Last fall I wrote two posts, Open borders and the economic frontier, part 1, and Open borders and the economic frontier, part 2, which were meant to be the first two parts of a three-part series, which sought to address the large macroeconomic question of whether open borders would cause the global economic frontier to move forward faster or slower. BK had made a loose empirical argument in the comments that the more productive ethnic groups seem to be most productive when they are in the majority, less so when they live as minorities among other ethnicities. In part 1, I gave the following reasons why I was skeptical of BK’s empirical analysis: (a) the weakness of the theoretical links between race, cognition, and GDP; (b) the fact that the poor/racially disadvantaged countries are concentrated in the tropics; (c) the pattern that the racial minorities BK’s argument focused on tend to have arrived under imperialist auspices as colonizers or settlers and are therefore particularly alienated from the population; and (d) the prevailing view in our culture that “racism” is scientifically erroneous. In part 2, I explained how, in my own theoretical work, growth is modeled as “exploration of the goods space” through an expansion of the “endogenous division of labor” as capital or population expands, and applied it to the data BK had highlighted. Using the EDOL model, I derived a new theory of relative growth and decline in the 20th century, namely, that the automotive revolution altered economic geography, reducing the economic distance between all manner of points connected by land, while increasing the distance between points separated by the sea: thus Britain and its empire suffered, while the United States and contintental Europe were big gainers, and the Soviet Union also benefited.

But let me step back a bit. Economic growth is very important because, in the long run, small changes in growth rates can alter the human condition enormously. If we can raise the long-run growth rate of per capita income merely from, say, 2% to 3%, that will make our grandkids 70 years from now twice as rich as they would have been if the slower growth trajectory had been sustained. In 700 years, this small change would make our descendants over a thousand times richer than they would have been. Lately, this important topic has become the subject of a large literature, mainly in economics though it spills over to other fields as well. I got a Masters in International Development and a PhD in economics and I’ve been exposed to a lot of this literature, yet I nevertheless feel inadequate to the task of summarizing it. Still, here’s a very rough typology of theories.

A few major theories in development economics

Policy

Some theories impute growth differences to differences in policy. In particular, there is a large literature on openness and growth, in which Sachs and Warner (1995) (ungated link) and Dollar and Kraay (2004) (ungated link) are among the more important contributions. Another policy explanation is that communism doesn’t work, and countries that got stuck under communist rule in the late 20th century ended up a lot poorer than countries that didn’t. Short of communism, large-scale government ownership of private enterprise tends to be bad for growth, as (arguably) do market distortions and big government generally. The Washingon Consensus that had a dominant place in development policy advice during the 1990s was a somewhat one-size-fits-all set of policy prescriptions by which development economists at the World Bank, IMF, and elsewhere tended to try to help developing countries grow.

Institutions

It’s a bit difficult to draw the line between policy and institutional explanations of the wealth and poverty of nations, but institutional explanations tend to dig a bit deeper. Whether to set interest rates or allow foreign trade are policy issues. Corruption and property rights are institutional issues. Policy and institutions both have to do with the behavior of governments, but one way to understand the difference is: if some policymaker can straightforwardly change it by fiat, it’s policy; if not, it’s institutions. Corruption can’t be changed by fiat because it’s against the law by definition, but its enforcement is constrained by a “who guards the guards?” problem. Corruption and clean government can both be equilibria. If a cop in the US doesn’t demand bribes, that’s as likely as not because he doesn’t think he can get away with it, and US citizens rarely offer bribes for the same reason. Societies might flip at random from one equilibrium to the other, as Hammond (2009) suggests. Good property rights enforcement requires judges to be not only honest but sometimes sophisticated, and it requires a credible long-term commitment on the part of the state not to expropriate legitimate owners.

Leading advocates of institutions as a theory of comparative development include Doug North, Dani Rodrik, and Daron Acemoglu. Founding figures in the study of economic institutions include Ronald Coase and Oliver Williamson. A crude ranking of these thinkers in terms of subtlety: Coase = Williamson > North > Rodrik > Acemoglu. I regard Acemoglu’s prominence in the field, in particular, as distressing. I was drawn to economics for two forms of rigor, and “rigor” is almost another word for “intellectual honesty,” that pervaded the field: (a) methodological individualism in theorizing; and (b) high econometric standards in empirical work. Acemoglu’s Why Nations Fail is destitute of both. To put the point bluntly and without excessive charity, Acemoglu wants to explain the wealth and poverty of nations as a function of democracy, blaming poverty on greedy elites that set up bad institutions from which democracy can save us, but since this story doesn’t really make sense (and so can’t be formulated as a rigorous theory) and isn’t supported by the data (Barro (1996) crunched the numbers and found that once suitable controls were in place “the overall effect of democracy on growth is weakly negative), he resorts instead to a kind of rambling and tendentious historical travelogue in order to make his case plausible by the accumulation of anecdotes. Why Nations Fail is just the sort of thing I went into economics (as opposed to history or other social sciences) to get away from. The actual economic content of “institutions” for Acemoglu consists mainly of property rights; for Rodrik, it includes public institutions able to deal with all manner of “market failures”; and in Doug North, the focus is on high transactions costs and the role of institutions in reducing them. Coase probed deeper questions of what property rights are and why there are firms, and Williamson shed more light on bargaining theory, long-term relationships versus spot transactions, and determinants of the size of firms and the structure of contracts.

Probably if you asked development economists for a one-word explanation of the wealth and poverty of nations, “institutions” would be the most common answer. But this is so for bad reasons. “Institutions” are (a) a vague category, and (b) not really economists’ forte. Whereas the Solow model has almost unanimously been taken to show (though I dissent) why capital accumulation can’t explain long-run growth, institutional theories of growth have never been made clear enough to be clearly tested and refuted.

Geography

Jeffrey Sachs is a champion of geography as an explanation of growth. From the abstract of Gallup, Sachs, and Mellinger (1999):

Location and climate have large effects on income levels and income growth through their effects on transport costs, disease burdens, and agricultural productivity, among other channels. Geography also seems to affect economic policy choices… At particular disadvantage are regions located far from coasts and ocean-navigable rivers, for which the transport costs of international trade are high, and tropical regions, which bear a heavy burden of disease.

In particular, poverty in sub-Saharan Africa, according to Sachs, is largely a function of (a) malaria, and to a lesser extent AIDS and other diseases, to which Africa is specially vulnerable, and (b) a not very sinuous coastline (in contrast to Europe and the Pacific Rim of Asia), lack of navigable rivers, and concentration of population in the interior, which make Africa’s population particularly poor in water transport links to the rest of the world.

Also, Jared Diamond’s Guns, Germs, and Steel, one of the most important books in development economics even though Diamond is not an economist, argues that the Eurasian landmass enjoys a long-standing lead in development because of the abundance of domesticable plants and animals, and the East-West orientation of the continent that helps them to spread (e.g., crops can move from China to Europe, since the climates are similar, but not from Aztec Mexico to Inca Peru). Since the Age of Discovery, Europe’s biological package of grains and domestic animals has spread to Australia, North America, the southern cone of South America, and other temperate regions, but not to the tropics, since wheat, say, or horses, aren’t well suited to tropical climates. As odd as it may seem to give this argument from biological endowments prominent mention a hundred years after the automobile replaced the horse, I think this is the strongest contender to the striking tendency for economic development to rise with distance from the equator. Obviously, it would need to be supplemented with a story about why development advantages from a hundred years ago have been so persistent. That said, disease burdens also tend to be higher in the tropics.

Culture

Development economists, whose natural intellectual posture is that of giving advice to policymakers, and who sometimes work for the World Bank or IMF and are really sitting across the table from people in ministers and giving them policy advice, have a bias in favor of policy explanations. To some extent, institutional and geographical theories of comparative development start to get favored when good policies don’t work. There was a global shift towards privatization, markets, and greater trade openness in the 1980s and 1990s. For the most part, it didn’t pay off the way economic advisers expected at the time. Easterly (2001) conveys well how frustrating development economics seemed fifteen years ago. Since then, growth in the developing world has started looking quite a bit better, and by now, I think it might be time to revisit the disillusionment with the “Washington Consensus” that prevailed a decade ago and get disillusioned with it. Maybe privatization, markets, and greater trade openness worked, after all, albeit with a bit of a lag. Still, policy is far from a complete explanation of the wealth and poverty of nations and needs to be supplement by deeper theories. Such as institutions. And geography. And also: culture.

Acemoglu dismisses culture as a theory of comparative development (and geography, too) in Why Nations Fail by pointing out that North Korea and South Korea have the same culture (and geography) yet wildly different economic outcomes. What a stupid argument! Obviously, growth can be crushed even in the most promising of circumstances by a ruthless totalitarian regime. It does not follow that, given the same kind of regime, culturally (or geographically) different regions would exhibit the same economic performance. From my experience with development economists, I have a sense that a certain self-censorship is practiced whereby cultural explanations are downplayed because they would be offensive. Racial explanations are a little different because I think most development economists are not even tempted to believe them: yes, racism is that dead. Cultural explanations, they are tempted to believe, and probably do believe at some level, but it’s not appropriate to say so. Of course, that’s both guesswork and overgeneralization.

Anyway, Deirdre McCloskey’s The Bourgeois Virtues and Bourgeois Dignity: Why Economics Can’t Explain the Modern World come to mind as sort of qualifying as a cultural theory of comparative development. Doubtless, there are more, and perhaps my policy-centered education and work in development gave me a bias against hearing them… but I still think culture is insufficiently studied in this regard. The subtitle of the latter suggests another reason why economists might be biased against cultural theories of development: it puts them out of a job, though institutional theories also tend to do that, suggesting that lawyers or political scientists might be better than economists at explaining the wealth and poverty of nations. Rodney Stark, in The Victory of Reason: How Christianity Led to Freedom, Capitalism, and Western Success and For the Glory of God: How Monotheism Led to Reformations, Science, Witch-Hunts, and the End of Slavery and elsewhere, makes a strong case that Christianity explains most of the distinctive Western traits that led to Western success, but development economists don’t seem to have noticed. Probably there are more cultural theories of comparative development out there but I’m drawing a blank.

Cultural explanations can be hard to distinguish from institutional explanations, on the one hand, and racial/genetic explanations on the other.

Race and genetics

BK, who knows more about this topic than I do, linked to Jason Collins’ blog Evolving Economics, which seems to focus on the intersection between evolutionary psychology and economic development. From the reading list compiled by Collins, important contributions include Greg Clark’s A Farewell to Alms, which argues that the English people starting in the Middle Ages were genetically altered through a pattern whereby the prosperous had more children, and what one might call bourgeois genes– low present bias in time preference, low propensity for violence– filtered down through the population, and Galor and Moav’s “The ‘Out of Africa’ Hypothesis, Human Genetic Diversity, and Comparative Economic Development,” which argues that genetic diversity helps development by endowing society with diverse capacities but hinders it through creating more conflict, so that regions with too much genetic diversity, such as Africa, and regions with too little, such as some Andean countries, do worse than countries with a happy medium of genetic diversity, such as the United States. In addition, there is the work of Garett Jones in articles such as “National IQ and National Productivity: The Hive Mind Across Asia.”

BK’s latest comments on part 1 highlight some new literature related to genetic explanations which I’ll respond to at the end of the post. Let me note here, though, that one thing I like about the genetic and geographic theories vis-a-vis the cultural, institutional, and policy theories is that they really try to get to the bottom of the chain of causation. If you ask questions like “Does democracy cause growth?” or “Does education cause growth?” you quickly run into problems of multi-directional causation. It’s plausible, prima facie, that democracy and education are causes of growth, but it’s also very plausible that when people get more prosperous, democracy and education are among the things that they demand. So even if there is a clear correlation, you have to strive mightily to establish the direction of causation. Geography and genes seem to undercut reverse causation critiques. If tropical countries are poor, that can’t be because their poverty caused them to drift into the tropics. If black people are poor, that can’t be because their poverty caused them to turn black. Yet subtler reverse causation is possible. Maybe more developed peoples– richer, more technologically advanced– seized all the best lands by superior strength, or knew enough to migrate to the places that were non-obviously most conducive to economic prosperity. Maybe more developed peoples practiced a form of exogamy cum monogamy cum openness that tended to produce a certain degree of genetic diversity, while polygyny cum tribalism produced different patterns of genetic diversity elsewhere. If so, correlations between genetics or geography and economic development would not be causal, and the conclusion that people who have genes or live in places historically associated with poverty are doomed to stay poor, would be mistaken. Also, genes or geography may be proxying for other things, including for each other. Perhaps people of European stock are rich for genetic reasons, and happen to like cool climates for cultural reasons, creating a misleading correlation between latitude and development. Perhaps people of European stock are rich because they live in cooler climates, creating a misleading correlation between European genetic traits and development. Fundamentally, geographical and genetic theories of comparative development are logically weak. If they are persuasive, it is because they fit the facts fairly well and seem to escape endogeneity critiques, but theory doesn’t pin down the actual causal story well at all.

Now, by way of summary, let me offer a critique of development economics that has perhaps already occurred to the reader. Most of these theories seem to involve economists writing about non-economics. Development economics explains the wealth and poverty of nations by first theorizing about genetics, or intelligence and cognition, or epidemiology, or climatology, or animal husbandry and horticulture, or legal systems, or intellectual history, and then offering rather loose and speculative stories linking their favorite non-economic aspect of human behavior and environment to the data about the distribution of global GDP. Economists have probably gotten more educated about history, law, biology, and other relevant fields, but they’re still basically amateurs. Acemoglu’s Why Nations Fail– that book is a pretty good summary of the failings of development economics– does, and doesn’t, exhibit an impressive grasp of history: impressive for an economist, yes, but not impressive for a world-historian. Why Nations Fail is greatly inferior to, say, Toynbee’s A Study of History, since Toynbee knows history much better than Acemoglu, and while Acemoglu knows more economics and statistics, he hardly makes any use of it in Why Nations Fail. Indeed, Acemoglu hardly seems to be aware of the public choice literature that systematically studies the workings of democracy using an economics toolkit and has achieved some real insight about it. That Jeffrey Sachs, who has done some of the best empirical work in development economics, has written a book called The Price of Civilization: Reawakening American Virtue and Prosperity, is frankly embarrassing. Is Sachs an expert on virtue? Not even close. Development economics has drifted away from the study of the things that economists really do more about than other people do, things like markets and prices and money and contracts and utility functions and competition.

This is odd, and requires explanation, and the explanation is that since Solow (1956), economists have largely believed that they couldn’t explain comparative development using their own tools. Deirdre McCloskey’s subtitle sums it up: “why economics can’t explain the modern world.” The Solow model supposedly proves that capital accumulation can’t explain long-run growth, because of (a) diminishing returns to capital, and (b) depreciation. The Solow model starts by assuming a constant returns production function: double capital and labor, and you double your output. This is an odd starting place since it’s common sense that production in bulk is usually more efficient, and economists ought to know this especially well since Adam Smith expressed it so brilliantly in the first chapters of The Wealth of Nations. The forgetting of this insight, at least in the inner sanctum of economic theory, was a curious side-effect of the refinement of the theory of (“perfectly”) competitive markets, which turned out to be inconsistent with increasing returns. Because the wealth and poverty of nations can’t be a function of capital accumulation, it must be a mysterious something which is usually assumed to be technology, as eventually captured in Romer (1991)‘s theory of “endogenous technological change.”

In my seemingly unpublishable theory most recently expressed in a paper titled “The Aggregate Production Function with Endogenous Division of Labor,” I attack the basic assumption of the Solow model directly. Using an agent-based simulation to get round the deep errors in the “orthodox” theory of perfect competition, I show how increasing returns in the aggregate production function can arise from the way growth of the capital stock, or the population, induces “exploration of the goods space,” i.e., the introduction of new goods that were already technologically available but not economically viable without sufficient capital and/or population. The EDOL model hasn’t changed that much since part 2, so I won’t explain it again, but the model makes economic development a good deal less mysterious. Deirdre McCloskey was mistaken in claiming that “economics can’t explain the modern world.” In a way, the EDOL model even suggests that, misled by Solow (1956), economists were mistaken in searching so far and wide for explanations of the wealth and poverty of nations. But that’s not to deny that they did a lot of good work and found out a lot of useful things. For the EDOL model doesn’t overturn all the theories above. It tends to ratify and further elucidate them.

For example, take institutions. The institutional literature is somewhat sprawling and vague, with leading figures like North, Rodrik, and Acemoglu having quite different agendas, and never even making it clear where and how they agree and disagree with each other. The EDOL model suggests how the institutional literature can be given a definite task. For institutions affect market size. If the legal system, combined with a law-abiding culture, provides an effective backstop to contracts of all kinds, explicit and tacit, then the set of people with whom one can do business is very large. If the legal system is useless you can only do business with people close enough to you to be trusted. Family and clan become more important, compromising individual liberty and narrowing markets, and lots of efficient specialization and trade fails to occur. Institutions also affect capital accumulation, since long-term investment requires pooling of resources and is greatly encouraged by firm security of property rights over extended periods of time against both direct and indirect expropriation (e.g., by unions or regulation).

Again, geographical theories of comparative development are helped in several different ways by the EDOL model. First, even small differences in environmental endowments, resulting in small surpluses, can spill over into capital accumulation, and growth rates compound. The Solow model would deny this, alleging that capital accumulation would run up against the limits of depreciation and diminishing returns, but the EDOL model gives a warrant to production functions for which the Solow ceiling is inapplicable. Second, and probably more importantly, geography can affect the size of networks of specialization and trade. Disease burdens do damage not only by reducing labor productivity directly, but by deterring people from forming cities, and access to the sea, as well as policy openness, matter because– this has long been recognized but the EDOL model gives it a better theoretical warrant than it has hitherto enjoyed– it allows people to plug into a global division of labor. Cultural factors may also encourage capital accumulation and market size. What matters is not only bourgeois inventiveness but bourgeois honesty in business, which facilitates doing business with strangers and thus enlarges trading networks, and bourgeois prudence and forethought, planning and saving for the future.

With all this in mind, let me respond to a few of BK’s quotes.

Acemoglu, Johnson, and Robinson’s results do not establish a role for institutions. Specifically, the Europeans who settled in the New World may have brought with them not so much their institutions, but themselves, that is, their human capital. This theoretical ambiguity is consistent with the empirical evidence as well.

Background: The Acemoglu et al. paper referred to here uses 17th-century settler mortality as an “instrumental variable” to explain modern development outcomes. This is one of the worst examples I know of far-fetched interpretation of an interesting econometric result, and I’m glad someone has taken them to task on it. Yes, if 17th-century settler mortality has proven to be a predictor of modern economic development, that’s because Europeans imported themselves and their human capital to places where they can live. And if, under open borders, Africans were able to import themselves and their generally low levels of human capital to temperate, rich countries, they would probably be less productive than the natives of those countries and lower average GDP per capita there; but it does not follow that the natives would be any worse off. They might enjoy large gains from living under better institutions and in more favorable locations, without harming natives on average, or even more likely, while helping the natives on average by providing cheap labor as domestic servants, drivers, factory workers, waiters and so forth. Gains from trade. To make African migration to the US a net negative for US natives, you’d have to have a story about how they greatly reduce “productivity.” Thinking about development through the lens of my EDOL model, I see no reason to expect that. Nor am I aware of any historical precedents for peaceful immigrants reducing productivity anywhere in the world. Another one:

“Michalopoulos and Papaioannou (2010) find that national institutions have little effect when one looks at the economic performance of homogeneous ethnic groups divided by national borders. …

Overall, their findings suggest that long-term features of populations, rather than institutions in isolation, play a central role in explaining comparative economic success.”

Hmm. This one seems to contradict things that I already know, e.g., that immigrants raise their wages a lot by migrating to rich countries. If Michalopoulos and Papaioannou merely mean that the relative performance of ethnic groups is persistent across states of migration, that would seem to have no ramifications for or against open borders. Another one:

“Putterman and Weil’s results strongly suggest that the ultimate drivers of development cannot be fully disembodied from characteristics of human populations. When migrating to the New World, populations brought with them traits that carried the seeds of their economic performance. This stands in contrast to views emphasizing the direct effects of geography or the direct effects of institutions, for both of these characteristics could, in principle, operate irrespective of the population to which they apply. A population’s long familiarity with certain types of institutions, human capital, norms of behavior or more broadly culture seems important to account for comparative development.”

Easterly and Levine (2012) confirm and expand upon Putterman and Weil’s finding, showing that a large population of European ancestry confers a strong advantage in development, using new data on European settlement during colonization and its historical determinants. They find that the share of the European population in colonial times has a large and significant impact on income per capita today, even when eliminating Neo-European countries and restricting the sample to countries where the European share is less than 15 percent—that is, in non-settler colonies, with crops and germs associated with bad institutions.

Again, I’m not sure whether these are stories about institutions being determined by the organic cultural traits of peoples, or stories about founder effects. If it’s about founder effects, it has no negative ramifications for open borders. If it’s about the organic cultural traits of peoples, it seems to contradict other things I know, e.g., the lack of any institutional debt of the United States to Africa or Germany. BK cites another source which seems to focus on explaining inequality by looking at the characteristics of pre-1500 populations. But I’m not really much concerned about inequality among citizens or residents of a given country, as opposed to global inequality, which I do care about.

I do tend to believe that development has deep roots, and that a culture => institutions causal link is important. But there is also an institutions => culture causal link: German-Americans are much more American than German. More fundamentally, I think the more mysterious factors in economic development have been considerably exaggerated. Much of American prosperity is just about having lots of capital and human capital and lots of people and a complex division of labor because of large cities and good infrastructure so that it can exploit the technology space, and immigrants won’t damage any of that. On the contrary, they’ll raise the return on capital and thereby encourage capital formation, while making cities bigger yet. To the extent that culture and institutions are important, I’ve yet to see evidence that founder effects can be overwhelmed by changing demography.