« The Quickening | Main | The Book Club »

Combinatorics

24 Apr 2008 03:24 pm

I commend this Noam Scheiber post from yesterday, and this one line at the end got me thinking of something:

There are really two broad swing groups: one working-class, the other affluent. In principle, you could win the general by winning one or the other, or some combination of the two.

Some combination, indeed. Political commentary sometimes proceeds as if we elect presidents via a demographic electoral college in which African-Americans are a mid-sized state like Wisconsin while white working class men and white working class women are, respectively, big states like Pennsylvania and Ohio. But of course it doesn't work that way. But of course it doesn't work that way. McCain gaining the support of 1,000 Kerry-voting Vermonters doesn't accomplish anything because he's hopelessly far behind in Vermont anyway. But persuading 1,000 Kerry-voting African-Americans does help him (assuming they don't live in Vermont) even though he'd still be hopelessly far behind.

Share This

Comments (6)

Um, why?

Put down the bong.

I think you're missing a paragraph here--I'm not following your logic here either.

But persuading 1,000 Kerry-voting African-Americans does help him (assuming they don't live in Vermont). . .

Considering there are only 4,000 or 5,000 African-Americans living in Vermont, and not all of them are voting age, landing 1,000 of them would be quite a feather in McCain's cap.

Coalition building is the key to all winning pols. Look at Bill Clinton...he didn't win either of his generals with a majority of the vote, he didn't win the white vote in either...the irony being he got an overwhelming majority of the black vote.

http://en.wikipedia.org/wiki/U.S._presidential_election,_1992

Obama threatens all the conventional wisdom with his demonstrated ability to bring entirely new factions into the equation....the old school pols don't know how to deal with.

He also threatens the media's conventional wisdom of fight the other side, target this group or that, etc.

*sigh* I just once wish someone in the status quo structure would get that his main draw is getting rid of all that.

Matt's post was a mess, but I think there is a good point in there somewhere.

As I would put it, all this talk of whether Person X will "win" Demographic Group Y tends to be useless at best, and actively misleading at worst. The real key to winning elections is to identify your marginal voters (the people who might vote for you, but might not) and persuade as many of those people to vote for you as possible. And when you slice and dice the electorate into tiny demographic chunks, it likely will turn out some of your marginal voters are in groups you will "lose" anyway, and some are in groups you will "win" anyway. But none of that really matters: persuading another marginal voter in a group you will "lose" is just as valuable as persuading another marginal voter in a group you will "win", and vice-versa.

Now I think part of Matt's post was supposed to convey the idea that the presidential election is actually not one big election, but many smaller elections with a weighted impact on the final result. And it could well be the case that in some of those smaller elections, there simply are not enough marginal voters around for you to win no matter how well you do among them, and vice-versa for your opponent. That is a decent point, although it is hard to know in advance how many marginal voters are around, and there are also downticket issues to consider.

But in any event, it is true that your first priority should be identifying and persuading marginal voters in contests where it is most likely there are enough of your marginal voters around to make a difference in the actual outcome. And again, it doesn't really matter how those marginal voters are distributed among the various possible demographic chunks, and thus I would agree that much of the commentary on elections today is pretty bad insofar as it based on the notion that candidates need to "win" various demographic groups.


Comments closed May 08, 2008.

Copyright © 2008 by The Atlantic Monthly Group. All rights reserved.