The Wealth of Employment Statistics

Featured photo:

Everybody, every day, hears about the national unemployment rate in the news. These statistics tend to be misinterpreted, and sometimes just misunderstood, so I decided to think about two questions, and then write this piece to help my fellow peers (and myself), who feel estranged from unemployment statistics. The two questions: one, how important is unemployment in the current economic environment; and two, how reliable are the methods used to produce the labor statistics we hear and talk about daily? I am going to focus on the rates produced and distributed by the United States, but most developed countries, including Canada, use a similar methodology.

The U.S. Department of Labor produces and releases monthly employment statistics that intend to reflect the nation’s employment gains and losses from the previous month. I find the actual percentages and numbers released to be somewhat alienating, as I can never form a real, material connection with a percentage statistic produced by a bureaucratic department. The real importance of the unemployment rates, though, is that they serve as a very useful measurement of national economic health, on both large and small scales. Rises in unemployment tend to coincide with economic recessions, and decreases in unemployment typically coincide with economic booms.

This concept is almost entirely undisputed. However, problems arise when the country’s policymakers and economists debate the best ways to accurately and effectively quantify unemployment data in order to produce statistics that will faithfully reflect economic health. So the first question we should ask is: what is the best method to produce employment data?

Before I delve into the current methodology for measuring employment, I want to discuss the basic economic ideas behind employment. Very generally, employment is loosely defined as the number of people of a certain age who have jobs, while unemployment is considered as the number of people of a certain age who don’t have jobs, but are actively searching for one. If you add the total employed to the total unemployed, you get the nation’s labor force. It follows that the unemployment rate will be the number of unemployed citizens divided by the total labor force, and this percentage is supposed to reflect how much of our workforce is, simply, out of work.


Now, there are several potential problems with this basic theory, and most deal with imperfect information. The issue is that this very basic analysis of the workforce just doesn’t satisfactorily reflect all the possible problems facing the United States’ labor force. Examples: discouraged workers who have exited the labor force (i.e., have stopped searching) due to poor conditions, even though they want a job; underemployed workers who may be working part-time but want (or need) to be working full-time; or underemployed workers who are working a retail job at Walmart when they have a PhD in Electrical Engineering. Our Department of Labor Statistics is well-informed of these potential problems, but simply does not have the resources to contact every single U.S. resident and learn the full extent of every laborer’s employment issues.

So what does our Labor Department do to combat imperfect information? The Bureau of Labor Statistics (BLS) has developed what they call “Alternative measures of labor underutilization,” wherein they produce 6 different unemployment rates, each of which measures underutilization by taking into account increasing amounts of variables. And, they avoid the impossibility of contacting every resident in the labor force by conducting the “Current Population Survey” in order to create a small-scale model that should, in theory, accurately represent the entire country. The survey, taken monthly by different households, asks questions relative to residents’ working activities (such as, whether or not they worked recently, whether they wanted to work recently but couldn’t, etc.). Essentially, the survey collects information on a sample-size of households from 800 different geographic areas to represent all 50 states and D.C. By diversifying the sample among a substantial amount of the potential labor force, they are able to create a basic source of data that should authentically reflect the entire nation’s workforce activities.

Now that I have described the basics of the survey, let’s move into the overall underutilization statistics it produces. Through the 60,000 households surveyed (an estimated 110,000 individuals), the BLS then produces labor underutilization rates in each of their 6 different categories, from U-1 through U-6. The BLS defines the 6 categories by these factors:

  • “U-1 — Persons unemployed 15 weeks or longer, as a percent of the civilian labor force.
  • U-2 — Job losers and persons who completed temporary jobs, as a percent of the civilian labor force.
  • U-3 — Total unemployed, as a percent of the civilian labor force (official unemployment rate).
  • U-4 — Total unemployed, plus discouraged workers, as a percent of the civilian labor force plus discouraged workers.
  • U-5 — Total unemployed, plus discouraged workers, plus all other persons marginally attached to the labor force, as a percent of the civilian labor force, plus all persons marginally attached to the labor force.
  • U-6 — Total unemployed, plus all persons marginally attached to the labor force, plus total employed part time for economic reasons, as a percent of the civilian labor force, plus all persons marginally attached to the labor force.”

When the survey is conducted precisely and uniformly, these six numbers usually conform with the ebbs and flows of the economic environment. U-3 is the most commonly cited measure, but also contains one recurring problem of basic economic theory, in that it doesn’t account for the breadth of possible underemployment problems. U-6 is the next most commonly cited measure, and accounts for some of the more intricate factors in the labor force.

There are many additional methods that have been proposed to help measure and produce unemployment rates that can exhibit overall economic conditions. One worth mentioning is the “Employment-population” ratio introduced by Princeton University Professor Paul Krugman. Krugman bypasses the labor force and simply divides the number of people employed by the population, i.e. the percentage of the population who have a job. Another big economic idea is comparing the unemployment rate (generally the U-3 rate) to what’s called the “natural rate of unemployment,” which can be defined as the rate when short-run business cycle factors are fully completed, i.e., when there is neither a short-run boom nor bust in the economy. This concept has to do with macroeconomic wage rates adjusting fully so that total supply and demand of labor are in equilibrium. In this, the only unemployment found is either frictional (the natural turnover of labor, accounting for the time it takes to find a new job), or structural (mismatch between skills in the labor force and industry demand for labor). Essentially, the “natural” rate of unemployment ignores short-term business cycle factors, and can show long-term declining or increasing unemployment (i.e. long-run economic health).

By now, I think we can agree that there is a wealth of information produced and provided in our economic environment on employment rates and cycles. The important question I’d like to raise is: how do we represent the total “slack” in the labor force and markets, to best give us a representation of economic welfare?

To my way of thinking, the important factor prudent to employment is the gross flows in and out of the labor markets during business cycles, or on a month-to-month basis. The funny thing I find in my analysis of most labor force statistics, is that they all tend to move with each other, at least in big swings. For instance, here is a chart (provided by the BLS) of all 6 measurements of labor underutilization since they started recording the data in 1994:


In this specific graph, shaded areas are meant to represent economic recessions, and it’s obvious that unemployment rose during these recessions in all 6 recorded statistics. There are slight differences. For instance, U-6 changes more drastically than U-1, but that is because it takes into account many more individuals in potentially in the workforce.

By analyzing all (or at least an adequate amount) of the raw information produced, I argue that you can find a mean average of the net change in activity in our labor force, and use that particular statistic as a gateway to the state of our nation’s economic welfare. For example, here is a table (provided by the BLS) recording the unemployment rates across rates U-1 through U-6 in 2016 and early 2017:


Looking at the changes in seasonally adjusted rates across all 6 statistics, they all dropped between .1 and .6 percentage points. According to the BLS, the labor force participation rate was around 63%, meaning these statistics apply to around 200 million American residents (.63 x Total U.S. population of 319 million). The average drop for these 6 measured unemployment rates was about .3 percentage points; logically, you can infer that a .3% drop in the rates would correlate to a net gain in employment activity for about 600,000 residents (.003 x 200 million), between people becoming newly employed, finding a job in which they no longer are considered underemployed, et cetera. This statistic would represent moderate improvements in overall economic well-being. For reference, only the month of February in 2017 showed an average drop of .1 percentage points among the 6 statistics, and the BLS estimates that just above 200,000 jobs were added in the month (.001 x 200 million = 200,000).

While this is my own analysis of unemployment data, I would say that the data I find most useful and important in our economic examinations are the statistics found from the Department’s general survey system (but then mapped out on the broadest possible level). Following these observations, I would analyze U-6 data recorded, making sure that is seasonally adjusted. The way I view it, the data that accounts for the highest amount of labor-force variables is the best indication of net flows in the workforce, and therefore a good indication of general economic welfare.

So, the next time you see journalists and media-members citing unemployment rates and pumping up the importance of those figures, ask yourself what statistics they have chosen and whether the hype is justified.