Having mentioned him just the other day, Scott now gives me an excuse for some good old-fashioned inter-blog commentary. He’s been playing around with some figures to show that being in the political centre has an advantage over the left or right, because the centre accrues less condemnation than either extreme.
The general concept sounds reasonable enough at first, although it’s complicated by there being not just one simple left-right axis, but many axes on different issues. Some tend to go together when you look at the broad population, so those become what we think of as the “left-right axis”, but an overall average isn’t a reliable guide to where a “left” or a “right” person would stand on any particular issue, and few people would be in the centre on all issues at once.
That aside, though, even on a simple single axis those in the centre could easily find themselves left high and dry, because politics isn’t about being the “least condemned”, it’s about building coalitions of interest. The people slightly to one side of centre find themselves faced with a choice of moving towards dead-centre to build a centre coalition, but having to put up with the slightly-opposite types who’ve joined them there, or moving towards the extreme a little more than they’d prefer in order to build a right or left coalition.
Take, for example, Scott’s ten-position scale: the people at positions 4 and 7 could decide to join with the 5s and 6s to form a centre coalition; if they do, they’ll have to put up with people three positions removed from them. But in that case they may as well join with the 1s (if they’re at 4) or the 10s (for the 7s), so it’s by no means certain which way they’ll go. The 5s and 6s could still form a sizable coalition, given that they probably represent the middle slice of a bell curve, but the 4s and 7s would be the ones tipping the balance in any three-cornered contest of left, right, and centre.
Given this, it makes more sense for the 4s and 7s to build coalitions around themselves. They’re never more than three positions away from their fellows on the left or right, and the extremists will always support them before they’d support someone in the dead-centre or on the opposite side. Yet the 4s and 7s also have a chance of drawing support from people just over the fence on the other side (a 6 might regard a 4 as a better ally than a 10), giving them the edge in a two-sided contest. If the 4s and 7s are both pursuing this strategy, the 5s and 6s then become the swing votes but never get the chance to build a strong coalition around themselves.
Any comparison with recent parliamentary elections is left as an exercise for the reader.
19 May 2005
In other words, when it comes to winning power and effecting change, having the least condemned position doesn’t mean that you’ll get the most support. If you have people with strong views to either side of you, each side is more likely to try and win your support for their side than to rally behind you in one happy centrist alliance. The people in the centre are the kingmakers, but that doesn’t make them the kings.
Added by Rory on 19 May 2005.
Interesting. I think he's right about the centre attracting less hostility. I also though like your 4s and 7s idea but suspect the economists were right first time around: given a unimodal distribution on a simple L-R axis, parties should go for the centre. (If it's bimodal, your model works.) Rory's point is relevant too, because intensity of preference also matters, but, I can't see an easy way to factor that in.
IOW it's not that simple...
Added by JayAnne on 22 May 2005.