>>12956437Hmm breaking into my old econometrics textbooks its says that most estimators encountered in statistics and econometrics can be written as functions of sample averages, in which case we can apply LLN and CLT.
Maybe the rub here is that its not a theoretic constraint, but a practical one. Economics especially has alot of missing, unrecorded and incomplete data sets, perhaps this case isn't about gaining statistical power by approach, but just the reality that we don't have the whole population data set, just sample fragments.
