Category Archives: poverty

Piketty’s Fence

Share Button

Most of the reviews of Thomas Piketty’s book Capital in the Twenty-First Century have already been written.  But I thought it might be best to read it all the way through before offering my own thoughts on this book, which startlingly rose to the top of the best seller lists last April.  It has taken me five months, but I finished it.

One of the things the book has in common with the Karl Marx’s Das Capital (1867) is that it serves as a rallying point for the many people who are passionately concerned about inequality, regardless whether they understand or agree with the specific arguments contained in the book in question.  To be fair, much of what Marx wrote was bizarre and very little was based on careful economic statistics.  Much of what Piketty says is based on careful economic statistics, and very little of it is bizarre.

A problem with having a debate about “inequality” is that there are many different measures of the distribution of economic resources and criteria for evaluating it.   Yes, we should not be content with summarizing all of economic performance into the single criterion of aggregate GDP.   But neither is it possible to summarize all other important dimensions of the distribution in a single second statistic to measure inequality.  Trends in within-country inequality can look very different from trends in cross-country inequality, for example.  And so forth.

Within the United States, income inequality by most measures has been rising since 1981, and by 2007 had approximately re-attained the peak of the early 20th century [Fig.1.1, p.24].  On this much we should be able to agree. The same is true within other major English-speaking countries (the UK, Canada and Australia).  In these countries, income inequality had declined sharply during the years 1914-1945, as it did also as in France, Germany, Japan and Sweden.  But in the latter group income remains far more equally distributed than it was at the peak of inequality a hundred years ago.

Economists, at least in the United States, have focused on one or the other of several kinds of inequality:

  • First is the difference in wages or incomes between “skilled” and “unskilled” workers, defined according to whether they are college-educated.  Here the higher wages of the better off group are often agreed to reflect the economic value of their skills in an increasingly technological economy, and the question is how to improve the skills of those on the other side of the fence.  (This has little to do with the gap between the upper 1% and the rest.)
  • Second is the very high compensation of corporate executives, and others in the financial sector.   The financial crisis of 2008 left many observers understandably doubtful of claims that this compensation is a return to socially valuable activities.
  • Third is the “winner take all” aspects of many professions, from dentists to university professors to sports and pop stars.   In a society that identifies, thanks to modern technology and culture, who is the best dentist in town or the best football player in the world, relatively small differences in abilities win far bigger differences in income than they used to.
  • Fourth is “assortative mating,” according to which highly accomplished professional men no longer marry their secretaries but instead marry highly accomplished professional women.

Piketty’s main concern is none of the above.   He discounts the skills gap explanation.  [“This theory is in some respects limited and naïve.” (p.30­5)]    He doesn’t say much about the “winner take all” or “assortative mating” phenomena.  He does have definite concerns about excessive executive compensation in areas like finance, but they are not the primary concern of this book.

The central concern of the book is none of these varieties of income inequality, all of which have to do with “earned income” (wages and salaries).  It is, rather, what he sees as a 21st century trend toward inequality of wealth, to be brought about by steady accumulation of savings among the better off, passed down along with accrued interest to the next generations in their inheritances.

This is a prediction about what will happen in the future more than an observation about a recent trend [figures 10.3-10.5; pages 344-349].   The increases in various measures of inequality since the 1970s have had more to do with earned income than with wealth.

It is true that the capital share of income (interest, dividends and capital gains) rose gradually in major rich countries during the period 1975-2007, and the labor share (wages and salaries) fell [Fig.6.5, p.222].   This trend would support Piketty’s hypothesis if it continued.  He deserves credit for pointing out the lack of foundations behind assertions that capital’s share will necessarily revert to a long-run constant, which used to be declared as approximately 0.25.

His vision is focused squarely on the truly long run, however: century-long trends, not decade-long fluctuations.  For example the Global Financial Crisis is a blip for him.   Incidentally, it is a blip that runs counter to his ultra-long-run hypothesis:  His statistics clearly show a discrete fall in inequality and in capital’s share in 2008-09, because asset prices plummeted.

Three century-long movements constitute the essence of the book:  a rise in inequality in the 19th century, a fall in inequality in the 20th century, and a predicted return to high inequality in the 21st century.

  • Piketty argues convincingly — not just with statistics but also with frequent citations of the novels of Honoré de Balzac and Jane Austen (as remarked by most reviewers) – that the first increase in inequality in France and Britain, mostly from 1800 to 1860, took the form of capital accumulation.   A fence divided the rich rentiers, who lived off their interest, from everybody else, who had to work for a living.
  • The most dramatic movement in Piketty’s graphs is the second one, the fall in inequality concentrated in the period from 1914 to 1950 [figures 8.1, 8.5, 9.2, 9.3 and 9.5, on pages 272, 291, 316, 317, and 317, respectively].  It is attributable to the destruction of capital in two world wars, inflation, the 1929 stock market crash, and a historic social movement in the 1930s and 1940s toward big government and progressive taxation.
  • What is surprisingly scarce in his numbers is evidence that the third movement, the renewed upswing in inequality, which he thinks may have already started since 1980 in the English-speaking countries, is due to a shift from labor back to capital.  The share of capital income in the UK and France remains far lower than it was in 1860. The increases in various measures of inequality since the 1970s have had more to do with shifts within labor’s share (between different categories of earned income) than with wealth. Today’s rich work, unlike those characters in Balzac and Austen and Balzac.

Thus the hypothesis is much less of an explanation for the past (whether the past 6 years, 35 years, or 150 years), than it is a prediction about the future:  a prediction that capital will accumulate and the rich will get richer through inheritance and capital income (”unearned income”) rather than through outlandish salaries and stock options (“earned income”).  For all the impressive collection of historical data, the prediction is based mainly on a priori reasoning:  income distribution must tend to inequality because savings accumulate.

But one could just easily find other a priori grounds for reasoning that countervailing forces might kick in if things get bad enough.  Democracy is one such force.  Progressive taxation arose in the 20th century, following the excesses of the Belle Époque.  A political trend of that sort could recur in this century if the gap between rich and poor continues to grow.

A few years ago, American voters and politicians were persuaded to reduce federal taxes on capital income and estates.  They phased out the estate tax completely (effective in 2010), even though this would benefit only the upper 1 per cent.   This is standardly viewed as an example of the rich manipulating the political economy for their own benefit.  Indeed, we know that campaign contributions buy some very effective advertising.  But imagine that in the future we lived in a Piketty world, a return to the golden age of Austen and Balzac, where inheritance and unearned income were the sources of stratospheric income inequality.  Would a majority of the 99 % still be persuaded to vote against their self-interest ?



Frank, Robert H., and Philip J. Cook. 2010. The Winner-take-all Society: Why the Few at the Top Get So Much More Than the Rest of Us. (Random House).

Goldin, Claudia, and Lawrence Katz. 2009. The Race Between Education and Technology (Harvard University Press).

Greenwood, Jeremy, Nezih Guner, Georgi Kocharkov, and Cezar Santos. 2014. “Marry Your Like: Assortative Mating and Income Inequality.” American Economic Review, 104(5): 348-53.

Piketty, Thomas. 2014. Capital in the Twenty-First Century (Belknap Press).


[A shorter version appeared at Project Syndicate.]

Share Button

Modi, Sisi & Jokowi: Three New Leaders Face the Challenge of Food & Fuel Subsidies

Share Button

In few policy areas does good economics seem to conflict so dramatically with good politics as in the practice of subsidies to food and energy.  Economics textbooks explain that these subsidies are lose-lose policies. In the political world that can sound like an ivory tower abstraction.   But the issue of unaffordable subsidies happens to be front and center politically now, in a number of places around the world.   Three major new leaders in particular are facing this challenge:  Sisi in Egypt, Jokowi in Indonesia, and Modi in India.

The Egyptian leader is doing a much better job of facing up to the need to cut subsidies than one might have expected.  India’s new prime minister, by contrast, is doing worse than expected – even torpedoing a long-anticipated WTO agreement in the process.  In Indonesia it is too soon to tell.

*   *   *


Egypt’s new president, Abdel Fattah al-Sisi, did something In July that few leaders in North Africa or the Middle East had been able to accomplish before: he sharply cut longstanding fuel subsidies and allowed prices to rise (by 41% to 78% and more).   Surprisingly few protests materialized.

Egypt’s food subsidy program badly needs reform as well.  The country has been spending over $5 billion a year on food subsidies.  The price of bread has been kept so low that, famously, it is often fed to animals. Past attempts to reduce such subsidies in North African countries have brought unrest and even government overthrow.  But it looks like the Sisi government is beginning reforms here as well.  Bread subsidies have already been cut by 13 per cent.

In a sense, he had little choice.  Egypt’s fiscal path under his predecessor’s policies was unsustainable.   Even with subsidy cuts, the current government only hopes to bring the budget deficit down to 10 per cent of GDP in the coming fiscal year, as opposed to 14% otherwise.   Still, who would have expected President Sisi, who took office in a fragile political environment, to start off with far more serious fiscal reforms than Indian Prime Minister Modi, who came to office amid hopes for sweeping economic reform and with a whopping democratic majority?

*   *   *


Also in July, Joko Widodo (“Jokowi”) was elected President in Indonesia, another country with a long history of fuel subsidies that it can no longer afford [currently costing $21 billion, or 13%-20% of government spending].  Outgoing President Yudhoyono took the first courageous step of raising fuel prices a year ago.  [The fuel subsidies problem is even worse in oil-producing countries like Iran, Saudi Arabia and Venezuela, where the right to dirt-cheap fuel is considered the national patrimony.  Indonesia itself is no longer an oil-exporter, in part because low domestic prices have encouraged rapid growth in fuel consumption.]  Jokowi does not take office until October, but his advisers favor cutting the remaining subsidies. He has forthrightly expressed an intention of doing so, gradually, over a 4-year period.

*  *  *

Market Failure vs. Government Failure

Why do economists feel confident that they are right to oppose these commodity subsidies?  Agricultural and energy markets tend to be relatively close to the ideal of perfect competition, with large numbers of consumers on the demand side and producers on the supply side.  (Where competition in commodities is imperfect, government itself is usually the problem, not the cure.)    Thus the classic economics argument applies:  setting the price artificially low, with the motive of benefiting consumers, works to discourage production and creates a gap of excess demand that usually has to be met by rationing.    Setting the price artificially high, with the motive of benefiting producers, discourages consumption and creates a gap of excess supply that often ends up in wasteful government stockpiles.   And either way, the policy is an invitation to corruption.

Skeptics of the Invisible Hand point out that, left to themselves, private markets can fail in a number of ways.  Two of the most prominent justifications for government intervention in the marketplace are environmental externalities such as air pollution and inequality in income distribution.  What is so striking about subsidies for food and fossil fuels is that, notwithstanding that the policies are often promoted in the name of the environment or equity, in practice they usually do little to help achieve these goals and often have the opposite effect.   Less than 20% of Egyptian food subsidies go to poor people.  In India, less than 0.1 per cent of rural subsidies for LPG (Liquefied Petroleum Gas) go to the poorest quintile, while 52.6 per cent go to the richest.  Gasoline (petrol) subsidies in most countries go to the middle class; they are the ones who drive cars, while the poor walk or take public transportation.  The same is true of other forms of fossil fuel subsidies:  worldwide, far less than 20% of the benefits go to the poorest 20% of the population. 


Typically they are also ruinous for the budget, as in all three countries considered here.

*  *  *


Can one single misguided policy do worse than simultaneously hurt economic efficiency, the environment, equity, corruption, the government budget, and the trade balance?  Yes it can.   It can derail the most important progress in multilateral trade negotiations of the last ten years!   This from a country, India, where a new prime minister was widely considered likely to tackle much-needed market reforms.

Agricultural subsidies sometimes seek to keep prices low (to benefit consumers at the expense of producers), especially in poor countries, and sometimes seek to keep prices high (to benefit producers at the expense of consumers), especially in rich countries. India’s policies try to do both. As a result, India has accomplished the extraordinary feat of rationing grain to consumers at artificially low prices through a card system and yet at the same time suffering excess supply from farmers, because they are paid high prices.   (Agricultural inputs are also subsidized — electricity, water and fertilizer — thereby delivering the requisite environmental damage.)  The government has had to buy up huge stockpiles of surplus rice and wheat that are rotting away [reportedlymore than 20 million tons of rice and 40 million tons of wheat reserves, at present].  The limited amount that is available for consumers is allocated in ways that are both corrupt and inconsistent with the stated goal of helping the poor.

The government would like to keep its subsidies and stockpiles. But it knows that this would violate international trade rules.  Modi has taken the unusual step of vetoing the Trade Facilitation Agreement that WTO member countries thought they had reached at a climactic summit in Bali in December.  This was to be the first substantive achievement in the long-suffering Doha Round launched in 2001.  It would have benefited all countries, including India.  But WTO agreements are supposed to require consensus.  Domestic political considerations evidently persuaded Modi to torpedo the agreement, which was to have been finalized by the end of July, if he couldn’t first extract a change in international rules to allow India permanently to keep its subsidies and its stockpiles.

* * *
What is a leader to do?

Of course the US and other rich countries have their own distorting subsidies in agriculture and energy. The US still subsidizes oil production and Europe still subsidizes coal.  The luckiest crops include cotton, sugar, dairy products, and grain.  The subsidies often hurt environmental quality, domestic consumer pocketbooks, and producers in developing countries, as well as the budget and general economic efficiency.    But at least rich country governments don’t usually set a particular low price for an important commodity and lead consumers to think they will never have to pay more.

Once subsidies are in place, they are extraordinarily difficult to remove.  People become accustomed to the idea that the government is responsible for the price of food or energy.  If world commodity prices go up, as they often have over the last decade, citizens who are accustomed to the domestic price being set in the market are more likely to accept the reality that their officials can’t wave a magic wand to insulate them from the unpleasant shock.   But people who are accustomed to the price being set by the government do hold it responsible.  In some countries, the removal of subsidies has led to civil unrest and even the overthrow of the government.

That is a strong reason not to adopt such subsidies in the first place.  But it doesn’t necessarily mean that, once in place, keeping them is the better option for the savvy politician.  If the alternative to raising the price is shortages or rationing through long lines, that can bring angry protestors out into the streets as well.  Similarly, the procrastinating leader is unlikely to look like a political genius if ever-widening gaps force an even bigger rise in the retail price when the day of reckoning comes.  (The gaps that tend to raise the necessary adjustment over time include declining domestic supply as producers respond to low price incentives; widening trade deficits, as the commodity shortfall is imported; and growing budget deficits, as the government pays for the price difference.)

Some subsidies do find their way to the poor.  [Cooking oil is an example.]  Ideally, as a matter of compassion as well as politics, other more efficient means of supporting incomes at the bottom will be instituted at the same time that subsidies for food and energy are cut.  Developing countries have learned a lot about efficient transfer mechanisms, from policy innovations such as the Conditional Cash Transfers of Mexico’s  Progresa/Oportunidad or Brazil’s Bolsa Familia and from technological innovations such as India’s Unique Identification system.  But in countries where the adjustment does not come until a budget crisis forces it, there may be no money left for transfers to cushion the pain.

The savvy politician should probably announce the unpleasant adjustment as soon as he takes office.  That seems to be the approach of Jokowi and Sisi.  Ironic that the third politician, Modi, the one who comes in with the biggest electoral mandate and the most hype about market reforms, is the one who is already falling short.

*   *   *
[A shorter version, “The Subsidy Trap,” appeared at Project Syndicate.  Comments can be posted there.]

Share Button

China Is Not Yet #1

Share Button

Widespread recent reports have trumpeted: “China to overtake US as top economic power this year.”  The claim is basically wrong. The US remains the world’s largest economic power by a substantial margin.

The story was based on the April 29 release of a report from the ICP project of the World Bank: “2011 International Comparison Program Summary Results Release Compares the Real Size of the World Economies.”     The work of the International Comparison Program is extremely valuable.  I await eagerly their latest estimates every six years or so and I use them, including to look at China.  (Before 2005, the data collection exercise used to appear in the Penn World Tables.)

The ICP numbers compare countries’ GDPs using PPP rates, rather than actual exchange rates.   This is the right thing to do if you are looking at real income per capita in order to measure people’s living standards.  But I would argue that it is the wrong thing to do if you are looking at national income in order to measure the country’s weight in the global economy.

The bottom line for China is that by both criteria, income per capita (at PPP rates) or aggregate GDP (at actual exchange rates), it still has a ways to go until the day when it surpasses the US.  This in no respect detracts from the country’s impressive growth record, which at about 10% per annum for three decades constitutes a historical miracle.


(click on the graph for larger image)

At current exchange rates, the American economy is still almost double China’s, 83% bigger to be precise.  But the cross-over day is not far off, as the graph shows.  If the Chinese real growth rate continues to exceed US growth by 5% per annum and the yuan appreciates at 3% a year in real terms (inflation is higher there), then China will pass the US by 2021. Soon, but not imminent.

The PPP-versus-exchange rate issue is familiar to international economists.   This annoying but unavoidable technical problem arises because China’s output is measured in its currency, the yuan, while US income is measured in dollars.  How should you translate the numbers so that they are comparable?  The obvious solution is to use the contemporaneous exchange rate.  (Multiply China’s yuan-measured GDP by the dollar-per-yuan exchange rate, so that is expressed in dollars.)

Someone then points out, however, that if you want to measure the standard of living of Chinese citizens, you have to take into account that many goods and services are cheaper there.   A yuan goes further if it is spent in China than if it is spent abroad.  Some internationally traded goods have similar prices.   T-shirts are virtually as cheap in the United States as in China, in part because we can buy them from there. (Oil is almost as cheap in China as abroad, because it can import oil.)   But haircuts, a service that cannot readily be traded internationally, are much cheaper in China than in the US.  For this reason, if you want to compare income per capita across countries, you need to measure local purchasing power, as the ICP does.

The PPP measure is useful for many purposes, like knowing which governments have been successful at raising their citizens’ standard of living.  A second application is estimating whether the country’s currency is “undervalued,” controlling for its productivity.  A third is judging whether it is reasonable to expect that the country has the means to start cutting pollution.   (The turning point for sulphur dioxide in international data has been estimated at roughly $10,000 per capita in today’s dollars.  China is now there.)

Looking at income per capita, China is still a relatively poor country, even by the PPP measure and even though it has come very far in a short time.  Its income per capita is now about the same as Albania’s, in the middle of the distribution of 199 countries (99th).

But Albania doesn’t often get headlines.   Why are we so much more interested in China?  Partly because it is such a dynamic economy, but not just that.  China has the world’s largest population.  When you multiply a medium income per capita times 1.3 billion “capita,” you get a large number.  The combination of a very large population and a medium income gives it economic power, and also political power.

Why do we consider the United States the incumbent number 1 power?   Partly because it is rich, but not just that.  If income per capita were the criterion, then Monaco, Qatar, Luxembourg, Brunei, Liechtenstein, Kuwait, Norway and Singapore would all rank ahead of the US.  For purposes of that comparison, it does not much matter whether you use actual exchange rates to make the comparison or PPP rates.  Relative rankings for income per capita don’t depend on this technical choice as much as rankings of size do.  (The reason is that the PPP rates are highly correlated with income per capita, the phenomenon known as the BalassaSamuelson relationship.)

If you are choosing what country to be a citizen of, you might want to consider one of these richest countries.   But we don’t consider Monaco, Brunei and Liechtenstein to be among the world’s “leading economic powers,” because they are so small.  What makes the US the #1 economic power is the combination of having one of the highest populations together with having one of the higher levels of income per capita.

So there is a widespread fascination with the question how China’s economic size or power compares to America’s, and especially whether the challenger has now displaced the reigning champ as #1.  It seems to me that PPP rates are not the best ones for making this comparison.  Why?

When we talk about size or power we are talking about such questions as the following.  From the viewpoint of multinational corporations, how big is the Chinese market?  From the standpoint of global financial markets, will the RMB challenge the dollar as an international currency?   From the viewpoint of the IMF and other multilateral agencies, how much money can China contribute and how much voting power should it get in return?   From the viewpoint of countries with rival claims in the South China Sea, how many ships can its military buy?   For these questions, and most others where the issue is total economic heft, you want to use GDP evaluated at current exchange rates.  It’s how much the yuan can buy on world markets that is of interest, not how many haircuts or other local goods it can buy back home.

It is sometimes objected that using the current exchange rate subjects the comparison to the substantial fluctuations that exchange rates often exhibit.   This is true.   But the large and variable measurement errors in the PPP adjustment are considerably worse.  There is a good case for using a five-year average of the exchange rate instead of the exchange rate in one particular year. It doesn’t make much difference for the US-China comparison during this period.  But even when exchange rate fluctuations seem large, the difference is relatively small compared to the other statistical issues at stake here.

[A shorter op-ed version of this article appeared at Project Syndicate. Comments can be posted there. A version at VoxEU will include references.]

Share Button