Despite mild weather, the winter of 2020-21 will be a terrible one for our nation. Health-wise, we face four more hard months of illness and death. And recent economic indicators of output and employment also show that we are in a difficult recession.
Economic indicators are complex and misunderstood. We get a flood of them, often overlapping and sometimes repetitious. Some sorting out is useful.
Yes, we just got employment numbers. Start, however, by going back a few days to the value of the goods and services we produce, or gross domestic product (GDP). This is basic because employment and spending result from it. If production drops, labor needed drops, less value is generated and less income can be spent.
Summing up everything produced is complicated and only done every three months. Results are not immediate as they are with employment, which is released within a few days after the end of a period.
GDP numbers are released in three estimates as data comes in. The first or “advance” estimate comes a month into the following quarter and the other two at roughly monthly intervals.
The advance estimate for July-September was on Oct. 29, the “second estimate” on Nov. 25 and the third will come Dec. 22. This three-step method is good for those in business and government who use it in decision-making. But the plethora of announcements for the same period can confuse people.
A further complication is that the numbers are reported in several different forms. One is simply the absolute dollar value of output. From that, one can compute the percentage change from the output level of the prior quarter.
However, there are seasonal variations. Some sectors, like tourism and construction are much stronger in the summer, consumer sales are big before Christmas. So, a statistical “seasonal adjustment” based on past patterns also is made.
This year, in raw-dollar terms, we produced $5.25 trillion in output from January through March. That fell to $4.90 trillion in April through June as the COVID-19 pandemic set in.
That drop of about 6.5 percent is huge. The worst quarter-to-quarter drop during the 2007-10 financial turmoil was only 3.6 percent. Back then, from the peak before to the absolute low point was 6.1 percent. So, yes, spring of 2020 was bad.
However, news reports made it sound worse than it was. For analysts looking at what is happening relative to long-term trends, working on a uniform “annual rate” basis is best. But changing quarter-to-quarter differences to what would be if that change kept on for a year is complicated. It is like converting from the monthly percent credit card to the effective annual rate. This is compound interest and 1.5 percent a month means 19.6 percent a year.
However, when 2020’s quarter 1 to quarter 2 drop was annualized, the figure was 31 percent. When reported in the media, it caused much alarm, more than justified. Then, when the third quarter estimates came out, we were told that output had increased 33 percent. So we were above where we started, right?
No, just as a weight drop from 200 lbs. to 180 is not the same percentage as gaining from 180 to 200. We are still about 1.8 percent below where we were before COVID-19 hit, with all the adjustments. In raw dollar terms, July-September 2020 had 2.9 percent lower output than the same months in 2019. And don’t expect the quarter we are in to be any better.
The slump is not evenly distributed across all sectors. Output in heavy industries and agriculture is down little, if at all. But air transport, hospitality, entertainment and others are off sharply. These are precisely sectors that use a lot of labor that gets paid relatively low wages.
In measuring employment, the most basic information is how many people have jobs. That is complicated because one must account for self-employed people, those in small businesses, workers in the gig economy and so on. Then you can calculate an unemployment rate, the percentage of people in the labor force who want jobs and are trying to get them, but don’t have them.
One can also calculate the number of jobs relative to the population and the number of people in the labor force, employed plus unemployed, relative to the population.
The Bureau of Labor Statistics comes at it from two angles: by surveying people, or “households,” and by surveying employers, or “establishments.” Both generate estimates of the number of jobs under various definitions. The household survey is the basis for the unemployment rate. Moreover, both are entirely separate from the numbers of people on or applying for unemployment compensation. If you get confused, it is OK.
Data for November are discouraging, even though the unemployment rate dropped slightly, from 6.9 percent to 6.7 percent. The most commonly-used job count was up 245,000 people. Yet that leaves us some 9 million jobs below where we were before this all began. And all indications are that both job numbers and the unemployment rate will reverse course, with jobs falling and the jobless rate rising. There will be much to follow and explain in coming months.