Thursday, July 04, 2013

Open borders in (at least) one (developed) country

Over at the Open Borders Blog Vipul Naik and co-bloggers have discussed an enormous variety of arguments for and against a system of open borders. Immigrants from poor countries earn much more when they move to rich countries, a "place premium." If this place premium is unaffected by very high levels of immigration, then open borders could greatly increase world GDP and eliminate almost all absolute poverty. Refugees would be guaranteed a place to escape bad conditions, governments would face incentives to improve their policies to keep their population from leaving.

So immigration is very much worth a look for those interested in effective altruism. This post covers a point that seems to me to have been relatively neglected among open borders advocates (although it has been discussed more by advocates of charter cities): it seems most of the expected benefits do not require a global system of open borders, just open borders in one or a few countries with the right properties, a much easier goal.

Place premium and population
Estimates of doubling GDP and eliminating poverty are based on the assumption that the place premium in rich countries is robust to very large increases in migration, e.g. because the place premium is caused by differences in legal systems and government, and those would not be much affected by large-scale immigration.

So the place premium would not be primarily a function of the relative proportion of natives and immigrants (over the relevant ranges), and could be retained even with a population that was 90% immigrant or more. Thus one or two large, developed countries might be able to meet the global demand for movement to a country with high place premium.

Consulting Wikipedia's table of countries and their population densities, if the United States, Canada, or Australia were as densely populated as South Korea, then they could each hold close to 5 billion additional people. At densities like those of New Jersey or Bangladesh the numbers could be increased by half again. A similar result would ensue if 10% of the area were as densely populated as the city-state of Singapore. The Canadian northern regions and Australian desert regions are less hospitable, requiring costly heating and water investments but nonetheless the potential for growth is clearly very high.

Such a heavily populated country would clearly have to import foodstuffs, but would likely increase global food production (through increased wealth and demand as the migrants become more productive thanks to the place premium of their new home).

At least one place to go
The Berlin Wall was a symbol of villainous oppression, keeping East Germans trapped in the clutches of Communist rule. Defenders of border walls and fences today say that they are oriented towards keeping intruders out, rather than trapping prisoners inside. And in ordinary life, it is much more objectionable to keep someone trapped in your basement than to prevent strangers from trespassing on your property. After all, in the latter case you're not stopping them from going anywhere else.

However, under the current international system many people who do not qualify as refugees, skilled immigrants, or other permitted classes may have no other country whatsoever which will accept them. So each developed country, by refusing to open its borders, is counterfactually responsible for denying potential immigrants the ability to move to any country with a high place premium. Some might argue that a country would only be responsible for its "fair share," corresponding to the immigration share it would receive under a global regime of open borders (similarly to claims about fair shares for foreign aid spending), but it is clear that the harm done to potential immigrants when they are denied entry is much greater in the absence of alternatives.

So if one large country opened its borders it would reduce the pressure on other countries to take more immigrants, and ensure that people in dire places always had a bolt hole. This would also offer a benefit of competitive government, encouraging policy improvements to prevent such emigration in less developed countries.

Value of information
The above analysis depends on the assumption that the place premium remains mostly intact even when immigrants are much more numerous than natives, and not selected. If this is not true, it would be extremely valuable information to know for countries elsewhere. If the place premium remains intact, then other countries could open their borders as well. If it falls, they could seek alternative arrangements to preserve the premium for large immigrant flows.

Such demonstration effects can be very important. The economic development of capitalist Singapore and Hong Kong is said to have played a role in leading Deng Xiaoping to implement his economic reforms in mainlaind China, one of the greatest success stories in poverty reduction in history.

Costs to natives
Natives may fear that large immigrant inflows will result in higher taxes, crime, unwelcome political change, falling wages, or other problems. These problems might be offset by, e.g. applying a surtax to immigrants and using the (potentially astronomical, for immigrant populations much outnumbering natives) revenue to compensate natives.

However, the total costs to natives, even if serious, would seem to scale with the size of the native population, e.g. if some harm befalls each of the 316 million current residents of the United States, that is worse in total than the harm befalling each of the 35 million current Canadians. Since the benefits to immigrants (assuming place premium is stable) are not so tied to the size of the current population, the cost-benefit ratio of open borders in a less populous country could be much improved even if there are harms to natives.

The case for subsidizing/supporting an open borders experiment
The above considerations suggest that the world could benefit a lot from trying open borders in one country, even if it thinks the global open borders is too risky and radical a move. The biggest problem is that countries tend to make decisions based on their own perceived benefit (or rather, electoral benefit to current politicians) rather than on global benefits. So there would be a case for large advanced economies like the United States and European Union to subsidize some small countries with high-quality institutions to open their borders (perhaps initially with maximum quotas of merely stupendous size to slow the transition).

The focus could be large developed but underpopulated countries like Australia and Canada. Those countries are already among the most receptive to immigrants on a per capita basis, although they focus on highly skilled immigrants and not the global poor and uneducated.

Or it could be small countries, too small to receive the world's immigrants (immigration would ultimately be choked off by rising land and other prices) but cheaper to subsidize. This idea would blend smoothly into the concept of charter cities.

Subsidies could include cash grants for taking immigrants from poor countries. This has past precedents, e.g. payments to other countries to hold refugees and asylum seekers in place of the funding country. Similarly, aid agencies could pay a per-migrant bounty to any country taking qualifying migrants from poor countries. A more exotic option would be promises to current natives that they could emigrate if things go badly for them (e.g. native Canadians prior to the reform could be guaranteed the right to move to the United States even as Canada opens its borders).

Immigration advocates could also concentrate their political advocacy for open borders on small countries where key decisions involve fewer people and limited activist and financial resources go further. Vipul Naik suggests that political activism in prominent countries like the United States may be more likely to reach international media, but the demonstration effects of open borders in one place seem likely to be stronger than more diffuse public opinion changes.

Such subsidies would be a very novel policy shift, but could fit into existing institutions. Aid budgets could increase incomes more with subsidies to migration than standard spending patterns, and there are parallels with carbon markets, which subsidize those countries which do the most to reduce carbon emissions, and past side-payments to affect migration.

More robustly, the broader points that most of the benefits of open borders can be achieved without global adoption, with even a single country making major changes, indicate that the task of open borders advocates is easier than one might think.


Erik said...

I think this is exactly how open borders would happen. One nation would be bold enough to try it; I doubt that many nations would simultaneously develop the political will to try it.

But that one nation might not give an accurate picture of how open borders would affect other nations that have very different cultures and economies.

Perhaps there are some very poor countries that would have little to lose from open borders, because things there could not get much worse.

For an opposite kind of example, consider the Japanese. They are known for having corporations that manufacture goods of exceptional quality and technological complexity. Many people appreciate using those products.

Why do the Japanese produce that result? I’ve seen statistics suggesting the Japanese, on average, have a high analytical or mathematical type of intelligence. But maybe there’s something else at work. For example, perhaps some of their people have a drive to be thorough and efficient in doing things and to constantly improve. Or maybe there are some subtle mental traits involved that I can’t even imagine.

And it’s difficult to know whether those traits are the result of culture or genetics or both.

Suppose the Japanese allowed mass immigration. Unless they sustained a high level of racism, the Japanese over the generations would presumably mate with the immigrants. Eventually, Japan might be largely a mixed-race society. Why should we assume that those future generations would behave like the present generations, and that Japan would continue to be a manufacturing powerhouse?

So Japan's experience with open borders could be very different than the first test nation's experience.

Carl said...

"Perhaps there are some very poor countries that would have little to lose from open borders, because things there could not get much worse."

I did write "one (developed) country."

Regarding Japan, in PPP GDP per capita, Japan is ranked in the 20s, behind Singapore, Hong Kong, Macau, Australia, Sweden, Germany, the Netherlands, and others. Since all of these different cultures were able to achieve development, I would expect that the mechanisms of development are not so culture-specific that the experience of one developed country would be irrelevant to others.

DB said...

While I strongly support the idea of a controlled open borders experiment, I think Canada is the worst possible location. Right now, it is the closest thing to a control for the lite (and admittedly, very far from theoretically ideal) open borders experiment which has already been forced on the American public. If the rest of the US goes the way of California, Canada is the most natural place for Americans preferring the less unequal society of their childhood to immigrate to, and I would not try to take away that option unless I had genuine hatred for the American middle class.

A set of much smaller countries (having more than one reduces the probability of path-dependent failure) would be better. My expectation is that any true-believer first attempt will (at least start to) fail spectacularly, but after some time for iteration a coherent set of policies will develop which supports freer migration than the current norm, and without the need for Singapore-style authoritarianism.

x0077FF said...

If you think of open borders as free trade in labor, it's clear why it's valuable to enact it unilaterally - same reasons why unilateral free trade is extremely valuable for the country enacting the free trade. Any country trying to attract new people by opening their borders - even a relatively undeveloped country - could see benefits just the same way you can see gains from trade.

Unless you're Liberia or one of the other worst-of-the-worst countries, you're at least an upgrade from somewhere, which is why countries like Algeria or Mali - despite being undeveloped - can still see an increase in GDP due to an influx of relatively cheap labor from Zimbabwe or Liberia or something - and that could free up relatively educated people for things like call centers, education, etc.

Anonymous said...

I am for immigration, but I findbthis piece poorly argued. Why wouldn't place premium decrease with use? What model is it based on? Why would the hottest region of Australia and the coldest region of Canada be a good place to locate people? Yes that puts them in the borders of a reasonable government but they can't work magic on the landscape. Why wouldn't immigration costs scale with the numbet of immigrants? And that's when I stopped reading...

Carl said...


I'm not assuming that place premium is unaffected by migrant flows. As I wrote above, the standard calculations of the benefits of open borders assume this is approximately true.

I wrote that if that is right, one only needs one or a few open-migration countries with good institutions to capture most of the benefits (benefits so large that desalinating water and the like could be paid for out of the proceeds) and if they are wrong, an experiment could reveal this and allow rethinking of plans.

I don't pretend to know with confidence what the actual response curve of place premium to different migration patterns would be.

Vollmer said...

Any thoughts on whether relatively stable and wealthy developing countries might be a) easier to lobby than developed countries, and b) more suitable for migration thanks to lower cultural and linguistic fragmentation?

Consider Botswana – a democratic country that has been relatively stable for decades and has a per-capita GDP higher than that of several Eastern European countries. Has anyone taken a closer look at such countries?