The American polity can endure and flourish with open borders

Post by Nathan Smith (regular blogger for the site, joined April 2012). See:

Very interesting discussion in a recent comments section. Let me start with a quote from an e-mail by Garett Jones to Vipul Naik:

I would emphasize a different conclusion: That the low-IQ immigrants will tend to worsen the institutions of the higher-IQ countries they move to. Low IQ immigrants will, to some degree, tend to make the country they move to more like the country they came from.

Partly this will be due to MRV and Caplan/Miller reasons: low IQ groups vote for bad policies.  Partly it’s because they will tend to elect individuals from their constituencies, which will, on average, tend to lower the average IQ of the legislature.  And partly it’s because the bureaucracy will tend to hire individuals from low-skill groups, which will lower government quality.

For these and other reasons, new low IQ citizens impose a tax on the nation’s institutions, and this institutional cost should be counted in a candid cost-benefit analysis.

*Shorter version: Good institutions are rare treasures, and institutions are endogenous with respect to (among other things) citizen IQ.  *

I would like to bracket the concerns in the second paragraph about voting, because I regard this as a solved problem, as far as the theory of open borders is concerned. Just because you let people in doesn’t mean you have to let them vote. There are already millions of Green Card holders in the US who can’t vote. The keyhole solution (Vipul’s term: I hadn’t thought of it at the time) which I advocate in Principles of a Free Society is open borders with (a) migrants preimbursing the government for their voluntary deportation if they become destitute (at least those under a new open-borders visa), (b) a surtax on (those) migrants, (c) mandatory savings, withdrawable only in the migrant’s home country, or forfeitable as part of (d) a path to “earned” citizenship once a migrant has saved a certain threshold amount in this mandatory savings account. Migrants would thus be given a substantial incentive to return home rather than to stay. Those who would choose to stay would presumably do so: (a) because their homelands were an especially bleak alternative; (b) because they foresaw high earnings in America so that forfeiting the savings account was worth it (these people would probably have relatively high IQs, on average); or (c) because they especially like, admire, and enjoy America (these people would presumably place a particularly high value on American institutions). To mix these migrants into the electorate is a very different, and doubtless much more favorable, prospect, than simply allowing anyone to come and vote. Of course, this is just one of many possibilities that would separate the right to come, live, and work from the right to vote. I made a similar point in the comments of the Garett Jones post, and added that “It’s even easier to maintain high hiring standards for the bureaucracy, which obviously doesn’t have to, and doesn’t, hire a representative cross-section of the resident population.” I also posed the question:

If I were to hypothesize that the maintenance of high-quality institutions depends mainly on the characteristics of an elite, and need not be much affected by adverse changes in the composition of the broad mass of the population, would the evidence that Garett has studied contradict me?

This was the jumping-off point for a very interesting debate between BK and John Lee. BK’s comments, in particular, are highly interesting and informative, yet I find myself unconvinced and dissenting at many points. BK answers my question:

Yes, if we are just referring to the overall demographics of a country. Note that across countries, smart people earn higher incomes as the proportion of smart people rises, not the absolute total.

Chinese-Singaporeans generate income almost twice as great in mostly Chinese Singapore as the large Chinese-Malaysian minority does in Malaysia (about $70,000 per annum vs about $38,000), even though there are less than 3 million Chinese in Singapore but almost 7 million in Malaysia. But the Chinese make up 75% of Singapore vs 25% of Malaysia…

There is a Chinese elite, but this isn’t enough to fix the institutions, which have to represent the general population. All this occurred in the context of strong legal discrimination in favor of Malay majority, racialized anti-business sentiment, and big gaps in political views between Chinese and non-Chinese Malaysians.

OK, but wait. There is a Chinese business elite in Malaysia, but the political elite is Malay. It is this Malay political elite that imposes “strong legal discrimination in favor of Malay majority, racialized anti-business sentiment,” etc. It is also important that (a) the Malays have deeper historic roots whereas the Chinese are relatively recent arrivals, and (b) the Malays are linguistically and religiously homogeneous (more or less, I think: BK and John Lee both know the region better than I do). If we’re looking for lessons from Singapore/Malaysia that cross-apply to a hypothetical open-borders United States, this argument would only be relevant if we’re supposing that voting immigrants would become the majority of the population, develop solidarity among themselves, and vote for “strong legal discrimination” and “racialized anti-business sentiment” against the offspring of today’s natives. With immigration tariffs and a gradual path to citizenship, you could more or less ensure that voting immigrants would never constitute a majority. Since immigrants would come from many different countries, it’s unlikely they’d develop solidarity among themselves except on wedge issues that related to them directly. Instead, they’d want to assimilate with American natives. Given that American society has a powerful absorptive capacity– if you’ve got fluent English and a college education and want to be a normal American, people will treat you like a normal American; and if you wereborn here, it’s taken for granted that you’re a normal American, never mind your background– any scenario resembling that in Malaysia is really quite implausible.

Also, I think the fact that immigrants would know they were immigrants makes a big difference. Malays in Malaysia think of the land as theirs. They’ve been there the longest. South Africa is in a similar situation, as far as I understand: black South Africans see themselves as the rightful owners of the soil, the whites as intruders. Russia, whose history I know better, stands in striking contrast to the 19th-century United States, because while they could both be described as multi-ethnic empires, in Russia the subordinate nationalities had never consented to be part of the Russian Empire, but for the most part had been simply conquered (it’s a little more complicated but never mind), whereas in the United States, the subordinate nationalities (if I may put it that way for the sake of the parallel) had in a real sense consented to rule from Washington by crossing oceans to immigrate. They were therefore much less inclined to question the legitimacy of the government and far more inclined to develop patriotic loyalty to the United States, superimposed on a lingering loyalty to their various mother countries.

BK asserts that “institutions… have to represent the general population.” No, they don’t. Certainly, it’s not a logical necessity. In England after the Norman Conquest, for a couple of centuries, institutions represented the Norman conquerors, not the Anglo-Saxon peasantry, just to give one example. Perhaps BK only meant that institutions do represent the general population in Malaysia. Fine, though even then, I find “the general population” a question-begging reification. Anyway, I do not think it’s the case, in general, that institutions depend all that much on the median voter even in democracies. Caplan’s post on affluence and influence summarizes recent research that politicians listen much more to the rich than to the poor. Meanwhile, courts are staffed by people with law degrees. They are only very indirectly answerable to the electorate. The Federal Reserve is autonomous, and listens to the economics professoriate more than to politicians. BK contradicts my claim about the bureaucracy, saying:

Er…bureaucracies and even private firms are required to hire representative cross-sections along several dimensions under affirmative actions and disparate impact standards. And institutions that are seen as unrepresentative take serious reputational penalties, while there are strong political pressures from subgroup activists for equal representation. This is true of the military, which is relevant for the risk of coups or civil wars, federal bureaucrats setting policy, teachers, PhD programs, science funding agencies, medical schools, law schools, university professors and researchers, and more.

I’m not sure that’s an accurate description of affirmative-action law, but at any rate, bureaucracies are certainly not required to higher a representative cross-section of the population in terms of ability. If you have to hire minorities, you’re certainly not barred from hiring the smartest minorities. Even if an ethnic group has lower than average IQ or education, there will be a few high achievers. Anyway, it’s certainly not infeasible to not extend affirmative-action practices to new immigrant groups; nor do I think it would be particularly difficult to avoid such an extension politically. American blacks (or African-Americans, though I don’t like that word, since it makes them sound like recent immigrants when they’ve been here, on average, longer than European-Americans like me) are a special case, because they never consented to come. Black Americans have an enduring identity in a way that Italian-Americans, Irish-Americans, German-Americans, etc., have never had. Proximity and undocumentedness give Mexican immigration a special character– one advantage of open borders is that it would be disproportionately favorable to non-Mexicans for whom undocumented entry by land is not an option, and would therefore likely be less divisive because immigrants coming from many source countries would have nothing to assimilate to except mainstream American culture– but even Mexican-Americans have nothing like the distinctive identity that blacks do. I highly doubt that an open-borders scheme such as that in Principles would threaten American meritocracy. (On the contrary.)

BK writes:

Using the above statistics, if the Chinese-Malaysians could have done as well as Singapore by also seceding from Malaysia into Chinese-dominated countries, total GDP of the region would rise substantially just from letting the Chinese-Malaysians free of the Malaysian electorate, even if incomes back in Malaysia plummeted. But it gets even better: Singapore lets in millions of guest workers from non-Chinese Malaysia, among other places, who send back huge quantities of remittances. Singapore generates more innovations in science and technology with positive spillovers for the rest of the world.

Basically, patterns like this seem to suggest that total GDP and welfare are much increased by international segregation by IQ and other characteristics contributing to productivity and performance, and that giving every country in the world demographics representative of the world would be devastating.

The claims that “total GDP of the region would rise substantially” with segregation and that “total GDP and welfare are much increased by international segregation by IQ” are interesting because they are the kinds of claims that restrictionists would need to make and defend if they wanted to have a really strong case against open borders. If Malaysian Chinese who control 60% of the Malaysian economy could be rendered almost twice as productive by political segregation, that would increase regional GDP by 10% or so even if Malays’ incomes fell to zero. But this is a very suspect extrapolation. First, Singapore is a densely populated city, while Malaysian Chinese are more spread out. City-dwellers always tend to be more productive. So if the Malaysian Chinese escaped Malay rule but were still spread out, we shouldn’t expect Singaporean levels of productivity. More generally, Singapore is probably the world’s most productive society today. That takes getting a whole lot of things right: location, personalities, finding certain economic niches, an institutional history going back to British rule and Stamford Raffles. With all due respect to the Malaysian Chinese, to assume that if they just had independence they’d replicate the world’s most productive society is surely too generous.

That said, no one is suggesting that we should give every country in the world demographics representative of the world population. Open borders certainly wouldn’t lead to that. The US’s internal open borders don’t lead to every city, county, and state having demographics representative of the US population, and moving from Boston to Tampa or Minnesota to Maui would be a lot easier than moving from Bangladesh to Oslo or Beijing to Fairfax, VA, even if all immigration restrictions were removed. (Culture, distance from family, etc.) Certainly, segregation by IQ contributes to productivity; but it doesn’t need to be government-imposed. Harvard segregates by IQ through its admission process. Top research labs segregate by IQ: they hire only very smart people. One doesn’t need to stop uneducated immigrants at the border in order to keep them out of faculty lounges or academic seminars or corporate boardrooms.

For all our dueling, BK and I might not differ all that much. I agree that there’s such a thing as the American polity, and that it’s important to preserve the integrity of the American polity. That polity consists, at the top, of highly educated judges and economists and legislative aides and bankers and the like, but it extends down to all sorts of average voters who watch presidential debates even though they’re not being paid for it, and debate politics on Facebook, and deliberate about whom to vote for, and get indignant, and donate to campaigns, even though single solitary citizens can’t have any effect, so it’s all individually irrational. As important as preserving the American polity is, I think some methods– deportations, separation of families by force, exclusion of individuals whose best chances of religious freedom or economic opportunity lie here– are unacceptable. Fortunately, they are not necessary. With careful policy design, the American polity can endure and flourish with open borders.

18 thoughts on “The American polity can endure and flourish with open borders”

  1. Forgive the length of this comment, as It tries to address many points addressed to me, and includes a look at South Africa and ZImbabwe in the spirit of the Malaysia/Singapore analysis.

    “I would like to bracket the concerns in the second paragraph about voting, because I regard this as a solved problem, as far as the theory of open borders is concerned…With careful policy design, the American polity can endure and flourish with open borders.”

    Neither Garrett nor I ever disputed that a keyhole solution of open borders which limited grants of citizenship and electoral participation could work, and work spectacularly, given that it was exogenously imposed. The concerns are about the alternative with political integration. And that alternative is relevant whenever we consider actions that affect the chances of both.

    The analysis re Malaysia and Singapore was clearly and explicitly about political merger, contrasted with SIngapore’s current policy of allowing unskilled workers in large numbers but not allowing them to become citizens.

    “The keyhold solution (Vipul’s term: I hadn’t thought of it at the time) which I advocate in Principles of a Free Society”

    Sounds fine, as long as the price of citizenship isn’t set too low.

    “There is a Chinese business elite in Malaysia, but the political elite is Malay.”

    This was responding to the idea that as long as you have some skilled people in your economy you can just hire them. In theory nothing is stopping the Malaysian government from adopting a meritocratic system for staffing government offices, except the Malaysian populace.

    “Perhaps BK only meant that institutions do represent the general population in Malaysia. ”

    Yes.

    ” (a) the Malays have deeper historic roots whereas the Chinese are relatively recent arrivals, and (b) the Malays are linguistically and religiously homogeneous (more or less, I think: BK and John Lee both know the region better than I do)”

    As I said in discussion with John, I agree that these are contributing factors in this and other specific cases, but that many diverse reasons get invented to attack market-dominant minorities, and so much of the blame has to be assigned to the disparities and the ability to track those disparities to recognizable groups. Simply observing big disparities in outcomes is enough to lend a lot of weight to political claims of wrongdoing, and political entrepreneurs can use religious, nationalist, racist, anti-capitalist, and many other rationales.

    “First, Singapore is a densely populated city, while Malaysian Chinese are more spread out. City-dwellers always tend to be more productive.”

    Malaysia is 70% urban, and the Malaysian Chinese are much more urban than the majority population in Malaysia.

    “More generally, Singapore is probably the world’s most productive society today. That takes getting a whole lot of things right…”

    As I said in the thread I agree we should regress downwards for these other causes. On the other hand, as I mentioned, there are other positive effects not included in the GDP figure such as extra R&D, innovation, contribution to global and regional public goods that affect worldwide welfare, and future generations.

    “an institutional history going back to British rule ”
    Agreed, although the rest of Malaysia also had British rule, if less micromanagement.

    “With all due respect to the Malaysian Chinese, to assume that if they just had independence they’d replicate the world’s most productive society is surely too generous.”

    I picked Singapore as an example because of the past integration, geographical proximity, and so forth, and only looked up the Malaysian Chinese statistics after forming the intention. I didn’t dredge through other options. I’ll try the analysis again for a different region, randomly selected to be Africa.

    The obvious data are the economic evolution of Zimbabwe (formerly Rhodesia) and South Africa after universal suffrage and the end of apartheid. This is complicated by the fact that both countries were suffering economically from crippling sanctions before majority rule, as well as internal racial conflict which were then lifted and replaced with foreign aid as part of an intentional effort to make post-suffrage conditions better than pre-suffrage conditions.

    In Zimbabwe the aggregate economic effects have not been good, despite the moral virtues of eliminating the old racist system (which economically inefficiently prevented smart black Rhodesians from competing with white skilled labor, in addition to its other nasty features):

    http://en.wikipedia.org/wiki/Economy_of_Zimbabwe

    The white population under Rhodesia enjoyed high incomes, but has mostly fled the country. Presumably they are earning Europe-typical incomes OECD countries unless they were killed, but were seriously disrupted in productivity for a long while. Incomes for the black population were substantially higher than elsewhere in Africa but have suffered since majority rule (although again, here the market-dominant minority was the immigrant population rather than the native population).

    In South Africa growth has been better than during the period of severe sanctions, but disappointing expectations, combined with big blows to life expectancy and human development, in part through policy such as denial that HIV causes AIDS..

    In 2009 Whites in South Africa earned about R136k Asian R56k, mixed-race or “coloured” 28k, and black 19k. The population numbers today (from wikipedia) are 9.2%, 2.6%, 8.8% and 79.4%. So rounding up 40% of income was going to whites, 5% to Asians, 8% to mixed-race people, and 48% to blacks. A just-finished census reports that income disparities and population ratios are about as estimated then, and GDP per capita overall is $11,900, with the white South African income.

    So, it seems that world GDP would not be much increased if the South African whites had all been shipped to the United Kingdom (although South African GDP in nominal terms has massively increased in the last year due to a resource price swing, which pushes it above expected GDP going forward, the overall figures are still good). Only 10% of the white population has left despite high crime and Africanization policies.

    These results are not that bad, and a bit better than I expected for South Africa. And in Rhodesia the richer minority was mostly able to flee without being killed and settle elsewhere. The black population of ZImbabwe suffered a lot after the end of Rhodesia, but they benefited from increased growth and prosperity under Rhodesia.

    One might worry that South Africa might also suffer from Zimbabwe-style demagoguery in the future, and that legacy institutions may be declining in quality, (much of the aggregate lag in Africa looks to be from the frequency of civil wars, expropriations, and other big-bang policies which are lumpy in time). Worry about political externalities affecting things like science policy, global warming policy, nuclear war and disarmament, and other global public goods/issues by lowering the average quality of the electorate in the most powerful countries would remain in force.

    But still overall it makes me somewhat reduce my estimates of the costs of political externalities for open borders with open citizenship. I would be interested in seeing more such analyses for other regions, ethnicities, cultures, institutional setups…

    “The US’s internal open borders don’t lead to every city, county, and state having demographics representative of the US population, and moving from Boston to Tampa or Minnesota to Maui would be a lot easier than moving from Bangladesh to Oslo or Beijing to Fairfax, VA, even if all immigration restrictions were removed.”

    If the U.S. wouldn’t look like the world, would the differences be in the direction of India, or Namibia, or China, or Brazil? In earlier comments I suggested that the migrant distribution would be more weighted towards the young and mobile, who are proportionally more numerous as you look at poorer potential big sources of immigration: Africa<India<China.

    I'm not sure what sort of divergence you're trying to claim here. Can you give a rough estimate of the percentage ix of incoming migrant flows over 30 years from Latin America, Africa, China, South Asia, the Middle East, and currently rich countries? Estimates for flows with easy citizenship vs keyhole solutions would be good too if you think they would differ.

    Regarding the situation in the U.S. I think it's important to mention some relevant facts. The U.S. makes enormous transfer payments from rich regions to poor (Social Security, Medicare and Medicaid, military bases, agricultural subsidies) which reduce the incentive to move, wage differentials are small, nominal internal open borders are offset by local exclusive policies like restrictive housing and environmental regulation that keep out poor migrants, and there is still huge internal migration.

    Also, given a situation of open international borders, wouldn't you want to move on to internal barriers to migration like restrictions on the construction of dense new housing?

    "If you have to hire minorities, you’re certainly not barred from hiring the smartest minorities."
    First, a disclaimer, I think that the "smart bureaucrat supply" is less important than the electoral challenge.

    This is true in many contexts, but false in many others. Frequently court rulings require that a valid test predictive of job performance be thrown out for disparate impact, but courts also take lawsuits for formal discrimination. So you get a result where the organization just has to lower standards until almost everyone can pass a test to reduce group differences. That was the situation in this recent supreme court case involving a test for firefighters:

    http://en.wikipedia.org/wiki/Ricci_v._DeStefano

    Similarly, in Texas and several other states inefficient proxies for race wind up being used, such the top 10% of high school students being automatically admitted to state universities, regardless of the absolute quality of the student or school (this produces racial balance because of de facto racial segregation of high schools). In fact a Supreme Court case up now was fighting over precisely the question of whether Texas universities that already have racial balance via these proxies can discriminate directly on the basis of race in order to get smarter and richer minorities.

    http://en.wikipedia.org/wiki/Fisher_v._University_of_Texas

    1. Next look: Latin America, first checking Brazil and Mexico, selected as the top two on this list:

      http://en.wikipedia.org/wiki/List_of_Latin_American_countries_by_population

      and making up over 300 million people by themselves. There, self-identified whites have GDP per capita far below the EU average, or Portugal or Spain (European ancestry is mostly from the colonial powers but not a huge majority).

      However both countries are heavily admixed, so self-identification translates to a different ancestry than places with less mixture. Brazil seems to be >(2/3) European genetically, with Mexico self-identified ‘Mestizo’ somewhat less (they make up about 80% of the population. So comparing self-identified Portugese-ancestry ‘whites’ does not give us a clear comparison group to match against Portugal.

      http://en.wikipedia.org/wiki/Race_and_ethnicity_in_Brazil#Admixture
      http://en.wikipedia.org/wiki/Race_and_ethnicity_in_Brazil#Racial_disparities
      http://en.wikipedia.org/wiki/Mexico#Population_genetics

      We can say that if one could proportionally scale up the population of Europe (weighted across European countries by their contribution to the Mexican and Brazilian populations, so especially Spain and Portugal) by the population of these countries multiplied by their European ancestry, total world GDP would be substantially increased, despite the fact that Portugal and Spain are relative economic laggards within Europe..

      If Garrett Jones is right that “max IQ” or “smart fraction” is important, i.e. variance, then admixture would tend to reduce national GDP, replacing a high variance bimodal distribution with a a single bell curve between the two previous peaks.

      1. “We can say that if one could proportionally scale up the population of Europe (weighted across European countries by their contribution to the Mexican and Brazilian populations, so especially Spain and Portugal) by the population of these countries multiplied by their European ancestry, total world GDP would be substantially increased, despite the fact that Portugal and Spain are relative economic laggards within Europe.”

        Oops, rather if that occurred and the populations of Mexico and Brazil were proportionately reduced and replaced with largely Amerindian+African populations, even if incomes in those countries plummeted.

    2. Next up: the Caribbean.

      http://en.wikipedia.org/wiki/List_of_Caribbean_island_countries_by_population

      The large majority of the population is in the three big countries: Cuba, Haiti, and the Dominican Republic.

      It’s hard to tell what the demographics of Cuba are without genetic studies, and it was hard to find income by race (which is less illuminating as a guide to productivity in a communist state, in any case):

      “There is disagreement about racial statistics. The Institute for Cuban and Cuban-American Studies at the University of Miami says that 62% is black,[167] whereas statistics from the Cuban census state that 65.05% of the population was white in 2002. The Minority Rights Group International says that “An objective assessment of the situation of Afro-Cubans remains problematic due to scant records and a paucity of systematic studies both pre- and post-revolution. Estimates of the percentage of people of African descent in the Cuban population vary enormously, ranging from 33.9 per cent to 62 per cent”.” http://en.wikipedia.org/wiki/Cuba#Current_demographics

      Nominal GDP in Cuba is about $5,000, PPP about twice that. European Union-typical, or Spain-typical GDP for the white population, with plummeting incomes for the rest, would clearly increase world GDP, but the exact amount is hard to determine.

      Haiti and the Dominican Republic are right next to each other, almost identical population size, but quite different demographics and GDP. Haiti is listed on the wikipedia as 85% black:15% mixed-race (the CIA factbook gives 95% black, 5% mixed-race and white), and has income of $738 (nominal) and $1,235 (PPP). The DR has income per capita of $5,638 nominal, about $10,000 (PPP), and “according to the CIA World Factbook, the Dominican population is 73% multiracial, 16% white, and 11% black.[1] The multiracial population is primarily a mixture of European and as much as 70% African, but there is as well a minor Taíno element in the population;[24] research published in 2010 showed that 15% of Dominicans have Taíno ancestry, and 70% have African genes.”

      Within both countries multiracial people earn significantly higher incomes than black, and Europeans and Asians the highest incomes. Again if we were to take the population of these countries, multiplied by the European genetic share, and give that hypothetical population EU-typical GDP, or Spain/Portugal typical GDP, total world GDP would go up even if the incomes of the residual plummeted.

      http://en.wikipedia.org/wiki/Dominican_Republic#Ethnicity
      https://www.cia.gov/library/publications/the-world-factbook/geos//ha.html
      http://en.wikipedia.org/wiki/Social_class_in_Haiti

  2. My university had a program with the Malaysian oil company to bring people over here to train to become engineers. So we had a lot of Malaysian and Chinese Malaysian people in my dorm. What stuck with me is just how much they hated each other. I mean a lot, worse then much the the racial animosity I was used to in America. If I ever needed a primer on why multi-ethnic societies tended to be miserable, especially those with a superior minority and inferior majority, this was it. I definitely don’t want America becoming like Malaysia and having the same problems I observed in that dormitory.

    All of these keyhole solutions are good, but they won’t last. Eventually a politician, race hustler, or something else will find a way to pick that lock and they will be able to vote and get the same rights and benefits. In my own state they are voting on giving some kind of benefits to illegals in this election. It’s inevitable.

    1. BTW, Lee Kuan Yew is a massive racists who hates AA and actively supports immigration of “good” races and tries to shut out “bad” races. Again, none of this is controversial in Asia. People simply get it and try to maintain their own people’s and cultures. Try to listen to the words of the greatest statesmen of our generation.

      “Lee says that ‘I started off believing all men were equal. I know now that that’s the most unlikely thing that’s ever been, because millions of years have passed over evolution, people have scattered across the face of this earth, been isolated from each other, developed independently, had different intermixtures between races, peoples, climates, soils…. I didn’t start off with that knowledge. But by observation, reading, watching, arguing, asking, that is the conclusion I’ve come to.”

      “Lee has been instrumental in encouraging the immigration of Chinese from Hong Kong. He defended this policy on the grounds that the birth rate of Singapore’s Chinese is lower than that of its Indians and Malays. ‘The numerical superiority of the Chinese must be maintained’, he said, ‘or there will be a shift in the economy, both the economic performance and the political backdrop which makes that performance possible.”

  3. Many thanks, BK, for these richly informative and interesting comments. I appreciate the time you’ve taken to look up data, but it’s not just that: it takes a good deal of insight to know what data to look for. If I may summarize the Jones/BK theory, it seems to be:

    1. High IQ (with race as a proxy) => better institutions => higher total factor productivity

    2. The boost to total factor productivity from raising average IQ is large enough that the global segregation created by border controls actually raises worldwide GDP.

    This is a difficult theory to test empirically– as with any empirical investigation of the determinants of global GDP, you have all sorts of problems of multicollinearity to sort through– but you’re making valiant efforts. My impression of the results of your investigation so far:

    1. The Malaysia/Singapore example is suggestive at best, as it seems much too generous to assume that Malaysian Chinese would be as productive as Singaporean Chinese, with its high population density and the genius of Lee Kuan Yew, absent the Malay political elite.

    2. South Africa, perhaps surprisingly, doesn’t yield strong support to the hypothesis, as South African growth rose after apartheid. Still, that may reflect resource price trends, and a question mark hangs over South Africa’s future.

    3. Rhodesia/Zimbabwe gives some support to the hypothesis, and I might even stress that more than you do. I met a lot of exiled Rhodesian whites while I was in Africa, and it was a haunting experience. A nation, shattered. The tragic struggle against majority rule, reviled by the world, yet receiving a strange, sad kind of half-vindication long afterwards as Mugabe ruined a once-flourishing country and made himself a byword for catastrophic incompetence and thuggish tyranny.

    4. You think Brazil/Portugal supports the theory. I looked at the figures a bit and I think it doesn’t. Brazil’s more than half as rich as Portugal and the whites there already make considerably more than the average. But there’s another element here. Portugal and Spain have been taken under the EU’s wing in the past few decades. Their institutions don’t just depend on Portuguese and Spaniards, but on the Europe-wide electorate and technocracy that is probably more answerable to Berlin and Paris and the Benelux countries than to Portugal and Spain. This has probably boosted Portuguese and Spanish incomes significantly. If you dock “natural” Portuguese GDP per capita a bit to remove EU influence, it starts to look like Brazilian whites do no worse for not being segregated.

    5. Spain/Spanish America may support your point better. However, what is one to make of countries like Argentina and Paraguay which are a lot poorer than Spain even though they are all-white? Racial admixture doesn’t seem a promising explanation of Argentina’s poor performance over the last century. And once we’ve established that, might the comparative poverty of white Colombians or Venezuelans, relative to Spain or Italy, reflect the same x factor that afflicts Argentina?

    In general, none of these examples answers the challenge I think I put a while back and may write up again sometime, namely: Are there any historical examples of low-IQ IMMIGRATION degrading institutions? My priors are that “founder effects” and the element of consent that occurs as part of the immigration process itself is very important. Long-time natives who are overrun by more productive foreigners are less likely to accept the rule of those foreigners and assimilate to their protocols and institutions and norms, than are people who immigrate to countries inhabited by more productive foreigners. I understand that episodes of mass migration into established productive polities may be rare enough that the scope for an empirical study is limited, which makes it attractive to look at multi-ethnic societies that have arisen in other ways. But the lessons may not cross-apply.

    Second, for these purposes, dealing with purchasing power parity concerns is crucial. I suspect a lot of things are a good deal cheaper in Brazil than in Portugal. If so, just comparing, say, the dollar incomes of Brazilian and Portuguese whites at international exchange rates would not be appropriate.

    Third, the theory behind high IQ => better institutions remains weak. One channel of causation, that low-IQ people vote for inferior policies, would seem to apply only inasmuch as countries are democratic. If the relative poverty of, say, Spanish America vis-a-vis Spain goes back to well before the time when democracy prevailed in these regions, some other theory of high IQ => better institutions would be needed.

    I should review Jones’ “Hive Mind” paper at some point. I know that IQ depends on environment at least to some extent, so simply treating it as exogenous may not be appropriate. In any case, this is very fruitful, and gives me much to look into further. It also highlights the importance of exploring the sustainability of keyhole solutions.

  4. ” If I may summarize the Jones/BK theory, it seems to be:

    1. High IQ (with race as a proxy) => better institutions => higher total factor productivity

    2. The boost to total factor productivity from raising average IQ is large enough that the global segregation created by border controls actually raises worldwide GDP.”

    Roughly, but I would add that there are also global spillovers from highly productive economies in the form of new technologies, works of art, institutions, FDI, etc. These effects provide the grist for catch-up growth by lower-IQ countries, and may be larger over time. A purely country-GDP-based analysis misses out on these factors: I have been using that to get started though, as it is more easily and rapidly testable.

    “You think Brazil/Portugal supports the theory. I looked at the figures a bit and I think it doesn’t. Brazil’s more than half as rich as Portugal and the whites there already make considerably more than the average.”

    As the links I included show, there was extensive non-Portugese European admixture in Brazil, from higher income countries such as Germany, Italy, etc. Thus my talk of weighting by source country.

    “Second, for these purposes, dealing with purchasing power parity concerns is crucial. I suspect a lot of things are a good deal cheaper in Brazil than in Portugal. If so, just comparing, say, the dollar incomes of Brazilian and Portuguese whites at international exchange rates would not be appropriate.”

    I checked PPP using the the wikipedia list (I wasn’t explicit in every paragraph above, but you can see some of the nominal vs PPP comparisons).

    “One channel of causation, that low-IQ people vote for inferior policies, would seem to apply only inasmuch as countries are democratic.”

    Dictators pay a lot of attention to the popularity of policies, even if they are more worried about revolution or internal coup than electoral defeat: consider the prevalence of fuel subsidies in authoritarian regimes. What do you think of the work of Bruce Bueno de Mesquita, by the way?

    http://en.wikipedia.org/wiki/Bruce_Bueno_de_Mesquita

    Also, from the perspective of dictators as stationary bandits, a higher IQ population probably increases the marginal return of improved institutions in growth to provide popularity (protection against revolution) and more wealth to steal.

    ” I understand that episodes of mass migration into established productive polities may be rare enough that the scope for an empirical study is limited, which makes it attractive to look at multi-ethnic societies that have arisen in other ways.”

    This does seem to be the case. It seems to me that this is a reason to have smaller-scale experiments to learn more about the effects, before taking big risks with an economy as big as the US or the EU (given the common labor market, it is tough to experiment with individual EU countries). Maybe try in Australia or Canada (they already have among the most welcoming migration policies, although these have a lot of skill/utility-based selection).

    “South Africa, perhaps surprisingly, doesn’t yield strong support to the hypothesis, as South African growth rose after apartheid.”

    Although this is confounded by the scale of sanctions, and the switch from sanctions to foreign aid at the end of apartheid (i.e. the international community invested a lot of effort to cause that shift and was underwhelmed). The South Africans were doing things like making liquid fuel from coal (at huge expense) to get around the sanctions, a nuclear program to fend off attack, and denied their best export markets. Also, I only alluded to it in the HIV policy context, but the human development index fall reflected a fall in life expectancy from 62 to 52.

    ” Racial admixture doesn’t seem a promising explanation of Argentina’s poor performance over the last century.”

    I agree Argentina is evidence in favor of other factors (Spanish colonial institutions, say), although it is not “almost all white” it has Arab, Afro-Argentinian, and self-reported “Mestizo” each in the millions, and the self-identified European population carries significant Amerindian ancestry. Average IQ estimate for Argentina in (noisy!) Lynn is in the mid 90s, significantly worse than the Argentinian-population-weighted IQ for the European source countries (Spain, but also substantial contribution from others such as Britain and Germany).

    “No recent Argentine census has included comprehensive questions on ethnicity, although numerous studies have determined that Argentines of European descent have been a majority in the country since 1914.[7]

    The term as used in this article refers to the cultural group which considers itself to be of European descent, and does not identify with any indigenous heritage. This group constitutes the majority of the Argentinian population, but a recent genetic study has shown that at least 53.7% of Argentinians have some Amerindian genetic heritage, and so the term “Argentine of European descent” must be understood as a cultural category, rather than one based purely on descent.[8]”

    http://en.wikipedia.org/wiki/Argentines_of_European_descent
    http://en.wikipedia.org/wiki/Demographics_of_argentina

    “And once we’ve established that, might the comparative poverty of white Colombians or Venezuelans, relative to Spain or Italy, reflect the same x factor that afflicts Argentina?”

    What I would like to see, and would find very convincing, is a methodologically high-quality set of econometric (higher-level “data overview”) studies testing using all the available data. This is something economists seem not to have done, probably for fear of giving offense within academia, which tends to react very negatively to IQ research, especially work with extensive relationship to race. I won’t be doing it (not my field, plus I’m anonymous), but perhaps some of the economists and mathematicians here will.

    1. Probably you would want to work with some population geneticists to get the ancestry measures right (and take things like inbreeding and uneven representation by sex in ancestry into account), and some psychometricians for the IQ data, validity, and stability.

  5. re: “I would add that there are also global spillovers from highly productive economies in the form of new technologies, works of art, institutions, FDI, etc. These effects provide the grist for catch-up growth by lower-IQ countries, and may be larger over time. A purely country-GDP-based analysis misses out on these factors.”

    But there are also spillovers from rich countries to poor countries via immigration. The most obvious example is remittances: migrants come to rich countries, earn money, and send it home, raising living standards there. Remittances as a form of external finance for development have received some attention and have desirable properties.

    Emigrants can also be crucial sources of investment and new ideas. The booming growth in India and China in recent years owes much to the Indian and Chinese diasporas. More generally, though it would be hard to prove, it seems that major migrant source countries to the United States tend to become durable democracies. Thus, 19th-century Europe sent enormous numbers of migrants to the US, then in the 20th century became democratic. Likewise, more recently, Mexico and the Philippines. In general, as Pericles said of 5th-century Athens, America is an education to the world, in both democracy and economic growth, and it is more effective when it lets in many “students.”

    Though on the other hand, disproportionate entrepreneurship among immigrants needs to be borne in mind, and it’s a reason to think migration would push the economic frontier forward faster. I think the reason for disproportionate immigrant entrepreneurship is just that throwing more ideas, more perspectives, more customs into the mix generates ideas. I don’t entirely like the meme that “ideas have sex” (long story), but I think exposing people to novelty and opening minds generates growth.

    So far, I have talked about *spreading* development. You might say that even if open borders would help to spread development, they might slow down the advance of the economic frontier. But for one thing, spreading development will tend to help the economic frontier move forward, too. It grows the market for new ideas, and it grows the number of minds who know the latest ideas and are ready to contribute. All this is admittedly rather vague. That’s the trouble with “spillovers”: almost by definition, they’re hard to track and quantify. If open borders would grow world GDP in static terms, as Clemens, Kennan, and others argue, they would probably push the economic frontier forward faster in dynamic terms as well. If, as you suggest, they would degrade institutions to the point of actually reducing world GDP in static terms, they might also slow the advance of the frontier.

    While there may not be a whole lot of examples of societies adapting to mass immigration, one big one is 19th-century America. And 19th-century America didn’t just raise a lot of people’s living standards because they got to live under better institutions than in the countries they came from. It was also super-innovative. Indeed, I think economic historians nowadays are starting to conclude that the golden age of open borders, plus another decade after it, were the most technologically progressive in the history of the world. Productivity had never grown faster, and much of the growth in the later 20th century was simply the fruits of the gradual commercialization of technologies born in that Golden/Gilded Age of economic progress: electricity, indoor plumbing, the light bulb, the automobile, the airplane, the radio. The precedent of 19th-century America seems at least as relevant as the examples of societies where a more productive racial minority is placed at a political disadvantage because of a legacy of conquest and/or because of being a historical newcomer.

    Small point: I found the Caribbean example somewhat unpersuasive because it’s located in the tropics. Worldwide, tropical regions have lower GDP. Singapore is obviously a striking exception, and I think Hong Kong is tropical too, but that’s it. The tropics are overwhelmingly poor. Of course, we don’t really know why that is. It could have to do with the disease burden, contemporary and/or historic, or the unsuitability of the advantageous middle-latitude crop packages evolved in Eurasia over millennia, a la Jared Diamond. Or maybe it’s just hard to work in hot weather. Whatever the reason, a racial/genetic explanation would have to compete with a latitude explanation. The same applies to much of Spanish America.

    On a quite different note, even in the cases where a lack of segregation arguably reduces total GDP, you seem to think, and I think too, that the withdrawal of the “market-dominant minority” would reduce living standards for the majority. It’s as if some sort of redistribution is going on, maybe because that’s literally what’s happening– the legal discrimination against the Malaysian Chinese is an example– or maybe through some sort of spillover effects. Now, even if we suppose for the sake of argument that you can raise total GDP through segregation, if you do so by making the productive minority richer and the poor majority poorer, it’s not obvious that that is a good thing. The marginal utility of a dollar tends to be higher for the poor. Total UTILITY, not an empirically observable quantity but one which can still be speculated about, may be raised, even if total GDP fell.

    This might have ramifications for the effect of segregation on the frontier, too. My impression is that the most powerful inventions haven’t catered to a small elite, but are for the common man. Thus, Henry Ford built the Model T so that any average farmer could buy it. Contrast modern health care innovation, which indeed extends and eases life for some, but is busting the budget in the process. If “frugal innovation” is more potent than expensive innovation, expanding the global middle class might be more effective for advancing the frontier than securing the wealth of a segregated elite. Admittedly, this is vague. Unfortunately, technological progress remains mysterious enough that it’s hard to talk about it with the analytical rigor or empirical groundedness one would like.

    Sorry for the length of this post, but one last point. While I’ve always preferred keyhole solutions and a certain degree of gradualism (though my scruples about deportation prevent from being too gradualist, for it’s just morally impermissible to sustain the present deportationist regime, almost regardless of what the consequences might be), this discussion has not made me more inclined “to have smaller-scale experiments to learn more about the effects, before taking big risks with an economy as big as the US or the EU.” Partly, that’s because I still have confidence in founder effects, consent-by-migration, keyhole solutions, and all that. If, to adapt asdf’s comment, it’s important that “the numerical superiority of [the American-born plus those assimilated to American culture/norms/protocols/productivity levels] must be maintained” lest “there… be a shift in the economy, [and a loss of] both the economic performance and the political backdrop which makes that performance possible,” then I think that numerical superiority (probably of resident population, more probably of voting citizens) can be maintained under open borders. The risk of a sort of institutional upheaval ending in a bad equilibrium– say, with a lot more corruption, tribalism, and populism– exists, but I don’t think it’s very large. But, also, with the exception of Zimbabwe/Rhodesia, the societies that BK uses as examplesare not such bad places. I suppose I’d rather be an American than a South African or Brazilian white, but South African and Brazilian whites don’t have it SO bad. South Africa represents a quite implausible worst-case scenario with open borders. One has to suppose that enough immigrants would flood in to outnumber the existing American population several-fold, drawn from the most impoverished populations on earth, with a certain solidarity against the native-born American population, and would take control of political institutions, leaving the native-born American population fairly impotent politically. Since the keyhole solution in Principles would certainly prevent that, one has to suppose either that pure open borders is adopted in a quite abrupt, unmediated fashion, or else that the keyhole solution for some reason proves unworkable. And even in THAT case, what you get is the situation of South African whites or Malaysian Chinese, a bit under siege, maybe hiring private security agencies, with little political voice, facing some legal discrimination, but still reasonably prosperous and able to maintain their ways of life. If that’s the worst-case scenario, while the upside is we’re likely to double world GDP, accelerating progress at the economic frontier, and eliminate extreme poverty worldwide… yeah, it’s a risk well worth taking.

  6. ” while the upside is we’re likely to double world GDP, accelerating progress at the economic frontier, and eliminate extreme poverty worldwide”

    The upside is very attractive.

    ” I think that numerical superiority (probably of resident population, more probably of voting citizens) can be maintained under open borders”

    I would enjoy seeing more detailed analysis. Especially taking into account the following (some of which I have explored in other comments):

    -income differentials
    -differential age structures by region
    -population trends that are significantly changing world demographics (and are clearest in the young and most mobile generations)
    -demonstration effects (friend migrates, sends word, so I follow, even if I am not one of the hundreds of millions who reported readiness to migrate in that survey)

    “But there are also spillovers from rich countries to poor countries via immigration. The most obvious example is remittances: migrants come to rich countries, earn money, and send it home, raising living standards there. Remittances as a form of external finance for development have received some attention and have desirable properties.”

    That is counted in GDP, at least as far as efficiency goes: if GDP and growth are raised by political separation, then trade, aid, and migration policies that don’t much risk slippage into political integration can do better. As they say, “growth isn’t everything, but in the long run it’s almost everything.”

    “The booming growth in India and China in recent years owes much to the Indian and Chinese diasporas”

    Yep. Although note that those diasporas were relatively selected: going through universities, the Canadian point-based immigration system, and so forth. They came back as investors, managers, representatives for FDI.

    “I don’t entirely like the meme that “ideas have sex” (long story)”

    I’m familiar with Romer’s work.

    “While there may not be a whole lot of examples of societies adapting to mass immigration, one big one is 19th-century America.”

    America had lots of cheap land (enabling fast population growth), and good institutions so the migrants were coming to it from Europe rather than the other way around. It got to have a huge share of the European-descended population. Why wouldn’t it be a locus of much of the innovation? The migration was from sources without very big cognitive or personality differences and high human capital. I guess I don’t get much from this example.

    “even in the cases where a lack of segregation arguably reduces total GDP, you seem to think, and I think too, that the withdrawal of the “market-dominant minority” would reduce living standards for the majority”

    I do. And it’s a serious cost (although the assumption of incomes falling to near-zero is an overestimate for analytical convenience).

    “Total UTILITY…total GDP…It’s as if some sort of redistribution is going on, maybe because that’s literally what’s happening– the legal discrimination against the Malaysian Chinese is an example– or maybe through some sort of spillover effects.”

    I think that if one doesn’t care much about long run impacts or future generations then open borders with citizenship is clearly far better,than the status quo. However, in the long run future generations will outnumber the current poor by many orders of magnitude, and from that point of view good stewardship of the safe growth of our civilization is most important. GDP and institutional quality weighted by power look like better proxies there.

    “powerful inventions…Admittedly, this is vague”

    I don’t think I buy the story, but it doesn’t seem worth arguing over much here.

    “And even in THAT case, what you get is the situation of South African whites or Malaysian Chinese, a bit under siege, maybe hiring private security agencies, with little political voice, facing some legal discrimination, but still reasonably prosperous and able to maintain their ways of life”

    I agree that totally open borders with political integration wouldn’t reduce high-performing populations to poverty. But I tend to think that growth and good government’s most important effects aren’t on the current generation or in the short-run.

    ” develop solidarity among themselves,”

    Class-based solidarity as poor disenfranchised migrants, while citizens get huge government benefits seems pretty plausible. And we do see in the U.S. and elsewhere reasonably successful coalition building along the lines of “unite to soak the rich, who happen to be visually distinguishable with high accuracy from the groups in the coalition.” It’s much less virulent than it could be because those coalitions also need to appeal to at least part of a market-dominant majority, and there are more tensions in an “anti-X” coalition than a “pro-Y, anti-X coalition”.

    “Sorry for the length of this post”

    :) I have been posting multiple epic comments per day on this blog, and read through the great majority of the archives. These are important issues, and I care a lot about having them clearly hashed out, especially among activists who are going to be working on them. I certainly don’t begrudge reading long responses to my long comments.

    1. Allow me to weigh in here on the point with 19th century America.

      “America had lots of cheap land (enabling fast population growth), and good institutions so the migrants were coming to it from Europe rather than the other way around.”

      Compared to Europe (and indeed most of the world) America still has lots of cheap land. The population density of the United States is still only at 91 people per square mile, which is less than the world average and less than the average of most sending countries. Indeed comparing house prices to income, the United States has one of the lowest ratios. Furthermore, the United States still has some of the strongest institutions in the world (which are currently declining because of native mismanagement, but are still some of the best in the world). So in that regard these are not differences with the 19th century.

      “It got to have a huge share of the European-descended population. Why wouldn’t it be a locus of much of the innovation? The migration was from sources without very big cognitive or personality differences and high human capital. I guess I don’t get much from this example.”

      This point is even weaker than the above in my opinion. On what basis are you asserting that European immigrants in the late 19th century were homogeneous and had high human capital? Literacy rates for Southern Italians, as an example, were under 16% in the late 19th century. Other countries in Eastern Europe sending immigrants had similarly low rates. As for not very large personality differences (I’m unsure on cognitive differences as to my knowledge no one was measuring IQ back then, but it would shock me if there were few differences between Northern Europeans with their greater literacy and more schooling than Southern and Eastern Europeans), the nativist complaints about Southern and Eastern Europeans would beg to differ. And indeed, in the sense that there was a real difference the nativists were almost certainly right. The cultures of Northern, Southern, and Eastern Europe are all very distinct. The types of governments they lived under, the degree of property rights they came from, the religions they followed (yes almost all were Christian but there are very significant differences between Catholic, Protestant, and Orthodox faiths).

      19th century workers were low skill, diverse, and coming into a country with very different institutions and culture from their home countries. There was no qualitative difference in the kind of immigration the US was receiving then and the kind you seem to be worried about receiving now. For any quantitative differences we need to see data (some of which just simply did not exist at the time admittedly) and then we need to question whether the quantitative differences are significant enough to warrant concern. I think in some ways the differences between migrants and US natives might even be less than the 19th century given the rise of a many shared aspects of global culture through the partial globalization we’re currently engaged in (not to mention that literacy rates around the globe are universally superior to the under 16% rate I cited for Southern Italians earlier).

      1. OK I’m still getting the hang of this system so it’s not allowing me edit to add in links. Instead I’ll just post them here.

        For the assertion of American population density: http://en.wikipedia.org/wiki/List_of_sovereign_states_and_dependent_territories_by_population_density

        For the assertion of American low property price to wage rates: http://www.numbeo.com/property-investment/rankings_by_country.jsp

        For the assertion about American institutions: http://www.heritage.org/index/ranking

        For the point about Southern Italian literacy: http://region-developpement.univ-tln.fr/fr/pdf/R33/Gagliardi.pdf

        Finally, for the point about current world literacy rates: http://en.wikipedia.org/wiki/List_of_countries_by_literacy_rate

      2. Good points, Chris. I will add that from a historical perspective, one reason I don’t find the criminality arguments persuasive is because all evidence on the linkage between crime and immigration in early 20th century America (in the run-up to the closing of the borders) suggests that the situation then was just as bad if not worse than it is now.

        Modern studies suggest that immigrants have lower levels of criminality than natives, but their descendants assimilate to native levels of crime (some would emphasise native levels of crime for one’s ethnicity). But most studies which look at late 19th or early 20th century immigration to the US find the exact same pattern. And some scholars suggest these studies are faulty, finding the rough opposite conclusion: that immigrants from this period had higher levels of criminality, and their descendants assimilated to lower native levels of crime. (I think Carolyn Moehling and Anne Morrison Piehl are the leading scholars making this argument.)

        I would add that almost all studies of modern immigration and crime that I’m aware of, in almost any locality, have similar findings — immigrants are less likely to commit crimes, but their descendants assimilate to higher, native levels of criminality. This seems like a strikingly robust conclusion to me, so in absence of concrete data on this from the 19th century, the heyday of open borders, it seems logical to assume that a similar pattern held true.

        If states could have open borders for over a century without falling apart (and, instead, generally progressing quite remarkably), it’s not clear to me why the “Immigrants/immigrants’ descendants are criminals/lazy/parasites/low IQ” argument should work well, because the exact same arguments held just as true in the heyday of open borders. Even if one grants that there are some difficult-to-estimate social costs from high levels of immigration (and I suspect virtually all of us would agree that there are, though we might dispute typical beliefs about these costs), it’s not clear from the evidence that these costs are high enough to generate the apocalyptic scenarios which tend to come up in these debates.

Leave a Reply