Mostly Harmless Econometrics: An Empiricist's Companion

This is a great book that I really wish had been written when I was first learning econometrics. It is probably best appreciated by someone who has already taken one or two econometrics classes and has had some exposure to applied econometrics research. If you haven't taken econometrics yet you will not really appreciate the 'paradigm shift' that they are trying to explain nor appreciate some of the humor and storytelling. But if you are like myself someone who first learned econometrics more than a decade ago you'll find yourself reading this book page by page with appreciation. I feel that it is helping me to develop a much more mature understanding of the field of applied micro econometrics (helping me to confirm hunches and general notions I already and also new gems of insight). There are many cookbooks that might explain the methods in more depth (e.g. Cameron and Trivedi Microeconometrics) but that's not the purpose of this book. It's purpose is to make you think more about 'research design' which is to say about the questions that you pose, and how you pose them, rather than the methods that you use to try to arrive at answers. For too long econometricians got lost in the details of methods without taking a step back to think about some really fundamental questions first. The book is filled with examples of both good and bad research and you'll be surprised at how bad some past very influential research looks in light of modern day paradigms. It's not that these earlier researchers didn't know enough math, it's that they used the math without clear enough purpose.This book will make you a better economist and beyond that make you see the world around you slightly differently. You'll end up with a keener eye for all those natural experiments happening all around you. If you are an advanced undergraduate, MA or starting PhD student with any pretensions of becoming engaged in original applied economics research this book should be a "must have" on your bookshelf.

OK, let's be honest. There is nothing "harmless" about this book. Contrary to what the title suggests, it requires some serious background expertise in econometrics. So if you are looking for an introduction to econometrics, this is not your book. Instead, try Mastering Metrics by the same authors or Wooldridge's Introduction to Econometrics. If you are a complete beginner, Mastering Metrics might be the best choice.Keeping this in mind, Mostly Harmless Econometrics is an excellent resource for those who have some background in econometrics and are interested in applying their theoretical knowledge to practical problems. I read this as part of my Ph.D. program in economics, and it was incredibly helpful. To be clear, you don't need a Ph.D. in economics to understand this, but having taken a class in econometrics or statistics is highly recommended.

I'm a PhD student in finance, and this book is phenomenal. Easy to read through, or to use as a reference on concepts (Greene is where you should go for the rigorous proofs, etc.), I actually enjoy picking it up for class. I bought this for a seminar course, which will be my 3rd or 4th econometrics course, and I'm looking forward to reading this text for the class. If you're going into academia, this will be a lifelong companion, or so my Prof says; I assure you he's correct

Excellent companion book for those doing regression analysis. Understanding regression and its pros and cons are vital, and this book teaches that well. Very technical, but well-written. I use it for my PhD Econometrics class. It won't replace a course in regression or econometrics, but it's an excellent refresher or a good way to parallel the material if you're getting into the weeds of regression analysis.

The first thing I want to say is this: If you plan on doing regression analysis in your research, stop what you are doing, and read this book first. I think this book represents THE current statement on how we should use regression. For Angrist and Pischke, regression is a technology for summarizing data. If regression is to be used for causal inference, then there is nothing in the specification of the model or the choice of estimator that can ultimately make the causal story persuasive. That is, you don't identify causal effects simply by including "control" variables in your regression. The identification comes from elsewhere---either a real or "quasi" experiment---and the regression is what you use to clean up the imperfections of the experiment and measure effects. Angrist and Pischke have done an enormous service to social science by writing a regression textbook that nonetheless emphasizes the primacy of design. This is a terrific corrective for the "101 flavors of regression" approach of textbooks to date.Even with this emphasis on design, Angrist and Pischke show us that are a lot of nuances to the way that regressions measure such effects---e.g., in the presence of effect heterogeneity---and that's what this book explores in exquisite detail. It's a hugely important book and a very serious and rigorous treatment, despite it's apparently causal style. They make some claims that may strike some as outrageous---e.g., always using OLS, even for limited dependent variables---but the rigor of their presentation means that the onus is on those who disagree to think harder about why, exactly, they would prefer, say, a more parametric approach.Nonetheless, it isn't a "5 star" book. It often feels a bit rough-draft-like. The presentation of technical material skips important steps rather haphazardly. I wonder if this was due to bad editing? Hopefully there will be a second edition that cleans up these rough edges, in which case it would be the ideal textbook on regression analysis.