Mark Penn in the New York Times:

Sure, young people voted heavily for Mr. Obama, but they voted heavily for John Kerry . . .

The exit polls:

Yes, maybe the exit polls were wrong here, but considering Penn’s article was otherwise peppered with exit poll statistics, so I don’t think that’s the issue.

So what happened?

My guess is that Penn had a storyline he wanted to focus on–the trends among minorities and upper-income voters–and so he wanted to dismiss any alternative explanations. But I think this is a misconception–the idea that a story has to be simple. Penn’s big story can be important but there’s no need for him to try to artificially strengthen it by dismissing the big changes in the under-30 vote. Lots of things are going on at once here.

P.S. Somebody pointed out to me that the only demographic group that Kerry won in 2004 was 18-29. I don’t think this gets Penn off the hook, though. Obama’s victory among the under-30s was so much bigger than Kerry’s, that I thought it was highly misleading for Penn to write, “Sure, young people voted heavily for Mr. Obama, but they voted heavily for John Kerry . . .” His first “heavily” is over a 30-point gap; his second “heavily” is a 10-point gap. He’s using English words to equate two numbers that are much much difference. I’m not saying Penn did this on purpose, but I’d still classify it as a mistake.

Andrew posted earlier about how wealthier voters again voted disproportionally Republican. What about states? Did rich states vote Democratic in 2008, as they did in 2004? Did poor states vote Republican? In short, yes, they did.

The two raw data scatterplots show the average per capita income of states in the election year (in 2006 dollars) on the horizontal axis, and the Republican share of the vote on the vertical axis. A best-fit line is drawn in each year. The income-vote slope decreases a small amount in 2008. Overall, though, rich states are still blue, and poor states are still red.

So how can rich states vote Democratic, while rich individuals vote Republican? And is this pattern new? Buy our book to find out!

Next step: multilevel models of individuals nested in counties and states. That’s going to have to wait until we get the exit poll data.

The first thing I did after getting back from Grant Park was to look at the exit poll results on income and voting and compare to 2000 and 2004.

Up through the middle class, there was little change:

But at the high end there was a flattening out, even a turn toward the Democrats, that we didn’t see before:

Is this a real change, perhaps a move to a new era in which, at a national level, the upper middle class and rich are divided evenly between the two parties?

Or maybe it’s coming from the economic crisis, which in financial terms is hitting higher-income voters harder. (After all, who has money in the stock market?)

Or maybe it has something to do with the candidates themselves. What would the breakdown of voting by income looked like had it been Clinton vs. Romney? (There might be some poll numbers on this from the primary election season, but I don’t know that I’d trust them much.) Maybe it would’ve looked more like 2004.

P.S. regarding the first line above: Boris and I didn’t have tickets. We just observed the crowd from a distance.

Regarding the probability-of-a-decisive-vote calculation, I’ve received several comments along these lines:

Given the virtual certainty that there would be a recount in a decisive state where the election was very close (not to mention decided by a single vote), and given the virtual certainly that the recount would not yield a second identical result, wouldn’t the true probability that one’s vote would make a difference in both the original count and the recount be many times larger than your estimates? I’m not sure how to calculate that higher probability — but I suspect it would range between 2 X and X-squared where X is the Gelman probability that your vote would be decisive.

My reply: No, that’s not right. See footnote 6 of the article, which points you to p.674 of my BJPS paper with Katz and Bafumi which addresses this issue. The key is that, at some point, the election is determined, and each vote slightly changes the probability of McCain or Obama is winning. Before the election, the exact vote margin is uncertain, and the probability that your vote is decisive must be calculated by averaging over all possible vote margins that might occur. The result is that our calculation works out.

From a recent email interview:

- Why the polls don’t matter while the elections are ‘predictable’? Is there too much confidence in a supposely ‘Obama landslide’?

Based on the economy, Obama was predicted to get something like 53% of the two-party vote.  Due to the U.S. electoral system, if you win 53%, you will win many many states and so the Electoral College will look more like a landslide.

- Do you give any credit to this Bradley Effect?

My guess is that this pattern is less of an issue in a highly contested national election.

- Your book ends with the myth that the rich elite votes Democrat and the poor vote Republican, but does it mean that American politics are in a way a battle of elites (those in the poorest and those in the wealthiest States)?

Yes, U.S. politics is in many ways a battle of elites:  the liberal elites in the Northeast, West Coast, and Upper Midwest, versus the conservative elites in the South and the rest of the country.

- If the image of the poor, religious, gun-loving Republican is a myth, why Republicans keep using the ‘real America’ theme and talk about values of small towns in the middle of the poorest part of the country? Is more a battle of rural vs. urban?

I am not sure here, but I suspect that rural images are popular even with many Americans who live in cities and suburbs.

- And what exactly makes the upper middle class in the wealthier States vote Democrats while their peers vote Republican in other States?

The upper middle class in wealthier states are more likely to be socially liberal (even if they are economically conservative).  In poorer states, the upper middle class tends to be both economically and socially conservative.

- Why the economic slowdown benefits Democrats? Do the richest not share the view that Democrats can handle better economic issues, as the polls generally show, or do they vote against their economic interests? (sort of the opposite that What’s the Matter with Kansas said)

A bad economy makes the current administration less popular among all groups.

- Are race or inmigration the main diving lines? And if so isn’t gonna affect disproportionatelly the Obama result?

Race is a big dividing line, but the line is already there.  At this point, people of all races are reacting to the positions of the candidates and new information.

- I was intrigued by your argument that religion matters nearly as much in European election than in American ones. How could it be when here religion is out of the political campaigns and discourse and there is no question whatsoever about the faith of candidates?

I’m not sure why it’s happened, but the two parties in the U.S. differ much more on social issues than they used to.

- I’d really appreciate your comments. Actually, I liked What’s the Matter with Kansas, but I always felt there was something wrong with the picture, as in Europe the richest tend to be more conservative (althought I also attributed to the fact that, in Europe, Democrats will be probably center-right and Republicans far on the right…). Anyway, thanks to you and your book for the clarification!

I hope this is helpful to you and your readers.  I liked “What’s the Matter with Kansas” also.  That book doesn’t claim to be a statistical analysis.  What it has is analysis of the rhetoric of the Republican party and lots of discussion of the battles within different Republican factions in Kansas.  It’s important stuff.  I think that quantitative studies such as ours and qualitative studies such as Frank’s are both important.

Nate Silver, Aaron Edlin, and I estimated the probability that a single vote in any state will be decisive in the presidential election. It was Aaron’s idea, Nate supplied the simulations, and I calculated the probabilities and made the graphs.
Here’s our article describing what we did, here’s the abstract:

One of the motivations for voting is that one vote can make a difference. In a presidential election, the probability that your vote is decisive is equal to the probability that your state is necessary for an electoral college win, times the probability the vote in your state is tied in that event. We compute these probabilities for each state in the 2008 presidential election, using state-by-state election forecasts based on the latest polls. The states where a single vote is most likely to matter are New Mexico, Virginia, New Hampshire, and Colorado, where your vote has an approximate 1 in 10 million chance of determining the national election outcome. On average, a voter in America has a 1 in 60 million chance of being decisive in the presidential election.

and here are some graphs:

There are more graphs if you follow the link to the article.

As Aaron, Noah, and I have discussed, it can be rational to vote even when the probability of decisive vote is 1 in 10 million.

P.S. Typo in Figure 1 caption above fixed.

P.P.S. See here for more discussion of why we can compute the probability of a decisive vote, even though the election might be decided by a recount

As David Greenberg, a historian at Rutgers, mentioned to me, Ronald Reagan’s famous question, “Are you better off now than you were four years ago?” echoes an earlier line from Franklin Roosevelt, from a 1934 radio address:

“But the simplest way for each of you to judge recovery lies in the plain facts of your own individual situation. Are you better off than you were last year?”

This reminds me of a point Aaron Edlin, Noah Kaplan, and I have made, which is that the evidence is (both from survey data and from theoretical considerations) that people vote based on what they think is good for the country, rather than what they think is their personal benefit. (This relates to the idea that it’s not rational to vote, with a probability of decisive vote being about 1 in 10 million, if your goal is to get a $300 tax cut or whatever, but it is rational to vote, with these same odds, if your goal is to make the country and world a much better place.)

Anyway, the Reagan quote is often taken as a symbol of selfishness, of people voting based on what makes themselves better off. But I’ve always taken Reagan’s statement as implicitly statistical or inferential: if your goal is to evaluate how the country is going, look to yourself and your neighbors and see how they are doing. “Are you better off than you were four years ago” is an estimate of “Is the country better off…”

This idea of personal-retrospection-as-inference is clearer in the original Roosevelt quote: “But the simplest way for each of you to judge recovery lies in the plain facts of your own individual situation…” As with Reagan, not an appeal to selfishness but rather an appeal to inference.

Here are two things I kept saying a couple of months ago that I think don’t need to be said any more:

Evidence that the state of the economy is crucial to voters

The nonpuzzle of the close election polls

I’ve written about rationality and voting before, but I still see some confusion out there (for example, in some of the discussion here). Our discussion is here (with longer article here). But let me try to clarify briefly right now.

I do not claim that voting is always rational or even mostly rationally motivated. What I do object to is the claim that voting is essentially never rational. For someone living in a swing state who cares about the outcome, I think voting in the presidential election is a very rational thing to do: a small cost and something like a 1-in-10-million chance of altering the history of the world. It’s certainly a bet I’d be glad to take.

But I’ll be voting in New York, where my vote has almost zero chance of making a difference, so why do I do it? Not for instrumentally rational reasons. I do it for the usual reasons of civic duty, supporting the legitimacy of the electoral process, etc. Again, I’m not saying that all voting or even most voting is rational but that voting can be rational in many important settings (including presidential voting in Ohio, and congressional and senatorial voting in all sorts of places).

I’d also like to address the objection that sometimes arises (for example, in here) that one vote never makes a difference, because if the election were decided by one vote, there’d be a recount anyway. This argument is wrong, as we discuss in the appendix to this article (turn to page 674) for details. We discuss how our decisive-vote calculations are reasonable, even for real elections with disputed votes, recounts, and so forth. We show this by setting up a more elaborate model that allows for a gray area in vote counting, and then demonstrating that the simpler model of decisive votes is a reasonable approximation.

This is cool (from France-Amerique).