Sunday, September 12, 2010

Inequality and Growth, Revisited

As politicians debate on whether to extend the Bush tax cuts and, if yes, to who, income inequality has been brought to the spotlight not just as a social plight but also as a structural impediment to growth that should be tackled—not least by repealing the tax cuts.

Inequality is all the more topical in light of data showing that the gap between America’s richest and its poorest has kept widening, and that the US is *the* most unequal country among the advanced economies.

I therefore thought of revisiting the theoretical and empirical findings on the link between inequality and growth and see what policy implications might come out of this exercise.

As is often the case in economics, the theoretical link between inequality and growth is actually ambiguous: For each argument pointing to a negative direction, there is usually an offsetting factor going in the opposite direction.

For example, one argument is that credit market imperfections lead poor households to forgo investing in higher education for their children. High inequality is then bad for growth because (given the diminishing marginal returns on education) the average productivity of the human capital in an unequal economy is low. This is because the poor under-invest in human capital even when return on their investment would have been high; while the rich “over”-invest in human capital even as the return on their investment becomes progressively lower.

Yet, the same credit market imperfections could make inequality good for growth. This can occur if investment and innovation require large start-up costs relative to a country’s median income. In that case, inequality in the form of capital concentration would help increase investment and thus raise economic growth. 1/

Empirically, this ambiguity is confirmed, in that no conclusive link is found between inequality and growth in a panel of countries. That said, as Robert Barro shows, the relationship becomes statistically significant when one splits the group into poorer and richer countries (i.e. when one controls for the level of income).

And as it turns out, the sign of the relation is different in the two groups: In countries with very low per capita incomes, Barro finds that inequality is bad for future growth. But in richer countries, higher inequality is associated with higher GDP growth. The reason for this difference may be that factors that drive the negative relation in poorer countries (such as credit market imperfections impeding higher education) are less relevant for richer countries, so that other, offsetting factors dominate the relationship in those countries.

Now, clearly, the US falls under the “rich” category. So is the rising inequality in America good for economic growth?

In attempting to answer the question, let me first throw in another economic relationship that has seen a high degree of empirical regulatiry in the data: That between the level of per capita income (not the growth) and the level of inequality.

This relationship has the shape of an inverted U (the so-called Kuznets curve): In low-income countries, economic development leads to a rise in income inequality (e.g. as some households shift from agriculture to higher-earning jobs in industry). But as the average income levels rise, inequality tends to decline again (e.g. as more and more households shift to urban areas and earn higher incomes in the industrial sector).

In theory then, the relationship between inequality and growth could be self-stabilizing: If rising inequality in the US were good for growth, the resulting increase in the level of the (average) per capita income should help bring inequality back down eventually (per the Kuznets curve).

But here are a couple of caveats: First, one of the theoretical channels between more inequality and higher growth is through a higher investment than would be achieved in a more equal society. But Barro fails to find a statistically significant empirical relationship between the level of inequality and investment growth.

Second (and perhaps a reason for the above) is the fact that studies that use national data to examine the relation between inequality and growth (within a country) ignore the impact of the globalization of labor and capital. But under globalization, the relationship could change.

For example, from the perspective of a country with a relatively low median income and with credit market imperfections, one may not longer need a high concentration of capital (ie wealth inequality) to start-up productive investments; foreign capital could fill the gap. In that case, inequality could only harm growth by adversely affecting the development of human capital (as explained earlier) and/or the political and social stability of the country.

Meanwhile, from the perspective of a richer country like the US, with sophisticated financial markets that allow its private sector to be a leading capital provider to the rest of the world, the implications may also be different: Higher inequality/concentration of capital could well lead to higher investment—but that investment may not happen inside the US, as companies now have a broader menu of opportunities to consider.

As a result, the feedback loop I conjectured above (higher inequality-->higher investment/growth-->higher average income level-->lower inequality) may not happen… at least within the US. Instead, it helps reduce the inequality between the average American household and the workers of the countries that are the beneficiaries of that investment.

What does this mean for policymakers? Globalization is hard to stop unless one is ready to face major disruptions in the global economy and setbacks in our standards of living. On top of that, you would be intellectually dishonest to be calling for its interruption, if you're the type who argues for more equality: From a strictly intellectual perspective, you shouldn’t care if the person whose income is converging with yours is American or Chinese.

Yet, one can’t deny that people (voters) care far more about their own financial situation than about China’s progress in eradicating poverty. And to the extent that dissatisfaction at home leads to legislative deadlock or even to social unrest, this would undermine the country’s productive potential, leading to lower growth, more capital flight and more inequality.

In my view, reconciling the globalization of investment opportunities with a self-stabilizing income-inequality dynamic at home need not be impossible. This should involve first and foremost a set of policies designed to making the home country a more attractive place for global investment—by raising the quality of human capital and by catalyzing the building of relevant infrastructure that facilitates the conduct of business. This goes back to my argument last week for focusing fiscal policy on increasing total factor productivity (here).

A second area for potential policy revisionism in this context (and one I’ve heard being discussed by some Washingtonians recently) is the way America taxes the profits its companies make abroad. I reserve the right for a separate piece on this as I don’t feel informed enough to have conclusive view. The one thing I would mention though is that my analysis will be heavily biased by my trust in the unfathomable creativity of American tax lawyers and accountants!

1/ For a more comprehensive account of the links between inequality and growth see Robert Barro’s “Inequality and Growth in a Panel of Countries” and Roland Benabou’s “Inequality and Growth”.


Priapus said...

How do you propose making one's country more desirable for investments?
No matter how much the make up artist is clever, he will not fool the eye of the shrewed Capital, so could that cry just end in a call for protectionism?
Opportunities present themselves by themselves, when you try to create them they usually tend to become disasters!
A drop of salty water will loose it's concentration once it falls in a lake full of fresh water!

bb said...

FYI: there are only a few countries in the world where education could be had only with credit, the U.S. and to some extent a few other countries. in the rest of the world, access to education is merit based.

Benign Brodwicz said...

And the macroeconomic figment of GDP is the only variable that matters? There is a wealth of evidence that disparities in SES (socioeconomic status) have severely negative consequences on mental and physical health.

The growth era is likely over. Once we hit some binding constraints on energy social justice will become the foremost issue worldwide, is my guess.

I respectively suggest you reframe this analysis on a broader, more humanistic reference frame.

Ignacio said...

I would distinguish among different ways to create inequality. For instance, If inequality is driven by rent-seekers profiting from a larger share of the economy, because they have dominant positions, surely growth will slow down.

Examples of this could be landowners, or banksters.

Anonymous said...

When someone tells me that the US should have a Carbon Tax because we consume a disproportionate amount of the world's resources, while China should be given an exemption because of its relatively low per-capita income, I respond,

How has per-capita median income done in the US over the last ten years? In China?

They respond, China "deserves" to get closer to our per-capita income, even if the US middle-class has to stagnate in the process.

I say that's unacceptable.

Ironically, more Dems feel comfortable with setting the table for global median wage convergence, even if it means that the US wage will stagnate, than Republicans, who continue to believe that the US should remain exceptional.

And, as a Dem, it pains me to concede that point.

santia said...

I like Dinar.and its revaluation of currency.

Steve Roth said...

Anonymous, the increasing prosperity in other countries could be a win-win, if the win on this side of the pond wasn't being funneled off to a tiny group of our fellow citizens.

Barro showed long ago that in prosperous countries, larger government (determined largely by the size of countries' social programs) if anything correlates with faster growth. But he's been pretty assiduous in burying that finding: