Democratic presidencies have lead to far higher growth rates post WW2

[eric82oslo]

While the 9 post World War 2 Republican presidential periods have averaged an annual GDP growth rate of just 2.5% (going from W's 2nd term 0.5% growth to Reagan's 2nd term 3.9% growth), the 7 first post World War 2 Democratic presidential periods on the other hand (Obama's 2nd term obviously not included yet) have shown an astonishingly higher 4.3% annual GDP growth rate. That is to say, the average GDP growth rate of a Democratic president has been considerably higher than the single most impressive presidential period of any Republican! The 7 Democratic presidential periods have seen average annual growth rates of between 2% (Obama's 1st term) to amazing 6.6% during Truman's 2nd term, closely followed by Kennedy's 5.8% and Johnson's 4.9%.

Here's the chart:



The Economist's article: http://www.economist.com/news/united-states/21611143-why-economy-has-grown-faster-under-democratic-presidents-timing-everything

So is this Democratic triumph of proportions due to just pure luck, as The Economist's writer suggests it might be, or is it due to smart economic policy proposals?

[RI]

Holy correlation and small sample size, Batman. The president's ability to affect economic growth is laughably small.

[bgwah]

Holy correlation and small sample size, Batman. The president's ability to affect economic growth is laughably small.

65 years is a small sample size? What would be sufficient then?

[Maxwell]

Holy correlation and small sample size, Batman. The president's ability to affect economic growth is laughably small.

This, and any Presidential effect is undoubtedly delayed some years.

[GaussLaw]

Holy correlation and small sample size, Batman. The president's ability to affect economic growth is laughably small.

65 years is a small sample size? What would be sufficient then?

Significantly more, considering that the number of Presidents here is fairly small.

Plus, Congress arguably is more important and this chart doesn't take that into account.

[King]

With the exception of the Bushes, I would say it's not significantly larger for the Democrats.

In today's context, however, there's no way the current Republican Party if elected can stimulate anything because they consider the stimulation of the economy by the government to be a myth conspiracy.

[Antonio the Sixth]

Love how growth was higher under Carter than in Reagan's first term and barely lower than in his second term, but somehow Carter was an "utter disaster" and Reagan was an "economic genius".

[○∙◄☻¥tπ[╪AV┼cVê└]

What's incredible is that from Hoover to Bush Jr. the average annual job growth of all 6 Democrats beat each of the 8 Republicans. BTW, the final numbers for Bush Jr. were 0.1. The early 2009 job numbers really hurt Obama, but he's about 0.7 now.

[King]

Love how growth was higher under Carter than in Reagan's first term and barely lower than in his second term, but somehow Carter was an "utter disaster" and Reagan was an "economic genius".

High standards. Hyperinflation and the energy crisis didn't make the economy feel strong, especially considering what Americans had been accustomed to out of an economy since WWII.

[GaussLaw]

Love how growth was higher under Carter than in Reagan's first term and barely lower than in his second term, but somehow Carter was an "utter disaster" and Reagan was an "economic genius".

High standards. Hyperinflation and the energy crisis didn't make the economy feel strong, especially considering what Americans had been accustomed to out of an economy since WWII.

Yeah.  The Dems have had an inflation problem, especially w/ Truman and Carter.

Jobs growth isn't the only thing.  You need a full economic index which includes inflation, unemployment, etc.

[King]

Old school Keynesians thought inflation was directly correlated to long term job growth (see the Phillips curve). It was a devastating error on their part and that eroded public trust in the left.

[MATTROSE94]

Holy correlation and small sample size, Batman. The president's ability to affect economic growth is laughably small.

65 years is a small sample size? What would be sufficient then?

Significantly more, considering that the number of Presidents here is fairly small.

Plus, Congress arguably is more important and this chart doesn't take that into account.

I agree. It would be interesting if there were a chart that took into account Congressional control and the rate of economic growth over the past few decades.

[King]

65 years is plenty of data. The federal government's influence on the US economy was relatively negligible before World War I anyway.

[RI]

I'll give another go at this. I don't see anything in the chart or in the link that specifies if this is real or nominal GDP growth, so generally we can assume the latter.

Carter's growth could be artificially higher from inflation. Indeed, let's look at Real GDP growth by term:

Truman 1: -2.2%
Truman 2: 6.0%
Eisenhower 1: 3.1%
Eisenhower 2: 2.1%
Kennedy: 5.7%
Johnson: 5.6%
Nixon 1: 3.4%
Nixon 2: 2.3%
Carter: 3.4%
Reagan 1: 3.5%
Reagan 2: 4.1%
Bush: 2.3%
Clinton 1: 3.5%
Clinton 2: 4.5%
Bush 1: 2.5%
Bush 2: 1.1%
Obama 1: 1.5%
Obama 2: 1.8%

Dem Avg: 3.3%, 4.0% without Truman 1
Rep Avg: 2.7%

Carter was not higher than either Reagan term once inflation is accounted for.

Obama, Clinton, Carter, Kennedy, and Truman 2 were all initially elected in the midst of economic downturns (<2.5%), which inevitably have recoveries eventually. They had much more opportunity for growth as a recovering economy can grow more quickly than a full employment economy due to unused resources.

Each Republican was initially elected during a time of economic growth (>3%), which always ends eventually and has a lower ceiling on growth.

If anything, I'd say presidential party choice is a lagging economic indicator with incumbency effects mixed in.

[AggregateDemand]

Whatever helps Dems sleep at night.

[Okay, maybe Mike Johnson is a competent parliamentarian.]

Going directly from several well-thought-out, pertinent posts to the R-TX Gruesome Twosome is certainly an interesting reading experience.

[Foucaulf]

I'll give another go at this. I don't see anything in the chart or in the link that specifies if this is real or nominal GDP growth, so generally we can assume the latter.

You (and everyone else) can just read the paper. It's not actually that technical: the authors took quarterly GDP data and regressed a bunch of time series on them, controlling for correlation between the datasets. (Figures are in real GDP by the way)

The statistical significance of different growth rates and recession times under Democratic and Republican presidencies suggest there's some relation between the party in charge and the economy. The problem is that common economic explanations for growth rate changes predict some of this differential, but not all - plus the predicted effects change depending on which time period you look at.

Without reading the paper too closely, I have two crackpot hypotheses. The first one is more precisely stated but I think the second is more accurate.


1) What if the target of government spending is a better predictor? By this I don't mean whether government spending is towards defense or welfare, but whether Democrats favour policies giving money to local government compared to Republicans. We may think that local government is more cash-starved and has a bigger bucket list of projects to build, so subsidies toward state/municipal governments as an incentive to drive policy creates more growth.

2) We know that, under the New Deal coalition, Democrats represent more marginalized social groups in society. We can believe that systematic discrimination depresses their wages and consumption. The hypothesis is that the times these groups give and support the Democrats are also the times when they begin exploring new technologies or professions that have given them a boost in wages, an effect which persists for a few years or so after the campaign is over.

[Gustaf]

I read this a while back and from what I can recall they did take into account things like congressional control and lags in policy effects. And that after the full analysis they decide that it's most likely more luck.

And yes, it is a pretty small sample size.