Chance Constrained Programming Formulations for Stochastic Characterizations of Efficiency and Dominance in DEA

Pareto-Koopmans efficiency in Data Envelopment Analysis (DEA) is extended to stochastic inputs and outputs via probabilistic input-output vector comparisons in a given empirical production (possibility) set. In contrast to other approaches which have used Chance Constrained Programming formulations in DEA, the emphasis here is on joint chance constraints. An assumption of arbitrary but known probability distributions leads to the P-Model of chance constrained programming. A necessary condition for a DMU to be stochastically efficient and a sufficient condition for a DMU to be non-stochastically efficient are provided. Deterministic equivalents using the zero order decision rules of chance constrained programming and multivariate normal distributions take the form of an extended version of the additive model of DEA. Contacts are also maintained with all of the other presently available deterministic DEA models in the form of easily identified extensions which can be used to formalize the treatment of efficiency when stochastic elements are present.     

Conditional mean functions of non-linear models of US output

We compare a number of models of post War US output growth in terms of the degree and pattern of non-linearity they impart to the conditional mean, where we condition on either the previous period's growth rate, or the previous two periods' growth rates. The conditional means are estimated non-parametrically using a nearest-neighbour technique on data simulated from the models. In this way, we condense the complex, dynamic, responses that may be present in to graphical displays of the implied conditional mean. First version received: Feb. 1999/Final version received: June 2001     

Skill-Specific rather than General Education: A Reason for US–Europe Growth Differences?

In this paper, we develop a model of technology adoption and economic growth in which households optimally obtain either a concept-based, general education or a skill-specific, vocational education. General education is costly to obtain, but enables workers to operate new production technologies. Firms weigh the cost of adopting and operating new technologies against increased profits and optimally choose the level of adoption. We show that an economy whose policies favor vocational education will grow slower in equilibrium than one that favors general education. More importantly, the gap between their growth rates will increase with the growth rate of available technology. By characterizing the optimal Ramsey education policy we also demonstrate that the optimal subsidy for general education increases with the growth rate of available technology. Our theory suggests that European education policies that favored specialized, vocational education might have worked well, both in terms of growth rates and welfare, during the 1960s and 1970s when available technologies changed slowly. However, in the information age of the 1980s and 1990s when new technologies emerged at a more rapid pace, they might have contributed to an increased growth gap relative to the United States.     

A child health production function estimated from longitudinal data. Cebu Study Team

Collaborating researchers used a multi equation model to analyze 3080 mother-infant pairs living on the island of Cebu in the central Philippines and to estimate a child health production function. The econometric methods used eliminated obstacles such as heterogeneity and endogeneity of significant explanatory factors. They also maximized the longitudinal quality of the data. The results showed that morbidity in 1 period reduced infant weight in following periods. For example, diarrhea and febrile respiratory infection reduced growth in the time period following the illnesses. Further effects of some contributing factors were great near birth not diminished with age. For example, breast feeding promoted growth and protected against infection, especially diarrhea and febrile respiratory infection, more substantially near birth than it did later. Further, in urban areas, preventive health care positively affected growth independently of its indirect influence on reducing respiratory infection. Yet it had little effect on diarrhea. Diarrhea had a strong adverse effect on growth. Exposure of the infant to enteric pathogens had the greatest influence on production of diarrhea. In urban areas, exposure included water supply contamination, exposure to feces, a more pathogenic diet, and community crowding. In rural areas, however, exposure included community crowding and increased rainfall which presumably washed feces into the water supply. Household crowding in both urban and rural areas and irritation from smoke in urban areas only tended to bring about febrile respiratory infection.     

Agency Structure and Firm Culture: OSHA, EPA, and the Steel Industry

We compare models of Occupational Safety and Health Administration(OSHA) and Environmental Protection Agency (EPA) enforcementand compliance for steel plants during the 1980s. We find thatOSHA and EPA respond similarly to plant-level compliance andmeasures of hazardousness, but differently to firm-level complianceand risks of plant closing, and we relate the differences tothe agencies' differing organizational structures. Plant-levelcompliance is affected by enforcement pressure, compliance costs,and the firm's overall compliance behavior in similar ways forthe two regulatory areas, but environmental compliance was alsosensitive to plant size and risk of closing. Finally, we findthat the likelihood that a plant was in compliance with oneagency seemed at most weakly related to whether it was in compliancewith the other, but that plants likely to receive enforcementattention from one agency were also more likely to receive enforcementattention from the other agency.     

Does seeing more deeply into a game increase one’s chances of winning?

The substantively rational value of the games studied in this paper does not help predict subject performance in the experiment at all. An accurate model must account for the cognitive ability of the people playing the game. This paper investigates whether the variation in measured rationality bounds is correlated with the probability of winning when playing against another person in games that exceed both players’ estimated rationality bound. Does seeing deeper into a game matter when neither player can see to the end of the game? Subjects with higher measured bounds win 63 percent of the time and the larger the difference the more frequently they win.