Michael Keane (economist)

Michael Patrick Keane (born 1961) is an American/Australian economist who is currently professor of economics and Australian Research Council laureate fellow at the University of New South Wales. From 2012 to 2017 he was the Nuffield Professor of Economics at the University of Oxford and a professorial fellow of Nuffield College.[1][2] He is considered one of the world's leading experts in the fields of Choice Modelling, structural modelling, simulation estimation, and panel data econometrics.[3][4][5][6][7]

Michael P Keane
Michael Keane at UTS in 2008
InstitutionUNSW Business School
FieldEconometrics
Alma materMIT, Brown University
Doctoral
advisor		Robert Moffitt
InfluencesKenneth Wolpin
John Geweke
James Heckman
ContributionsChoice Modelling, Structural Modelling, Simulation Methods, Panel Data Econometrics
Awards
Information at IDEAS / RePEc

He is also one of the world's leading economists by many measures of research productivity.[8][9] Keane works in numerous areas including labor economics, econometrics, consumer demand models, marketing, industrial organization, health economics, and trade.

He is currently a chief investigator of the Australian Research Council Centre of Excellence in Population Ageing Research (Cepar).[10] From 2006–10 he was Co-Director of the Centre for the Study of Choice (CenSoC) at UTS.[11] Keane became a dual citizen of Australia in 2010.

Education and career

Keane was born in Suffern, New York, United States, and graduated from Xavier High School in Manhattan in 1979. He received a B.S. degree from the Massachusetts Institute of Technology in 1983, and a Ph.D. from Brown University in 1990.

In 1993, he became a tenured associate professor at the University of Minnesota, and was promoted to full professor in 1996. He subsequently held full professor positions at New York University (1998–2001) and Yale University (2000-2006).

In 2006, he moved to Australia to take up an Australian Federation Fellowship at the University of Technology Sydney.[12] In 2011, he became an Australian Laureate Fellow at the University of New South Wales.[13]

Keane was elected a Fellow of the Econometric Society (2005),[14] to the Council of the Econometric Society (2009), and a Fellow of the Academy of Social Sciences in Australia (2012). He was the recipient of the John D.C. Little award for the Best Paper in Marketing (1996) and the Kenneth J. Arrow Award for Best Paper in Health Economics (2008).[15] In 2004–05, Keane was the Goldwater Chair of American Institutions at Arizona State University and, subsequently, has been a regular visiting professor there.

Contributions

Keane's work is notable for the fact that it spans a very wide range of substantive and methodological areas. He is best known for work on the following topics:

Discrete choice models

Keane's work on recursive importance sampling (the "GHK" algorithm), contained in his thesis (1990) and published in 1993–1994, made it feasible to estimate a much larger class of discrete choice models than was previously possible. In particular, his thesis developed a fast algorithm for the highly accurate calculation of areas of polyhedrons in very-high-dimensional spaces. While primarily a result in applied mathematics, this result is very useful in economics (and other social sciences) because the choice probabilities in discrete choice models generally have this form.[16] The GHK algorithm is now included in many popular econometrics software packages, including SAS, Stata, GAUSSX, Matlab and R-Cran-Bayesm,[17] and is a standard topic in graduate econometrics texts.[18]

Consumer demand and marketing

A 1996 paper with Tulin Erdem in Marketing Science presented what is now the main economic model of advertising and consumer learning. This paper received the John D.C. Little Award for the Best Paper in Marketing in 1996, and it has had a major impact on the fields of marketing and industrial organization. There is now a large literature on consumer learning based on the Erdem-Keane framework.[19] Erdem and Keane (among others) have argued that their framework can provide an economic explanation for the phenomenon known as brand equity, based on incomplete information and risk aversion.[20][21] The November 2013 issue of Marketing Science contains an extensive review of the large literature based on the Erdem-Keane framework.[22]

Dynamic life-cycle models

In a series of joint papers with Kenneth Wolpin, published between 1994 and 2010, Keane developed a major line of research on dynamic life-cycle models of career (i.e., school and work) choices.[23][24] This line of research is notable both for the methodological contributions on how to estimate these types of models, and for its substantive economic contributions. Methodologically, their method of approximating the solution to computationally intensive dynamic programming problems led to a great expansion in the class of such models that are feasible to implement empirically (i.e., their method made it possible to estimate models with many more choices and state variables than was possible previously). Substantively, their seminal 1997 paper on "The Career Decisions on Young Men" presented the so-called "90 percent result"—i.e., that most of what matters for lifetime earnings has already happened by age 16. This result helped to shift the focus of the human capital literature away from college education and towards early childhood education. This is now a very active area of research in economics, which has been pursued by both Keane and Wolpin and, quite notably, by the Nobel Prize–winning economist James Heckman,[25] among others.

Welfare and transfer programs

His 1998 paper with Robert Moffitt, entitled "Multiple welfare program participation and labor supply," has had great influence on subsequent models of transfer/welfare programs. This was the first paper to account for the very complex budget constraints created when people may participate in several government welfare programs simultaneously. The model predicted that welfare caseloads would drop substantially in response to earnings subsidies (like the Earned Income Tax Credit).

Human capital and labor supply
Michael Keane at the Becker-Friedman Institute, 2013

In recent years, Keane has argued persuasively that, due to human capital effects, labor supply elasticities are much larger than the previous consensus of the economics profession would suggest. These views are presented in Imai and Keane (2004), Keane (2010) and Keane and Rogerson (2012).[26][27] If correct, his views imply that welfare losses from income taxation are much higher than was previously thought. Recently, Keane gave a keynote lecture summarizing this work at the 2015 annual meeting of the Royal Economic Society.[28] His Cowles lecture at the 2011 summer meeting of the Econometric Society also dealt with this topic.

Modelling expectations

Keane's papers with David Runkle (1990, 1998) are considered fundamental contributions in the literature on how people form expectations. These papers showed that the widespread empirical failure of "rational expectations" was in fact due to a set of econometric and data problems (such as the failure to account for aggregate economic shocks and the effects of data revisions).

Panel data econometrics

The recursive importance sampling algorithm developed in Keane's 1994 Econometrica paper made it possible to estimate panel data discrete choice models with complex serial correlation patterns. This approach is now widely used to model discrete dynamic processes in marketing and labor economics. Keane's 1992 Journal of Business and Economic Statistics paper with David Runkle developed a new approach for estimating linear panel data models in cases where the available instruments are predetermined but not strictly exogenous. This is a very common case that includes all dynamic panel data models as a leading example. Chamberlain (1982) noted that the Keane-Runkle approach was not fully efficient because it fails to use all available instruments. Keane and Runkle (1992) responded that the use of additional instruments would be unwise as it would generate bias due to the "many instrument problem." Nevertheless, the development of more efficient panel data estimators based on more instruments became a major research program in the 90s. Examples of this line of research are well-known papers by Arellano-Bond (1991), Ahn-Schmidt (1995), Arellano-Bover (1995) and Blundell-Bond (1998). For a review of the literature see Baltagi (2005) chapter 8.[29] More recent work, such as Ziliak (1997),[30] supports Keane and Runkle (1992)'s original argument that use of additional instruments may cause severe bias.[31]

Econometric methodology

Keane is well known as a champion of the "structural econometrics" school, which emphasizes the important role of economic theory in empirical work. This contrasts with the "experimental school" which has become very popular in the last 20 years. The latter seeks to use "natural experiments" to substitute for economic theory. He has written a number of articles on the importance of theory and the limitations of experiments (see Keane 2010a, 2010b).[32]

Other areas

In addition, Keane has done significant work in many other areas, such as health economics, child development, international trade, political economy, experimental economics, and development economics.

Selected papers
  • Testing the rationality of price forecasts: New evidence from panel data, (with David Runkle), American Economic Review, 80:4, (1990), 714–35.
  • On the Estimation of Panel Data Models with Serial Correlation when Instruments are Predetermined but not Strictly Exogenous, (with David Runkle), Journal of Business and Economic Statistics, 10:1, (1992), 1–9.
  • Simulation Estimation for Panel Data Models with Limited Dependent Variables, in The Handbook of Statistics, G.S. Maddala, C. R. Rao and H.D. Vinod editors, North Holland publisher (1993).
  • A Computationally Practical Simulation Estimator for Panel Data, Econometrica, 62:1, (1994), 95–116.
  • The Solution and Estimation of Discrete Choice Dynamic Programming Models by Simulation: Monte Carlo Evidence, (with Kenneth Wolpin), Review of Economics and Statistics, 76:4, (1994), 648–72.
  • Decision Making under Uncertainty: Capturing Dynamic Brand Choice Processes in Turbulent Consumer Goods Markets (with Tulin Erdem), Marketing Science, 15:1, (1996), 1–20.
  • The Career Decisions of Young Men, (with Kenneth Wolpin), Journal of Political Economy, 105:3, (1997), 473–522.
  • A Structural Model of Multiple Welfare Program Participation and Labor Supply, (with Robert A. Moffitt), International Economic Review, 39:3, (1998), 553–89.
  • Are Financial Analysts' Forecasts of Corporate Profits Rational? (with David Runkle), Journal of Political Economy, 106:4, (1998), 768–805.
  • A Model of Health Plan Choice: Inferring Preferences and Perceptions from a Combination of Revealed Preference and Attitudinal Data, (with Katherine Harris), Journal of Econometrics, 89, (1999), 131–57.
  • Mixture of Normals Probit Models, (with John Geweke), in Analysis of Panels and Limited Dependent Variable Models, Hsiao, Lahiri, Lee and Pesaran editors, Cambridge University Press, (1999), 49–78.
  • Bayesian Inference for Dynamic Discrete Choice Models without the Need for Dynamic Programming, (with John Geweke), in Simulation Based Inference and Econometrics, Mariano, Schuermann and Weeks editors, Cambridge University Press, (1999), 100–31.
  • The Effect of Parental Transfers and Borrowing Constraints on Educational Attainment, (with Kenneth Wolpin), International Economic Review, 42:4, (2001), 1051–103.
  • Intertemporal Labor Supply and Human Capital Accumulation, (with Susumu Imai), International Economic Review, 45:2, (2004), 601–42.
  • Behavior in a Dynamic Decision Problem: An Analysis of Experimental Evidence Using a Bayesian Type Classification Algorithm, (with Daniel Houser and Kevin McCabe), Econometrica, 72:3, (2004), 781–822.
  • Accounting for the Growth of MNC-Based Trade using a Structural Model of US MNCs, (with Susan Feinberg), American Economic Review, 96:5, (2006), 1515–58.
  • Sources of Advantageous Selection: Evidence from the Medigap Insurance Market, (with Hanming Fang and Dan Silverman), Journal of Political Economy, 116:2 (2008), 303–50.
  • Structural vs. Atheoretic Approaches to Econometrics, Journal of Econometrics, 156:1, (2010a), 3–20.
  • A Structural Perspective on the Experimentalist School, Journal of Economic Perspectives, 24:2 (2010b), 47–58.
  • The Role of Labor and Marriage Markets, Preference Heterogeneity and the Welfare System in the Life-Cycle Decisions of Black, Hispanic and White Women, (with Kenneth Wolpin), International Economic Review, 51:3, (2010), 851–92.
  • Child Care Choices and Children’s Cognitive Achievement: The Case of Single Mothers, (with Raquel Bernal), Journal of Labor Economics, 29:3, (2011), 459–512.
  • Labor Supply and Taxes: A Survey, Journal of Economic Literature, 49:4, (2011), 961–1075.
  • Micro and Macro Labor Supply Elasticities: A Reassessment of the Conventional Wisdom (with Richard Rogerson), Journal of Economic Literature, 50:2, (2012), 464–76.

References
  1. "Academic Experience" (PDF). ifs.org.uk. Retrieved 27 September 2018.
  2. "Michael Keane arrives". business.unsw.edu.au. 25 August 2017. Retrieved 27 September 2018.
  3. "Professor Michael Keane". assa.edu.au. Retrieved 10 September 2017. According to the Academy of the Social Sciences in Australia, " He is considered one of the world's leading experts in the fields of Choice Modelling, structural modelling, simulation estimation, and panel data econometrics."
  4. "Nuffield Professorship of Economics". ox.ac.uk/. Retrieved 10 September 2017. According to the University of Oxford Nuffield Professorship announcement, "He is a world leader in choice modelling, the statistical technique that involves developing mathematical models to predict how individuals or companies make different types of decisions."
  5. "Michael Keane". Bfi.uchicago.edu. Retrieved 21 February 2015. According his profile at the Becker-Friedman Institute at the University of Chicago, "Michael Keane is a distinguished behavioral economist and econometrician and a world leader in choice modeling."
  6. "Prestigious joint appointment for Michael Keane". business.unsw.edu.au. Retrieved 10 September 2017. According to the UNSW Business School appointment announcement, "He is widely known for his seminal contributions in empirical microeconomics and econometrics. His methodological innovations are used extensively in a variety of applied fields including labour economics, health economics and marketing."
  7. "Michael P Keane". ifs.org.uk. Retrieved 10 September 2017. According to the Institute for Fiscal Studies profile, “Keane is best known for work on simulation methods (e.g., the “GHK algorithm”) and for contributions to the theory and application of dynamic discrete choice models."
  8. "As of September 2017 Keane has over 16,000 citations on Google scholar". Retrieved 10 September 2017.
  9. And, according to the IDEAS/RePEc October 2013 rankings of economists by several measures of research productivity, he ranks 110th among all economists in the world, 5th in the United Kingdom, 37th in the world over the past 10 years, and 1st among all economists in the 1990 PhD cohort. In terms of total journal pages, weighted by journal impact factors and number of co-authors, he is 23rd in the world and 5th in the world over the last 10 years.
  10. "Cepar Homepage". Cepar.edu.au. Retrieved 2014-01-15.
  11. "UTS: About – Study of Choice". Censoc.uts.edu.au. 2012-10-15. Archived from the original on 2014-01-26. Retrieved 2014-01-15.
  12. "UTS tops economics rankings in Australia". Theaustralian.com.au. 2013-08-29. Retrieved 2014-01-15.
  13. "Australian Laureate Fellows at UNSW". Research.unsw.edu.au. 2013-06-19. Archived from the original on 2014-01-16. Retrieved 2014-01-15.
  14. "Welcome to the website of The Econometric Society An International Society for the Advancement of Economic Theory in its Relation to Statistics and Mathematics". Econometricsociety.org. 2003-01-03. Archived from the original on 2008-12-10. Retrieved 2014-01-15.
  15. "Arrow Award Winners | iHEA | International Health Economics Association". Archive.healtheconomics.org. 2007-07-10. Archived from the original on 2014-01-02. Retrieved 2014-01-15.
  16. See Skrondal, Anders and Sophia Rabe-Hesketh (2004), "Generalized Latent Variable Modeling," London: Chapman & Hall/CRC, pp. 173–4, where they state "In econometrics the GHK simulator is very popular for models with multi-normal latent responses, for instance probit panel (longitudinal) models (e.g., Keane, 1994, Geweke, et al., 1994). This is probably because the simulator has been shown to outperform other simulators although it is relatively easy to implement."
  17. "Cappellari L. and Jenkins, S.P. (2003), "Multivariate probit regression using simulated maximum likelihood," The Stata Journal, Vol. 3:3, pp. 278–94". Stata-journal.com. Retrieved 2014-01-15.
  18. Greene, William H. (2003). "Basic textbook discussion of the algorithm" (Fifth ed.). Econometric Analysis. pp. 931–33. Retrieved 21 February 2015.
  19. "Branding and Brand Equity : SAGE Knowledge". Knowledge.sagepub.com. 2009-06-22. Retrieved 2015-02-21.
  20. "Google Scholar Citations". 1998-12-31. Retrieved 2015-02-21.
  21. Tülin Erdem; Joffre Swait (June 2004). "Brand Credibility, Brand Consideration, and Choice". Journal of Consumer Research. The University of Chicago Press. 31 (1): 191–198. doi:10.1086/383434. JSTOR 383434.
  22. Ching, A. T.; Erdem, T. L.; Keane, M. P. (2013). "Invited Paper—Learning Models: An Assessment of Progress, Challenges, and New Developments". Marketing Science. 32 (6): 913–938. doi:10.1287/mksc.2013.0805.
  23. Heckman, J.; Lochner, L.; Todd, P. (2006). "Handbook of the Economics of Education". p. 441. Retrieved 21 February 2015. Keane and Wolpin (1997) and Eckstein and Wolpin (1999) pioneered the estimation of dynamic discrete choice models for analyzing schooling choices.
  24. Belzil, Christian (2007). "The return to schooling in structural dynamic models: A survey" (PDF). European Economic Review. Scienedirect.com. 51 (5): 1059–1105. doi:10.1016/j.euroecorev.2007.01.008. S2CID 10552331. In his survey of the human capital literature, Belzil (2007) states that "The first stage consists of seminal work on schooling and earnings by Becker (1964, 1967) and Mincer (1958). The second stage, which culminates in Willis and Rosen (1979), is largely influenced by the econometric self-selection literature (Heckman, 1976). Finally, the recent literature, stimulated by Keane and Wolpin (1997), uses stochastic dynamic programming techniques, and forms a third stage ..."
  25. "The Economics and Econometrics of Human Development by James Heckman". YouTube. 2013-11-07. Retrieved 2015-02-21.
  26. Prescott Nobel Lecture, In discussing the long-standing conflict between the typically small empirical estimates of labor supply elasticities and the large elasticities implied by macro models, Prescott (2006) states: "Recently, Imai and Keane (2004) resolved the conflict... [they] took into consideration the value of human capital people acquire when working ... This adjustment dramatically increases the estimate of the labor supply elasticity ..."
  27. Becker-Posner Blog "Is Raising Marginal Tax Rates on Higher Income Individuals a Good Idea?" July 15, 2012, In discussing Keane and Rogerson (2012), Becker states: "... on the job investments in human capital ... bias downwards the micro estimates of labor supply elasticities that are based on variations in hours worked over the lifecycle. This bias is especially large at younger ages where most of these investments occur since the true earnings at these ages is much larger than the observed wages at these ages because younger workers are raising their future wages through investments in their human capital. This bias helps explain why computed elasticities of labor supply are usually greater for older workers ... These and other corrections to simple interpretations of the micro evidence on labor responses to changes in tax rates clearly suggest that aggregate labor responses to tax rates may be quite large."
  28. http://www.fsmevents.com/res/2015/session02/
  29. Baltagi, Badi (2005). "The Econometric Analysis of Panel Data" (3rd ed.). Chichester: John Wiley & Sons. Retrieved 21 February 2015.
  30. Ziliak, James P (1997-02-02). "Efficient Estimation with Panel Data When Instruments Are Predetermined: An Empirical Comparison of Moment-Condition Estimators". Ideas.repec.org. Retrieved 2014-01-15.
  31. See the well-known reference on panel data by Baltagi (2005), pp. 151–52, where he states: "Ziliak (1997) performs an extensive set of Monte Carlo experiments for a dynamic panel data model ... Ziliak finds that the downward bias of GMM is quite severe as the number of moment conditions expands, outweighing the gains in efficiency. Interestingly, Ziliak finds that the forward filter 2SLS estimator proposed by Keane and Runkle (1992) performs best in terms of the bias/efficiency tradeoff and is recommended."
  32. See also the discussion by Rosenzweig, Mark; Wolpin, Kenneth (2000). "Natural 'Natural Experiments' in Economics". Journal of Economic Literature. 38 (4): 827–74. doi:10.1257/jel.38.4.827. JSTOR 2698663.
This article is issued from Wikipedia. The text is licensed under Creative Commons - Attribution - Sharealike. Additional terms may apply for the media files.