gdp and presidents

Doing some homework, I found 3 surprising trends in U.S. gdp growth since 1959. 1) Democrats are better. 2) Transition years are worst. 3) We steadily approach zero growth by 2030. I didn't expect any of those results. The common narrative says that only Republicans are good for business, and American prosperity is always improving. It turns out that both points are demonstrably (and blatantly) false.

I am quite pleased to see that the trend for the last half-century is towards a zero-growth economy. I've been arguing for years that we should plan for zero growth, as our productivity is inhibited by the harsh reality of dwindling resources. It seems that we're getting there anyway, dragged down to zero growth against our collective will. It bodes poorly for our financial system, though, since it requires growth for it to function at all. We're losing our ability to feed perpetual growth.

I was compelled to look up the numbers when the Wall Street Journal published an article that blamed Obama for poor gdp performance when compared to prior presidents. Noticeably absent from their graph of data was any reference to Bush... either of them. Cherry-picked data is a good warning sign that someone may be misleading you. So I searched for more data. I don't necessarily accept that gdp is the best measure of an economy. Other people think it is, however, so I wanted to look at the numbers that are important to them.

I started with the data in the Wall Street Journal's table, then I added all of the years that were available in a 2009 report from the U.S. Department of Commerce. I pulled from Wikipedia a simple list of presidents with their political party and year of inauguration. I assigned a political party to each calendar year, so it was easy to determine credit or blame for the relevant political party. My first collection included numbers straight from those other tables. Afterwards, I pondered various kinds of leniency, and the results just enhanced the original finding. Click the photos twice to see their largest versions.

Data set 1: "Strict Accountability" I assigned a political party to each calendar year strictly by which president entered office that year. It's the easiest way to determine "responsibility" for the economy, if not the smartest way. Results: Democrats perform better than Republicans. The slow trend across half a century is toward zero-growth gdp. The 2009 recession is an obvious blemish on the national record.
Data set 2: "Shifted Accountability" I allowed that the first year of a presidency is actually following the previous president's budget. There is fair reason to think that the newcomer has little influence over existing expenditures, so I swapped the assignment of those first years to the outgoing president (rather than the incoming). Results: Democrats perform slightly better than before, winning again over Republicans. The worst 3 performances of the last half-century belong to Republicans.
Data set 3: "Pardoned Transition" I considered that those transition years were more accurately a mixture of differing policies, so I just assigned them all to a mythical "Transition" regime. I figured that exception would be fair to all incoming presidents. Results: Democrats perform slightly better than before, winning again over Republicans. Everyone gets excused for the 2009 recession, but the next largest negative-growth year is solidly Republican. Transition years perform worse than either political party. I can't tell if a bad economy leads to a party change ("It's the economy, stupid."), or if a party change leads to economic uncertainty and poor growth. Whichever leads the other, the association is clear; transition years are worse for all of us. The maximum, minimum, average, and median are all lower during a political transition year.
Data set 4: "Pardoned Transition + Pardoned Outliers" I also allowed for "an off day". I eliminated the single best and single worst year remaining for each political party, removing their most unusual performances from their trends. Results: Democrats perform slightly better than before, winning again over Republicans. The only negative-growth years during the last half-century are all Republican.

I prefer to think that Congress (which writes the laws) holds more influence over the economy than the president. It might be wishful thinking on my part, though, rather than reality. I would also like to see how these trends hold up when comparing gdp growth to income inequality or changes in gas prices. I figure that income inequality is a dead weight slowly dragging on the economy, while gas price shocks may have more immediate effects on gdp.

Here is an animated GIF including the 4 datasets described above. On the right, you can see how each new leniency improves the performance of the Democratic presidency. The Republicans are more consistent, yes, but they are consistently less productive than Democrats. That conclusion should be no surprise since the Congressional Research Service exposed the lie that "trickle-down economics", a staple of Republican policy for decades, ever actually helps saving, investment, or productivity growth. It's the transition years, though, that gather the lowest gdp changes.


I close this blog post with a message for Microsoft. I disapprove that SQL Server 2012 Express contains no inherent aggregate function for computing MEDIAN value, as it does for AVERAGE, MIN, and MAX. (I used the versatile PERCENTILE_CONT, but it's still more complex than a "MEDIAN(column)" function should be.) I also could find no inherent aggregate function for computing a TRENDLINE on my data or graph. Why, Microsoft? Why? I'm sure I used handheld calculators from Texas Instruments with those features... 20 years ago.