THE SILENCE OF THE LAMBDAS - A TEST OF THE ALMOST IDEAL AND ROTTERDAM MODELS

During the past decade, the Linear Approximate (LA) Almost Ideal Demand System and the Rotterdam model have been adopted by agricultural economists as the demand systems of choice in most applications. The apparent explanation is that the two models are both (second-order) locally flexible and compatible with demand theory, they have identical data requirements and are equally parsimonious with respect to parameters, and both are linear in the parameters. While the two models are thus equally attractive in most respects, and indeed appear very similar in structure, they lead to different results in some applications. This article develops a test of each against the other. In an illustrative application to U.S. meat demand, the Almost Ideal model is rejected while the Rotterdam model is not.