Knowledge that Transforms

Existing conditions for gainful trade under conditions of variable returns to scale are merely sufficient. Basing ourselves on the possible non-uniqueness of the production equilibrium under variable returns, we develop a globally valid necessary and sufficient condition for gainful trade. In particular, it is shown that trade may be gainful without changes in product prices, even if trade results in a loss of productivity.

This paper provides sufficient conditions for the existence of a unique Cournot equilibrium. Previous uniqueness results have depended on an assumption of non-degeneracy of equilibrium. As we illustrate, this assumption often fails in multi-stage games with proper Cournot subgames. Since our uniqueness results do not depend on this assumption, they are more widely applicable.

This paper derives, and then estimates, a model of employment where unions and firms bargain over wages and possibly employment, and efficiency wage considerations may be important. It illustrates the difficulties associated in interpreting many existing attempts to discriminate between alternative models. The results (based on over 200 U.K. firms) suggest that employment is negatively related to the firm's own wage and the change in the own wage relative to outside opportunities. The latter may be an efficiency wage effect. Various financial factors are also seen to have a significant effect on employment.

The paper attempts to explore whether lagged variables that help predict stock returns are merely proxying for mis-measured risk. Therefore, three different ways of measuring risk are employed (i.e. semi-parametric, GARCH and lagged squared returns). In an application to Japanese data, four key predictor variables are shown to have non-trivial additional forecasting power irrespective of how we measure risk. Interestingly, unlike the U.S., the level of the lagged dividend yield is not positively correlated with returns in either Japan or South Korea.

This paper concerns a Bayesian game model of the Buyer's Bid Double Auction, which is a procedure for organizing trade that selects a market-clearing price from a list of offers/bids. Strategic misrepresentation by traders may make the outcome of trade inefficient. Satterthwaite and Williams (1989) showed that misrepresentation and inefficiency quickly converge to zero as the number of traders on each side of the market increases. This is extended here to cases in which the number of buyers may differ from the number of sellers. The existence of equilibria in a generic instance of the model is also proven.

We study a model in which a customs union trades with countries which behave strategically. Provided that the members of the customs union are similar but not identical, we show that both in the case in which intra-union transfers are allowed as well as in the one in which they are not, one country may want to delegate authority for making union policy to its partner. The delegation decision depends on whether the policies used by union and non-union countries are strategic substitutes or complements, and on which union member is more “aggressive”.

The paper presents a theory of exploration for exhaustible resources. Exploration serves both to find new reserves and obtain information about potential reserves. Although any realized time path of the resource price fails to obey the Hotelling rule, the expected resource price does increase exponentially through time at the rate of discount. We obtain sharper results when the resource is assumed to be distributed randomly throughout the exploration region and follows a Poisson process with parameter λ. Consumption and exploration decisions are investigated for both the case where λ is known and the case where λ is unknown.

Contrary to the belief that capital over-accumulation is impossible in an economy with land—i.e. a non-reproducible factor of production, this paper shows that the possibility of dynamic inefficiency depends on the income share of land, and examines the historical tendencies of the income share of land in the U.S.

The Lucas Critique of Econometric Policy Evaluation argues that the parameters of econometric models are subject to theoretical cross-equation restrictions which follow from the fact that the endogenous variables of the models are chosen optimally by forward-looking agents. In this paper I argue that these facts alone are insufficient to generate such restrictions. I present an example of a model in which there exist multiple stationary rational expectations equilibria one of which is supported by a process-invariant forecast rule. Immunity to the Lucas Critique is proposed as a selection criterion in models with multiple equilibria.