Scenario:

3 people are working on a paper:

1.) A Professor acting as the PI. Not really engaged in writing the paper; he/she just wrote the grant. Will take a look a drafts of the paper and make comments.

2.) A grad student who is writing the paper and determining the analytic strategy. They will take a look at the log files, and say how to construct tables.

3.) A research analyst who is cleaning the data, and writing the do file. They are doing most of the analytic work, but don't have a strong background in econometrics.

I've seen this pattern repeated in a couple of different places, and it always strikes me as being really inefficient. You basically have 2 3 write the whole paper, then 1 makes some off hand comment ("Oh, re-do everything and add a fixed effect for y.") that might entail rewriting all of the code, or finding a new set of data. 1 2 might not even understand what the data looks like.

Ideally, you'd have 1 come in a bit earlier - maybe look at the do files, or the actual data. But 1 usually has no interest in doing this sort of grunt work (and pays 2 3's salaries).

Anyone else use a different set u[p? How does it work for you?