Comparing two countries in which the poor have equal real incomes, the one in which the rich are wealthier is likely to have a higher infant mortality rate. This anomalous result does not appear to spring from measurement error in estimating the income of the poor, and the association between high infant mortality and income inequality is still present after controlling for other factors such as education, medical personnel, and fertility. The positive association of infant mortality and the income of the rich suggests that measured real incomes may be a poor measure of social welfare. Countries with unequal income distributions have higher rates of infant mortality than countries with similar levels of national product per capita but more equal income distributions [Flegg, 1982; Rogers, 1979]. This finding is not surprising. Infant mortality is concentrated among the poor. If the rich are richer and the average level of income is the same, the poor are poorer. What is surprising, however, is that infant mortality appears to be posi-tively related to the income share of the rich (the upper 5 percent of the income distribution) when the incomes of the poor (the lowest 20 percent) are equalized among countries. In this paper I first document this striking empirical regularity. To a first approxima-tion, an increase in the wealth share of the rich while the incomes of the poor are held constant leaves the poor's command over resources unchanged. So why is such an increase associated with a decrease in the poor's ability to achieve the valued outcome—low infant mortality? The association presumably arises because an unequal income distribution is associated with another factor making for high infant mortality. This paper considers and tests several possible explanations, such as the provision of medical services, the degree of urbanization, the extent of female literacy, and differences in the composition of births among different income groups. None of these factors adequately accounts for the positive association between the incomes of the rich and infant mortality.

Professional forecasters may not simply aim to minimize expected squared forecast errors. In models with repeated forecasts the pattern of forecasts reveals valuable information about the forecasters even before the outcome is realized. Rational forecasters will compromise between minimizing errors and mimicking prediction patterns typical of able forecasters. Simple models based on this argument imply that forecasts are biased in the direction of forecasts typical of able forecasters. Our models of strategic bias are rejected empirically as forecasts are biased in directions typical of forecasters with large mean squared forecast errors. This observation is consistent with behavioral explanations of forecast bias.