Customer Reviews

Analysis of Panel Data (Econometric Society Monographs)

5 star:		(5)
4 star:		(2)
3 star:		(0)
2 star:		(0)
1 star:		(0)

 

 

Cheng Hsiao's 1986 monograph "Analysis of Panel Data" drew the attention of many to the inherent advantages of using longitudinal (ie. panel) data sets, where the same subjects are observed at different periods in time. Substantially as a result of Hsiao's excellent presentation of key issues, the volume of research using panel data increased dramatically during the next fifteen years.

It is especially timely that Hsiao has written an updated version of this classic. All the essential concepts from the first edition have been retained; in addition, the author does a very good job of providing coverage of the many advances which have been made in analysis of panel data since 1986.

Of special note, in relation to the second edition, are the chapters on dynamic models (Ch. 4) and discrete response models (Ch. 7). These two areas have been particularly in need of an updated treatment, since a great deal of recent research has taken place into the properties of estimators in these situations. Hsiao's treatments of these are correct up to and including 2001, which provides a major service for researchers interested in these topics.

An alternative book, by Badi Baltagi, exists. However, Hsiao's book is superior and should be purchased by all who wish to pursue modern empirical work.

This book covers panel data econometrics theory very well. It's probably the best in this field. However, this book lacks data and more importantly computer codes on how to implement panel data models. The author didn't provide a website where you can get the data in the book, which makes it difficult for applied econometricians to learn and implement. It would a lot more useful if the author can provide some support on how to implement his models in either packages (e.g., Stata or SAS) or programming languages (e.g., Gauss or Matlab). I found biometricians and social scientists do a much better job in both research and implementation of panel data models than econometricians, although they have different names for the same thing (e.g., mixed effects, multilevel, or HLM, etc.)

This book is a classic on panel data econometrics. Nevertheless, the materials in this book do not account for nonstationary panel data. I am anxiously waiting for Arellano's Panel Data Econometrics, which contains up-to-date treatment.

If you need some kind of Panel data theory for your reference, this one is quite essential. The only cons is it lacks some empirical studies to increase the idea.

This neat monograph is necessary for empirical research in panel data analysis. In particular, discussions about fixed effects,random effects, discrete data and dynamic models have been very helpful for my empirical analysis.

Good Book.

Important Academic treatment on longitudinal panel data. It is really a shame that longitudinal panels and longitudinal analysis is not used more often in market research. It is so much more powerful in many ways than dynamic panels, though there is obviously a need for both. My complaint would be that there wasn't much on static panels, and of course the copy write date, and thus lack of information on online panel data, either static or dynamic. So much of the what has been written on marketing research really needs to be updated now that almost everything is being done Online.

If you are looking for something current that deals with Online Panel data you may want to check out Andrei Postoaca's (IPSOS Interactive Services CEO), The Anonymous Elect: Market Research Through Online Access Panels which was just released by Springer.


Tom Anderson
Anderson Analytics, LLC

This neat monograph is necessary for any research in panel data analysis.