At the time this article was written Peter McCormick was professor of political science at the University of Lethbridge.

This article explains the recall idea, looks at how it has been used in other countries, outlines how it would might work in Canada and offers some suggestions about what difference recall would make to Canadian politics. This article, based on research done for the Lortie Commission on Electoral Reform is an edited version of a presentation to the Standing Committee on Procedure and House Affairs on March 15, 1994.

The recall of elected members is not a new idea, nor is it particularly complicated. It is simply a device whereby elected officials can be subjected at any time, to the review of the people who put them in office. When I began to study recall it was considered a terribly exotic idea that had no application to Canadian circumstances, but in the past few years it has become part of the political debate in Ottawa and in several provinces.

Part of the interest is because, as we have seen, a lot can happen these days in the four or five years that is a government's mandate. Two very obvious points are worth emphasizing. First of all, the recall is a device linking officials to the people who elected them. You can only recall your own representative, not anybody else's, no matter how angry that person might make you as a voter. Second, you can only recall your elected representative, not the entire government. It is a device that focuses on the elected member.

Has anybody ever tried it? There are 15 states in the United States that have the recall in place for elected officials. A majority of states have the recall for municipal officials. The recall was first adopted in Oregon in 1908, and the most recent adoption was Georgia in 1978. Two states have been considering recall provisions within the last five years. So it is still a going issue in the United States. It is not a leftover from pre-World War I times.

The recall has not been exercised very often. The best count I have been able to come up with is seven state representatives, one state senator, two elected state cabinet officials and one governor. There was also a second governor on the verge of being recalled when the Senate jumped the gun and impeached him. I suppose we could call that a half success for the recall.

I would argue that it is a device that takes the individually elected representative far more seriously than most devices and most of the arguments we are getting these days.

The low number of recalls derives largely from the fact that the normal term of office in the United States is two years. Therefore, it allows less time for voter resentment to build. This creates an earlier threshold, and often it makes more sense just to wait for the term to run out and then use the next election to replace the person.

In Canada only one province, Alberta, has ever adopted the recall. The act was approved in 1936 and repealed just over a year later when the premier's own constituents were on the verge of collecting the very high required number of signatures. The whole episode is too bizarre to go into in detail but one of the key factors was the Premier Aberhart's rather brusque approach toward constituency service. He was so dismissive, telling them to get out, not to waste his time, he was busy. It was not an attitude which I would recommend to any of today's politicians.

Premier Aberhart's own explanation, that he was being done in by eastern moneyed interest, made very good sense from the point of view of Social Credit ideology but it is completely at variance with the United States experience with recall or referendum, which is that money does not very often buy results. If anything, it is the reverse. Our own experience with the Charlottetown referendum, where the high-spending side went down to defeat, could be matched by dozens of similar anecdotes from the United States.

Two Models of Recall

How does the recall work? Part of the confusion in talking about the recall is that there are two models. They have rather different threshold numbers and implications.

The first model, the most common one, has a three-stage process. First, there is a petition that must be signed by a specific percentage of the electorate. Second, there is a vote on the question of recall, where the citizens of the constituency are asked whether or not they want to recall the representative. Third, if that second vote was positive, you would then have a by-election to fill the vacancy that resulted. This three-stage process is the most common method used in the United States. It is the one that was suggested in a piece of legislation introduced into the Alberta legislature in 1993.

The second model is a two-stage process. It has the petition trigger the actual by-election directly with no intervening vote. If you get enough signatures the seat is vacant and a by-election must take place. This second model is used by a minority of the US states that have this provision in place. It was also recommended by a British Columbia select committee that was considering the question of recall last fall.

The appropriate number of signatures is a key point but it clearly makes a difference which of the two routes you are talking about. If the effect of the petition is simply to trigger a vote on whether or not to have a recall, it makes sense for the threshold to be relatively low. The normal figure in the United States is 15%. On the other hand, if it is the petition itself that triggers the by-election that creates the vacancy in the seat that removes the member, then it makes sense to have the threshold considerably higher. The threshold in Alberta in 1936 was two-thirds of the eligible voters in the constituency. The number suggested by the British Columbia select committee was 50%. As a slight variant, instead of counting the registered voters you could count the actual voters. A majority of US states use this slightly lower criterion. It is a percentage of the actual number of voters in the election that put the person in office, not a percentage of the total electorate, some of whom will not have exercised their franchise.

There is often but not always a time limit. In the United States, with two-year terms, the time limit for collecting signatures is not as big a deal as it would be here. The implicit time limit is the next general election and that is never more than two years away. In states that have a time limit, 60 days is normal, although some have a higher number of 180 or 270. There is a wide variety of numbers, but many states have a time limit on the collection of signatures.

Fourth and obviously most critically, what difference would it make? I would like to begin by dealing with the negatives first. These are three things that I think it would not do.

For one thing, we have to take with a very large grain of salt the idea that the recall could be casually used by other political parties for their immediate partisan purposes. This does not take into account the way Canadians relate to parties these days. The people who vote for a party in a general election are not that party's army that they can snap their fingers and mobilize for action. It takes more than that to trigger voters to action. We see this in the recent excitement in Markham, Ontario. The thing that is so obvious about Markham is that it is not Reformers who want a second chance at the seat and it is not poor-loser Conservatives, it is Liberals who are agitating for the chance to recall their own representative, and I think that catches an extremely important dimension of the recall.

Recall is the way that the people who put the person in office can call that person back to account. It is not a way that other parties or other voters can sneak around and try to cancel out the previous election.

Second, we have to reject the argument that voters would be discarding representatives every second week, that there would be recalls going on endlessly and constantly. In a democracy it should not be necessary to belabour this point. If we can trust the electors to show some wisdom and some judgment in electing people in the first place, surely it is not unreasonable to say they will exercise similar wisdom and judgment in how often and to what purposes they use the recall. If we can not trust them to use the recall, then it is not clear why we trusted them to do the electing in the first place.

Recall is often a demand for an explanation. The circulation of a recall petition is a way of alerting a representative to the fact that his electors are annoyed. If it brings an MP hotfooting from Ottawa to hold a number of town fora or something, that would be a good thing. Usually that will bleed off enough of the discontent that the petition would fail.

Another strong feature of the recall is that in a way it is a self-validating protest. Either they do get the signatures or they cannot. If the cannot, then the representative is enhanced in his claim to represent the riding and the complaint has been revealed to be a minority in the riding.

Third, although Prime Ministers and senior cabinet ministers are attractive targets the Aberhart experience demonstrates that we should not over-react to this prospect. The US experience can tell us nothing about this particular problem. One possibility might be to exempt such individuals from the recall. My own feeling again is that would not be necessary. Without wanting to be too cynical, I think most citizens are perfectly aware of the very real, often very material, and often extremely physical advantages that flow from having a senior cabinet minister, or the Prime Minister, represent your particular constituency, I think that does something to tilt the scales the other way and balance the odds.

Putting these bogeymen aside, the real consequences of the recall will be twofold. First, elected members would know that they had to take their electors seriously all the time, not just once every four or five years. That, in turn, would encourage a definition of democracy that defined leadership in terms of persuading voters to follow; it would not just be making the tough decisions, but making the tough decisions and explaining why they were necessary, and making that accounting when the voters asked for it. That is a dimension of leadership, the public education dimension, which would be emphasized by the recall.

Just as important, I see the recall as a device not to limit private members, not to reduce their role, but to increase it. I think a lot of private members would welcome the chance to have a second master to play off against the caucus master they now clearly have and that they would benefit from that opportunity.

In the longer run, I think the only real road to a more effective Parliament is a feistier set of backbenchers, and I value the recall for the chance that it might contribute to exactly that outcome.

What we are talking about here is a conception of democracy and of representation. If we think of democracy as a way for one party to win the big prize every four years, and the critical question being an electoral system that structures voter preferences so as to generate that choice, then our electoral system is simply horrible. The mathematical problems in the single-member riding are abysmal. The luck of the draw that occurs in riding after riding makes it a difficult way to transmit that kind of popular feeling.

If what we have in mind is allowing voters to cast carefully nuance choices, weighing parties and candidates and issues, then our electoral system is totally inappropriate for the occasion. We could easily improve on it.

However, if the purpose of an electoral system, if the purpose of democratic representation, is to link a specific representative to a clear and identifiable group of voters and to provide for accountability, responsiveness, leadership between those two, then our electoral system has real advantages. Indeed that is the only way you can defend it as a method of democratic representation.

One of the big arguments for the recall is the way it builds on that very real strength and logic which already exists in our electoral system.