Free exchange

TODAY'S recommended economics writing:

• China to alter taxes in attempt to cut wealth gap (Financial Times)

• Game theory and the budget (Marginal Revolution)

• Toward a general theory of public options (Mike Konczal)

• Last mill standing (Etsy)

• Paul Ryan's plan to deepen the recession (Matt Yglesias)

Over the next few days, our bloggers will be discussing "Poor Economics", a new book by Abhijit Banerjee and Esther Duflo on their work in experimental economics.

IN THIS week's issue of The Economist we review "Poor Economics", the new book by Abhijit Banerjee and Esther Duflo. (The book's website is here.) The two MIT economists are best known for their randomised controlled trials (RCT) in developing countries. Over on Marginal Revolution, Tyler Cowen says the book is "self-recommending". But let me recommend it anyway.

The book arrives hard on the heels of "More Than Good Intentions", by Dean Karlan of Yale and Jacob Appel of Innovations for Poverty Action (a book I haven't yet seen). Readers can therefore look forward to a feast of results from randomised trials, which seem to be growing in audacity even as they gain in popularity.

But "Poor Economics" is more than just a compendium of the randomistas' greatest hits. For one thing, it contains some well-observed reporting. Whatever else they do, field experiments force economists to get out and about. The book benefits from the insights of bureaucrats and charity-workers whom the authors enlisted to carry out their experiments. The pair also learn alot from talking to the guinea pigs themselves.

Even their more formal evidence is not limited to RCTs. The book's origins, according to the authors, date back to a 2006 article describing the economic lives of the poor, which drew mainly on old-fashioned household surveys. (See this Economics focus from the time.)

Randomistas are sometimes accused of being atheoretical: they crank out their trials and let the results speak for themselves. They say they are being open-minded; their critics say they are being empty-headed. I'll explore the backlash against the randomistas in my next post. But for now, let me just say that "Poor economics" is not mindless. In fact, although it is pitched at the laity, it still manages to reveal something of the way economists think. The authors ask the sort of questions that only occur to dismal scientists. In the Indian city of Chennai, for example, they find fruit sellers who buy 1,000 rupees of produce from a wholesaler each morning on credit, paying 4.69% interest at the end of the day. By the power of compound interest, a 4.69% daily rate equates to a terrifying annual rate of 1.8 billion percent. A gross injustice, crying out for the intercession of microcredit.

But the remorseless power of compounding cuts both ways. Suppose the fruitsellers went without six cups of tea, saving themselves 15 rupees. They could use that money to buy their morning produce, borrowing 15 rupees less from the wholesaler. That would cut the amount they have to repay at the end of the day by 15.7 rupees. That means that on the following morning, they could borrow 15.7 rupees less, shaving 16.4 rupees off their end-of-day repayment, and so on. After 94 days (by my calculations), they could afford to buy their morning produce from the wholesaler in cash, escaping usury altogether.

In a similar vein, the book argues that many of the big debates in poverty and aid—debates that divide left and right, inspiring each side to fits of righteous anger and indignation—boil down to disagreements about the shape of a function. They plot two graphs of the poor's prospects (see below).

The one on the right has a toppled L-shape. The one on the left resembles an elongated S. If you believe the poor's prospects are L-shaped, you will tend to believe that they can gradually pull themselves out of poverty, since a small investment (in their livelihood, their education or their health) will yield disproportionate returns, paving the way for larger investments in the future. If, on the other hand, you believe their prospects are S-shaped, you will conclude that the poor are trapped. They need a "big push" to get them out of a rut and onto the sunlit uplands.

In my experience, this way of thinking about things irritates many non-economists, who would rather have an animated fight over political profundities than discuss whether the world is quadratic or cubic. But the economist's way is probably more fruitful.

OVER the past few days a spirited debate over the status of the American government, and in particular its ability to borrow, has rumbled around Washington. This debate was obviously triggered by S&P's announcement Monday morning that it was cutting its outlook on American government debt to "negative", signaling that there is a one in three chance of a downgrade in America's credit rating by 2013. I found the attention to this statement mystifying. S&P didn't say anything about America's debt situation that wasn't already well known. The market reaction has been about as benign as one could hope for; relative to the Friday close, equities are higher, bond yields are lower, and the dollar is basically flat. Markets shrugged.

A lot of pundits found it very difficult to accept that markets shrugged. They seemed to be labouring to explain that the S&P's warning represented the beginning, at long last, of an American debt reckoning. I can understand this impulse. America's debt situation is not sustainable, and something must eventually be done about it. But Congress, which would be the institution to do the something that needs doing, is unlikely to take difficult deficit cutting decisions unless its feet are being held to the fire by bond markets. And so the fact that bond markets steadfastly refuse to hold Congress' feet to the fire is frustrating for those who'd like to see the government budget put on a sustainable path.

And so here everyone is, struggling to piece together explanations for why this or that piece of news does mean that crisis is around the next corner, despite the pacific appearance of Treasury markets.

Ezra Klein, for instance, offers that the S&P forced Britain into austerity, and he suggests that only Britain's lightning-quick response to the agency's warning saved it from crisis. This seems dubious to me. S&P threatened Britain way back in 2009. When the country's debt became an election issue in the spring of 2010, it was against the backdrop of actual crisis elsewhere in Europe. British bond yields declined as trouble in Greece grew, but there was a palpable feeling that markets might turn on anyone next, and that—the real threat of market pressure—resolved the issue in favour of austerity. Well, that, and the fact that in Britain a party can fail to win an electoral majority and still wind up with more or less unfettered authority.

Others argue that Treasury prices are unreliable. They're distorted, it's said, by Fed purchases, and by the buying of foreign sovereigns. But this is not a particularly convincing argument. Buying patterns indicate that there's growing demand for Treasuries and that the end of Fed purchases won't lead to a big jump in yields. Other sovereigns, notably Japan and China, are big buyers of American debt. Japan has essentially said that it's not worried about American borrowing, but China is increasingly vocal about its concerns. The problem is that China has very little leverage.

The issue isn't simply, as Buttonwood suggests, that China feels the need to manage its currency against the dollar. It's that it holds a massive pile of American government securities already, and if it takes a step that leads to a big fall in the dollar or a big decline in Treasury prices, then it will cost itself a fortune. And it's that it has few good options for places to park the reserves that it continues to accumulate.

So let's take a step back and think about the situation in which America finds itself. The country has a debt problem. America's ratio of gross government debt to GDP currently stands at about 99%. That's not an absurdly high level for a rich country at the present time. It's about 24% above Germany's ratio and 20% above Britain's. It's 82% of Italy's debt ratio, 66% of Greece's, and less than half of Japan's.

This is a high debt level, historically speaking. It's also above the 90% threshold identified by Carmen Reinhart and Kenneth Rogoff, above which debt impairs growth. And where most large European countries are expected to have falling debt-to-GDP ratios by about 2013, America's ratio is projected to continue growing for the foreseeable future.

And that's a problem. America's deficit will narrow to close to primary balance by about 2016 which would stabilise debt if it were sustained. But there is a risk that America won't get to primary balance, and there's a risk that by the time it does debt will be large enough to spook markets. And then there is the really big risk that America will fail to get growing health care costs under control, such that beyond 2016 deficits once more gap out to levels that would push the debt-to-GDP ratio into dangerous territory.

Against this, one must set some offsetting pieces of information. American economic growth is generally quite hardy relative to that in other rich countries. The IMF projects that America will grow consistently faster than Europe over the next few years, and by 2016 America is forecast to grow at twice the German pace. This will make it easier to handle a given debt load. America's demographics are relatively better than those elsewhere, and if immigration picks up post-crisis, the fiscal situation could begin to look much better. America is the issuer of the world's dominant reserve currency and the world's most plentiful safe asset. Foreigners hold huge stocks of dollars and Treasuries and other dollar-denominated securities, and they therefore have an incentive to support the dollar and Treasury prices. And many of the other available safe assets aren't particularly attractive right now; as Gillian Tett put it in a conference last week, American yields are low not because American debt is winning a beauty contest but because it's losing an ugly contest.

If America were to finally get its act together and address the fiscal situation, it would have a lot of room to make things better in a relatively easy fashion. America's tax system is woefully inefficient. Improvements to the efficiency of the tax code would allow relatively small tax increases to generate substantial amounts of revenue. Other parts of the budget look hugely bloated relative to peer nations. American military spending dwarfs the spending of its allies and rivals alike. America spends a fortune on health care without doing much better on health outcomes. Its budget fixes would be painful in the sense that established interests hate to see change, but they would not be painful in the sense that there's no fat to cut.

American debt is not out of control, in other words, and it's fundamentally affordable, and there are good reasons to expect that its creditors are prepared to give America a lot of room to get its spending under control before giving up on dollars and Treasuries. The only question is: can America's political system use this room to make the necessary policy changes?

This, of course, is precisely the problem S&P identified in its Monday announcement. And it's a common thing to fret over. On Monday, Buttonwood wrote:

[I]t resembles one of those Greek myths when the hero's power is accompanied by a curse; in this case, a political system that is not designed for serious deficit-cutting (the point made by S&P). The world's dominant power tends to think its financial strength will never drain away. But Spain, having absorbed all that gold and silver from Latin America, still defaulted on its debts in the 16th century; Louis XIV, the sun king whom other monarchs dreamed of emulating, set France on the road to financial ruin; and Britain started the 20th century with a huge empire and piles of overseas assets but was rationing food in peacetime by the late 1940s.

Of course, it was scarcely a decade ago that America was running actual surpluses and not long before that that bitterly opposed politicians were cutting the deals that made those surpluses possible. The suggestion is that American politics has become so polarised that such deals are no longer imaginable. That's possible, but it's certainly not obvious. Only weeks ago, Congress passed an extremely contentious bill to fund the government through the end of the fiscal year, and did so on a bipartisan basis. Leaders of both parties indicate that the debt ceiling can and will be raised, and leaders of both parties have put out ambitious plans to address long-run debt issues. There's a long way between what's said and proposed and a piece of signed legislation, but this looks to me like a remarkable state of affairs given that debt yields have rarely been lower.

Some writers worry that views of American creditworthiness may turn on a dime, such that there will be no time to act between the moment when yields begin rising and the moment they hit unsustainable levels. That's not impossible, but it does seem highly unlikely. American yields didn't soar overnight in the early 1990s, and Greek yields crept up for months before spiking last April. The existence of a broad array of parties interested in an orderly American budget fix makes it likely that parties would intervene to slow rising American yields if they did spike rapidly, in order to buy Congress time to take action. But I find it hard to imagine things progressing that far. Whenever Treasury yields climb the least bit, even if its only from epically low levels to abnormally low levels, pundits and politicians begin squawking and demanding immediate budget steps. And indeed, these squawking pundits have been quite effective; it's striking how much attention is being paid to deficit issues rather than, say, unemployment. If the 10-year returned to its 2007 level, all hell would break loose in Washington, even though the yield in 2007 was really unusually low.

None of this is a call for complacency. The Economist has been regularly and rightly calling on Congress to put in place a credible plan for medium-term deficit reduction, and Congress should do this. The tail risks associated with a mishandled debt situation are sufficiently nasty that America's leaders should not be playing games.

But I also don't see that it makes much sense to describe the present situation as anything other than what it actually is. America's debt position is an uncomfortable one that requires some action. And it's very unpleasant to watch America's petty political leaders wriggle their way toward that action. But there are good reasons that yields are low, and it isn't that markets have lost their minds. And there are good reasons to think that Washington will ultimately do what it has to. I just don't see how hysterics over a non-event like the S&P news are going to improve the odds that good policy gets made.

KEVIN DRUM draws attention to an analysis of catch-up growth episodes by Stuart Staniford, which suggests that the Eichengreen-Park-Shin paper on growth slowdowns is too pessimistic about China. Mr Staniford writes:

To try to get a better grip on the situation, I did two things. Firstly, to formalize the instinct that the US has been at/near the productivity frontier at most times, I expressed every country's GDP/capita as a fraction of the US value in the same year. Then I started kicking countries out of the sample, unless they met the following criteria: they started out the sample clearly less productive than the US (I took less than 60% as my threshold), and ended up significantly more productive, relative to the US, than they had started out. Ie, we want countries where it's somewhat plausible that there's a story of underdevelopment, period of rapid catchup, followed by slowing growth once the country is a fully developed country with modern capital infrastructure and levels of productivity.

And Mr Drum summarises the findings:

Long story short, Stuart produced the chart below, which suggests China can keep growing at a fast pace until its per capita income is somewhere in the $25,000 range, which is probably still 15-20 years away. I don't have the chops to adjudicate this, but I thought it was worth highlighting a contrarian opinion anyway. China might very well slow down in the next five or ten years anyway, since it faces multiple constraints (resource scarcity, productivity limits, demographics), but the $17,000 limit is just a guess, and you should probably put some fairly large mental error bars around it.

So, a few points. First, it seems a little unfair to call the $17,000 threshold a guess. What the authors did was assemble a sample of growth episodes and determine when the probability of a growth slowdown was highest. They helpfully provide a chart of the distribution of slowdown events:

As you can see, there are experiences in which rapid growth persisted well after per capita output hit $17,000, but these are not the norm. The six high-income outliers shown above are, in order of ascending income, Puerto Rico, Japan, Hong Kong, Ireland, Singapore, and Norway (included despite the indication that oil-exporters are excluded). The only large, populous, manufacturing-oriented economy in the bunch is Japan, and I'm not sure that China hopes to emulate the Japanese experience.

Second, the $17,000 figure is not the only guidepost for which the authors find a significant probability of slowdown. They also identify a ratio of per capita output to the leader of about 58% to be important, as well as a share of manufacturing in total employment of 23%. On the first measure, as Mr Staniford indicates and as the authors acknowledge, China has a way to go. Hard data aren't available on the second, but extrapolations from previous data releases suggest that China is at or close to the manufacturing employment threshold. So, two of the three indicators point to a near-term growth slowdown for China.

Finally, the Eichengreen-Park-Shin paper tests which economic variables have a significant influence on the probability of a slowdown. They find that a high level of trade openness is helpful in delaying a slowdown, and they attribute the anomalous performance of places like Hong Kong and Singapore to this factor. Factors that bring forward the moment of deceleration include a high old-age dependency ratio, an undervalued currency, and an extremely low level of consumption in output. China suffers from these afflictions in spades (though to be fair Japan did as well, but again, China would probably prefer to avoid Japan's ultimate growth fate).

Obviously, China is a unique case, and one simply can't be sure how the above factors will affect its economy. If I had to guess, I would say that by mid-decade China's growth rates are likely to slow from over 9% a year to something like 6% a year. And I would further say that what happens to China's economy after that will depend a great deal on how its financial and political systems handle that slowdown.

TODAY'S recommended economics writing:

• Fear of a weak China (Reihan Salam)

• Stocks, flows, and Pimco (Paul Krugman)

• Jim Manzi on the budget (Modeled Behavior)

• Tales from the stoop (American Prospect)

• Paul Ryan's reverse Robin Hood budget (Wall Street Journal)

THIS week's Economics focus discusses a new working paper by Barry Eichengreen, Donghyun Park, and Kwanho Shin (previously mentioned here at Free exchange) focusing on middle-income growth slowdowns. The authors find that an economy experiencing rapid catch-up growth will often see a decline in per capita growth rates around the time it hits a per capita output level of about $17,000. The authors further note that China is approaching this threshold and may hit it by 2015.

For further debate on the question, we turned to the economists at Economics by invitation and asked them whether a slowdown lurks in China's near future. The responses have been very interesting. Michael Pettis writes:

I would certainly argue that the consensus medium-term growth predictions for China are wildly optimistic. My own guess is that over the next decade annual growth will average 5% or less, although it will be heavily frontloaded—higher in the first few years and lower later on.

The reason for this slowdown is that true Chinese growth rates during the past decade have probably been overstated by a significant amount. I say this because much of the growth in the past decade has come about in the form of massive increases in investment, much of which has gone to fund projects and manufacturing capacity that are not economically viable. Very low interest rates, an undervalued currency, and other direct and indirect subsidies have made these projects seem viable, but only because the true costs have been disguised. The problem has been made worse by an incentive structure that concentrates the rewards of investment in the jurisdiction that initiates the investment while spreading the costs through the entire banking system.

If it is true that capital is being misallocated, then it is also true that initially GDP growth will be inflated by the extent of the overinvestment, but in later periods, as the resulting debt is repaid, the losses will show up as lower GDP growth.

Lant Pritchett adds:

I recently did a study examining the growth consequences of sudden large democratisation (a shift in the POLITY index of more than 6 points). Of the 22 cases that experienced rapid democratisation with above average growth: (a) all but one had a growth deceleration, (b) the average deceleration was 3.5 ppa, and (c) the predicted deceleration was increasing with growth—roughly, post-democratisation countries reverted to world average growth. So in "predicting" China's growth, whether this is "conditional" or "unconditional" on democratisation makes a big difference.

Second, the likely path to a long, sustained, rapid boom is to have had sufficiently tragic events and perverse policies that forced a country far from its potential. One way to think of China's boom since 1978 is a rapid make-up from the negative shocks the country experienced at least from 1911 onwards (invasion, civil war, totalitarian regimes with perverse policies). After all, the historical question about China has always been why it was not one of the world's leading countries given its tremendous strengths on many fronts.

Stephen Roach, Laurence Kotlikoff, Gilles Saint-Paul, and Michael Heise also weigh in, commenting on the challenges China faces in rebalancing its economy and how rich China can expect to grow before converging to advanced world growth rates. Do click through and have a read.

WHEN we talk about the debt crisis in Europe, we tend to focus on the specific details—a relative loss of peripheral European competitiveness, accumulation of debt, rising bond yields and contracting economies. But the bigger story is a simpler one: The euro zone's political institutions did not keep up with its economic institutions. A piece in this morning's Financial Times beautifully captures the dynamic:

The True Finns party won 19 per cent of the vote, increasing its number of seats in the 200-strong parliament to 39 from just five at the previous election amid public anger over the succession of taxpayer-funded bail-outs for crisis-hit eurozone countries.

The result raised fears that Helsinki could block the  Portuguese deal because Finland, unlike other eurozone countries, requires parliamentary approval to take part in bail-outs, which can go forward only with unanimous EU support.

How long could America maintain its dominance—or, indeed, its union—if the fact of a secessionist party winning 19% in a Maryland election could prevent the union from undertaking a step critically important to the stability of the American economy?

REIHAN SALAM is a sceptic of the value of investments in American intercity rail and I am more optimistic concerning their likely benefits. But he makes a good point here:

A shockingly large number of people, including Obama, seem to believe that had the federal government not stepped up to the plate in the postwar era and invested vast sums in highways and putting a man on the moon, the United States would have wound up an economic backwater. But perhaps not building a huge network of highways would have kept American families in more compact, walkable neighborhoods. Instead of sprawling suburbs and SUVs, we’d have more high-rises and bike lanes. The Interstate Highways helped supersize America’s government, by centralizing authority in D.C., and our waistlines, by encouraging us to drive and to fatten up on fast food. It’s not obvious to me that we’re better off as a nation plagued by high taxes and heart disease.

What would the American economy have looked like without a massive government investment in highways? It's very hard to say, and it's certainly possible that the net effect of their construction is large and positive. But I also think that observers significantly overestimate the value of highways, because they fail to take into account the fact that in a world without them markets would have optimised to the non-highway status quo.

As it turns out, this is not the first time a debate like this has occurred. Back in 1944, economist Leland Jenks published a paper called "Railroads as an economic force in American development". Here's how economist Robert Fogel describes Mr Jenks' work:

Leland Jenks's article describing the pervasive impact of the railroad on the American economy first as an idea, then as a construction enterprise, and finally as a purveyor of cheap transportation, has become a classic of economic history. The particular contribution of the Jenks article was not the novelty of its viewpoint, but the neat way in which it summarized the conclusions both of those who lived during the “railroad revolution” and those who later analyzed it through the lens of elapsed time. Out of this summary the railroad emerges as the most important innovation of the last two thirds of the nineteenth century. It appears as the sine qua non of American economic growth, the prime force behind the westward movement of agriculture, the rise of the corporation, the rapid growth of modern manufacturing industry, the regional location of industry, the pattern of urbanization, and the structure of interregional trade.

It was a big deal, in other words. And this was a very common assessment of the value of the railroad. But Mr Fogel disputed this conclusion. To gauge the importance of the railroad, one can't simply look at the way growth proceeded in the wake of its construction and attribute some or all of that growth to the new infrastructure. Rather, you need to compare "social savings" from a new technology relative to alternatives. That is, how much did the railroad reduce costs compared to other transport technologies?

Shipping via railroad was substantially cheaper than transport via wagon. But it wasn't much cheaper than shipping via canal. Now, land use changed in response to the boom in railroad construction, and a great deal of new land was brought into production. But in a but-for world, it's reasonable to assume that canal construction would have continued and markets would have optimised around canal infrastructure. In the end, Mr Fogel estimates that "the social saving attributable to the railroad in the interregional transportation of agricultural products was about 1% of national income".

That's not very much. Of course, Mr Jenks also points to the important economic role of construction of the railroads. That might well have applied to canals to some extent, as well, but that would require a canal building boom comparable to the investment in rail. Mr Jenks also talks about rail as an idea, and it's more difficult to analyse this potential contribution.

But the lesson is clear. Highway construction generated some positive effects and some negative effects. We tend to focus on the positive effects and remark on how constrained the economy might have been without a highway boom. But absent a highway boom something would have been built and markets would have optimised to that something. It's not clear that the savings from highways are so substantial that the American economy is clearly better off as a result of the system's construction. Highways obviously had a large effect as an idea, and they made direct contributions to the economy as a construction enterprise, but the net addition to growth through trade is uncertain, and probably much smaller than most people assume.

That doesn't mean that infrastructure isn't worth building. New infrastructure will quite often yield real cost reductions, and my assessment is that many new projects, particularly those along corridors that are already busy and congested, would probably produce benefits. I suspect that the potential benefits would be clearer if roadways were kept free of congestion via tolls; toll revenues along some well-traveled routes would meet or exceed the cost of construction of new infrastructure capacity.

But as someone who takes the need for new infrastructure seriously, I think it is worth being realistic about what new investments can deliver. Higher incomes, in some cases? Yes. Economic revolutions? No, not for the most part.

THE Financial Times has a story by Robin Harding above the fold this morning, headlined, "Fed to signal end of monetary easing". Sounds like big news! What's it all about?

When the rate-setting Federal Open Market Committee meets on April 27, it is unlikely to limit its options by ruling out asset purchases beyond the second $600bn “quantitative easing” programme – or “QE2” – that is due to finish by the end of the second quarter.

Fed officials, however, know that announcing more asset purchases at the last minute would disrupt markets. Silence on a follow-up “QE3” at next week’s meeting would therefore signal that their current intention is to complete the $600bn QE2 programme and then stop.

I understand what Mr Harding is saying here. Based on recent statements from key people on the FOMC it seems clear that the Fed's intention is to complete QE2 and then stop. So long as inflation expectations are rising, I would be very surprised to see additional Fed action. (I don't necessarily agree that rising inflation expectations should stay the Fed's hand; I simply think that they will.)

And yet, it's too strong to say that the Fed is signaling an end to its easing cycle. It would be more accurate to call the impending end of QE2 a pause. The end of the initial round of asset purchases did not represent the conclusion of the easing process; when global developments undermined expectations in America, the Fed responded. I am sure that the FOMC is keeping a close eye on falling projections for first quarter output and on the potential threats to a self-sustaining recovery represented by Europe, commodity prices, and the global swing toward policy tightening.

Meanwhile, the Fed doesn't actually need a lot of lead time before easing again. As we observed last fall, the change in outlook—the signal—does most of the policy work. When Ben Bernanke hinted that QE2 was forthcoming in a speech last August, markets responded immediately and strongly, even though the actual policy announcement didn't occur until November.

If the June Fed meeting arrived and the Fed determined that QE3 might be necessary, it could hint as much and the impact would be immediate. Given that the FOMC can influence the economy relatively quickly, there's little reason for Fed officials to indicate that they're doing anything other than what they are, in fact, doing—which is pausing while the sustainability of rising expectations remains uncertain.

TODAY'S recommended economics writing:

• Saving the WTO from the Doha round (Vox)

• IMF believes Greece should consider debt restructuring (Wall Street Journal)

• A shot at a sane budget (New York Times)

• Chinese recycling and US interest rates (Michael Pettis)

THE American Economic Association has announced that Jonathan Levin, an economist at Stanford University, has one the prestigious John Bates Clark medal, which is given to the most promising economist under 40. Previous winners include Milton Friedman and Paul Krugman; last year's winner was Esther Duflo.

Here's the AEA's introduction of Mr Levin:

Jonathan Levin is a leading scholar in the fields of industrial organization and microeconomic theory, whose work stands out for its combination of theoretical depth, empirical methods, and compelling applications. He has conducted influential research on the economics of contracting, the organization and design of markets, subprime lending, and on empirical methods for studying imperfect competition. His research is methodologically broad, and often combines a sophisticated grasp of economic theory with careful empirical analysis. He has been a leader both in developing new methods in industrial organization and microeconomic theory, and in producing path-breaking applied research.

You can read more here and here. Or here, at Mr Levin's website.