First, the authors say: "In this paper, we simply sought to dispel the myth that globalization generates no losers." In my experience, If you start out seeking to confirm your biases, you generally will.
Second, all their evidence is correlation not causation. I'm sure the correlations are strong, but there is no way to prove that H-1B workers actually caused wages to go down. There are too many external factors.
Finally, unless there is some relation shown between the companies' profitability and the wages they're paying, you can't rule out the possibility that the less expensive H-1B workers and lower wages for American workers didn't actually save the American worker's jobs by keeping costs down and their employers in business. Lower wages and more H-1B workers could just mean that the company is doing what it has to do to stay afloat. Less pay is better than no job at all. That's just one of many possible scenarios.
The paper isn't out yet, but as of now, their argument seems pretty weak.
If you're worried about San Fran being the next Detroit, start looking at that fact that you and other hackers depend on a specific region to sustain their salaries and arguably inflated sense of self worth. Beyond physical networking, where you are doesn't matter like it used to.
I know that's not what you meant, but it's the same syndrome that contributed to the downfall of Detroit auto workers. They all said "But this is Detriot, the home of the Big 3! United States Auto Town, we can't fail" Meanwhile the rest of the world was busy building cars cheaper and better everywhere else while the folks in Detroit tried to protect their salaries without improving the product.
Find a place where your rent is cheap, your staff is cheap and you can build something fantastic (and then pay everyone more) without worrying about what the next 'hot' place to live or have an office is.
I suggest you read up something about the raise and fall of US automotive industry: Detroit used to be no less "startupish" than SV is today: thousands of engineering startups have flourished and vanished over relatively short period of time. The history of Detroit is one of several reasons why I don't take Paul's views on special role of SV seriously: Valley isn't something we haven't seen before. If anything, it is declining:
Valley startups no longer represent the cutting edge of computing innovation at large [as they did in the 80s], they have mostly escaped competition by hiding in relatively narrow niche of "English-speaking consumer Internet" very much like Detroit has been hiding behind big SUVs. And "release early, release often" web sites are no different from body-on-frame SUVs: their tech is so well understood and simple that Chinese can produce cheaper and faster copycats of facebooks and twitters in a matter of months within an original launch.
When they'll learn to market them properly, they will build their own data centers in San Antonio, TX and the similarity with Detroit will be complete.
Detroit became entirely dependent on a centralized system powered by a few massive companies. Silicon Valley has some of that, but for the most part is highly distributed. The thousands of startups that feed off the big auto companies can't exist without them. Silicon Valley doesn't die if Google and Yahoo fail to innovate.
That English speaking "niche" currently represents the vast majority the potential revenue. It's a lot harder to innovate than it is to internationalize your software, which virtually every successful company does.
The idea that the value of Twitter or Facebook is in the 5,000 lines of Ruby or PHP that it took to create them is seriously mistaken. Anyone (Chinese or not) can easily clone Twiter or Facebook. Just like almost any competent artist can clone a great painting. That's not the hard part.
Yeah? How about a 30% cut, all your benefits, and a clause in your employment contract where you lose your apartment (wherever it may be) when you quit?
That 6% number --- which is misleading on its face --- also ignores the fact that many H1Bs are FTEs for body shop firms that lease them to big companies.
What a fatuous comment. Automobiles didn't themselves lock foreign workers into abusive contracts. H1-B worker visas aren't an evolution of the market; visa and foreign labor terms and conditions are as old as foreign relations.
that's a reason why abusive contracts are bad, not a reason why allowing foreign workers into the united states is bad.
we should allow more foreign workers into the united states for the same reason we should let automobiles overtake horse and buggy - it's more efficient.
You're arguing with a straw man. Nobody is arguing that foreign workers shouldn't be allowed into the country.
It's a market inefficiency for companies to be allowed to "capture" employees and artificially suppress their wages by keeping them off the open employment market under threat of deportation.
I don't understand your assertion -- the deportation thing is BS, but working conditions seem to be favorable enough that people are willing to leave their families behind and work in the USA.
I don't really want to argue about immigration on here. Suffice it to say I welcome any skilled people who want to contribute to the GDP and body of technical expertise in my country.
The point I was hoping to make is that even if a glut of talent depresses salaries it's worth it in order to stay globally competetive.
It's impossible to extract the deportation thing from the debate. If immigrants were working on a level playing field with natives, this would just be about xenophobia and protectionism. But they aren't. They work under unfair terms that depress their wages and, by extension, the wages of the entire field.
It's not simply that a Bulgarian or Thai developer will do the same job for less money; it's that the system is rigged so that they essentially have to work for less money.
If that were the equation, I'd be on board with that point of view. The problem is that the H1B is increasingly used to outsource jobs rather than bring talented engineers to the US. Seven of the top 10 users of the H1B visa are India-based outsourcing companies, and even many of the US-based users are really there to facilitate outsourcing. These companies use the visa to cycle someone through the job in the US for a few years, do "knowledge transfer" where an American is allowed another 6 months of employment and a severance under the condition that they train their replacement, and then move the job and worker overseas.
SO in this case, you'll get your 6% pay cut and the outsourcing of an industry - all under the guise of making sure that talented foreign engineers are allowed to come to the US and create wealth (which is an indisputably wonderful thing for the US and would be possible without all this abuse if the system were reformed).
Wow, I had the exact same thought (ie. comparing hackers to Detroit auto workers) when I read the article. Though, I thought I was being profoundly pessimistic.
So, to add more pessimism (for fun)... we don't have the bargaining power that the UAW had, and I don't think we have any members of congress that will consider software professionals to be a serious constituency. So, without legislation, we would definitely spiral down faster as an industry.
Even if the H1-B visa program did have some negative result on the average programmers' income, it doesn't affect any of us. The average programmers' income is only relevant to the mediocre among us. Just be the best so you don't have to worry about averages.
I don't care where my competitors were born; I want to get job offers and win contracts based on my merits, not based on my mother's geography. If there's a guy in Thailand or India that is better for a position than I am, then he deserves the job and I want him to have it.
Literally everybody I know that has a H1-B visa is making this country better. If I had to make a prioritized list of people to throw out of the country, most of the people on the list would be American citizens. If I got to choose 10 things to change about this country to increase my net income, the H1-B visa program wouldn't even be a consideration--it doesn't make sense to complain about an speculation of a 6% cut due to the H1-B program when the government is taking 6% (sales tax) + 6% (state income tax) + 40% (federal income tax) + 2% (property tax) and when we are paying a huge percentage of our income to insurers who turn around and rip us off when we make claims.
If you really want to know salaries, take a look at the H1-b job databases (they are required to report salaries for specific job titles). I don't look because I don't want to know (plausible deniability). I'm warning you - you don't want to know what some people earn (when I looked X years ago, my eyes almost burned out)
If you really want to increase software engineer wages, put a legal cap on the number of CS degrees that can be handed out every year, like med school! I can't say what this will do for the health of the US software industry, but I'm sure it will drive wages up.
Would be nice to see the actual paper, and understand how they went from correlating resume's from "a leading online job search site" to employers and the wages they pay their H1-B employees.
I'm assuming that someone wants us to draw the conclusion that this is a bad thing.
But I don't see why our salaries should be kept at a high level, at the expense of both employers (and thus businesses and consumers) and other humans.
Also, it seems to ignore the demographic problems that society is starting to see. Without a broader base of employed people, we can't keep paying retired people their Social Security and Medicare. An influx of employment-age people is necessary to keep these systems afloat. So in the long run, having these Visa workers will help ensure that I get out of those systems at least some of what I'm putting in.
Interesting point. But, I would guess the resulting wage deflation of opening up our markets completely (increasing labor supply) would outpace the deflation of medical expenses, and I doubt the latter would ever catch up completely. Therefore, those expensive government programs would have to be funded with much less tax revenue due to lower wages. I don't think that would fix the problem at all.
Isn't this the site where entrepreneurs go, those people that compete on being better. Not the site where salary-slaves go to whine about having to defend their salaries?
It's simple supply and demand. H-1B increases the supply of programmers, which means that the price (wages) paid to them will decrease.
As for the whole "dispels the myth that globalization generates no losers," thing -- that's not really how globalization works. Labor and capital mobility, along with low impediments to trade allows for specialization, which may hurt specific industries like manufacturing in the developed world but helps others like services.
The extra programmers that were hired because they were 6% cheaper brought a lot into the US economy, and they brought a lot more to their home countries in forms of remittances. For some of these developing countries remittances are 25%+ of GDP, meaning that they are completely dependent upon them.
And of course, there is no discussion of methodology nor data.
Not that simple. The fact of the matter is that the H-1B is linked to only a single company and cannot be transferred, meaning that any foreigner wanting a legitimate shot at citizenship cannot afford to leave their job for up to a decade or more (until green card).
IMHO that is why the wages are lower - would you give your employee a raise if you knew that he had absolutely no way to quit your lousy job for something better?
"The fact of the matter is that the H-1B is linked to only a single company and cannot be transferred"
Incorrect. You can transfer your H1-B quite easily. However you can't transfer an L1-B (intra-company transfer from a foreign country). But even in that case, you can apply for an H1-B while you are in US and then change jobs once you have received your H1-B.
Also, all the H1-B holders I know get very good salary - often higher than 'the market'. When you leave your home, get a visa and come to work half-a-world away, you are not going to dick around. You are going to do everything you can to maximize what you came here for.
H1Bs live in two parallel universes. People in one never meet the people in the other.
The H1Bs who work at Google probably get paid a pretty decent wage (though even then, the restrictions on how and where they can work, as well as the difficult in changing professions, probably does still drag down wages relative to what they would earn if they were totally free agents).
The H1Bs at the body shops are often paid miserable wages, and these visas are used without question to fire Americans and replace them with foreign nationals. Keep in mind that 7 of the top 10 users of the H1B visa are actually Indian Outsourcing companies.
Here's a link to the story about the usage of H1B (the link is msnbc, but the original story was on businessweek)
http://www.msnbc.msn.com/id/17048048/
"infosys was certified by the department of labor meeting the prevailing wage requirement to bring 100 computer programmers in at $9.15 an hour. HCL America got 75 programmers at $24,000 a year."
The real debate is how prevalent the two camps are - Microsoft and Google insist that the H1Bs are exceptional engineers who couldn't be found "at any price" in the US, and who create huge amounts of wealth. They would tend to claim that the abuses are minor, the kind of thing that happens in any complex system.
Opponents, on the other hand, tend to say that the abuses are widespread, and that very talented engineers are a small minority of the H1B workers.
For the record - I believe that talented engineers from overseas are exceptionally valuable to the US and should be allowed in without all the indentured crap. However, I believe that recent evidence demonstrates that the abuses are extremely widespread, and that the H1B has greatly reduced the appeal of engineering careers at a time when the US desperately needs to ensure a steady supply from within its own population.
My first real job (I'm a native-born US citizen) was in such a body shop. Everybody there got paid chicken feed.
In fact, the INS would check that the green card applicants were getting paid market wages. When an employee reached that point in the process, his salary was actually increased as appropriate. But for those of us that had no such immigration issues, we never got such an upward adjustment.
(Things may have changed since then. That was the early '90s, the time of a big downturn in the industry)
I got paid a little extra to do technical interviews for this purpose. A green card applicant had to prove that there were no available US citizens that could do the job, so we'd run an ad in the Times, soliciting applicants for that job. I would then give them technical interviews so difficult that they couldn't pass, thus proving that our employee was the only one qualified.
I know an H1B that switched jobs 3 times in the US. The rule is you need employment within X days of the loss of your job but you are NOT tied to a specific company.
First, the authors say: "In this paper, we simply sought to dispel the myth that globalization generates no losers." In my experience, If you start out seeking to confirm your biases, you generally will.
Second, all their evidence is correlation not causation. I'm sure the correlations are strong, but there is no way to prove that H-1B workers actually caused wages to go down. There are too many external factors.
Finally, unless there is some relation shown between the companies' profitability and the wages they're paying, you can't rule out the possibility that the less expensive H-1B workers and lower wages for American workers didn't actually save the American worker's jobs by keeping costs down and their employers in business. Lower wages and more H-1B workers could just mean that the company is doing what it has to do to stay afloat. Less pay is better than no job at all. That's just one of many possible scenarios.
The paper isn't out yet, but as of now, their argument seems pretty weak.