GOP, Dems, and National Income:

There's an interesting posting on Dani Rodrik's always-interesting blog about the difference that the President's party affiliation makes to the distribution of income during the four years of the president's term. Citing a new book by political scientist Larry Bartels, the data apparently show that when a Republican president is in power, people at the top of the income distribution experience much larger real income gains than those at the bottom--a difference of 1.5 percent per year going from the bottom to the top quintile in the income distribution. The situation is reversed when a Democrat is in power: those who benefit the most are the lower income groups.

Not surprising, I suppose -- but what is surprising is that "compared to Republicans, Democratic presidents generate higher income gains for all income groups (although the difference is statistically significant only for lower income groups)." Without having read the Bartels analysis, I'm not sure I believe it -- but if it is true, it's pretty interesting, and troublesome (for those of us who from time to time pull the GOP lever in Presidential elections).

I'd rather have smaller income gains attached to lower tax rates than higher income gains attached to higher tax rates.
4.2.2008 9:56am
Zathras (mail):
This is consistent with the little-known fact that there is no empirical correlation in the U.S. between tax rates(even on capital gains) and economic growth:

Capital gains rates display no contemporaneous correlation with real GDP growth during the last 50 years. Although the effect of capital gains on economic growth may occur with a lag, Burman (1999)1 tests lags of up to five years and finds no statistically significant effect.
4.2.2008 10:02am
anon252 (mail):
Repeat after me: correlation does not equal causation, and is even less probative when there are many potential confounding factors and a small sample size. Note all the studies about the stock market going up or down depending on which conference wins the Super Bowl and the like.
4.2.2008 10:08am
Jacob T. Levy (mail) (www):
It's only a surprise because the Carter-Reagan switch holds such a large place in many people's historical memories. But if you think in broadest brush:

2000s: stock market bust and recession followed by tepid recovery and then a housing bust and a new recession, GOP
1992-2000: recovery that became a boom, without a recession in the period: Democratic
1988-1992: flattening growth that turned into a recession: GOP
1983-88: boom, GOP
1980-82: deep recession, GOP
1976-80: stagflation, Democratic
1970-76: stagflation, GOP
1960-70: generally a boom, with 8 years Democratic and 2 Republican.

The Clinton and Kennedy-Johnson eras outweigh the Carter term, 16 years to 4. The Reagan boom of 83-88 outweighed by the presidencies of two Bushes, Nixon, and Ford. In 1984 people were better off than they'd been four years before, and again in 88, but that period can't sustain the GOP average over decades.
4.2.2008 10:14am
Chris Bell (mail) (www):
Trickle up economics?
4.2.2008 10:16am
Jacob, your outline shows the problems with assuming that correlation is equal to causation. Was it Reagan's fault that Carter left him a pile of garbage? Was it W's fault that Clinton's tech bubble popped? Leaving aside the one-sided analysis (such as your description of each of the Bush terms), you're not addressing the causative question.
4.2.2008 11:05am
TheOneEyedMan (www):
I'll second the correlation is not causation comment.
But also, there is a policy overlap problem. The first few years of a presidency have as much or more to do with the last guy as the current president's polices.
So even if there was a causal relationship between Republican and Democratic policies, this wouldn't measure it.
4.2.2008 11:18am
Jumping in with autolykos:

1960-70: Generally a boom, but at least partially attributable to significant spending on Vietnam War combined with the massive expansion of government under the "Great Society".
1980-82: An intentional recession. Remember Paul Volcker, the "Inflation Slayer"?

Speaking for myself, I'd rather have a more moderate economy without a war. I'm also glad that an adult came along to fix the economic mess of the late 1970s and early 1980s.
4.2.2008 11:19am
donaldk2 (mail):
It has never been explained to my satisfaction what difference it makes. Is it asserted that windfalls to a relative few IN ANY WAY impoverish the members of the lowest quintile? Show me please just how this works.

Mr. Levy writes of busts and recessions. It is not yet evident that we are suffering a recession. There is no increase in unemployment even in the medium term. Let's wait a bit before convening the funeral procession.

The most common and pernicious of mankind's failings is
envy. It will never subside whatever is the distribution of wealth, nor of any other good, for that matter.
4.2.2008 11:31am
Ever heard of a thing called the "omitted variable bias"? Not sure the variable for whether the President is a Democrat or a Republican captures everything (or anything) about what's going on in the economy.
4.2.2008 11:52am
eric (mail):
I think it is somewhat strange to just look at who the President is. What about which party controls Congress? I would like to see the correlations there.

I think divided government is probably the best for the economy. Just my opinion.

Blaming a bad economy on the President is foolish; giving a President all the credit for a good economy is foolish.
4.2.2008 11:55am
Jacob T. Levy (mail) (www):
I didn't claim causation, though of course presidents and presidential candidates are quick to do so as the opportunity presents itself. As I recall-- I haven't looked at this stuff in a while-- the results are robust to 1- or 2-year lag effects, though that doesn't show causation either.

But there's a myth about even what the correlations are, a myth that's tied up with the memory of Carter-Reagan. The myth is strong enough that even someone as well-read as David is surprised by this result, though it's been in social science circulation for a while.

In any event, donaldk2 misses the point. The gross correlations show that *every* income class does better under Democrats. This isn't a "soak the rich" result. It does mean, according to those who put a lot of weight on it, that the emphasis on getting windfalls to the rich under Republican administrations doesn't help the economy in aggregate, and doesn't even help the rich as much as broad-based economic growth does.
4.2.2008 12:03pm
ak47pundit (www):
Is there any analysis on the net income under each President/party rather than just the gross income? It doesn't help that your income grows due to inflation or even real economic growth if such increase is just taken away by more taxes.
4.2.2008 12:13pm
Gaius Marius:
Personally, I prefer greater investment returns than increased income. I consider myself middle class (perhaps upper middle class) who has been maxing out 401K and Roth IRA contributions (for both the spouse and I) for several years to put myself on track to realize investment returns that far exceed my annual income thanks to the power of compound interest.
4.2.2008 12:23pm
Frog Leg (mail):
A single correlation still contains more evidence than a thousand non-empirical theories.

It appears nonetheless that many people here are not willing to allow any ugly fact to slay their beautiful theories, to paraphrase Huxley.
4.2.2008 12:48pm
As I recall-- I haven't looked at this stuff in a while-- the results are robust to 1- or 2-year lag effects, though that doesn't show causation either.


If we step away from the purely statistical analysis and look at the actual causative mechanisms, most of the things that affect income over the long term explicitly take more than one or two years to manifest.
4.2.2008 12:59pm
Zathras (mail):
JBL: And your evidence for that is....?
4.2.2008 1:11pm
Autolykos said:

Was it Reagan's fault that Carter left him a pile of garbage?

Carter inherited a pile of garbage - remember Nixon's wage price controls? Carter appointed Paul Volcker who raised short term rates to unprecedented levels in an effort to slay the inflation dragon. Most educated observers believe Volcker's solution kept inflation under control until very recently, although at great short-term cost to the economy.
4.2.2008 1:16pm

It appears nonetheless that many people here are not willing to allow any ugly fact to slay their beautiful theories, to paraphrase Huxley.

Well, any person who allows any single fact to change their entire worldview is a windsock. This, like any other fact/study, should be evaluated, given appropriate weight and placed into your own personal economic paradigm.

I don't know enough about the ommitted variables to speak definitively about this study or even to judge its merits, but I know enough about economic history to state that extremely liberal states perform more poorly than extremely conservative ones. Heck, the French, Russians, English and even Chinese have figured this out, so I don't know why people expect die-hard big government liberals like Barack Obama to yield positive results (which seems to be what people are implicitly suggesting).

I'd like someone to posit a theory for these results. They otherwise look too anamolous to give much credence.
4.2.2008 1:36pm
Professor Post,
Why would this be troublesome just because you occassionally vote for a Republican? There's no rule that we have to vote on the economy, and there's plenty of other issues to vote on. Don't get me wrong, I'm a committed liberal who doesn't think the Republican position on most of those issues makes sense, but if you think it does, why would this make you change?
4.2.2008 1:41pm
CJS (mail):
Are we really trying to analyze the correlation of income to the President without considering the Congress or the Fed?
4.2.2008 1:48pm
Soronel Haetir (mail):
And then, for some of us, an economic downturn is actually good for business. I own a small retail liquor store in a town where the grocery stores are not allowed to sell beer or wine and a state that does not impose price controls. I'm basically immune. People always find a way to feed their vices.
4.2.2008 2:00pm
astrangerwithcandy (mail):
"Are we really trying to analyze the correlation of income to the President without considering the Congress or the Fed?"

What they said.
4.2.2008 2:30pm
Is the explanation not obvious.

1) Republicans put the policies in place that lead to prosperity.
2) They get voted out and Democrats get the credit for the resulting prosperity.
3) Democrats, being smarter than all of us, and the market, enact wealth-destroying policies.
4) They get voted out and Republicans take the blame for the downturn.

The US economy is a very large ship. It cannot turn on a dime. Change takes time. It take a long-time to see the effects of many policies.

Look no further than 2001 for evidence. Democrats love to howl about the "Bush recession" from March '01 to November '01. Ignoring the question of whether that actually constituted a recession, it's absurd to think that Bush could have started a recession in March '01, only 6 weeks into his first term. It's nearly equally absurd to think he could have even started one in November in that same year.

Such is the cycle.
4.2.2008 2:37pm
lostmycookies (mail):
Troubling? What is troubling? First, if people are unshackled and allowed to reach their true potential, whatever that is, of course the measurable differences will be extreme.

But what really matters is how completely bogus such stats are. All measures of utility are subjective AND such studies do not account for individuals. Take me, for example. If you measured my income 15 years ago, I was bottom 5th percentile. Dad was dead, I dropped out of highschool to work, etc. Now I am a hot shot attorney with a big firm making lots of money. Such income mobility is simply not measured by bogus aggregated data formulations. As has been pointed out by others, what is really interesting is the mobility between the "income groups" (whatever the f'ing heck that is).
4.2.2008 2:39pm
mcb (mail):
exactly, why is this troubling? clearly lower taxer rates better allow the wealthy to use income to re-engage the market, of which they will beneift. those in lower income levels, use lower tax rates to pay off debt, etc.

so what?
4.2.2008 3:49pm
Bski (mail):
Anyone know what the results are if the results are similar if we look at who is in charge of the legislature rather than the executive?
4.2.2008 7:46pm
Bski (mail):
Anyone knowif the results are similar if we look at who is in charge of the legislature rather than the executive?
4.2.2008 7:46pm
I think there is a dynamic of conservative advocacy working better in opposition to a liberal president. In some cases, even better from a minority party. Certainly the recent issue of Republicans wanting to keep getting elected and abondoning principle so blatantly has hurt economic policy.
4.2.2008 10:08pm
Zathras -

Capital investment cycles, for both individuals and businesses, are usually at least three years.

Education usually takes more than a year to complete.

Job experience usually takes at least a year to fully aquire and realize - I would guess that regular salary review periods of less than a year are relatively rare. Even without other training, increased experience tends to increase wages for up to ten years.

Short-term hiring for seasonal or temporary demand can be done fairly quickly, but any significant long-term increase in staff is almost always part of a long-term strategy.

Strategic planning is almost by definition longer than one or two years.

Significant life changes, like moving, can easily take at least a year and are usually part of a much longer-term outlook.

Liquid investments such as stocks and bonds are usually made with a medium to long-term goal.

Significant changes in agricultural production naturally take at least a year.

Land development processes take more than a year, often more than two.

Construction projects usually take more than a year, and are almost always done with a much longer outlook in mind.

4.3.2008 2:58pm
BobPM (mail):
Quoting Mr. Bartels 2/2004 paper: Partisan Politics and the U.S. Income Distribution, he addresses JBL's 3 year lag theory. It is also notable that the graphs are already based on a one year lag.

"Indeed, averaging across entire administrations, the partisan difference in real income growth for the working poor was much larger when the same party remained in power (2.8 percentage points) than in cases of partisan turnover (1.1 percentage points). Democratic presidents who succeeded themselves or other Democrats produced average real income growth of 2.9 percentage points for families at the 20th
percentile of the income distribution; Republican presidents who succeeded themselves or other Republicans produced average growth of 0.1 percentage points. Clearly these differences cannot be attributable to short-term corrections of the other party's misguided policies."
4.3.2008 4:18pm
Actually I don't have a 3-year lag theory. My theory is that income is affected by a large variety of factors, and that those factors have different cycles. Some are almost immediate, some take a year or two, and some take 40 years or more. Mr. Bartels' 2004 paper doesn't really address that issue. It does establish that the Democrats do better with the short to medium-term variables, and in that respect his results are very robust. For long-term effects, the data sets he's using don't have enough information to establish a correlation one way or the other. The 2004 paper also doesn't address causative mechanisms. Hopefully the book does; I'm eager to read it. His work is not comprehensive, but it's certainly valuable.

To be clear, I'm not making a partisan argument here. I've voted for both Republicans and Democrats, and frankly I don't think either party has done a particularly good job. I just wanted to make the point that statistics alone aren't sufficient to determine policy; some understanding of the causative mechanisms is also necessary.
4.3.2008 6:49pm