A mathematician reads the election

Nice to know that dsquared and CT are ahead of the curve. Not that I agree, really.

It’s an interesting thesis, but it doesn’t hold up when you look beyond the presidential numbers. There has been a slow shift towards the Republicans in the Senate and the House as well. Statistically, that makes quite a difference.

…although, I would add… I have a lot of time for Paulos and have read his book (and do have a maths degree, from Oxford). There are a lot of spurious statistics out there being used to pollute our minds with rubbish ideas. But I don’t take his views on the election. Insofar as apparently “random” events decide the US election, valuable lessons can be learnt from seeing what those events were and making sure they happen to you next time.

I don’t know if I agree with Paulos’s thesis, although it’s very consoling. But The gains made by the Republicans in the House and Senate cannot be used as evidence to the contrary. In the Senate the problem is simple: little, almost uninhabited states like North Dakota get the same two Senators as New York or California. And most of these little empty states vote Republican. This is a structural problem that the Democrats have to solve, but it says nothing about the relative popularity of the parties. The Democratic minority in the Senate probably represent at least half the population. As to the House, weren’t the small Republican gains all the result of blatant, and very possibly illegal, gerrymandering in Texas?

I don’t know if I agree with Paulos’s thesis, although it’s very consoling. But The gains made by the Republicans in the House and Senate cannot be used as evidence to the contrary. In the Senate the problem is simple: little, almost uninhabited states like North Dakota get the same two Senators as New York or California. And most of these little empty states vote Republican. This is a structural problem that the Democrats have to solve, but it says nothing about the relative popularity of the parties. The Democratic minority in the Senate probably represent at least half the population. As to the House, weren’t the small Republican gains all the result of blatant, and very possibly illegal, gerrymandering in Texas?

I take the author’s caution against drawing conclusions from these close elections, but…

Since the ’64 Civil Rights Act, white flight to the suburbs and the Dixiecrats leaving the Democratic party, Republican presidents have won 7 elections, while the Dems have only won three. And the Democratic presidents were both southern governers.

This small dynamic might be enough to turn these otherwise close elections. But there are surely other small factors which, when isolated, could be argued as decisive.

Paulos is writing at a time when Democrats are wondering whether we need to radical reform the party or compromise our principles to remain electable.

Paulos is asking evidential questions. Should we conclude that the Democratic party is in terrible trouble because we narrowly lost this election? Do these results prove that we need to do something dramatically different next time? Not necessarily.

People are talking about this election like it was an ignominious defeat. It wasn’t. We nearly beat a wartime incumbent in a non-depression economy. We must have done a lot of things right.

Obviously, mistakes were made. But we only lost by about 3 million votes. We’ve got to keep things in perspective. We don’t need a scorched earth policy to pick up another 3-4 million votes in 2008.

I want radical reform in the Democratic party, but not as a reaction to a relatively minor electoral setback. If Bush had won in a landslide, I’d be wondering about the future of the party. Losing by a 3% margin is hardly a resounding rejection the Democratic party and its values, despite what the mean spirited pundits would have us believe.

Interesting idea, but I think its premises…

There is a large bloc of people who will vote for the Republican candidate no matter what, and a similarly reliable Democratic bloc of roughly the same size. There is also a smaller group of voters who either do not have fixed opinions or are otherwise open to changing their vote.

…are becoming less true, if they were ever true. Both parties are coalitions of interest groups. The constant aim of politics (at least in the American winner-take-all system) is to broaden the coalition without creating or exacerbating internal inconsistencies, which give the other party opportunities to split off disaffected groups. Thus the system tends toward 50/50 stability.

Case in point: the GOP used to be the party of fiscal discipline. But partly in an effort to win over some traditional Democratic interest groups (like seniors and African-Americans), they’ve spent huge amounts of Federal dollars things like on Medicare and education. So in response, the Democrats appeal to Republican voters disgusted by the soaring Federal deficit.

I think the blocs that will vote Democratic or Republican every time, regardless of candidate or platform, are getting smaller, while the bloc of undecided voters is getting larger. But I also think “undecided” misrepresents a large chunk of this bloc: many voters are more conflicted than undecided. For instance, some voters (including me) were turned off by Bush’s social agenda, but supported his foreign agenda. These voters had to decide which issue was more important to them.

Lastly, Paulos’s article ignores a rule that I learned in college poli sci class and have found quite instructive ever since: most people who develop an attachment to one party or the other do so because of some major life event. Millions of Americans became lifelong Democrats because of the Depression and FDR’s New Deal; millions of others left the Democrats for good because of the civil rights struggles. I believe 9/11 was just such a momentous event, and may have altered the political balance in this country for generations.

One other thing: I believe the closeness of the election can be chalked up to the weakness of Bush as a candidate. If it had been, say, John McCain in the White House these past four years, and events had unfolded exactly the same, the Republicans would have won by 10 or 15 points, rather than 3 and a half. Thus I’d say the only real strategy for the Democrats is to sit back and wait for the GOP to overreach, which they almost certainly will do. If they pass the FMA, or overturn Roe, they will push a lot of “national security” Republicans into the Democratic camp, and the pendulum will swing right back the other way. That’s about all the Democrats can do, anyway.

Chance the Gardener: your argument is undermined by the fact that the results were closer (both in absolute numbers and in percentages) in other states, and not always in Bush’s favor. Take Wisconsin: Kerry won the 10 electoral votes there by about 14,000 votes — less than half a percent of the total. To say that Ohio’s 130,000 is the only margin that matters is silly.

Paul Gottlieb – Just what is the Democratic Party expected to about Senate representation? It’s constitutionally guaranteed. Removing Senate representation from a state requires the express consent of the State. Even if you had permission from an individual state, other states of similar population makeup are required to pass any constitutional amendment.

Then there is the political component. Forget party viability for any political party even seriously suggesting this. Who in their right mind would vote for the party that advocates removing elected representation?

Just thought I’d chime in since I have a rare chance to say that I’m in complete agreement with Brett Bellmore: this election was a get-out-the-vote contest, not a convince-the-ditherers contest. But I think this supports, not debunks, JAP’s conclusion: if the difference in this election is that Democrats weren’t able to match the Republican 72-hour GOTV storm, then Paulos is perfectly right to say there’s little evidence for a “seismic shift” promising an era of Republican dominance. The question for Democrats now is then “What’s the best way to knock on doors and make sure people vote?” and maybe “How can we block redistricting and other procedural changes designed to keep Democrats out of office?” not “How can we overhaul our party so as to appeal more strongly to religious voters/ Hispanics / religious security soccer Hispanic moms?”

I think Paulos’s _own_ reason for his conclusion is weaker — I’m just not confident there are enough data points for him to place any weight on the discrepancy between the regression analysis and the actual result. But I haven’t thought much about that analysis, so maybe it’s sturdier than I imagine.

No one’s mentioned what I thought was the best thing in the article — JAP’s observation that we’re biased against believing that really important things happen by chance, even when this is true. I’ll bet ogmb upthread is right to say that the weather in Cleveland was a bigger factor in the election than almost anything people have been writing op-eds about.

“if the difference in this election is that Democrats weren’t able to match the Republican 72-hour GOTV storm, then Paulos is perfectly right to say there’s little evidence for a “seismic shift” promising an era of Republican dominance.”

If Democrats couldn’t match Republican get out the vote efforts this year, with one of the more unpopular Republican presidents in recent history, with a war going on that is highly unpopular with Democratic voters, with the largest youth vote in the history of the US, with more money coming into the Democratic coffers than ever before, how can you fail at least suspect that something is going on?

Look, Bush is a lousy, maladroit President who’s pushing lots of policies that lots of people disagree with. Yet the Dems still managed to lose. That’s pretty sad.

My own hunch: it had less to do with slightly more than half the people voting for Bush than it did with slightly more than half the voters not liking what the Dems were selling.

You really should do better than that.

So, yeah, sure: you do deserve to have some doubts, a crisis of confidence, and a deep self-reevaluation. Can’t hurt. Lots of people out there just don’t like you. It may be worth wondering why.

And can I humbly suggest that deciding that the reason they dislike you is that they’re stupid is a wrong, or at least unproductive, tack to take? (And that your tendency to call them stupid may be part of the reason they dislike you?)

I didn’t vote for Bush myself, so don’t look at me that way.

It’s true that the election was very close, but then the candidates’ positions were also very close–forced towards the center by the need to appeal to independent, moderate and undecided voters. The interesting question is, “to what degree is each party reconciled to the centrist positions that both major candidates were obliged to embrace?”

More here.