Eric S. Maskin (2014) - Why Haven’t Global Markets Reduced Inequality in Developing Economies?

Good morning to everyone. I must say it’s a real privilege to have the chance to talk to you here. As Peter said: this is inequality morning. And you’ve just heard a talk by Joe on the subject. I’m going to be talking on inequality too, but emphasising rather different aspects. In particular I’m going to be looking at the connection between globalisation and inequality in developing countries. I’m sure I don’t have to persuade you that there has been an enormous increase in globalisation over the last 20 years or so. We’ve seen more trade of goods and services between countries. We’ve also seen - and this is a point that I’m going to be emphasising later on – an internationalisation of the very production process. And there are various reasons why this huge surge in globalisation has occurred. There’s been a decline in transport costs, certainly. But, perhaps even more important, there’s been a decline in communication costs. It’s now possible to send messages around the world for practically zero cost. And there has also been - through trade agreements, through the creation of the EMU - a removal of trade barriers. Well, this globalisation has come with many promises. Its proponents have suggested that many good things will have happened because of globalisation. In particular it’s been argued that globalisation should bring prosperity to developing countries, to emerging economies. There’s an important sense in which this promise has actually been delivered. So if we look at many developing countries - and China and India are only the most spectacular examples –, we see very significant increases in GDP per capita over this period of globalisation. But there’s another promise that the proponents of globalisation made which does not seem to have been delivered on so well. And that is to reduce the gap between the haves and the have-nots in order to reduce inequality in emerging economies In fact, we’ve seen just the opposite in many countries around the world. And, again, India and China provide perhaps the most prominent examples: very large increases in inequality in both those countries, but in many other developing countries too. There’s been lots in the news recently about inequality. Much of this has to do with the Piketty book. But if you look at the press stories on inequality, most of the emphasis has been on inequality in rich countries. And Joe’s talk looked primarily at the US/European countries. What I’d like to do today is to look at the increase in inequality in emerging economies. You might ask, well, why should we be concerned with this inequality? Perhaps many of you won’t need persuasion that inequality there matters. But let me just mention 3 pretty obvious reasons why we should care about it. The first is a moral argument. The idea that human beings ought to be treated equally. And so very egregious differences in income offend our egalitarian impulses. But if you don’t accept that, you may nevertheless believe that eradication of poverty is a good idea. And, in fact particularly in developing countries, reducing inequality closely correlates with the eradication of poverty. But even if you don’t accept that argument, there is a much more practical reason for being concerned with inequality. Which is that there’s a well-established correlation between inequality and political and social instability. Countries which have very wide dispersion of income have more social and political unrest. So for all these reasons I think inequality is an important problem. But that generates the question: Should we be surprised that, thanks to globalisation, we’ve seen an increase in inequality in emerging countries in the last 20 years? Well, if our perspective is the theory of comparative advantage, which is the best established theory of international trade we have, the answer is actually 'yes'. Because the theory of comparative advantage predicts just the opposite. The theory of comparative advantage goes back at least 200 years to Ricardo and, at least in previous globalisations did a good job of explaining or predicting what would happen to income inequality. This theory says that when you have an increase in free trade, if you reduce the barriers to free trade, that should induce a fall in inequality in emerging economies. And that's such an important prediction that it’s worth spending a few minutes seeing why that’s so. So let’s have a quick look at the theory of comparative advantage. At least in the version due to Heckscher and Ohlin, this theory assumes that the important differences between countries, when it comes to trade, is their relative endowments of the factors of production. So the important differences are factor ratios. And the particular factors that I want to concentrate on today, because I’m interested in income inequality, is labour. I want to look at high-skill labour and low-skill labour. Let’s compare a rich country with a poor country, an emerging economy. If we focus on labour, we will expect the rich country to have a higher high-skill to low-skill ratio than the poor country. The reason the rich country is rich is because it has more high-skill workers, it has more human capital. And that means that the rich country has a comparative advantage in producing goods where a lot of high-skill work is required. Take computer software for example. Computer software is mainly produced by people with rather high skills: software engineers. By contrast with the emerging country which has a comparative advantage in producing goods where skill doesn’t matter so much. And agricultural goods often fall into this category - rice might be an example to keep in mind. To understand how globalisation affects production and inequality, let’s do a thought experiment. We’ll look at what goes on before globalisation, before there’s any trade between the rich and poor country. And we’ll compare that to what goes on after globalisation. And the difference between the 2 can be attributed to globalisation. Before the rich and poor country can trade, consumers in the rich country will want to consume both software and rice. So both software and rice will have to be produced by companies in the rich country. It can’t be imported from the poor country. And the same thing is true in the poor country. It will produce both software and rice. But there’s a sense in which, according to the theory of comparative advantage, producing software in the poor country is making inefficient use of the labour force. The labour force in the poor country is better suited to producing rice – to the extent that the poor country is producing software that’s going to be reducing demand for low-skill workers. They’re not much needed for software. And that means that demand for low-skill labour is reduced, wages for low-skill labour are reduced. And just the opposite is true for high-skill workers: They’re very much in demand for software. They see their wages at a high level. Well, what happens once the door to trade between the rich country and poor country opens up? Well, now the rich country doesn’t have to produce so much rice anymore. It can shift production to software and import a lot of its rice from the emerging country. The emerging country will do just the opposite. It will concentrate on rice and import its software from the rich country. Now, what effect will this have on inequality? Well, the emerging country is now producing more rice, less software than before. That means that the demand for low-skill workers is higher than before, because low-skill workers are really needed for rice. So that’s going to raise the demand for low-skill workers. Since not so much software is being produced, the demand for high-skill work goes down. And we see a fall in inequality: low-skill wages rise, high-skill wages fall, inequality comes down. What I’ve just described is the standard prediction of the theory of comparative advantage. And probably the major reason that proponents of globalisation gave for explaining why globalisation would have a salutary effect on inequality in emerging economies. But that’s not what happens. Interestingly that prediction was valid for most of the other globalisations that have occurred in the last several hundred years. The current globalisation is by no means the first. There was a major globalisation at the end of the 19th century. Thanks to reductions in shipping costs, sending goods and services across the Atlantic, trade between Europe and North America increased enormously. At that time Europe had a relative abundance of low-skill labour and the US had a relative abundance of high-skill labour. Because of this increase in trade between the US and Europe, inequality in Europe fell - just as the theory predicted it would. So that was a great success. And the theory of comparative advantage worked well for a number of other globalisations as well. But it’s been less successful for the current globalisation. There are several predictions that haven’t turned out to be true. If you go back to the building block of the theory, the factor endowment ratios. The theory predicts that the more difference 2 countries factor endowment ratios are, the more trade should take place between those 2 countries, because the greater the gains to trade. But in fact that prediction hasn’t worked so well recently either. There are some countries - African countries, sub-Saharan African counties like Malawi are a good example - that have been basically left out of recent globalisation altogether and do not trade very much with the rich countries of the world. But more important for my talk today, as I’ve shown you, the theory of comparative advantage implies a decrease in inequality in emerging economies which did not take place. That led Michael Kremer, who is development economist at Harvard, and me to think about alternative hypotheses. And what I’d like to show you for the rest of this talk is an alternative theory. A theory that we’ve developed, which we think is not a substitute for the theory of comparative advantage, but I think it complements the theory and does a better job of understanding the effect of globalisation on inequality. And the basic point we try to make is that the reason this globalisation is different is because is that the current globalisation has involved an internationalisation of the production process itself. So to give a typical example: look at computers. Computers are very much the international product par excellence. They might be designed in the US, programmed in Europe, and assembled in China. And this internationalisation of production is the key to this alternative model. In this model we emphasise that there are not just 2 skill levels to look at. In fact, for my purposes today I will look at 4 skill levels. But everything I say can be extended to as many as you like. I’d like to do the same kind of thought experiments that I did for comparative advantage, that is I want to concentrate on 2 countries: one rich and one poor. There are going to be 4 skill levels: A, B, C and D. The rich country will have workers of skill level A and B. The poor country will have workers of skill level C and D, where A is bigger than B, B is bigger than C, C is bigger than D. And we want to do the same sort of pre-globalisation, post-globalisation thought experiment that we did before. In this model wages/incomes will depend on how workers of different skill levels are brought together to produce output. It’s going to be a matching model. You should think of the production process as consisting of different tasks. And we can choose a worker, in principle a worker of any skill level to perform any task. However, some tasks are more sensitive to skill than others. There’s going to be a managerial task which is very sensitive to skill, a subordinate task which is less sensitive to skill. And the amount of output you get depends on the skill level of the manager and the subordinate in your company. For purposes of illustration let’s look at a very simple production function, where output is the square of the manager skill level. The fact that there’s a square there emphases that the output is quite sensitive to the manager's skill level times the subordinate's skill level. So if the manager's skill level is 4, the subordinate's skill level is 3, total output would be 4 square times 3 or 48. Now, there are different ways that workers of different skill levels could be matched. Let’s take a little example where there are 2 workers of skill level 3 and 2 workers of skill level 4. One way of matching them is to match a 4-worker with a 3-worker and the other 4-worker with the other 3-worker. That’s what I’ll call cross-matching, where you have workers of different skills working together in the same firm. In this case you get total output of 96. Or you could match the 4-worker with the other 4-worker, the 3-worker with the other 3-worker, which I’ll call self-matching or homogeneous matching. In this case you get output of 91. If you have an efficient labour market, and we’re going to assume that the labour market is efficient, in this example you’re going to get this kind of matching, cross-matching. But if we change the numbers a bit. If instead we have 2 2-workers and 2 4-workers, we’ve reduced the skill levels of a couple of the workers. So there’s a bigger gap between the high-skill and low-skill workers than before. Now when we compare cross-matching and homogenous matching we get higher outputs with homogeneous matching. An efficient labour market will lead to this kind of arrangement. So the point I’m making is: the kind of matching you get is determined endogenously by the distribution of skills of people who are available to work. And there are 2 forces pulling you in opposite directions. On the one hand because the 2 tasks, the managerial task and the subordinate task, are differentially sensitive to skill. The managerial task is much more sensitive to skill than the subordinate task. We’re going to want to have cross-matching. Because in cross-matching you’re putting the more able worker, the higher-skill worker, in the more skill-sensitive role. On the other hand if skill levels are too different, if the gap is too big, then self-matching, homogenous matching, is better. Because there’s no point in having a very high-skill manager, if the subordinate is of such low skill that he drags down the productivity of the manager. So these 2 forces, pulling in opposite directions, will be balanced in the equilibrium of the model. So now let’s apply this to our 2 countries. I’m going to use particular numbers for the various skill levels. But in fact the qualitative conclusion I’m going to draw is very robust to a wide range of parameters. And let’s compare pre-globalisation to post-globalisation. In pre-globalisation, if you’re a producer it’s not possible to employ workers from 2 different countries. All production must be domestic. So that means that in our example we’ll get A-workers matched with B-workers in the rich country, C-workers matched with D-workers in the poor country. But what happens when it now becomes possible for firms to employ workers simultaneously from different countries, so we can have international production. Well, the matching pattern changes quite dramatically. And we now have the B-workers matched with the C-workers. The D-workers, who are the lowest skill workers, are left to their own devices. Now what implications does this change have for inequality? Before globalisation - so we are back in this picture. Before globalisation the D-workers had the benefits of being matched with C-workers. And that was a good thing for the D-workers. Because if you’ve ever worked with someone more highly skilled than yourself, you understand that your productivity is enhanced by working with someone who is better than you. And that was true for the D-workers as well: Their productivity was enhanced. But after globalisation the D-workers were left to their own devices. The C-workers went off with the B-workers. That was good for the C-workers because they had this nice international employment opportunity. The C-workers see their wages rise. The D-workers see their wages fall, or at least stay stagnant. So that induces the increase in inequality that I’ve been talking about. It’s the fact that the C-workers get the international opportunities and the D-workers don’t. Well, what lesson do we learn from this? If we accept this model - and I should emphasise that there’s lots more work to be done to develop the model to make it genuinely susceptible to empirical testing. The model as I’ve laid it out is too stylised a model at this point to lend itself well to empirical work directly. But if you accept the model then the reasonable policy approach is to try to raise the skill level of D-workers, through job training, through education. So that they have international matching opportunities too. Of course that’s expensive. Providing education and training is expensive. And there’s always the question, who is going to pay for it? Unfortunately, the workers themselves are not likely to be able to pay for it. We’re talking about D-workers, who are some of the poorest people in the world. They can’t afford to. And one problem with saying, 'well, let the market take care of it', is that producers too are not going to have sufficient incentive to do this investment in education and training. Because if you hire me, say I’m a D-worker, you hire me to work for your company and you give me a job training. Well, in the end you’re going to have to pay me more because there’s a competitive labour market. In other words some of your investment is going to be lost to the higher wages that you have to pay me. In fact, I can leave you once I’m trained and work for your competitor, in which case your investment is lost altogether. So producers are not likely to make the investment either, or at least not sufficiently. That means that some third party is going to have to do it. And the obvious candidates are governments. Domestic governments can invest in education and training. International agencies like the World Bank, NGOs. Foreign aid can take the form of subsidies for education and training. But the point I’m making is that it’s not going to happen by itself. There is going to have to be come concerted effort to deal with the D-workers of the world. And so in conclusion, I don’t want you to take away from this talk the idea that I’m anti-globalisation. Because I started out by showing you something that you probably already knew. Which is that on average developing countries have benefited from globalisation, that is GDP per capita has significantly risen as a result. Even if we could stop globalisation it would be counterproductive to do that because of the average gains that have been achieved. The problem is with the distribution of those gains. And what I would like to suggest is that we allow the low-skilled workers, the D-workers, the poorest people in the world, to share in the benefits of globalisation as well. Thank you very much.

Eric S. Maskin (2014)

Why Haven’t Global Markets Reduced Inequality in Developing Economies?

Eric S. Maskin (2014)

Why Haven’t Global Markets Reduced Inequality in Developing Economies?

Abstract

The theory of comparative advantage predicts that globalization should cause income inequality in emerging economies to fall. However, this has not happened in the current era of increasing international trade (although the prediction held up well for previous globalizations). In this lecture, I sketch an alternative theory - developed in collaboration with Michael Kremer - that seems to fit recent history well. Our model conceives of globalization as an increase in international production (Computers provide a good example: they may, for instance, be designed in the U.S., programmed in Europe and assembled in China). We show that when the barriers to international production come down, moderately-skilled workers in emerging economies get new employment opportunities and unskilled workers don’t. It is this disparity, we argue, that accounts for rising inequality in many developing countries.

Content User Level

Beginner  Intermediate  Advanced 

Cite


Specify width: px

Share

COPYRIGHT

Content User Level

Beginner  Intermediate  Advanced 

Cite


Specify width: px

Share

COPYRIGHT


Related Content