Log in

No account? Create an account
There's Only One Type of Immigrant - Arvind Narayanan's journal [entries|archive|friends|userinfo]

There's Only One Type of Immigrant [Jun. 13th, 2011|12:48 pm]
Arvind Narayanan
[Tags|, , , , ]

Late one evening many years ago, as a fresh graduate student and recent immigrant, I was sitting in my cubicle working when my colleague Justin walked in to pick something up from his desk. As he was leaving, noticing that I was the only one left, he quipped, “Ah, Arvind, doing the jobs that Americans won’t do.” 

At that moment I was enlightened.

Before I go on, let me address those of you who I’m sure are sitting there offended on my behalf. One of the awesome things about Justin is that he is not hindered by political correctness, and I’m glad that he’s my friend and says things like this.

Everyone “knows” that there are two types of jobs held by immigrants, roughly on opposite ends of the skill/wage spectrum: blue-collar and white-collar, or low-skilled and high-skilled. The former mow lawns, clean homes, and drive taxis; they are able to get those jobs because they accept lower wages. The latter are doctors, scientists and engineers; they are able to get those jobs because their skills are in shortage. If you’re on the Left, you might also believe that this is because the US sucks at math and science education.

Nonsense. There’s only one type of immigrant.

Here’s my theory. In any country that’s under immigration pressure (more people trying to get in than get out), through a combination of market forces and laws, native-born residents are able to pick more desirable jobs on average, where desirability is a composite that incorporates all the relevant factors—how much fun the job is, how much education/training it requires, and of course, pay. The negative correlation between desirability of a job category and the percentage of immigrant workers is strong, but not perfect.

The most important insight from this theory—and the main reason for this post—is that poor math and science education in the US is not a causing a lack of native-born workers in tech jobs, but is a consequence of it.*

In other words, educational opportunities in math and science exist aplenty, but students choose not to avail of them (not paying attention in math class, not taking AP classes or math-y majors in college, etc.), since the job prospects aren’t very appetizing, say compared to a career in law.

Of course, students don’t always consciously make these decisions; the actual processes by which demand influences supply are complex, and may include such factors as stigmatization of certain fields. For example, this recent article describes how the financial success of Silicon Valley is making computer science cool again, leading to increased enrollment. The rejuvenation arguably also improves the quality of undergraduate CS programs (and longer term, via a trickle-down effect, K-12 math and science).

This view also explains many other things—the existence of numerous high-skill professions with negligible immigrant representation, such as airline pilot and astronaut, and the average-to-poor social status of immigrant-dominated high-skill job categories. It also partially explains the lack of American chess players: it’s not that Americans don’t become chess players because they suck at it; they suck at it because they don’t want a career in chess.

The broader principle that demand for jobs determines supply (i.e., educational opportunities) as well as social norms, rather than the other way around, applies equally well to other countries. For example, Indians value tech skills and (some) Indian engineering colleges are exceptionally good because those are the most valuable job sectors in India. Unfortunately, Indians who move to the US usually raise their kids based on the same norms, which is not only economically suboptimal but also inflicts a toll on emotional wellbeing. (See also: Asian parent stereotypes; Tiger parenting.)

* About a year ago I read an article that discussed the supply-determines-demand vs demand-determines-supply debate in the context of math/science education in America. It noted that the latter view is still in the minority, but is winning converts among experts. I’m really bummed that I’m unable to find that article any more. If someone could fish it up for me, I’d be very grateful.

Commenters on the linked post provided convincing explanations of why Americans don’t want a career in chess.

[User Picture]From: dannyman
2011-06-14 09:13 pm (UTC)


"At either end of the economic scale, immigrants are doing the jobs that Americans don't want to do."

(Full disclosure, I read every word.)

(Speculation: H1B/immigration process is the white collar equivalent of paying a coyote to guide you through the desert at night.)
(Reply) (Thread)
[User Picture]From: arvindn
2011-06-14 09:50 pm (UTC)

Re: tl;dr

I did start with the tl;dr before elaborating :-)
(Reply) (Parent) (Thread)
From: jake7891
2011-06-19 10:18 pm (UTC)
> “Ah, Arvind, doing the jobs that Americans won’t do.”

Presuming you were working on the computer software industry, what *kind* of jobs (in that industry) are the immigrants expected to do while Americans are not? I can understand the taxi driver jobs, but just not in the software industry.
(Reply) (Thread)
[User Picture]From: arvindn
2011-06-19 11:36 pm (UTC)
Justin's comment was in reference to graduate school—where 70% of those in my (CS) program were foreigners. Scientific research as a whole is immigrant-dominated.

The software industry also has a high immigrant ratio, although less skewed than science. I'm not saying Americans aren't expected to do these jobs, I'm saying they don't want to do it. It's obviously not black-and-white; I'm talking about statistical averages here.
(Reply) (Parent) (Thread)