I have seen that many applied microeconomist have knee jerk reactions against structural models such as the ones applied in macro or i.o. estimating parameters to a fit of system non-linear equations to data seems dubious at best to these people.
Yet causal estimates from parameters to fit an assumed to be linear single equation models is some how convincing. I can kind of get this on the grounds taht at least they are making fewer assumptions.
What I for the life of me cannot figure out how they believe that causal estimate from parameter estimate coming from syntethical constructed control group ( a convex combination of data not effected by policy variable to some how resemble the area that your policy variable effects) is some how more believable.
