The economics of open borders

by John Q on December 21, 2016

A colleague recently sent me a paper on the economics of open borders, by John Kennan, which I hadn’t known of before, though it came out in 2013.
Kennan’s conclusion is striking

Liberal immigration policies are politically unpopular. To a large extent, this is because the beneficiaries of these policies are not allowed to vote. It is also true, however, that the enormous benefits associated with open borders have not received much attention in the economics literature.20 Economists are generally enthusiastic about free trade. But if free movement of goods is important, then surely free movement of people is even more important.
One conclusion of this paper is that open borders could yield huge welfare gains: more than $10,000 a year for a randomly selected worker from a less-developed country (including non-migrants). Another is that these gains are associated with a relatively small reduction in the real wage in developed countries, and even this effect disappears as the capital–labor ratio adjusts over time; indeed if immigration restrictions are relaxed gradually, allowing time for investment in physical capital to keep pace, there is no implied reduction in real wages.

So, is Kennan right about the benefits of open borders? And if so, is there a way of transferring some of those benefits to already-resident wage earners who would otherwise lose, or at least not gain, from expanded migration?

On the first question, I’ll offer a bold Maybe. Kennan’s core assumption is that immigrant workers with a given level of education and (I think) experience will have the same productivity as already resident workers. So a move from a low productivity country to a high productivity country produces a big increase in their effective labor capacity. That benefits those workers, but also produces a shift in global income from labor to capital since the supply of labor has increased.

There’s room to debate this assumption, and there are special cases where it clearly doesn’t apply, such as that of professionals whose qualifications aren’t recognised in their new country. But Kennan makes a good case that it isn’t far from the truth.

Moreover, in a world where more than a billion people travel internationally each year, it’s inevitable that vast numbers of people are going to have close relationships of all kinds with citizens of other countries. Restrictions on movements across borders impose costs on all those people ranging from minor to calamitous.

Supposing that Kennan is right about the economics, what can be done to spread the benefits of open (or less tightly closed) borders more broadly and thereby, potentially, get increased political support. Since owners of capital benefit from open borders, an obvious possibility would be to increase the rate of tax on capital income and redistribute the income to labor. That seems neat enough in the abstract, but there’s no obvious (to me) way of putting it together as a political package.

The other way to spread the gains would be to tax, or otherwise capture, some of the benefits gained by immigrant workers. For example, new immigrants could be obligated to make a contribution, say through a tax surcharge, to a fund representing their share of the existing infrastructure of the destination country. Again, it seems neat enough in the abstract, but there are obvious difficulties. In particular, migrants who could have entered anyway would be significantly worse off.

Still, it seems unlikely that support for the migration policy status quo, let alone an expansion of existing flows, will be an adequate response to the rise of rightwing identity politics (what I’ve previously called tribalism), in which opposition to migration is a central feature. The more people who see freer movement as benefitting them and their families, the better will be the chances of mobilising support for a diverse and tolerant society.



Tom Davies 12.21.16 at 4:46 am

A surcharge seems like a good idea: focussed on a particular issue, transparent (although how the compensation would be distributed seems harder — how do you identify people who have been made worse off by immigration?). You could grandfather existing immigrants out of the surcharge, and cap it at some dollar amount, so that it is proportionally less of a burden on the higher skill immigrants we let in anyway.

I think increased immigration is an underrated way for richer nations to improve human welfare.


afinetheorem 12.21.16 at 5:43 am

Jess Benhabib has two very nice, theoretically rigorous, papers on this exact question: is on the political economy of migration, and on optimality .


Sebastian H 12.21.16 at 5:53 am

This is definitely approaching it from the right angle–large immigration flows act like globalization. They improve overall average GDP but definitely hurt certain sectors of workers in ways that thus far in the experiment suggests that they never recover.

I wonder about the effect of big city housing costs. They act as a barrier to moving to a better job. Is this something that we should be worried about as part of the immigration issue?


ZM 12.21.16 at 6:15 am

I think if you want to improve the economic inequality between countries, there are better ways than open borders. If the aim is to decrease economic inequality, you could make policies to reach this outcome that are more targeted than open borders, for example you could implement financial transfers between countries, or you could implement international minimum wages that could be phased in over 10-20-30 years, etc.

If the other problem you want to address is mobility for people who want to immigrate for personal reasons, you can just improve the access to immigration within the normal migration system, and increase migration quotas in line with some sort of expectation of what a optimum maximum population would be within a set period.

Also there is already a problem with gentrification in many cities, and associated issues of people having to move further away from family and friends, and not enough affordable housing, and homelessness — open borders would increase all these problems I would think as it would take out all the regulations. And we already have problems with people from poor countries or poor areas being under pressure to migrate for work and financial reasons, and open borders would exacerbate that problem as well.

I think refugees need to be able to migrate the most urgently, but it would still be better for them that there were specific policies for refugee migration that would allow the high numbers of refugees to migrate to safety either temporarily or permanently, rather than open borders.

Most of the bloggers here are not in favour of laissez faire free trade, so I don’t see why open borders are favoured when “open trade” isn’t?


ZM 12.21.16 at 6:21 am

Another thing is infrastructure, it would be difficult to forecast infrastructure needs if migration is unregulated. It would take several decades to settle into a sort of equilibrium and until then you couldn’t do very good projections of future infrastructure needs. In Victoria we already have had population growth that has outpaced infrastructure, and there are big problems particularly with transport but also with other infrastructure needs.


Chris Bertram 12.21.16 at 9:14 am

I blogged about a bunch of work, including a draft version of the same paper by Kennan, a few years ago John:


reason 12.21.16 at 9:19 am

I’m a bit suspicious that this sort of analysis suffers from large measurement biases. As an environmentalist, I’m concerned that we may be increasing a statistic (GDP) that is just a measure of an extent of how much higher a proportion of consumer is now being captured in the market, and not a measure of actual welfare. I remember very well as young economist wondering when I heard a more senior economists complaining that Australians didn’t want to work but just wanted to lie about on the beach. And then I thought about how northern Europeans pay large amounts of money in order to be able to lie about on the beach. Maybe the beach occupying Australians were being rational and the economist was not being rational. Having more crowded beaches does not show as a minus on GDP as far as I know.


reason 12.21.16 at 9:20 am

P.S. GDP may also measure how rapidly our natural capital is being converted to perishable goods but not measure how rapidly it is being eroded.


Chris Bertram 12.21.16 at 9:34 am

On the immigration surcharge thing, I can see its attractions as a policy, but let me just comment on it from the point of view of principle, not to advocate any particular solution but to notice some things:

The surcharge is supposed to be a payment towards the existing infrastructure, from which the new entrants benefit. But native-born citizens, who benefit from the infrastructure built up by previous generations get the same benefit as a free gift! That already presupposes some quite strong claims about who is entitled to what, and who is entitled to exclude whom from access.

(Adding in some plausible history, we might further note that the existing domestic infrastructure hasn’t, in many cases, been built up simply from the unaided efforts of the ancestors of the natives but often reflects the efforts of the colonised or dominated ancestors of the would-be immigrants.)

Then notice also another asymmetry, that the proposal is to charge the incomers for the benefits they derive from the infrastructure, whilst allowing the natives to benefit for free from human capital that has been created elsewhere through educational and training programmes. That issue, the so-called brain drain problem (I’m not a fan of the term or many of the associated claims btw) forms the basis for a quite different set of proposals for taxing immigrants, the so-called Bhagwati tax. So the poor migrants get hit by taxation proposals from both sides, as it were!


reason 12.21.16 at 9:35 am

P.P.S. Don’t get me wrong here, I’m not totally against open borders in all circumstances (in fact in a fairly equal world with a stable population I would be all for it), but I see a distinct danger of very rapid immigration:
1. In a world with a rapidly increasing population and a resource base coming under increasing stress it acting merely to spread misery faster and to stop experiments in sustainability.
2. It undermining social and democratic structures.


Tom Davies 12.21.16 at 9:42 am

“Chang, who opposes open borders” … “Currently a reader in the Political Economy of Development at the University of Cambridge” Open borders for me but not for thee?


ZM 12.21.16 at 9:45 am

Chris Bertram,

“So the poor migrants get hit by taxation proposals from both sides, as it were!”

Another issue with the tax is that it would make migration more difficult for lower income people who could’t afford the tax. Countries like the UK are already targeting their migration intake to higher income earners where possible, and a tax would encourage that policy.

I would rather make improve inequality between countries, and then experiment with freer migration after that. Since I think there would be less incentive to migrate if countries were more equal, and then freer migration would be more likely to run smoother.


Tom Davies 12.21.16 at 9:46 am

And (didn’t see comment 7) — while yes, a surcharge isn’t strictly fair, it is a far lesser evil than not letting immigrants in at all.

I think it’s plausible that remittances would more than make up for the investments home countries had made in ‘their’ immigrants.


SamChevre 12.21.16 at 11:14 am

On a surcharge–I proposed one years ago in the context of the proposed “amnesty,” that I think avoids some of the problems Chris Bertram notes above.

What if there were a minimum tax per immigrant per year, equal to the average taxes paid by citizens? This would be a “pay your share of current costs” tax, not an additional tax. For the US, it would be roughly $10,000 a year for the federal government (20% of per capita GDP)–and any taxes paid to the federal government (FICA and income tax) would be credits against it.


Dipper 12.21.16 at 12:16 pm

It is worth considering the world’s economy as an engineering system that responds to forces placed upon it. One of the features of making migration difficult, through either bureaucratic or financial resistance, is that it dampens the response of the system to external forces. Open borders removes that damping and allows much faster response. Like most things in life, that has both good and bad consequences, but one of the consequences is the system becomes less stable.

As an example, it is possible to get as many workers as you want into the UK within a few days. If you have a warehouse that needs staff they can be here from anywhere in the EU, all well educated, speaking English, with accommodation and ready to work. I’m not saying its a good or a bad thing. But its a thing that has consequences.

One organisation that has resisted Open Borders is the Corbyn clique in the Labour Party. Entry into the employment opportunities within that sector of the economy appears to be only open to relatives, children of political allies, school mates and children of celebrity chums.


BenK 12.21.16 at 1:50 pm

‘Not allowed to vote’ is political smokescreen. If those people have for some reason not established voting in the country they currently live in, then yes, they haven’t paid the price for liberty there. If they can, then they are voting – just not where they would apparently prefer to be voting.

As for the productivity argument – as usual, political theorists underestimate the value of extended family and long term inter-family arrangements in creating ‘social capital’ for productivity and stability. Mobility has its place, and in a time when most Americans never went more than 25 miles from their birthplace during their entire lives an increase in mobility increased overall productivity. However, there are many reasons to believe that individual mobility is costing communities dearly in these present times; and that bad government and market oriented policies which are exacerbating the problem.


William Meyer 12.21.16 at 2:46 pm

A surcharge might be a useful approach. I will say, echoing Reason, that massive waves of immigration into a region change a lot of things, and not necessarily in ways that the natives would view as positive. I’ve lived in L.A. for the past 35 years, and during that time millions of immigrants have come into this metropolitan area. Traffic problems, noticeable in L.A. when I arrived in 1981 but something that could be reasonably dealt with, got much, worse. The public school system went from fair to actively problematic. In both cases, the problem was made much worse by lagging public investment, particularly in transportation systems. Maybe L.A. was uniquely dysfuntional politically, but I would suspect that most regions would see degradations in public goods in times of massive in-migration. The significant investment required for massive population growth will, in all likelihood, not be made, especially in a timely way, and the sort of planning that would actually be necessary for a pleasant transition to a more populous future seems likely to be beyond the capabilities of most cities, at least the ones I’ve lived in. The inevitable resulting problems will not endear the newcomers to the natives, even if those problems are solely the fault of the immigrants.


William Meyer 12.21.16 at 2:48 pm

Sorry, my mistake, the last sentence should read “even if those problems are NOT solely the fault of the immigrants.


divelly 12.21.16 at 3:50 pm

Reminds one of the old story of the Capitalist who berates the local for quitting fishing after catching enough for today’s dinner so he can lay about playing guitar and drinking beer on the beach.
“You should fish from dawn to dusk 7 days a week.Sell your surplus.Buy another fishing boat.Do this for 30 years.”
“What for?”,asks the local.
“So you can retire and lay about ,play guitar and drink beer on the beach!”


Omega Centauri 12.21.16 at 5:10 pm

I think its a very difficult sale poloitically. But, you already know that.

There also is the issue of potentially large scale population flows from less “productive” areas to
more “productive” areas. We have this same issue within countries, such as rust belt to coastal cities in the US, and we’ve seen political consequences -Trumpism become ascendant. So in addition to the problems of infrastructure and gentrification on the recieving end of these net flows, we have issues in the regions that are being left-behind. Our current reactionary politics seems to be one of the consequences of this difficult issue.


Chris Bertram 12.21.16 at 7:25 pm

@divelly it is from Adam Smith, Theory of the Moral Sentiments, part 3, ch. 3 :

“What the favourite of the king of Epirus said to his master, may be applied to men in all the ordinary situations of human life. When the King had recounted to him, in their proper order, all the conquests which he proposed to make, and had come to the last of them; And what does your Majesty propose to do then? said the Favourite.—I propose then, said the King, to enjoy myself with my friends, and endeavour to be good company over a bottle.—And what hinders your Majesty from doing so now? replied the Favourite.”


Stephen 12.21.16 at 8:10 pm

Moral sentiments of less desirable people: I propose to enjoy myself by being revenged on and utterly destroying my enemies, and to be good company over the finest available bottle with those who dare not contradict me.

Not my sentiments, not CB’s but …


Matt 12.21.16 at 10:43 pm

I think we should be extra skeptical of any paper that claims that the “economics of open borders” hasn’t received “much” attention. Maybe not as much as many other things, and there may be some hedging about what, exactly, fits, but the economics of migration has received _lots_ of attention. In The US, the National Academy of Science did a huge study on it in the mid 80’s, and updated it again just recently. Jagdish Bhagwati has written quite a bit on it, both popular and formal. George Borjas has written a lot on it (most of it not good, in my opinion, but a lot on it.) Lots and lots of people, including some very famous economist, have responded to Borjas. Paul Krugman has written on it. One of my mentors, Howard Chang, a lawyer-economist at Penn Law, has written a lot on it. Etc. So, already we know that there is something a bit fishy here.

Next, this sort of thing typically assumes, for its strong conclusions, that everyone will move to where he or she will get the “highest” return for his or her skills. We know this is false, because it doesn’t even happen within any particular country, where there are no restrictions, no “surcharge” to pay, and fewer cultural barriers. So, the gain will certainly be much smaller than is projected.

I’d also suggest that this bit from John, Moreover, in a world where more than a billion people travel internationally each year, it’s inevitable that vast numbers of people are going to have close relationships of all kinds with citizens of other countries. Restrictions on movements across borders impose costs on all those people ranging from minor to calamitous.

Would need to be _much_ more rigorous to do any work. I travel quite a bit, yet unless “close relationships” means “people I know somewhat”, this isn’t true for me. Is it true for “vast” numbers of people? I’m not sure. It’s too flabby to do work now. And, do we have in mind visits, temporary residence, permanent residence (with or without access to full membership?) Etc. There are really a huge number of details here, and the absolutely must be worked through, carefully, before you can say anything useful. I’m in favor of reducing most barriers to movement. But, the arguments, if they are to be any good, really do need some care.


Chris Bertram 12.22.16 at 12:00 am

I’m puzzled by your last paragraph Matt, given what I know about your work. I don’t know how large a number has to be to be “vast”, but the spouses separated from one another and the children separated from one parent by the UK’s spousal visa income requirements already number in the 10s of 1000s. Add to that elderly dependent relatives who are separated from children, lone refugee children separated from family members in other countries. And then multiply all this separation by the number of countries that make things difficult for people. I think that probably adds up to a vast number of people in close relationships with others who are separated by border regimes and who are currently incurring costs that are often calamitous. Don’t you?


Matt 12.22.16 at 2:41 am

Hi Chris – yes, the cases you mention are interesting and important ones. It goes a little way towards making John’s too flabby to work statement a bit better. But even in these cases, it’s important to work through what’s wrong with the different examples. (This is what I try to do in my work, and it’s why I’m annoyed by what seems to me to be handwaving that blurs and distorts more than it helps.) I would insist that “making this difficult” for people, or causing them to “incur costs” isn’t a good way to think about these issues at all. (I will go see my parents for the first time in over a year next week. It will be difficult and I will incur may costs to do so. Nothing interesting follows from that at all, I think.) And, John’s categories include may more than those you mention. What follows for them? Why are borders, and not other types of boundaries relevant here? (Suppose my best friend is admitted to Harvard and I am not. But I’d like to study with him! Is it unfair that I’m not allowed to? Why not?) There are answers here, but we’ll not get at them from the approach in the post, I think, and especially not if we follow the approach in this paragraph. The issues need to be dug in to, even though that take time.

(I might note that I’ve just finished a semi-popular short piece on thinking about immigration post-Trump and post-Brexit. I started it by thinking about some of your discussion of Joe Carens’ book from a few years ago, and tried to think about reasonable strategies for working towards fairer immigration policies in our dark times. One thing I suggested was fighting against needlessly mean (in both senses of the word) restrictions like the too-high social support requirements for family members in the UK. So, I see that as a real problem. But, I don’t think that helps rehabilitate the claims made, or suggested, in this paragraph. If and when the piece comes out, I’ll send it to you.)


John Quiggin 12.22.16 at 3:22 am

Matt @24 It was an aside of course, but one that I didn’t think needed a detailed exposition. The calamitous cases Chris mentions are well known, as is the fact that lots of people suffer no, or only trivial, problems of this kind. Rather than multipy such trivial examples as you do in @26, why not explain why you think the calamitous cases are rare, or need to be explained in detail?


Faustusnotes 12.22.16 at 5:10 am

Matts response is to glib. My own family was driven into poverty by separation in the 1980s and the long term pressures on all of us of that experience were huge. Yes nothing follows from that if you’re a wealthy academic, but quite a lot follows from it if you’re not.


Dave 12.22.16 at 10:05 am

I worry that “free movement of people” tends to have massive social costs that get swept under the rug when the issues are discussed in a purely economic framework. For most of human history, the vast majority of people lived in extended family groups in villages, towns, or temporary encampments where they knew their neighbors, had relatively small social worlds, and didn’t travel more than a few days from home. Cities, as we know them, (and their accompanying social maladies) are really only two or three centuries old and post-industrial cities are an even newer phenomenon.

What people in the urban professional class tend to forget, however, is that the old model of village life never went away. In truly rural or otherwise undeveloped areas, it’s mostly stayed the same, and other cases it was remapped onto urban neighborhoods or desperately clung to in “small towns” that are, in fact, larger than most Medieval cities.

Now, these people have a problem, which is that they’d very much like to maintain a traditional village lifestyle (well, some of them just want to escape or move to the city and get rich, but I’ll get to that), but neither industrial nor post-industrial capitalism has had any patience for people who want to stay put. Industries and opportunities have concentrated in large urban agglomerations, but exactly which industries and which cities shifts every generation or two. Plants close down or move to other countries, higher education pulls millions of people far from home, entire fields of employment vanish or emerge from whole cloth and it’s impossible to keep up. So, we as individuals can, at any time, be forced into a terrible dilemma. Either move away from the life you know and the people who keep you happy, healthy, safe, and sane, or forfeit your “optimal” career and some share of prosperity and human capital.

Depending on what class you are, the values you hold, and what the costs and benefits of moving away really are, there may be no choice at all. Really, there are two kinds of migrants. There are the desperate, who migrate for negative reasons, and the ambitious, who do it for positive ones (with plenty of overlap) and only the latter is really making a choice as such. The outcomes are different too. Refugees and economic migrants occasionally become rich and successful, but usually they’re just looking for security. Whereas people who move around a lot to get the best education and the best jobs, are often massively rewarded, but too many such people creates a culture of anomy and alienation where no one knows their neighbors and everyone seems to be from somewhere else.

What’s worse, the rootless urban professionals have money, which means they can buy or rent homes anywhere, displacing the old residents. This has a double whammy effect, not only do the neighborhoods get new people who don’t quite fit in, but the old “villagers” get forced out (and in many cases become rootless transplants in some other town), so communities enter a state of perpetual social flux where there aren’t enough old timers left to assimilate the new arrivals and the social fabric disintegrates as natural communities are replaced by a massive web of voluntary ones that often don’t (especially if there’s a class or language barrier) and leave some people with no community at all.

I grew up in the Southern California suburbs in wake of the Sunbelt migrations and massive immigration from Latin America and had the utterly peculiar experience of being one of only a tiny fraction of the population whose grandparents (well, two of them) also grew up there. Growing up, it seemed like most of my teachers (and really a huge chunk of the professional class in general) were from either the East Coast or the Midwest and many of them had strange notions about what it meant to be Californian, having moved here for the sunshine or the surfing or the jobs or the “vibe” and more able to see the place as an ideal than a reality.

People don’t realize the extent to which generations of migration can isolate people from the land, but I saw it. People that luxuriantly watered their lawns despite the climate and planted gardens full of plants from all over the world while treating the native plants like weeds. The tragedy of people in brushfire country not even realizing that having wooden shingles is a bad idea. People mocking the native California accents, affecting them badly to fit in, or refusing to acknowledge that we had one at all (we have several). Or, take the baffling experience millions of California kids get this time of year where adults around them act like our Christmas is somehow “wrong” because there’s no snow and we don’t have a “real” winter.

There was and is wanton disregard for tradition or the environment. The old growth oaks that once covered much of SoCal were cut down for wood and cattle land and now most people have no idea they were ever there, huge tracts of “empty” desert were flooded with saltwater when the Salton Sea was created, the LA river was turned into a storm drain, massive population increases and utterly unrestricted suburban sprawl has destroyed most of our wetlands and turned the Coastal Sage Scrub into one of the most threatened ecosystems on Earth. All that and CalTrans still plants invasive, flammable Eucalyptus by every freeway. These were largely the work of generations of short sighted, greedy migrants who didn’t understand or value the land, but will be borne by generations to come. Our land is being paved, poisoned, and pumped dry and most people don’t even see it because there aren’t enough people around who still remember when it was any different.

If open borders means that places all over the world start getting flooded with migrants and disrespected and debased the way Southern California has been, then I have no choice but to oppose it.


reason 12.22.16 at 10:47 am

A small point: – instead of adding additional taxes, one way to get the same net effect is to have a basic income with a long residency requirement for non-citizens.


Matt 12.22.16 at 3:29 pm

John – there is a lot of space between “not rare” and “vast”, isn’t there? That space needs to be looked at carefully, and not used as a hand-wave. That’s my point.

Faustunotes – I’m sorry to hear that. In published work, I’ve argued for strong rights for family migration schemes. Without knowing more about your situation (not that I’m asking for details now) I can’t say more about, but, for example, the sorts of public support systems I’ve argued for (and that exist in many countries) can be easily met by lots of people – the US requires 125% of the poverty level for a family of the appropriate size, for example. That is arguably a top acceptable level. That meant that I was able to sponsor my wife when I was a grad student making $15,000 a year (in 2003), not at all a “wealthy academic”. So, again, it’s important to get the details right, to criticize particular cases, and not draw strong conclusions from hand-waving generalizations. Failing to do this won’t lead to any good work.


Chris Bertram 12.22.16 at 7:08 pm

Matt, I think your quibbling with John on “vast numbers” is pretty silly here. You are an American, and the US is a continental power with a large population. Perhaps it is rare for Americans to have close relationships (let’s set the bar at good friendships) with people outside the borders of their country. But many of us live in smaller countries and on continents with lots of borders. I think you’ll find that when you tot up all the Europeans and Latin Americans with cross border relationships (to name but two continents) and add in all the people who belong to ethnicities that stretch across many borders, you’ll get to a pretty high number. Think of the million Poles in the UK, for example, they will predictably have “close relationships” with … people in Poland, and will British retirees in Spain with people in the UK, and Irish people in the UK with people living in Ireland ….


Alesis 12.22.16 at 8:55 pm

The ever present struggle with taking the empirical body of knowledge on gains from migration and making it into policy is that the only salient objections to migration are decidedly non economic. Sure they pretend at an economic basis with admirable dedication to the act but the bottom line is even if you prove that net wages for every single individual would go up from migration it would till have exactly the same opponents you started with.


WLGR 12.22.16 at 9:53 pm

At this point I really, really have to emphasize the plug for John Smith’s Imperialism in the Twenty-First Century because he addresses this kind of neoclassical boosterism more or less directly, from a leftist point of view. If he was reacting to this post Smith would zero in on the key premise underlying Kennan’s model: the idea that global wage differentials inherently reflect global differences in the productivity of labor between nation-states, known as the Balassa-Samuelson hypothesis. Kennan seems to handwave away the idea of actually defending it by deferring to “large bodies of evidence”, evidence whose interpretation within a more-or-less standard neoclassical framework he takes as a given — although notice how he hedges his initial claims more carefully (“cross-country differences in income levels are associated with differences in productivity”, and “large differences [in productivity] remain after adjusting for differences in physical and human capital endowments“, emphasis mine) before moving on to construct a model where “relative wages are used below to measure cross-country differences in labor efficiency”, plain and simple. Nice trick!

In any case, here’s Smith:

The North-South purchasing power anomaly is sometimes called the Penn effect, after the Penn World Table, which has gathered comparative price data from most countries in the world since 1950. This effect is inversely correlated with per-capita GDP; as Figure 5.2 (page 143) clearly shows, the poorer the nation, the bigger the gap. Mainstream neoclassical economics advances two chief explanations for this anomaly, the Balassa-Samuelson hypothesis, which hinges on differences in labor productivity between rich and poor countries; and an alternative model, proposed by Jagdish Bhagwati, Irving Kravis, Richard Lipsey, and others, which claims to circumvent differences in labor productivity and accounts for the anomaly as the consequence of differences in “factor endowments,” that is, the relative abundance of capital and labor in the two countries. Since their arguments are tautological, they arrive at the same conclusion. In the former approach, the relative productivity of labor and capital determines the demand for these two factors and, in conjunction with their supply, determines their equilibrium (market-clearing) prices. In the second approach, different factor endowments affect the supply and demand in markets for labor and capital, determining marginal productivities, so arriving at the other’s starting point.

According to both approaches, the purchasing power anomaly arises because of the low wages of workers providing services (for example, a bus journey or a haircut), resulting in the prices of these services being typically much lower in, say, Bangladesh than in Belgium. But equilibrium exchange rates do equalize the prices of internationally tradable goods—in other words, they assume that strong PPP holds in the tradable goods sector. Service sector wages are low in Bangladesh because wage levels in the service sector are determined by wage levels in the tradable goods sector. This occurs because labor is intersectorally mobile but not internationally mobile; in other words, workers can freely move between the tradable and non-tradable sectors within nations, equalizing wages between them, but cannot freely move across the borders between nations, especially those between hard-currency and soft-currency nations. it therefore turns out that the suppression of the free international movement of labor, the great exception to the principle of globalization and whose cardinal importance is stressed in this book, is also at the heart of the purchasing power anomaly.

In sum, the Balassa-Samuelson hypothesis says that the purchasing power anomaly results from the lack of correspondence between the similar levels of productivity of service workers in Belgium and Bangladesh and the vast differences in their wages. The contrary argument advanced here is that it is the oversupply of labor, not its productivity, that is the prime determinant of Southern wage levels. wages of service providers and incomes of petty entrepreneurs are kept low not by the paltry productivity of workers in the tradable goods sector, as mainstream theory has it, but by the destitution of a large part of the working population. This is why a haircut or a bus journey in Dhaka is so much cheaper than in Amsterdam, even though a pair of scissors or a bus may cost the same in both countries, and may even have come off the same production line. Furthermore, local capitalists are not the prime beneficiaries of the super-profits generated by this expanded employment of low-wage labor. Instead, intense competition among Southern exporters leaves them with only a minor share of the proceeds, the rest passed on to their Northern customers through ever-lower export prices. The purchasing power anomaly results not only or mainly from conditions in goods and Forex markets but is fundamentally the product of conditions in labor markets and in the sphere of production where this labor is put to work. The enormous growth in the relative surplus population combines with suppression of international labor mobility to exert a tremendous downward pressure on all wages and on the incomes of small producers, maintaining or widening still further the distance between real wages in the imperialist nations and in the Global South.

In Smith’s telling the suppression of international labor mobility is actually central to explaining not just global wage differentials perceived to result from differences in productivity, but also the data by which labor productivity between countries is measured in the first place. The neoclassicals’ trick here is to take the international division of labor that emerges from what Smith calls “global labor arbitrage” (e.g. outsourcing) and remove it from their conceptual category of production altogether, instead regarding it through the lens of international trade as if workers on a factory floor were constantly “trading” their partially-assembled products to others further down the assembly line. Here’s Smith again:

Statistics on labor productivity, obtained by dividing the value added of a firm, industrial sector, or nation by its total workforce, are highly deceptive. Much of the alleged increase in labor productivity in the imperialist nations is an artifact resulting from the outsourcing of low value-added, labor-intensive production processes to low-wage countries. As Susan Houseman has argued, “when manufacturers outsource or offshore work, labor productivity increases directly because the outsourced or offshored labor used to produce the product is no longer employed in the manufacturing sector and hence is not counted in the denominator of the labor productivity equation.” This is extremely important, because “the rate of productivity growth in U.S. manufacturing increased in the mid-1990s, greatly outpacing that in the services sector and accounting for most of the overall productivity growth in the U.S. economy.” Thus she argues, “To the extent that offshoring is an important source of measured productivity growth in the economy, productivity statistics will, in part, be capturing cost savings or gains to trade but not improvements in the output of American labor.” Houseman believes this solves “one of the great puzzles of the American economy in recent years … the fact that large productivity gains have not broadly benefited workers in the form of higher wages. … Productivity improvements that result from offshoring may largely measure cost savings, not improvements to output per hour worked by American labor.”

Thus, when a firm outsources labor-intensive production processes, the productivity of the workers who remain in its employment rises, even though nothing about their specific labor has changed. Outsourcing therefore has what might be called a “ventriloquist effect” on measures of productivity. But this only scratches the surface of the productivity paradox. Labor-intensive production processes are practically synonymous with low value-added production processes, yet the more labor-intensive it is, that is, the larger living labor is relative to dead labor, the greater is its contribution to value and surplus value—but much of this is captured by capital-intensive capitals, showing up as a much higher value added per worker.


John Quiggin 12.22.16 at 11:17 pm

@Dave You are arguing against internal freedom of movement. Do you support systems of internal passports, as in the Soviet Union or the hukou system in China?


John Quiggin 12.23.16 at 3:14 am

WLGR: I’ll look for this book. But on an initial reading of your first quotation, it seems to me that Smith is just restating the factor endowment model. What does “surplus labor” mean, if not a high ratio of labor to capital? Does he spell out the distinction somewhere else?


ZM 12.23.16 at 3:26 am

John Quiggin,

China has a very large population, hukou is problematic and has some undesirable impacts, but China needs to get all the provinces and cities more equal before they can change the hukou system. At the moment the inequality between provinces and cities in China is very very great compared to inequality between States and cities in Australia.


ZM 12.23.16 at 3:27 am

Although inequality has decreased as more people have been lifted out of poverty in the last 10-20 years.


Dave 12.23.16 at 6:06 am

@ John Quiggin 34

That’s a very good question and it does show why one should always consider the full ramifications of ones’ arguments. I would say that policies against internal migration are not limited to Communist dictatorships — that’s what serfdom was, after all. I’m enough of a liberal to find that sort of thing oppressive, but I do think it had a certain social utility (of course, letting people move and travel has advantages too).

It’s not exactly freedom of movement I’m arguing against so much as the notion that population centers have an essentially unlimited ability to absorb newcomers. From 18th Century Manchester to the American West to exploding Chinese industrial cities today, boom towns are notorious for their environmental devastation and social dysfunction.

Even in a modern era where resource extraction and heavy industry are less dominant economic drivers than they once were, the combination of free movement of capital and free movement of labor is a consistent recipe for explosive, unplanned, and unsustainable growth in whatever areas are deemed economically valuable. The boom bust cycle of capitalism maps onto the landscape itself and the effects for both the natives and the newcomers can be devastating.

What I would argue though is that free movement of people is a problem only insofar as there is free movement of capital. You won’t have millions of people flood a region if that region hasn’t already been flooded with millions of jobs. This would require a new international regulatory framework to put the brakes on massive industrial and commercial development and a rejection of the current extreme growth bias in economic thought. In effect, I think that it should be businesses that have to apply for those permits or internal passports, rather than individuals.

Many municipalities already do this through the use of building permits, but their efforts are compromised by an imperative to expand their tax base and competition between municipalities that gravely limits their effective bargaining power. In a free trade, open borders world, I can see the same thing happening at the national scale, forcing whole countries to compete with one another for jobs and labor and hastening the rate of neo-colonial resource plundering.

My worst case scenario is something like this. Lets say a fairly small — but not necessarily tiny — country like Uruguay adopts global open borders. A little while later, they make the shocking discovery that they’re sitting on some of the largest reserves of, oh let’s say, rare earth metals in the world.

Now, these metals are incredibly valuable so getting the capital to open mines and ore processing centers isn’t a problem, the bigger issue is that Uruguay only has 3.4 million people, most of whom already have jobs, so the tens of thousands of employees needed to build the new mining industry will mostly be coming from elsewhere (or the mining companies will start by hiring Uruguayanos, but then they’ll need migrants to fill the jobs the natives vacated). It doesn’t stop there though, because the great new mining industry will produce secondary industries such as cell phone manufacturing, service jobs for the growing population, construction jobs to expand the national infrastructure, and on and on and on. These jobs bring in new migrants, who help grow the economy, and attract new migrants in a feedback loop that only ends when the bubble bursts or wages collapse. How big does Uruguay get before the boom goes bust? Does it double in size? Triple? Does Montevideo become one of the biggest cities in South America, with sprawling, polluted, slums to match? What happens to the reasonably stable, reasonably prosperous, reasonably progressive little country that was there before? Would Uruguay still be Uruguay at that point?

These are the things I worry about.


hix 12.23.16 at 7:14 am

Poor former farmers that move to big cities typically wont drive cars, wont handle big industrial manichery and will only heat /cool tiny lifing spaces. So they are probably not a significant factor for the environmental issues in say big Chinese cities.


Neville Morley 12.23.16 at 7:39 am

@Dave #28: “Cities as we know them (and their accompanying social maladies) are really only two or three centuries old”. No. Ancient Rome had a population of 750,000-1,000,000, based almost entirely on migrants, and more than large enough to create any number of social maladies; at least five other cities in Mediterranean with populations in the hundreds of thousands; series of cities in China with populations similar to Rome. Evidence suggests significant levels of mobility, not just for elite. I don’t think this necessarily has any bearing on the modern situation (capitalism, technology, yadda yadda), but certainly the historical evidence doesn’t support your implied “large-scale migration is unnatural” thesis.


Igor Belanov 12.23.16 at 7:58 am

Dave @ 38

“Would Uruguay still be Uruguay at that point?”

What a daft question. When did the ‘model’ Uruguay exist, the one that we are supposed to preserve for all eternity? Now? Before the Uruguayan nation-state was formed? Before Columbus?

The irony is that the effort needed to prevent change would in all likelihood just lead to other changes of a more dysfunctional nature.


engels 12.23.16 at 12:21 pm

OT and possibly an ignorant question but does anyone know of any meaningful national or cultural difference between Uruguay and Argentina?


bob mcmanus 12.23.16 at 12:27 pm

Factor endowments equals they have poor people for cheap labor and we have rich people who create, consume and finance and the origins of the difference is like shrouded in mystery?

John Smith of WLGR is Marxian. Apples profits are generated by the workers at Foxconn in China, not the designers in San Jose. The surplus accruing to intellectual property is mostly a product of past and present Imperialism.


Matt 12.23.16 at 1:30 pm

Chris – maybe it’s silly, but, if my work on immigration has tried to show anything at all, it’s that to make a contribution on the subject, it’s important to get the facts right, not make assumptions about movement we know are not true (people will move to where they get the best return on their skills, etc.), not assume away other difficulties, and not blur cases together through hand-waiving (“vast numbers”, “relationships”, etc.) All of that’s done here, and even more so in the paper under discussion. I find it really annoying. Maybe I shouldn’t let it bother me, but it seems to me to be typical “assume a can-opener” level of discussion, at the very best, and not helpful.

By the way, at least according to Wikipedia, there are 830,000 Poles in the UK_ so well south of a million still. It’s a lot, but, lots less than, say, the 2.9 million Russian-born people in the US, a population I’m very familiar with, so I don’t really need the lecture here.


Dave 12.23.16 at 2:03 pm

@ Neville 40
You’re right, of course. I recognize now that my argument was a bit fuzzy and verges into begging the question (“Modern cities, as I’ve chosen to define them, only existed under capitalism, therefor urban dysfunction is all capitalism’s fault, QED, etc.”). “Bigger than Cleveland” is not a universal definition of what a city is and I shouldn’t have treated it as one.

I’m not trying to say massive migration was unnatural though. I’m of the opinion that anything humans do is natural, if that helps. Nor do I think migration, even of the large-scale variety is wrong, but rather that it can be immensely harmful if there are no systems in place to mitigate its social and environmental effects. So discussing policy that would tear down all political barriers to migration as if it were mainly an issue of wages and productivity struck me as reductive. Even on purely economic terms, the way migration contributes to urban sprawl outpacing infrastructure is a huge issue that I frequently see overlooked.

@Igor 41:
I’m not arguing against change, but rather change that comes so quickly it creates a schism between the past and the present. The Gold Rush changed California from predominantly Spanish-speaking to majority Anglo in just a few years (and also killed tens of thousands of Indians in the process), so even if a place still has the same name following migration, it might not be pronounced the same way.


engels 12.23.16 at 3:32 pm

Also would be interesting to see numbers on marriages to foreign nationals by country—a quick google didn’t turn it up.


Chris Bertram 12.23.16 at 4:26 pm

@engels – pretty sure that this is because the stats don’t exist for many countries. The British government simply has no idea how many of its nationals are married to EU nationals (at least for England and Wales, there may be some record-keeping in Scotland).


engels 12.23.16 at 6:55 pm

Chris, that makes sense—there’s a bit about estimates here though:


Dipper 12.23.16 at 7:36 pm

@47. The UK has no idea how many EU citizens are here full stop. Here’s an excerpt from Wikipedia.

The 2001 UK Census recorded 36,555 Portuguese-born people resident in the UK. More recent estimates by the Office for National Statistics put the figure at 107,000 in 2013. The 2011 Census recorded 88,161 Portuguese-born residents in England and Wales. The censuses of Scotland and Northern Ireland recorded 1,908 and 1,996 Portuguese-born residents respectively. Other sources estimate the Portuguese community to be larger, with the editor of a Portuguese-language newspaper putting the number of Portuguese passport holders in London alone at 350,000. According to academics José Carlos Pina Almeida and David Corkill, writing in 2010, estimates of the Portuguese population of the UK range from 80,000 to 700,000.

I mention it because informal information from someone at the embassy puts the number closer to 1 million. Many of them well have been born here. But nevertheless the point is that the estimates are all over the place.

Again, its not necessarily a good or a bad thing, but as a scientist with a bit of a measurement fixation I find the fact that no-one has any idea to be quite disturbing.


engels 12.23.16 at 8:31 pm

E.g. on US vs Europe:

“4.6% of Americans were married to a foreigner in 2010, up from 2.4% in 1970”
“in France the proportion of international marriage rose from about 10% in 1996 to 16% in 2009. In Germany, the rise is a little lower, from 11.3% in 1990 to 13.7% in 2010. Some smaller countries have much higher levels. Nearly half the marriages in Switzerland are international ones, up from a third in 1990. Around one in five marriages in Sweden, Belgium and Austria involves a foreign partner”


Collin Street 12.23.16 at 8:56 pm

OT and possibly an ignorant question but does anyone know of any meaningful national or cultural difference between Uruguay and Argentina?

Argentina has a much, much more active domestic sailing scene than Uruguay does. It’s not much, I know.


engels 12.23.16 at 10:08 pm

It might seem problematic for John’s argument that Switzerland, which has a very high rate of international marriages (50%), also has a very high degree of popular support for immigration controls (over 50% in 2014 referendum)…


engels 12.23.16 at 10:09 pm

Thanks Colin


ZM 12.24.16 at 1:05 pm

“I’m not trying to say massive migration was unnatural though. I’m of the opinion that anything humans do is natural, if that helps.”

Sound like a good advertising pitch, eg “Margarine, the natural alternative to artificially churned cows’ cream”.


hix 12.24.16 at 3:03 pm

How do those cited statistics work? Just foreigner/national couples? Or also foreigner/foreigner? And what about dual citicenship holders?


engels 12.24.16 at 7:26 pm

I suppose it’s possible half the Swiss citizenry are married to foreign nationals and the other half wants them deported…

Comments on this entry are closed.