David R. Henderson  

Immigration: Anu Bradford's Creative Suggestion

Naik on Bastards and Stigma... What is a 'fiscal union'?...

One of the biggest issues separating those, like Bryan Caplan and me, who want to allow much more immigration and those who want to limit it to its current level or lower is the issue of the welfare state. Even Milton Friedman, as various commenters have noted over the years, had doubts about legal immigration because of his concern about the United States being a welfare magnet.

Co-blogger Bryan has been a strong advocate of betting, often quoting Alex Tabarrok's claim that "betting is a tax on bulls**t." Is there a way to use betting to resolve the welfare magnet issue? Yes, there is. Although she doesn't quite state it as a bet, Anu Bradford, a professor of international economic law at Columbia Law School, proposes such a bet. She calls her proposal "reversible bonds," and gives the following example and explanation:

For example, if an American company wanted to hire a Colombian worker, it would post a bond, the amount of which the US and Colombia had set. If the worker later became unemployed, the bond would be released to the US to cover welfare benefits. Likewise, if the worker committed a crime, the bond would cover deportation costs. If, however, the migrant remained an employed and law-abiding resident for a specified period, the bond would be divided between the company that had posted it (as a reward for having screened a productive foreign worker) and the Colombian government (as compensation for productivity loss and the costs of educating and training the worker).

Finally, if the migrant returned to Colombia after a specified period of working productively in the US, neither country would need compensation. The US and the employer would have benefited from the employee's labor and tax revenue, while Colombia would benefit from the worker's return, presumably with more money and enhanced skills. In this case, the bond would be released to the worker, creating an incentive for return migration and cultivating "brain circulation" instead of brain drain.

Being a libertarian and, frankly, being in the mainstream on the issue of freedom of emigration, I would change her proposal in an important way. Sorry, Colombian government; if someone leaves your country, you get nothing. In "Liar, Liar," a lawyer played by Jim Carrey gave blunt advice to his thieving client, "Stop breaking the law, a**hole." I think readers can see the connection. You don't want people to get "free" education and then leave your country? Then don't pay for their education.

I would also change it in another important way: there's no reason why the entity posting the bond couldn't have been the immigrant. If the U.S. government had given this option to me in 1977, when I was immigrating, and required a bond of, say, $20,000--that's $77,000 today--I would have said "Done." Although the number is inherently arbitrary, one way to generate the "right" amount is to take the present value of average welfare payments for an 8-year period and then have the bond returned in 8 years. I think the U.S. government would have millions of takers--or, more accurately, payers.

Comments and Sharing

COMMENTS (16 to date)
nl7 writes:

So this is basically indentured servitude, right? The worker gets a windfall at the end of service, forfeits beforehand if he fails to complete the service, and the employer must make a large up front commitment draw him there.

Of course, in that historical situation it was necessary to entice poor people to say farewell to their homes and families forever and come to a possibly dangerous new world. The bounty (typically a land grant) was incentive to overcome the risk, sort of akin to today paying oil rig workers or war-zone Arabic translators exorbitant amounts to work in hazardous or uncomfortable conditions.

But immigrants over the last couple centuries have wanted desperately to come here, so desperately that billions of dollars of enforcement did not stop them. So the idea of a bounty makes little sense to me as an attractor. They will come without it, so the compliance burden is only on US employers.

I also think it's very problematic because the cost of the bond is likely to be fixed regardless of the migrant's productivity, meaning it is disproportionately huge for low productivity migrants. Yet it's unskilled migrants who (before the recession) were the most controversial and the hardest to police. So they are the least likely to be benefited by this policy and the black market continues.

This is an interesting idea, but I think it's more a philosophical gesture than a real policy suggestion.

RPLong writes:

Your last point is excellent, and it makes me wonder to what extent this sort of enterprise could be used to reduce the national public debt.

Clay writes:

Stats say that assimilated descendants of certain groups of immigrants have high unemployment/crime/dysfunction even when the immigrants themselves have low unemployment/crime/dysfunction. Such a bond system wouldn't address that.

Tom in Miami writes:

Common David. When you were trying to immigrate, before you took your first American job at Univ of Rochester, you were buying groceries on a credit card to be paid back with your first check. Where would you have borrowed $20K for a bond? On the other hand, the Univ of Rochester might well have paid the bond to lock you in for 5 years at a low salary.

One thing that I like about the employer having to make the commitment is that it is a barrier to those who would immigrate to America because this is a more rewarding country in which to be anything: including a criminal, a con artist, or a beggar. There are some immigrants whom I think we have a legitimate reason to keep out, and I like the idea of employers being the ones who make that call rather than bureaucrats in the INS.

David R. Henderson writes:

@Tom in Miami,
"Common" David? I'm common? Hmmm.
By the way, folks, I'm pretty sure, given "Tom in Miami's" awareness of my purchasing habits with groceries, "Tom in Miami" is my best friend from grad school.
Now to the issues:
Where would you have borrowed $20K for a bond?
Mainly from my father--and it would have been a loan in those days. That was a large % of his net worth.
To put it in perspective, a number of Chinese people caught in a ship off the coast of Santa Cruz in the late 1980s or 1990s, had allegedly paid about $40K in those year's dollars. People can be amazingly resourceful when the stakes are high.
On the other hand, the Univ of Rochester might well have paid the bond to lock you in for 5 years at a low salary.
I would have taken it. Also U. of R. could have lent me some of the money with a pay deduction scheme. You might think "moral hazard" and it's a reasonable thought. But if you look at the huge investment the U of R made in me without any commitment on my part to pay it back, the order of magnitude is about right.

Koz writes:

Ideas like this tend to be interesting in the abstract but don't demonstrate what I take their proponents want to show. And in fact it really doesn't meaningfully rebut Milton Friedman's point.

Just like we could conceivably cut the welfare state, we could create some kind of bond scheme to compensate for it.

But those favoring a liberalized immigration policy don't have nearly enough political influence to implement such things, and make no effort to argue that such things would occur of their own accord.

Therefore it's a forensic failure on their part. Given that, they should engage the restrictionist side of the immigration argument with more nuance than they tend to do.

Alex Nowrasteh writes:

Good idea. Bonding is not new to this debate, but it's usually used as a way to incentivize the guest worker to leave at the end of his term of work. The trick is also to make sure the migrant, who has the most power over his ability to abscond, is the one who absorbs the cost of doing so. A bond combined with deducted migrant pay in an interest bearing account that he forfeits if he doesn't leave, commits a crime, whatever, is another way. I wrote about it on page 8:


John Thacker writes:

Another idea, based on something Canada does: state-level work visas, instead of work-based. You could change employers, but only within the same state.

Thomas Sewell writes:

The way I've always written a similar proposal is a bond posted by the immigrant (could be subsidized by a company, or whomever) purchased from a bonding company to cover the entire net cost from their moving to the United States.

By net cost, I mean +taxes paid, -welfare received, -criminal damage done, -direct government benefits received.

The primary factor I believe makes my proposal much better is that rather then a set bond amount, the competing bonding companies would be on the hook for the total net cost, but could set the price to the immigrant to whatever is mutually agreed upon.

You'd have to create requirements similar to an insurance company's capital requirements for the bonding companies, but competition would help ensure the immigrant would be charged a fair and individual rate, minus transaction and tracking costs.

Thus if immigrants from a particular country are a higher risk of doing massive criminal damage, that would be priced into the cost of their bonds. If immigrants from another country, or without a degree, or without a job offer, are higher risk of ending up on welfare, their bond price would be appropriately more expensive.

If, as the pro-immigration side proposes, immigrants are a net benefit, then the bonds will be very inexpensive on average, just covering administrative costs.

If, as the anti-immigration side suggests, immigrants are a net drain, that drain will be paid for by the immigrant.

Either way, the "good" immigrants will be more encouraged to come than the "bad" immigrants and the bonding companies will be incentivized to gather all the most relevant information to telling the two apart.

Thomas Sewell writes:

Going back a couple years ago to the last time we had this discussion reminds me of another nice feature.

If the net cost to the government is positive, then the bonding company could actually be owed payments, thus making the bonding companies willing to pay high-benefit individuals to immigrate.

David J. Duecker writes:

The whole scheme presumes a program that requires administration. Note that I haven’t even suggested oversight. So what mechanism is suited to this accounting and how likely is it that abuse would soon follow implementation? Clearly a new program demands more staff and structure.

Are we not only encouraging growth in bureaucracy but incentivizing other unforeseen consequences? We would not be having these conversations if we all operated in a free and unfettered economy. Rather than fashioning gimmicks to game the incentives and institutions lets promote a policy that we all know will work.

Even simpler; have each new entrant to the U.S. sign up for a special Social Security card. When an employer employed anyone with such a card they withhold some percentage of their earnings and send it to the government for safekeeping. When the person goes back home he gets the withheld funds, less any costs he's imposed on the taxpayers, such as food stamps, medical care....

David R. Henderson writes:

@Thomas Sewell,
Thanks for the link. Actually, that was a different discussion. Good point about high-benefit people, though.

aki writes:

[Comments removed pending confirmation of email address. Email the webmaster@econlib.org to request restoring this comment. A valid email address is required to post comments on EconLog and EconTalk.--Econlib Ed.]

Thomas Sewell writes:

Yeah, it wasn't the exact same discussion. Just related to immigrant admission costs, etc... which is why the bonding point came up at that point as well.

In retrospect, you'd also have to deduct some sort of pro-rated cost for shared costs like national defense, etc... before allowing any refunds to high benefit individuals.

A bonding scheme isn't an ideal solution, but it's a better solution than most being proposed and has the advantage of potentially being politically feasible.

Alex writes:

Has anyone looked at how the cost of welfare compares to the cost of foreign aid, from the pov of improving people's lives and generating political and economic stability across the globe?

Comments for this entry have been closed
Return to top