Nathan Smith has proposed a scheme for a keyhole solution: DRITI, that has as its main features return deposits, surtaxes, and savings accounts. While I found it interesting, it looked a little complicated to me. In a similar vein, Anu Bradford suggests that a bond be posted by companies for the immigrants they hire. I liked that idea, too, but I share David Henderson’s reservations regarding the exact scheme.

Now, allow me to offer my own proposal. I am for real open borders. So it is only meant as a compromise that could work with most people being skeptical or even hostile to open borders. For want of a better name, let me call it an “immigration-backed bond” (unfortunately, “immigration bond”, which would be a better term, already refers to bonds that immigrants have to post to get out of detention).

The main points that I would like to change are these:

• The above proposals rely on individual and private contracts. It would be preferable to exploit economies of scale and create a market with public information.
• A product that needs to attract a lot of capital has to be reasonably simple. Even better if it conforms to some standard template.
• There is often a complaint that proponents of open borders want the benefits for themselves, but risks are socialized. It would be nice if you could instead put your money where your mouth is.
• Self-interest has been a reliable motivation for ages. So those who supply the capital should reap a profit commensurate with risks taken.
• To increase political feasibility, I would also like to create a material interest on the part of the government.

The basic structure of an “immigration-backed bond” may look complicated, but only relies on rather standard features that have been around for a long time, e. g. for corporate and mortgage-backed bonds. It is a bond that has a certain maturity, e. g. in ten years, and is then paid back. In the meantime it pays a regular coupon. The coupon depends on a pool of immigrants and how they behave. It might pay back before maturity (in part or whole), i. e. there is a “callable” provision. And in extreme cases what you receive at maturity could be less than 100%. Of course, the coupon would have to be high enough to compensate you for the risks you take.

How does this relate to the immigrants?

For every, let’s say, $20,000 in bonds issued, there is a visa for one immigrant. The visa is “shall issue” and not “may issue”, i. e. the government can only refuse to issue it if there are some predefined reasons (e. g. no clean criminal record, affiliation with a terrorist organization). “Shall issue” is in line with a prima facie right to immigrate, whereas the current “may issue” is not. The bond insures risks from the immigrant, and the $20,000 serve as collateral if there are any problems that result in expenses to the government or third parties (e. g. fines, damages, cost of repatriation).

Unlike with current visas, the immigrants keep their visas as long as they pay up. So losing a job or changing employers would not be a problem per se. Or someone else might pay for them, e. g. a charity, family members or a spouse. If the immigrant does not incur any such costs, or less than some limit, they become permanent residents at maturity. If they default on payments or cross the limit, they can be sent back to the country they came from.

It still may not be clear how this would work on an operational level. So let me elaborate. To make it very concrete I suggest a rather specific implementation. However, I don’t think that the specifics are in any way optimal. They are only meant to show that there is no inherent problem.

The main part would be to set up a special legal entity (usually called a “Special Purpose Vehicle”) that obtains a certain amount of money from investors, let’s say, 500 million dollars for ten years, which would correspond to a pool of 25,000 immigrants. Half a billion dollars would be more than enough for a liquid market in the bond. In principle, the Special Purpose Vehicle could directly handle everything and have a dedicated staff for that purpose. However, that is not particularly likely because it is inefficient to start all over again for each new bond. Instead most operations would probably be outsourced to external parties who receive a fee from the Special Purpose Vehicle for their services.

As for the external parties involved I could imagine the following:

• Investment banks set up the Special Purpose Vehicle and take care of bringing the bond to market.
• A bank takes over the financial side of operations.
• Rating agencies might supply a rating for the risks involved.
• An accountancy firm supervises the different operations.
• Probably a law firm takes care of handling charges from the government and re-couping them from the immigrants.
• One or more human resource consulting firms originate the pool, i. e. select the immigrants (and probably also match them with job offers).
• Some other party (e. g. a telephone operator) collects payments from the immigrants.
• Insurance companies might also be involved, e. g. insuring extreme risks.

What is perhaps unclear is how some party (probably a human resource consulting firm) would originate the pool. Here’s how this could work:

The Special Purpose Vehicle comes with a definition for the characteristics of the immigrants. That’s because investors probably want to have an idea of what the risks are. The human resource consulting firm would then try to find people with the relevant profile in the usual way via ads or going through their databases (and probably also match them with respective job offers). Or it could be the other way around: the human resource consulting firm has a larger pool of potential immigrants, they “repackage” part of their larger pool and offer it to someone who wants to set up a Special Purpose Vehicle. An accountancy firm certifies they follow a defined protocol. To make the originator interested in delivering a good result, they could be asked to take on some of the risk themselves, e. g. with payments staggered until maturity, dependent on whether there are unexpected losses.

How would the definition be set for the pool? The main point would be to target a certain level of risk, and select immigrants accordingly. So any set of attributes that has predictive power, such as age, gender, education, etc. might be interesting as well as indicators for commitment (acquisition of relevant language skills, knowledge about the target country). The originator could also ask the immigrants to post some collateral themselves, or that someone vouches for them (e. g. their prospective employers, domestic citizens in the receiving country, immigrants with a good track record in some other pool, etc.). So parameters like degree of self-collateralization or percentage vouched for by reliable parties could also play a role.

What happens with the money?

Obviously, it would be stupid to leave it in a bank account at money market rates, so it should be invested in something. And here is where I would rely on the interest of politicians and the government: the money is invested in newly issued government bonds. And to make this even more attractive for the government, you could give them better than market conditions, e. g. no coupons or even negative coupons that are paid and not received by the Special Purpose Vehicle. Technically, you would perhaps want to keep the two things separate, so you have standard government bonds that can be sold in the market, plus a regular or lump payment to the government.

Up to this point, this looks like a money-losing operation. Holders of “immigration-backed bonds” would only have government bonds with lower or even negative coupons minus additional costs. That’s why the Special Purpose Vehicle has a claim on the immigrants for regular payments that cover all costs so far (market interest rates as an opportunity cost, expenses for origination and operation, extra payment to the government, a risk premium for investors).

It may be the business of the Special Purpose Vehicle to collect payments from the immigrants. However, it would probably outsource operations to some other party. Any company in the business of collecting money over a wide area would do. One idea would be to collect it via phone bills. Alternatively, contributions could be tacked onto tax payments as a surtax that is routed to the pool (minus some compensation to the government for supplying the infrastructure and costs for collecting the surtaxes). Since immigrants might die or leave the country before the ten years are over, there could also be a so-called “callable” provision, i. e. early payback of part of the principal, either by lot or pro rata. The government might also have to cede any extra payments after that point, so the underlying government bonds can be sold in the market. Or you could do without such a provision and investors would have to bear the risk of the respective defaults.

Could this work in the real world? Just a back-of-an-envelope derivation to see whether the dimensions are okay: If the pool were pretty risky, e. g. on a par with a junk bond with a rating of BB, you would want a premium of perhaps 5% per year to compensate for this. This is a truly risky pool, as over ten years you would expect charges and defaults of about $10,000 dollars per immigrant. You would also want a market rate for a riskless security over ten years. Now those are very low at the moment. A realistic long-term level would perhaps be something like 4%. Then there may be additional costs for origination and operation of 2% and a further 2% to get the government interested. This adds up to 13%. Now 13% of 20,000 dollars would translate to 2,600 dollars a year per immigrant or somewhat more than 200 dollars a month. My guess would be that this is more of an upper bound.

The government would get 10 million per year and half a billion in capital for ten years. The originator and operator (and their subcontractors) would obtain an annual $10 million together. Again this is only meant as a rough estimate and more of an upper bound to see whether something like this could work.

Would a typical government be interested? I think it would. For an extra 100.000 immigrants in one year, the government would have $40 million in extra money per year and an extra $2 billion in government bonds issued. For an extra 100,000 immigrants per year, you would have to do this for ten years in a row and then revolve. Now it is an extra $400 million per year and $20 billion in government bonds issued. Many governments in the Eurozone would have been glad to get their hands on such cheap refinancing. Of course, in that case the underlying government bonds would account for much of the risk premium.

With this rough calculation, my conclusion is that “immigrant-backed bonds” could be feasible and interesting for all parties involved. However, there are all kinds of loose ends and potential problems:

• Unlike with mortgages that can be originated over an extended time frame and then repackaged, here many things have to happen within a short period of time, e. g. 25,000 applicants with a job offer right now. There could be some leakage with offers or applications being withdrawn at the last minute. However, one solution would be to have an excess of applicants and jobs, so there are always enough to fill the pool.
• It could be tricky to define when and how much to pay to the government. As for fines and damages, this would be the same as what the immigrants have to pay directly. But how about expenses for court cases, police enforcement, etc.? What about if these charges materialize only after the bond has matured? And how about welfare payments? There could be some moral hazard problems here.
• Is insurance limited to $20,000 per immigrant (or whatever it is, as the figure was purely arbitrary) or is there a real insurance via the pool where also extreme risks are insured (e. g. one of them perpetrating a major terrorist attack)? In the latter case: What if the Special Purpose Vehicle goes bankrupt? A typical solution would be to insure such extreme risks with a reinsurance company.
• Asymmetric information and adverse selection: the immigrants who are selected could not immigrate in other ways, so there might be some hidden risks.
• The originator could be corrupt. Supervision by an accountancy firm might not be sufficient to detect this. However, one remedy is to make them share in the risk via staggered payments that depend on results. The prospect of repeat business and protecting your reputation could be additional motivations for diligent origination.
• “Asset-backed bonds”, after which this proposal is modelled, have fallen out of favor with investors and regulators in the wake of the financial crisis, but have made a modest comeback since then. So there might be a certain reluctance to participate in an offering, especially in the beginning.
• It is easy to shoot down such a scheme with moralistic hyperbole: “You are treating immigrants like a commodity.” There is also a lot of potential for harrowing stories how immigrants are “exploited” or about “human trafficking.”
• There might be perverse incentives for politicians to close down other venues because “immigration-backed bonds” are more lucrative. So there could be some crowding-out.
• It might be hard to end such a scheme if you want to go for real open borders. On the upside, that would also apply if the government wanted to abolish it for more closed borders.

I am admittedly not an expert on structuring such products, so there might be even major improvements or else arguments why the whole concept would not work. That’s the reason I would like to offer this proposal for debate. However, if “immigration-backed bonds” came to pass as the best possible alternative under given circumstances, I would certainly be interested in investing my own money in them, just to show that I put my money where my mouth is – and to make a profit from a great opportunity to increase wealth.

[I would like to thank the Open Borders team for valuable input on a previous version of this post. I hope I now address most of the points they have raised. As always: all errors are mine. And for full disclosure, I would like to add that I work in the financial sector, but have no professional stake in what I suggest. And I am certainly not speaking on behalf of my employer.]