10-13-2008 04:24 PM
Before we got started with this Current Events forum, there was some discussion in the Commuty Room about the relative wealth of members of political parties and the demographics that they appeal to - or are percieved to appeal to.
Today in the B&N Review is a book review that I think is particularly pertinent to that discussion - the review concerns Andrew Gelmans recent book Red State, Blue State, Rich State, Poor State .
Here's the review itself:
Red State, Blue State, Rich State, Poor State: Why Americans Vote the Way They Do
By ANDREW GELMAN
Andrew Gelman's Red State, Blue State, Rich State, Poor State can be summed up with extreme concision: Rich people vote Republican, but rich states vote Democratic. Poor people vote Democratic, but poor states vote Republican. That's pretty weird.
But to Gelman, it's worse than weird. It's unknown. Even by the people who are paid good money to know things like that. And it's that troubling fact which seems to have spurred Gelman to write this book. His introduction begins with a quote from pundit Tucker Carlson: "Here's the fact that nobody ever, ever mentions -- Democrats win rich people. Over $100,000 in income, you are likely more than not to vote for Democrats." As Gelman goes on to explain, there are no fewer than two large falsehoods in that statement. First, lots of people mention that "fact." Gelman quotes a number of them at length, ranging from left-wing populists like Thomas Frank to right-wing culture warriors like Michael Barone. Second, the statement is false. As Gelman says, "If poor people were a state, they would be 'bluer' even than Massachusetts; if rich people were a state, they would be about as 'red' as Alabama, Kansas, the Dakotas, or Texas."
There are a variety of reasons for the confusion, but the empirically interesting cause comes in the form of what Gelman calls his "supergraph." This graph is important for the book, so it's worth describing in some detail. One axis tracks the income of voters; the other axis tracks the probability of voting for Bush. Gelman plotted three states across these metrics. One, Mississippi, is a low-income state. The next, Ohio, is a middle-income state. And the third, Connecticut, is high-income. "Our original thinking," writes Gelman, "was that some states were more Republican than others, but that there would be a consistent pattern of income and voting within each state." It was not to be. Connecticut exhibits a gentle slope. The rich are more Republican, but not by much. Ohio is a steeper incline. The rich are a fair bit more Republican than the poor. And Mississippi shoots sharply upward. The rich are far more Republican than the poor. Why should rich states be less sharply polarized than poor states?
It's an interesting puzzle, and one that Gelman spends much of the book trying to solve. The basic answer is that Marx was wrong: Class ain't everything. "With growing political polarization," writes Gelman, "parties have started to take ownership of social and religious issues, such as abortion, that used to straddle the political divide…low-income voters are pretty similar in their political attitudes, but high-income voters show much more geographic variation, with rich red-staters being clearly more conservative on social issues than rich blue-staters." The rich in red states, as it turns out, are likely to be evangelical Christians. The rich in blue states are not. This matters. Race is a big part of it, too. African Americans are disproportionately poor. They're also disproportionately Democratic. When you're looking at Mississippi, say, this will make the class divide look incredibly sharp. But, in part, you're seeing a racial divide.
"The division," concludes Gelman, "is not simply between rich and poor voters, or religious and secular, or rich and poor states. Each of these factors is important, but the true differences lie in subsets of the population. Rich versus poor voters in poor states, high-income religious versus high income secular voters, red states versus blue states among rich voters…the rich-poor divide is a superposition of many changes."
At the most basic level, this is an argument for complexity. The country is not as simple as some would have it, and if that means political discussion segments need to be lengthened from two minutes to four minutes, then tough. But it's also an argument for data, and for increased rigor among the chattering class.
Gelman, for one, is bothered by the fact that so simple a question -- How do rich people, on average, vote? -- is so frequently misunderstood. Searching for an answer, he settles on the existence of "availability bias." Availability bias is the term used for the human tendency to generalize based on nearby information. For instance: If I were governor of Alaska, a state that runs on oil revenues, I might think economic governance in general was quite easy and required few tradeoffs and I was fully prepared to do that task on a larger scale. I'd be wrong, but it would have tracked with the nearby information. Similarly, most journalists are well educated, have high incomes, and live in rich states (broadly speaking, California, New York, Maryland, DC, and Virginia). And most of their friends and colleagues vote Democratic. But this is not the norm. As Gelman notes, "richer counties tend to support the Democrats within the media center states of Maryland, Virginia, New York, and California but not, in general, elsewhere. And richer voters support the Republicans everywhere, but this pattern is weaker -- and thus easier to miss -- within these richer states."
What Gelman does not engage is that this mistake is also quite convenient. Journalists, after all, have access to the same exit polls Gelman is using. They could look this up. So, too, could producers and editors. Those who misstated the facts could be censured. But they are not. Rather, they advance rapidly through the ranks. This is due to the profession's preference for narratives it understands, knows, and promotes. One such narrative is that Democrats are the party of elites. Start with the presumption that that's true, and though most rich people actually vote for Republicans, the idea that they vote for Democrats is, in Stephen Colbert's memorable term, truth-esque. That's not availability bias. It's professional bias. And you can't solve it with a graph.
Ezra Klein is a staff writer at The American Prospect.
What do you all think of some of the claims made in this book as explained in this review? Based on what you see here - or perhaps on the synopsis on the product page - do you think this book brings a good approach to examining the voting trends of various states?
I certainly think its true that looking at political parties purely in terms of wealth is misleading - as Gelman points out, the wealthy in some states are highly religious while in other states they tend not to be, and this makes up for a signifigant amount of the political difference. Another question worth investigating: can we really say that either political party is "the party of the elite" when the political members of both political parties can easily be described as the "social elite" and both parties number among their supporters signifigant numbers of people with lower incomes?
10-13-2008 05:18 PM
I haven't read the book, but as one who used to work professionally as a pollster and statistician, I can already tell you that if the review is correct, Gelman is a bad statistician.
I hate to clue him in, but the ballot in this country is secret. He has no way of knowing how any particular person voted.
Even exit polls are not reliable, since a) people lie, b) I don't know of a single exit poll that has ever interviewed every voter in any given precinct, c) some people who are interviewed refuse to reveal their votes, d) they disregrd absentee ballots, and e) they rely on the people the qestionerschoose to interview, and as has been shown over and over, there is an inherent bias in the people pollsters choose to interview (and the people willing to stop and talk to the pollsters). And, f), very few if any exit polls ask the income level of the person being interviewed.
Also, although we have average incomes for census blocks, census blocks are not identical to voting precincts, so it's very hard to say specifically what the income levels and distributions in any precinct are.
The best we can really says is that in precincts which are partly congruent with census blocks with a higher or lower level of personal incomes the voters voted a certain way in a certain election. Since even the census blocks with the wealthier average incomes have a minority of wealthy people (of course, it depends on how you define wealthy!), you really don't know how those individual wealthy people voted.
I could go on, but I'll just recommend, very strongly, that everybody who has ever read a statistic in the paper or a magazine or has listened to one on TV or radio and has been tempted to believe it should read a relatively short, very informative, and quite amusing book: "How to Lie with Statistics" by Darrell Huff. You will actually enjoy reading the book, and will never look at a statistic the same way again. I guarantee it.
I think, therefore I drive people nuts.
10-13-2008 08:05 PM
some of these people vote in particular ways because they don't declare all of their income or they shelter it off shore or they attach many of their expenses to their business, so as we say in private, they can afford to live like the republicans they criticize but they vote like democrats because much of their income is off limits or hands off when it comes to the government getting its rightful share. they can afford to vote to give away their money because they are not being assessed on the full amount. of course that means that you and i, if we are honest, have to pick up their shortfall. was any of that covered in the book or is this question totally irrelevant?
this is not a sarcastic question. i am seriously interested if anyone has ever looked into the voting habits of people who are in the gray area (although perhaps legally) and are basically cheating the government and the rest of us or are they invisible by nature of what they are doing? i am sure we all know some of these people who profess to want to pay more taxes because they are not paying more taxes on their entire income. is a blue state vs a red state possibly a condition also of people who have no way to hide income vs people who do? of course, there are some legitimate people with little or no income, but they, like the buffets and gates of the world are usually outliers in these kinds of studies.
once i had a friend who said to me...do you know what it would cost me if i had to pay for my own insurance, gas, mortgage etc.? i just kind of looked at her and said duh! what a question! most of us do pay for our own insurance, (home, medical, car) and have no way to hide that or our other expenses inside a business expense.