1. Melinda Pitts, John Robertson, and Ellyn Terry have a chart that seems to me to show that much of the decline in labor force participation in recent years can be accounted for by population aging, disability, and young people attending school.
Pointer from Mark Thoma.
2. James Pethokoukis has a chart showing that the number of people on food stamps has remained really, really high.
I would interpret (1) as saying that we are in a “nothing to see here, move along” sort of labor market. Given that the unemployment rate is normal, if labor force participation is just following natural demographic trends, then the economy is pretty much ok.
I would interpret (2) as saying that there is something to see here. With unemployment down, we should be seeing people fall off the food stamp rolls.
Are senior citizens, people on disability, and young students going on food stamps in droves? Are people who are still in the labor force and working staying on food stamps in droves?
I am not trying to make a point. I genuinely do not know how to connect these dots in the data. For those of you who follow algebra, we have
FS = food stamp recipients
POP = total population
UNEMP = employedunemployed
LF = Labor force
Then FS/POP = (LF/POP)(UNEMP/LF)(FS/UNEMP). We know that FS/POP is unexpectedly high, but LF/POP is low, and UNEMP/LF has come down. So FS/UNEMP must be quite high, right?
I suspect two things. First, change in composition of workforce means that those no working are more likely to need food stamps The linked article notes (as shown many labor force participation rates by age graphs) that older folks are more likely to be in the labor force. The ones that aren’t in the labor force have retired and don’t qualify for or don’t need/apply for food stamps. Because they are now a smaller portion of the not-working, food stamp rates are up. Second, post-recession jobs have been lower-wage. E.g., http://www.chicagotribune.com/business/breaking/chi-jobs-wages-inequality-mayors-20140811-story.html. So there are more working poor qualify for and apply for food stamps.
The Disability number is an alarm bell. Are people really in worse health than before?
Also, they should confine their analysis to the 24-54 demographic which is surely the bellwether for the health of the employment market.
Much is not most. Half still means much is not. Wages have fallen so even those that are working are more likely to be on food stamps. Instead of arguing for wage flexibility we should be arguing for increased wage support. Another reason talk of the disincentives to work is misguided when it is the reduced incentives to work and the one is not the obverse of the other.
It’s kind of like that old song: “I wonder who’s kissing her now” that’s goes on to “but I really don’t want to Know.”
But what would be potentially useful:
A Scale of use of “benefits” (food stamps, e.g.) by:
Age groups-
Unemployed- By lengths of unemployment
Employed – by full or part time
Students – In what classes of Ed.
But, they really don’t want to know.
Great source: “The Great Society at Fifty” by Eberstadt
typo: “UNEMP= *un*employed”
To me the ‘alarm bell’ is that the model seems to not hinge on what happens regarding food stamps and those not in the labor force.
Switching notation a bit where “~” is “not” (or “UN”).
You’re reaching for population statistics. What’s actually described is FS/POP for the segment of the population that is ~EMP and LF. This form must be integrated (summed over all discrete independent categories) to get the desired answer:
P(FS) ==
FS/POP = LF/POP * ~EMP/LF * FS/~EMP +
LF/POP * EMP/LF * FS/EMP +
~LF/POP * FS/~LF
(collapsing the meaningless categories EMP/~LF and ~EMP/~LF)
Upshot: the model doesn’t help; as you observe in logic/prose, who knows where FS goes (given just these axioms)