Through the Hourglass

From his title on, Dani Rodrik is at pains to identify himself as a neoclassical economist, bred in the bone. He writes, “If I often depart from the consensus that ‘mainstream economists’ have reached in matters of development policy, this has less to do with different modes of analysis than with different readings of the evidence and with different evaluations of the ‘political economy’ of developing nations.” Not to start an argument, if the book were about professional cooking, he might have called it One Chemistry, Many Recipes (and Plenty of Chefs). True, economics is not very much like chemistry, but the reason for Rodrik’s emphasis on the primacy of theory, I think, has less to do with the presence of economics’ many competitors in the development game – political scientists, sociologists, lawyers, business executives, savants of all sorts — than with what happened in mainstream economics itself in the twenty-five years since he began his career.

Rodrik was just finishing graduate school when something called “the new growth economics” took off. To all intents and purposes, it began with a famous lecture in Cambridge, England, by Robert Lucas, of the University of Chicago. Lucas’ reputation was as a monetary theorist, not an expert on growth. He had never traveled abroad before he spoke. Yet “The Mechanics of Economic Development” is the fifteenth most frequently cited of all papers in major economics journals since 1970, according to one careful survey. It was the first salvo in a barrage of papers that brought about a profound reshaping of the concerns of young economists. The new growth literature returned the wealth of nations (and lack thereof) to the center of their theoretical and empirical investigations, relegating to a lesser position the interest in distribution that had been dominant since Ricardo, and overshadowing for more than twenty years developments in traditional macroeconomic problems such as inflation, unemployment and the business cycle.

There are many versions of what the new growth economics was all about, including those of Paul Romer, “The Origins of Endogenous Growth,” Journal of Economic Perspectives, Winter,1994, and Olivier Blanchard, Chapter 30, “The Story of Economics,” Macroeconomics, 5th edition, Prentice-Hall, 2007. Here’s my interpretation: Lucas’ powerful rhetoric depended on asking a series of questions, not answering them. In Robert Solow’s model of economic growth, dominant for thirty years, nothing much could be done about growth, aside from sound macroeconomic management. In Lucas’ view, which was expressed in the formalisms that were becoming standard in economics in the 1980s, technological improvements (and the institutional arrangements that made them possible) could not be taken for granted. He pointed to immigration flows to pose a puzzle: why did people throng to cities, and to highly developed countries, instead of the other way around? Why didn’t capital move freely to less-developed areas around the world? Lucas was interested in trade and growth. The hypothesis he advanced was that the engine of growth was human capital accumulation; that people went to cities because that was where the valuable skills were (just as money was in the banks).

Of course, practically as Lucas spoke, the pattern was reversing. Highly skilled people and capital were going back to poorer areas – to Ireland, to China, to India. Soviet-style communism was crumbling. Globalization was swinging into high gear. But meanwhile he had opened the door to a different answer, advocated chiefly by his student, Romer: that knowledge, nonrival and often partially excludable, was important, too, and that it possessed quite different economic attributes than goods do. Economists argued about it for a time, before the best among them agreed to stipulate that knowledge had an economics of its own, quite different from the economics of things. They left behind a handful of researchers to work it out, and moved on to the investigation of the institutional apparatus and social framework that created and distributed new ideas around the world.

Some years later, David Kreps described this sort of an evolution in terms of an hourglass, an image he attributed to Romer. Before World War II, Kreps said, economics had consisted of many different fields that had grown up in substantial isolation from one another: trade, development, economic history, labor markets, public finance and so on. These semi-autonomous discussions, with their focus on typologies and institutions, sounded like so many regional dialects, some of them all but incomprehensible to others. But after the war, as the advancing salient of economics – mainly macroeconomics and general equilibrium theory — embraced more formal methods, mathematical models in particular, the new techniques gradually were extended into one area after another (a movement often experienced as “colonization” by those previously working in the field). Gradual standardization of methods was the result. Previously disparate sub disciplines gradually came to sound more like branches of a single common (methodological rather than topical) tongue, wrote Kreps. And when, as was always the case, the new dialect of mathematical modeling lacked the appropriate vocabulary to discuss important topical features of the landscape, “rather than speak in an unfashionable dialect, some things were just not discussed,” at least by those who felt as if they were on their field’s cutting edge. Gradually techniques were developed to bring these familiar topics under the lens of the new mathematical economics and broaden the discourse.

Hence the image of an hourglass. The vertical axis represents time, and the horizontal axis the scope or breadth of economics. As time passes, we first see a narrowing of topical concern as the language is unified and then a widening of concerns as the language develops. Throughout, it is important to note, the discipline of economics did not entirely abandon subjects such as institutional economics. Pockets of resistance to the evolution persisted; in some overseas locations and in some domestic departments, they dominated. The hourglass describes roughly the development of orthodox or mainstream economics, and, (at that), primarily in the United States. (Daedalus, Winter, 1997)

The excitement over the new growth theory has now passed on to “new” institutional economics and, more generally, to the “new” (meaning more formal) political economy. But while it lasted, the excitement had at least two unfortunate side-effects. The first was to greatly overshadow development economics as a discipline. Beginning in the years after World War II, development economics was an important new field, populated by original scholars who were thought of as giants – Albert Hirschman, Alexander Gerschenkron, W.W. Rostow, W. Arthur Lewis, Theodore Schultz, John D. Black, Peter Bauer. But while those scholars attracted plenty of fans, they didn’t train many students of the next generation who could compete as stars with those who identified themselves as (mathematically trained) growth economists; Mancur Olson, Peter Timmer and Anne Krueger were probably the best-known among those who won prominence. Others, like Henry Rosovsky, left the field for other responsibilities.

The other side-effect in the 1980s and ’90s was to add a little momentum to the impulse to send a lot of bright young theorists out into a world that was rapidly opening up. They were to serve as “country doctors.” The content of their black bags consisted mainly of a understanding of the fundamentals of sound macroeconomic management described in 1990 by John Williamson as “the Washington consensus” and ordinarily interpreted as One Size Fits All. Central to the bedside manner of these advisers was their ability to speak fluently of the latest advances on the research frontier; detailed knowledge of the country in question was not required. Jeffrey Sachs, Lawrence Summers and the late Rudiger Dornbusch were among the most prominent. Countless teams were mobilized by the International Monetary Fund and the World Bank as well. And even in a case where advice was being given by a theorist with intimate knowledge of the local economy, Andrei Shleifer, a Harvard University professor who had grown up in the former Soviet Union, the project collapsed amid charges of corruption among its leaders, a disaster for both US foreign policy and for Harvard.

Rodrik’s heart has been in development since he started, at Princeton’s Woodrow Wilson School of Public Policy more than twenty-five years ago, but he recognized immediately that he couldn’t hang out a shingle except as a full-fledged practicing member of the community of mainstream economics – hence the Princeton PhD. Throughout the years in which the triumphant “Washington consensus” dominated, Rodrik taught, published and worked around the world on development problems, mostly behind the scenes. He retained his Turkish citizenship. In 1996, he moved from Columbia University back to Harvard, where he had been an undergraduate, and quickly built the Kennedy School’s MPA in International Development into a leader in the field.

Rodrik’s task now, as I understand it, and that of the many other economists in his community around the world, is to rebuild development economics on the firmer foundation of the new, more cosmopolitan economics – all the various “new” economics that have emerged since the 1970s, when game theory and decentralized stochastic general equilibrium models became part of the normal graduate student toolkit. He co-chairs Harvard’s Growth Lab with Philipe Aghion, Harvard’s leading growth theorist, and the two teach development together. In Many Recipes, he repeatedly demonstrates his familiarity with the finer points of theory (and they become fine very quickly), taking account, for example, explicitly and implicitly, of a couple of important applied papers in the lit – Murphy, Shleifer and Vishny on the concept of the “big push,” Romer on the significance of an export-processing zone in Mauritius.

He tackles the literature even more directly in “Why We Learn Nothing Regressing Economic Growth on Policies,” a paper too technical to be included in the book but available on his website, even more powerful for its elegant take-down of the economists’ equivalent of digging holes and filling them in. These cross-country regressions were a outgrowth of the new theories of growth (if particular policies could affect the growth rate, why not compare their effects in countries all around the world? Especially since a great compilation of data on production, income and prices in 188 countries, knows as the Penn World Table, had just become available?

Rodrik’s chapter on growth diagnostics, with Ricardo Hausmann and Andrés Velasco, is the heart of the book. There may be other ways to write the decision tree they have devised to identify the constraint on growth of a given country and identify the most binding one – it seems to me that human capital formation is under-stressed – but as a means of disciplining conversation about the experiences of disparate countries, I suspect it cannot be surpassed. It is, in embryo, a contribution equivalent in many ways to the Penn World Table, which also was devised by a handful of economists working at a distance from Washington, in that case at the University of Pennsylvania. Growth diagnostics deserves to be taken up and systematically elaborated by the World Bank, the United Nations or some other agency concerned with global development, s a means of getting beyond the platitudes of “the Washington consensus.”

I especially like the concept of national “self-discovery” that Rodrik identifies as being the nub of the problem in, say, El Salvador. Clearly innovation is what is required if low levels of entrepreneurship and private investment are to be reversed – new markets for new goods, especially since the successful products can be scaled up for sale in the world market (which is the essence of new growth theory). But new ideas are easy to imitate, hard to protect. What it that El Salvador can uniquely trade in the world economy? Not just coffee, surely; that grows just as well in a dozen other countries around the world. Rodrik is particularly good at drawing out the implications in chapter four, “Industrial Policy for the Twenty First Century.” Most significant instances of product diversification are the result of collaboration between governments and the private sector, he says, in East Asia and in Latin America. (He might have added that the same is true of the United States and Europe, too.) What’s needed is to understand much better the differences between success and failure.

And this brings me back to Robert Lucas. What was so startling in 1985 about “The Mechanics of Economic Development” was the frank and surprising admission by the leading theorist of the University of Chicago that building new skills was at the heart of the problem. Sure, Lucas imagined an austere world, in which there were only two goods, potatoes and computers. And yes, the two models he brought to bear were highly mathematical. But both led ineluctably to a possible role for government in fostering growth. In the first model, a subsidy to schooling would enhance development, he noted. In the second, an industrial policy focused on “picking winners” might improve matters. Picking winners was easy in the model he had written down, Lucas wrote: “If only it were so in reality!”

In those days, Harvard’s Kennedy School of Government and the business school across the river had a joint seminar discussing industrial policy that produced, among others, Robert Reich and Ira Magaziner. But Lucas’ talk changed all of economics. On my reading, Rodrik understands that – it is part of what he means when he writes that “social phenomena can best be understood by considering them to be an aggregation of purposeful behavior by individuals – in their roles as consumer, producer, investor, politician and so on – interacting with each other and acting under the constraints that their environment imposes.” Precisely because that he understands that mainstream economics, Chicago and Cambridge, now presupposes that industrial policy is as much a part of sound macroeconomic management as monetary or industrial policy, he is a regular rock star among development economists, quite able to stand toe-to-toe and slug it out on practical matters with any theorist.