While free speech is obviously an important right that needs judicial nurture and protection, equating democracy with individualsâ ability to purchase unlimited quantities of partisan advertising is not the only interpretation of a free and democratic societyâŠA more robust model of democracy should seek to prevent those with wealth from influencing, through sheer purchasing power of commercial advertising, which issues are deemed to be important.
Janet L. Hiebert, 19981
Pressed about why Bush was declining to abide by the spending limits that will apply to most of his rivals, Mindy Tucker, a spokeswoman for the Bush campaign, said: âI think Governor Bush feels that this is part of the democratic process that people can contribute to a campaign to express their opinion. There are 74,000 who have heard his message and to convey their opinion, they have made a contribution in sums ranging from $5 to $1,000.â
Don Van Natta Jr, 19992
The manner in which election campaigns are conducted and financed is a significant factor in the health of liberal democracies. Regrettably, on two continents there are signs of a malaise. In the United States, which is in the throes of a presidential election campaign, the issue of election financing has dogged the primaries and is likely to cloud the general election. The reasons are the disturbingly large amounts of money that have been raised and spent by the candidates thus far, which led Arizona Senator John McCain among others to make campaign finance reform a prominent topic in his failed bid for the Republican Partyâs nomination. Meanwhile, in Germany former Chancellor Helmut Kohl and his Christian Democratic party are found to have been so embroiled in questionable financial practices that the party itself is in a state of near collapse and Kohlâs reputation is in tatters.
In all liberal democracies, the financing of election campaigns is regulated by the state. The political parties, the candidates and their multitudes of advisers understandably are preoccupied with the details of the particular regulatory scheme under which they operate. From the standpoint of the public and the public interest in a healthy, lively democracy, however, it is essential to keep in sight the overriding purposes of the regulation. What  might these purposes be? At the risk of oversimplification, we have identified them as follows:
In the remainder of the paper, we examine the efforts made in Canada at the federal level to pursue these purposes as well as the controversies that surround such efforts. Since the United States is an innovator in campaign strategy and campaign finance, and at the same time an influential example for Canadian campaign strategists, we look at some of the current practices there. In the conclusion, we highlight the problems that are likely to prove especially thorny for governments and citizens to address.
It is important to emphasize at the outset that in Canada the consensus on the federal regulatory scheme that appeared to prevail for many years is no longer intact. Instead, the scheme is under attack in the courts and in the media. Even as we write, the federal Parliament is attempting to respond to the criticism by revamping important elements of its campaign finance law. In the face of the crumbling consensus on the old scheme and the unceasing and costly electoral innovations that develop from campaign to campaign, it is essential for citizens to maintain a firm grip on the financial foundations of open, fair and free elections. Certainly it was the concern for precisely those foundations that led to the establishment of the modern campaign finance regimes in Canada and the United States. We begin with a brief reminder of the origins of the Canadian regime, and include some comparisons with the American system, and then turn to the first of the purposes identified above, namely, the avoidance of corruption.
It is often stated that the Canadian system limits expenditures rather than contributions while the American system limits contributions rather than expenditures. While the comparison as stated is hopelessly oversimplified, it does contain an element of truth that is rooted in the origins of the two schemes. The Canadians were worried about the spiralling costs of election campaigns. The Americans were reeling from campaign contribution scandals.
As early as 1963 Canadian politicians were concerned about the costly effects of television on the conduct of election campaigns. In that year the Liberal Party included a promise of election finance reform in its campaign platform, and a year later the Liberal government appointed an Advisery Committee to Study Curtailment of Election Expenditures, chaired by Alphonse Barbeau.3 Thus began the decadeâs worth of work by policy advisers and legislators that culminated in the election-expenses regime enacted by Parliament in 1974.
The Election Expenses Act featured three important objectives. One was to strengthen public confidence in elections by guaranteeing the transparency of the financial activities of the parties and the candidates through strict disclosure and reporting requirements. Another was to promote fair electoral competition by limiting the advantage to a competitor of having more money than anyone else. This was the purpose of the use of public funds to subsidize the election costs of the candidates and the political parties. And it was the purpose of the spending restrictions, which featured then â and still feature: (1) limits on the amount of money that candidates and registered parties can spend on specified âelection expensesâ; (2) limits on the amount of money that individuals and groups who are not electoral competitors can spend (the category is referred to variously as third parties or independent spenders); and (3) rules governing the use of broadcast media by the candidates and the political parties. A final objective was to increase public participation in politics through the offer of tax rebates for donations by citizens to the candidates and the political parties.
As is well known, Canada possesses a system where political parties predominate, and the election-expenses regime has reflected that fact from the outset. For example, there are limits on the election expenses of the candidates and the political parties, but not on contributions to them. Only contributions from foreign sources are outlawed. The limits on election expenses help to equalize the competition between the electoral competitors, and to assist them in avoiding undue or easy reliance on the contributions of big business and wealthy individuals. The limits also minimize one advantage â money â that wealthy independent candidates might use to break into the competitive party circle. In the United States, by contrast, independent candidates face no such obstacle in their efforts to spend their way to office. The notable pioneer of this category is Ross Perot, who made a bid for the presidency in the 1992 election. Perot was thought to have spent some $60 million of his own money on his bid, or more than half as much as his Democratic and Republican opponents together.4
The Canadian assumption of party democracy was also clear in the provisions on third-party or independent spending. Third parties were left free to publicize their position on issues (advocacy spending), but they were prohibited from incurring election expenses, that is, from promoting or opposing the candidates and the parties. The prohibition was designed to ensure that the prospects of participants in an electoral competition were not unfairly harmed by the impact of unaccountable and unregulated money. It was also intended to keep the candidates and the parties from soliciting the aid of third-party spending and thereby escaping the restrictions on their own campaign expenses. However, if third parties did incur such expenses, and were prosecuted for doing so, they could mount a defence to the effect that the effort was aimed at gaining support for a public policy stance and was undertaken âin good  faith.â
In 1983 Parliament removed the good-faith defence, a move that triggered a hostile reaction from at least one interest group, the National Citizensâ Coalition, which went to court to have the provision thrown out as an unwarranted violation of the freedom of expression. Thus began a series of court battles over the idea of limiting independent spending. We will consider this issue more closely in the sections on fair competition and free expression. For now the point is simply to delineate the contours of the Canadian system, including the emphasis on the control of campaign spending by the electoral competitors and by third parties.
As indicated earlier, the American system took shape largely5 in the aftermath of the 1972 presidential election, when it was discovered that individuals had made contributions in the order of one to two million dollars to Republican President Richard Nixonâs successful re-election campaign that year. Further investigations into the Watergate scandal revealed violations of existing election law, including illegal contributions to the Republican campaign from corporations and foreign nationals.6 President Nixon resigned, and in 1974 the Congress passed a comprehensive set of amendments to the existing election campaign law, among them: strict disclosure provisions on contributions and expenditures; limits on contributions to the candidates and the political parties; limits on the campaign expenditures of presidential and congressional candidates, the political parties and independents;7 a system of public financing for presidential campaigns; and the establishment of an independent commission to administer the new scheme.
As is evident from this brief sketch, the Americans initially were as concerned about the corrupt ion implications of the large campaign contributions of the wealthy as they were about the cost of campaigns, and as a result the 1974 American scheme was a comprehensive package of contribution restrictions as well as spending restrictions. However, the scheme lasted only two years, at which point it was gutted by the Supreme Court of the United States. In its landmark ruling in Buckley v. Valeo,8   the court noted that the primary purpose of  the scheme was to forestall corrupt ion in elections, a purpose which it accepted as an important and valid use of Congressâ power. However, the court also emphasized the need to consider carefully the burdens that the corrupt ion purpose might impose on the all-important First Amendment freedoms of speech and association. In the event, and in the name of those freedoms, it threw out the spending limits imposed on congressional campaigns; threw out the limits on presidential campaigns, except for candidates who choose to accept public monies; and threw out the limits on independent spending. The court could not see how spending limits would insulate campaigns from corrupt ion. On the other hand, it upheld the limits imposed on contributions to the candidates and the political parties, and the disclosure provisions, precisely because it could see a direct link between them and the anti-corrupt ion purpose.
A striking aspect of the Supreme Courtâs decision was the insistence that in the modern age money is inextricably related to free speech. The court did not say that money is speech as opposed, say, to property, but it came rather close to that, as the following passage from the decision suggests:
The expenditure of money simply cannot be equated with such conduct as destruction of a draft cardâŠ.[T]his Court has never suggested that the dependenceof a communication on the expenditure of money operates itself to introduce a nonspeech element or to reduce the exacting scrutiny required by the First AmendmentâŠ
A restriction on the amount of money a person or group can spend on political communication during a campaign necessarily reduces the quantity of expression by restricting the number of issues discussed, the depth of their exploration, and the size of the audience reached. This is because virtually every means of communicating ideas in todayâs mass society requires the expenditure of money.9
When all is said and done, the Buckley decision removed the brakes on campaign spending in the United States. With each passing set of elections, congressional and presidential, spending accelerates. Moreover, there is now thought to be more unregulated spending (soft money) than regulated spending (hard money). The candidates are required to report campaign contributions and expenditures, data which are made available to the public by the Federal Election Commission (FEC), which administers the legislation. However, monies that are donated to political parties in some of the states need not be reported at all. Further, there is no limit on the amounts that the  political parties can spend on such activities as get-out-the-vote and voter registration drives and the preparation of campaign materials, so long as the activities in question are not co-ordinated with the campaigns of the candidates.10 The so-called soft-money exception is regarded as an enormous loophole in the American regulatory scheme.
It is fair to conclude that the Canadian and American systems took divergent paths from the start. The Americans began by limiting campaign spending as well as contributions. Now they are left with the contribution limits in place, but no spending limits except for presidential candidates who apply for federal matching grants. In addition, âsoft moneyâ innovations effectively enable the political parties to contribute unlimited support to the candidates so long as the support takes the form of spending that is uncoordinated with the campaigns of the candidates. Not to put too fine a point on it, the American system is well financed and there is increasingly wide concern about it. The noted American constitutional scholar, Ronald Dworkin, has written that âmoney is the biggest threat to the democratic process.â11 The Canadians chose to ignore the idea of contribution limits from the start, and instead clamp down on the spending of the candidates, the political parties and independent individuals and interest groups (third parties). The Canadian system is not thought to be awash in money. Indeed, it is subject to a different criticism altogether, namely, that it places too much of a burden on the freedoms of expression and association, especially in relation to independent spending.
As Dworkinâs pithy comment indicates, campaign finance regulation is not an end in itself. The regulation is a means to an end, the end being a healthy democratic process, the elections of which produce legitimate or widely-accepted results. Indeed, it is precisely because of the instrumental nature of the regulation that it is possible to sort through some of the difficult developments and issues that arise today. With this in mind, we turn to the objectives identified at the outset of the working paper, all of them important to the conduct of legitimate democratic elections. The first, the avoidance of corrupt ion, is not the least significant for being self evident.
Corrupt ion seemingly is a permanent feature of election campaigns. Everyone is familiar with tales of old scandals (treating, or buying votes) and concerns about new ones (e.g., buying pin numbers used in televoting). There are endless variations on the theme, from vote buying to fraud. At issue here, however, is  corruption in terms of the financing of election campaigns. The great fear is not so much vote buying as politician buying.
Election campaigns are costly affairs. There is always the possibility of the individual of independent means who can finance his own campaign. Ross Perotâs 1992 presidential candidacy has been noted. The most recent example in the United States is Steve Forbes. Forbes dropped out of the race to gain the 2000 Republican presidential nomination, but not before spending some $32 million, most of it his own money.12Â (In his 1996 bid, the figure is thought to have been some $37 million.13) Moreover, Forbes himself would often point out the virtue of this feature of his campaign, stating it as proof that he was not in the hands of âthe interests.â Still, Forbes is the exception, not the rule. As a rule, the candidates and the political parties need to raise money in order to prosecute a campaign. That being so, there are two specific concerns. One is that donors will âbuyâ influence with their donations. The other is that the candidates and the parties will âauctionâ future favours.
Whether donors try to buy influence or the electoral competitors implicitly auction future considerations, the net result is a politically corrupt process that favours large campaign contributors over everyone else. In the words of the Royal Commission on Electoral Reform and Party Financing (RCERPF), it is a matter of âundue influence.â14 As might be expected, making contributions to electoral competitors in order to obtain a reward or advantage of some kind is outright bribery and therefore illegal in most countries.15 Short of the criminal sanction for such behaviour, however, there are two main techniques designed to discourage such transactions: reporting requirements and public disclosure of the reports, a process sometimes referred to simply as disclosure; and contribution limits.
Disclosure is the preferred Canadian alternative. The RCERPF considered disclosure to be a âcardinal principle of the present federal regulatory framework for party and election finance.â16 There is no need to undertake an exhaustive review of the details of the Canadian disclosure requirements. They apply to the registered parties and the candidates, both of whom are required to submit reports of their activities at specified intervals. The reports must include an audited account of election and other expenses, and the source and amount of all contributions of more than $100. The reports are available for public inspection, and summaries of reports following general elections are published in local newspapers. It is important in this context not to neglect mention of the position of the âofficial agent.â The political parties need to appoint official agents as do each of the candidates. The official agent must attest to the accuracy of the submitted reports, and is made subject to legal sanction for violations of the law. The position therefore is a key component of the enforcement of the disclosure rules.
In its proposed amendments to the election act, the federal government is seeking to strengthen the disclosure provisions by extending them to third parties. Third parties who incur election advertising expenses of at least $500 must register with the Chief Electoral Officer (CEO), appoint a financial agent, and file a report with the CEL within four months after election day. The report must contain details of the advertising expenses and details of the contributions donated for election advertising purposes, including amounts and the identity of those donating more than $200. Finally, those who incur $5000 or more in election advertising expenses must appoint an auditor to report on the election advertising report.17 The idea of extending the disclosure provisions to third parties might seem somewhat draconian, especially since it places administrative responsibilities on citizens who have been free of them, to date. However, third-party spending is a growing phenomenon, albeit a slowly growing one.18 And the federal government also is proposing to raise significantly the old ceiling on third-party spending. Moreover, since the Canadian system restricts the expenditures of the candidates and the political parties, and imposes disclosure restrictions on them, there seems to be a prima facie case for treating third parties the same.
In the United States, there is heavy reliance on disclosure requirements to maintain the integrity of the federal electoral system. The designated campaign committees of the candidates, the campaign committees of the parties, and the political action committees (PACs) must file detailed reports of expenditures and contributions at specified intervals with an independent regulatory agency, the Federal Election Commission (FEC). The PACs are a uniquely American phenomenon. They are committees that unions, corporations and other groups form for the purpose of channelling money to the campaigns of the candidates and political parties that they prefer. The Canadian governmentâs proposal to apply disclosure provisions to third parties â individuals or groups â that spend on election advertising is not unlike the subjection of the American PACs to disclosure provisions. On the other hand, it is important to stress that in the United States individuals and groups that spend money on election advertising independently of the campaigns of the candidates and the political parties bear no obligation to disclose anything about their activities.
As discussed earlier, the American regime features contribution limits and the Canadian regime does not. It is important to stress that the American limits on contributions made by individuals, PACs and political parties to the candidates are conceived as an anti-corrupt ion mechanism. The idea is simply that limits on contributions will lessen the impact of any one contributor on a candidate. Moreover, the Supreme Court upheld the limits on contributions precisely in the light of this anti-corrupt ion purpose. The court opined that âCongress was surely entitled to conclude that disclosure was only a partial measure, and that contribution ceilings were a necessary legislative concomitant to deal with the reality or appearance of corrupt ion inherent in a system permitting unlimited financial contributions, even when the identities of the contributors and the amounts of their contributions are fully disclosed.â19
Before leaving the issue of disclosure, it is worth stressing how important it is for the effectiveness of disclosure rules that the information disclosed be made easily available to the public. The FEC works hard at making information about campaign finance available to the public and simple to assess. Long before the advent of the commercial internet, Lisa Young observed that âthe American system is designed to make information available in such a manner as to allow voters to use [the] information in deciding how they will vote.â20 In its internet service, this is truer than ever.21 For example, in the presidential party primaries for the contest in November 2000, the FEC has maintained running totals of the candidatesâ contributions and expenditures, updated monthly. The Canadian office of the Chief Electoral Officer has also gone âonlineâ to provide citizens with relevant information about campaign finance.22
Anti-corruption measures like disclosure and contribution limits cannot be confused with fairness. In the United States the Supreme Court, in upholding measures to prevent corruption and the appearance of corruption, has defined corruption narrowly as contributions âgiven to secure a political quid pro quo from current and potential office holders.â23 Theoretically speaking, then, the absence of corruption does not imply that elections are fair. Fairness implies more, and we turn to it now.
The RCERPF presented the most comprehensive account of the fairness concept in connection with Canadaâs election finance regime. In its Report, the commission situated Canadian elections within the countryâs tradition of democratic rights and freedoms, a tradition that was amplified by the adoption of the Canadian Charter of Rights and Freedoms in 1982. The Charter establishes the principle of equality as an inextricable component of democratic rights and freedoms. The commission reasoned that in elections, equality in the exercise of democratic rights and freedoms is possible only if electoral processes have the âproperty of fairness.â It continued:
Fairness is thus the central value that must inform electoral laws if they are to promote the desired outcome of the equality of citizens in the exercise of their democratic rights and freedoms. In this sense, fairness gives meaningful effect to rights and freedoms by setting a standard that the law must meet in regulating behaviour or providing benefits. Electoral laws are fair only to the degree that they promote the meaningful exercise of the rights and freedoms essential to a healthy electoral democracy.24
The value of fairness thus defined is given effect in a number of ways, from the constitutional provision of the universal franchise to the freedom to stand as a candidate for office to financial provisions, the last of which are of concern here and which we will take up shortly. Before doing so, however, it  is worth emphasizing that the meaning of fairness necessarily incorporates some notion of equality. Once that is admitted, then the fairness value is bound to embody concerns that move beyond merely procedural ones such as the provision that each voter may vote only once. The Supreme Court of the United States recognized precisely this point in Buckley in its consideration of the arguments for expenditure limits on third parties or, as the court referred to them, independent expenditure limits.
The court noted that the limits applied only to individual expenditures that advocated the election or defeat of an identified candidate. Expenditures made to promote particular points of view on public policies were subject to no such limits. From the courtâs perspective, this made the provision inadequate in terms of the anti-corruption justification. The court accepted that large individual expenditures, like large individual contributions, pose dangers of corruption. But it reasoned that the absence of limits on advocacy spending meant that some individuals would undoubtedly spend large sums of money for advocacy purposes and find a way to do so in such a manner as to benefit their preferred candidates. Thus the spending limits could not be justified on the ground of avoiding corruption. That left the government with its second line of argument, namely its âancillaryâ interest in injecting some measure of equity into the capacity of individuals and groups to contribute to the outcome of elections. The court rejected this equity argument in unmistakable terms:
âŠthe concept that government may restrict the speech of some elements of our society in order to enhance the relative voice of others is wholly foreign to the First Amendment, which   was designed âto secure the widest possible dissemination of information from diverse and antagonistic sourcesâ and âto assure unfettered interchange of ideas for the bringing about of political and social changes desired by the peopleâ.25
It is worth pondering the American courtâs rejection of the equity purpose on the ground that it necessarily founders on the rock of the first-amendment freedoms of speech and association. In the contest between equality and freedom, freedom wins. The Canadian concern for the kind of equity consideration that the American court spurned clearly implies a different, more balanced approach to these two desiderata. It is also a controversial approach, even in Canada, and it comprises the three prongs cited at the outset of the paper: subsidization of the election costs of the candidates and the political parties; limits on the election expenses of the candidates and the political parties; and limits on the election advertising expenses of third parties.
In terms of the electoral competitors, subsidies and spending restrictions combine to establish a financial floor and ceiling. For example, a candidate is eligible to obtain a reimbursement of half of his election expenses, provided that he obtains at least 15 percent of the valid votes cast in the electoral district. On average, the same candidate in the 1997 general election was restricted to  some
$60,000 in election expenses. A registered political party receives subsidies as well, an important example being the free radio and television advertising time allocated in accordance with a formula based upon factors set out in the governing legislation.26 There are spending restrictions too. For example, in the 1997 general election, any party that ran a candidate in each of the total of 301 electoral districts could spend up to $11,358,749.04. 27
The combined effect of a financial floor and ceiling is an attempt to locate the electoral competitors on something of a level playing field. But no more than that. It is neither expected nor assumed that these competitors will run financially equivalent campaigns. And they certainly do not. In the 1997 general election there were 10 registered political parties. The Liberal party spent the most, just over $11,247,141, very close to the limit. The Green party was at the bottom, recording no election expenses, followed by the Marxist-Leninist party at $375. In other words, the strongest party and its candidates will raise more money and run more expensive campaigns â but within limits. Conversely, the weakest party and its candidates will have difficulty raising money and most probably will run a cheap campaign, quite possibly an ineffective campaign â but a campaign nonetheless. In other words, the equity concern is not blind to electoral realities. The provisions just discussed speak to the effort to keep the electoral competition between the candidates and the political parties open to all contenders. The rationale behind the regulation of election advertising by third parties is exactly the same. But it does not look the same, and therefore needs to be spelled out. The rationale for limits on third-party advertising is to keep the regime governing the candidates and the political parties intact. There are a couple of considerations here. One is that the electoral competitors (a small, finite number) whose financial hands are tied behind their back can hardly be expected to withstand an onslaught of independent advertisers (a potentially vast number). Under the legislative rules in place (but not enforced28), a third party (anyone who is not a candidate or a political party) could spend up to $1000 on advertising expenses. In his report on the 1997 general election, the Chief Electoral Officer, Jean-Pierre Kingsley, restated this consideration in the light of the unenforced third-party spending restriction:
In the long run, it can be expected that this situation, if not remedied, will erode the financial foundation of the electoral system. Both parties and candidates will feel at a disadvantage compared with third parties, who will be able to organize and fund their activities in the shadows without any limits on the expenses they may incur while pursuing their goals.29
A different consideration is the prospect of collusion between the electoral competitors and third parties. The thinking here is that third parties not subjected to spending limits are bound to be tempting targets for cash-starved electoral competitors. The competitors, it is assumed, would seek ways to get friendly third parties to mount advertising campaigns that are helpful to them. They would learn to collude with third parties on parallel campaigns that look independent of one another, but in fact are not. In this way, the electoral competitors would manage to get around the spending limits that are imposed upon them. Indeed this consideration was likely the major factor that led the Barbeau Committee back in 1966 to recommend strict limits on third-party expenditures. In effect the aim of the committeeâs recommendations was as much to protect the public and interest groups from possible predatory activities of parties and candidates seeking to circumvent spending limits as to protect parties and candidates from interest  groups.
It is evident from the considerations just reviewed that the spending restrictions on the electoral competitors and the spending restrictions on third parties form an integrated package. In effect, they support   one another. They also arguably infringe upon the expressive freedoms that are guaranteed in the Canadian Charter of Rights and Freedoms. Since elections are the democratic hallmark of modern liberal democracies, this is hardly a trivial matter. Therefore, securing these freedoms is also an important purpose of election law.
In Canada the clash between the value of fairness in the sense of equitable electoral competition and the fundamental freedoms of expression and association has centred on the restrictions on third-party advertising. Let us recall that in 1983, Parliament removed the good-faith defence that enabled third parties to claim that their advocacy advertising (on which there are no limits) had strayed unintentionally into partisan advocacy (at that time prohibited altogether). Outraged by the prohibition and the removal of this particular line of defence, one third party, the National Citizensâ Coalition (NCC), went to court to argue that the prohibition violated the groupâs rights of expression and association guaranteed in the Charter of Rights and Freedoms. Thus began the legal consideration of this matter which, in court after court, continues to this day.
Since the legal approach is an important ingredient of public policy, it is useful to review the relevant cases. In doing so, however, the architecture of the Charter must be kept in view. In essence, the rights and freedoms that are guaranteed in the Charter are subject to âsuch reasonable limits prescribed by law  as can be demonstrably justified in a free and democratic society.â Plaintiffs need to show that the legislation or executive action in question in fact infringes their guaranteed rights and freedoms. If they succeed in that, then it  is the responsibility of the respondent (the government) to show that the infringement is justifiable under the âreasonable limitsâ clause.
In each of the cases involving the limits on third-party advertising, the plaintiffs have been successful in showing that the limits are breaches of the freedoms of expression and association. The action has always turned on the effort of the government to persuade the court that the breach is justified. In the light of the decision handed down in the very first case brought by the NCC, this has entailed the need to demonstrate that third-party advertising is somehow harmful. In that case, Judge Medhurst of the Alberta Court of Queenâs Bench could find no compelling evidence that third-party spending was an abuse of the spending regime that needed to be curtailed. âThere should be actual demonstration of harm or a real likelihood of harm to a society value,â he opined, âbefore a limitation can be said to be justified.â30
Ever since, the âharmâ question has dogged governments that attempt to defend third-party spending restrictions in court. The Canadian government declined to appeal Judge Medhurstâs decision, and although technically it applied only in Alberta, in fact it had nation-wide application because the Chief Electoral Officer decided not to prosecute interest groups that defied the spending restrictions in the 1984 federal election campaign.31 The record of election expenses in the campaign revealed no unauthorized or partisan spending on the part of advocacy groups.32 However, the 1988 election was a different story. By Canadian standards there was massive third-party advertising in the âfree-tradeâ election, advertising that was directly related to the candidates and the political parties because their respective positions on the proposed free-trade agreement with the United States were so clearly articulated. By one estimation, third parties on both sides spent $4.73 million on advertising in the print media alone, a picayune figure by American standards, except that it was equal to 40 percent of the money spent on advertising by the three main parties in the election.33
The Conservative government that won the 1988 general election established the aforementioned royal commission, which undertook its wide-ranging examination of electoral matters and, on the issue of third-party advertising, took the view that an outright prohibition of partisan advertising was an unnecessary as well as unconstitutional denial of freedom of expression in any meaningful sense. Accordingly, the commission recommended that third-party advertising directed at a candidate or a political party in an election be restricted to $1000, no pooling among groups to be allowed.34 The government accepted the recommendation and made the change effective in May 1993, at which point the NCC, finding the $1000 spending limit as objectionable as the initial blanket prohibition, returned to court. The Alberta Court of Queenâs Bench was once again the choice of venue.
By this time the courts had been interpreting the Charter for a decade, and in aid of that enterprise they had developed the Oakes test35 as a way of formally assessing the validity of a governmentâs justification of the infringement of rights and freedoms. The first step of the test is to consider whether the objective of the legislation is sufficiently compelling and substantial to justify the restrictions in question. If it is, then the next stage of the test is to determine whether the restrictions used (the means) are rationally and proportionally connected to the objective (the end). Accordingly, Judge MacLeod applied the test to the federal governmentâs justification of the $1000 limit â the justification essentially being the need to promote fair competition among the electoral competitors, which was said to require restrictions on the competitors and third parties.36 But Judge MacLeod, like Judge Medhurst in 1984, could not accept the justification because he could find no convincing evidence that third-party spending matters. According to the testimony on the only two quantitative studies on the point before the court, he wrote, one of the studies was inconclusive, while the other study determined that third-party advertising had no effect on voter intentions.37
This time the federal government decided to appeal the trial courtâs ruling to the Alberta Court of Appeal, which dismissed the appeal in a ruling handed down in June 1996. In it Justice Conrad agreed with Judge MacLeod that there was no compelling empirical evidence to support the claim that restrictions on third-party spending are an essential part of a regime to promote fairness in electoral competition. More importantly for our purposes here, she wrote at length about the ramifications of the restrictions for the fundamental freedoms of expression and association, which freedoms, she noted pointedly, are guaranteed to individuals under the Charter, not to political parties. Radically, she concluded that the real purpose of the restrictions on third parties was to maintain an electoral system that âgives a privileged voice to [registered] political parties and official candidates within those parties.â38
Justice Conrad appeared to imply that in an election the electoral competitors are pitted against the voters. The individual voter, she imagined, faced with advertising from the candidates and the parties, wants a counterweight in the form of information and independent advice on the politicians from interest groups or community and religious leaders. She conceived these commentators to be objective, or at least âwithout the self-interest involved in candidate and party advertising.â39 From this standpoint, she found the noncumulative limit of
$1000 on independent spending to be too slight to sustain any meaningful counterweight.
Basically, Justice Conrad had defined third-party spending limits as unwarranted and even dangerous limits on free expression that leave voters defenceless against the onslaught of partisan campaigns. Many agreed with her, including of course the president of the NCC, David Somerville, whose reaction to the decision was understandably enthusiastic: âThey just picked up on all the arguments weâve used for years to argue against the gag law. This is beyond all our expectations.â40 The federal government appeared to throw in the towel, since it decided not to appeal the decision.41 Then the action shifted to  Quebec.
Quebec provides for the regulation of expenditures during provincial general elections and referendums. The general purpose of the regulatory scheme is the same as that of the federal scheme, namely, to ensure that fairness as conceived in terms of the competition between the candidates and the political parties is maintained as a central value of the democratic electoral process. As matters transpired, it was the third-party spending restrictions of the law governing referendums that were moving through the Quebec courts at about the same time that the corresponding federal provisions were being tested in the Alberta courts. However, by contrast with the Alberta outcomes, both the Quebec Superior Court42 and the Quebec Court of Appeal43 sustained the validity of the restrictions, which meant that the trial and appellate courts in the two provinces were at odds with one another. Since the plaintiff in the Quebec case, Robert Libman, chose to appeal the Quebec Court of Appealâs ruling to the Supreme Court of Canada, the judicial conflict was bound to be resolved one way or the other. And it was â in favour of the idea of restrictions, although not the particular ones in place in Quebec at the time.
Like all of the courts before it, the Supreme Court agreed that the impugned third-party spending restrictions were violations of the fundamental freedoms of speech and association. But it found the restrictions to be justifiable as part of a regulatory scheme designed to promote the objective of fairness within a democratic, electoral process. The Court quoted approvingly the words of Professor Peter Aucoin, who served as the Quebec governmentâs expert witness: âThe purpose of spending limits in an election or a referenda [sic] campaign is to promote fairness as a primary value or objective of the democratic process.â44 And herein lies the significance of the decision. Parsing this purpose, the Court stated that in providing for the control and the use of money, and thereby minimizing the impact on the referendum process of its uneven distribution among the members of the society, the legislation was egalitarian, and had the effect of keeping debate open and inspiring public confidence that the process is not dominated by the âpower of money.â45
What about the fundamental freedoms of expression and association? The courtâs attempt to balance them against the fairness objective surfaced in its rejection of the severe restrictions that Quebec placed on third-party spending in referendum campaigns.46 Ironically, the court recommended that the Quebec government consider the federal provisions â the $1000 spending limit for third parties that was recommended by the royal commission, adopted by the federal government, and dismissed by the Alberta Court of Appeal. Indeed, the court went to the trouble of citing in full the provisions of the Canada Elections Act that limit the advertising expenses of third parties to $1000.
The upshot of the Libman decision is that campaign finance regulation designed to promote the fairness objective is constitutionally justifiable so long as third parties are permitted to spend reasonable amounts on election advertising. The benchmark for what is reasonable was stated to be the federal limit  of $1000. However, this is not the end of the legal story. The opposition to limits on third-party advertising is intense in some quarters, as indicated by the flavour of the responses to the Libman ruling. David Somerville, the former president of the NCC, called it âdangerous, bad lawâ and then stated: âThe decision essentially, gratuitously, advises [Quebec Premier Lucien] Bouchard, [Prime Minister Jean] ChrĂ©tien and others how they can construct an election gag law which the court will find acceptable.â47 In the same newspaper on the same day, an unhappy editorial writer echoed the theme: âThe Supreme Court killed only the most extreme provisions of this ludicrously restrictive law. Worse for freedom of expression, it effectively revived a federal election-spending law that is nearly as bad as Quebecâs.â48
Opponents of such restrictions have fought and won a round recently in a British Columbia court room. In Pacific Press,49 the Supreme Court of British Columbia struck down the limits imposed on third-party advertising in that province, which, at $5000 for an individual or an organization, were among the most liberal in the country. Judge Brenner accepted that the objective of the spending limits is fairness in the electoral process. However, he did not accept the argument that fairness requires the imposition of limits on independent spending as well as limits on the spending of the candidates and the political parties. He suggested that the link between independent and partisan spending is a presumption only. â[I]s it valid to assume,â he queried, âthat the goal of fairness requires restrictions on third party spending in the same fashion as it requires restrictions on party and candidate spending?â50 And he answered his query: âIn my view the evidence in the case at bar does not support such a conclusion.â51 Instead of hard evidence, the government could offer only âtheoretical abstractionsâ and âunproved hypothesesâ about future possibilities.52
The Supreme Court of British Columbia followed the Supreme Court of Canada in accepting fairness as a valid reason for limiting campaign expenditures, but not expenditures that take the form of third-party advertising. The British Columbia court found such advertising to have no impact on voting at all, much less a harmful impact. That being so, the spending restrictions left intact in the province are the restrictions on the electoral competitors. There is no doubt that the legal chapter on third-party advertising restrictions is far from closed. Nevertheless, it is worth pausing to consider the tenability of pursuing the fairness objective through restrictions on the electoral competitors, while third parties are left free to spend as they  please.
The argument has already been made in court that there is no substantial or pressing need to restrict the campaign expenditures of the candidates and the political parties.53 Moreover, should third-party election advertising intensify in subsequent elections, it will not be long before the electoral competitors reach the conclusion that it is patently unfair to tie their financial hands, and no one elseâs. Or will they reach such a conclusion? Might there be another alternative? There certainly is â collusion. The electoral competitors might give tacit encouragement to well-heeled organizations that support them to run parallel advertising campaigns, campaigns that are friendly to them and harmful to their opponents. As already indicated, this was the very concern of the authors of the federal campaign finance regime. Restrictions on third-party advertising were considered essential to maintain the restrictions on the competitors. Is this still a valid consideration?
Obviously it is difficult to answer the question on the basis of Canadian experience, because for the most part comprehensive restrictions have been in place since 1974. And even when third-party advertising restrictions have not been enforced, the tradition of spending presumably is not as robust as it would have been had there been no such restrictions in place. However, the federal government has proposed new and far less onerous restrictions in Bill C-2, which may lead to significant increases in third-party spending in the future. In the meantime, the place to look is the United States, where questions about collusion have been raised in the presidential campaign this year.
It is important to stress that in the United States, individuals can spend as much as they like on advertising in elections so long as they are spending independently of the candidatesâ campaigns. Otherwise, their spending is a contribution, not an expenditure, and subject to the contribution limits established under federal law. However, the concern arises that such spending might tally a little too closely with the campaign of the candidate whom the independent supports. In other words, how âindependentâ is independent spending?
The issue arose in the bitter contest for the Republican partyâs presidential nomination between front runners Texas Governor George Bush and Arizona Senator John McCain. A notable example occurred during the campaign leading to âSuper Tuesdayâ (March 7), when twelve states held Republican primaries or caucuses that together would produce 588 of the total of 2,067 delegates to the Republican National Convention in August at which the partyâs nominee is chosen. In New York, a private group purchased an ad that was run state-wide in which Senator McCainâs environmental record was savaged. McCain contended that the ad was co-ordinated by the Bush campaign. He was quoted as saying: âHereâs a guy [Bush] who gave waivers to 85 polluters, all of whom gave to his campaign. Somebody is putting in $2-million to try to hijack the campaign here in New York.â54 And he continued: âItâs everything Iâve been fighting against. Two million dollars in the last few days in this campaign can make a difference in a race that is a statistical dead heat.â55
The same issue also has arisen in somewhat different form in connection with the political parties and the candidates, and was the subject of a Supreme Court decision in Colorado.56 In 1986, and before either political party had selected its candidate for the United States Senate seat that was open that year, the Federal Campaign Committee of the Colorado Republican party purchased advertising to attack the individual who was expected to be the Democratic partyâs choice. The FEC brought suit against the committee on the ground that the advertising was co-ordinated rather than independent, and therefore amounted to a contribution to a congressional candidateâs campaign (albeit one yet to be formally chosen). Being a contribution, it was subject to limits that had been breached. The Supreme Court majority found that there was no factual basis for the claim that the advertising was co-ordinated. Instead, it ruled the advertising to be an independent expenditure entitled to first-amendment protection and therefore subject to no limits at all.
The Colorado case is relevant because it suggests that the distinction between âindependentâ spending and co-ordinated spending is not a meaningful one in practice. Thus even though he agreed with the outcome of the decision, Justice Thomas reached the conclusion that âthere is no constitutionally significant difference between campaign contributions and expenditures.â57 According to him, contributions and expenditures are both forms of speech:
Whether an individual donates money to a candidate or group who will use it to promote the candidate or whether the individual spends the money to promote the candidates himself, the individual seeks to engage in political expression and to associate with like-minded persons. A contribution is simply an indirect expenditure; though contributions and expenditures may thus differ in form, they do not differ in substance. As one commentator cautioned, âlet us not lose sight of the speech.â58
Of course, if contributions and expenditures both are forms of speech, then they are protected by the first amendment, and any curbs on them are subject to the strictest scrutiny. On Justice Thomasâ scrutiny, there ought to be no  curbs: âI am convinced that under traditional strict scrutiny, broad prophylactic caps on both spending and giving in the political processâŠare unconstitutional.â59
While the substantive distinction between expenditures and contributions might be dubious, the legal consequences of the distinction in the United States are critical, since independent spending there is subject to no limitation, while co-ordinated spending, being considered a contribution, is. And collusion has the effect of transforming what looks like independent spending into co-ordinated spending.
Readers might well question the relevance for Canada of the difficulties that the Americans have in maintaining a distinction between contributions and independent spending. The answer is the concern about creating a situation that invites collusion to get around the law. It cannot be repeated too often that, from the standpoint of the electoral competitors, there is never enough money to prosecute a campaign. They can always use more, so long as it is applied effectively on their behalf and against their opponents. It is a short step from welcoming independent spending on oneâs behalf to tacit encouragement that such spending follow a strategic course. Such was the essence of Senator McCainâs complaint, cited earlier. Of course there was another aspect to his complaint, that being the effectiveness of independent spending. He assumed that it was effective, and he did lose the New York Republican primary to Governor Bush. In Canada, however, there is considerable debate about the effectiveness of third-party advertising. Can such expenditures move the vote in one direction or another? Have they done so in the past? Presumably if such expenditures have little effect, the concerns about regulating these expenditures and about collusion pretty well disappear. Thus it is essential to consider the effects of third-party spending.
The question needs to be situated in the more general literature on the impact of campaigning and political advertising. It is a topic that has been studied for more than half a century, beginning with the first modern voting behaviour studies by Paul Lazarsfeld et al. in the 1940s and 1950s and later by Angus Campbell et al. in their classic study The American Voter.60 Among their findings was the observation of little actual change in the preferences of voters during the course of a campaign. As David Butler and Donald Stokes note in Political Change in Britain, the impact of campaigns themselves tends to be limited: âTime and again it has been shown that relatively few votes are changed and that these are largely in mutually cancelling directions.â61 Butler and Stokes were working within what is referred to as the âparty  identificationâ paradigm, which assumes that linkages between voters and parties are relatively enduring and unchanging. In an era when levels of party identification in several countries were high, levels of support for the major parties remained relatively stable. However, even before the era of voter dealignment and electoral volatility that arrived in the 1970s and 1980s, Butler and Stokes noted that although campaign effects are considered to be marginal when parties are evenly balanced, small shifts can be decisive. Secondly, they noted that the assessment of campaign effects was usually made when both parties were running active if not similar campaigns. What would happen, they asked, if one of the parties unilaterally decided not to wage a campaign? The answer, they noted, could well be defeat, as happened in Thomas Deweyâs non-campaign in the 1948 American presidential  election.62
With the arrival of dealignment and increased electoral volatility in the 1970s, the party identification model fell into disfavour and voting studies increasingly emphasized the importance of party leaders, party policy, and campaign efforts as determinants of electoral outcomes. It was in this period that analysts began taking an interest in the nature and volume of campaign communications, including political advertising. The interest manifested itself in part in the form of analyses of the impact of political issues and leadership debates. It also appeared in the form of examinations of the role of advertising, a matter that was coupled with the concerns on the part of some parties and politicians about the role of money in politics and the possible need to restrict political expenditures. Much of this work was done by economists who saw the study of the impact of advertising as an extension of marketing studies. Thus one of the earliest Canadian studies was done by an economist, Kristian Palda, who used data from the Quebec provincial elections of 1966 and 1970.63 These elections were among the first to be held under electoral law requiring full disclosure of expenditures by parties and candidates. Somewhat to his surprise, he discovered that election expenditures did have an effect: âGuarding against simultaneous equation and specification biasâŠ, the [regression] estimates showed advertising expenditures in various forms (as well as other campaign expenditures) to be a powerful conditioning variable.â64
These earlier studies were focused mainly on constituency or district level contests and relied on aggregate data. Like Paldaâs pioneering effort, these studies uncovered largely positive effects, although the effects were not necessarily straightforward. That is to say, the expenditure of additional moneys did not necessarily have the same effect for different types of candidates or in different contexts. Increasingly, as Ansolabehere et al. have noted, researchers were moving away from the âhypodermic model,â which assumed that an increase in advertising would have the same independent and linear effect for all candidates. Instead, there was increasing emphasis on context, the manner in which a particular campaign âresonatedâ with other forces, and the nature of strategic interactions among candidates.65 For example, Gary Jacobson noted the paradox that while additional expenditures appeared to help challengers, the same was not true for incumbents.66 Indeed the latter were likely to render themselves worse off by spending more money, a finding that was not unimportant for resolving the debate as to whether limits should be placed on election expenditures. It seemed that limits would favour incumbents over challengers and thereby reduce the likelihood of turnover in the legislature, an outcome that could be construed as less rather than more desirable.
It was recognized, however, that assessing the effects of campaign expenditures was no easy matter, given that expenditures are merely one among a number of factors that determine electoral outcomes. The task at hand is in good part one of disentangling the effects of these separate factors and assigning causal weight to them. In addition to controlling for exogenous factors there is also the issue of endogenous factors, that is, those factors internal to the relationship. In the case of the Jacobson model, for example, Green and Krasno argued that it failed to take into account the dynamics of the perceived need for increased expenditures by the incumbent in relation to the quality of the challenger: often incumbents would spend more not in order to increase their advantage but in order to fight off an effective challenger.67 What was being measured, therefore, was not so much the impact of money but the effectiveness of the challenger. Subsequent and more sophisticated work, using multi-equation rather than single equation models (as originally used by Jacobson68), have been better able to capture the effects of these endogenous variables. Thus Kenny and McBurney, using both a multi-equation model and individual level data were able to show that, in their model, expenditure terms for both challenger and incumbent were significant.69 Also, through the use of individual level data they were able to explore the pathways through which the influence of expenditures travelled and where it had the greatest effect. They found that it was individuals lacking higher education and having weak convictions who were most likely to be influenced by campaign expenditures.
Munroe Eagles (1993), using Canadian data and multivariate models incorporating a variety of political and socio-demographic variables, also found that campaign spending by local parties and candidates contributed âsignificantly to explaining patterns of voter support.â70 In the 1984 and 1988 elections he found that additional expenditures were more likely to help NDP and Liberal candidates rather than Conservative candidates. In more recent work on US presidential elec-tions, an arena where it was thought that television advertising and other campaign-expenditure activities have only limited effects, it has been shown that increases in expenditures and campaign appearances at the state level by presidential candidates can affect statewide voter preferences.71 Other recent research has explored the effects of advertising, particularly negative advertising, on dependent variables beyond voting preference. Thus Ansolabehere et al., using an experimental design, found that negative advertising was correlated with both a drop in vote intentions (turnout) and an increase in political cynicism.72
In brief, the research to date on campaign expenditures in a variety of countries indicates that such expenditures can have a powerful impact. At the same time, although campaign expenditures can help mobilize or âmove the voteâ they will rarely do so on their own. Typically such expenditures work in conjunction with other factors such as the quality of the candidate. It is difficult if not impossible for a weak candidate to simply âbuyâ an election through extraordinary expenditures. It should be further noted that most research is focused on expenditures made or controlled by the candidates and their organizations. Notably lacking in the general literature, primarily American in focus, are studies of the phenomenon here under examination, namely independent or third-party expenditures. The findings that do exist, however, both Canadian and American, are quite revealing. And given the centrality of independent expenditures on both sides of the border it is worth looking at them quite closely.
Except for a very recent study, to be noted later, the few studies on thirdparty expenditures that exist are Canadian. The lack of such studies in the United States became evident in the 1993 Somerville case noted earlier involving the NCC. One of the expert witnesses for the NCC, Neil Nevitte of the University of Calgary, was only able to cite one study based on American experience by Page et al.73 in his review of the literature. This study subsequently played a prominent role in helping the court decide that third-party expenditures had no effect on voters, despite the fact that the study bore only on the role of special interest groups in agenda setting. Indeed, the study made no reference whatsoever to voting behaviour. Presumably the intention was to have the court draw the inference that if positions on specific issues promoted by special interests tend to be viewed in a negative light by the general public, the same would hold true for political parties or candidates endorsed by those interests. By implication, therefore, third-party advertising will tend to be counter-productive, that is, moving voters in a direction other than what is intended. Whether the court was  willing to make this inferential leap or whether it was even aware that such an enormous leap was being made, is not clear.74
The fact that this was the only American study that Nevitte could find to cite at the time is reflective of the fact that in the US third-party expenditures have been, until recently, of only limited importance, both in terms of volume and in terms of effectiveness. While one can find individual instances of interest groups becoming directly involved in specific election campaigns, taken as a whole such spending as a proportion of all election spending is miniscule, at least up to the early 1990s.75 It is only during the recent primary campaigns that the issue has really come to the fore. Where such spending has occurred in the past, as in the 1982 elections when a number of liberal US senators were targeted by the National Conservative Political Action Committee (NCPAC), it has been decidedly limited in effectiveness. In this instance, as Jacobson notes, only one of the nine targeted incumbents was defeated, someone who was already in difficulty for other reasons.76 Furthermore, âthere is considerable evidence that NCPACâs tactics actually backfired in 1982 as targetted candidates used the specter of NCPAC to raise money and rally support.â77 It is also important to distinguish actual independent third-party expenditures from the contributions made by PACs to the war chests of candidates, a much more common activity. While such contributions may well influence the candidates in question, either during or after the conclusion of the campaign, election spending would still be under the control of the candidates and their  organizations.
In Canada, as noted earlier, the nature of the electoral law regime with its stress on expenditure limits has ensured that the third-party expenditure issue has been salient for some time. This was particularly the case in 1988 with what has been called the âfree trade election.â More so than in most previous elections the electoral contest revolved around one major issue â the free trade agreement (FTA) with the United States negotiated by the incumbent Conservative government. For many groups and individuals outside the political parties the stakes were seen as sufficiently high to warrant becoming directly involved in the election campaign through paid advertising promoting or opposing the FTA. There were groups on both sides, though according to Janet Hiebert the side favouring the FTA, mainly business groups, outspent the opponents approximately four to one.78 The 1988 election was also the subject of a careful study of the effects of election advertising by the Canadian Election Survey team, using, for the first time, rolling polling data garnered through telephone interviews. Two studies arose out of this survey that attempted to examine the effects of third-party advertising. First, Richard Johnston, the lead researcher of the 1988 team,  produced a memo79 for the Royal Commission on Electoral Reform and Party Finance in which he suggested that âthird-party expenditures might have helped boost Conservative vote intentions late in the 1988 campaign.â80 These effects appeared most pronounced amongst those already favouring the FTA. At the same time, Johnston entered a number of qualifications, including the possibility that âother, as yet unspecified variables might lie behind the impact attributed to third-party advertising.â81 Johnstonâs 1990 memo subsequently constituted important evidence for Janet Hiebert in reaching her conclusion that the âprincipal effect of the advertisements [by third-parties] was to mobilize those Free Trade supporters intending to vote for the Liberal party to support the Conservative party.â82
Later, however, Johnston and his fellow researchers backed away from the tentative conclusions that Johnston had reached in 1990. Taking into account the effect of party advertising, as well as news coverage, the four researchers came to the conclusion that âthe net impact of third-party FTA advertising was essentially nullâ83 or, as they put it in the more technical language of statistics: âthird-party advertising coefficients defy substantive interpretation: some are large and significant but the pattern is offsetting and the total coefficient effectively zero.â84 This phrase, however, lends itself to rather different interpretations. Neil Nevitte, in explaining this conclusion for the court in Somerville, states it means there were no effects, period. By contrast, a rather different interpretation was put forward by Bakvis and Smith in their analysis of the Somerville decision. They argued that âit was not a matter of no effects but, rather, no cumulative effects.â In other words the positive and negative effects of third-party FTA advertising essentially cancelled each other out. Furthermore, they note that âwhile it is true that the data disclose nothing firm about the effects of third-party advertising, still from a social science perspective it would be unwise to reject altogether the possibility of such effects.â85
Johnston himself later offered testimony that seems to provide some support for this position. In his written submission to the British Columbia court in Pacific Press, Johnston, acting as an expert witness for Pacific Press, found Nevitteâs characterization of the Johnston et al. findings âto be succinct and to the point,â by implication rejecting the Bakvis and Smith argument.86 However, under cross-examination Johnston conceded that another interpretation was possible â a perverse effect wherein âpositive third-party advertising of the FTA reduce[d] Conservative vote share.â87 This admission did not sway Judge Brenner from his assessment that the essential thrust of Johnstonâs testimony was âthat third party spending had no effectâ and that this âwas not challenged by any other empirical study or the testimony of any other expert witness  who testified at trial.â88 But given that perverse effects , in the judgeâs words, amount to âone possible reading of the data,â it is worth stating that even perverse effects â that is, third-party pro-FTA advertising hurting rather than helping the pro-FTA Conservatives â are effects nonetheless. The perverse-effects interpretation also appears to represent a confirmation of sorts of the supposition originally put forward by Nevitte in the 1993 Somerville case, namely, that communications from special interest groups are often viewed in a negative light by the general public. In other words, it might have been the case that voters, upon being exposed to third-party advertisements favouring the FTA, saw them as self-serving, suspicious, irritating, or all three, and thus voted in   a direction opposite to that intended by the  ads.89
This line of argument, the âNevitteâ argument if we can call it that, receives further corroboration in a more recent study by Tanguay and Kay based on a comparison of the 1993 and 1997 elections.90 The 1993 election was relatively quiescent as far as third-party activity is concerned while in the 1997 election groups such as the National Citizensâ Coalition were active in a number of ridings in endorsing or opposing specific candidates. Essentially the authors compared the swing for or against candidates targeted by three groups â the National Citizensâ Coalition, Campaign Life, and Catholic Insight â with the mean province-wide swing for the parties of the candidates. According to Tanguay and Kay, the effects of the advertisements of these three organizations were negligible. The candidates blessed with positive endorsements from Campaign Life and Catholic Insight did slightly worse and slightly better respectively, while the candidates opposed by the National Citizensâ Coalition actually did about five per cent better compared to the provincial swing. The study lacked controls and other refinements. As the authors note, a number of other organizations, such as the Canadian Police Association, were active in many of the same ridings targeted by the National Citizensâ Coalition, but no efforts were made to disentangle the effects of the different groups. In the case of the five per cent swing in favour of the candidates targeted by the NCCâs âOperation Pork Chop,â the authors discount this result, speculating instead that an American style âsophomore surgeâ among other factors might have been at work in several ridings. No effort was made to test for such an effect, however. Tanguay and Kay use their finding of essentially trivial effects, or at least trivial relative to the effects of incumbency and party campaigning, to argue that the problem of third-party advertising in elections is vastly âoverestimated.â91
Given the lack of controls in the Tanguay and Kay study, it is probably best not to take speculation based on their findings too far. Nonetheless, the five per cent swing in favour of candidates targeted in the NCCâs negative campaign   is intriguing insofar as it fits with both the Nevitte argument and the Johnston et al. interpretation of their 1988 data (as described by Johnston in his cross-examination in the B.C. case), namely that third-party advertising can lead to perverse effects. It also fits with the commentary of a key campaign strategist who claims that pro-FTA advertising probably hurt rather than helped the Conservative cause.92 Particularly in the light of Johnstonâs testimony, we would want to be careful not to dismiss prematurely the five per cent perverse effect uncovered by Tanguay and Kay. At the same time, it is also important to look at the most recent American study of the phenomenon, a study that suggests independent expenditures by a sophisticated and well-funded organization can move votes in a desired direction.
As noted earlier, in the US the vast bulk of election expenditures tend to be made by parties and candidates, drawing in good part on contributions from PACs. So-called âsoft moneyâ that is used for unregulated expenditures is another avenue that various special interests can pursue to funnel money into election campaigns. Still, limitations in the form of contribution limits exist, and given that the actual dollar limits have not changed since the 1960s, many groups, as well as candidates and parties, are beginning to find them so constraining to the point that in recent years some have begun exploiting the thirdparty route. The 1994 congressional elections stand as a turning point, with groups in the name of âpolitical educationâ making independent election expenditures, while in reality pursuing partisan objectives. The former Speaker of the House of Representatives, Republican Newt Gingrich, is generally regarded as having pioneered the technique, his efforts culminating in the successful 1994 congressional campaign, in which Republicans gained a majority in the House for the first time since 1954. Not surprisingly, the line between efforts to promote political education, like Gingrichâs college courses and the Progress and Freedom Foundation, a conservative think tank associated with him, and outright political action committees, like Gingrichâs GOPAC committee, is easily blurred.93 Challenges by aggrieved Democrats, arguing that these educational campaigns were little more than legal fictions designed to circumvent the limits on contributions imposed by legislation, found little favour in the courts, which essentially confirmed the legality of these educational activities. Two years later the tables were turned, however, as one of the Democratsâ biggest allies, the AFL-CIO, decided to launch its own âvoter educationâ campaign. This campaign, involving substantial sums of money, targeted several Republican candidates. Given its scope and the sizeable amounts of money involved, this educational campaign also served as an ideal natural experiment. Gary Jacobsonâs assessment of the role of the AFL-CIO in the 1996  congressional elections constitutes the first serious study in that country of direct third-party expenditures.94
The AFL-CIO targeted 64 House Republicans, using one of two approaches. The first approach, âvoter video guides,â involved âTV ads that made explicit comparisons between the issue positions held by the Republican and his or her opponentâ and also came closest to looking like regular political ads. It was considered the heaviest weapon in the AFL-CIOâs arsenal. The second type of âeducationalâ spot merely âexposed the Republicanâs voting record but did not mention the challenger.â95 The analysis drew distinctions between freshmen and senior Republicans as well as between the two different kinds of advertisements. The results were striking. Of the 44 Republican freshmen who were targeted, 12 (27 percent) lost; of the 27 not targeted, all were successfully re-elected. Those freshmen Republicans subjected to the video guide were much more likely to lose compared to those targeted by the milder ads. Using an OLS regression model, incorporating controls for amounts spent by both incumbents and challengers and whether the challenger had held elective office, among other factors, Jacobson also discovered that AFL-CIO targeting had virtually no impact on senior Republicans. In other words the third-party effect was restricted to those districts held by freshmen incumbents. Amounts spent by and the level of experience of challengers were also important factors. For example, none of the challengers spending less than $700,000 were successful, with or without help from the AFL-CIO.
To summarize to this point, our analysis of the limited number of studies suggests that in the Canadian context third-party campaigning efforts tend to have a perverse effect, that is, the outcome is opposite to that intended by the group in question. The reasons for this are not clear. It could be, as Nevitte argues, that messages from special interest groups tend to be seen in a negative light by voters who then act accordingly at the ballot box. It could also be that the relative inexperience of such groups leads them to use inappropriate strategies â ads high in vitriol and low in warmth, for example. The reliance by third-party groups on print media, bill boards and the like and their lack of access to the medium of television may also play a role, but this would tend to limit their impact altogether as distinct from having an effect, positive or negative. It is also worth keeping in mind the possibility that the primary aim of groups such as the Canadian Police Association or the NCC may not necessarily be to move voters in a particular direction. Groups such as these have complex constituencies and memberships, and their advertising campaigns may be crafted largely with a view to demonstrating to their membership that they are capable of taking a tough line, regardless of whether this tack may turn off  the broader voting public. In other words, one should not automatically assume that all groups behave rationally in the sense of wanting to persuade citizens in a particular way. For some groups, at least, the public statement of their positions appears to be the primary motive.
Yet the very possibility that lack of experience may account for the pattern in Canada so far suggests that it would be unwise to assume that Canadian third-party advertising will always tend to have either no effect or a perverse effect. The study by Jacobson of the role played by the AFL-CIO in the 1996 House elections in the US indicates that where an interest group is quite serious about moving the vote, has considerable resources, and has access to the medium of television, the results can be quite dramatic. Tanguay and Kay argue that âinstitutional differences between the Canadian and American political systems,â which tend to link constituency-based vote swings to national or regional swings, will always limit the opportunity for interest groups in Canada to âcherry pickâ and target specific ridings.96 Yet the case of the AFL-CIO involved more than cherry-picking of specific districts. The campaign succeeded in a fairly broad, albeit carefully defined, category of districts, namely those held by freshmen Republican incumbents.
Tanguay and Kay also suggest that in Canada where most ridings have three or four parties competing for votes, the dynamics and effects of thirdparty spending differs from those prevailing in a strict two-party system. That is to say, with three or more parties votes that are lost to a candidate because of negative advertising by an independent group will not automatically flow to the candidate favoured by that group. In other words, the effects are much less predictable. But this argument would not necessarily hold in the case of advertising that appears to be positively linked to one of the candidates or parties, that is, if it looks like it came from, or appeared very similar to, the advertising from the party or candidate itself. In other words, if there was collusion between the third-party and a political party or candidate, the effects could well be different.
In general, Jacobsonâs study suggests that with sufficient resources (estimated to be $30 million in the case of the AFL-CIO) and the right formula (ads that looked like political advertising, a particular class of incumbents), a campaign national in scope can succeed. In other words, although specific conditions and factors in Canada differ from those in the United States, such differences do not rule out the possibility that a well-resourced group, serious about moving the vote and operating on a regional or national basis, and quite possibly working in collusion with a political party, will succeed in finding a formula that proves to be effective.
In Canada most of the debate, and virtually all of the court cases, have centred around the notion of electoral fairness and how third parties, as independent entities, may alter the balance between parties and outside actors. What tends to be neglected in this debate is the other important dimension, which sees third parties not as competitors or opponents of political parties but as potential colluders. That is to say, interest groups may have as their objective the successful election of a specific candidate or political party, which they will then pursue by paying for advertising outside of the official campaign of the party or candidate, but which will nonetheless be similar to, or co-ordinated with, the advertising of the party or candidate. Conversely, a candidate or party may seek out the support of interest groups as a way of being able to put more resources into their election campaign.
Up to this point, there has been little or no evidence of collusionary activity between third parties and political parties along the lines, for example, of the AFL-CIO and the Democrats in the United States. This is due in part to the fact such behaviour in Canada is expressly forbidden by law and in part to the fact that the restrictions on third-party advertising have not been enforced since 1984. Also, as we have argued, it appears that third parties in Canada are still  at an early stage in developing effective advertising strategies. Once they find their feet, so to speak, and realize that effective advertising often involves coordinating their strategy with that of the parties or candidates they intend to support, the dynamic of electoral politics with respect to the role of third parties might well change.
Collusionary activities between parties and outside interests are potentially far more harmful to the integrity of the electoral finance system than independent third-party expenditures. Collusion raises the spectre of influence peddling â witness the current travails of Helmut Kohl and the CDU in  Germany
â the role of offshore money, and, generally, the role of big money in politics. In effect, it means the end of effective expenditure limits in election campaigns. Given the party-centred nature of Canadian electoral politics, collusion is also more likely to occur between political parties and outside interests rather than between individual candidates and outside interests. Again, it is just as likely to involve parties approaching outside interests as vice versa. In the case of the former, it will likely mean the strengthening of parties vis-Ă -vis outside groups.
To be sure, as Justice Conrad noted in the Somerville case 1996, such collusion is still illegal under Canadian law, even with the suspension of limits on third-party expenditures. Yet  the incentives for parties and interest groups to circumvent the law will be very strong. Much like attempting to prove price fixing practices among businesses under competition law, prosecuting and obtaining convictions for collusionary activities are very difficult pursuits. Further, the current limits on the parties and the candidates may come under legal challenge should limits on third parties be removed, the argument being that the parties and candidates, themselves constrained by strict limits, would face unfair burdens if confronted with unrestrained independent spending by third parties. A Charter challenge along these lines would not be implausible.97
The critical question for the moment, however, is to what extent the provisions on spending limits in the new Canada Elections Act, Bill C-2, affect the role of third parties and the possibility of collusion between parties or candidates and outside interests. Bill C-2 contains some intriguing changes to the independent-spending regime as part of a comprehensive package of amendments to the countryâs electoral law. The largest change is the proposal for a spending limit of $150,000, of which no more than $3,000 could be spent in each constituency on election advertising for or against the candidates. The full
$150,000 could be spent on a national or regional advertising campaign for or against a particular political party. Third parties that spend more than $500 on election advertising would be made subject to registration and reporting requirements, including the requirement to report donations amounting to more than $200, dating back to six months before the election is called. The expenditure limits are vastly more liberal than before, and certainly offer an opportunity for interest groups to participate in elections by mounting effective advertising campaigns.
The first point to note with respect to the new legislation is that it will very likely be challenged in the courts, notwithstanding the much more generous limits on independent expenditures. The NCC, among other organizations, has given every indication that it will mount a challenge. Certainly the victory for the independent spenders in Pacific Press this year is a strong incentive for such a course of action. Nonetheless, in light of the Libman decision there are grounds for thinking that the new restrictions in Bill C-2 would survive a judicial challenge. And even if the proposed restrictions were not to make the judicial cut, surely the disclosure provisions would pass judicial muster, precisely because the restrictions are so generous. Disclosure provisions are a necessary condition of any kind of meaningful regulation of matters such as barring the use of foreign money in election campaigns. For a court to decide that it would be inappropriate for the Canadian government to bar the use of foreign money in Canadian elections is almost unthinkable.
The final point to note is whether the proposed limits are high enough to invite collusion. It is always difficult to predict future outcomes. Such outcomes are in good part dependent upon the learning and experience of those participating in the new regime. For example, while we have discussed the possibility of collusion between the electoral competitors and independent spenders, there is always the possibility of co-ordinated campaigns among the independent spenders themselves. Three or four business organizations that spent the maximum permitted to them could mount an effective advertising campaign with minimal coordination effort. In the meantime, there remains the fact that any limit is bound to be challenged by an organization like the NCC. Given the nature of this complex problem, and the passionate beliefs it invokes, the federal governmentâs efforts in Bill C-2 to strike a new balance can be seen as a reasonable step forward. In short, although the risk of collusion is definitely there, it may be a risk worth taking in the effort to generate a policy that is acceptable to all sides.
Jennifer Smith is a Professor of Political Science at Dalhousie University. She writes about various aspects of the electoral system, and also on issues related to parliamentary government, federalism, and Atlantic Canada.
Herman Bakvis is a Professor of Political Science and Director of the School of Public Administration, Dalhousie University. He publishes in the areas of intergovernmental relations, policy development and electoral law.