Research design in total institutions: problems,pitfalls, and possible solutions

Research in total institutions is fraught with problems and pitfalls rarely discussed in methodology and research design textbooks but regularly encountered in the literature. Five of the major relevant problems in research designs for this area are considered, examples of the pitfalls are given, and suggestions for minimizing or overcoming them are offered.     

Gross substitutability of point-to-set correspondences

In this paper the notion of gross substitutability for the multi-valued case is studied. It is proved that, if in a pure exchange equilibrium model gross substitutability and some auxiliary conditions prevail, then (a) the set of equilibria is a Cartesian product of a convex set of equilibrium resource allocations and a convex cone of equilibrium prices; hence all equilibria are equiadvantageous for every trader; (b) the weak axiom of revealed preference holds in any equilibrium; (c) any equilibrium is stable with respect to reallocations of initial resources. Some situations in which Walras' law does not hold are considered as well.     

Business climate as a leading indicator? An empirical investigation for west Germany from 1978 to 1990

For twenty-five years now, the leading indicator “Business Climate” of the Ifo Institute has been applied in the Federal Republic of Germany. The intent of this note is to provide answers to two questions by means of econometric methods. The first question is: Is the business climate really a leading indicator? The second question is: How much time is the indicator ahead? For verification and in order to respond to both questions, the Granger causality test is used in the version of Hsiao.     

The Valuation of Assets under Moral Hazard

The design of managerial incentive contracts is examined in a setting in which economic agents are risk averse, and the actions of managers can affect asset returns which contain both systematic and idiosyncratic risks. It is shown that in the absence of moral hazard, owners of assets will insure managers against idiosyncratic risks, but with moral hazard, contracts will depend on both systematic and idiosyncratic risks. The traditional recommendation of asset pricing models, namely, to focus only on systematic risks, is thus proved to be valid only when there is no moral hazard. The major empirically testable predictions of the model are (1) managerial incentive contracts will generally depend on systematic as well as idiosyncratic risks, (2) idiosyncratic risks will generally be important in investment decisions, (3) the managers of firms with relatively high levels of idiosyncratic risks will have compensations that are less dependent on their firms' excess returns, and (4) the compensations of managers of larger firms will be relatively more dependent on the excess returns of their firms.     

Unbiased product estimators

Summary The paper presents a comparative study of product estimators proposed byRobson [1957] andMurthy [1964]. It is seen that the Robson's estimator gives a better performance.