50/50 World
Kevin Drum observes a whole series of very close elections -- USA, Germany, Italy, now Mexico -- and wonders "Have we gone from a 50-50 nation to a 50-50 world? What's going on?" Three theories.
One -- it's a totally coincidence and this is just random variance. I remember playing poker once and the flop came out as the ace, king, and queen all of the same suit. "What are the odds?" I said, prompting a statistician to observe that the odds of any particular three cards turning up on the flop are equally low. Certain patterns are more salient than others, but these things happen.
Two -- more elections are being held because democracy is spreading and the overall number of countries is increasing. This increases the odds that you'll see an outlier phenomenon during any given six month period and the string of close elections is basically an illussion.
Three -- very close elections are predicted by the median voter theorem. It's possible that over time the level of technological and other competence is increasing, which leads actual party behavior and election outcomes to more-and-more closely conform to the theoretical model.
I think this is another case of the plural of "anecdote" turning into "data". Bigger wins don't make headlines. Michele Bachelet won the Chilean Presidency with 53.4%, Evo Morales won in a multiparty landslide, Alivaro Uribe won big in Colombia, etc. The orange-banana constitutional referendum in Kenya was defeated by almost a 20-point margin. Representation in India's parliament is 215-182 for the Congress Party and allies, with 136 others.
There's a fourth close election, Costa Rica's, which generated exactly zero domestic press. But this sort of facile analysis looks like style section trend journalism.
July 3, 2006 11:08 AM | Reply | Permalink
Interesting thoughts, but they are based on a flawed premise. The election in Mexico is too close to call last I heard, but it is not 50/50.
The election in Mexico is a three-party race, and the two frontrunners appear to be in a dead heat (both have declared victory) with results in the mid 30s (around 35% and 36% respectively).
July 3, 2006 11:43 AM | Reply | Permalink
I go for a middle voter theorem. Except that justification may be more complicated.
Middle voter theorem will not work in every society. Two parties adjust their programs and campaign tricks in the competition for the "middle voter": this means that the people in the "middle" do not perceive very strong reasons to follow either of the two parties involved (or three, for that matter) and thus they can be swayed by adjustments of political programs and campaign tricks. Moreover, the political rules have to be conducive to a two party system.
I have an alternative theory for even split in American politics. Note that this split is quite persistent ever since Gingrich revolution. Many political commentators dwelled on assorted weaknesses of Democrats during this period: lack of coherent program and leadership, terrible strength of "social issues" politics, increasingly lopsided money advantage by GOP. Yet, the advance of GOP stopped as soon as they reached 50%.
What I perceive is that GOP had an agenda that could not be popular elsewhere and had, and has, a potential to alienate the majority. To those people political capital is not valuable in itself but as a tool in an agenda. If they would get lopsided majorities in both houses of Congress they would know that they could push even more of their agenda with impunity --- which would mean that some of their political capital was wasted.
My impression was that Berlusconi had the same approach to politics -- a tool to obtain other goals, not necessarily popular ones.
July 3, 2006 2:55 PM | Reply | Permalink
I think Nicholas is right that there are plenty of big wins going on, but the number of very close national elections (Canada and the UK in the last year as well) does seem striking. I would point out that the median voter theory - in order to be predictive- requires both that politics be one dimentional (you can fit everyone on one line, rather than two or three) and that the distribution be stable over that line over time (otherwise parties can't discover where the "median" is). It is possible that as new kinds of politics (feminism, globalization, environmentalism, religious fundamentalism) create alternative spectrums along which voters identify, and that is making it much harder to cobble together large coalitions. Alternatively, I suspect that we're in a transitional moment: the neo-liberal politics of the Reagan-Thatcher period are largely exhausted, but no new politics has emerged that can consistently capture large majorities (in part because desired polciy results - like low unemployment and increased political accountability- are difficult to achieve). But I would also note that other than in Mexico, we have recently seen the emergence of a new populist left capable of winning impressive majorities, precisely by being willing to challenge neoliberal economic orthodoxy while also deepening democracy by replacing outmoded constitutional strucutres. Something like this may begin to catch on in the older social democratic and left parties in the rich capitalist world.
July 4, 2006 8:13 AM | Reply | Permalink
The UK wasn't that close - Labour won a comfortable majority of seats, and won the popular vote by 3%. Britain's certainly seen a lot closer elections (as well as ones where the party with a popular plurality failed to win a majority).
July 4, 2006 3:52 PM | Reply | Permalink
The median voter theory really does not work well in the US, since the GOP has both campaigned and governed very hard right, and it woes its bare ascendency not to true majorities of the American people, but to wisdespread gerrymandering, including the fact that the elctoral college and the Senate grant excessive weight to minorities of the American people. In effect the GOP has moderated its stances only insofar as needs to in order to win just enough of that crucial minority's support, but it is still well to the right of the center of the American electorate.
July 4, 2006 6:00 PM | Reply | Permalink
I think there is something to be said for the Median Voter Theory here, especially with the rise in polling and related technologies. It might be hard to compare the larger, older, wealthier democracies with more recent, smaller and poorer democracies.
If both sides are better able to accurately identify the median voter, then they will increasingly get to the middle. They will simultaneously work to polarize both sides, as explained by Univ. of Chicago Law prof. Cass Sunstein on a recent episode of 20/20.
July 6, 2006 5:07 PM | Reply | Permalink
Median voter theory still works in a three-party system.
July 6, 2006 5:08 PM | Reply | Permalink