Gerhard Tintner, Recent Econometric Studies in Agricultural Marketing: Discussion (1955)

Tintner’s journal discussion assesses three papers through their statistical assumptions, theoretical connections, and treatment of technological choice. His brief contribution welcomes modern econometric methods while distinguishing economic usefulness from statistical reliability.

He praises Suits’s application of econometrics to concrete data, but questions the practice of testing several models before selecting the most adequate:

I think it should be emphasized that such a procedure, however commendable from the economic point of view, introduces statistical difficulties because biases arise.

This qualification bears directly on the reported standard errors: choosing a model through preliminary comparisons complicates the interpretation of its statistical results. Tintner also suggests, cautiously, that the model’s moderate forecasting success may reflect linearity assumptions unlikely to hold across the entire range of the estimated functions.

Turning to Tolley and Harrell, he proposes connecting their decision model with economic models advanced by Shackle and Kalecki and with similar work by statisticians Blackwell and Girshick, and Savage. The emphasis shifts from criticism to situating agricultural decision analysis within a broader theoretical conversation.

Finally, Tintner links Black’s methods to linear programming. Both move analysis into the technological domain and address selection among alternative technologies:

This is a great progress compared with traditional methods of the economic analysis of production.

His concluding endorsement identifies technological choice as an important extension of production analysis. Across the three comments, methodological progress requires attention both to statistical limits and to the range of decisions a model can represent.