Article published in Liberty Free Press, November 21, 2000
Said Jean Chrétien, "there is nothing more democratic than calling an election." The statement is either pure PR rhetoric, or extremely naïve. Under the label of "public choice" or "social choice" theories, economic research over the past 50 years has shown how political processes in general, and elections in particular, are very imperfect methods of expressing, and reconciling, individual preferences.
Nobel Prizewinner Kenneth Arrow mathematically demonstrated that a collective choice cannot be both rational and democratic. In the 18th century, Condorcet had already discovered the "paradox of voting." For example, it is mathematically possible that the electorate would prefer Stockwell Day to Paul Martin, and Paul Martin to Jean Chrétien, but Jean Chrétien to Stockwell Day. By the very nature of the process by which individual preferences are aggregated in collective choices, election results are often intransitive and incoherent.
In simple, one-dimensional issues, another type of problem is described by the "median-voter theorem": in order to maximize his support and be elected, a politician has to represent the preferences of the median voter, i.e., the one at the center of the political spectrum. This explains why Stockwell Day appears more and more indistinguishable from Jean Chrétien -- think about the health issue. When issues are very divisive, the main outcome of electoral processes is to create frustrated majorities, because only the extreme center is satisfied.
Matters don't get better when one realizes that much public policy is determined not so much by elections as by special-interest activism. Politicians need the support of vocal pressure groups, and they will buy it with public favours and subsidies. Politicians buy their electoral clientèles with taxpayers' money. This leads to contradictory interventions, and to the state being generally biased in favour of concentrated, well-organized interests (like trade unions or subsidized "non-government" organizations), and against the general citizen-taxpayer.
The more all-encompassing the state is, the more powerful the civil service becomes. You need numerous bureaucrats to write and enforce numerous laws and regulations. According to an 1990 index, there is more than 1,500 federal statutes on the books. In the 12 months of the last parliamentary session, 41 bills were tabled by the government in the House of Commons. MPs do not write these laws, they have no incentive to read them and, anyway, even specialists often disagree on what they mean.
Not only does the bureaucracy becomes the de facto legislator, but bureaucrats form a major special-interest group by themselves. The fact that the federal bureaucrats make up 2% of the Canadian electorate only shows a tiny part of their influence on public policy. If one considers all levels of government, one out of six electors is a bureaucrat. Not all bureaucrats are die-hard statists, but their having the right to vote creates a strong statist bias in the electoral process. Moreover, it is in each bureaucrat's career interest that his bureau grows to solve more "social problems."
Another crucial finding of the economic analysis of politics is the fact of "rational ignorance." A rational voter will remain ignorant of political issues for two reasons. First the cost of acquiring knowledge is high, if only because of the sheer volume of laws and regulations. At two minutes per page, a citizen would spend four weeks full-time reading the 4,800 pages of new federal laws and regulations adopted in 1999 only.
Understanding the consequences of the complex, and often murky, policy packages presented by different candidates requires much more than the knowledge of unreadable laws. For example, does the citizen want to analyze the impact of government surpluses, expenditures or taxes? Under the four titles "Taxation, Subsidies, and Revenues," "Fiscal Policies and Behavior of Economic Agents," "National Government Expenditures and Related Policies," and "National Budget, Deficit, and Debt," the September 2000 issue of the Journal of Economic Literature indexes 161 new technical articles and 16 new scholarly books.
The citizen's rational ignorance towards the complexity and opaqueness of public policy explains at least partly why electoral campaigns are not about understanding the issues, but about politicians' image, empty rhetoric, and simplistic advertising. Actually, much political advertising would fall in the misleading or fraudulent category if it were held to the same standards as commercial advertising.
The second reason for the voter's rational ignorance is that, even if he spent enough time and other resources to cast an informed vote, his influence on the outcome of the election would be nil. If you get information to purchase a car, you can then use it to get your chosen car. But if you go to the trouble of finding political information, you will get what the majority of voters (or what the ruling party, or what the bureaucrats) decide, anyway. This is a special case of what economists call "rational apathy," i.e., the fact that an ordinary citizen (we are not talking about journalists, university professors, or apparachicks) has no reason to get involved in politics, because he cannot change the outcomes.
An individual voter has an influence on election results only in case there would have been a tie without his vote. This is virtually impossible with a large number of voters. With many political parties, the math gets messy as we face the so-called "classical occupancy problem" in probability theory. Using an approximation formula proposed by statistician Fred Huffer from Florida State University, assuming 15,000,000 voters of which 90% are pre-committed, and taking recent opinion-poll data to represent voting intentions, we can calculate that the probability of a tie between the Liberal Party and the Canadian Alliance is between 0 (the lower limit, given certain assumptions) and, at most, 5 chances out of 10 raised at the 18,173th power (that is, 5 over 10 followed by 18,172 zeros).
To put this probability in perspective, consider the following. The probability of winning $50,000 at the Québec lottery is approximately 1 chance on 600,000. The probability of winning a Loto-Québec jackpot is one chance on a few million. By contrast with the number of lottery tickets sold, the number 10 at the 168,173th power is an incredibly large number: physicists estimate the total number of elementary particles in the universe to be of the order of at 10 at the 100th power. For all practical purposes, the probability of a tie in the coming federal election is nil. So is the influence of any individual vote.
Citizens who vote do it for many reasons -- to express an opinion, to support a team, to avoid disapproval from their peers, to do their moral duty -- but not to have an influence on the election results. The ordinary citizen commands only one vote, and a single vote has less influence on the election than buying one less tomato has on tomato prices -- "I bought one less tomato so that the poor get them for less." This explains why voters' participation, about 75% in federal elections, has remained fairly constant over the last 100 years, with a downward trend since the 60s, even if the role of the state and, thus, the apparent stakes have become much more important.
To this bleak picture of popular sovereignty, the British electoral system adds problems of its own. Once in parliament, elected representatives follow the party line, period. The Prime Minister is a dictator we elect every four years or so. To be fair, he shares his dictatorial power with faceless bureaucrats and brainwashed judges.
All that did not matter much when the issues revolved around settling minor differences about government management, given some basic agreement on a minimal number of fundamental values. But this is not the case any more. There cannot be near universal agreement when the state wants to set the standards of happiness. Politics is becoming more conflictual and more divisive. Democracy may be useful to settle small disagreements, but it cannot resolve conflicts of fundamental values. In this perspective, calling an election serves more to keep minorities in their place than to "consult the people," whatever that means.
Political democracy is not an end, but a means to run, and limit, government in the interests of everybody. If democracy means that every individual counts for one, it is embodied not so much in political democracy as in each individual expressing freely his preferences on the web of non-coercive human relations -- of which the market is the quantitative part. Individual choices come first; collective choices are only, at best, a surrogate.