The New Yorker
By: John Cassidy
February 26, 2014
In the ongoing debate about rising income inequality, two questions are often raised: one from the left--Is rising inequality impeding economic growth? And the other from the right: Does tackling inequality, which usually involves some form of redistribution, reduce growth?
The questions reflect differing concerns and differing world views. In his 2012 book, "The Price of Inequality," Joseph Stiglitz, the liberal Columbia economist, argued that recent trends in income distribution threatened not just economic growth but the very fabric of democracy. On the other side of the ideological divide, conservative economists claim that tackling inequality--by, for instance, raising taxes on the rich and using it to finance government programs for the poor--has adverse effects on incentives and restricts growth, which is counterproductive for everybody.
Economic theory isn't much help in resolving this debate. It's pretty easy to build mathematical models in which high rates of inequality, by generating higher rates of saving and investment, are associated with higher rates of G.D.P. growth. (Nicholas Kaldor demonstrated how to do it back in the nineteen-fifties.) It's a bit trickier, but also perfectly possible, to construct models in which high rates of inequality--by, for example, constricting saving and investment (especially human-capital investment) among the poor--lead to lower rates of growth. (Brown University's Oded Galor is a leader in this area.)
The debate, then, turns on data. And, thanks to three researchers at the International Monetary Fund, we've got some striking new findings that answer the second question, whether tackling inequality reduces growth, with a firm no. Countries that take redistributive measures in order to attenuate inequitable market outcomes do not, on average, tend to grow less rapidly than other countries. Indeed, the contrary is true. They tend to grow a bit more rapidly.
The research paper, "Redistribution, Inequality, and Growth," has been posted on the I.M.F.'s Web site and authorized for distribution by Olivier Blanchard, the I.M.F.'s chief economist. (I am grateful to Jim Tankersley, of the Washington Post, for drawing my attention to it on Twitter.) Its authors--Jonathan D. Ostry, Andrew Berg, and Charalambos G. Tsanarides--begin by pointing to previous empirical findings that sustained economic growth seems, on average, to be associated with more equal income distribution. But this might not, in itself, make the case for using distribution to attain that equality, they explain: "In particular, inequality may impede growth, at least in part, because it calls forth efforts to redistribute that themselves undercut growth. In such a situation...taxes and transfers may be precisely the wrong remedy."
To figure out whether this is happening, the authors exploit a new dataset that distinguishes between the level of "market inequality" in a given country and the level of inequality after taxes and transfers. The difference between the two types of inequality provides a measure of how much redistribution is being carried out; the authors then use statistical regressions to analyze whether there is any relationship between redistribution and growth rates. Here are two of their conclusions:
Redistribution appears generally benign in terms of its impact on growth; only in extreme cases is there some evidence that it may have direct negative effects on growth.
The combined direct and indirect effects of redistribution--including the growth effects of the resulting lower inequality--are, on average, pro-growth.
The two accompanying charts, which I took from the paper, illustrate these findings. Each of the dots represents a country. The first chart shows a negative relationship between initial levels of inequality (as measured by the Gini coefficient) and subsequent rates of G.D.P. growth. The second chart shows a (slightly) positive relationship between the level of redistribution and subsequent rates of growth.
As with any statistical exercise of this type, some notes of caution should be attached to these findings.
To begin with, they are based on cross-country growth regressions, and it's long been known that many of the results generated by this method of analysis are fragile. It remains to be seen how this study will hold up to inspection by other researchers. Arcane disputes are likely, possibly even within the I.M.F.'s research department. (In the past, other I.M.F. economists have challenged the claim that growth and inequality are negatively correlated.)
The authors, however, are careful not to oversell their conclusions. In the introduction to their paper, they acknowledge the limitations of their data and the statistical tools they are working with. But they also point out that they utilized the best information available from a large number of countries, and that their results hold up even when their analysis is restricted to more recent data, which is the most reliable.
In any case, the big contribution of the paper isn't to suggest that a given re distributive policy--a tax-financed program to encourage kids in developing countries to stay in school, say--will deliver another x per cent of G.D.P. growth. Such a conclusion is beyond our current (and possibly our future) knowledge base. Where the paper will hopefully make a mark is in challenging the presumption, which until pretty recently was all too common in economics, that countries that try to reduce inequality inevitably pay a price in terms of lower growth. Once again, here is what the authors say: "We should not assume that there is a big tradeoff between redistribution and growth: the best available macro-data do not support that conclusion."