Some Thoughts on Unemployment
Some Thoughts on Unemployment
It's higher than we say, and benefits aren't the problem.
Note: any and all opinions are entirely my own and do not necessarily reflect those of the Cleveland Fed, the FOMC, or any other person or entity within the Federal Reserve System. I am speaking exclusively for myself in this post (as well as in all other posts, comments, and other related materials). No content whatsoever should be seen to represent the views of the Federal Reserve System.
Hindsight is 20/20. Though policymakers underestimated the pace of vaccination (despite an explicit goal of 100 million doses in the first 100 days, which even at the time was known to be undershooting what was possible), pandemic unemployment insurance benefits were likely extended past their necessity. But is this holding the economy back? Are people sitting at home sponsored by the federal government with unemployment benefits instead of looking for jobs? Perhaps some, but not many, and not most.
Currently, benefits (true name: Federal Pandemic Unemployment Compensation, or FPUC) expire in September. How did we get here from March 2020, though? The CARES Act, passed in the opening act of Covid-19 in the United States, shelled out an additional $600 per week until July 31st. All benefits were paid in addition to state benefits. In December 2020, the “$900 Billion Coronavirus Relief Package” (as it is unattractively referred to) reinstated $300 in additional benefits, set to expire in March 2021. The American Rescue Plan (ARP) was signed into law by President Biden in March of this year, and extended the $300 benefit through the summer until the first week of September 2021.
In the figure above I show two measures of unemployment (measured along the right axis) along with the size of the FPUC since March 2020. The future is unwritten with regards to the unemployment situation, which is why I’m writing this. I will seek to explain two things: first, what’s the deal with the two measures of unemployment, why does it matter? And second, are benefits keeping people from seeking work?
The first question is more easily discussed. Many economists, from the Federal Reserve to the ivory tower, have pointed out that the unemployment rate is misleading due to changes in the size of the labor force. Not only is it misleading, but it is also an understatement of the real situation.
The unemployment rate is calculated by the number of unemployed people divided by the size of the labor force. Who counts as unemployed? From the Bureau of Labor Statistics:
Persons are classified as unemployed if they do not have a job, have actively looked for work in the prior 4 weeks, and are currently available for work. Persons who were not working and were waiting to be recalled to a job from which they had been temporarily laid off are also included as unemployed.
The labor force is thus all of the unemployed plus all of the employed. Those not in the labor force include students, retirees, stay-at-home parents, and most importantly for our purposes, those who are unemployed but not seeking employment.
In February 2020, right before the pandemic began to require vast shutdowns of many, if not most, sectors of the economy, the labor force totaled 164,448,000 people with 5,716,000 unemployed; a 3.5% unemployment rate. Once the pandemic struck, the composition of the labor force changed: many individuals on the employed-side of its composition shifted to the other side. However, some individuals left the labor force altogether. This might have occurred for a number of reasons. Perhaps they were on the verge of retiring anyway, or they went back to school (a pattern of previous recessions), or suddenly had to be home due to online schooling. In any event, two things happened: the number of unemployed individuals rose (and the number of employed declined), while the labor force as a whole declined. Employment and the labor force are shown in the graph below; the gap between them is the unemployment level.
This occurred after a record run of job gains. America had not gone as long without a recession -- from mid-2009 until early 2020 -- since before World War 2. Even still, this recession was not caused by some natural occurrence of business cycles, but rather by a pandemic. The smaller labor force and larger unemployment level is not due to an inherent excess in or a cycle of the economy, but rather due to a microscopic virus. Perhaps, then, it is important to consider an unemployment rate that takes account of a reduced labor force. We can construct a counterfactual for the labor force had the pandemic not occurred by extrapolating the trend from January 2019 through February 2020 forward. In the graph below, we see three new features. First, the potential size of the labor force had the pandemic not occurred; second, the labor force participation gap, which is the difference between the potential and actual size of the labor force; and third, true unemployment, which takes account of the fact that the labor force has declined.
In these terms, for the labor market to truly recover we must not just close the unemployment gap to zero but rather close the true unemployment gap to zero (not literally zero, but close to zero). Otherwise, all the workers who fell out of the labor force over the last year and a half will be left behind and our economy will be left understaffed and running below potential. This has long-running implications due to hysteresis, as workers lose skills and find it more difficult to re-enter the labor force over time.
Speaking of understaffed -- are benefits part of the problem? Are they keeping people out of employment? To start, let’s get a sense of the numbers here. First, let’s see the number of continuing claims -- people who have already filed for unemployment initially and are continuing to do so and claim benefits -- along with the number of the unemployed.
Next, let’s look at the percentage of the labor force receiving benefits, in terms of both the actual and potential labor force.
In the first graph, we can see the situation where the labor force decreases but “true” unemployment is higher than the reported number. This is likely occurring when continuing claims rises above the unemployment rate (as in summer 2020). In the second graph, we can see that at its peak in May 2020, nearly 15% of the labor force was receiving unemployment benefits, and that currently this number is around 2%, no matter how you define the labor force. Of course, the fact that one is claiming unemployment benefits does not mean that they are in the labor force, but this remains a useful metric nonetheless. As a proportion of all the people working or unemployed (per the official definition), only 2% are receiving benefits. As a proportion of the unemployed, a little over a third are receiving benefits.
So are benefits discouraging work? Let’s take a look at one more graph before fully answering that question. Here we’ll reintroduce the benefits to the chart, with the ratio of continuing claims/unemployed, the unemployment rate, and the insured unemployment rate.
Now, let’s finally start to answer the question. From the graph above, there is a steep drop in the claims/unemployed ratio between the cutoff of the first $600 benefit and the resumption of it at $300. A hasty observer might jump to the conclusion that the cutoff was to blame, but if this were the case one would surely expect the drop-off to not take until September to begin, and surely it would not begin to taper off at such a high level (around 60%). Additionally, the claims-unemployment ratio did not begin to rise when the $300 benefit was restored, nor did the claims-labor force ratio.
Additionally, as Skanda Amarnath has noted, two things remain important. The first is that while the labor force participation rate has flatlined since last summer, the employment-to-population ratio has been on a steady incline. Unemployment is tricky to measure, especially during a pandemic where there are many ways in which one could be unemployed (even the BLS maintains several unemployment measures as it is, each with different definitions). Referencing the employment rate (and therefore unemployment rate) to the entire population instead of to the labor force (which includes the difficult-to-define unemployed component) is a better image of the situation.
Amarnath’s second point, relevant to last month’s job report, is that job gains were highest in lower wage sectors, with the slowdown in job growth actually occurring in mid-wage industries. If the $300 benefit were truly keeping people from returning to work, this probably would not be the case; it should be the higher-waged professions, who already paid above unemployment benefits to begin with, that had more job gains.
Two San Francisco Federal Reserve economists corroborated this story in a working paper published last summer and recently updated. They found that of the one-in-four unemployed individuals who received a job offer they would typically accept, only one-in-seven of them would deny it due to unemployment benefits. In other words, imagine 28 unemployed people. Seven of them, a quarter, are expected to receive job offers they would generally accept. Only one of those seven would deny that offer due to benefits. In fact, their model finds that food service workers, who are anecdotally the ones with the shortest supply, are among the likeliest to take offers. This was modeled after the $600 benefit, too; surely the effect is lower on $300.
With all of this considered, it is unlikely that unemployment benefits are to blame for any “shortage” in the labor market. They definitely do play a part, and certainly are not helping, but it seems relatively clear to me that they are at the very least not responsible for most of the labor situation. This is not to say that the extra unemployment benefits are necessary at this stage; in my view, in most cases they are not. States can (and are) making the choice to offer them at their discretion. It is my view that Democrats were too pessimistic on the success and speed of vaccines distribution during the ARP’s composition between January and March. However, the benefits cannot be to blame for everything, or even most of everything.
Policymakers should remember the lessons learned during the ten-year recovery period from the Great Recession. It is not merely the unemployment rate that matters, but rather the deviation of the labor force from the trend. The true gap accounting for potential should be closed, not just the unemployment gap. Otherwise, workers are left behind, and the economy is left running below potential.