It is becomino recocnized in the United States that nutrition in the broadest sense, that is, all substances that are ingested, is a very important factor in health, perhaps the most important over which mia,n has atny control. To some extent this insight results from observation of relations between geochemistrv and health.' For example, the relationship between iodine deficiency and enildemiic goiter has been recognized for many years. Similarly the relationislhips between iron deficiency and anemiiia and betweeni flouride deficiency and clental caries are well known. Others include zinlc deficiency and dwarfism, and lithium deficiency and mental disorders. As such relationships become better understood, large opportunities to improve health through preventive measures become available. It is most unfortunate that medical research has devoted itself almost entirely to cures rather than understanding the etiology of disease. Even nmore urgent perhaps than understanding how natural geochemlistry influences health and disease is finding out how human interventions in the chemistry of the environment ar-e related to disease. T here are many such effects possible and a number have occurr-ed on a relatively large scale. Dramatic instances of heavy mletal poisoning have happened, mlost notably in Japan. A number of combustion products discharged to the atmosphere from stationary anid mlobile sources are implicated in luing disease. But the suspected relationshlips between environmental contamination and cancer are currently receiving the most attention because in many ways that is our most important disease. Cancer incidence has been rising at an average rate of approximately 1 percent a year for somne time. Heart disease is still the largest killer, accounting for about 40 percent of total deaths in 1975 according to the National Center for Health Statistics. But the death rate from heart disease went down while cancer deaths, about 20 percent of the total, went up more than 2 percent in 1975. Contrary to the impression which might be gained from some of the more dramatic stories in the press, the long term rise has been dominated by only one form of this disease, lung cancer. Since 1930 lung canicer has been rising steadily for both men and women but much more rapidly for the former. The recent upturn in lung cancer rates for women reflects the delayed (relative to men) popularity of cigarette snmoking amiiong women. Thus, the increase in female smoking in the 1930's and 1940's may shortly cause a dramatic increase in female lung canicer rates. Stomach cancer has been trending pretty steadily clownward, although quite recently cancers of the digestive tract have shown an increase and the others have been constant. The increase in lung cancer appears, as mentioned, to be strongly related to smoking. Lest one feel complacent about the situation, however (all we have to do is stop smoking and anyway smokinig is voluntary, one might say), one should recall three sets of facts. First, the cancer rate is high in the United States, accounting for a large portion of mortality, and death from cancer is often especially painful. Second, *Prl)fcssor of Econiomics and Associate Professor of Econoimiics. respectively The Univei-sity of New Mexico. 'For a review ot the evidence, see Geochemnisatrs atid th, Enovironmenot, I, National Academy of Sciences 1974.

The volume of new issuances in secondary loan markets fluctuates over time and falls when collateral values fall. We develop a model with adverse selection and reputation that is consistent with such fluctuations. Adverse selection ensures that the volume of trade falls when collateral values fall. Without reputation, the equilibrium has separation, adverse selection is quickly resolved, and trade volume is independent of collateral value. With reputation, the equilibrium has pooling and adverse selection persists over time. The equilibrium is efficient unless collateral values are low and originators' reputational levels are low. We describe policies that can implement efficient outcomes.

Few health policy issues have been both subjects of more consensus and yet more debate than that of expanding insurance coverage of the uninsured. On the one hand, surveys show a majority of the public identifies the absence of insurance coverage for fifteen

The double auction trading institution (DA) has been highly efficient across diverse marginal-cost market structures, whether human subjects or "zero-intelligence" robots populated those markets. Accordingly, many researchers suspect that DA performance transcends market structure and agent strategy. But we show that (1) large avoidable costs undermine the efficiency and stability of human subject DAs, and (2) these low human efficiencies are simultaneously well above zero-intelligence efficiencies. Our results dramatically illustrate the potential havoc wrought by highly competitive institutions when they must cope with nonconvex technologies.