Response to Paul Collier: Chapter 1

Post by Nathan Smith

Paul Collier’s Exodus: How Migration is Changing Our World is probably the best book on migration from the restrictionist side that currently exists. Though, that is not saying much. It is pretty strong on the economics, and while I find Collier’s ethical attitudes weird, repugnant, and indefensible, they serve as a useful window on the way a lot of people think. Exodus is a refreshing contrast to books like Victor Davis Hanson’s Mexifornia: A State of Becoming. The arguments in Hanson’s book are too thoroughly flawed to be answered. You’d have to rip them to shreds, almost sentence by sentence, to avoid leaving the impression that anything in them is valid. Any reader who would be a worthy interlocutor in a learned conversation would have seen through books like these. My advice to writers like Hanson is to read Collier’s book and spend a couple of weeks contemplating its intellectual merits, and then ask themselves seriously whether they can emulate them sufficiently that their future writings will be net positive contributions to public debate. If Collier sets the standard that future restrictionist writings will be expected to live up to, the quality of public discourse about immigration will be vastly improved.

Interestingly, Exodus is responding in part to open borders as a political cause, even if it’s a cause that his implicit interlocutors don’t usually embrace explicitly. Whereas others will speak loosely of “the open borders lobby” as an epithet to characterize mainstream people who, in fact, want a lot of immigration restriction, Collier is a development economist who has some idea what real open borders would mean, and knows that there is a case for it. He seems to know about the double world GDP literature. So far, the debate has been conducted within the restrictionist end of the spectrum, with advocates of more migration sometimes mistaking themselves for open borders advocates because they’re naïve about how radical open borders really is. Collier thinks about migration in the context of the global struggle against poverty. He doesn’t pretend the rest of the world isn’t there. He doesn’t adopt a principle of moral indifference to the rest of mankind. That’s a big improvement over previous restrictionist literature.

At present, then, Exodus is the argument to beat on the question of open borders. For that reason, I thought it deserved, not just a book review, but a thoroughgoing engagement with the argument. That said, Collier gave me very little reason to change my mind about supporting open borders, though he might have convinced me to shift my position on a few aspects of the question in subtle ways. There are two main reasons that Collier is unconvincing. First, he has the wrong ethics: he knows about “utilitarian universalism” but is constantly engaged in inadequately motivated attempts to substitute manifestly inferior ethical ideas. Second, his policy imagination is very deficient. My greatest regret is that Collier doesn’t engage with DRITI. Again and again, I found myself saying, “Yes, that’s a problem, but DRITI solves it.”

Chapter 1 sets the stage for Collier’s book with a lot of reflections on the peculiar character of the public debate about immigration. For example, he writes that…

In the liberal policy circles that on most policy issues provide the most informed discussion, migration has been a taboo subject. The only permissible opinion has been to bemoan popular antipathy to it.

This gives us a glimpse of who Collier’s imagined interlocutors are: a policy elite, enlightened in their own eyes, not daring to think the immigration through, but feeling superior for disdaining anti-immigration attitudes among the people. Of course, that’s not me, nor, I assume, will many readers of Open Borders: The Case fit the description. Consequently, one has the feeling of eavesdropping on one side of a conversation between two groups of wrong-headed people. Collier promises to present evidence but says his “more ambitious aim is to induce people to reexamine the inferences they draw from their values,” but I think it would be truer to say that thinking about immigration tends to expose conflicts and contradictions within people’s scheme of values, e.g., between their desires to help the poor and to promote freedom, and to achieve domestic equality and avoid seeing poverty on the streets. Collier could identify different conflicts and contradictions. For example, when Collier writes…

Often politicians talk tough and act soft, and more rarely the opposite. Indeed, sometimes they appear to be embarrassed by the preferences of their citizens.

… he may be identifying a contradiction between left-liberals’ commitments to (a) openness to immigration, and (b) democracy and respect for the will of the people. But the conflict could be resolved in different ways: by yielding to popular demands for greater immigration control, or by deciding that there are, after all, moral principles that transcend the will of the people and should be obeyed even in defiance of a democratic majority. By the way, against the claim that politicians talk tough and act soft, the Obama Administration stands as an enormous exception: it poses as pro-immigration, yet is on track to deport 2 million people.

Collier is not a “citizenist,” indifferent to world poverty. He writes…

There is a clear moral obligation to help very poor people who live in other countries, and allowing some of them to move to rich societies is one way of helping. Yet the obligation to help the poor cannot imply a generalized obligation to permit free movement of people across borders. Indeed, the people who believe that poor people should be free to move to rich countries would likely be the first to oppose the right of rich people to move to poor countries: that has uncomfortable echoes of colonialism.

Here a promising argument gets nipped in the bud by an attempt by Collier to get his imagined interlocutors on his side. Again, he’s not talking to me. I would assert freedom of migration in both directions: from poor to rich, and from rich to poor countries. My view of colonialism is much more nuanced than blanket condemnation: for starters, colonialism bears little of the blame for the Democratic Republic of the Congo, and deserves much of the credit for Singapore. Anyway, the problem with colonialism wasn’t with Europeans moving to, but with Europeans ruling, and in particular with Europeans often ruling unjustly, large swaths of Africa, Asia, and the Americas.

It’s still true, as Collier says, that “the obligation of the rich to help the poor, and the rights of freedom of movement between countries” are distinct issues. That said, Collier’s claim that “there are many ways of fulfilling our obligation to the poor” is questionable. Foreign aid has a mixed record. Sometimes it works well, but it seems highly doubtful that it could be scaled up enough to achieve the amount of poverty alleviation that open borders promises. Anyway, if we’re not doing as well as we should be in the struggle against world poverty, a claim that we could do more seems like a poor excuse to close the borders. It’s as if, when you tell me I should spend less on beer so as to stop running up my credit card debts, I say that I don’t really have to do that, because I can get a second job, and then I don’t get a second job.

The abstract for Exodus at contains the following sentence: “Ultimately the danger is that both host and countries of origin may lose their national identities– an outcome that would be disastrous, Collier argues, as national identity remains a powerful force for good.” This claim intrigued me. I tend to feel that explicit, emotive nationalism is mildly politically incorrect among the intelligentsia, both on the political left and on the libertarian right, but that there must be a stronger case for nationalism than one hears in the public arena. I have an inkling of what that case is, myself, such as the following half-formed thoughts. Democracy requires a loyal opposition, and it is linked to nationalism because the nation gives the loyal opposition something to be loyal too. Democracy seems unworkable in theory because of cycling and Arrow’s impossibility theorem, but it works in practice because of altruism among voters that is related to nationalism. Or, democracy seems unworkable because of rational voter ignorance, but it works in practice because people care about their native country and get emotionally involved in its history, which motivates them to stay abreast of politics, spread the world, proselytize for political views, and generally, to devote private resources to sustaining healthy democratic dialog. I was hoping Collier would expand on these inchoate, suggestive thoughts that have occurred to me over the years. That hope was mostly disappointed. For example, Collier writes:

While nationalism does not necessarily imply restrictions on immigration, it is clearly the case that without a sense of nationalism there would be no basis for [immigration] restrictions.

No, it isn’t. One could advocate immigration restrictions on pure labor protectionist grounds, without the slightest inkling of nationalist or patriotic feeling. This would be obvious if Collier analyzed immigration through explicit public choice lenses. I’m not sure whether Collier failed to think this through, or whether he’s aware of many analytical possibilities but is leaving them out to focus on the ones that he thinks are most plausible, for the benefit of a general readership.

This passage is closer to what I was looking for:

In the many societies that have never had a strong national identity, its absence is usually a matter of regret and concern. In Canada, Michael Ignatief recently ignited a storm by admitting that the long attempt to forge a translinguistic sense of common identity between the Quebecois and Anglophone Canadians had failed. In Africa, the weakness of national identity relative to tribal identities is widely regarded as a curse that it is the task of good leadership to rectify. In Belgium, which currently holds the world record for the longest period without a government– because the Flemish and Walloons could not agree on one– there has not even been an attempt to forge a common identity… Both Canada and Belgium manage to sustain high incomes despite their weak national identities, but their solution has been complete spatial segregation between the different language groups, combined with radical decentralization of political authority to these subnational territories. For practical purposes of public service delivery, Canada and Belgium are four states with cohesive identity, not two states without it.

Very interesting, but why? Collier continues:

Nationalism has its uses. Its potential for abuse cannot be forgotten, but a sense of shared identity turns out to enhance the ability to cooperate. People need to be able to cooperate at various different levels, some below the level of the nation and some above it. A shared sense of national identity is not the only solution to achieving cooperation, but nations continue to be particularly salient. This is evident from taxation and public spending: although both functions occur at many levels of government, overwhelmingly the most important is national. So if a shared sense of national identity enhances the ability of people to cooperate at that level, it is doing something truly important.

I am open to the idea that nationalism enhances the ability to cooperate and that such enhanced cooperation is crucial to economic productivity and psychic well-being. Open to the idea; but unconvinced. I need evidence. The kind of evidence that would be relevant would have to be at the level of civil society, voluntarism, and especially, spontaneous encounters among strangers. If it’s super important for people to be able to spontaneously cooperate with random people they meet on the street, and if cultural homogeneity hugely enhances that kind of spontaneous cooperation, you’d have a case against open borders.

The fact that taxation and public spending are usually done at the national level, as an argument against open borders, lacks logical traction. First, it’s very far from clear that most taxation and public spending has anything to do with raising economic productivity and general welfare. Much of it is mere redistribution, shrinking the pie but making it more evenly divided among insiders. One objection to that is that when most insiders are better off than most outsiders, equity is not particularly served by redistribution among insiders. But even if you do think, for some reason, that redistribution among native citizens is desirable, that can still happen, even if you open the borders to non-welfare-eligible immigration. In other cases, taxation and public spending contribute to public goods, in the strict sense (which ought to be the only sense) of goods that are non-rival and non-excludable. Unfortunately, the term “public goods” tends to be abusively extended to many goods which are rival and/or excludable for which the nifty conceptual toolkit economists have developed is irrelevant. In that case, calling something a “public good” is just a way of lobbying for the state to provide it, with this lobbying effort masked by spurious economic analysis. But if we’re talking about public goods in the strict, valid, appropriate, legitimate sense, immigration can’t reduce the supply of them, because they’re non-rival.

Collier makes a sharp distinction between nationalism, which he’s tentatively for, and racism, which he’s definitely against. Thus, he writes:

By far the most potent spillover to support for freedom of movement between countries as a natural right comes from opposition to racism. Given the histories of racism in both Europe and America it is both unsurprising and fully warranted that opposition to racism is so impassioned… In Britain, one high-profile anti-immigration speech in the 1960s clearly crossed [the] line [into racism]: opposing the immigration of people of African and South Asian origin in lurid terms of impending interethnic violence. That foolish speech by a long-dead minor politician, Enoch Powell, closed down British discussion of migration policy for over forty years: opposition to immigration became so indelibly linked to racism that it could not be voiced in mainstream discourse…

I only decided to write this book once I judged that it is indeed possible to distinguish between the concepts of race, poverty, and culture. Racialism is a belief in genetic differences between races: one for which there is no evidence… A refusal to countenance racially based differences in behavior is a manifestation of human decency. A refusal to countenance culturally based differences in behavior would be a manifestation of blinkered denial of the obvious… If [income and cultural differences are] assumed to be code for racism, then it is best that debate not be attempted, at least in Britain: we may still not be free of the long shadow of Enoch Powell.

I find all this very interesting, and very strange. First, is it really true that there is “no evidence” for “genetic differences between races?” What about “the bell curve?” And IQ and the Wealth of Nations? There seems to be at least some evidence for it. That aside, what exactly is the moral objection to racism? Is it merely that there is “no evidence?” Surely not. If racism were merely a gap in some people’s knowledge, an erroneous belief in differences that aren’t there, how could one call “a refusal to countenance” racism as “a manifestation of human decency?” I don’t regard people as offending against human decency if they erroneously believe that the US economy has a big inflation problem right now, or that foreign aid is 5% of GDP, or that the Earth is frequently visited by UFOs. I might pity their ignorance a bit; I might mildly censure their lack of due diligence in forming opinions; but I won’t be morally indignant.

Surely, our culture censures racism not because it is erroneous, but because we regard it as unjust. But why? Special historical reasons may really be the most important factor, but at the level of principles, I think our aversion to racism springs from an aversion to people’s fortunes in life depending on the mere accidents of birth, rather than on their merits. “I have a dream that my four little children will one day live in a nation where they will not be judged by the color of their skin but by the content of their character,” said Martin Luther King. Amen! But a person’s place of birth is also a factor outside their control. Why is it worse for a person’s station in life to be determined by the color of their skin, than for it to be determined by their place of birth? Given that Collier endorses people’s station in life being determined by their place of birth, I don’t understand the basis for his opposition to racism. Or at any rate, I don’t understand how he can justify being so emphatic about it that he thinks a major public policy issue shouldn’t even be discussed until suspicions of racism can be put to one side.

Of course, I do agree that culture, race and income are not the same thing. Probably Collier and the mainstream agree that racism is despicable but birthplacism, so to speak, is somehow OK. But I don’t think they have any good reason, or, really, any reason worthy of the name, for thinking so. Race is certainly correlated with culture; and birthplace, too, is merely correlated with culture. If we want to exclude people with undesirable cultural characteristics (whatever that means), why is using race as a proxy for culture any less legitimate than using birthplace, or national origin, as a proxy for culture?

This is one of many cases where I find Collier’s ethical premises to be indefensible, groundless, alien. I can’t see how he can imagine that he’s justified in thinking what he thinks.

While I am often unable to understand Collier’s ethics, he roughly understands mine, or rather, he roughly understands the “utilitarian universalist” perspective that is the presumptive ethical framework in normative economics. Thus, he writes:

For the neediest sections among the indigenous population the net effects of migration are often probably negative… To move from description to evaluation we need both an analytic and an ethical framework. In the typical work of advocacy on migration both the analytics and the ethics trivialize the problem because all the important effects appear to work in the same direction, with opposing effects being dismissed as “controversial,” “minor,” or “short term.” But any honest analysis must recognize that there are both winners and losers… If some people win while others lose, whose interest should prevail? Much economic analysis of migration comes to a clear and powerful answer: the winners gain much more than the losers lose, so hard luck on the losers. Even with the simple metric of monetary income, the gains far outweigh the losses. But economists usually move on from money to the more sophisticated concept of “utility,” and by this metric the overall gains from migration are even larger. For many economists, that answer settles the matter: migration policy should be set so as to maximize global utility.

Excellent, but I would nonetheless add a couple of qualifications. First, there’s a certain illegitimacy in interpersonal utility comparisons. It seems like a reasonable rule of thumb that the marginal utility of a dollar decreases with income, so there’s good reason to think “double world GDP” understates the utility gains from open borders; but this can’t really proved, or if you prefer, given a rigorous grounding in revealed preference. Second, it’s possible for policy to “tax the winners and compensate the losers,” which in theory allows open borders to be Pareto-improving; in practice, it won’t quite do that (it’s implausible that no one will be harmed) but it could come pretty close.

Anyway, since Collier understands the utilitarian universalist case for open borders, what’s his excuse for not accepting it. He writes that…

Nations are important and legitimate moral units… The very existence of nations confers rights on their citizens, most especially on the indigenous poor. Their interests cannot lightly be dismissed through the invocation of gains in global utility.

As an advocate of the Pareto criterion, I agree that neither the interests of the indigenous poor, nor anyone else’s, should be lightly dismissed. And I would accept, in some sense, the claim that “nations are important and legitimate moral units.” Indeed, I might possibly accept, in some sense, that the “existence of nations confers rights on their citizens,” though I’d have to think the matter through; it seems a highly tenuous claim. But anyway, what rights? And how does it confer them? And why should they include the forcible exclusion of foreigners from a country’s territory? Of course, even if they do include that, I would still say that natives are better off taxing than restricting migration, and they should let the market realize the vast gains available from labor mobility and then transfer some of the surpluses to themselves, rather than imposing extremely inefficient restrictions on the movement of workers to where they will produce the most value. Collier and I could argue whether that’s a good idea or not from a strictly economic standpoint; but I think he’d find that ethically objectionable, and again, I can’t understand why. Ethically, not only are we far apart, but I don’t know how to bridge the gap, because I find strange, unsupported ethical claims like “the very existence of nations confers rights on their citizens, most especially on the indigenous poor” lurking around every corner.

2 thoughts on “Response to Paul Collier: Chapter 1”

  1. A lot of interesting stuff here. I agree that the part starting with “it is indeed possible to distinguish between race, poverty, and culture” is particularly problematic.

    I basically agree with your complaints, but would like to also point out another problem with the statement: it implies it is better to discriminate based on culture and economic class than it is to discriminate based on race.

    Whatever we are using as a proxy for culture, why do we have to discriminate based on culture in the first place? What is wrong with simply trying to evaluate people as individuals?

    I wonder what Collier would say about this phenomenon:

Leave a Reply