The overall US unemployment rate fell from 13.3% in May to 11.1% in June. That is very good news. After falling 14.7% in the previous month, the highest unemployment rate since World War II, this suggests that the US labor market, severely damaged by the coronavirus pandemic, appears to be on the way to a rapid recovery.
As good as the news may seem, the improvement in the unemployment rate was even better than the headlines suggest. Even more threatening is that the improvement appears better since last month's comparison point was much higher than 13.3%.
The United States Department of Labor Statistics (BLS) reported in a footnote to its July 2 press release that the overall unemployment rate is likely to be the actual unemployment rate due to a misclassification error in which some temporarily unemployed workers were counted as employed in their monthly household survey data underestimated. In its press releases of April 3, May 8 and June 5, the BLS included a similar footnote to the employment situation in March, April and May, respectively.
Here is the explanation for the error. The BLS conducts a household survey in the middle of each month and uses the results to calculate the monthly unemployment rate, which it reports shortly after the end of the month. The BLS stated in a footnote that it instructed its interviewers to classify people who were absent from work due to temporary business closures related to coronaviruses as "unemployed due to temporary layoff". Accordingly, a respondent who stated that she had not worked during the week but expected to be called back to her job should have been classified as "layoff unemployed". This would be in line with the usual practice of being considered unemployed who have been made redundant but who expect to be recalled (regardless of whether they are looking for another job).
But as is so often the case, there is a complication. The correct classification of a person who is temporarily absent from their job depends on the reason for their absence. She would be counted as employed if she were temporarily absent due to bad weather, illness, industrial action, personal reasons or vacation. These reasons seem clear enough.
Here's the complication. She may also be registered as employed, but may be absent from work for another reason. For example, she might have been on leave to look after a loved one who has COVID-19. When “another reason” is given, the interviewer usually records a short description. When reviewing these descriptions, the BLS found that a significant number of respondents in their four surveys from March to June were classified as employed but unemployed rather than unemployed due to layoffs.
Correcting the misclassification error would have increased the overall unemployment rate by almost a full percentage point in March, by almost five percentage points in April, by three percentage points in May and by about one percentage point in June, all not seasonally adjusted. It is important that the misclassification errors have decreased in recent months, which gives reason to hope that they will soon no longer affect the unemployment rate.
Nevertheless, the errors affect the last four monthly unemployment rates to different degrees. And the misclassification errors are large enough to affect our interpretation of the data, as I will explain. The overall unemployment rate decreased by 1.4% in May (from 14.7% to 13.3%) and by a further 2.2% in June (to 11.1%). This indicates an acceleration in the recovery of the labor market.
There is even better news here. As the attached graph shows, the correction of the misclassification error in May shows a drop in the rate that is well above the headline rate, a decrease of 3.4% compared to a decrease of only 1.4% based on the headline unemployment rate and a decrease of 4.2% in June. versus just a 2.2% decrease implied by the overall unemployment rate.
Unfortunately there is no free lunch here. The graph also shows that the correction of errors indicates an increase in the unemployment rate in March by more than the headline, an increase of 1.9% versus 0.9% and an increase in April of 14.3% versus 10.3% . The corrected unemployment rate is higher every four months. And the rate peaked at 19.7% in April, five percentage points above the headline rate. It is also far higher than the peak unemployment rate during the Great Recession of 10.0% in October 2009. Where are all the adjustments going to be? The corrections are still leaving the US with an unemployment rate of 12.1% in June, one percentage point above the headline.
The graphic also shows an important feature of the COVID-19 world. Volatility, in this case the monthly change in the unemployment rate, has increased and is even higher than the headlines suggest. It is much more volatile than ever after the Second World War. The one-month decrease of 4.2% in June and 3.4% in May (both corrected) is the largest since 1948. The next largest one-month decrease was 1.5% in November 1949. The two-month decrease of 7.5% (corrected and compound) between April and June are the largest since 1948. The next largest two-month decline was 1.3% between October and December 1949.
The greater volatility in the unemployment rate is worrying. This means that the risk of a sharp reversal in the unemployment rate is greater if the reappearance of the corona virus leads to more states reintroducing restrictions such as the closure of bars, gyms and restaurants. There is a risk that the labor market and the economy will suffer more damage than the overall unemployment rate of the past four months suggests. The BLS reported that total employment outside agriculture increased by 7.5 million in May and June when the states reopened their economies. These jobs are potentially at risk if states step down. And the US unemployment rate could possibly rise above the April peak of 14.7%, especially if the number of unemployed is correctly counted in the coming months. With the June household survey conducted between June 7th and 13th, and many states pressing the pause button in the second half of June, unemployment may have increased.
Those who refuse to wear masks or practice safe social distancing not only put many of their fellow citizens at risk of getting COVID-19, but also increase the likelihood that state and local governments feel compelled to introduce stricter restrictions that could force thousands of employers to go out of business or close their businesses a second time, causing millions of newly hired workers to suffer from job losses again.
John D. Finnerty is a professor of finance at Fordham University.
More opinion in the capital::
- The investment industry claims to support diversity, but the numbers don't add up
- America's billionaires have a hoarding problem
- Why COVID-19 didn't stop the digital transformation in medium-sized companies
- What J.K. Rowling's missteps in terms of gender
- What I learned when I was trained as a coronavirus contact tracer