Hacker News Reader

Comment by femiagbabiaka | Hacker News Reader

The idea that those on H1B visas are bringing wages down for software engineers is a popular one. Is it supported by data?

If it isn’t, this is just thinly veiled xenophobia.

Personally, I think the more diverse groups have an opportunity to come to America legally, the better.

View on HN

Look at it from the other direction: large companies have fought giving H1-B workers better protections. They wouldn’t be doing that if they weren’t profiting from workers with less negotiating ability.

That’s also my position: don’t ban them, give them flexibility to switch jobs like everyone else so companies can’t hold the threat of deportation over them if they want a better deal.

Agreed — more freedom, not less, is the ideal solution here.

You replied to a comment arguing against positions made from speculation without data by supplying more speculation without data.

There's some data out there but it's also hard to compare reliably since you need to account for differences in skill level when comparing salaries.

My point was that the large companies which have the best data, and whole teams of people doing the math on this issue, have consistently fought to preserve the old ways of doing it and since they're for-profit corporations it seems unlikely that they're doing this for any reason other than the econ 101 explanation that they profit from workers who can't negotiate as well.

I think when you're contesting a well known economic principle (increased supply tends to decrease prices given equal demand) the impetus is on you to prove the claim.

That said, this would be difficult to demonstrate with data because the effect manifests as salaries growing slower than they would without H1Bs. Obviously, there are a million other variables involved, so without some kind of controlled study it's unlikely anybody can be more specific than "increased labor supply decreases wages".

Software engineers are being hired at much higher rates than any other profession in the US and that has been going up for decades.

https://www.bls.gov/ooh/computer-and-information-technology/...

And so has our pay. But, like I said, there are million other variables besides the size of the labor pool that affect salaries. In general, though, more candidates available for a position means a lower salary.

And there are more serious problems with H1B than just lowering salaries. It's very difficult for H1B workers to switch jobs once they're here, and employers (especially contracting shops) use that as leverage against them.

And don't even get me started on H2B workers, who often make minimum wage, have to live in employee housing, and have even less recourse than H1B workers because they make far less money.

We need better immigration policies, and I don't think tying immigration status to job status is the best way forward.

You understand that the predictions for industry growth on those pages are virtually the same as the number of CS bachelor's degree graduates every year.

We don't need to important more labor that that which is getting educated inside of the United States short of those who demonstrate extraordinary ability (i.e. those who come into the country on EB1 rather than on H1B).

> You understand that the predictions for industry growth on those pages are virtually the same as the number of CS bachelor's degree graduates every year.

they're not...

> We don't need to important more labor that that which is getting educated inside of the United States short of those who demonstrate extraordinary ability (i.e. those who come into the country on EB1 rather than on H1B).

We can't produce enough internally, so we do.

You're wrong, look at the data. There are 75,000+ graduates of bachelors and masters programs in the US each year, with many more in related majors and in bootcamps, where the BLS data predicts an industry growth for software developers (including standard, applications, and systems) of around 600,000 workers between 2016 and 2026.

You can make the argument that many of those with degrees aren't as skilled as companies would like, and in turn they seek employees from abroad, but it's blatently false that the US isn't graduating enough engineers to meet demand.

That's a lie that's being propagated by tech companies. The US educates more than enough engineers to keep up with demand. The argument about "structural imbalance" in the job market has been debunked over and over again. I assume that it's not lack of google coverage stopping from researching the issue so linking you a few papers won't change your mind one bit, so I won't bother wasting time doing that.

Your look at the situation is naive at best. The fact that wages are growing doesn't mean they couldn't grow more, or that are growing for everyone.

WSJ > https://www.wsj.com/articles/a-new-look-at-the-h-1b-visa-pro...

Fortune > http://fortune.com/2017/02/15/h1-b-silicon-valley-wages/

Science Magazine > http://www.sciencemag.org/careers/2015/05/economists-h-1b-vi...

There's huge age discrimination at work in the tech industry, and older, very qualified workers struggle to find good jobs because tech employers prefer young and naive serfs.

Every time we have this conversation we have to do this silly song and dance routine instead of, you know, doing a quick google search.

WSJ > https://www.wsj.com/articles/a-new-look-at-the-h-1b-visa-pro...

Fortune > http://fortune.com/2017/02/15/h1-b-silicon-valley-wages/

Science Magazine > http://www.sciencemag.org/careers/2015/05/economists-h-1b-vi...

I particularly appreciate the part where you passively-aggressively calle me a xenophobe not knowing I'm an immigrant too. Classy move.

That paper discussed in Science Magazine is very interesting. Here's part of the abstract from 'The Effects of High-Skilled Immigration Policy on Firms: Evidence from Visa Lotteries' (2016) - https://gspp.berkeley.edu/assets/uploads/research/pdf/h1b.pd...

"We compare winning and losing firms in the Fiscal Year 2006 and 2007 lotteries for H-1B visas, matching administrative data on these lotteries to administrative tax data on U.S. firms, and to approved U.S. patents. Winning additional H-1B visas causes at most a moderate increase in firms’ overall employment, and these H-1Bs therefore substantially crowd out firms’ employment of other workers. Additional H-1Bs generally have insignificant and at most modest effects on firms’ patenting and use of the research and experimentation tax credit. There is some evidence that additional H-1Bs lead to lower average employee earnings and higher firm profits."

However anecdotal this claim may be, it seems fairly intuitive given the basic economic model of supply and demand. Can you explain how greatly increasing the supply of developers by importing them from overseas will not affect the demand, and therefore negatively affect the cost of labour?

Qualified supply is nowhere close to matching demand. Ask anyone that has to do recruiting for their teams.

Source: personal experience

Talking about demand exceeding supply is somewhat nonsensical without reference to a price.

What you're saying is "at the price my company is willing to pay, we're unable to find qualified candidates." That means you can either pay more to attract qualified candidates, or do without.

Alternatively, we can raise the supply curve, which has the affect of lowering wages, and might allow your company to attract talent without paying more.

Ok, fair enough, my comment was perhaps flippant.

My company pays at least market rate and can afford to go higher if the candidate negotiates, so I don't think that's the issue. At least, we never had a candidate refuse a position due to money concerns. The skills we look for are not that specialized either, it's not like I'm looking for ML PhD's with a background in medicine.

Anecdotally, it just feels like almost everyone who is already good, is already happily employed. So either we need to head-hunt candidates, hope to coincidentally catch them while they're between jobs, or hire very junior candidates and train them up.

Of course, the issue might also be just that we need better recruiters to populate the candidate pool :)

I don't know which market you are hiring in, but for jobs which pay well into 300K+ (many multiples of median household income across the country), where engineers are pampered with free meals and massages, there are shortages in qualified supply.

This is only true for experienced professionals, and is something that will change in the next ten years given current metrics. Hiring new graduates whose qualification is that they have a computer science degree isn't difficult given that 60,000 graduate with a BS and 15,000 graduate with an MS every year.

100k this year.

>> Qualified supply is nowhere close to matching demand.

This is wrong on so many levels.

1. "Qualified" supply means experience. That means you arent hiring 22yos working at subsistence salaries shacking up with three buddies in a double-bunk apartment. It means you have to pay a "qualified person salary." I'm a qualified person, I have 18yrs experience at top firms. But guess what...175k/yr in SF/SV doesnt cut it, it gets you a 1br apt. You cant raise 2 kids in a 1br apartment. Where would the kids sleep? in the bathroom? on the dining table? Also, option compensation doesnt count, as I cant pay rent with options.

2. If you want "qualified talent" perhaps SF/SV would not age discriminate. But they do. They want young people willing to earn so little they are forced into having 3 roommates.

3. There are tons of qualified workers. Most of us dont want to move to SF/SV because the prevailing wage is too low. Pay the right wage (i.e., perhaps something that one can raise a family on) and your "shortage will magically disappear.

>> Qualified supply is nowhere close to matching demand.

I think what this typically means is: we're looking for talented developers willing to earn so littler they are forced to bunk with multiple roommates and/or live in roach infested apartments, something only desperate overseas workers are willing to do.

That's nonsense. As an industry we have an enormous age and gender bias. I know plenty of older folks that are great software engineers and can't find work because nobody will give them a chance to interview.

here for some actual facts: https://archives.cjr.org/reports/what_scientist_shortage.php

> I know plenty of older folks that are great software engineers and can't find work because nobody will give them a chance to interview.

maybe tell them about triplebyte

Yeah, it's true about the hiring biases in general.

Where are they looking? Have you shown them "Who's hiring" HN threads?

Software engineers have a habit of assuming problems are more intuitive than they actually are.

Even with H1B, there are literally not enough engineers in the Valley to saturate the demand from companies. BLS estimates that the number of employed software engineers will go up 24% next year, that’s insane.

Also, I’m not seeing this supposed depression in salaries. The median salary for a software engineer is 103k a year: https://www.bls.gov/ooh/computer-and-information-technology/.... That number has been growing expontially in the last decade and a half.

So there’s data that shows that engineers are getting hired more and paid more than ever before. I can’t find any to prove that H1Bs are dragging down salaries.

>Even with H1B, there are literally not enough engineers in the Valley to saturate the demand from companies.

Why is this? This sounds like a huge U.S. talent development problem. How does this visa change help to fix this issue?

>I can’t find any to prove that H1Bs are dragging down salaries.

The very fact that the number of employed software engineers is expected to go up by 24% in a single year is the evidence you're looking for that there is a massive supply distortion in the labor market.

> Why is this? This sounds like a huge U.S. talent development problem. How does this visa change help to fix this issue?

It doesn’t, I’m against it.

> The very fact that the number of employed software engineers is expected to go up by 24% in a single year is the evidence you're looking for that there is a massive supply distortion in the labor market.

Yes... because the demand for software engineers is increasing every year and can’t keep up with supply.

>and can’t keep up with supply.

yes...the very definition of a labor distortion that drives down wages.

I feel like we can’t be reading the same thing. You’re saying that a market where there is more demand than supply is one where prices are lower? Again, I’m going to need a citation for that.

where did I say that prices are lower? I’m expressioning skepticism that a whopping 24% increase in hiring is due only to demand increases, which you heavily implied is the case. How much more would wages rise here if we weren’t artificially increasing the supply like this? We could certainly find out, if we wanted. But that would require some policy changes that are not popular to employers who are, incidentally, not interested in letting wages rise.

It makes for a good story but it's actually bullshit. There's no labor shortage in the sciences. There's just a shortage of companies willing to pay living wages.

https://archives.cjr.org/reports/what_scientist_shortage.php

>> Even with H1B, there are literally not enough engineers in the Valley to saturate the demand from companies.

By "not enough engineers" I'm assuming you mean not enough engineers willing to work for so little they are forced to bunk with 3 roommates?

I know plenty of engineers. Most dont want to work as SWEs because the median salary barely affords a studio apartment, especially after California taxes.

Also, what about the 40+yr olds who are oddly purged from the workforce regularly. If there were a shortage, elderly workers would be retained, not purged in favor of cheaper, younger workers.

Tons of engineers go into consulting, real estate, finance, etc because engineers simply dont make enough, especially given the SF/SV center-of-gravity of jobs. There isnt some magical shortage of workers. There is a shortage of workers willing to work for near subsistence wages [in SF/SV prices]

Yes. There's value in being able to change jobs as you see please. That value is about 1.3 to 1.5x of not having the freedom to do so.

Workers under H1B transfer jobs less often, but its not impossible or even uncommon, except under the Trump administration.

From an economics perspective, unless the hiring of more software engineers results in an increased demand for software engineers (i.e. there being a significant latent demand), H1B absolutely then lowers the earnings potential of engineers. Especially those earning more towards the middle and bottom of the market.

I'd also add that BLS data suggests that this induced demand isn't true. As the projected employment growth for engineers between 2016 and 2026 is 600,000 where there are around 60,000 people graduating with computer science bachelor's degrees every year in the US and even more getting masters or doing bootcamps.

There aren't 60k people graduating with computer science bachelor's degrees every year in the US from any source I can find: https://datausa.io/profile/cip/110701/.

And the projections are for 300k, not 600: https://www.bls.gov/emp/tables/fastest-growing-occupations.h....

where are you getting your data from?

There are more job codes for software engineers than just the one listed on that fact table.

https://www.bls.gov/ooh/computer-and-information-technology/...

Additionally, you can see what degrees people are graduating with here https://nces.ed.gov/fastfacts/display.asp?id=37

So yes my numbers are correct if not lower than they should be for both.

Any data they aren’t? We have at least one, highly publicized case at Disney that suggests that’s exactly what happens. Otherwise, why fire experienced workers and force them to train people who didn’t know the job? That makes no sense unless cost was the motivator. It certainly wasn’t about higher skill levels.

Norm Matloff at UC Davis has blogged about this for a decade. A sampling of one paper has several references about the bias toward much younger and cheaper H1-B candidates in academia and professional workforce. He also argues diversity is better.

https://normsaysno.wordpress.com/2016/12/10/reforming-h-1b-a...

Thanks for the link. It looks like he wrote for U of M Journal of Law Reform on this topic: http://heather.cs.ucdavis.edu/Mich.pdf. Until I read through it and cross check his sources I can't judge the validity of course, but after reading over the first two pages his arguments seem compelling.

A link to a 2000 profile in Salon on him: https://web.archive.org/web/20150307053811/http://www.salon....

> Personally, I think the more diverse groups have an opportunity to come to America legally, the better.

Is this supported by data?

If it isn't, this is just thinly veiled parroting of propaganda.

I'm not against any color but blindly saying a star trek bar is better than a more homogeneous bar is just parroting

There’s a huge difference between stating my personal opinion and making factual claims. My opinion is that more diversity is good and makes the world better.

Also: > I’m not against any color but...

lol.

Personal opinion- I think homogeneous societies cooperate much better and I personally believe that so-called forced diversity of racial backgrounds is an effort to lower the camaraderie in a team and keep the loyalty to the firm and not the team members and thus bringing down wages by competition.

I am all for diversity of ideas and background. Something which would make a companies more productive at the same wage.

I also believe that no race is superior to another- everybody has different goals. And so long as one doens't try and interfere in another's society, the world would be a very smooth place.

Homogeneity is good only if you are moving towards the optimal goal. Diversity is good to bring in new ideas which helps adjust the goal and keep the society robust against failure. Like autocracy vs democracy. If you have a benevolent godly dictator, that's perfect. But we accept that no such ideal exists, so democracy is the compromise. Diversity in the workplace exerts the same goal. It lowers efficiency, but efficiency is pointless, even counter-productive if it is towards the wrong goal.

I don't see what you added to my comment. I already said diversity of ideas and background is good- diversity purely based on color to make a star trek movie is not.

Racial diversity is a proxy to background/idea diversity since different races come from different backgrounds and have different ideas of reality. Organizations are pushing all kinds of diversity, not just one based on color. Race just happens to be easy to measure. That doesn't make it wrong, it is somewhat in the right direction though it might not be optimal.

This is the most racist thing I have read today. Different races come from different backgrounds and have different ideas of reality? I think you meant different cultures which is a VERY different thing from race.

Either way, why use a proxy and not the real thing: idea diversity. Then what you propose is laziness at best.

Also, what do you mean by right direction? Diversity of ideas means diversity of goals right?

Reply on news.ycombinator.com