Egypt's puzzling income inequality

The Arab Spring is the result of growing inequalities and iniquities and has been a wake-up call for the leaders of other countries where a few privileged dominate the masses. That could be a short summary of the commentary coming out the mainstream media about what happened in Tunisia, Libya, Syria and in particular Egypt. And it is all wrong.

Indeed, Vladimir Hlasny and Paolo Verme point out that there is really nothing special about the income distribution in Egypt, and if anything it has become more egalitarian during this millennium. They can turn the data whichever way, same result. However, it appears from the World Value Survey that the tolerance for inequality has sharply declined. And this change must be coming from other factors that the income distribution.

Are economists really uneasy about studying inequality?

Economists are often misunderstood. People do not understand what we do. They think we spend all our time forecasting the stock market. And the whole profession has been accused of not foreseeing the recent recession. Some economists have redirected this criticism and made a name for themselves by complaining that economic models do not take this or that into account. That is true, but this is often irrelevant, as models are abstractions and they cannot take everything into account. You want to build the right model for the particular question at hand. I have mentioned a few of those essays on this blog, in part because they frustrate me as they are ignoring the very literature they are calling for. There is a lot more to Economics than the principles with perfect markets we tend to teach as an introduction to the field.

The latest paper to frustrate me is by Brendan Markey‐Towler and John Foster. They claim the Economics profession is uncomfortable with issues about inequality to the point of ignoring them. To support this, they quote extensively from the introduction of the Handbook of Economic Inequality, which of course is going to try to make the case that inequality is underrepresented in the literature. Why so? Markey-Towler and Foster claim this has to do with the profession's adherence to Arrow-Debreu markets, welfare theorems, the Hicks-Kaldor efficiency-equity trade-off, and Arrow's impossibility theorem. Because the profession is so enamored in these theorems, it views the impact of inequality to be political only, but of no economic consequence. Never mind that you can still have inequality in such economies. Never mind that every issue of the top journals has papers with such properties and inequality. Never mind that many papers go through great lengths in trying to model observed inequality while studying many issues. I agree not every paper does this, far from this, but then not every answer hinges on inequality. Again, models are an abstraction, and one cannot include everything. One keeps what is most likely to matter. Occam's razor is still valid today.

Markey-Towler and Foster have this distorted and unfortunately common view that economists believe markets are always complete and perfect, and thus inequality cannot happen. This sounds a lot like those who criticize Economics after taking one class, where they learned that free markets and free trade are good. But economists have long realized that things are much more complicated than that, and the study of the departures from this perfect world dominates current research in Economics. In fact, read this blog and you should see that I hardly mention such perfect markets. The authors' solution? Complex systems theory, which I liken to modeling the actors of an economy being linked by a giant plumbing system with leaks, plugs and bottlenecks. That sounds much like the frictions, information asymmetries and imperfect competition we put in our models, except that complex systems theory is much more detailed, requires gigantic amounts of data to calibrate or estimate, and has gone nowhere so far. So researchers had to resort to heroic assumptions to show something could happen, without any ability to validate it empirically.

I do not think this is the way to go, and we can agree to disagree on that. But I take offense at the idea that economists are somehow uncomfortable, even scared of dealing with inequality. That is just not true.

Much of observed income mobility is measurement error

Much of the discussion about income inequality and poverty is vacuous if it does not take into account some form of dynamics. Are the poor of today also the poor of tomorrow? If yes, we have a problem, if not, we have much less of a problem. It thus become quite important to measure properly income mobility, how people move from one part of the income distribution to an other. That is easier said than done. One can take two snapshot of the the income distribution, and then calculate some correlations. But there could be measurement error, always a problem, and income changes may be temporary, artificially biasing upwards mobility indicators.

Tom Krebs, Pravin Krishna and William Maloney make significant progress in this measurement by putting some structure to it. For example, a permanent change in income should be reflected in a change in consumption, while a temporary income change does not. Thus, building a consumption-saving model with an income process that has permanent and temporary components, tying this with income data from households in Mexico to calculate income mobility with the new methods and old ones. It turns out that most of the usually measured income mobility comes from measurement error or temporary income. In other words, there is at least in Mexico much less income mobility than we think there is. This is bad news for economies where we know income shocks have a tendency to be temporary, such as the US. The American Dream is farther that you think (previous exhibits I and II).

Poverty exacerbates self-control issues

Conservative pundits tend to blame the poors' behavioral issues for their fate. Liberals tend to claim that the poor do not have a fighting chance to make it from the day they were born. And of course neither is completely right, the truth is somewhere in between. It is the economists' task to find where that is.

Douglas Bernheim, Debraj Ray and Sevin Yeltekin bring an interesting piece to this question by integrating self-control issues into a standard model of savings. They are careful to limit the time-inconsistency to eliminate extreme cases where someone would never save (which belong to state assistance anyway) or would never use assets. Then it turns out that self-control issues are amplified by initial poverty. The reason that this poverty trap emerges is that if you have few prospects of assets in the future, there is little to lose from deviating from optimal, time-consistent behavior. If you have more assets or you know you should have more in the future, there is a lot to lose from falling in the trap. This implies, as whenever there is a trap, that one can help these poor people by providing them with enough support to get above the threshold level, and then they can happily fend for themselves.

How econophysics describes the income distribution

It has been a while since I last discussed a paper from econophysics, where it appears there is a substantial literature trying to describe the distribution of income. It turns out to be quite difficult, because the goal is to do this with a single equation. What one would want to do with that equation is not clear to me, but anyway.

Maciej Jagielski and Ryszard Kutner claim success with this endeavor by essentially dividing up the distribution in three parts, fitting each to a different distribution function, and then rejoining them into a single equation. But what income are they taking about, you may ask? They look at European income in 2006 and 2008, and take the data from the SILC EU project. That still does not determine what income they are considering, as the dataset allows multiple different ways to define income. It is not even clear whether this is income before or after taxes and whether it includes capital gains.

One problem the authors realized is that they need oversampling for to incomes. To take care of this, they look at the European billionaires on the Forbes list of the richest people over several years, conclude that changes in wealth must be "income" and take that, dropping all negative incomes along the way. Then they notice a large discontinuity from merging the two dataset and decide to divide the top incomes by 100 to make the joint distribution continuous. Oh boy. And this is the dataset they used for their study, believe it or not.

The puzzling evolution of income inequality

There is good evidence now that income inequality is decreasing in some countries at an unprecedented pace. Income is the outcome of a series of factors, including human capital. So, what is happening with the human capital inequality? And what about countries?

Amparo Castelló-Climent and Rafael Doménech use the latest update of the Barro and Lee dataset on human capital to figure this out. They find that overall, income inequality has not changed in sixty years, but human capital inequality has significantly decreased. This has come through important improvements at the bottom of the distribution. The authors conclude that improvements in literacy are not sufficient to reduce income inequality. I would say that this is even worse. Indeed, the Barro and Lee dataset only counts years of education. But we know that the marginal contribution of an additional year of education decreases over an individual's schooling career. Thus improvements in literacy in developing countries should have had even better results in terms of income. At the aggregate level, the authors show the opposite, though. Puzzling.

Strange facts about inequality over the business cycle

The last recession has renewed interest about the distributional impact of business cycles. Clearly, not everyone is affected in the same way by a recession, and the massive policy interventions of the last few years also affected people differentially. For example, what what do we really know about the dynamics of inequality?

Virginia Maestri and Andrea Roventini use the database from a special issue of the Review of Economic Dynamics (which I discussed before). While the sample length is unavoidably short, including only a couple of recessions for every considered country, some general lessons can be learned: income inequality is counter-cyclical while consumption inequality is pro-cyclical. That is going to be tough to replicate in a theory. Imagine a recession. I can imagine that inequality becomes more severe because the people how typically have low incomes are more likely to lose their job. But why would consumption inequality be reduced? Unemployed workers do have lower consumption, and may in fact be more severely affected due to the lack of appropriate savings. So it must be that the consumption of the richer ones drops like a stone, and I do not see a model delivering this.

PS: I realize there was a large drop in income for the richest ones in this recession, and they recovered quickly. But this was not a typical recession.

PS2: And I still hate it when a paper starts on page 9, and the six last pages are back-cover material. What a waste.

On measuring income standards

How do you measure living standards? The usual way is to look at income, because it is the easiest to find. But of course, this only pertains to market income and does not include other potential forms of income, such as informal income, income in goods, and imputed income from home production. All these can be important for poor households, and a better way may be to look at consumption, especially if you want to look at a proxy for permanent income rather than a temporary measure like income.

Mike Brewer and Cormac O'Dea do this for the United Kingdom. They show that the correlation between income and consumption is actually quite low, and defining poverty with income measures can be quite misleading, and they suggest this is due to underreporting of income while consumption expenditures are rather accurate, at least for the poor. This implies that if one looks at consumption expenditures to identify the poor, one should not focus so much on retirees. Indeed, they enjoy substantial housing services, need to sustain fewer people and can draw on the sale of assets. Absent a measure of consumption, using income plus imputed income from housing services does a good job, though.

How to best tax by gender and marital status

It is well known that income taxes are distorting in a way that is not welfare improving, as it discourages labor supply. But as we need government revenue to finance public goods and there is support for some redistribution, we have to live with income tax. Of course, one can discuss whether it would be better to crank up sin taxes to provide revenue and provide lump-sum subsidies (or taxes) to provide for redistribution, but let us suppose we have only labor income tax available. Then it is obvious that tax rates need to be differentiated by labor supply elasticity. That is difficult to elicit from individuals, but there are some individual characteristics that can help here.

Let us focus on gender and marital status. Women have a higher labor supply elasticity, especially when married. One has therefore to be careful not to tax them too much, or they drop out of the labor force. The same applies to a much lesser degree to married men. Gender is mostly unalterable, thus it should be easy to tax by gender, but politics get in the way. It is easier to differentiate taxes by marital status, but the latter is unfortunately endogenous. All this is probably why many countries tax differently by marital status, but not by gender.

Spencer Bastani studies the question using a model where, unfortunately, marriage is exogenous and characterized by perfect assortative matching: the most productive men marry the most productive women. When married, spouse bargain over each one's consumption and labor hours. Critical here is the exogenous bargaining power of the husband. Note that it is assumed that marriage always remains viable, there is no divorce, even though divorce is a threat point. In the end, men should be taxed more than women, unless the bargaining power of the husband is high (where lump-sum payments to women are likely to be higher, so they do not lose completely. And remember, this bargaining power is exogenous, and can be changed by law). If household production has a significant public good component, there should be redistribution from couples to singles. Welfare gains from such taxation are important, especially if wage gaps between genders are large. And this is the situation where it is the most feasible: women are then secondary earners, and one can tax secondary incomes differently without causing too much of a commotion.

Banking crises and income inequality

With the Occupy X movement, discussion about the unequal distribution of income has flared up. At the same time, we are still not over the banking crisis. Several people have linked the two, saying that the large banking sector has lead to more income inequality and that the rich have benefited form the crisis at the expense of the poor. We probably do not yet the data to corroborate any of this, but we have data that allow to look at income inequality through other banking crises.

This is what Luca Agnello and Ricardo Sousa set out to do with a panel dataset from OECD and non-OECD countries. They find some regularities: there is a run-up of income inequality before the crisis hits especially in non-OECD countries; it declines fast thereafter, especially in OECD countries; better access to credit reduces income inequality; and the size of government has no impact on income inequality. The estimates of the paper are rather crude, there is just a lag on the Gini coefficient. I am sure one can tease out more interesting dynamics with a structural vector auto-regression. But the results are still interesting as is.