More on Rethinking Immigration
I really enjoyed Giovanni Peri and Giamarco Ottaviano's Rethinking the Effects of Immigration on Wages. The argument was fairly straightforward. I'll leave out the math and subtleties, and try to give a reasonably lay abstract. The motivation for the paper comes from the fact that in the last 30 years, the share of immigrant workers in our labor force has increased from 5% to nearly 15%. And political groups left and right are quick to blame immigrants for "stealing our jobs", and so on. Which begs the question - are they?
The paper answers two questions:
(1) What has been the effect of immigration on average wages?
(2) What has the distribution of these effects looked like? i.e. if we partition the population into groups by education level (super low, low, medium, high), how have these wages fared due to immigration?
So let's consider the previous literature on this subject:
(a) The Pre-Moderns: these guys assumed that foreign labor and domestic labor were perfectly substitutable and that labor was homogenous. Pretty shitty.
(b) The Moderns: they began to utilize the argument of imperfect substitution. They split roughly into two groups:
i. Researchers considering foreign vs. native as imperfect substitutes, but disregarding education and experience.
ii. Researchers considering experience and education level in calculating imperfect substitutions, but disregarding weighting this on foreign-ness or national-ness.
Lastly, it is worth noting that almost all moderns regarded capital stock to be fixed in the short run.
What did Peri say in response to all of this?
Well, first, he said that partitioning the levels of imperfect substitution into (b)i or (b)ii was a mistake. Instead, we should use a nested function, that adjusted first for education level, second for experience given an education level, and third on whether or not someone was foreign given an education level and experience.
Second, Peri argued that this notion of a fixed capital stock makes little sense, because we do not magically come up with capital every 10 years (census years) en masse. Instead, firms see that immigrants are entering yearly, and respond to this by continuously investing in more and more capital.
It turns out that these two changes create a rather general analysis of the problem - and in my opinion a reasonably "complete one" relative to what existed before. In particular, I liked the methodology revision.
As for what happens - well it turns out that if immigrants come in with levels of education and experience close to most of society, obviously these guys are too substitutable and so wage rates drop. Conversely, if they have rather different levels of education and experience, obviously they aren't so substitutable and in fact wage rate can potentially go up.
As for empirics, average wage did go up, and in the 4 rough blocks of education levels (college grad, college dropout, high school grad, high school dropout) here is the following order of who did best to who did worst due to immigration:
1. College Dropouts (+++)
2. High School Grads (++)
3. College Grads (+)
4. High School Dropouts (-)
So here is a good question to end on - can anyone give any intuition as to why this ordering makes sense?
The paper answers two questions:
(1) What has been the effect of immigration on average wages?
(2) What has the distribution of these effects looked like? i.e. if we partition the population into groups by education level (super low, low, medium, high), how have these wages fared due to immigration?
So let's consider the previous literature on this subject:
(a) The Pre-Moderns: these guys assumed that foreign labor and domestic labor were perfectly substitutable and that labor was homogenous. Pretty shitty.
(b) The Moderns: they began to utilize the argument of imperfect substitution. They split roughly into two groups:
i. Researchers considering foreign vs. native as imperfect substitutes, but disregarding education and experience.
ii. Researchers considering experience and education level in calculating imperfect substitutions, but disregarding weighting this on foreign-ness or national-ness.
Lastly, it is worth noting that almost all moderns regarded capital stock to be fixed in the short run.
What did Peri say in response to all of this?
Well, first, he said that partitioning the levels of imperfect substitution into (b)i or (b)ii was a mistake. Instead, we should use a nested function, that adjusted first for education level, second for experience given an education level, and third on whether or not someone was foreign given an education level and experience.
Second, Peri argued that this notion of a fixed capital stock makes little sense, because we do not magically come up with capital every 10 years (census years) en masse. Instead, firms see that immigrants are entering yearly, and respond to this by continuously investing in more and more capital.
It turns out that these two changes create a rather general analysis of the problem - and in my opinion a reasonably "complete one" relative to what existed before. In particular, I liked the methodology revision.
As for what happens - well it turns out that if immigrants come in with levels of education and experience close to most of society, obviously these guys are too substitutable and so wage rates drop. Conversely, if they have rather different levels of education and experience, obviously they aren't so substitutable and in fact wage rate can potentially go up.
As for empirics, average wage did go up, and in the 4 rough blocks of education levels (college grad, college dropout, high school grad, high school dropout) here is the following order of who did best to who did worst due to immigration:
1. College Dropouts (+++)
2. High School Grads (++)
3. College Grads (+)
4. High School Dropouts (-)
So here is a good question to end on - can anyone give any intuition as to why this ordering makes sense?
posted by arun at 10:35 PM
0 Comments:
Post a Comment
<< Home