PEOPLE RESPOND TO INCENTIVES. Freakonomics author Steve Levitt was tonight's guest on Milt Rosenberg's radio seminar, and his concluding remark was a prediction that future economics research would feature a higher ratio of empirical to theoretical work? Why? Many of the useful series are now available on-line. Professor Levitt noted that what used to take several days to transcribe, using lots of copy cards and lots of keyboarding, now can be downloaded over lunch.
Specification, however, continues to require careful thought. Oh, and nonlinear estimators sometimes won't converge, and seminar presenters will continue to fret about "wrong signs" and t-statistics.