Applied panel data analysis: short panels

Econometrics as a subject is often misconstrued as an esoteric subject reserved for an elitist few who may want to pursue it only for the purpose of meeting academic requirements and pursuing purely an academic career. The truth of the matter is that with the considerable progress that has been made over the last few decades in computational facilities, econometrics is beginning to find increasing relevance in applications to real world problems. Its tools and techniques of analysis are beginning to find direct applications in addressing challenges of decision making that hitherto had been left to rule of thumb and back of the envelope calculations. The analyst in business, and government, where competing ideas and questions demand answers that must be confronted with analytical rigour, is now more able to do so with greater ease than in the past, with advances in applied econometrics. Applied econometric analyses help us to distinguish between sound, feasible and implementable solutions and esoteric fantasies. It is increasingly becoming easier to find reasonable quantitative answers to quantitative challenges. The well equipped analyst, policy adviser or policy maker no longer needs to rely on mere rule-of-thumb approach, or simply stumbling through dark alleys approach to tackling complex real world problems that confronts him or her in modern day business and government environment. Econometrics is finding increasing applications in assisting the decision maker to establish rational and sound basis for making well informed choices in complex situations. This manual is one in the intended series of hands on training material, aimed at equipping the practicing or would-be applied econometrician, who needs to gain competence in the application of econometric tools in the day-to-day analysis of real world problems at his or her work place. The level of previous knowledge of econometrics assumed in the series is exposure to basic undergraduate course in econometrics or its equivalent. The series are designed to accommodate the needs of the average would-be applied econometrician from a wide variety of backgrounds in the Science, Applied Sciences and the Social Sciences. The training material differs from the conventional text book in econometrics in a number of ways. First, it seeks to integrate real-world questions and data with development of econometric theory. Secondly, it seeks to encourage direct hands-on training on how to apply relevant techniques and models to tackling real–world problems employing appropriate and readily available software. We all know that it is one thing to learn and master econometric principles, techniques and models in a typical textbook approach and yet, have difficulty in relating these in application to real world problems. Our objective in these series is to provide practical manuals that would assist the average would be applied econometrician to bridge this gap more easily, and demystify econometrics as an elitists concern of a select few. In this particular volume, attention is focused on Applied Panel Data Analyses. It starts from simple static model constructs and progresses to more sophisticated dynamic model constructs and implementation in STATA soft-ware environment. In the current volume, attention is focused on how static and dynamic panel data models with “large (N) entities” observed over “small (T) time horizon” can be specified, estimated, tested, and find applications in real-world analyses. For the static panel data models, distinctions are made between ‘pooled panel data models’, ‘fixed effect models’ and ‘random effect models’. In a similar vein, issues relating to specification, estimation, tests and applications of dynamic panel data models are covered, leaving out the issues of non-stationarity that largely occur in “large N, and large T panels”. This is done because panels in developing economies are more characterized by the former than the latter