Discussion: Empirical Analysis of Agricultural Commodity Prices: A Viewpoint

I congratulate Professors Tomek and Myers on a fine review of "structural" and "time series" methods as applied to the econometric analysis of agricultural commodity prices. Further, I second their list of recommendations for improved practice in the field of price analysis. My discussion covers two points. First, I discuss their paper from the perspective of Kuhn's model of "Scientific Revolutions". Perhaps the two major modeling efforts described in their paper can be understood as alternative paradigms. Viewing the subject from this perspective may help us see the future of applied price analysis. Second, I comment on their use of the word structure. I argue that it will be difficult for us to obtain structure from either Tomek and Myers' "structural models" or their "time series" models. This may show up in our work as fragile parameter estimates and is due to the "omitted variables" problem. While I offer no solution to the problem, I make one suggestion which may help in some settings