Ed Dolan's Econ Blog

What Does the U-6 “Broad Unemployment Rate” Really Tell Us?

During the recent deep recession and slow recovery, U-6, an alternative measure of unemployment issued by the Bureau of Labor Statistics, has received increased attention. People often refer to U-6, which includes several groups of workers in addition to the officially unemployed, as the “broad unemployment rate.” No doubt, part of its popularity stems from the simple fact that the broad measure of joblessness makes the employment situation look worse than the standard one, and bad news attracts readers and viewers. Beyond that, thought, just what does U-6 really add to our understanding of labor market conditions?

What U-6 tells us

The main contribution of U-6, as I see it, is to remind us that those whom the BLS defines as unemployed—those who are not working for pay even an hour a week but have looked for work within the last four weeks—are not the only ones who suffer when labor markets are weak. U-6 brings in two groups of people who feel  labor market distress even though they do not fit the official definition of unemployed:

  • Those who are working part-time “for economic reasons,” that is, who would prefer to work full time but are not doing so because they cannot find a full-time job or because their employer has cut their hours.
  • Those who are “marginally attached to the labor force,” that is, who would like a job, are available for work, and have looked for work in the past year but not in the past four weeks. “Discouraged workers” are a subset of the marginally attached who have not recently looked for work because they think none is available. The remaining marginally attached workers have other reasons for not actively looking for work. The BLS adds marginally attached workers to both the numerator and the denominator of U-6.

Popular accounts often focus on the discouraged worker component of U-6, but in practice, as the next chart shows, involuntary part-time work is the most important added category. The top line in the chart shows the gap between U-6 and U-3. The next line, which shows involuntary part-time workers as a percentage of the labor force, almost entirely fills the gap. It is responsible both for most of the rise in U-6 during the recession and most of its gradual decline during the recovery. The third line, which shows marginally attached workers, contributes relatively little either to the U6-U3 gap or to its change over the cycle.

Furthermore, it is worth noting that discouraged workers, as defined by the BLS, make up only about a third of all marginally attached workers, whatever the state of the business cycle. In March 2007, when unemployment was at its prerecession low of 4.4 percent, marginally attached workers accounted for 0.6 percent of the labor force, of which just .17 percentage points represented discouraged workers. More recently, in July 2013, the marginally attached were .99 percent of the labor force, with discouraged workers accounting for .37 percentage points. As the next chart shows, the number of discouraged workers peaked at about half a percent of the labor force in December 2010. The remaining marginally attached workers did not look for work for other reasons than a belief that there was no work to be found, that is, for reasons such as ill health, family responsibility, school, or transportation problems.

What U-6 does not tell us: Those who want a job but are not looking at all

Although U-6 covers involuntary part-time workers, discouraged workers, and other marginally attached workers, it misses two other groups who suffer economic stress as the result of a weak labor market.

One of those groups consists of people who say they want a job but who are excluded from U-6 because they have not looked for work in the past 12 months or are not currently available for work. The number of such people, shown in the next chart, is equal to the total number who are not in the labor force but want a job minus the number of marginally attached workers. Some people think that anyone who wants a job but does not have one should be included in a broad measure of unemployment. For example, we could construct a hypothetical U-7 that included in its numerator the officially unemployed, the involuntary part-timers, and all those outside the labor force who want a job, and in its numerator, the civilian labor force plus all those not in the labor force who want a job. However, two considerations would limit the usefulness of such a measure.

First, as the chart shows, the number of those who want a job but do not qualify as marginally attached is not especially large nor does it vary much over the business cycle. For that reason, adding these people would not greatly change the picture of the employment situation that we get from U-6.

Second, these people, despite their declared desire for a job, are even more loosely attached to the labor force than those the BLS officially classifies as “marginally” attached. Keep in mind that the BLS standards for actively looking for work are rather low. For example, answering one job ad a year, contacting an employment agency even once, or simply asking a friend or a relative about job openings once a year is enough to keep a person in the marginally attached category. Perhaps we could describe those who do not meet that standard, but who still say they want a job, as “attached by a thread.”

Some of those hanging to the labor force by a thread may be suffering great economic distress but are so discouraged about work prospects that they will not lift a finger to look for a job. Others presumably have other sources of income (family, savings, government assistance, or whatever) and want to work only in the sense that if just the right job fell into their lap, they would take it. The BLS does not provide the data we would need to make such distinctions.

What U-6 does not tell us: Those who are working full time but underemployed

Another group who suffer distress in a weak job market, but are not covered by U-6, consists of people who are working at a job that does not make full use of their potential. Both anecdotal evidence and systematic studies suggest that the number of such persons is substantial and is higher now than before the recession.

For example, a study by Richard Vedder, Christopher Denhart, and Jonathan Robe that was released earlier this year by the Center on College Affordability and Productivity shows surprisingly large numbers of college graduates in occupations that do not usually require a college degree, such as taxi drivers (15 percent), retail clerks (24 percent), and firefighters (18 percent). All together, the Vedder study claims that as many as 48 percent of all college graduates are employed in occupations that require less than a college degree.

Vedder et al. point out that we can view these data in two ways. On the one hand, they can be seen as a a cyclical phenomenon, in the sense that a weak job market pushes college graduates into occupations that do not make use of their skills. On the other hand, the data can be seen as a symptom of secular overinvestment in education. They suggest that a “college for all” mentality may be running up against diminishing returns, so that the number of graduates (at least in some fields) is growing faster than demand regardless of the state of the business cycle.

Whatever the reason, though, there is some sense in which a person with an accounting degree who is working full-time as a sales clerk is as much underemployed as a classmate who is working short hours as an accountant. In principle, it would be nice to have some “U-8” or other measure that captured that kind of underemployment, but there do not seem to be any easily available data that would allow us to construct such an index.

The bottom line

When all is said and done, what we really learn from looking closely at U-6 is that unemployment is inherently a difficult concept to pin down. Different measures each catch a different group of people who fail to find what they want or need in the labor market. The broader measures are better in the sense that they miss fewer people who suffer economic stress because of a weak labor market. At the same time, the broader the measure, the more it dilutes the common sense notion of unemployment as a situation of looking for work and being unable to find it. In the end, we learn more by looking at a number of different indicators than by trying to synthesize them into one “true” unemployment rate.


17 Responses to “What Does the U-6 “Broad Unemployment Rate” Really Tell Us?”

windrivenSeptember 16th, 2013 at 10:49 am

Perhaps because it is Monday morning as I read this I struggle with this:

"One of those groups consists of people who say they want a job but who are excluded from U-6 because they have not looked for work in the past 12 months or are not currently available for work."

There are two sets here: those who have not looked for work in 12 months and those not currently available for work. In the first case the 'want' of a job is necessarily a very soft want indeed. I might say I want a pickle but if I don't check the refrigerator for a jar of pickles my want is clearly marginal.

It is the second that most confuses me, those who want a job but are currently unavailable for work. Is this meant to corral those in, say, a full body cast who nonetheless would love a job welding doodads on an El Dorado?

I've phrased my question lightly but it is a serious question. I don't understand how those sets should be included in any unemployment number. It seems that both sets would more accurately be said to want additional disposable income rather than a job.

Ed Dolan EdDolanSeptember 16th, 2013 at 12:24 pm

I agree, I am confused about the category of "want a job" but "not available for work." If you check BLS table A-38, you will see this is currently 658k out of 7013k who are out of the labor force but want a job. It does not include people who have temporary illness or transportation problems; they are counted as "other marginally attached" on the theory that they are "available to take a job" even if they won't be able to start work until they get over the flu or get their car fixed or whatever.

Aneta NascuSeptember 17th, 2013 at 4:36 am

There are two kind of unemployed people:
1.those who didn’t have a job before crisis occurred (2007-2009), most of them beneficiary of sub-prime mortgage loans or bank loans, who defaulting to pay the installments and interest rates related to the loans and mortgages got from banks, generated the lack of liquidity at those banks. This category does not want to get a job as they are used to take advantage of the government social support.

2.Those who always had a permanent job having expertise in what they did. In 2008 the banks short on liquidities postponed the loans to corporate customers which were forced to:
• Reduce volume of merchandise and services to be bought from suppliers;
•reduce production,
•reduce the amount of products sold on the market;
Suppliers, distributors and customers, beneficiaries of first company’s output, did the same as well. All of them reduced business activity and lay off part of employees or worst got bankrupt. And……

Aneta NascuSeptember 17th, 2013 at 4:37 am

…..This part of lay off employees has been sallow by the crisis tide without having any contribution to the facts which generated the lack of liquidity in banks and disturbances in all financial system.
This part of lay off employees use to be high and good rated bank customers, having a job, equity owners, and paying back all loans got from the financial market. They were financial crisis’s victims and they are depressed, without hope. Even they got a job they cannot get loans to recover what they lost for the next 5-7 years, and their emotions and disappointment could be inherited by the young people too.
In order to restart the economy engine there is need to pay more attention to the human emotions, invest in changing confidence, temptations, envy, resentment, and illusions—and especially by changing stories about the nature of the economy and government rules and regulations.There are solutions to fix the damages

TejasSeptember 17th, 2013 at 5:32 am

I suppose as long as the different measures of unemployment move in the same direction there is no real problem. In this case the level of the unemployment rate(s) would be less important than the direction in which the unemployment rate(s) move. However, as soon as we focus on the level of any of these rates – as the Fed is doing now – the trouble starts. (Interesting and useful analysis – thanks a lot)

George N WellsSeptember 17th, 2013 at 11:11 am

What I deduced from reading this article is that there are many interpretations of the data on employment and even more argument about the data. While this argument may keep economists and pundits employed, the discussion does little to actually resolve the socio-economic issues at play.

I will posit the following: "I believe that Capitalism, as it is currently practiced, is fundamentally incapable of providing employment to everyone on the planet that needs/wants to be employed." Comments/responses will be appreciated

Ed Dolan EdDolanSeptember 17th, 2013 at 11:50 am

>"I believe that Capitalism, as it is currently practiced, is fundamentally incapable of providing employment to everyone on the planet that needs/wants to be employed."

You are certainly right about this. The reason is that capitalism is based on a system of free labor, under which a person can be employed only if both the worker and the employer agree. There are two kinds of situation in which such an agreement is not reached.

First, the worker thinks the terms of the employer's offer are not good enough, and therefore continues the job search in the hope of finding a better offer. Even if everyone who searches eventually finds a better offer, there is always a certain level of "frictional" unemployment.

Second, the employer may think that the worker's reservation wage is higher than the worker's productivity. In that case, the capitalist employer, in pursuit of profit, will not hire the worker at a loss-making wage.

As far as I know, there are only two economic systems that guarantee work to all who "need/want" to be employed.

One is peasant agriculture. The distribution of means of production is such that everyone can grab a hoe and chop weeds as long as their marginal product is even microscopically above zero. That eliminates the second source of unemployment (although not perhaps the first). I suppose hunter/gatherer society also fits this pattern.

The other system of guaranteed employment for all is slavery, where again, the opportunity cost of employing a slave, at the margin, is zero, so that anyone who has a marginal product even microscopically above zero can work.

George N WellsSeptember 17th, 2013 at 1:15 pm

Mr. Dolan,

We are in agreement. That means that a percentage of the working age population is always going to be under/unemployed by choice or as a result of larger economic issues. That leads to the questions:

1. What is an acceptable level of unemployment at both the national and international levels?

2. What is the responsibility of government to their citizens/residents who are under/unemployed?

3. What are the social/economic consequences of a larger than acceptable portion of the world-wide population being under/unemployed?

Clearly the creation and maintenance of an international Plutocracy isn't the answer, but considering that there is a lot of valuable work that can, and needs to be, done on the planet we do need a different approach.

o2generateSeptember 17th, 2013 at 1:03 pm

I wonder how much effect employment in the military absorbs potential employees in search of a job? Not simply the high school graduates who opt for this government pay, but more highly skilled people who are under-employed in military activity, or, underpaid as military doctors.

Ed Dolan EdDolanSeptember 17th, 2013 at 4:08 pm

With an all-volunteer military, I don't see the logic of treating military service any differently than any other job. Right now, serving military are not counted in the labor force, so they are neither employed or unemployed. If they were treated like other workers, they would all be in the labor force and employed. The effect would be a slight decrease in the unemployment rate.

TonyFebruary 29th, 2016 at 9:45 am

This discussion reminds me of Winston Churchill's famous quote "The inherent vice of capitalism is the unequal sharing of blessings. The inherent virtue of socialism is the equal sharing of misery!" Pick your poison, for me I pick capitalism any day of the week, but I'm afraid our electorate is moving more and more towards sharing their misery.