Fair:

Here's a pdf of historical data with the current U-3 totals going back to 1942. Also, here's historical data back to '94 under U-6.

Why shouldn't people in the military be considered in the labor force? It may pay poorly (for most recruits), but it's a paying job (with significant benefits). Aren't people in other government jobs considered as being in the labor force?

Am I misunderstanding what happened and who you're talking about?

Evan, that one is more defensible, although it's clear that the intention was to "lower" unemployment by moving numbers. The Reagan administration recognized that by moving an uncounted but 100% employed group into the labor force, he pushed the percentage of unemployed people down (four unemployed people out of ten is a larger percentage than four of eleven). No one reduced unemployment by this technique, but the numbers went down.

Should serving a hitch in the military be considered regular employment? It's a fine question to ask. But let's not pretend including military is comparable to not including it. Calculate the statistic consistently.

Jeff, did you run across anything on what the differences between European unemployment numbers and US unemployment numbers come down to in specifics? One of the things that's always bugged me about people touting the "low" US figure has been the apples-to-oranges nature of such figures.

For that matter, apart from the military, you have to wonder how the comparatively high incarceration rate in the US would affect the unemployment rolls. The US rate in 2006 was over 700 per 100,000 vs. an average in Europe of just over 100 per. There are about 2.25 million prisoners in the US now; if the incarceration rate was the same as Europe, nearly two million of those would be out on the street, presumably in need of jobs (not to mention all of the corrections employees who would be out of work).

By my calculation, if we weren't warehousing so many people in prisons, that could account for more than a 1% uptick in the unemployment figure all by itself.

Jeff, did you run across anything on what the differences between European unemployment numbers and US unemployment numbers come down to in specifics?

Not in this round of research, but I have seen references to this before. I think your suspicion is well warranted--as I recall, other countries have their own calculations and it is indeed apples to oranges.

The prison thing: good point (but dark).

The prison thing: good point (but dark).

Yeah, well, consider the source.

Most of the stat I have seen do not include those of us who are self-employed or contract workers. And there are a heck of a lot of us out there.

Back in '03 when I lost my contract job, I applied for unemployment and continued to fill out the paperwork even though I wasn't eligible for any payments - it meant I was counted in their numbers.

Great post Jeff.

Prisons get even darker if you were to put into the mix prison workers who work for contract businesses (e.g. Oregon's "Prison Blues") or displace potentially fully-waged workers (as with Portland Parks' Bureau employment of prison labor).

<hr/>

On the military question, I think there may be a bit more to it than just the comparability issue.

What is the purpose of the labor force statistic?

One of its purposes may be to measure capacity that the economy has for production of goods and services for civilian uses. For measuring such capacity, and utilization of that capacity, it might make sense to regard people on full-time military service, and those is prison, as not in the labor force. Certainly a national guard soldier who is deployed overseas is not available to his or her employer, who much find a replacement or increase workload on other workers.

It seems likely that such a way of looking at things was especially at work during World War II and the period immediately after, and that it persisted during that peak era of U.S. manufacturing production. During the Great Depression, unemployment as non-utilization of labor power was a key element in the vicious cycle that made the Depression so hard to break.

Lend-lease and World War II did two things. One was that it dramatically raised the demand for labor for production, especially as the total war economy kicked in. Then it took about 8 million adult men out of the domestic workforce -- say roughly 60 million males in the population, perhaps 40 million working age, maybe fewer, call it 40 million, that's 20% of the working age male population in the military.

The pressure on capacity was so great that the waged workforce got redefined to include many women who by convention and custom and Depression era discrimination had not participated.

As the end of the war approached and for the first few years afterward there was a great fear that demobilization would re-introduce the high unemployment of the 1930s. We could analyze that as servicemen "losing" their military jobs, I suppose, but that was not how it was viewed -- rather, they were re-entering competition for normally productive domestic economy jobs.

With the decline of U.S. manufacturing and the shift to service sector jobs, conceptualization of military personnel as "providing a service" gets easier than when the paradigmatic idea of labor is "making something."

If we're thinking about unemployment mainly from the point of view of asking, do people have an occupation that provides them food, shelter etc. in return for work, including the military makes fine sense.

But if we're thinking about "the labor force" as that part of the population available for domestic economic activity that contributes to economic growth and productivity growth, it may not make so much sense.

It would be interesting to know if military personnel are included in the workforce for purposes of calculations of productivity, and how the "services" they produce are valued.

<hr/>

"Frictional unemployment" is tricky IMO for two reasons. One is that the Clinton method of counting doesn't include workers who on Jeff's chart are "part time for economic reasons," which I have also seen described as "would like to work full time if such work were available." Throughout the Clinton era jobs being destroyed by technological replacement or relocation of production outside of the country tended to be full time and higher paying, and to be replaced by jobs with lower wage rates and often fewer hours.

The other is that the idea of "the 'normal' local variations in business conditions" is dodgy in creating a national geographical mean "frictional unemployment" statistic. It disguises economically depressed geographical areas, whether it be Oregon timber counties or deindustrialized former manufacturing areas like the Monongahela Valley or inner city areas distant from large scale employers like the South Bronx.

Real "frictional unemployment" would need to be understood by looking at the lowest unemployment rate areas, IMO. If the national rate under Clinton was defined as 4%, that meant that there were many counties where it was 3% or 2%.

IMO real "full employment" nationally would need to apply that standard everywhere.

A final question about such "frictional unemployment" has to do with the bargaining power of workers and the share of increased productivity they are able to command. In part because of Nixon-Ford-Carter & early Reagan-era inflation and stagflation, control of inflation became the central regulatory function of the Fed in the 1980s, and employment levels that were "too high" were treated as a bad thing economically.

But, as current circumstances that are beginning to resemble the late 1970s suggest, this focus on "too much employment" as a driver of inflation has a strong ideological element that really is about the rates of profit for employers (share of productivity increases).

The NPR news today is reporting that the Producer Price Index is rose at an annual rate over 9% last month, while unemployment is rising, looking a good deal like "stagflation." Then, as is typical, the report goes on to say that if energy (fuel) and food prices are left out, the PPI was rising at about 2%. But eventually those energy costs work themselves into other dimensions of producer prices, having no cause in employment levels. On the contrary, they are likely to lead to cutting back employment and intensifying labor demands on remaining workers as means to control costs.

Meanwhile fuel and food costs for consumers (more of whom are losing jobs or working fewer hours) are not "secondary" items to be considered apart from "core" CPI inflation, as a practical matter of living life. On the contrary, they are among the most prominent, salient, frequent and hard to limit costs.

<hr/>

The relative weakness of labor's capacity to demand a larger share of increased productivity also has a bearing on the redefinition of the workforce to include many more women who are married and/or have children than before World War II. There is something of a myth that all of the female war workers left wage work after World War II. Actually female labor force participation was rising substantially through the 1950s, and may in some ways have been a cause as much as an effect of the emergence of the women's movement in the 1960s and 1970s.

The necessity for a two-income family to maintain what is regarded as a decent standard of living also needs to be factored into the picture. This may relate in complex ways to the "underemployment" (working less than desired hours) piece of things.

I believe U6 captures the prison population.

Extra points for the use of "capture" and "prison population" in the same sentence.

Yeah, well, consider the source.

That made me laugh out loud--

Jeff, I dare you to publish the record high number of Blacks and College age people that are unemployed.

I'm pretty sure these stats are available, so you are surely as able as I to post them yourself.

As someone who is self employed, I'm outside of all or most of those stats. Jeff, do you know if the relevant agencies track those trends/stats also?

Unfortunately, no. I'm not really an economist, I just play one (badly) on the blogs. You should really be asking Chris, who is ten times more informed than I. Pretty much my strategy is to post something to get him thinking and then let him take the wheel. Which he did, ably as always, above.

I particularly like this analysis, which captures a way of thinking about why a statistic is important (namely, to reveal):

But if we're thinking about "the labor force" as that part of the population available for domestic economic activity that contributes to economic growth and productivity growth, it may not make so much sense.

Indeed.

Jeff, do you know the work of Jacob Hacker? He's coming to Reed this fall to give a talk (he's also appearing at the City Club).

We also have Austin Goolsbee, Obama's chief economic advisor, coming into town.

Both have research and writing on the political annd economic impacts of rising inequality.

Let me know if you want a VIP slot (meaning nothing more than a chance to go to dinner).

Paul --

I hope you'll give us advance notice here at BlueOregon about those two visits. Hacker's health care plan, after all, is the starting point that informed Wyden's, Obama's, Edwards's, Clinton's, and HCAN's health care plans....

On self-employed, yes, you'd be counted as "in the workforce" -- there is sort of a continuum between self-employed contractors and workers in skilled trades, which has been changing in some ways recently with so many new fields for such contracting in new technology areas. So, Jenni is a self-employed contract web-designer (to possibly oversimplify a bit) but she could also do the same or similar work as a regular employee of a business, or as an employee of a temp agency -- and if her contracting business got big enough she might also act like a construction contractor who hires skilled trades workers on job-work basis, and hire someone else with the skills temporarily to help out.

I think contract workers who have no work may end up effectively similar to "discouraged workers," i.e. those who could be in the workforce but have given up looking for work, and are thus counted as "not in the workforce." I believe that in employment surveys if you define yourself as "self-employed" you have to show a certain amount of income from the claimed employment in the past year to be counted as "really in that field."

Jenni's engagement with the government employment apparatus registers her as "looking for work" and thus in the workforce. Rick Hickey's point about those whose unemployment benefits have run out goes somewhat to the same point -- to get those benefits you have to certify you're looking for work and be prepared to document it, and once they run out, people may not maintain contact with the unemployment office, though they could still use their job-seeking services.

However, Rick Hickey's claim that a large proportion of U.S. unemployment can be attributed to illegal immigration is pretty much spurious and serves to obscure the whole range of reasons for it. He has something of a point in some fields of work, e.g. construction and landscaping and domestic work and unskilled day labor, where there seems to be some real displacement. The idea that illegal immigration has a big effect on college grad employment is just weird.

A lot of African-American unemployment is related to U.S. residential segregation, which has intensified since the 1960s, and mismatch between where people live and where jobs are. It may be that some of the pattern in the last 15 or 20 years of black people moving from northern cities "back" to the south (in some cases this is children of parents who had moved north) relates to finding access to jobs. But I think one of the dynamics involved is that a higher proportion of new immigrants / illegal immigrants are more mobile, at least to start with, having already chosen to leave their homes and families for labor migration. We don't see the unemployed or underemployed in their home countries who don't choose labor migration; the residential separation or "segregation" is at a greater distance and across a border. But with black workers in the U.S., we do see those who don't want to leave their families or communities, but can't find work in those areas, as well those who may engage in labor migration.

I take issue with your conclusion that this is a "bipartisan collusion" designed to hide the real numbers.

I think you're missing the key political point here: no president can survive with high unemployment. It's true that you can make arguments for why different groups should be included or not, but that conceals the very strong impetus for keeping the ranks of the "unemployed" as exclusive as possible.

I'd also argue that those discussions about what to include are secondary; by changing the formulation, we lose the statistic's central virtue: comparability over time. In fact, for scholarly or policy reasons, you could calculate the statistic in any way you wish. But the only reason to use this for government purposes is to reveal to citizens a statistic they can understand in historical perspective.

The collusion comes because both parties decide it's better to have a fake stat than stand up to the reality of returning to historical standards and having 10% unemployment hooked to your administration.

Kari Every year I promise to advertise our series on B.O. and every year I end up missing deadlines.

I'll try to do better. The whole series is described here http://web.reed.edu/public_policy_series/

• Austin Goolsbee on economic policy
• Spencer Overton on voting rights and vote suppression
• Jacob Hacker on health care
• Howard Wolpe on international relations and Africa
• Caitlin Baggott on voter participation

Paul, I'd definitely be interested in attending these, but I don't want to damage the prestige of the event. Reed might get flack for identifying a blogger as "important." Your call.

The standard unemployment figure excludes the institutional population from both the dividend and the divisor. While Europeans jail significantly smaller proportions of their 18-64 population, they don't generally have proportionately smaller institutional populations. The other measure economists use to assess employment trends is the percentage of the civilian noninstitutional population 18-64 that is employed. The difference between US and european levels on this measure is consistent with differences in measured unemployment. Just as we don't count some folks as unemployed that Europeans do, they don't count a lot of folks we do. There is very little reason to believe that the standard unemployment measure is biased across developed countries (US 5.5%, UK 5.3%, Can 6.2%, Euro area 7.2%) or for the US over time (25 years ago when reported civilian unemployment was 10%, 57% of the working age civilian population was employed; with 5.5% reported unemployment 62.5% of the population is employed).

Since the 1960s the largest group entering the labor force in the US are the spouses of high income men, who themselves also tend to earn high incomes.

WHo made Kevin Phillips an economic authority?