Presidential Unemployment

A while back, I was discussing politics with someone online (I know, I know...not the best move, perhaps) and he made the claim that Trump was not given enough credit for the amazingly strong economy, pointing to the fact that unemployment was at an historic low. Since that's a claim that can be pretty easily verified—at least the part about unemployment being at historically low levels—I went out to grab some data.

The Bureau of Labor Statistics has the official unemployment rate each month going back to January of 1948. As I write this in July 2018, the data goes up through last month. The data set starts partway through the Truman administration and ends part way through the Trump administration. So any conclusions we make about Truman or Trump will have to have the caveat that they're not fully covered by the data.

Anyhow, the first thing I noted was that the unemployment rate was low, but it kind of depended on what you meant by "historic low." For example, 37 months of Truman's administration were equal to or lower than it is now (the June 2018 rate is 4.0%). Presidents Eisenhower, Johnson, Nixon, and Clinton also saw multiple months with lower unemployment. Still, that's fewer than half of the presidents during that time period.

The average of all the monthly rates is 5.8%. The lowest rate was 2.5% for two months in 1953 (Eisenhower's 4th and 5th month in office). So it's certainly below average, and ever so slightly closer to the minimum than the average, but clearly not any kind of record. Heck, the entire decade of the 1950s averaged only slightly higher than what we've got now.

Looking at the chart, unemployment had clearly been dropping for more than Trump's time in office. The most recent peak of unemployment happened in the first year of Obama's administration, and then with few minor exceptions, dropped consistently and continually for the rest of his presidency. The slope of the curve doesn't change when Trump takes office, so it really looks like he's benefiting from the momentum of Obama-era policies. If Trump's policies had had a positive effect, you'd see the rate dropping faster once he came into office. But it doesn't. It just follows roughly the same slope as Obama's drop, or if anything it might be slightly less steep.

The chart also got me thinking about the cyclical pattern of unemployment. Clearly there are ups and downs and I wondered if those correlated with different presidents. So I broke this down by presidency. Because there seemed to be a lag (Eisenhower's lowest point was a few months into office; Obama's highest rate was in his 9th month), and because the data was by month, I "rounded up" each presidency to the next month. So the outgoing president claims all of January, and the new president "starts" on the first of February.

This showed me a really interesting pattern.

Every Democratic president saw unemployment lower at the end of their term. And with two exceptions, every Republican saw unemployment higher when they left office.

The two Republican exceptions are Ronald Reagan and Donald Trump. As mentioned before, Trump's only about a third of the way into his presidency, so a lot can still change...especially since there does seem to be a period of inertia at the beginning of a president's time in office. Although Reagan did see a drop of 2% comparing his first February and his last January, in between saw the highest unemployment rate in the entire data set. (Was it the highest ever? Undoubtedly not—the Great Depression isn't part of this data, since the survey it's based on wasn't even created until the mid 1940s.)

Let's set aside Trump for a moment and look only at the former presidents. There was something of a range of different lengths in office, since there were assassinations, resignations, and so forth. Dividing the total change for the president by the number of months he was in office, gives us an average monthly change. Ranked by this number, Obama is the clear, hands-down winner. During his presidency, unemployment went down by the greatest amount: 3.5%. The next two most beneficial (i.e., lowered the rate) presidents were Clinton at 2.9% and Johnson at 2.1%. Kennedy, because of the tragically shortened presidency, sneaks in between Obama and the other two when viewed on a per-month basis. All four lowered unemployment by 0.03% per month and differed only in the next smaller decimal place.

On the other side of things, the worst former president for unemployment was Eisenhower at an even 4%. George W. Bush came in second, seeing unemployment rise on his watch by basically as much as Obama lowered it: 3.6%. His father George H. W. Bush and Richard Nixon tie for third at 2.1%. Weirdly, on a monthly average basis, Ford jumps to the front of the line. But the two presidents Bush and Eisenhower still round out the top four in increasing unemployment.

By far the smallest change in a presidency was Jimmy Carter. If I counted him from January to January, the change would be zero, but Feb to Jan results in a drop of only 0.1% total. The next closest is Truman at 0.5%. And this is a good place to bring Trump back in. Why? Because his impact so far is 0.9%, so he'd be third for the smallest impact on the unemployment rate...at least as it currently stands.

Bringing it back to the comment that started this little analysis: should Trump get more credit for historically low unemployment rates? Well, probably not. They're low right now, but not historically. The drop in unemployment we're seeing now has been going on since Obama's first year of his first term. That, and the impact since Trump has taken office is the third smallest since 1948. So for those reasons, I don't think Trump can reasonably take credit for today's unemployment rate.