Introduction; Towards Observed Preferences in Econometric Decision Models

Abstract

This paper contains some of the remarks I made during the opening of the conference on econometric decision models at the University of Hagen, June 19–20, 1981. Its main message is concerned with a variation of a theme by W. Leontief (1971), namely: In econometric decision models, use observed preferences of the decision maker instead of theoretically assumed (hypothetical, “plausible”) preferences.

In recent developments in theory and methods, there are two main lines along which this can roughly be accomplished:

• a) Some methods are directed towards numerically determining the parameters of the scalar-valued preference function of econometric decision models from data (time series observations, cross section observations from surveys, interviews etc.). Papers of problem area 2 deal with this line of research.

• b) Interactive methods of vector optimization avoid the necessity of explicitly specifying a scalar-valued preference function. They are more or less well adapted to the needs of decision makers working with econometric equation systems. For practical decision making, this approach has important advantages in comparison with control theoretical decision models in which an explicitly specified scalar-valued preference function is required. Papers of problem area 3 deal with this line of research.