Friday, 30 September 2011

If you like their voters so much, why don't you join them?

So, Ed Miliband MP, leader of the Labour Party, has declared "I want Conservatives voting for us – that's how we win elections".

It's worth noting that "getting Conservative voters" has generally worked out far better for the Conservatives than it has for Labour, and the Conservatives haven't managed to get a convincing election victory since 1987.

I've previously noted that charging past the centre ground in an attempt to give your opponents no space for their own policies is extremely risky: any political opponent thinking long-term will respond by becoming more extreme so that the centre of politics gets shifted towards them.

However, it's not even good short-term strategy at the moment.

There's some interesting recent polling on how likely people are to vote for various parties. It's only a rough guide, of course - note that at least 43% of people will always vote either Labour or Conservative, but only 36% would never vote Lib Dem. So at least 7% of those polled will always vote Labour or Conservative, except for those times not counted in "always" when they occasionally vote Lib Dem. However, with those caveats...

42% of people would never vote Conservative. 42% of people is enough to win a UK election on its own with a massive majority - and in practice Labour could pick up considerably more than that from people who might in theory vote Conservative but aren't doing so this time.

The calculation in "Principles or Power" regarding two-party elections (and, with First Past the Post, most crucial marginal seats can be simplified to two-party) means that it can be more effective in terms of short-term votes to move towards your opponent than away from them - every voter you convince directly from your opponent is worthwhile provided you don't lose two voters to apathy or minor parties in the process. The 2:1 ratio is what makes this so powerful for short-term electoral success.

Labour, however, has been moving towards the Conservatives for quite some time now, and is running in to diminishing returns.

YearConservative votersLabour votersTotal voters

During the entire protracted recovery from their disastrous defeat in 1997 to their partial victory in 2010, the Conservatives gained only a million votes.

Labour lost almost five million in the same time period.

Very roughly, their net losses were:

  • 1 million to the Conservatives
  • 1.5 million to the Lib Dems
  • 1.5 million to declining turnout1
  • Less than 1 million to minor parties

Conservative voters who voted Conservative in 1997 are probably completely unwinnable for Labour. So, at most, Labour can recover 1 million votes by getting former Conservative voters - and that's in a best-case landslide victory scenario. They would need, on 2010 figures, to win almost all of them to have a plurality of the popular vote, assuming they don't win over anyone else.

They're not, as far as I can tell, trying to win over anyone else with their policy announcements.

The alternative strategy has far more space to work - there are almost 4 million voters there, most of whom left Labour for a left-of-centre party2 or now don't vote at all.

In a situation with turnout generally declining, there's a huge opportunity to not only win, but win by a landslide, by getting those voters back3. The Conservatives - thanks to Thatcher - are vastly unpopular with most of the country; despite their many failings, Blair and Brown did not make Labour "toxic" in the same way.

These are the voters who would respond most to a party in complete opposition to Conservatism - not to a faint attempt to cosy up to it4 and pick its pockets.

Sadly, deep down, "just to the left of the Conservatives, but definitely not the Conservative Party", is where Miliband and the rest of the senior figures in Labour want to be.


1 Some of this decline in turnout, of course, will be because many of those who voted Labour in 1997 were dead in 2010. But in normal conditions they should have been replaced by newer voters, since the size of the eligible electorate has increased slightly. As it happens, turnout has plummetted, as fewer voters see the point any more.

2 No arguing over whether the Lib Dems count as left-of-centre... For the period 1997-2010 they were generally left of Labour on a wide range of issues: I doubt there was that much Labour to Lib Dem swing from people who thought the Lib Dems were to the right of Labour on their preferred issues.

3 While FPTP gives disproportionate importance to the votes in a few Conservative-Labour marginal seats, a party that is a few million votes ahead in total is still unlikely to lose the marginals.

4 That's not to say that Labour shouldn't be trying to convince swing voters and moderate Conservative voters that Labour's policies are good for them, or be explaining its policies in terms that resonate with them. Absolutely they should. There's a big difference between that and changing policies to the right to be closer to what those voters currently want, though.