Doing macroeconometrics

I just discovered – courtesy of the great man himself – David Hendry’s free textbook Introductory Econometrics: A New Approach on doing time series econometrics on macro data. I’ve not yet read the book, still less tried the exercises (which require OxMetrics of PcGive), but thoroughly approve for two reasons. First, it’s free – based on his lectures to 2nd year PPE undergraduates at Oxford, and assuming just a little statistical knowledge.

Secondly, as the introduction puts it: “[M]uch of the huge variation over long time periods in many aggregate variables does not fall under the purview of economic analysis, but is due to extraneous forces such as wars, changes in legislation, shifts in social mores, and technological, medical and financial innovations, which in turn are only partly affected by
economics. The current vogue for seeking micro-foundations for such variables in terms of a ‘representative’ agent who is simultaneously, employed, unemployed, growing up and retired, rich and poor, etc. sits uneasily with the historical evidence.”

Yep. Even the Bank of England is tiptoeing away from ‘micro-foundations’, it seems. But there are plenty of diehards still adhering to these models, and, worse, teaching them.