Three inequality tales

Being at talks and panel discussions at Columbia (yesterday) and Harvard JFK School today stimulated me to make three following observations, all on income inequality. I couch it into three inequality tales.  

           Inequality spillovers

            I  presented this paper yesterday, but will not discuss it now (for a summary of the paper, see here). In the discussion after the presentation I went back to something that impressed me in Stiglitz’s paper on which I wrote a blog several months ago. Suppose that income inequality increases in only one county in the world, which is sufficiently big. Let’s call it China. Then suppose that people from China simply in order to diversify their portfolios, or because they like New York, or because they are worried about political developments at home, decide to start buying real estate in the United States. The owners of the plush properties in the United States, for which the demand goes up, are rich Americans. Then suddenly since their assets go up in price, wealth inequality in the US increases, and if returns to wealth are more or less fixed, income inequality increases as well. There is thus what I call an “inequality spillover”. Unlike in the pre-globalization era when, at least in principle, you could have had inequalities in different countries  “kept” in separate compartments, this may no longer be possible during globalization.

            Could this be one of the forces behind the noted convergence of countries’ inequality levels (see the graph below, indicating that countries with higher inequality before 1980 had smaller increases or even declines in inequality since)? I doubt it. The "spillover" forces are not strong enough. They are still marginal. But in some countries they can indeed exert a more important influence. Portugal has recently liberalized its residence rules in an effort to attract rich non-EU citizens. The unintended result may be an increase in wealth and income inequality domestically. 

            Inequality of opportunity

            In a very nice presentation at JFK School of Government at Harvard today, Chico Ferreira showed the graph below (coming from this paper), plotting the estimated inequality of opportunity (from household surveys) against the correlation between parents’ and children’s incomes. Now, both measures have something to do with unequal opportunities, but they are very different and are calculated from entirely different data sources—so the fact that they move together is important. The inequality of opportunity measure (on the vertical axis)  is the share of all income inequality accounted for by the differences in “exogenous circumstances”, that is race, sex, parental background etc. This is the “bad” inequality, inequality which in principle should be zero. The variable on the horizontal axis is the correlation of income between  parents and  children. The higher the correlation, the less we have equality of opportunity. Ideally, we would expect that the distribution of children’s income be the same for all levels of parental incomes and that the correlation be zero. 

 

            The implication is that the work which tries to decompose inequality into  “good” and “bad” inequality (which I think is one of the major new developments in inequality economics) is showing very promising results. First, two papers by Gustavo Marrero and Juan Gabriel Rodriguez, here and here, showed that “bad” inequality is negatively correlated with economic growth in Europe and the US (and conversely, that “good” inequality is positively correlated with growth). Second, high correlation between this measure of “bad” inequality and inter-generational income correlation shows that we are really up to something: we are measuring something that does make sense and is proxying inequality of opportunity.

            The janissaries

             Chico Ferreira ended his talk by pointing out that if you followed Roemer and Rawls and wanted to equalize opportunities, it is not sufficient that the kids from both privileged and poor backgrounds go to the same schools. What is necessary is to have the very opposite of what we have today: to have kids from poor backgrounds attend better schools. And then he highlighted the importance of parental background by saying that hypothetically equalization of chances would be furthered if Harvard students in the audience would marry students from less privileged and poorer backgrounds, attending less fancy schools. But of course, Chico said, that would imply violation of human rights. So, some measures to reduce inequality of opportunity simply cannot be implemented.

           That last comment made me think of radical historical attempts to equalize opportunities. I thought of at least three. Plato’s idea that children should  be taken away from the parents, and be raised and educated by the state. It would certainly reduce inequality of opportunity and would break correlation between parental and children’s incomes too. It was implemented, partly, in Sparta wherefrom Plato took it for his "Republic" and "Laws". 

            The second radical policy was Ottoman policy of abduction of Christian male children (devşirme in Turkish), conducted in some parts of the Empire. The boys were then trained in the capital to become either an elite military corps (the janissaries, meaning literally “the new army”) or imperial administrators. It provided an extreme form of upward mobility for many of them; some became the viziers (prime ministers of the empire) most notably, Sokollu Mehmed Pasha (in Serbo-Croatian-Bosnian Mehmet Pasa Sokolovic), a kid abducted from his parents in Bosnia around 1510. Mehmed Pasha served for 15 years as Prime Minister under Suleiman the Magnificent and two other sultans. The policy nicely illustrates the trade-off:  yes, you can equalize chances of the kids (for what opportunities to rule would a Christian kid born in a small Bosnian hamlet ever have in life?), but you have to run roughshod over human and parental rights. You make parents possibly forever unhappy, and destroy the entire cultural milieu of a community. But  you do achieve upward mobility.

            The third example is from Communism. There too, there were strong “affirmative action” policies that favored children from workers’ and peasants’ backgrounds. The idea was precisely to compensate such families for unfair treatment they had under bourgeois or semi-feudal governments. I think that it generally worked. If you look simply at Soviet leaders from the late 1950s to Gorbachev, and then more recently to most post-Soviet leaders, they all came from modest backgrounds. This was precisely the result of affirmative action policies. It was no longer the children from affluent or intellectual backgrounds who, like in the early stage of the revolution, were in power. And obviously, aristocracy was gone.

            The same is true in China today. However, these families have now created a new class structure where their own children remain on top. It is notable that both the current Chinese president Xi Jinping and the man who could have become president, Bo Xilai are “princelings”, i.e., come from prominent second-generation Communist families. This point suggests how in order to counteract privilege, affirmative action needs to be permanently reset in motion  (the Cultural revolution comes to mind), but the previous stories also show how equalizing opportunities can, in its turn, be unfair and unjust. A theme for philosophers.