Hacker News Reader

Actually, I think the real problem is that H1Bs are given out via a queue rather than an auction.

With a queue, there is no incentive to note apply for H1B's. Do the paperwork, you get someone to fill some position (often cheaper than a recruiter). With an auction, H1Bs will be given out to the employers who value the employee most highly rather than simply a random selection of whoever got their application in by Jan 1.

As for job mobility, I think requiring the new company to pay off the pro-rated auction cost to the old company is fair (i.e., 1.5 years into a 3 year visa, new company pays old company for half the visa cost).

I like the auction idea in principle.

However the devil is in the unintended consequences of the details. If you transfer the auction cost around then you've effectively made an H-1B into a regular employee, part of whose salary gets paid to the government. This encourages underpaying H-1Bs relative to local competition, and as long as there is a significant market of low paid skilled people you're putting large downward pressures on American salaries.

The version I suggested gives incentives to bring in good people, but also provides market incentives to pay competitive salaries to anyone who is brought in.

> However the devil is in the unintended consequences of the details. If you transfer the auction cost around then you've effectively made an H-1B into a regular employee, part of whose salary gets paid to the government. This encourages underpaying H-1Bs relative to local competition,

Nope. You forgot the money being paid to the govt. That increases the costs to the company, which is the number that they care about.

If the auction price is reasonably high, that pretty much guarantees that H1Bs will not be used for low-wage occupations.

Suppose that the auction price is $100k. If the going rate for a particular job is $120k, no company is going to hire an H1B for more than $20k. Even if said company finds someone for $15k, the cost to the company is only $5k below market. If they find someone for $30k, that's like paying a not-H1B $130k.

>> However the devil is in the unintended consequences of the details. If you transfer the auction cost around then you've effectively made an H-1B into a regular employee, part of whose salary gets paid to the government. This encourages underpaying H-1Bs relative to local competition,

> Nope. You forgot the money being paid to the govt. That increases the costs to the company, which is the number that they care about.

Read more carefully. I did not forget the money being paid to the government. Instead I was describing a net effect where employers would keep their internal cost the same and drop the salary offering by the amount of the fee. The net result is then exactly that part of what would have been the H1B's salary goes to the government instead.

By contrast if the fee is only owed by the original employer then any subsequent employer can offer the same salary that they would for a local and have the same total cost for the employee. Market pressures will push them to do that. This will push the original employer to pay prevailing market rates or better to keep the employee around long enough to recoup their initial investment.

> Read more carefully. I did not forget the money being paid to the government. Instead I was describing a net effect

and ignoring the H1B holder, which is why I provided an example.

Very few H1B holders who have skills that a US company values at $120k will take a job for $20k without some angle.

You suggest one such angle, namely that such employees go to another employer who doesn't pay the fee to govt, and suggest that original employers will therefore pay more to avoid losing the benefit from paying the fee.

Let's extend that example. Suppose that it takes $100k/year to keep an H1B holder from jumping ship. Add that to the $100k fee and that employee costs the original employer $200k. Since the agreed-on market rate for that job was $120k, the original employer isn't going to pay $100k year. Instead, they'll avoid the risk of losing the $100k fee and hire locally.

That's why the fee will discourage H1Bs for low-wage jobs.

If the fee is paid by all employers (and presumably refunded when an H1B jumps ship), there's no poaching bonus so every employer is essentially the first employer.

First a pretty obvious point. The point of the H1B visa is to make it easier to import specialty workers whose skills are in high demand. By definition those jobs are going to be high wage jobs. Therefore not covering low wage jobs is a benefit - it ensures that the visa is only covering the kind of people it is supposed to cover.

Moving on, where did you ever get a $100K figure from? I haven't even really thought through what the bond should be. I kind of like the idea of setting a cap on the numbers then letting the figure be set in an auction.

But that said your extension makes a pretty big mistake. You're adding a $100K one time fee to the $100K/year salary and are coming out to $200K. But you're adding apples and oranges. If the employer only expects to keep the H1B for 1 year, your math is right. But suppose the employer reasonably expects to keep the H1B for 4 years. Now we're at $125K, which is only $5K over the market rate.

Now an obvious question is how long the employer can expect to keep an employee for. There are two possible attitudes here. One is that turnover is out of your hands. The other is that it is under your control. I submit that the proposal is going to appeal more to companies that think the second and believe they can keep people happy. I happen to think that this is a good thing because I prefer companies with that trait and like rewarding them.

Furthermore I believe that the proposal is going to be more appealing for companies with candidates that they reasonably believe are better than the local applicant pool. Again I believe that this is a good thing since that is exactly the kind of person and situation that this program is theoretically for. Except that today we verify this based on the testimony of the hiring company, whereas my proposal makes it in the best interest of the company to act this way.

> Moving on, where did you ever get a $100K figure from?

It came from the example.... As to whether it's reasonable, it came from an auction, which is the premise of the whole discussion. The number of H1B slots to be auctioned is a decision. We could set that number low enough such that only those companies that are willing to pay a lot to hire H1Bs will be able to.

Yes, the employee may stay longer than a year. Or not.

Your arithmetic doesn't make sense. If the employee stays 4 years, that's $25k/year. It's a $120k job, so the employer won't be willing to pay more than $95k. Now we're getting into the range where someone might be willing to take it, but there's a lot of employer risk in that number.

>> Moving on, where did you ever get a $100K figure from?

> It came from the example.... As to whether it's reasonable, it came from an auction, which is the premise of the whole discussion.

Let me review your logic here. You got $100K from your example based on your theoretical auction and then you're arguing that $100K is too much and no company in their right mind would pay that much. From that you conclude that my proposal is bad.

Well if no company in their right mind would pay that much, then they wouldn't bid up the auction that much. Therefore what you've just argued is that your hypothetical auction came up with an unrealistically high number. At which point you're arguing against a straw man that you set up rather than my actual proposal.

> Let me review your logic here. You got $100K from your example based on your theoretical auction and then you're arguing that $100K is too much and no company in their right mind would pay that much.

Not at all. I showed that the auction would raise the threshold at which companies would choose H1Bs. The amount clearly depends on what the auction price turns out to be.

And, yes, we could restrict the number of H1Bs being auctioned so the price ends up around $100k. Or, we could set the number so the price is $100.

> From that you conclude that my proposal is bad.

Not at all. I merely accurately explained how things would work.

I disagree with a lot of what you said. First, the system is already "market-based solution": getting a visa for an employee isn't a rocket science. There's competition for good H1b talent. And H1b by itself isn't that strong of a chain that holds them hostage. Changing employers while on H1b is not terribly hard.

Moreover, in my recruiting experience, you (as a company) have to compete pretty hard to get the best foreign CS grads - they know they're good and the fact that they need a visa after graduation doesn't make them less valuable.

The shady part of H1b business is a bunch of "body shops" that bring in sub-par talent here (with poor english or/and poor technical skills) and abuse them until those workers improve their marketability on US job market and leave.

In short, I see nothing wrong with the H1b system except for general slowness of the INS bureaucracy and poor enforcement of H1b abuse (just like with most immigrant-related laws, enforcement is a problem).

Full disclaimer: I myself was employed on H1b by a 1-st tier US tech firm and saw absolutely no discrimination in salary or any other matter.

You are absolutely right that companies have the opportunity to bring in the best in the world as well as the opportunity to bring in people whose expectations are lower. We agree that the problem arises with the ones who go after people whose expectations are lower. My further opinion, which you might not share, is that those abuses are both widespread and problematic. And therefore I was focusing on that part of the issue. But you're right that I shouldn't tar all employers who hire H-1B people with the same brush.

The solution I proposed is intended to address this by making resolving abuse cases resolvable by market pressures rather than government bureaucracy. Where the program works I absolutely want to see it continue. This country was built on immigrants, and I firmly believe that its ability to attract top talent worldwide is its greatest asset.

The auction idea is good, but there should also be some assurance companies don't get H1Bs for significantly less than the rate they would pay locals. H1B visas should never be used as cost-cutting mechanism. Perhaps a good one would be to fix the compensation of the H1B worker to the average local compensation for the same or comparable position.

OTOH, that would probably drive some companies out of the US because all their competition is doing the same. One must measure the possible consequences carefully.

The assurance in my proposal lies in the fact that if you don't pay market wages then someone else will hire the H1B away from you.

It will only work if the employee is desirable for other companies. It would not prevent H1B from being used as a cost-cutting measure - it would only prevent it from being very effective at that.

If the person signing the check can't (or doesn't have to) tell the difference between a good and a bad professional, 5K saved are 5K saved.

A much better way are the points based systems found in Australia, Canada, and the UK. The details vary but it works like this: the government creates a system where each candidate is scored against. The final score needs to be higher than a threshold for the candidate to be granted a work permit and residency visa.

The government decides what to score based on its idea of what's a useful addition to the economy. In the UK these include the highest degree earned, the last 12 months' earnings (weighted by country of earnings), English proficiency, etc.

It also can add bonus points to professions it wants to attract - say nurses or doctors. The UK used to give automatic full points to an MBA holder under the HSMP visa scheme! PhDs didn't get this red carpet.

> A much better way are the points based systems found in Australia, Canada, and the UK.

The US has a points-based immigration classification. It's pretty much a disaster.

I have no idea how well UK et all run their points based schemes, but it doesn't actually matter. The relevant question when proposing one for the US is how well the US would run one. Maybe this time would be different but....

A bond, hmmm, well if the bond automatically went into the bank account of the local immigration attorney, this idea might stand a chance.

My friend on H1B makes $80/hr, so not all of them are "cheap".

No, not all are cheap. The more interesting question is what a comparable skillset in an American would cost. For a random web developer, $80/hr seems high. It would be low for an experienced DBA.

Furthermore the fact that companies have incentives to abuse the system doesn't mean that companies do abuse the system. Your friend may be a fairly treated example. But his existence doesn't change the fact that many cases exist where the system is abused.

True but most foreigners view the H1B as the first step to getting a green card - something highly highly desirable among the scores of white collar foreigners working in the US.

From my experiences, I know quite a few who after getting their H1B feel "locked" down with the company that hired them. This in turn allows the company to abuse the shit out of them because they know that the only option for most of these people, especially in the current economic climate, is to stfu and keep on taking the abuse lest they want to return back to sweatshop call centers with 1/10th the pay.

Then these guys get further screwed when it comes to the annual review & promotion process when their american counterparts (i.e. white males w/o fobby accents) get bumped ahead of them.

A blanket statement for sure; I realize there are more nuances to this but Ive seen it happen ALL-THE-TIME.

Plenty of them are. Here's a datapoint I can contribute:

Lucent in 2001 (yes, a very bad time). I'm 40 (although I look like I'm in my early 20s), making $80K.

My best coworker is a brilliant native of a Caribbean island; he got a EE from a respectable US university and was getting his Masters at night (showed me his MOSIS chip at one point).

For the job we're doing, in his own areas he's as competent and productive as I am. Immigration law requires the company to post his salary; he's making $45K.

If he was just getting paid what I was at my second job, the first one at a market rate, he'd be getting $63K (that's adjusted for inflation, but you should subtract some for higher health care costs (they've risen faster than CPI inflation)) and he'd be worth every penny.

He knew he was being exploited and wasn't happy about it, but it was useful experience and a good path to a Green Card, at least when he signed on (Lucent of course continued its decline and canceled the project we were working on, which was important enough that it canceled their future as well).

The near indentured servitude aspect of the H1-B system is reason enough to seriously reform it ... but as noted, the necessary docility of those who have one is too attractive to too many companies.