Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Japanese direct foreign investment : sources and sustainability Price, Tracy J.E. 1989-12-31

You don't seem to have a PDF reader installed, try download the pdf

Item Metadata

Download

Media
[if-you-see-this-DO-NOT-CLICK]
UBC_1989_A4_6 P74.pdf [ 12.96MB ]
[if-you-see-this-DO-NOT-CLICK]
Metadata
JSON: 1.0097354.json
JSON-LD: 1.0097354+ld.json
RDF/XML (Pretty): 1.0097354.xml
RDF/JSON: 1.0097354+rdf.json
Turtle: 1.0097354+rdf-turtle.txt
N-Triples: 1.0097354+rdf-ntriples.txt
Original Record: 1.0097354 +original-record.json
Full Text
1.0097354.txt
Citation
1.0097354.ris

Full Text

JAPANESE DIRECT FOREIGN INVESTMENT: SOURCES AND SUSTAINABILITY by TRACY J.E. PRICE B.A., UNIVERSITY OF VICTORIA, 1984 A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTERS OF SCIENCE in THE FACULTY OF GRADUATE STUDIES Faculty of Commerce and Business Administration We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA August 1989 @ Tracy J.E. Price, 1989 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department of Cornr^exe-cg Piroo SusirQe^-s «oro\tvv.S"re-ftTi or\J The University of British Columbia Vancouver, Canada Date S~>HPT S /i^'sq  DE-6 (2/88) ABSTRACT In 1984, Japan became the world's leading creditor nation. Although most of this capital has been in the form of foreign portfolio investment (FPI), foreign direct investment (FDI) has contributed significantly to the total. The rapid acceleration of Japanese FDI is evidenced by the fact that the nation's accumulated foreign direct assets in 1988 exceeded $96 billion (in 1980 U.S. dollars) roughly 3 times the 1984 total and 6 times the 1980 total. This startling change in Japan's role in global FDI raises two important questions. Firstly, why has Japan been able to substantially accelerate its foreign direct investments in the 1980s? This study suggests that the country's rapid expansion in FDI is the result of macro-economic developments which have taken place in Japan since 1973. These developments include the transition of the country from a high-growth to slow-growth economy after the first oil crisis; the resultant decline in capital formation requirements and sustained savings surpluses in the private sector; and the committment of the Liberal Democratic Party (LDP) to fiscal austerity after 1978. The second question concerns the future sustainability of Japanese FDI. As described in this study, the answer to the second ii question depends largely on the answer to the first; namely, that the future rate of Japanese FDI will be determined by the extent to which the macro-economic developments noted above prevail. In format, this study first provides a historical perspective of Japanese offshore direct investment, concentrating on the changing level, nature and motivation of Japanese FDI in the post World War II period. The study then provides a brief analysis of modern FDI theory and its inadequacy for explaining past Japanese FDI or for predicting its future sustainability. After introducing the theoretical rationale behind the study's two main contentions, the analysis then turns to an identification of the forces which are generating the huge amounts of capital currently available for offshore investment. In particular, the study suggests that the dramatic appreciation of Japanese land prices has been a primary cause of excess savings in the private sector. It is these excess savings, coupled with the LDP's committment to balanced budgets after 1978, that have sponsored Japan's remarkable increase in FDI in the 1980s. iii The study also examines the results of a regression model developed to test the above macro-economic contentions. The model provides some evidence to support the macro-economic rationale used in the study but is limited by statistical problems with the data. Finally, the study examines the issue of sustainability and concludes that, based on likely macro-economic developments in the medium term, Japanese FDI will be sustained at its present high level. iv TABLE Of CONTENTS 1.0. INTRODUCTION 1 1.1. THESIS STATEMENT AND FORMAT 1 1.2. THE CURRENT STATUS 3 2.0 DEFINITIONS AND DATA SOURCES 5 2.1. DEFINITIONS OF FOREIGN INVESTMENT 5 2.2. SOURCES OF STATISTICAL DATA 8 3.0. HISTORICAL REVIEW OF JAPANESE FOREIGN DIRECT INVESTMENT 11 3.1. JAPANESE FDI PRIOR TO WORLD WAR II 1(1870 to 1940) 3.2. JAPANESE FDI; 1951 to 1965 14 3.3. JAPANESE FDI; 1966 to 1973 7 3.4. JAPANESE FDI; 1974 to 1980 23.5. JAPANESE FDI; 1981 to 1984 33 3.6. JAPANESE FDI; 1985 TO THE PRESENT 38 4.0. TOWARDS A THEORY OF JAPANESE FOREIGN DIRECT INVESTMENT 52 4.1. CONVENTIONAL FDI THEORY 53 4.2. CLASSICAL THEORY 54 4.3. OFFENSIVE FDI THEORY - THE INDUSTRIAL ORGANIZATION APPROACH 6 4.3.1. Economies of Scale 57 4.3.2. The posession of unique skills or special  expertise 54.3.3. Differentiated Products 58 4.3.4. Internalization4.3.5 Applicability of Offensive FDI theory in the Japanese FDI Context 50 4.4. DEFENSIVE FDI THEORY 61 4.4.1. Market Imperfections Created by Governments 61 4.4.2. New Markets 2 4.4.3. Access to New Technologies 64.4.4. Risk Diversification 64.4.5. Product Cycle Theory 3 4.4.6. Follow the Leader 4 4.4.7. Applicability of Defensive FDI models in the  Japanese Context 64.5. A MODEL OF JAPANESE FDI 5 v 5.0. MACROECONOMIC BALANCES OF THE JAPANESE ECONOMY 67 6.0. INTERNAL MACROECONOMIC COMPONENTS OF THE JAPANESE 75 6.1. THE PRIVATE SECTOR BALANCE 86.1.1. Private Sector Savings 78 6.1.1.1. The Need to Provide for Retirement 80 6.1.1.2. The High Cost of Housing 83 6.1.2. Private Sector Investment 87 6.1.2.1. Rising Industrial and Commercial Land Prices 88 6.1.2.2. Rising Real Wages 89 6.1.2.3. Reduced Productivity Growth 90 6.1.2.4. Environmental Constraints6.1.2.5. The Impact of the Oil Crisis 91 6.1.2.6. The Investment Climate since 1973 93 6.1.3. The Private Sector Surplus since 1986 96 6.1.4. Implications of Private Sector Savings  Surpluses 99 6.2. THE PUBLIC SECTOR BALANCE 100 6.2.1. Government Expenditures after 1973 101 6.2.2. Government Revenues after 1973 103 6.2.3. The Pressure for Fiscal Austerity 105 6.2.4. The Government Balance since 1984 107 6.4. THE DOMESTIC SECTOR - SUMMARY 109 7.0. EXTERNAL MACROECONOMIC COMPONENTS OF THE JAPANESE ECONOMY 111 7.1. NET FOREIGN DIRECT INVESTMENT 117.2. NET FOREIGN PORTFOLIO INVESTMENT 113 7.3. NET SHORT TERM CAPITAL INVESTMENT 8 7.4. NET MONETARY MOVEMENTS 119 7.5. SUMMARY OF THE CAPITAL ACCOUNT BALANCE 118.0. A STATISTICAL MODEL OF JAPANESE DIRECT FOREIGN INVESTMENT 121 8.1. RATIONALE AND METHODOLOGY OF STATISTICAL APPROACH 123 vi 8.2. NET FOREIGN DIRECT INVESTMENT 125 8.2.1. The NFDI Model; Hypotheses and Rationale 126 8.2.2. The NFDI Model; Data Sources 129 8.2.4 Results of the NFDI Model 130 8.3. THE PRIVATE SECTOR BALANCE 131 8.3.1. (S-I) Model; Hypotheses and Rationale 132 8.3.2. (S-I) Model; Data Sources 135 8.3.3. (S-I) Model; Results 138.4. NET FOREIGN PORTFOLIO INVESTMENT 138 8.4.1. NFPI Model; Hypotheses and Rationale 139 8.4.2. NFPI Model; Data Sources 141 8.4.3. NFPI Model; Results 142 8.5. LIMITATIONS OF THE NFDI RESULTS 143 8.5.1. Multicollinearity among the Explanatory  Variables 148.5.2. Auto-Correlation of the Residuals 144 8.5.3. The Problem of Endogeneity 148.5.4. Structural Change 148.5.5. Problems in the Data 6 8.6. SUMMARY 147 9.0. JAPANESE FDI; THE FUTURE 148 9.1. AVAILABILITY OF FUNDS9.1.1. Private Sector Balance 149 9.1.1.1. Household Savings 149.1.1.2. Private Residential Investment. 152 9.1.1.3. Corporate Savings 153 9.1.1.4. Corporate Investment 154 9.1.1.5. Summary of Private Sector Balance 155 9.1.2. Government Sector 156 9.1.2.1. Expenditures9.1.2.2. Income 159.1.2.3. Summary of Government Sector Balance 158 9.1.3. Summary of Macroeconomic Balance 158 vii 9.1.4. Other Factors 159 9.1.4.1. Real Estate Market 159.1.4.2. Stock Market Earnings 160 9.1.4.3. Off-Shore Earnings 169.2. THE REQUIREMENTS OF THE CURRENT ACCOUNT BALANCE 161 9.3. THE ALLOCATION PRIORITIES OF FOREIGN INVESTMENT FUNDS 162 9.3.1. The Allocation of Funds to Direct Foreign  Investment 163 9.3.1.1. Comparative Returns 164 9.3.1.2. Comparative Production Costs 165 9.3.1.3. Trade Frictions and Barriers to Market entry 166 9.3.1.4. Procurement of Resources 167 9.3.1.5. Longterm Strategic Plans 8 9.3.2. Allocation of Funds to Foreign Portfolio  Investment 170 9.3.3. Allocations in a Competing Environment 171 9.4. ACCEPTABILITY 172 10.0. CONCLUSIONS 7 BIBLIOGRAPHY 179 APPENDICES Appendix 1 Data Used in Regression Models 188 Appendix 2 Results of Regression Models 189 viii LIST OF TABLES SECTION 1: 1.1 Japanese Net Foreign Investment, F.Y. 1960-1988 SECTION 3; 3.1 Number of Overseas Offices of Selected Sogo Shosha Prior to WWII 13 3.2 Approval of Investment Projects in East Asia, 1960-1974 19 3.3 Supply Sources of Selected Resources, 1969 21 3.4 Japanese FDI Classified by Industry, F.Y. 1965-F.Y. 1972 23.5 Japanese FDI by Region, F.Y. 1965-F.Y. 1973 22 3.6 Japanese FDI by Region, F.Y. 1974-F.Y. 1980 5 3.7 Japanese FDI Classified by Industry, F.Y. 1972-F.Y. 1980 25 3.8 Japanese FDI by Region, F.Y. 1981-F.Y. 1984 31 3.9 Japanese FDI Classifed by Industry, F.Y. 1984-F.Y. 1987 34 3.10 Japanese FDI by Region, F.Y. 1984-F.Y. 1987 35 SECTION 6; 6.1 Japanese Private Sector Savings-Investment Balance, F.Y. 1960-F.Y. 1987 67 IX Following Page SECTION 6 CONT'D Following Page 6.2 Inflation Results and Time Deposit Rates by Type of Institution, 1969 to 1981 70 6.3 Life Expectancy and Labour Participation Ratios of Selected Countries 72 6.4 Residential Land Price Indices, Japan, F.Y. 1960-F.Y. 1987 74 6.5 Average Prices of Residential Land of Six Largest Cities6.6 Commercial Land Price Indices, Japan, F.Y. 1960-F.Y. 1987 77 6.7 Industrial Land Price Indices, Japan, F.Y. 1960-F.Y. 1987 8 6.8 Real Wage Indices and Compound Annual Growth Rates, Japan, F.Y. 1960-F.Y. 1987 79 6.9 Average Prices of Commercial Land of Six Largest Cities 83 6.10 Average Prices of Industrial Land of Six Largest Cities6.11 The Private and Public Sector Balance, F.Y. 1960-F.Y. 1987 87 6.12 Budgeted Government Spending by Category, Real Percentage Change over Previous Fiscal Year, F.Y. 1971-F.Y. 1986 89 6.13 Budgeted Government Spending as a Percentage of GNP, F.Y. 1970-F.Y. 1986 90 6.14 Growth of Budgeted Government Spending, by Category, F.Y. 1970-F.Y. 1986 90 6.15 Percentage Real Increase in Tax Burden, F.Y. 1961-F.Y. 1987 94 6.16 The Domestic Sector and Current Account Balance, F.Y. 1960-F.Y. 1987 94 x SECTION 7: 7.1 Japanese Net Portfolio Investment, F.Y. F.Y. 1987 7.2 Investment in Foreign Securities by Japanese, F.Y. 1981-F.Y. 1988 7.3 Real Discount Rates, U.S. and Bank of Japan, Averages, F.Y. 1961-F.Y. 1988 7.4 Investment in Japanese Securites by Non-Japanese, F.Y. 1981-F.Y. 1988 7.5 Japan's Balance of Payments, F.Y. 1961-F.Y.1987 Following Page 100 100 102 103 104 SECTION 8: 8.1 Results of Net Foreign Direct Investment Model (1) 8.2 Results of Net Foreign Direct Investment Model (2) 8.3 Correlation Matrix for the Private Balance Regression 8.4 Results of Private Balance Regression (1) 8. 5 Results of Regressions of Individual Explanatory Variables on Private Balance 8.6 Results of Private Balance Regression (1) Corrected for First Order Auto-Correlation 8.7 Results of Regressions of Individual Explanatory Variables on Private Balance, Corrected for First Order Auto-Correlation 114 114 119 119 119 120 120 8.8 Results of Net Foreign Portfolio Investment Model 124 8.9 Results of Net Foreign Portfolio Investment Model, Corrected for First Order Auto-Correlation 124 8.10 Testing for Structural Change, 1960 to 1971 and 1972 to 1987 127 XI LI8T OF FIGURES Following Section 5; Page Figure 5.1. Balance of Payments Schematic 62 Section 8: Figure 8.1. Macro-Economic Rationale Behind Statistical Approach 109 xii ACKNOWLEGEMENT My deep thanks are extended to Dr. James Brander for his enthusiasm, support and encouragement throughout the course of this project. I would also like to thank Dr. Peter Nemetz and Dr. Elan Vertinsky for their valuable contributions to the paper and the time and effort they devoteded on my behalf. The two years which I have spent at U.B.C. completing this degree have been both worthwhile and informative; my thanks are extended to the administrative staff of the Department of Commerce and to all the professors whose courses I had the good fortune to take. Finally, I must acknowledge the encouragement and unfailing patience of my family. To my parents, who have given me constant support throughout this project and my life, and to my husband, whose friendship and sense of humour has never faltered, I extend both my gratitude and love. xiii 1.0. INTRODUCTION 1.1. THESIS STATEMENT AND FORMAT In each of the past 4 years, Japan has been the world's largest exporter of capital. Although most of this capital has been in the form of foreign portfolio investment (FPI), foreign direct investment (FDI) has contributed significantly to the total. The startling change in Japan's role in world FDI can be gauged by the fact that the country's accumulated direct foreign assets amounted to over $96 billion U.S. (in 1980 dollars 1) by the end of fiscal 1988; which was almost 3 times the 1984 figure of $33 billion U.S. and over 6 times the accumulated direct investment assets held in 1980. Both the size and the changed focus of Japan's FDI are likely to be of pivotal importance in shaping the world economic order of the 1990s. Any assessment of this impact must address two key questions: (1) What factors are producing the current excess of capital in the Japanese economy which is available for offshore investment? (2) Given (1) , is the current rate of Japanese foreign direct investment sustainable? The answers to these questions need to be extracted from a complex Unless otherwise stated, all figures referring to the level of cumulative and annual Japanese FDI flows will be in constant 1980 dollars. - Page l -matrix of economic parameters, disparate corporate cultures and overlying political considerations. Nevertheless, the general shape of the answers seem clear. The Japanese economy will be able to sustain high rates of foreign investment and a growing proportion of this investment will be directed towards asset acquisition. As a consequence of this increased involvement in off-shore ownership and operations, the major Japanese companies will develop a far greater degree of internationalism than has been the case hitherto. This later trend, with its concomitant strains on the traditional Japanese management system, will be a major test of the resilience and adaptability of the Japanese industrial society. This study seeks to explain the rapid increase in Japanese FDI in the context of macro-economic developments which have taken place in Japan since 1973. In format, this paper will first provide a historical perspective of Japanese offshore direct investment, concentrating on the changing levels, nature and motivation of Japanese FDI, particularly since 1973. In section 4, the study will provide a brief analysis of current FDI theories and their inadequacy for explaining past Japanese FDI or for predicting its future continuation. Section 5 will introduce the macro-economic basis for the study's contention that structural changes in the domestic economy are driving Japan's current FDI initiatives. The study will then identify the forces which are generating the huge amounts of capital currently available for - Page 2 -offshore investment and will seek to establish the contribution of each of these forces. In section 8.0, the results and limitations of a regression model developed to support the foregoing macro-economic discussion will be examined. Finally, section 9.0 will evaluate the long-term sustainability of the capital generating factors described in the macro-economic and statistical sections of this study. As will be discussed in section 9.0, the future levels of Japanese FDI will depend not only on the extent to which current macro-economic conditions prevail, but also on the desirability of continued Japanese offshore investment and its future acceptability in terms of the recipient countries. Before the supporting discussion of this thesis, it may be of value to provide a background brief of the current circumstances of Japanese foreign investment. 1.2. THE CURRENT STATUS As noted above, Japan became the world's leading creditor nation in 1984 and, spurred by aggressive portfolio, direct and, increasingly, foreign real estate investments, has maintained this position since. Although this study is focussed on Japan's FDI, it is of interest to take a brief look at the country's overall foreign investment practice. In fiscal year (F.Y.) 1988 (ending March 31, 1989), Japan's net overseas assets exceeded $260 billion - Page 3 -US having risen by a record $46 billion over F.Y. 1987 (2) . Total assets overseas increased by 37.1% to $1,321 billion US while total debts rose 41.4% to $1,059 billion US. Japan's total foreign investment in 1988 was about $109 billion U.S, of which $78.7 billion was in the portfolio classification, which resulted in an accumulated net FPI of over $293 billion U.S.(3). Between 1984 and 1988, the sum of the annual net portfolio investments amounted to $313 billion U.S.; the corresponding figure for the preceding 24 years was minus $20.2 billion U.S. Table 1.1 shows the pattern of Japan's foreign investment from 1960 to 1988 and clearly indicates the rapid expansion of both FPI and FDI since 1983. Although currency appreciation has over-stated this growth, it can also be seen that total foreign investment has increased from 0.5 percent of GNP in F.Y. 1983 to 4.2 percent of GNP in F.Y. 1988. 2 "Japan World's Leading Creditor," The Province. May 28, 1989. 3 Both foreign portfolio and foreign direct investment will be defined in Section 2.1. - Page 4 -TABLE S.i: JAPANESE NET TOTAL OVERSEAS INVESTMENT AS A X OF 6NP, FISCAL YEAR i960 TO 1987, BILLIONS OF U.S. t *1 BILLIONS OF U.S. $ TOTAL NET NET NET FQREI6N F0REI6N OVERSEAS X OF DIRECT X OF PORTFOLIO I OF YEAR 6NP INVESTMENT 6NP INVESTMENT 6NP INVESTMENT 6NP 1960 45.02 0.09 0.2 0.07 0.2 0.02 0.0 1961 55.15 0.02 0.0 0.03 0.1 -0.01 0.0 1962 60.17 -0.06 -0.1 0.01 0.0 -0.07 -0.1 1963 71.09 -0.13 -0.2 0.02 0.0 -0.15 -0.2 1964 82.39 -0.10 -0.1 0.05 0.1 -0.15 -0.2 1965 91.15 -0.05 -O.i 0.03 0.0 -0.08 -0.1 1966 106.81 0.10 0.1 0.08 0.1 0.02 0.0 1967 125.89 0.09 0.1 0.08 0.1 0.01 0.0 1968 148.24 -0.21 -0.1 0.14 0.1 -0.35 -0.2 1969 172.94 -0.79 -0.5 0.14 0.1 -0.93 -0.5 1970 202.92 0.00 0.0 0.25 0.1 -0.25 -0.1 1971 233.52 -0.66 -0.3 0.14 0.1 -0.80 -0.3 1972 312.26 0.46 0.1 0.54 0.2 -0.08 0.0 1973 396.55 3.66 0.9 1.94 0.5 1.72 0.4 1974 467.07 2.56 0.5 1.69 0.4 0.87 0.2 1975 377.11 -1.06 -0.3 1.53 0.4 -2.59 -0.7 1976 574.24 -0.87 -0.2 1.90 0.3 -2.77 -0.5 1977 703.15 0.99 0.1 1.63 0.2 -0.64 -0.1 1978 982.53 5.17 0.5 2.36 0.2 2.81 0.3 1979 1013.25 3.89 0.4 2.66 0.3 1.23 0.1 1980 1081.25 -7.32 -0.7 2.11 0.2 -9.43 -0.9 1981 1177.42 -2.96 -0.3 4.71 0.4 -7.67 -0.7 1982 1093.56 3.26 0.3 4.10 0.4 -0.84 -0.1 1983 1196.25 6.10 0.5 3.20 0.3 2.90 0.2 1984 1276.34 29.93 2.3 5.97 0.5 23.96 1.9 1985 1344.74 46.83 3.5 5.08 0.4 41.75 3.1 1986 1985.34 116.29 5.9 14.25 0.7 102.04 5.1 1987 2428.01 109.41 4.5 18.61 0.8 90.80 3.7 1988 2889.06 121.30 4.2 33.76 1.2 87.54 3.0 Footnote: *1 + signs refer to net outflows and - siqns refer to net inflows - Page 4a -2.0 DEFINITIONS AND DATA SOURCES The study makes extensive use of statistical data in describing and analyzing the subject of Japanese foreign investment. Before entering the main body of the discussion, it will be helpful to define the principal terms used with reference to foreign investment and to comment on the sources of the data used. 2.1. DEFINITIONS OF FOREIGN INVESTMENT From the balance of payments perspective, international capital movements which involve transactions with credit maturities of less than one year are referred to as short term movements, while transactions which have infinite maturities, or maturities exceeding one year (such as stocks and physical assets) are deemed as long term movements. Generally, short term capital movements serve to smooth out short term fluctuations in the balance of payments by responding swiftly (overnight) to changes in relative interest rates and exchange rates. Although important in the overall monetary sense, short term movements are, necessarily, of a transient nature and do not confer a lasting impact on either the recipient or donor economies. They have no part in the discussion at hand. On the other hand, long term capital movements, represent an extended transfer of capital to the recipient countries and are, therefore, more important in the context of long term economic - Page 5 -development. These long term capital movements can be further sub divided into indirect and direct assets. The former, also referred to as foreign portfolio investment (FPI), includes purchases of foreign equities, bonds and other similar instruments. In general, foreign portfolio investment does not confer any direct management rights on the foreign investor. Such investment is made purely on the basis of expected returns. Conversely, foreign direct investment (FDI) does assign management rights to the investor and allows the investor to influence, in part or in total, the conduct of the recipient organization's business. The line of demarcation between portfolio and direct investment is obviously an arbitrary judgement which is established more for the purposes of foreign investment accounting than as a true attempt to separate passive and active investment. For Japan, the demarcation point is 10 percent; i.e. the acquisition of 10 percent, or more, of a foreign company's equity is considered to be direct foreign investment. Below that level, the transaction is generally recorded as foreign portfolio investment. Japan's Ministry of Finance (MOF) uses a somewhat broader net to encompass FDI. In addition to the 10 percent rule, the MOF also includes investments in foreign corporations with which the donor has established a permanent, economic working relationship. Further, it classifies loans as FDI when such loans are made to companies in which the lender has 10 percent or more equity interest. - Page 6 -Obviously, if "direct" is intended to convey an ability to influence, the use of an arbitrary level of investment is not technically sound. Kojima's (4) description of direct foreign investment provides a more comprehensive definition: "...direct foreign investment should be understood as the transmission of management resources in a package of capital, management ability, and technical expertise to a host country. The management resources are organizations that exhibit various capabilities in the process of corporate management, consisting outwardly of the nucleus of managers but encompassing in a wider sense, managerial knowledge including patents, technical know-how and marketing techniques, market positions in regards to sales, materials procurement, and capital raising, trademarks and goodwill and organizations for information gathering and research and development." Notwithstanding its arbitrary nature, the use of statistics based on the MOF's definition of FDI is probably a reasonable basis for measuring Japanese overseas investment which has influence on the conduct of the recipient companies. In many cases, the Japanese participation is clearly dominant or exclusive; e.g. recent investments in North America and Europe by the automobile industry. Where small equity investments are concerned, influence (if not control) is often exercised through technical assistance, marketing or licensing agreements or, as provided for under the MOF interpretations, loans. Kiyoshi Kojima, "Japanese-Style Direct Foreign Investment," Japanese Economic Studies. New York: ME Sharpe Inc., Spring 1986, p. 58. - Page 7 -2.2. SOURCES OF STATISTICAL DATA Two sources of Japanese DFI statistics are available; one is based on notifications (and licenses before December 1, 1980) and the other is based on implementation. The former refers to applications for FDI which are approved by the Ministry of Finance in accordance with the Foreign Exchange and Trade Control Law (1949 and 1980). The second source is the Balance of Payments Statistics compiled by the Bank of Japan. As the former tends to include all reported or licensed investments including, those which are not eventually executed and investments which are divested in the future, the MOF data are an over-estimation. Conversely, the Bank of Japan statistics tend to under-state Japanese FDI, covering only those transfers of monies (in the form of investments in, or loans to, branch offices or subsidiaries) where the Japanese investor holds 10% or more (25% before 1980) of the common stock. According to Hamada (5) , the resulting difference between the MOF and Bank of Japan statistics suggests that "there exists a substantial lag in the actual transfer of funds behind authorization." Ozawa (6) notes that between 1951 and 1965, the ratio of actual FDI to the amount approved by the MOF was 44.4%; and in the periods 1965 to Koichi Hamada, "Japanese Investment Abroad'" in P. Drysdale, ed., Direct Foreign Investment in Asia and the Pacific, Toronto: University of Toronto Press, 1972. Taken from Terutomo Ozawa, Multinationalism, Japanese Style. Princeton: Princeton University Press, 1979 p. 237. 6 Terutomo Ozawa, Multinationalism. Japanese-Style. Princeton: Princeton University Press, 1979, p. 237. - Page 8 -1970 and 1971 to 1979, the ratios were 38.5% and 34.9, respectively. Ozawa attributes the decline in the execution ratio to the increase in the average size of Japanese overseas investments which has prompted longer periods of preparatory activities. (7) On balance, and despite the propensity to over-estimate, it appears that the MOF statistics are the best available estimates of Japanese FDI. The MOF statistics make a number of exclusions which are important in the context of overall Japanese FDI. These exclusions include: (1) Direct investments financed through local borrowing; (2) Technology transfers; and, (3) Reinvestment of retained earnings by Japanese subsidiaries abroad. (8) (1) and (3) are particularly relevant since the size of many Japanese foreign subsidiaries enables them to raise capital overseas on the basis of their own assets. According to Ozawa (9) , therefore, "the amount of direct foreign investment approved by the Japanese government serves as a good proxy for the actual amount invested because the amount of capital raised and reinvested overseas, so far, roughly matches the amount of delayed or 7 Ibid.. p. 238. 8 OECD Economic Surveys. Japan. 1987/1988. Paris, 1988, p. 65. 9 Ozawa, Multinationalism. Japanese-Style, p. 238. - Page 9 -cancelled capital outflows from Japan." In addition to the above exclusions, since 1980, the MOF have dropped foreign real estate acquisitions from the FDI statistics. Under the new Foreign Exchange and Trade Control Law of 1980, notifications of foreign real estate purchases were declared non-mandatory. Given the phenomenal increase in foreign real estate holdings by Japanese firms and individuals, especially since 1984, this exclusion probably represents a serious understatement of the volumes of investment between 1980 and 1987. Further, as the inventory of revenue earning investments has increased, the amount of money available for reinvestment by the overseas subsidiaries is likely to have experienced corresponding growth. Given the above, and because their better availability, this study has generally relied on MOF data. - Page 10 -3.0. HISTORICAL REVIEW OF JAPANESE FOREIGN DIRECT INVESTMENT Direct foreign investment is not a new phenomenon among Japanese strategies. Such investment has been practiced since the early days of the Meiji period and has formed a vital part of Japan's past policies of industrial and political expansion. To establish a proper historical perspective, within which the current FDI status can be evaluated, this section will provide an outline of FDI in each of six periods; ranging from 1870 to the present. For each period, the discussion will present a brief description, not only of the investments made, but also the causal factors behind such investment. From the sequence of period snapshots a picture emerges of an eminently practical approach to using FDI as a tactical device towards furtherance of the strategic objectives of the day. 3.1. JAPANESE FDI PRIOR TO WORLD WAR II (1870 TO 1940) Japan had accumulated approximately Yen 219 billion, in nominal terms, in overseas assets by 1945; which was roughly 3 times the value of GNP in 1944 (10) . Geographically, this was distributed between Manchuria (61.1%) and China (30.1%) with the remaining 8.8% concentrated in other, mainly Southeast Asian, nations (11) . Although the large majority of these investments K. Yasumuro, "The Contribution of Sogo Shosha to Multinationalization", in A. Okochi and T. Inoue, edit., Overseas  Business Activities. Tokyo: University of Tokyo Press, 1984, p.77. 11 Ibid. . p.77. - Page 11 -were of a military nature, a number of important trends were established in the pre-war period which later re-emerged after 1951 when FDI was once again permitted. At the forefront of much of the non-military investments made in Asia prior to World War II were those undertaken by the Japanese textile industry. The textile industry had developed major export markets in Asia after World War I when the British textile producers largely dis-invested in this region. The primary market was China and, by 1915, 80% of Japanese textile exports were earmarked for Chinese markets (12) . Subsequently, however, these export markets were threatened by the rapid expansion of Chinese cotton spinning production capacity, which rose from 650 thousand spindles in 1915 to 2.1 million spindles in 1921. This expansion was supported by protective measures, introduced by the Chinese government after 1919, which included the revision of custom rates on imported yarn and textile products. (By 1930, these rates were as high as 40 to 70 percent 13) . In response to these combined threats, Japanese cotton spinners established local mills in Shanghai and Tsingtao between 1917 and 1922. By 1926, Japanese cotton spinners accounted for 36% of the total production capacity in China, controlling T. Inoue, "A Comparison of the Emergence of MNC manufacturing,11 in Overseas Business Activities, p. 13. 13 Yasumuro, "The Contribution of Sogo Shosha to Multinationalization," p. 71. - Page 12 -approximately 1.3 million spindles ( ). Thus, these early investments by the Japanese textile industry represented the first example of FDI undertaken for the purpose of protecting important Japanese export markets. A second important characteristic of the pre-World War II period was the FDI undertaken by the large Japanese general trading companies (sogo shosha). The sogo shosha were the trading arms of the powerful Japanese financial and industrial conglomerates known as zaibatsu which have contributed extensively to both the pre- and post-World War II economic development of Japan. As global traders, the sogo shosha were instrumental in sourcing the raw materials necessary for Japan's industrialization and in establishing overseas export markets for Japanese products. To support their trading business, the sogo shosha established a network of overseas offices. Mitsui Bussan, the largest trader in the pre-WW II period, established 46 branches in Asia (primarily China) ; 5 in Europe, 2 in the United States and 1 in Australia between 1877 and 1914 to facilitate its trade in textile raw materials and end products (15) . By 1919, 37% of Mitsui Bussan's employees served in overseas branch offices (16) and by 1939, the 14 Inoue, p. 13. 15 Yasumuro, "The Contribution of the Sogo Shosha to Multinationalization," p. 67. 16 Ibid. . p. 67. - Page 13 -company had established 91 branch offices worldwide ( ) . The early "internationalization" of Mitsui Bussan was matched by other large trading companies, notably Mitsubishi Shoji (active in the mineral trade); C. Itoh and Marubeni (textiles), Iwai (chemicals and heavy industrial products); Nissho (foodstuffs) and Kanematsu (wool). Table 3.1 indicates the number of overseas offices established by the leading trading companies prior to World War II. In addition to establishing overseas branch offices to support their trading business, the sogo shosha also participated in overseas joint ventures with Japanese industrial enterprises and local partners. The majority of these joint ventures were in manufacturing or extractive industries with the sogo shosha accepting a minority interest in order to secure the trading business generated from the investment. While small in number (at the end of 1942, Mitsui Bussan had 8 joint ventures with Japanese partners and 3 with Japanese and local partners 18) , these investments represented a precedent for what would become an important characteristic of Japanese FDI in the 1960s and 1970s. 3.2. JAPANESE FDI; 1951 to 1965 At the end of 1945, Japan was stripped of its entire stock of overseas investments and FDI was forbidden until 1951. After 1951 Yoshihara Kunio, The Sogo Shosha. Oxford: Oxford University Press, 1982, p. 17. Yasumoro, "The Contribution of the Sogo Shosha to Multinationalization," p. 76. - Page 14 -TABLE 3.1: NUMBER OF OVERSEAS OFFICES OF SELECTED S060 SHOSHA PRIOR TO tt.U. II NAME TOTAL AS OF HITSUI BUSSAN 91 1939 MITSUBISHI SHOJI 46 1941 C.ITOH 31 1941 IWAI 24 1941 NiSSHO 23 * 1941 •Footnote: includes 4 lanufacturinq subsidiaries in China. Source: Yoshihara Kunio, Sogo Shosha, Oxford: Oxford University Press, 1982, paqes 17, 42, 53, 59 and 69. - Page 14a -modest amounts of Japanese FDI occurred but the average between 1951 and 1965 was less than US $70 million, in nominal terms, per year. There were a number of reasons behind this. First, FDI was strictly regulated to prevent the outflow of scare foreign reserves. At the end of 1945, the yen had been pegged at 360 to the dollar. During the early part of this period the rate represented an over-valuation of the yen, making it difficult for Japan to sell enough products overseas to pay for its much-needed imports of basic foodstuffs and raw materials. Under the Bretton Woods system, the onus of adjustment was largely on the deficit countries through either reductions in the current account deficit or devaluation. For the Japanese, struggling for re-acceptance in the international community, a devaluation of the yen would have represented a major national humiliation. Furthermore, the Japanese need to import capital goods to develop their heavy industrial capacity was also a strong argument against devaluation. To protect the yen, the Foreign Exchange and Trade Control Act of 1949 made all international transactions subject to licensing approval by the Ministry of Finance. As noted by Uno (19) , a second factor behind the lack of Japanese FDI between 1951 and 1965 was that investment opportunities in Japan abounded. Much of the early part of this period was spent re-building war-torn factories and basic social Kimio Uno, Japanese Industrial Performance. Amsterdam: Elsevier Science Publishers B.V., 1987, p. 402. - Page 15 -infrastructure. Later, supported by a variety of government industrial policy measures, Japanese manufacturers of low technology, consumer goods expanded plant capacities for the purpose of export. In the second half of the period, government industrial policy encouraged investment in domestic heavy industries such as iron and steel and petrochemicals. Supported by protective tariffs and preferential and low interest financing, these investments tended to be highly profitable. When coupled with the general lack of financial resources available, the abundance of domestic investment opportunities tended to preclude the demand for FDI. Another compelling reason for Japan's limited FDI during this period was the lack of financial, managerial and technological sophistication in Japanese industries. (One notable exception, however, was the textile sector). Isolation imposed by war had left many Japanese industries technologically backward; as a result, the period between 1951 and 1965 was characterized by massive efforts to import and adapt Western technologies to Japanese manufacturing processes. Despite the relatively small scale of Japanese FDI between 1951 and 1963, the period is important because of the resumption of several pre-war FDI practices noted in section 3.1. First, the trading companies re-grouped after 1951 and once again established overseas branch offices in principal export and import markets. Second, the Japanese textile industry, the most mature and - Page 16 -competitive of Japan's industries in the immediate post-war period, faced rising tariff barriers in Southeast Asia where host governments were implementing import substitution policies. The tariff barriers, when combined with lower costs of production in these countries, threatened the international competitiveness of Japan's textile industry. As a result, textile producers such as Toyobo, Kanebo, and Kurabo invested offshore in Latin America (primarily Brazil) and Southeast Asia. Finally, the period also witnessed the resumption of investments in overseas resource developments for the purpose of export to Japan. These included four large, government-supported projects: Alaska Pulp (1953); the Nippon Usiminas (Brazil) steel project (1957) ; Arabian Oil (1958); and the North Sumatra Oil development (1960) ( ) . 3.3. JAPANESE FDI; 1966 to 1973 Total cumulative Japanese FDI increased rapidly during this period, rising from US $2833 million in 1966 to US $10,030 million in 1973; that is, at a compounded annual growth rate of 43.4%. The average annual flow jumped to greater than US $750 million between 1966 and 1973. This rapid increase can be attributed to several factors. First, after 1965, the balance of payments turned to surplus reducing the pressure on the central authorities to restrict capital outflows. As a result, the period witnessed a gradual liberalization of foreign direct investment, beginning in October 1969 when case by case screening by the Ministry of Finance Uno, Japanese Industrial Performance, p. 401. - Page 17 -was replaced by automatic approval from the Bank of Japan for investments under US $200,000. In September 1970, the limit for automatic approval was raised to US $1 million and it was completely eliminated in July 1971 (21) . According to Yoshihara (22) , this gradual approach reflected both the government's concern that the BOP surplus would be sustained and that FDI not be injurious to domestic industries. By 1972, the effects of liberalization were clearly felt; FDI increased by $3.9 billion in F.Y. 1972 and by $4.3 billion in F.Y. 1973 with the combined total for these two years exceeding the total for the entire 1951 to 1971 period (23) . In addition to liberalizing capital outflows, the Japanese government implemented several measures to encourage FDI. Included among these were the provision of foreign currency loans to residents after August 1972 (designed to eliminate foreign exchange risks); the lowering of interest rates on monies borrowed for FDI through the Export-Import Bank of Japan by 1% in November, 1972; the provision of bank loans through the Export-Import Bank for operating funds as well as machinery and equipment procurement after November 1972; tax incentives allowing for deductions from profits to be used as reserves against FDI losses and credits on Ozawa, Multinationalism Japanese-Style, p. 16. 22 Kunio Yoshihara, Japanese Investment in Southeast Asia. Honolulu: University Press of Hawaii, 1978, p. 3. 23 Uno, Japanese Industrial Performance, p. 403. - Page 18 -taxes paid abroad; and, finally, improved insurance schemes for overseas investments. In addition, in the early 1970s, the government established the Institute of Developing Economies; expanded the Japan External Trade Organization's (JETRO) information service; subsidized overseas missions of the Japan Chamber of Commerce and other private organizations and concluded commercial treaties with foreign countries to expedite the granting of business visas to Japanese managers ( ). The pro-FDI stance of the Japanese government reflected their desire to hold down the appreciation of the yen and to relieve perceived domestic pressures which threatened Japan's export competitiveness. In 1971, the world moved to floating exchange rates and the yen appreciated by 20% between 1971 and 1973. Domestic pressures included rising labour shortages and associated wage hikes; accelerating land costs; pollution problems; and shortages of new sites for industrial production. Wage and land price increases had been particularly severe. In terms of the former, manufacturing wages rose at an average rate of 10% between 1960 and 1965. In the second half of the 1960s, rapidly escalating wage increases exceeded productivity gains and, in 1972, wages increased by 20% (25) . Industrial land prices rose 66% between 1966 and 1973. Finally, pollution problems such as the Minimata mercury poisoning incident in the late 1960s galvanized public support for Yoshihara, Japanese Investment in Southeast Asia, p. 12. Ibid.f p. 4. - Page 19 -increased investments in pollution control and for full consultation between industry and the communities which could be affected by potential developments. These "push" factors driving Japanese manufacturers offshore were reinforced by several "pull" factors. Japan's export success in the late 1960s and early 1970s had met with rising criticism and the introduction of trade restrictions in several of her principal export markets in Southeast Asia. At the same time, many Southeast Asian countries were pursuing export promotion strategies designed to encourage foreign capital and technology investments. These policies included the establishment of export processing zones allowing for the duty-free import of inputs, unrestricted repatriation of foreign earnings and the provision of tax incentives to foreign investors. After 1970, for example, the government of Korea allowed for 100% ownership of companies by foreign investors, exempted new companies from taxation for the first 5 years of operation and gave a 50% tax reduction on corporate profits for the succeeding 3 years (26) . Although formal diplomatic relations were not established between Japan and Korea until 1965, by 1974 the number of Japanese investment projects in Korea far exceeded those in Taiwan, Singapore and Hong Kong (see Table 3.2). A second influence which encouraged offshore Japanese manufacturing investment in Southeast Asia was the dismantling of Yoshihara, Japanese Investment in Southeast Asia, p. 18. - Page 20 -TABLE 3.2: APPROVAL OF INVESTMENT PROJECTS IN EAST ASIA, 1960-1974 YEAR TAIWAN KOREA H0N6 KONG SIN6AP0RE Before 1960 3 - 1 1 1960 4 - 1 -. 1961 2 - 4 1 1962 9 - 1 -1963 8 - 7 8 1964 - - 4 -1965 9 - 3 3 1966 24 10 4 1967 46 3 2 4 1968 80 8 1 3 1969 73 15 6 10 1970 50 58 5 7 1971 16 51 10 14 1972 13 113 16 26 1973 68 290 19 38 1974 (1-3) 21 43 7 2 TOTAL 426 581 97 121 MEDIAN (YEAR) 1969 1973 1971 1972 indicates no investients SOURCE: Kunio Yoshihara, Japanese Investient in Southeast Asia, Honolulu: The University of Honolulu, 1978. p.18 - Page 20a -trade barriers against imports from developing countries after 1970. Japanese tariffs, notoriously high in the 1950s and 1960s were gradually relaxed after 1967 with the signing of the Kennedy Round of GATT. Between 1968 and 1971, tariffs on over 2000 commodities were lowered by 50% and an additional 20% reduction was made on 1900 goods after 1972 (27) More importantly, after August 1971, Japan allowed the duty-free import of manufactured goods from less developed countries, up to certain prescribed • • • • 28 • limits. According to Yoshihara ( ) , imports under this scheme amounted to $760 in 1972 and $806 million in 1973. These tariff reductions provided an incentive for Japanese firms to locate production offshore in nearby Southeast Asia and export to Japan as well as third countries. Finally, Japan's rapid industrial restructuring during the 1960s towards chemical and heavy industries, produced an insatiable demand for raw materials at stable prices. For example, between 1964 and 1968 Japan's demand for petroleum grew at an average annual rate of almost 18% (29) . The demand for copper, aluminum, nickel and crude steel grew at 11.7, 21.0, 25.2, and 20.0 percent, respectively over the same period. By 1969, Japan's share of the 27 Ibid. . p. 5. 28 • • • Yoshihara, Japanese Investment in Southeast Asia, p. 6. 29 Ozawa, Multinationalism. Japanese-Style, p. 159. - Page 21 -OECD coking coal trade exceeded 41%, up from 15.8% in 1965 ( ) and the country was also a major importer of iron ore. In addition, Japan had developed a pronounced dependency on imported foodstuffs. By 1974, Japan imported 95% of the wheat consumed, 96% of the soybean and 82% of the barley (31) . This strong dependence on overseas resource supplies, coupled with Japan's high economic growth rates during the 1960s (averaging in excess of 10% per annum between 1964 and 1968), provided a significant incentive for extractive investment offshore. Table 3.3 indicates the percentage of Japanese resource consumption in 1969 provided by Develop & Import (D&I) arrangements. The D&I projects were generally undertaken by consortiums of Japanese buyers which provided financial backing (principally through long-term loans and purchase agreements, but also through equity participation) to develop overseas resources. Often, the consortiums would consist of several buyers from the same keiretsu or, large oligopolistic industrial group, originating from the former zaibatsu conglomerates. However, in the case of resource developments such as coal or iron ore, inter-keiretsu consortiums were also active; joint purchasing by the Japanese steel industry was one such example of this type of activity. The projects were Japan's share of OECD trade in other raw materials showed similar increases with iron ore rising from 23.7 to 39.3%; timber from 15.4 to 29.9%; copper from 9.5 to 19.1%; and in crude oil from 12.6 to 15.6%. Ozawa, Multinationalism. Japanese-Style, p. 160. 31 Ibid. . 160. - Page 22 -TABLE 3.3: SUPPLY SOURCES OF SELECTED RESOURCES, 1969 DEVELOP-AND-IMPORT DOMESTIC (D 4 1) REGULAR SUPPLY INVESTMENT ABROAD IMPORT PURCHASE RESOURCE I % I COPPER 28.3 16.1 (11.6)* 55.6 LEAD 44.7 6.3 49.0 ZINC 51.1 2.5 46.4 ALUMINUM (BAUXITE) - 9.6 90.4 NICKEL - 6.4 93.6 IRON ORE 14.0 82.5 (75.2)* 3.5 COAL 22.4 68.6 (61.5)* 9.0 CRUDE OIL 0.5 10.5 89.1 NATURAL GAS 95.6 4.4 URANIUM - 0.0 100.0 •Footnote: Percentages in parentheses show those iiports developed under longteri loans instead of direct investment. Source: Terutomo Ozawa, Multinationals, Japanese Style, Princeton: Princeton University Press, 1979, p. 176. - Page 22a -usually organized by the keiretsu major trading company and financed by the group's bank (for example, the Mitsubishi Group would be led by Mitsubishi Trading Company and financed by Mitsubishi Bank.) These consortium-backed investments reduced the financial risk of overseas developments. As noted by Ozawa (32) , "Group investment...plays an important risk-sharing function, because large-scale investments, particularly in resource development, call for huge sums of capital outlay involving extremely high risks...the linkage-sharing function played by group investment appears to be a positive and unique characteristic of Japan's system-focused strategy for overseas extractive ventures." The importance of overseas extractive investments is corroborated by statistics about the type of FDI undertaken during this period. As indicated in Table 3.4, mining investments increased at an average annual rate of 26% between F.Y. 1965 and F.Y. 1972, representing the single largest proportion of Japanese FDI (34%) in F.Y. 1972. However, some of the strongest rates of growth during the period F.Y. 1965 to F.Y. 1972 were witnessed outside the primary sector investments. This was most notable in overseas investments in chemical manufacturing, electrical machinery (including consumer electronics) production, and in the finance-insurance and commerce sectors with compounded annual rates of growth of 42, 37, 27 and 26% respectively. FDI by the textile sector also realized an annual rate of growth in excess of 20%. The rapid growth in manufacturing investments was led by labour intensive industries Ozawa, Multinationalism. Japanese-Style, p. 187. - Page 23 -TABLE 3.4: JAPANESE FDI CLASSIFIED BY INDUSTRY, ON A REPORTED BASIS, F.Y. 1%5 TO F.Y. 1972, CUMULATIVE TOTALS FOR SPECIFIED FISCAL YEAR, MILLIONS OF 1980 U.S. * ANNUAL X OF I OF RATE OF YEAR 1965 TOTAL 1972 TOTAL GROWTH INDUSTRY Manufacturing: Food Stuffs 74 2.8 185 1.5 13.9 Textile 173 6.5 776 6.1 23.9 Pulp fc Lufiber Products 205 7.7 552 4.4 15.2 Chemicals 22 0.8 263 2.1 42.2 Iron I Ferrous Metals 168 6.4 448 3.5 15.0 General Machinery 75 2.8 239 1.9 18.0 Electrical Machinery 35 1.3 321 2.5 37.1 Transport Machinery 160 6.0 265 2.1 7.5 Others 59 2.2 218 1-7 20.6 Sub-Total 972 36.7 3267 25.9 18.9 Primary Industries: Agriculture-Forestry 41 1.6 155 1.2 20.7 Fisheries 24 0.9 91 0.7 20.9 Minina 848 32.0 4241 33.6 25.9 Sub-Total 914 34.5 4487 35.5 25.5 Tertiary Industries: Construction 46 1.8 88 0.7 8.9 Commerce 296 11.2 1410 11.2 25.0 Finance-Insurance 191 7.2 1009 8.0 26.8 Others 231 8.7 2375 18.8 39.5 Sub-Total 766 28.9 4882 38.6 30.3 Total 2652 100.0 12636 100.0 25.0 - Page 23a -(in the case of consumer electronics and textiles) and resource-intensive industries (in the case of chemicals). Investments in the commerce and finance-insurance sectors were undertaken largely to support export offensives and, in the case of the latter, to provide financial services to the growing number of Japanese companies moving abroad. Regionally, as indicated in Table 3.5, Japanese FDI was concentrated in North America (24%) and Asia (23%) with Europe and Latin America accounting for 19 and 18%, respectively. While all regions experienced phenomenal rates of growth in Japanese FDI, Oceania and Europe, with virtually no Japanese investment in 1965, saw cumulative totals grow at annual rates in excess of 60%. Oceania realized a significant proportion of Japanese mining investments while Japanese commercial and financial investments rapidly increased in Europe during this period. 3.4. JAPANESE FDI; 1974 to 1980 The rapid expansion of Japanese FDI witnessed between 1965 and 1973 was interrupted by the oil crisis of 1973 and the ensuing global recession. Although net FDI averaged about $2 billion per annum between 1974 and 1980, net FDI did not recover their 1973 levels until after 1981 (see Table 1.1). The decline in Japanese FDI was largely induced by the world recession which followed the 1973 oil shock. This reduced the profitability of Japanese businesses both at home and abroad and - Page 24 -TABLE 3.5: JAPANESE FDI BY RE6I0N ON A CUMULATIVE BASIS, F.Y. 1965 TO F.Y. 1973, MILLIONS OF 1980 U.S. $ ANNUAL I OF I OF RATE OF YEAR 1965 TOTAL 1973 TOTAL 6R0WTH RE6I0N N. AMERICA 666 25.4 4069 24.0 25.4 LATIN AMERICA 776 29.6 2993 17.6 18.4 ASIA 519 19.8 3947 23.3 28.9 EUROPE 69 . 2.6 3299 19.4 62.1 OCEANIA 19 0.7 1058 6.2 64.9 AFRICA 30 1.2 418 2.5 38.8 MIDDLE EAST 541 20.7 1183 7.0 10.3 TOTAL 2622 100.0 16969 100.0 26.3 - Page 24a -provided an uncertain environment for foreign investment. However, other factors contributed to the decline of Japanese FDI. For example, in Southeast Asia rising nationalism resulted in anti-Japanese demonstrations in Bangkok and Jakarta during the visit of Prime Minister Kakuei Tanaka in 1974. The anti-Japanese sentiment was largely spawned by the visible concentration of Japanese investment in this region. As indicated in Table 3.6, American investment was comparable to Japanese FDI in Southeast Asia at the end of 1974, ($3,533 million in total US FDI vs $3,764 million in total Japanese FDI, nominal terms). However, the number of projects undertaken by Japanese companies in countries for which information is available (33) , exceeded the corresponding number of American projects by 2.7 times. This profusion of projects tended to make the Japanese economic presence more conspicuous than that suggested by its total value of investment. The concentration of Japanese firms in Southeast Asia resulted from the "bandwagon effect of investment,"(34), where the relocation of one Japanese firm was usually followed by several in the same industry. According to Murakami this "follow the leader" investment pattern resulted in the transplanting of excessive competition by the same firms in Japan to ASEAN and NIC nations. Because Japanese Indonesia, Taiwan, Hong Kong and South Korea. 34 Hikoji Katano, Atsushi Murakami, Kiyoshi Ikemoto, edit, Japan's Direct Investment to Asean Countries. Kobe: Research Institute for Economics and Business Administration, Kobe University, 1978, p. 8. - Page 25 -TABLE 3.6: JAPANESE FDI.BY REGION ON A CUMULATIVE BASIS, F.Y. 1974 TO F.Y. 1980,. MILLIONS OF 1980 U.S. $ ANNUAL I OF I OF RATE OF YEAR 1974 TOTAL 1980 TOTAL GROWTH REGION N. AMERICA 4120 23.8 9798 26.8 15.5 LATIN AMERICA 3434 19.8 6168 16.9 10.3 ASIA 4267 24.6 9830 26.9 14.9 EUROPE 2989 17.3 4471 12.3 6.9 OCEANIA 1023 5.9 2525 6.9 16.2 AFRICA 421 2.4 1445 4.0 22.8 MIDDLE EAST 1067 6.2 2259 6.2 13.3 TOTAL 17321 100.0 36496 100.0 13.2 - Page 25a -investments were generally undertaken in labour intensive industries, domestic firms were often unable to develop competitively. This problem of over-concentration was not limited to Southeast Asia. After 1974, there was increasing criticism of Japanese acquisitions in Hawaii following a surge of investment in hotels, golf courses, condominiums and other real estate between 1970 and 1973 (35) . Despite the problems resulting from global recession and rising nationalism among host countries, Japanese FDI recovered in the latter half of the 1970s, jumping from $20,257 million, on a notifications basis, in 1975 to $36,496 million in 1980. The increase in FDI was attributable to the recovery of the Japanese economy after 1975; the sustained appreciation of the yen throughout the decade (with the exception of 1975); and continually accelerating domestic costs for land, labour and energy. In addition, Japan became the subject of greater import restraints by the industrialized nations. The most notable of these were the import restraints introduced on colour television imports by the United States between July 1977 and June 1980 (36) . Local production of colour televisions, initiated in the United States in the first half of the 1970s, was rapidly accelerated. Sony was the first electronics firm to undertake FDI in the US in 1972 with 35 James C. Abegglen & George Stalk Jr., Kaisha: The Japanese  Corporation. Tokyo: Charles E. Tuttle Company, 1985, p. 257. 36 Uno, Japanese Industrial Performance, p. 405. - Page 26 -construction of a plant in San Diego, but Matsushita soon followed, acquiring Motorola's TV division in 1974. A final stimulant to renewed FDI in this period was provided by the greatly increased degree of international sophistication of Japanese companies, in terms of the managerial, technological and the financial resources available for overseas investment. Regionally, the increase in Japanese FDI was most notable in North America (from $4977 million in 1975 to $9798 million in 1980); Asia (from $5361 million in 1975 to $9830 million in 1980); Oceania (from $1182 million in 1975 to $2525 million in 1980) and in the Middle East (from $1240 million in 1975 to $2259 million in 1980). By 1980, Japanese cumulative investment was evenly divided between North America and Asia (27%). (See Table 3.6). In terms of sectoral developments, between 1974 and 1980, tertiary sector investments became more significant while investments in the overseas resource developments, as a percentage of overall investment, declined. In 1972, tertiary investments accounted for 39% of total investments, followed closely by primary investments (36%) and manufacturing investments (26%). By 1980, tertiary investments had expanded to 44% of total investments with manufacturing and primary investments accounting for 34% and 22%, respectively. Industries realizing the largest gains during the period were chemicals (up 898%) ; iron and ferrous materials (up 485%); electrical machinery (up 392%); construction (up 350%); commerce (up 284%) ; transport machinery (up 268%) ; and textiles (up - Page 27 -11.1%) . Although mining investments still represented the largest proportion of total cumulative investment in 1980, its share of total investment had fallen to 19% in FY 1980, from 34% in FY 1972. (See Table 3.7). The relative decline of mining investments and of the primary sector overall, was due to the slower growth of the Japanese economy after 1975 and the declining importance of heavy industry. Although Japanese real GNP continued to grow at around 5% after 1977, this represented a halving of the growth rates experienced during the 1960s and early 1970s. Concentration in heavy industry and raw materials processing industries had been sustainable while Japan was able to access the necessary resource inputs under stable price conditions. However, these conditions changed after the late 1960s. According to Abegglen (3r) , "It was clear by the late 1960s that raw materials processing industries would come under pressure to reduce production in Japan. The choice was either to source semi-finished or finished product from off-shore, or move off-shore to control the sourcing by investing in facilities in foreign countries. This shift in structure away from on-shore processing was likely for a variety of reasons; pressures from material-supplying countries to add value to the materials in the source country before their export; costs in Japan from very stringent pollution control requirements; and energy and land cost disadvantages in Japan. The probability of this industrial structure change became an inevitability as energy prices exploded in the 1970s." The rising factor input costs and structural changes in the economy had a profound effect on Japan's petrochemical industry. After 1974, foreign chemical manufacturing investments increased Abegglen & Stalk, Kaisha: The Japanese Corporation, p. 244. - Page 28 -TABLE 3.7: JAPANESE FDI CLASSIFIED BY INDUSTRY, ON A REPORTED BASIS, F.Y. 1972 TO F.Y. 1980, CUMULATIVE TOTALS FOR SPECIFIED FISCAL YEAR, CONSTANT 1980 DOLLARS, MILLIONS OF U.S.* ANNUAL I OF I OF RATE OF YEAR 1972 TOTAL 1980 TOTAL GROWTH INDUSTRY Manufacturing: Food Stuffs 185 1.5 587 1.6 15.6 Textile 776 6.1 1637 4.5 9.8 Pulp & Lumber Products 552 4.4 758 2.1 4.0 Chemicals 263 2.1 2626 .7.2 33.3 Iron & Ferrous Metals 448 3.5 2619 7.2 24.7 General Machinery ,239 1.9 . 894 2.4 17.9 Electrical Machinery 321 2.5 1579 4.3 22.0 Transport Machinery 265 2.1 979 2.7 17.7 Others 218 1-7 894 2.4 19.3 Sub-Total 3267 25.9 12573 34.4 18.3 Primary Industries: Agriculture-Forestry 155 1.2 609 1.7 18.7 Fisheries 91 0.7 301 0.8 16.1 Mining 4241 33.6 7071 19.4 6.6 Sub-Total 4487 35.5 7981 21.9 7.5 Tertiary Industries: Construction 88 0.7 396 1.1 20.7 Commerce 1410 11.2 5409 14.8 18.3 Finance-Insurance 1009 8.0 2426 6.6 11.6 Others 2375 18.8 7712 21.1 15.9 Sub-Total 4882 38.6 15943 43.7 15.9 Total 12636 100.0 36497 100.0 14.2 - Page 28a -substantially, as Japanese petrochemical manufacturers responded to increased demands by the oil-producing nations for greater value-added production at home. Rising costs of production domestically and increased uncertainty about the reliability of supply sources further reinforced the need to relocate production offshore. The expansion in petrochemical investments in the Middle East, especially, was accompanied by large increases in government aid and loans to these regions. This assistance was intended to ensure an uninterrupted flow of oil supplies since Japan, even by 1978, was still dependent on the Middle East for over 85% of its oil supplies. Loans extended by the Japanese government to the Middle East increased to $3 billion (nominal dollars) in 1973, up • • • 38 5 times from the corresponding amount extended in 1972 ( ) . The increase in government aid and promises for corporate investment was openly welcomed by the Middle East host countries and after January 1974, Japan was placed on OPEC's list of "friendly nations." After reaching an economic and technological agreement with the Japanese government in February 1974, the Saudi Arabian Oil Minister, Sheik Ahmed Zaki Yamani is reported to have stated that "Japan in the No. 1 position both to help us and to be the recipient of Saudi Arabian oil on a long term basis." (39) As usual, the soga shosha were at the forefront of this surge in investment in the Middle East. The trading companies either Ozawa, Multinationalism, Japanese-Syle. p.144. Ibid.. p. 145. - Page 29 -participated in, or acted as intermediaries for practically all the Japanese industrial projects in the region (40) . Included among these projects was the turnkey contract signed by SONATRAC, Algeria's state-owned oil-gas corporation, with C.Itoh and Toyo Engineering to build a huge maintenance and repair complex for Algeria's pipeline, oil refining, petro-chemical, LNG and mining industries (41) . Other large projects organized by the trading companies included the $200 million direct-reduced iron plant established, in 1974, at Alexandria as a joint venture between the Egyptian government (50% ownership) and C.Itoh, Korf-Stahl of West Germany, and Compahia Vale do Rio Doce of Brazil (42) . The Middle Eastern projects were not without their difficulties; in the early 1970s, Mitsui Bussan agreed to organize the construction of the world's largest petrochemical project in Iran in exchange for an oil concession right in the Lorestan area southwest of Teheran (43) . The project, in which Mitsui Bussan invested in excess of $500 million (the largest single foreign investment of the Mitsui group), faced construction delays and finally, an indefinite postponement of completion after the fall of the Shah in 1979. Other large-scale investments undertaken by Japanese companies *u Ibid., p. 145. 41 Ibid. . p. 146. Ozawa, Multinationalism. Japanese Style, p. 147. 43 Yoshihara, The Sogo Shosha. p. 119. - Page 30 -during this period, included major projects in Brazil and Indonesia. Brazil has long been a popular host to Japanese FDI, initially in the textile and other light manufacturing sectors during the 1950s and 1960s, but later in shipbuilding and iron & steel. However, after 1974, further large investments were undertaken in the resource processing industries including a large electric power and aluminum project in the Amazon basis. The Japanese consortium, originally consisting of 5 major smelters but later expanding to include participation by the governmental agency OECF (the Overseas Economic Cooperation Fund) and 32 private corporations, has a 49% ownership in the project. The remaining 51% is controlled by the state mining company Companhia Vale do Rio Doce (44) . In addition, other large investments have been made in steel making (Kawasaki Steel's 24.5% participation in the Tubarao steel-making project in Espirito Santo State); pulp and paper (Oji Paper's investment in the Cinebra pulp project in the state of Minas Gerais); and fertilizers (Sumitomo Chemicals and Ataka of Japan's investment with a local partner to build a million ton capacity plant in Recife). In Indonesia, also a popular host country to Japanese investment, the $800 million Asahan electric power and aluminum project is another example of resource-based investment undertaken after 1974. The refinery is 90% owned by Japanese interests, however, initial plans were to transfer 25% of the shares of the project to the Indonesian government 10 years Ibid.. p. 135. - Page 31 after its completion. In addition, control of the 430,000 kilowatt hydro-electric power plant is to revert to the Indonesian government after 30 years. A final characteristic of Japanese FDI undertaken between 1974 and 1980 is the large number of overseas investments made by small to medium sized firms. At the end of the 1970s, small to medium scale enterprises accounted for 41.8% of the total number of outstanding Japanese manufacturing projects overseas (45) . However, when examined on a country by country basis, the ratio of participation by these firms was higher than that of large companies in some countries. In Taiwan, they represented 58.6% of the total investment projects and in South Korea, they represented 70% (46) . In 1979, 82.3% of the total value of manufacturing investments made by small to medium size firms were located in Asia. According to Ozawa, this predominance of small and medium sized firms in Asia reflected not only the traditional domination of such enterprises in Japan's industrial structure but also the "unique process by which the labour-intensive, low-productivity end of the dual industrial structure (was) being gradually pushed out of Japan and sent to more labour-abundant neighboring countries through direct foreign investment" (47) . These Yoshihara, The Sogo Shosha. p. 26. Small to medium size firms refer to those companies having employees of 300 or fewer or paid in capital of Y100 million or less. 46 Ibid. . p. 28. i 47 Yoshihara, The Sogo Shosha. p. 28. - Page 32 -projects were concentrated in the textiles, sundries, metal products and relatively unsophisticated light manufacturing industries, accounting for 65.5% of the Japanese FDI in textiles, 50.2% of her investments in electrical appliances and 60.7% of her investments in sundries being located in Asia, during 1978. This large overseas investment presence by small and medium scale companies was in direct contrast to typical US multinational investments. The latter was generally undertaken by large, technically and financially, sophisticated firms which demonstrated strong competitive advantages in (primarily) growth industries. On the other hand, a large proportion of Japanese investment prior to the 1970s was undertaken by firms which were relatively unsophisticated in terms of financial and managerial resources and was directed towards industries which were declining in Japan. 3.5. JAPANESE FDI; 1981 to 1984 A number of important trends occurred during this period, including the rapid escalation of annual FDI flows. In 1981, Japanese cumulative net FDI reached $27 billion with net annual investment flows exceeding $4 billion for the first time (see Table 1.1). On a notifications basis, cumulative FDI increased from $44 billion in F.Y. 1981 to $66.6 billion in F.Y. 1984. By the end of F.Y. 1984, annual FDI flows had increased to slightly over $10 billion. The increase in the volume of FDI can be largely attributed to the continuing appreciation of the yen against the - Page 33 -US dollar; decreasing factor cost differentials and rising trade frictions between Japan and her industrialized trading partners; and increased incentives for FDI in many developed countries. Japanese FDI, however, did not grow as fast as in previous periods, averaging around 18% per year. This was due, in large part, to the world recession following the second oil shock in 1979. One of the most significant aspects of this period was the increased concentration of Japanese FDI in North America. At the end of 1980, as noted previously, Japanese FDI was evenly divided between North America and Asia with each accounting for 27% of the total investment and Latin America and Europe accounting for 17 and 12%, respectively. By F.Y. 1984, 30% of total Japanese FDI was held by North America; the corresponding Asian share had fallen to 25%. Latin America and Europe had increased their shares marginally to 18% and 13%, respectively (See Table 3.8). A large proportion of the incremental investment earmarked for North America was in the commercial and financial sectors, however, manufacturing investments also increased sharply. By F.Y. 1984, North America accounted for 29.4% of total Japanese manufacturing investment compared to 19.3% in F.Y. 1980. (Japanese manufacturing investment in Asia accounted for 36.4 and 32% of total manufacturing investment in F.Y. 1980 and F.Y. 1984, respectively 48) . At the end of F.Y. 1984, Japanese manufacturing investments Jetro, White Paper on World and Japanese Direct Foreign Investment, Tokyo, 1986, p. 8. - Page 34 -TABLE 3.8.: JAPANESE FDI BY REGION ON A CUMULATIVE BASIS, F.Y. 1981 TO F.Y. 1984, MILLIONS OF 1980 U.S. * ANNUAL I OF I OF RATE OF YEAR 1980 . TOTAL 1984 TOTAL GROWTH REGION N. AMERICA 9798 26.8 20028 30.1 19.6 LATIN AMERICA 6168 . 16.9 12146 18.2 18.5 ASIA 9830 26.9 16816 25.2 14.4 EUROPE 4471 12.3 8463 12.7 17.3 OCEANIA 2525 6.9 3468 5.2 8.3 AFRICA 1445 4.0 2982 4.5 19.9 MIDDLE EAST 2259 6.2 2730 4.1 4.9 TOTAL 36496 100.0 66633 100.0 16.2 - Page 34a -in the United States numbered 440, were located across 40 states and involved firms which employed approximately 80,000 people. The manufacturing investments in North America were characterized by investments in joint-venture, technology-intensive fields. Between 1984 and 1985, the number of cases of US-Japan industrial cooperation increased to 652, with manufacturing related joint ventures accounting for 79%, or 505 cases. Of these, the greatest number of cases were in high-technology related fields with 59 cases in computer-related industries; 51 in semi-conductors and ICs; and 25 in new materials ( ). During the early 1980s the Japanese auto industry became increasingly active in United States investments, as a means of circumventing tighter restrictions on Japanese auto imports into North America. Honda was the first auto manufacturer to begin production with the 1982 opening of its Marysville plant in Ohio. In the early 1980s, investments by consumer electronics and automotive manufacturers, coupled with the demands by recipient countries for greater local procurement, also led to related investments by components-manufacturing industries (e.g., auto parts, and electronics components). The early 1980s also witnessed an increase in the number of acquisitions in the North American financial industry by the large Japanese banks. By 1984, almost one quarter of the world's largest Jetro, White Paper on World and Japanese Direct Foreign Investment, p. 12. - Page 35 -banks were Japanese with the largest, Dai-Ichi Kangyo (assets in excess of $108 billion), being ranked seventh in the world. The high-profile takeovers of several American banks were an indication of the growing power and international confidence of Japanese financial institutions. Included among these acquisitions were Fuji Bank's $450 million acquisition of the financially troubled Heller Financial Services Group; the $282 million takeover of the Bank of California by Mitsubishi Bank; and Sumitomo Bank's takeover of Union Bank of California in 1984. Japanese investments in Europe also increased significantly during this period, rising at an average annual rate of 19%. Previously, Japanese FDI in Europe had been primarily in the tertiary sector (commerce, finance, insurance and branch offices). Early manufacturing investments were mainly in sales companies and were designed to promote exports and serve as a foothold to develop distribution and marketing channels, if and when local production was undertaken. However, in the early 1980s, several European nations implemented incentive schemes to attract foreign direct investment in the hopes of reducing chronic unemployment problems. In addition, by the end of the period, further steps were being studied towards removing non-tariff barriers and unifying the EEC into a single market. These movements encouraged increased Japanese manufacturing investment and, by F.Y. 1984, Japanese manufacturing ventures in - Page 36 -Europe totalled 188 ( ) . The principal recipients of this investment were the United Kingdom, France and West Germany. In the U.K. the ventures included investments made by Brother and Sharp Industries (electronic typewriters), Tabuchi Electric (small transformers for Video Tape Recorders (VTRs) and Nihon Radiator/TI Silencer (mufflers). The French investments included those by Honda Motor (power mowers) , Canon (electronic typewriters) and Sony (compact disc players). In West Germany, the Sanyo Group (tuners for VTRs), Matsushita Electric (electronic components), Hitachi (VTRs), Matsushita Group (car radios and stereos) and Hitachi Koki (electric tools) were among the principal investors (51) . The average annual growth of Japanese FDI in Asia declined marginally to 14% during the period 1981 to 1984 from 15% over the previous period. However, in absolute terms, total cumulative investment increased 83% from $9,830 million in F.Y. 1980 to $16,816 million by F.Y. 1984. The stagnant growth rate may be attributed to the fact that the majority of labour-intensive, low technology industries had already relocated from Japan during the 1970s. However, increased investments were undertaken in response to meet the demands of burgeoning consumer markets due to rapid economic growth. In addition, increasing costs in the NICs were also resulting in the shift of investment towards lower cost Jetro, White Paper on World and Japanese Direct Foreign Investment, p. 14. 51 Jetro, White Paper on World and Japanese Direct Foreign Investment, p. 15. - Page 37 -producers such as Thailand, Malaysia and China. By the end of 1984, China was receiving vigorous attention, with 741 additional cases of Japanese investment in F.Y. 1984 (52) . 3.6. JAPANESE FDI? 1985 TO THE PRESENT The most recent period of Japanese FDI has been characterized by a continuation of trends started in the early 1980s, but on a much larger scale. The cumulative total of Japanese FDI grew by almost 71% from $66.6 billion in F.Y. 1984 to $130 billion in F.Y. 1987. This increase represented a compounded annual growth rate in excess of 22%. (See Table 3.9). The indicated net FDI for F.Y. 1988 showed a further quantum jump from $18.6 billion U.S. in 1987 to over $33 billion U.S. in 1988. The increase in the volume of Japanese FDI, especially since 1984, has been due to a number of factors. First, between 1984 and 1987, the yen appreciated against the dollar by over 80%. This appreciation has both increased the relative costs of producing in Japan (thus reducing the international competitiveness of Japanese exports and increasing the appeal of imports) and has made overseas investment, especially in North America, very attractive. In addition, the precipitous rise in the yen has been accompanied by mounting trade actions by developed countries against Japanese local procurement practices and the emergence of potentially protectionist trading blocks in Japan's major regional export Ibid., p. 13. - Page 38 -TABLE 3.9: JAPANESE FDI CLASSIFIED BY INDUSTRY, ON A REPORTED BASIS, F.Y. 1984 TO F.Y. 1987, CUMULATIVE TOTALS FOR SPECIFIED FISCAL YEAR, MILLIONS OF 1980 U.S. * YEAR INDUSTRY Manufacturing: Food Stuffs Textile Pulp k Lunber Products Cheaicals Iron fc Ferrous Metals General Machinery Electrical Machinery Transport Machinery Others Sub-Total Priiary Industries: Agriculture-Forestry Fisheries Mining Sub-Total Tertiary Industries: Construction Coiaerce Finance-Insurance Services Transportation Real Estate Others Sub-total Total ANNUAL RATE OF I OF I OF GROWTH 1984 TOTAL 1987 TOTAL 1984-87 961 1.3 1398 i.i 13.3 1974 2.7 2127 1.7 2.5 993 1.4 1384 1.1 11.7 3714 5.1 4744 3.8 8.5 4419 6.1 5700 4.5 8.9 1765 2.4 2969 2.4 18.9 3298 4.5 6469 5.1 25.2 2901 4.0 5131 4.1 20.9 21633 29.8 35278 2B.0 17.7 745 1.0 413 0.6 11138 15.3 11696 9.3 1.6 12296 16.9 11696 678 0.9 1025 0.8 14.8 11062 15.2 15196 12.1 11.2 8670 11.9 26014 20.6 44.2 3958 5.4 8161 6.5 27.3 4199 5.8 9014 7.2 29.0 2078 2.9 10812 8.6 73.3 8047 11.1 8762 7.0 3.0 38692 53.3 79005 62.7 26.9 72622 100.0 125980 100.0 20.2 - Page 38a -destinations. These combined forces have pushed Japanese FDI to record levels, drastically increasing the profile of Japanese overseas investment practices both at home and abroad. The role of North America as the primary destination of Japanese FDI was further emphasized after 1984. By F.Y. 1987, this market accounted for 36.1% of the value of Japan's total cumulative investment and approximately 36.2% of the total number of cases of FDI (see Table 3.10). According to the U.S. Commerce Department, in 1987, Japan passed Great Britain as the largest foreign direct investor in the United States (53) . The onslaught of Japanese FDI has largely resulted from the rapid appreciation of the yen against the dollar, the dominant position of the United States among Japan's export markets and the strong performance of the American economy. In addition, investment in the United States has been encouraged by the development of state-based incentive programs for FDI such as government subsidies to foreign firms locating in depressed regions. Further, the unitary tax systems (54) were repealed in several states, including California; thus removing a contentious obstacle to foreign investment. Canada has also experienced an increase in Japanese FDI because of an improved "Japanese are Biggest Foreign U.S. Investors,", The  Financial Post. June 28, 1989. 54 The unitary tax system, adopted in several states, sums the total earnings of the firm uses the ratio of assets, wages and turnover of the parent to calculate the corporation tax levied on the subsidiary operating within that state. Foreign investors charge that the system results in double taxation and cumbersome administrative and clerical effort to collect the required data. - Page 39 -TABLE 3.10: JAPANESE FDI BY REGION ON A CUMULATIVE BASIS, F.Y. 1984 TO F.Y. 1987, MILLIONS OF 1980 U.S. i ANNUAL RATE OF X OF I OF GROWTH YEAR 1984 TOTAL 1986 TOTAL 1984-86 RE6I0N N. AMERICA 20028 30.1 33729 35.3 29.8 LATIN AMERICA 12146 18.2 18371 19.2 23.0 ASIA 16816 25.2 19647 20.6 8.1 EUROPE 8463 12.7 13049 13.7 24.2 OCEANIA 3468 5.2 4720 4.9 16.7 AFRICA 2982 4.5 3317 3.5 5.5 MIDDLE EAST 2730 4.1 2720 2.8 -0.2 TOTAL 66633 100.0 95553 100.0 19.8 - Page 39a -foreign investment climate and its potential as a stepping stone to the larger US market. The recent Free Trade Agreement concluded between Canada and the United States, in particular, has been cited as an important future stimulant to Japanese investment in Canada (55) • Europe has also been an increasing recipient of Japanese FDI, accounting for 13.5% of the value of cumulative investment in F.Y. 1987 and 10.8% of the total cases. Although these proportions increased only marginally over F.Y. 1984 (in F.Y. 1984, Europe accounted for approximately 13% of total Japanese cumulative investment) , the rate of growth of Japanese FDI in Europe has averaged 31% annually since 1984 (versus 24% for North America). The rapid rate of growth in Japanese FDI in Europe has been fuelled by two important factors. Firstly, the number of trade actions against Japanese imports and perceived "screwdriver" assembly plants has increased dramatically. In terms of the former, in 1987, a 20% dumping tariff was applied to all Japanese copiers, thus forcing local production by Canon Inc., Konica Inc., Ricoh Co., Minolta Camera Co., Toshiba, Matsushita Electric Industrial Co. and Sharp Corp (56) In addition, in November of 1988, a 47% anti-dumping tax was placed on Fujitsu computer Report prepared by the Gleneagles Group, Japanese Direct  Foreign Investment in North America - A Canadian Perspective. Vancouver, January 1988, p. iv. 56 "Challenge and Dilemma for Corporate Japan, the EC Gears up for 1992," TOKYO Business Today. Tokyo, April 1989, p. 26. - Page 40 -printers and a 20 to 34% duty on all other Japanese printers ( ) . Severe penalties have also been levied against Japanese manufacturers in Europe which do not meet mandatory local procurement requirements. Under EC local content rules, at least 40% of the value of the product must be generated from local production; the percentage is even higher for certain products (e.g., automobiles, 60%) and certain countries (e.g., France, automobiles, 80%). Japanese subsidiaries producing electronic typewriters, photocopiers and electronic scales have been particularly affected since 1987 (58) . Secondly, the commitment by the 12 member EC trading block to complete unification of the European Community market by 1992 has also prompted investment by Japanese firms seeking to establish a foothold before "Fortress Europe" emerges. Although down-played by the European Parliament, 1992 has sparked strong anti-protectionist fears among Europe's major trading partners. The broadly-defined reciprocity concept introduced in 1988, (whereby trading partners would be allowed equal access to European markets, if and only if, European companies have the same degree of access to foreign markets) has led to concern that certain European markets previously open to foreign exporters will be closed after 1992. Despite reassurances to the contrary, the ill-defined nature 57 Ibid. . p. 26. "Japan Copiers Made in U.S. May Be Hit With EC Duties," The Asian Wall Street Journal. February 8, 1989, p. 4. Page 41 -of 1992's impact on the international community and the threat of exclusion from the world's largest common market, has been a strong impetus to Japanese FDI in Europe since 1985. With the rise in relative importance of industrialized nations as hosts to Japanese FDI has been a concomitant decline in the proportion of Japanese investment in the developing countries. The decline has been most notable in Asia. The region in F.Y. 1987 accounted for only 20.6% of total Japanese investment, down from 25% in 1984 (see Table 3.10). Nevertheless, investment in Asia has been increasing in absolute terms since 1984, albeit at a slower rate than was previously the case, (8%/annum versus 14%/annum between 1980 and 1984). As a result, total cumulative investment increased from $16.8 billion in F.Y. 1984 to $20.6 billion in F.Y. 1987. An interesting feature of the investment patterns in Asia since 1984 has been the increasing shift of investment from the NICs to relatively lower cost producers such as Thailand, Malaysia, the Philippines and China. Yamaha's investment sequence provides an example of this shifting focus. In 1979, Yamaha went to Singapore to produce tennis rackets; in 1982, to Taiwan to produce golf clubs; in 1987, to Thailand to produce skis; and in 1988, to Indonesia to produce electronic organs (59) . Manufacturing production in some instances has also been scaled down, or shut completely, in the NICs. Uniden, a manufacturer of telecommunications equipment, is closing factories in Taiwan and "On the Move Again," The Economist. November 5, 1988, p. 83. - Page 42 -Hongkong and moving to the Philippines and China; Asahi Optical has reduced the number of employees at its Pentax Camera plant in Hongkong to 150 from 350 (60) . Africa and the Middle East have experienced declines in both the proportion of total and rate of growth of Japanese FDI in their regions. Latin America, on the other hand, has increased both its growth rates and overall proportion of Japanese FDI since 1984. This, however, is attributed to the rise in Japanese banking and insurance investments in such "tax havens" as the Bahamas, Cayman and Panama rather than increases in manufacturing investment. The decreasing importance of the developing regions is due to several reasons. First, political and economic instability in the developing countries, primarily the Middle East, Africa and Latin America, greatly increased the risks associated with foreign investment. Second, stricter government controls on FDI, particularly with regard to resource investments, also deterred foreign investment by the Japanese. Third, structural changes in Japan leading to lower requirements for offshore resources further inhibited Japanese FDI in the resource sectors of developing nations. Fourthly, as previously mentioned, most of the offshore labour-intensive manufacturing investment by Japan had been completed by the late 1970s. Finally, there is an emerging trend of Japanese FDI in ventures which are primarily targeted towards profit maximization. In the case of manufacturing companies this "On the Move Again," The Economist, p. 83. - Page 43 -means adopting the multinational companies' approach of making maximum permitted use of comparative advantage to increase returns, rather than being influenced primarily (as they were in the past) by domestic (Japanese) market considerations. The situation is even more clear-cut in the real estate and financial areas. In the former case, investments are being made because real estate, especially in North America, is considered to be under-valued, with good appreciation potential. Shuwa Corporation (Shigeru Kobayashi) which, by 1987 had purchased about $1.5 billion U.S. of American real estate, now earns more rental income from its U.S. assets than it does from its Japanese holdings. This shift has resulted from the massive increases in Japanese real estate prices (which have provided huge amounts of collateral), the relatively low cost of borrowing in Japan and the increased purchasing power of the yen in North America (61) . The investments in the financial sector reflect the enormous financial power of the Japanese banks, insurance companies and security firms and the natural progression towards translating this power to an international market scale. The FDI of the 1970s was characterized by the transfer of labour- and resource-intensive manufacturing production offshore. However, during the 1980s, and particularly since 1984, the trend has been towards technology-intensive manufacturing industries. Examples of this trend is provided by the automobile, electronics 61 "The Shuwa Shogun", Tokyo Business Today. Tokyo, March 1987, p.28-32. - Page 44 -and office and automation industries. Until recently, the technology-intensive industries have tended to remain in Japan because they have been able to absorb the increasing costs of local factors of production through rapid productivity improvements. In fact, many of these industries would prefer not to invest offshore. As noted by President Jun Kobayashi of Tokyo Electric, "To be quite honest, it would be so simple to just produce everything in Japan, and then sell it around the world. The fact than we cannot forces us to adopt new strategies " (62) . However, during the 1980s, many of these industries have been forced to move offshore because of rising trade restrictions in important export markets. The FDI of the Japanese automobile industry in North America and Europe is a clear example of this. Startling productivity gains in this industry at home (around 45% since 1985) have meant that exports are still more profitable than overseas production despite the 80% appreciation of the yen over the past 4 years. According to an article by Kevin Done of the Financial Times (63) , the exchange rate break-even point for cars is now Y105 to the dollar whereas five years ago it was Y160-170 to the dollar. (By 1991, this break-even point could be as low as Y95 to the dollar.) However, as a result of the voluntary export restraints (VERs) introduced in the early 1980s, local production in the United States by Japanese auto-"Challenge and Dilemma for Corporate Japan," TOKYO Business  Today. April 1989, p. 28. 63 Kevin Done, "Car Wars After the Yen Shock," The Financial Times, London, May 12, 1989. - Page 45 -manufacturers could reach between 1.8 to 2.2 million units per year by 1992. Honda, the largest Japanese manufacturer operating in the U.S., currently produces over 500,000 cars per year. As noted above, another important trend in Japanese FDI since 1984, has been the emergence of the financial-insurance industry as the dominant overseas investor. By 1987, finance-insurance investments accounted for the largest single proportion of total Japanese FDI (17.1%). Again, these investments have been concentrated in the United States and Europe. Japanese banks now own 5 of the largest 11 banks in California, controlling $363 billion in banking assets (w). Accompanying the investment surge in the finance-insurance industries has been a boom in real estate acquisitions. Foreign real estate holdings by Japanese increased at a phenomenal pace between 1985 and 1987, rising from $2.9 billion in F.Y. 1984 to almost $11 billion by F.Y. 1987. These acquisitions have often involved high-profile purchases such as the $610 million acquisition by Mitsui Real Estate Development Co. Ltd of the Exxon Building in New York and the $620 million purchase of the Arco Plaza in Los Angeles by Shuwa Corporation (65) . According to James D. Noteware, the national director of real estate services for Laventhol & Horwath, the future Japanese activity in US real estate John Woodruff, "Banking on the Golden State," The Japan  Times. May 22, 1989, p. 6. 65 "Overseas Investment", Japan Economic Almanac. 1987. Tokyo, 1988, p. 32. - Page 46 -will be diversified away from "high-profile trophy office buildings" to suburban office buildings, industrial/business parks, resorts and special facilities such as nursing homes and casinos as smaller, more flexible Japanese investors become active in the market (bb) . The Japanese have also been active in real estate investment in Europe and Australia. In Australia, the boom has been concentrated in the Gold Coast tourist region, where the Japanese spent over Australian $1.2 billion acquiring hotels, resorts, and future tourism-related development sites in F.Y. 1988. Japanese investors now control 57% of the value and 64% of the number of rooms of the seven largest international hotels in the Gold Coast area (67) . Three other important characteristics of Japanese FDI since 1984 are important signals of the future pattern of investment towards the internationalization of Japanese company profit strategies. First, mergers and acquisitions are becoming an increasingly popular method of entering foreign markets, reflecting the increasing aggressiveness and international confidence of Japanese companies. Between 1984 and 1987, the number of annual M&As undertaken by Japanese firms increased from 44 to 228. M&As in the U.S. increased from 33 in 1984 to 120 in 1987; in Europe, from 12 in 1985 to 36 in 1987; and in Australia from 1 in 1984 to "Japanese Likely to Diversify US Investment Portfolio," World Property. London: RICS Journals Ltd., April 1989, p. 21 67 "The Overwhelming Japanese," World Property. London: RICS Journals Ltd., May 1989, p. 32. - Page 47 -16 in 1987 ( ) . Most of the M&As occur with businesses similar to those conducted by the Japanese buyer; examples of which include Sony's $2 billion takeover of CBS Records, Bridgestone*s $2.6 billion acquisition of Firestone Corp and Dainippon Ink's $500 million purchase of Reichold Chemicals. Other M&As involve the diversification of the Japanese companies main line of business, such as the $2.15 billion purchase of Intercontinental Hotels from Grand Metropolitan PLC of Britain by Seibu Saison Group, a major private retailer in Japan. Both types of M&As are undertaken to further the globalization of the Japanese buyer; either by providing overseas production bases, securing strong local markets or providing access to important new technology. A second important trend is the rising movement of R&D functions offshore. Many Japanese companies believe this is the key to becoming truly global corporations. Sony, for example, currently has laboratories in West Germany, New Jersey and Britain and is planning to double its scientific staff overseas during the next two to three years (69) • The final important trend is "re-importing" and third country exporting from the industrialized countries. "Re-importing" or "Reverse Exporting" refers to the importing of production from Japanese subsidiaries abroad back to Japan. In the past, most "Rapid Increase in Japanese Overseas M&A," TOKYO Business  Today. Tokyo, January, 1989, p. 21. 69 "...As big Japanese firms attempt to really go global," The  Globe and Mail, July 3, 1989, p. B4. - Page 48 -foreign production was earmarked for local or third country markets and if re-importing was practiced, it was generally between Japan and Southeast Asia. However, in 1985, the U.S. Department noted that Japanese companies operating in the U.S. exported products worth $22.75 billion, of which $15.85 billion was destined for parent companies in Japan (70) . Honda Motor Inc. currently exports U.S.-made cars to Japan and is considering third country exports to Europe. Ricoh Co., exports copiers from its California plant to Europe and other electronics and machinery manufacturers are considering similar plans (71) . The trends in Japanese FDI since 1984 are important harbingers of the future strategy of Japanese investment. According to Hollerman (72) , current Japanese FDI is part of a general plan to shift from an economy based on manufacturing exports to a "headquarters" economy fuelled by service account surpluses. This strategy is designed to lessen direct trade frictions between Japan and her major trading partners; as noted by Hollerman, "Japan's trade balance will turn to deficit; its service account ...to surplus. Japan's strategy will have contrived to smother its bilateral trade friction with the U.S. while promoting multilateral "Overseas Investment," Japan Economic Almanac. 1988. Tokyo, p. 42. 71 "Japan Copiers Made in U.S. May Be Hit with EC Duties," The  Asian Wall Street Journal. Tokyo, February 9, 1989, p. 4. 72 Leon Hollerman, Japan's Economic Strategy in Brazil:  Challenge for the United Statesf Lexington, Ma: Lexington Books, 1988. - Page 49 -friction between the U.S. and the new Japans." While perhaps not, as some suggest, part of a Machiavellian conspiracy, Japanese overseas investment is an important component of the nation's economic strategy. The de-industrialization of Japan, despite longterm employment fears at home, has been promoted by the government through aid packages to declining industries and worker re-training schemes. As in the 1960s and 1970s, Japan refuses to be fettered by declining industries at home, preferring to move them offshore with surprising ease and panache. This globalization process is aided by Japan's capital surplus position, which places it in an enviable position. As in the 1960s, when Japan could shop the world for the cheapest resource inputs, in the 1980s they can shop the world for the least cost production sites; the highest yielding assets; and the most auspicious markets. In addition, Japanese firms have reached a level of sophistication such that they can undertake international ventures and investments with increasing confidence. Only two forces are likely to impede this foreign "buying binge". They are: the lack of a truly international mindset and the potential for host country protectionism against Japanese FDI. The former refers to Japan's relative difficulty in dealing with non-homogeneous workforces; promoting foreign nationals to executive positions; and adapting to local cultures etc. With regard to the latter, the concentration and tide of Japanese investment has already raised demands for protectionist action in - Page 50 -the United States, Europe and Australia. In addition, there is a dawning realization that the former protectionist actions taken to limit imports from Japanese may have had the even more adverse consequence of introducing the Japanese "Trojan horse" into the indigenous economies. As noted in one Asian Wall Street Journal article, "...European industrialists must confront their worst nightmare; because of their own policies, they will soon have to compete in their own back yards with the very Japanese industrial giants they were trying to keep out" f73) . In Europe, there is growing support for the elimination of host country subsidies for foreign investment. "Competition from Japan Begins in the Back Yard," Asian  Wall Street Journal. March 4, 1989, p. 13. Page 51 -4.0. TOWARDS A THEORY OF JAPANESE FOREIGN DIRECT INVESTMENT The foregoing sections have provided a historical perspective of the developments in Japanese direct foreign investment particularly since World War II. In less than 40 years, Japan has become one of the world's most important sources of offshore direct investment, with the largest increase in FDI taking place in the last 8 years. As will be discussed in subsequent sections of this study, the rise in Japanese direct foreign investment can be attributed largely to macro-economic developments which have taken place in Japan since 1973. However, while Japan's rapid increase in offshore investment is of interest, from a global perspective the more pertinent issue may be the likely sustainability of Japanese FDI flows. The latter depends greatly on the continuation of the macro-trends alluded to above, a subject which will be addressed in section 9.0. The emphasis on macro-economic developments in this study stems largely from the inability of modern FDI theory to adequately explain either the past patterns of Japanese offshore investment or to predict its likely continuation. Thus, before addressing the study's macro-economic rationale for Japanese foreign direct investment, a brief analysis of modern FDI theories and their shortcomings in the Japanese context is provided below. Page 52 4.1. CONVENTIONAL FDI THEORY During the past 30 years, there has been a substantial increase in the literature dealing with the subject of FDI. Modern theories have tended to concentrate on firm-specific motivations for undertaking foreign investment abroad as opposed to exporting or licensing. These theories have two drawbacks in terms of their applicability to Japanese FDI. First, existing theories have been developed primarily on the basis of Western multinationals; in this context, Japanese multinationals have been regarded as either special cases or newcomers which will eventually emulate their Western counterparts. However, as noted by Ozawa, the industrial pattern of Japanese multinationals may warrant a different theoretical approach (74) . In addition, current theories focus on micro-economic explanations for FDI, attributing FDI to the particular circumstances of firms or industries. On the other hand, Japanese multinationals have historically been strongly influenced by prevailing macro-economic factors in both their own economy and in the host country. On this basis, micro-economic theories of FDI may not be the most useful for explaining Japanese offshore investment. In the following section, current theories relating to FDI and their applicability to the Japanese context are examined. In general, these theories classify FDI as either offensive or defensive strategies undertaken to support the firm's goal of profit maximization. In the past, Japanese Ozawa, Multinationalism. Japanese Style, p. 41. - Page 53 -multinationals have tended to undertake FDI for defensive reasons; to ensure continued access to raw material supplies; to circumvent foreign government-imposed trade barriers; and to protect the comparative advantages of firms operating in international markets. Increasingly, however, offensive strategies have become more and more important in explaining Japanese foreign direct investment. An examination of these FDI theories provides a useful background to the likely evolution of Japanese FDI in the future. A brief explanation of the classical model is provided as a theoretical background for modern FDI theory. 4.2. CLASSICAL THEORY The classical theory of international capital flows is derived from the Hechsher-Ohlin-Samuelson (H-O-S) factor-proportions model of international trade, modified to allow for international capital mobility. The modified version of the H-O-S model assumes perfect competition in capital markets. Under this assumption, capital flows to where it earns the highest rate of return. The differences in national rates of returns on capital occur because of differing factor proportions and prices across countries. Under the additional assumption that the marginal productivity of capital in each country is equal to the interest rate on bonds, countries with a relative abundance of capital (and therefore, a relatively lower domestic marginal productivity of capital) will buy bonds from capital-scarce countries leading to a transfer of real Page 54 -capital. Barring market imperfections, these capital flows will tend to equalize capital endowments over time and remove a basis for trade. Although useful as a possible explanation for foreign portfolio investment, classical theory fails to explain the economic rationale behind foreign direct investment. For example, if capital markets are perfectly competitive, entrepreneurs in capital scarce countries could sell bonds to investors in capital rich countries and use the proceeds to establish operating facilities at home where the rates of return are higher. Presumably, these local entrepreneurs could earn a higher rate of return than foreigners investing in their country because of their superior knowledge of the local market. This would tend to preclude direct investment in productive facilities by foreigners. However, in the 20th century, foreign direct investment has been substantial and, in the case of the United States, the world's largest cumulative foreign direct investor, has accounted for a larger proportion of international capital flows than portfolio related investments. In addition, the theory does not address why capital flows occur in both directions. If the productivity of capital is higher in the capital-scarce country then there should be no flow of capital from the capital-scarce country to the capital-rich country. The problems associated with the classical theory of international capital flows have prompted modern economists to - Page 55 -develop theories of FDI which are based on global oligopolistic competition in imperfect markets. These theories explain FDI as offensive and defensive strategies used by multinational corporations to maximize global profit positions. 4.3. OFFENSIVE FDI THEORY - THE INDUSTRIAL ORGANIZATION APPROACH Theories explaining offensive FDI emphasize industrial organization as a basis for offshore investment. This framework has been developed by Hymer, Kindleberger and Caves and postulates that FDI occurs in industries which are characterized by oligopolistic markets in both the domestic and foreign economy. FDI, on this basis, is undertaken by a few relatively large firms and yields products which are highly differentiated in terms of market acceptability or technical superiority. The theory assumes that the investing firms are able to exploit some firm-specific, competitive advantage in global markets. These competitive advantages compensate the MNC for the additional costs associated with operating in foreign markets such as lack of knowledge regarding local customs, markets, and legal matters and greater communication and control costs. Further, these advantages allow the firm to earn a higher rate of return than it would undertaking projects of similar size and risk in the home market. Because these competitive advantages generate monopoly profits, the investing firm will be more inclined to set up wholly-owned subsidiaries rather than license or export so as not to share the - Page 56 rents with local interests. Some of these competitive advantages are discussed briefly below. 4.3.1. Economies of Scale The existence of economies of scale in production, marketing, research and development, transportation and purchasing is regarded as a possible rationale for FDI. Firms undertake FDI to expand markets and therefore, spread their fixed costs over a larger volume of production. Empirical studies tend to support the need for multinationals to be large to succeed. In a comparison of US multinational and domestic manufacturing concerns, Horst found that the firm size was the only statistically significant variable. However, these studies do not determine causality, i.e., is an MNC larger and more profitable than domestic firms because it is a multinational or is it a multinational because it is large and more profitable. 4.3.2. The posession of unique skills or special expertise Other theories postulate that MNCs undertake FDI because they posess special skills which can be tranferred internationally. These special skills include managerial and marketing expertise, technological know-how and financial strengths which allow MNCs to overcome the superior knowledge of competing host country firms. Empirical studies regarding the significance of - Page 57 -managerial/marketing expertise in the foreign investment decision are limited; however, there have been numerous studies which underline the importance of R & D as a characteristic of multinational firms. These include Raymond Vernon's product cycle theory, in which MNCs are the originators of new technologies because of emphasis on R&D functions. 4.3.3. Differentiated Products According to Richard Caves, multinational firms tend to operate in marketing or research intensive industries. Because of this, multinationals are able to develop firm-specific advantages by producing or marketing differentiated products. To maximize the return on the large fixed costs associated with R&D and marketing, the multinational may want to market these differentiated products globally. Caves notes that direct foreign investment will not be used in every case to penetrate foreign markets. However, if the firm's main competitive strength lies in research, marketing, and managerial expertise rather than in any specific differentiated product, then direct foreign investment may be used to expand overall production. 4.3.4. Internalization According to Buckley, Casson and Dunning, the existence of imperfect markets and competitive advantages is not sufficient to induce foreign direct investment. Rather, the competitive - Page 58 -advantages must be firm-specific. Financial strength and economies of scale are not unique to any one firm. Some technology can be licensed, bought or copied and differentiated products can lose advantage to modified alternatives with appropriate marketing. Buckley et al. postulate that only posession of proprietary knowledge and control of human capital can generate sufficient competitve advantages to warrant FDI. In this sense, production by overseas subsidiaries is preferrable to licensing or joint ventures because the latter would jeopardize the information monopoly of the MNC and its ability to gain monopoly profits. 4.3.5 Applicability of Offensive FDI theory in Japanese FDI  Context The industrial organization approach postulated by Hymer, Kindleberger and Caves is not sufficient to explain the large majority of Japanese overseas direct investment undertaken in the past. As outlined in Section 3.0, a vast proportion of Japanese manufacturing FDI has been undertaken by small and medium scale operations in Southeast Asia and Latin America. These firms produce primarily standardized, low-technology products in which there is little opportunity for economies of scale or product differentiation. In addition, as noted by Ozawa f75) , the market imperfections which give Japanese firms advantages in these countries originates from the underdeveloped nature of the host Ozawa, Multinationalism. Japanese Style, p. 44. - Page 59 -economies rather than oligopolistic characteristics of the investing Japanese companies. This is in direct contrast to the industrial organization approach which stipulates that FDI only occurs when oligopolistic market structures for a particular industry exist in both the investing and host country. The fact that Japanese firms derive their quasi-advantages from the backwardness of the host country illuminates an important concern for many Japanese firms operating abroad. If such is the case, then the advantage of the investing Japanese concern must inevitably be transient, especially if one impact of FDI is to further the economic development of the host country. However, Ozawa notes that there is one "lasting advantage" of Japanese firms. This is the world-wide marketing networks developed by many Japanese companies. These networks can also be accessed by small/medium concerns through partnerships with the large sogo shosha. This represents a strong advantage of Japanese firms operating in underdeveloped countries where the global marketing skills of local firms is often weak. Although industrial organization theories are not applicable to a significant proportion of past Japanese overseas investment, recently, there has been an increase in Japanese investment which does fit the theoretical framework. Such firms as Sony, Toyota, Mitsubishi Heavy Industries, Nippon Steel, Matsushita and Honda operate in global oligopolistic markets and are using FDI as an offensive strategy to foster global profit maximization and the - Page 60 -evolution towards true multinational status. 4.4. DEFENSIVE FDI THEORY Some FDI theories postulate that foreign investment occurs because MNCs must protect their global profit positions from external threats. Principal motivations behind defensive FDI include: 4.4.1. Market Imperfections Created by Governments Market imperfections resulting from tariff or non-tariff barriers, preferential puchasing policies, tax incentives, capital market and exchange market controls, can lead to defensive foreign direct investment. In addition, the formation of protectionist, multi-nation trading blocs, such as the the European Economic Community, can also induce FDI by non-member countries. Nevertheless, government policies leading to the creation of protected markets will only lead to FDI if the markets are sufficiently large or protected to compensate the MNC for the additional costs of operating abroad. 4.4.2. New Markets MNCs may reach a stage where opportunities at home are limited. Home markets may be characterized by product saturation, intensive competition, or changing consumer tastes. The firm is faced with the alternative of expanding into new products at home - Page 61 -or seeking new foreign markets for the existing product. Penetrating foreign markets may be preferrable because of the firm's built-up experience in producing and marketing the product. If transportation costs for the product are high then FDI may be the most appropriate method of accessing local markets. Other incentives for off-shore investment in search of new markets include following domestic clients abroad. This is particularly relevant to the service industries such as banking, insurance, and finance which have undertaken FDI to counter efforts by foreign service firms to attract the business of domestic clients. 4.4.3. Access to New Technologies An increasingly important motivation for FDI is to access technology in foreign countries, especially for foreign MNCs undertaking FDI in the United States. Defensive FDI of this type usually takes the form of joint-ventures with local companies. Technology acquired from these joint venture investments can then be used to improve production processes in all global subsidiaries. 4.4.4. Risk Diversification Risk diversification theory explains FDI by MNCs as a means of reducing systematic risk. This occurs because project returns in one country will not be perfectly correlated with returns from projects in other countries due to differing economic environments. - Page 62 -By diversifying internationally, therefore, a firm can reduce the volatility of its cash flows. Stable cash flows reduce the need to set aside funds in liquid, lower-yielding bank accounts to cover unexpected shortages; allow for uninterrupted dividend payments; and enable the firm to enjoy a lower cost of funds from shareholders or banks. 4.4.5. Product Cycle Theory Raymond Vernon's product cycle theory is perhaps one of the most recognized theories describing defensive FDI. According to Vernon, FDI is a natural stage in the life cycle of a product. Vernon postulates that oligopolistic competition, economies of scale and other imperfections in product and factor markets lead firms in advanced countries to undertake extensive R&D, leading to the creation of new, technologically-advanced gOods. These goods are first introduced into the home market because close coordination of the production and marketing divisions of the firm is required. After a short time lag, the product is then exported. As it reaches maturity, competition from near similar products will reduce the profit margin of the product both in home and export markets. In order to maintain its profit margins, the firm will undertake FDI to establish manufacturing facilities in areas which provide the lowest unit cost of production. Thus, defensive FDI occcurs because of the need to protect profit margins. - Page 63 -4.4.6. Follow the Leader A final theory describing defensive FDI is Frederick Knickerbocker's follow the leader model of foreign investment. Knickerbocker contends that when one firm in an oligopolistic industry undertakes a FDI, other firms in the same industry make defensive FDIs in the same market. These investments are undertaken to prevent competing firms from enjoying competitive advantages such as economies of scale through FDI. 4.4.7. Applicability of Defensive FDI models in the Japanese  Context Defensive FDI models can be used to explain particular cases of past Japanese foreign investment. Japanese firms have undertaken FDI in the past to circumvent government imposed trade barriers; to develop new markets; and to access foreign technologies and strategic resources. In addition, Japanese manufacturing investment by small/medium scale firms in Southeast Asia and Latin America has also exhibited a follow the leader pattern of investment. However, unlike Knickerbocker's theory, these firms have tended to produce low-technology, standardized products and operate in industries which are not oligopolistic. A modified version of the product cycle theory may be more appropriate for explaining Japanese manufacturing FDI. The product cycle theory is based on the investing firm developing new technologies which are successfully introduced at home and transplanted abroad. In contrast, Japanese industry "has not - Page 64 -introduced any significant innovations that would invite massive imitations overseas as envisaged by the model.11 (76) As importers of technologies, Japanese firms were concerned about reducing production costs, not spreading the overhead of large R&D expenditures. In the 1950s and 1960s, Japanese firms accumulated valuable experience in importing technologies and adapting them to a relatively labour intensive environment. When Japanese production costs became too high this experience was then used to transplant production to lower cost countries. In recent years, many of these technological importers have graduated from interceptors of Western technologies to innovators; in particular, Japanese firms have developed efficient, highly productive manufacturing processes which can be transplanted to relatively capital intensive regions like North America and Europe. Others are now overcoming technology trade barriers by investing in joint ventures in the United States to gain access to new technologies. Thus, in the future, the product cycle theory may be more relevant to Japanese firms as innovators rather than importer-modifiers of old technologies. 4.5. A MODEL OF JAPANESE FDI Both offensive and defensive based models of FDI fail to explain adequately the rise in Japanese FDI since the 1970s and, in particular, the dramatic growth since 1984. This failure occurs Ozawa, Multinationalism. Japanese Style, p. 52. - Page 65 -because of the concentration on firm-specific motivations for undertaking FDI. Conversely, Japanese economists, such as Kojima and Ozawa, have tended to attribute past Japanese FDI to macro-economic developments at home and abroad. These developments include the dependency of Japan on foreign markets both as outlets for her exports and as important sources of vital productive inputs; the rising costs of industrialization at home including massive factor price increases and environmental pollution; increasing protectionism in foreign markets; and the desires of developing nations to attract foreign capital to further economic development. Macro-economic factors are suggested as the reason for offshore investment by Japanese firms which exhibit few of the characterisitics of true multinationals. Yet, as noted by Ozawa C77) , "...ironically, from a macroeconomic viewpoint, they are judged both appropriate and ready to be transplanted overseas." Thus, in order to explain the change in the level and direction of Japanese FDI flows since the 1970s, an understanding of the macro-economic factors stimulating offshore investment is required. An examination of macro-economic developments in Japan indicates that the increase in FDI on a national basis can be largely attributed to the economic slowdown after 1973 and the inability of the domestic economy to absorb the surplus of savings in the private sector during the 1980s. These issues will be discussed fully in Section 5.0. Ozawa, Multinationalism. Japanese Style, p. 40. - Page 66 -5.0. MACROECONOMIC BALANCES OF THE JAPANESE ECONOMY Japanese net longterm capital investment [net foreign direct investment (NFDI) plus net foreign portfolio investment (NFPI)] has experienced dramatic growth during the past few years, rising from U.S.$17.7 billion in 1983 to U.S.$121.3 billion in 1988, with a peak of U.S. $144.7 billion in 1986. To understand the sourcing of these funds, and to judge the ongoing sustainability of this magnitude of investment, it is necessary to examine the underlying macroeconomic features of the Japanese economy. This examination will also provide a perspective from which to view the statistical analysis contained in section 8.0. of this report. The macro-economic features of a national economy can be summarized by an accounting identity which defines the nation's fundamental domestic and external balances. The identity also provides a useful perspective from which Japanese FDI in the post 1973 era can be examined. This perspective is clouded by two characteristics of identities; namely, they do not provide quantitative precision, and they do not explain causality. The first difficulty is not important within the context of this paper, since our approach will, in any event, involve certain simplifying assumptions. On the other hand, causality is important to the conclusions being sought. However, as long as the discussion occurs within the context of what actually happened in the Japanese economy after 1973, reasonable inferences can be made about inter-factor dependencies with respect to the direction of causality. - Page 67 -The macro-economic accounting identity can be derived as follows. Gross national product which is the sum of all goods and services produced in an economy is made up of private-sector consumption (C) and investment (I); government spending (G); and the difference between the exports of goods and services produced in an economy (X) and the imports of goods and services produced in other countries (M). This can be stated as: GNP can also be defined in terms of how the total income produced in an economy is spent, or: GNP =C+I+G+ (X-M) (1) GNP = C + S + T (2) where: C = total income consumed 8 = total income saved, and T = total income paid to the government in taxes. Equating (1) and (2): C + I + G + (X-M) = C + S + T Which may be reduced to: - Page 68 -(S-I) = (G-T) + (X-M) (3) According to (3) , surpluses or deficits between private sector savings and investment (S-I) must be offset by surpluses or deficits in the government sector- (G-T), or in the current account balance (X-M). Any excess of investment over savings (i.e. S-I < 0), must be offset by either a government sector surplus (i.e. G-T < 0) , or a current account deficit (i.e. X-M < 0) , or a combination of both. Similarly, a private sector saving surplus (i.e. S-I > 0) must be balanced by either a government sector deficit (i.e. G-T > 0), or a current account surplus (i.e. X-M > 0), or both. A second macro-economic concept of importance in the following discussion is the Balance of Payments (BOP). The latter must always sum to zero; what exits the country must equal what enters. As such, although subsets of the BOP accounts can be in surplus or deficit, the BOP must balance. Under the Japanese accounting framework, the Balance of Payments is determined by the sequence outlined in Figure 1. For the BOP to balance, the current account balance must, necessarily, be offset by the capital account balance which equals net long term and short term capital, errors and omissions, and official and private monetary movements. This can be illustrated by an example. If Japan has a current account surplus it means that Japan is paying out less foreign exchange for its imports than - Page 69 -it was receiving for its exports. This excess of foreign currency can be used in either of two ways: (1) Japanese corporations, individuals or public interests could purchase interest bearing foreign debt instruments such as foreign equities or bonds; or (2) they could hold the excess foreign currency which is a non-interest bearing debt instrument of foreign governments. - Page 70 -Less Exports (f.o.b.) Imports (f.o.b.) FIGURE 1 BOP Schematic Equals Merchandise Trade Balance Plus Service Exports Plus Service Imports Plus Unilateral Transfers Equals Current Account Balance Plus Longterm Capital Balance Equals Basic Balance Plus Short Term Capital Balance Plus Errors & Omissions Equals Overall Balance Minus Official Monetary Movements Minus Private Monetary Movements Equals • ZERO • - Page 71 -In both cases, the transactions would equal a capital outflow; the first on either the long term or short term capital balance, depending on the duration of the debt instrument; and the second on the private monetary movements. The argument works in reverse for current account deficits; the shortage of foreign currency can be made up either by selling domestic equities, bonds etc. to foreigners (representing a capital inflow on the long or short term balance) or by foreigners holding the debtor nation's currency. From the above, it is clear that the capital account is merely the flipside of the current account balance. This can be expressed as: (X-M) = Lc + SC + E + F (4) where: Lc = Net Long term capital Sc = Net Short term capital E = Errors & Omissions P = Official and Private Monetary Movements Errors and omissions (E) represent a balancing item which brings the BOP to zero after the effects of current account balance, the long term and short term capital balances, and the official and private monetary movements are considered. Because E is a random quantity, used only to balance (X-M) and (Lc+Sc+F), it cannot be considered a causal factor determining the level of - Page 72 -NFDI and has, therefore, been omitted from our discussion. Therefore: (X-M) = Lc + Sc + F and since, based on the discussion in section 2.0, Lc is equal to Net Foreign Portfolio Investment (NFPI) and Net Foreign Direct Investment (NFDI), then: (X-M) = NFPI + NFDI + SC + F (5) Substituting (5) into equation (3), gives the expression: NFDI + NFPI + SC + F = (8-1) - (G-T) (6) According to this equation, total net external investments are equal to the private sector savings balance minus the government sector balance. The excess monies available in the internal sector (either because of a private sector surplus or government surplus or both) are available for net direct, net portfolio and net short term capital investments plus official and private monetary movements. Presumably, if we can identify the factors which generate excess capital in the domestic economy and the factors which determine the distribution of this excess capital among the various external investment options, we can develop some - Page 73 -understanding of the forces determining the level of NFDI. Using identity (6) as a base, the next two sections will examine the internal [(S-I) and (G-T)] and external [NFDI, NFPI, Sc and F] macroeconomic components of the Japanese economy. - Page 74 -6.0. INTERNAL MACROECONOMIC COMPONENTS OF THE JAPANESE ECONOMY Since 1973, Japan has been generating large surplus savings in the private sector which, until the late 1970s, were more or less absorbed by government deficits. With the implementation of fiscal austerity programs after 1978, government deficits no longer fully absorbed the excess savings in the private sector and the surplus was increasingly channelled offshore in long and short term capital investments. In order to determine whether or not this situation can continue to persist, it is important to examine the factors which have been causing the private sector surpluses and the evolution of Japanese fiscal policy since 1973. The sustainability of the post 1973 trends will determine the excess capital available for future offshore investment. 6.1. THE PRIVATE SECTOR BALANCE The 1973 oil crisis was a watershed not only for the Japanese economy as a whole but as a force for generating large amounts of surplus capital in the private sector. Japan's high growth rates had begun to decline from 1970 onwards but it was the 1973 crisis which finally terminated the long cycle of economic growth. In 1974, the economy declined 0.4% in real terms, thus registering its first real negative growth since the end of the war. After 1974, economic growth reflected the maturity of the economy, with an average real growth rate of 3.8% per year between 1974 and 1985. - Page 75 -This economic slowdown had significant impact on the private sector savings and investment patterns and was the single most important factor contributing to the rapid increase in Japanese foreign investment after 1980. Table 6-1 shows the breakdown of total private savings and investment as a percentage of GNP between 1960 and 1987 for both the household and corporate sectors. The pattern of savings/investment balances can be conveniently separated into three time periods: that is, from 1960 to 1964; from 1965 to 1974; and from 1975 onwards. During the first period, private sector investment tended to exceed savings, sometimes by large margins. In 1961 the net deficit peaked at the equivalent of 7.4% of GNP. Between 1960 and 1964, this average deficit was equal to about 3.2% of GNP. The deficit was financed from two sources; an excess of government savings over investment, which averaged about 1.6% of GNP; and small current account deficits. The private sector deficits reflected large increases in capital formation during the 1960s with annual corporate investments between 1960 and 1964 averaging about 24.4% of GNP. During the second period, between 1965 and 1974, private sector savings and investment remained more or less in balance; the former averaged about 31.23% of GNP and the latter about 31.27% of GNP. In relative terms both the average annual corporate investment and savings declined marginally during this period (investment dropped to 23.8% of GNP from 24.4% in 1960-1964 and - Page 76 -TABLE 6.1: THE JAPANESE PRIVATE SECTOR SAVINGS-INVESTMENT BALANCE, F.Y. 1960-1987 PERCENTAGE OF 6NP CORPORATE HOUSEHOLDS TOTAL PRIVATE SECTOR rEAR SAVINGS INVESTMENT BALANCE SAVINGS INVESTMENT BALANCE SAVIN6S INVESTMENT BALANCE 1960 16.5 24.6 -8.1 11.8 6.6 5.2 28.3 31.2 -2.9 1961 16.5 29.9 -13.4 12.0 6.1 5.9 28.5 36.0 -7.5 1962 15.6 21.8 -6.2 12.3 6.7 5.6 27.9 28.5 -0.6 1963 15.9 23.7, -7.8 11.6 7.1 4.5 27.5 30.8 -3.3 1964 16.0 21.9 -5.9 11.2 7.3 3.9 27.2 29.2 -2.0 1965 15.8 19.1 -3.3 11.6 8.4 3.2 27.4 27.5 -0.1 1966 17.5 20.4 -2.9 11.9 8.3 3.6 29.4 28.7 0.7 1967 18.4 23.8 -5.4 13.0 8.9 4.1 31.4 32.7 -1.3 1968 19.7 24.1 -4.4 13.2 8.9 4.3 32.9 33.0 -0.1 1969 19.8 25.3 -5.5 12.7 9.2 3.5 32.5 34.5 -2.0 1970 18.7 27.5 -8.8 14.6 6.5 8.1 33.3 34.0 -0.7 1971 16.1 24.8 -8.7 14.9 5.2 9.7 31.0 30.0 1.0 1972 16.9 23.9 -7.0 15.1 5.3 9.8 32.0 29.2 2.8 1973 15.4 26.2 -10.8 16.9 5.4 11.5 32.3 31.6 0.7 1974 10.3 23.1 -12.8 19.7 8.3 11.4 30.0 31.4 -1.4 1975 8.4 17.9 -9.5 20.6 8.9 11.7 29.0 26.8 2.2 1976 9.1 16.2 -7.1 21.3 9.8 11.5 30.4 26.0 4.4 1977 9.5 15.2 -5.7 20.0 9.4 10.6 29.5 24.6 4.9 1978 11.4 13.9 -2.5 19.4 10.0 9.4 30.8 23.9 6.9 1979 11.7 16.1 -4.4 17.3 9.2 8.1 29.0 25.3 3.7 1980 11.3 17.0 -5.7 17.1 8.1 9.0 28.4 25.1 3.3 1981 10.6 16.9 -6.3 17.3 7.3 10.0 27.9 24.2 3.7 1982 11.2 16.2 -5.0 16.0 7.1 8.9 27.2 23.3 3.9 1983 11.0 15.3 -4.3 15.9 6.6 9.3 26.9 21.9 5.0 1984 11.4 15.9 -4.5 15.3 6.4 8.9 26.7 22.3 4.4 1985 11.8 17.2 -5.4 15.0 5.6 9.4 26.8 22.8 4.0 1986 - - - - - - -. - 5.8 1987 - - - - - - - - 4.5 - Page 76a -savings dropped from 16.1% of GNP to about 15.8%). The major factor behind the balancing in the private sector was the surge in surplus savings in the household sector which averaged approximately 7% of GNP throughout the period. Thus the household sector provided an important source of investment funds for the corporate sector during the high growth era. Finally, from 1975 onwards, private savings has exceeded private investment by a substantial margin; with an average annual savings surplus of 4.2% of GNP between 1975 and 1985, and a peak of 6.8% in 1978. GNP growth rates were halved in the latter part of the 1970s but this had little impact on the ratio of private savings to GNP, which exhibited only marginal declines. During the period the average savings rate was about 28.4% of GNP, compared to around 31.2% of GNP between 1965 and 1974. Private investment, however, fell substantially, dropping from an average of 31.3% between 1965 and 1974 to only 24.4% after 1974. An understanding of the reasons for the change from private sector savings deficit to surplus is crucial to any assessment of the sustainability of the current level of Japanese FDI. If the surpluses are maintained, the Japanese can, if they so choose, continue to invest heavily in foreign assets; with the choice being between direct and portfolio investments. To gain such an understanding, we will examine both components of the private sector balance, i.e., private sector savings and private sector investment. - Page 77 -6.1.1. Private Sector Savings Table 6.1 shows that the two components of private sector savings, i.e., households and corporations, demonstrated markedly different responses to the post-1973 economic slowdown. Corporate profits fell dramatically in the period of economic stagnation which followed the 1973 crisis. This caused corporate sector savings to drop to an average of 10.7% of GNP after 1974, down from the 17.9% average between 1965 and 1974. Conversely, household savings increased sharply after 1973; rising from an average of 13% of GNP between 1960 and 1973 to over 20% between 1974 and 1978. Thereafter, the rate declined, averaging about 16% from 1979 to 1985. Contrary to conventional theory, the high inflation of 1973-75 and the accompanying recession caused the acceleration of Japanese household savings rates during the 1974 to 1978 period. The rapid inflation resulted in substantial losses in the real value of accumulated savings in the household sector. Bank time deposits make up the largest proportion of financial assets; and the interest rates on these instruments were, and remain, controlled at very low levels by the Ministry of Finance (MOF). As indicated in Table 6.2, negative real interest rates on savings deposits continued until 1978, with even the highest yield - Page 78 -TABLE 6-2: INFLATION RATES AND TIME DEPOSIT RATES BY TYPE OF INSTITUTION, 1969 TO 1981. PERCENT POSTAL SAVINGS COrMERCIAL BANKS TRUST BANKS MAX I HUM niNinufi HAXIHUt1 niNinun MAXIHUtt niNinun INFLATION ) 3 YEARS LESS THAN M ) 3 YEARS *1 LESS THAN *i : ) 3 YEARS M LESS THAN M (EAR RATES 1 YEAR 1 YEAR 1 YEAR 1969 5.2 5.50 4.20 5.50 4.00 7.27 5.50 1970 7.7 5. 75 (4) 4.25 (4) 5.75 (4) - 7.47 (3) 5.75 (4) 1971 6.1 6.00 (2) - 6.00 (2) - 7.17 (9) -1972 4.5 - - -• - 7.12 (4) -5.50 (8) 4.00 (8) 5.50 (8) 3.75 (7) 6.82 (7) 5.25 (7) 1973 11.7 6.00 (4) 4.25 (4) 6.00 (4) 4.00 (4) 7.12 (4) 5.75 (4) 6.50 (7) • - ' 6.50 (7) - 7.42 (7) 6.00 (7) 6.75 (10) 4.50 (10) 6.75 (10) . A-25 (10) 7.72 (9) 6.25 (9) • - - - - 8.52 (12) -1974 24.5 7.50 (1) 5.25 (1) 7.50 (1) 5.25 (1) - 7.25 (1) £.00 (9) 6.00 (9) 8.00 (9) 5.50 (9) 9.02 (9) 7.75 (9) 1975 11.8 7.00 (11) 5.00 (11) 7.00 (11) 4.50 (11) 8.82 (8) 6.75 (11) - - - 8.32 (11) -1976 9.3 ': - - - - - -1977 8.1 6.00 (5) 4.25 (S) 6.00 (5) 3.75 (5) 7.52 (5) 5.75 (5) 5.50 (9) 3.75 (9) 5.50 (9) 3.25 (9) 6.72 (9) 5.25 (9) 1978 3.8 4.75 (4). 3.00 (4) 4.75 (4) 2.50 (4) 6.22 (4) 4.50 (3) 1979 3.6 5. SO (5) 3.75 (5) 5.50 (5) 3.25 (5) 6.82 (5) 5.25 (5) 6.25 (8) 4.50 (8) 6.25 (8) 4.00 (8) 7.32 (8) 6.00 (8) 1980 8 7.25 (3) 5.50 (3) 7.25 (3) 5.00 (3) 7.92 (2) 7.00 (3) 8.00 (4) 6.50 (4) 8.00 (4) 6.00 (4) 8.62 (3) 7.75 (4) r 7.25 (12) 5.50 (12) 7.25 (12) 5.00 (12) . 7.92 (12) 7.00 (3) 1981 4.9 6.50 (4) 4.75 (4) 6.50 (4) 4.25 (4) 7.62 (5) 6.25 (4) Footnotes: *1 Figures in brackets refer to ionth in which interest rates were changed. Source: Bank of Japan, Econoiic Statistics Annual, 1985, p. 177-78; 1979, p. 73-74. - Page 78a -instruments losing 30% of their value between 1973 and 1978 ( ) . According to Lincoln, the decline in the real value of savings prompted Japanese households to temporarily increase savings rates to recover these losses. This portfolio refurbishing, and the general disinclination to consume triggered by the uncertainty following the 1973 oil crisis, served as a powerful stimulus for increased savings after 1975. However, this was a temporary phenomenon and Japanese household savings rates as a percentage of GNP have been falling gradually since 1978. Nevertheless, current household savings rates are still higher than the average of the pre-1973 period and exceed those of all OECD nations except Italy (re) . (Even this exception may be somewhat illusory, since the Italian rate is greatly distorted by the existence of a large "underground", and undeclared, sector of the Italian economy.) In absolute terms, the 1985 level of household savings (nominal) was more than 336% that of 1972; even after applying the GNP deflator, the 1985 available pool of household savings was 66% higher than that of 1972. The factors behind the high household savings rate in Japan have been a subject of lengthy debate with no clear consensus. Two principal reasons put forward are uncertain retirement incomes Edward J. Lincoln, Japan: Facing Economic Maturity. Washington: The Brookings Institution, 1988, p. 152. 79 Atsushi Maki, "Why is the Japanese Household Savings Rate So High?," Keio Business Review. No.24. 1987. Tokyo: Keio University Press, 1987, p. 2. Page 79 which, in turn, is linked to inadequate social security pensions and extended longevities, and the high cost of housing. Both of these reasons are highlighted below. 6.1.1.1. The Need to Provide for Retirement Insufficient social security pensions have often been cited as a principal factor behind high Japanese savings rates. The argument was more applicable during the 1950s and 1960s than it is today. Social security contributions as a percentage of GNP in Japan are now on a par with the U*S., the U.K. and Canada, although they are substantially below the levels in France, Sweden, Italy and West Germany (80) . However, the downward pressure of improved social security benefits on the savings rate has been negated by two factors. First, the mandatory retirement age for most Japanese companies is 55, although in recent years many companies have • 81 raised the age to 60 ( ) . Second, the average longevity in Japanese is the highest in the world with 1986 figures of 75 and • 82 81 years for men and women, respectively ( ). These two factors Maki, "Why is the Japanese Household Savings Rate so High?," Keio Business Review, p. 3. As a percentage of GNP, social security contributions are 18.2% for France; 16.2% for Italy; 13.8% for Sweden; 12% for West Germany; 7.8% for Japan; 6.5% for the U.S. and U.K.; and 5.6% for Canada. 81 Kazuo Sato, "Savings and Investment," The Political Economy  of Japan: Volume 1 - The Domestic Transformation, edited by Kozo Yamamura and Yasukichi Yasuba, Stanford: Stanford University Press, 1987, p.163. 82 Japan 1988. An International Comparison, Tokyo: Keizai Koho Centre, 1988, p. 8. - Page 80 -have lengthened the period which must be covered by savings generated in the principal working years. Although the average labour force participation by males over 65 is high by Western standards (see Table 6.3), Sato notes that the growing inverted pyramid structure of Japanese demographics will inevitably result in a larger number of people competing for the limited number of jobs available to the over 60 age-category of workers (83) . The rapid aging of Japan is also expected to place a greater burden on the working population to support an increasing number of retired workers. As noted by Sato f84) , "The social security burden on workers will rise to an intolerable level if the present benefit level is maintained. A debasement in the benefit structure is inevitable if the social security system is to remain solvent." These views are based on a classical concept of age-capability relationships and do not take into account the changing nature of technology and related labour demands. Nevertheless, the perception of a need to provide for a longer post-employment period may tend to support continued high savings rates in Japan. The contrary argument has also been advanced; namely, that the rapid aging of Japan will result in a gradual decline in the savings rates. This view holds that retired or elderly workers are dis-savers and that, as the proportion of elderly to the Sato, "Savings and Investment", The Political Economy of  Japan. Volume 1. p. 163. 84 Ibid. . p. 164. - Page 81 -TABLE 6-3: LIFE EXPECTANCY AND LABOUR PARTICIPATION RATIOS OF SELECTED COUNTRIES LABOUR FORCE PARTICIPATION RATIO LIFE 1982 (PERCENT! 1 EXPECTANCY COUNTRY (YEARS) HALE FEMALE HALE 65+ YEARS JAPAN 73.8 (1981) 89.2 55.9 38.3 UNITED STATES 69.9 (1979) 88.1 63.1 : 17.8 CANADA 70.2 (1977) 86.2 59.7 13.8 FRANCE 70.1 (1980) 79.8 56.0 5.9 GERMANY 69.9 (1981) 80.6 49.6 6.0 ITALY 69.7 (1977) 81.3 .40.0 11.7 UNITED KINGDOM 70.4 (1978) 89.2 58.1 7.8 Source: Kazuo Sato, 'Savings and Investment" The Political Economy of Japan, Vol 1, p. U7. - Page 81a -general working population increases, the proportion of households which are either not adding to, or drawing down, their savings stocks will also rise. The case for this view is by no means proven. As people age, their purchasing needs (especially for "big ticket" items such as houses, cars, etc.) may be significantly reduced. While affluent seniors in North America are a prime target for the marketing of expensive goods and services, Japanese seniors, used to a more frugal life style, may not follow the same pattern. Thus, with the probability of extended working life spans, older Japanese may continue to do what they have done throughout their adult lives; i.e., save. In addition, Hayashi (85) , notes that there is little cross-sectional indication of wealth being run-down during the retirement period. Helliwell ( ) , suggests that this may be partly explained by the importance of bequests and inter-vivos gifts in the extended family system in Japan. In view of this, the downward pressure on savings of the changing demographic structure of Japan may, at least, be offset to a considerable degree. The Economic Planning Agency, in its official long term forecast, has estimated the size of household Fumio Hayashi, "Why is Japan1s Savings Rate So Apparently High?", NBER Macroeconomics Annual. 1986. Cambridge, Ma.: National Bureau of Economic Research, 1986, p. 196-197. 86 John F. Helliwell, "Some Comparitive Macroeconomics of the United States, Japan, and Canada," Discussion Paper No. 87-04, University of British Columbia Department of Economics, Vancouver, March 1987, p. 10. - Page 82 -savings in the year 2000 with and without demographic changes. Their findings indicate that the average savings rate, as a ratio of GNP, may decline by only 2 percentage points if demographic • 87 changes are factored in ( ) . 6.1.1.2. The High Cost of Housing The target-motive of owning a home and the high cost of housing relative to household income have been considered important stimuli for high household savings rates. The Japanese owner-occupancy rate has been remarkably stable for many years and, at a level of about 60 percent, is comparable to the United States where the cost of home ownership is considerably less f88) . Further, despite the massive concentration of the population in the six largest urban areas of Tokyo, Yokohama/Kanagawa, Osaka, Nagoya, Kyoto and Kobe, the trend towards urban, rather than suburban living continues. Illustrative of this trend is the growing concentration of multiple unit structures in high density urban areas (representing 38.4% of total housing in these areas versus 26.9% for the nation as a whole) where such units have doubled in • 89 number during the last 20 years ( ) . This concentration of population has placed strong pressure on the available stock of Lincoln, Japan: Facing Economic Maturity, p. 79. 88 Maki, "Why is the Japanese Household Savings Rate So High?," p. 4. 89 Ibid. . p. 4. - Page 83 -usable residential land and prices have exhibited massive increases, particularly in the six largest urban areas listed above. Table 6.4 shows the residential land price indexes for the nation, the six largest cities, and all other districts. For the nation as a whole, average residential land prices have been growing at 13% per year since 1960. The same figure for the 6 largest cities and all other districts is 14% and 12%, respectively. However, since 1985, residential land prices have been increasing at 18% per year. As indicated in Table 6.5, the average price of residential land in Tokyo has, on average, risen at 24 percent per year between 1980 and 1987, and 73 percent between 1985 and 1987. Residential land prices in Kanagawa (which comprises Yokohama and Kawasaki city) to the south of Tokyo, have also realized substantial gains, rising at an annual average of 17% between 1980 and 1987, and 32% between 1985 and 1987. These two areas account for almost 15 million people, or roughly 12% of the total population of Japan. Rising land prices since 1960, especially in the urban areas, have led to greater proportions of household income being allocated to savings for down payments. A 1983 survey indicated that the Japanese average acquisition price for a new home was 5 times the annual average household income with 39% of - Page 84 -TABLE 6.4: RESIDENTIAL LAND PRICE INDICES, JAPAN, F.Y. 1960-1987 COMPOUNDED ANNUAL RATE OF 6R0WTH ALL OTHER SIX LARGEST ALL OTHER SIX LARGEST fEAR AVERA6E DISTRICTS CITIES PERIOD AVERAGE DISTRICTS CITIES 1960 6.1 6.9 5.2 1960-1965 24.2 20.9 28.0 1961 8.5 9.5 7.5 1962 11.3 12.0 10.6 1963 13.5 13.8 13.1 1964 15.8 15.6 16.0 1965 17.9 17.8 17.9 1965-1970 13.5 14.9 12.0 1966 18.8 19.0 18.5 1967 20.4 21.0 19.7 1968 23.4 24.5 22.2 1969 27.8 29.3 26.2 1970 33.6 35.7 31.5 1970-1975 16.1 16.2 15.9 1971 39.9 42.3 37.4 1972 45.8 48.4 43.1 1973 60.9 62.2 59.5 1974 75.0 78.7 71.3 1975 70.8 75.7 65.9 1975-1980 7.2 5.7 8.7 1976 72.0 76.8 67.2 1977 75.1 79.8 70.3 1978 78.8 83.6 74.0 1979 86.5 89.6 83.3 1980 100.0 100.0 100.0 1980-1985 6.3 6.6 6.0 1981 111.3 112.0 110.6 1982 120.3 122.6 117.9 1983 126.1 129.7 122.5 1984 130.6 134.4 126.8 1985 135.9 137.9 133.8 1985-1987 10.0 2.7 17.0 1986 143.6 140.4 146.7 1987 164.4 145.4 183.3 1960-1987 12.5 11.5 13.6 - Page 84a -TABLE 6.5: AVERA6E PRICES OF RESIDENTIAL LAND OF SIX CITIES, PER ONE METRE SQUARED, 1000 YEN LARGEST YEAR TOKYO OSAKA KANA6AHA 1980 197 96 92 1981 225 112 106 1982 249 128 126 1983 268 150 157 1984 279 157 160 1985 297 165 163 1986 431 170 170 1987 890 184 283 SAITAMA KYOTO NAGOYA AVERAGE 77 71 53 98 99 80 63 114 115 110 76 134 127 132 92 154 128 140 96 160 129 145 97 166 121 123 94 185 158 128 98 290 RATE OF INCREASE 1980-1987 24.0 9.7 17.4 10.8 8.8 9.2 16.8 1985-1987 73.1 5.6 31.8 10.7 -6.0 0.5 32.2 Source: Ministry of Finance, Japan Statistical Yearbook, 1980-1987. - Page 84b -the purchase price being self financed ( ) . In Tokyo, the 1983 ratio was 7.9 times annual income. Savings rates have had to accommodate to these greater equity demands. In the future, it is uncertain what the impact of the current level of housing prices will be on savings rates. The phenomenal increase in land prices since 1985, especially in the greater Tokyo area, has led to rising frustration among young urban workers who face growing obstacles to owning a home in the future. Indeed, the rise in land prices has sponsored a sharp demarcation line between landowners and the rest of Japan. As noted by one writer: ...Japan, which has prided itself on its egalitarian society, is suddenly becoming a nation of haves and have-nots. The haves are those with land; the have-nots, those without it. (91) A 1987 survey by the Ministry of Health and Welfare indicated that nearly 50% of the nation's 38 million households are having trouble making ends meet despite average annual incomes of Canadian $50,000 (92) . Although mortgage interest rates are low (around 6%) , the average salaried worker can expect to take a lifetime to pay off his mortgage; with a good chance that his children may have to finish the job. The high cost and increasing difficulty of home-ownership Sato, "Savings and Investment," The Political Economy of  Japan. Volume 1. p. 608. 91 Edith Terry, "Why the Japanese Can't Relax," The Globe and  Mail Report on Business. August 1989, p. 23. 92 Angela Bianchi, "Home Ownership an Elusive Dream for Most," The Financial Post. July 24, 1989, p. 26. - Page 85 -could be a crucial influence on the level of future Japanese savings rates. If potential buyers abandon the goal of home-ownership, monies which would have previously been set-aside for down payments may be diverted to other consumption purposes. This could result in a fall in the domestic savings rate. On the other hand, if the goal of home-ownership persists as a general objective, an even greater proportion of household income may have to be saved for this purpose, thus leading to an increase in the household savings rate. In summary, the principal factors influencing the generally sustained high rates of Japanese household savings rates appear to be a continued concern for post-retirement welfare and rising housing costs. Savings rates reached their highest levels just after the 1973 oil shock, fuelled by a general uncertainty as to the future well-being of the economy and Japanese living standards. Since 1985, household savings rates have exhibited only marginal declines despite very significant increases in housing costs. This suggests that either house ownership (and, hence, a need to save for down payments) remains as a general goal, or that there has been an off-setting trend towards reduced confidence in the future. The latter hypothesis is not supported by the findings of a recent survey undertaken by the Bank of Japan. The survey indicated that business-confidence is at an all-time high (93) . If Ian Rodger, "Many Pressures Strain New Era," The Financial  Post. July 24, 1989, p. 24. - Page 86 -Japanese households share this confidence, it could be expected that consumption would be stimulated and savings reduced. Indeed, despite problems relating to home-ownership and generally high living costs (the average Japanese household spends more than $1500 per month on basic food supplies), a public opinion poll in 1988 indicated that more than 67% of those polled were either "fully" or "generally" satisfied with their way of life (94) . Thus the marginal decreases (relative to GNP) in household savings rates during recent years could be interpreted as being the net result of decreases due to higher public confidence in the future being largely off-set by increased savings to provide for the higher down-payments needed to buy housing in today's inflated market. 6.1.2. Private Sector Investment The second contributor to the large private sector surpluses after 1974 was the substantial decline in corporate investment. The latter dropped to an average of 16.2% of GNP between 1975 and 1985 after attaining an average rate of 23.8% between 1965 and 1974. Accompanying this decline, corporate savings also fell during the period from 16.9% of GNP between 1965 and 1974 to 10.1% between 1975 and 1985. Despite this drop, the corporate sector savings deficit narrowed from about 6.9% of GNP between 1965 and 1974 to 5.5% of GNP after 1975. (See Table 6.1). Thus, corporations were no longer absorbing the excess savings in the Terry, "Why the Japanese Can't Relax," p. 20 - Page 87 -household sector. As a result, the private sector, in aggregate, turned to surplus after 1974. A number of factors contributed to the downturn in corporate investment after 1973. It is important to note that, while oil supply and price problems were the immediate cause of slower growth, precursors of this reduced growth rate had already manifested themselves prior to 1973 and would have inevitably prevailed, even in the absence of the oil crisis. An analysis of real GNP growth rates leading up to 1973 attest to this point. Although growth rates in excess of 12 percent were achieved during the 1960s, the 1970-73 average fell to 7.5 percent(95). The precursors included escalating land costs; rising shortages of labour and, hence, rising real wages; the closing of the technology gap with the West; and more stringent environmental regulations. These factors have prevailed in the slower growth period of the 1970s and 1980s and, as such, warrant brief discussion. 6.1.2.1. Rising Industrial and Commercial Land Prices The rapid escalation of industrial land price rises in the 1960s and early 1970s increased the costs of developing new industrial and commercial sites and reduced the expected return from new operations. Between 1960 and 1970 land prices increased at phenomenal rates, averaging 20% and 25% for commercial and industrial land, respectively (see Tables 6.6 and 6.7). The rates Lincoln, Japan: Facing Economic Maturity, p. 3. - Page 88 -TABLE 6.6: COMMERCIAL LAND PRICE INDICES. JAPAN, F.Y. 1960-1987 ALL OTHER SIX LARGEST YEAR AVERAGE DISTRICTS CITIES 1960 10.9 10.5 11.2 1961 16.1 14.2 17.9 1962 20.9 17.6 24.2 1963 23.9 20.7 27.1 1964 27.2 23.2 .31.1 1965 30.1 26.4 33.8 1966 31.3 28.1 34.5 1967 34.1 31.9 36.3 1968 37.2 35.5 38.8 1969 43.1 41.7 44.4 1970 50.6 49.8 51.4 1971 56.7 56.9 56.5 1972 62.5 63.2 61.7 1973 76.8 76.5 77.0 1974 90.1 91.4 88.7 1975 85.1 88.1 82.1 1976 85.4 88.4 82.4 1977 86.7 89.4 83.9 1978 88.5 91.2 85.7 1979 92.1 93.8 90.3 1980 100.0 100.0 100.0 1981 107.5 106.9 108.1 1982 115.0 113.2 116.7 1983 121.3 117.9 124.6 1984 128.5 121.3 135.7 1985 139.1 124.5 153.6 1986 163.1 128.3 197.9 1987 191.8 136.9 246.7 COMPOUNDED ANNUAL RATE OF GROWTH ALL OTHER SIX LARGEST PERIOD AVERAGE DISTRICTS CITIES 1960-1965 22.6 20.2 24.7 1965-1970 10.9. 13.5 8.7 1970-1975 11.0 12.1 9.8 1975-1980 3.3 2.6 4.0 1980-1985 6.8 4.5 9.0 1985-1987 17.4 4.9 26.7 1960-1987 10.8 9.6 • 11.7 - Page 88a -TABLE 6.7: INDUSTRIAL LAND PRICES, JAPAN, F.Y. i960-19B7 ALL OTHER SIX LARGEST YEAR AVERAGE DISTRICTS CITIES 1960 8.8 9.7 7.9 1961 14.7 14.6 14.8 1962 20.6 18.9 22.3 1963 24.3 22.4 26.1 1964 28.5 26.6 30.4 1965 31.2 29.3 33.1 1966 32.2 30.3 34.0 1967 34.0 31.7 36.3 1968 38.0 35.1 40.9 1969 40.6 40.3 40.9 1970 47.6 47.3 47.8 1971 55.0 54.1 55.9 1972 62.3 61.8 62.7 1973 78.9 76.8 80.9 1974 95.1 94.9 95.3 1975 88.3 89.9 86.7 1976 88.7 90.4 86.9 1977 89.3 91.1 87.5 1978 90.4 92.2 88.5 1979 93.4 94.7 92.0 1980 100.0 100.0 100.0 1981 106.7 106.7 106.7 1982 112.7 112.7 112.6 1983 116.9 117.1 116.7 1984 120.5 120.4 120.6 1985 124.1 123.1 125.0 1986 128.4 125.5 131.2 1987 . 141.2 128.8 153.6 COMPOUNDED ANNUAL RATE OF GROWTH ALL OTHER SIX LARGEST PERIOD AVERA6E DISTRICTS CITIES 1960-1965 28.8 24.7 33.2 1965-1970 8.8 10.1 7.6 1970-1975 13.2 13.7 12.6 1975-1980 2.5 2.2 2.9 1980-1985 4.4 4.2 4.6 1985-1987 6.7 2.3 10.9 1960-1987 10.4 9.7 11.2 - Page 88b -of increase were even higher between 1960 and 1965, averaging 29% for industrial land (with the six largest city areas witnessing an annual growth rate of 33%), and 23% for commercial land (with the six largest urban areas realizing an annual growth rate of 25%) . This rapid escalation continued into the early 1970s, with prices for industrial land increasing by 17, 15.5, 13.3 and 26.6 percent in 1970, 1971, 1972 and 1973, respectively. The corresponding increases for commercial land were 17.4, 12, 10.2 and 22.8 percent over the same period. 6.1.2.2. Rising Real Wages In addition to escalating land costs, Japanese corporations also encountered rising real wage costs. During the previous period of rapid economic growth, much of the under-utilized labour in Japan had been absorbed, leading to tightened labour market conditions and rising real wages. Real wages grew at compounded annual rates of 6% between 1960 and 1970, 8% between 1965 and 1970, and in 1971, 1972, and 1973, they accelerated by 8, 10.4 and 9.5%, respectively (see Table 6-8). For unionized workers the 1974 "shunto" (spring wage offensive) resulted in an average, nominal wage increase of 32.9%, following an increase of 20.1% in 1973 (96) . These high wage settlements served to reinforce the inflationary effects of rapidly increasing energy prices after 1973, and worked to retard economic activity. Lincoln, Japan: Facing Economic Maturity, p. 32. - Page 89 -TABLE 6-8: REAL WAGE INDICES AND COMPOUNDED ANNUAL GROWTH RATES, F.Y. 1960-1987 REAL COMPOUNDED WAGE ANNUAL RATE YEAR INDEX PERIOD OF GROWTH 1960 38.4 1960-1970 0.06 1961 40.6 1962 41.9 1963 42.6 1960-1965 . 0.04 1964 45.1 1965 46.3 1966 48.9 1965-1970 0.08 1967 52.6 1968 56.8 1969 62.4 1970 67.8 1970-1980 0.04 1971 73.2 1972 80.8 1973 88.5 1970-1975 0.07 1974 90.5 1975 92.9 1976 95.6 1975-1980 0.01 1977 96.1 1978 98.4 1979 101.7 1980 100.0 1980-1987 0.01 1981 101.1 1982 101.8 1983 104.0 1980-1985 0.01 1984 104.8 1985 104.9 1986 107.4* 1987 109.7 1965-1987 0.02 1960-1980 0.04 - Page 89a -6.1.2.3. Reduced Productivity Growth The impact of higher wage costs was exacerbated by slower productivity gains. Rapid wage increases during the 1960s had been largely offset by improved productivity. Much of this improvement had been effected by the importation and adaptation of technology from the West. This process had enabled Japanese industry to make significant improvements in productivity at relatively low cost, without having to make corresponding expenditures on research and development. By 1973, Japan was more or less technologically equal to her Western trading partners, making it much more difficult to obtain the low cost, foreign technology which had supported previous investments. As noted by Lincoln (97) , The impact of technological equality on capital formation was straightforward: because rapid increases in productivity based on imported technology were no longer possible, corporations could no longer expect the high levels of profit from new investment to which they had become accustomed. With lower expectations, investment de-accelerated. What had once been a powerful incentive for investment quickly evaporated in the mid-1970s. 6.1.2.4. Environmental Constraints A further impediment to corporate investment after the advent of the 1970s was increasing public pressure on the central government to provide better environmental protection. The deadly Minimata mercury poisoning incident of the late 1960s and a host of other pollution-related incidents served to galvanize public Lincoln, Japan: Facing Economic Maturityf p. 44. - Page 90 -pressure for more stringent regulation of industry pollution levels and new developments. The government responded with legislation in 1970 which led to a quadrupling of public spending on pollution control between 1970 and 1973; to 430 billion yen, or $1.6 billion go t at 1973 exchange rates ( ). Industry was also forced to increase pollution control spending to comply with government directives. By 1974, total private sector expenditures on pollution control was roughly two times the level of central government spending and on a par with that of local governments. These investments in pollution abatement, while socially imperative, increased the cost per unit of plant capacity and thereby acted as a dis-incentive to investment. 6.1.2.5. The Impact of the Oil Crisis Superimposed on these other factors, the higher energy costs after 1973 posed a difficult (and for some industries, terminal) impediment to capacity expansion. Oil price increases brought a premature end to Japan's strategy of relying on heavy industry for economic growth. The increases in oil prices were of particular concern to Japan given its severe dependence on outside oil supplies. Energy-intensive industries which had grown up behind a wall of protective barriers, lost their international competitiveness and were forced to down-size or even close. The Japanese aluminum industry is a case in point; investment in Ibid.. p. 49. - Page 91 -aluminum refining had been substantial during the 1950s and 1960s because of tariff barriers and high expected profits from utilizing state of the art technology. By 1984, because of the combined oil shocks, capacity in the aluminum industry had been greatly reduced and imports represented 71% of total aluminum consumption in Japan versus 7% in 1965 (") . Lincoln, Japan; Facing Economic Maturity, p. 47. - Page 92 -6.1.2.6. The Investment Climate since 1973 The impact of the above factors on capital formation after 1973 was formidable. As noted, corporate sector investment fell substantially after 1973 in response to the combined effects of higher energy prices and the induced recession of 1975-1976. The accelerator principle meant that reduced economic growth required greatly reduced levels of corporate fixed investment. From 1966 to 1972 real GNP and corporate investment grew at rates of 9.4 and 15.0 percent, respectively. However, between 1974 and 1985, the growth of real GNP was reduced to an average of 3.8% and that of corporate investment to 3.9% (10°) . In the immediate aftermath of the 1973 oil shock, the previous high rates of investment resulted in over-capacity in most industries. This removed most of the incentive for further investment. Indeed, the average capacity utilization index (1980 = 100) fell from 106 for the period 1965 to 1973, to 96 for the period from 1974 to 1984. Corporate profits also dropped substantially, thereby restricting the capacity to make new investment. The average ratio of recurring profits (101) to net worth decreased from 25.6% between 1965 to 1972 to 21.7% between 1974 and 1984. These figures, however, should be treated with some caution. Net worth is, as noted by Lincoln, often grossly under-valued in Japan because corporations carry land Lincoln, Japan: Facing Economic Maturity, p. 81. 101 "Recurring profits" is a before-tax measurement which includes non-operating income and expenses but excludes extraordinary gains and losses. - Page 93 -values at purchase cost on their books; this overstates the profits to net worth ratio. On the other hand, the average debt to equity ratio of Japanese companies was higher during the 1974-84 period than during the 1965-72 period, so the ratio of profits to assets used may not have reflected the same decrease. The 1980s have seen a general recovery of the Japanese economy, especially since 1986. Corporate investment still remains substantially below its pre-1973 levels although in the last two years, corporate fixed investment has shown an upward trend, rising • 102 at rates in excess of 17% percent ( ). Nevertheless, the factors which led to slower capital formation in the 1970s are still largely applicable today and, in many respects, have been exacerbated by the rapid appreciation of the yen after 1985. As noted in Table 6.8, real wages have risen only marginally since 1975 but the 80% appreciation of the yen since 1985 has made Japan a relatively high-cost labour country even compared to the United States and most West European nations. The problems of commercial and, to a lesser extent, industrial land prices have also intensified since 1985. Commercial land prices, which had been relatively quiescent during the 1970s and early 1980s, increased sharply between 1985 and 1987. Commercial land prices in the six largest urban areas rose at an average compounded rate of 63% per year (see Table 6.9) . As shown in Table 102 * * Quarterly Economic Review. August 1989. Vol.19. No. 3. Nomura Research Institute, Tokyo, 1989, p. 41. - Page 94 -TABLE 6.9: AVERAGE PRICES OF COMMERCIAL LAND OF SIX CITIES, PER ONE METRE SQUARED, 1000 YEN LARGEST EAR TOKYO OSAKA KANA6AWA SAITAMA KYOTO NA60YA AVERAGE 1980 636 353 261 202 201 129 297 1981 713 399 291 251 271 145 345 1982 866 460 320 285 330 185 406 1983 1117 684 395 324 433 236 532 1984 1333 753 419 333 459 260 593 1985 1894 855 464 351 481 308 726 1986 4211 1159 628 392 460 367 1203 1987 6493 2025 1279 658 606 473 1922 RATE OF INCREASE 19B0-19B7 39.4 28.3 25.5 1B.4 17.1 20.4 30.6 1985-1987 85.2 53.9 66.0 36.9 12.2 23.9 62.8 Source: Ministry of Finance, Japan Statistical Yearbook, 1980-1987. - Page 94a -6.9, the average price of commercial land increased by 85%, 66% and 54% in Tokyo, Kanagawa, and Osaka, respectively during this 3 year period. The average cost per square metre for commercial land in Tokyo rose to 6.5 million yen in 1987, or about $45,000 U.S. Increases in industrial land prices also underscore the problems facing firms wishing to make investments in new plant locations; as prices per square metre rose by an average of 48% and 28% per annum in Tokyo and Kanagawa, respectively between 1985 and 1987 (see Table 6.10). These land price increases, in conjunction with the appreciation of the yen, have made land in the greater Tokyo area the most expensive in the world by a large margin. The high factor prices for labour and land have, since 1974, acted as powerful limits to capital investment. However, another prime reason for reduced investment has been the structural changes taking place in the Japanese economy. Recognizing the unfavorable prognosis for continued reliance on the heavy industrial sector, Japan has shifted its emphasis to an economy based on technology intensive manufactured products and service industries. This structural change has led to a very significant reduction in the investment requirements of the capital intensive heavy industries. It reflects, once again, the Japanese ability to adapt to radical adjustments in the domestic economic environment and clearly signals the maturation of the country's industrial economy. Given this re-direction of the Japanese economy, it seems reasonable to project that the stabilization of capital formation rates (at - Page 95 -TABLE 6.10: AVERA6E PRICES OF INDUSTRIAL LAND OF SIX LAR6EST CITIES, PER ONE METRE SQUARED, 1000 YEN ;AR TOKYO OSAKA KANA6AHA SAITAHA KYOTO NA60YA AVERAGE 1980 136 70 51 33 50 36 63 1981 152 79 54 45 54 39 71 1982 175 93 55 50 72 45 82 1983 208 113 85 61 102 56 104 1984 208 112 87 63 106 56 105 1985 212 124 99 67 100 58 110 1986 243 123 112 64 82 58 114 1987 462 136 162 81 89 59 165 RATE OF INCREASE 1980-1987 19.1 10.0 18.0 13.7 8.6 7.3 14.7 1985-1987 47.6 4.7 27.9 10.0 -5.7 0.9 22.5 Source: Ministry of Finance, Japan Statistical Yearbook, 1980-1987. - Page 95a -between 15 and 17% of GNP) experienced since 1975 will be continued into the future. 6.1.3. The Private Sector Surplus since 1986 Since 1986, the private sector savings surplus has narrowed, from a peak of about 32.3 trillion yen in 1986 to 17.8 trillion yen in 1987; and it is projected to fall further in F.Y. 1988 and 1989 to 15.1 and 13.3 trillion yen, respectively (103) . The narrowing of the private sector savings surplus is a direct result of Japan's increasing emphasis on domestic demand led growth, rather than export led growth. Since 1986, the Japanese economy has been one led overwhelmingly by domestic demand. In F.Y. 1986 and F.Y. 1987, domestic demand contributed 2.8 and 5.8 percentage points to real economic growth, respectively ( ). Net exports, in the same two years, contributed minus 1.4 and minus 1.0 percent, respectively. Personal consumption has been leading the domestic demand based economic growth and has contributed 1.9 and 2.5 percent to real economic growth in F.Y. 1986 and F.Y. 1987, respectively. This consumption boom has been sponsored by improved employment; higher nominal wages and bonuses; increased leisure time; and strong housing investment due to the speculative land boom of 1985-87. In 1989, household residential 103 Nomura Quarterly Economic Review Vol. 19. No. lf Tokyo: NRI & NCC Co. Ltd, February 1989 p. 43. 104 Nomura Quarterly Economic Review. Vol. 19. No.l. p. 15. - Page 96 -investment is expected to moderate somewhat, with residential housing starts easing from 1.65 million units in F.Y. 1988 to 1.53 • • • • 105 • • million units in F.Y. 1989 ( ). Despite the 3% consumption tax imposed earlier this year, massive cuts in personal income taxes amounting 3.3 trillion yen are expected to maintain personal consumption at high levels over the next year, or so; thereby contributing a projected 2.8% to real economic growth in F.Y. 1988 and F.Y. 1989 (106) . This consumption boom has led to a reversal of the trend towards lower corporate investment and there were sharp increases in such investment during 1987 and 1988. The increase in corporate investment has been due to the improved corporate profits which resulted from the expansion of domestic demand and the recovery in exports after 1987 (pre-tax profits before extraordinary items for all industries were up by 28.8% in 1988, compared to 1987); low inflation; continued low interest costs; and unexpectedly large productivity gains which have offset recent rises in real wage rates. According to the Nomura Research Institute, planned investment for F.Y. 1989 has risen to its highest level in 15 years, a projected 10.5% increase over F.Y. 1988 (107) . However, the longevity of this increased consumption cycle is not clear, since it may have been fuelled, in part, by the wealth 105 Ibid. . p. 15. 106 Ibid. . p., 15. 107 • • Nomura Quarterly Economic Review. Vol. 19. No. 1. p. 15 - Page 97 -effect of increased stock prices and, especially, of soaring land • 108 prices ( ). In the longer term, it seems probable that the trend towards reduced investment intensity will be confirmed by the structural change which Japanese industry is undergoing. This change away from heavy industry as the vehicle of growth towards advanced technology intensive industries and the service sector, is likely to reduce the need for the huge, concentrated investments associated with blast furnace steel mills, refineries and similar installations. Further, the other side of the corporate sector balance is now improving. Corporate profits have significantly increased in recent years (especially in 1988) and, given the recent emphasis of the profit motive by most large Japanese companies, corporate savings can be expected to improve. Thus, the recent pattern of reduced corporate sector deficits is likely to continue and, indeed, strengthen. In this regard, it is worthy of note that by the end of fiscal 1988, the average ratio of net worth to total capital in the manufacturing sector was 31.1%, the first • • • 109 time that this ratio has exceeded 30% ( ). In general, it seems the factors which have contributed to the private sector balances of recent years appear likely to have future validity. "The Consumption Boom in Japan", Tokai Monthly Economic Letter, No. 125, Tokai Bank, Tokyo, June 1989, p.5. "Corporate Profit Performance", NRI Quarterly Review, Nomura Research Institute, Tokyo, August 1989, p.20. Page 98 6.1.4. Implications of Private Sector Savings Surpluses Despite its recent narrowing, the private sector savings surplus continues to be high, at around U.S. $100 billion. This has significant implications for the remaining macro-economic balances of the Accounting Identity introduced in 5.1. As noted previously, private sector savings surpluses must be accommodated by government deficits or by current account surpluses. The latter, as defined in the balance of payments accounting framework, necessarily requires an equal capital account deficit. As will be shown in the following discussion, the private sector saving surplus up to 1982 was, to a large extent, absorbed by government deficits. Since 1983, however, the surplus has been assimilated by widening current account surpluses. These developments are discussed in the following section. - Page 99 6.2. THE PUBLIC SECTOR BALANCE Table 6.11 shows that, between 1975 and 1983, the private sector surplus was principally absorbed by increasingly large government (central, municipal and prefectural) fiscal deficits. Although bond issuances were legislatively permissible after 1949, the Japanese government practiced balanced budgets until 1966, when the first long-term public bonds were issued. Because of a tendency to underestimate economic growth, the government sector remained largely in surplus throughout this period, with an annual average surplus of about 2% of GNP. This situation changed after 1975, with fiscal deficits averaging 4% of GNP between 1975 and 1983 and peaking at 5.5% of GNP in 1978. Since 1984, the deficit has continued to narrow and is projected to be about 300 billion yen in 1989, down from Yen 19,100 billion in 1983. Two important points emerge from Table 6-11. First, although the government deficit widened considerably after 1974, government investment as a percentage of GNP remained largely unchanged. During the 1970s, government investment averaged about 6.2% of GNP, only slightly higher than the 5.2% average of the 1960s. Second, government savings which averaged about 6.9% of GNP in the 1960-74 period, declined sharply after 1975 and averaged only 2.7% of GNP from 1975 to 1983. The post-1973 deficits were caused by sustained, moderate government spending and lowered tax revenues. Nominal expenditure growth remained more or less constant, averaging 17.9% between F.Y. - Page 100 -TABLE 6.11: THE PRIVATE AND PUBLIC SECTOR BALANCE, F.Y. 1960-1987, PERCENTA6E OF 6NP TOTAL PRIVATE SECTOR YEAR SAVINGS INVESTMENT BALANCE 1960 28.3 31.2 -2.9 1961 28.5 36.0 -7.5 1962 27.9 28.5 -0.6 1963 27.5 30.8 -3.3 1964 27.2 29.2 -2.0 1965 27.4 27.5 -0.1 1966 29.4 28.7 0.7 1967 31.4 32.7 -1.3 1968 32.9 33.0 -0.1 1969 32.5 34.5 -2.0 1970 33.3 34.0 -0.7 1971 31.0 30.0 1.0 1972 32.0 29.2 2.8 1973 32.3 31.6 0.7 1974 30.0 31.4 -1.4 1975 29.0 26.8 2.2 1976 30.4 26.0 4.4 1977 29.5 24.6 4.9 1978 30.8 23.9 6.9 1979 29.0 25.3 3.7 1980 28.4 25.1 3.3 1981 27.9 24.2 3.7 1982 27.2 23.3 3.9 1983 26.9 21.9 5.0 1984 26.7 22.3 4.4 1985 26.6 22.8 4.0 1986 - - 5.8 1987 - - 4.5 BALANCE TOTAL PUBLIC SECTOR AVAILABLE FOR FOREIGN SAVINGS INVESTMENT BALANCE INVESTMENT 7.7 4.7 3.0 0.1 8.5 4.8 3.7 -3.8 8.2 5.7 2.5 1.9 7.5 5.3 2.2 -1.1 6.6 5.3 1.5 -0.5 6 5.4 0.6 0.5 5.6 5.4 0.2 0.9 6.3 5.1 1.2 -0.1 6.9 5.1 1.8 1.7 7.5 5 2.5 0.5 6.8 5.1 1.7 1 7 5.8 1.2 2.2 6.2 6.3 -0.1 2.7 6.9 6.4 0.5 1.2 6.3 6 0.3 -1.1 3.3 6 -2.7 -0.5 2.1 5.8 -3.7 0.7 2.5 6.3 -3.8 1.1 1.5 7 -5.5 1.4 2.5 7.2 -4.7 -1 2.7 7.1 -4.4 -1.1 3.3 7.1 -3.8 -0.1 3.2 6.8 -3.6 0.3 2.8 6.4 -3.6 1.4 3.9 6 -2.1 2.3 4.8 5.6 -0.8 3.2 - - -1.1 4.7 - - -0.2 4.3 - Page 100a -1970 and 1973 and 17.4% between F.Y. 1974 and 1979 (11°) . However, because of high inflation during this period, the growth in real expenditures decreased from an average of 13.9% to an average of 8.5%. The combined influence of sustained nominal expenditures and decreased tax revenues on the government balance is discussed below. 6.2.1. Government Expenditures after 1973 Some authorities have attributed the Japanese public sector deficit to the growth in social security spending which occurred after 1970 (111) . As can be seen in Table 6.12, in real terms, social security expenditures increased at an average of 15% per annum during the 1970s, while the general account budget grew at an average of 9.3% per year. The increase in social security spending reflected the political commitment of the LDP to improved social welfare conditions in Japan; a commitment which emerged after the public dissent of the late 1960s. Prior to 1973, old age pension benefits were only 20% of the average salary but the ratio jumped to 43% after 1974 and an indexation provision for inflation was added (112) . The commitment to improved social conditions was also reflected in expanded public works programs during the 1970s, Lincoln, Japan: Facing Economic Maturity, p. 92. 111 Yukio Noguchi, "Public Finance," The Political Economy of  Japan. Volume 1. p. 187-219. 112 Ibid. . p. 205. - Page 101 -TABLE 6.12: GROUTH OF BUDGETED GOVERNMENT SPENDING BY CATEGORY, F.Y. 1970-1986 REAL PERCENTAGE CHAN6E OVER PREVIOUS PERIOD YEAR 1972 1974 1976 1978 1980 1981 1982 1983 19B4 1985 1986 CATE60RY ' SOCIAL WELFARE: 30.2 36.3 33.8 26.7 13.8 4.0 1.3 0.2 3.6 -0.1 0.9 Social security 30.4 51.2 39.1 31.5 15.5 5.1 -0.2 -5.4 -5.5 2.5 3.5 health 20.9 -4.2 2.2 8.7 6.3 -1.3 0.3 1.1 10.5 2.9 5.4 EDUCATION/SCIENCE 26.4 24.3 14.9 14.2 11.1 1.5 -1.6 -0.8 1.3 -2.5 -1.7 GOVT. PENSIONS 10.3 18.9 43.6 21.2 15.4 6.5 3.0 -0.8 -1.5 -2.7 -2.6 DEFENSE 24.7 9.4 7.7 11.0 13.1 3.8 3.8 6.8 6.0 5.3 116.7 PUBLIC WORKS 67.7 -17.6: 10.9 38.0 9.6 -0.3 2.1 -1.5 -5.6 -1.0 -75.2 F0REI6N AID 14.4 3.6 -5.2 39.0 27.1 8.1 9.6 3.9 7.4 5.7 4.5 SMALL BUSINESS 26.0 9.9 22.9 40.1 -1.2 -1.2 -3.3 -2.6 -5.3 -8.8 -6.1 ENERGY 45.2 13.3 10.5 -1.1 8.1 1.0 -1.7 FOODSTUFF CONTROL 4.1 38.2 -21.1 -5.8 -4.4 2.8 -2.6 -8.8 10.5 -13.8 -14.1 OTHER 41.5 20.7 11.6 9.2 8.4 -2.1 -2.4 -2.2 -2.3 -1.2 -7.0 REVENUE SHARING 22.8 28.1 -15.2 31.4 27.4 7.9 -9.6 -4.4 18.7 3.5 3.2 DEBT SERVICE 41.9 37.0 87.6 58.1 58.9 17.4 1.9 17.4 11.6 8.6 8.7 TOTAL 34.4 16.0 12.7 26.9 18.7 5.0 -0.9 1.6 5.3 1.7 -1.7 - Page 101a -although the latter tended to closely follow economic cycles. Noguchi attributes increased social welfare and public works spending to the favorable fiscal conditions of the early 1970s. By 1970, Japan had attained a "fiscal affluence", within which increased government expenditures could be supported by the "natural" increases in tax revenues which resulted from sustained economic growth (113) . That this situation did not endure after 1974 is evidenced by the increase in government bond issues in the latter half of the 1970s. As a result, the largest increase in public expenditures between 1970 and 1986 was that relating to debt-service, which grew at an average real rate of over 39% per year. Despite the significant annual increases in real government spending, expressed as a percentage of GNP, actual program expenditures remained surprisingly constant. Average government expenditure as a percentage of GNP grew from an annual average 14.6% of GNP during the 1970s.to 17.1% after 1980 (see Table 6.13). However, most of this increase was due the rising cost of debt-service, which represented the largest single expenditure item on the general account by 1985 (see Table 6.14). 113 Ibid.. p. 206. - Page 102 -TABLE 6.13: 60VERNMENT SPENDING BY CATEGORY AS A PERCENTAGE OF GNP, F.Y. 1970-1986 YEAR CATE60RY SOCIAL WELFARE: Social security Health EDUCATION/SCIENCE 60VT. PENSIONS DEFENSE PUBLIC WORKS F0REI6N AID SHALL BUSINESS ENERGY FOODSTUFF CONTROL OTHER REVENUE SHARING DEBT SERVICE TOTAL 1970 1972 1974 1976 1978 1980 1981 1982 1983 1984 1985 19B6 1.5 2.5 2.2 0.9 1.4 1.4 0.2 0.3 0.2 1.3 2.1 1.7 0.4 0.6 0.4 0.8 1.3 0.9 2.0 4.2 2.2 0.1 0.2 0.1 0.1 0.1 0.1 0.0 0.0 0.0 0.5 0.6 0.5 1.3 2.3 1.8 2.3 3.5 2.9 0.4 0.7 0.6 11.1 18.5 13.9 2.8 3.2 3.3 1.8 2.1 2.2 0.2 0.2 0.2 1.8 1.8 1.9 0.6 0.6 0.7 0.9 0.9 0.9 2.3 2.9 2.8 0.1. 0.1 0.2 0.1 0.1 0.1 0.0 0.1 0.2 0.4 0.3 0.3 1.8 1.8 1.8 2.3 2.8 3.2 1.1 1.6 2.2 14.3 16.4 17.6 3.3 3.3 3.2 2.3 2.2 2.0 0.1 0.1 0.1 1.8 1.8 1.7 0.7 0.7 0.7 0.9 0.9 1.0 2.8 2.7 2.6 0.2 0.2 0.2 0.1 O.l 0.1 0.2 0.2 0.2 0.3 0.3 0.2 1.7 1.6 1.5 3.3 2.9 2.7 2.6 2.5 2.9 17.9 17.2 17.7 3.1 3.1 3.0 1.6 1.8 1.8 0.1 0.1 0.1 1.6 1.5 1.5 0.6 0.6 0.6 1.0 1.0 2.1 2.3 2.2 0.5 0.2 0.2 0.2 0.1 0.1 0.1 0.2 0.2 0.2 0.3 0.2 0.2 1.4 1.3 1.2 3.0 3.0 3.1 3.1 3.2 3.4 16.8 16.4 16.3 - Page 102a -TABLE 6.14: BUD6ETED GOVERNMENT SPENDIN6 BY CATE60RY, F.Y. 1970-1986 REAL EXPENDITURES, BILLIONS OF YEN YEAR 1970 i972 1974 1976 1978 1980 1981 1982 1983 1984 1985 1966 CATE60RY SOCIAL WELFARE: 2350 3060 4172 5582 7073 8049 8369 8474 8494 8803 8793 8666 Social security 1298 1692 2558 3559 4681 5405 5682 5670 5364 5070 5196 5378 Health 273 330 316 323 351 373 368 369 373 412 424 447 EDUCATION/SCIENCE 1983 2507 3116 3580 4087 4541 4611 4538 4503 4560 4444 4369 GOVT. PENSIONS 623 687 817 1173 1422 1641 1748 1800 1786 1760 1713 1668 DEFENSE 1217 1517 1659 1786 1983 2243 2328 2416 2580 2735 2881 6243 PUBLIC WORKS 2996 5024 4138 4590 6335 6942 6922 7069 6960 6567 6503 1612 FOREIGN AID 194 222 230 218 303 385 416 456 474 509 538 562 SMALL BUSINESS 104 131 144 177 248 245 242 234 228 216 197 185 ENER6Y 0 0 0 0 294 427 484 535 529 572 578 568 FOODSTUFF CONTROL 702 731 1010 797 751 718 738 719 656 725 625 537 OTHER 1948 2757 3327 3713 4053 4395 4301 4198 4105 4011 3961 3683 REVENUE SHARING 3475 4269 5468 4638 6096 7769 8386 7582 7249 8602 8899 9164 DEBT SERVICE 602 854 1170 2195 3470 5513 6474 6598 7747 8647 9389 10207 TOTAL 16194 21759 25251 28449 36115 42868 45019 44619 45311 47707 48521 47666 Source: Lincoln, Japan: Facing Econoaic Maturity, p. 92 - Page 102b -6.2.2. Government Revenues after 1973 As noted above, the primary cause of rising government deficits during the 1970s was the decline in tax revenues. Both nominal and real receipts decreased markedly after 1975. In real terms, national tax receipts grew at an average of 11.7% between 1960 and 1973, fell consecutively in 1974 and 1975 by 7.2 and 14.5%, respectively, and averaged 6.5% growth thereafter (see Table 6.15). Thus, tax revenues were permitted to lag behind moderate increases in real spending. Lincoln (m) points out that: The most striking feature (of government fiscal policy), in fact, is that all categories of expenditures continued to rise relatively rapidly in nominal terms, while taxes were allowed to lag behind. Thus, Japan did not get itself into large deficits because of a major burst of new spending so much as from a lack of revenue to support its relatively modest real increases in spending. The real terms shortfalls in tax revenues initially occurred because of the economic recession which followed the oil shock of 1973. However, over the longer term, the principal reason for lagging tax receipts lay in the psyche of the Japanese public. Throughout the 1970s and early 1980s, tax reform was strongly opposed by powerful lobby interests. In 1979, the Ministry of Finance proposed that a 5% value added tax be imposed on all firms with sales in excess of 20 million yen. Opposition from small businesses, historically strong backers of the LDP, was so intense that prior to the 1979 election for the Lower House of the Diet, the Liberal Democratic Party (LDP) was forced to promise that it Lincoln, Japan: Facing Economic Maturity, p. 96. - Page 103 -TABLE 6.15: PERCENTAGE REAL INCREASE iN TAX.BURDEN, F.Y. 1961-1S87 PERCENTAGE INCREASE OVER PREVIOUS F.Y. (REAL TERNS) YEAR 6NP NATIONAL INCOME TOTAL TAXES NATIONAL LOCAL TAXES TAXES 1961 13.5 10.0 14.1 14.6 12.9 1962 5.4 8.7 6.2 3.7 12.6 1963 13.0 11.3 9.4 9.3 9.8 1964 11.1 8.7 10.7 10.8 10.6 1965 17.3 19.6 12.3 10.1 17.4 1966 11.6 10.1 7.2 6.5 8.8 1967 10.9 12.0 13.4 12.9 14.3 1968 11.9 12.6 14.8 15.1 14.1 1969 11.2 10.7 3.0 -2.4 14.1 1970 9.0 13.5 25.3 32.4 12.8 1971 5.7 2.2 4.2 2.8 6.9 1972 9.8 9.5 14.9 16.5 11.7 1973 8.2 6.2 18.1 19.7 14.9 1974 -2.4 0.8 -3.3 -7.2 5.0 1975 1.9 3.4 -12.3 -14.5 -8.1 1976 6.2 3.8 8.6 8.2 9.5 1977 4.8 4.4 5.5 3.7 8.7 1978 4.5 4.1 14.9 20.2 6.1 1979 4.3 3.9 6.7 4.3 11.3 1980 6.3 7.9 9.3 9.5 9.1 1981 2.6 1.4 11.5 13.9 7.1 1982 3.0 2.1 -2.4 -5.8 4.1 1983 3.5 4.4 5.9 5.9 5.7 1984 5.4 3.8 6.6 6.3 7.0 1985 4.2 4.5 5.5 4.8 6.8 1986 1.4 2.1 2.0 1.4 3.1 1987 4.4 4.4 3.8 5.1 1.8 Average increase over period: 1960-1973 1976-1987 11.8 6.5 11.7 6.5 12.4 6.7 1960-1974 1976-1987 17.7 9.6 17.3 9.6 18.8 9.8 - Page 103a -would not introduce the tax. In spite of this, the LDP lost its majority in the lower house and was only able to retain power with the support of a number of independents (115) . Other attempts to correct tax loopholes met with similar opposition. The attempt to introduce the "green card" system was a particularly illustrative example of the opposition to change. Prior to 1987, savers were not taxed on interest income earned by savings accounts which together totalled less than 3 million yen. Commercial banks were required to file information on all savings accounts to allow the authorities to determine if an individual's total accounts exceeded 3 million yen. However, the postal savings system, which has, by far, the country's largest deposit base, was only required to file reports on accounts which exceeded this amount. This loophole enabled depositors to open numerous accounts. (It is reported that the number of accounts in the postal savings system exceeds the population of Japan). In 1979, the Ministry of Finance (MOF) attempted to correct this weakness by proposing that every saver be issued a green identity card that would have to be produced whenever opening a tax-qualified savings account. However, the proposed reform was strongly opposed by the Ministry of Posts and Telecommunications which controls the postal savings system and by the public, who decried the proposal as an unwarranted intrusion of the government into their personal affairs (116) . After three 115 Noguchi, "Public Finance," p. 209. 116 Lincoln, Japan: Facing Economic Maturity. f p. 106. - Page 104 -years of unsuccessful attempts to have the program introduced, the MOF advanced a compromise proposal that a low rate of tax be charged on all interest income. This proposal was finally implemented in 1987. 6.2.3. The Pressure for Fiscal Austerity As a result of the failure of tax receipts to match spending increases, government revenue shortfalls were met with increasingly large bond issues after 1973. Between 1973 and 1986, the proportion of bonds, in total government borrowing, increased from 69% to 81% (117) . By 1978, the MOF began to voice increasing concern over the size of the deficits and to pressure the government to return to a balanced budget fiscal policy. The Ministry voiced several concerns, including the possibility that future fiscal policy could be constrained by the increasing debt-service requirements, a non-discretionary component on the general account budget. Additional arguments encompassed the burden placed on future generations; the potential "crowding-out" effect of private investment; the inflation potential of monetized government debt; and the removal of the natural limitation of government spending to government tax receipts. As Lincoln has pointed out (118) , many of these arguments were either not applicable, or tended to be less important than 117 Ibid. . p. 140. 118 Lincoln, Japan: Facing Economic Maturity, p. 101-102. - Page 105 -construed by the Ministry of Finance. In view of the large private-sector surplus, the "crowding-out" argument was particularly weak; even at their height in the late 1970s and early 1980s, government deficits failed to fully absorb private sector surpluses. However, regardless of the strength of the counter arguments, the end of the 1970s brought a general, public and private sector consensus that government deficits should be curtailed. Indicative of this, in July 1980, newly-elected Prime Minister Suzuki announced that the issuance of deficit financing bonds would be eliminated by F.Y. 1984. Several measures were introduced to achieve this goal. In F.Y. 1980, the ceiling for annual increases in budget requests for each ministry and government department was lowered from around 20% to 10%. Permissible increases were eliminated in F.Y. 1982; and after 1984, a "minus-ceiling" was introduced for certain expenditures (119) . In addition, the government established the Ad Hoc Council on Administrative Reform in March 1981, with the mandate to assess government activities for methods of rationalizing expenditures. The Council, commonly known as Rincho, was headed by Toshio Doko, the retiring chairman of the Keidanren, a federation of economic organizations in Japan and the "voice of big business in 120 t Japan"( ). The council recommended a policy of "fiscal reconstruction without tax increases" in which government deficits 119 Noguchi, "Public Finance," p. 211. 120 Lincoln, p. 117. - Page 106 -were to be curtailed by decreases in expenditures rather than tax reform. To this end, the council proposed that the social security system be revised and that government-run monopolies such as the Japan National Railway, Nippon Telephone and Telegraph Corporation, and Japan Tobacco and Salt be privatized (121) . The fiscal austerity program proved successful in reducing government bond issues; however, the target of eliminating all issuances by F.Y. 1984 was not achieved and has since been postponed to F.Y. 1990 (122) • 6.2.4. The Government Balance since 1984 Since 1984, a number of changes have taken place which indicate that the government is reconsidering its fiscal austerity program. This time, the pressure for change has been from external, rather than internal, forces. In particular, the large sustained surpluses in the Japanese current account balance has led to mounting pressure from the country's leading trade partners for fiscal expansion. In addition, the appreciation of the yen since 1985 by 80% has ushered in a new phase of Japanese macro-economic growth. Although the current account surplus has failed to decline substantially in dollar terms (due to the short term effects of yen appreciation through the J-curve effect, significant productivity gains in some key export industries, and the continuing popularity Lincoln, Japan: Facing Economic Maturity, p. 118. Noguchi, "Public Finance," p. 211. - Page 107 -of Japanese exports), the prospect of reduced current account surpluses in the future must be considered from the perspective of the macro economic identity presented earlier. As evidenced in Table 6.16, the private sector surplus was increasingly absorbed by large current account (merchandise trade) surpluses after 1980 as the government implemented its program of fiscal reform. As merchandise trade surpluses decline in the future, the economy may have to adjust by either reduced surpluses in the private sector or larger government deficits. This adjustment, however, could be postponed or reduced by offsetting changes in the services trade balance, which has posted increasingly large receipts of foreign investment earnings since 1980. The latter has important ramifications for the sustainability of Japanese foreign direct investment in the future and will be discussed in Section 9.0. To some extent, some adjustment in the domestic sector has already started to occur in response to declining merchandise trade balances. Government expenditures on social infrastructure have been undertaken to stimulate domestic demand in light of the reduced demand in the export sector. These expenditures were encouraged by the Maekawa report of April 1986 which recommended that Japan reduce its reliance on exports for economic growth and move towards a domestic-demand driven economy. However, movements away from fiscal austerity have met with strong opposition from the Ministry of Finance and the Bank of Japan. In addition, certain actions taken by the government in 1989 would seem to indicate some - Page 108 TABLE 6.16: THE DOMESTIC SECTOR AND CURRENT ACCOUNT BALANCE, F.Y. 1960-1987, PERCENTAGE OF 6NP PERCENTA6E OF GNP TOTAL PRIVATE SECTOR TOTAL PUBLIC SECTOR CURRENT ACCOUNT /EAR SAVIN6S INVESTMENT BALANCE SAVINGS INVESTMENT BALANCE BALANCE 1960 28.3 31.2 -2.9 7.7 4.7 3.0 0.0 1961 28.5 36.0 -7.5 8.5 4.8 3.7 -1.8 1962 27.9 28.5 -0.6 8.2 5.7 2.5 0.0 1963 27.5 30.8 -3.3 7.5 5.3 2.2 -1.5 1964 27.2 29.2 -2.0 6.8 5.3 1.5 0.0 1965 27.4 27.5 -0.1 6 5.4 0.6 1.2 1966 29.4 2B.7 0.7 5.6 5.4 0.2 0.9 1967 31.4 32.7 -1.3 6.3 5.1 1.2 -0.3 1968 32.9 33.0 -0.1 6.9 5.1 1.8 1.0 1969 32.5 34.5 -2.0 7.5 5 2.5 1.2 1970 33.3 34.0 -0.7 6.8 5.1 1.7 1.0 1971 31.0 30.0 1.0 7 5.8 1.2 2.5 1972 32.0 29.2 2.8 6.2 6.3 -0.1 2.2 1973 32.3 31.6 0.7 6.9 6.4 0.5 0.0 1974 30.0 31.4 -1.4 6.3 6 0.3 -1.0 1975 29.0 26.8 2.2 3.3 6 -2.7 -0.1 1976 30.4 26.0 4.4 2.1 5.8 -3.7 0.7 1977 29.5 24.6 4.9 2.5 6.3 -3.8 1.5 1978 30.8 23.9 6.9 1.5 7 -5.5 1.7 1979 29.0 25.3 3.7 2.5 7.2 -4.7 -0.9 1980 28.4 25.1 3.3 2.7 7.1 -4.4 -1.0 1981 27.9 24.2 3.7 3.3 7.1 -3.8 0.5 1982 27.2 23.3 3.9 3.2 6.8 -3.6 0.7 1983 26.9 21.9 5.0 2.8 6.4 -3.6 1.6 1984 26.7 22.3 4.4 3.9 6 -2.1 2.8 1985 26.8 22.8 4.0 4.8 5.6 -0.8 3.7 1986 - - 5.8 - - -1.1 4.3 1987 - - 4.5 - - -0.2 3.6 - Page 108a -confusion as to the future direction of fiscal policy and efforts to reduce the private sector savings surplus. In terms of the latter, some of the tax reforms undertaken after 1987, have been directed at reducing the tax advantages of personal savings (i.e., the 20% tax levied on all interest income from savings accounts) . At the same time, the consumption tax introduced in the spring of 1989 is biased against personal consumption. These reforms mean higher tax receipts which, when coupled with only moderate increases in government expenditures, seem to indicate that the LDP is unwilling to shift away from fiscal austerity. 6.4. THE DOMESTIC SECTOR - SUMMARY The preceding sections have highlighted the factors behind the enormous pool of capital which has been available for offshore investment since the early 1980s. This pool can be traced to the private sector savings surplus where sustained high levels of household savings have more than offset the savings/investment deficits of the corporate sector. Up to 1986, declining savings rates in the private sector were more than matched by reductions in investment in the corporate sector so that net deficits are well below the pre-1975 levels. Since 1987, corporate investment has increased sharply and is expected to remain high in 1989. However, in the longer term, the structural change of the Japanese economy towards technology and service industries will reduce investment intensity and this, together with improved profits and higher - Page 109 equity ratios, will probably limit the size of corporate sector deficits. As will be discussed in Section 9.0., these forces are likely to maintain the private sector surpluses at reasonably high levels. Government actions since 1978 have contributed to the recycling of this excess capital in the domestic economy offshore. Notwithstanding the proposals of the Maekawa Report of 1986 and the recent modest stimulations of the economy, government practice suggests that fiscal austerity and balanced budgets are the principal policy aims. The recent, and politically unpopular consumption tax is the most evident manifestation of these objectives. The inability of the domestic economy to absorb the excess savings of the private sector has been manifested in the huge increase in deficits on the capital account balance and cocommittant surpluses on the current account balance. As the current situation in the private and government sector balances appears unlikely to change dramatically in the near future, the disposition of these excess savings among various foreign investment uses becomes important in determining the sustainability of recent increases in foreign direct investment. As such, Section 7.0. will examine the external components of the macro-economic accounting identity from the perspective of the capital account balance. - Page 110 -7.0. EXTERNAL MACRO-ECONOMIC COMPONENTS OF THE JAPANESE ECONOMY The four external macroeconomic components of the Japanese economy are: * Net Foreign Direct Investment * Net Foreign Portfolio Investment * Net Foreign Short term Capital Investment and, * Net Official and Private Monetary Movements. Each of the above components of the overall economic balance will be examined briefly. 7.1. NET FOREIGN DIRECT INVESTMENT Japanese FDI is the focal subject of this paper. The net direct investment balance of Japan is a reflection of the power of the Japanese economy and, if continued at the current high levels, will be a powerful determinant of future economic developments well beyond Japan's own borders. Section 4.0. described the practice of Japanese foreign direct investment since World War II and, especially, since the mid-nineteen eighties, when foreign direct investment accelerated to dimensions of global strategic importance. Net foreign direct investment must also consider the impact of foreign direct investment in Japan. To date, such investment has been relatively small; for the period F.Y. 1950 to F.Y. 1987, - Page 111 -cumulative direct foreign investment in Japan totalled approximately $8.5 billion (123) . Until 1980, foreign investment in Japan was strictly regulated by government authorities initially to protect the country's precarious balance of payments position and then to protect "infant industries" from foreign competition. In the 1980s, however, foreign investment has been completely de regulated although, as noted by Higashi and Lauter (124) , foreign hostile acquisitions of Japanese companies are still discouraged. Despite the lessening of regulatory restrictions on foreign investment in Japan, FDI still remains relatively weak in the 1980s. The reasons behind this mirror the factors responsible for Japan's FDI push offshore; high relative labour costs, exorbitant land prices and land use restrictions, and high energy costs. As a result, foreign investment to date has consisted primarily of investments in distribution, wholesaling and retailing as opposed to the set-up of production facilities, etc. There is some evidence that this may be changing as foreign companies seek new methods of penetrating Japanese markets and accessing Japanese technology. As noted by Higashi and Lauter (125) , during the latter Nippon. Business Facts and Figures. 1989. Jetro, Tokyo, 1989, p. 49. Chikara Higashi and G. Peter Lauter, The  Internationalization of the Japanese Economyr Boston: Kluwer Academic Publishers, 1987, p. 16.1. 125 Ibid. . p. 161. - Page 112 -half of the 1980s, More foreign companies decided that the only way to stay in the highly competitive Japanese marketplace was to establish a manufacturing and maybe even a research and development base. They concluded that investments in distribution facilities alone would not be sufficient in the future because not only does a production and research base enhance competitive abilities but it also makes the acquisition of the increasingly significant Japanese research findings much easier. Although foreign investment may increase in Japan as a result of these issues in the future, the likelihood is that FDI in Japan will continue to be dwarfed by Japanese foreign investment abroad. Foreign investment will remain limited by the high costs of operating in the Japanese market. 7.2. NET FOREIGN PORTFOLIO INVESTMENT Net foreign portfolio investment is, by far, the largest component of total Japanese net investment outflows. Purchases of foreign securities rose from U.S. $10.3 billion in 1982 to U.S. $296.9 billion in 1985 and to U.S. $1,440.6 billion in 1988; during the same time frame, net purchases increased from U.S. $6 billion in 1982 to U.S. $88.8 billion in 1988, with a 1986 peak of U.S. $100.1 billion in 1986(126). Net purchases of Japanese securities by non-Japanese, during the same period, ranged between U.S. $1.2 billion in 1981 and minus U.S. $2.8 billion in 1988. 126 NRI Quarterly Review. Nomura Research Institute, Tokyo, August 1989, p.67 - Page 113 -NFPI accounts for about 70 to 75% of total Japanese net foreign long term investments. Despite this, NFPI will be given only cursory examination in this analysis because, within the context of the paper, NFPI is only of interest as a competing utilization of excess Japanese savings. Furthermore, portfolio investments are predominantly made for the sole purpose of return and, in the long run, are of much less strategic importance than direct foreign investment, which seeks to shape and control policy of foreign based operations. Although real Japanese net foreign portfolio investment has increased at a compounded annual rate of 134% since 1982 (see Table 7.1), it tended to fluctuate substantially prior to 1982. As indicated in Table 7.2, 91.4% of Japanese net foreign portfolio investment since 1980 has been directed towards bonds and only 8.6% has been invested in foreign securities. Nevertheless, securities investment showed a remarkable jump in F.Y. 1987, rising to U.S. $16.9 from approximately U.S. $7 billion in F.Y. 1986, although it dropped back to U.S. $3 billion in F.Y 1988. The United States remains the largest recipient of Japanese net foreign portfolio 127 investment. According to Jetro ( ), nearly 70% of the total 1988 purchases were U.S. securities. There are several factors behind the rise in net foreign portfolio holdings by the Japanese. From a macro-economic viewpoint, the rise in Japanese NFPI coincided with the easing of Nippon 1988. Jetro, Tokyo, 1989, p. 94. - Page 114 -TABLE 7.1: REAL JAPANESE NET FOREIGN PORTFOLIO INVESTMENT, F.Y. 1960-1987, U.S.* BILLIONS ! REAL NET ! PORTFOLIO ! INVESTMENT YEAR ! US$ BILLIONS I I 1960 I -0.06 ! 1961 ! 0.03 i 1962 ! 0.20 ! 1963 ! 0.41 ! 1964 ! 0.39 ! 1965 ! 0.22 ! 1966 I -0.05 ! 1967 ! -0.02 ! 1968 ! 0.82 i 1969 ! 2.09 IRATE OF 6R0WTH 1970 ! 0.52 IPER ANNUM 1971 ! 1.58 11983-1987 134X 1972 I 0.15 ! 1973 i -2.84 ! 1974 ! -1.19 ! 1975 ! 3.29 ! 1976 ! 3.29 ! 1977 ! 0.72 ! 1978 ! -3.01 ! 1979 I -1.28 ! 1980 i 9.43 ! 1981 ! 7.43 ! 1982 ! 0.80 i 1983 ! -2.74 1 1984 ! -22.35 i 1985 ! -38.34 ! 1986 ! -92.01 ! 1987 ! -82.10 ! - Page 114a -TABLE 7-2: INVESTMENT IN FOREIGN SECURITIES BY JAPANESE,F.Y. 1981 U.S. t MILLIONS U.S. DOLLAR MILLIONS STOCKS 1 1 BONDS NET 1 1 NET YEAR PURCHASE SALE PURCHASE {PURCHASE 1 SALE PURCHASE 1981 937 697 240 1 ! 9399 3591 5808 1982 1127 976 151 ! 16970 10904 6066 1983 2105 1447 658 ! 22906 10399 12507 1984 1569 1519 50 ! 56347 29576 26771 1985 5484 4489 995 ! 291376 237859 53517 1986 20917 13869 7048 i 1346989 1253965 93024 1987 70935 54061 16874 ! 1274201 1200944 73257 1988 76560 73564 2996 ! 1364061 1 1 1278248 85813 - Page 114b -the government's fiscal deficit position and the release of large amounts of excess private savings for offshore investment. Many commentators point to the 1980 revision of the 1949 Foreign Exchange Control Law as a factor behind the increase in offshore securities investment after 1980. Under the 1949 law, all foreign exchange transactions were controlled unless otherwise excepted. The revised law of 1980 permitted all transactions unless excepted. However, the government did retain extensive powers to control transactions which might damage the domestic economy ( ). In reality, there had been de facto liberalization of foreign exchange transactions since the early 1970s but the new law was important in formalizing and codifying administrative practice. Lincoln notes (129) : By confirming the validity of previous changes, the revision also legitimized continued liberalization; in effect it was a public statement that a new consensus had formed, an acknowledgement that in turn contributed to the momentum for further change in the same direction. Administrative encouragement of the outflow of capital was a necessary plank of the government's policy of fiscal austerity. In addition, Japanese offshore investment also helped to maintain the low value of the yen which had been depressed as a result of the 1979 crisis; which improved the competitiveness of Japanese exports and generated offsetting surpluses on the current account balance. Lincoln, Japan: Facing Economic Maturity, p. 250. Ibid.. p. 250-251. - Page 115 -Macro-economic developments external to Japan also contributed to the increase in net NFPI. These developments included the U.S. government's decision to pursue fiscal expansion after 1980 and the consequent need of the U.S. economy to attract large amounts of foreign capital to finance the resulting deficits. Unlike Japan, the U.S. private sector savings rate is very low (about 3% of GNP) ; thus, fiscal expansion had to be accommodated by an increase in capital inflows. This was achieved by a sharp increase in U.S. interest rates in the early 1980s. Coupled with strong economic growth and low inflation, its higher relative interest rates made the U.S.A. an attractive recipient country for foreign investment (see Table 7.3). As of March 1987, approximately 60 percent of Japanese foreign bond holdings were dollar denominated, principally in U.S. Treasury issues. (130) . Return on foreign portfolio investment requires consideration not only of interest earned but also of the exchange rate at which earnings can be converted for repatriation. It is, therefore, somewhat surprising that the most substantial increases in Japanese FPI came after 1984 when the yen was appreciating rapidly. Indeed, it has been reported that the 23 major Japanese life insurance companies (which, together, hold about 30% of all foreign securities holdings) wrote off foreign exchange losses of $14.8 130 • • • • Aron Vmer, The Financial Samurai. The Emerging Power of  Japanese Money. London: Kogan Page Limited, 1988, p. 164. - Page 116 -TABLE 7.3: REAL DISCOUNT RATES, U.S. AND BANK OF JAPAN, AVERAGE F.Y. 1961-•1988 REAL U.S. REAL REAL DISCOUNT B.O.J. U.S. - JAPAN RATE DISCOUNT DISCOUNT RAT (AVERAGE) RATE DIFFERENTIAL YEAR 1961 4.3 -0.64 -5.41 1962 1.2 3.04 1.84 1963 l.B 1.29 -0.48 1964 2.0 2.22 0.22 1965 2.5 11.21 8.67 1966 1.4 0.51 -0.90 1967 1.6 -0.48 -2.27 1968 0.5 1.01 0.53 1969 1.4 1.31 -0.12 1970 1.1 -1.37 -2.49 1071 0.2 -0.13 -0.34 1972 0.5 -0.72 -1.21 1973 -0.2 -5.62 -5.41 1974 -2.7 -11.58 -8.84 1975 -1.2 -1.16 -0.01 1976 -0.1 -0.62 -0.50 1977 -0.9 -1.56 -0.64 1978 0.8 -1.32 -2.07 1979 1.5 3.26 1.72 1980 1.9 3.41 1.50 1981 3.3 2.30 -1.02 1982 3.9 3.16 -0.71 1983 4.6 4.24 -0.36 1984 5.1 3.77 -1.37 1985 4.7 3.41 -1.30 1986 3.6 1.16 -2.45 1987 2.3 2.77 0.45 1988 2.8 2.14 -0.66 - Page 116a -billion for the year ending March 31, 1987 ( ) . The largest increase in foreign securities holdings occurred in 1986 (net purchases of $92 billion) when the yen appreciated from 238.54 yen to the dollar to 168.52 yen to the dollar. This increase may have occurred because Japanese investors mistakenly assumed that the appreciation of the yen had peaked. Nevertheless, the losses accrued in F.Y. 1986 did not prevent a further net portfolio investment of $82 billion in 1987. Viner observes (132) : The decline of the dollar has simply reduced the foreign currency expense of buying dollar assets while simultaneously reducing the chances that the holders will incur exchange rate losses in the future. The other side of the NFPI equation is the inflow of foreign capital into Japan. Prior to 1980, the inflow of either portfolio or direct investment was generally discouraged by regulatory restrictions. The new Foreign Exchange Control Law of 1980 removed • • 133 • • many of these restrictions ( ). Nevertheless, foreign investment in Japan has remained relatively small (see Table 7.4). From the portfolio investment perspective, this lack of interest may be attributed to the lower relative interest rates on Japanese Viner, The Financial Samurai, p. 184. 132 Ibid. . p. 184. 133 Michele Schmiegelow, "The Reform of Japan's Foreign Exchange and Foreign Trade Control Law: A case of Qualitative Economic Policy," Japan's Response to Crisis and Change in the  World Economy. edit, by Michele Schmiegelow, New York: M.E. Sharpe, Inc., 1986, p.19-22. - Page 117 -TABLE 7.4: INVESTMENT iN JAPANESE SECURITIES, BY NON JAPANESE. F.Y. 1981-1988. YEN BILLIONS YEN BILLIONS STOCKS I I 1 ' ..... BONDS NET 1 I NET YEAR PURCHASE SALE PURCHASE 1 PURCHASE SALE PURCHASE 1981 5399.6 4576.7 822.9 1 ! 5179.4 3953.2 1226.2 1982 4000.0 6320.2 -2320.2 ! 6970.8 5846.1 1124.7 1983 7683.7 6640.9 1042.8 ! 9746.5 9213.6 532.9 1984 8440.6 10157.6 -1717.0 ! 14776.5 13960.4 816.1 1985 9381.0 10160.6 -779.6 ! 25193.7 24044.6 1149.1 1986 17026.9 20617.0 -3590.1 i 41143.3 41528.7 -385.4 1987 23371.5 30822.9 -7451.4 ! 43908.6 42915.6 993.0 1988 21995.2 21744.7 250.5 i 36088.0 1 38893.6 -2805.6 Source: Quarterly Econosic Review, August 1989 Nooura Research Institute, Tokyo, p. 67 - Page 117a -government bonds and the lower earnings potential of Japanese stocks (price earnings ratios average about 60 to 1 in Japan). The weakening value of the yen prior to 1984 has also been suggested as a contributing factor. However, the subsequent rapid appreciation of the yen has not been accompanied by a corresponding rise in portfolio investment by foreigners in Japan; suggesting that the latter may have had minimal impact on foreign long-term capital inflows (134) . 7.3. NET SHORT TERM CAPITAL INVESTMENT As indicated in Table 7.5, net short-term capital movements are an erratic component of the balance of payments. These capital movements fluctuate substantially with changes in international short-term interest rate differentials and, particularly, with changes in foreign exchange rates. The net flows can be considerable; in 1987 and 1988, for example, net short term capital flows were $20.5 and $31.1 billion, respectively (135) . The relatively large positive inflows of capital during these two years presumably reflected foreign expectations of the continued appreciation of the Japanese yen/U.S. dollar exchange rate. Although numerically substantial, short-term capital flows have not been considered extensively in this study due to the 134 Lincoln, Japan: Facing Economic Maturity, p. 249. 135 Quarterly Economic Review. August 1989. Vol. 19. No. 3. Nomura Research Institute, August 1989, p. 52. Page 118 -TABLE 7.5: JAPAN'S BALANCE OF PAYMENTS. F.Y. 1961-F.Y. 1987, MILLIONS OF U.S. $ 1 1 1 CURRENT ACCOUNT BALANCE 1 1 1 1 1 IL0N6-TERM SHORT-TERM BASIC OVERALL j TOTAL TRADE UNREQUITTED! CAPITAL CAPITAL ERRORS ' BALANCE ' BALANCE YEAR ! (A) BALANCE EXPORTS IMPORTS SERVICES TRANSFERS 1 (B) (C) (D) (A+B) (A+B+D) - j 1 1 1961 i -982 -558 4149 4707 -383 1 1 -41 I -11 21 20 -993 -952 1962 : -48 401 4861 4460 -420 -29 ! 172 107 6 124 237 1963 : -780 -166 5391 5557 . -569 -45 ! 467 107 45 -313 -161 1964 : -480 377 6704 6327 -784 -73 i 107 234 10 -373 . -129 1965 I 932 1901 8332 6431 -884 -85 i -415 -61 -51 517 405 1966 ! 1254 2275 9641 7366 -886 -135 : -808 -64 -45 446 337 1967 ! -190 1160 10231 9071 -1172 -178 : -812 506 -75 -1002 -571 1968 : 1048 2529 12751 10222 -1306 -175 1 -239 209 84 809 1102 1969 ! 2119 3699 15679 11980 -1399 -181 ! -155 178 141 1964 2283 1970 ! 1970 3963 18969 15006 -1785 -208 ! -1591 724 271 379 1374 1971 ! 5777 7787 23566 15779 1758 -252 ! -1082 2435 527 4695 7657 1972 ! 6624 8971 28032 19061 -1883 -464 ! -4487 2137 1966 2137 6240 1973 ! -136 3688 36264 32576 -3510 -314 ! -9750 2407 -2595 -9886 -10074 1974 ! -4693 1436 54480 53044 -5842 -287 I -3881 1778 -43 -8574 -6839 1975 : -682 5028 54734 49706 -5354 -356 1 -272 -1138 , -584 -954 -2676 1976 : 3680 9887 66026 56139 -5867 -340 I -964 111 117 2696 2924 1977 ! 10918 17311 79333 62022 -6004 -389 : -3184 648 657 7734 9039 1978 ! 16534 24596 95634 71038 -7387 -675 ! -12389 1538 267 4145 5950 1979 ! -8754 1845 101232 99387 -9472 -1127 : -12618 2377 2333 -21372 -16662 1980 i -10746 2125 126736 124611 -11343 -1528 ! 2324 3141 -3115 -8422 -8396 1981 i 4770 19967 149522 129555 -13573 -1624 ! -9672 2265 493 -4902 -2144 1982 i 6850 16079 13663 119584 -9848 -1381 : -14969 -1579 . 4727 -8H9 -4971 1983 ! 20799 31454 145468 114014 -9106 -1549 : -17700 23 ! 2055 3099 5177 1984 ! 35003 44257 168290 124033 -7747 -1507 ! -49651 -4295 . 3743 -14648 -15200 1985 i 55019 61601 180664 119063 -4745 -1837 : -73177 -1475 , 4034 -18158 -15599 1986 : 94139 10648 211293 109645 -5135 -2374 ! -144680 899 . 5698 -50541 -43944 1987 ! 83474 94034 233435 139401 -5591 -3867 ! -119465 20502 ! -1490 -35991 -16979 - Page 118a -unpredictability of their movements and generally transient nature of their impacts. 7.4. NET MONETARY MOVEMENTS Net monetary movements include official interventions on the foreign exchange markets (accumulation of foreign reserves) and private holdings of foreign exchange and gold. Between F.Y. 1980 and F.Y. 1988, these monetary movements in aggregate totalled approximately -$103.7 billion. Of this total, slightly greater than 70% was accounted for by official (net) positive puchases of foreign currencies and gold (136) . The balance reflected (net) positive holdings of foreign currencies and gold in the private sector. The high variability and liquidity of these movements suggests that they will not be significant factors in determining the sustainability of direct foreign investment. For that reason, this paper has not considered official and private monetary movements in its analysis. 7.5. SUMMARY OF THE CAPITAL ACCOUNT BALANCE As indicated in the previous sections, the principal components of the capital account balance are net foreign portfolio and direct investment and short term capital movements, with the remaining Quarterly Economic Review. Vol. 19. No. 3. Nomura Research Institute, Tokyo, August 1989, p. 52. - Page 119 -components acting as primarily balancing items. Of the former, the most important from a numerical standpoint is net foreign portfolio investment, representing about 70% of Japanese capital outflows. In order to understand the determinants of Japanese FDI, therefore, two issues become important. First, the factors behind the domestic private sector savings-investment differential must be understood and, secondly, the factors responsible for determining the level of net foreign portfolio investment must also be considered. Both these issues are treated in section 8.0, where the results of a regression model, based on the structure of the macro-economic accounting framework, are analyzed. Page 120 -8.0. A STATISTICAL MODEL OF JAPANESE DIRECT FOREIGN INVESTMENT In the preceding macro-economic discussion, it was postulated that the rapid growth in the amount of Japanese capital available for offshore investment, particularly since the mid-1980s, is attributable to structural changes in the Japanese economy since 1973. These structural changes have included the halving of economic growth rates and a consequent decrease in capital investment requirements; a sustained tendency towards large private sector savings imbalances; and the unwillingness of the government to absorb excess savings from the private sector by deficit financing after 1978. On this basis, section 7 delineated foreign direct investment as the residual of the total foreign investment funds available after allowing for other competing offshore investments such as net foreign portfolio investment and short-term capital investments. This macro-economic approach provides a perspective of the source of funds used in direct and other foreign investment activities. In contrast, popular explanations of Japanese FDI have concentrated on the motivations for such investment including; the appreciation of the yen, protectionism in Japan's major export markets; and the reduced competitiveness of Japanese domestic production in some industries. More recently, the huge escalations in land prices have allowed Japanese investors, both individual and corporate, to use domestic land holdings as collateral for both direct foreign investments and expanded equity capital. The use of Japanese real - Page 121 estate as security for foreign purchases is typified by the tactics of Shigeru Kobayashi, chairman and founder of the Shuwa Corporation, who owns over $1.5 billion of U.S. real estate. Kobayashi notes that his recent increase in U.S. real estate acquisitions is "partly due to the soaring land prices in Tokyo, partly due to the appreciation of the yen and partly due to luck." («7, The motivating influences noted above have validity as determinants used to allocate foreign investment resources between direct, portfolio and short-term foreign investments. However, they do not provide any insight into the source of the funds; nor do they offer any projections as to the sustainability of Japanese direct foreign investment in the future. The macro-economic approach described in this paper addresses these factors. It provides, in one package, a quantitative method for assessing and projecting the pool of capital surplus to domestic needs and providing a framework within which the allocational influences of exchange rates, trade policies and other factors can be judged. The balance of this section will explain the results of a statistical model developed to test the validity of some these factors as explanatory variables of foreign direct investment by Japan. 137 nTne shuwa Shogun," Tokyo Business Today, Tokyo, March 1987, p. 28. - Page 122 -8.1. RATIONALE AND METHODOLOGY OF STATISTICAL APPROACH In section 5, an identity was introduced to explain the amount of capital available for foreign investment in terms of the amount generated and used in the domestic economy. This identity was, NFDI + NFPI + SC + F = (S-I) - (G-T) (6) Since NFDI is the focal point of this" discussion, it is useful to reform the identity as follows, NFDI = (S-I) - (G-T) - NFPI - SC - F (7) From this perspective, the amount of capital dedicated to NFDI is dependent on the excess capital generated in the domestic sector, and the amount of the excess which is earmarked for net foreign portfolio investment (NFPI), net short-term capital investment (Sc) and net monetary flows (F). Each component of this identity is determined by the interaction of a number of fundamental economic indicators. For example, it could be hypothesized that the private sector balance (S-I) is determined by the combined effects of interest rates, changes in income, and land prices on both savings and investment. Similarily, net foreign portfolio investment might be determined by the combined interaction of international interest rate differentials, the exchange rate etc. Because we have postulated that the components have some relationship with the level of net foreign direct investment, we can also hypothesize that the underlying determinants of the components may also explain, in part, NFDI. - Page 123 -Based on this reasoning, we have regressed several perceived determinants of the components of identity (7) on net foreign direct investment. (The rationale behind the approach is depicted in Figure 8.1.) The hypotheses concerning the relationships of these variables with NFDI and the results of the model are explained in section 8.2. Because of the volatility of Sc and nature of F as a balancing item on the capital account, the determinants of these components will not be incorporated in the NFDI regression model. In addition, the government balance was included in the model as an exogenous factor, although this is somewhat problematic given the probability of interdependence between the government and private sector balances. Therefore, the primary focuses of the NFDI regression model are the determinants of the private sector savings-investment balance and net foreign portfolio investment. This focus is justified by the macro-economic developments in Japan since 1973, specifically the exogenous increase in the excess savings generated in the private sector and the dominance of net portfolio investment in total Japanese offshore investment. Presumably, if we can determine the factors responsible for the generation of excess capital in the domestic sector and for its partial allocation to NFPI, some inference can be made regarding the forces behind NFDI. In order to test the hypotheses concerning the relationship between perceived underlying determinants and NFDI, regression models were also developed for the individual components of the - Page 124 -FIGURE 8.1 Macro-Economic Rationale Behind Statistical Approach NFDI = (S-I) - (G-T) - NFPI - Sc - F ****** * * -*— =*= * :*= COMPONENT REGRESSIONS *** * * * * * * * * * * * F*= (S-I) ******* * * *= =*= * * ***** **** * * =* (N PI) FUNDAMENTAL DETERMINANTS Interest Rates Changes in Income Land Prices =*= * * * ** =*= * * ****** **** * * =*= *** * * * * * * * * * * * *= International Interest Rate Differentials Exchange Rates NET FOREIGN DIRECT INVESTMENT =*= * * * ** - Page 124a -identity; specifically those for (S-I) and NFPI. The rationale and results of these models are contained in sections 8.3. and 8.4., respectively. It should be noted that the NFDI model represents a first attempt at explaining the factors behind foreign direct investment. A more formal approach would have to account for the problem of interdependence between the explanatory variables, a problem which is common among regressions using macro-economic data and which will be referred to in section 8.5.3. However, as a supplement to the foregoing macro-economic discussion, the model provides some useful, if albeit uncertain, results. The regressions and their results are discussed below. 8.2. NET FOREIGN DIRECT INVESTMENT As stated earlier, the independent variables which have been regressed against net foreign direct investment have been chosen because of their perceived explanatory power in determining the components of identity (7). However, the independent variables real wages, real land prices, and the Yen/U.S. dollar exchange rate are also popularly regarded as factors behind the rise in Japanese foreign direct investment. The hypotheses and results of the model are listed below. - Page 125 -8.2.1. The NFDI Model; Hypotheses and Rationale Before examining the hypotheses behind the NFDI model, it should be noted that both NFDI and NFPI (to be discussed in section 8.4.) are entered into the regression model as they appear in the balance of payments accounts i.e., negative numbers for net outflows and positive numbers of net inflows. The assumptions behind the NFDI model include: * Real Wages (W): Rising real wage costs in Japan has been proposed as a rationale behind the relocation of labour intensive industries from Japan to low labour cost countries. Thus, Japanese FDI would increase and foreign FDI would decrease, leading to a larger net outflow of foreign direct investment (i.e., NFDI would become more negative) . This would suggest a negative relationship between NFDI and real wages. * Real Land Prices (L): Rising real land prices (associated with land shortages, environmental constraints, etc.) have also been suggested as a reason for the relocation of some Japanese industries offshore. In more recent years, it has also been contended that Japanese land assets have been used increasingly as collateral for purchases of foreign real estate. These two contentions would support a negative relationship between NFDI and real land prices. Page 126 -* The Government Balance (GOV): The Government (spending minus revenues) balance is entered into the model as an exogenous factor. This approach can be rationalized on the basis that the government sector absorbs, in part, some of the excess capital generated in the private sector (which in this model is described by real wages and real land prices). Therefore, the government balance is a determining factor in the amount of capital available for total offshore investment. As the government balance increases (i.e., the government deficit increases), the amount of capital available for total offshore investment declines. Under such conditions, foreign direct investment by Japanese will decrease, resulting in NFDI becoming less negative. Thus, the government balance should be positively correlated with NFDI. * The Exchange Rate (EXCH): It is a popular contention that the appreciation of the yen is the primary reason for the increase in Japanese offshore direct investment in recent years. (This phenomenon also occurred to some extent in the early 1970s when the yen appreciated rapidly after the introduction of floating exchange rates.) Based on this contention, Japanese NFDI should be positively correlated with the exchange rate. - page 127 -* The Bank of Japan-U.S. Discount Rate differential (DISC): As will be discussed in section 8.4.1., if the spread (SPR) between the Bank of Japan discount rate and the U.S. Federal Reserve discount rate (i.e., the Bank of Japan discount rate minus the U.S. Federal Reserve discount rate) is increasing, FPI by Japanese will tend to decrease and FPI by foreigners in Japan will tend to increase. This should result in a positive relationship between SPR and NFPI. Since NFPI is a competing outlet for Japanese excess capital in the Japanese economy, the opposite should be true for Japanese net direct foreign investment (i.e. Japanese net direct foreign investment should be negatively correlated to SPR.) Based on the above, the NFDI model can be written as: DIRt = Qtt - Bn(W)t - B2(L)t + B3(GOVT)t + B4(EXC) - fl5(SPR)t + et where DIR is net foreign direct investment, GOVT is the government balance and e« N(0,a2). - Page 128 -8.2.2. The NFDI Model; Data Sources Data for net foreign direct investment was taken from the IMF's International Financial Statistics. 1988 for Japan. The data, as noted, has been entered into the model as it is represented in the balance of payments statements (i.e. negative numbers for net outflows and positive numbers for net inflows) and is in billions of yen (see Appendix 1). Real wages and land prices were taken from the Japan Statistical Yearbook (years 1970 through 1988). Both indices use 1980 as the base year. The real wage index is based on the nominal wage index (compiled from the average cash earnings of establishments with 30 or more regular workers) divided by the corresponding consumer price index for the year in question. The land price index represents the average land prices of the six largest urban areas in Japan (138) . For EXCH, data were taken from the IMF's International Financial  Statistics. Foreign Exchange Rate Supplement 1985 for data prior to 1985 and the Nomura Research Institute's Quarterly Economic  Review, for 1986 and 1987. Data for the Bank of Japan discount rate were also compiled from the Japan Statistical Yearbook. The regression uses the average rate for the year in question. The discount rate data for the U.S. was taken from the Business  Statistics. 1986 published by the U.S. Department of Commerce, 138 • These areas include: Tokyo, Kanagawa (Yokohama and Kawasaki City), Osaka, Kyoto, Sapporo and Nagoya - Page 129 -Bureau of Economic Analysis and the monthly publication Economic  Indicators (June 1989) prepared for the Joint Economic Committee by the Council of Economic Advisors. Both the Bank of Japan and U.S. discount rates were deflated by their respective GNP deflators to remove the effect of inflation (see Appendix 1). All variables, with the exception of the real wage index, were deflated using the corresponding GNP deflator for each year to remove distortions associated with inflation. 8.2.4 Results of the NFDI Model The results for the NFDI model are listed in Table 8.1. The signs of the estimated coefficients support the hypotheses of 8.2.1.; all estimated coefficients with the exception of the government balance were significant in terms of their t-values. Because of the insignificance of the government balance, a second regression was run omitting the latter as an explanatory variable. The results of the second regression are listed in Table 8.2. The estimated coefficients for the remaining explanatory variables changed marginally over Table 8.1. In addition, both the F-ratio and Durbin Watson statistics improved with the omission of GOVT (from 18.8 to 24.1, in terms of the former and from 1.56 to 1.63 in terms of the latter). The Durbin Watson statistic of 1.63 suggests the possibility of some positive auto-correlation in the error term; however, the D-W statistic is in the indeterminate range of 1.29 to 1.74 (for a 5% confidence level given the number - Page 130 -TABLE 8.1.: RESULTS OF NET FOREIGN DIRECT INVESTMENT MODEL (1) DEPENDENT VARIABLE: DIR INDEPENDENT VARIABLES: WAGE, LAND. GOVT, EXCH, SPR VARIABLE ESTIMATED STANDARD STANDARD. T-RATIO SEQUENTIAL SIMPLE NAME COEFFICIENT ERROR COEFFICIENT R-SQUARED R-SQUARED WAGE -19.14 7.69 -0.7756 -2.49 0.6043 0.6043 LAND -5.16 3.14 -0.2656 -1.64 0.6445 0.548 GOVT 0.01417 0.0228 0.0979 0.62 0.7193 0.0994 EXCH 3.67 1.62 0.4099 2.27 0.765 0.6699 SPR -6212.5 2685.6 -0.5459 -2.31 0.811 0.295 CONSTANT 287.43 908.85 0 0.32 R-SQUARED: 0.811 ADJUSTED R-SQUARED: 0.767 DURBIN-WATSON: 1.56 F-RATIO: 18.88 - Page 13.0a -TABLE 8.2.: RESULTS OF NET FOREIGN DIRECT INVESTMENT MODEL (2) DEPENDENT VARIABLE: DIR INDEPENDENT VARIABLES: WAGE, LAND, EXCH. SPR VARIABLE ESTIHATED STANDARD STANDARD. T-RATIO SEQUENTIAL SIMPLE NAME COEFFICIENT ERROR COEFFICIENT R-SQUARED R-SQUARED WAGE -17.7 7.24- -0.7173 -2.45 0.6043 0.6043 LAND -5.77 2.94 -0.2969 -1.96 0.6445 0.548 EXCH 3.96 1.52 0.4447 2.62 0.7331 0.6699 SPR -6992.3 2342.4 -0.6145 . -2.99 0.8077 0.295 CONSTANT 225.8 891.29 0 0.25 R-SQUARED: 0.808 ADJUSTED R-SQUARED: 0.774 DURBIN-WATSON: 1.63 F-RATIO: 24.14 - Page 130b -of observations and explanatory variables). Thus, the hypothesis of no autocorrelation cannot be rejected. Turning to the simple R-squared results, the exchange rate would appear to have the largest explanatory power for Japanese net foreign direct investment. The interest rate differential has the least explanatory power. The overall R-squared for the regression is .81. Corrected for the distortion of the number of regressors, the adjusted R-squared is .77. These results suggest that there is a statistically significant negative relationship between net foreign direct investment and real wages, real land prices and the Bank of Japan-U.S. Federal Reserve discount rate spread and a positive relationship between net foreign direct investment and the exchange rate. 8.3. THE PRIVATE SECTOR BALANCE Regression models were also developed for specific components of identity (7), namely (S-I) and NFPI. These regressions were conducted to test the model hypotheses contained in section 8.1, that explanatory variables used in NFDI model should be determinants of the components of identity (7) . In order to reinforce the rationale behind the NFDI statistical approach, therefore, these regressions should show strong relationships between the components (S-I) and NFPI and the fundamental determinants used in the NFDI model. Page 131 -The following sections examine the rationale and results behind the model developed for the private sector balance. It is important to determine the factors behind the private sector savings and investment differential from the perspective of determining the amount of excess capital in the domestic economy which is available for offshore investment. Three explanatory variables were selected to be regressed against (S-I), namely; * the real wage rate index (W); * the real land price index (L); and, * the real Bank of Japan discount rate (D). 8.3.1. (S-I) Model; Hypotheses and Rationale Real wages were chosen as an explanatory variable because of their positive impact on real incomes and household savings. (Corporate savings, or retained earnings, could be negatively impacted by rising real wages; however, rising real wages should also be suggestive of economic growth and rising corporate profits.) From the perspective of investment, rising real wages increase operating expenses and reduce profit expectations of firms. As such, rising real wages should have a negative impact on planned investment. Based on this, it was hypothesized that the real wages should be positively correlated to the private sector balance. The discount rate should also have a positive correlation with the private sector balance. Classical economic theory postulates - Page 132 -that present consumption is dependent on current and future income, the rate of interest and family preference. If the real interest rate rises, the present value of future income declines and current consumption will decrease. The corollary to this is increased current savings. In terms of investment, rising real interest rates result in a higher cost of funds thereby deterring current investment. The relationship of land prices with the private sector balance is less intuitive. The level of savings could be stimulated by rising residential land prices as potential home-buyers set aside more money for larger down-payments. Conversely, Japanese land and housing prices have increased so dramatically over the last four years that home-ownership may be out of the reach of an increasing proportion of the population. If the latter is correct and potential home-buyers abandon the goal of home-ownership, savings previously^ set-aside for this purpose may be released for consumption, thereby reducing the savings rate. However, as noted in Section 6, savings rates have declined only marginally in spite of land price increases in excess of 70% per year in Tokyo since 1985. On balance, it is assumed that home ownership remains a fundamental objective of Japanese society and that real land price increases should have a positive impact on savings. The impact of rising land prices on investment is clearer. Dramatically higher land prices act as a deterrent to new developments at home. On the other hand, companies with large land - Page 133 -holding inventories have seen a massive appreciation of their asset base; and a corresponding increase in the market assessment of their stock values. The large steel-making companies offer clear examples of land rich companies which have experienced huge increases in their market assessed value, while having poor profit performance in their ongoing operations. In many cases, land-rich companies have taken the opportunity to make new share issues, thus using their land-fuelled prices to generate new equity funds. Some of these new funds have been directed towards FDI. Higher land values have also provided the collateral for direct foreign investment in real estate both by corporations and individuals. In fact, loans made by Japanese financial institutions for foreign real estate investments are typically secured by mortgages on domestic property, with the foreign real estate providing only secondary security (139) . Thus, we have assumed that high land prices are positively correlated with savings and negatively correlated with domestic investment. Higher land prices have increased the need for down-payment related savings and have retarded plant development at home because of increased start-up costs. At the same time, however, foreign direct investment has become more attractive because the high value of domestic real estate acts as a pool of ready collateral for offshore investments. On this basis, the (S-I) 139 • • William Krueger, "Japanese Real Estate: At Home and Abroad," Tokyo Business Today, Dec. 1988. p. 37. - Page 134 -balance should be positively correlated with rising land prices. Given the above hypotheses, the original model for S-I was, PRIVt = ot + B^(Ml)t + B2(L)t + B3(D)t + et (1) where PRIV is the private sector savings-investment balance, W is real wages; L is real land prices; and D is the discount rate. The error term, e, is distributed normally with mean 0 and constant variance a2 . 8.3.2. (S-I) Model; Data Sources All data for the following regressions are listed in Appendix 1. The data for the private sector balance were taken from annual data provided by the Japanese Economic Planning Agency in the Annual Report on National Income Statistics (years 1970 and 1971) and the Annual Report on National Accounts (years 1984 and 1987). Data for 1986 and 1987 were taken from the OECD Report on Japan, 1987 edition. The data cover the period from fiscal year 1960 to fiscal year 1987 and are recorded in billions of yen. The explanatory variables, real wages, real land prices and the Bank of Japan discount rate are as described above in 8.2.2. 8.3.3. (S-I) Model; Results Results for the (S-I) regression proved inconclusive due to limitations in the data set. A number of factors could explain these results. Included among these is the strong - Page 135 -multicollinearity between the explanatory variables as is evidenced in the correlation matrix for regression (1) listed in Table 8.3. (Correlations between the discount rate and wages and land prices were particularly high i.e., -.95 and -.76, respectively). Multicollinearity is a problem because it leads to high variability in the OLS estimators. This means that the estimated coefficients are imprecise and cannot be used reliably to test the underlying hypotheses of the model. However, it is also likely that the model is mis-specified in some way (i.e. that the explanatory factors don't fully incorporate the true relationship between savings and investment). In addition, important determinants may have been omitted from the model which might bias the results. The results of regression (1) are shown in Table 8.4. Contrary to the original hypotheses, the estimated coefficients for all explanatory variables showed a negative relationship with the dependent variable (S-I). The t-value for wages was insignificant (-0.10729). In addition, the Durbin Watson statistic was low (1.072) suggesting some first order auto-correlation of the error term. Further regressions, contained in Table 8.5, which regressed the explanatory variables independently against (S-I) showed strong positive relationships between wages and (S-I) and between land and (S-I), but a negative relationship between (S-I) and the discount rate. However, because of the wide deviations of these - Page 13 6 -TABLE 8.3.: CORRELATION MATRIX FOR THE PRIVATE BALANCE REGRESSION WAGES LAND DISCOUNT RATE PRIVATE BALANCE WAGES LAND 1.00 0.80 0.80 1.00 -0.95 -0.77 0.85 0.61 DISCOUNT PRIVATE RATE BALANCE -0.95 0.85 -0.77 0.61 1.00 -0.92 -0.92 1.00 - Page 136a -TABLE 8.4.: RESULTS OF PRIVATE BALANCE REGRESSION 1 DEPENDENT VARIABLE: PRIV INDEPENDENT VARIABLES: UA6E, LAND, DISC VARIABLE ESTIMATED STANDARD T-RATIO PARTIAL STANDARD. ELASTICITY NAME COEFFICIENT ERROR CORR. COEFFICIENT AT MEANS UA6E -7.0493 65.703 -0.10729 -0.0219 -0.026339 -0.1118 LAND -49.478 26.176 -1.8902 -0.36 -0.23476 -1.0379 DISC -133000 27517 -4.8333 -0.7023 -1.123 -2.8197 CONSTANT .24561 7751.5 3.1686 0.5431 0 4.9693 R-SQUARED: 0.8654 R-SQUARED 0.84B5 ADJUSTED: DURBIN WATSON: 1.072 F-RATIO: 51.421 - Page 136b -TABLE 8.5.: RESULTS OF REGRESSIONS OF INDIVIDUAL EXPLANATORY VARIABLES ON PRIVATE SECTOR BALANCE DEPENDENT VARIABLE: PRiV INDEPENDENT VARIABLES: WAGE VARIABLE ESTIMATED STANDARD T-RATIO PARTIAL STANDARD. ELASTICITY NAME COEFFICIENT ERROR CORR. COEFFICIENT AT MEANS WAGE 227.34 27.698 8.2078 0.8494 0.84943 3.6055 CONSTANT -12878 2279.5 -5.6494 -0.7423 0 -2.6055 R-SQUARED: 0.7215 R-SQUARED 0.7108 ADJUSTED: DURBIN WATSON: 0.8825 F-RATIO: . 67.369 - Page 136c -TABLE 8.5 CONT'D: RESULTS OF REGRESSIONS OF INDIVIDUAL EXPLANATORY VARIABLES ON PRIVATE SECTOR BALANCE DEPENDENT VARIABLE: PRIV INDEPENDENT VARIABLES: DISC VARIABLE NAME DISC CONSTANT ESTIMATED STANDARD COEFFICIENT ERROR T-RATiO PARTIAL CORR. STANDARD. ELASTICITY COEFFICIENT AT MEANS -108670 16329 9234 1100 -11.758 -0.9176 -0.91757 -2.3038 14.845 0.9458 0 3.3038 R-SQUARED: 0.8419 R-SQUARED 0.8359 ADJUSTED: DURBIN WATSON: 0.9032 F-RATIO: 138.484 - Page 136d -coefficients from those produced in regression (1) , the results are also suspect. Again, these regressions also had low Durbin-Watson statistics. Because of the low Durbin Watson statistics, the above regressions were all corrected for first order auto-correlation. The results are contained in Table 8.6. and 8.7. Although the Durbin Watson statistic of the aggregate regression (Table 8.6.) improved substantially, the t-value for wages was insignificant. Only the discount rate showed strong significance in the corrected aggregate regression. In terms of the corrected individual regressions (Table 8.7.), real wages showed a significant positive relationship with the private sector balance as predicted by the hypothesis of 8.3.1. However, once corrected, the individual regression of land on PRIV became insignficant (t-value of 0.18). The results for the discount rate were more interesting, demonstrating a strong negative relationship between the private balance and the discount rate. These results necessitate some re thinking of the original hypothesis that the discount rate should have a positive relationship with the private sector balance. Savings rates in Japan have been relatively constant (with the exception of the years immediately following the 1973 oil crisis). In addition, interest yields on savings have remained regulated at low levels. Given this, the negative parameter estimate is more likely explained by the relationship between investment and the - Page 137 -TABLE 8.6.: RESULTS OF PRIVATE BALANCE REGRESSION 1 CORRECTED FOR FIRST ORDER AUTO-CORRELATION DEPENDENT VARIABLE: PRIV INDEPENDENT VARIABLES: WAGE, LAND, DISC VARIABLE ESTIMATED STANDARD T-RATIO PARTIAL STANDARDIZED ELASTICITY NAME COEFFICIENT ERROR CORRELATION COEFFICIENT AT MEANS WAGE LAND DISC CONSTANT 5.8604 -37.283 -120690 21213 77.296 28.665 28854 8342.6 .075818 -1.3006 -4.1826 2.5427 .0155 .2566 .6493 .4607 .021897 -.1769 -1.0191 0 .092942 -.78205 -.25586 4.2918 R-SQUARED: R-SQUARED ADJUSTED: .8891 .8753 DURBIN-WATSON: 2.0393 - Page 137a -TABLE 8.7.: RESULTS OF-REGRESSIONS OF INDIVIDUAL EXPLANATORY VARIABLES, CORRECTED FOR FIRST ORDER AUTO-CORRELATION DEPENDENT VARIABLE: PRIV INDEPENDENT VARIABLE: WAGE VARIABLE ESTIMATED STANDARD T-RATIO PARTIAL STANDARDIZED ELASTICITY NAME COEFFICIENT ERROR CORRELATION COEFFICIENT AT MEANS WAGE 224.19 47.143 4.7556 .6821 .83767 3.5555 CONSTANT -12360 3887.7 -3.1793 -.5291 6 -2.5008 R-SQUARED: .8044 R-SQUARED .7969 ADJUSTED: DURBIN-UATSON: 1.9135 - Page 137b -TABLE B.7. CONT'D: RESULTS OF REGRESSIONS OF INDIVIDUAL EXPLANATORY . VARIABLES, CORRECTED FOR FIRST ORDER AUTO-CORRELATION DEPENDENT VARIABLE: PRIV INDEPENDENT VARIABLE: LAND VARIABLE ESTIMATED STANDARD . T-RAT10 PARTIAL STANDARDIZED ELASTICITY NAME COEFFICIENT ERROR CORRELATION COEFFICIENT AT MEANS LAND -6.8307 37.145 -.1839 -.036 -.03241 -.14328 CONSTANT 6519.5 6192.4 1.0528 .2022 0 1.319 R-SQUARED: .7565 R-SQUARED .7471 ADJUSTED: DURBIN-WATSON: 1.8045 TABLE 8.7. CONT'D: RESULTS OF REGRESSIONS OF INDIVIDUAL EXPLANATORY VARIABLES, CORRECTED FOR FIRST ORDER AUTO-CORRELATION DEPENDENT VARIABLE: PRIV INDEPENDENT VARIABLE: DISC VARIABLE ESTIMATED . STANDARD T-RATIO PARTIAL STANDARDIZED ELASTICITY NAME COEFFICIENT ERROR CORRELATION COEFFICIENT AT MEANS DISC -106930 13066 -8.1838 -.8487 -.9029 -2.267 CONSTANT 16393 1616 10.144 .8935 0 3.3166 R-SQUARED: .8811 R-SQUARED .8765 ADJUSTED: DURBIN-WATSON: 2.0337 - Page 137c -discount rate. Several factors could be at work. First, the discount rate may not be an appropriate proxy for the interest rates charged for corporate borrowing. Second, the Japanese discount rate might be positively correlated with other world interest rates but demonstrate a lower variance over time. Thus, although Japanese domestic interest rates may have been rising, interest costs at home would have remained relatively low promoting domestic versus foreign investment. Finally, assuming again that Japanese discount rates were positively correlated with world interest rates, if the latter were rising, this might imply unstable economic conditions abroad (i.e. impending recession, etc) . Assuming investors are risk-averse, under these conditions, domestic investment would be preferred to foreign investment. 8.4. NET FOREIGN PORTFOLIO INVESTMENT Two explanatory variables were chosen to be regressed against net foreign portfolio investment (described as POR in the regression model). Again, the variables include annual data extending from fiscal year 1960 to fiscal year 1987. The variables chosen were: * The Yen/U.S. dollar (average annual) spot rate, (EXCH); and, * The difference between the Bank of Japan and U.S. Federal Reserve (real) discount rates, (SPR). - Page 138 -8.4.1. NFPI Model; Hypotheses and Rationale Because of the large proportion of U.S. securities in Japanese foreign portfolio investment (approximately 60-65%), the Yen/U.S. dollar exchange rate and the differential between the Bank of Japan and the U.S. Federal Reserve discount rates were chosen as possible explanatory variables for Japanese NFPI. The spot rate was used as the explanatory variable although forward rates could be considered more appropriate as offshore portfolio investment is motivated by expectations of future, not current, exchange rates. By definition, portfolio investment includes instruments with maturities of 1 year or more. However, the market for forward rates becomes thin after 180 days and past data on forward rates in excess of 12 months is difficult to find. In addition, academic studies ( ) suggest that forward rates incorporate other factors in addition to exchange rate expectations (such as a premium for aiding the hedging of exchange rate risk, transaction costs, and government intervention) which can result in substantial deviations between forward and actual future spot exchange rates (141) . Empirical studies indicate that the spot rate tends to follow a random walk process such that the "market's best forecast of the See, for example, Richard M. Levich, " Are Forward Exchange Rates Unbiased Predictors of Future Spot Rates," Columbia Journal  of World Business 14. No.. 4. (Winter 1979). p. 49-61. Richard M. Levich, "Evaluating the Performance of the Forecasters," International Financial Management, Donald R. Lessard edit., New York: John Wiley and Sons, 1985, p. 218. - Page 139 -future spot rate is (approximately) the current spot rate." ( ) On the other hand, given that the explanatory variable chosen incorporates only average annual data, the spot rate may not adequately describe the actual international financial arbitrage effect. Turning to the interest rate differential, discount rates were used because of the difficulty in finding similar interest-bearing instruments over the period in question in the U.S. and Japanese capital markets. The discount rate serves as a basic indicator of all other interest rates in the economy and can, therefore, be viewed as an appropriate proxy. The relationship between the Yen/U.S. dollar exchange rate and NFPI is not intuitively obvious. If the current spot rate is useful in predicting future spot exchange rates, as suggested by the random walk hypothesis, as EXCH rises (implying a depreciation of the yen) then FPI by Japanese will increase (and FPI by foreigners in Japan will decrease) resulting in a larger outflow of net Japanese foreign portfolio investment (i.e. Japanese net foreign portfolio investment will become more negative). This would tend to support a negative correlation between the spot exchange rate and NFPI. This rationale might explain continued record amounts of investment by Japanese in foreign securities during 1986 (net purchases of $92 billion) despite the 30% Jacob A. Frenkel, "Flexible Exchange Rates, Prices and the Role of *News'," International Financial Managementf Donald R. Lessard, edit., New York: John Wiley & Sons, 1985, p. 131. - Page 140 -appreciation of the yen during the same year. Japanese investors mistakenly assumed that the yen appreciation had peaked. In general, assuming the random walk theory holds, the relationship between the spot exchange rate and NFPI should be negatively correlated. Turning to the interest rate differential (the BOJ discount rate minus the U.S. Federal Reserve discount rate), if SPR is increasing (implying a widening spread between the Bank of Japan and U.S. discount rates) , foreign portfolio investment by Japanese will tend to decrease and portfolio investment in Japan by foreigners will tend to increase. This should lead to a positive increase in Japanese NFPI. Similarily, if SPR is decreasing (implying a narrowing differential between the BOJ and U.S. discount rates), foreign portfolio investment by Japan will tend to increase and portfolio investment in Japan will tend to decrease resulting in a larger outflow of foreign portfolio investment. Thus, the relationship between NFPI and SPR should be positive. Thus, the model for NFPI can be expressed as: PORt = Qtt + fi1(SPR)t - fi2(EXCH)t + et (2) where POR is net foreign portfolio investment and e « N(0,a2 ) . 8.4.2. NFPI Model; Data Sources Data for NFPI were taken from the IMF's International Financial  Statistics. 1988 for Japan. The data, as discussed, are entered - Page 141 -into the model as they appear on the balance of payments (i.e. negative numbers for net outflows and positive numbers for net inflows) and is in billions of yen. The data for the explanatory variables, EXCH and SPR are as described above in section 8.2.2. 8.4.3. NFPI Model; Results The results of the regression (2) are listed in Table 8.8. The signs of the estimated coefficients were opposite to what was predicted in section 8.4.1. Because of the low Durbin Watson statistic of Table 8.8. (.68), the regression was corrected for first-order auto-correlation. These results are contained in Table 8.9. The Durbin Watson statistic remained low (1.35), suggesting second-order correlation. The estimated coefficients for both explanatory variables differed significantly from the regression contained in Table 8.8. (although the signs remained the same) and the T-value for SPR became insignificant. These results are largely attributable to the severe auto-correlation of the error terms and, because of this, cannot be used reliably to prove or disprove the hypotheses contained in Section 8.4.1. Page 142 -TABLE 8.8.: RESULTS OF NET FOREIGN PORTFOLIO INVESTMENT HODEL DEPENDENT VARIABLE: POR INDEPENDENT VARIABLES: SPR, EXCH VARIABLE NAME EXCH SPR CONSTANT ESTIMATED STANDARD COEFFICIENT ERROR 56.65 -38939 -18092 11.499 14637 3459.3 T-RATIO PARTIAL STANDARD. ELASTICITY CORR. COEFFICIENT AT MEANS 4.9264 0.7018 0.9673 -13.117 -2.6604 -0.4697 -0.52237 0.0040636 -5.2299 -0.7226 0 14,113 R-SQUARED: R-SQUARED ADJUSTED: DURBIN WATSON: F-RATIO: 0.5043 0.4646 0.6774 12.175 - Page. 142a -TABLE 8.9.: RESULTS Of HE I FOREIGN PORTFOLIO INvtSTMENT HODEL CORRECTED FOR FIRST ORDER AUTO-CORRELATION DEPENDENT VARIABLE: PGR INDEPENDENT VARIABLES: SPR, EXCH VARIABLE ESTIMATED STANDARD T-RATIO PARTIAL STANDARD. ELASTICITY NAME COEFFICIENT ERROR CORR. COEFFICIENT AT MEANS EXCH 22.168 10.314 2.1493 0.3949 0.37852 -5.1327 SPR -737.86 18196 -0.0406 -0.0081 -0.0099 0.0001 CONSTANT -9391.1 4180.5 -2.2464 -0.4098 0 7.3256 R-SQUARED: 0.8032 R-SQUARED 0.7874 ADJUSTED: DURBIN WATSON: 1.3507 - Page 142b -8.5. LIMITATIONS OF THE NFDI RESULTS As indicated in the preceding sections the results for the component regressions did not generally support the rationale underlying the choice of explanatory variables for the NFDI model. In addition, although the results listed in Table 8.1 appear to corroborate the hypotheses of section 8.2.1., a number of important qualifications must be noted. These qualifications restrict the value of the approach as a model for Japanese foreign direct investment. 8.5.1. Multicollinearity among the Explanatory Variables A correlation matrix for the explanatory variables used in the NFDI model is provided in Appendix 2., page 2.2. As noted earlier, real wage and land prices are closely correlated (.80) as are real wages and the spot exchange rate (-.84) and the interest differential and real wages (-.89). Multicollinearity between the explanatory variables can result in large standard errors in the OLS parameter estimates so that the precision of the coefficients may be suspect. The presence of multicollinearity, therefore, makes it difficult to disentangle the marginal effects of explanatory variables on net foreign direct investment. - Page 143 -8.5.2. Auto-Correlation of the Residuals As noted in 8.2.3., the Durbin-Watson statistic for the second NFDI model is 1.63. While this statistic is near the upper limiting distribution, there is still some suggestion of positive auto-correlation in the residuals. The existence of the latter tends to bias the R-squared upwards and may suggest some mis-specification in the model. With auto-correlation present, the standard deviations of the coeficients are under-estimated and thus, the coefficients of the explanatory variables may appear significant when they are not. 8.5.3. The Problem of Endoqeneity The approach has used the Ordinary Least Squares (OLS) technique to estimate the regression coefficients of the model. The OLS technique assumes that the explanatory variables are distributed independently of the residuals; if this is not the case then the regression coefficients can be biased, even asymptotically. Since the model described in this study makes use of macro-economic data, there is a strong likelihood that the regressors are not independent of the residuals and that a more formal approach using simultaneous equations is warranted. 8.5.4. Structural Change If significant structural changes occur in an economy over a period under consideration, the regression coefficients may not be - Page 144 -stable over time and, therefore, may not be reliable estimators of the true effects of the independent variables. Given the macro-economic developments in the Japanese economy described in Section 6, it was decided to test for structural changes. In particular, regressions were run on foreign direct investment using the previously listed regressors for 2 time periods: 1960 to 1971 and 1972 to 1987. The demarcation year of 1971 was chosen for two reasons: (1) in 1971, the world moved from a fixed to floating exchange rate system, with the yen appreciating by almost 20% between 1970 and 1973; and (2) by 1971, the internal pressures associated with a decade of rapid economic growth (i.e., high land prices, higher real wages, higher inflation, etc.) were manifested in the dipping of Japanese real growth rates from 10% during the 1960s to 4.3% in 1971. Although 1973 could have also been chosen as the demarcation year, as noted in Section 6.0, the evolution towards slower growth in Japan began earlier. The regression results of these two time periods are shown in Table 8.10. The regression coefficients differ significantly between periods, with the signs of the LAND, EXCH, and SPR changing over the two time frames. To further illustrate this problem, the individual regressors were plotted against foreign direct investment for the two time frames. The plots of these data are contained in Appendix 2, pages 2.33 through 2.36 for 1960 to 1971 and pages 2.37 through 2.40 for 1972 to 1987. These plots indicate that the location of the sample data fluctuated considerably over - Page 145 -TABLE 8.10: TESTING FOR STRUCTURAL CHANGE - 1960 TO 1971 DEPENDENT VARIABLE: DIP INDEPENDENT VARIABLE:. WAGE, LAND, EXCH, SPR VARIABLE ESTIMATED STANDARD STANDARDIZED T-RATIO PARTIAL ELASTICITY . NAME COEFFICIENT ERROR . COEFFICIENT CORRELATION AT WEANS WAGE LAND EXCH SPR CONSTANT -6.8838 1.4295--7.334S 116.25 1.342 .60897 3.0645. 398.57 1170.7 -1.547S .72221 -.44694. .10435 0 -5.1058 2.3474 -2.3934 .29677 .8879 .6637 .6709 .1115 .6703 4.8159 -1.055 35.862 .079313 -38.093 R-SQUARED: P-SOUARED ADJUSTED: F:RAT 10: DURBIN WATSON .8886 .8249 13.956 - Page 145a -TABLE 8.10 CONT'D: TESTING FOR STRUCTURAL CHANGE - 1972 TO 1387 DEPENDENT VARIABLE: DIR '. INDEPENDENT VARIABLE: WAGE, LAND, EXCH, SPR VARIABLE ESTIMATED STANDARD STANDARDIZED T-RATIO PARTIAL ELASTICITY •NAME COEFFICIENT ERROR COEFFICIENT CORRELATION AT MEANS WAGE LAND EXCH SPR ' •CONSTANT -62.145 -IE.316 -2.2909 -£96.92 7978 15.643 2.718 589.1 2581.8 -.79489 -.64866 -.22792 .029726 0 -3.9727 -2.66 -.84288 -.12505 3.091 .7676 .6257 .2463 .0377 .6817 6.3883 2.3558; .59713 -.026904 -8.3143 R-SE'UARED: R-SQUARED ADJUSTED: . FiRATIO: DURBIIM WATSON .8239 .7598 12.863 .279 - Page 145b -the two periods. As such, the use of a simple linear regression equation to model the data is questionable. The large deviations of the coefficients suggest either of two problems: (1) that structural change did occur in the Japanese economy making it difficult to use regression analysis to model direct foreign investment, or (2) the model is mis-specified. In either case, the regression results listed in Table 8.7 are necessarily suspect. 8.5.5. Problems in the Data The problems listed above are manifestations of the limited information contained in the data set. A number of problems with the data exist, some of which have already been alluded to. These problems include the use of annual averages for the spot rate and interest rates. As noted, the former was probably not appropriate as an explanatory factor for NFPI because of its inability to describe the actual international arbitrage effect. From the perspective of NFDI, given its longer time frame, the use of the annual average spot rate is not unreasonable. The use of indexes for the real wages and land prices may also be inappropriate. In the case of real wages, the index incorporates the averages for a broad spectrum of industries which may cloud the relevant data if the impact of real wages on foreign direct investment is sector-specific. For example, the real wage index includes wages for sectors which do not engage in direct foreign investment such as agriculture and certain service sector - Page 146 -industries. Finally, the land index incorporates the average and land values (in the case of real land prices). These averages may understate the true explanatory power of each variable. In addition, the data for the government and private sector balances came from two or more sources which could have resulted in further measurement problems. Finally, the data size was relatively small including only 28 (annual) observations. 8.6. SUMMARY A number of conclusions can be gathered from the foregoing statistical attempt to explain Japanese net foreign direct investment. Based on the regression results contained in Table 8.1., there is some support, albeit weak, for the hypothesis that the determinants of the components of identity (7) are also determinants of Japanese net direct foreign investment. However, the model suffers from a number of problems which are particularly manifested in the results of the component regressions. These problems, as discussed, lie largely with the data set and the statistical problems noted in Section 8.5. and not necessarily with the underlying hypotheses of the model. The results and their associated problems underscore the limitations of the data used in the model and the difficulties of using the OLS technique with macro-data to determine the factors behind Japanese net direct foreign investment. - Page 147 9.0. JAPANESE FDI; THE FUTURE The future levels of Japanese direct foreign investment will be shaped by four considerations: a. the availability of funds for foreign investment; b. the requirements of the Current Accounts balance: c. the allocation priorities for these funds; and, d. the acceptability of Japanese foreign direct investment in the recipient countries. 9.1. AVAILABILITY OF FUNDS Any discussion of the future availability of Japanese foreign investment funds requires that we return to the macroeconomic identity introduced in section 5.0. It will be recalled that this identity can be stated as: NFDI + NFPI + SC + F = (S-I) + (G-I) + E where: NFDI = net foreign direct investment NFPI = net foreign portfolio investment Sc = short term capital investment F = private and government monetary flows (S-I) = private sector balance (savings - investment) (G-I) = government sector balance E = errors and omissions. - Page 148 -In the case of Japan, errors and omissions are random in nature, usually quite small and, for current purposes, can be ignored. Private and government monetary flows may also be disregarded. Thus the total, internally-funded amount of money available for foreign investment (short and long term) can be assumed to be equal to the net private sector and government sector balances. 9.1.1. Private Sector Balance The private sector balance reflects the levels of household savings, private residential investment, corporate savings and corporate capital outlays. 9.1.1.1. Household Savings As in the past, future levels of household savings will be a function of the strength of the economy (GNP) , real wage rates, the propensity to consume, the general perception of future security, the cost of real estate and the inherent tendency of the Japanese to save. The general feeling is that the economy, buoyed by capital spending and strong domestic consumer demand, will stay reasonably healthy over the medium term, at least. Key industries, such as automobile manufacturing, have made very significant productivity gains to counter-balance the impact of the high yen. In fact, it has been stated that the automobile industry could be competitive in world markets even if the exchange rate rose as high as 95 yen - Page 149 -to the U.S. dollar( ) . The programs of rationalization and diversification of the established, heavy industrial companies are continuing, despite a number of initial set-backs. For example, since 1987 Nippon Steel has closed 4 of its 8 steelworks, will reduce its 1991 workforce to only 46% of the 1987 level, and expects a half of its 1995 revenue to accrue from non steel-making activities, as opposed to the current 20 percent (UA). High utilization rates, strong corporate profits and a fairly tight labor market appear likely to continue the upward pressure on wages. However, the expanded use of automation, the decline in the ratio of organized labor, and the current round of efficiency enhancing investments will probably keep such increases within reasonable bounds. At the same, inflation is well under control (despite the appreciation of the yen) and this situation will be supported by the increased availability of imported goods. All in all, the prognosis would seem to be for a controlled, steady growth in real wages (145) . Increased domestic consumption is being encouraged by a number of government initiatives. The recent reduction in personal income Kevin Done, "Car Wars After the Yen Shock", The Financial  Times. London, May 12, 1989. "Japan's smokestack fire-sale"; The Economist; 19-25 August, 1989; p.51. "Labor supply and wage costs"; Japan Times. August 19, 1989; p. 7. - Page 150 -taxes and MITI1s support of tax incentives to promote imports ( ) are examples of such initiatives. Conversely, the 3 percent consumption tax, introduced in April, will tend to restrict consumption. On balance, it can be expected that consumer spending will remain strong, but not more so than the anticipated growth in real purchasing power. Furthermore, the restraint shown last winter, during the long illness of the Emperor, is a clear indication that frugality is not far below the surface of the average Japanese consumer's buying habits. As noted in section 6, the general level of confidence in Japan, both in business and at the individual level, is high; in addition there has been a significant improvement in the social services programs. Nevertheless, the in-grained habits of prudently providing for the future will die hard in Japan. Finally, land prices in the Tokyo metropolitan have recently begun to rise again, after slight declines during the past 2 years (147) . The uncertain impact of higher real estate prices on savings has already been noted. Savings could go up as potential buyers put aside money for down payments which increase in lockstep with housing prices. On the other hand, higher prices could force potential buyers out of the market. However, during the third quarter of fiscal 1988, housing investment increased by 5.6 "MITI supports tax incentives to promote imports"; Japan  Times; August 18, 1989; p.12 "Sumitomo Realty and Development Co.", Japan Times. Tokyo, August 19, 1989, p.9 - Page 151 -146 147 percent, despite a significant reduction (to 1.62 million) in housing starts (U8). Overall, it is reasonable to assume that the current levels of real estate related savings will be maintained. In summary, household savings in Japan should remain high for the foreseeable future. During the past few years, when Japanese foreign investment has exploded, the level of household savings has stayed fairly constant at around 15 to 16 percent. At this time there is little, if any, evidence to suggest that this will change. 9.1.1.2. Private Residential Investment. Notwithstanding the very high cost of housing, home ownership in Japan remains at a level similar to that in the U.S.A. and is a basic life-style objective for Japanese people. It seems likely that this priority will be continued in the future and that savings and other consumption patterns will reflect this fact. As noted in the above discussion of household savings, the housing starts in the third quarter of fiscal 1988 amounted to an annualized rate of 1.62 million, about 9.3% below the comparable period in 1987. However, the average size of housing in 1988 was about 3.2% higher than in 1987 and total expenditures were about 6.5% higher than the 20.8 trillion yen spent on residential housing in 1987. For fiscal 1989 projected expenditures are about 5% lower than fiscal 1988 "The Japanese Economy"; NRI Quarterly Economic Review  Vol.19. No.2; Tokyo, NRI & NCC Co.Ltd.; May 1989; p.2 - Page 152 -( ) . In general, future residential construction may be somewhat reduced from the high levels of 1987, largely because the rental housing boom has run its course. Over the longer term, the "aging" of the Japanese population should introduce some structural reduction in the number of new homes required, but the latent demand is such that this structural process will occur over an extended period. 9.1.1.3. Corporate Savings The corporate pre-tax profit performance of Japanese companies in fiscal 1988 for all industries was 27.4% higher than in 1987 and a recent survey indicated that 79% of companies contacted expected further increases in 1989. (150) In addition to strong current operations (due to robust consumer demand, high levels of corporate investment and, in some industries, notable improvements in productivity), many of the companies have taken advantage of the high market prices of their stock (the average P/E ratio for the NRI 400 Composite was 59.1, as of March 1989 151) . These high prices, often supported by the huge increases in the value of land holdings, has enabled companies to reduce debt loads by issuing Nomura Quarterly Review Vol.19 No. 2; Tokyo: NRI & NCC Co. Ltd., May 1989; p.46. 150 Quarterly Economic Review. May 1989, Nomura Research Institute, Tokyo, p. 19-21. NRI Quarterly Economic Review. Vol. 19 No.2; Nomura Research Institute; Tokyo; May 1989; p.66. - Page 153 -treasury stock on advantageous terms. The steel industry is a good example of this trend, raising 210 billion yen in 1987 (152) . Stronger equity positions and high current earnings have enabled companies to make significant increases in their retained earnings. Overall, the majority of Japanese companies are stronger today than they were before the rapid appreciation of the yen, and the recent round of efficiency increasing investments will add to this strength. Further, the profit motive appears to be assuming a more dominant role in the strategic planning of many Japanese companies. Thus it is reasonable to assume that retained earnings (corporate savings) will continue to improve. 9.1.1.4. Corporate investment As noted previously, corporate investment has in the past couple of years reversed the downward trend of the post-1974 period. Within Japan, the motivation for new investment seems to be shifting from increased capacity (to meet foreign and, recently, strong domestic demand) to production efficiency and the development of new businesses (153) . The level of investment will continue to be supported by the search for productivity and labor saving improvements and by the recent cut in corporate income tax "Japan's smokestack fire-sale"; The Economist; The Economist Newspaper Ltd, London; 19-25 August, 1989; p.51-52. "The Japanese Economy"; NRI Quarterly Review Vol 19 No.2; Nomura Research Institute, Tokyo; May 1989; p.6. Page 154 -rates. However, the expectation is that the current spending boom will gradually slow down. MITI has already suggested that car makers reduce their investment plans because of potential over capacity (154) . Finally, the higher levels of corporate savings will mean higher equity shares in future capital investments, with a corresponding reduction of demand on the household savings surpluses. 9.1.1.5. Summary of Private Sector Balance In summary, the Private Sector balance can be expected to continue in strong surplus for at least the next few years. Household savings, despite government encouragement to increase domestic demand, is not likely to decrease significantly. Residential construction, with pent-up demand tending to be off set by high prices, should remain at, or near current levels. There is a good chance that the trend towards stronger profits and higher corporate savings will continue. On the other hand, corporate investments can be expected to remain stable, or gradually decline. Much of the investments which will be made will be in the area of improved efficiencies, which will tend to support profits and corporate liquidity. "MITI to warn carmakers to go slow"; The Japan Times; Tokyo; August 17, 1989; p.12. - Page 155 -9.1.2. Government Sector As noted in section 6.2, the latter half of the 1970s and early 1980s eighties were characterized by large government deficits which soaked-up much of the private sector surpluses. Since 1978, the government has implemented fiscal austerity measures designed to achieve fiscal balance. In the medium term, the most important concerns about Japanese fiscal policy will be the extent to which expenditures are controlled, income collection improved, and borrowing strategy rationalized. 9.1.2.1. Expenditures Given the strong performance of the Japanese economy, the use of public works spending as a stimulant can be expected to decline over the near and longer term periods. Conversely, there will be a gradual increase in spending on social welfare due to expanded participation in the program of retirement benefits and the need for ongoing improvements to the system. The net effect should be gradual and controlled increases in government expenditure, at growth rates well below those of the private sector. 9.1.2.2. Income The government income in Japan has long been plagued with an inequitable and inefficient system of tax collection. Personal income tax burdens have been most unfairly distributed, with wage earners bearing a disproportionate share of the payments made. - Page 156 -Small businesses, professional people and, especially, farmers have • • • • 155 not paid their fair share because of income concealment ( ). The 3% consumption tax, introduced in April was intended to be evasion-proof and to raise about 10 percent of the required government budget. Public opinion has been strongly opposed to the tax and, given the recent difficulties of the LDP party, it is not clear how strongly the Kaifu administration will maintain the tax's integrity. If the tax is maintained in its present broad form, the government's progress towards a balanced budget should be maintained. In the alternative, the government will have to find other sources of income or face growing deficits. All indications, especially the reactions of the Ministry of Finance and Keidanren (the leading business association), suggest that the government will display resolve in this matter. In terms of the mechanics of raising debt funds, the government will move away from longterm bonds towards increased issues of treasury and financial bills. This will reduce the problems now being encountered as large amounts of 10-year bonds fall due and, at the same time, help to satisfy foreign demands that Japan's capital markets be further developed. "Deficit-ridden Japan caught between tax revolt and demographics"; Globe and Mail; Toronto; August 22, 1989; p. B7. - Page 157 -9.1.2.3. Summary of Government Sector Balance Based on the available evidence, the most probable scenario for the Government Sector balance is one of gradual increases in expenditure, largely in the area of social services, and continuing attempts to improve the fairness and efficiency of the tax collection system. The government will move towards a better balance of financing instruments rather than the previous concentration on longterm bonds. Overall, the government posture is likely to be one of fiscal responsibility and an objective of balanced budgets. Thus, it is not likely that the government sector will resume its former role of absorbing a large part of the private sector surplus. 9.1.3. Summary of Macroeconomic Balance In terms of the availability of internally generated funds for Japanese foreign investment, the macroeconomic identity identifies the supply side as being the sum of the private sector and government sector balances. The preceding discussion suggests that neither of these balances is likely to experience sufficient change to materially affect the availability of funds for overseas investments. Household savings, the primary source of surplus funds, is not expected to diminish to any great extent. Residential construction will remain fairly constant, with a probability of slight declines from the 1987 peak. In the corporate sector, a number of factors point to strong profit - Page 158 -performances, with a consequent improvement in retained earnings and corporate liquidity. Corporate investments should remain buoyant, but it is improbable that the corporate sector balance will deteriorate. In the government sector, modest increases in social services expenditures should be offset by improvements in the tax collection systems, especially with the introduction of the visible and unavoidable consumption tax. 9.1.4. Other Factors In addition to the macroeconomic considerations, a number of underlying determinants will impact on the availability of funds for foreign investment. These factors include: a. the high prices of Japanese real estate; b. the huge capital earnings which have accrued from stock market appreciations; and, c. the earnings of off-shore investments. 9.1.4.1. Real Estate Market The unparalleled appreciation of Japanese real estate during the past few years has been extensively discussed in prior sections. This expanded, and un-measured, base of collateral available to Japanese investors has been a significant facilitator of Japanese foreign investment in the past couple of years. The continuing Page 159 -disparity between foreign (especially, North American) and Japanese real estate prices and interest rates is likely to encourage more Japanese foreign investment, both in fixed assets and interest earning securities. 9.1.4.2. Stock Market Earnings Very large profits have been earned on the Tokyo Stock Exchange in recent years, with the weighted average stock price in the First Section rising from 528.67 yen in 1984 to 1,333.72 yen in 1988. It seems probable that Japanese investors have used at least some of the gains earned on the Tokyo Exchange to buy more modestly priced, higher return stocks in foreign countries. 9.1.4.3. Off-Shore Earnings The large, and growing, inventory of direct foreign investment assets owned by Japanese investors will earn increasing amounts of profits. These profits (as will be discussed below) will, if repatriated, reduce the services deficit of Japan's Current Account balance, thus providing an equivalent off-set to any decrease which might occur in the country's Merchandise Trade. To the extent that these earnings are not repatriated, they will provide funds for Japanese foreign investment which are not accounted for by the macroeconomic identity. Clearly, should Japanese FDI continue at current rates, a point could be reached where the investments are self-sustaining. - Page 160 -9.2. THE REQUIREMENTS OF THE CURRENT ACCOUNT BALANCE As discussed in section 5.0, any surplus in the domestic economy, i.e., the sum of the private and government sectors, must be balanced by a current account surplus. Under the macroeconomic concept of Balance of Payments (BOP), the value of what leaves the country must equal the value of what enters the country. Thus any current account surplus must be balanced by an equal deficit in the capital account. The capital account is composed of long and short term foreign investment, government and private monetary flows plus errors and omissions. As discussed earlier, we can, for the purposes of this discussion think of the capital account as the outflow of long and short term investments. The current account surplus of the BOP is composed of the Merchandise Trade Balance (MTB) and the Services Sector Balance (SSB) . In the case of Japan, recent MTBs have been heavily in surplus; while SSBs have been in deficit. The SSB deficit exhibited a significant increase in 1988 as a result of greatly expanded foreign travel expenditures by the Japanese. Apart from this increase in foreign travel, the SSB deficit has been declining in recent years as foreign earnings grew. These earnings could be returns on short term or portfolio investments, or they could reflect profits earned by foreign direct investment assets. The former type of earnings are likely to be repatriated to Japan. However, the FDI profits may be, and often are, retained off-shore to expand the capital base of the company that produced them, or - Page 161 -to fund other foreign direct investments. The matter of repatriated earnings may assume some importance in Japan's future Balance of Payments strategies. As they grow, such earnings may reach a level whereby Japan can satisfy the wishes of its trading partners for reduced MTB, while still retaining a healthy Current Account Balance because of surplus Service Sector balances. This trend, which is already underway, will allow Japan to utilize its current MTB surpluses as the "Trojan Horse" by which Japanese companies are invited to compete in the "home" markets of Japan's trading partners. 9.3. THE ALLOCATION PRIORITIES OF FOREIGN INVESTMENT FUNDS The allocation of foreign investment funds is fundamentally concerned with net longterm investments. Short term investments cycle rapidly and the measurement of the net levels of such investment are of limited value. Thus, should any allocation procedure be required, it must select between FPI and FDI. The merit of FPI may be judged solely on the basis of dividend or interest returns, while the merit of FDI requires assessment of a broader range of returns. In considering the impact of any allocation process on either type of longterm investment, it is important to consider the relative sizes of the two demands for funds. Large amounts of FPI are a phenomenon of the post-1984 period. Prior to 1984, net FPI was often in deficit and seldom exceeded NFDI. However, since - Page 162 -1984, FPI has dominated net Japanese foreign investment and in 1988 amounted to about 72% of the total net longterm investment outflow. Of the US$89 billion of NFPI in 1988, only US$3 billion, or 3.4%, was associated with the purchase of stocks. In terms of total 1988 purchases, stocks accounted for only US$76.6 billion, or 5.3% of the US$1,440.6 billion of total FPI purchases. Thus, about 70% of all the 1988 net longterm foreign investment was associated with the purchase of bonds; largely in the U.S.A., and entirely motivated by the anticipations of higher returns. Performance-related foreign portfolio investments were restricted to about 2.5% of total net longterm investment in the stocks of foreign companies and about 28% of total net longterm foreign investment in the direct acquisition of control, or influencing, positions in foreign companies. 9.3.1. The Allocation of Funds to Direct Foreign Investment There are five principal types of considerations which cause funds to be allocated to foreign direct investment, namely: a. the comparative returns on foreign and Japanese assets; b. the comparative production costs of foreign and Japanese plant locations; c. trade frictions and barriers to market entry; - Page 163 -d. the procurement of resources which are not available in Japan and for which direct investment offers advantages of supply, security, or cost; and, e. strategic planning objectives. 9.3.1.1. Comparative Returns In a number of areas the cost of comparable assets is very much higher in Japan than it is in other countries, such as the U.S.A. Comparative real estate prices are a good example of this situation. Japanese investors are buying foreign real estate, especially in North America, because prices are lower and returns very much higher. The higher returns are particularly attractive when combined with the lower interest rates offered by Japanese financial institutions, often with Japanese real estate holdings being used as collateral. Individual Japanese are also finding foreign real estate to be very captivating. It was recently reported that wealthy Japanese buyers have shifted their targets from Hawaii to southern California, where they are buying showcase homes which cost $4 million, or more, each (156) . The current advantages of off-shore asset purchases should persist. There are no indications of dramatic reversals in the Japanese economy, the value of the yen, or the cost of real estate in Japan. Japan Times; Tokyo; August 16, 1989;p.10 - Page 164 -9.3.1.2. Comparative Production Costs This has been one of the classical justifications for foreign direct investment. In the past, this rationale was used to justify the relocation of labor intensive, modest technology production to low wage areas. More recently, the rapid appreciation of the yen brought dollar-terms wages in Japan to the North American levels. This, together with the high cost of other factor inputs, such as energy and land, was thought to make a number of industries (even in more advanced technology sectors) non-competitive and force the relocation of much of Japan's manufacturing capacity. This so-called "hollowing-out" process impacted a wide range of production, from chop-sticks to automobiles. In reality, actual experience has been somewhat different than the first gloomy predictions. Many industries have indeed become non-competitive and have re-located. On the other hand, other industries have made very substantial gains in productivity which have more than off-set the currency appreciation. Perhaps the best example of this latter situation is the automobile industry. It has been reported that the Japanese automobile manufacturers could now remain competitive even if the yen appreciated to a level of 95 yen to the U.S. dollar. Profits in the industry have been at record levels in the past couple of years and the companies have enormous reserves of liquid assets. Toyota is reputed to be holding U.S.$12 billion in such assets. - Page 165 -Overall, Japan has already shed much of its labor and energy intensive industry. Heavy industries such as ship-building and steel-making have reduced capacity and sought diversification both in terms of product and overseas investment. It is reasonable to believe that, in the absence of any further major changes in input costs, the surge of comparative cost driven relocation spawned by the yen appreciation is over. Comparative cost considerations will continue to motivate some relocation, but it is likely to be on a gradual basis. Further, the planned concentration on advanced technology and knowledge intensive industries will tend to reduce the problem of comparative input costs. 9.3.1.3. Trade Frictions and Barriers to Market entry Again, this has been one of the classical reasons for foreign direct investment. As barriers to trade are raised, or threatened, Japanese companies have relocated plants in their major market countries. The production of television sets and automobiles in North America and Europe are principal examples of FDI motivated by this consideration. The recent spate of plant development in North America by the Japanese automobile manufacturers was widely perceived as a means of avoiding legislated barriers against Japanese imports. At the time of the investment decisions, the impact of yen appreciation was also a significant consideration but, as discussed above, this - Page 166 -factor is probably of less importance today. As will be discussed below, it now appears that strategic considerations were also instrumental in the car makers' decisions. The influence of trade frictions on the Japanese FDI decision making process is likely to strengthen in the future. Due to the success of the efficiency enhancement programs, stabilization of the yen and strong non-price competitiveness, Japanese exports are projected to increase by 11% in dollar terms, during fiscal 1989 (157) . Although imports have substantially increased, the trade surplus for fiscal 1989 is expected to amount to about U.S.$97.3 billion, and the current account surplus to about U.S.$77.5 billion. These continuing surpluses will lend further support to the protectionist groups in the United States and Europe. 9.3.1.4. Procurement of Resources This consideration was the motivation for investments in iron, coal and non-ferrous metals before 1973; in oil and liquefied natural gas after 1973; and, currently, in pulp production in Canada. In recent years, this motivation has also been expressed by FDI aimed at the acquisition of technology and knowledge related resources. Japanese investment in American software companies illustrate this approach. Even more recently, Japanese companies are funding research centres which are located on, or close to, 'The Japanese Economy"; NRI Quarterly Review, Vol.19  No.2; Nomura Research Institute; Tokyo; May 1989; p.19 - Page 167 -American university campuses such as the University of California at Irvine and Princeton (158) . This use of financial power to acquire, control, or monitor technology research and development is likely to play an increasing role in future Japanese FDI planning. 9.3.1.5. Longterm strategic Plans This factor is becoming increasingly important in the FDI decision making of Japanese companies; especially the large companies which intend to expand their trans-national influence and operations. There are strong indications that large Japanese companies are moving away from the defensive, reactive foreign direct investment decisions of the past towards offensive, overseas development strategies. Banking and financial services, automobiles, steel-making and advanced electronics appear to be target sectors for the Japanese companies. All five major steel producers in Japan, in concert with diversification and rationalization of domestic production capacity, have established joint venture operations with steel companies within the U.S.A. These joint ventures, which involve both financial and technical inputs from Japan, will mitigate complaints against Japanese imports and establish a Japanese presence in the internal steel industry and market of America. "Advanced Bio Class? That's Over in Hitachi Hall"; Business Week; New York; August 7, 1989; p.73. - Page 168 -The major Japanese banks, which are the 10 largest banks in the world, have embarked on a program of FDI which is intended to establish their pre-eminence as global financial institutions. It has been reported that Japanese interests now control more than 25% of California's banking assets (159) . In another development, the Victor Company of Japan (JVC) has invested U.S.$100 million in a Hollywood based film making joint venture. This investment could satisfy a number of requirements for JVC. It should be profitable; it will secure a source of films for its domestic (Japan) videocassette distribution system; and, above all, it provides JVC with a "window" on the film making centre of the world and a base from which it could use financial power to establish a major position in a global, lucrative service industry. Sony, which is JVC's biggest competitor in the videocassette market, bought CBS's record division in 1987 and is reported to be shopping for a film studio. It is interesting to speculate that the two Japanese entertainment giants may fight their future strategic battles through American subsidiary proxies. Finally, the Japanese automakers have established, or are building, 10 plants in North America which will have a combined 1994 capacity of over 2.3 million units per year. In 1989, Japanese plants will produce about 14.7% of the North American output. It is difficult to judge whether this Japanese presence in the North American automobile market is the result of a series Japan Times; Tokyo; August 16, 1989; p.7. - Page 169 -of ad hoc responses to American government and union pressures, or if it reflects a set of carefully orchestrated, strategic plans to compete with the American automobile companies on their own ground. The latter seems to be the more logical conclusion. Toyota et al. have firmly established themselves, at the invitation of the host countries, in the largest automobile market in the world; and it has been suggested that every three cars produced in their plants will displace one import and two Detroit cars (16°). 9.3.2. Allocation of Funds to Foreign Portfolio Investment As previously stated, FPI decisions are based entirely on assessments of the various returns options available to the investors. These assessment take into consideration the security of the investment, the currency exchange risks involved, and the anticipated, or defined, net return. Analysis of Japanese investments in the recent past clearly indicates that the preference is for low risk instruments (bonds), in a secure host country (e.g., the U.S.A.) where the interest rates are high and the yen/dollar relationship is deemed to now be relatively stable. This trend can be expected to continue. "Shaking Up Detroit"; Business Week; N.York; August 14, 1989; p.74. - Page 170 9.3.3. Allocations in a Competing Environment Given the current projections as to future economic growth in Japan and the country's trade balance, it appears unlikely, from a macroeconomic perspective, that there will be restrictions of FDI due to shortages of investment funds. Should the availability of investment funds decrease, however, it is probable that the reduction would impact FPI, rather than FDI. The rationale for this judgement is based on four points, namely: 1. The volume of FDI is only a fraction of the total net foreign investment. 2. The aggregate of all return considerations (i.e., financial, strategic and political) relating to a foreign direct investment should be competitive with bond interest rates. 3. Pre-1984 experience indicated that FDI was the dominant off-shore investment process; NFPI grew only because of the large surpluses in the Japanese economy; and, 4. Much of the Japanese FPI is controlled by the country's large life and general insurance companies which - Page 171 -usually have close keiretsu affiliations with the large companies which will make most of the FDI investments. To summarize, it is unlikely that there will be any significant reduction in the funds available for Japanese foreign investment in general. In the unlikely event that such a reduction does occur, it would not be reflected in an equivalent reduction in the funds available for Japanese FDI. 9.4. ACCEPTABILITY The preceding discussion in this section has determined that Japanese foreign direct investment is unlikely to be impeded by either macroeconomic considerations, or by preferential allocation of foreign investment resources to other types of foreign investment. It has also been concluded that FDI will continue to be an attractive option for Japanese companies; indeed, that it has become, and will continue to be, a central instrument in the strategic approach of many large companies as they move towards full trans-national status. The remaining question is: will the host countries and communities continue to welcome Japanese FDI? The answer to this question is both complex and uncertain. It may be re-phrased as two questions: will there be legislated opposition to Japanese FDI? will there be un-official resistance to Japanese FDI? - Page 172 -Legislated opposition to Japanese investment is, within the North American and European context, likely to be restricted to specific sectors. Examples of such intervention include defense- sensitive industries (e.g., the review of Fujitsu's acquisition of Fairchild), transportation (usually subject to limitations for any foreign buyer) and, possibly, restrictions on real estate holdings by non-residents (e.g. Nebraska's law against foreign ownership of farmland). However, in most cases, the host country's espousal of market-place economics, regional competition for job-creating investments and the exposure of its own investors to similar restriction in other countries, will tend to act as a brake on such direct legislation. Non-legislated opposition to Japanese FDI is a broader topic. Clearly, as the increasing presence of Japanese companies could result in an escalation of anti-Japanese sentiment. A great deal will depend on the corporate and community practices of the Japanese operated companies. To a large extent, this problem is an integral part of the larger issue of how Japanese companies will accommodate to the new roles of true multi-national corporations. This accommodation will have to address the current centralization of authority in the Japan head offices, or in the hands of officials despatched from such head offices. The problems associated with this adjustment are very complex and will be difficult of solution. Nonetheless, solutions will have to be found if Japanese companies are to adopt multi-nationalism and be - Page 173 -accepted in the host countries. To-date the score card is mixed. In terms of employee relations, most Japanese firms have done an excellent job. They have, as in Japan, made great efforts to persuade employees that they have common interest with the company. Most Japanese companies in Europe and North America appear to have established very good employee relations; the recent defeat of the UAW's attempt to organize the Nissan plant in Smyrna, Tennessee attest to this. On the other hand, the tax exemptions and outright grants extracted by the Japanese automakers have generated significant adverse reaction at both the community and state levels in the United States. Unions generally resent the non-existent, or reduced influence, roles they play in Japanese-run auto plants. Similarly, local parts manufacturers clearly dislike the tendency of the Japanese companies to invite Japanese parts suppliers to establish plants in close proximity to the automobile plant and then to favour such companies in their procurement policies. The principal battleground on which this issue of Japanese FDI acceptability will be fought is, of course, the United States. Xenophobic forces are at work in the U.S.A., and the country's long-held "open-market" principles are under attack. The facts are perhaps not as bad as some of the "revisionists" have led the American public to believe. In 1988, accumulated foreign direct investment in the U.S.A. ($329 billion) did surpass accumulated American direct investment in other countries ($327 billion). Page 174 However, since these investments were stated at cost, the evaluation was weighted towards recent purchases in the U.S.A. by foreigners, as compared to past investments by American buyers of foreign assets. Further, it should be noted that, despite the recent buying spree, foreigners only own about 3% of total American assets (which is much less than in many European countries) and less than 1% of American real estate. Among the foreign direct investors in the U.S.A., Japan is still a somewhat distant second to Britain (161) . Nevertheless, it is Japanese FDI which is perceived as the principal threat. Compared to other sources of FDI, Japan has more money to invest; Japanese economic power is greater and, currently and prospectively, more competitive with American industry and commerce; and Japanese investments in steel, car-making, electronics and banking are more visible and are perceived as having broader potential impact on the strategic options of the American economy. The situation is rather ironic. To rectify its huge merchandise trade deficit, the United States must increase its exports of manufactured products. But, this increase cannot be achieved without significant improvements in production efficiency; and, since the capital required to finance such improvements is not available in the U.S.A., this mandates increased FDI. The Japanese 161 "Xenophobia rules", The Economist. London, August 26 -September 1, 1989, p. 68-70. - Page 175 -companies have both the financial and technical strengths needed to meet the American goals. However, the quid pro quo for such assistance will be to allow Japanese-owned, American-based operations to compete for the full American market (not a restricted import quota) on an equal basis with American companies. The acceptability of the remedy will largely depend on the sensitivity with which it is administered. - Page 176 -10.0. CONCLUSIONS This study has attempted to explain the factors behind the rapid acceleration of Japanese foreign direct investment in the 1980s. The approach has been to combine macro-theory with the actual developments in the Japanese economy since 1973 to explain the current capital surplus position of Japan. In particular, on-going savings surpluses in the private sector, when coupled with the LDP's committment to balanced budgets after 1978, have been suggested as a primary determinant behind the growing capital outflow from Japan after 1983. It has been contended that these surpluses have been generated by several factors including: 1) the transition of Japan from a high-growth to mature economy and the resultant decline in corporate capital formation requirements; 2) rising relative real wages as a result of the 80% appreciation of the yen since 1984; and 3) the dramatic escalation of Japanese land prices. The latter, while deterring domestic investment in land-intensive industry at home, has also created a huge collateral base from which potential Japanese investors can finance offshore investments. The study has also attempted to support its macro-economic rationale by developing a regression model designed to test the relationship of Japanese net FDI with real wages, real land prices, the Japanese Yen/U.S. dollar spot exchange rate, and the Bank of Japan and U.S. Federal Reserve discount rates differential. - Page 177 -Although supporting the macro-economic contentions to some degree, the results are limited by problems with both the methodology and data used to develop the model. Nevertheless, as an exploratory examination of the determinants of Japanese foreign direct investment, the model retains some merit. Finally, the study has examined the future sustainability of Japanese foreign direct investment. This analysis, again, has relied on macro-economic theory and our analysis of current trends in the Japanese economy to conclude that the present level of Japanese foreign direct investment is sustainable. This conclusion is based on the premise that the factors which have created Japan's current excess capital position will prevail in the medium term. In addition, in the unlikely event that future surpluses are reduced, the study concludes that Japanese foreign portfolio, rather than foreign direct, investment will be impacted. Finally, it is suggested that the sustainability of Japanese FDI may be determined more by its future acceptability with recipient countries. The latter may represent the most serious impediment to the continuing drive of Japanese companies offshore, although this will be tempered in most cases by the host countries' committments to unrestricted capital flows and domestic employment considerations. Page 178 -BIBLIOGRAPHY BOOKS Abegglen, James C. & Salk, George Jr. Kaisha: The Japanese  Corporation. Tokyo: Charles E. Tuttle Company, 1985 Berenson, Mark L., & Levine, David M. Basic Business Statistics  Concepts & Applications, Englewood, N.J.: Prentice-Hall Inc., 1979 Bronte, Stephen, edit. Japanese finance Markets & Institutions. London: Euromoney Publications, 1982 Dore, R.P. Structural Adjustment in Japan. 1970-1982,. Geneva: International Labour Organization, 1986 El-Agraa, Ali M. Japan's Trade Frictions. Realities or  Misconceptions. London: MacMillan Press, 1988 Frenkel, Jacob A. " Flexible Exchange Rates, Prices and the Role of 'News'," International Financial Management. Donald R. Lessard, edit., New York: John Wiley & Sons, 1985 Hamada, Koichi," Japanese Investment Abroad," in P. Drysdale, ed, Direct Foreign Investment in Asia and the Pacific. Toronto: University of Toronto Press, 1972 Hayashi, Fumio. " Why is Japan's Savings Rate So Apparently High?", NBER Macroeconomics Annual. 1986. Cambridge, Ma: National Bureau of Economic Research, 1986 Higashi, Chikara and Lauter, G.Peter. The Internationalization of  the Japanese Economy. Boston: Kluwer Academic Publishers, 1987 Hollerman, Leon. Japan's Economic Strategy in Brazil: Challenge  for the United Statesf Lexington, Ma: Lexington Books, 1988 Inoue. T., "A Comparison of the Emergence of MNC Manufacturing,11 in Overseas Business Activities. Tokyo: University of Tokyo, Katano, Hikoji; Murakami,Atsushi; Ikemoto, Kiyoshi; edit., Japan's Direct Investment to Asian Countries. Kobe: Research Institute for Economics and Business Administration, Kobe University, 1978 Katano, Hikoji. Japanese Enterprises in Asian Countries.  Statistical Outlook. Kobe, Japan: Research Institute for Economics and Business Development, Kobe University, 1981 - Page 179 -Levich, Richard M. " Evaluating the Performance of the Forecasters," International Financial Management, Donald R. Lessard, edit., New York: John Wiley and Sons, 1985 Lincoln, Edward R. Japan: Facing Economic Maturity. Washington: The Brockings Institute, 1988 Literman, David K. & Stonehill, Arthur I. " The Foreign Investment Decision," Multinational Business Finance. Reading Ma.: Addison-Wesley Publishing Company, 1986 Madura, Jeff. "Direct and Indirect Foreign Investment," International Financial Managementf St.Paul: West Publishing Co., 1986 Noguchi, Yukio. " Public Finance," The Political Economy of  Japan. Volume 1 edit, by Kozo Yamamura & Yasukichi Yasuba, Stanford: Stanford University Press,1987 Okochi, Akio, & Inoue, Tadakatsu edit. Overseas Business  Activities, Tokyo: University of Tokyo Press,1984 Ozawa, Terutomo. Multinationalism. Japanese Style. Princeton: Princeton University Press, 1979 Robinson, Richard D. Direct Foreign Investment. Costs and  Benefits New York: Rogers Publishers, 1982 Root, Franklin K. "The Nature and Scope of Multinational Enterprise," International Trade & Investment. 5th Edition. Cincinnati: The Warton School of Pennsylvania; South west Publishing Co., 1984 Sakiguchi, Sueo. Asian - Japan Relations. Investment. Proceedings of a Workshop and a Conference Organized by the Japan Center for International Exchange, Singapore: Dec. 5-6, 1981. Oise, Japan: May 20-23, 1982. Singapore: Institute of S.E. Asian Studies, 1983 Sato, Kazuo. "Savings and Investment," The Political Economy of  Japan: Volume 1 - The Domestic Transformation. Edited by Kozo Yamamura and Yasukichi Yasuba, Stanford: Stanford University Press, 1987 Schmiegelow, Michele. "The Reform of Japan's Foreign Exchange and Foreign Trade Control Law: A case of Qualitative Economic Policy," Japan's Response to  Crisis and Change in the World Economy, edit, by Michele Shchmiegelow, New York: M. E. Sharpe, Inc. 1986 - page 180 -Schofield, Norman et al. Advanced Statistical Methods in the  Social Sciences. New York: Praegar Publishers, 1986 Tsumumi, Yoshi. Soaoshosha. Montreal: Institute for Research on Public Policy, 1984 Uno, Kimio, Japanese Industrial Performancef Amsterdam: Elsevier Science Publishers B.V., 1987 Viner, Aron. The Financial Samurai. The Emerging Power of  Japanese Money. London: Kogan Page Limited, 1988 Yoshihara, Kunio, Japanese Investment in South East Asia. Honolulu: University Press of Hawaii, 1978 Yoshihara, Kunio, The Sogo Shosha. Oxford: Oxford University Press, 1983 Yasumuro, K., " The Contribution of Sogo Shosha to Multinationalization," in A. Okochi and T. Inoue, edit., Overseas Business Activities., Tokyo: University of Tokyo Press, 1984 MAGAZINES Abegglen, James C. "Going Abroad," TOKYO Business Today. January 1988 Abegglen, James C. "Economics: Stone Bridges," TOKYO Business  Today. April 1988 "Average Savings Now at 8.19 Million Yen," TOKYO Business Today, May 1988 Cubitt, Guy. "Tokyo's Real Estate Turmoil," TOKYO Business Today. May 1987 Caplen, Brian, edit.," A New Sun Rising in the East," Asian  Business. September 1988 "Challenge and Dilemma for Corporate Japan: The EC Gears for 1992," TOKYO Business Today. Tokyo: April 1989 "Corporate Profit Performance," NRI Quarterly Review. Nomura Research Institute, Tokyo: August 1989 Hijikata, Takeshi. "Need for Tax Reform and Its Direction," Keidanren Review No. 110. April 1988 - Page 181 -Hines, Mary Alice. "Investing in Japanese Real Estate," Real Estate  Review Vol. 18 No. 4. Winter 1989 Hirata, Ikuo."Overseas Investment," Japan Economic Almanac 1987. Tokyo, 1987 Itoh, H. "Japan's Billionaires," TOKYO Business Today. October 1987 "Japanese Direct Investment in U.S. Gains Momentum," Keidanren  Review No. 113. October 1988 "Japanese Find it Tough Abroad," Asian Business. October 1988 "Japanese Investors Look to Italy," World Property. March 1989 "Japan Investors Turn to Foreign Stocks," The Japan Economic Journal, June 17, 1989 "Japanese Likely to Diversify US Investment Portfolio," World Property. London: RICS Journals Ltd., April 1989 "Japanese Manufacturers Go Local," TOKYO Business Today. July 1988 Kobayashi, Shigeru. " The Shuwa Shogun," TOKYO Business  Today. March 1987 Kojima, Kiyoshi, " Japanese - Style Direct Foreign Investment," Japanese Economic Studies. New York: ME Sharpe Inc., Spring 1986 Kosai, Yutaka. "The Japanese Economy In 1989: Constraints and Challenges," Journal of Japanese Trade & Industry. No.l 1989, Tokyo, 1989 Krueger, William. "Japanese Real Estate: At Home and Abroad," Tokyo Business Today. December 1988 Krueger, William. "Tokyo," World Property. January 1989 Levich, Richard M. "Are Forward Exchange Rates Unbiased Predictors of Future Spot Rates," Columbia Journal of  World Business 14. No.4. Winter 1979 Maki, Atsushi. "Why is the Japanese Household Savings Rate So High," Keio Business Review. No. 24 1987. Tokyo: Keio University Press, 1987 Page 182 -Murphy, Terence. "The War Between the States," TOKYO Business  Today. April 1987 Namai, Toshishige."Japan's Drug Companies Venture Abroad," TOKYO  Business Today. June 1989 Nomura Quarterly Economic Review Vol. 19. No.l, Tokyo: NRI & NCC Co. Ltd., February 1989 Nomura Quarterly Economic Review Vol 19 No. 3. Nomura Research Institute, Tokyo: August 1989 Oda, Takeshi. "Overseas Investment," Japan Economic Almanac. 1986 Ogura, Masao. "A New Stage of Inventory," TOKYO Business Today, September 1988 "On the Move Again," The Economist. November 5, 1988 Ota, Toshiki. "The Costly Takeover of Firestone," TOKYO Business  Today. June 1988 "Rapid Increase in Japanese Overseas M&A," TOKYO Business Today. Tokyo: January, 1989 Robins, Brian. "All Eyes on the New WorLd Leader," Triple A  Banking Journal. December 1987/January 1988 Rudnitsky, Howard et al. "Land of the Rising Stocks," Forbes, May 18, 1987 Sharp Jr., William E. "EC-Japanese Economic Relations: Trade Friction From Europe," TOKYO Business Today, December 1986 Sneider, Daniel. " The Maekawa Report: Promises Japan Can't Keep?" TOKYO Business Today. June 1986 Statistics. BOJ Economic Statistics Monthly. Japan: Research St Statistics Dept., December 1988 Takahashi, Masami. " Overseas Investment," Japan Economic  Almanac 1988 Takahashi, Yuichi. "Japanese Companies Overseas Ventures: Automobile Industry Leads Investment in Developed Nations," Japan Economic Almanac. 1986 "The Consumption Boom in Japan," Tokai Monthly Economic Letter. June 1989 Page 183 "The Overwhelming Japanese," World Property. London: RICS Journals Ltd., May 1989 "The Shuwa Shogun," Tokyo Business Today. Tokyo: March 1987 "Tokyo in the Making," TOKYO Business Today. May 1987 Ungphakorn, Peter. "Japanese Eye Thailand Investments," TOKYO  Business Today. April 1988 Williams, David. "Nissan's Great British Experiment," TOKYO  Business Today. June 1988 Woronoff, Jon. "End of buy America Era," Asian Business. December 1988 PAPERS Helliwell, John F., " Some Comparative Macroeconomics of the United States, Japan and Canada." Department of Economics Discussion Paper No. 87-04 . Vancouver, University of B.C. March 1987. Japan 1988, An International Comparison, Tokyo: Keizai Koho Centre, 1988 Jetro, White Paper on World and Japanese Direct Foreign Investment, Tokyo 1986 Sazanami, Yoko. " Recent Trends in Japanese Trade and Capital Flow - Japan vs Western Europe and Asia," Keio Economic Studies  Vol. XXV No. 1 1988 NEWSPAPER ARTICLES "...As big Japanese firms attempt to really go global," Globe & Mail from The Economist. July 1989 "Auto Parts Maker Has U.S. Plans," The Japan Times. May 20, 1989 Bianchi, Angela. " Home Ownership and Elusive Dream for Most," The Financial Post. July 24, 1989 "Brother Sets Up Malaysian Plant," Japan Times. May 20, 1989 "Canadian Auto Parts Venture Set Up," The Japan Times. Osaka Jiji Press May 17 1989 - Page 184 -"Deadlocked Talks Leave U.S., Japan at Odds Over Trade," The  Financial Post. Reuters, June 16, 1989 Done, Kevin. " Car Wars After the Yen Shock," The Financial  Times. London: May 12, 1989 Done, Kevin. " GM, Isuzu to Invest 70 million pounds in Four-Wheel Drive Project," The Financial Times. May 12, 1989 "Government Says 2.2% Growth Shows Japanese Economy in on' Ideal Path1," The Globe & Mail. Reuter, June 20, 1989 "Growth Will Hit 4.75% in'89 Then Diminish, OECD Says," Japan Times, May 16,1989 Hudson, Richard L."Competition from Japan Begins in the Back Yard," Asian Wall Street Journal, February 26, 1989 "Japan Copiers Made in U.S. May Be Hit With EC Duties," The  Asian Wall Street Journal. February 8, 1989 "Japanese are Biggest Foreign U.S. Investors," The Financial  Post, June 28, 1989 "Japanese Internationalization: Banking on the Golden State," The Japan Times May 22, 1989 "Japanese Investments in Foreign Stocks," The Japan Times, courtesy of The Nomura Securities Co. Ltd., May 20, 1989 "Japan's Global Trade Surplus narrows as U.S. Gap Shrinks," The Financial Post. United Press International, June 1 5 , 1989 "Japan World's Leading Creditor," The Province. May 28, 1989 Kingston, Jeff." The Brazil Problem and Japan's Effective Solution to It," Japan Times. May 20, 1989 Lewington, Jennifer. " An Uncertain Welcome for Investors in U.S.," The Globe & Mail. ????? Lewington, Jennifer. "Foreigners Invest $65 Billion in U.S.," Globe & Mail May 31, 1989 Lewington, Jennifer. " National Security1 Can Block Foreign Takeover," The Globe & Mail. July 3, 1989 Lubin, Joann S. et al. "Toyota's Plan Would Pressure European Firms, The Asian Wall Street Journal, February 27, 1989 - Page 185 -McCooey, Christopher. 11 Japanese Direct Investment in Europe: Why is the U.K. so Attractive?" The Japan Times. May 15, 1989 Montagnon, Peter. " Buoyant Economy Brings Inflation Concern," The Financial Post, July 24, 1989 Rodger, Ian. " Many Pressures Strain New Era," The Financial  Post. July 24, 1989 Rodger, Ian. "Shakeout Seen as Foreign Investors Forsake Japanese Stock," The Financial Times. December 9, 1987 Shirouzu, Norihiko. "Toyota Weighing Options on Making Cars ni Europe."The Asian Wall Street Journal. January 27-28 1989 Terry, Edith. "High-Flying U.S. Dollar Leaves Japan's Buoyancy Up in the Air," The Globe & Mail. June 20, 1989 Terry, Edith. " Why the Japanese Can't Relax," The Globe &  Mail Report on Business. August 1989 "Thermos Sold to Firm in Japan," The Globe & Mail, June 19, 1989 "Though Rich, Japan is Poor in Many Elements," The Asian Wall  Street Journal. March 3, 1989 "TSE's Big Four Securities Companies Post Record High Operating Revenues," The Japan Times. May 20, 1989 "U.S.- Japan Talks on Trade Friction End with Little Progress Made," The Financial Post. June 15, 1989 Woodruff, John. " Banking on the Golden State," The Japan  Times. May 22, 1989 "World's 'Top 50" List dominated by Japanese Financial Houses," Mainichi Daily News. June 23, 1987 REPORTS & YEARBOOKS Business Statistics. 1986. Washington: U.S. Department of Commerce, Bureau of Economic Analysis, 1986 IMF. Foreign Exchange Rate Supplement. 1985. IMF, 1985 International Financial Stastics. 1988. IMF, 1988 Page 186 -Japan Economic Almanac. 1985 Ministry of Finance, Japan Statistical Yearbook. Tokyo: 1965 through 1988 Miti, Japan's Overseas Investments. 1977 edition NECD Economic Surveys. Japan 1987/1988. Paris, 1988 Nippon.Business Facts and Figures.1989 Jetro, Tokyo: 1989 OECD Economic Survey of Japan. 1963 through 1973, 1987 Report prepared by the Gleneagles Group, Japanese Direct  Foreign Investment in North America - A Canadian  Perspective. Vancouver: B.C. Canada, January 1988 The Comptroller General of the United States. Industrial Policy:  Japan's Flexible Approach, June 23, 1982 The Japan Economic Journal. Tokyo: 1985 - Page 187 -APPENDIX 1: DATA USED IN REGRESSION MODELS YEAR D1R POR PRIV GOV UAGE LAND EXCH DISC SPR 1960 -79.37 -22.20 -1492 -1569 38.4 22.9 360 0.219 0.115802 1961 -32.50 11.76 -4321 -144 40.6 45.0 360 0.215 0.124615 1962 -11.36 71.02 -306 1196 41.9 62.2 360 0.188 0.099004 1963 -19.02 146.74 -2226 1389 42.6 70.7 360 0.158 0.063131 1964 -46.88 140.63 -1545 1966 45.1 79.7 360 0.172 0.064732 1965 -30.39 80.11 -91 3762 46.3 72.9 360 0.152 0.032551 1966 -76.32 -18.42 708 4505 48.9 89.7 360 0.145 0.022121 1967 -71.78 -9.90 -1458 4448 52.6 88.1 360 0.144 0.030532 1968 -117.65 296.47 -125 3085 56.8 91.1 360 0.146 0.044101 1969 -112.11 751.12 -2652 5554 62.4 99.8 360 0.141 0.001011 1970 -187.50 187.50 -1065 875 67.8 86.0 360 0.131 -0.002780 1971 -96.65 550.30 1610 3570 73.2 120.1 349.33 0.108 -0.000090 1972 -305.97 44.78 4948 5261 80.8 128.0 303.37 0.093 -0.00307 1973 -935.54 -829.75 1147 892 88.5 149.3 291.7 0.121 -0.000300 1974 -675.79 -347.47 -2613 6156 90.5 146.1 292.08 0.127 -0.014200 1975 -771.28 1306.23 4183 14663 92.9 99.2 396.79 0.083 -0.033530 1976 -667.85 973.90 8888 13713 95.6 117.3 296.55 0.077 -0.012750 1977 -491.03 192.83 10371 8666 96.1 113.7 268.51 0.048 -0.036120 1978 -531.55 -632.09 15038 11985 98.4 111.7 210.44 0.037 -0.070800 1979 -605.40 -280.37 8761 10913 101.7 116.3 219.14 0.065 -0.072040 1980 -478.00 2138.00 8090 10666 100.0 100.0 226.74 0.073 -0.061180 1981 -1006.78 1639.53 9311 9538 101.1 105.1 220.54 0.053 -0.085000 1982 -971.46 198.86 10108 9260 101.8 110.1 249.08 0.048 -0.054920 1983 -717.66 -650.61 13416 9685 104.0 114.5 237.51 0.047 -0.034590 1984 -1322.76 -5308.77 12444 5819 104.8 119.0 237.52 0.047 -0.035060 1985 -1272.73 -9145.09 11782 2391 104.9 125.9 238.54 0.046 -0.023420 1986 -2165.01 -15505.86 19477 3305 107.4 141.2 168.52 0.027 -0.028430 1987 -2434.00 -11874.32 16004 7345 109.7 178.3 144.64 0.023 -0.025480 - Page 188 -READU) DIR POR PRIV GOV WAGE LAND EXCH DISC SPR STAT/ all PCOR Ols PRIV WAGE LAND DISC/ ANOVA LIST auto PRIV WAGE land disc/ ANOVA LIST OLS PRIV WAGE/ ANOVA LIST OLS PRIV LAND/ ANOVA LIST OLS PRIV DISC/ ANOVA LIST auto PRIV WAGE/ ANOVA LIST auto PRIV land/ ANOVA LIST auto PRIV disc/ ANOVA LIST OLS POR EXCH SPR/ ANOVA LIST auto POR EXCH SPR/ ANOVA LIST OLS DIR WAGE LAND EXCH SPR gov/ ANOVA LIST OLS DIR WAGE LAND EXCH SPR/ ANOVA LIST plot d1r wage plot dlr land plot dir exch plot dir spr sample 1 12 OLS DIR WAGE LAND EXCH SPR/ ANOVA LIST plot dir wage 1 plot dir land ^ plot dir exch a, plot dir spr CTQ sample 13 28 fi> plot dir wage plot dir land 00 plot dir exch plot dir spr OLS DIR WAGE LAND EXCH SPR/ ANOVA LIST 1 END > T) T3 PI Z o t—I •x to |_OLS PRIV WAGE LAND DISC/ ANOVA LIST REQUIRED MEMORY IS PAR= 4 CURRENT PAR= 40 OLS ESTIMATION 28 OBSERVATIONS DEPENDENT VARIABLE = PRIV .. .NOTE . .SAMPLE RANGE SET TO: 1, 28 R-SQUARE = 0.8654 R-SOUARE ADJUSTED = VARIANCE OF THE ESTIMATE = 0.70725E+07 STANDARD ERROR OF THE ESTIMATE = 2659.4 MEAN OF DEPENDENT VARIABLE = 4942.6 LOG OF THE LIKELIHOOD FUNCTION = -258.376 0.8485 MODEL SELECTION TESTS - SEE JUDGE ET.AL.M985, P.242) AKAIKE (1969) FINAL PREDICTION ERROR- FPE = 0.80828E+07 (FPE ALSO KNOWN AS AMEMIYA PREDICTION CRITERION -PC) AKAIKE (1973) INFORMATION CRITERION- AIC = 15.903 SCHWARZf1978) CRITERION-SC = 16.094 REGRESSION ERROR TOTAL ANALYSIS OF VARIANCE SS 10910E+10 16974E+09 12608E+ 10 DF 3 . 24 . 27. - FROM MEAN MS 0. 36367E + 09 0. 70725E + 07 0. 46695E + 08 F 51.421 REGRESSION ERROR TOTAL ANALYSIS OF VARIANCE SS 17750E+10 16974E+09 19448E+10 DF 4 . 24 . 28 . - FROM ZERO MS 0.44376E+09 0. 70725E*07 0.69456E+08 F 62.744 VARIABLE NAME ESTIMATED COEFFICIENT STANDARD T-RATIO ERROR 24 DF WAGE -7.0493 65.703 -0.10729 LAND -49.478 26.176 -1.8902 DISC -0.13300E+06 27517. -4.8333 CONSTANT 24561. 7751.5 3.1686 PARTIAL STANDARDIZED ELASTICITY CORR. COEFFICIENT AT MEANS -0.0219 -0.26339E-01 -0.11180 -0.3600 -0.23476 - 1.0379 -0.7023 -1.1230 -2.8197 0.5431 O.OOOOOE+OO 4.9693 Hello/Bonjour -PNPN Welcome to i SHAZAM - Version 6.1 - OCT 1988 SYSTEM=G PAR= 40 |_READ(4) DIR POR PRIV GOV WAGE LAND EXCH DISC SPR ...SAMPLE RANGE IS NOW SET TO: 1 28 I STAT/ ALL PCOR NAME N MEAN ST. DEV VARIANCE MINIMUM MAXIMUM DIR 28 -579.79 630.08 0. 39700E + 06 -2434.0 - 11.360 POR 28 - 1282.0 4 127.5 0.17036E»08 - 15506. 2138.0 PRIV 28 4942.6 6833.4 0.46695E+08 - 4321 .0 19477 . GOV 28 5674.8 4355.1 0.18967E+08 - 1569.0 14663. WAGE 28 78.386 25 532 651 .89 38.400 109.70 LAND 28 103.67 32.423 1051 . 2 22.900 178.30 EXCH 28 296 .82 70.477 4967 .0 144.64 396 . 79 DISC 28 0.10479 0.57701E-01 0.33294E-02 0.23000E-01 0.21900 SPR 28 0.13379E-03 0.55371E-01 0.30659E-02 -0.85003E-01 0.12462 CORRELATION MATRIX OF VARIABLES - 28 OBSERVATIONS DIR 1.0000 POR 0.81828 1.0000 PRIV -0.79668 -0.60430 1.0000 GOV -0 '.28174 0.15525 0.54537 1 .0000 WAGE -0.77732 -0.40860 0.84943 0.66819 1.0000 LAND -0.74031 -0.50618 0.60793 0.40339 0.79624 1 .0000 EXCH 0.81846 0.60326 -0.88974 -0.41175 -0.83544 -0.68226 1.0000 . DISC 0.75416 0.45591 -0.91757 -0.67748 -0.94627 -0.76905 0. •83314 1 .0000 SPR 0.54244 0.15175 -0 . 71657 -0.76141 -0.89098 -0.69886 0 . 69691 0 . 904 19 1.0000 DIR POR PRIV GOV WAGE LAND EXCH DISC SPR |_AUTO PRIV WAGE LAND DISC/ ANOVA LIST REQUIRED MEMORY IS PAR= 5 CURRENT PAR= 40 DEPENDENT VARIABLE = PRIV . .NOTE. .R-SQUARE,ANOVA,RESIDUALS DONE ON ORIGINAL VARS LEAST SQUARES ESTIMATION 28 OBSERVATIONS BY COCHRANE-ORCUTT TYPE PROCEDURE WITH CONVERGENCE = 0.00100 ITERATION RHO LOG L.F. SSE 1 0.00000 -258.376 0.16974E+09 2 0.39454 -255.829 0.14065E+03 0.43018 -255.768 0.13986E+09 4 0.43428 -255.763 0.13979E+05 0.43477 -255.763 0. 13978E + 09 LOG L.F. •255.763 AT RHO 0.43477 RHO ESTIMATE 0.43477 ASYMPTOTIC VARIANCE 0.02896 ASYMPTOTIC ST.ERROR 0.17019 ASYMPTOTIC T-RATIO 2.55466 R - SQUARE = 0.8891 R-SQUARE ADJUSTED = 0.8753 VARIANCE OF THE ESTIMATE = 0.58243E+07 STANDARD ERROR OF THE ESTIMATE = 2413.4 MEAN OF DEPENDENT VARIABLE = 4942.6 LOG OF THE LIKELIHOOD FUNCTION = -255.763 MODEL SELECTION TESTS - SEE JUDGE ET.AL.( 1985. P.242) AKA'IKE ( 1969) FINAL PREDICTION ERROR - FPE = 0.66564E+07 (FPE ALSO KNOWN AS AMEMIYA PREDICTION CRITERION -PC) AKAIKE (1973) INFORMATION CRITERION- AIC =' 15.709 SCHWARZI 1978) CRITERION-SC = 15.899 REGRESSION ERROR TOTAL ANALYSIS SS 1 1210E+ 10 13978E+09 12608E+10 OF VARIANCE DF 3 . 24 . 27 . • FROM MEAN MS 0.37366E+09 0.58243E+07 0.46695E+08 REGRESSION ERROR TOTAL ANALYSIS OF SS 18050E+10 13978E+09 19448E+10 VARIANCE DF 4 . 24. 28 . • FROM ZERO MS 0.45125E+09 0.58243E+07 0.69456E+08 VARIABLE ESTIMATED STANDARD T-RATIO PARTIAL STANDARDIZED ELASTICITY NAME COEFFICIENT ERROR 24 DF CORR. COEFFICIENT AT MEANS WAGE 5.8604 77.296 0.75818E-01 0.0155 0.21897E-01 0.92942E-LAND -37.283 28.665 -1.3006 -0.2566 -0.17690 -0.78205 DISC -0.12069E+06 28854. -4.1826 -0.6493 -1.0191 -2.5586 CONSTANT 21213. 8342.6 2.5427 0.4607 O.OOOOOE+00 4.2918 RVATION OBSERVED PREDICTED CALCULATED NO . VALUE VALUE RESIDUAL 1 - 1492.0 -5969.3 4477.3 2 -4321 .0 -6546.3 2225.3 3 -306.00 -3815.5 3509.5 4 -2226 .0 -251 .04 - 1975.0 5 - 1545.0 -2575.9 1030.9 6 -91.000 412.02 -503.02 7 708 .00 493.44 214.56 8 - 1458.0 679.52 -2137 . 5 9 - 125.00 235.49 -360.49 10 -2652.0 430.54 -3082.5 1 1 - 1065.0 2405.3 -3470.3 12 1610.0 3739 .0 -2129.0 13 4948.0 5289.5 -34 1.48 14 1147.0 4 57.35 689.65 15 -2613.0 - 196.41 -2416.6 16 4183.0 8008.6 -3825.6 17 8888.0 7842.5 1045.5 18 10371 . 1 1874. - 1503.1 19 15038. 13420. 1618.2 20 876 1 .0 9445.0 -684.00 21 '8090.0 9199.5 -1109.5 22 9311.0 1 1599. -2288.4 23 10108. 12012. - 1904 .0 24 13416. 11912. 1504.2 25 12444 . 1 1684. 760.47 26 11782. 11474. 307.58 27 19477 . 13227 . 6250.2 28 16004. 1 1907 . 4097 . 1 DURBIN-WATSON = 1.0720 VON NEUMAN RATIO = 1.1117 RHO = 0.39454 RESIDUAL SUM = 0.63665E-11 RESIDUAL VARIANCE = 0.70725E+07 SUM OF ABSOLUTE ERRORS= 55461. R-SOUARE BETWEEN OBSERVED AND PREDICTED = 0.8654 RUNS TEST: 13 RUNS, 13 POSITIVE, 15 NEGATIVE, NORMAL STATISTIC = -0.7468 COEFFICIENT OF SKEWNESS = 0.6962 WITH STANDARD DEVIATION OF 0.4405 COEFFICIENT OF EXCESS KURTOSIS = 0.1341 WITH STANDARD DEVIATION OF 0.8583 GOODNESS OF FIT TEST FOR NORMALITY OF RESIDUALS - 10 GROUPS OBSERVED 0.0 0.0 2.0 7.0 6.0 7.0 2.0 3.0 1.0 0.0 EXPECTED 0.2 0.8 2.2 4.5 6.3 6.3 4.5 2.2 0.8 0.2 CHI-SQUARE = 4.4915 WITH 4 DEGREES OF FREEDOM |_OLS PR IV WAGE/ ANOVA LIST REQUIRED MEMORY IS PAR= 4 CURRENT PAR= 40 OLS ESTIMATION 28 OBSERVATIONS DEPENDENT VARIABLE = PRIV .. .NOTE . .SAMPLE RANGE SET TO: 1, 28 R-SQUARE = 0.7215 R-SQUARE ADJUSTED = 0.7108 VARIANCE OF THE ESTIMATE = 0.13503E+08 STANDARD ERROR OF THE ESTIMATE = 3674.6 MEAN OF DEPENDENT VARIABLE = 4942.6 LOG OF THE LIKELIHOOD FUNCTION = -268.551 MODEL SELECTION TESTS - SEE JUDGE ET.AL.U985, P.242) AKAIKE ( 1969) FINAL PREDICTION ERROR - FPE = 0.14468E+0B (FPE ALSO KNOWN AS AMEMIYA PREDICTION CRITERION -PC) AKAIKE (1973) INFORMATION CRITERION- AIC = 16.487 SCHWARZ( 1978) CRITERION-SC = 16.582 4> ANALYSIS OF VARIANCE - FROM MEAN SS OF MS F REGRESSION 0. 90968E+09 1 . 0.90968E+09 67, , 369 ERROR 0. ,35108E+09 26. 0. 13503E + 08 TOTAL 0. 12608E+10 27 . 0.46695E+08 ANALYSIS OF VARIANCE - FROM ZERO SS DF MS F REGRESSION 0. 1 5937E+10 2. 0.79685E+09 59 .013 ERROR 0. 35108E+09 26 . 0.13503E+08 TOTAL 0. 19448E+10 28 . 0.69456E+08 VARIABLE ESTIMATED STANDARD T-RATIO PARTIAL STANDARDIZED ELASTICITY NAME COEFFICIENT ERROR 26 DF CORR. COEFFICIENT AT MEANS WAGE 227.34 27.698 8.2078 0.8494 0.84943 3.6055 CONSTANT -12878. 2279.5 -5.6494 -0.7423 O.OOOOOE+OO -2.6055 RVATION ' OBSERVED NO. VALUE 1 - 1492 .0 2 -4321 .0 3 - 306.00 4 -2226.0 5 - 1545 .0 6 -91.000 7 708.00 8 - 1458.0 9 - 125.00 10 -2652.0 1 1 - 1065.0 12 1610.0 13 4948 .0 14 1147.0 15 -2613.0 16 4183.0 17 8888.0 18 10371 . 19 15038. 20 876 1 .0 21 8090.0 22 9311.0 23 10108. 24 13416. 25 12444. 26 11782. 27 19477. 28 16004. PREDICTED CALCULATED VALUE RESIDUAL -5846.6 4354.6 -4281 .7 -39.308 -2744.0 2438.0 1 168 . 2 -3394.2 -3115.2 1570.2 729.25 -820.25 432.38 275.62 880. 19 -2338.2 -477.96 352.96 556.34 -3208.3 1075. 1 -2140. 1 2539.0 -929.01 4594.5 353.52 1239.1 -92.122 .788.30 -3401 . 3 6521.6 -2338 .6 6412.8 2475.2 12083. - 1712.4 12562. 2475.5 10445. - 1683.7 8883.2 -793.24 10981 . - 1670.4 10964 . -855.86 1 1097 . 2319 . 2 12385. 58.815 1 1898. - 115.51 13406. 6070.9 15109. 894.66 DURBIN-WATSON = 2.0393 VON NEUMAN RATIO = 2.1148 RHO = -0.08905 RESIDUAL SUM = -1893.3 RESIDUAL VARIANCE = 0.59737E+07 SUM OF ABSOLUTE ERRORS= 49172. R-SQUARE BETWEEN OBSERVED AND PREDICTED = 0.8874 RUNS TEST: 19 RUNS, 12 POSITIVE, 16 NEGATIVE, NORMAL STATISTIC |_OLS PRIV LAND/ ANOVA LIST REQUIRED MEMORY IS PAR= 4 CURRENT PAR= 40 OLS ESTIMATION 28 OBSERVATIONS DEPENDENT VARIABLE = PRIV ...NOTE..SAMPLE RANGE SET TO: 1. 28 R-SQUARE = 0.3696 R- SQUARE ADJUSTED = 0.3453 VARIANCE OF THE ESTIMATE = 0.30569E+08 STANDARD ERROR OF THE ESTIMATE = 5529.0 MEAN OF DEPENDENT VARIABLE = 4942.6 LOG OF THE LIKELIHOOD FUNCTION = -279.990 MODEL SELECTION TESTS - SEE JUDGE ET.AL.M985, P.242) AKAIKE (1969) FINAL PREDICTION ERROR- FPE = 0.32753E+08 (FPE ALSO KNOWN AS AMEMIYA PREDICTION CRITERION -PC) AKAIKE (1973) INFORMATION CRITERION- AIC = 17.304 SCHWARZf1978) CRITERION-SC = 17.399 (D REGRESSION ERROR TOTAL REGRESSION ERROR TOTAL VARIABLE NAME SS 0.46596E+09 0.79480E+09 0.12608E+10 SS 11500E+10 79480E+09 19448E+10 ESTIMATED COEFFICIENT STANDARD ERROR VARIANCE FROM MEAN DF MS F 1 . 0.46596E+09 15.243 26 . 0. 30569E + 08 •2 7 . 0.46695E+08 VARIANCE FROM ZERO OF MS F 2 . 0. 57498E + 09 18.809 26 . 0.30569E+08 28 . 0.69456E+08 T-RATIO PARTIAL STANDARDIZED ELASTICITY 26 DF CORR. COEFFICIENT AT MEANS LAND 128.13 CONSTANT -8341.1 32.818 3559.2 3.9042 -2.3435 0.6079 0.60793 2.6876 -0.4176 O.OOOOOE+OO -1.6876 RVATION OBSERVED PREDICTED CALCULATED NO. VALUE VALUE RESIDUAL 1 - 1492 .0 -5407.0 3915.0 2 - 4321 .0 -2575.3 - 1745.7 3 -306.00 -371.53 65.533 4 -2226 .0 717.55 -2943.6 5 - 1545.0 1870.7 -3415.7 6 -91.000 999.44 - 1090.4 7 708.00 3152.0 -2444.0 8 - 1458.0 2947.0 -4405.0 9 - 125.00 3331 .4 -3456.4 10 -2652.0 4446 . 1 -7098. 1 1 1 - 1065.0 2677 .9 -3742.9 12 1610.0 7047 . 1 -5437. 1 13 4948.0 8059.3 -3111.3 14 1147.0 10788. -964 1 .4 15 -2613.0 10378. - 12991 . 16 4183.0 4241 . 1 -58.071 17 8888.0 6688.3 2199.7 18 10371 . 6227.1 4143.9 19 15038. 5970.8 9067.2 20 876 1 .0 6560.2 2200.8 21 8090.0 4471.7 3618. 3 22 9311.0 5125.2 4185.8 23 10108. 5765.8 4342.2 24 13416 . 6329.6 7086.4 25 12444 . 6906. 1 5537.9 26 11782. 7790.2 3991 .8 27 19477 . 9750.6 9726.4 28 16004. 14504. 1499.9 DURBIN-WATSON = 0.6367 VON NEUMAN RATIO = 0.6602 RHO = 0.67252 RESIDUAL SUM = 0.72760E-11 RESIDUAL VARIANCE = 0.30569E+08 SUM OF ABSOLUTE ERR0RS= 0.12316E+06 R- SQUARE BETWEEN OBSERVED AND PREDICTED = 0.3696 RUNS TEST: 5 RUNS, 14 POSITIVE, 14 NEGATIVE, NORMAL STATISTIC = -3.8516 COEFFICIENT OF SKEWNESS = -0.3221 WITH STANDARD DEVIATION OF 0.4405 COEFFICIENT OF EXCESS KURTOSIS = -0.0304 WITH STANDARD DEVIATION OF 0.8583 GOODNESS OF FIT TEST FOR NORMALITY OF RESIDUALS - 6 GROUPS OBSERVED 1.0 2.0 11.0 10.0 4.0 0.0 EXPECTED 0.6 3.8 9.6 9.6 3.8 0.6 CHI-SQUARE = 1.9482 WITH 2 DEGREES OF FREEDOM |_OLS PRIV DISC/ ANOVA LIST REQUIRED MEMORY IS PAR= 4 CURRENT PAR= 40 OLS ESTIMATION 28 OBSERVATIONS DEPENDENT VARIABLE = PRIV ...NOTE..SAMPLE RANGE SET TO: 1. 28 R-SQUARE = 0.8419 R-SQUARE ADJUSTED = 0.8359 VARIANCE OF THE ESTIMATE = 0.76649E+07 STANDARD ERROR OF THE ESTIMATE = 2768.6 MEAN OF DEPENDENT VARIABLE = 4942.6 LOG OF THE LIKELIHOOD FUNCTION = -260.623 MODEL SELECTION TESTS - SEE JUDGE ET.AL.M985. P.242) AKAIKE (1969) FINAL PREDICTION ERROR- FPE = 0.82124E+07 (FPE ALSO KNOWN AS AMEMIYA PREDICTION CRITERION -PC) AKAIKE (1973) INFORMATION CRITERION- AIC = 15.921 SCHWARZf1978) CRITERION-SC = 16.016 ANALYSIS OF VARIANCE - FROM MEAN SS DF MS F REGRESSION 0. 10615E+10 1 . 0.10615E+10 138. 484 1 ERROR 0. 19929E+09 26. 0.76649E+07 "0 TOTAL 0. 12608E+10 27 . 0.46695E+08 03 OQ ANALYSIS OF VARIANCE - FROM ZERO CD SS DF MS F REGRESSION 0. 17455E+10 2. 0.87274E+09 113. 862 ERROR 0. 19929E+09 26. 0.76649E+07 co TOTAL 0. 19448E*10 28. 0.69456E+08 I VARIABLE ESTIMATED STANDARD T-RATIO PARTIAL STANDARDIZED ELASTICITY NAME COEFFICIENT ERROR 26 DF CORR. COEFFICIENT AT MEANS DISC -0.10867E+06 9234.0 CONSTANT 16329. 1100.0 - 1 1 . 768 14.845 -0.9176 -0.91757 -2.3038 0.9458 O.OOOOOE+OO 3.3038 RVATION OBSERVED PREDICTED CALCULATED NO. VALUE VALUE RESIDUAL 1 - 1492.0 -7468.6 5976.6 2 -4321.0 -7033.9 2712.9 3 -306.00 -4100.0 3794.0 4 -2226.0 -839.99 - 1386.0 5 - 1545.0 -236 1 .3 816.31 6 . -91.000 - 187.99 96.995 7 708.00 572.66 135.34 8 - 1458 .0 681.33 -2139.3 9 - 125.00 464.00 -589 .00 10 -2652.0 1007.3 -3659.3 1 1 -1065.0 - 2094.0 -3159.0 12 1610.0 4593.3 -2983.3 13 4948 .0 6223.3 - 1275.3 14 1147.0 3180.6 -2033.6 15 - -2613.0 2528.6 -5141.6 16 4183.0 7309 .9 -3126.9 17 8888.0 7961 .9 926.08 18 1037 1 . 11113. -742.22 19 15038. 12309. 2729.5 20 876 1 .0 9265.9 -504.91 21 8090.0 8396.6 -306.58 22 9311.0 10570. - 1258.9 23 10108. 11113. - 1005.2 24 13416. 1 1222. 2194. 1 25 12444. 11222. 1222. 1 26 11782. 11331 . 451.45 27 19477 . 13395. 6081.8 28 16004 . 13830. 2174.1 DURBIN-WATSON = 0.9032 VON NEUMAN RATIO = 0.9367 RHO = 0.45778 RESIDUAL SUM = -0.77307E-10 RESIDUAL VARIANCE = 0.76649E+07 SUM OF ABSOLUTE ERRORS= 58622. R-SQUARE BETWEEN OBSERVED AND PREDICTED = 0.8419 RUNS TEST: 9 RUNS, 13 POSITIVE, 15 NEGATIVE, NORMAL STATISTIC = -2.2956 COEFFICIENT OF SKEWNESS = 0.5115 WITH STANDARD DEVIATION OF 0.4405 COEFFICIENT OF EXCESS KURTOSIS = 0.2115 WITH STANDARD DEVIATION OF 0.8583 GOODNESS OF FIT TEST FOR NORMALITY OF RESIDUALS - 6.GROUPS OBSERVED 0.0 5.0 10.0 10.0 1.0 2.0 EXPECTED 0.6 3.8 9.6 9.6 3.8 0.6 CHI-SQUARE = 6.0268 WITH 2 DEGREES OF FREEDOM I_AUTO PRIV WAGE/ ANOVA LIST REQUIRED MEMORY IS PAR= 5 CURRENT PAR= 40 DEPENDENT VARIABLE = PRIV ..NOTE..R-SQUARE,ANOVA,RESIDUALS DONE ON ORIGINAL VARS LEAST SQUARES ESTIMATION 28 OBSERVATIONS BY COCHRANE-ORCUTT TYPE PROCEDURE WITH CONVERGENCE = 0.00100 ITERATION 1 2 3 4 RHO 0 .00000 0.55096 0.55668 0.55688 LOG L.F. -268.551 -263.787 -263.789 -263.789 SSE 0. 35108E + 09 0 . 2466 1E + 09 0.24656E+09 0.24656E+09 LOG L.F. = •263. 789 AT RHO 0.55688 l T3 Co OQ (T> NJ O o RHO ESTIMATE 0. 55688 ASYMPTOTIC VARIANCE 0.02464 ASYMPTOTIC ST.ERROR 0.15697 R-SQUARE = 0.8044 R-SQUARE ADJUSTED = VARIANCE OF THE ESTIMATE = 0.94832E+07 STANDARD ERROR OF THE ESTIMATE = 3079.5 MEAN OF DEPENDENT VARIABLE = 4942.6 LOG OF THE LIKELIHOOD FUNCTION = -263.789 MODEL SELECTION TESTS - SEE JUOGE ET.AL.M985, AKAIKE (1969) FINAL PREDICTION ERROR- FPE = ASYMPTOTIC T-RATIO 3.54773 0.7969 P.242) 0.10161E+08 (FPE ALSO KNOWN AS AMEMIYA PREDICTION CRITERION -PC) AKAIKE (1973) INFORMATION CRITERION- AIC SCHWARZI 1978) CRITERION-SC = 16.229 16.134 ANALYSIS OF VARIANCE - FROM MEAN SS DF MS REGRESSION 0. 10142E+10 1 . 0. 10142E+10 ERROR 0. 24656E+09 26. 0.94832E+07 TOTAL 0. 12608E+10 27 . 0.46695E+08 ANALYSIS OF VARIANCE - FROM ZERO SS DF MS REGRESSION 0 . 16982E+10 2 . 0.84910E+09 ERROR 0. 24656E*09 26 . 0.94832E*07 TOTAL 0 . 19448E+ 10 28. 0.69456E*08 VARIABLE NAME ESTIMATED COEFFICIENT STANDARD ERROR T-RATIO 26 DF PARTIAL STANDARDIZED ELASTICITY CORR. COEFFICIENT AT MEANS WAGE CONSTANT 224 . 19 • 12360. 47.143 3887 .7 4.7556 3.1793 0.6821 -0.5291 0.83767 O.OOOOOE+OO 3.5555 •2.5008 00 O OBSERVATION NO. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 OBSERVED VALUE - 1492.0 -4321.0 •306.00 - 2226 .0 - 1545 .0 -91.000 708.00 - 1458 .0 - 125.00 -2652.0 PREDICTED VALUE - 375 1 .2 - 1065. 1610. 4948 . 1147 -2613. 4183. 8888. 10371. 15038. 8761 .0 8090.0 9311.0 10108. 13416 . 12444 . 11782. 19477 . 16004 . - 1999 - 3558 - 1328 - 1924 - 1588 -345.22 604.63 - 121.90 1351 .5 455.79 1876 .0 4395 . 3 7031 .6 4402 .0 2596.5 6686.7 9081.8 10361 . 13413. 9123.9 9209.1 9908 .6 10758. 12505. 11886. 12066. 16554. CALCULATED RESIDUAL 2259.2 -2321.1 3252.5 -897.98 379. 15 1497.0 1053.2 -2062.6 -3.1017 -4003.5 - 1520.8 -266.01 552.66 -5884.6 -7015.0 1586.5 2201.3 1289.2 4677 . 1 -4651 .5 • 1033.9 101.91 199.36 2657.7 -61.136 -104.39 7411.3 •550.43 DURBIN-WATSON = 1.9135 VON NEUMAN RATIO = 1.9844 RHO = 0.03240 RESIDUAL SUM = -1258.1 RESIDUAL VARIANCE = 0.95440E+07 SUM OF ABSOLUTE ERR0RS= 59494. R-SOUARE BETWEEN OBSERVED AND PREDICTED = 0.8039 RUNS TEST: 14 RUNS, 14 POSITIVE, 14 NEGATIVE, NORMAL STATISTIC = -0.3852 |_AUTO PRIV LAND/ ANOVA LIST REQUIRED MEMORY IS PAR= 5 CURRENT PAR= 40 DEPENDENT VARIABLE = PRIV . .NOTE. .R-SQUARE,ANOVA,RESIDUALS DONE ON ORIGINAL VARS LEAST SOUARES ESTIMATION 28 OBSERVATIONS BY COCHRANE-ORCUTT TYPE PROCEDURE WITH CONVERGENCE = 0.00)00 ITERATION RHO LOG L.F. SSE 1 0.00000 -279.990 0.79480E+09 2 0.67252 -269.046 0.35599E+03 . 0.83467 -267.663 0.31576E+09 4 0.88346 -267.493 0.30836E+05. 0.89334 -267.481 0.30719E+09 6 0.89496 -267.480 0.30701E+07 0.89521 -267.480 0.30698E+09 LOG L.F. = -267.480 AT RHO = 0.89521 ASYMPTOTIC ASYMPTOTIC ASYMPTOTIC ESTIMATE VARIANCE ST.ERROR T-RATIO RHO 0.89521 0.00709 0.08422 10.62936 R- SQUARE = 0.7565 R- SQUARE ADJUSTED = 0.747 1 VARIANCE OF THE ESTIMATE = 0.11807E»08 STANDARD ERROR OF THE ESTIMATE = 3436.1 MEAN OF DEPENDENT VARIABLE = 4942.6 LOG OF THE LIKELIHOOD FUNCTION = -267.480 MOOEL SELECTION TESTS - SEE JUDGE ET.AL.U985. P.242) AKAIKE (1969) FINAL PREDICTION ERROR - FPE = 0.12650E*08 (FPE ALSO KNOWN AS AMEMIYA PREDICTION CRITERION -PC) AKAIKE (1973) INFORMATION CRITERION- AIC = 16.353 SCHWARZl1978) CRITERION-SC = 16.448 ANALYSIS OF VARIANCE - FROM MEAN SS DF MS REGRESSION 0. 95378E+09 1 . 0.95378E+09 ERROR 0. 30698E+09 26 . 0. 1 1807E + 08 TOTAL 0. 12608E+10 27 . 0.46695E+08 ANALYSIS OF VARIANCE - FROM ZERO SS DF MS REGRESSION 0. 16378E+10 2. 0.81889E*09 ERROR 0. 30698E+09 26. 0. 1 1807E + 08 TOTAL 0. 19448E+ 10 28. 0.69456E+08 VARIABLE NAME ESTIMATED COEFFICIENT STANDARD T-RATIO ERROR 26 DF PARTIAL STANDARDIZED ELASTICITY CORR. COEFFICIENT AT MEANS LAND CONSTANT •6.8307 6519.5 37. 145 6192.4 -0.18389 1.0528 -0.0360 -0.32410E-01 -0.14328 0.2022 O.OOOOOE+OO 1.3190 I_AUTO PRIV DISC/ ANOVA LIST REQUIRED MEMORY IS PARr 5 CURRENT PAR= 40 DEPENDENT VARIABLE = PRIV ..NOTE..R-SQUARE,ANOVA,RESIDUALS DONE ON ORIGINAL VARS LEAST SQUARES ESTIMATION 28 OBSERVATIONS BY COCHRANE-ORCUTT TYPE PROCEDURE WITH CONVERGENCE = 0.00100 ITERATION 1 2 3 4 RHO 0.00000 0.45778 0.47391 0.47490 LOG L.F. -260.623 -256.806 -256 . 767 -256.765 SSE 19929E+09 15046E+09 14994E+09 14991E+09 LOG L.F. = •256.765 AT RHO = 0.47490 l CO fD ro o u> RHO ESTIMATE 0.47490 ASYMPTOTIC VARIANCE 0.02766 ASYMPTOTIC ST.ERROR 0. 16631 R- SQUARE = 0.8811 R- SQUARE ADJUSTED = VARIANCE OF THE ESTIMATE = 0.57659E+07 STANDARD ERROR OF THE ESTIMATE = 2401.2 MEAN OF DEPENDENT VARIABLE = 4942.6 LOG OF THE LIKELIHOOD FUNCTION = -256.765 MODEL SELECTION TESTS - SEE JUDGE ET.AL.M985, AKAIKE (1969) FINAL PREDICTION ERROR- FPE = ASYMPTOTIC T-RATIO 2.85547 = 0.8765 P.242) 0.61777E+07 (FPE ALSO KNOWN AS AMEMIYA PREDICTION CRITERION -PC) AKAIKE (1973) INFORMATION CRITERION- AIC SCHWARZI1978) CRITERION-SC = 15.731 15.636 ANALYSIS OF VARIANCE - FROM MEAN SS DF MS REGRESSION 0 . 1 1 108E+ 10 1 . 0.11108E+ 10 ERROR 0 . 14991E+09 26 . 0. 57659E + 07 TOTAL 0. 12608E+10 27 . 0.46695E+08 ANALYSIS OF VARIANCE - FROM ZERO SS DF MS REGRESSION 0. 17949E+ 10 2. 0.89743E+09 ERROR 0. 14991E+09 26 . 0.57659E+07 TOTAL 0. 19448E+10 28. 0.69456E+08 VARIABLE NAME ESTIMATED COEFFICIENT STANDARD T-RATIO-ERROR 26 DF PARTIAL STANDARDIZED ELASTICITY CORR. COEFFICIENT AT MEANS DISC CONSTANT •0. 10693E+06 16393. 13066. 1616.0 -8. 1838 10.144 -0.8487 -0.90290 0.8935 O.OOOOOE+OO -2.2670 3.3166 OBSERVATION OBSERVED PREDICTED CALCULATED NO. VALUE VALUE RESIDUAL 1 - 1492.0 -7024 .6 5532.6 2 - 4321 .0 -3969.5 - 351.52 3 - 306.00 -2629.0 2323.0 4 -2226.0 1114.5 -3340.5 5 - 1545.0 -2817 . 7 1272.7 6 -91.000 355. 19 -446 . 19 7 708 .00 778.58 -70.583 8 - 1458.0 909.49 -2367.5 9 - 125.00 - 383.78 258.78 10 -2652.0 885.47 -3537.5 1 1 -1065.0 500. 78 - 1565.8 12 1610.0 3206.0 - 1596.0 13 4948.0 4912.3 35.652 14 1147.0 •* 2741.8 - 1594.8 15 • -2613.0 1717.0 -4330.0 16 ' 4183.0 4941 .0 -757.97 17 8888.0 6575.6 2312.4 18 10371 . 1 1606 . - 1235.3 19 15038. 12014. 3023.9 20 8761 .0 10678. - 1916.9 21 8090.0 8263.4 -173.37 22 9311.0 10490. -1178.5 23 10108. 10588. -480.42 24 13416. 10820. 2596. 1 25 12444. 12340. 103.87 26 1 1782 . 1 1985. -203.46 27 19477 . 13652. 5825. 1 28 16004. 16769. -765. 18 DURBIN-WATSON = 2.0337 VON NEUMAN RATIO = 2.1090 RHO = -0.11673 RESIDUAL SUM = -2627.5 RESIDUAL VARIANCE = 0.60314E+07 SUM OF ABSOLUTE ERR0RS= 49196. R-SQUARE BETWEEN OBSERVED AND PREDICTED = 0.8764 RUNS TEST: 18 RUNS, 10 POSITIVE. 18 NEGATIVE, NORMAL STATISTIC = 1.7435 |_01S POR EXCH SPR/ ANOVA LIST REQUIRED MEMORY IS PAR= 4 CURRENT PAR= 40 OLS ESTIMATION 28 OBSERVATIONS DEPENDENT VARIABLE = POR . . .NOTE . .SAMPLE RANGE SET TO: 1, 28 R-SQUARE = 0.5043 R-SOUARE ADJUSTED = 0.4646 VARIANCE OF THE ESTIMATE = O.91210E+O7 STANDARD ERROR OF THE ESTIMATE = 3020.1 MEAN OF DEPENDENT VARIABLE = -1282.0 LOG OF THE LIKELIHOOD FUNCTION = -262.509 MODEL SELECTION TESTS - SEE JUDGE ET.AL.M985, P.242) AKAIKE (1969) FINAL PREDICTION ERROR- FPE = 0.10098E+08 (FPE ALSO KNOWN AS AMEMIYA PREDICTION CRITERION -PC) AKAIKE (1973) INFORMATION CRITERION- AIC = 16.127 SCHWARZ(1978) CRITERION-SC = 16.270 l T3 Co 0°. ft) to o REGRESSION ERROR TOTAL REGRESSION ERROR TOTAL ANALYSIS OF VARIANCE SS DF 0.23195E+09 2. 0.22803E*09 25. 0.4599.8E + 09 27. ANALYSIS OF VARIANCE SS 27797E+09 22803E+09 0.50599E+09 DF 3. 25. 28 . - FROM MEAN MS 0.11598E+09 0.91210E+07 0.17036E+08 - FROM ZERO MS 0.92656E+08 0.91210E+07 0.18071E+08 F 12.715 F 10.159 VARIABLE ESTIMATED STANDARD T-RATIO PARTIAL STANDARDIZED ELASTICITY NAME COEFFICIENT ERROR 25 DF CORR. COEFFICIENT AT MEANS EXCH 56.650 11.499 4.9264 0.7018 0.96730 -13.117 SPR -38939. 14637. -2.6604 -0.4697 -0.52237 0.40638E-02 CONSTANT -18092. 3459.3 -5.2299 -0.7228 0.OOOOOE+00 14.113 I <D Ol OQ fD O RVATION OBSERVED PREDICTED CALCULATED NO. VALUE VALUE RESIDUAL 1 -22.220 -2206.9 2184.7 2 11 . 760 -2550.1 2561 .8 3 7 1 .020 - 1552.8 1623.8 4 146.74 - 155.95 302.69 5 140.62 -218.25 358.87 6 80.110 1034.8 -954.70 7 - 18.420 144 1.0 - 1459.4 8 -9.9000 1113.4 -1123.3 9 296.47 585.09 -288.62 10 751.12 2262.9 -1511.8 1 1 187.50 2410.7 -2223.2 12 550.30 1701 .5 -1151.2 13 44.780 -786.00 830.78 14 -829.75 - 1554.9 725. 19 15 -347.47 -992.43 644.96 16 1306.2 5692.5 -4386.3 17 973.90 -795.63 1769.5 18 192.83 - 1473.9 1666.7 19 -632.09 -3413.2 2781 . 1 20 -280.37 -2872.0 2591.7 21 2138.0 -2864 . 3 5002.3 22 1639.5 -2288.2 3927.8 23 198.86 - 1842.6 2041 .4 24 -650.61 -3289.7 2639. 1 25 -5308.8 -3270.7 -2038.0 26 -9145.1 -3666.2 -5478.9 27 - 15506. -7437.8 -8068.0 28 - 1 1874. -8905.6 -2968.7 DURBIN-WATSON = 0.6774 VON NEUMAN RATIO = 0.7025 RHO = 0.65692 RESIDUAL SUM = -0.31832E-10 RESIDUAL VARIANCE = 0.91210E+07 SUM OF ABSOLUTE ERR0RS= 63305. R-SOUARE BETWEEN OBSERVEO AND PREDICTED = 0.5043 RUNS TEST: 6 RUNS, 16 POSITIVE, 12 NEGATIVE, NORMAL STATISTIC = -3.4291 COEFFICIENT OF SKEWNESS = -0.8703 WITH STANDARD DEVIATION OF 0.4405 COEFFICIENT OF EXCESS KURTOSIS = 1.0257 WITH STANDARD DEVIATION OF 0.8583 GOODNESS OF FIT TEST FOR NORMALITY OF RESIDUALS - 6 GROUPS OBSERVED 1.0 2.0 9.0 14.0 2.0 0.0 EXPECTED 0.6 3.8 9.6 9.6 3.8 0.6 CHI - SQUARE = 4.6546 WITH 1 DEGREES OF FREEDOM I_AUTO POR EXCH SPR/ ANOVA LIST REQUIRED MEMORY IS PAR= 5 CURRENT PAR= 40 DEPENDENT VARIABLE = POR ..NOTE..R-SQUARE,ANOVA,RESIDUALS DONE ON ORIGINAL VARS LEAST SQUARES ESTIMATION 28 OBSERVATIONS BY COCHRANE-ORCUTT TYPE PROCEDURE WITH CONVERGENCE = 0.00100 ITERATION RHO LOG L.F. SSE 1 0.00000 -262.509 0.22803E+09 2 0.65692 -251.901 0.10475E+03 0.85839 -250.374 0.91378E+08 '4 0.90181 -250.424 0.90589E+05' 0.90871 -250.452 0.90545E+08 6 0.90982 -250.457 0.90540E+07 0.91001 -250.458 0.90540E+08 LOG L.F. = -250.458 AT RHO = 0.91001 ASYMPTOTIC ASYMPTOTIC ASYMPTOTIC ESTIMATE VARIANCE ST.ERROR T-RATIO RHO 0.91001 0.00614 0.07835 11.61447 R-SQUARE = 0.8032 R-SQUARE ADJUSTED = 0.7874 VARIANCE OF THE ESTIMATE = 0. 36216E+07 STANDARD ERROR OF THE ESTIMATE = 1903.0 MEAN OF DEPENDENT VARIABLE = -1282.0 LOG OF THE LIKELIHOOD FUNCTION = -250.458 MODEL SELECTION TESTS - SEE JUDGE ET.AL.U985, P.242) AKAIKE (1969) FINAL PREDICTION ERROR - FPE = 0.40096E + 07 (FPE ALSO KNOWN AS AMEMIYA PREDICTION CRITERION -PC) AKAIKE (1973) INFORMATION CRITERION- AIC = 15.203 SCHWARZt1978) CRITERION-SC = 15.346 ANALYSIS OF VARIANCE - FROM MEAN SS DF MS REGRESSION 0. 36944E+09 2 . 0.18472E+09 ERROR 0. 90540E+08 25. 0.36216E+07 TOTAL 0. 45998E+09 27 . 0.17036E+08 ANALYSIS OF VARIANCE - FROM ZERO SS DF MS REGRESSION 0. .41545E+09 3. 0.13848E+09 ERROR 0 .90540E+08 25. 0.36216E+07 TOTAL 0. . 50599E + 09 28 . 0.18071E+08 VARIABLE ESTIMATED STANDARD T-RATIO PARTIAL STANDARDIZED ELASTICITY NAME COEFFICIENT ERROR 25 DF CORR. COEFFICIENT AT MEANS EXCH 22.168 10.314 2.1493 0.3949 0.37852 -5.1327 SPR - 737.86 18196. - 0.40551E-01-0.0081 -0.98984E-02 0.77005E-CONSTANT -9391.1 4180.5 -2.2464 -0.4098 O.OOOOOE+OO 7.3256 Co OO (b to o OO RVATION OBSERVED PREDICTED CALCULATED NO. VALUE VALUE RESIDUAL 1 -22.220 - 1496.0 1473.8 2 11 .760 - 161.36 173.12 3 71 .020 - 105.62 176 .64 4 146.74 -42.420 189 . 16 5 140.62 1.2179 139.40 6 80.1 10 20.468 59.642 7 - 18.420 -48.508 30.088 8 -9.9000 - 151.38 141.48 9 296.47 - 147.99 444.46 10 751.12 171.71 579.41 1 1 187.50 559.31 -371 .81 12 550.30 - 194.65 744.95 13 44.780 -664. 10 708.88 14 -829.75 -459.71 -370.04 15 - 347.47 -999.58 652.11 16 1306.2 1757 .8 -451.58 17 973.90 - 1100. 1 2074.0 18 192.83 29.225 163.60 19 -632.09 - 1393.3 761 .22 20 -280.37 -802.04 521.67 21 2138.0 -497.85 2635.8 22 1639.5 1437.0 202.54 23 198.86 1702.9 - 1504. 1 24 -650.61 -435.1 1 -215.50 25 -5308.8 -960.50 -4348.3 26 -9145 . 1 -5186.0 -3959. 1 27 - 15506. - 10238. -5267.5 28 -11874. - 15149. 3274 . 7 DURBIN-WATSON = 1.3507 VON NEUMAN RATIO = 1.4008 RHO = 0.28829 RESIDUAL SUM = -1341.2 RESIDUAL VARIANCE = 0.36935E+07 SUM OF ABSOLUTE ERRORS= 31635. R-SQUARE BETWEEN OBSERVED AND PREDICTED = 0.8010 RUNS TEST: 9 RUNS, 20 POSITIVE, 8 NEGATIVE, NORMAL STATISTIC = 1.6319 I_OLS DIR WAGE LAND EXCH SPR GOV/ ANOVA LIST REQUIRED MEMORY IS PAR= 5 CURRENT PAR= 40 OLS ESTIMATION 28 OBSERVATIONS DEPENDENT VARIABLE = DIR .. .NOTE. .SAMPLE RANGE SET TO: 1, 28 R-SQUARE = 0.8102 R-SQUARE ADJUSTED = 0.7671 VARIANCE OF THE ESTIMATE = 92457. STANDARD ERROR OF THE ESTIMATE = 304.07 MEAN OF DEPENDENT VARIABLE = -579.79 LOG OF THE LIKELIHOOD FUNCTION = -196.437 MODEL SELECTION TESTS - SEE JUDGE ET.AL.M985, P.242) AKAIKE (1969) FINAL PREDICTION ERROR- FPE = 0.11227E+06 (FPE ALSO KNOWN AS AMEMIYA PREDICTION CRITERION -PC) AKAIKE (1973) INFORMATION CRITERION- AIC = 11.622 SCHWARZt 1978) CRITERION-SC = 11.907 ANALYSIS OF VARIANCE T3 Co K> O REGRESSION ERROR TOTAL REGRESSION ERROR TOTAL SS 0.86850E+07 0:20340E+07 0. 10719E + 08 DF 5. 22. 27 . ANALYSIS OF VARIANCE SS 0. 18097E+08 0.20340E+07 0.20131E+08 DF 6. 22. 28. - FROM MEAN MS 0. 17370E+07 92457. 0.39700E+06 - FROM ZERO MS 0.30162E+07 92457. 0.71898E+06 F 18.787 F 32.623 VARIABLE NAME ESTIMATED COEFFICIENT STANDARD ERROR -RATIO 22 DF PARTIAL STANDARDIZED ELASTICITY CORR. COEFFICIENT AT MEANS WAGE -19.010 7.6969 -2.4698 -0.4659 -0.77033 2.5701 LAND -5.1087 3.1415 -1.6262 -0.3276 -0.26288 0.91351 EXCH 3.6925 1.6209 2.2780 0.4369 0.41302 -1.8903 SPR -6169.2 2693.2 -2.2906 -0.4388 -0.54214 0.14236E-02 GOV 0.13932E-01 0.22929E-01 0.60760 0.1285 0.96295E-01 -0.13636 CONSTANT 265.74 907.83 0.29272 0.0623 0.OOOOOE+00 -0.45833 I TJ CD-ffO ft) O I RVATION 08SERVED PREDICTED CALCULATED NO. VALUE VALUE RESIDUAL 1 -79.370 11 . 783 -91.153 2 -32.350 - 177.46 145. 1 1 3 - 1 1 .360 - 113.37 102.01 4 - 19 .020 53.889 - 72 .909 5 - 46 .870 -41.447 -5.4229 6 - 30.390 194.03 -224.42 7 - 76.320 133.47 - 209.79 8 -71 .780 18.626 -90.406 g -117.65 -179.24 61 .590 10 -112.11 -29 .919 -82. 191 11 - 187.50 - 103.85 -83.650 12 -96.650 - 399. 16 302.51 13 - 305.97 -711.74 405.77 14 -935.54 - 1088.0 152.43 15 -675.79 -949.20 273.41 16 -771.28 - 125.65 -645.63 17 -667.85 -786. 17 118.32 18 -491.03 -806.93 315.90 19 -531.55 -794.67 26 3. 12 20 -605.40 -856.06 250.66 21 -478.00 -782.84 304.84 22 - 1006 . 8 -721.50 -285.28 23 -971 . 46 -844.39 - 127.07 24 -717.66 - 1070.9 353.26 25 - 1322.8 - 1160.0 - 162.74 26 -1272.7 - 1313.0 40.238 27 -2165.0 - 1653.6 -51 1 .44 28 -2434.0 - 1936.9 -497.07 DURBIN-WATSON = 1.5372 VON NEUMAN RATIO = 1.5942 RHO = 0.19193 RESIDUAL SUM = -0.12506E-11 RESIDUAL VARIANCE = 92457. SUM OF ABSOLUTE ERR0RS= 6178.3 R-SQUARE BETWEEN OBSERVED AND PREDICTED = 0.8102 RUNS TEST: 13 RUNS, 14 POSITIVE. 14 NEGATIVE, NORMAL STATISTIC = COEFFICIENT OF SKEWNESS = -0.6103 WITH STANDARD DEVIATION OF 0.4405 COEFFICIENT OF EXCESS KURTOSIS = -0.0970 WITH STANDARD DEVIATION OF 0. -0. 7703 8583 GOODNESS OF FIT TEST FOR NORMALITY OF RESIDUALS - 10 GROUPS OBSERVED 0.0 1.0 2.0 3.0 8.0 6.0 7.0 1.0 0.0 0.0 EXPECTED 0.2 0.8 2.2 4.5 6.3 6.3 4.5 2.2 0.8 0.2 CHI-SQUARE = 4.3793 WITH 2 DEGREES OF FREEDOM I_OLS DIR WAGE LAND EXCH SPR/ ANOVA LIST REQUIRED MEMORY IS PAR= 5 CURRENT PAR= 40 OLS ESTIMATION 28 OBSERVATIONS DEPENDENT VARIABLE = DIR ...NOTE..SAMPLE RANGE SET TO: 1, 28 R-SQUARE = 0.8071 R-SQUARE ADJUSTED = 0.7735 VARIANCE OF THE ESTIMATE = 89921. STANDARD ERROR OF THE ESTIMATE = 299.87 MEAN OF DEPENDENT VARIABLE = -579.79 LOG OF THE LIKELIHOOD FUNCTION = -196.670 MODEL SELECTION TESTS - SEE JUDGE ET.AL.M985. P.242) AKAIKE (1969) FINAL PREDICTION ERROR- FPE = 0.10598E+06 (FPE ALSO KNOWN AS AMEMIYA PREDICTION CRITERION -PC) AKAIKE (1973) INFORMATION CRITERION- AIC = 11.567 SCHWARZt 1978) CRITERION-SC = 11.805 REGRESSION ERROR TOTAL ANALYSIS OF VARIANCE SS 0.86508E+07 0. 20682E + 07 0. 10719E + 08 DF 4. 23. 27. • FROM MEAN MS 0.21627E+07 89921 . 0.39700E+06 F 24.051 REGRESSION ERROR TOTAL ANALYSIS OF VARIANCE - FROM ZERO SS DF MS 0.18063E+08 5. 0.36127E+07 0.20682E+07 23. 89921. 0.20131E+08 28. 0.71898E+06 F 40. 176 VARIABLE NAME ESTIMATED COEFFICIENT STANDARD T-RATIO ERROR 23 DF PARTIAL STANDARDIZED ELASTICITY CORR. COEFFICIENT AT MEANS WAGE LAND EXCH SPR CONSTANT -17.618 -5.6972 3.9990 -6944 . 4 205.81 7.2466 2.9472 1.5191 2339. 1 890.00 -2.4312 - 1 .9330 2.6324 -2.9688 0.23125 -0.4522 -0.3738 0.4812 -0.5264 0.0482 -0.71393 -0.29316 0.44731 -0.61026 2. 3819 1 .0187 -2.0473 0. 16024E-C O.OOOOOE+00 -0.35498 I 09 fD RVATION OBSERVED PREDICTED CALCULATED NO. VALUE VALUE RESIDUAL 1 - 79.370 34.278 - 113.65 2 -32.350 - 191.59 159.24 3 - 1 1 .360 - 134.63 123.27 4 - 19.020 53.722 -72.742 5 -46.870 -52.709 5.8395 6 - 30.390 188. 36 -218.75 7 -76.320 119.28 - 195.60 8 -71.780 4.7931 -76.573 9 - 1 17 .65 - 180.52 62.872 10 -112.11 -29.520 -82.590 1 1 - 187.50 - 19.685 - 167.81 12 -96.650 -370.45 273.80 13 -305.97 -712.43 406.46 14 -935.54 - 1035.3 99.801 15 -675.79 -954.35 278.56 16 - 771.28 - 170.70 -600.58 17 -667.85 -872.31 204.46 18 -491.03 -810.41 319.38 19 -531 .55 -830.93 299.38 20 -605.40 -871 .87 266.47 21 - 4 78-.00 -794.07 316.07 22 - 1006.8 -701 .93 - 304.85 23 -971.46 -837.48 - 133.98 24 -717.66 - 1088.8 371 . 10 25 - 1322.8 - 1125.2 - 197.59 26 - 1272. 7 - 1243.0 -29.737 27 -2165.0 - 1619.4 -545.57 28 -2434.0 - 1987 . 3 -446.68 DURBIN-WATSON = 1.5992 VON NEUMAN RATIO = 1.6584 RHO = 0.16496 RESIDUAL SUM = -0.96634E-12 RESIDUAL VARIANCE = 89921. SUM OF ABSOLUTE ERRORS= 6373.4 R-SQUARE BETWEEN OBSERVED AND PREDICTED = 0.8071 RUNS TEST: 13 RUNS, 14 POSITIVE, 14 NEGATIVE, NORMAL STATISTIC = -0.7703 COEFFICIENT OF SKEWNESS = -0.4631 WITH STANDARD DEVIATION OF 0.4405 COEFFICIENT OF EXCESS KURTOSIS = -0.4634 WITH STANDARD DEVIATION OF 0.8583 GOODNESS OF FIT TEST FOR NORMALITY OF RESIDUALS - 10 GROUPS OBSERVED 0.0 2.0 1.0 4.0 7.0 5.0 7.0 2.0 0.0 0.0 EXPECTED 0.2 0.8 2.2 4.5 6.3 6.3 4.5 2.2 0.8 0.2 CHI-SQUARE = 5.7034 WITH 3 DEGREES OF FREEDOM I_PLOT DIR WAGE REQUIRED MEMORY IS PAR= 3 CURRENT PAR= 40 FOR MAXIMUM EFFICIENCY USE AT LEAST PAR= 3 28 OBSERVATIONS *=DIR M=MULTIPLE POINT 400.00 317.95 235.90 153.85 71 . 795 - 10.256 -92.308 -174.36 -256.41 -338.46 -420.51 -502.56 -584.62 -666.67 -748.72 - 1405 1 -1487 2 - 1569 2 - 1651 3 -1733 3 -1815 4 -1897 4 -1979 5 -2061 5 -2143 6 -2225 6 -2307 7 -2389 7 - 247 1 8 -2553 8 -2635 9 -2717 9 -2800 .0 fD ho CO -830.77 -912.82 -994.87 - 1076.9 - 1 159 .0 - 1241 .0 - 1 323. 1 M 30.000 40.000 50.000 60.000 70.000 80.000 90.000 100.000 110.000 WAGE |_PLOT DIR LAND REQUIRED MEMORY IS PAR= 3 CURRENT PAR= 40 FOR MAXIMUM EFFICIENCY USE AT LEAST PAR= 3 28 OBSERVATIONS •=DIR M=MULTIPLE POINT 400.00 317.95 235.90 153.85 71.795 - 10.256 -92.308 * .... - 174.36 -256.41 -338.46 -420.51 -502.56 -584.62 -666.67 -748.72 - 1241 .0 - 1323. 1 - 1405. 1 - 1487.2 - 1569.2 - 1651 . 3 - 1733.3 -1815.4 - 1897 . 4 - 1979 . 5 -2061 .5 -2143.6 -2225 .6 -2307.7 -2389 . 7 -2471 .8 -2553.8 -2635.9 -2717.9 -2800.0 oo fD -830.77 \ -912.82 ' v 994. 87 >1076 .9 -1159.0 60.000 80.000 100.000 120.000 140.000 160.000 180.000 20.000 40.000 LAND |_PLOT DIR EXCH 00 fD N> i—1 Ln I REQUIRED MEMORY IS PAR= 3 CURRENT PAR= FOR MAXIMUM EFFICIENCY USE AT LEAST PAR= 28 OBSERVATIONS *=DIR M=MULTIPLE POINT 400.00 317 .95 235.90 153.85 71 . 795 - 10.256 -92.308 - 174.36 -256.41 -338.46 • -420.51 -502.56 -584.62 -666 .67 -748.72 -830.77 -912.82 -994.87 - 1076.9 - 1 159.0 - 1241 .0 - 1 323. 1 - 1405. 1 - 1487.2 - 1569.2 - 1651 . 3 - 1733. 3 -1815.4 - 1897 . 4 - 1979.5 -2061 .5 -2143.6 -2225.6 -2307.7 -2389.7 -2471 .8 -2553.8 -2635.9 -2717.9 -2800.0 40 3 M M 120.000 160.000 200.000 240.000 280.000 320.000 360.000 400.000 440.000 EXCH I_PLOT DIR SPR REQUIRED MEMORY IS PAR= 3 CURRENT PAR= 40 FOR MAXIMUM EFFICIENCY USE AT LEAST PAR= 3 28 OBSERVATIONS •=DIR M=MULTIPLE POINT 400.00 317.95 235.90 153.85 7 1 . 795 - 10.256 -92.308 - 174.36 -256.41 . -338.46 -420.51 -502.56 -584.62 -666.67 -748 . 72 -830.77 -912.82 -994.87 - 1076.9 - 1 159.0 - 1241 .0 - 1323. 1 - 1405. 1 - 1487 . 2 - 1569.2 - 1651 .3 - 1733. 3 - 1815.4 - 1897 . 4 - 1979 . 5 -2061 .5 -2143.6 -2225.6 -2307.7 -2389.7 -247 1 . 8 -2553.8 -2635.9 -2717 .9 -2800.0 -0.090 -0.060 -0.030 0.000 0.030 0.060 0.090 0. 120 0. 150 SAMPLE SPR 12 I_OLS DIR WAGE LAND EXCH SPR/ ANOV*A LIST REQUIRED MEMORY IS PAR= 4 CURRENT PAR= 40 OLS ESTIMATION 12 OBSERVATIONS DEPENDENT VARIABLE = DIR ...NOTE..SAMPLE RANGE SET TO: 1, 12 R-SQUARE = 0.8886 R-SQUARE ADJUSTED = 0.8249 VARIANCE OF THE ESTIMATE = 447.40 STANDARD ERROR OF THE ESTIMATE = 21.152 MEAN OF DEPENDENT VARIABLE = -73.447 LOG OF THE LIKELIHOOD FUNCTION = -50.4140 MODEL SELECTION TESTS - SEE JUDGE ET.AL.M985, P.242) AKAIKE (1969) FINAL PREDICTION ERROR- FPE = 633.82 (FPE ALSO KNOWN AS AMEMIYA PREDICTION CRITERION -PC) AKAIKE (1973) INFORMATION CRITERION- AIC = 6.3978 SCHWARZI1978) CRITERION-SC = 6.5998 CD CO fD ho I REGRESSION ERROR TOTAL REGRESSION ERROR TOTAL ANALYSIS OF VARIANCE SS DF 24976. 4. 3131.8 728108. 11. ANALYSIS OF VARIANCE SS 89710. 3131.8 92842. DF 5. 7 . 12. FROM MEAN MS 6244 .0 447.40 2555.2 FROM ZERO MS 17942. 447.40 7736.8 F 13.956 F 40.103 VARIABLE NAME ESTIMATED COEFFICIENT STANDARD T-RATIO ERROR 7 DF PARTIAL STANDARDIZED ELASTICITY CORR. COEFFICIENT AT MEANS WAGE LAND EXCH SPR CONSTANT -6.8838 1 .4295 - 7.3348 118.28 2797 . 8 1.3482 0.60897 3.0645 398.57 1170.7 -5.1058 2.3474 -2.3934 0 . 29677 2.3900 -0.8879 0.6637 -0.6709 0.1115 0.6703 - 1.5478 0.72221 -0.44694 0.10435 O.OOOOOE+OO 4.8159 - 1 .5055 35.862 -0.79813E-01 -38.093 . OBSERVATION NO. OBSERVED VALUE PREDICTED VALUE CALCULATED RESIDUAL 2 3 4 5 6 7 8 9 10 1 1 12 •79.370 -32.350 - 1 1 . 360 - 19.020 -46 . 870 -30.390 -76.320 -71 .780 - 1 17 .65 -112.11 - 187 .50 -96.650 60.597 43.107 30.497 27.408 31 . 562 53.350 48.465 75.228 98.246 129 .46 186.80 96.650 - 18.773 10.757 19.137 8. 3875 - 15.308 22.960 -27.855 3.4478 - 19.404 17.346 0.69530 -0.13856E- 12 DURBIN-WATSON = 2.8240 VON NEUMAN RATIO = 3.0807 RHO = -0.46827 RESIDUAL SUM = -0.32994E-11 RESIDUAL VARIANCE = 447.40 SUM OF ABSOLUTE ERRORS= 164.07 R-SQUARE BETWEEN OBSERVED AND PREDICTED = 0.8886 , RUNS TEST: 9 RUNS. 6 POSITIVE, 6 NEGATIVE, NORMAL STATISTIC = 1.2111 COEFFICIENT OF SKEWNESS = -0.2933 WITH STANDARD DEVIATION OF 0.6373 T) COEFFICIENT OF EXCESS KURTOSIS = -1.2265 WITH STANDARD DEVIATION OF 1.2322 CD CD GOODNESS OF FIT TEST FOR NORMALITY OF RESIDUALS - 10 GROUPS OBSERVED 0.0 0.0 1.0 3.0 2.0 3.0 3.0 0.0 0.0 0.0 EXPECTED 0.1 0.3 1.0 1.9 2.7 2.7 1.9 1.0 0.3 0.1 CHI-SQUARE = 3.2742 WITH 3 DEGREES OF FREEDOM ro 00 |_PLOT DIR WAGE l •n CD OQ fD hO REQUIRED MEMORY IS PAR= 3 CURRENT PAR; FOR MAXIMUM EFFICIENCY USE AT LEAST PAR= 12 OBSERVATIONS *=DIR M=MULTIPLE POINT 30.000 23.846 17 .692 1 1 .538 5. 3846 •0.76923 -6.9231 - 13.077 -19.231 • -25.385 ' - 31.538 -37.692 -43.846 -50.000 -56.154 -62.308 -68.462 -74.615 -80. 769 -86.923 -93 .077 -99.231 - 105.38 - 1 1 1 .54 - 117.69 - 123.85 - 130.00 - 136. 15 -142.31 - 148 . 46 - 154.62 - 160.77 - 166.92 - 173.08 - 179.23 -185.38 - 191.54 - 197.69 -203.85 -210.00 40 3 35.000 40.000 45.000 WAGE 50.000 55.000 60.000 65.000 70.000 75.000 |_PLOT DIR EXCH REQUIRED MEMORY IS PAR= 3 CURRENT PAR= 40 FOR MAXIMUM EFFICIENCY USE AT LEAST PAR= 3 12 OBSERVATIONS •=DIR M=MULTIPLE POINT 30.000 23.846 17.692 1 1 . 538 5.3846 -0 . 76923 -6.9231 - 13.077 -19.231 -25.385 -31.538 -37.692 -43.846 I -50.000 -56 . 154 ^ ' -62.308 -68.462 fD -74.615 -80.769 5s? -86.923 -130.00 - 136 . 15 -.142. 31 - 148.46 - 154.62 - 160.77 - 166.92 -173.08 -179.23 - 185.38 - 191.54 - 197.69 -203.85 -210.00 348.000 349.500 351.000 352.500 354.000 355.500 357.000 358.500 360.000 O -93.1 -99.; - 105 -111 -117 - 123 077 231 i.38 .54 .69 I. 85 M EXCH |_PLOT DIR LAND l T> CO OO CD REQUIRED MEMORY IS PAR= 3 CURRENT PAR: FOR MAXIMUM EFFICIENCY USE AT LEAST PAR= 12 OBSERVATIONS •=DIR M=MULTIPLE POINT 30.000 23.846 17.692 1 1 . 538 5.3846 -0.76923 -6.9231 - 13.077 - 19.231 -25.385 -31.538 -37.692 -43.846 -50.000 -56.154 -62.308 -68.462 -74.615 -80.769 -86.923 -93.077 -99.231 - 105.38 - 1 1 1 . 54 -117.69 - 123.85 - 130.00 - 136. 15 -142.31 - 148.46 - 154.62 - 160.77 - 166.92 - 173.08 - 179 . 23 - 185.38 - 191 .54 - 197.69 -203.85 -210.00 40 3 15.000 30.000 45.000 LAND 60.000 75.000 90.000 105.000 120.000 135.000 |_PLOT DIR SPR 03 09 fD K) REQUIRED MEMORY IS PAR= 3 CURRENT PAR: FOR MAXIMUM EFFICIENCY USE AT LEAST PAR= 12 OBSERVATIONS •=DIR M=MULTIPLE POINT 30.000 23.846 17.692 11.538 5.3846 -0.76923 -6.9231 -13.077 - 19.231 -25.385 -31.538 -37.692 -43.846 -50.000 -56.154 -62.308 -68.462 -74.615 -80.769 -86.923 -93.077 -99.231 -105.38 -111.54 -117.69 -123.85 - 130.00 -136. 15 - 142.31 - 148.46 -154.62 - 160.77 - 166.92 - 173.08 - 179.23 -185.38 - 191.54 -197.69 -203.85 -210.00 40 3 -0.020 0.000 0.020 0.040 0.060 0.080 0. 100 0. 120 0. 140 SPR SAMPLE 13 28 PLOT DIR WAGE T3 Cu QQ fD K> CO REQUIRED MEMORY IS PAR= 3 CURRENT PAR FOR MAXIMUM EFFICIENCY USE AT LEAST PAR= 16 OBSERVATIONS •=DIR M=MULTIPLE POINT -300.00 -361 .54 -423.08 -484.62 - 546. 15 -607.69 -669.23 -730.77 -792.31 -853.85 -915.38 -976.92 - 1038.5 - 1 100.0 -1161.5 - 1223. 1 - 1284.6 - 1346.2 - 1407 . 7 - 1469.2 - 1530.8 - 1592.3 - 1653. 8 - 1715.4 - 1 776 .9 - 1838. 5 - 1900.0 - 1961.5 -2023.1 -2084.6 -2146.2 -2207.7 -2269 . 2 -2330. 8 -2392. 3 -2453.8 -2515.4 -2576.9 -2638.5 -2700.0 80.000 84.000 88.000 WAGE I ..PLOT DIR LAND •D Co 09 fi> K> REQUIRED MEMORY IS PAR= 3 CURRENT PAR= FOR MAXIMUM EFFICIENCY USE AT LEAST PAR= 16 OBSERVATIONS •=DIR M=MUL TI PL E POINT -300.00 - 361.54 - 423 .08 - 484 .62 -546.15 -607 . 69 -669 .23 -730.77 -792.31 -853.85 -915.38 -976.92 - 1038.5 - 1100 .0 -1161.5 - 1223.1 -1284.6 - 1 346.2 - 1407 . 7 - 1469 . 2 - 1530.8 • 1592.3 -1653.8 - 1715.4 - 1 776 .9 - 1838 . 5 - 1900.0 - 1961 .5 -2023. 1 -2084.6 -2146.2 - 2207 . 7 - 2269 . 2 -2330.8 -2392. 3 -2453.8 -2515.4 -2576.9 -2638.5 -2700.0 40 3 87.500 100.000 112.500 125.000 137.500 150.000 162.500 175.000 187.500 LAND I_.PLOT DIR LAND REQUIRED MEMORY IS PAR = 3 CURRENT PAR= 40 FOR MAXIMUM EFFICIENCY USE AT LEAST PAR= 3 16 OBSERVATIONS •=DIR M=MUL TIPLE POINT - 300.00 • 361.54 -423.08 -484.62 -546.15 -607.69 -669.23 -730.77 -792.31 -853.85 -915.38 -976.92 -1038.5 -1100.0 -1161.5 1 -1223.1 n3 -1284.6 03 - 1346.2 0? -1407.7 ro -1469.2 N3 -1530.8 -1592.3 - 1653.8 -1715.4 -1776.9 - 1838.5 - 1900.0 - 1961 .5 -2023 . 1 -2084.6 -2146 . 2 -2207 . 7 -2269.2 -2330.8 -2392.3 -2453.8 -2515.4 -2576.9 -2638.5 -2700.0 87.500 100.000 112.500 125.000 137.500 150.000 162.500 175.000 187.500 LAND PLOT OIR EXCH 0) 00 fD REQUIRED MEMORY IS PAR= 3 CURRENT PAR = FOR MAXIMUM EFFICIENCY USE AT LEAST PAR= 16 OBSERVATIONS •=DIR M=MULTIPLE POINT 40 3 -300.00 -361 .54 -423 -484 -546 -607 -669 -730 -792 -853 -915 -976 - 1038 - 1 100 - 1 161 - 1223 - 1284 - 1346 - 1407 . 7 - 1469 . 2 - 1530 - 1592 - 1653 - 1715 - 1776 - 1838 - 1900 - 1961 -2023 -2084.6 -2146.2 -2207 -2269 -2330 -2392 -2453 -2515 - 2576 -2638 - 2700 .08 .62 . 15 .69 . 23 . 77 .31 .85 . 38 .92 .5 .0 .5 . 1 .6 . 2 120.000 160.000 200.000 240.000 280.000 320.000 360.000 400.000 440.000 EXCH I..PLOT DIR SPR Co OO fD fo REQUIRED MEMORY IS PAR= 3 CURRENT PAR= 40 FOR MAXIMUM EFFICIENCY USE AT LEAST PAR= 3 16 OBSERVATIONS •=DIR M=MUL TIPLE POINT -300.00 - 361.54 -423.08 -484 .62 -546 . 15 -607.69 -669 . 23 -730.77 -792.31 -853.85 -915.38 -976.92 -1038.5 - 1 100.0 -116 1.5 - 1223 . 1 - 1284 .6 - 1 346.2 - 1407 . 7 - 1469 . 2 - 1530.8 - 1592. 3 -16 5 3.8 - 1715.4 -1776.9 - 1838 . 5 - 1900.0 -1961.5 -2023. 1 -2084.6 -2146.2 - 2207 .7 -2269.2 -2330.8 - 2392 . 3 -2453.8 -2515.4 -2576.9 -2638.5 -2700.0 -0.088 -0.075 -0.063 -0.050 -0.038 SPR .025 -0.013 -0.000 0.012 |_OLS DIR WAGE LAND EXCH SPR/ ANOVA LIST REQUIRED MEMORY IS PAR= 4 CURRENT PAR= 40 OLS ESTIMATION 16 OBSERVATIONS DEPENDENT VARIABLE = OIR ...NOTE..SAMPLE RANGE SET TO: 13, 28 R-SOUARE = 0.8239 R-SQUARE ADJUSTED = 0.7598 VARIANCE OF THE ESTIMATE = 84976. STANDARD ERROR OF THE ESTIMATE = 291.51 MEAN OF DEPENDENT VARIABLE = -959.55 LOG OF THE LIKELIHOOD FUNCTION = -110.506 MODEL SELECTION TESTS - SEE JUDGE ET.AL.M985, P.242) AKAIKE (1969) FINAL PREDICTION ERROR- FPE = 0.11153E+06 (FPE ALSO KNOWN AS AMEMIYA PREDICTION CRITERION -PC) AKAIKE (1973) INFORMATION CRITERION- AIC = 11.600 SCHWARZ(1978)"CRITERION-SC = 11.842 ANALYSIS OF VARIANCE I CD 00 CD N) K> tsj oo I I REGRESSION ERROR TOTAL REGRESSION ERROR TOTAL SS 0.43721E'07 0.93473E+06 0.53068E+07 DF 4 . 1 1 . 15. ANALYSIS OF VARIANCE SS 0. 19104E+0B 0.93473E+06 0.2OO39E+08 DF 5. 1 1 . 16 . - FROM MEAN MS 0. 10930E*07 84976 . 0.35379E+06 - FROM ZERO MS 1 0. 38208E + 07 84976. 0.12524E+07 F 12.863 F 44.963 VARIABLE NAME ESTIMATED COEFFICIENT STANDARD ERROR T-RATIO 1 1 DF PARTIAL STANDARDIZED ELASTICITY CORR. COEFFICIENT AT MEANS WAGE LAND EXCH SPR CONSTANT •62.145 • 18.316 •2.2909 •698.92 7978.0 15.643 6.8857 2.7 180 5589 . 1 2581.8 -3.9727 - 2.6600 -0.84288 -0.12505 3.0901 -0.7676 -0.6257 -0.2463 -0.0377 0.6817 -0.79489 -0.64866 -0.22792 6.3883 2.3558 0.597 1 3 -0.29726E-01 -0.26904E-01 O.OOOOOE+OO -8.3143 RVATION OBSERVED PREDICTED CALCULATED NO. VALUE VALUE RESIDUAL 13 -305.97 -80.638 -225.33 14 -935.54 -924.49 - 11.050 15 -675 . 79 -981 .33 305.54 16 - 771.28 -479.51 -291 .77 17 -667 .85 -782.02 114.17 18 -491 .03 -666.58 175.55 19 -531.55 -615.61 84.058 20 -605.40 -924.01 318.61 21 -478.00 -544.81 66.807 22 - 1006.8 -675.73 - 331.05 23 -971 .46 -897 . 22 -74.242 24 -717.66 - 1102.2 384.57 25 - 1322.8 - 1234. 1 -88.697 26 - 1272.7 - 1377. 1 104.40 27 -2165.0 -1648.8 -516.19 28 -2434 .0 -2418.6 -15.366 DURBIN-WATSON = 2.2790 VON NEUMAN RATIO = 2.4309 RHO = -0.16683 RESIDUAL SUM = -0.21032E-11 RESIDUAL VARIANCE = 84976. SUM OF ABSOLUTE ERRORS= 3107.4 R-SQUARE BETWEEN OBSERVED AND PREDICTED = 0.8239 RUNS TEST: 9 RUNS. 8 POSITIVE, 8 NEGATIVE. NORMAL STATISTIC COEFFICIENT OF SKEWNESS = -0.4141 WITH STANDARD DEVIATION OF 0.5643 COEFFICIENT OF EXCESS KURTOSIS = -0.2418 WITH STANDARD DEVIATION OF GOODNESS OF FIT TEST FOR NORMALITY OF RESIDUALS - 10 GROUPS OBSERVED 0.0 0.0 1.0 3.0 4.0 4.0 3.0 1.0 0.0 0.0 EXPECTED 0.1 0.4 1.3 2.5 3.6 3.6 2.5 1.3 0.4 0.1 CHI-SQUARE = 1.5062 WITH 3 DEGREES OF FREEDOM | _END 0.0000 .0908 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Country Views Downloads
China 183 4
United States 36 0
Russia 14 0
Japan 4 0
United Kingdom 3 0
Benin 3 0
Thailand 2 0
Turkey 2 0
India 2 3
Republic of Korea 2 1
Poland 2 0
Italy 1 0
Norway 1 0
City Views Downloads
Beijing 34 0
Hangzhou 33 0
Unknown 22 18
Nanjing 18 0
Guangzhou 15 1
Changsha 10 0
Shanghai 10 0
Cabot 10 0
Chongqing 10 0
Hefei 9 0
Shenzhen 9 3
Harbin 8 0
Clarks Summit 7 0

{[{ mDataHeader[type] }]} {[{ month[type] }]} {[{ tData[type] }]}
Download Stats

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0097354/manifest

Comment

Related Items