Hacker News new | ask | show | jobs
by davidw 3797 days ago
The real fix to this is to make sure that people brought over with an H1-B can easily change jobs. That way they're not beholden to some low-wage job, and companies that bring people in merely for low wages will face high rates of churn.
9 comments

In my opinion, the real fix is to make the H-1B 'lottery' into an auction. Instead of accepting 65,000 H-1Bs at random--accept the 65,000 H-1Bs with the highest wages. That way we are getting the immigrants with the highest valued skills and stopping companies like Cognizant and Infosys that game the immigration system by applying for the cheapest H-1Bs possible.
What is special about the number 65,000?

Absolutely nothing. It's central planning for something that doesn't require it: http://johnhcochrane.blogspot.com/2014/06/the-optimal-number...

That's a pretty ignorant article. The limiting factor for immigration is how quickly you can Americanize immigrants, not land area.

As for why it's imperative to Americanize immigrants. Look at how much trouble the U.S. has had cultivating democracy in places like Iraq. Most people acknowledge that the Iraqi people aren't ready for Democracy. Do you think they would be any more ready for democracy if you moved them en masse to Minnesota?

> Most people acknowledge that the Iraqi people aren't ready for Democracy.

This is a really contentious statement that you just drop in both as fact and as the basis of your entire argument.

No, I would not say that 'the Iraqi people aren't ready for Democracy' is a true statement. I also wouldn't say that what 'most people [in the US]' think about the Iraqi people's readiness for democracy [in Iraq] is really relevant to the question of whether they are ready for democracy [in Iraq]. I also wouldn't say that the question of whether they (as a group) are 'ready' for democracy in Iraq is relevant to answering the question of whether individual Iraqi people are 'ready' for democracy in Minnesota (whatever that means).

> Look at how much trouble the U.S. has had cultivating democracy in places like Iraq.

It also doesn't help that we have a long history of doing the exact opposite of 'cultivating democracy' in the Middle East (ie, going in and deposing democratically-elected leaders so that we can install dictators that are friendly to the US).

The US has had trouble cultivating democratically elected regimes that are friendly towards us in countries like Iraq. In the US, we tend to conflate 'democratic government' with 'government that shares our objectives and goals'.

It's an idea that was invoked by the American left to explain Bush's failure in Iraq, and eventually conceded by key figures on the right as well: http://www.businessinsider.com/donald-rumsfeld-admits-that-e.... So I'm not sure it's all that controversial.

And we're not talking about whether individual Iraqis in Minnesota are ready for democracy. We are talking about the link posted earlier in the thread, which suggested that the optimal number of immigrants in the U.S. would be two billion. That's not encouraging immigration of selected individuals, it's endorsing wholesale migration of huge populations.

> which suggested that the optimal number of immigrants in the U.S. would be two billion.

It suggested no such thing. Please reread it:

"What is the optimal number of imported tomatoes? Soviet central planners tried to figure things out this way. Americans shouldn’t. We should decide on the optimal terms on which tomatoes can be imported, and then let the market decide the number."

In 1945, most Americans would have believed that the Japanese people and the German people were not ready for democracy. Luckily the Truman administration and its allies were not as feckless and irresponsible as the Bush administration.
> how quickly you can Americanize immigrants, not land area.

Well that's an easily defined metric, isn't it...

There's nothing that says you must let immigrants become citizens right away, or give them all the same privileges citizens are entitled to. For instance, it wouldn't be at all unfair to kick people out for certain classes of crime.

> There's nothing that says you must let immigrants become citizens right away [...]

There's nothing that says you must let immigrants come.

Fixed that for you.

Actually, there is: economics. You can't have a successful country and exclude everyone who wasn't smart enough to be born there. In a world of N billion people, a lot of talented people are going to be born abroad. Keep them out, and they'll concentrate in other places.

Do you realize how much of the tech world was built be people from all over? HTTP, Linux, Google, Java, C++ and so on.

> There's nothing that says you must let immigrants become citizens right away, or give them all the same privileges citizens are entitled to.

You already have a process of green card + a bunch of time leading to eligibility. Isn't that good enough?

> For instance, it wouldn't be at all unfair to kick people out for certain classes of crime.

you mean prior to them getting citizenship right?

I'm not talking about privileges. I'm talking about getting immigrants to buy into the attitudes and values that make America worth living in. I'm talking about integrating immigrants into American society as neighbors so they're not living in their own neighborhoods where they can insulate themselves from the prevailing culture. There is a limit on how quickly you can do this.
> the attitudes and values that make America worth living in.

I'm going to go out on a limb and say that codifying exactly what those are and if someone is going to be able to accept them in X months/years is not going to be an easy task.

I am deeply saddened that this borderline racism passes for civil discourse these days. :(
I am amused that you think your feelings about perceived "racism" have any place in this civil discourse.
Funnily enough, his argument is that it doesn't mean much for the whole economy for there to be more immigrants. It sure as hell means a lot for individual communities! He seems to approach the economy as a race to the bottom / free market fiasco that has ended up concentrating all the profit in those who already own the land and capital.

It is good for a laborer when there is more available work than there is labor - they can command better rates and generally have a better quality of life (see: almost everyone in the tech industry right now). It is good for someone with lots of capital to have as much excess labor as possible (assuming there's still people that can afford to buy that person's products). When a person with capital can make cheaper products because the labor is cheap, it doesn't necessarily mean that all products become cheaper. You end up with the walmart economy where you have the people who own and manage in their mansions and everyone else unable to generate any meaningful wealth.

If you upgrade the H1-B's to green card after 12 months, that "65,000" number is irrelevant.

Once you remove the "indentured" part from H1-B's, we'll never hit the cap again.

Its super easy to game even your suggestion. Pay $100K, get $40k back in cash. your suggestion may end up making it worse for legit H1b employers, especially startups, that'll get suffocated with such high costs.

Ability to switch jobs at will, can bring up the paid salary and it puts the onus on employer to ensure employee is treat equal to other other employees in US. If employee is not treated/paid well enough, employee can move to another company in US.

Whether 65,000 is the right number or not isn't the important bit. While it may seem unpalatable on its face, granting the visa to the candidates with the highest paying offer is the simplest, least prone to abuse way I can think of to bring about the stated purpose of the H1B visa.
Except this leaves out any organization that pays below market rate for exceptional talent. That might sound like a good thing in the face of it, but consider such organizations include: national labs, early stage startups, software foundations, etc.

Also, unless you cap it in a per industry basis (which makes the whole thing very inflexible to changes in the economy) you just created an insurmountable problem for those firms that want to hire, say, a Catalan interpreter. Required skill uncommon in the U.S.? Check. Easy to get a foreign worker with that skill? Check. But now you need to pay them a software engineer salary or higher.

As a matter of fact, given how small the cap is, you could conceive that the only software engineering jobs that would be hiring internationally would be in high frequency trading and other such areas of the industry that pay higher salaries. Or, software engineering, but just in the bay area (other places pay less because of adjusting for the cost of living). No matter the rest of the considerations associated with why someone chooses a particular job.

Academia, publicly funded research labs, non-profits... are already cap-exempt.

Other than that, you're absolutely right. This would mostly benefit software companies in the Bay or NYC, and the ones that pay way above average at that. I'd argue it'd be better than current situation though. Indian employees are effectively tied to their employer due to the Green Card backlog. It's significantly less true for the H-1B holders hired by Facebook, Google, Amazon and the like.

Thanks for the info. I didn't know about cap-exemption for national labs. I knew academia didn't face an H1B problem, but tenure-track professors can go through EB-1 immigration anyways, so I always assumed that was the difference. But now that I check, apparently even if you are a lab technician or hired developer, you should be cap-exempt as long as you are working on publicly funded science of some sort.
<Academia, publicly funded research labs, non-profits... are already cap-exempt.>

And there are other skills-specific visa categories. For example, a professional hockey player on a North American roster from wherever is never denied a visa because of any other visa quota (H-1 variants, L variants, etc.)

Without a number, there is no scarcity, and if there's no scarcity, there's no bidding. So with a bidding system, you absolutely do need someone in power to pick a magical number.
Conversely, without a magic number, you don't need a bidding system. So what do you need?
Just why should the companies be able to bid on the labour? If there is a labour shortage in an industry, surely one can come up with a number of slots and then hand our green cards to the immigrants who can pick the employer that offers the best conditions.
Idea of curating immigration is practical in other countries (Canada, Australia), but somehow is not politically feasible in the US.

American electorate is somehow more placated by having a random number generator yield winners in a green card lottery.

Quite a clever idea. It might be more honest of the true intent of the visa to make it a reverse auction and accept the lowest wage bids.
As someone who went through the process, I can tell you that the mere existence of a maximum annual number was insulting.

If I get a job, and my company NEEDS me, and I'm GOOD at what I'm doing, I should get IN; period. There should be no crap involved. How the hell is it good for a company to be told that it can't grow because it happened to petition the 65,001st person out of 65,000 that year? How the hell is it good for the prospective employee to have to try again later? And in extremely-fast-moving industries such as tech, even the arbitrary time frame of "one year" is an insanely long time.

The only limit that kind of makes sense is a thorough investigation of the type of person you're bringing in (e.g. university degrees or other background, some indication of what they're bringing to the country as a whole). It may make sense to force companies to prove that no U.S. citizen can do the job but this system has been gamed for years, as companies produce vague job descriptions just like they post vague patent descriptions.

>Limits of any kind make no sense at all.

They are trying to protect US citizens from losing their jobs (or getting wages cut) due to a sudden availability of cheap labor. A government is rightfully concerned with making sure its own citizens are gainfully employed. Citizens vote, foreigners don't. Every person within those borders who does not have a job puts a drain on the rest of the country. Someone without a job outside of those borders does not. So bringing someone across those borders while an unemployed person is within them is a net economic negative.

Look at the biggest H-1B recipients: http://www.myvisajobs.com/Reports/2015-H1B-Visa-Sponsor.aspx It's all low quality outsourcing shops like Infosys and Cognizant.

These laws/protections are intended to prevent a race to the bottom. If the public subsidizes the operation of a company through security, education, and infrastructure, the community that makes that investment is entitled to ensure that the fruits of that investment go to other members of the community.

The thing is, "protection of US citizens" assumes a lot about what actually happens to companies, and it assumes a lot about what the citizens themselves are doing.

First, are you willing to pay $500 more for every product, and $10 more for every meal? Companies have to compete, and they are responding to what is necessary to survive. If a company employs 5,000 U.S. citizens and can't compete, it may stumble and lay off 1,000 U.S. citizens, or fail entirely and shed 5,000 U.S. jobs, all because it wasn't allowed to bring in a few immigrants to grow a little.

Entire companies (and successful companies, like Google) have been started by immigrants, creating potentially thousands of jobs for U.S. citizens. There is no reason to automatically fear immigrants; many of them are brilliant people.

A lack of a paid job does not make you a "drain" on society, either! What about children? What about volunteer work in communities? For that matter, I have met some astoundingly lax and irresponsible people over the years that have paying jobs, to the point where I almost thought of them as a net negative to the company.

>First, are you willing to pay $500 more for every product, and $10 more for every meal? Companies have to compete, and they are responding to what is necessary to survive. If a company employs 5,000 U.S. citizens and can't compete, it may stumble and lay off 1,000 U.S. citizens, or fail entirely and shed 5,000 U.S. jobs, all because it wasn't allowed to bring in a few immigrants to grow a little.

You're playing fast and loose with the facts. $500 for every product and $10 for every meal? That doesn't seem like an intellectually honest scenario. Where is that based in reality? Skilled immigration and non-skilled immigration are totally different. No one here is begrudging the immigrants picking fruits and doing farm labor. The issue is when companies lie/cheat/commit fraud to outsource jobs that Americans do want and are qualified/willing to do, all for the sake of driving down wages.

Not to mention most of those 'products' are manufactured in China already so there isn't going to be much price increasing there. Second, what a ridiculous scenario where a company has to choose between hiring 'a few' immigrants to save 'thousands' of US workers. Again, where are you drawing these examples from? Clearly not reality.

You didn't acknowledge what I said and again shifted the goal posts to another issue entirely. This is a pointless conversation if you can't even listen to the other side.

Throwing up walls isn't going to protect people from the US. It's a facile and ultimately unproductive response that plays to fear. The answer is to make US workers competitive in a global marketplace. And it absolutely is possible to compete on things other than price.

And while his numbers are invented, the concept is spot on: producing some things is expensive in the US, perhaps too much so to be competitive.

Nobody says the company can't grow; just that they need to hire folks legally able to work in the United States.

Ask an Indian entrepreneur how easy it is to hire an US national with origins in Pakistan.

There is no limit to growth there's just a limit to how little a developer will accept a job for. That is the sole limit H1B is trying to work on.

I assure you and anyone reading this, your skills are not unique nor is you talent unmatched. And even if they were, there is no way it could be discovered in our broken hiring processes. The chief attraction in these 65000 cases, so long as some vaguely plausible skill exist, is price I'm afraid.

I'm against foreign employees in all cases except when there is no available qualified individuals in the country of employment. It's ridiculous to screw over your own people for foreign help at a cheaper rate.
In a situation like this large direct employers usually take advantage of L-1 visa, which is reserved for moving existing employees between foreign and US-based offices. Outsourcing companies are at disadvantage with L-1, but so are smaller startups who don't have foreign offices.
Why is it better to let corporations decide who gets to live in the US and become a resident or citizen, rather than doing this by lottery?

I mean, in the absence of open borders, we have a limited number of spots available. Why should someone who wants to open a sandwich shop have lower priority than a programmer? Why should we skew things to accomodate Disney's desire to fire middle aged programmers? I think it's pretty clear we aren't talking about best and brightest here.

While I don't love the idea, I would concede that focusing on skilled immigration does make a certain amount of sense. But even then, I see no reason to give corporations the power to micromanage who gets in and who doesn't. They are, of course, free to hire any of the immigrants who come here - in fact, it would be very illegal, and rightly so, to discriminate against immigrants.

This.

My first experience with H1-B workers was working with contractors at a state agency (in Virginia). They were nice, but not highly skilled overall, and terrified of losing the job as they didn't have any guarantee of getting another U.S. job. This, in turn, made work life worse for everyone, as they'd never complain about anything on the job, even when it was egregious and not just software devs being whiny.

My second experience with H1-B workers was when I moved to Seattle. My colleagues were often H1-B, and were vastly skilled (still nice). Good workers (generalizing, of course), but perfectly willing to complain if the workplace or work was subpar, because they knew they could land another job in the U.S. more easily than our employer could replace any of us.

That said, the paperwork process was amazingly bad, and many of them got stuck in jobs (even avoiding promotions) because they had started the citizenship process, and for at least one phase changing your job starts the process over. Ugh.

The difference between the two, and the impacts it had on the workplaces were stark. My leaving the first job was in the middle of most of their non-H1-B talent leaving (you could call that a self-correction, but that just left those that stayed worse off). My second experience left me rethinking everything I thought I learned from the first experience.

I concur. As someone who is here in US on this visa (fortunately, my workplace presence resembles your seattle experience), I can totally understand what you are pointing out.

Something that'll genuinely solve all the problems (qualified americans losing jobs, outsourcing companies abusing loopholes, salaries racing to the bottom), would be to have a concept of temporary green card.

Instead of a 3 year H1b, have a 3 year green card (lets call this the Super H1b). Once an immigrant is here in the US on a Super H1b visa, he/she will be free to take up any employment if need be, even with multiple employers.

This will ensure that an employer sponsoring the Super H1b visa will be paying at or above market rate salary, because the employer truly needs this employee.

Are there chances of this visa being abused? sure. But its way low, much more beneficial (to americans, to the immigrant and also to the genuine Super H1b sponsors).

@ergothus : Do you think having such a visa would have helped (considering your experiences in Virginia and Seattle?)

I'm hardly qualified to address the nuances of the immigration system: were it up to me, anyone willing to work and pay taxes would be welcome here (I'm particularly baffled that the U.S. will further the education and/or experience of people, then make it hard for them to stay), so it that sense your proposed visa is still too much paperwork, too many unnecessary rules.

That said, it certainly sounds like an improvement over the status quo: slightly better quality of life for the Seattle-like folks, and real options to reduce employers taking advantage of desperate employees.

No, the real fix is to make sure the actual goals of the program align with what actually happens. At this point, saying that each H1-B worker must make $117,700[1] or more to prove they are skilled beyond US workers seems to be a fair change. After all, if you can hire US citizens cheaper, you are not fulfilling the goals of the program[2].

1) 10x poverty guidelines in US for 2015.

2) http://www.dol.gov/whd/immigration/h1b.htm

> At this point, saying that each H1-B worker must make $117,700[1] or more to prove they are skilled beyond US workers seems to be a fair change

That's not enough for some places. I know mid level engineers at some of the big companies make more than that, for example.

I believe so many companies don't hire people for culture fit but then hire an h1b worker. If they had to pay market rate or better like you say it would be great for those of us who can't seem to make it past the interviews even when qualified.
it is not fair. Its not fair for genuine h1b sponsoring companies in geographic locations where cost of living is low.

$117,700 in San Francisco is not the same as $117,700 in Austin.

Again, any minimum salary requirement, becomes very easy to game (employee can be expected to pay back in cash, for the difference). This will put genuine companies (that cannot do such shady deals) at a great disadvantage

Perhaps I should have put the full poverty level tables which are calculated at https://aspe.hhs.gov but 10x the local poverty level seems more than fair.

> Again, any minimum salary requirement, becomes very easy to game (employee can be expected to pay back in cash, for the difference). This will put genuine companies (that cannot do such shady deals) at a great disadvantage

Paybacks are illegal now and should stay that way. A couple of criminal prosecutions would do wonders for fraud.

Kickbacks certainly are illegal (and it should remain that way).

But, just because it is illegal, doesn't mean it doesn't happen.

Robberies happen but we don't stop trying to put criminals away or give up and make it legal to rob people.
> make it legal to rob people.

Civil forfeiture comes scarily close.

The real fix to this is to make sure that people brought over with an H1-B can easily change jobs

When I was on an H1B visa I changed jobs multiple times, very easily. Part of the problem is that people (including potential employers) aren't aware how easy it is.

I think he meant, change jobs without needing to file for an H1b transfer or any documentation. Once a high skilled immigrant is in US, instead of giving work permit to a specific employer, make it so that its a work permit, with a time limit.

3 years, immigrant can work anywhere, any number of employers.

Guess how many illicit h1b employers will be interested in bringing an immigrant, if they are not guaranteed that the immigrant will stay with them?

The transfer documentation is not that onerous, to be honest. Allegedly it's a gatekeeper against H1 program turning into a spout for unqualified immigration where "an employer" hires you for extremely high salary, and you quit on day 2 to start an exciting career in dishwashing.
The simplest way is to convert to a green card after being employed for 12 months. Employer is responsible for all legal fees, background check fees, etc. over the course of your employment.

If you quit at day 2, the employer is still on the hook for all the fees.

It can be gamed, but if you make the fees something like 30-40K, genuine employers will simply shrug as they'll amortize that over 3-4 years. Sweatshops, however, will lose bad.

My understanding is that current employment-based green card process is constrained entirely by slow turnaround by Department of Labor and USCIS, not some company shenanigans or legislative provisions for artificial delays, at least according to http://www.uscis.gov/sites/default/files/USCIS/Resources/E2e... And that slow turnaround is predicated by DoL / USCIS budgets and their ability to hire and train proper employees.
You are 100% correct. HOWEVER, you don't see the companies lobbying to increase the funding to DoL/USCIS.

The reason is that the choke point benefits these companies. If you cleared the backlog so that a green card was a 12 month process, the companies would quit asking for H1-B's unless they really wanted them.

not true. these green cards are given (in addition to employment based GC requirements) based on the country of birth (to maintain diversity).

If US gives X green cards to nationals of Y countries (X>>Y) in a year, each country is allocated X/Y number of green cards.

This puts Indians, Chinese and other high population countries at a disadvantage.

An Indian applying for green card in EB-3 category (minimum qualifications being undergrad degree + 3 yrs experience i think) has to wait for ~12 years from the date his gc process was started. EB-2 (advanced degree and/or 5+ yrs experience) is ~10 years (these are my ballpark numbers)

Its not the turnaround time, but the concept of diversity based green cards that slows down the process.

I think the problem may be with those who intend to apply for permanent residency while legally present on an H1B.

Once you apply for a green card, you have to maintain legal immigration status continuously. Furthermore, nationals of India and China have to wait longer, due to country quotas and relative demand. There is something called a priority date, which is somewhat akin to taking a number instead of waiting on queue, except it's really more like a number for the secondary queue to get a number for the real queue.

There is an I-140 form that is part of the process, which is filed by the employer. If this form is withdrawn, perhaps in response to changing jobs, it can reset priority dates or cause extensions to be denied.

i-140 can transferred between employers and you retain your PD. Typically, sponsoring companies do not revoke i-140.
The previous employer can choose to withdraw the I-140 petition, and typically does so because keeping I-140 petitions around for ex-employees can create problems for the I-140 petitions of current and potential new employees.
Totally agree with this! This will make the employers think twice before sponsoring H1b, and will ensure that they treat (in terms of pay and work load) their immigrant employees equal (to americans).

Until this happens, Americans are going to be at the receiving end.

Even limiting number (drawn out of thin air) of visas, or increasing the cost of the visa, doesn't solve any issues. If anything, it'll make it worse for the legit h1b sponsoring company to bear the costs and handle the immigration process.

Absolutely! They can be abused to work long hours because they have their visa hanging over their head. Let labor be free to change jobs and suddenly the economic incentives are greatly reduced.

Who wouldn't want to hire a bunch of people that are beholden to you to stay in the country? Remove that from the equation and wages will move back up to market.

To me the real fix is to abolish the entire system. It was a half baked program cooked up by industry lobbyists that invented and perpetuated the myth of a "STEM" shortage. There is no shortage and the character and education system of the US produces plenty of well trained and proven trainable people.
Another one is denying the visas to a company which is obviously replacing people 1:1. So they can't argue that they cannot find qualifying candidates just because they just so happen to have fired them to make them unqualified.

This isn't a cold trail of off shore shell accounts and obfuscating financials dealings. The trail is pretty clear.

Easier than current H1 transfers? You're introducing a loophole where unqualified labor will be hired (and let go on day 2) to switch to even lower-wage jobs.
Auction + transfer tends to guard against this. Don't shell out for labour which walks.