According to the BLS, the U.S. economy shed 23,000 jobs in July. As if that’s not bad enough, May employment gains were revised down from 129,000 to 66,000 while June was revised down from 57,000 to 20,000.
At the same time, the same BLS report revealed that the unemployment rate fell from 4.19% to 4.09%. To be sure, the headline jobs numbers and unemployment rate are from different surveys: the jobs numbers from an Establishment Survey, unemployment from a Household Survey. But this is only part of the story since the household survey has shown even larger employment losses.
There have been recent discussions in the news regarding “breakeven” job gains. This discussion involves some math, but is nonetheless helpful since it points to reasons why low — or even negative — job changes can be consistent with an unchanging unemployment rate. To start, the population (P) (defined as those 16 years of age and not in an institution) can be divided into three mutually exclusive groups: employed (E), unemployed (U) and nonparticipants (N):
The labor force participation rate is given by
Now, suppose that the labor force participation rate is constant over time while population grows at rate n%:
For the participation rate to remain unchanged, it must be that the labor force, U+E, also grows at rate n%. Next, define breakeven employment gains as the increase in employment required to keep the unemployment rate unchanaged. Since the unemployment rate is given by
it follows that for the unemployment rate to remain fixed, it must be that the number of people unemployed, U, grows at the same rate as the labor force, U+E, and therefore, employment (E) also grows at the same rate, n.
What do these back-of-the-envelope derivations imply for breakeven job gains? Since 1970, U.S. population growth has averaged maybe 1% per year, or 0.08% per month. Since U.S. employment is around 159 million, breakeven job gains works out to nearly 132,000 per month (that is 0.08% of 159 million). However, U.S. population growth has slowed to perhaps 0.6% per year since 2000; the associated breakeven employment gain is close to 79,000 per month. Some commentators have suggested that U.S. population growth is actually negative. If population falls 0.5% per year, then the U.S. labor market can actually lose 66,000 jobs per month yet the unemployment rate will remain unchanged!
There are, of course, other things going on. For example, for the purposes of illustration, we worked with a constant labor force participation rate. Between 1960 and 2000, the labor force participation rate rose from less than 60% to around 67%. Such an increase would lead to a higher level for breakeven job gains. Since 2000, the labor force participation rate has fallen below 62% which, in contrast, would lower the value of breakeven job gains. The current labor force participation rate is 61.5%, suppose the labor force participation rate fell by 0.5 percentage points to 61%, the breakeven employment gain would be -23,900.
With all of the above in mind, it’s not clear that the labor market is displaying weakness when the U.S. job market is shedding jobs.
