In the run-up to the Nobel Prize in economics, usually someone is willing to go out on a limb. A few days after dark horse Trygve Haavelmo won in 1989, New York Times economics columnist Peter Passell polled senior economists to write a memorable column about the speculation that had taken place on the eve of the award (“The Morning Line on the Next Nobel”). That prompted Eugene Garfield, proprietor of a pioneering scientific data bank in Philadelphia, to write an article a year later naming likely winners of the prize based on a citation count.
For a few years then, newspaper reporters took turns trying to scoop the announcement. University of California at Berkeley professor David Romer ran a poll for a time. An electronic prediction bourse operated briefly at the University of Frankfurt. Last year, Steven Levitt coyly reported (on his Freakonomics blog with Stephen Dubner) a list of forty names “found on blackboard” at the University of Chicago. And annually since 2002, Thomson Scientific has published three predictions beforehand, based on citation counts — without much success. (Thomson bought Garfield’s index of footnotes, which is now known as the ISI Web of Science.)
Take Thomson’s guesses this year. Oliver Hart, of Harvard University; Bengt Homstrom, of the Massachusetts Institute of Technology; and Oliver Williamson, of the University of California at Berkeley for their work on firms and contracts? Eminently plausible.
But Jagdish Bhagwati, of Columbia; Avinash Dixit and Paul Krugman, of Princeton, for their work on trade? An improbable threesome, at least based on what the Swedes have done in the past. Nor is Harvard econometrician Dale Jorgenson a likely winner by himself. The Thomson press release failed to mention the ultimate winner, Columbia’s Edmund Phelps, though the company counts him as one of its successful predictions because Garfield noted in 1990 that he was a possible laureate. (They don’t give you a sixteen-year leeway in Las Vegas!)
In contrast, Passell’s thoughtful reporting in 1989, “unscientifically culled from a dozen phone chats with economists who, sensibly, prefer to remain anonymous,” identified many laureates of the next few years, including Gary Becker, Robert Lucas, Joseph Stiglitz and Ronald Coase. Of his “A” list, only Richard Musgrave and Edmond Malinvaud failed to be selected.
Efforts to guess a specific prizewinner get it backwards. We care about the Nobel award in economics because it tells us what the Swedes (having proven themselves very astute students of the history of thought) think about what’s been happening in economics. Citation counts tell us something of the same thing, but from a rather different vantage point. Citations — footnotes, or references in older parlance — are the means by which economists embed their new results in the work of those who have gone before. In effect, they are the votes of individual scientists about the recent contributions to their fields that have mattered most. Citation etiquette is subtle. The abundance or lack of cites can be misleading, especially in subfields and sub-subfields. But citation counting has turned out to be a useful tool for identifying the most important work in any science.
An authoritative citation survey for economics appeared in September as a working paper of the National Bureau of Economic Research. “What Has Mattered to Economics Since 1970,” by E. Han Kim and Adair Morse of the Ross School of Business of the University of Michigan and Luigi Zingales of the University of Chicago’s Graduate School of Business, is a particularly useful attempt to identify the most important work of the past thirty years. It is easy enough to imagine ways in which the Royal Academy of Sciences in Sweden may differ from the rough guidance that is implicit in the citation ranking. It is hard to imagine their ranking being superseded.
“Citations are one way that past research echoes through time,” they write. “Although the number of academic citations accumulated by a published research paper is an imperfect measure of quality or influence of that paper, citation counts do have certain virtues. They are not subjective. They are widely used in studies of academic productivity. They are reasonably comprehensive across subject areas in economics.”
So Kim, Morse and Zingales searched every article published in 41 prominent economics journals between 1970 and 2002 (rather than the unwieldy 175 economics and 40 finance journals included in the Thomson ISI/Social Science Citation Index), with a view to preparing a list of articles published in these journals that have acquired 500 or more citations.
Such cite-worthiness entitles a paper to be described unequivocally as “famous.” It doesn’t mean that the author eventually will win a Nobel Prize. But there is no doubt that their list of famous papers makes fascinating reading.
“Within these 146 papers,” they write, “an elite group of 11 economists authored or co-authored at least three papers. Robert Barro, Eugene Fama and Joseph Stiglitz have six each. Michael Jensen follows with five; Robert Lucas and David Kreps with four; and Robert Engle, Lars Hansen, Robert Merton, Edward Prescott and Stephen Ross have three each.”
The single most frequently cited article? Econometrician Halbert White’s 1980 paper on robust standard errors, “A Heteroskedasticity-Consistent Covariance-Matrix Estimator and a Direct Test for Heteroskedasticity.” White, a fascinating character virtually unknown outside his field, is a professor at the University of California at San Diego.
(I would reproduce Kim, Morse and Zingales’ Table 2 here as a service to readers, but as widely as EP circulates among the community of those who are those interested in what is going on in technical economics, this is one of those cases where copyright protection makes good sense. To help the authors get some the credit that is due them, then, take the trouble of exercising your NBER membership rights or sign up here for a free trial subscription to its working papers, then download w12526, thereby enhancing the standing of “What Has Mattered to Economists” among the NBER’s most popular papers. You won’t be sorry that you did. )
For instance, among the authors of the top fifteen papers, eight have already won Nobel Prizes (Daniel Kahneman, Clive Granger/Robert Engle, James Heckman, Fischer Black/Myron Scholes, George Akerlof, George Stigler and Robert Lucas.). Two papers of general interest in those top fifteen not honored by the Swedes stand out. They are a 1976 article by Harvard Business School’s Michael Jensen and the late William Meckling of the University of Rochester on managerial behavior and the theory of the firm, and a 1972 paper by UCLA’s Armen Alchian and Harold Demsetz on production, information costs and economic organization. Also among the authors of the top fifteen papers are three econometricians, thoroughly famous within the profession, who have not been celebrated in the wider world: David Dickey, of North Carolina State University, and W.A. Fuller (unit roots), and MIT’s Jerry Hausman (specification tests).
A number of other possibilities stand out in the top fifty papers, after those first fifteen,: a pair of papers by Paul Romer of Stanford’s Graduate Business School, “Increasing Returns and Long-Run Growth” (16) and “Endogenous Technological Change” (22); Lars Hansen of the University of Chicago, “Large Sample Properties of a Generalized Method of Moments Estimators” (17); a trio of papers by Gene Fama of the Graduate School of Business of the University of Chicago, “Efficient Capital Markets — Review of Theory and Empirical Work” (20), “Agency Problems and the Theory of the Firm” (34), and, with Michael Jensen, “Separation of Ownership and Control” (43); Christopher Sims of Princeton, “Macroeconomics and Reality: (23); and Sam Peltzman of the Graduate School of Business at the University of Chicago, “Toward a More General Theory of Regulation” (27).
Others include a pair of papers by Robert Barro of Harvard, “Are Government Bonds Net Wealth?” (28) and “Economic Growth in a Cross-Section of Countries” (33); MIT’s Holmstrom, “Moral Hazard and Observability” 39; Berkley’s Williamson, “Transaction Cost Economics” (41); Quantitative Financial Strategies’ Sanford Grossman and Harvard’s Hart, “The Costs and Benefits of Ownership,” 48; and Stanford’s Paul Milgrom and R..H. Weber, “A Theory of Auctions and Competitive Bidding.”
Included in those second 35 are several papers by other Nobel laureates: Robert Merton, Joseph Stiglitz, Finn Kydland/Edward Prescott and Gary Becker. On the list as well are two star economists who died much too young: Sherwin Rosen and Rudiger Dornbusch. The authors of the next hundred papers on the list contain nearly as many famous and potentially famous names.
There are plenty of highly-cited papers that didn’t make the list, of course, mainly those published before 1970. There are some, too, that were published in journals other than those the authors of the study deemed leading. For example, Richard Thaler, of the Graduate School of Business at the University of Chicago, would have had two hits if “Toward a Positive Theory of Consumer Choice” (631 cites) hadn’t appeared in the Journal of Economic Behavior and Organization and “Mental Accounting and Consumer Choice” (541 cites) in Marketing Science.
But star-spotting is not the really interesting thing in “What Has Mattered to Economics Since 1970.” The large sample of 146 famous papers permits Kim, Morse and Zingales to document some major shifts of interest that have taken place in technical economics in the last thirty years: “Theory loses out to empirical work, and micro and macro give way to growth and development in the 1990s,” they write. (Here the skein of work that flows out of just one thesis paper is remarkable: Paul Romer’s “Increasing Returns and Long-Run Growth” (1640 citations) was the starting point for Robert Lucas’ “On the Mechanics of Economic Development” (1772 cites); Romer’s revision of his original approach, “Endogenous Technological Change” (1333 cites); Robert Barro’s, “Economic Growth in a Cross-section of Countries” (1111 cites) and N. Gregory Mankiw, David Romer and David Weil’s “A Contribution to the Empirics of Economic Growth” (792 cites).
Moreover, “While we do not witness any decline in the primacy of production in the United States, the concentration of institutions within the US hosting and training authors of the highly-cited articles has declined substantially.” Thus 85 percent of their list of famous papers was written by researchers while they were working at universities in the US. The share doesn’t decline over time, they say; instead, it increases slightly in the last five years.
On the other hand, the concentration of talent in the three or four best departments declined substantially over thirty years, the authors note. In the early 1970s, almost a quarter of the most highly cited papers were written at the University of Chicago, and only 15 institutions were represented at all. By the end of the century, they write, no institution accounted for more than 16 percent of the market, and 26 institutions contributed to the production of highly-cited articles. The share of English university PhDs with top papers had risen slightly, from 6.3 percent at the beginning of the period to 6.6 at the end, about the same as that of Princeton PhDs.
Forecasting individual Nobel prizes in September is a mug’s game — sillier even than pedantically insisting that the prize is not really a Nobel Prize at all because it was created in 1969 as The Sveriges Riksbank Prize in Economic Sciences in Memory of Alfred Nobel. It spoils the fun of the annual surprise. But attempts to understand what has been happening in technical economics are always welcome, and to this enterprise, Han Kim, Adair Morse and Luigi Zingales have contributed a very useful tool.