Customer Reviews

Econometric Analysis (4th Edition)

5 star:		(9)
4 star:		(4)
3 star:		(3)
2 star:		(3)
1 star:		(0)

 

 

I bought this book in order to understand better my PhD courses. I must admit that after a few tries, I went back to Davidson and Mackinnon manual and to Hamilton's Time Series Bible. It could be seen as an intermediate level book, which is a main problem: undergraduates (except for the advanced ones) won't feel comfortable with the level of the book; applied econometrists will think perhaps that greene's book covers all the main topics but not profoundly enough (In the unit root chapter for example, the demonstration of the unit root test -Dickey-Fuller- is not present). As some of the other reviewers, I agree with the fact that, covering so many topics, the organization of the book could be improved. The mathematical tools explanation is a very clever idea; everybody lacking mathematical knowledge should read the first chapters. Some of my teachers appreciate (and I agree with them, without being an expert of that topic) particularly the panel regression chapter. I think this book fits better the graduate needs, especially those being, or wanting to be applied econometricians. If you are looking for an advanced theoretical manual, Greene's is not the best choice; if you are interested in having a -slightly more than- fairly good understanding of what is done currently in the field, then grab this book. But be aware, if you want to feel comfortably when reading it, you should be familiarized with the field. I recommend Johnston and Dinardo's Manual and Walter Ender's Manual (for time series only). Finally, it must be said that the fourth edition doesn't differ too much from the third one.

I've just bought the 4th edition. At my first glance it differs not much from the 3rd one, except: [1] a CD-ROM is attached; [2] typos have been corrected; [3] the WHOLE content is blue-printed instead of dark-printed, which I find uncomfortable (not used to) in reading it.

The rest is similar to its 3rd edition -- poor in organization, as I mentioned before. In my professor's and my views, its 2nd edition is much better.

I own many econometrics books and this is by and large the worst I had (I sold it at once after the course I HAD to buy it for, while I kept all others). It seems it covers everything while, in fact, almost all results are just presented rather than analyzed, let alone analyzed in depth. The best thing about the book is probably the table of contents, as one gets an overall idea of what econometrics deals with today. In fact, no other book aims at covering so much. Otherwise, it looks, feels and performs like an undergraduate text ...

On top of that, this edition in particular is especially bad: the blue ink stains your hands, and there are MORE typos than in the previous one. If you absolutely want to get it, buy the third edition (if you find one used).

If you want more serious general econometrics, have a look at Davidson and MacKinnon's book. If you want asymptotic theory, get White's. For time series, Hamilton's or Brockwell and Davis'.

I agree that this is not the best econometrics textbook out there, though for a long time it seemed to be the only complete one. Fortunately, Ruud has recently put out a better textbook.
Although this text is pretty complete for an intermediate level graduate text, it does not make a very good reference. Its index is incomplete and difficult to use. Even worse, Greene has a nasty habit of referring to formulae, definitions and proofs from previous section without even giving a page number in order to find the reference. Additionally I do not like the way example 10.8 might be right after line 10.35 or whether or not something has a number seems to be haphazard, which makes it very very difficult to find anything referenced from previous chapters. Finally, Greene does not use standard mathematical notation, so one has to figure out what his notation (he never redefines anything) his while trying to find something. As a reference, I am much much happier with Ruud _An introduction to classical theory_ and wish it had been out when I first took graduate econometrics.

Greene's book provides a good starting point for graduate-level econometrics. The book is really comprehensive, including all major subjects of modern econometrics. The only problem I have is most references in the book are a bit out-of-date. Since econometric techniques evolve so fast, the author should include more recent journal articles instead of those written in 1970s or 1980s. Most examples in the book are also very old (1930s, 40s, etc.) In addition, I would prefer more applications in finance instead of boring consumption/production/labour economy stuff, but others may not agree. The last point is just my personal preference.

I am already through the education process (PhD in hand :) and have a Masters in statistics, so my primary use for this book is as a reference work to regression models, broadly defined. It serves this purpose quite well and I've gotten a fair bit of mileage out of it. Even in a thorough statistics education (like mine, I guess), there are inevitably topics that the instructor didn't have time for, didn't explain well, you didn't get the first time, or you simply haven't used since then, and Greene's book does an admirable job of filling these cracks. Of course, it's an *econometrics* textbook, so (a) it definitely is written from that perspective and (b) there is little or no treatment of topics such as multivariate analysis, experimental design, or nonparametrics that economists don't seem to make much use of. There's a lot more out there than regression models--readers should not deceive themselves that if they know everything in Greene they're "done." This is particularly true for students in disciplines other than economics.

I imagine this book would be rather tough going for a novice, though, which is why I dock it a star. I think it would be an great second book to own (but own Kennedy's book A Guide to Econometrics too) but I would use another text for an intro class.

The coverage of this book is vast and comprehensive. It includes almost all of the important topics that practitioners and graduate students need to know.

Because of the scope of this books, I think it is better used as a reference book rather than a text. To save space, many major and crucial results are only stated without proofs (This drives serious econometrics students crazy). Also, equations pop-out from nowhere sometimes and I have to scratch my head like crazy to figure them out. For instance, the material in Chapter 14 Random Effect section is poorly explained.

In addition, real-data examples are of no help. The author just throws in a lot of numbers to make you wondering around how he got those numbers.

I believe there is a better text out there. Consult your professors before you buy this book. It is not cheap at all. But if you need a good reference on your book shelves, this is the right one.

I took the Ph.D. core econometrics course some years ago using Greene's Econometric Analysis. The book at that time was the first edition. It was an excellent book which had already become a dominant work in the field. Now, with the fourth edition, the scope of the book is enormously expanded. Upon purchasing the fourth edition recently, I was amazed at the coverage; it seems doubtful that any other book could offer the reader such a comprehensive presentation of topics of current interest in econometrics. Virtually everything the reader could want is in this book presented in clear and concise language. I recommend it wholeheartedly.

I learned my econometics from that textbook. It is not an easy read, however, it is an effort well invested. It is an excellent reference for up-to-date econometric techniques (If you wish rigor, you might read directly Econometrica.) I believe the author did a great service to the econometric community (including all those using it even outside the realms of Economics) by collecting and presenting systematically the tools of the modern econometric theory. It comes with the Limdep software, which I find very useful, especially for panel data analysis.

This is an indispensable book for applied econometrics students. It has lots of applied and a reasonable degree of theoretical stuff. What is missing in many advanced econometrics textbooks, including the popular books Davidson and MacKinnon (1993) and Hamilton (1994), is some data sets and some results based on them so you can learn econometrics by actually doing it, not by simply having to trust the author's results. I mean, after all, what is the point of putting actual data examples in a book if the reader cannot replicate them because he does not have the data? In the preface to his important 1971 textbook Principles of Econometrics Theil remarks that "...nobody should believe that he will be able to handle statistical data in economics without touching actual data." I cannot agree more.
I use Ox and Gauss for my econometric analyses and write lots of codes. The very first thing I need is a clear explanation of the theory and some results to check the accuracy of my codes. In those respects Greene's book is the champion. You have the theory, the data sets and the results. Replicate them and learn. I believe this is how you can learn applied econometrics.
Besides Greene's book, I could also recommend "Patterson (2000) An Introduction to Applied Econometrics" for applied time series. You can get the data sets used in this book by contacting the author.
Finally, answering some critics. First, I do not think that evaluations of this book based on some course experiences are unbiased. If you use a book to pass a Ph.D. course its usefulness depends very much on the structure of the course. And some books are better suited to the course than others. Actually, no book can beat a good set of lecture notes. Second, if you think you are a serious econometrics student, go to the original papers for proofs and the sources of the mathematical equations. If you look for them in textbooks you are in the wrong place. Third, the author does not just throw in a lot of numbers. He gives you the theory and the results. The rest is your job. If you cannot figure out how he got those numbers, get a tutor.