I just finished Larry Bartels' new book, Unequal Democracy: The Political Economy of the New Gilded Age, and I want to share my thoughts on it. Economists and political scientists have known for a long time that income growth rates are higher for every percentile under Democratic presidents than under Republican presidents, and that income inequality increases under Republican presidents and decreases under Democratic presidents. This has always been treated as a curious statistical fluke, because the other factors affecting income growth (e.g., technological change, globalization) surely dwarf the effects of a president. In Unequal Democracy, Bartels undertakes to prove that this is not merely a statistical fluke. Bartels argues that the dramatic increase in income inequality over the past 50 years is a direct result of the policies of Republican presidents, not impersonal (or inevitable) market forces. After reading the book, I'm still not 100% persuaded, but not because Bartels' case was weak—it wasn't. To the contrary, it was a tour de force: cogent, methodical, data-driven, and statistically sophisticated. I just don't think there's enough data to arrive at a definitive answer. Specifically, I think monetary policy played a big role in income growth rates over the past 50 years; we simply don't have enough data to tease out the effects of monetary policy on income growth. However, Bartels' argument was much more convincing than I expected. For my money, the most convincing piece of evidence was the fact that almost the entire difference in income growth under Democratic vs. Republican presidents occurs during the second year of each administration (the year you would expect a president's policies to start having an effect). Here's the chart Bartels provides:

As you can see, Democratic presidents have consistently produced extremely strong income growth in their second ("honeymoon") year, while Republican presidents have consistently produced extremely weak income growth in their second year. Moreover, this doesn't merely reflect a cyclical pattern in income growth, because it holds in honeymoon years with and without partisan turnover. That's remarkable. Income growth during the first, third, and fourth years of each administration was virtually identical, regardless of party. This implies that presidents have little impact on income growth in those years, but substantially affect income growth in the second year of their administrations. The fact that Democrats have systematically produced higher-than-average income growth in the second year of their administrations, and Republicans have systematically produced lower-than-average income growth in the second year of their administrations, strongly suggests that the difference is attributable to the president's party affiliation. The book is an absolute treasure trove of interesting statistical findings like this, many of them based on original work. Not all the evidence in the book is favorable for the Democrats either—for example, Bartels finds that both Democratic and Republican senators are completely unresponsive to the preferences of low-income constituents. This makes Bartels' argument all the more credible. Highly recommended.

5 comments:

Robert said...

A couple of economists once wrote a new Keynesian type model with the implication that there should be a recession in the second year of a Republican administration. The idea is that wage contracts are written for an expected policy between that of Democrats and Republicans and Republicans have less expansionary policies so etc.

The wrote it in 89 or 90 and nailed the Bush sr recession.

By the time Junior came around they had tenure.

They were Alberto Alesina and Nouriel Roubini (whatever happened to him).

HA said...

(Sorry accidentally posted on the wrong blog post)

Nice write up. One major problem with Bartels's work is that he employs a one year lag to begin measuring the success of policy implemented in the honeymoon period. While this certainly makes sense for Democratic policy--direct forms of wealth transfer should take much less time to impact growth at all levels and inequality--it doesn't seem reasonable for Republican policies, which rely heavily on longer term market forces. Bartels provides some interesting evidence that a longer lag may be needed for Republican policy when he notes that all income categories seem to do better under Republican presidents in their 4th year, though he seems unsure of what to attribute this 4th year bump. State-level analysis on income inequality and economic liberalism generally finds the opposite to be true--free-market policies reduce the income gap.

Marmalade said...

"State-level analysis on income inequality and economic liberalism generally finds the opposite to be true--free-market policies reduce the income gap."

Actually, that would seem to confirm Bartel's analysis. I'm not sure what you mean by free-market policies, but I know that Democratic states have lower income inequality on average (source: The Spirit Level). Also, Democratic states tend to give more money to the federal government in taxes than they receive in federal benefits.

http://benjamindavidsteele.wordpress.com/2010/02/05/demographics-red-states-blue-states/

Jeffrey R Weber said...
This comment has been removed by the author.
WOW Gold said...

A few financial experts as soon as submitted a fresh Keynesian sort model while using the effects there really should be a recession from the second 12 months on the Republican operations. The thought is that often salary legal papers are authored for an predicted insurance plan amongst that from Democrats and Republicans in addition to Republicans tight on expansionary insurance plans therefore for example.

Cheap RS Gold
RS Gold