The General Election: Numbers Crunched (1)

I've been reflecting on the results of the voting at the General Election, and looking at the numbers*. Here are some of my thoughts.

1. Turnout
Turnout was 65.1%, which was the highest this century. (2001 was 59.4%, 2005 was 61.4%.) However, it is also the 3rd lowest since 1945. The fourth lowest was in 1997 with 71.3%. So it seems that while there has been more interest in this election among the British people than there was for the previous two, levels of political interest and enthusiasm are, historically speaking, low. Why was turnout higher? I don’t know, but I do wonder if the “prime ministerial debates” (which, in my opinion, were not a good thing) were a major contributory factor.

My own opinion is that a low turnout is not, in itself a bad thing, but is a sign of a deeper underlying problem. I don’t think it is a good idea to try to boost turnout by artificial means, since this may treat the symptom, but does nothing about the underlying lack of enthusiasm for the political process.

2. Opinion Polls

The exit poll turned out to be fairly accurate - predicting that the Conservative Party would have 307 MPs elected, Labour 255, and the LibDems 59. In the event, the totals were 306, 258, and 57. The exit poll, was however, widely disbelieved at the time - the LibDems were expected to do considerably better. Dan Hannan’s comment was typical “I’m going to stick my neck out here. I believe the exit poll predictions will move during the night: the Tories and LibDems will do better, Labour worse.” This was largely because most opinion polls had given the impression that the LibDems would do considerably better than that, and might win about 80 seats. However, it turns out that the exit polls were the ones to be believed - they recorded the people who actually voted. Toby Young reckons that many people recorded by the ordinary polls as LibDem supporters in the run up to the election were people who were not actually very committed to voting.

Which brings us to the opinion polls themselves. Here is what the polls were showing. (All figures taken from ukpolling report).
The first column shows an average of the polls taken is the last 3 months of 2009, the second column those taken in the first three months of 2010, the third column shows the state of the polls in the first couple of weeks of April (the opening days of the campaign), before the first “prime ministerial debate”, which led to a surge in support for the LibDems, and the fourth column shows the final 18 polls of the campaign - those that should have been expected to predict the final results most accurately. The actual voting turned out to be somewhere between the figures suggested by the polls at the beginning of the campaign and those at the end. In other words, the campaign had changed minds, but not as far as the polls led people to believe. Or, to put it another way, the LibDem surge was very real - it just wasn’t quite as big as we thought. But on the whole, the polls were (once again) shown to be a fair, if not infallible, guide to how people would vote.

3. The major parties: short term perspective

Which parties had a good election? It depends how one judges. One yardstick is to compare performance with expectations - that is, to compare the actual share of the vote cast for parties on election day with the share of the vote predicted by opinion polls in recent months. But which opinion polls? How recent? Let’s try a variety of approaches.

Very short term (a couple of months): Assuming that opinion polls were basically accurate, the Conservatives were down about 2 percentage points on where they were a couple of months before the election, the Labour were down about 1.5 points, and the LibDems were up about 4 points. In other words, the LibDems had a very good campaign, while Labour and (especially) the Tories had disappointing campaigns.

Medium short term (6 months): Compared to the opinion polls 6 months ago, the Conservatives were down 4 points, Labour up 1, LibDems up 5.5, and minor parties down 2 points. Very disappointing for the Conservatives and the minor parties, not bad for Labour, and again, very good for the LibDems.

Long short term: Over the course of the past few years, if one excludes Gordon Brown’s prime ministerial honeymoon, the Conservatives have averaged about 40% in the polls, Labour 30%,and the LibDems about 17 or 18%. So on that basis, again, disappointing for the Conservatives, respectable for Labour, and very good for the LibDems.

Summary: Looking at things from a short term perspective, the Conservatives had a disappointing election, Labour had a reasonable election, and the LibDems had a very good election. Alas, because the opinion polls had raised expectations to unreasonable heights, it didn’t actually look that good at the time.

4. The major parties: longer term perspective

How well did the parties do compared with previous elections? There are various measures one could use to assess performance. Much of the time, comparisons are made on the basis of the share of the vote received. My opinion is that this assumes either that turnout is the same for every election, or that variations in turnout are meaningless. I don’t accept either assumption. I would expect that if every party fought a poor campaign, turnout would be low - whereas if all parties fought good campaigns, turnout would be high. In other words, the success of a party should not be measured by its share of the vote, but by how many people it can persuade to vote for it. However, just to compare raw numbers is not helpful, since the size of the electorate varies. So the best measure of the success of a party is the proportion of the electorate that it can persuade to vote for it.

Here, then, are the figures for the three main parties since 1970.

Clearly, by this measure, the election was a disaster for the Labour Party - a result even worse than their defeat under Michael Foot in 1983. One crumb of comfort for Labour is that it wasn’t quite as bad as the result the Conservatives had in 2001 under William Hague.

For the Conservatives, the picture is more mixed. It was their best result since the Conservative victory in 1992. However, it was worse than any result the Conservatives achieved between 1945 and 1997. David Cameron may be on the verge of becoming Prime Minister, but these are not good days for the Conservative Party.

For the LibDems, a good result in terms of votes - but not a great one. Not only was it not as good as the performances of the Liberal / SDP Alliance in 1983 and 1987, it actually fell short of the Liberal performance in February 1974.

Summary: No comfort for Labour, a little comfort for the Conservatives, and a fair amount of comfort for the LibDems.

*(Numbers are incomplete, since the constituency of Thirsk and Malton has yet to vote.)