Larry Bartels on Income Inequality

LARRY BARTELS ON INCOME INEQUALITY….Do Republican economic policies make income inequality worse? Or is growing inequality due primarily to wider societal trends, rather than anything the GOP does? In an email to Brad DeLong, Paul Krugman provides some evidence to back up his claim that “political and institutional change” does indeed play a large role, and one piece of his evidence is Larry Bartels’ 2004 paper demonstrating that “there’s a strong correlation between party control of the White House and inequality trends even in the short run.”

Krugman calls Bartels’ results “mysterious,” and he’s right. They’re also fascinating. I blogged about his paper twice last year, and it’s worth revisiting those posts in light of the discussion going on in the blogosphere on this topic. Besides, I haven’t posted any charts and graphs for a while.

The posts were originally written on May 9 and May 10 of 2005. You can read them again under the fold.


REPUBLICANS vs. DEMOCRATS ON THE ECONOMY….Did you know that Democratic presidents are better for the economy than Republicans? Sure you did. I pointed this out two years ago, back when my readership numbered in the dozens, and more recently Michael Kinsley ran the numbers in the LA Times and came to the same conclusion.

The results are simple: Democratic presidents have consistently higher economic growth and consistently lower unemployment than Republican presidents. If you add in a time lag, you get the same result. If you eliminate the best and worst presidents, you get the same result. If you take a look at other economic indicators, you get the same result. There’s just no way around it: Democratic administrations are better for the economy than Republican administrations.

Skeptics offer two arguments: first, that presidents don’t control the economy; second, that there are too few data points to draw any firm conclusions. Neither argument is convincing. It’s true that presidents don’t control the economy, but they do influence it ? as everyone tacitly acknowledges by fighting like crazed banshees over every facet of fiscal policy ever offered up by a president.

The second argument doesn’t hold water either. The dataset that delivers these results now covers more than 50 years, 10 administrations, and half a dozen different measures. That’s a fair amount of data, and the results are awesomely consistent: Democrats do better no matter what you measure, how you measure it, or how you fiddle with the data.

But it turns out there’s more to this. Via Brendan Nyhan, I recently read a paper by Princeton’s Larry Bartels that adds some fascinating details to this picture.


The first thing Bartels did was break down economic performance by income class. The unsurprising result is shown in the chart on the right.

Under Democratic presidents, every income class did well but the poorest did best. The bottom 20% had average pretax income growth of 2.63% per year while the top 5% showed pretax income growth of 2.11% per year.

Republicans were polar opposites. Not only was their overall performance worse than Democrats, but it was wildly tilted toward the well off. The bottom 20% saw pretax income growth of only .6% per year while the top 5% enjoyed pretax income growth of 2.09% per year. (What’s more, the trendline is pretty clear: if the chart were extended to show the really rich ? the top 1% and the top .1% ? the Republican growth numbers for them would be higher than the Democratic numbers.)

In other words, Republican presidents produce poor economic performance because they’re obsessed with helping the well off. Their focus is on the wealthiest 5%, and the numbers show it. At least 95% of the country does better under Democrats.


But this raises an interesting question: if 95% of the country does better under Democrats, and if economic performance is the most important factor in most presidential elections, then how do Republicans ever get elected? The most common hypothesis ? spelled out in detail in last year’s What’s The Matter With Kansas? ? is that cultural issues often override economic considerations. But Bartels proposes a surprising alternative explanation illustrated in the two charts below.

The top chart shows income growth during non-election years, and it displays the usual characteristics: under Democrats, income growth is strong overall and the poor do a bit better than the well off. Under Republicans, income growth is weak overall and is tilted heavily in favor of the already prosperous.

But now look at the bottom chart. It shows economic performance during election years and it’s a mirror image of the top chart: Republicans produce better overall performance, and they produce especially stupendous performance for the well off. Democrats not only produce poor overall performance, they produce disastrous performance for the well off, who actually have negative income growth.

In other words, voters aren’t necessarily ignoring economic issues in favor of cultural issues. Rather, Republicans produce great economic growth for all income classes in election years, and that’s all that voters remember. They really are voting their pocketbooks.

Bartels doesn’t essay an explanation for this. Do Republican presidents deliberately try to time economic growth spurts ? and are Democratic presidents too lame to do the same? Is it just luck? Or is the difference somehow inherent in the different ways that Democrats and Republicans approach the economy (with Democrats typically focusing on employment and Republicans on inflation)? At this point, your guess is as good as anyone’s.

Bottom line: if you’re well off, vote for Republicans. But if you make less than $150,000 a year, Republicans are your friends only one year in four. Caveat emptor.


WHY I LIKE DEMOCRATIC PRESIDENTS….Yesterday morning I presented some graphs from Larry Bartels’ paper about economic growth under Democratic and Republican presidents. As long as I have the paper handy, here’s another one.

This graph displays income inequality under Democratic and Republican presidents since 1947. Bartels uses a very simple measure of inequality: the income of the 80th percentile family divided by the income of the 20th percentile family (raw data here). From 1947 through 1969 this ratio was steady at about 3:1, but since then it’s risen to about 4:1.

But the graph also shows something else: Democratic presidents tend to promote policies that either keep income inequality in check or lower it a bit (Jimmy Carter is the exception), while Republican presidents pursue policies that make income inequality worse. The upper and lower lines are guesstimates of what income inequality would be if we had followed only Republican policies or only Democratic policies since 1947. Pure Democratic rule would have produced a slight decrease in inequality, while pure Republican rule would have produced a staggering increase in the ratio to 6:1.

This is important because it’s at the heart of the difference between liberal and conservative views of what’s good for the economy. I won’t try to pretend that I can prove this, but I believe pretty strongly that the single most important economic indicator you can look at is the health of the working and middle classes, the (approximately) middle 60% of the country. Why? Because if unemployment is low and middle class incomes are growing, then everyone wins. The poor win because a healthy middle class is more likely to support safety net and anti-poverty programs, and the rich win because a healthy middle class drives overall economic growth.

Conservatives drive up income inequality because they focus primarily on the well off, which benefits only the well off. Liberals keep income inequality in check because they focus (or should focus) primarily on the working and middle classes, which benefits everyone. And that’s the underlying reason that Democratic presidents are better for the economy than Republican presidents. If you keep the unemployment level low and middle class incomes growing, the rest of the economy will pretty much take care of itself.