Alan Blinder: "Is History Siding With Obama’s Economic Plan?"

Princeton economics professor Alan Blinder’s article in today’s New York Times provides a useful summary of a new book by a Princeton colleague, Larry Bartels, which finds consistent differences in economic performance and income inequality trends between Democratic and Republican administrations.

From the New York Times:

Many Americans know that there are characteristic policy differences between the two parties. But few are aware of two important facts about the post-World War II era, both of which are brilliantly delineated in a new book, “Unequal Democracy,” by Larry M. Bartels, a professor of political science at Princeton…

I call the first fact the Great Partisan Growth Divide. Simply put, the United States economy has grown faster, on average, under Democratic presidents than under Republicans….

Data for the whole period from 1948 to 2007, during which Republicans occupied the White House for 34 years and Democrats for 26, show average annual growth of real gross national product of 1.64 percent per capita under Republican presidents versus 2.78 percent under Democrats.

That 1.14-point difference, if maintained for eight years, would yield 9.33 percent more income per person, which is a lot more than almost anyone can expect from a tax cut…

The second big historical fact, which might be called the Great Partisan Inequality Divide, is the focus of Professor Bartels’s work.

It is well known that income inequality in the United States has been on the rise for about 30 years now — an unsettling development that has finally touched the public consciousness. But Professor Bartels unearths a stunning statistical regularity: Over the entire 60-year period, income inequality trended substantially upward under Republican presidents but slightly downward under Democrats, thus accounting for the widening income gaps over all…

The Great Partisan Inequality Divide is not limited to the poor. To get a more granular look, Professor Bartels studied the postwar history of income gains at five different places in the income distribution.

The 20th percentile is the income level at which 20 percent of all families have less income and 80 percent have more. It is thus a plausible dividing line between the poor and the nonpoor. Similarly, the 40th percentile is the income level at which 40 percent of the families are poorer and 60 percent are richer….The 95th percentile is the best dividing line between the rich and the nonrich that the data permitted Professor Bartels to study. (That dividing line, by the way, is well below the $5 million threshold John McCain has jokingly used for defining the rich. It’s closer to $180,000.)

The accompanying table…tells a remarkably consistent story. It shows that when Democrats were in the White House, lower-income families experienced slightly faster income growth than higher-income families — which means that incomes were equalizing…

The table also shows that families at the 95th percentile fared almost as well under Republican presidents as under Democrats (1.90 percent growth per year, versus 2.12 percent), giving them little stake, economically, in election outcomes. But the stakes were enormous for the less well-to-do. Families at the 20th percentile fared much worse under Republicans than under Democrats (0.43 percent versus 2.64 percent). Eight years of growth at an annual rate of 0.43 percent increases a family’s income by just 3.5 percent, while eight years of growth at 2.64 percent raises it by 23.2 percent.

The sources of such large differences make for a slightly complicated story. In the early part of the period — say, the pre-Reagan years — the Great Partisan Growth Divide accounted for most of the Great Partisan Inequality divide, because the poor do relatively better in a high-growth economy.

Beginning with the Reagan presidency, however, growth differences are smaller and tax and transfer policies have played a larger role. We know, for example, that Republicans have typically favored large tax cuts for upper-income groups while Democrats have opposed them. In addition, Democrats have been more willing to raise the minimum wage, and Republicans have been more hostile toward unions.

The two Great Partisan Divides combine to suggest that, if history is a guide, an Obama victory in November would lead to faster economic growth with less inequality, while a McCain victory would lead to slower economic growth with more inequality. Which part of the Obama menu don’t you like?

Print Friendly, PDF & Email


  1. David

    I read the original at NYTimes and was impressed at the statistical accuracy which matches much of the public’s mood…..

    Nowever, reason seems to not motivate voters most of the time and also the poor vote less in percentage. But since we are all feeling poor…..maybe we’ll all vote less? Interesting though that all demographics do well under Democrats. Another plus for Keynesians.

  2. jest

    not to nitpick, but i find blinder’s conclusion disingenuous, because it focuses solely on the presidential party, but not who is in charge of congress.

    that’s a really important point b/c clinton had a republican congress, while i believe reagan had a democratic one.

  3. Demand Side

    I’ve been writing on this for years. A Democrat in the White House means stronger growth, lower unemployment, higher employment growth, stronger profits and healthier investment.

    Check out the tables at, left links column: “Demand Side, the book,” (a pdf) pp. 53-74.

    “Political business cycle,” is a term that has traditionally meant the efforts by the incumbent to juice the economy in time for the election. I see some correspondents have confused that with the fact that egalitarian policies lead to stronger growth.

    Demand Side

  4. Anonymous

    Could this be a statistical regularity relating to the business cycle? A plausible causitive channel could be that people are more likely to elect a Republican government when the economy is bad, hence Republican presidential economic growth is lower because they serve around recessions.

    Indeed, if you want to take it to an extreme, looked at this way the Democrats’ good economic record could be viewed as riding on the coat tails of the reforms introduced by Republican administrations during the bust years. This is certainly the perception I have of our Labour government in the UK: benefiting from the long-term structural reforms of the previous Conservative administration under Thatcher.

  5. laughingsong

    Jest, Anon 1:56AM, Anon 5:38AM,

    If you would like to see a similar series of analyses that discuss factors such as party of Congress, spillover from previous administrations, etc., go here (apologies for the long snd ugly link!):

  6. Anonymous

    Hmmmm…pardon my cynical hesitancy, but Bartels’ ‘analysis’ is based on the data coming from where? The government? That would be the same bogus stats that are regularly out of sync with just about everything because of the manipulation that started back in the 60’s and widely used by both parties? Sounds like GIGO to me.

  7. Anonymous

    Ah yes, even before I started reading the comments I was sure there would be a “well, that’s because Republican presidents set up the Democratic presidents with a good economy.”

    Tsk, tsk.. That Americans can’t see beyond the thought-depriving two-party system that pits them against each other is so sad. We deserved Bush and I wouldn’t be surprised in the least if McCain and his cheerleader take the oval office.

  8. Anonymous

    Too bad none of this matters. Nobody is going to stop the collapse that’s ahead. This is one year where partisan arguments are basically moot. Once the truth sinks in and the pain begins in earnest, people will vote for whichever party wasn’t in office when the public finally caught on to what was happening–or rather, the public began to dramatically feel what was happening in their own lives.


  9. Clayton

    “I call the first fact the Great Partisan Growth Divide. Simply put, the United States economy has grown faster, on average, under Democratic presidents than under Republicans…”

    This is probably the most ignorant and oft-cited fallacies of modern politics. The problem is not the statistics (which I have no doubt are absolutely accurate), but the difference between a correlation and a causal relationship.

    As one anonymous poster already pointed out, this can be easily explained by a different causal relationship that arises in voting patterns.

    – When the economy is good, people feel rich and vote for Republcians with low taxes and few government programs.
    – When the economy is poor, people feel poor and vote for redistributive policies and proactive government support.

    I think everyone will agree that this would tend to be the bias of voters. These two, simple facts explain the trends:

    – On average, Republicans come into office during strong economies and leave during weak ones…
    – On average, Democrats come into office during weak economies and leave during strong ones.

    We need NO CAUSAL RELATIONSHIP between political parties and economic performance to produce this relationship. The selection of this fact ex-critical thinking is just the typical confirmation bias at work.

  10. Zippy in Annapolis

    Proves nothing other than there is a business cycle–not what might influence the cycle. Sophistry. I bet Blinder builds models in his spare time!

  11. Yves Smith

    Cactus at Angry Bear, who is a statistician and does do models in his spare time, looked at the question of Presidential party and economic performance from quite a few angles, He looked at lag effects and if memory serves me correct, also conjoined Presidential party and control of Congress. Go to Angry Bear and search for “cactus”.

  12. Lune

    Even as a lifelong Democrat, I’m not entirely convinced of Blinder’s statistics. I don’t know enough about economics to be able to say with any certainty whether the correlation he cites is indeed a cause-effect relation as well.

    Nevertheless, I think his conclusions are correct primarily because, especially WRT to income disparities, that is the stated goal of the two parties. That is, democrats have always had the goal of reducing income disparities, and republicans have always had the goal of increasing it. Why is it surprising that they were successful at their stated goals?

    Furthermore, for all those that believe this is just business cycles and nothing more, there is a logical conundrum based on which premise you start with:

    1) Business cycles are not influenced at all by presidential policies. That would imply that business cycles have no reason to conform to 4 or 8 year cycles, and therefore, wouldn’t correlate exactly with presidents coming in and out of power. If they don’t occur with the same periodicity, then over time, both Dems and Repubs would equally experience all parts of the business cycle. Thus, if Dems turned in a higher growth rate despite being exposed equally to all parts of the business cycle (over the span of 50 years), then it speaks well to their capabilities.

    2) OTOH, you may believe that business cycles occur in 4 or 8 year periods because they are influenced by Presidential policies (e.g. the usual presidential "goosing" of the economy before elections). In which case, again, the Dems seem to have an advantage since they do a better job with their cycles than the Republicans do.

    I don't think it's that simple. Let's look at the record:
    1980 -> lousy economy -> Reagan elected.
    1996 -> great economy -> Clinton re-elected
    2004 -> lousy economy -> Bush re-elected.

    That means that 3 of the last 7 elections went opposite to your theory. In many elections, the economy may not be the most important issue (Nixon's impeachment weighing in Carter's election, 9/11 helping Bush get re-elected).

    Furthermore, to the average voter, taxes are always too high, and govt programs are always not enough. It doesn't matter whether the economy is doing well or not. Who wins an election is usually based on which guy is able to lie to the public more convincingly when telling them that he will cut taxes and increase spending while lowering the deficit. Sometimes that's the republican, and sometimes that's the democrat. But they usually say the same thing, while trying to paint their opponent as doing the opposite (with the possible exception of Mondale who famously said he'd raise taxes, then went on to lose every state except MN and DC)

  13. S

    Statistics without perspective are meaningless. As are faulty correlations. Just ask the monolines. Policy is like a snowball rolloing downhill. FReally, has economic/trade policy changed directionally (notwithstanding pace) all that much since the 50s? Talking about inequity without a discussion of the wage arbitrage and trade policy is the sound of one hand clapping. Statistics are meaningless without context. The growing inequality has as much to do with globalization policy whcih is championed by both sides for different reasons: The demos like to think they are lifting the world out of poverty (while putting their own constinuency into it) , while the pro bsuiness republicand love the margin potential the tradeoffs brings.

  14. Francois

    “Congress passes legislation, not the President”

    Yes, but the President chooses those who’ll head the go-vermin Agencies during his mandate. They are the ones making sure that the presidential agenda get done.

    Plus, there is not data in Blinder’s study to support his main conclusion.

    Now, it is known that “certain factions” in this country do not want equal opportunity for all. They’ve been actively working against the average American in order to accumulate more power, more wealth, and the ability to tap into a docile and unprotected workforce.

    Now, does the majority of the electorate wants that? If so, please make room for China and Co. They’ll be more than eager to take our place as #1.

  15. Anonymous

    The recent (Clinton / Bush) inequality data doesn’t support this theory. Check out the Census Gini data.

    Families –
    Households –

    Some data points

    Year Families Households
    2007 0.432 0.463
    2001 0.435 0.466
    2000 0.433 0.462
    1993 0.429 0.454
    1992 0.404 0.433

    So you can see from 1992 (last year of Bush Senior) to 2000 (last year of Clinton), family and household Gini’s rose. From 2000 to 2007 they have been flat. This appears to have been a consequence of a substantial decline for both family and household Gini’s in 2007.

    Why the decline in 2007? I don’t know. However, the overall results don’t fit the easy narrative of rising inequality under Bush. Of course, there is other data to support the general thesis of rising inequality (or so I have read).

  16. Clayton


    First, I’m talking about *bias* not absolute choices.
    – There will always be other factors at play including, in your examples, the war and a natural preference to an incumbant during strong economic times.
    – Also, selecting small sample sizes (7 elections) makes the influence of other factors more likely to muddy the biases.

    Second, notice that you cited 2 REelections as violating the theme.
    – As REelections, they are more likely to be influenced by incumbant related factors.
    – Incidentally, they also won’t impact the trends generated from entrance to exit stats. In this case, the fact that a good economy keeping a democrat in office doesn’t necessarily violate the idea that people will flip parties when given a choice of candidates without any incumbant bias.

    What say you?

Comments are closed.