Accuracy and Other Illusions by Joseph Sobran

After the most grotesque presidential election in American history, everyone is proposing reforms to prevent the recurrence of a situation so freakish that it won't recur in a thousand years.

It was freakish in three distinct ways.

First, the country was divided evenly, in both the popular vote and electoral votes. This is rare enough.

Second, the electoral vote totals depended on a single state that was also evenly divided. This was the most improbable part.

Third, the governor of the crucial state happened to be the brother of one of the presidential candidates.

What are the odds against all three of these conditions occurring together?

A million to one?

But it was the second condition that was most problematic. When two candidates are separated by a handful of votes, there is no meaningful or satisfactory way to determine the "real" winner. The margin of victory is bound to be much smaller than the number of votes lost by mischance, error, and other accidental factors we usually forget – including the subjective judgments of those who count the votes.

The O.J. Simpson murder trial reminded us that criminal justice is a messier process than we like to assume. Before that, most of us took it for granted that when a criminal leaves physical evidence at the crime scene, the police will find it, the prosecution will present it, and the jury will convict. But we found that clever lawyers can work malign wonders, especially when the prosecutors are inept. There turned out to be plenty of room for manipulation of seemingly clear-cut facts.

Similarly, the Florida vote shook us out of the pleasant assumption that the modern voting machine guarantees an automatic tabulation of the "will of the people." There was enough confusion even before the lawyers got into the act with their dizzying prestidigitation. Raw numbers create the illusion of exactitude where the reality is ambiguous.

But Al Gore's partisans kept insisting that there was a "real" winner in Florida who could be ascertained by (selective) recounts. They also assumed that Gore's thin edge in the national popular vote constituted a definitive "will of the people" – though this edge too was well within the normal margin of error, even setting aside the vote fraud endemic to big cities where one party is dominant.

If only one in every hundred votes cast across the country was lost, miscounted, or stolen, we can't say with any assurance that Gore "won" the popular vote. An absolutely complete, honest, and accurate count, if it were possible, could have gone either way.

All these considerations should cause us to reopen some old questions. What is the moral basis of majority rule, anyway? Does the fact that one party wins more than 50 per cent of the vote (and none did this year) entitle it to impose its will by force? Can a majority of the voters authorize the winner to coerce the entire population? Can sheer numbers make right what is intrinsically wrong?

A popular vote may provide a useful mode of succession. It may be preferable to hereditary rule or to a raw, violent struggle for power. But it can't authorize a government to expand its powers beyond the bounds of natural justice. It can't justify taxing some people for the benefit of others. The majority has no more right to rob the minority than to exterminate it.

The more excessive the powers of government, the more bitter elections are bound to be. If government were limited to a few modest powers, the stakes in any election would be small, and it wouldn't be worthwhile to spend huge sums of money to help one candidate win.

Those who demand campaign finance reform are approaching the problem from the wrong end. Big government creates pressure for lavish campaign spending. Limited government doesn't.

Our liberty should never depend on who wins an election. That's what the Constitution is for: to ensure that no matter who wins, our freedom is not at risk. Unfortunately, this is far from the case. In fact both major parties now stand for predatory government.

When your real problem is constitutional, you can't solve it by improving methods of counting votes.