Which good econometrics textbooks would you recommend?
Edit: there are quite a few books out there, with varying levels of mathematical sophistication. It would be good to get some idea of how technical the book you're recommending is.
Definitively Econometric Analysis, by Greene. I'm not an econometrician, but I found this book very useful and well written.
Depends on what level you're after. At a postgraduate level, the one i've most often seen referenced and recommended, and have therefore looked at most myself, is:
Wooldridge, Jeffrey M. Econometric Analysis of Cross Section and Panel Data. MIT Press, 2001. ISBN 9780262232197
Most of what little I know about econometrics I learnt from this book. 776 pages without a single graph.
"Mostly Harmless Econometrics: An Empiricist's Companion" (Angrist, Pischke 2008) is a less technical and entertaining summary of the field. I wouldn't describe it as a beginner book, but it's well worth reading once you understand the basics.
It depends on what you really want, (GMM, time series, panel...) but i can recommand those two book : Fumio Hayashi's Econometric and Davidson and McKinnon Econometric Theory and methods and For a course in econometric time series, Hamilton's Time Serie Analysis is great
I would definitely recommend M. Verbeek's A Guide to Modern Econometrics. Woolwridge is too worldly (and this long-windedness loses the reader's focus too early in the chapters). Greene (i'm referring to the 5th edition) often gets lost in minutiae: i.e. strives to catalog formulae that are orthogonal to the main subject of the chapter (good for references, but again, not ideal for learning).
i've not read the Hayashi (thought i suspect it's a bit outdated now). Hamilton, is really focused on...TSA so it's a bit off the mark for general econometrics.
I really like Kennedy's A Guide to Econometrics, which is unusual in its setup, since every topic is discussed on three different levels, first in a non-technical way, then going into details of application and finally going into theoretical details, although the theoretical parts are a bit superficial.
I am an econometrics lecturer. Definitely, the best book depends on what you want and the level that is suitable for you. However, my first option is "Basic Econometrics" written by Gujarati. The fourth edition of that textbook provides a good and well-written overview of the subject (Gujarati, 2002). Sadly, I cannot say the same regarding the fifth one (Gujarati and Porter, 2008). I use that textbook in undergraduate and postgraduate courses. For the undergraduate ones, I use it as the main textbook. For the postgraduate ones, I use it in addition to "Econometric Analysis" by Greene (2012). I believe necessary to do this in order to develop the econometric intuition behind the formulae. Econometrics involves intuition, art and technique. Currently, no book provides them in an adequate nor equilibrated manner.
One at a somewhat lower level of mathematical sophistication than Wooldridge (less dense, more pictures), but a bit more up to date on some of the fast-moving areas:
Murray, Michael P. Econometrics: A Modern Introduction. Addison Wesley, 2006. 976 pp. ISBN 9780321113610
Seems that it's not available for preview on the web and the publisher is out of stock, but you can view pdfs of 11 web extensions to get an idea of its style.
(Disclaimer: I'm not an economist.) I gather you might like to have a range of possibilities listed, however, most of the answers focus on more advanced texts. Should someone want a very introductory text, I can recommend:
Gujarati, D., & Porter, D. (2008). Basic Econometrics. McGraw-Hill/Irwin.
This is very basic (i.e., little math), very comprehensive (~1k pages--there's a ~30p chapter on every conceivable topic that can be read very quickly), and very clear. Moreover, huge numbers of these seem to have been printed and used in classes a few years back, so you can pick up an old copy for a few bucks in a used book store in any college town.
I like Cameron and Trivedi's Microeconometrics. It strikes a nice balance between breadth, intuition, and rigor (if you follow up on the references). The target audience is the applied researcher. Their Microeconometrics Using Stata is also quite good if you're a Stata user, though it covers less ground. At advanced undergrad or MA level, I love Johnston and DiNardo's Econometric Methods. I hope there's a new edition soon. Kennedy's book is great for intuition if any of the above are tough going.
I prefer the fourth edition of "Basic Econometrics", among other reasons, because the text is completely self-contained. The fifth edition requires to have access to the web in order to replicate the exercises contained in the text ( Users of the previous edition did not have this problem because the book was packed with an additional cd). This change, that seems a simply technological update, extremely reduces the possibility to replicate the textbook exercises for third-world students (like the ones I teach). Indeed, in my university exists the proposal to eliminate "Basic Econometrics" as textbook due to this reason among others (an additional one refers to the difficulties that local lecturers have to obtain access to the Online Learning Center).
Another problem of the fifth edition relates to the the writing style. According to the authors, one "improvement" of the new edition was the simplification of the analyses included in several chapters (see preface). However, I do not agree with them: The explanations contained in previous editions definitively are better than the ones contained in the fifth one. Evidently, I am aware that it can be argued that this situation may be the result of translation problems (I use the Spanish version of the book). However, users of the fifth edition in English have pointed out the same situation (see comments in amazon.com).
Despite all these problems, I believe that "Basic Econometrics" is the best undergraduate textbook in the market. Even the fifth edition is a good one. Hopefully the next edition may be as good as the fourth one.
Hashem Pesaran's book looks very promising. It covers such topics as dependencies in panel data and others that I haven't seen in other books.