The problems associated with interpersonal comparisons are particularly intractable. This paper presents a procedure whereby the relative importance of any particular individual varies over the set of social states. In one sense, the stronger (relative to some norm) a person feels about any particular pairwise decision, the larger his say in that outcome. This procedure leads to a nested sequence of aggregate partial orderings which reflects this strength of preference. Under the assumptions presented it is also possible, given any two social states, to characterize the minimal amount of interpersonal comparison which is necessary in order to arrive at an aggregate ordering.

[A one-way expected utility representation has the expected utility of one probability measure greater than the expected utility of another probability measure whenever the first is preferred to the second. It requires preferences to be acyclic but not necessarily transitive, and does not require indifference to be transitive. Preference axioms which are sufficient for one-way expected utility for sets of simple probability measures have been presented before (see [8]). This paper uses additional axioms to extend the one-way representation to sets of discrete and more general probability measures.]

Generalizing the Sharpe-Lintner capital asset pricing model, Dieffenbach [4] presents a model of securities markets in a private enterprise economy in a multiperiod competitive equilibrium with uncertainty. Risk premiums on securities depend on the covariances of holding period returns with the return on the market portfolio and with a multiperiod cost-of-living index. This paper develops a quantitative theory of that relationship suitable for empirical estimation and testing. Whether the Arrow-Pratt relative risk aversion of a representative investor is greater or less than one is important in the theory; the empirical results for the United States suggest that this value exceeds one. A theoretical and empirical application of the theory to the term structure of United States Treasury securities concludes the paper. Mean observed returns are consistent with theoretical predictions for medium and long term securities, but the differences of mean observed returns among bills of different maturities exceed the theoretical predictions.

[Models of the behavior of populations of self-reproducible natural resources in an economic framework have rarely anticipated the consequences of different forms of production functions. This paper investigates sufficient conditions for extinction in a very general model as well as a model having a specific production function. In the second section additional considerations relating to extinction are deduced as well as the existence of a watershed level of population. These conclusions are exemplified using data from one particular population of red deer.]

Linear decision rules for controlling complex systems are often obtained by matrix inversion, but transform methods offer an alternative approach that yields insights into the structure of the decision problem of maximizing expected payoffs under constraints.