Angrist and Pischke are on a tear. They’re bringing econometrics to the masses with their new book, and the editors of the Journal of Economic Perspectives have seen fit to publish a debate around their article assessing the state of econometrics. A&P claim, and I more or less agree, that microeconometrics has undergone an inspiring “credibility revolution.”
The best summary I’ve found of their article is by Austin Frakt, here. Arnold Kling comments here. Andrew Gelman reviewed their textbook positively and constructively here.
Angrist’s website gave ungated links to most of the comments on his paper:
Michael Keane, Edward Leamer, Aviv Nevo and Michael Whinston, Christopher Sims, and James Stock
Added 6/3:
Permutations 
Thanks for the citation. Just a quick comment to say that I’ve also posted on Leamer’s and Nevo’s responses, and I’ll be posting a review of Angrist’s and Pischke’s book in a couple of weeks. Relevant URLs follow:
http://theincidentaleconomist.com/leamers-econtalk-interview/
http://theincidentaleconomist.com/io-vs-labor-economics/
I wouldn’t say Angrist and Pischke’s “Mostly Harmless Econometrics” is really suitable for “the masses”, at least if you include the average sociologist in those masses. I think it’s still far too technical for that…
Rense, the truth is I haven’t even looked at their book, just read positive reviews. Now that you mention it, Andrew Gelman did mention that it included proofs which probably wouldn’t be included in a text accessible to most sociologists. Another recent book that I haven’t looked at, that is arguably a substitute, is Morgan and Winship’s Counterfactuals and Causal Inference.
Any other suggestions Rense?
No, sorry… indeed it would be nice to see some more accessible econometrics texts.
Of course, there is a lot of overlap in the content of books on regression and econometrics. I recommend Raudenbush & Bryk’s Hierarchical Linear Models book, and Gelman and Hill’s Data Analysis Using Regression and Multilevel-Hierarchical Models. Both of them, but especially G&H, use as little math as one can use without completely compromising the student’s ability to understand the tools they are using.
I suppose that an accessible textbook written by an econometrician wouldn’t have as much material on multi-level modeling and would have more on things like instrumental variables. I’m sure they are out there, but for better or worse I only own graduate level econometrics textbooks.