The political scientist Paul Musgrave on Juneteenth used the occasion to write about long-term racial differentials in unemployment levels. Remembering the Global Financial Collapse — a term that seems to have supplanted the Great Recession — and the nearly 10% overall unemployment rate by the fall of 2009, which was big — he then pointed out that the “average unemployment rate (seasonally adjusted) for Black Americans from 1972 until 2023 has been 11.6 percent.” Given the typical definition of a generation as 20 years, that’s 2.5 generations worth of crushing unemployment.
White people have overall experienced much lower rates of unemployment. Let’s go back to November 2009 and see how this plays out. Instead of the national 10% seasonally adjusted unemployment rate, let’s go with the not seasonally adjusted 9.4% number, because seasonally adjusted didn’t exist for Asians at the time, let alone some other groups not even mentioned. For whites, it was 8.6%. Blacks experienced 15.2% unemployment. Hispanic and Latino ethnicity, 12.2%. Asian unemployment was 7.3%.
Musgrave was correct in his thought that a long-term view is important. But so is a shorter-term one, because there is where we can start to see change that might get masked in the greater timespan. Below is a graph I put together on the Federal Reserve Bank of St. Louis’ FRED site, again using not seasonally adjusted data. (A side note: black unemployment was not tracked until the beginning of 1972. Hispanic tracking didn’t start until 1973. Asian, not until 2000. What data organizations and government don’t track can say as much sometimes as what they do.)
Over time, unemployment rate gaps between races have shrunk considerably. However, the results still aren’t where they need to be for anyone who thinks that systemic fairness or even, at a selfish market-level view, having as many people available for a hungry labor market is important.
An interesting take is to look at the ratios of group unemployment rates compared to those of whites at the same time. Here is a graph.
Musgrave did something similar for black and white unemployment rates from 1972 to 2023. I’ve chosen to look at the data a little differently, using non-seasonally adjusted data and only from 2000, when there was data for blacks, Hispanics, and Asians as well.
There has been compression of the ratios, but only so much. Asians remain below white unemployment rates. Hispanics saw one of the least amounts of additional unemployment in May 2023. Only August 2019 equaled it, with May 2006 and September 2019 bettering it.
Black unemployment, however, is again on a relative upswing compared to white. In the unsmoothed numbers, it has returned to 80% over white unemployment. The ratios are jittery, and this is why people often look at seasonally adjusted numbers for some sense of continuity and the ability to consider longer-term trends. It’s like looking at moving averages, another form of smoothing that can bring insights that up-and-down jumping won’t show.
The bigger question is what will happen in the next recession, as there is always another one in the future. In recessions, unemployment rises. Depending on the severity, the unemployment rate gaps can grow enormous.
Now would be the time to prepare, to think in advance of putting aside funds to address a problem. Unfortunately, that seems unlikely. Typically, such spending only occurs after a disaster of large enough scope that it rises above normal politics and leaves politicians in a spot where they have to do something.
If a recession seems mild, chances are there wouldn’t be significant help offered. Hopefully, the ratio compressions would not significantly reverse with a national number masking the increased impact on specific groups. But that will have to be seen.