Hashem Pesaran Talks about Analyzing Panel Data in Economics
Fast Moving Front Commentary, January 2012
Article: A simple panel unit root test in the presence of cross-section dependence|Pesaran, MH|J APPL ECONOM
Authors: Pesaran, MH
M. Hashem Pesaran talks with ScienceWatch.com and answers a few questions about this month's Fast Moving Fronts paper in the field of Economics & Business.
Why do you think your paper is highly cited?
It deals with an important practical problem that arises in the analysis of panel data in economics; the methodology is simple and coherent, and it is easy to apply.
Does it describe a new discovery, methodology, or synthesis of knowledge?
It describes a new methodology of testing for unit roots in the presence of error cross-sectional dependence. Most economic data either arranged by countries, regions, or industries display a considerable degree of interdependence. My paper provides a simple approach to dealing with such dependencies when testing for highly persistent time effects.
Would you summarize the significance of your paper in layman's terms?
"The changing nature of cross section dependence over time, is also of interest as it can shed light on herding behaviour and contagion."
It deals with an important topic in panel econometric modeling, that of testing for the stochastic behavior of individual panel members in the realistic case where they are subject to a common unobserved factor such as technological innovations, taste changes or evolving institutional features that are difficult to measure.
How did you become involved in this research, and how would you describe the particular challenges, setbacks, and successes that you've encountered along the way?
My involvement in research on panel data models in economics started with a paper I published in the Journal of Econometrics with Ron Smith in 1995.
This paper focused attention on dynamic panels with heterogeneous slopes where the cross section and time series dimensions of the data are relatively large. However, it did not address the problem of cross-section dependence. The importance of allowing for cross-section dependence in the analysis of panels became clear to me during a period of three years that I worked in the private sector, analyzing the interactions across markets and economies.
Modeling of both observable and unobservable interactions across individuals, markets, and economies is important from a number of different perspectives. It is important for valid statistical inference. It is vital for risk analysis where the degree to which risk can be diversified critically depends on the degree of cross-section dependence.
Allowing for dependence across individuals, countries, or markets creates a number of challenges for statisticians and econometricians, as well as for the economic theorist where network interactions are of interest. So far I have been fortunate to have been able to make a number of significant contributions in this area.
Where do you see your research leading in the future?
An important focus will be in exploring the nature of cross-section dependence in panels. More specifically identifying whether cross-section dependence is best characterized by spatial (or network) patterns, by unobserved common factors, or by both. The changing nature of cross-section dependence over time is also of interest as it can shed light on herding behavior and contagion.
KEYWORDS: PANEL UNIT ROOT TEST, CROSS-SECTION DEPENDENCE, HETEROGENOUS PANELS, COINTEGRATION, POWER.