Growth econometrics may benefit if it attempted to link new research with past research more thoroughly than it does at present, and establish new abstract generalisations for existing research. I have written before on GLE about how there are a handful of overarching theories which together encompass most of econometrics.
I am thinking particularly about testing and estimation of the model Y = X.B + error where Y is a vector of current growth rates by year, X is a transposed vector of lagged determinants, and B is a coefficient vector. X can include constants and lagged terms from Y. There have been recent developments on testing whether the B term is constant across countries, and estimation and comparison of country values when the terms are not constant.
Enough already. I shall return to this theme, when the time is right.
Subscribe to:
Post Comments (Atom)
No comments:
Post a Comment