Today featured a lackluster set of Friday job numbers. Payroll employment was up 115,000 jobs in April, and the unemployment rate went down to 8.1% from 8.2%. Government jobs were down 15,000, including 5,200 state and local education workers.

There’s three ways of parsing the jobs numbers. One way is to focus on the jobs created – where are they, what industries are they in, and how much wage growth and hourly gains are there? The second is to focus on unemployment – who is unemployed, how long have they been unemployed, and what characteristics do they have? And the third is to look at secondary unemployment characteristics, the numbers that try to interrogate the boundaries between the unemployed and those “not in the labor force.” We’ll spend some time in the next week talking about how to think of this third category.

For instance, is this even an interesting question? Matt Yglesias makes a case that it is overblown, arguing that it is really catching longer-term patterns and needs to be put in the context of the global economy. I agree in the sense that I think 8.1% unemployment is sufficient for serious reaction. But I think digging into this is important for both economic and political reasons. We’ll start a reply with this post.

It is the case that the size of the labor force hasn’t grown (as raw number of people, not a percent) since the recession started. Though we don’t know what the “true” size of the labor force should be at full employment, it should be a bigger number than it was in 2007. That’s a problem, because conventional unemployment can’t capture that.

And it is still true that the unemployed are more likely to drop out of the labor force than find a job. This is a brand-new phenomenon in the post-Great Depression economy.

Though the drop-out rate is within a longer historical range as a percent of the unemployed (which is in the chart above), the number of unemployed people doubled during this recession. This channel is undertheorized in normal economics – why would someone looking for a job decide to stop looking, given that they were willing to look at one point? For those concerned about the long-term costs to our economy of hysteresis, this is a problem. We aren’t seeing an uptick in those moving from “Not in the Labor Force” to “Unemployed,” and thus no increase in unemployment, which we had wondered if we were going to see as the economy picked up.

Many of those who are “Not in the Labor Force” want a job but are declining from actively searching for them. This number went up in the recession and is hovering at a high rate:

This is the categorization of “want a job now,” but “U-5” unemployment also captures some of these changes. That additional 1.5 to 2 million unemployed workers would give us a higher unemployment rate. It isn’t increasing in the past year, but it isn’t decreasing either.

An aging population should create decline in the employment-to-population (the percent of people working) and labor force participation (the percent of people working or looking for work) rates. How does this look when we just look at 25-54 year olds? Here is the labor force participation rate for 25-54 year olds:

And the employment-to-population ratio:

There’s a question as to what extent the recession is speeding up already occuring trends (retirements in an aging population, increased schooling, fewer men working) or has caused these trends to happen (or at least overshoot) as a result of being away from full employment. But retirement and schooling is less of an issue in the 25-54 year old range, and yet we see dramatic results here. Will these go back to their previous levels? Probably not. But I believe they’d go back at least a little at full employment. And that needs to be accounted for.

Cross-Posted From The Roosevelt Institute’sNext New DealBlog

The Roosevelt Institute is a non-profit organization devoted to carrying forward the legacy and values of Franklin and Eleanor Roosevelt.