Wednesday, April 30, 2008

Prof. Larry Bartels on Unequal Democracy


Larry Bartels of Princeton delivered a talk on his new book Unequal Democracy: The Political Economy of the New Gilded Age on Monday. The talk was interesting in that a respected academic came to a data driven but very partisan conclusion: if you're concerned about income inequality, elect Democrats.

You can see his most startling graph on the right. It clearly states that income growth is higher for all groups under Democratic presidencies than under Republican presidencies and that this growth is much more equal. A number of objections immediately jump to mind but from what I've seen so far this conclusion is robust, Prof. Bartels provides a response to some of the criticisms at Dani Rodrik's blog.

Why would voter's continue to vote for Republican presidents against their own economic interests? Prof. Bartels believes that they do vote with their economic interests, but only for the last year. The structure of Republican policies Bartels finds is that they lead to lower growth in earlier years of the presidency as spending and programs are cut. But this leads to higher growth towards the end of the term (and the upcoming election) as the economy rebounds from its bitter medicine. Democrats, however, unleash spending and new programs at the beginning of their term. By the time the end of their term has rolled around, the economy has begun to slow as the effects of the stimulus wear out and inflation kicks in.

Unfortunately for Democrats, according to Bartels (with support from Brookings' William Galston and Thomas Mann and over the objections of someone from Pew) voters only really remember the last year when assessing their economic fortunes. Thus, Democrats lose and Republicans win.

A couple of other interesting facts and figures:

1. Low income voters are more likely to support Democrats. It's high income voters in "red" states that swing them to Republicans.

2. Information matters: the more information self-identified liberal voters consume the more likely they are to correctly identify that it has increased in the United States.

3. Information distorts: the more information self-identified conservative voters consume the more likely they are to incorrectly deny that income inequality has increase in the United States.

4. Information doesn't matter: No matter ideological preference and amount of information consumed and preferences regarding income inequality, about 2/3rds of Americans oppose the inheritance tax.

Thursday, April 17, 2008

Thursday, April 10, 2008

"Predictably Irrational"

I've been meaning to link to this review
of a couple of books on behavioral economics. "Predictably Irrational," a new book by Dan Ariely at MIT sounds like it has some particularly interesting experiments.

In one study, he asked students to look at the last two digits of their social security numbers and then bid on various items. Their social security numbers had marked effects on their bids.

The students whose Social Security number ended with the lowest figures—00 to 19—were the lowest bidders. For all the items combined, they were willing to offer, on average, sixty-seven dollars. The students in the second-lowest group—20 to 39—were somewhat more free-spending, offering, on average, a hundred and two dollars. The pattern continued up to the highest group—80 to 99—whose members were willing to spend an average of a hundred and ninety-eight dollars, or three times as much as those in the lowest group, for the same items.


This effect, which Ariely calls "anchoring," and which retailers such as Tiffany's have been acquainted with for decades (and probably longer) blows conventional economics out of the water. Clean downward sloping demand curves a la Econ. 101 assume rationality on the part of consumers, that they trade off the benefit of consuming the good against the benefit of the other goods they could consume for the same price. If they aren't cold calculators all the time, companies can rely on tricks such as putting other high numbers in the store to set the customer's "anchor" and engine of the free market economy is reduced to a sputter.

This is, however, more of a problem for a lot of academic economists than anyone else. The big money today isn't made on trying to produce commodities that consumers examine with steely eyes and then make a decision based on price. The game is to find a niche demographic and tailor your product to fit their needs. I didn't buy my Mac based on processing power, I bought it because my wife has one, sleek marketing, and because it doesn't feel (and perform) like a hunk of junk.

The key is differentiation, a good businessman doesn't just compete on price. That means that all those pretty supply and demand curves that we were all taught in Econ. 101 are virtually non-existent (they're also pretty damn hard to examine empirically too.) Perhaps this is why economists don't run the world but rather tell others how to?