Saturday, June 13, 2015

Do low education foreigners work more and harder than native born Americans with low education attainment?

The world is complex and data is interesting. From Need a high school diploma to work full-time? Natives, yes. Immigrants, no. by Brad Hershbein and Elisa Jácome. This is a great object lesson in being careful about averages and maintaining contextual awareness. Sometimes we get lost in the abstract. Abstract ideas can be appealing but misleading.
The share of men with a high school degree or some college (but not a BA) who worked full-time and year-round fell from 76 percent in 1990 to 68 percent in 2013. Interestingly, for men without a high school degree, the share was low but unchanged, at 55 percent. Why did men without a high school degree not experience the same declines in full-time year-round employment?
For a national economy, the erosion of labor force participation in general is of course concerning. it is hard to continue increasing productivity if fewer people are working. The loss of full-time workers is even more concerning because full-time work is what tends to drive higher efficiency and effectiveness.

So why did the labor force participation rate for less educated remain level while it was falling for others?
It turns out that overall employment rates for men without a high school degree mask important differences between immigrants and native-born individuals. Immigrants now account for a much larger share of men without a high school degree, doubling from about one-quarter to one-half between 1990 and 2013.

US-born men without a high school degree have seen a dramatic reduction in their employment rate—from 72 to 58 percent. The fall was also dramatic for the share working full-time and year-round, sliding from 54 to 42 percent.

In sharp contrast, the employment rate of immigrant men without a high school degree rose from 80 to 87 percent, and the share working full-time and year-round rose from 57 to 66 percent.
Well that's ideologically inconvenient. Many people want to believe that a bad economy hurts the poor and least educated the worst and that that in turn causes all sorts of sociological undesirable consequences such as crime. Empirically that is not true but that is what is believed.

In this case though, the data is indicating that the least educated and most impoverished (foreigners coming in with little to their name and usually with the additional burden of language and culture differences) are not only improving their labor force participation dramatically but doing so at the same time that the native born educationally comparable are letting their own labor force participation drop dramatically.

Hershbein and Jácome ask the relevant question.
Why are less-educated immigrants more likely than US-born peers to work?
They ask a good question but they have no idea what the answer might be. They trot out some mumbo jumbo about country of origin rapid pace of technological change, trade and globalization, and the growing generosity (relative to possible earnings) of safety-net programs without really developing an argument to explain the conundrum.

Right leaning ideologues might instinctively resort to "It's culture" with hoary stereotypes of hardworking high work ethic poor foreigners and low work ethic, lazy welfare layabouts. And undoubtedly that could be some element of the answer.

I suspect that the real answer is different and more complex. I would look at this using a somewhat different set of abstractions to get at a plausible answer. I suspect, even though we are controlling for education level as the variable of interest, we are not actually comparing apples and oranges.

Instead of looking at education level, let's look at whether the two populations (native and foreign born) are also alike in other pertinent attributes such as age, non-cognitive skills, health, mental health, substance abuse, etc..

The first aspect to consider is that education attainment is a product of both opportunity (usually much less available among poor foreign immigrants but universally available in the US) and personal behavior (self-control, self-discipline, perseverance, etc., the non-cognitive skills identified by James Heckman). Those who fail to complete high school in the US usually fail owing to behavioral/non-cognitive skill issues. Poor foreigners who fail to complete high school fail for economic reasons or simple lack of availability. The upshot is that foreign and domestic born might look the same in terms of low education attainment, they may have dramatically different non-cognitive skills.

Second. Domestic born individuals who have failed to achieve high school completion cover the entire age gamut from 18 to 75. If you measure labor force participation rates across a wide age range, you are going to get a far different answer than if you measure across a narrow age range. If all foreign born low education attainment rate individuals are between 15 and 35, you would expect their average participation rate to be much higher simply because of their average age.

Third. Self-selection among poor immigrants. Yes, they probably do have a high work ethic. They probably also have lower mental health issues, lower substance abuse, etc. In other words, the act of immigration is both difficult and risky, tending to weed out the lower functioning and the less able.

There are probably other ways in which low education attainment native born and foreigners differ but I think these might be the most significant: the foreign born low education attainment individuals have better non-cognitive skills, they are younger and healthier, and the very act of immigration self-selects for some degree of ability and accomplishment. The native borns who have lower education attainment are probably, on average, older, have lower non-cognitive skills, have more health issues, have more mental health issues, and have a broader array of social pathologies.

If that is true, then the mystery disappears. It is not that poor foreign immigrants with low education attainment are "better" people than our own native born low education attainment. It is that they have a materially different array of other attributes (health, youth, non-cognitive skills, etc.) which independently correlate with high labor force participation rate.

I doubt what I am describing is the full story but I suspect that the fact that we are not comparing apples to oranges is the root of the conundrum.

No comments:

Post a Comment