New York Times columnist Peter Coy raised an interesting point about the September jobs data. Most of us were celebrating the 254,000 job gain reported for the month, in addition to the upward revision of 72,000 jobs to the prior two months’ data.
Perhaps even more importantly, the household survey reported the unemployment rate fell 0.1 percentage point to 4.1 percent, a level that is very low by any historical standard. This partially reversed a 0.6 percentage point run-up from January to July. At least there is no clear upward trend in the unemployment rate.
But Coy pointed to another item in the jobs report. There were 3.45 million people (2.0 percent of the workforce) who reported being unemployed in August, who still reported being unemployed in September. Coy points to this, as well as some other data, and cautions that Vice-President Harris should be cautious about touting the current state of the labor market.
Coy’s column raises two distinct questions.
First how good is the labor market. He rightly points out that it is not as strong as in 2022-23, when we were seeing labor shortages in many areas. The second question is how important the weakening of the labor market is likely to be politically in the fall elections.
How Good/Bad Is the Labor Market?
The first question is how much we should be concerned about the weakness identified by Coy. The point here is not that 4.1 percent of the labor force is unemployed, but rather that workers no longer feel good about their job prospects and are therefore reluctant to quit their current jobs.
There is evidence to support this view. The quit rate, the percentage of workers who voluntarily quit their jobs, was 1.9 percent in August. (We don’t have September data yet.) That is down from a peak of 3.0 percent in the fall of 2021 and winter of 2022.
A separate measure, the percentage of unemployment due to people who voluntarily quit their job tells a similar story. These are people who feel sufficiently confident in their labor market prospects that they are prepared to quit their current job before they have a new job lined up. This figure stood at 12.1 percent in September, down from a peak of 16.0 percent in September of 2022.
Clearly people are less anxious to leave their jobs now than they were earlier in the recovery, but it is worth asking how these numbers look in a longer time frame. Measured as a share of the labor market, the 3.45 million people who were unemployed in both August and September is somewhat higher than the lows seen just before the pandemic, but roughly the same as the numbers were saw in 2018, which was generally regarded as a very good labor market.
It’s also lower than the lowest levels reached in the recovery from 2001 to 2007. The share is also roughly equal to share hit in 1999, although somewhat higher than the low points hit the following year. Until the current recovery, the late-1990s boom was considered the strongest labor market since the 1960s, so being slightly worse than the best year hardly sounds like a serious indictment.
The picture looks somewhat worse with the quit rate. Before the pandemic the quit rate peaked at 2.4 percent in January and July of 2019. It averaged 2.3 percent in both 2018 and 2019. In the recovery from 2001-2007 it peaked at 2.3 percent in September of 2005 and then averaged 2.2 percent in 2006, edging down slightly the following year before the collapse of the housing bubble led to the Great Recession. The series begins in December of 2000, just before the economy went into recession, but it managed to hit 2.4 percent before a weakening labor market sent it downward.
The share of unemployment due to voluntary quits tells a similar story. That peaked at 15.2 percent in June of 2019. The year-round average for 2019 was 13.6 percent. The peak in the recovery before the housing bubble collapse was 12.3 percent in May of 2006, only slightly better than the September reading, but we had year-round averages of 13.3 percent and 13.7 percent in 1999 and 2000, respectively. Clearly people were more willing to quit their jobs in prior business cycle peaks.
But there are two sides to these low rates of labor turnover. One can be the story of a weak job market discouraging people from taking chances. The other is that people might be relatively satisfied with their current jobs and therefore see less reason to leave.
That explanation should not sound far-fetched. We had record rates of turnover from 2021 until the middle of 2023. Many people who had been in bad jobs left them for jobs they expected to be better. This is likely a major reason why the Conference Board found a record level [2] of workplace satisfaction in its survey last year.
It is also worth noting that wages continue to grow at a healthy pace. The annualized rate of wage growth over the last three months was 4.3 percent, up slightly from the year over year rate of 4.0 percent. With the current pace of inflation this corresponds to real wage gains of close to 2.0 percent, better than workers have seen through most of the last half century.
There are two points to be made about this pace of wage growth. First, if the labor market is really weakening sharply, it’s hard to see why the pace of wage growth would be accelerating. The data are sufficiently erratic so that we may not want to make much of the modest rise in the annualized rate over the last three months, but we sure can’t say that the pace of wage growth is slowing. The other point is that if workers are seeing healthy real wage growth in their current jobs, it is not surprising that they would be less likely to look to quit.
To be sure, there is little doubt that the labor market has weakened from where it was earlier in the recovery. To my view, I would be happier with an unemployment rate closer to the 3.4 percent low hit in April of 2023. But if we look back over the last three decades, the labor market is still looking very good.
By most measures it is hugely better than the average over this period and only exceeded by the peaks of the business cycles in 1999 and 2000, and in 2019. If we were just comparing the September 2024 labor market to prior decades, and not the extraordinary labor market earlier in the recovery, we would almost certainly be touting its strength.
The Politics of the Current Labor Market
The Coy piece initially sent me up the wall since it runs 180 degrees opposite to what the political experts were telling us when the labor market was at its peak strength earlier in the recovery. When people like me pointed to the extraordinarily low unemployment as evidence of the success of Biden’s policies, we were treated to a line like “unemployment only affects a few million workers, whereas inflation affects everyone.”
This was a profoundly stupid thing to say since it completely misunderstands how the labor market works. We may only see 6-8 million people unemployed at a point in time, but in a normal month 6 million people lose or leave their job. This means that over the course of year tens of millions of people are changing jobs. This means a very large segment of the population is directly affected by the strength of the labor market.
I’m glad to see Coy recognizing the point that the strength of the labor market has an impact far beyond just whether or not people are unemployed. It would have been nice to see more people making this argument back in 2021-2023.
What Do We Tell Harris?
Fortunately, I don’t have to decide what Harris should say about the labor market and the economy, I’m just an economist, not a political consultant. I will say that by any historical standard the labor market is doing pretty damn good. It could be better, but a low unemployment rate and rapidly rising real wages is a better story than any incumbent administration could tell since – 2000, oh well.