Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Influential, efficient, or both : does committee review of legislation matter in Westminster parliaments? Whelan, Colin 2013

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

Item Metadata


24-ubc_2013_fall_whelan_colin.pdf [ 403.09kB ]
JSON: 24-1.0074182.json
JSON-LD: 24-1.0074182-ld.json
RDF/XML (Pretty): 24-1.0074182-rdf.xml
RDF/JSON: 24-1.0074182-rdf.json
Turtle: 24-1.0074182-turtle.txt
N-Triples: 24-1.0074182-rdf-ntriples.txt
Original Record: 24-1.0074182-source.json
Full Text

Full Text

INFLUENTIAL, EFFICIENT, OR BOTH:DOES COMMITTEE REVIEW OF LEGISLATION MATTER IN WESTMINSTER PARLIAMENTS?by COLIN WHELANB.A., Simon Fraser University, 2011A THESIS SUBMITTED IN PARTIAL FULFILLMENT OFTHE REQUIREMENTS FOR THE DEGREE OFMASTER OF ARTSinTHE FACULTY OF GRADUATE AND POSTDOCTORAL STUDIES(Political Science)THE UNIVERSITY OF BRITISH COLUMBIA(Vancouver)August 2013? Colin Whelan, 2013AbstractThis paper examines the legislative impact of committee review in Canadian provinces.Traditionally, legislation in Westminster systems has been scrutinized by a Committee of theWhole House. Today, four of Canada's ten provinces have moved this committee stage intothe  separate committee system. The British House of Commons and Canada's FederalParliament have also made similar reforms. These reforms consistently received cross-partysupport, as reformers believed that the changes would both increase efficiency and alsocreate more legislative influence for committees. I test for these effects by measuring boththe ability of committees to amend legislation and the efficiency of the legislative process inCanada's provinces from 1983 to 2013. Results show that while legislation is amended morefrequently when reviewed in the committee system, there are no measurable gains inefficiency. This finding has important implications for questions about legislative-executivebalance and the effects of institutional reform in Westminster-style parliaments.iiPrefaceThis thesis is original, unpublished, independent work by the author, Colin Whelan.iiiTable of ContentsAbstract......................................................................................................................................iiPreface......................................................................................................................................iiiTable of Contents......................................................................................................................ivList of Tables..............................................................................................................................vAcknowledgements...................................................................................................................vi1. Introduction............................................................................................................................12. The reforms in context...........................................................................................................52.1 The arguments for reform in Canada..............................................................................52.2 The motivations for reform in Canada............................................................................73. Political science theory and committee power....................................................................103.1 Perspectives in the literature.........................................................................................103.2 Informational committee power in Westminster parliaments.......................................123.3 Committees and efficiency in Westminster parliaments...............................................164. Data......................................................................................................................................194.1 Control variables...........................................................................................................194.2 Sources of data..............................................................................................................214.3 Descriptive statistics......................................................................................................234.4 Informational versus policy effects...............................................................................275. Results..................................................................................................................................325.1 Modelling amended legislation.....................................................................................325.2 Modelling legislative efficiency....................................................................................366. Conclusion...........................................................................................................................40References................................................................................................................................42ivList of TablesTable 1: Average amendment rates and efficiency by province, 1983-2012...........................24Table 2: Measures of professionalization in Canada's provinces, 1983-2013.........................26Table 3: Amendment rates by recorded division (4 provinces), 1983-2013............................27Table 4: Amendment rates by bill importance (all provinces), 1983-2013.............................28Table 5: Unpassed bills by Province, various time periods.....................................................30Table 6: Regressions, amended bills in Canada's Provinces, 1983-2013................................33Table 7: Predicted probabilities from Model 2 .......................................................................35Table 8: Regressions, legislative efficiency in Canada's provinces, 1983-2013......................38vAcknowledgementsMany thanks to my supervisor, Prof. Christopher Kam, for great advice and many, manyhelpful suggestions. Thanks also to Prof. Gerald Baier for his role as my second reader.In general, I am very grateful to the entire faculty and staff of the Political Sciencedepartment at UBC who were nothing but friendly and helpful during my time here. Thanksin particular to all those Professors who I had the privilege of taking a class from or workingwith.Finally I want to acknowledge the unending support from Mom, Dad, and Anna throughoutmy entire university career.vi1. IntroductionIn Westminster parliaments committee stage of legislation traditionally took place in aCommittee of the Whole House. Over the last number of decades however, both theCanadian and the British House of Commons have moved committee stage out of the wholehouse and into the committee system of the legislature. Similar reforms have beenundertaken in 4 of the 10 provinces in Canada. Reformers consistently claim that committeereview not only improves the efficiency of the legislative process, but increases the realinfluence of the legislature relative to the executive. Typical views of Westminsterparliaments see them as highly dominated by executives, and would suggest that any realincrease in legislative influence is unlikely. However the difference between the two types ofsystems is clear in the fact that 15% of legislation is amended in provinces that review billsusing Committees of the Whole,  whereas provinces using a committee system amend 45%of all bills they pass, a threefold difference. The question of whether this increase isattributable to the practice of committee review is the focus of this article. My findingssuggest that controlling for other factors, committee review does have a real impact on theability of Westminster style parliaments to scrutinize a bill, doubling the rate at which thelegislatures bring successful amendments to legislation. On the other hand, increases inefficiency are not seen in the data.The common view of Westminster legislatures is that they play little role in shapinglegislation. Despite this assumption, Martin and Vanberg argue that ?scholars knowremarkably little about the extent to which legislatures in parliamentary systems matter?(2005, p. 14). Little work exists systematically testing the effect of legislatures by examiningactual legislative outputs (Arter, 2006; Cairney, 2006). Typically, evaluations of thelegislative influence of parliament take a ?veto-players? approach and focus on dividedgovernment in the form of coalitions or minority governments (e.g.: Damgaard & Jensen,2006; Martin & Vanberg, 2005; Newell, 2006). The lack of knowledge about legislativeinfluence is especially acute regarding Westminster committees, which are typically1evaluated not on their ability to influence legislation directly, but instead on their influenceon other policy actors such as the executive, bureaucracy, or interest groups (Hamm, 1983;Monk, 2010).This is due in part to the fact that Westminster committees are particularly weak relative toalternative constitutional arrangements (Lees & Shaw, 1979, pp. 398?404). Single partymajorities use party discipline to tightly control the legislative agenda. Nonetheless, there issubstantial room for variance within the Westminster system as to how much powercommittees are given and what role they play. Specifically, the ability to review legislation isnot a given. In the UK, legislative review was devolved from Committee of the Whole House(COWH) in 1882. However, these ad-hoc standing committees operated by much the samerules as in COWH until 2006 when they were given the ability to call witnesses and hearevidence. In Canada, the Parliament's standing committees1 weren't given the power toreview legislation until 1968, before which time all review was conducted in COWH.Currently, only four of ten Canadian provinces send bills into the committee system in lieu ofCOWH review. Proponents of committee review are seeking to increase efficiency in thelegislature, but they also consistently claim that committee review enhances the power of theprivate members to influence legislation as it passes though the house. Despite these claims,almost no research has examined if this institutional reform is successful in its stated goals.It is clear that in theory the differences between review in COWH and proper committeereview are substantial. It is efficient because having multiple committees enables multiplepieces of legislation to be considered simultaneously. However, it also provides a verydifferent context in which the review takes place. The differences are of three types:membership, resources, and examination powers. Committees consist of different membersfrom the whole house, and are significantly smaller. As such members may devotesignificant time and attention not just to the bill at hand, but to the issues which their1 In the UK permanent committees are referred to as select committees where as standing committees are adhoc. In Canada standing committees are the parliament's permanent committees. For the duration of thispaper committee in general  refer to a sub-set of members of the house with more informal rules of debate,and the power to call witnesses and receive written evidence. Whereas I  use the abbreviation COWH torefer to a Committee of the Whole House.2committee examines in general (i.e. they can specialize). Committees are also given access toresources, mainly research staff, that is not available in COWH. If opposition or governmentmembers need research assistance before questioning a bill during Committee of the Wholethey must draw on their caucus' resources to do so. Finally, committees that consider a billhave the power, and in some cases the obligation, to conduct additional types of examinationbeyond just clause-by-clause analysis. They can call expert witnesses, or conduct publichearings to gauge both the professional and public opinion on the bill. Such a process wouldbe completely foreign to proceedings in COWH. Overall, committees can work as a smallergroup of more knowledgable members, and do so using a greater array of tools andresources. On the other hand, it is not surprising that governments are protective of their legislativeagenda. Debating the changes at the federal level in 1968 Ged Baldwin, the house leader forthe official opposition, outlined this reality with some flair:When  the  government  sees  fit  to  introduce  a  measure  andbrings  it  to  the  stage  of  second  reading,  obviously  it  is  ameasure that has worked its way up from the bowels of the civilservice, has received final approval so far as the government isconcerned and probably has political sex appeal. Once the billcomes  to  second  reading  it  is  a  beloved  child  of  thegovernment. The government, as is the custom of all parents,will  not  tolerate  any  suggestion  that  there  is  any  hint  ofimperfection or deformity in this child (Canada. Parliament. House of Commons, 1968, p. 401)Given this, it is reasonable to ask if enabling committees to review legislation does in factenhance the scrutiny that legislation receives. Further, if this is the case it is unclear whygovernments would empower these committees in the first place.Work that has been done examining committee review has never directly tested its effects ascompared to review by COWH. Thompson (2012) who updated the work of Griffith (1974)on the British House of Commons, examines amendments to bills made by committees.3However her work focuses only on change over time, as opposed to differences betweeninstitutions. Levy (2009, 2010) has an excellent analysis of reforms made in the U.K. in2006. Her analysis concludes that they were effective at increasing legislative scrutiny; butthis conclusion is based on the impressions of committee members given in a series ofinterviews. A number of analyses were also conducted following the 1968 Canadian reformsand came to mixed, though mostly positive, conclusions (Hockin, 1970; Mallory & Smith,1972; Thomas, 1978). These were primarily based on case studies however. This paper exploits variance in the use of committee review across Canada's provinces todirectly test its effect. I use a new dataset of all government bills considered in the legislativesessions of Canadian provinces since 1983. These data are well suited to test the effect ofcommittee review. The high level of similarity across provinces, both in terms of thestructures and rules of the legislatures, as well as social and economic contexts in which theyoperate, allows for a more meaningful comparison than can be made in cross-nationalresearch. In addition to the differences across provinces, Saskatchewan changed their systemduring the period studied, allowing for a comparison within that province over time.The next two sections of the paper outline the context of the reforms in Canada, and thetheory that speaks to them in the political science literature. Two perspectives emerge both inpractice and in theory. Reformers in practice see committee review as a means to bothefficiency and legislative influence. Section 2 below outlines how with both aims in mind,these changes were well supported by both government and opposition members. Section 3shows that the arguments made by both kinds of reformers are well supported in thetheoretical literature, at the expense of other kinds of explanations. Based on these findings, Idevelop a series of hypotheses in subsections 3.2 and 3.3 to test the effects of committeereview. The remainder of the paper describes the data and carries out the tests. I find thatclaims about legislative influence increasing are well supported, but there is no evidence ofan increase in efficiency.42. The reforms in contextLees and Shaw state that, ?the advantages of devolving some of the heavy work of a largeorganization onto committees .. .are well understood but not often articulated? (1979, p. 366).These advantages are consistently cited as twofold:  a more efficient use of legislative timeand greater role for the legislature in shaping bills. These dual reasons have been cited eachtime a Canadian legislature has devolved legislative review to committees.2.1 The arguments for reform in CanadaThe moves to committee review in Ontario, Quebec, and the federal parliament came aboutat similar times in the late 1960s and early 1970s. The roles of government were quicklymultiplying with more legislation, and more complex legislation, being introduced eachsession, the need for a division of labour and a greater degree of specialization arose(Kornberg & Mishler, 1976, pp. 29?33; Lees & Shaw, 1979, pp. 214?216). Federally, theSpecial Committee on Organization and Procedure was struck to examine the issue. Theyjustified the need for greater legislative efficiency in their 3rd report by arguing that membersneeded as much time as possible to be in their constituencies and Ministers to run theirdepartments (Special Committee on Organization and Procedure, 1968, pp. 763?764). InOntario, the change was suggested by the Camp Commission, which undertook a broadreview aimed at modernizing the legislature's practices. In their words, ?a major problemfaced by the Ontario Legislature is how to deal effectively with an increasing volume ofcomplex, provincial business" (Commission on the Legislature, 1973, p. 70) and committeereview was the means to address this. The situation was no different in Quebec, where,Massicotte argues, ?contrary to the impression that prevailed in certain quarters, this reformwas designed less to restore the influence of private members than to free the floor of theChamber, which had been visibly overburdened since the early 1960s" (Massicotte, 1989, p.87). In each case, efficiency was the primary motivation.5If the impetus of committee review was simply to move more legislation through the housefaster, we would expect it to be a controversial measure. In fact, while opposition membersexpressed some wariness on this front, they nonetheless supported the reforms because theysaw them as an important increase in the influence of the legislature. In the House ofCommons, speaking on behalf of the NDP, Stanley Knowles argued that ?this procedureshould also enhance the prestige of members, in short emphasize their role as legislator andnot as civil servant" (Canada. Parliament. House of Commons, 1968, p. 412). Saskatchewanadopted committee review in 2003, and supporting the change in the house, a member of theopposition explained that, ?whenever legislation is presented to the Assembly, basically it?s afait accompli. ... and the public, who are just finding out about any particular piece oflegislation, have very limited opportunities to have any input into this,? however he added,?these recommendations, Mr. Speaker, when implemented,  have that opportunity for thepublic to come forward and have a say during the committee sessions. I believe that?s a majorchange? (Legislative Assembly of Saskatchewan, 2003, p. 357). While a governmentmember claimed the goal of the reform was to ?[provide] a more active role for privatemembers on both sides of the House to engage in policy discussion, to have a more activerole in legislation, in the study of Bills" (Legislative Assembly of Saskatchewan, 2003, p.356). In Ontario and Quebec, the desire for enhanced scrutiny was further emphasized whenadditional resources and powers were extended to committees in each province at the start ofthe 1980s (Cannon, 1990; Dunn, 1990).The use of committee review in Manitoba actually focuses on enhancing scrutiny to the pointthat efficiency is not a motivating goal. It is not clear when committee referral began inManitoba, but Anstett and Thomas claim that it is ?a long-standing practice of the Manitobalegislative process.? It is also clearly an important one given that, ?Manitobans viewcommittee stage as their point of access to the law-making process. There is never a questionof whether there should be public input, just a question of when,? however; regardingefficiency, ?this question is of some relevance because of the absence of a sessionaltimetable? (Anstett & Thomas, 1989, p. 100). In other words, the drag committee review6creates on the sessional calendar is a frequent concern.There are multiple ways in which advocates think committee review will increase legislativeinfluence. Anstett and Thomas explain that one effect of committee review is to expose thegovernment's agenda to the influence of public opinion. They claim that the prospect ofsignificant criticism in a public committee hearing means that, ?the leverage available to thepublic and special interest groups is greater in Manitoba [compared to provinces withoutpublic hearings], if for no other reason than the government's desire to avoid embarrassment?(1989, p. 101). The federal parliament's Special Committee on Organization and Procedureexplained that there is also room for influence through specialization. Their goal was that:[Standing  committees]  would  develop  areas  of  subjectspecialization.  We  would  expect  debate  in  the  StandingCommittees to be well-informed and pertinent; their membersto  become  influential  in  the  areas  of  their  specialisedexperience; and their reports to the House to assume a criticalsignificance related more closely to the national interest  as awhole than to simple political differences.(Special Committee on Organization and Procedure, 1968b, p.434)Note that in neither case is the committee able to change the policy of the government justbecause it disagrees with them. Instead, influence comes from the new methods to bring tobear public opinion or expert analysis.2.2 The motivations for reform in CanadaDespite this increase in legislative influence, the governments enacting these reforms werenever forced to do so. The reform was enacted by a minority government in Saskatchewan,whereas majorities were in power in Ontario, Ottawa, and Quebec. That being said, thereview process federally was first proposed by a minority government, and in Ontario BillDavis' Progressive Conservative majority government was ?urged on by a number of its7frustrated backbenchers? (White, 1980, p. 358). However, even if they felt some pressure, itwas the governments themselves who tended to champion the reform process to begin with.If backbench pressure was the concern, empowering institutions like cabinet committees orcaucus meetings is a much less risky or public way to allow backbench influence (Malloy,2004). Why governments would freely give up such influence is an important puzzle. It may be that the claims of reformers are misguided or disingenuous, and there simply is noincrease in influence. The picture of diligent and specialized committee work is often metwith criticism.  Rush argues that despite the claims of reformers, efficiency gains havealways been the primary motivation (1982). Paquin explains that in Quebec, ?committeesmake little use of their powers and are not as autonomous as expected. More than the lack oftime or resources, blame rests with the control exerted by political parties and the executive"(Paquin, 2012). Nonetheless, whether or not specialization, resources, and examinationpowers are realized to their fullest potential, they are inevitably much greater in committeesystems compared to COWH. As such, the difference between the two ought to bemeaningful, all else being equal.Alternatively, the increase in influence could be the price governments have to pay for anincrease in efficiency. Given the wide range of alternatives open to these governments, this isalso unlikely. For example, the British system until 2006 gained efficiency by runningmultiple committee sessions, but without giving additional powers to the committees, norany degree of specialization. A similar tactic was employed in British Columbia in 2012.Facing a legislative backlog mid-way through the session, the government decided to splitthe chamber into two COWH sessions in parallel, so that, in addition to the Committee ofSupply, 3 chambers could operate at once. As the government house leader explained at thetime, ?I don't believe that two people asking questions of each other, across from each other,need to take up the time in this Legislature. There's other work to be done? (British ColumbiaLegislature, 2012).This approach is obviously distinct from that taken by legislatures thathave adopted committee review.8In practice, the behaviour of reformers fits neither explanation. They have consistentlyproposed reforms citing both goals, and in doing so have successfully drawn support bothfrom  government and opposition parties. Governments proposing these reforms had otheroptions open to them if simply increasing efficiency or placating their own backbench wasthe primary goal.93. Political science theory and committee power3.1 Perspectives in the literatureThe two arguments made by reformers are well reflected in the political science literature.Cox argues that managing plenary time is the most basic challenge faced by legislatures, andcommittees are a natural response to the challenge of increasing demands on plenary time(2008).  However, new institutionalist literature suggests a number of other types of benefitsthat empowered committees bring about. Mattson and Str?m (1995, pp. 250?255) provide asummary of these arguments and find three basic view points. Strong, enduring committeescould serve as enforcement mechanisms to maintain policy coalitions over time (see also:Shepsle & Weingast, 1987; Weingast & Marshall, 1988; Weingast, 1989). They could serveas tools to solve information problems; Krehbiel (1992) shows that if legislatures haveimperfect information to begin with, then empowering committees can create the appropriateincentives for members to invest in specialization. Lastly, they could simply serve asinstruments of partisan coordination (Cox & McCubbins, 1994, 2007). In this viewindividual member's power in committee is dependent on their party. Because of this, theparty can threaten to withdraw that power as an effective tool to maintain discipline.Committees are powerful in order to increase the value of this ?bond? between party andmember. This breakdown is useful to analyze the role committees might play in theWestminster context.These three motivations do not all fit the Westminster system though. Clearly, partydiscipline is strong enough in the absence of committee review that such legislatures do notseverely lack in partisan coordination. Additionally a member's committee specific capital inCox and McCubbin's theory relies on institutions such as seniority and enduring committeesthat are absent in Canadian legislatures.2. In Westminster systems strong parties also act ascoalitional bargaining solutions, the role committees play in the Weingast and Shepsle view.2 In Canada standing committees are re-struck with new members and new mandates after each election10Weingast himself makes this point in discussing the differences in parliamentary systems(1989, p. 700). Even in situations where parties are not perfect policy coalitions, institutionsother than parliamentary committees exist to correct for that; typically cabinet committeesand caucus meetings (Malloy, 2004). Information problems, on the other hand, are unlikelyto be addressed by party discipline. This provides a more plausible reason that parliamentarygovernments might devolve some genuine power to committees. The information perspective (Gilligan & Krehbiel, 1990; Krehbiel, 1992) starts from theview that legislators want to make good policy to increase their chances of getting re-elected,but have imperfect policy information. To incentivize some members to invest inspecialization, the legislature gives sufficient power to committees that members can haveinfluence in return for their role as specialists. One of Krehbiel's most important conclusionsis that this ?informational committee power? creates  positive-sum outcomes, specifically, ?ininstances of informational committee power, a committee credibly transmits privateinformation to get a majority to do what is in the majority's interest? (1992, p. 77).In Canada, if committees passed amendments to legislation because they disagreed with thepolicy, this would require the government members on the committee to register theirdisagreement publicly. It may occasionally be the case that the government backbenchers ona committee disagree sufficiently with the government to dissent. If the amendment madeonly a marginal change to the legislation this would be dissent with low visibility and neednot significantly harm the government's public image. Research shows that not all dissent iscreated equal in this regard (Kam, 2009, pp. 130?160). But these circumstances would berare. The strength of party discipline in the Canadian case is simply too high to reasonablysuspect that any ambitious government committee member would routinely go against theparty position in the exercise of committee work. Additionally, any change made by thecommittee can be overturned at report stage if the government so desires.However, if an information perspective is adopted, the willingness of a government todelegate powers to a parliamentary committee makes more sense. The work of the committee11is more likely to improve the bill in ways that are beneficial to the government either becauseit is objectively better policy, or because it is more in line with public opinion. However,there are a number of key differences between the committee-legislature relationship inKrehbiel's analysis and the situation in Canada. Not only does the existence of a strong partysystem with high levels of party discipline alter the incentives, but committees in Canada aremuch less powerful overall, and most importantly do not have the ability to proposelegislation, only to alter government proposals.3 Nonetheless, the fundamental points that canbe translated from Krehbiel's arguments are a) where informational asymmetries exist,opportunities are available for positive-sum gains from specialization; and b) committees arewell-situated to act as policy specialists. The information perspective helps explain whysome legislatures would adopt this approach and not others. If the committees' enhancedscrutiny is about bringing to bear better information, then the efficiency and scrutiny effectsof committee review address the same problem: an increasingly complex policyenvironment.3.2 Informational committee power in Westminster parliamentsAssuming that  the government does not have perfect information prior to introducing a bill,there is typically room for improvement to the government's benefit. They could lackinformation about what the optimally efficient policy is (as Krehbiel assumes) or about thepublic popularity of the policy. With only COWH review, investing in specialization, or evenscrutiny of a single bill, is costly for members. Not only do they have to rely on their ownresearch or use up caucus resources, but the time allocated on the floor is likely to be limitedrelative to what would be enjoyed in a committee system. Committee review, by providingresources and powers to committee members, subsidizes the cost of specialization andscrutiny. Any influence exchanged by the government further incentivizes investment inspecialization by members. If committees have influence in Westminster parliaments, it islikely because they allow members to act as information specialists, helping the legislature as3 In fact, because after second reading the legislation has been approved ?in principle? the committee is barredfrom making any amendments that would fundamentally change the principle of the bill12a whole create positive-sum benefits from legislative scrutiny. Scholars that have studied these reforms previously have also observed information effects.Hockin (1970) and Mallory and Smith (1972) both argue that committee review made animpact in the Canadian House of Commons in cases of constructive work across party linesto improve the quality of legislation. Levy also reports that information effects are the modethrough which reforms in the U.K. have had an impact (2009, 2010). This logic was wellexpressed by an opposition member in the Canadian House of Commons, who argued thateven though ?there are innumerable cases of statutes coming out of this parliament and out ofprovincial legislatures which are half-baked, ill-conceived, ill-prepared and which do notattain the object for which they are designed? committee review would result in ,?anopportunity to put in a more concrete and eloquent way, a way which  receive some attentionfrom the government, proposals which  better the legislation enacted by this parliament"(Canada. Parliament. House of Commons, 1968, p. 402).In order to study the impact of committee review, measures are needed for each effect.Looking first at legislative influence, I measure whether or not the legislature amends a billafter first reading. Not all amendments are highly substantive. However; if we are comparingthe rate of amendments across jurisdictions there is little reason to worry that the need forhousekeeping or technical amendments would be systematically different given thecommittee system. What is crucial, is to show that any observed committee effect comesfrom informational power and not because committees alter policy to suit their preferences.If the latter is true, a problem emerges for identifying the effect of committee review bylooking at rates of amendment. This is because to the extent that the government has anyknowledge about the policy preference of the committee they can exercise their agendapower to anticipate and offset this effect. This strategic interaction would alter the rate ofamendments in ways that mask any true effect of committee review.To see this, consider a scenario in which the government proposes a different policy than thatpreferred by a majority on the committee. Assuming the committee has some influence over13the policy, then to the extent that the government knows a) the majority preference of thecommittee and b) the willingness of the committee to act on this preference (i.e. given thecost of dissent to government members), they can use this information when deciding whatto propose. Suppose committees amend proposed legislation in the direction of theirpreferences, but the extent of such amendments is limited by the costs of dissenting.Knowing this, the government could simply propose a more extreme policy than they prefer,such that the committee's amendments would bring it in line with their true preference. Inthis case amendments would be observed but they would not represent any genuine power ofthe legislature to change government policy. Alternatively, say there is a cost to thegovernment's public image when backbenchers amend its bills against its wishes. In thiscase, they might anticipate the compromise in advance and arrange it with the committeemembers without the display of proposal and amendment occurring in the public forum ofthe legislature. In this case the committee system would in fact impact the policy, but such aneffect would not be measured. In either case, observed amendment rates would not reflect thetrue impact of committee review.On the other hand, if committees do not generate policy contention, but instead act as sourcesof information, the observed legislative impact should reflect the true effect of committeereview. By changing the information available to government, they actually change thepolicy preferences of the government. This effect, by definition, could not be anticipated inthe government proposal. This is true whether they alter the politically preferable policy byenhancing public input or media scrutiny or by altering the assessment of which policy ismost optimal through expert witness testimony or some level of committee expertise itself.Therefore, if the role played by committees is an informational one, any impact of committeereview is directly observable.For this reason, some hypotheses are required that distinguish between information effects,and effects arising from policy disputes. Recall from above that the claim that committeeswield informational power in the Westminster context derives from the claim that thecommittee system subsidizes scrutiny of bills, and incentivizes further specialization by14members. Nonetheless, members in systems of COWH review can choose to invest someamount of resources into enhanced scrutiny where they think it will make a difference. If weidentify types of legislation that members facing COWH review are likely to invest in, wecan hypothesize that these differences in scrutiny would be attenuated in the provincespracticing committee review. This is because as scrutiny becomes less expensive, the rangeof bills facing a high level of scrutiny should increase. This hypothesis can be tested in two ways. First with regard to whether or not legislation ispolitically contested. If legislation is controversial it is more likely to be visible to the publicand in the media, and members are more likely to invest costly time and informationresources into scrutinizing it. These bills should then receive more amendments in COWHreview provinces, compared to uncontroversial bills, but such a difference should be muchsmaller in committee review provinces. Second, I would contend that new statutes are morelikely to attract additional scrutiny than amending bills. Though amending bills may containimportant policy changes, they often address minor issues or housekeeping matters. Newstatutes are, in general, more likely to reflect new and important policy issues. Again, thiswould mean higher amendment rates among new statutes compared to amending ones underCOWH review, with these differences attenuated under committee review.If informational power causes governments to agree to amendments, it should also causethem to back off of legislation after committee stage more frequently. This is especially trueinsofar as committee review increases the government's exposure to public opinion on theissue. Because third reading is little more than a formality in most cases, a bill that haspassed through committee stage is very unlikely to not pass third reading for time reasons.Rather, this will almost always indicate that the government has decided to back off of, ordelay to a future session, the passage of the bill.  On the other hand if committees seek toamend legislation to suit their policy preferences, governments are free to deny the changesat report stage and pass the unaltered legislation. Thus, if governments in provinces that usecommittee review are more likely to let legislation lapse after it has passed throughcommittee stage, this would suggest that the effect of committees is information based.15To sum up, if committees pass amendments because they generate new information about thequality of the legislative proposal, then the observed effect of committee review onamendment rates would be the true effect, ceteris paribus. The discussion in this sectiongenerate four testable hypotheses. The first addresses the fact of committee power, and theremaining three test that this fact results from an information effect:1.  Amendments  hypothesis:  Ceteris  paribus,  legislation  isamended  at  a  higher  frequency  in  provinces  that  practicecommittee review of legislation2a.  Contentious  legislation  hypothesis:  if  amendments  aredriven by the informational role of committees, then committeereview  attenuates  differences  in  amendment  rates  betweencontentious and non-contentious legislation compared to reviewin committee of the whole2b. New statutes hypothesis: if amendments are driven by theinformational  role  of  committees,  then  committee  reviewattenuates differences in amendment rates between new statutesand amending legislation compared to review in committee ofthe whole3. Retraction hypothesis: if committee power is informational,then provinces that practice committee review  more frequentlyabandon legislation after committee stage.If the latter three hypotheses are supported by the evidence, then modelling the firsthypothesis produces informative results. If they are not, the models are more suspect.3.3 Committees and efficiency in Westminster parliamentsThe second effect to measure is the impact of committee review on efficiency. The claim thatcommittees are a tool to increase efficiency is in line with Cox, who argues that committeesare created in the first place to address bottlenecks in plenary time as legislatures specialize(Cox, 2008). However, committees are not the only means to this end. Regarding the need to16pass a greater volume of legislation, governments have alternative options. Moststraightforwardly, they can sit for more days. Typically, of course, the introduction ofcommittees seeks in part to avoid this option. Another way to decrease the load on memberstime is to increase the number of members themselves. This allows the house to sit for longerhours or more days but with less total stress on the time of each individual member. As in thecase of the British House of Commons before 2006, or British Columbia in 2012, legislaturescan split into additional sections without giving these sections new powers of scrutiny. Eachof these changes may negate the need for a move to committee review. Most importantlythough, if stresses on plenary time are high in any one session a government can use theinstitutions of closure and time allocation.4 Because the typical situation is one of majoritygovernment, this tool consistently means ?prime ministers seem to be able to get thelegislative results they want" (Malloy, 2004, p. 209). The ability to force legislation throughthe house in the event that time runs low means governments are able to pass as muchlegislation as they feel is necessary. Any of these institutions could act, in part, as substitutesfor committee review.In addition, committee review in general may lead to increased lengths of time spent incommittee per bill, and may make it more difficult to expedite the committee stage oflegislation when the government desires to do so. In this way committee review may alsohinder efficiency by throwing up new roadblocks in the way of bills passing through thehouse. What this means is that any increase in efficiency from committee review need not bereflected in the volume of legislation passed. The observed effect may be positive, it may benegative, or it may be nil. In none of these cases would this necessarily reflect the true effect.A more direct test of the efficiency gains from committee review is whether it reduces thereliance on institutions like time allocation to push a government's agenda through the house.As one federal MP pointed out (as an argument in support of committee review), during the4 The main difference between closure and time allocation is that the latter requires unanimous consent,whereas closure can be imposed by a majority. Nonetheless time allocation is much more common. This islikely because the threat of closure is a strong incentive for opposition parties to extract what littleconcessions they can in agreeing to a schedule for time allocation. In either case, the effect is that a greatvolume of legislation can be passed in a shortened period of time.17period of 1966-67, ?the house sat for 250 days and passed, I believe, 83 bills. A survey ofthem can be a very interesting exercise. 78 of these measures were each passed in one day. Ittook one day each for these measures to pass all stages ... five bills of that session took 80days to pass" (Canada. Parliament. House of Commons, 1968, p. 402). If the time it takes topass bills is constant throughout the session, this is indicative of an orderly and well managedlegislative agenda. If, on the other hand, a large percentage of bills are passed right at the endof a session, or just before sessional breaks, this reflects a reliance on expediting the agendaas time runs out. In other words, committee review can be seen as efficient not through directeffects on the volume of legislation, but rather to the extent that lessens the need to rushlegislation through at the end of a sitting.Thus, to examine efficiency, I test the following hypothesis:4.  Concentration  of  passage  hypothesis: The  timing  ofpassage of bills in legislatures that practice committee reviewpass is less concentrated than the timing in legislatures that useCOWH to review legislation 184. DataThe hypotheses from section 3 are tested using an original dataset on legislation in Canada'sprovinces from 1983-2013. Data was collected at the bill level on the progress of legislationin each session. It was also collected at the session level to measure a number of importantcovariates.4.1 Control variablesDespite the similarity across Canada's provinces, a number of control variables are stillimportant to model the impact of committee review. Three groups of control variables areconsidered below. Of primary theoretical concern is legislative professionalization. Literature on U.S. StateHouse's concludes that the degree to which a legislature is professionalized significantlyeffects its partisan make-up (Fiorina, 1994, 1999; Meinke & Hasecke, 2003), incumbentturnover (Berry, Berkman, & Schneiderman, 2000), and the efficiency and level of scrutinyin the legislative process (Squire, 2007). Professionalism is hypothesized to impact thequality of legislators who make-up the legislature, the investment they make in legislatingwell, and the tools available to them. Moncrief shows that legislatures in Canadian provincesvary significantly with regards to professionalization (1994). Professionalism in thisliterature is defined by the compensation available to members, the degree to which alegislator's job is a full-time commitment, and the resources available to members (Squire,2007). The size of the legislature is also thought to be an important factor (Fiorino &Ricciuti, 2007; Pettersson-Lidbom, 2012; Stigler, 1976).19The move to committee review is itself one part of professionalizing these legislatures.Typically it was done in conjunction with other moves meant to modernize the practices ofthe institutions. If differences in amendments are merely a function of a more professionaland engaged legislature as opposed to the specific result of committee review, results mightbe spurious. As such, measures of professionalization are absolutely essential to test theeffect of committee review on both efficiency and scrutiny.The second major concern is the political context in which bills are considered. The first partof political context I consider is the stage in the electoral cycle. This means both how manysessions have passed since the last election and whether the current session is the last beforean election takes place. Research demonstrates that electoral timing affects government'sincentives regarding fiscal policy (Alesina, Cohen, & Roubini, 1993; Blais & Nadeau, 1992)including in Canadian provinces in particular (Petry, Imbeau, Cr?te, & Clavet, 1999). Assuch, the models must allow for similar behaviour regarding parliamentary strategy and thewillingness to allow amendments to pass. Another determinant of parliamentary strategy isthe status of the government (minority or majority) and if they have a majority, then its size.The control of a government over the legislative agenda is obviously a direct consequence ofits control of the legislature. Any test of committee review's effects on legislative efficiencyand scrutiny must account for this.Last, measures of population and GDP are necessary to ensure that effects correlated with thecommittee variable are not merely reflecting differences in the size of the economy orpopulation of the provinces. As populations and economies grow this is a direct driver of theneed for more government policy, and a greater level of policy complexity facing thelegislature. This is especially important because population size and economic capacity aretwo areas where there are substantial differences across Canadian provinces. Thus, given thatmuch of the leverage in this data relies on cross-sectional comparisons, properly controllingfor these differences is paramount.204.2 Sources of dataData on the progress of bills in provincial legislatures is available as far back as 1983 forevery province and territory. For most provinces the data are available from the provinciallegislature's website from between 10 and 20 years back.5 During periods not covered by theonline data, data was collected from Provincial Pulse (CCH Canadian Limited, 1982) whichwas published between 1982 and 1986 and the Provincial Legislative Record (CCH CanadianLimited, 1987) published between 1987 and 2009. The total dataset comprises 17,605 billsspanning 276 legislative sessions in 10 provinces. Data for each bill covers the name of thebill, whether it reached second reading, if it was amended, whether the bill was passed, thedate it was passed, and if it was not passed whether it was carried over to the next session (aprocedure used rarely, and only in some legislatures). The data include only governmentbills. To control for professionalization I collect data at the session level on sitting days, legislaturesize, and the salaries of legislators. The data on salaries is not available for every yearunfortunately. Moncrief reports data for 1980 and 1990 (Moncrief, 1994, p. 37), Fleming'sCanadian Legislatures reports data for 1997 (Fleming & Glenn, 1997), and the Albertalegislature collected data for the year 2010 (Legislative Assembly of Alberta, 2010). Thisgives information on salary at four points in time throughout the study period. For otheryears, each session was coded at the salary level of the year with data closest in time to theyear in which the session began. Data on sitting days are based on Hansard records from eachprovince. To generate the other control variables, I also collect data at the session level on thepartisan make-up of the legislature, and the population and GDP of the province at the timeof each session. Population and GDP data come from Statistics Canada's CANSIM database5 The URLs of the provincial legislature web pages that contain the data are as follows:;;;;;;;;; All data were collected May and July 2013.21(Statistics Canada, 2011, 2012).In order to test hypotheses 2a and 2b I require data on the contentiousness of bills and theirtype. For the latter, coding is based on the name of the legislation. Amending statutes usetitles along the lines of ?An Act to Amend Statute X?; ?Statute X Amendment Act?; or?Statute X (amendment).? These are coded as such. As mentioned above, supply bills mustalso be coded separately, this coding is based on the keywords ?appropriation(s) act?;?supply?; and ?loan act?. The number of bills is small enough that error checking could bedone to correct for any non-supply legislation erroneously captured by this method.6 Finallyit is important to differentiate omnibus legislation which is a distinct category. Coding ofomnibus bill titles is based on the descriptions given by Massicotte (2013). These includereferences to ?various statutes? or ?various measures?; and the specific titles: ?MiscellaneousStatutes Amendment Act?; ?Budget Measures Implementation Act?; and ?Statute LawAmendment Act.? Though coding based on the name does not allow categories as fine-grained as would be ideal, it still tells us about important differences between bills.The measure of contentiousness is whether a bill faced a recorded division at second or thirdreading. Most motions in Westminster parliaments pass or fail on voice votes, but when agroup of members wants to ensure that the vote is formally counted and recorded they havethe right to call for a division. Divisions are used in cases where there is either genuineuncertainty that a measure will pass, or when one party caucus wants the record to reflecttheir opposition to a motion proposed by a different party. In either scenario, divisionsindicate a heightened level of political contestation surrounding the vote. Unfortunately,information on divisions is not included in most of the data sources used for this study.However it is included for the online portion of the data from Alberta, Ontario, and Quebec.In addition, British Columbia's Hansard Indexes include excellent summaries of recordeddivisions for all but two sessions in the data set (Legislative Assembly of British Columbia,2013). As such data on divisions are available for 64 sessions across 4 provinces. This6 For example ?An Act to secure the supply of hogs to a slaughterhouse enterprise in the Abitibi-T?miscamingue region? from the 2nd session of Quebec's 36th National Assembly was caught and corrected.22includes two that practice committee review (Ontario and Quebec) and two that do not(Alberta and BC). Though data from all provinces would be preferable, the comparison is auseful one as Alberta and BC are the 3rd and 4th largest provinces after Ontario and Quebec interms of both population and GDP.Testing the concentration of passage hypothesis requires a measure of concentration. Iconsider two. A simple measure is the proportion of legislation passed in the final 5 days7before a session is prorogued. This measure is helpful because it measures not justconcentration, but concentration at the end of the session specifically. It is alsostraightforward to calculate and interpret. However concentration can occur not only at theend of the session, but leading up to mid-session breaks. A measure of overall concentrationis necessary to capture this effect. A widely used measure of concentration is an inverseSimpson index. This measure is a technique well known in political science from its use tocalculate the effective number of political parties (Laakso & Taagepera, 1979). Equivalently,I calculate the effective number of third reading days in a session.8 Unlike the first measure,this does not tell us if the passage is at the end of the session, but it incorporates informationabout every bill into its measurement without the need for to choose a cut-off point. Bothmeasures give important information about the concentration of passage.4.3 Descriptive statisticsTable 1 shows descriptive statistics for the dependent variables across all 10 provinces. Thedata for Saskatchewan is reported both before and after 2003, which is when that provinceadopted committee review. The provinces that practice committee review are bolded in thetable. 7 As with any similar measure, the choice of cut off is arbitrary. In this case, 5 days is meant to represent thelast week's worth of sitting days. Testing with other cut-offs obtained the same results8 An inverse Simpson index is the inverse of the sum of the squared proportion of observations described byeach category. In this case, each bill is an observation, and each day of a session is a category. So themeasurement is calculated for every session as the inverse of sum of the squared proportion of bills passedon each day of that session. This calculation is also identical to the inverse of the Herfindahl index widelyused in economics literature.23Table 1: Average amendment rates and efficiency by province, 1983-2012Amendments EfficiencyProvinceAmendmentRateBills passed persitting day% Passed last 5daysEffectivenumber of 3rdreading daysAlberta 23.3% 0.73 27.7% 7.33British Columbia 25.4% 0.62 16.1% 11.21New Brunswick 11.9% 1.31 27.4% 6.72Newfoundland 11.4% 0.79 37.6% 5.25Nova Scotia 11.7% 0.88 28.8% 7.41Prince Edward Island 8.0% 1.11 34.9% 3.87Saskatchewan pre-03 10.7% 0.92 28.8% 8.38Saskatchewan post-03 17.9% 0.76 14.9% 8.76Ontario 59.7% 0.31 8.5% 18.17Quebec 62.6% 0.88 10.2% 18.84Manitoba 33.2% 0.57 59.8% 2.84OVERALL 23.2% 0.82 28.4% 8.39Provinces in bold are those that practice committee review; data are pooled at the session level beforeaveragingLooking at amendment rates the difference is clear. With the exception of Saskatchewan, theprovinces with the highest amending rates are those that use committee review. Importantly,we see that although Saskatchewan under committee review does not amend legislation asfrequently as other provinces, there is a large increase before and after the adoption ofcommittee review. The table also demonstrates that Ontario and Quebec amend significantlymore legislation than any other province. Because they are also the largest and mostpopulous provinces this makes it especially important to ensure that any results are notdriven by the uniqueness of Ontario and Quebec relative to other provinces.Turning to the measures of efficiency, the effect of committee review is less clear. In additionto the measures of concentration, a measure of the volume of legislation passed (bills persitting day) is included. Quebec has a reasonably high rate of bills passed per sitting day, but24it is still lower than the rate of passage in New Brunswick and PEI. Ontario, on the otherhand, has the lowest rate of bills passing per sitting day of any province, and the secondlowest is in Manitoba. This is not surprising. As discussed above, it does not mean thatcommittee review is ineffective at moving more legislation through the chamber compared toa hypothetical status quo, but it confirms that provinces have more than one way ofincreasing the efficiency of their legislative process. The measures of concentration are a better test of the efficiency effect of committee review.In this case the results appear more positive. Ontario and Quebec pass much fewer legislationin the last 5 days of their sessions than other provinces and have much lower levels ofconcentration of third readings than other provinces.9 Though Saskatchewan did not see amarked decline in the concentration of passage after changing to committee review, they arepassing a significantly lower portion of their legislation at the end of their sessions. Manitobais the clear exception though. As mentioned above, Anstett and Thomas suggest that theprocess of public hearings in Manitoba is a drag on the efficiency of the process. This resultis clear here as Manitoba passes a very high 59.8% of legislation at the end of the session,and has the highest concentration of third reading dates, effectively passing all their bills onjust a few days each session.Looking at provinces that do not practice committee review, Alberta and British Columbiastand out as having higher rates of amendment despite the lack of committee review. Onepossibility is that being the third and fourth most populous provinces, each with a large andimportant economy, they likely have more professionalized legislatures. This reaffirms theneed to control for professionalization. Table 2 summarizes the measures ofprofessionalization. The measures show clear differences in professionalization across provinces. After Ontarioand Quebec, B.C. and Alberta have the most seats in their legislatures. and B.C. legislators  9 Because this measure is the ?effective number of third reading days? higher numbers indicate lowerconcentration.25Table 2: Measures of professionalization in Canada's provinces, 1983-2013Province Sitting DaysAverageSeatsChange inSeats 1983-20132010 Salary (CAD 2002)Change inSalary 1980-2010Alberta 55.0 83 +4 $67,961 $19,210B.C 60.8 74.5 +25 $88,593 $33,340Manitoba 69.4 57 0 $74,420 $32,290NewBrunswick51.3 56 -3 $73,930 $25,530Newfoundland 60.1 50 -4 $82,938 $39,196Nova Scotia 67.0 52 0 $75,338 $24,229Ontario 145.4 118 -18 $101,371 $26,549P.E.I. 37.1 29 -5 $56,834 $15,394Quebec 131.3 124 +3 $87,299 -$1,880Saskatchewan 74.3 61 -6 $76,597 $20,810Sitting Days and Average Seats are the averages by session for the whole time period. Salaries are in 2002CAD and include both member's base salary and their indemnities; sources for the salary data are discussedabove. receive the third highest compensation after Ontario and Quebec. Alberta actually pays abelow average salary to its members. With the exception of Ontario and Quebec there is notsubstantial variance in the average length of sessions from province to province. Mostlegislatures are in session for about 60 sitting days at a time. Ontario and Quebec on the otherhand tend to hold sessions across multiple years, with only a few sessions between eachelection. The third and fifth columns show changes over time. Here we see that while mostprovinces have held the size of the legislatures relatively constant over time, BritishColumbia's has grown substantially, while Ontario's has shrunk slightly. In terms of salary,most legislatures have seen substantial increases over time, with the exception of Quebecwhich has had a consistently high salary for members since the beginning of the studyperiod.264.4 Informational versus policy effectsThe next section will test these differences in full regression models, but first it is importantto test hypotheses 2 and 3. These hypotheses predict particular differences that committeereview would produce if its effect was an informational one. If these hypotheses are notconfirmed then it is problematic to claim differences in amendment rate are reflective of thetrue effect of committee review.Hypotheses 2a and 2b stem from the assumption that committee review provides aninexpensive way to apply scrutiny to the a bill. If this is the case, then that factors whichaffect the investment in scrutinizing bills made by members in COWH provinces would beattenuated under committee review. Table 3 examines this for the case of politicallycontentious bills. Table 3: Amendment rates by recorded division (4 provinces), 1983-2013COWH ReviewDivision Chi-Square TestAmended No Yes ?2 p-valueNo 1087 141 28.91 0.000Yes 346 97% 24.1% 40.8%CommitteeReviewDivision Chi-Square TestAmended No Yes ?2 p-valueNo 292 53 2.13 0.145Yes 853 198% 74.5% 78.9%The table shows a cross-tabulation of bills passed that were amended and those that faced arecorded division at second or third reading for each type of committee system. In the27provinces without committee review (Alberta and B.C.), bills that face division are almosttwice as likely to be amended. In contrast, this difference disappears in Ontario and Quebec.Bills in these provinces are amended frequently but there is no difference in this frequencybetween bills facing division and those not. While legislators may invest in scrutiny inAlberta and B.C. to influence contentious legislation, in Ontario and Quebec, committeereview reduces the need to make such a choice and all bills receive higher levels of scrutiny.The chi-squared tests bear this out. While the distribution of amendments is significantlycorrelated with division in the case of COWH review, they are unrelated in the committeereview provinces. Table 4 repeats this analysis for bill type instead of division. This table includes omnibuslegislation as a separate category. This is done for two reasons. First, omnibus legislationtypically consists of amending legislation, but may also include new statutes. Second, whilesome omnibus bills are likely to be of particular interest, others are quite routine. As such,including them in one or the other basic category may dilute the results. Legislation that Table 4: Amendment rates by bill importance (all provinces), 1983-2013COWH ReviewBill type Chi-Square TestAmendedAmendingStatuteNew Statute Omnibus ?2 p-valueNo 5149 2394 111 99.09 0.000Yes 784 584 52% 13.21% 19.61% 32.30%CommitteeReviewBill type Chi-Square TestAmendedAmendingStatuteNew Statute Omnibus ?2 p-valueNo 1585 548 96 163.41 0.000Yes 1156 788 239% 42.17% 58.98% 71.34%28repeals existing acts is excluded from these counts.The table shows similar results to those in Table 3. In provinces without committee review,new statutes are amended about 1.5 times more often then amending legislation and omnibusbills are amended almost 2.5 times more often. In these provinces, new statutes and omnibusbills both draw significantly more scrutiny than amending legislation. There are stilldifferences in the committee review provinces, but they are somewhat smaller. New statutesare amended about 1.4 times more frequently then amending legislation, and omnibus billsare amended only about 1.7 times more frequently. While new statutes and omnibuslegislation are amended more frequently than amending statutes in all provinces, thedifferences are smaller under committee review than under COWH review. However, forthese data the relationship between bill type and amendment is statistically significant in thechi-squared tests for both types of provinces.An informational perspective on committee power also implies governments  abandonlegislation as a result of committee scrutiny. This requires looking at the set of legislation thatpassed committee stage. Unfortunately, these data are only available for the online portion ofthe sample, as the data from CCH does not describe passage of committee stage inparticular.10 Table 5 summarizes how frequently governments neglect to pass legislation afterit has completed committee stage.These events are exceedingly rare. This is not surprising given the intense controlgovernments exercise over the legislative agenda in Westminster legislatures, and the factthat for most legislation third reading is little more than a formality. This is a strong reminderthat whatever increase in power committee review gives to Westminster parliamentarycommittees, they are still weak actors relative to the government itself. Nonetheless the table 10 This sub-sample also excludes Prince Edward Island for which there is no data. In all it represents 109 of the274 sessions in the full dataset.29Table 5: Unpassed bills by Province, various time periodsProvinceNumber of?retracted? billsTotal bills to passcommittee stageAlberta 1 545British Columbia 0 447New Brunswick 0 503Newfoundland 1 514Nova Scotia 0 868Saskatchewan pre-03 0 57Saskatchewan post-03 5 538Ontario 0 311Quebec 35 1120Manitoba 13 368The data in this table covers the most recent sessions going back as far as datais available. In total it covers 102 sessions across 9 provinces.shows that while these events are all but unheard of under COWH review, when they occur itis in provinces with committee review.  A brief example shows on the one hand that this effect can be driven by committee process.But, on the other hand, the results do not guarantee that the government will not ultimatelyget its way. Bill 80 was introduced in Saskatchewan in 2009. It was a controversial piece oflabour law, changing union regulations in the construction industry. The bill met significantopposition through committee stage (?Meeting on Bill 80 heated,? 2009), and was ultimatelynot passed in the legislative session in which it was introduced. However, the governmentreinstated the bill in the following session, and pushed it through without amendment (Hall,2010). In this case committee review had no substantive effect on the content of the bill thatwas ultimately passed. However, an editorialist for Saskatoon's Star-Phoenix newspaperargued that, after the process it went through, passing the unamended law did seriouspolitical damage to the government, while minor amendments would have largely placatedthe opposition and avoided the damage (Mandryk, 2010). Likely, other governments in30similar circumstances would allow such amendments. In general, the fact that nearly all suchevents occur in provinces with committee review provides strong evidence for hypothesis 3.315. ResultsProperly testing the effects of committee review requires multivariate analysis. The primaryconcern is controlling for differences in professionalization across legislatures. As such,limited models are fitted in each case controlling for only this set of covariates. Full modelsalso include the controls for political context as well as population and gdp. The results ofthis section show that committee review predicts a substantial increase on legislativescrutiny, but has no measurable effect on the concentration of bill passage.5.1 Modelling amended legislationThe first set of models addresses legislative scrutiny. Because the dependent variable is acount of the number of bills amended, ordinary least squares regression is not appropriate.The first choice for count data is a poisson regression model. However testing showed thatthe amendments variable was significantly overdispersed. Because a key assumption of apoisson model is that the mean and variance are equal (conditional on the covariates), amodel that relaxes this assumption must be used if the data is overdispersed. The mostcommon choice is the negative binomial regression model (Long, 1997, p. 230). Thenegative binomial model adds a variance term to the estimated equation that accounts foroverdispersion in the data. Model 1 in Table 6 shows a negative binomial regression of the number of amended bills onthe use committee review, controlling for professionalization. Because not all sessions passedthe same number of bills overall, an exposure variable must be used to scale the model. Inthis case the exposure is the number of bills passed. Including the exposure variable causesthe count model to effectively model the rate of amended bills passed out of all bills passed(Winkelmann, 2008, p. 74). This model tests the hypothesis that the effect of committeereview on amendments remains even when controlling for other aspects of32professionalization. Coefficients in this model, and all count models hereafter, are reported asincident rate ratios.11 In this model, and all models hereafter hypotheses are tested usingrobust standard errors that are (where appropriate) clustered at the province level.Table 6: Regressions, amended bills in Canada's Provinces, 1983-2013Full Sample Restricted Sample SK OnlyModel 1 Model 2 Model 3 Model 4 Model 5CommitteeReview1.834*** 2.072*** 2.117*** 2.114*** 6.729***Seats 1.015*** 1.002 1.024*** 1.007 ---Compensation 0.999 0.999 0.999*** 0.999 0.999***Sitting Days 1.002** 1.002*** 1.002 1.004** 0.999Pre-election --- 0.851* --- 0.775** 0.663Session --- 1.069* --- 1.093* 1.193*Minority --- 0.637 --- 0.473 1.478Majority Size --- 1.001 --- 1.001 0.998Population (ln)1 --- 1.471** --- 1.241 1.365***GDP pc --- 1.000 --- 1.000 0.999Exposure Bills passed Bills passed Bills passed Bills passed Bills passedN 269 269 234 234 28? Parameter 0.33 0.27 0.36 0.31 0.055? = 0  p-value 0.000 0.000 0.000 0.000 0.000*: p < 0.1; **: p < 0.05; ***: p < 0.01; Coefficients are reported as Incident Rate Ratios (see note), standarderrors are robust and (except in model 5) clustered by province1: For the sake of comparability across models, coefficients for for logged population variable are displayed asthe IRR of a one standard deviation change instead of a unit change, these standard deviations, in order, are1.206, 1.0007, and 0.015. Which correspond to population changes, relative to the mean, of 3,069,636;1,678,822 and 15,446 in each sample respectivelyIn this dataset, the average session in a province with COWH review passes about 50 bills,11 An incident rate ratio (IRR) is the ratio of the predicted rates of the dependent variable before and after aone unit change in the independent variable. In other words, if the outcome occurs at twice the rate after aone unit change in some independent variable, the IRR of that variable is 2. If it is half as likely, the IRR is0.5.33and an average of 7 of these bills are amended. Model 1 predicts that implementingcommittee review doubles this, an increase of 7 additional amended bills. However thenumber of bills that doubling the rate of amendment represents depends on the baseline. Inthe average P.E.I. session it means an increase from 3 out of 30 bills to 6, whereas in B.C. itwould predict that of a similar 30 bills passed, 20 would be amended instead of 10.Model 2 adds the additional control variables. A likelihood ratio test confirms that this fullmodel is a better fit of the dependent variable than the limited model. With the additionalcontrol variables in the model, the effect of committee review remains the same: twice thenumber of bills are amended when committee review is used. This is clearly a substantivelylarge effect compared to the other variables in the model. For example, a 50% increase inpopulation (e.g. from 1 million to 1.5 million) predicts only 1 more bill amended on top ofthe 7 in the hypothetical average COWH session. Another way of interpreting the effects ofthe model is to examine the probability of specific counts. Table 7 breaks down the predictedprobability of various numbers of amended bills. It separates the predictions out by thenumber of bills passed and the use of committee review, other variables are held at theirconditional means for each cell.These probabilities reaffirm the importance of the role committee review plays. Even in asession that passes between 75 and 100 pieces of legislation, there is a 45% chance that aprovince without committee review will amend less than 10 of those bills. The chance of thishappening in a province with committee review is only 4%. In fact, no matter how muchlegislation is passed, less than 10 amended bills is the modal category for provinces lackingcommittee review. On the other hand, in committee review provinces there is a significantchance of a large portion of the bills passed being amended. This is true no matter how manybills are passed.Looking briefly at the other independent variables, although the professionalization variablesshowed some effects in the limited model, they are mostly insignificant in Model 2. Sitting 34Table 7: Predicted probabilities from Model 2 Probability of amending N bills:Bills passed 0-10 10-20 20-30 30-40 40-50COWH 0-25 98.1% 2.7% 0.1% n/a n/aComm. Rev. 0-25 78.1% 20.1% 4.9% n/a n/a0-10 10-20 20-30 30-40 40-50COWH 25-50 86.0% 16.0% 1.7% 1.7% 0.0%Comm. Rev. 25-50 41.4% 40.0% 16.6% 6.3% 2.5%0-10 10-20 20-30 30-40 40-50COWH 50-75 69% 30.5% 5.6% 1.0% 0.2%Comm. Rev. 50-75 16.9% 31.7% 23.5% 14.3% 8.7%0-10 10-20 20-30 30-40 40-50COWH 75-100 44.8% 39.0% 15.8% 5.7% 0.2%Comm. Rev. 75-100 3.8% 13.1% 17.1% 16.3% 13.7%days is statistically significant, but substantively the predicted effect is very small. Forexample, in Ontario and Quebec, which hold significantly longer sessions than otherprovinces, the average session is 139 sitting days. In all of the other provinces the averagesession is 59 sitting days. The model predicts that the effect of this difference is a 17%increase in the number of bills amended, in other words about 1 additional bill over theaverage 7 in COWH provinces. Other than committee review, the most significant effect in the model is that of the electoralcycle. Again, if the average COWH session passes 50 bills, the model predicts the firstsession would see about 7 amended, but this would rise to 9 by the fifth (unless it was anelection year). The average committee review session on the other hand, while still passingabout 50 bills, sees 31 of these amended in a first session, rising to 41 by the fifth. Theexception is the pre-election session. These sessions have a 15% decrease in the number ofbills amended. If the fifth sessions above were also the final sessions of the parliament, thiswould mean a reduction to 8 and 35 bills amended respectively. These effects, while certainlymeaningful, also reiterate the degree to which the effect of committee review is35overwhelmingly important.Models 3 through 5 serve as robustness checks. Because of the concern that the results couldeasily be driven by the uniqueness of Ontario and Quebec, models 3 and 4 repeat the analysiswith Ontario and Quebec dropped from the sample. As such, these models capture the effectof committee review only in Manitoba and post-2003 Saskatchewan. Nonetheless, the resultsare remarkably consistent. Most other coefficients remain consistent across the models, withthe most notable difference being that the effect of population drops to insignificance. Model5 tests that the results aren't driven merely by cross-sectional variance. By limiting thesample to only Saskatchewan, I test only for the effect of change over time within a province.This model excludes the variable for number of seats because it has changed very littlewithin the province in this time period, however this omission does not affect the results. Theresults are again consistent with the previous models. In all, these models strongly suggestthat committee review increases the ability of the legislature to impact legislation throughamendment.5.2 Modelling legislative efficiencyThe second set of models tests hypothesis 4: that committee review reduces the need forexpedited bill passage. Given that Anstett and Thomas' discussion of low efficiency inManitoba was confirmed by the descriptive statistics, Manitoba is not counted as acommittee review province in models 3 through 6, but instead is controlled for by a separatedummy in the models.12  Like the number of bills amended, the number passed in the last 5days of the session is still count data. For this data tests showed similar overdispersionproblems and so negative binomial regressions are used again. To model the measure ofeffective concentration regular OLS is appropriate. In all cases the limited model, controllingonly for professionalization, is fit first, and the fuller model afterwards. Once again standard12 The effect of instead modelling Manitoba as a committee review province is fairly predictable: significantcoefficients on the committee review variable of the same direction and size of those seen on the Manitobadummy variables in Table 8.36errors are robust and are clustered by province.There are some changes in the covariates used in these models compared to the previoussection. The basic models in each specification include a new term measuring the proportionof omnibus legislation each session. This controls for the additional stress larger pieces oflegislation inevitably put on the session's schedule. For similar reasons, in the full models, acontrol is also included for wether or not the session includes the passage of a budget, whichlikely affects the time remaining to pass other legislation. Additionally, I add a dummyvariable on whether the session is the first  after a change in government. This controls forany effect that might arise from a high volume of policy change desired by a newgovernment. A final difference in this set of models is that sessions in which minoritygovernments were defeated are excluded. This is because the legislative calendar in thesesessions is truncated unpredictably, meaning there is no ?end? of the session at which time arush could occur. The results are given in Table 8. In models 1 and 2, coefficients are againIRRs, meaning that numbers above 1 indicate an increase in the number of bills passed at theend of the session, and thus an increase in concentration. In the OLS regressions, models 3and 4, it is negative effects that show an increase in concentration, as the effective number ofthird reading days declines. Across the board, Manitoba has significantly higher levels of concentration. In the average50 bill session, model 2 predicts that provinces without committee review pass 12 bills in thelast 5 days, and those with it pass 8. These predictions are not statistically different from oneanother at a 95% confidence level. The model predicts that Manitoba, on the other hand,would pass 31 of these 50 bills at the end of the session. The effect is similar in model 4,while outside of Manitoba, the predicted effective number of third reading days is 9 for bothCOWH and committee review provinces, it is 3.3 for Manitoba itself. Thus, while committeereview in general has no measurable effect on efficiency in this data, there is somethingparticular about Manitoba that causes problematic delays.Unlike the models of legislative influence, these models suggest that other aspects of  37Table 8: Regressions, legislative efficiency in Canada's provinces, 1983-2013Passed last 5 days (negativebinomial)Effective Third ReadingConcentration (OLS)Model 1 Model 2 Model 3 Model 4CommitteeReview10.957 0.673 -2.365 -0.276Seats 1.004 1.030*** 0.012 -0.082Compensation 0.999 0.999*** 0.0001 0.0001**Sitting Days 0.989*** 0.986*** 0.102*** 0.104***Budget Session --- 1.160 --- 1.196OmnibusProportion0.021*** 0.029* 16.909** 8.467Pre-election --- 1.220 --- -1.188*NewGovernment--- 1.145 --- -1.472Minority --- 0.417* --- -1.795Majority Size --- 0.986*** --- 0.047*Population (ln) --- 0.600* --- 2.596*GDP pc --- 0.999 --- -0.0001ManitobaDummy3.315*** 2.680*** -6.57*** -5.60***Exposure: Bills passed Bills passed --- ---N 262 262 262 262? Parameter 1.360 1.247 --- ---? = 0  p-value 0.000 0.000 --- ---R2 --- --- 0.594 0.641*: p < 0.1; **: p < 0.05; ***: p < 0.01; Coefficients in models 1 and 2 are reported as Incident Rate Ratios,standard errors are robust and clustered by province1: This excludes Manitoba, whose effect is captured separatelyprofessionalization are relevant for efficiency. Not surprisingly, more sitting days decreasesthe reliance on a last-minute rush, and the overall concentration of passage. Salary also has amodest effect in the model of effective concentration. The number of seats in the legislature,on the other hand, predicts a statistically significant increase in the number of bills passed at38the end of the session. Though this result is not repeated in model 4. Ultimately, what theseresults suggest, in conjunction with the results for committee review, is that various attemptsat professionalizing Canada's provincial legislatures have had mixed success. More researchshould be done to separate out these effects.Other factors that predict more effective use of legislative time are a larger population, alarger government majority, and the existence of a minority government. In the averagesession, model 2 predicts a minority would pass 10 bills in a rush at the end, whereas agovernment with a 10% majority (e.g. 55 seats out of 100) would pass 20, and a governmentwith a 30% majority (e.g. 70 seats out of 100) would pass 15. However, while the result formajority size holds in model 4, the effect of minority governments is no longer significant. In all, these models provide no evidence that committee review helps provinces to deal withlegislation more efficiently. There are a number of ways of getting at the issue of efficiency,and this approach has examined only the one most in line with the present theory. However,two other pieces of evidence are suggestive of the possibility that committee review has notbeen as helpful as it was thought to be. First, provinces that use committee review usesignificantly more omnibus legislation than those which do not. In provinces with committeereview, 8% of legislation is coded as omnibus legislation, whereas the number is 2.5% inother provinces. Given that omnibus legislation is an institution that both helps governmentspass a greater volume of policy per bill and also is a means to deal with more involved andcomplex policy problems, this may suggest that committee review, whether or not it hashelped, has not been sufficient to address these problems. Provinces that use committeereview are also the only provinces in this time period that have carried over legislation fromone session to the next. Manitoba carries over 2.4% of bills introduced, Ontario 4.2%, andQuebec 7.6%. Saskatchewan did use the practice prior to introducing committee review in2003, but only on 0.1% of bills. Since 2003 it has been used on 1% of all bills introduced.This again suggests the possibility that committees can create a burden on getting legislationpassed that is not seen in provinces that use COWH.396. ConclusionA number of important conclusions and avenues for future research emerge from thesefindings. There is strong evidence that committee review creates a higher level of legislativescrutiny in Canadian provinces where it is present. This undermines typical views ofWestminster committees that cast them as ineffective and unimportant. Nonetheless, it isimportant to put this finding in context. Committees in these legislatures are not makingpolicy through this process. The government still maintains the power to propose legislation,and if it ever wants to can easily override any concerns that committees might have. But thefact that committee review appears to double the rate at which bills are amended stronglysuggests that bills receive better scrutiny in these legislatures, with more opportunity forpublic input.This conclusion leaves unaddressed some implications that are important to test in furtherwork. Are the amendments passed by these committees mostly insubstantial? It is possiblethe increased scrutiny amounts to little more than a glorified error-checking process. If so theclaim of an important and influential role for the legislature would be undermined. A furtherapproach would be to examine if the apparent difference in scrutiny manifests in measurablepolicy differences between these provinces. For example, do legislatures that practicecommittee review pass legislation with more public support than those which do not?Assessing the practical impact of the effects observed in the present study would not besimple. It is clear though, that the effects exist, and are substantial.The results as regards efficiency are less clear. Though the analysis presented here shows thatcommittee review has not guaranteed a higher level of efficiency, that does not mean it hasno effect. What it suggests is that governments, in the absence of the ability to devolve workto a parallel committee system, have found other mechanisms to enhance the efficiency oftheir legislative process. Most of the variance in efficiency was explained by the political40context, and structural features like the population of the province.In Westminster parliaments, legislatures are the only directly elected branch of government,and democratic accountability mechanisms must run through legislative elections. Whilemany scholars see these elections as merely a means to choosing an executive, the questionof whether legislatures have an influence beyond that role is an important one. This datashows that the use of standing committees to review legislation in Canada's provinces hasmaterially increased the influence of these legislatures. This result contributes to a growingresearch agenda demonstrating the previously under-estimated role legislatures can have inparliamentary systems. Though policy in Westminster systems always runs through thegovernment first, the legislature's role in refining and scrutinizing it cannot be written off.41ReferencesAlesina, A., Cohen, G. D., & Roubini, N. (1993). Electoral business cycle in industrialdemocracies. European Journal of Political Economy, 9(1), 1?23. doi:10.1016/0176-2680(93)90027-RAnstett, A., & Thomas, P. G. (1989). Manitoba: the role of the legislature in a polarizedpolitical system. In G. Levy & G. White (Eds.), Provincial and TerritorialLegislatures in Canada (pp. 90?109). University of Toronto Press.Arter, D. (2006). Introduction: Comparing the legislative performance of legislatures. TheJournal of Legislative Studies, 12(3-4), 245?257. doi:10.1080/13572330600875423Berry, W. D., Berkman, M. B., & Schneiderman, S. (2000). Legislative Professionalism andIncumbent Reelection: The Development of Institutional Boundaries. The AmericanPolitical Science Review, 94(4), 859?874. doi:10.2307/2586212Blais, A., & Nadeau, R. (1992). The electoral budget cycle. Public Choice, 74(4), 389?403.doi:10.1007/BF00137686British Columbia Legislature. (2012, May 15). Official Report of Debates of the LegislativeAssembly (Hansard). 2011 Legislative Session: Fourth Session 39th Parliament.Retrieved from, P. (2006). The analysis of Scottish Parliament committee influence: Beyondcapacity and structure in comparing West European legislatures. European Journal ofPolitical Research, 45(2), 181?208. doi:10.1111/j.1475-6765.2006.00295.xCanada. Parliament. House of Commons. (1968). House of Commons debates. (Vol. 28thParliament. First Session. Vol. 2). Ottawa: Queen?s Printer.Cannon, L. (1990). Rules of Procedure in the National Assembly: The 1984 ReformExperience. Canadian Parliamentary Review, 13(2). Retrieved from42 Canadian Limited. (1982). Provincial pulse (Vols. 1-3). Don Mills, Ont: CCH CanadianLimited.CCH Canadian Limited. (1987). Provincial legislative record. Don Mills, Ont.Commission on the Legislature. (1973). Report no. 4. Ontario Legislative Assembly.Cox, G. W. (2008). The Organization of Democratic Legislatures. In D. A. Ritchie & B. R.Weingast (Eds.), The Oxford Handbook of Political Economy. Oxford UniversityPress.Cox, G. W., & McCubbins, M. D. (1994). Bonding, Structure, and the Stability of PoliticalParties: Party Government in the House. Legislative Studies Quarterly, 19(2), 215?231. doi:10.2307/440425Cox, G. W., & McCubbins, M. D. (2007). Legislative Leviathan: Party Government in theHouse. Cambridge University Press.Damgaard, E., & Jensen, H. (2006). Assessing strength and weakness in legislatures: Thecase of Denmark. The Journal of Legislative Studies, 12(3-4), 426?442.doi:10.1080/13572330600877486Dunn, C. (1990). Executive Dominance in Provincial Legislatures. Canadian ParliamentaryReview, 13(1). Retrieved from, M. P. (1994). Divided Government in the American States: A Byproduct ofLegislative Professionalism? The American Political Science Review, 88(2), 304?316.doi:10.2307/2944705Fiorina, M. P. (1999). Further Evidence of the Partisan Consequences of LegislativeProfessionalism. American Journal of Political Science, 43(3), 974?977.doi:10.2307/299184343Fiorino, N., & Ricciuti, R. (2007). Legislature size and government spending in Italianregions: Forecasting the effects of a reform. Public Choice, 131(1-2), 117?125.doi:10.1007/s11127-006-9108-yFleming, R. J., & Glenn, J. E. (1997). Fleming?s Canadian Legislatures. University ofToronto Press.Gilligan, T. W., & Krehbiel, K. (1990). Organization of Informative Committees by aRational Legislature. American Journal of Political Science, 34(2), 531?564.Griffith, J. A. G., & Planning, P. and E. (1974). Parliamentary Scrutiny of Government Bills.Allen & Unwin [for] PEP and the Study of Parliament Group.Hall, A. (2010, May 20). Controversial bills passed into law. Leader Post. Regina, Sask.,Canada. Retrieved from, K. E. (1983). Patterns of Influence among Committees, Agencies, and InterestGroups. Legislative Studies Quarterly, 8(3), 379?426. doi:10.2307/439591Hockin, T. A. (1970). The Advance of Standing Committees in Canada?s House ofCommons: 1965 to 1970*. Canadian Public Administration, 13(2), 185?202.doi:10.1111/j.1754-7121.1970.tb00016.xKam, C. (2001). Do Ideological Preferences Explain Parliamentary Behaviour? Evidencefrom Great Britain and Canada. The Journal of Legislative Studies, 7(4), 89?126.doi:10.1080/714003894Kam, C. (2009). Party Discipline and Parliamentary Politics. Cambridge University Press.Kornberg, A., & Mishler, W. (1976). Influence in Parliament, Canada. Durham, NC: DukeUniversity Press.Krehbiel, K. (1992). Information and Legislative Organization. University of MichiganPress.Laakso, M., & Taagepera, R. (1979). The ?Effective? Number of Parties: A Measure with44Application to West Europe. Comparative Political Studies, 12(1), 3?27.Lees, J. D., & Shaw, M. (1979). Committees in legislatures: a comparative analysis.Durham, NC: Duke University Press.Legislative Assembly of Alberta. (2010). Member Indemnity and Expense AllowanceComparisons (p. 1). Retrieved from,%202010.pdfLegislative Assembly of British Columbia. (2013). Hansard Services: Previous Sessions.Retrieved July 2, 2013, from Assembly of Saskatchewan. (2003, April 3). Debates and Proceedings 4thSession 24th Legislature. Retrieved from, J. (2009). Strengthening parliament?s powers of scrutiny? An assessment of theintroduction of Public Bill Committees (Report No. 145). London, UK: TheConstitution Unit, Department of Political Science, UCL. Retrieved from, J. (2010). Public Bill Committees: An Assessment Scrutiny Sought; Scrutiny Gained.Parliamentary Affairs, 63(3), 534?544. doi:10.1093/pa/gsp050Long, J. S. (1997). Regression Models for Categorical and Limited Dependent Variables.California: SAGE.Mallory, J. R., & Smith, B. A. (1972). The legislative role of parliamentary committees inCanada: the case of the Joint Committee on the Public Service Bills1. CanadianPublic Administration, 15(1), 1?23. doi:10.1111/j.1754-7121.1972.tb01225.xMalloy, J. (2004). The executive and parliament in canada. The Journal of LegislativeStudies, 10(2-3), 206?217. doi:10.1080/135723304200032231945Mandryk, M. (2010, May 21). NDP victory comes at price. Star - Phoenix. Saskatoon, Sask.,Canada. Retrieved from, L. W., & Vanberg, G. (2005). Coalition Policymaking and Legislative Review.American Political Science Review, 99(1), 93?107. doi:10.1017.S0003055405051518Massicotte, L. (1989). Quebec: the successful integration of French culture and Britishinstitutions. In G. Levy & G. White (Eds.), Provincial and Territorial Legislatures inCanada (pp. 62?89). University of Toronto Press.Massicotte, L. (2013). Omnibus Bills in Theory and Practice. Canadian ParliamentaryReview, 36(1). Retrieved from, I., & Str?m, K. (1995). Parliamentary Committees. In H. D?ring (Ed.), Parliamentsand Majority Rule in Western Europe (pp. 249?307). Frankfurt/New York: CampusVerlag/St. Martins Press.Meeting on Bill 80 heated. (2009, November 27). Star - Phoenix. Saskatoon, Sask., Canada.Retrieved from, S. R., & Hasecke, E. B. (2003). Term Limits, Professionalization, and PartisanControl in U.S. State Legislatures. Journal of Politics, 65(3), 898?908.doi:10.1111/1468-2508.00218Moncrief, G. F. (1994). Professionalization and Careerism in Canadian ProvincialAssemblies: Comparison to U. S. State Legislatures. Legislative Studies Quarterly,19(1), 33?48. doi:10.2307/439798Monk, D. (2010). A Framework for Evaluating the Performance of Committees inWestminster Parliaments. The Journal of Legislative Studies, 16(1), 1?13.doi:10.1080/13572330903541904Newell, J. L. (2006). Characterising the Italian parliament: Legislative change in longitudinal46perspective. The Journal of Legislative Studies, 12(3-4), 386?403.doi:10.1080/13572330600877437Petry, F., Imbeau, L. M., Cr?te, J., & Clavet, M. (1999). Electoral and Partisan Cycles in theCanadian Provinces. Canadian Journal of Political Science/Revue canadienne descience politique, 32(02), 273?292. doi:10.1017/S0008423900010490Pettersson-Lidbom, P. (2012). Does the size of the legislature affect the size of government?Evidence from two natural experiments. Journal of Public Economics, 96(3?4), 269?278. doi:10.1016/j.jpubeco.2011.07.004Rush, M. (1982). Studies of parliamentary reform. The Journal of Commonwealth &Comparative Politics, 20(2), 138?154. doi:10.1080/14662048208447404Shepsle, K. A., & Weingast, B. R. (1987). The Institutional Foundations of CommitteePower. The American Political Science Review, 81(1), 85?104. doi:10.2307/1960780Special Committee on Organization and Procedure. (1968). Third Report. In Journals of theHouse of Commons of Canada (Vol. Second Session, 27th Parliament Vol. 4, pp. 760?767). Ottawa, Ont: Queen?s Printer.Squire, P. (2007). Measuring State Legislative Professionalism: The Squire Index Revisited.State Politics & Policy Quarterly, 7(2), 211?227. doi:10.1177/153244000700700208Statistics Canada. (2011). Gross domestic product (GDP), expenditure-based, provincialeconomic accounts. CANSIM. Retrieved July 4, 2013, from Canada. (2012). Estimates of population, by age group and sex for July 1, Canada,provinces and territories. CANSIM. Retrieved July 4, 2013, from, G. J. (1976). The Sizes of Legislatures. The Journal of Legal Studies, 5(1), 17?34.doi:10.2307/72407247Thomas, P. G. (1978). The Influence of Standing Committees of Parliament on GovernmentLegislation. Legislative Studies Quarterly, 3(4), 683?704. doi:10.2307/439621Thompson, L. (2012). More of the Same or a Period of Change? The Impact of BillCommittees in the Twenty-First Century House of Commons. Parliamentary Affairs.doi:10.1093/pa/gss016Weingast, B. R. (1989). The Political Institutions of Representative Government:Legislatures. Journal of Institutional and Theoretical Economics (JITE) / Zeitschriftf?r die gesamte Staatswissenschaft, 145(4), 693?703. doi:10.2307/40751252Weingast, B. R., & Marshall, W. J. (1988). The Industrial Organization of Congress; or, WhyLegislatures, Like Firms, Are Not Organized as Markets. Journal of PoliticalEconomy, 96(1), 132?163. doi:10.2307/1830714White, G. (1980). The Life and Times of the Camp Commission. Canadian Journal ofPolitical Science / Revue canadienne de science politique, 13(2), 357?375.doi:10.2307/3229729Winkelmann, R. (2008). Econometric Analysis of Count Data. New York: Springer..48


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



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"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items