What Are Elections Actually For?

This sounds like a stupid question. It is not. It is one of the deepest questions you can ask about our system of government, and the standard answer, the one taught from primary school to the political science department, is that elections are how the people express their opinion. That answer rests on very little evidence and even less reasoning.

So let us list what elections could be for, and check each one.

Extracting Information About Who Should Lead

The first candidate purpose: the public knows who would govern well, and elections extract that knowledge.

This is hard to take seriously. The same system tells us the public cannot be trusted with the actual issues, which is why representatives decide them. But it trusts the public to see the true character of a politician through months of advertising, staged appearances and a flag behind the podium? You cannot have it both ways. If the public cannot evaluate a tax policy, it certainly cannot evaluate a person whose entire profession is being evaluated favourably.

Only in rare cases does an election recognise real merit or a real fraud. And even then the signal is muddled, because a public that has given up on the system may deliberately pick the fraud in the hope that he breaks something. That is a vote, but it is not the extraction of information about who should lead.

Extracting Information About the Issues

The second candidate: even if the person does not matter, the vote tells the government what the public wants done.

Four things stand in the way, and any one of them is enough.

The important subjects are not on the table. The distribution of income, the wars we conduct, the power of large organisations over the legislature: no party runs on changing these, and so no vote touches them.

The subjects that are on the table arrive bundled. You cannot vote for one party's position on immigration and the other's on health. You take the package, and packages produce fifty-fifty results even when every item inside them has a clear majority.

The public is not informed about the subjects, and it is rational for it not to be, since one vote in four years on a package cannot reward the effort of learning.

And the government and the media lie to the public about the facts, so what information does get extracted is about a picture that was painted for the occasion.

Punishing the Last Administration

The third candidate is the one that survives: elections let the public punish whoever made a mess.

This is real. Most people who have thought about it agree that the public is competent at exactly this. It cannot evaluate a monetary policy, but it knows when a son died in a war that should not have been fought, or a mother died in a pandemic that was handled badly. It cannot say what should have been done; it can say that whoever did this must go.

Punishment is a deterrent, and a deterrent is worth having. The knowledge that a visible disaster ends careers is probably the most useful thing the electoral mechanism produces.

But notice how little it is, and notice the condition attached. Punishment only works if the alternative is actually different. If the two parties agree on all the important questions, throwing one out and bringing the other in changes the faces and not the policy. Then even the punishment is theatre: the public expresses its anger and the system absorbs it without changing anything.

Informing the Public

The last candidate purpose: the campaign educates. People hear the arguments, weigh them and emerge better informed.

No. Most campaigns are lies, and the rest are slogans. A campaign is optimised to move a vote, not to transfer knowledge, and the two goals conflict at almost every point. Nobody has ever come out of an election season knowing more about the budget than they went in.

Where This Leaves Me

My own position is that elections are not a meaningful part of our system of government. They land near fifty-fifty, which makes them close to a coin toss decided by the last week's news. The important questions are never put to the vote, and the parties agree about them anyway, so it does not much matter who wins.

You do not have to agree with that. But if you think elections are important, and you hold the assumption underneath that view, that the public and its opinions matter, then you have to contend with the list above. Of the four things an election could do, three fail on inspection and the fourth works only under conditions we do not have.

A system whose central ritual does so little is a system that has misdescribed itself. That is the real answer to the stupid question. Elections are for making the arrangement look like it consults the governed, so that the governed do not ask what would actually consult them.