Econometrics have experienced phenomenal growth in the past 50 years. There are six volumes of the Handbook of Econometrics, most of it dealing with post 1960’s research. A lot of the recent growth reflects the rapid advances in computing technology. The broad availability of micro data bases is a major advance which facilitated the growth of panel data methods (see Chapter 12) and microeconometric methods especially on sample selection and discrete choice (see Chapter 13) and that also lead to the award of the Nobel Prize in Economics to James Heckman and Daniel McFadden in 2000. The explosion in research in time series econometrics which lead to the development of ARCH and GARCH and cointegration (see Chapter 14) which also lead to the award of the Nobel Prize in Economics to Clive Granger and Robert Engle in 2003. It is a different world than it was 30 years ago. The computing facilities changed dramatically. The increasing accessibility of cheap and powerful computing facilities are helping to make the latest econometric methods more readily available to applied researchers. Today, there is hardly a field in economics which has not been intensive in its use of econometrics in empirical work. Pagan (1987, p. 81) observed that the work of econometric theorists over the period 1966-1986 have become part of the process of economic investigation and the training of economists. Based on this criterion, he declares econometrics as an “outstanding success.” He adds that:
The judging of achievement inevitably involves contrast and comparison. Over a
period of twenty years this would be best done by interviewing a time-travelling
economist displaced from 1966 to 1986. I came into econometrics just after the beginning of this period, so have some appreciation for what has occurred. But because I have seen the events gradually unfolding, the effects upon me are not as dramatic. Nevertheless, let me try to be a time-traveller and comment on the perceptions of a 1966’er landing in 1986. My first impression must be of the large number of people who have enough econometric and computer skills to formulate, estimate and simulate highly complex and non-linear models. Someone who could do the equivalent tasks in 1966 was well on the way to a Chair. My next impression would be of the widespread use and purchase of econometric services in the academic, government, and private sectors. Quantification is now the norm rather than the exception. A third impression, gleaned from a sounding of the job market, would be a persistent tendency towards an excess demand for well-trained econometricians. The economist in me would have to acknowledge that the market judges the products of the discipline as a success.
The challenge for the 21st century is to narrow the gap between theory and practice. Many feel that this gap has been widening with theoretical research growing more and more abstract and highly mathematical without an application in sight or a motivation for practical use. Heckman (2001) argues that econometrics is useful only if it helps economists conduct and interpret empirical research on economic data. He warns that the gap between econometric theory and empirical practice has grown over the past two decades. Theoretical econometrics becoming more closely tied to mathematical statistics. Although he finds nothing wrong, and much potential value, in using methods and ideas from other fields to improve empirical work in economics, he does warn of the risks involved in uncritically adopting the methods and mind set of the statisticians:
Econometric methods uncritically adapted from statistics are not useful in many research activities pursued by economists. A theorem-proof format is poorly suited for analyzing economic data, which requires skills of synthesis, interpretation and empirical investigation. Command of statistical methods is only a part, and sometimes a very small part, of what is required to do first class empirical research.
In an Econometric Theory interview with Jan Tinbergen, Magnus and Morgan (1987, p. 117) describe Tinbergen as one of the founding fathers of econometrics, publishing in the field from 1927 until the early 1950s. They add: “Tinbergen’s approach to economics has always been a practical one. This was highly appropriate for the new field of econometrics, and enabled him to make important contributions to conceptual and theoretical issues, but always in the context of a relevant economic problem.” The founding fathers of econometrics have always had the practitioner in sight. This is a far cry from many theoretical econometricians who refrain from applied work.
Geweke, Horowitz, and Pesaran (2008) provide the following recommendations for the future:
Econometric theory and practice seek to provide information required for informed decision-making in public and private economic policy. This process is limited not only by the adequacy of econometrics, but also by the development of economic theory and the adequacy of data and other information. Effective progress, in the future as in the past, will come from simultaneous improvements in econometrics, economic
theory, and data. Research that specifically addresses the effectiveness of the interface between any two of these three in improving policy — to say nothing of all of them — necessarily transcends traditional subdisciplinary boundaries within economics. But it is precisely these combinations that hold the greatest promise for the social contribution of academic economics.
1. See the interview of Professor L. R. Klein by Mariano (1987). Econometric Theory publishes interviews with some of the giants in the field. These interviews offer a wonderful glimpse at the life and work of these giants.
2. Simultaneous equations model is an integral part of econometrics and is studied in Chapter 11.
3. Tjalling Koopmans was the joint recipient of the Nobel Prize in Economics in 1975. In addition to his work on the identification and estimation of simultaneous equations models, he received the Nobel Prize for his work in optimization and economic theory.
4. I encountered this attack by Keynes on Tinbergen in the inaugural lecture that Peter C. B. Phillips (1977) gave at the University of Birmingham entitled “Econometrics: A View From the Toolroom,” and David F. Hendry’s (1980) article entitled “Econometrics – Alchemy or Science?”