The Nobel Prize

2000 Daniel L. McFadden (b. 1937)

Daniel McFadden was born in Raleigh, North Carolina, USA in 1937. He grew up in rural isolation as a consequence of his parents’ decision to opt for an unconventional life on a remote farm with no electricity or running water. McFadden’s mother had been trained as an architect and his father, despite little formal education, had kept the books at a bank. After helping out on the farm and with few other readily available diversions, McFadden relaxed by reading books from his father’s library. He remembers being a good student and he pushed beyond the limitations of his high-school curriculum by taking correspondence courses in algebra and geometry. McFadden was helped in this endeavour by his mother who had found work as a high-school mathematics teacher. At age 16 he passed the entrance examination for the University of Minnesota and was awarded a BS in physics with High Distinction in 1957, when he was 19 (Nobel Foundation, 2004).

McFadden continued at Minnesota as a graduate student of physics but in 1958 he secured a place on a Ford Foundation sponsored behavioural science training programme and began a much more eclectic programme of PhD courses in psychology, sociology, economics, anthropology, political science, mathematics and statistics. During this period, McFadden ‘developed an interest in mathematical models of learning and choice’ and found that some economists at Minnesota – Professors John Chipman and Leo Hurwicz – were working in this area; accordingly, he elected to enrol for an economics PhD (Nobel Foundation, 2004). He was awarded a PhD in behavioural science (economics) in 1962. As part of his doctoral training McFadden spent a summer externship at Stanford University, where he worked for Kenneth Arrow and Marc Nerlove. At Stanford he also ‘had a brief interaction with Professor Hirofumi Uzawa that proved to be a pivotal point in my research training, giving me a dissertation topic, and, most importantly, a flash of understanding of how to use mathematics as a research tool’ (Nobel Foundation, 2004).

McFadden’s first academic post following the completion of his PhD was at the University of Pittsburgh where in 1962–63 he was a Mellon post-doctoral fellow and Assistant Professor of Economics. In 1963 he moved to the University of California at Berkeley as Assistant Professor of Economics. In 1966 he was promoted to Associate Professor of Economics and, following an appointment as visiting associate professor at the University of Chicago in 1966–67, he became Professor of Economics at Berkeley in 1968. In 1977–78 he was Irving Fisher Research Professor at Yale University. McFadden left Berkeley in 1978, taking up the post of Professor of Economics at the Massachusetts Institute of Technology (MIT), where, in 1984, he was given the James R. Killian Chair. In 1986 he became director of the Statistics Center at MIT, a position he held until 1988. After spending 1990 as a Sherman Fairchild Distinguished Fellow at the California Institute of Technology, McFadden left MIT in 1991 to return to Berkeley as Professor of Economics and holder of the E. Morris Cox Chair. At Berkeley he has also been director of the Econometrics Laboratory (1991–95 and 1996–date) and chair of the Department of Economics (1995–96).

McFadden’s professional and public service duties have included service on a number of advisory boards including: the Economics Advisory Panel of the National Science Foundation (1969–71); the Advisory Committee on Transportation Models; the Metropolitan Transportation Commission (1975); the National Academy of Sciences (NAS) Committee on Energy Demand Modeling (1983–84); and the NAS Commission on Science, Engineering and Public Policy (1995–date). McFadden was president of the Econometric Society in 1985, vice president of the American Economic Association in 1994 and its president-elect in 2004.

McFadden’s offices and honours include election as a fellow of the Econometric Society in 1969 and the award of the John Bates Clark Medal by the American Economic Association in 1975. He was elected to the American Academy of Arts and Sciences in 1977 and received the MIT Outstanding Teacher Award in 1981. In 1986 he was awarded the Frisch Medal by the Econometric Society and won the Richard Stone Prize in Applied Econometrics in 2000–2001. In 2000, McFadden was awarded the Nobel Memorial Prize in Economics ‘for his development of theory and methods for analysing discrete choice’ (Nobel Foundation, 2004).

In the introduction to his Nobel lecture, McFadden (2001, p. 351) notes that, prior to the 1960s, consumer theory in economics ‘was usually developed in terms of a representative agent, with marketlevel behaviour given by the representative agent’s behaviour writ large’ (emphasis in original). This approach was consistent with an analysis of choice and decision expressed in continuous form as, for example, in the familiarly referenced trade-off between the consumption of guns and butter: more of one can be consumed but only at the expense of forgone consumption of the other. Choice, however, can take discrete as well as continuous forms. Decisions whether or not to work, marry, to live in a particular location, to select a particular mode or mix of transportation, all involve discrete choices: thus, a person is either married or not married, tradeoffs between these two states are not possible. McFadden’s Nobel award recognises the unique contribution he has made to the empirical study of discrete choice: his work resulted in ‘new statistical methods based on a new economic theory of discrete choice’ (Nobel Foundation, 2004). Indeed, his 1974 paper, ‘Conditional Logit Analysis of Qualitative Choice Behaviour’, ‘was immediately recognised as a paradigmatic breakthrough’, while related work (Domencich and McFadden, 1975a; McFadden, 1975b, 1976) – partly with Thomas Domencich – ‘fundamentally changed the way econometricians and empirical researchers thought about the econometric analysis of individual behavior’ (Manski, 2001, p. 224).

One of the key milestones in the development of McFadden’s work was a request by a Berkeley graduate student, Phoebe Cottingham, for advice on the analysis of data on freeway routing choices by the California Department of Highways. Drawing on work by L.L. Thurstone, Jacob Marschak and R. Duncan Luce,

McFadden constructed an econometric model of discrete choice between alternative freeway routes. Although Cottingham’s thesis was completed before McFadden’s model could be computed, he was later able to use it to analyse her data (McFadden, 2001). Although described in McFadden’s (1974) paper as a conditional logit model, it is now generally known as the multinomial logit (MNL) model. McFadden’s analysis of discrete choice was the subject of rapid and wide application. His own research has included travel demand forecasting and his Nobel citation singles out in particular work on the design of the Bay Area Rapid Transport (BART) system in San Francisco (see McFadden et al., 1977; 1978a); more recent research, also acknowledged in his Nobel award, has examined the housing decisions of elderly people (see, for example, McFadden and Feinstein, 1989). Further innovations in econometric modelling have included the development, with Jeffrey Dubin, of a means of analysing discrete and continuous decision making in the context of consumer products (McFadden and Dubin, 1984).

McFadden’s approach to economics has been to consistently focus on applied problems. In his Nobel autobiography he reflects that

[A] common theme of my research has been an emphasis on tightly binding economic theory and the problem of measurement and analysis, and on developing theoretical and statistical tools that expand the options available to applied economists. I have a strong appreciation for elegant and innovative mathematics and statistics, but as a matter of scientific priority try to keep my research focused on concrete applications, and provide templates for applied economists to follow.

There is no doubt that he has been highly successful in the latter endeavour. As Manski (2001) has noted, in addition to those areas McFadden himself has actively researched, his work has been widely applied in fields such as marketing, telecommunications and industrial organisation.

Beyond his work on discrete choice, McFadden has made contributions to a number of other areas of economics, including the analysis of production (McFadden, 1963; Fuss and McFadden, 1978b), economic theory (McFadden, 1969), and growth theory (McFadden, 1973). In addition, with fellow Nobel Laureate Daniel Kahneman and others, he has explored the effects of Kahneman and Tversky’s ‘anchoring’ heuristic on the willingness to pay for public goods (Green et al., 1998; and see entry on Kahneman in the present volume).

As with many other Laureates, McFadden is concerned to acknowledge the importance of the contributions of others to his own success. Notably, like Kahneman, he cites the work of Kahneman’s long-time collaborator Amos Tversky. McFadden donated his Nobel Prize money to the East Bay Community Foundation, directing that it should be used for the promotion of arts and education.

Main Published Works
(1963), ‘Constant Elasticity of Substitution Production Functions’, Review of Economic Studies, 30, June, pp. 73–83.
(1969), ‘On Hicksian Stability’, in J.N Wolfe (ed.), Value, Capital and Growth, Edinburgh: Edinburgh University Press, pp. 329–51.
(1973), ‘On the Existence of Optimal Development Programmes in Infinite Horizon Economies’, in J. Mirrlees and N.H. Stern (eds), Models of Economic Growth, London: Macmillan, pp. 260–82.
(1974), ‘Conditional Logit Analysis of Qualitative Choice Behaviour’, in P. Zarembka (ed.), Frontiers in Econometrics, New York: Academic Press, pp. 105–42.
(1975a), Urban Travel Demand (with T. Domencich), Amsterdam: North-Holland; reprinted by The Blackstone Company: Mount Pleasant, MI, 1976.
(1975b), ‘The Revealed Preferences of a Government Bureaucracy: Theory’, Bell Journal of Economics and Management Science, 6, Autumn, pp. 401–16.
(1976), ‘The Revealed Preferences of a Government Bureaucracy: Empirical Evidence’, Bell Journal of Economics and Management Science, 7, Spring, pp. 55–72.
(1977), ‘Demand Model Estimation and Validation’ (with A. Talvitie, S. Cosslett, I. Hasan, M. Johnson, F. Reid, K. Train), in Urban Travel Demand Forecasting Project, Final Report, vol. V, Institute of Transportation Studies, University of California, Berkeley.
(1978a), ‘Modeling the Choice of Residential Location’, in A. Karlqvist, L. Lundqvist, F. Snickars and J. Weibull (eds), Spatial Interaction Theory and Planning Models, Amsterdam: North-Holland, pp. 75–96.
(1978b), Production Economics: A Dual Approach to Theory and Applications, vols I and II (ed. with M. Fuss), Amsterdam: North-Holland.
(1984), ‘An Econometric Analysis of Residential Electric Appliance Holdings and Consumption’ (with J. Dubin), Econometrica, 52, March, pp. 345–62.
(1989), ‘The Dynamics of Housing Demand by the Elderly: Wealth, Cash Flow, and Demographic Effects’ (with J. Feinstein), in D. Wise (ed.), Issues in the Economics of Aging, Chicago: University of Chicago Press, pp. 33–87.
(1998), ‘Referendum Contingent Evaluation, Anchoring and Willingness to Pay for Public Goods’ (with D. Green, K. Jacowitz and D. Kahneman), Resource and Energy Economics, 20, June, pp. 85–116.
(2001), ‘Economic Choices’, American Economic Review, 91, June, pp. 351–78.

Secondary Literature
Heinrich, C.J. and J.B. Wenger (2002), ‘The Economic Contributions of James J. Heckman and Daniel L. McFadden’, Review of Political Economy, 14, January, pp. 69–89.
Manski, C.F. (2001), ‘Daniel McFadden and the Econometric Analysis of Discrete Choice’, Scandinavian Journal of Economics, 103 (2), pp. 217–29.

No comments: