stands the middle-of-the-road position required of premiers: "Because he [Abe] had not been associated with any blamable developments and thus had incurred no blame himself, i t i s presumable that he was chosen for the premiership as the man least l i k e l y to arouse f r i c t i o n in any ?ft quarter. Events just prior to Abe's appointment, however, suggest that his ministry could not have been received a l l that neutrally. The signing of the Molotov-Ribbentrop Pact had vindicated the Navy's and Foreign Ministry's opposition to the pro-German policy of the Hiranuma govern-ment, and the President of the Privy Council and Lord Privy Seal regarded Hirota of the Foreign Ministry as the most suitable successor. I n i t i a l l y the Army made no objection. But suddenly on August 2k, a movement began 213 among the Army General Staff against Hirota, and the next day they came up with Abe. Although Saionji favoured Ikeda Seihin, the President of the Privy Council decided to support Abe, and Saionji gave in. 29 Abe was the compromise and last-minute choice of the Army as well the man who robbed the Foreign Office of i t s chance to put right what i t regarded as an erroneous foreign policy. In forming his cabinet, he sought solely to appease the pro-German forces, apparently oblivious of the Army-House conflict which had undone a l l his predecessors in the 1930's. He acceded to the Army's demand not to select the Foreign Minister from the Foreign Office, and gave two seats to one of the men who had started the anti-British movement, Nagai Ryutaro, an anti-mainstream leader in the Minseito. In a l l , there was nothing about the composition of the cabinet, besides the Prime Minister's past, to suggest that i t was born of an anti-German reaction. Nor was there reason to believe i t could win the confidence of the Lower House. If the party maverick, Nagai, was unlikely to bring Abe MinseitS support, the Seiyukai dissident, Kanamitsu Tsuneo, was less l i k e l y to further liaison with that party. Supported mainly by the pro-German forces in the Army at a time of public h o s t i l i t y towards Germany, the Abe cabinet failed to create any balance between r i v a l institutions, and was soon overthrown by those i t excluded.3° When the previous Army-led ministry under Hayashi f e l l after irre-concilable conflict between the House and the Army, the ensuing f i r s t 21k Konoe cabinet was based on a third force and had been reasonably success-f u l . So had the two Navy-led cabinets which came between the House and the Army after a period of party-led cabinets. It is not surprising therefore, that a Navy man was chosen to succeed Abe. That neither a party-led nor an Army-led government could survive in the conditions of the 1930's was gradually gaining acceptance. Like Konoe, Hirota, and Hiranuma, Yonai was expected to win the confidence of both antagonists. Baba Tsunego also understood the problem: "The main concern of the Lord Keeper of the Privy Seal, the GenrS, and others in choosing the last Premier was to find a man who would not be opposed by the Army yet who could steer Japan's course in foreign and domestic affairs in collaboration with the Army rather than in opposition to i t . " Baba then noticed: "In the course of i t s organization, the Yonai Cabinet showed more friendship and respect for the p o l i t i c a l parties than any previous bureaucratic cabinet."31 Why then did this ministry last only six months? Was there anything about the coalition of forces behind i t to suggest the likelihood of such an outcome? Yonai's relationship with the House was closer than that of a l l the successful third-force cabinets of the decade. He included four party men and a l l received the blessing of their parties. But he failed to win the confidence of the Array. If the Abe cabinet, which was expected to capitalise on the Army's embarrassment in 1939 over i t s pro-German policy, could not resist i t s growing anti-British sentiment, how could a cabinet, . 215 which had as i t s Prime Minister and Foreign Minister men who were long time opponents of an Axis with Germany, possibly win the confidence of the Army? But for the opposition of Yonai and Arita Hachir5, the Hiranuma government would have concluded a military alliance with Germany, something the Army could not forget. Although the Army did not want one of i t s Generals to lead another cabinet for fear of another anti-Army reaction, i t was so incensed by Yonai's appointment that the Army Minister had to be summoned to the palace and be told to cooperate with the new government. As could have been expected, the Army soon found a way to unseat Yonai without appearing disloyal to the Throne.32 A theoretical point w i l l help explain why Yonai was succeeded by Konoe. When the national consensus breaks down in a separation-of-powers system, i t becomes extremely d i f f i c u l t to find a leader who can win over the warring institutions and introduce real shifts in policy. In America, many people in recent years have come to believe that only Ted Kennedy can do i t . In Japan i n the 1930's, the belief that only Konoe could gradually gained wider acceptance. In 19^0 Konoe came to power not merely as the sole person who could act as a successful third- force between the Army and the House, but as the leader of a popular campaign to unite the members of these in s t i t u -tions in a single mass organization. For the f i r s t time, the goals of the parties and the Army appeared, i f not similar, at least not contra-dictory, because both supported the new Konoe-led p o l i t i c a l organization. 216 The precedents of the Liberals inviting Ito to lead them in 1900, and the Progressives inviting Katsura to do so in 1913, provide the clearest i l l u s t r a t i o n of the party leaders' motives. On each of these occasions the problem was to obtain party control of the Cabinet. Because only a particular type of personality could become Prime Minister, the parties f e l t that'.the reason for their long exclusion from control over important Cabinet seats was, as the old parliamentarian Okazaki Kuni-suke said:33 There i s not anyone in our p o l i t i c a l parties . . . who has ever mastered this art of riding the steed of State. Without such mastery, the only alternative seems to be a coalition cabinet. When Yonai was appointed in 19^ -0, many party members were f i n a l l y per-suaded that a new leader was needed to restore predominant party control of the Cabinet. Nakajima Chikuhei, one of the f i r s t party leaders to throw in his lot with the new party movement, deliberately alluded to the precedent i n 1900 and said he would merge with Konoe just as Itagaki had merged with Ito.3^ By the end of the 1930's, the supporters of the new party took a less c r i t i c a l view of the Army's foreign and domestic policies, largely because they realised that no party could restore Lower House influence to what i t had been in the 1920's unless i t compromised with the Army in defend-ing Japan's international interests. That different views on how to revive party control of the Cabinet were at the heart of party and factional cleavages in the House is obscured by the nature of these cleavages: both parties were divided into a majority which opposed many Army proposals to 217 extend economic planning and controls, and a minority which supported them. But the minority's major purpose was to revive party fortunes by having a new Konoe-led party enact radical economic reforms. Because the Army was the only institution which had a domestic and foreign prog-gramme that could excite the imagination of the people on whom the party men ultimately depended, the cleavage took the misleading form of pro-and anti-Army.35 As in 1900 and 1913> the main supporters of a. new party in the late 1930's were younger members who had become impatient with the old tactic of "total opposition." In 19^0 when the House demanded Abe's resignation, both parties were split down the middle, and pressure to replace the Abe ministry with one based on Konoe's new party came from the younger members who had lost confidence in their leaders ' a b i l i t y to revive party morale and restore party fortunes. Within three months, both major wings of the Seiyukai under Kuhara Fusanosuke and Nakajima Chikuhei came out in favour of the new party. Because President of the MinseitS, Machida Chuji, refused to climb on the band wagon, he had to witness a huge defection from his party under Nagai soon after Konoe announced to the press that he intended to form a new party.3^ If Konoe came to power in 19^ 0> not merely as a third-force between the Army and the House, but with the positive support of the House, his assumption of office and proposal for a new p o l i t i c a l party was equally welcome to the Army. By January 19^0, the Army had been fighting i n 218 China for two and a half years with only mixed success, and had to see i t s major concern, domestic economic reform in order to strengthen the country's military capabilities in case of war with Russia, pushed into the background. But by this time the Army appreciated that no govern-ment could enact a policy to control and plan the economy without.the sanction of the Lower House of the Diet. Without Diet support, Hayashi had failed to pass the Electric Power Control B i l l , whereas soon after-wards, Konoe, with such support, had not merely passed this b i l l but the National Mobilisation Billaas well. The Army's greatest institutional deficiency was i t s lack of a legitimate popular base, without which i t could not really compete with the House for the affections of the public. If the implementation of Army proposals required popular support,.the Army had no alternative to forming an alliance with the House. As in 1925, when General Tanaka became President of the Seiyukai, personality was the basis of the compromise in 194-0, and the Army Vice Minister could confidently say': "The Army is united in backing Prince Konoe as the next premier."37 Because the preparatory commission on Konoe's new organisation, the nature and composition of which i s discussed in the section on new devices to avoid deadlock, included his C i v i l Service-dominated cabinet, the C i v i l Service was also a major influence behind his government.3° Throughout the 1930's, more dramatically so at the beginning and end of the decade, the Army and the House were the most influential 219 institutions responsible for the coalition of forces behind each ministry. While the coalition usually took the form of some third • force assuming the central positions in the government to mitigate Army-House conflict, there were two Army-led ministries, and one that had the whole-hearted support of the Army and the House, Once i t was admitted that Army-led cabinets were as provocative as House-led ones, and that third-forces could not always be perfectly impartial, pressures for Army-House cooperation increased and led to a truce in the late 1930's. The basic cause of the conflict was also the main reason for the new sp i r i t of cooperation. Inter-imperialist rivalry and domestic economic conditions in the 1930's made i t very d i f f i c u l t for the Army and the House to avoid encroaching on each other's legitimate "functions", because high levels of expenditure on armaments were bound to strain the economy and provoke interference in Army matters by that institution whose function it,was to rat i f y a l l expenditure. Because the Army found such expenditures essential to the responsible exercise of i t s function, i t could not help interfering in House affai r s . Each was incensed by the other's interference, particu-l a r l y when the veto power was freely used. The Army vetoed party premiers, the House vetoed Army premiers. After a while both sides saw that real power was shifting i n favour of third-parties, to their mutual disadvantage. They had become mainly negative forces-, dominant in bringing about the demise of cabinets, but unable to raise many that could translate their 220 own preferences into national policy. There were therefore in s t i t u -tional pressures for them to bury their differences, and in order to realise their own interests rather than simply veto those of each other, to raise cabinets of their own, as they did in July 19^0 and partly in April 1937* Cabinet-making in the 1930's required a greater willingness to compromise than ever before. No longer could a Genro-House compromise assume the predominant influence in raising cabinets, but both the Genr5 and the House were compelled to strike the best bargain they could in a more even competition involving institutions which in previous years had been content to take a back seat. The only cabinet that could survive in a separation-of-powers system in which most branches have more or less equal power was the kyokoku i t c h i cabinet. In the 1930's pressures for this type of cabinet did not merely come from a threatening international environment but from previous experiences with the Meiji Constitution. Conditions in the 1930's altered the institutional balance of power and brought institutions, which in previous years only occasionally upset the balance appropriate to those years, to the forefront of the competition in the 1930's. The Army and the Lower House, far from being the pivotal forces in Cabinet-making in the 1930's, had weaker incentives to support cabinets than the institutions which acted as third forces. A cabinet which i s not the main creation of•any institution i s unlikely to receive 221 strong support from that institution. Every third-force ministry in the 1930's was brought down by the House or the House and the Army, while both Army-led ministries were brought down by the House. C ab inet-Ere aking, 1932-1940 Cabinet-breaking power, while not as evenly distributed as Cabinet-making power, was nevertheless not the preserve of any single institution. Usually at least two institutions had to combine to turn out a government, although the Lower House's dominant influence in the dismissal of cabinets made the 1930's resemble the 1890's more than any other period. The reason i s that the sole institution with a legitimate popular base played a more limited role in raising cabinets in any other period excepting the 1890's. In the present section, once again only the institutional press-ures causing the demise^of cabinets, beginning with the Saito cabinet and excluding the second Konoe cabinet, are identified. Soon after the rise of the Saito cabinet, Mori Kaku of the Seiyukai and Lieutenant General Suzuki Teiichi of the Army were cooperating to pul l i t down. Until Saito's ultimate f a l l in July 1934, allegedly over corruption in the Finance Ministry, the Army and the Seiyukai plotted his resignation. Education Minister Hatoyama Ichiro said in September 1932 that the price of Seiyukai support was a promise to resign in December, and the President of the Seiyukai, Suzuki Kisaburo, made support of the budget conditional on a promise to resign after the session. But when the time came, Takahashi, the Seiyukai Finance Minister, refused to 222 f u l f i l his promise to resign, and the party's plans were foiled. In June the next year, the Seiyukai came out with a public statement of str i c t neutrality towards the government. In May 1933} when Lieutenant General Matsui Iwane called for a reorganisation of the cabinet, Saionji believed the Seiyukai was relying on the Army to take the odium of toppling the cabinet. By December, relations between the Army and Takahashi, whom the Seiyukai had v i r t u a l l y disowned, had reached breaking point. On the 20th, Generals Hayashi SenjurS and Mazaki Jinzaburo urged the Army Minister to resign, because nothing could be expected of the SaitS ministry. When the scandal over corruption in the Finance Ministry involving the Imperial Rayon Company fi n a l l y broke, Saito took the opportunity to step down. The real 39 pressures, however, came from the Army and the Lower House. ' An almost identical story can be told of the Okada cabinet, the f a l l of which Yamazaki also classifies as due to "taking responsibility for some incident," in this case the February 26 Incident.^"0 Again the real forces causing the cabinet's f a l l , rather than merely occasion i t , came from the Army and the House. Both plotted Okada's dismissal soon after his appointment. The Seiyukai expelled those party members who accepted cabinet seats, and took up i t s position firmly as an opposition party. The Araki Sadao and Mazaki Jinzaburo faction of the Army began i t s efforts in September, and by July 1935 was causing the Prime Minister and Army Minister considerable trouble, the more so when i t began to cooperate with 223 a section of the Seiyukai. In the September issue of Contemporary Japan, Baba Tsunego report-ed: "The Seiyukai was almost frantic in i t s machinations to p u l l down the Okada cabinet before the fated day [the election^ . O n January 31 1936, the Seiyukai proposed a vote of no-confidence, but the govern-ment had already decided to dissolve. Baba reported that " i t was being generally surmised" that the Okada government would "be deserted even by the Minseito and thus be compelled to r e s i g n . T h e cabinet f i n a l l y resigned after the February 26 Incident, in which Mazaki and his asso-ciates were heavily implicated, and which Okada interpreted "as motiv-ated by Army resentment at the 'Navy cabinets' which had . . . been used as a counter force to check the Army's program. "^"3 The f a l l of the Hirota ministry, the third attempt to mitigate conflict between the House and the Army by means of a third-force, resulted from a head on c o l l i s i o n between these two institutions, a c o l l -ision which ended in victory for the House. The trouble began when Army Minister Terauchi Hisaichi proposed to prohibit Dietmen from receiving Cabinet positions and flared up after the signing of the Anti-Comintern Pact on November 25 1936. At the party conventions in early 1937, the government was severely c r i t i c i s e d for submitting to the Army and i t s foreign policy, and on January 21 Hamada Kunimatsu echoed these feelings in the House. Terauchi reacted strongly and demanded that the House be dissolved, but because the four party ministers as well as the Navy Minister 22k would not comply, Hlrota f e l t he had no choice hut to resign. The House and the Navy were therefore the breakers of Hirota's cabinet.^" The demise of the Hayashi cabinet, the f i r s t to be led by someone who was solely an Army spokesman, was a clear-cut case of the House's cabinet-breaking power. Even Yamazaki regarded i t as one of five such cases.^"5 The Army had got the Cabinet to dissolve the House, and the parties responded by conducting their•election campaign under the slogan, "the dissolution was unconstitutional." The result was a massive defeat for the government, which resigned soon after the parties held a meeting calling on i t to do s o . ^ Yamazaki regards the resignation of the f i r s t Konoe cabinet as an instance of "a cabinet change to f a c i l i t a t e a policy c h a n g e , b u t this says nothing about the kinds of forces that necessitated a policy change and made i t d i f f i c u l t for the cabinet to continue. An examination of the main institutional pressures which Konoe had to endure for a year and a half reveals once again that the Army and the House were the institutions whose cooperation had been most d i f f i c u l t to retain. At the heart of the problem was the unanticipated China Incident, which from the beginning led to strained relations between the Prime Minister and the Army. The Army consistently acted on i t s own, often without even informing the Prime Minister. By the middle of November 1937, Konoe wanted to resign, because the Army was opposing the equal distribution, regardless of rank, of money to the families of the war dead and was calling for the reorganisation 225 of the cabinet.^ u When the Emperor asked Konoe whether he could control the Army, Kon-oe replied that the only way to do so was to merge the parties, implying, i t seems, an intention to balance the two main institutional forces more equally by strengthening the weaker one. But by this time the House had become an independent source of trouble. Konoe's National Mobilisation B i l l was c r i t i c a l l y questioned in the House, and although i t was even-tually passed, together with the Electric Power Control B i l l , i t had generated a great deal of heat, and the House had forced the government to back down on a number of points, for example, prohibition of public meetings and control of newspapers. When Konoe added insult to injury by rousing the ambitions of many party members through his encouragement of rumours that he intended to form a new p o l i t i c a l party only to drop the subject as suddenly as he had raised i t , his relationship with the House reached an a l l time low. With grumblings in bureaucratic circles and no prospect to end the China Incident and improving relations with the Army, Konoe decided to step down. The Army, the House, and less so the bureaucracy, brought the Prime Minister to this point. Yamazaki's classification of the Hiranuma ministry's resignation as another instance of "a cabinet change to f a c i l i t a t e a policy change" is quite accurate in this case, and i t is d i f f i c u l t to isolate particular institutional forces. Baba Tsunego also wrote, "resignation en bloc [hadj become necessary owing to the conclusion of the German-Soviet non-aggression treaty."^^ The government was almost entirely occupied throughout i t s 226 seven month term in an attempt to conclude a military alliance with Germany and to reconcile the position of the Army, which wanted to include Britain and France as countries against which the alliance was directed, with that of the Wavy and Foreign Office, which would only contemplate an alliance directed solely against Russia. Hiranuma leaned in favour of the Army's position, and anti-British demonstrations were financed by the Army and believed by the public to have had government support. If definite institutions are to be isolated as the ones which Hiranuma could not face after the discrediting of his policy, the Wavy and the Foreign Ministry are obvious. But equally important was the govern-ment's loss of public confidence, and the House could not have tolerated the existence of a government which was alienated from the House's own constituency.51 General Abe also resigned because he lost public confidence, but this time the House acted more directly. The issue was the supply and price of rice, about which the government seemed unable to do anything, and the Agriculture and Forestry Minister suggested resignation. Although Abe tried to keep on good terms with the orthodox party leaders, the middle-echelon members wanted to unseat him, and on January 7 19-40, they proposed a resolution of non-confidence, which was signed by 276 members. Wagai Ryutaro, Kanemitsu Tsuneo,jand Akita Kiyoshi, the three party members of the government, advocated dissolution, but the Army refused, because i t feared that an election would only s t i r up anti-Army sentiment in the 227 country. The Army's refusal to side with the supporters of dissolu-tion meant that the Prime Minister had no alternative but to bow to the wishes of the House, and Berger concludes that "a Diet centred campaign had resulted in the resignation of an army-supported govern-ment ." The Army1s role was minor.52 The Yonai ministry, which was supposed to command the support of the Army and the House, resigned when Yonai recognised that he had lost the confidence of both. His popularity i n the House decreased i n almost direct proportion to the mounting campaign to have Konoe lead a new p o l i t i c a l party. When Konoe publicly announced his decision to do so, the orthodox party leaders, who were on close terms with the government, could no longer contain the feeling in the House in favour of a new Konoe ministry. The Army, availing i t s e l f of the opportunity to get r i d of a cabinet i t had disliked from the beginning, forced the Army Minister to resign and refused to nominate a successor.53 The f a l l of Konoe's second cabinet in July 1941 took place under conditions of such ominous war clouds that i t is beyond the scope of this study. It belongs more to the category of Chamberlain's replacement by Churchill on the eve of the European war, than to the interplay of institutional forces with which this study is concerned. No attempt has therefore been made to f i l l in the missing symbols in Appendix I. Throughout the 1930's, the Army and the Lower House were the major forces responsible for the rise and f a l l of cabinets. In raising cabinets, their influence usually took the form of making third-force ministries 228 the only ones li k e l y to have any measure of success, vhile in bringing cabinets down, their influence was more direct. Table k indicates the breakdown of Appendix I into each institu-tion's major and minor roles in Cabinet-making and -breaking in the years 1932-19U0. The minor roles assigned to the GenrS after Hayashi's appoint-ment, roles which are assigned to the Jushin (Senior Retainers) in the Appendix, are explained in the next section on the functions of in s t i t u -tions. That section also explains who the Jushin were and why the roles of the Genro and Jushin are grouped together in Table k-. Table h. Major 'and Minor Roles in Cabinet -Making and -Breaking, 1932-19^ -0 • Cabinet-Making Army House Bureaucracy GenrS Wavy Council Peers Major 9 7 4 3 3 2 1 Minor 0 0 1 6 0 0 0 Cabinet-Breaking Major h 8 1 0 2 0 0 Minor 1 0 1 0 0 0 0 In 1932-19^ -0 the Army was the major maker of cabinets, the Lower House their major breaker. The period differed from the 1890's in two main respects. F i r s t l y , the House was more prominent in Cabinet-making than i t had been in the 1890's, and the Army more prominent in Cabinet-breaking that the GenrS had been in the 1890's, mainly because governments in the 1930's were headed by third-forces. Secondly, other institutions were more involved in raising and toppling cabinets than in the 1890's, 229 although not much more so in the latter role. An important similarity in the two periods is that the struggle between the two dominant i n s t i -tutions led to a compromise at the end of the period: the GenrS-House compromise of 1900, and the Army-House compromise of 19^0. Although the Army and the Lower House were the main influences on the Cabinet's existence in the 1930's, only the two Army-led governments were not strongly backed by at least three institutions and not brought down by major pressures from at least two. While in the l890's the Cabinet could be described as negatively responsible to the Lower House, which, on the whole, was the Cabinet's sole breaker, the typical cabinets of the 1930's could not be brought down by any single institution. The existence of the Cabinet was separate from any one institution and the description of the Cabinet as a balanced federation of institutions remains accurate. The Navy was:: sufficiently conspicuous in the raising of the Saito, Okada, and Yonai ministries, and the breaking of the Hirota and Hiranuma ministries, to be considered a force that always had to be reckoned with. Its absolute control over the Navy Minister's seat also indicated i t s unchallenged participation in every "federation." The roles of the GenrS, the C i v i l Service, the Privy Council, and the Upper House, however, were different and require individual mention. The Functions,of the Separate Branches One of the main characteristics of the years 1932-19^0 was each institution's growing willingness to confine i t s e l f to the exercise of 230 i t s constitutionally determined functions alone. Only when these functions overlapped did serious deadlock occur, as in the case of the Army, the Navy, the Foreign Office, and the Lower House. The comparative absence of Genro, Privy Council, and Upper House interference in the 1930's resulted largely from the failure of other institutions to encroach on matters these institutions regarded as their legitimate functions. By the 1930's they a l l came to accept a narrower definition of what their prerogatives were and confined themselves to acting within the" Constitution. The present section examines the functional specificity of. the Upper House, the GenrS, and the Privy Council. It also discusses the role of the C i v i l Service, the "Army-ization" of which parallels i t s "party-ization" in previous periods, but whose control of certain Cabi-net seats also reveals that the C i v i l Service confined i t s e l f to exercis-ing specific functions. It begins with the House of Peers. The part played by the House of Peers in the institutional struggles of the 1930's contrasted sharply with that played i n the 1920's and indicated that the Peers wanted no more than to perform their legitimate consitutional function, to safeguard the interests of the aristocracy and to act as a "check to hasty, one-sided . ... l e g i s l a t i o n . " ^ In the 1920's, when the Lower House could play a major part in i n i t i a t i n g l e g i -slation, the Peers were active in preventing the bourgeoisie and working-class from swamping the interests of the aristocracy. In the 1930's, 231 however, when such i n i t i a t i v e as the Lower House did possess was "checked and balanced" by the greater influence of the Army, the Peers f e l t less called on to perform this function. The only time any govern-ment faced any serious resistance from the Peers was when i t suggested something that came too close to home, like Peerage reform, which even Konoe was unable to implement.55 Another indication that the Peers were prepared not to interfere in matters that were the prerogatives of other institutions provided they themselves were not affected was their a b i l i t y to retain their share of representatives in the Cabinet. While the total number of Peers who served in this period was about the same as the number who served in any other period, Appendices II to V show that the Kenkyukai received a greater share of Upper House seats in the 1930 's than ever before. Eighteen of the Peers' Uo seats in the 1930's went to the Kenkyukai, while 17 of the 49 in the previous period did. Prime Ministers were always conscious of the reasons for allocating seats to the Upper House. For example, even the Peer-dominated f i r s t Konoe ministry appoint-ed Hatta Yoshiaki in a reshuffle in order to further liaison with the Peers. To the very end, the Upper House jealously guarded i t s prerog-atives and was not prepared to dissolve i t s factions in response to Konoe's c a l l for a new p o l i t i c a l order i n 1940.57 The position of the C i v i l Service in the 1930's was very similar to i t s position in the 1920's, except that the major factions into which i t 232 divided, apart from those along department lines, were pro- and anti-that institution which i n this period was exercising a major influence on appointments to the Cabinet and to certain-high C i v i l Service posi-tions. When Yamagata and the Seiyukai vied for this power, bureaucrats divided into two corresponding camps; when the DSshikai and the Seiyukai did, they were associated mainly with one or other party; and when the Army came to assert a major influence, they received a great incentive to become what Spaulding calls "revisionists", or c i v i l servants who "usually supported army policies."5® What Spaulding describes as the "military penetration of the C i v i l Service" was accomplished in stages. The f i r s t took place under Okada, who removed from the Foreign and Overseas ministries almost a l l powers in Manchuria and the leased territory and transferred them to the Man-.churian Affairs Bureau, which was largely controlled by the Army. The second instance of active-duty officers receiving positions that would previously have gone to civilians was the appointment of Army men to the Cabinet Investigation Bureau, also by the Okada cabinet. The c i v i l servants who received positions i n these bureaux were predictably "revisionists," as weretthose appointed to the heir of the Investigation Bureau, the Planning Board, which Konoe set up in 1937* Men who climbed onJ.the revisionist band wagon and who later received Cabinet positions included GotS Fumio, Hirota KSki, F u j i i Masanobu, Kido KSichi, Matsuoka Yosuke, Hoshino Naokiy. Yoshida Shigeru, and Aoki Kazuo.^9 233 But when the Army's encroachment on what the C i v i l Service regarded as i t s peculiar preserve began to threaten C i v i l Service sovereignty in these matters, factional cleavages disappeared and the bureaucracy united to defend i t s institutional interests. Even Hirota strongly resisted the establishment of the Asian Development Board in 1938» and when the Army wanted to set up a Ministry of Foreign Trade, which would have swallowed up the Foreign Ministry's Bureau of Commercial Affairs, the Foreign Ministry o f f i c i a l s revolted. The centre of the dispute was the latter's desire to retain control over appointment and dismissal of commercial attaches. Because the Foreign Office had been defeated over the Asian Development Board issue, i t was adamant this time. As many as 131 senior o f f i c i a l s submitted their resignations, and anti-Army feeling ran high. In the end, the government was forced to capitulate.^ 0 The C i v i l Service was not prepared to surrender so much of i t s power of promotion and appointment to the military. This was again illustrated by i t s successful resistance to a joint Army-Navy plan in 1936 to create a Cabinet Personnel Bureau to control C i v i l Service appoint-ments. The entire bureaucracy rose in opposition to such centralised control, and jealously guarded the tradition of each ministry's sover-eignty i n these-matters. Every, cabinet from that^of Hirota to Yonai tried and failed to implement this plan, which Konoe's second ministry ultimately abandoned.^1 Although the bureaucracy was "Army-ized" in the 1930's, i t s t i l l 234 retained the same degree of esprit de corps i t displayed in previous periods. Foreign Ministers were regularly career men from the Foreign Office, even though men like Hirota and Matsuoka were revisionists. But i n 1918-1932 Shidehara Kijuro had been associated with the Minseito and Uchida Yasuya with the Seiyukai, and in the 1890's Sone Arasuke and Aoki Shuzo were in the Yamagata clique. The Justice Ministers were reg-ularly Justice Ministry o f f i c i a l s and, particularly after the "Minobe Affair" in 1935, when Prof. Minobe Tatsukichi was dismissed from Tokyo University because he regarded the Emperor as an organ of the State rather than as the State i t s e l f , there was a tendency for Imperial Uni-versity Professors or o f f i c i a l s from the Education Ministry to occupy the corresponding Cabinet seat. Besides this obvious control of the Cab-inet, there was occasional evidence that bureaucrats had a crucial voice in the choice of ministers. For example, Harada said that Commerce and Industry Minister Kobayashi Ichizo in Konoe's second cabinet "was the 6? nomination of Vice Minister Kishi and no one else." Kishi Nobusuke was also crucial in having Fujiwara G-injiro appointed to this position under Yonai.°3 when Abe was looking for a Justice Minister in 1939, he asked Shiono Suehiko, a leading bureaucrat and ex-minister himself, to nominate someone and appointed Shiono's nominee, Miyagi Chogoro.0^" The C i v i l Service reached the height of i t s power in the 1930's: It was one of the third-forces that came between the Lower House and the Army in raising the f i r s t Konoe and Hiranuma cabinets; i t s Foreign 235 Ministry performed the same function in the Hirota cabinet and played a major role in unseating Hiranuma; i t was a major force behind Konoe's second cabinet; and i t helped bring down Konoe's f i r s t cabinet and raise Hayashi's cabinet. Besides this greater-than-ever influence on Cabinet-making and -breaking, the C i v i l Service had a greater influence on Cabi-net-composition than ever before. Appendices IV and V show that 27 of the 68 seats occupied by bureaucrats in 1932-1940 went to men who rep-resented the C i v i l Service alone, while 54 df the 62 corresponding seats in 1924-1932 were occupied by bureaucrats who were also members i n other institutions. The C i v i l Service was the main institution to capitalise on Army-House conflict i n the 1930's. The Privy Council in the 1930's, like the House of Peers, was much less ambitious about the range of functions in which i t was prepared to exercise a veto. Because the ra t i f i c a t i o n of foreign treaties did remain one which i t regarded as peculiarly i t s own, i t was. an active participant in foreign treaty disputes. For example, in June 1936, i t issued a rebuke to the government for•including a clause in the treaty on extra-territorial rights in Manchuria, and the government apologised. The President of the Privy Council was also one of the few privileged people to have taken part in the War Councillor's Meeting before the Emperor to discuss China policy in December 1937*^ But there is no record in the 1930's of the Privy Council ever venturing to exercise a veto over the Army, Navy, or Foreign Office, even though the President 236 of the Council remained an active participant i n the inter-institutional discussions on foreign affairs. The Privy Council's function to interpret the Constitution also became much narrower during the 1930's but also the one in which i t was prepared to assert i t s e l f more forcefully. By the 1930's the Constitution was generally regarded as workable only i f each of the separate i n s t i t u -tions confined i t s e l f to the exercise of i t s own legitimate functions alone. But because functions can never in practice be neatly compart-mentalised in this way, the practical problem of interpreting the Con-stitution centred on the d i f f i c u l t y i n appointing as Prime Minister a man who could preserve the institutional peace, usually by forming as balanced a "federation" as possible. The President of the Privy Council was always most active in times of Cabinet changes. In the consultations over candidates for the premiership he and the Lord Privy Seal were the sole participants who had a legitimate institutional base. The GenrS institution would disappear after Saionji's death, and the ex-premiers who were frequently brought in never had much influence precisely because they lacked an institutional base. A proposal in the early 1930's to make these men Privy Councillors would have increased the influence both of the Council and the ex-premiers and was rejected by Saionji for this very reason.^7 It i s frequently forgotten"by those who analyse Japanese p o l i t i c s in terms of feudal forces that no matter how prestigious a man be in 237 the social hierarchy, he is powerless without a hold on some po l i -t i c a l institution. Konoe, who desperately wanted to succeed Saionji, f e l t this acutely and proposed that the President of the Privy Council, Konoe himself, convene the Jushin (Senior Retainers) to select the new Prime Minister. When Saionji rejected the idea, Konoe decided to resign from the Presidency of the Council and make a new party his new base, bringing to mind ItS's decision in 1900.^ ® The role of the President of the Privy Council i n the group of men whom Titus calls "Negotiators", whose task i t was to select the most "balanced" man to form the most balanced cabinet, can be better understood in relation to the'role of the other "Negotiators," who in the 1930's included the GenrS, former prime ministers, and the palace o f f i c i a l s , mainly the Lord Keeper of the Privy Seal. During the 1930's, the institutional force which in previous periods was referred to as the Genro, sometimes simply meaning Yamagata, at others Saionji, was referred to as the Jushin and included a l l those persons whom Titus regards as "Negotiators." By the 1930's the Jushin had also learnt to live with the separa-tion-of-powers and to confine themselves to the exercise of certain functions alone. Of one of these functions, Titus writes: "The institutional separation of Court and Government . . . produced a parallel division of function: the emperor and Court ritualized, the Government decided and executed."^9 The paradox of an i n f a l l i b l e Emperor 238 who was to wield imperial prerogatives was solved by recourse to separation-of-powers theory. Titus continues His personal w i l l , which might be f a l l i b l e , was not the "imperial w i l l " , which was by definition the i n f a l l i b l e w i l l of the imperial ancestors. This in turn meant restricting the emperor to acts of formal r i t u a l , such as rites and the formal sanctioning of Govern-ment decisions. In so far as Court o f f i c i a l s were also Jushin, they also exercised the function of "negotiating," particularly after 1936, when the Genro ceased to be an independent force in po l i t i c s and became merely one of them. Of this function, Titus writes The Negotiators would advise the emperor to appoint as Prime Minister that person most able to "cope with the situation" or to r a t i f y that policy most "in line with the times" at any given time. In doing so they were influenced by their estimate of "public opinion." . . . By public opinion was meant the views of the persons who counted, not the public at large. . . . What counted was the weight assigned . . . to any one or combination of the components of imperial prerogative—Imperial Diet, Foreign Ministry, army, navy. The imperial prerogative was exercised by separate institutions, each-with i t s own function, while the "Negotiators" tried to appoint as Prime Minister the person most able to reunite these components into something that could be regarded as the single voice of the Government for the Emperor to sanction. If the "Negotiators" were to choose the best person, i t was essential that they did not lean in favour of any particular•institution. Until 1936, however, this was not the case, because u n t i l then Saionji, like his predecessors, did have fixed ideas on which institutions were to have 239 the major share of influence in normal times. In 1932, although he consulted representatives of a l l institutions, he did so individually in order to keep the ultimate decision i n his own hands.^2 In 193^ -, Saionji retained the i n i t i a t i v e by seeing to i t that the conference of Jushin, which this time included only the Lord Privy Seal, the President of the Privy Council, and former Prime Ministers (and therefore Taka-hashi and Wakatsuki), was firmly under his management and accepted the person on whom he and Saito had previously decided.^ i n February 1936", Saionji did not convene the Jushin, but merely consulted the President of the Privy Council and the Lord Privy Seal, and made his recommend-at i o n s . ^ Hayashi's appointment was Saionji's last act as an independent force i n the selection of Prime Ministers. Although he did not come to Tokyo because of illn e s s , he was consulted by the Lord Privy Seal, who also consulted the President of the Privy Council. Saionji nominated Ugaki, Hiranuma, and Hayashi. Because only his third choice was accept-able, Saionji only played a minor part in the rise of the Hayashi cabinet. When Konoe was nominated in 1937* a change i s believed to have taken place. The Lord Privy Seal was the person to whom the Emperor put the question of a successor, whereas in the past i t had been put to the Genro through the Lord Privy Seal. While the Lord Privy Seal consulted the Genro, he recommended Konoe on his own responsibility, and according to Yamazaki, "the GenrS's opinion was used only as a reference."^ 2k 0 Just before the f a l l of the Yonai ministry, a contemporary wrote :' ' The Hiranuma, Abe, and Yonai Cabinets a l l came into being after p o l i t i c a l conferences in the capital while Prince Saionji remained at his Okitsu v i l l a and was kept informed of the proceedings by his private secretary. . . . (T)here has developed what .virtually amounts to a new Genro, or group of advisers for the selection of Premiers. The key men in this group were the Lord Privy Seal and the President of the Privy Council, who were always consulted, and former premiers, who only sometimes participated. Hiranuma and Abe were chosen without the assistance of the former premiers, and in Abe's case the President of the Privy Council tipped the balance, because he refused to support the Genro's choice, Ikeda Seihin.^® Yonai and Konoe were nominated at a formal Jushin .kaigi, which was attended by the President of the Privy Council, the Lord Privy Seal, and former prime ministers, while Saionji was con-sulted through his secretary.^ Yonai was chosen against the wishes of the President of the Privy Council, and Konoe against those of the Genro, indicating that the Jushin had come to acquire a corporate existence and could overrule in the name of the whole the wishes of any single p a r t i c i -pant. The Privy Council did not take over entirely.the function abrogated by Saionji, but had to share i t with the Lord Privy Seal, and on occasion, former premiers. Because the Jushin were themselves also institutional representatives, they could exercise less discretion than the Genro had, and they played only minor roles in Cabinet-making. Unlike the GenrS u n t i l 193^, the Jushin only tried to assess institutional influence and to appoint the person most lik e l y to retain the maximum support. Neither the GenrS nor 241 the Jushin played any role in Cabinet-breaking i n 1932-19^0. They always wanted governments to remain in office as long as possible and rarely had any idea of successors u n t i l governments actually f e l l . After 1936, when the Jushin ceased to lean in favour of any part-icular institution but made i t their task to determine the national consensus, a parallel change took place at Court. Until then, Saionji had not only been much more than a barometer to assess the institutional balance of power but able to impress on Court o f f i c i a l s ' h i s views on the most desirable institutions to lead the Cabinet. The Lord Privy Seals Makino Shinken, Ichiki Kitokuro and Yuasa Kurahei were a l l constitutional Q Q monarchists like Saionji. Keeping pace with the new development in the nature and role of the Jushin, a new group of palace o f f i c i a l s gradually came into being and achieved pre-eminence in 1940 when Kido became Lord Privy Seal. Under these new men, Titus says, "a careful balance of institutional forces at the side of the throne was maintained, not to secure a partisan D-i stance but to ensure the accuracy of the 'national consensus'." To live with the separation-of-powers the Court i t s e l f had to acquire a balance of institutional representatives, a development that began in 1929 and reached i t s culmination in 1940. After 1929 only Admirals were 8? appointed to position of Grand Chamberlain, Titus points out why: Such appointments reveal the nature of the institutional balance required at the Emperor's side. The post of Chief Alde-de-Camp was always held by an army general. It appeared advisable by 1929 2k2 to counter the voice of the army at court, since by that time the army had mounted i t s institutional campaign to declare the Imperial Wil l in both domestic and international aff a i r s . The Army's voice needed to be countered because the advice of the Chief Aide-de-Camp invariably reflected the views of the Army. Examples are the Chief Aide's advice to dismiss Minobe Tatsukichi in 1935 and to exercise restraint rather than strong disciplinary action against the Army after the February 26 Incident.®3 Titus shows that during the 19 years spanning the period 1926-45 the four major palace positions were held by the following institutional Ph representatives : Imperial Household Minister: Foreign Office: 9 years. Home Office: 9 years. Finance Ministry: 1 year. Grand Chamberlain: Foreign Office: 2 years. Wavy: 17 years. Chief Aide-de-Camp: Army: 19 years. Lord Privy Seal: Foreign Office: 9 years. Home Ministry: h years. House of Peers: 5 years. He concludes:®5 A clear pattern of bureaucratic representation'at the emperor's side emerges. These might be called the four major constituents of modern Japanese bureaucratic p o l i t i c s : the army, navy, Foreign Ministry, and Home Ministry. This balance of forces among imperial aides, Titus points out, helped the Jushin maintain a balance in the cabinet :^ Those close to the throne, however, never permitted the Military complete domination. Institutional balances were s t i l l maintained in the composition of the Cabinet, though the military represent-atives carried the most weight. 2h3 Because military domination over other institutions in the 1930's was less than Genr5 domination in the early years and party domination in the 1920's, inter-institutional coordination was weakest during these years. The result was a proliferation of new devices to promote harmony, devices which f a l l into three categories: i ) those initiated by the government, i i ) those spontaneously created by the institutions in need of coordination, and i i i ) joint attempts. New Devices to Avoid Deadlock Although each of the three types of dei ex machina existed in previ-ous periods, they had been less developed and less frequently used. The 1930's witnessed a greater proliferation of their number and a higher development of their purposes and characteristics than ever before. A discussion of how they functioned belongs to the years 1932-19^0, when the absence of a single dominant institution made their role more import-ant. Only typical examples of each type are presented. Government-sponsored coordinating devices were almost invariably different kinds of•committees consisting of representatives from each of the main institutions. The f i r s t such committee was formed In June 1917 hy Terauchi, who set up the Advisory Council on Foreign Relations. According to Colegrove, i t s "ostensible purpose was to formulate a con-sistent foreign policy; i t s real purpose was to minimise opposition to whatever policy the cabinet should pursue."^ Consisting of representa-tives of the parties, bureaucracy, Peers, Privy Council, Army and Navy, 2hk 88 i t attempted to create a national consensus. A number of almost identical devices were created in the 1930's. The f i r s t was Okada's Naikaku Shingikai (National Policy Deliberation Council), which included party men, Jushin, Peers, military men, and representatives of big business.' According to Harada, i t s purpose was "to educate" these men with conflicting views in order to minimise opposition from the institutions they•represented.°9 Under Konoe's f i r s t ministry, the main coordinator was the group of Sangi (Cabinet Councillors), which, according to Ogata Taketora, was created "to strengthen his cabinet" by enabling him "to confer with leaders of the nation representing as many different interests as poss-i b l e . " - ^ It also consisted of representatives from the Army, Navy, parties, bureaucracy, and business. The inclusion of party men in his cabinet, Ogata says, was considered'insufficient to assure the Cabinet of the support of the Diet."^ 1 Other schemes introduced by Konoe were less comprehensive and only usually included representatives of the armed services, the zaikai (the financial world), and the C i v i l Service. Examples are the Kikakuin (Planning Board), the Liaison Conference, and the Five Minister Confer-ence, which included the Prime Minister, the Army, Navy, Foreign, and Finance Ministers.92 Th e reason for the exclusion from these bodies of representatives from the two Houses of the Diet seems to have been that they a l l dealt almost exclusively with foreign policy, particularly the 245 war in China. As on past occasions when the country was at war, the two Houses made no attempt to interfere with i t s conduct, and there-fore were not in need of coordination. The second category of expedients to prevent deadlock, inter-institutional factional alliances, resulted from the chronic faction-alism a l l institutions endured under the separation-of-powers. Because no single one of them, no matter how disciplined, could ever hope to translate i t s desires into government decisions without a great deal of compromise with others, different factions within institutions naturally advocated different kinds of cross-institutional alliances. If the concurrence of a l l institutions was essential before the govern-ment could act, one way, perhaps the best, to gain general agreement was to have groups of people within each one who worked for i t s acceptance of the desired policy, much the way Yamagata did. There are so many cases of the use of this tactic in the 1930's that the concliision i s hard to resist that institutional forces were one of the main causes. Only a few of the many examples w i l l be given. The Tokonami Takejir5 and Kuhara Fusanosuke factions of the Seiyu-kai were among the f i r s t to learn to live with the separation-of-powers and recognise that their party would never achieve the degree of control over the Cabinet i t had i n the 1920's. The most i t could expect were a few Cabinet seats, and rather than uncompromisingly oppose a l l cabinets, certain party factions sought to participate in cabinets through cooperating 246 with factions i n other institutions. In the early days of the Sait5 cab-inet, Tokonami was in contact with the Hiranuma faction, which i t s e l f was influential in the Privy Council, the Army, and the C i v i l Service, parti-cularly the Justice Ministry. Hayashi, a member in the Army's Kodoha, was also in touch with Hiranuma at this time, and by 1935 other members of the Kodoha were involved: Araki Sadao and Mazaki Jinzaburo. J In December that year, Kuhara, in cooperation with a group in the Army, planned to submit a non-confidence motion in the government over i t s handling of the controversy over the Emperor organ theory and to bring down the cabinet.9^ In January 1938, Matsuoka Yosuke, a Foreign Ministry "revisionist," was in association with Kuhara and General Minami Jiro, and in December 1939, an alliance between Kuhara and Generals Mazaki and Araki was reported to have been formed.95 While the precise tracing of factional alliances i s beyond the scope of this study, the tendency for House factions to favour alliances with Army men in this period above ones with one another is easy to discern. In the 1920's, factions in institutions with less control over the Cabinet than the parties tended to form alliances with the parties, whereas in the 1930's, those with less influence than the Army favoured cooperation with the Army. Table 5 indicates that fewer bankers, bureaucrats, and military men joined the parties in the 1930's than in the 1920's. Table 5_. Institutional background of Lower House Members, 1928-40.9°" Election Ex-Officials Ex-Military Ex-Bankers Businessmen 1928 kl k 5 92 1930 36 3 2 82 1932 39 1 1 - 79 1936 27 0 0 72 1937 9 3 1 72 2hj While corresponding figures are not available to indicate the increased gravitation of men from other institutions towards the Army, the "Army-ization" of the bureaucracy has already been noted. A good example of r i v a l Foreign Office factions which cooperated with rivals in another institution was the Shiratori Toshio faction, which worked with certain Army and Navy men to oppose Admiral Nomura KichisaburS's appointment as Foreign Minister under Abe, and the Tani Masayuki faction, which supported Nomura.97 Another example of gravitation towards the Army i s that Finance Ministers, who were typically bankers, in the late 1930's were no longer also party members but were on close terms with the Army, for example, Yuki ToyotarS and Ikeda Seihin. A f i n a l example is that the men whom the parties t r i e d to recruit as Presi-dents had become military men like SaitS Makoto and Ugaki Kazushige.9® That these cross-institutional alliances partly served to harmonize relations among institutions i s indicated by the development for this very purpose of one of them into the third type of coordinating device, a cross-institutional mass movement. The Taisei Yokusan Kai (imperial Rule Assistance Association) of 19^ +0, was l i t t l e more than an expanded version of the Konoe faction of the early 1930's, which was a loose association of leaders of major institutions, who r a l l i e d round Konoe as the man most l i k e l y to help further their purposes. In 1931 Konoe was in contact with Shiratori of.'.the Foreign Office and Mori Kaku of the Seiyukai in an effort to make Hiranuma premier. 2k8 He was also in contact with Army leaders like Suzuki Teiichi, Nagata Tetsuzan, and Obata ToshishirS, and when he met Araki Sadao, he became closely associated with the K5d5ha for many years. His association with party leaders has already been mentioned, and as a member in the Upper House since 19l6 and i t s President since 1 9 3 3 , he had extensive contacts among the Peers. In later years, when the Army and the parties urged him to form a new party, he was always careful to say that what he wanted was not a new party, but a new " p o l i t i c a l order" (Shin s e i j i  t a i s e i ) . What he envisaged was a kyokoku (whole nation) p o l i t i c a l system, and his new organization was to include, as he put i t , "a legis-lative - branch, an administrative branch, and, in a sense, the supreme command."99 Kuhara saw eye to eye with Konoe much more than did the other party leaders, and the members of the national and local councils which Kuhara advocated were to be selected from the two Houses of the Diet, the bureau-cracy, and the armed services.-'-^ Both men regarded the new organization as primarily an organ of coordination with a mass base. Konoe said i t would absorb the unions, agricultural groups, and commercial associations, and would include Diet and non-Diet leaders. It would also work closely with the Army.-^l Konoe never intended i t to be a Diet-centred p o l i t i c a l party but a cross-institutional mass movement to support a government that was the victim of conflicting institutional pressures. This intention i s revealed in the l i s t of 2k persons whom Konoe nominated to serve on the 2k9 preparatory commission. There were six members of the Lower House, three Peers, three high ranking bureaucrats, the Chairman of the National Council of Village and Town Mayors, two Army Generals, two Admirals, two businessmen, four leaders of the press, and one or two others. Later the members of Konoe's cabinet were also appointed. But even before the new organization was formally launched, Konoe had begun to lose interest in i t , probably because i t was d i f f i c u l t to translate the theory into practice. The f u t i l i t y of trying to coordinate conflicting institutional pressures by including their representatives i n the same cabinet should have taught the advocates of the new p o l i t i c a l order the f u t i l i t y of any organization based on the same principle. None of the expedients in the f i r s t category had done much to mitigate deadlock, and conflicting pressures in Konoe's new organization paralysed i t as much as they did the Cabinet. The Diet, the Home Ministry, and the Army soon discovered that they had opposite views on what i t was supposed to be.l°3 Berger says that as a result of the new organization,-'-0^ neither the p o l i t i c a l system nOr the balance of power among the elites underwent further change. . . . Perhaps the most striking feature of the IRAA was the minimal impact i t had on p o l i t i c a l institutions and relations among e l i t e s . The relationship of the cabinet, Diet and military services was ultimately l e f t unchanged by the IRAA. In i t s e l f the IRAA did not enable either the cabinet or the military to impose new constraints on the independence of one another, nor did i t permit either oftthem to deprive the Diet of i t s constitutional prerogative to approve budgets and represent the people. . . . The new order also had l i t t l e effect on the oper-ation of the cabinet system and the continuing need for coalition among the e l i t e s . 250 The sudden demise of this f i n a l attempt to cover up somehow the separation-of-powers only emphasises that everyone had really learnt to live with i t . There could be no solution u n t i l the Cabinet i t s e l f ceased to be a federation of representatives from antagonistic i n s t i -tutions and included only like-minded men. This in turn would only be possible once the power to make and break i t was the sole possession of only one institution, and the Cabinet became responsible to i t alone. 251 CONCLUSION Throughout the period I89O-I9U5, the most striking characteristic of the Japanese Cabinet remained i t s chronic weakness. It could ordin-a r i l y be unseated by almost any two of a variety of institutions and a new cabinet could "rarely be raised by fewer than three. Its existence was therefore quite separate from each individual institution, and to describe i t as "fused" with or responsible to any other institution would be incorrect. Its coming to office was quite separate and independent of each individual institution and i t s tenure in office usually depended on the collective w i l l of a l l but one: i f two conspired, the Cabinet would ordinarily f a l l . Unlike the American executive, which cannot be removed by any one or even a combination of other institutions except in the gravest of constitutional crises, the Japanese executive suffered without an existence completely independent of other institutions. It ended with a l l the disadvantages of an executive in a separation-of-powers system and none of those of an executive in a parliamentary system, because i t could be unseated, not by one other institution, but by a variety of combinations of other institutions. Its existence was precarious because, not having a completely independent existence like the American executive and not being the sole creation of only one other institution like the British executive, no other institution had a strong incentive to support i t . Its precarious existence then further decreased incentives to support i t and i t s existence became more precarious. 252 The prewar Japanese polity was neither a pure separation-of-powers type, nor a pure parliamentary type. The government was divided into separate branches, each of which had a corresponding identifiable function, and each of which possessed a limited veto power over the others. The members of most branches were usually members of only one branch, although this was less true of the two Houses of the Diet, the Cabinet, and the C i v i l Service. Ex-officials, whose influence in their ministries did not usually cease with their departure from government service, were frequently members of one or other House. Cabinet members were always members of one or more other institutions, and many members of the Upper House were also members of parties in the Lower House. Only the last kind of overlapping membership also exists i n America. But deadlock was as much a characteristic of prewar Japanese government as American government. That separation-of-powers theory provides the best understanding of the prewar Japanese system is also indicated by no single institution's complete control over appointments to any other institution. Because none was totally dependent on any other, each had a separate existence, although the degree of separateness varied. A l l except the Genro and the Cabinet had an independent power base, which guaranteed that they and their constituencies alone, excepting the Upper House 31$ of whose members were imperial appointments, determined their own composition. The Genro had no independent power base and would either have become 253 "fused" with some other institution or disappeared completely after Saionji's death. The position of the Cabinet was different and reveals that certain parliamentary forces were present, even though the drafters of the Constitution firmly intended that no other institution affect the Cabinet's composition or existence. Because in practice each in s t i t u -tion, excepting the GenrS and the Privy Council, regularly controlled a few Cabinet seats, a "part" of the Cabinet was "fused" with each i n s t i -tution and was responsible to i t . In a sense the Cabinet was "concurrently responsible" to a number of institutions, each of which possessed the entire power to appoint part of i t . But no single institution could raise an entire cabinet, and the Cabinet was s t r i c t l y speaking not responsible to any other institution. Although each institution could also veto Cabinet policy, only in the 1890's and on a few other occasions could only one institution, the Lower House, dismiss an entire cabinet, which was occasionally "fused" with the House and responsible to i t in this negative sense. And because the Cabinet could usually be dismissed by any two of a number of i n s t i t u -tions, one could stretch the language of parliamentary theory and say that the Cabinet was "concurrently responsible" to any two institutions, although the term "responsible" is best reserved for an institution which possesses the entire power of appointment and dismissal. On balance, although the prewar Japanese polity i s more comprehensible 25 k i n terms of separation-of-powers theory, to c a l l i t a hybrid system may be more accurate, because i t also had certain characteristics of parlia-mentary systems. A Cabinet that is responsible to only one institution can easily retain the support of that institution. A Cabinet that is "responsible" to different combinations of a l l institutions cannot win the whole-hearted support of any. A President who is "responsible" to no other institution can ordinarily continue in office without the support of any. Because the prewar Japanese Cabinet was "responsible" to a l l , had the whole-hearted support of none, and needed the support of a l l , i t acted as i f and was i n fact responsible to none. The result was that i t s average l i f e span was only l.k years, and i f the second Ito and f i r s t Katsura ministries are excluded, the average was just over a year. What guarantees the longevity of B r i t i s h cabinets is that i f an i n s t i -tution i s the sole maker of a cabinet, at least in peace time i t w i l l not be that cabinet's breaker. This mechanism could not come to the rescue of the Japanese Cabinet, which was not made solely by one institution. What guaran-tees the longevity of the American executive i s that i f an institution has no power to make a President, i t ordinarily has no power to break him either. This mechanism was also of no use i n Japan, because a l l institutions had some power to make the Cabinet. What determined the rise and f a l l of the Japanese Cabinet was that i f an institution did not take part in making a cabinet, i t would try to break that cabinet, although i t did not necessarily follow that i f an institution conspired to break a cabinet, i t would participate 255 in making the next cabinet. Until 1932, the only cabinets whose f a l l resulted from a major use of influence by the same institution that played a major part in raising them were the second Matsukata, f i r s t and second Okuma, and f i r s t Wakatsuki cabinets. On each occasion the institution was the Lower House, but only the second Matsukata cabinet was brought down by the same party that played a major role in raising i t . In the years 1932-1940, the rise and f a l l of a l l cabinets, excepting the two Army-led ones, resulted from a major use of influence by the Lower House. But the Lower House was not the pivotal force in raising any of these cabinets, a l l of which were formed by prime ministers from institutions other than the Lower House. Because the separation-of-powers system in Japan was never pure and the Cabinet was "responsible" to a l l institutions and yet to none, there was a greater likelihood in Japan that one institution would encroach on the functions of another. The greater any institution's power to make and break the Cabinet, the greater i t s a b i l i t y to usurp functions not constitutionally i t s own. Because in different periods, different institutions could exercise more than their f a i r share of Cabinet-making and -breaking power, they could also exercise more than their legitimate share of government functions. In any separation-of-powers system, the nature of the dominant problem facing the nation at any time w i l l tend to determine the balance 256 of power among the various institutions. In America, the Supreme Court lost influence i n the 1930's when the Congress and the President advocated social reform in response to a national mood overwhelmingly in favour of such a change. In the 1950's, however, the Court could gain influence when i t sought to extend the rights of Black Americans, because this was a time when the nationrwas deeply concerned to imple-ment principles relating to the rights of man. On the question of States' rights, as long as the economy was not inter-dependent to a degree that laissez-faire philosophy was seriously questioned, the institution that championed States' rights most staunchly, Congress, gained the upper hand. But once the need for national economic policy became more urgent, the balance shifted in favour of the President, the only institution i n a position to undertake the kinds of measures that would?.;satisfy the public mood. The same applies to foreign policy, an area in which the President alone is well situated to act decisively. During the years of post Second World War American imperialism accom-panied by an aggressive national mood, the President achieved a degree of influence i n foreign policy unequalled by any previous institution in any issue area. The pattern of institutional rise and f a l l in Japan was also largely determined by the nature of the dominant problem confronting the nation in successive periods, although the hybrid nature of-the Japanese system without i t s American stabilisers permitted institutional influence to 2 5 7 fluctuate much more than i t can in America. In the lfiQO's, when State-building was the dominant problem, i t was hardly surprising that the revolutionaries who set up the new State were the dominant institution. In the next period, the absence of a single a l l pervasive problem was paralleled by the absence of a single dominant institution. Before industrialisation under capitalism had reached the stage of unleasing a discontented urban proletariat and before any real problems in the nation's foreign relations had emerged, no overriding concern could t i p the balance in any institution's favour. But after the First World War, the economy received a boost that almost overnight brought to the fore a discontented urban working class and a greedy capitalist class whose conflicting interests determined the nature of the major national problems during the 1920's. Because the Lower House of the Diet was controlled by the capitalist' class, other institutions, which were equally opposed to the demands of the new proletariat, were content to leave the House as the major battleground for the time being, provided of course there was not too much encroachment on what they regarded as their special prerogatives. But towards the end of the 1920's and throughout the 1930's, when foreign relations and colonial wars absorbed the nation's consciousness, the balance was bound to shift in favour of those institutions constitutionally entitled to deal with these matters: the Army, Navy, and Foreign Office. But why did so much influence go to the Army, rather than to the 258 Navy or Foreign Office? Partly for the same reason that no matter how popular the actions of the Supreme Court in the 1950's and 1960's, this institution could never hold the balance of influence for long, because i t has no direct ties with the public at large. In Japan, while the Navy and Foreign Office had no such ties with the people, the Army did, mainly because many of i t s recruits were peasants, s t i l l the largest group in Japanese society, and the Army was the only institution to undertake a radical defence of peasant interests, at least u n t i l 1936. The Army also gained more influence than the Navy and the Foreign Office because, like the C i v i l Service in domestic affairs, the Army actually executed the policy of expansion i n Asia and the Navy was only occasionally involved. Because the Lower House had a legitimate and secure popular base, i t could become the Array's main competitor for the position of the dominant institution, even in times when the nation was concerned mainly with foreign affairs and colonial wars. In the long run therefore, whether or not the Americans had introduced constitutional changes after the war, the departure of foreign affairs from the centre of the stage would have meant a gradual evolution of the Japanese system from the separation-of-powers type to the parliamentary type. The system's great thorn in the flesh, the absence of a secure basis for the executive's existence, would also have allowed i t to develop in a direction denied the American and post-1962 French F i f t h Republic systems, whose executives 259 are popularly elected and "responsible" to no other institution and so unable to become responsible to only one of them. The Japanese exec-utive, because i t was "responsible" to a l l , would have stood a chance of becoming responsible to only one. 260 FOOT-NOTES INTRODUCTION 1 For example, Uyehara Etsujiro, The P o l i t i c a l Development of Japan, I867 - I 9 0 9 (London, 1910). 2 Thomas Francis Mayer-Oakes trans, and intro., Fragile Victory: Prince Saionji and the London Treaty: Issue from the Memoirs of Baron  Harada Kumao (Detroit; Wayne State Univ. Press, 1968), p. 2 0 . 3. Frank 0 . Miller, Minohe Tatsukichi: Interpreter of Constitutiona- lism in Japan (Berkeley: Univ. of California Press, 19^5 )> p. 135-k Ito Takashi, "Conflicts and Coalitions in Japan, 1930: P o l i t i c a l Groups and the London Naval Disarmament Conference," in Sven Groennings, E.W. Kelly, Michael Leiserson, eds., The Study of Coalition Behaviour: Theoretical Perspectives and Cases from Four Continents (New York: Holt, Reinhart & Winston, Inc., 1970), p. 162. 5 Vere H. Redman, "How the Cabinet i s Controlled," Contemporary Japan, 1, No. 3 (1932) , 407. 6 Ibid., p. 418 7 M.J.C. Vile, Constitutionalism and the Separation of Powers (London: Oxford Univ. Press, 1967), p. 13. 8 Ibid., p. 18. 9 Ibid. 10 Maurice Duverger, P o l i t i c a l Parties (London: Methuen, I 9 6 U ) , pp. 393 f f • 11 Leon D. Epstein, P o l i t i c a l Parties in Western Democracies (London: P a l l Mall Press, 1967), p. 36T 2 6 l 12 Ibid., p. 35-13 Ibid., p. 340. Ik A.H. Birch, Representative and Responsible Government: An Essay on the British Constitution (London: George Allen and Unwin Ltd., 1964), pp. 72 f f . 15 Walter Bagehot, The English Constitution, intro. by R.H.S. Crossman ( l 8 6 6 ; rep. London: Cox and Wyman Ltd., 1963), pp. 65-68. 16 Ibid., pp. 7k, 79. 17 For example, books on particular cabinets include, Hamaguchi Nai-kaku Hensanjo, Hamaguchi Naikaku (Tokyo, 1929); and Noi Hideichi, Konoe  Waikaku no Shutsugen ni atari (Tokyo: Teito Wichi Nichi Shinbunsha and JitsugyS no Sekaisha, 1937)• General works on the Cabinet and minister-i a l changes include, Hirota Naoe, Naikaku1K5tetsu Gojunenshi (Tokyo: Shun'y5d5, 1930); and Maeda Renzan, Rekidai Naikaku Monogatari, 2 vols. (Tokyo: J i j i Tsushinsha, 1961). The only research known to me that deals specifically with problem in Bagehot's t-erms is that of Vere H. Redman, "How the Cabinet is Controlled." 262 CHAPTER 1 1 Article LV. 2 Ito Hirobumi, Commentaries on the Constitution of the Empire of  Japan, trans, by l t d Miyoji, 3rd ed. (1889: rpt. Tokyo: Chuo Daigaku, 19317, P. 85. 3 Ibid., pp. 92-93. k Ibid., p. 93. 5 Herschel Webb, The Japanese Imperial Institution in the Tokugawa  Period (New York: Columbia Univ. Press, 1968). 6 Chitoshi Yanaga, "The Theory of the Japanese State," Diss. Univ. of California 193k, pp. 163-67. 7 M.J.C. Vil e , Constitutionalism and the Separation of Powers (Oxford: The Clarendon Press, I967): pp. 212 f f . 8 Sir Alpheus Todd, Parliamentary Government in England, was perhaps their main source on English constitutionalism. See George Akita, Foundations of Constitutional Government in Modern Japan, 1868-1900 (Camb. Mass.: Harvard Univ. Press, 1967), p. 11. In his preface Todd wrote, "the great and increasing defect in a l l parliamentary governments, whether provincial or imperial, i s the weakness of executive authority." Quoted in Akita, p. 206, note 33-9 Ronald Butt, The Power of Parliament (London: Constable & Co., I967), pp. 61-96. 10 The Meiji leaders did not want a strong Emperor. It5 modelled himself on Bismarck and envisaged a Bismarckian Emperor who was willing to defer to the Chancellor. 263 11 Johannes Siemes, Hermann Roesler and the Making of the' Meiji  State (Tokyo: Sophia Univ. Press, 1966), p. 47. 12 Quoted in Joseph Pittau, P o l i t i c a l Thought in Early Meiji Japan (Cam. Mass.: Harvard Univ. Press, 1967), p. 178. 13 Johannes Siemes, p. 85. 14 Ibid., pp. 86-87. 15 Ibid., p. 88 16 Ibid., p. 96. 17 Imperial Ordinance no. 22, April 28 1888, Chapter II, Article VI, rpt. in W.W. McLaren, "Japanese Government Documents," Transactions of  the Asiatic Society of Japan, 42, part 1 (May 19l4), 128~. 18 The Constitution did not establish the practice that high ranking naval and army officers alone could be appointed Service Ministers. It had been pursued by custom for many years before i t was legalised in the Imperial Ordinance regarding the Organisation of the Navy Department, No. 194, May 19 1900. 19 See p. 43 below. 20 Uyehara Etsujiro, The P o l i t i c a l Development of Japan: I867-I910 (London: 1910), p. 129. 21 For example, Hugh Borton, Japan's Modern Century, 2nd ed. (New York: The Roland Press Co., 1970), p. 16JI 22 Ito Hirobumi, pp. 89-90. 23 Ibid., pp. 94-95. See also Siemes, pp. 94-95. 26k CHAPTER 2 1 Uyehara, pp. 195-96. 2 Akita, p. 24-7, note k; Tetsuo Najita, Hara Kei in the Politics of  Compromise, 1905-193-5 (Cam. Mass.: Harvard Univ. Press, 1967), p. 232, note 2; Koshaku Matsukata Masayoshi Den, ed. Tokutomi Iichiro (Matsukata Masayoshi Denki Hakkojo, 1935), I, 375-76; Ito Hirobum Pen, ed. Shumpo Ko Tsuishokai (Tokyo: Shumpo Ko Tsuishokai, I9U0), III, ik^-^k. 3 Akita, pp. 90-91, 105. k Ninon Kokkai Nanajunenshi, ed. Maezawa Hiroaki (Tokyo: Shinbun Godo Tsushinsha, 1953), II, 62, hereafter cited as Nihon Kokkai. 5 Roger Hackett, Yamagata Aritomo in the Rise of Modern Japan, 1838-1922 (Cam. Mass.: Harvard Univ. Press, 197I), p. 177; Inukai Bokudo Den, I, as cited in Nihon Kokusei Jiten, comps. Nihon Kokusei Jiten Kan-kokai (Tokyo: Maruzen, 1953), II, 686, hereafter cited as NKJ; Akita, 126; Nihon Kokkai, I, 136; Tokyo Nichi Nichi Shinbun, Sept. 25 1596, as cited in NKJ, II, 687. 6 H.N.G. Bushby, "Parliamentary Government in Japan," The Nineteenth  Century (July 1899), p. ikk; See Also Akita, p. 128. 7 Ito Hirobumi Pen, III, 370-90. 8 Hackett, p. 185. 9 Tokyo Asahi Shinbun, June 29 1898, as cited in NKJ, III, kO; Joyce C. Lebra, ^Japan's First Popular Statesman: A Study of the P o l i t i c a l Career of Okuma Shigenobu ( I 8 3 8 - I 9 2 2 ) , " Piss. Radcliffe College 1958, p. 199. 10 Ito Hirobumi Pen, III, kok; Hackett, pp. 188-91; J i j i Shinpo, November 8 I898, as cited in NKJ, III, 135. 11 Naikaku Kanbo, Naikaku Seido Nanajunenshi (Tokyo: OkurashS insatsu-kyoku, 1955), p. kk. 265 12 Hackett, pp. 136-37, 1^1, 157j Robert Scalapino, Democracy and  the Party Movement in Prewar Japan: The Failure of the First Attempt (Berkeley: Univ. of California Press, 1953), p. 157. 13 Nihon Kokkai, I, 43-44. Ik Tokyo Nichi Nichi Shinbun, July 29 1892, as cited in NKJ, I, 67I; Yamada Tokazo^ "Shinbunshi yori mitaru Ninon ni okeru Naikaku seiritsu no keishiki," Kokka Gakkai Zasshi,38, No. k (192k), 56; Gikai Seiko Nana- junenshi, ed. Shugiin Sangiin (Okurasho ins atsukyoku, 1963), XII, 80-84, hereafter cited as Gikai Seido. Vol. X l l is t i t l e d Kenseishi Gaikan. 15 Nihon Kokkai, I, 59-60. 16 Ibid. 17 Akita, p. 100 18 Nihon Kokkai, I, 60, Akita, pp. 102-C4; Tokyo Nichi Nichi Shinbun, July 31 1892, as cited in NKJ, I, 671. 19 Uyehara, p. 223. 20 I d d i t t i Smimasa, The Life of Marquis Shigenobu Okuma: A Maker of  New Japan (Tokyo: Hokuseido Press, 19^0), p. 279. 21 Akita, pp. 111-115; Nihon Kokkai, I, 7U ff,. 22 Scalapino, p. 167. 23 Nihon Kokkai,!, 134; Akita, pp. 120-21; Tokyo Nichi Nichi Shinbun, August 29 I896, as cited in NKJ, II, 120. 266 2 4 Akita, p. 120. See also Nihon Kokkai, I, 13U; Koshaku Yamagata  Aritomo Den, ed. Tokutomi IchirS (Tokyo: Yamagata Aritomo Kokinen Jigyokai, 1953), H I , 284-286. 25 Tokyo Asahi Shinbun, December 28 1897, as cited in NKJ, II, 706-07; Lebra, pp. 178-88; Nihon Kokkai, I, l40-Ul, 148-51; Akita, p. 259, note 8, p. 260, note 10, p. 126; Hara Satoshi (Kei) Nikki, ed. Hara Keiichiro (Tokyo: Kangensha, 1950-1951), November"!) I 8 9 7 , 11:1, 229. 26 Tokyo Asahi Shinbun, June 8 and 11 1898, as cited in NKJ, III, 27-30; Hackett, pp. 181-82; Nihon Kokkai, I, I58-I6O; Kokumin Shinbun, June 27 I898, as cited in NKJ, III, 39-40. 27 Lebra, pp. 199-206, Otsu Jun'ichiro, Dai Nihon Kenseishi, IV, as cited in NKJ, III, 120-129; Nihon Kokkai, I, 165-173-28 Hackett, p. 208. See also J i j i ShinpS, November 30 1898, as cited in NKJ, III, 242-43; Yamagata Aritomo Den, III, 350-57; Ito Hirobumi Den IIl7~4~63-468; Hackett, pp. 187-208; Uyehara, p. 242. 29 W.W. McLaren, A P o l i t i c a l History of Japan During the Meiji Era, 1867-1912 (London: George Allen and Unwin Ltd., 1916), pp. 228-29. 30 Uyehara, p. 152. 31 • See note l6 above. 32 Nihon Kokkai, I, 74; Gikai Seido, XII, 93; Hackett, p. 155-33 Nihon Kokkai, I, 78; Gikai Seido, XII, 96. 34 Tokyo Nichi Nichi Shinbun, March 21 I897, as cited in NKJ, II, 695. 35 See Appendix II for the dates of a l l ministerial appointments, including reshuffles, and a l i s t of a l l Cabinet members in 189O-I9OO. 267 36 Kono Togama and Soejima are possible exceptions, because their appointment was partly to f a c i l i t a t e relations with the Lower House in order to keep the whole cabinet i n office. See Akita, p. 100. 37 Nihon Kokkai, I, 133; Gikai Seido, XII, 107; Hara Kei Nikki, February 3 I896, 11:1, 132. When Nomura Yasushi was dismissed to make way for Itagaki, he was temporarily replaced by Yoshikawa Kensei to make the reason less obvious. In a letter to Yamagata, Kiyoura Keigo acknow-ledged the truth. See Yamagata Aritomo Den, III, 282-84. 38 Akita, p. 85. 39 See note 37 above. ho J i j i Shinpo, September 30 I 898 , as cited in NKJ, III, 242-4-3; Yamagata Aritomo Den, III, 354-57. 41 Uyehara, . p. l49. 42 Ibid., p. 159. ^3 Harold S. Quigley, Japanese Government and P o l i t i c s : An Introduct- ory Study (London: The Century Co., 1932), p. 86. 268 CHAPTER 3 1 Quoted in Peter Duus, Party Rivalry and P o l i t i c a l Change in TaishS Japan (Cam. Mass.: Harvard Univ. Press, 1968), p. 101. 2 Charles Burton Fahs, "The Japanese House of Peers," Diss. Worth-western Univ. 1933, P- 15^. 3 Hirota Naoe, Naikaku Kotetsu Gojunenshi (Tokyo: Shun'yod-O, 1930), p. 443; Ito Hirobumi Den, III, 468-77; Akita, pp. 137 f f . 4 Hamada Kenji, Prince Ito (Tokyo: Sanseido Co. Ltd., 1936), p. 128; It5 Hirobumi Den, III, 494-95, 503-512•; Nihon Shinbun, February 26 and 28, March 8, 12., 15, and l6 1901, as cited in NKJ, IV, 22-24, 27, 30-31; Tokyo Asahi Shinbun, April 17 and May 3 1901, as cited in NKJ, IV, 35-36; Hirota, p. 453-5 Tokyo Asahi Shinbun, May 21 1901, as cited in NKJ, IV, 4 l ; Hirota, pp. 461-69; J i j i Shinpo, May 23 1901, as cited in NKJ, IV, 41; Takekoshi Yosaburo, Prince Saionji, trans. Nariaki Kozaki (Kyoto: Ritsumeikan Univ., 1936), p.. 182-84. 6 Tokyo Asahi Shinbun, October 28 1902 and December 10 1902, as cited in NKJ, IV, 159, 162; It5 Hirobumi Den, III, 574 f f . 7 Uyehara, p. 252, McLaren; P o l i t i c a l History, p. 285. 8 Tokyo Asahi Shinbun, July 4 1903, as cited in NKJ, IV, 185; Takekoshi, pp. 203-04; Ito Hirobumi Den, III, 589-94. 9 Quoted in Hackett, p. 222. 10 Najita, p. 29; See also McLaren, P o l i t i c a l History, pp. 300-04; Hirota, pp. 506, 511-12; Najita, pp. 28-31. 11 Kokumin Shinbun, January 15 1908, as cited in NKJ, V, 47-49, Hackett, pp. 239^h~0. 269 12 Najita, pp. 46-56. 13 Tokyo Nichi Nichi Shinbun, July 5 1908, as cited in NKJ, V, 58; Ito Hirobumi- Den, III, 794-95; Takekoshi, p. 238. 14 Ito Hirobumi Den, III, 795-96j Tokyo Nichi Nichi Shinbun, July- 10 1908, as cited in NKJ, V, 59; Najita, pp. 72-73. 15 McLaren, P o l i t i c a l History, p. 336-38. 16 Hirota, pp. 599-603; Kokumin Shinbun, January 19 1911, as cited in NKJ, V, 558-59-17 Takekoshi, p. 251; Tokyo Asahi Shinbun, August 1, l6 and 26 1911, as cited in NKJ, V, 568-69; Hirota, pp. 602-03. 18 Najita, pp. 72-86; Takekoshi, p. 251; Hackett, p. 246. 19 Hirota, p. 635; Najita, p. 96; Hackett, pp. 251-53-20 Hirota, pp. 624-37; Najita, pp. 87-98. 21 Koshakuu Katsura Taro Den, as cited in NKJ, V, 907-09; Tokyo Asahi  Shinbun, December 4 and 6 1912, as cited in NKJ, V, 908-09; Hackett, pp. 251-561. 22 Hirota, pp. 659-668; Tokyo Asahi Shinbun, December 11 1912, as cited in NKJ, V, 910; Hackett, pp. 255-56; Najita, pp. 115-117. 23 J i j i Shinpo, December 22 1912, as cited in NKJ, VI, 6; Hackett, p. 257. 24 Najita, pp. 119-4-1. 270 25 Hackett, pp. 258-62; Takekoshi, pp. 273-78; Najita, pp. 144-57; Hirota, pp. 680-8l. 26 Najita, p. 154. 27 Najita, pp. 157-162; Tokyo Asahi Shinbun, February 6, 7, 10, and 13 1913, as cited in NKJ, VI, 22, 26-29, 32; Hara Kei Nikki, February 8, 9, 10, and 11 1913, V, 186-93. 28 Najita, p. 167. 29 Tokyo Asahi Shinbun, February 13, 18 and 20 1913, as cited in NKJ, VI, 33, 35; Hirota, pp. 683-84, 692-93; Najita, pp. 164-75. 30 Tokyo Nichi Nichi Shinbun, March 27 and 30 1914, as cited in NKJ, VI, 122-23; Tokyo Asahi Shinbun, April 1, 9, and 12 19l4, as cited in NKJ, VI, 124, 126, 129-30; Lebra, pp. 218-39; Najita, pp. 190-93-Hackett, pp. 269-96; Duus, pp. 93-95; Najita, pp. 196-97. 32 Hirota, pp. 714-22; Hackett, pp. 302-04; Tokyo Asahi Shinbun, October 2 19l6, as cited in NKJ, VI, 423; Hara Kei Nikki, July 26 and 27 and August 4, 9, 10, and l6 19l6, VI, 456-57, UoTP f l . Tokyo Asahi Shinbun, October 5 and 8 19l6, as cited in NKJ, VI, 423, 425-26; Hirota, pp. 725-28; Hackett, pp. 306-07; Hara Kei Nikki, October 1, 3, k, 5, 6, and 7 1916", VII, 18-30. 3k Hackett, p. 307; Duus, pp. 96-97* According to Hackett, the phrase used by Terauchi was heikS j i h e i , but I have not been able to verify this. Terauchi's biographer quotes the Prime Minister saying chusei fuhen (^ji.vp./^) • See Kuroda KSshiro, Gensui Terauchi Hakushaku Den (Tokyo: Gensui Terauchi Hakushaku Denki Hensanjo, 1920), p. 824. 35 Kokumin Shinbun, August 15 and September 22 1918, as cited in NKJ, VI, 705; Hackett, pp. 312-17; Hirota, p. 737; Duus, p. 101. 36 Hirota, p. 443- See Appendix III for the institutional qualifications of a l l ministers in 1900-1918. 271 37 Hirota, p. 668. 38 For example, Hackett, pp. 209-10. 39 See note 4 above. ho Fans, p. lV7. 4 l Ibid., p. Ikj. See also pp. 126-27. 42 Najita, p. 54 and p. 2kk, note 62; Takekoshi, p. 236. 43 Hirota, pp. 626-27. kk Ibid., p. 666. 45 Kanpo Gogai, December 21 1912, as cited in NKJ, VI, 5. 46 Hara Kei Nikki, February 10 1913, V, 189-92. 47 Scott Flanagan, "Crises in the P o l i t i c a l Development of Modern Japan, 1880-1945: A Study of Coalition Behaviour and P o l i t i c a l Change," Diss. Stanford Univ. 1972, pp. 223-24. 48 Hackett, pp. 251-53. 49 Robert M. Spaulding, Imperial Japan's Higher C i v i l Service Examinations (Princeton: Princeton Univ. Press, 1967), pp. 111-26, 366-68. 50 Najita, p. 38. See also pp. 34-45. 272 51 Harvey C. Mansfield, " P o l i t i c a l Parties, Patronage, and the Federal Government Service," i n Wallace S. Sayre ed., The American Assembly: The  Federal Government Service, 2nd ed. (Englewood C l i f f s : Prentice Hall, Inc., 1965). 52 See Richard E. Neustadt, "Politicians and Bureaucrats," in David B. Truman ed., The American Assembly: Congress and America's 'Future (Englewood C l i f f s : Prentice Hall, 1965); and Richard E. Neustadt, "White House and Whitehall," in Anthony King, ed., The British Prime Minister (London: Macmillan and Co. Ltd., I969). 53 Najita, p. 178. See also p. 177. 5 k Hirota, pp. 444, 452. 55 Hackett, pp. 212-13. 56 Najita, p. 10. 57 Ibid., p. 6. 58 Duus, p. 84. 59. 5tsu Jun'ichiro, Dai Nihon Kenseishi, VII, as cited i n NKJ, VI, 18. 60 Richard Neustadt, Presidential Power: The Politics of Leadership (New York: John Wiley and Sons, Inc., i 9 6 0 ) , pp. 107, l 86T~ 61 See Philip Williams, Crisis and Compromise: Politics" in the Fourth  Republic (London: Lonmans, Green and Co. Ltd., 1958). 62 Duus, p. 97. 63 Uyehara, p. 254. See also Hirota, p. 524. 273 6h Duus, pp. 19-21. 65 Uyehara, pp. 2kr(-k&. See also p. 260. 66 McLaren, P o l i t i c a l History, p. 283. 67 Duus, p. 25. 68 Ibid., p. 26. 69 Ibid., p. 27. 70 Najita, p. 157. 71 Duus, p. 2kk. 27 k CHAPTER 4 1 Scalapino, pp. 393 f f . 2 Duus, pp. 122-25. 3 Maeda Renzan, Rekidai Naikaku Monogatari (Tokyo, J i j i Tsushinsha, 196l), II, 206-21; Hackett, pp. 318-22; Duus, pp. 102-04; Hara Kei Nikki, September 20, 22, 23, 25 19l8, VIII, 11-30. 4 Duus, pp. 136-42. 5 Matsumoto Gokichi S e i j i Nisshi, eds. Oka Toshitake and Hayashi Shigeru (Tokyo: Iwanami Shoten, 1959), P« 126. 6 Tokyo Asahi Shinbun, November 13 1921, as cited in NKJ, VII, 87-88; Maeda, II, 308-14. 7 Maeda, II, 331-56; Shinobu Seisaburo, Taisho Seijishi (Tokyo, 1952), TV, 1085-86; Tokyo Asahi Shinbun, June 6 and 7 1922, as cited in NKJ, VII, 504-06. 8 Shinobu, IV, 1088. 9 Matsumoto S e i j i Nisshi, p. 184. See also Maeda, II, 358-68; Shinobu, TV, 1088-91; Matsumoto S e i j i Nisshi, pp. I76-87; Duus, pp. 167-68, 270-71 note 10; Tokyo Asahi Shinbun, June 12 1922, as cited in NKJ, VII, 506-07. 10 Shinobu, TV, 1104. 11 Matsumoto S e i j i Nisshi, p. 255' 12 Maeda, II, 382-87; Shinobu, IV, 1108-11; Matsumoto S e i j i Nisshi, pp. 250-60; Duus, pp. 168-69; Tokyo Asahi Shinbun, August 28 and 29 1923, as cited in NKJ, VII, 606-O7. 275 !3 Maeda II, 395-96; Matsumoto, p. 288; Otsu Jun'Ichiro, Dai' Nihon  Kenseishi, IX, as cited in NKJ, VII, 735-37; Tokyo Asahi' Shinbun, December 23, 28 and 30 1923, as cited i n NKJ, VII, 738-39. 14 Tokyo Asahi Shinbun, January 2 , 4 and 5 1924, as cited i n NKJ, VII, 740-41; Maeda, II, 398-411; Matsumoto S e i j i Nisshi', pp. 288^59; Tokonami Takejir5 Den, as cited in NKJ, VIII, 7. 15 Tokyo Asahi Shinbun, January 21 and June 8 1924, as cited in NKJ, VIII, 14, 29-30; Maeda, II, 4 l 3 - l 8 ; Shinobu, TV, 1144-47. 16 Shinobu, IV, 1147. 17 Matsumoto S e i j i Nisshi, pp. 321-22; Shinobu, IV, 1147-49; Maeda, II, 420-27; Tokyo Asahi Shinbun, June 10 1924, as cited in NKJ, VIII, 30. 18 Asahi Shinbun S e i j i Keizai Bu, ed., S e i j i Keizai-Sosho - 5 - Sumitsuin  Mondai (Tokyo: Asahi Shinbunsha, 1930), pp. 55-81, hereafter cited as Sumitsuin Mondai. 19 Tokyo Asahi Shinbun, January 21 1927, as cited in NKJ, VIII, 391-92; Kenneth Colegrove, 'The Japanese Privy Council," in American P o l i t i c a l  Science Review, 25, No. 4 (Nov. 1931), 885-88; Duus, pp. 227-30. 20 Harada Kumao, Saionji Ko to Seikyoku (Tokyo: Iwanami Shoten, 1950), I, 3-13; Arthur E. Tiedemann, "The Hamaguchi Cabinet, First Phase July 1929-February 1930: A Study in Japanese Parliamentary Government," Diss. Columbia Univ. 1 9 5 k , PP. 55-65; Tokyo Asahi Shinbun, May 23, 24, 25, 26, and June 29 1929, as cited in NKJ, VIII, 580-82, 609. 21 Tokyo Asahi Shinbun, November 22 and December 11 and 12 1931, as cited in NKJ, IX, 258-59, 2 6 l , 263; Harada, II, l 6 - l 8 , 66-87, 102-03, 111-17, 153-57. 22 Matsumoto S e i j i Nisshi, p. 567. 23 Hara Kei Nikki, May 13, l 4 , and 15 1920, VIII, 5 k 3-^6. 276 24 Maeda, II, 4 2 7 . 25 Kenneth Colegorve, "Parliamentary Government in Japan," American  P o l i t i c a l Science Review, 21, Wo. 4 (Nov. 1927), 846. 26 Tiedmann, p. 76. 27 Shinobu, IV, 1091. 28 Fahs, p. 173. 29 See note 13 above. 30 Yoshihara Toshi, Ryunin Mondai to Shinsetsuron (Tokyo: Hisano Shoten, 1932), pp. 8-10. 31 Duus, pp. 196 f f . 32 Redman, p. 4l7. 33 Matsumoto S e i j i Nisshi, p. 544. 3k Harada, II, 121 f f . , 130, I38, 151-53, 158-59, 162, 164-67, 204, 220, 224; Tokyo Asahi Shinbun, May 17 1932, as cited i n NKJ, IX, 372-77. 35 Harada II, 291, 293• See also note 34 above. 36 Colegrove, "The Japanese Privy Council," APSR, Nov. 1931, pp. 882-83. 37 Duus, pp. 200, 205. 38 Ibid., pp. 231-35. 277 39 Sumitsuin Mondai, p. 101. 4o Ibid., pp. 102-107. See also Mayer-Oakes for a detailed account of the entire a f f a i r . 4 i Robert Spaulding, "The Bureaucracy as a P o l i t i c a l Force," in James Morley ed., Dilemmas of Growth in Prewar Japan (Princeton: Princeton Univ. Press, 1971), p. 66. k2 See Appendix TV for the institutional qualifications of a l l ministers in 1918-1932. h3 Maeda, II, 220, 225. 44 Matsumoto S e i j i Nisshi, p. 290. k5 Maeda, II, 224-25. 46 Spaulding, "The Bureaucracy as a P o l i t i c a l Force," p. 53* 4 7 Duus, p. 104. 48 Scalapino, pp. 393 f f . 278 CHAPTER 5 1 The general information in this section comes from: William Lockwood ed., The State and Economic Enterprise in Japan: Essays in' the P o l i t i c a l  Economy of Growth (Princeton: Princeton Univ. Press, 19637; James Morley ed., Dilemmas of Growth in Prewar';'Japan (Princeton Univ. Press, 1971); G.C. Allen, A Short Economic History of Modern Japan (London: George Allen and Unwin, 1962); and George 0. Totten, The Social Democratic Movement in Prewar Japan (New Haven: Yale Univ. Press, I966) . 2 Sir Ivor Jennings, The Queen's Government (Penguin Books, 195 k)> pp. 118-19. • 3 Yamazaki Tansho, Naikaku Seido no Kenkyu (Tokyo: Takayama Shoin, 1 9 k 2 ) , p. 375-h Ogata Taketora, "Behind Japan's Greater Cabinet," Contemporary  Japan, 6, No. 3 (Dec. 1937), 383. 5 Spaulding, "The Bureaucracy as a P o l i t i c a l Force," p. 57-6 Ogata, "Behind Japan's Greater Cabinet," p. 387. 7 Harada II, 311, VII, 203. 8 Iwabuchi Tatsuo,"Prince Fuminaro Konoye," Contemporary Japan, 5, No. 3 (Dec. 1936), 369. 9 David Anson Titus, "Palace Polit i c s in Prewar Japan: The Context of Imperial Involvement in Politics and Palace Leadership i n the Showa period, 1929 -19 kl," Diss. Columbia Univ. 1970, p. 239. 10 Kyokuto Kokusai Gunji Saiban Kenkyukai ed., Kido Nikki (Tokyo: Heiwa Shobo, 1 9 k 7 ) , PP- 22-25, hereafter cited as Kido Nikki; Gordon Mark Berger, "The Search for a New P o l i t i c a l Order: Konoe FuminarS, the P o l i t i c a l Parties, and Japanese Polit i c s during the Early Showa Era," Diss. Yale Univ. 1972, pp. 86-87; Harada, II, 2 5 3 - 5 k , 285-298; Berger, pp. 86-87. 279 11 Iwabuchi, "Prince Fuminaro Konoye," p. 370. 12 M i l l e r , p. 176. Kojima Se i ich i , Okada Naikaku to Senkyuhyakusanjugonen- (Tokyo: Sen'kura Shobo, 193 k), p. 8; Baba Tsunego, "Problems of the Okada Cabinet," Contemporary Japan, 3, Wo. 2 (Sept. 193 k), 217; Harada, I I I , 328-36, 2 k7- k8, TV, 5-10; Tokyo Asahi Shinbun, July 6 and 7 193k, as cited i n NKJ, IX, 535. Ik Iwabuchi, "Prince Fuminaro Konoye," p. 370. 15 Harada, V, 1 3 - l k . 16 Baba Tsunego, "Hirota's 'Renovation' Plans," Contemporary Japan, 5, No. 2 (Sept. 1936), 167-72; Harada, V, 13-19; Kido Nikki , 34-36; Tokyo Asahi Shinbun, March 6 and 7 1936, as cited in NKJ, X, 79, 8 l . 17 Baba Tsunego, "The Anti-Comintern Pact i n Domestic P o l i t i c s , " Contemporary Japan, 5, No. h (March 1937), 5k3« 18 Harada, V, 210-11, 2 k2- k3; Berger, pp. l k l - 5 k . 19 Berger, pp. 151-15k; Harada, V, 227-33, 2 kl-251; Gikai Seido, IX, 5k6-4-7; Tokyo Asahi Shinbun, January 30 and February 8 1937, as cited in NKJ, X, 286, 381. V o l . IX of Gikai Seido i s v o l . II of Teikoku Gikaishi. 20 Royama Masamichi, "New National Government," Contemporary Japan, 7, No. 2.(Sept. 1938), 222. 21 Berger, p. 130. 22 Gikai Seido, IX, 58k.. 280 23 Berger, pp. 129-36, 158-59, 171-72; Harada, V, 313-23; Ogata, "Behind Japan's Greater Cabinet," pp. 385-87; Tokyo Asahi Shinbun, June 2 1937, as cited in NKJ, X, 394. 24 Yoshioka Bunroku, "Reshuffling the Wartime Cabinet," Contemporary  Japan, 8, Wo. 1 (March 1939), 35-25 Harada, IV, 201, 210; Scalapino, p. 360. 26 Baba Tsunego, "War and Parliamentary P o l i t i c s , " Contemporary Japan, 8, No. 2 (April 1939), 185. 27 Harada, VII, 246-259-28 Iwabuchi Tatsuo, "Japan's New Premier, General Nobuyuki Abe," Contemporary Japan, 8, No. 8 (Oct. 1939), 959-29 Harada, VIII, 55-67; Iwabuchi, "Japan's New Premier," p. 959. 30 Harada, VIII, pp. 63, 75. 31 Baba Tsunego, "Stage-Setting in Japanese P o l i t i c s , " Contemporary  Japan, 9, No. 2 (Feb. 1940), 138, 139-32 Harada, VII, 311, 332, 338, 353 f f . , VIII, l4l-42, 157-73; Baba Tsunego, "Stage-Setting in Japanese P o l i t i c s , " pp. 131-39-33 Quoted in Baba Tsunego, "The P o l i t i c a l Outlook," Contemporary  Japan, 4, No. 4 (March 1936), 515. 34 Berger, pp. 301-02, 323-35 Ibid., pp. 155, 306. 281 36 Ibid., pp. 322, 133, 3^ 2, 483. 37 Quoted in Berger, p. 382. 38 Harada, VIII, 290-300; Kido Nikki, p. 65; Hosokawa Takamoto, "Exit the Yonai Cabinet," Contemporary Japan, 9, No. 8 (Aug. 1940), 979-83; Berger, pp. 347-54, 379-93, 424; Gikai Seido, IX, 631, 653. 39 Harada, II, 311, 325, 336, 368, 4l6, III, 53, 95-96, 78, 198.199, 209-10; Gikai Seido, IX, 425-26, 444-45, 447, ^63; Tokyo Asahi Shinbun, May 19 and 20 1934, as cited i n NKJ, IX, 528-529. 40 Yamazaki, p. 358-41 Baba Tsunego, "On the Eve of the General Election," Contemporary  Japan, 4, No. 2 (Sept. 1935), 196. 42 Baba Tsunego, "The P o l i t i c a l Outlook," p. 518. 3^ Yale Candee Maxon, Control of Japanese Foreign Policy: A Study of  Civil-Military Rivalry, 1930-1945~[Berkeley: Univ. of California Press, 1957), P- 108; Harada, XV, 72-75, 290-91, 319, 395-96, 4-08-11; Kido  Nikki, pp. 33-34; Gikai Seido, IX, 457-61, 496. 44 Harada, V, 71, 119, l6l, 194, 209, 217-18, 221, 233-^ 0; Berger, 140-49; Gikai Seido, IX, 545-46, 557-58; Baba Tsunego, "The Anti-Comintern Pact i n Domestic P o l i t i c s , " pp. 336-39; Tokyo Asahi Shinbun, January 22, 23, and 24 1937, as cited in NKJ, X, 273-77-k5 Yamazaki, p. 339-46 Harada, V, 252, 264-68, 286, 292, 295, 299, 303-08; Gikai Seido, IX, 547-48, 583; Baba Tsunego, "Konoye and Recent P o l i t i c a l Changes," Contemporary Japan, 6, No. 2 (Sept. 1937), 221-22; Tokyo Asahi Shinbun, March 31, April 11, May 3, June 1 1937, as cited in NKJ, X, 387-88, 393-hi Yamazaki, p. 353-282 48 Harada, VI, 47, 6 l , 66, 137-38, 142-43, 14-7. 49 Harada, VI, 34-35, 47, 6 l , 137-38, 142-43, 252, 272, 274, 312, VII, 36, 41-42, 105, 129, 141-42, 195; Kido Nikki, pp. 39-53; Berger, 196, 233; Gikai Seido, " IX, 63O-3I, 653; Tokyo Asahi Shinbun, January 5 1939, as cited in NKJ, X, 496-97. 50 Baba Tsunego, "The New Cabinet's Foreign Policy," Contemporary  Japan, 8, No. 8 (Oct. 1939), 947-'Harada, VII, 17, 22, 268-70, 332-39, 347-55, 375-83, VIII, 48-54. 52 Berger, p. 323. See also Baba Tsunego, "Stage-Setting i n Japanese P o l i t i c s , " pp. 137-38; Harada, VIII, 64, 68, 81-82, 110, 115, 117, l42, 146-47, 150-51; Berger, pp. 319-330; Gikai Seido, IX,.I707. / Maxon, pp. lUl-48; Harada, VIII, 278, 282-90; Berger, pp. 282-83; Kido Nikki, p. 65. 54 See page 34 above. Baba Tsunego, "Konoe and Recent P o l i t i c a l Changes," p. 2l6. 56 Harada, VII, l 8 l . Takasugi Kojiro, "The Diet under the New P o l i t i c a l Structure," Contemporary Japan, 9, No. 11 (Nov. 194o), l4o8. 58 Spaulding, "The Bureaucracy as a P o l i t i c a l Force," pp. 56 f f . 59 Ibid., pp. 65-69. 60 Harada, VIII, 96-IOO. 283 6 l Spaulding, "The Bureaucracy as a P o l i t i c a l Force," pp. 73-75 j Tokyo  Asahi Shinbun, November 10 1936, as cited in NKJ, X, 2 6 k - 6 5 . 62 Harada, VIII, 298. 63 Ibid., p. 165. 6k Ibid., p. 63. 65 Harada, V, 100. 66 Harada, VI, 209. 67 Harada, II, 29-30. 68 Berger, p. 3 k 5 ' 69 Titus, p. Ik. 70 Ibid., p. 74-. 71 Ibid., p. 382. 72 Harada, II, 253, 287-91. Baba Tsunego, "Problems of the Okada Cabinet," pp. 217-18; Harada III, 3 k 7 - k 8 . 7 k Harada, V, 1 3 - l k ; Baba Tsunego, "Konoe and Recent P o l i t i c a l Changes," p. 217. 284 75 Harada, V, 241-46. 76 Yamazaki, pp. 3l6-l8; Baba Tsunego, "Konoe and Recent P o l i t i c a l Changes," pp. 217-18. 77 Charles Nelson Spinks, "Japan's New Genro," Contemporary Japan, 9, No. 7 (July 1940), 844. 78 Harada, VII, 246 f f . , VIII, 55-56, 60-61. 79 Yamazaki, pp. 319-23; Harada, VIII, 159-64; Kido Nikki, p. 65. 80 Titus, p. 150. 81 Ibid., p. 151. 82 Ibid., p. 195. 83 Ibid., pp. 216-17. 84 Ibid., p. 242. 85 Ibid., p. 243. 86 Ibid., p. 377. 87 Kenneth Colegrove, "The Japanese Privy Council," APSR, 25, No. 4 (Nov. 1931), 889, note 2. 88_ Otsu Jun'ichiro, Dai Nihon Kenseishi, VIII, as cited in NKJ, VI, 695; Tokyo Asahi Shinbun, June 3 1917, as cited in NKJ, VI, 694-95. 285 89 Harada, IV, 236; Gikai Seido, TX;:495-96\ 90 Ogata Taketora, "Behind Japan's Greater Cabinet," p. 380. 91 Ib id . , p. 386. See also Harada, VI , 115, VII I , 183. 92 Maxon, pp. 124-25, 150-56; Harada, VII , 57-60. 93 Harada, I I I , 175, VI , 201. Harada, IV, 395. 95 Harada, VI , 215, VIII , 142-44, 151-52. 96 Berger, p. 103. 97 Harada, VII I , 88-89. 98 Harada, IV, 391-92, 4o8. 99 Harada, VII I , 275. See also pp. 269-70. 100 Berger, p. 304. 101 IIbid,,ppp.3358 r59. 102 Ib id . , pp. 422-24. 103 Ib id . , pp. 457-66. 104 Ib id . , pp. 486-87. 286 BIBLIOGRAPHY APPENDICES GLOSSARY BIBLIOGRAPHY Japanese Sources Asahi Shinbun Seijikeizaibu, ed. Seijikeizai S5sho -5- Sumitsuin  Mondai (Library of Politics and Economics -5- Problems of the Privy Council). Tokyo: Asahi Shinbunsha, 1930. Date Genichiro. Nihon Ke