A couple of weeks ago, Donald Trump retweeted an image with nine charts about the US economy that purported to show a central thesis: The Obama presidency has been really bad for America. Student loans have skyrocketed; food stamps, health care costs, and federal debt all rose swiftly; and labor force participation, median family income, and homeownership all fell dramatically.
It’s hard to see the labels on either the vertical or horizontal axes, so it’s difficult to discern exactly what each graph is showing. As Philip Bump at the Washington Post has ably shown, many of the graphs cut off halfway through the Obama presidency, cut off the vertical axis, or mix time frames.
But most politicians are wont to take the broader view that the country’s economic picture is tied to the exact time a president is in office. Unfortunately, that thesis is inherently flawed. A president can’t control what is happening in the economy (or the world) prior to his (or possibly her) election to office.
President Obama was elected on November 5, 2008, but wasn’t sworn into office until January 20, 2009. Does that mean we should attribute everything in the economy—good and bad—to him starting in November or January?
A careful look at unemployment rates shows how that approach misses the big picture. A year before Obama was elected (in December 2007, when the recession officially began), the unemployment rate hovered around 5 percent and stood at exactly 5.0 percent in April 2008. In May, the unemployment rate began to tick up steadily, reaching 6.8 percent in November, the month Obama was elected. It continued to climb thereafter, reaching 7.8 percent in January 2009.
The unemployment rate would, of course, continue to climb as the economy soured, ultimately reaching a peak of 10 percent in October 2009 before beginning its slow decline back down to its current rate of 4.7 percent.
Data on participation in the Supplementation Nutrition Assistance Program (SNAP), formerly known as food stamps, tell a similar story about the importance of context. Again, participation was edging up prior to the presidential election and continued to rise after Obama took office. Two pieces of context are key.
First, such increases are a primary feature of many safety net programs: they are designed to expand to meet need during economically troubled times.
Second, there was legislation that served to further increase participation: the economic recovery act of 2009 (signed by Obama) intended to increase SNAP participation (and thereby reduce food insecurity during the recession).
What should politicians take from these examples? Yes, both SNAP participation and the unemployment rate continued to rise during the Obama presidency, but as we’ve seen, he inherited an economy that was already tanking when he took office.
We can also see that both series have declined over the past two years; perhaps not as fast or as low as one might like, but they have declined. Whether those changes are due in part or at all to the president is also a matter of debate. We can certainly argue about whether participation is too high and that it did not start to decline sooner, but that’s a policy argument, not a data argument.
So let’s zoom out a bit and see what we can learn from SNAP over the past 40 years. Here, I’ve placed SNAP participation on the horizontal axis and the unemployment rate on the vertical axis (each point represents a year). I’ve also added a color dimension—red shows years in which a Republican was in the Oval Office and blue shows years in which a Democrat was in the Oval Office.
If you look carefully, you can see trends in and around presidential terms that you can’t necessarily pin on that president.
Take, for example, the transition between Presidents Ford and Carter. The unemployment rate dropped between 1975 and 1977, while SNAP (then food stamps) participation increased slightly. Once Carter took office in January 1977, the unemployment rate continued to decline for another couple of years. Is that to his credit or to existing trends and policies?
Similarly, after President G.W. Bush took office in January 2001, the unemployment rate started to increase swiftly; but the Internet bubble collapsed starting around 1999, so is it fair to assign blame for the early-2000s recession to Bush?
This is all to make a simple point: The economy is much too complex to simply pin a president’s swearing in as a turning point. That nuance, though important, is not something our presidential candidates typically deem worthy.
As we move into the general election phase and the candidates and parties try to assign blame for this and that to the other, let’s keep in mind that those dates are often arbitrary and have little meaning. Maybe nuance would be a good thing.