At the time this article was written James Girling was  a lawyer with the Ontario Ministry of the Attorney General and on secondment as Legal Counsel to the Cabinet Office. From July 1997 to June 1998, he was Legal Adviser on the Ontario Referendum Project and, in that position, was responsible for the development of the legal policy and the draft legislation used in the Government’s public consultation on provincial and municipal referenda in Spring of 1998. This article is based on his presentation to the 37^th Canadian Regional Conference held in Toronto from July 18 - 24, 1998.

This article looks at arguments for and against the use of referenda as a way of keeping governments accountable. It concludes that there is no simple answer to the question as to whether referenda makes governments more accountable but outlines a number of factors that should be considered when deciding if a referendum is an appropriate public policy instrument.

In the British context, the first modern proposal for a national referendum came from the great constitutional expert, A.V. Dicey who in 1890 advanced the idea of a referendum on the Government’s policy of Irish Home Rule. He believed that the majority of the voters would oppose it. No referendum was held, but as history has shown, the issue never really went away.

Dicey believed that a referendum could and should be used to make the Government more accountable, at least on this policy. To the extent that Dicey believed that constitutional change was at issue, he was the harbinger of the use to which referendums have most often been put in the Commonwealth, that is, to determine how the people are to be governed. Examples of constitutional referenda are legion:

• The 1975 referendum in the United Kingdom on whether or not the U.K. should remain in the Common Market.
• The 1979 referendums in Scotland and Wales on the devolution of legislative authority from Westminster, referendums which were repeated with different results within the last year.
• The 1992 referendum in Canada on proposed changes to the national constitution
• The 1995 referendum in Bermuda on independence from Britain, and
• The 1998 referendum in Ireland and Northern Ireland on the peace agreement

Even those referendums which are not strictly constitutional in nature tend to be on matters of high political, and often moral, emotional or psychological importance for the electorate, and hence for the elected.

In Canada, the only two other direct national votes on issues that did not strictly fall into the constitutional category, but were none the less issues considered to be of high significance in their day were the 1898 and 1942 votes on prohibition and conscription, respectively. They may not at first blush appear to non-Canadians to have the same degree of controversy as the 1981 Italian and 1993 Irish referendums on abortion, but in the greater scheme of things may seem to be of more significance than local referenda on the fluoridation of water supplies or on liquor licensing within municipal boundaries, as is provided under legislation in Ontario.

All these examples do, however, have one thing in common; they arose from attempts by politicians of different political stripes and backgrounds to give their electorate a direct vote on issues considered by the voters to be of fundamental importance to the way they live. This direct voting on an issue has given rise to the description of referenda as “direct democracy”. It was the Government which chose, in each case, to make itself more accountable by submitting to the electorate a policy decision on a specific issue. What those Governments did not do was to submit all issues to popular referenda. For, as both Canada and Quebec have learned, the mere fact that a Government has called a referendum, has campaigned in favour of the proposal and even expects to win the vote is no guarantee of success. When a Government is left with the choice of what issues to submit to voters, it still controls the agenda to a degree.

But, referenda at the discretion of Governments are not the only type of referendum models which are currently in use. In addition, there are mandatory referenda where, by ordinary statute or constitution, referendums must be held in specified circumstances, such as:

• Approval for changes to the constitution, as in Australia, Ireland, British Columbia and Alberta;
• Prior approval of the raising of public money by way of issuing bonds, as in many local and state US jurisdictions;
• Prior approval to the introduction of certain types of bills, as under the Alberta Taxpayer Protection Act in relation to the levying of provincial sales taxes and in Manitoba in relation to increases in certain taxes;
• Prior approval before a law already enacted may come into force, in effect ratification of a law, as in certain US jurisdictions;
• In response to a citizens’ petition to enact law (often called an initiative), as in British Columbia and Saskatchewan;
• In response to a citizens’ petition to amend a law already enacted by way of initiative, as in California where, because a positive vote in an initiative actually creates substantive law, that law may only be amended by further initiatives;
• In response to a citizens’ petition to recall or remove from office an elected representative before the end of his or her term, as in British Columbia and Kansas.

The British Columbia example is perhaps unusual in a Commonwealth context because the recall mechanism appears to have its greatest appeal in jurisdictions which, like the US states, have governments of fixed terms. This is distinguishable from governments in the British parliamentary tradition whose terms are subject to votes of no confidence to test popular support for the government before the end of its maximum allowable term under its constitution.

In recent years, there has been movement by some provincial governments to adopt laws setting out the conditions under which referendums are to be held. Initiative laws in both British Columbia and Saskatchewan and provisions for mandatory referendums on tax issues in Alberta and Manitoba have been followed in Ontario by the release of a 1998 government consultation paper on proposals for provincial and municipal referenda.

Some might argue that these efforts to make governments more accountable to their electorates through referenda are attempts by politicians to counter what is perceived to be, rightly or wrongly, a growing cynicism and dissatisfaction with government among voters. But, if that is the objective, the question arises what makes the referendum exercise credible. With so many referenda having been held around the world and with so many referendum laws on the books, there are a number of issues which may need to be addressed before assuming that referendums can automatically make governments more accountable. Here are a few.

The subject matter of the referendum: Effective referendums are those that deal only with matters over which a government has power and authority. Other structures of government still need to be preserved, such as maintaining the division of powers between the legislative and the judicial functions. Therefore, it is arguable that referendum questions should not be allowed to interfere with judicial or quasi-judicial decisions or decision-making.

The form of the question: The usefulness of a referendum question in governmental decision-making is fundamentally linked to the question itself. Referendum questions need to be neutral, clear, concise and capable of being answered definitively by the answers proposed. A question that is inherently biased or attempts to collapse too many issues into one question could be misleading. Such a question might not only produce ambiguous results which may be hard, if not impossible to implement, but could also undermine the validity and credibility of the referendum process itself. For example, questions having multiple choice answers have the potential for splitting the vote, with no one answer attracting a clear majority of votes cast. Therefore, it is essential that questions be vetted to eliminate these and other structural problems from the very beginning of the process, before too much time and public resources are devoted to them.

The desirability of resolving conflicts between questions in advance of the vote: There may be occasions when a question which meets the criteria for clarity and simplicity is placed on the same ballot as another question on the same or similar topic. As a result, the two questions may produce results which are in conflict with one another or which lead to confusion. This is especially an issue in the case of citizens’ initiatives. In such circumstances, it is important to have a means of resolving the problem, whatever it may be in advance of the vote.

Confidence in the running of the vote and its tally: Often votes have been criticised as being in the control of the government which is not neutral on the outcome. Therefore, it has been found desirable to have an independent office responsible for arbitration and administration of the referendum.

Thresholds of support before holding an initiative vote: One of the challenges of allowing citizens to initiate their own referendums is to ensure that taxpayers’ money is not spent on frivolous proposals or matters of limited interest to the community as a whole, while maintaining an accessible process (i.e. not creating too high a threshold to initiate a referendum). For this reason, the establishment of a minimum threshold of support as evidenced by the collection of the signatures of a certain percentage of the eligible voters may be advisable. Given the range in size and diversity of most jurisdictions, it may be that a simple percentage threshold does not accommodate that diversity, and that other factors, such as geographic or electoral distribution, should be considered in determining whether or not a particular minimum threshold of support is sufficient to require a referendum to be held.

Campaign rules: There may need to be a balance between allowing anyone to initiate a referendum with as few restrictions as possible and preventing persons with unlimited resources to buy signatures in order to advance their own interests.

Given the expenditure of public money on a referendum as well as the seriousness of the consequences of a successful vote on a question, certain safeguards may need to be an audit of the petition to ensure not only that there are the minimum number of signatures, but also that the signatures are valid and those of eligible voters.

A carefully crafted, verifiable process is necessary to ensure that the rules are neither flouted nor abused and that, when the rules are breached, appropriate sanctions can be levied.

Threshold for implementation of the result: One of the differences between an ordinary election and a referendum is that the referendum produces direct, immediately tangible results. With referendums, the voters address issues of interest directly and, in a manner of speaking, instruct their government as a whole on how to act. For this reason, the degree of support necessary in an referendum to require implementation may need to be higher than a simple majority vote such as is the case in an ordinary election vote.

Binding effect: The credibility and viability of referendums as a means of involving voters directly in government decision-making is related to ability of the voters to have confidence that their wishes will be implemented. The binding effect of referenda is what distinguishes them from merely consultative or advisory votes, sometimes called plebiscites in English.

Even once the various safeguards mentioned are place, one is still left with the traditional arguments, both for and against the use of referenda as a means of making governments more accountable.

The main arguments raised in favour of the use of referenda are:

• Referendums are the purest from of democracy, dating back to the ancient Greeks, if not further;
• Putting questions directly to the voters limits the influence of lobbyists, power brokers and organizations devoted to promoting special interests by influencing government and elected representatives;
• Referenda legitimize political decisions taken by governments on behalf of the people
• Referenda promote individual responsibility for the governance of the people;
• Involving people directly in deciding their future counters voter apathy and cynicism.

The main arguments raised against the use of referenda are:

• Issues in referenda votes are reduced to overly simplified choices, thereby limiting the ability to arrive at compromises accommodating greater popular support;
• Voters lack expertise to weigh all the consequences of a particular proposal
• Voters lack sufficient information or time to acquire and assimilate sufficient information to make an informed decision;
• Referendum campaigns have the potential to be dominated by monied interests;
• Reliance on the referendum mechanism acts as a disincentive for governments and politicians to make hard decisions.

“Will referenda make government more accountable?” There is no single answer because the answer in one jurisdiction, at one point in its history, may well be different from the answer for another jurisdiction or for the same jurisdiction at another time. The debate will continue.