To make high-quality research more accessible and easier to explore.

Fields:
10 results ✕ Clear filters

Fisher's Tests Revisited

Econometrica 1976 44(2), 247
[This paper is concerned with Fisher's tests for index numbers. In particular, uniqueness and inconsistency theorems are proved. Beyond that, Fisher's system of tests is weakened considerably. Without any regularity assumption (such as differentiability or continuity) it is shown that every subset of the system of weakened tests is consistent while the whole system is inconsistent. The question of how far the whole system must be weakened to obtain a consistent set of tests is also considered.]

Risk Aversion in the Small and in the Large

Econometrica 1976 44(2), 420
This paper concerns utility functions for money. A measure of risk aversion in the small, the risk premium or insurance premium for an arbitrary risk, and a natural concept of decreasing risk aversion are discussed and related to one another. Risks are also considered as a proportion of total assets.

Turnpike Theory

Econometrica 1976 44(5), 841
[Support prices are derived for weakly maximal paths in an optimal growth model which is time dependent but without uncertainty. The notion of "reachable" stocks and paths is defined and used to derive turnpike theorems by the value loss method. The proofs do not depend on the presence of optimal balanced paths nor on the usual transversality conditions. The theorems are extended to the classical model which has a non-trivial von Neumann facet.]

A Problem on Rankings by Committees

Econometrica 1976 44(2), 241
[We consider how large a committee must be before it is possible to achieve simultaneously any two rankings of m alternatives by two seemingly consistent procedures.]

Capital Aggregation in a General Equilibrium Model of Production

Econometrica 1976 44(6), 1179
[In this paper, we focus on capital aggregation in a general equilibrium model of production. Various potential aggregates involving intrasectoral and intersectoral, as well as full aggregation are discussed in connection with the various aggregation procedures. It will be shown that the satisfaction of the Gorman conditions allows for full aggregation within a general equilibrium model of production. We shall derive new conditions for aggregation using a composite commodity approach that appears to be somewhat weaker than the conditions associated with restrictions-on-functional-form theorems. Our main conditions relate to the equality of sectoral labor shares. The data for testing those conditions appear to be readily available. It is shown that the equal labor share condition can be applied to models with joint and nonjoint products. In addition, the conditions for aggregation are derived for a model with many primary inputs and also for a model with unequal rates of depreciation. Two sections are devoted to the main correspondences between certain aggregation procedures in the literature from the point of view of a general equilibrium model. The implications of our analysis for the form of the unit cost function and of the aggregate production function are discussed. In particular, if our aggregation condition holds, then the aggregate production function can be Cobb-Douglas, if one of the sectoral forms is also Cobb-Douglas, irrespective of the forms of the other sectoral production functions.]

Some Experimental Results on the Statistical Properties of Least Squares Estimates in Control Problems

Econometrica 1976 44(6), 1289
The statistical properties of the certainty equivalence control rule and of the least squares estimates generated by this rule are examined experimentally in a linear model with two unknown parameters. It is found that the least squares certainty equivalence rule converges to its true value with probability one and is asymptotically efficient, having an asymptotic distribution with a variance as small as any other strongly consistent rule. However, while a linear combination of the parameter estimates is consistent, the evidence does not confirm that the individual estimates themselves are consistent. If these converge to their true values at all, they do so very slowly (on the order of (log t)').