The remorseless decline of tribal socialism

My copy of Dan Hannan and Douglas Carswell's book, The Plan, arrived today. Haven't read much yet and don't agree with all that I've read, but all the same, if you haven't got a copy, you should. It's well worth the read, and more right than wrong.

Something struck me about a graph of voter turnout in UK general elections, early in the book. It looked like Conservative wins were generally on higher turnouts than Labour wins. I thought I'd check.

What follows is not strongly statistically significant, but it's not insignificant.

My impression was confirmed by the figures. Since the war, the average turnout at elections where the Conservatives won was 76.4% and when Labour won was 72.5%.

In collating the figures and looking for an explanation, some other trends suggested themselves. The Labour vote appeared to be much more consistent than the Conservative vote. And indeed it is. The standard deviation in the Conservative vote is 1,932,000, whereas the standard deviation in the Labour vote is 1,393,000. The standard errors from the linear regression lines show similar but even stronger outcomes (1,955,000 for Conservative, 1,224,000 for Labour).

And even this overstates the variability of the Labour vote, due to two extraordinary performances (one good, one bad) in 1983 and 1997 (the reasons for which are well known and not representative of Labour's performance under normal circumstances). Without these two, the standard deviation in their vote is only 1,122,000, whereas you can take the two most extreme performances by the Tories out of the equation and the variability of their other performances would still be such that their standard deviation would be around 1,600,000, much higher than Labour's all-inclusive figure.

The consistency of the Labour vote made it relatively easy to spot another trend. Their vote is heading quite remorselessly downwards (with the exception of 1997), even as the voting population increases. So, statistically, is the Tories' vote (though marginally less so). But in the Tories' case, this trend is dominated by their dismal performance in the last three elections, again for reasons that are well known. Taking those performances out of the equation, the Conservative trend is somewhat upwards, whereas the Labour trend remains unremittingly downhill.

Some highly-speculative conclusions suggest themselves to me.

1. The Labour vote is more tribal and dependable (and, I am tempted to say, unthinking).
2. Elections are won or lost by the extent to which non-Labour voters are sufficiently alarmed by Labour to turn out for the Tories, or sufficiently disillusioned with the Tories to stay at home and let Labour win by default. With the exceptions of '83 and '97, the electorate seem to swing to or from the Tories, much more than they swing to or from Labour.
3. Labour may not be able to rely on disillusion with the Tories bringing them back to power in the future, as their core vote is declining, and they don't usually pick up much of a swing vote. Recent events won't have helped.
4. Going by the trend, even with a fair wind and a competent and likeable Prime Minister, they would be doing well to pick up 10,000,000 votes at the next election. If they are out for more than one term, their core vote may be down to 9,000,000 and that will rarely be enough to win an election, even against an unpopular Tory party, and even if the Tories don't correct the constituency boundaries that are currently gerrymandered to Labour's advantage.
5. Labour are unlikely to have a fair wind between now and the election, and they bottled the chance to ditch their incompetent and contemptible Prime Minister, so 8 to 9 million votes looks like a realistic expectation for the coming election.
6. If the last three elections were representative of a sustained trend, but assuming that Cameron will come in above trend, the Tories could expect upwards of 11 million votes.
7. But the last three elections look like aberrations rather than part of a trend to me. If so, the Tories have a decent chance of being in the 12 to 14 million range for votes. Psephologists may doubt the likelihood of such a strong swing, but 1997 demonstrated that something of that scale is possible, and Brown's government is every bit as unpopular as Major's was, Brown is personally much less well-liked, and Labour won't be able to point to three years of strong economic performance and good guardianship by a judicious Chancellor.

Unless the Tories self-destruct or Brown gets his economic miracle, this highly-speculative, statistically-weak analysis hints that Labour may be about to get drowned by a democratic tsunami greater even than the media have been starting to suggest.

Good on the British people. It may be slow progress, but it seems that each time we give the socialist alternative a chance and discover what a disaster it is in the long-run, a few more people learn the lesson and don't forget it again. They may still be susceptible to all sorts of woolly, interventionist, paternalist nonsense from wet, one-nation Tories, but at least they have recognized that the socialist snake-oil doesn't work. I haven't felt so optimistic in a long time - not about the political options that we have at the moment, but about the fundamentally-sound economic sense of a good swathe of the British public, despite all the best efforts of our intellectual class. It confirms my belief that British voters could be persuadable to support a genuinely-economically-liberal party, if one were to present itself to them.