MIT's Model Economics Department

Samuelson did stay at MIT, and the department, bolstered by a number of appointments of technically strong scholars, became, by the late 1950s, one of the three or four most distinguished research departments of economics in North America. In another decade, it would become the most highly regarded economics department in the world… MIT’s rise to prominence coincided with the remarkable transformation of American economics in the postwar period.

—E. Roy Weintraub, “Introduction: Telling the Story of MIT Economics in the Postwar Period.”1

MIT and the Transformation of American Economics presents the edited proceedings of a History of Political Economy conference held in 2013 at Duke University. It offers perspectives on the rapid evolution of economics into a technical discipline and the role that the Massachusetts Institute of Technology economics department played in that process. I came away from this collection of essays with many insights and some profound questions.

The insights include the following:

• World War II influenced the path of economics. The use of technical economic tools during the War, notably linear programming, made the U.S. government inclined to support economic research of the mathematical and statistical character epitomized by MIT.2
• MIT was able to hire prominent economists more aggressively than competing departments, in part because many of those departments had informal quotas limiting the number of Jewish professors that they would employ.3
• Paul Samuelson’s textbook was viewed at the time by conservative businessmen as antithetical to free-market principles, and they applied pressure on MIT and Samuelson to modify its contents.4

My questions, some of which are only tenuously addressed in the book, include the following:

• Was the transformation of economics from a philosophical discipline to a technical discipline inevitable, or was it influenced by idiosyncratic factors, including the intellectual charisma of Samuelson and Robert Solow?
• In this transformation, was important intellectual perspective lost? If so, can it be recovered?

Eminent MITorians

From 2008 through 2014, four Nobel Prizes in economics were awarded to MIT PhD’s: Paul Krugman, Peter Diamond, Robert Shiller, and Jean Tirole. Previously, from 1969 through 2007, I count eight other Nobel laureates who either earned their doctorates at MIT, taught there for many years, or both: Samuelson, Lawrence Klein, Franco Modigliani, Solow, Robert Merton, George Akerlof, Joseph Stiglitz, and Eric Maskin.

The conference volume includes an essay by Andrej Svorencik5 that offers some striking statistical evidence concerning MIT’s dominance in economics, such as:

• As of 2012, the John Bates Clark Medal, the most prestigious award for an American economist, had been given 16 times to someone with MIT affiliation out of a total of 35 awards.
• Since 1980, 27 percent of the economists elected as Vice-Presidents of the American Economic Association have been affiliated with MIT. Over that same period, MIT economists made up 34 percent of the executive committee of the AEA.
• Between 1977 and 2011, MIT economists accounted for 34 percent of the President’s Council of Economic Advisers.

Prominent policy makers with MIT Ph.D.s include former Federal Reserve Chairman Ben Bernanke, Federal Reserve Vice Chairman Stanley Fischer (who also served as the governor of Israel’s central bank and as Deputy Managing Director for the International Monetary Fund), European Central Bank President Mario Draghi, and IMF Chief Economist Olivier Blanchard.

Svorencik conjectures that MIT has the highest rate of academic retention of its Ph.D.’s, meaning the proportion of its graduates who had careers as professors. Another channel of influence is leading economics textbooks, many of which were authored by either MIT faculty or graduates.6

Seeing the Economy Like a State

Several essays in the volume describe the distinctive approach to economics that emerged at MIT and spread throughout the profession, thanks to its textbooks and well-placed graduates. Concerning Solow’s growth model, one of the emblematic examples of the MIT style, Verena Halsmayer writes,

In making dynamic economic issues intelligible to his engineering students, Solow had formulated the collapsed production function, the essential component of his design model. He used this simple model as a teaching device. One concern about the use of design models as teaching objects was that students and teachers too often stuck with these simple models…

Building simple, transparent, and exploratory design models, and conducting econometric applications with a focus on clear-cut definitions and closed systems, became the “proper” way of doing economics and was associated with MIT from the 1960s on.7

Halsmeyer goes on to write,

The kind of technical expertise that Solow and others provided came with left-liberal Keynesian views. Planning to dampen cycles, to create economic growth, low inflation, and low unemployment, was a rather neutral concept; it was seen as an organizational necessity rather than a political option.8

Paul Samuelson shared this belief in the non-ideological nature of MIT economics. Harro Maas writes that in Samuelson’s view,

… the technical economist, in the postwar era, had brought rigor and precision to economic policy. Economics freed from ideology enabled the economist to, in Theodore M. Porter’s phrase, “speak precision to power.”9

Maas elaborates,

Samuelson did not consider it his task to be partisan for one particular line of economic policy, or to compromise between opposite policy positions, but to step back, or better, “step aside,” and offer analytical perspective from which to choose.10

Similarly, in his essay on the creation of Samuelson’s textbook, Yann Giraud concludes,

What we see emerging… is the figure of the economic expert, who is able to acknowledge the various political positions but is not taking a stand when giving policy recommendations. Therefore, it is possible to interpret the preceding narrative as an important episode in the construction of the image of the economist as expert… Although Samuelson clearly intended his textbook as a policy-relevant treatise, there is no reason to think that he ever quite considered his own policy orientation as strongly partisan from a political point of view… Samuelson’s main concern in revising his text was to show that his policy recommendations were not tied to predetermined political convictions…

Far from being anecdotal, therefore, the controversies that surrounded the publication of Economics in its early years should rather be seen as foundational in the identity of what historians are trying to identify as “MIT economics,” which Robert Solow once defined as “an atmosphere of rigor-with-policy-relevance.”11

In Seeing Like a State, political philosopher James C. Scott argues that many political institutions, such as population censuses, property records, or economic data, are designed to benefit rulers by making the society “legible” to them. Similarly, I think one could argue that the essence of MIT economics is to take a “seeing like a state” perspective on the economy. Its tenets include:

• The Keynesian model, particularly in its attempted implementation in large computer models of the economy by Klein, Modigliani, and others, is supposed to allow fiscal and monetary authorities to regulate total employment and output.

• The Phillips Curve, first studied for the United States by Samuelson and Solow, appeared at first to provide policy makers with a sort of menu of choices for unemployment and inflation.12
• Samuelson, along with Abram Bergson, is credited with developing the concept of a “social welfare function,” which is an attempt to characterize economic policy as a mathematical optimization problem.
• Samuelson’s concept of “revealed preference” can be seen as an attempt to arrive at objective characterizations of individual choice.
• The neoclassical production function, a concept that is central to both MIT microeconomic and macroeconomic theory, is often used as a tool to interpret the distribution of income in terms of “labor’s share” and “capital’s share.”

The production function also tends to tie economic output to “objective” measures of cost. Originally inspired by Solow’s work on Wassily Leontief’s input-output analysis,13 the production function sees the value of output as the sum of the cost of inputs. It leaves out, for example, the way that the process of free exchange creates value. More deeply, it ignores the irreducibly subjective aspect of cost.14

Although these concepts are offered as ideologically neutral, they are anything but. They assume or imply that policy makers can possess a level of economic understanding that is beyond what is really possible. Friedrich Hayek’s 1974 Nobel Prize lecture, “The Pretence of Knowledge,”15 offers a direct critique of the approach to economics popularized by MIT.

Triumph and Tragedy

If we use an evolutionary scale for evaluation, then MIT economics has been an enormous success. The philosophical and non-mathematical discussions that dominated economics up until the second World War now occupy only a minor niche. Instead, the ecosystem of academic publication and policy discussion has been overrun by the technicians and model-builders who follow in the footsteps of Samuelson, Solow, and their MIT descendants.

One of the reasons that Harvard allowed MIT to abscond with Samuelson was that his style of economics in the 1940s was still considered peripheral. Older economists thought of his mathematical approach as a peripheral footnote, without real-world significance. Fortunately for Samuelson and MIT, the U.S. Department of Defense saw value in mathematical optimization techniques. As a result, MIT did not depend on the older economic establishment for funding the growth of its economics department. One way to think of the rise of MIT economics is as an element of the post-war military-industrial complex.

How might history have been different? What if MIT economics had remained peripheral?

The MIT-influenced mainstream view is that the older approach to economics was woolly-headed and lacking in rigor. The use of models imposes discipline and clarity where confusion once reined. When placed next to well-executed technical papers, the older verbal and philosophical discussions are not competitive. “It takes a model to beat a model” is an often-used catch-phrase among MIT graduates.

My own view is that while there is much in the pre-1940 economic literature that is of little value, the turn to technical economics has had its costs. I align myself with Hayek’s “pretence of knowledge” critique. In fact, I would say that since Hayek gave his lecture, in some ways the situation has worsened. As Samuelson and Solow gave way to later-generation MIT economists, like Fischer and Krugman, the modeling hubris has gotten worse. The conference volume includes several mentions of Solow’s misgivings about the applicability of simple models, misgivings that do not seem to trouble subsequent generations.

In education circles these days, a popular buzzword is “critical thinking.” This phrase has no standard definition, but to me it means asking the question, “How do we know?” How do we know that a neoclassical production function can be used to characterize trends in income distribution and productivity growth for the U.S. economy? How do we know that the relationship between inflation and unemployment is described by the Phillips Curve?

My sense is that the MIT-dominated profession has experienced a decline in critical thinking. Instead, once a modeling assumption has appeared often enough in the literature, it no longer is questioned. This creates an element of arbitrariness and path dependence to the professional consensus about the equations used to characterize the economy.16 I see this as unhealthy. To me, it is possible that the older Harvard professors who thought that Samuelson’s place was on the margins of the profession were right.