Comment by thephyber
2 days ago
What do you mean “fair”? What happens in the years/decades between when this hypothetical system is enacted and when the US can train up sufficient workers to substitute the labor force we currently have with H1-B?
Your proposal will mean 99% of all of the H1-B allocation will go to hedge fund quants and 1% maybe go to an AI researcher, but all of the materials science (eg. Cutting edge battery tech), semiconductor fabrication, neuroscience, pharmaceutical research etc will have to go without the skilled workers they currently get from visas. This is a recipe for the Boeingization of the US economy.
exactly wrong. Americans are dissuaded from going into these highly skilled fields because anyone talented enough to do those things realizes they can make much more building SAASes or working on wall street.
the Boeingization of the economy is mbas and bean counter middle management realizing that an H1-B is much cheaper than a citizen and opting to buy that labor, even if it's worse quality. as management, you put an ass into a seat, so job accomplished, here's your accolade.
Your comment has a thin veneer of truth, but actually has very little to do with H1-B system.
Boeing is full of bean counters now, but they are optimizing things like opening factories in poor non-unions states (the South Carolina factory has had lots of whistleblowers screaming about lack of training and pressure to build faster than is safe), convincing the FAA to let Boeing employees do regulatory review on their own company, etc. few or none of Boeing’s problems are solved by eliminating/reducing H1-Bs for that company/industry, which is why I chose them as the example.
“Americans are dissuaded…”
This has an emotional appeal to intuition, but I don’t think it’s what’s causing Americans not to compete for jobs/industries that heavily use H1-B. If it was, there would simply be a market competition and those programmer salaries would drop. Instead, I think Americans have been convinced by Theil types to avoid US universities (either for cultural reasons or ROI reasons). You seem to be making an argument that the ROI would be better if H1-Bs were scarce, but that wouldn’t change the fact that tuition in elite US institutions is expensive and seats are scarce+competitive. Without also changing either the university system to seat more students or companies to hire from different signals (instead of highly prizing the bland name of the university), American job applicants won’t be dissuaded from getting those degrees.
Arguably H1-Bs have done the most damage to US programmers, but there are several other structural problems regarding programmer hiring in the US. The big tech collusion to reduce employee poaching (not current, but recent past), application process (“resume firewall”, ghost jobs, deluge of automated applications), the interview process (we seem to have optimized for gotcha questions and LeetCoding tests, rather than real world requirements), high interest rates (higher than the recent past) have squeezed VC funding and closed the wallets of employers, and the race to replace/augment salary employees with AI agents. All of these are structural problems that arguably do more dissuading than the visa system.
Or... those other parts of the economy increase salaries for skilled labor?
If we can only bring 85,000 people into the country on one type of visa, doesn't it make sense to prioritize those that will bring the most value (tax revenue, in this case)?
And if that's not enough people... raise the limit? And be confident that a raised limit is still keeping a high quality bar on entrants?
Option 1: you give a visa to a quant with 2M/y today’s salary
Option 2: you give a visa to a PhD to work for 150k/year in a small biopharma startup that thinks it has the solution to cancer.
This salary stacked ranking optimizes for today’s worth of work. Not its potential.
Are there 85,000 new hedge fund quants that need to be hired each year? I guess it is more like 1,000. The number of people employed as quants at hedge funds is incredibly small.
You could make multiple pools, having separate ones carved out for research and advanced technology.
A lot of H1Bs are not working on anything you described though.
"materials science (eg. Cutting edge battery tech), semiconductor fabrication, neuroscience, pharmaceutical research "
This is a beautiful fantasy for H-1B, that is totally disconnected from reality. What is that 1% of the H-1Bs currently? It is mostly IT and software slop jobs.
Here are the top 40 employers, it isn't going to hurt research in the US to cut them to zero.
Amazon.Com Services
Cognizant Technology Solutions
Ernst & Young
Tata Consultancy Services
Google
Microsoft
Infosys
Meta Platforms
Intel
Hcl America
Amazon Web Services
IBM
Jpmorgan Chase
Walmart
Apple
Accenture
Capgemini
Ltimindtree
Deloitte Consulting
Salesforce
Qualcomm
Tesla
Amazon Development Center
Wipro
Fidelity Technology Group
Tech Mahindra
Compunnel Software Group
Deloitte Touche
Mphasis
Nvidia
Adobe
Bytedance
Goldman, Sachs
Cisco
Linkedin
Pricewaterhousecoopers Advisory Services
Paypal
Ebay
Servicenow
Visa USA
For non-slop jobs, give them a green card and fast track to citizenship. For an IT consultant, no thanks.
source: https://www.myvisajobs.com/reports/h1b/
Intel H1Bs are engaged in semiconductor research.
Well it's Intel... not really selling your case here.
Doubtful.
> Your proposal will mean 99% of all of the H1-B allocation will go to hedge fund quants and 1% maybe go to an AI researcher, but all of the materials science (eg. Cutting edge battery tech), semiconductor fabrication, neuroscience, pharmaceutical research etc will have to go without the skilled workers they currently get from visas. This is a recipe for the Boeingization of the US economy.
If they're that necessary, let companies hire them on green card visas.
This is just an argument against allowing the market to set wages, which you could make if you wanted to but it is not a strong one.