The rise and fall and rise (literally) of the most important curve in economics

| | Comments (0) | TrackBacks (0)

The Atlantic
By: Kentaro Toyama
April 30, 2012

The rise and fall and rise (literally) of the most important curve in economics

Happy birthday, Simon Kuznets! (Also, you might be wrong.)

Today is the birthday of Simon Kuznets, who was awarded the 1971 Nobel Prize in Economic Sciences "for his empirically founded interpretation of economic growth." He passed away in 1985, but like many of his fellow Nobelists, Kuznets left a prolific legacy. He is widely noted as the architect of the national income accounts that allowed accurate estimates of Gross National Product. He laid much of the foundation for the study of business cycles. And, he proposed the idea that with industrialization, nations experience a rise and a subsequent decline in economic inequality that is often characterized as an inverted "U."

This last hypothesis bears his name as the "Kuznets curve," and it has had lasting influence. But was he right? Does continued economic growth eventually lead to decreasing inequality?

There is both evidence and explanation that bears him out. For example, the graph below based on work by economists Thomas Piketty and Emmanuel Saez shows the income share of the top 10% of households in the United States between the years 1917 and 1970. It shows a clear inverted-U, with an initial rise, a peak in 1928, and an overall decline through 1970. Similar curves are found for most Western European countries, though their timing varies. Many developing countries see greater inequality than developed countries, hinting that inequality decreases as a country becomes developed.

Income shares of the top 10% of households in America. From Emmanuel Saez, 2012)

Kuznets called the reduction in inequality "a puzzle," since it seemed to counter forces that ought to increase inequality. First, there is a rich-get-richer phenomenon, as upper-income groups are able to save more and therefore invest more in income-earning assets. Second, increasing industrialization causes more of the population to move from low-income rural agriculture to high-income, higher-variance urban employment. Both trends ought to increase inequality.

So, what might cause inequality to decrease? Kuznets and others have offered a range of explanations. First, a dynamic economy may see more movement into and out of the higher income brackets, effectively eroding and diluting the richest households' share at the top. Second, industrial pay may saturate at the top, as more workers enter into it. Third, the greatest inequality may happen when a population straddles two sectors with differential income, so as the agricultural sector shrinks to nothing, inequality also declines.

The explanations are plausible enough that the Kuznets curve has had a sturdy following. Most often, it is invoked to support growth-focused agendas in the face of rising inequality - in effect, why worry about inequality, if more economic growth will eventually eliminate it? On the surface, the Kuznets curve seems to justify laissez-faire economics and trickle-down policies.

However, Kuznets himself admitted that his conjecture was "perilously close to pure guesswork." He might have been right only for the scope of data he had access to in the 1950s. Since then, we know that many oppressive regimes see growth accompanied by continuously increasing inequality, and fast-growing East Asian countries often saw inequality decline throughout. The inverse-U is not universal.

Even the developed countries that saw an initial Kuznets curve have witnessed a literal turn of events since the 1970s. When plotted through 2010, Piketty and Saez's data shows U.S. income inequality rising starkly since 1970, and OECD data see similar trends throughout Western Europe.

Mostly recently, economist Jordi Guilera proposes that Kuznets's inverse-U should be extended to an "N" shape. Using data from Portugal, where inequality has charted an N-shaped course, Guilera argues that in a post-industrial society, more people shift into high-skill sectors that pay a premium for greater skill. In short, more and more of us are in what is more and more a high-skill meritocracy - with high variance in "merit," there is high variance in income.

What does this mean for policy? One clear solution is that if income inequality is based on inequality of skill, we have to ensure high-skill education and training for everyone.

More broadly, though, whether the Kuznets curve should really be a U, an N, or even an M, it doesn't advocate a hands-off approach to inequality. Inequality happens. And, policy is key to managing it. In 2002, economists Daron Acemoglu and James Robinson proposed that the Kuznets curve is just one possible outcome, one that happens exactly when government leaders respond sensibly to rising inequality and social unrest with greater political inclusion and redistributive policy. The alternatives are an "autocratic disaster" or revolution. Their theory seems to be borne out by recent events in the Middle East.

Kuznets might have agreed wholeheartedly. In fact, he was far from recommending a laissez-faire approach to the economy. In his 1955 paper, he highlighted the role of "legislative interference and 'political' decisions" on counteracting inequality and warned about the tendency of policy to be captured by wealth.

He concluded his paper thus: "Effective work in this field necessarily calls for a shift from market economics to political and social economy."

0 TrackBacks

Listed below are links to blogs that reference this entry: The rise and fall and rise (literally) of the most important curve in economics.

TrackBack URL for this entry:

Leave a comment

About this Entry

This page contains a single entry by CFED published on May 2, 2012 4:22 PM.

New research suggests federal education grants are working was the previous entry in this blog.

Should colleges earn money from prepaid student debit cards? is the next entry in this blog.

Find recent content on the main index or look in the archives to find all content.