On May 2, there will be a federal  general election—the fourth in seven years.

Politicians often argue that the proper, and best, means by which citizens can hold them accountable is the general election. In general, many politicians view success in a general election as a sort of general expiation for their many sins and a sign of approval for their other acts.

I argue that general elections are a crude tool for citizens to hold their representative, let alone the government of the day, accountable for several reasons.

Infrequent and Irreversible: First, general elections are infrequent (data below), and their outcome is almost always irreversible until the PM calls the next one (up to a constitutional limit of five years). While people buy a house infrequently, if they are unhappy with their choice, they can reverse the decision within a few months—albeit at some cost. While marriage implies a long-term commitment, a spouse can move out and a divorce can be obtained, again at some cost in both emotional and financial terms.

Between 1949 and 2004, federal general elections occurred approximately every 3.3 years. From mid-2004 to May 2, 2011, there will  have been four general elections, and the first three resulted in minority governments. In all but three cases since 1949 (involving defeat on a  non-confidence vote), the timing of general elections was at the discretion of the Prime Minister. In what is a dynamic environment, the relative infrequency of elections makes the opportunity for citizens to signal their preferences to government between elections much more important.

No matter how badly an MP performs, local citizens have no means to shorten his/her tenure, by means of  recall, for example.

How to Measure Performance? Second, any notion of accountability requires that the principal (in this case, the citizen) measure the performance of the agent (in this case, their MP, or the party in power). Obviously, measuring the performance of a government is a huge and complex task. What criteria should be used? Given the great scope of government activities, scores or possibly hundreds of measures are needed. But that requires a huge amount of information (see below). Then there is the problem of “adding up” the scores on the host of performance measures. What weights are to be used? But the hardest issue of all is to compare the past performance of the party in power to (a) its expected future performance, and (b) the expected performance of the next best alternative party.

There is some evidence to suggest that general elections are retrospective referenda on the performance of the economy. The idea is that each voter asks: Am I and my family are better off in economic terms than I was when the present government was elected? If the answer is yes, then the voter reasons that success in managing the economy in the past is likely to continue in the future if the party in power is re-elected.

The Information Problem: Assessing the performance of the government and of other parties requires a great deal of information that is costly to collect, and much of it is not in the public domain. While Canada has had the Access to Information Act since 1983, it is of very limited use—due both to its poor design and also to the frequent efforts of  governments to undermine it. In the latter regard, Prime Minister Stephen Harper has been notorious.

It is hard to over-emphasize the importance of open government and access to information for citizens to evaluate the government during its mandate.

The Bundling Problem: Holding elected representatives accountable to citizens is complicated by the bundling problem. In general elections in the Westminster model, voters essentially vote for the party they would like to form the next government. But a party is a complex “bundle” including the attributes of its leader, ideology, and policy promises. The bundle is an all-or nothing choice for voters.

The bundling problem on election day is greatly exacerbated by the dynamic environment (uncertainty) in which governments operates. In particular, it is quite likely that even a few months after an election the government will be faced with a number of issues not discussed during the election campaign. (It is common for contending parties to avoid discussing issues that are expected to require painful decisions until after they are in office). Thus the bundle of a party’s policy promises during an election campaign, it is incomplete. Thus, voting necessarily contains a large (implicit) act of faith in the skills of the party leader.

There is also the bundling problem of all the things that the government did, or didn’t do, in the past. Do voters remember it all or even the most important matters?

Parties as Teams: Under the Westminster model, parties are often described as teams competing for office. But they are rather unusual teams. First, the actual composition of the prospective team depends on the outcome of the nomination race in each of the 308 constituencies. Note, however, that the party leader must sign the nomination papers of the candidate, and it is becoming more common for the leader to appoint some candidates directly, although this usually causes controversy at the constituency level. Second, the composition of the Parliamentary team (the party’s MPs) depends on the outcome of 308 separate campaigns.  Third, the composition of the operative team (i.e., cabinet) of the winning party is solely the prerogative of the party leader, now the PM.

In practical terms, the Cabinet is the real team—the one that exercises power over both the executive and the legislature. But imagine what it would be like to bet serious money on a football game if the bettors did not know the names and attributes of team members, but knew only the names of all those who had tried out for the team! This is the problem that electors face when voting for a political party as a team.

Once the Cabinet is selected, the extent to which ministers have autonomy depends almost entirely on the PM. If he wants to operate a “court government” and retain power in his own hands (with a few advisors), he is free to do so. The last three PMs have practised court government.

The Aggregation Problem: Then there is what is called the aggregation problem—how individual votes are eventually translated into a government. Each individual can cast a vote in a geographic constituency, but the decision as to who will be the MP is effectively a collective one in which each voter’s preference has equal weight. The winner is the candidate who obtains a plurality. This is seldom a majority of votes cast because three or more parties field a candidate.

The problem is repeated at the next level of aggregation: the voters in each constituency elect an MP, but the formation of a government is effectively another collective decision (it would be better to call it an arithmetical artifact). The first-past-the-post system increases the odds of one party obtaining a majority of seats even if it has less than a majority of the popular vote where there are three or more parties. The convention, however, is that the party with the plurality of seats forms the government.

The problem continues. In operational terms, a government consists of a subset of MPs chosen by just one MP (the Prime Minister) to form the Cabinet. (And he can even choose Senators to sit in the Cabinet.) Each member of the Cabinet is said to be jointly and individually responsible for the actions of all other members of the Cabinet. But enormous power is concentrated in the hands of the PM. If his party has a majority, the PM is effectively an elected king for up to five years (see Hall & Stanbury, The Hill Times, Jan. 18, 2010).

The aggregation problem goes even further. The Government has, in theory, control over a huge, sprawling and complex organization (loosely, the public service). It is composed of hundreds of organizations engaged in an extraordinary variety of activities. Thus, the government is said to be responsible for—even if it did not itself create—thousands of policies. Therefore, it is entirely possible that many of these policies do not reflect the preferences of a majority of voters. But they are stuck with them—and not just until the next election. Why? Because changing the party in power may not result in any noticeable changes in many policies since the new government does not deem changing these policies sufficiently important to its gaining and retaining sufficient political support.

Interpreting the Results of a General Election: It is a great stretch to interpret the results of a general election as a referendum on each party’s set of policy promises. Alternatively, is a general election a sort of referendum on the perceived attributes of each party leader? Citizens may be saying that even an elaborate set of policy promises is of far less import in an  uncertain world than the attributes of the leader of the party in power, particularly when extraordinary power is concentrated in the person of the PM in Canada.

Do general elections indicate “the will of the people”? Hardly! According to Profs. Benn and Peters (The Principles of Political Thought, New York, The Free Press,1965, p. 408), “the electors’ will is limited to the single question decided at the polls: Who shall govern? Nothing can be inferred beyond that.” Nor can one logically infer that when a party forms a majority government that a majority or even a plurality of voters voted for it because they supported its bundle of policy promises. So talk about a “mandate” from the voters is really not justified in most cases. Reporters and commentators make a logical error when, in analyzing the election results, they say the public wanted a minority government, or the public wanted to punish a certain party or its leader. This treats millions of voters as if they were a single individual with one mind controlling all the pencils in all the polling booths.

What About Non-Voters? The accountability problem is greatly complicated by the fact that about one-third of electors do not vote in Canadian federal general elections. The turnout on Oct. 14, 2008, was only 59 per cent. How do we interpret that fact? Do non-voters represent the triumph of narrowly-construed rationality? Are non-voters saying a plague on all your houses? Are non-voters simply saying that they are prepared to accept the decisions of their fellow citizens?

Conclusions

The grand political bargain is that those in office to whom this authority (including the power of coercion) is delegated by ordinary citizens are to be held accountable for its use. However, the key mechanism for holding elected representatives accountable to citizens (the general election) is grossly deficient:  it is infrequent, crude, indirect, and lacks the means to inflict serious sanctions on those directly responsible, no matter how bad their performance. And sometimes they are blamed for things over which the Government has little or no control.

W. T. Stanbury is professor emeritus, University of British Columbia. He is indebted to B. Thomas Hall for helpful comments.

news@hilltimes.com

The Hill Times