The latest government report on employment can be counted on to raise questions with Answer Desk readers about just who is officially considered employed and who isn't. Mary in Kansas City is wondering if people whose unemployment benefits run out are counted. And Tony in New Jersey is wondering where military workers fit in.
Is the figure used for total unemployment derived from the various states' statistics regarding the number of people who are receiving unemployment compensation? If so, is there an adjustment to that figure to include people who have been unemployed so long that their unemployment benefits have run out?
-- Mary, Kansas City, Mo.
The data are not collected from unemployment claims. The Labor Department conducts two separate surveys — one in which they call workers directly and the other in which they collect information from employers on employment, hours and earnings. That sample includes about 160,000 businesses and government agencies at 400,000 separate workplaces.
So while you may quibble with the government’s data definitions, it’s pretty clear how they go about collecting the numbers. And while they have recently become the subject of much political spin, the numbers aren’t supposed to be a straw poll on the success of the government’s economic policies. They’re intended for use by economists, analysts and forecasters to track changes in labor status over time.
That’s why the numbers are also “seasonally adjusted” to take into account annual events that don’t reflect long-term job gains or losses. Without these adjustments, for example, the unemployment rate for the retail industry would go down in November as temporary workers are hired for the holidays and then jump up in January when those temporary jobs end. Adjusting the numbers gives you a better read on longer-term industry trends.
While the monthly report may not best describe what you or I would call “unemployment,” it's one of the best sets of data available to track changes in the economy. That’s because the size of the sample is large enough to pick up broad changes in hiring patterns. And because the numbers are available on the first Friday of the month, they’re the “freshest” of the major economic reports available.
If you wanted to, you could certainly define unemployment differently. A lot depends on who you choose to include (or exclude) from the labor force you’re using as a base. Should students be in or out? How about retirees working part-time but not currently employed? Furloughed auto workers? Illegal immigrants? In the end, the reason for including or excluding various categories of workers has more to do with improving the sensitivity of the survey as an economic barometer than it does with adhering to a definition of employment that you or I might find more accurate.
The “household” survey, for example, asks if someone is “actively” looking for work, and if the answer is that they've given up looking, they're excluded from the count of the total labor force. In a sense, they're invisible. You can argue — as many readers have — that these people should be counted as “unemployed.” Since all this data is available, you can include these “discouraged workers” in your definition of unemployment, and many economists and analysts do so.
As for the quality of jobs created, it’s not entirely clear that high-paying jobs are being replaced by low-paying jobs. Overall, wages are rising. What does seem clear is that the gap between the lowest- and highest-paid continues to widen.
Are the members in the armed services counted as employed or unemployed? If they are counted as unemployed, why?
-- Tony T., Maywood, NJ
They're neither. The employment figures count only civilians, so military workers don't even show up in the official tally of the total work force.
In the 1980s, members of the armed services were included in the numbers based on the recommendation of a presidential commission. The thinking was that, with the change to an all-volunteer army, the military work force was a lot more like the civilian work force, so it should be included. Since all military workers are by definition “employed,” including them had the effect of dropping the unemployment rate buy about a tenth of a percent.
But when the employment survey was overhauled in January 1994, military workers were dropped for three reasons, according to Sharon Cohany, an economist with the Bureau of Labor Statistics.
First, the count of military workers wasn’t as accurate as the BLS needed them to be. Different branches of the services reported differently, so the data was inconsistent from one month to the next. And with the rapid deployment of the Gulf War, some soldiers posted overseas were still being counted as U.S. residents long after they were no longer part of the U.S. workforce.
Second, the military employment rolls weren’t broken down the same way as the civilian data. So when it came time to break out a specific age group or region within the workforce, the military numbers didn’t match up with the civilian numbers.
Finally, very few of the people who use the employment data —analysts, academics and economists — actually used the military numbers. One reason is that including or excluding military workers makes very little difference in the unemployment rate. For one thing, these workers make up a relatively small portion of the total workforce. And changes in military employment occur very gradually and don’t necessarily tell you much about changes in the wider economy.
None of this is meant to diminish the importance of the work performed by American service men and women. If you want to add them back in, just knock a tenth of a point off the “official” unemployment rate, and consider them counted.