At the time this article was published Lucinda Flavelle and Philip Kaye were Research Officers in the Research Service the Ontario Legislative Library.

Party discipline is a mechanism political parties use to keep their members functioning as a cohesive group rather than as a collection of disparate individuals. It encourages party loyalty among members who may be tempted to act individually. The most significant manifestation of party discipline occurs in legislative voting: members of a party will vote together on the floor of a legislature.

Disciplined voting is particularly common in parliamentary regimes, where the norm in legislative voting is that of a highly unified government majority facing a cohesive opposition party (or parties). Legislative voting in the Canadian House of Commons has generally followed this pattern since the turn of the century.

Traditionally, divisions in the Canadian House have been considered by governments and opposition parties to be tests of the legislature's confidence in the government. This particular confidence convention, which has been accepted in Canada for almost a century, is rooted in the notion of responsible government. Responsible government holds that the government remains in office only as long as it has the support of the legislature. In a system of responsible government, the executive and legislative branches are fused, and the cabinet is responsible to the legislature for its actions. If the government loses a major vote in the House, the convention has usually meant that the government must resign or seek a dissolution. Thus, backbench members support the executive so that their party can remain in power. By the same token, opposition members seek united action in order to bring down the government.

Voting discipline did not always prevail in Canadian politics. Only in the last two decades of the nineteenth century did party lines, and thus, party loyalties, begin to coalesce. Until that time, a substantial number of elected members were independents, or 1ocise fish," as John A. Macdonald labelled them. Members were not particularly concerned about acting cohesively to achieve party goals; their primary aim was to be on the winning side so that they would secure as many benefits as they could for their constituents.

A similar situation prevailed in the United Kingdom and Australia. Governments in Britain were regularly defeated in the second half of the nineteenth century, and often changed without elections. The rudimentary beginnings of party discipline in Australia dated from about the same time. There, disciplined voting was evident in the "pledge" system. Beginning in the 1880s, party members would pledge themselves in support of, for example, free trade or protectionism, and would lose party en0dorsement if they voted against their pledged stand. Lists of pledged candidates were published by the parties to aid voters.

Students of party politics associate the development of party discipline in these. nations with the expansion of the electorate and changes in electoral practices. In Canada, for example, property requirements limited the size of the electorate prior to 1885. Moreover, organizing voters was difficult. Until 1878, the open ballot meant that voters had to publicly declare their support of a candidate; elections were non simultaneous and the government could schedule difficult constituency elections to take place after those in "safe" constituencies, so as to make use of bandwagon effects; and the government could gerrymander constituency boundaries to minimize opposition voting strength. Changes in these electoral practices encouraged the development of a disciplined party system. In order to fulfill election promises to newly enfranchised groups, parties needed individuals who would act in concert with their party colleagues.

Samuel Beer, writing about the British House of Commons in 1965, noted that cohesive voting "bad increased until in recent decades it was so close to 100 per cent that there was no longer any point in measuring it.` Cohesive legislative voting is the norm on the floor of the German Bundestag, the Italian Chamber of Deputies, and the Australian Parliament, in addition to the Canadian national and provincial Parliaments. In fact, in Ottawa, party discipline is so complete that a party has never been forced out of office because of a breach in voting discipline.

Although party discipline has governed members' legislative voting for decades, defections from loyal party voting have occurred in all legislative parties in democratic countries. These defections tend to take place more frequently in parties with large majorities where a deviation from the party line would not affect the legislative outcome, or in parties on the extreme right or extreme left. Since the middle of this century, members of the British Parliament have indicated an increased willingness to vote against their parties. This decline in the extent of voting discipline was particularly evident in the 1970s. Between 1972 and 1979, the government suffered sixty-five defeats, many of which were due to backbench dissent. (However, only three of these defeats were followed by votes of confidence, and on only one occasion did the prime minister request a dissolution.)

Philip Norton, a noted student of British politics, suggests that the increased determination of backbench members to reject the advice of whips or party leaders brings into sharp relief the constitutional myth that a governing party must resign or request a dissolution if it is defeated in any division. He suggests three categories of government defeats, each requiring a different response. A defeat on a vote of confidence means the government must resign or seek a dissolution. A defeat on an issue central to government policy means the government must either seek a vote of confidence or resign. A defeat on an item not central to government policy means the government only has to consider whether to accept the defeat or seek its reversal. It is the last type of defeat which Norton argues has formed the majority of defeats at Westminster. And more often than not, the government of the day has accepted the defeat, either outright or in modified form.

Nevertheless, members who vote against their party are usually perceived as having committed a serious breach of party discipline. In one study of the British House of Commons, MPs were asked to rank "acts of disloyalty" in order of seriousness. The items to be ranked included, among others, cross-voting, abstaining, writing critical letters or articles in the press, making critical speeches (both inside and outside Parliament), and expressing private dissent to the whips. Party leaders and backbenchers both rated cross-voting as the most serious violation of party discipline.

Party discipline in legislative voting is maintained by the party whips. According to Professor Martin Westmacott, their duties, among other things, are to contain dissent, to promote cohesion, to serve as a "sounding board" for the concerns of backbenchers and to determine whether accommodation can be reached when the party position and that of an individual member come into conflict. They are responsible for ensuring that party members are present for all important divisions in the House and that members vote along party lines.

If members fail to follow the party whip, various sanctions may be imposed. Dissidents may be ostracized by party associates; refused party funds and organizational support in election campaigns; passed over for promotion to cabinet; denied decent office accommodation and adequate staff; overlooked as members of prestigious parliamentary committees or of travelling parliamentary delegations; denied the opportunity to ask a question during Question Period; refused party assistance in performing services for constituents; or expelled from the caucus. The most drastic measure available in theory to a party leader to discipline errant members is to request a dissolution of Parliament and call an election. Few members enjoy "safe" seats, and no member relishes the prospect of fighting for a job.

Some observers believe the ability of the whip and party leaders to maintain party discipline has been overestimated. The whip, after all, has few formal powers, especially in Canada. Some dissident members are promoted to the front-bench based on sheer ability; others are moved into cabinet posts, which makes it much more difficult for them to criticize government policy. Efforts to thwart re-nomination in the constituency may not be well received. In Britain, for example, local associations tend to continue to support members who vote against their party. For these reasons, whips and party leaders prefer to use quiet diplomacy and persuasion, rather than sanctions, to make sure the party's position is implemented on the floor.

In assessing the motivation of members, one can conclude that disciplined voting is, in large part, a self-imposed discipline: members perceive it as serving their self interest. Government members will almost invariably support their party, even if they object to a proposal, when the alternative is losing a vote of confidence, losing office, and possibly losing their seats. Second, members usually want to support their party. They recognize and accept the powers of the parliamentary leadership. The party leader, after all, is to a large extent responsible for the party's success at the polls and many members ride on the coattails of their leaders into office. Third, many ministers have been promoted to cabinet because they have the support of backbenchers. Hence, backbenchers are likely to support them in legislative voting. Moreover, members of a party tend to share common beliefs, values, and goals, and are united in their concern for the viability of the party. Finally, backbenchers normally aspire to ministerial posts and believe that dissent will damage their chances of promotion to the front bench.

The Case for Party Discipline

For the government, party discipline provides an assurance that it will be able to have its legislative programme passed intact (at least with a majority government). A break in party ranks, on the other hand, might cost the party control of office or, at a minimum, significant political embarrassment.

The voting predictability which results from party discipline also allows ministers to know the fate of their bills. As one British cabinet minister remarked, it is carrying democracy too far if you do not know the result of a vote before the meeting. Further voting predictability ensures stable government. Frequent elections are not well liked by any electorate. Assured support means that the government can get on with the business of governing. Related to this idea of stable government is the way party discipline provides legitimacy to the party system and parliamentary government. In a 1962 study of the 25th Canadian Parliament, MPs were asked to reply to the question, "What are the advantages of going along with your party?" Forty-eight percent of the legislators answered that they were motivated by a desire to see the party achieve its goals and to maintain its viability. Others fourteen percent emphasized the viability of the parliamentary system itself and the need to secure its continued maintenance. (Twenty-tree percent said that they were motivated to act cohesively because it was personally advantageous both to themselves and their party. Only eleven percent replied that there were no advantages in voting with their party.)

Party discipline also means that voters are provided with clear alternatives. Under a parliamentary system, the government party must formulate policies and the opposition parties, in criticizing these policies, are expected to present alternatives. In so doing, all parties have a strong concern with issues. The demand for internal party discipline reinforces the consistency of each party's position with respect to issues.

Relaxing party discipline would mean extra organizational efforts on the part of already overworked party leaders. The argument has been made that even if the formal strictures of party discipline were removed, the leaders of the various parties would still want to elicit support by approaching members on an individual basis. To accomplish this goal, each party would have to identify the positions of individual members. Such a development would require that considerably more time and effort be spent on organization.

Relaxing disciplined voting would mean a move in the direction of a congressional system. Members would be free to negotiate among themselves for support on votes (as are Members of the United States Senate and House of Representatives). Their voting records would also make them more vulnerable to lobbyists and special interest groups.

Finally, it can be argued that party discipline is a prerequisite for the smooth functioning of the modern administrative state. The role which governments play in social and economic matters requires an enormous amount of planning. But lengthy parliamentary debate does not lend itself to an efficient decision-making process.

The Case Against Party Discipline

One rationale for relaxing party discipline on the floor of the House focuses upon "conscience issues." Questions of morality, it is argued, ought to be decided according to the consciences of individual members, and not according to the dictates of the party whips. Such questions are alleged to be above partisan politics.

In practice, however, over the past two decades only two issues have been handled by free votes in Ottawa: the capital punishment debates and one of the flag debate motions. Issues such as abortion, homosexuality, and divorce were handled as whip votes.

Some MPs have claimed they would like to act more independently when the party position placed them in direct conflict with the interests of their constituents. Relaxing party discipline would mean that members in such a position could vote according to the wishes of their constituents. A 1983 Gallup poll showed that 49.5 percent of the respondents thought that Members of Parliament should vote as their constituents would want. Only 7.9 percent thought that members should vote as their party required. More than one-third (38.3 percent) thought that members should vote according to their own judgement. Thus, there seems to be clear public support for relaxing the principle of party discipline in legislative voting.

On some issues, relaxing party discipline can help ease tension in the legislature. An illustration is the 1964 decision of Prime Minister Lester B. Pearson to make the vote on a new flag a free one. He told the House of Commons that "in a question so personal, so close to their conscience, so deep in their emotions, as a flag, a member of the Liberal party, in voting against the design submitted by the government, would not be read out of the party for so doing."

Relaxing party discipline could mitigate voter cynicism about parliamentary government. According to some voters the main role of Canadian legislative bodies is to legitimize decisions which have already been made by a small core group, the cabinet. If the outcomes of votes were less predictable, the level of parliamentary debate would be raised, and the media and the public would likely follow those debates more closely. The institution of Parliament would experience enhanced credibility, and it would be perceived as a more effective instrument of accountability to the electorate.

In 1980 Robert Stanfield, a former leader of the federal Progressive Conservative party, called for more free votes in order to 1ncrease the dignity of the backbenchers." Mr. Stanfield conceded that in any session certain bills are the core of the government's legislative programme and the government must secure their passage. On the other hand, there are many bills which are not really part of the core programme. Mr. Stanfield continued: 1f we could reach the point where we only had a party line vote when the government indicated it was a vote of confidence, this would increase the dignity of the backbenchers both on the government side and in the opposition there would be free votes in the opposition too. A member would then feel less like a number, less like a cipher. It would be good therefore for both government and opposition backbenchers and it would be good too for the perception of members of Parliament by the public. It would make them look bigger, more independent, more important, more self-reliant. I think it should be given consideration."

Rigid discipline is a significant barrier to the expression of the wide variety of interests and opinions which is at the foundation of a dynamic parliamentary system. Most importantly, it can serve to stifle effective representation of important regions. As a result, voters may turn to provincial governments or third parties to serve their interests.

Less rigid discipline could possibly improve public policy. Governments may be more open and more willing to reveal the information upon which policy decisions are based. It also could make governments more willing to accept backbench amendments. In one study of amendments in the British Commons, 99.9 percent of ministerial amendments were approved by the Commons while only 9.5 percent of backbench amendments were approved. Governments could be forced to modify, withdraw or rethink measures because of backbench pressure.

Finally, relaxed voting discipline could be useful to ministers who wish to overrule the views of their senior officials. A minister who disagrees with his officials on a measure can use the argument that "the House will not accept this" to win his or her point.

The Ontario Experience

An analysis of recorded votes during the 32nd Parliament of the Ontario Legislature (19811984) reveals a high degree of party cohesion. Ninety-four percent of all the recorded votes were party votes. (A party vote is a vote in which at least 90 percent of the members of each party vote together). If one looks only at the more significant votes, i.e., votes respecting government bills and committee reports, the degree of party solidarity is even higher.

A party is more likely to be divided on a resolution or a private member's bill. Here, confidence in the government is not at issue. Indeed, during the fourth session of the 32nd Parliament, the only recorded votes revealing splits within a party were votes on resolutions sponsored by Jim Foulds, MPP, a member of the New Democratic Party (May 10, 1984) and by Ernie Eves, MPP, a member of the Progressive Conservative party (June 7, 1984). Mr. Foulds' resolution held that the cost of medically necessary travel in excess of 200 miles should be fully insured under OHIP and that the government should institute such coverage for residents of Northern Ontario within two years. Mr. Eves' resolution focused on the parole system. It urged the federal government to

consider increasing the minimum period of time that must be served by an inmate to be eligible for parole. Votes on these two resolutions make up only four percent of all recorded votes during the fourth session. This voting pattern party unity on over 90 percent of recorded votes underscores the significance of party discipline in Ontario.

Two recent developments may have an impact on this pattern. In the May 1985 "agreement for a reform minority parliament signed by Premier David Peterson and New Democratic Party leader Bob Rae, the leaders agreed that only the budget as a whole or a specifically worded motion of confidence will be considered matters of confidence. To date, this change has not noticeably affected voting patterns in the House.

The second development is the release of a report by the Legislature's Standing Committee on Procedural Affairs and Agencies, Boards and Commissions in the fall of 1985. The report proposed, among other things, that a government resign or request a dissolution only if defeated on general budget or supply bills or specifically worded motions of confidence. This proposal mirrors the recommendations of the Special Committee on Reform of the House of Commons which was accepted, in principle at least, by the federal government. The Ontario committee recommended that its proposal be implemented by March 3,1986 for a one-year trial period.