Network News

X My Profile
View More Activity

Tab dump

1) Myths of murder and multiple regression: A critique of the original Freakonomics.

2) If you build a coverage mandate, will people use it?

3) Behind the public option: The subsidies fight.

4) Damon Linker angrily misreads an old post of mine.

5) Peter Orszag takes aim at Fred Hiatt.

Recipe of the day: Chinese-style beets.

By Ezra Klein  |  October 26, 2009; 6:40 PM ET
Save & Share:  Send E-mail   Facebook   Twitter   Digg   Yahoo Buzz   StumbleUpon   Technorati   Google Buzz   Previous: An interview with Sen. Sherrod Brown: 'Reid listened to his senators'
Next: A morning with Jonathan Cohn


I think Peter Orszag's answers are fair and am sure Fred will respond to that.

My questions:
- Yes there are these 2 provisions (excise tax and independent commission); but the real doubt is those will not survive the remaining Congressional Process. We are looking for commitment from White House that President will stick for these real meaningful commitments and will NOT accept anything without those.

- Beyond that, Sen. Reid's attempt to add-on $250Billions via a separate bill to reimburse doctor fee shortfall - how are going to avoid any such things in future? Why did White House not ask Congress to include this cost in the comprehensive reforms which are under way? What can be done more, legislatively, so that any such short cuts are not taken in future? Is White House lobbying for any such additional warranties? Or (now it is our time to ask Peter), say it clearly that White House does not see this as a problem and simply accepts the sovereignty of the Congress in making up any such bills in future even though those are fiscally irresponsible. Peter, come clean on that.

Otherwise, quite an impressive rebuttal from White House.

Posted by: umesh409 | October 26, 2009 8:02 PM | Report abuse

That's not a critique of Freakonomics, it's a critique of quantitative social science.

Posted by: SimonCox | October 26, 2009 8:26 PM | Report abuse

Speaking of freakonomics, I got this month's issue of Gourmet in the mail this weekend, and it had a great gift subscription offer!

Posted by: bdballard | October 26, 2009 9:57 PM | Report abuse

No, it's a critique of a conception of quant. social that I'm unsure of. If you take regressions to be giving you causation then sure that's a critique. But I'm not sure why you're doing that in the first place. All a regression can really do is tell you how well your model predicts the data you have collected. This can then be suggestive for other data, but it cannot nail down causal inference. And, obviously, no model will be "complete" - you'll never get that, there is always the risk of omitted variable bias, etc. etc. If you good causal inference conduct an experiment (although that obviously has its own problems of validity, both internal and external). Maybe this is a timing thing but I'm somewhat shocked that this is actually a critique that holds water. I can distinctly remember my methods courses nailing this point. Then again I've learned this stuff much more recently, ergo after the Freedman critique offered in the article (and partially from a student of Freedman, so, well, there's that).

Posted by: y2josh_us | October 26, 2009 9:58 PM | Report abuse

The comments to this entry are closed.

RSS Feed
Subscribe to The Post

© 2010 The Washington Post Company