This is the full text of my interview for a Korean paper Hankyoreh. The questions were formulated by Professor Kang-Kook Lee from Ritusmeikan University in Kyoto.
1. First, could you please introduce your book briefly. What was your motivation to write this book, and what is your main argument in it?
My motivation was to present a picture of the world and the distribution of income and economic power in it during the era of globalization. This is a very remarkable period in terms of its effects on income distribution, not solely within countries but between countries as well. It is probably the greatest reshuffle of global income positions with people from formerly poorer countries going up in the global income distribution since the Industrial revolution. I thought it was very important to describe and analyze these changes.
Since the book deals with the world, it is, if I may say so myself, “rich” in implications and messages. I would single out three key arguments. First, for inequality within countries, I argue that it follows the Kuznets waves of rising and decreasing inequality where the increases are driven by the interaction of technological change and globalization. Second, I argue that huge inequality between countries’ mean incomes (or GDPs per capita) implies the existence of a “citizenship rent”, an “unearned” income which depends on one’s chance of having been born in a rich country as opposed to being born in a poor country. Migration is an attempt by people from poor countries to appropriate that rent too. So migration today should be seen to derive from globalization conducted under the conditions of very uneven income levels among the countries. Third, I argue that the effects of globalization will always be unequal. Even if everybody gains, the gains will be perforce different as is the case with all dramatic political and economic shifts. It is naïve to believe that everybody will benefit equally. In addition, differences in gains, given that the world is politically organized in nation-states, may produce political problems that can threaten the sustainability of globalization.
2. Your argument about the causes for rising inequality is so extensive and far-reaching. Mainstream economists emphasize that technology plays the most important role, while you argue that technology, opening and policy are all important and interdependent. Could you explain how your argument is different from mainstream explanations?
The gist of my argument is that the three generally considered the most important factors that increase inequality—globalization, technological change and economic policy—can be distinguished conceptually but cannot be broken apart in an empirical analysis. This is because they are interrelated: openness to international trade and free movement of capital affect policy decisions like the maximum tax rate that can be imposed on capital; the type of technological change we witness is also affected by globalization. I would like the explain this last point. Technological progress is not an abstract thing that is unrelated to the real world. The types of innovations that are actually implemented depend upon the prices of the factors of production (capital and labor). And they in turn depend on whether there is globalization or not. Globalization has made it possible to produce many new goods, from laptop computers to smart phones, using the cheap labor in Asia. Without globalization, these goods would not have been produced in such a particular shape and form because the cost of labor would have been higher. The existence of globalization influences the type of technological progress that is profitable and that therefore gets implemented.
Because of the interdependency of globalization, technological change and economic policy, if we were theoretically to assume that one of these three, say globalization, does not hold, the other two may not hold either. Thus a neat decomposition into globalization, technological change and policy with one accounting for x percent of inequality change, the other for y, and the third for 1-x-y, simply cannot be done.
3. Populist politicians blame globalization for rising inequality and many people accept it. How true do you think is this argument? It seems like that people believe globalization or policy could be controlled to some extent, while technology is comparatively exogenous, and that’s why technology is not political concern. In that sense, isn’t technology qualitatively different from other factors?
I agree that in principle technology seems exogenous. I emphasize the word “seems” because in reality, as I already mentioned, the choice between different technological innovations is made based on the actual prices and profitability of making such goods or services. Take the example of supersonic passenger airplane. We have had the required technology for at least half a century. But the use of this technology was always limited and after the crash of Concorde (which had nothing to do with technology as such) no further flights were undertaken. Why? Because supersonic passenger jets are not profitable. We see on this example, and I think one could find many others, that the technology that actually gets implemented depends on the type of the political and economic system that we have and thus on the prices of labor and capital. And globalization affects both of them. One type of technology will be profitable with globalization, another type of technology would be profitable without globalization.
4. The concept of Kuznets waves is truly interesting. But I think that forces for these waves, up and down, could be different historically between the first and second wave. (for example, aging population is now a force to raise inequality, different from the past) What could be the difference in these forces over history and why?
Yes, I agree that some of the forces behind the first and the second modern Kuznets waves could be different. Demography as you mention is one of them. Education is another. During the downswing phase of the first Kuznets wave, which in rich countries lasted from around the 1940s to the 1980s, expansion of education played an equalizing role by reducing the wage gap between the more and less skilled labor, The average number of years of education in the most advanced countries increased from 6 or 7 to over 13. But today the increase in education levels cannot play the same role simply because most that can be done is go from an average of 13 years of education to say, an average of 14 or 15. In a situation where most people are highly educated you cannot expect to obtain big effects on inequality by making the additional small groups better educated. Similarly, trade unions have had a large role in reducing inequality in the past. But the new type of employment structure, away from factories and towards services with much smaller size of units, makes unionization of workers more difficult and hence the role of unions and collective bargaining between employers and employee less important.
5. As to the prospects of inequality in the US in the 21st century, you are concerned about a perfect storm due to several factors such as rising capital share, correlation of high labor and capital income, rising power of the rich and so on. You also write that you are skeptical about redistribution and education. Although it is extremely hard to predict the future, what could be the most plausible force to possibly reduce inequality in the future in the US? In particular, you sound somewhat politically pessimistic about the future of inequality in the US. But we saw that so many people supported Sanders (and Corbyn), demanding equal distribution. Do you think that there is no much hope for changes in politics for equal distribution in the US? I mean, what kind of political efforts should be made by the liberals and left for prosperity for all?
You are right that I am somewhat pessimistic regarding the short-term prospects for the reduction of income and wealth inequality in the United States. I am even more pessimistic now as we face the Trump administration (one should remember that I wrote the book before Trump even began his run for the Republican nomination). But over the medium term I see at least two important factors that could check the increase of US inequality and perhaps reverse it. The first is political one, reflected in, as you mentioned, the support received by Bernie Sanders. I believe that in the next electoral cycle anti-inequality policies advocated by Sanders may be expected to obtain the support of many voters whether it is Sanders or somebody else who champions them.
The second element is the gradual erosion of large rents that have accrued to the entrepreneurs and capitalists that have introduced new ideas and technologies as the new competitors emerge. Here take the examples of Nokia and Blackberry that were the leaders in their respective fields only to be displaced by new producers and then driven out of business. Similar fate may befall other technological leaders. Its upshot would be the reduction in large rents received and which clearly increase inequality.
6. Recently, many people are concerned about technological progress such as robot and artificial intelligence could lead to very high unemployment and inequality. What do you think of the shock of this technology to inequality? Opposite to this, you present a possibility of technological change that favors low-skilled workers, reducing inequality, (which reminds me of a recent model presented by Acemoglu.) Could you give us more extensive explanation about it?
I would divide your question and my answer into two parts. First, the fears of technological progress displacing all labor are, I think, exaggerated, Such fears have been with us from the very first technological revolution two hundred years ago. And while each change has substituted some labor and negatively affected some occupations, overall it created more jobs. I do not see any reason why this may not happen again. Second, I see some technological advances that indeed can reduce incomes of the rich in developed economies and thus reduce inequality: doctors, accountants, architects, designers, professors can be in many instances replaced by cheaper labor working remotely in emerging market economies, This is indirectly pro-low skill change because it reduces the gap between wages of highly skilled and low skilled labor.
7. You say that, in the 21st century, reduction in inequality in endowment such as assets and education will be more effective than redistribution since taxing capital is harder with globalization. I am not sure how feasible this is in reality. Probably you believe that the power of the government has weakened significantly against capital but the role of the government in terms of tax and government spending was not so much reduced in reality in spite of globalization. What do you think?
I think that there are areas like the equalization of access to high quality education and deconcentration of asset ownership, where the government can play a role unconstrained by globalization. If you have a political will to help small investors by giving them substantial tax benefits or if you help worker ownership of shares through the so-called Employee Stock Ownership Plans (ESOPs), you can reduce the concentration of income from capital and there is nothing that globalization can do to help or made such a policy more difficult. So I think that in some policy areas which have to do with distribution of income, governments can be more active and interventionist regardless of globalization. In others, like taxation of capital or taxation of high-skilled labor, globalization does place limits on what governments can do.
8. You show that the recent fall in global inequality is mainly owing to economic growth in Asia such as China and India, while other developing countries in Africa are behind. Then we can say that gains of globalization are not evenly distributed, not only within a country but also in the global economy. What should be done to change this in developing countries and in the global level?
Yes, the gains are not evenly distributed and are unlikely ever to be. This is because globalization is a huge and multifaceted development that is never going to affect everybody equally: some will gain more, others will gain loss, and some might even lose in absolute terms.
Regarding African countries, it is a different and difficult question as to what they could do to benefit more from globalization. I do not address this in my book. But if we look at the Asian examples one could say that for Africa improvements in skills and greater incentives to attract foreign capital should help. I know however that this is too general to be really helpful and applicable to individual countries though.
9. In the book, you do not discuss implications of inequality to growth much, but you say that the decline in middle class led to changes in consumption pattern and social expenditure, which may hamper growth. It could be consistent with studies that report that inequality is bad to long-run economic growth. What do you think of the effect of inequality on growth?
This is a huge area of research that was very active in the 1990s but has since gone into a decline because the results from the 1990s were inconclusive, The key relationship between income inequality and growth was all over the place: from the authors who argued that inequality is good for growth to those that found no relationship and those that found a negative relationship. I think that we are now on the verge of another wave of such studies because of much better data that we currently have. In the past the studies were very rough: they just looked at the growth rate of GDP and Gini or another aggregate inequality index. But now we have detailed micro (household-level) data for a number of countries and over a long period of time. In a recent paper that I wrote with Roy van der Weide from the World Bank, we looked at different types of inequalities (among the poor, the middle class and the rich) and how they affect income growth at different parts of distribution. It is a much richer way to look at the question. And we are not the only ones who on such much richer data find that higher inequality reduces subsequent growth rate of the poor.
10. Now, let’s talk about globalization, and Brexit and election of Trump in view of your book. These events made your ‘elephant curve’ so famous. Do you believe that those are caused by the revolt by losers in globalization in developed countries? It should be noted that these events occurred in Anglo-Saxon countries where the role income distribution is limited, while inequality has risen higher. Then, could the future of inequality and globalization in other developed countries be different from Anglo-Saxon countries?
The low growth of middle class incomes in rich countries that is highlighted by the “elephant chart” is not the only reason behind the results of the Brexit referendum and the election of Donald Trump but I do think, based on the empirical evidence regarding the votes, that it played a role. It would be of course wrong to say that it alone explains the outcome, There are other reasons. But some of them too relate to globalization like for example migration fears which played a role in the United Kingdom Leave campaign.
10. What do you think of criticism against the elephant curve by a paper at Resolution Foundation that population growth in China played an important role? (with constant population, the result seems somewhat different)
The key point that the Resolution Foundation paper tries to make is that different rich countries had different outcomes when it comes to their middle class income growth---so that the dip around the 80-85th global percentile, where the rich countries’ middle class are, is not universal. But while the middle classes of different countries did indeed experience different income growth (e.g. UK middle class incomes grew faster than American and German) they all still grew much slower than incomes in Asia. And it is the gap between these two that gives the “elephant chart” its distinct shape. And in particular, the middle class incomes in three largest rich counties (US, Japan and Germany) grew very little or almost not at all.
China of course plays a key role in the “elephant chart”. The chart looks different if you take China out, but other than highlighting the extraordinary importance of China, it is not clear to me that that makes much sense. Finally, when you keep the positions of the underlying country fractiles at their 1988 positions in global income distribution (thus effectively controlling for population dynamics), you still get the elephant-alike chart with the highest growth rates around the middle and the top of the global income distribution.