I have come across quite a number of assertions that the wrong metrics are being used to assess on thing or another. Last week I read an article highly critical of our withdrawal from Afghanistan claiming that the metrics that were used there were those that were relatively easy to gather (as we did in Vietnam) rather than actual metrics of success.
Consider this passage from a New Yorker article by E. Tammy Kim:
Though twice as many Ohioans work in health care, education, and retail than in manufacturing, factory labor persists as a Rust Belt obsession.
One of the possible interpretations of that is that the author is using the wrong metric, i.e. number of jobs rather than the number of jobs that would have been supported with the natural market clearing price of health care and education. I have seen proposals for using R0 or Re rather the number of cases of COVID-19, new cases of COVID-19, or deaths due to COVID-19 to determine whether the COVID-19 pandemic has ended.
In this post I’m going to reflect on how we’ve used the wrong metric to measure the strength of the labor force for decades and the influence that has had on policy. For the last 90 years politicians have been using “unemployment” as a measure of the strength of the labor force. For subsistence farmers the term has no meaning—they are always busy, i.e. “employed”. Starting in 1880 (in the Federal Census) questions about individuals’ “profession, occupation, or trade” began to appear in the Census. In the 1930 census 20% were listed as “Class A unemployed”, i.e. able to work, not working, and seeking work, having risen from 5%.
Since then the unemployment rate, i.e. the number of those unemployed as a percentage of the total potential labor force, has been a cardinal metric for the strength of the labor force. As just one indication of that between 1980 and 2005 more than 1,000 papers were published in economics journals with “unemployment” in their titles. Politicians frequently cite the unemployment rate either favorably or unfavorably depending on the rate, which party is in power, and the party to which the speaker belongs.
I would claim that the headline unemployment rate (everything from U1 to U7) has been moot for decades, has resulted in a misallocation of government resources, and should be replaced by another metric that includes both underemployment and overemployment.
There are three varieties of underemployment: visible underemployment, invisible underemployment, and the sort of underemployment resulting from individuals being unable to secure jobs in their chosen fields. Visible underemployment is when an individual wants to employed fulltime but is only working part-time. Invisible underemployment characterizes those who have stopped looking for employment because they have been unable to secure a job in their chosen field. I’m going to use the term in an even more restricted sense: those who with college degrees who are working at jobs that do not require college degrees.
I’m going to use the term “overemployment” to describe individuals who hold full-time jobs as well as one or more part-time jobs.
The alternative metric I would propose would be something on the order of total individuals of working age less the number of those working full-time but not more than full-time less the number of underemployed. I think that measure makes intuitive sense, much more than the unemployment rate. Furthermore it emphasizes jobs that aren’t menial jobs and pay a living wages rather than maximizing the number of bad jobs which I what I think the employment rate has come to overemphasize.
I wish I had a catchy name for that metric but I don’t. If I were an economist, I might be trying to calculate that metric and compare it with the unemployment rate going back to about 1980. I think it would be eye-opening. If I were a political scientist, I might be trying to determine the impact of using that metric on policy if minimizing that had been the objective rather than minimizing unemployment.
