A liberal physicist plays the market

Before I get into explaining the subject line, I feel I should make a couple of caveats up front:

  • I am a northeast academic, and as a result, a dirty hippy. If anything, I’d prefer that the Democratic party move a little to the left. You may proceed to make any assumptions about my prejudices that you like.
  • I’m a physicist, not an economist. When I play with numbers, I’m just looking for novel ways to mash them up that I haven’t seen before.
  • Correlation does not imply causality. I know it, you know it. That said, if all of the good things in life are plentiful under one party (on the whole) and scarce under another, then you’d be forgiven if you started to read something into it.

When I’m feeling masochistic, I watch fox news (capitalization intentionally omitted out of contempt), or turn on AM radio. I am meant to believe that the Republicans are the wise stewards of the economy, and that the Democrats, taken individually or collectively are profligate spenders, terrible for working men and women, and ultimately fiscally irresponsible. Nevertheless, there are countless statistics that suggest a Democrat in the White House is far better for the economy than Republicans by almost any standard.

This article from currencythoughts.com, for example, shows that since the 60’s the GDP, employment, and the stock market have grown much faster under Dems than Republicans, and that the dollar has been stronger. Or consider this article from Talking Points Memo, which shows that income disparity between rich and poor increases under Republicans, and decreases under Democrats. Indeed, by almost any standards: inflation, budget deficits, you name it, Democrats are better for the economy than Republicans.

Even though I’m ideologically pre-disposed to simply accept this as irrefutable proof that liberals are awesome at everything, I have to be intellectually honest. It’s actually pretty inconclusive to give just give credit to a president for everything that happens under his/her watch. What if a president takes office and the stock market skyrockets on day 1? He couldn’t possibly have had anything to do with it.

But what if we asked a question along the lines of:

How much does the stock market (or employment, or the GDP, etc.) rise, on average, if a particular party has held the presidency for at least a month? A year? Two years? With a single governing philosophy at the helm for a long time, presumably we should see what the true effect of that ideology will be.

So here’s what I did. I downloaded the Dow Jones Industrial Average for every trading day since 1929 inclusive (all of the data that Yahoo could give me), and calculated the daily change. I then noted which party was office on each day. For a series of time intervals, I asked:

“What is the average daily stock market rise during this period, assuming that a (Democrat/Republican) has been in office for at least T years?”

The Democrats are in solid, the Republicans dashed.

The results are pretty unambiguous. If you knew nothing else about the world except for the fact that a Democrat has been in office for a few years, you could make a fair bet that (at least since 1929) the market would be doing better than if a Republican is in charge. I’d be very interested in seeing what happens if we ran a similar analysis on employment, GDP, inflation, the deficit, income disparity, and so on, but I’m far too lazy to try to track down and process those numbers right now.

Even with the DJIA numbers, it doesn’t take too much to try to poke holes in the “Democrats are great” hypothesis. But there are at least a few questions that I’m prepared to deal with right away (and by the way, I’m happy to share my IDL code with you, if you’d like to tinker with it):

  • These results might overemphasize one off events like the stock market crash of 1929 when a Republican (Herbert Hoover) was in the White House.True, but I’ve run the numbers by clipping outliers at the 5-sigma level, and gotten essentially the same result.
  • Perhaps these results are just due to a couple of presidents being lucky (Clinton) or unlucky (W), and thus dramatically skewing the overall numbers.Could be, although I broke the sample into halves, pre-1960, and post, and each of the two time periods showed the same result. At some point, you can’t just call it, “luck.” Besides, even if all Dems were idiots, but somehow magically ended up with a country filled with peace and prosperity, for most of us, that would be good enough.

There are a bunch of things I either can’t explain, don’t have a good answer for, or might be interesting to look at in their own right. For example:

  • The president doesn’t make the budget in a vacuum. What effect does control of Congress have on all of this?
  • Even though there are huge number of days in the sample, the number of economic cycles are small. Could this simply be a shot noise effect?
  • What happens after 3 years? This is strange. If a single party has been in office for more than about 3 years or so, it doesn’t matter who’s in charge.I could imagine a few responses to this: 1) After a few years, a president is out of good ideas, has dealt with the major catastrophes in the economy, and we simply regress to the mean, 2) Maybe my hypothesis is wrong. Maybe Republicans (gasp!) are as good as the Democrats (or better), but there is just a long lag before the market realizes it.

    I simply don’t know, but I’d be interested in hearing your thoughts.


This entry was posted in Uncategorized and tagged , . Bookmark the permalink.

6 Responses to A liberal physicist plays the market

  1. Emily Joy says:

    This is very interesting. I’ve heard this a lot, but never explained so clearly. I hope someone does try the other calculations too- it’s intriguing.

  2. dave says:

    Of course you’d say that! You’re my wife.

  3. Josh says:

    What happens if you ignore the president and run it for control of Congress? Solid and split?

    fox news (capitalization intentionally omitted out of contempt)

    e.e. is confused and a little hurt.

  4. Josh says:

    Also: you’re doing stock market results but income growth works too.

  5. dave says:


    I like Yglesias’s commentary on the growth of income. BTW, since the Congress has so seldom been split over the last century, it’s more interesting to see what happens with different permutations of Presidential party and Congressional party when both houses are held by the same party. It’s a bit complicated to put in time delays, but for overall performance we get, in decreasing order:

    P=Dem, C=Rep: 11% per annum
    P=Dem, C=Dem: 5.5% per annum
    P=Rep, C=Dem=: 1% per annum
    P=Rep, C=Rep: -9.3% per annum

    Which kind of suggests that Republicans should NOT be given the sort of unbridled control of the country that they were over 6 of the last 8 years.

  6. Grad Student says:

    Great post,

    I wonder if your result is due to sociology rather than the relative ability of dems/repubs to manage the economy. Perhaps people elect democrats in times of prosperity and repubs when the economy sucks. Then after ~3 years the economy reaches an equilibrium.

    Great blog by the way,
    -Grad student

Leave a Reply to dave Cancel reply

Your email address will not be published. Required fields are marked *