Contributor Post Created with Sketch. Are Democratic Presidents Really Better for the Economy Than Republican Ones?

 

HillaryClintonPartisans are likely to focus on these stats in Paul Krugman’s latest column: Since 1947, “the economy grew, on average, 4.35 percent per year [under Democrats]; under Republicans, only 2.54 percent.”

To his credit, Krugman does not make strong claims about the superiority of Democratic economic management (though he does mock current GOP policy proposals). Neither does the 2014 study, from where those numbers come, “Presidents and the U.S. Economy: An Econometric Exploration.” The study’s authors, Alan Blinder and Mark Watson, sort of give it the old ¯\_(ツ)_/¯:

Democrats would no doubt like to attribute the large D-R growth gap to macroeconomic policy choices, but the data do not support such a claim. If anything, and we would not make too much of small differences, both fiscal and monetary policy actions seem to be a bit more stabilizing when a Republican is president — even though Federal Reserve chairmen appointed by Democrats preside over faster growth than Federal Reserve chairmen appointed by Republicans by a wide margin.

It seems we must look instead to several variables that are mostly “good luck,” with perhaps a touch of “good policy.” Specifically, Democratic presidents have experienced, on average, better oil shocks than Republicans (some of which may have been induced by foreign policy), a better legacy of productivity shocks, more favorable international conditions, and perhaps more optimistic consumer expectations (as measured by the Michigan ICE). These factors together explain slightly more than half of the 1.80 percentage point D-R growth gap. The rest remains, for now, a mystery of the still mostly-unexplored continent.

Thoughts:

1) Policy matters. But sometimes it takes a while to matter. And the world is a complicated place. So teasing out cause and effect is difficult.

What was the impact of the Obama stimulus? Maybe good, maybe not so good. But it was happening at the same time as we had changing monetary and regulatory policy. Or take the 1981 Reagan tax cuts. Was their biggest impact on the 1980s or on the 1990s? And were those impacts different? Maybe they boosted demand in the 1980s with the supply-side effects taking longer to develop.

2) There have been a dozen presidents since World War II. Is that a huge sample size, especially given changing demographics and macroeconomic conditions?

3) Luck matters. Here is the FT’s Gideon Rachman on the Clinton Boom:

The Soviet Union had collapsed in 1991, just a year before Mr. Clinton was first elected. Throughout his eight years as president, there was no serious competitor to the US for the role of global superpower. … The name Osama bin Laden had yet to impinge on the public consciousness. …Mr Clinton’s economic inheritance was similarly golden. The frightening deficits of the Reagan years disappeared in the 1990s, partly because of sensible fiscal decisions taken by President George HW Bush. By the time Mr Clinton took office, the US economy was already recovering strongly. He was the lucky beneficiary of a surge in American productivity, following the transformation of the workplace by computers. With unemployment at just 4 per cent and inflation under control, there was exuberant talk of a “New Economy”. Given this fortunate combination of circumstances, is it any wonder that the president had time for dalliances in the Oval Office?

4) Does partisan labeling really reveal policy, especially as parties evolve over the decades? Many conservatives have attached themselves to the policies of Democrats JFK (big tax cuts) and Clinton (free trade, balanced budgets, investment tax cuts) and rejected those of Republicans Richard Nixon and George H. W. Bush. As I have written:

In a 2008 WSJ piece, investment strategist Donald Luskin noted that since 1948, the total return of the S&P 500 had averaged 16% with a Democrat in the White House and 11% with a Republican. But swap Clinton and JFK for Nixon and Bush I, and you find that the market is up an average of 15% under the GOP and 11% under the Dems.

There are 13 comments.

  1. John Walker Contributor

    James Pethokoukis: 2) There have been a dozen presidents since World War II. Is that a huge sample size, especially given changing demographics and macroeconomic conditions?

    Precisely. A sample size of twelve, even without considering other variables, has such a large statistical error that any conclusion from the data is likely not to be significant.

    • #1
    • November 2, 2015, at 3:24 PM PST
    • Like
  2. Gazpacho Grande' Coolidge

    The other piece is that even if it is a D in office, so what? Clinton was more or less forced to sign the tax cut that he later took credit for, because of the Republican majorities. Every one of his 8 budgets he submitted to Congress called for deficit spending, but the economy was booming, and revenues exceeded costs, resulting in surpluses. That’s not by design, that was a Democratic accident.

    The GDP growth in the 90’s under Clinton probably skews this metric higher. Just guessing. But there’s more to it than just who’s in office, on either side.

    • #2
    • November 2, 2015, at 3:27 PM PST
    • Like
  3. I Walton Member

    I had an undergraduate micro professor who said the government could globber an economy in one direction or another but couldn’t really do what the Fed and Keynesians believed it could do, but most professors believed the models reflected reality. That was well over 40 years ago, now the whole notion should be seen as absurd. To the extent governments stops doing harmful or threatening things, which includes threatening talk as well as harmful actions, they can affect the growth positively, and to the extent they do things that clog up the information system which a market is, or create uncertainty, impose dead weight costs, and raise risks, they negatively impact investment and hence growth. Even with this, other things drive economies up or down in its larger moves. These positive and negative affects take time, so that trying to tease out anything but the really lumpy stupidity or radically positive moves (which are not spending) is extraordinarily difficult.

    • #3
    • November 2, 2015, at 3:41 PM PST
    • Like
  4. Gazpacho Grande' Coolidge

    Federal spending is a component of GDP. If Democrats spend more – like Uncle Barry’s doing – GDP goes up.

    That doesn’t mean the economy’s doing better.

    • #4
    • November 2, 2015, at 3:54 PM PST
    • Like
  5. Son of Spengler Contributor

    Results of the study are skewed by outsized growth under JFK. Give me a current Democrat offering JFK’s tax cuts and regulatory modesty, and then we can talk about which party is more reliable in fostering economic growth.

    • #5
    • November 2, 2015, at 4:30 PM PST
    • Like
  6. Sabrdance Member

    Larry Bartels really should get more flak than he does for starting this off with Unequal Democracy. He at least had the decency to admit in the book that his model is heavily reliant on the specification. Meaning, things like choice of start year, whether you include the first year of a president’s term as the previous president’s, and so on are really driving the results. And I’m sure that a book released in an election year and obviously intended to convince people to vote for Democrats because they were good on the economy was totally above board in selecting that model which is what really drives the results.

    • #6
    • November 2, 2015, at 4:41 PM PST
    • Like
  7. JimGoneWild Coolidge

    Clinton received:

    • Start of Economic Recovery (soft landing by Bush)*
    • Cold War Peace Dividend*
    • Savings & Loan Bailout Ending
    • Downsizing benefits from end of Desert Storm
    • Internet age brought in by Deregulating the Phone System
    • Cell Phone Bandwidth Auction Revenue (about $40 Billion)
    • Tax Revenue surge from Reduced Rates on Investment*
    • Welfare Reform
    • Continued Strong Dollar

    Thank god, Bill and Hillary scared the hell out of voters and in came a Republican Congress. Congress stopped all the crazy schemes coming out of the White House, like taxing toilets, HillaryCare, and forcing telephony equipment makers to put in backdoor access to everything.

    *Mentioned above.

    • #7
    • November 2, 2015, at 4:46 PM PST
    • Like
  8. drlorentz Member
    drlorentz Joined in the first year of Ricochet Ricochet Charter Member

    John Walker:

    James Pethokoukis: 2) There have been a dozen presidents since World War II. Is that a huge sample size, especially given changing demographics and macroeconomic conditions?

    Precisely. A sample size of twelve, even without considering other variables, has such a large statistical error that any conclusion from the data is likely not to be significant.

    That’s way too sweeping a statement. A sample of 12 is plenty if the correlation coefficient is high and the effect is strong. How many samples you need depends on the variance of the data. You can’t rely on gut feel to decide. The authors of the paper claim in the abstract that the effect is “both large and statistically significant.” So much for the feelz.

    Before assessing the significance of an effect, one has to formulate a well-defined hypothesis to test. To that point, #1 matters more. The authors use the time-frames of the presidential terms, which does not account for any policy lags. A more competent analysis would

    1. Do a cross-correlation between economic time series and presidential terms
    2. Do more to disentangle confounding factors, possibly a multi-variate fit.

    This is not a first-class work. Krugman likes it because of confirmation bias. That shows he’s not first class either. Then again, socialist are supposed to be classless, right?

    • #8
    • November 2, 2015, at 4:58 PM PST
    • Like
  9. Zafar Member

    John Walker:

    James Pethokoukis: 2) There have been a dozen presidents since World War II. Is that a huge sample size, especially given changing demographics and macroeconomic conditions?

    Precisely. A sample size of twelve, even without considering other variables, has such a large statistical error that any conclusion from the data is likely not to be significant.

    If you’re comparing years (growth rate per year) then your sample size in 68 (? years between 1947 and now) with the President’s party being one of the variables. Which would still be a small sample, right?

    • #9
    • November 2, 2015, at 6:22 PM PST
    • Like
  10. Hammer, The Member

    John Walker:

    James Pethokoukis: 2) There have been a dozen presidents since World War II. Is that a huge sample size, especially given changing demographics and macroeconomic conditions?

    Precisely. A sample size of twelve, even without considering other variables, has such a large statistical error that any conclusion from the data is likely not to be significant.

    significant, schmignificant. I heard Hillary say it at the dem debate and it got a huge applause. I’ve got liberal friends who will quote it with all the snarkiness of Jon Stewart. Why? Because snark doesn’t bother to ask questions. Unfortunately, it’s easy and it sounds good. Just ask all those women who make 72c to every dollar men make.

    • #10
    • November 2, 2015, at 8:36 PM PST
    • Like
  11. Sabrdance Member

    Zafar:

    John Walker:

    James Pethokoukis: 2) There have been a dozen presidents since World War II. Is that a huge sample size, especially given changing demographics and macroeconomic conditions?

    Precisely. A sample size of twelve, even without considering other variables, has such a large statistical error that any conclusion from the data is likely not to be significant.

    If you’re comparing years (growth rate per year) then your sample size in 68 (? years between 1947 and now) with the President’s party being one of the variables. Which would still be a small sample, right?

    No, the fixed effect for each president is going to eat degrees of freedom like Pac Man (I’m going to assume no statistical idiocy like not controlling for autoregression among Presidents and years). The model might converge without chicanery, but it is still not worth drawing conclusions from even if you’re popping p<0.05. Secular trends in the data unrelated to Presidential behavior, random noise, and uncontrolled autocorrelation are far more likely to be driving the results, and a single outlier (like Nixon or Clinton) could be the difference between significance and non-significance.

    Never forget: Pollsters can’t count, and Presidency scholars are even worse. (Don’t get me started on Congress scholars.)

    • #11
    • November 2, 2015, at 10:00 PM PST
    • Like
  12. Gazpacho Grande' Coolidge

    Even a correlation isn’t causation, and policies, as implemented, are so gray in terms of scope and implementation timelines that to claim, on either side, that a D or an R in the White House means a better economy is kind of crazy.

    The test that can be made will never be made. The test would be deconstructing half the government’s spending and regulation – let’s say by 25%. Then see what happens for five years. Then you’d have an indicator of which party’s (supposed) policies are better for the economy.

    • #12
    • November 3, 2015, at 3:28 AM PST
    • Like
  13. Peter Murphy Member

    There is another explanation for the D-R gap. Let’s call it (in the spirit of James Piereson’s Shattered Consensus) regime dynamics. With the onset of the New Deal the post-1932 regime of American political economy was Democratic, dominated by big-spending good-times liberalism. Unsurprising big spenders, both in government and in the wider economy, flourish in good times. Republicans clean up the hang-overs from the good times. They govern in recessions and lower-growth eras. This was not always so. If we go back to before FDR, to an era that was spiritually dominated by Republicans rather than Democrats, Republican presidencies outdid Democrat ones. Blinder and Watson consider data going back to 1875. If you include the FDR years then Democrats still outdo Republicans. But if the FDR years are excluded, that is if we stick just with the Republican-dominated era of limited government typified by Coolidge, then ‘growth was actually higher under Republicans. So one might say that higher growth under Democrats began with Hoover.’ The lesson of this for Republicans is to stop being the party that voters turn to when the economy begins to slow and economic exuberance wanes—and start creating their own affirmative can-do political economy based on the kind of small-government high-growth ethos that in the 60 years before FDR had Republicans dominating the spirit of the age and the partisan gap.

    • #13
    • November 4, 2015, at 4:29 AM PST
    • Like