Hacker News Reader

Comment by rayiner | Hacker News Reader

> It proposes investments in computer science and engineering education . . . and attaching a green card to the diplomas of foreign-born students earning STEM degrees.

These sound good in theory, but in effect they're an industry-specific labor subsidy. When the government "invests" in STEM education, what happens is that the companies that need to hire engineers save money both as a result of reduced training costs and as a by-product of increased supply driving down wages.

It's worth keeping in mind that STEM education isn't necessarily a public good.[1] The benefits of education aimed at developing industry-specific skills accrues mostly to those in an employer-employee relationship. Under those circumstances, it's most economically efficient to force the employer and the employee--rather than the public--to bear the cost of that education.

[1] https://en.wikipedia.org/wiki/Public_good

View on HN

Agreed. People here are mostly thinking "brilliant computer scientist" or "will cure cancer," but many universities have degree programs that are revenue generators, requiring full tuition payments from students and attracting many foreign nationals, which may not be held to the same standards as other programs. These programs serve plenty of students from China and Saudi Arabia whose parents are making an investment in getting an American option for the their child or the family. This is a very rational decision, but I don't think that it serves US interests to give an automatic green card to a kid from (pardon my apparent condescension) Directional State U because his parents paid enough to get him to scrape through a master's in geology or CS.

I'm at a big state U, mildly insulated from the pressures to develop a program at which anyone can succeed, but I have a lot of academic friends who have encountered significant pressure from administrations to keep pass rates high in revenue-generating STEM programs (many of them masters' programs). Attaching a green card to the diploma is a boon to these programs and some non-US students: all the incentives are there to just keep taking that full tuition and graduating students without working them too hard. Win-win, and it cuts down on the end-of-semester "I have to leave the country if I fail!" conversations, which I've had several times and are always unpleasant. These micro-level problems could lead to a flooding of the market with people whose educations just aren't that great.

Coming from a family of immigrants, I'm all for making green card processes more efficient. Knowing America's education system, attaching a green card to diplomas of STEM graduates is pay-to-play and that's not cool.

> "brilliant computer scientist" or "will cure cancer,"

These are already covered by eb-2 anyways.

So? An increase in citizens trained in a highly intellectually demanding field and encouraging immigration of some of the best and brightest of other nations makes it more than worthwhile to pursue.

> So?

So, the point of these programs isn't citizen enrichment. The point is job training.

Consider: philosophy is also intellectually demanding, but I don't see "Philosophy for All" educational initiatives.

I think a lot of people who oppose CS Ed investments are self-interested, but that doesn't mean there aren't genuine problems with government policies that tie themselves too closely to today's labor market.

If we're not careful, "learn to code and you'll have a job" will be tomorrow's equivalent of "get a college degree and you'll have a job".

If we're not careful, "learn to code and you'll have a job" will be tomorrow's equivalent of "get a college degree and you'll have a job".

It already is. It's not a matter of if the bottom will drop out of STEM=JOBS, but a matter of when.

If this plan goes thru, maybe a university should undertake some sort of advanced accreditation process to be able to award green cards. I would hazard a guess that some universities (state unis?) aren't necessarily attracting the best and brightest. Anecdotal, but I took a few masters CS courses at an average state uni, and the foreign students had a similar academic rigor and aptitude as the general uni population for the most part, displayed through mean test and assignment scores.

I could be totally wrong but it might be dangerous to assume we are attracting only the best and brightest, no? Seems like we are attracting anyone on the spectrum who can dole out some money to a university and get a stem degree.

Even if it were only one in a hundred foreign students that turned out to actually be exceptionally talented, taking them all is a drop in the bucket relative to overall incoming immigration, both legal and not. (Even if the ratio were one in a thousand, it would still not be all that bad; the STEM student population is not all that large.)

Moreover, even if most foreign born students in a STEM program were merely as good as native students, as you suggest, that still puts them intellectually and educationally well above the average citizen, let alone the average immigrant.

There is no way that it does not make sense to offer foreign STEM students green cards, so long as the university is reasonably accredited.

Isn't that a good thing? What would be some negative consequences?

Wages going down and competition going up for people in the tech industry.

I worry more about quality going down than either of those. Companies will amass huge armies of truly awful engineers and it will become a profession of attrition.

Not to mention that we will motivate foreign students to study mostly only STEM and thus deprive ourselves of the diversity of thought and the artistic and critical labor output of foreign workers in non-STEM domains.

Makes me think of Arcade Fire: "One day they will see it's long gone."

Not just the tech industry too. There are plenty of areas of STEM which are totally saturated with smart effective highly educated people and the result is low wages and poor employment outcomes. Just look at chemistry or biology.

But fewer complaints of H1-B wage slavery (note: I'm exaggerating somewhat), which also suppresses salaries. Hard to say which is more significant an effect.

The tech industry isn't zero-sum though, even from just a purely selfish "can I get a job" POV. More talent in the pool will arguably make the industry bigger, rather than diluting the value of that talent. There's just so much room to grow.

It favors companies that hire engineers (vs. those that don't) and may lower engineer wages. Whether or not these are negative depend on your view.

What's the negative view on "it favors companies that hire engineers over those that don't"?

It is a market distortion. It favors the industry that is getting government support, taking away valuable resources from other industries that would do better in a free market.

Practically, HN is pro-engineering, so we likely approve of such moves. Similar subsidies in other industries might not be viewed as favourably.

Personally, I am not against it, because as someone who has many of his peers playing a zero-sum game in mathematical finance, I've seen some of the downsides the top talent being allocated where the free market desires them most.

It's not just a pro- versus anti-engineering issue. A lot of engineering fields are rigorous and difficult while not being very highly paid.[1] Part of the reason is the market distortion you point out: both on the demand side (e.g. aerospace is a heavily regulated industry with few opportunities to build "unicorns"), and on the supply side (e.g. the government invested heavily in churning out aerospace degrees during the Cold War).

[1] I made more money out of school in software, with no relevant degree, than I would have in aerospace, with a B.S. in the field.

Ah, in the "engineers vs other professions" sense. I was parsing it as "hiring vs non-hiring".

The Econ 101 answer is that we'll have some number of people going into STEM fields would be more productive to the economy (and probably personally better off) as lawyers or cops or bakers or whatever. More realistically, it's likely to cause average salaries for engineers to decline somewhat as supply catches up to demand.

I'd ask the manager of my local Starbucks, who has a PhD in Biology.

It's government trying to pick winners, which rarely works out that well. I think government should stick to basics like infrastructure, education and making sure everyone has decent health care; that kind of thing that companies are less suited for.

Nothing you said is restricted to STEM education. Do you favor eliminating subsidies for K-12 as well? If not, why not?

I think education occupies a range on the continuum between public and private goods. Basic literacy probably yields positive externalities, but specialized, employment-targeted skills mainly benefit the employee and employer.

That said, I think K-12 could probably be 1-8 and yield similar positive externalities.

> These sound good in theory

I disagree.

> but in effect they're an industry-specific labor subsidy.

What do you mean, "but in effect" -- that's clearly why they are being advocated.

> When the government "invests" in STEM education, what happens is that the companies that need to hire engineers save money both as a result of reduced training costs and as a by-product of increased supply driving down wages.

This is win-win (for domestic labor and capital [which is less tied to the domestic market], respectively) when you are training your citizens to take in-demand jobs. Its lose-win (on the same terms) when you are subsidizing and incentivizing foreign students to consume domestic educational resources to take in-demand jobs.

Reply on news.ycombinator.com