Editor's Note: Andrew G. Berg and Jonathan D. Ostry are, respectively, ASsistant Director and Deputy Director in the Research Department of the International Monetary Fund. The views expressed here are those of the authors and should not be attributed to the IMF.
Looking back at Irving Kristol's 1980 essay "Some Personal Reflections on Economic Well-Being and Income Distribution," as Foreign Affairsrecently did, provides a useful intellectual lens from the past to focus the economic conversation today. Kristol argued that economic inequality was "but one manifestation of how nineteenth-century ideologies - and most especially the socialist ideologies - have so decisively shaped modern social science." Moreover, he wrote, income distribution does not really change over time so it is, as a subject for study, inconsequential.
Fortunately, economists failed to take his advice; recent studies of inequality reveal the limitation of Kristol's historical perspective. Kristol narrowly focused on one long spell of stable and relatively even distribution. But a careful look at the varying levels of inequality in different countries demonstrates just how much societal divides in wealth really matter. Countries with high inequality are far more likely to fall into financial crisis and far less likely to sustain economic growth.
It is a coincidence that just as Kristol was writing, the United States was set to undergo a dramatic economic transformation. In the 30 years that followed (see chart above), income inequality grew significantly, rising gradually in the early 1980s, and then later more sharply. So, at least in some sense, Kristol's argument can be forgiven as a victim of circumstance.
Still, economic inequality was a significant phenomenon long before 1980. Kristol chides the National Bureau of Economic Research for its excessive concern with income distribution in the 1920s. But as a recent study by the economists Thomas Piketty and Emmanuel Saez shows, income inequality in the United States reached unprecedented heights in the 1920s.
Indeed, a growing body of economic literature suggests that inequality can lead to system-wide financial crises. For example, in his book Fault Lines, the economist Raghuram Rajan points at the political and economic pressures that in recent decades led high-income individuals in the United States to save and low-income individuals to borrow and spend. Financial institutions and regulators encouraged this process, creating a dangerous credit bubble.
Our own work on the relationship between income distribution and growth has found that other inequality produces other negative results. Some years ago, we set out to understand what sustains long periods of high growth, or "growth spells." Other economists, notably Lant Pritchett, had come to realize that countries rarely enjoy a steady economic climb. Periods of growth are often punctuated by collapses and stagnation. Moreover, spurring growth is often not the hard part; even in the poorest regions, growth takeoffs are relatively common. The hard part seems to be keeping growth spells going.
What stood out most in our research was that growth spells were much more likely to end in regions with less-equal income distributions. The effect is large. If Latin America, for example, could bridge half of its inequality gap with East Asia, its growth spells would last twice as long as they do now. Remarkably, inequality made a big difference in our results regardless of the other variables we included or exactly how we defined growth spells, a claim that we cannot make for the other factors seen as conventional drivers of good growth performance.
An equal income distribution is of course not the only thing that contributes to economic health, but from our analysis it belongs in the pantheon of well-established growth factors such as the quality of a country's political institutions or its openness to trade.
The immediate role for policy, however, is less clear. Increased inequality may shorten growth duration, but some inequality, as Kristol would surely note, is integral to the effective functioning of a market economy and to providing the incentives needed for investment and growth. And poorly designed efforts to lower inequality could grossly distort incentives and thereby undermine growth, hurting even the poor. It would thus be a mistake to conclude that policies should be geared to reduce inequality at the expense of pursuing all other social and economic goals.
There nevertheless may be policies that both lessen inequality and promote growth, such as better-targeted subsidies, improvements in economic opportunities for the poor, and policies that promote employment, such as labor market training. And even when there are tradeoffs between growth and equality in the short run, much of this new research argues that attention to inequality can bring significant long-run benefits for overall economic growth. Growth may simply not be sustainable in the face of high or rising inequality.
Kristol concludes his 1980 essay by sympathizing with the misfortunate economists who would like inequality to matter, noting, however, that an excessive attachment to ideology is destructive to the integrity of economics as a scientific discipline. On this latter point, we are inclined to agree.