This paper argues that long-run trends in geographic segregation are inconsistent with models where residential choice depends solely on local public goods (the Tiebout hypothesis). We develop an extension of the Tiebout model that predicts as mobility costs fall, the heterogeneity across communities of individual public good preferences and of public good provision must (weakly) increase. Given the secular decline in mobility costs, these predictions can be evaluated using historical data. We find decreasing heterogeneity in policies and proxies for preferences across (i) a sample of U.S. municipalities (1870–1990); (ii) all Boston-area municipalities (1870–1990); and (iii) all U.S. counties (1850–1990).

In many empirical studies of the effect of social programs researchers assume that, conditional on a set of observed covariates, assignment to the treatment is exogenous or unconfounded (aka selection on observables). Often this assumption is not realistic, and researchers are concerned about the robustness of their results to departures from it. One approach (e.g., Charles Manski, 1990) is to entirely drop the exogeneity assumption and investigate what can be learned about treatment effects without it. With unbounded outcomes, and in the absence of alternative identifying assumptions, there are no restrictions on the set of possible values for average treatment effects. This does not mean, however, that all evaluations are equally sensitive to departures from the exogeneity assumption. In this paper I explore an alternative approach, developed by Paul Rosenbaum and Donald Rubin (1983), where the assumption of exogeneity is explicitly relaxed by allowing for a limited amount of correlation between treatment and unobserved components of the outcomes. The starting point of the sensitivity analysis is the assumption that the exogeneity assumption is satisfied only conditional on an additional unobserved covariate. Making assumptions about the effect of the unobserved covariate on the outcome and its correlation with the treatment, I trace out the set of possible values for the treatment effect of interest. By considering a sufficiently large set of possible correlations with outcomes and treatment, one can recover the bounds on the treatment effect derived by Manski (1990). The approach here, in the spirit of Rosenbaum and Rubin (1983) and Rosenbaum (1995), is to allow only a limited amount of correlation and to judge the sensitivity of average treatment-effect estimates to such correlations. There are two novel features of the proposed analysis. First, rather than formulate the sensitivity in terms of coefficients on the unobserved covariate, the sensitivity results are presented in terms of partial R values, which may be easier to interpret. Second, the partial R values of the unobserved covariates are compared to those for the observed covariates in order to facilitate judgments regarding the plausibility of values necessary to substantially change results obtained under exogeneity. The proposed sensitivity analysis is conceptually related to the practice of assessing sensitivity of estimates by comparisons with results obtained by discarding one or more observed covariates (James Heckman and V. Joseph Hotz, 1989; Rajeev Dehejia and Sadek Wahba, 1999; Jeffrey Smith and Petra Todd, 2001). The attraction of the sensitivity analysis is that it is more directly relevant: one is not interested in what would have happened in the absence of covariates actually observed, but in biases that are the result from not observing all relevant covariates.

I agree with the point made by Edoh Y. Amiran and Daniel A. Hagen (2003) that there can be a substantial, or even infinite, divergence between the WTA and WTP for a public good even where there is a nonzero elasticity of substitution between market goods and the public good, provided that the indifference curves are asymptotically bounded with respect to market goods in the manner they describe. This is an important point. They are also correct to point out that the elasticity of substitution is a local concept, whereas their asymptotic boundedness condition applies also for discrete changes. My 1991 paper used a local analysis because it was following the structure of the analysis in Robert D. Willig (1976) and Alan Randall and John R. Stoll (1980); I wanted to show that, while Randall and Stoll appeared to extend Willig’s local result on WTA versus WTP from price changes to changes in the quantity of a public good, the relevant elasticity was in fact different and involved the substitution elasticity as well as the income elasticity. I view these points by Amiran and Hagen as not two separate results but essentially the same result: their asymptotic boundedness condition generalizes my zero elasticity of substitution condition to discrete changes. The asymptotic boundedness condition can be expressed as follows: assuming a bivariate utility function u( x, q), and given a reference point ( x*, q*) associated with a reference utility level u* u( x*, q*), there exists some q q* such that, for all q q , there exists no x such that u( x , q) u*. In other words, no amount of x can substitute for the reduction in public good from q* to q q . In the case of a zero elasticity of substitution, q q* but, as Amiran and Hagen show in their Theorem 1, this is unnecessarily restrictive when dealing with a discrete reduction in q. Furthermore, their Theorem 2 can be viewed as a special case of their Theorem 1 in which q 0, which makes q an essential commodity. It is well known in consumer theory that the WTA to avoid the loss of an essential market good is infinite; their Theorem 2 extends this result to the case of an essential nonmarket good. But, as their Theorem 1 shows, essentialness is not necessary for an infinite WTA. The boundedness condition is the key, and this implies a fundamental lack of substitutability between money (market goods) and the public good.

This paper extends the standard principal–agent model to allow for subjective evaluation. The optimal contract results in more compressed pay relative to the case with verifiable performance measures. Moreover, discrimination against an individual implies lower pay and performance, suggesting that the extent of discrimination as measured after controlling for performance may underestimate the level of true discrimination. Finally, the optimal contract entails the use of bonus pay rather than the threat of dismissal, hence neither “efficiency wages” nor the right to dismiss an employee are necessary ingredients for an optimal incentive contract.

This paper analyzes the relationship between economic reform and the democratic process to ask the following question: Does the likelihood of adoption of an economic reform increase with an increase in the efficiency benefits from that reform? A priori we might expect that this likelihood should increase monotonically, for two reasons. First, economic reform results in an increase in the size of the national pie. If the government has the ability to make compensatory tax-transfers, an increase in each citizen's income is possible. Second, a larger economic reform results in a greater number of winners. This might also be expected to reinforce the political support for economic reform. So we should expect a reform with greater efficiency benefits to a larger population to have a greater chance of adoption by the government. However, in this paper we show that this intuition is mistaken. In particular, we demonstrate that there exists a certain non-monotonicity between the distribution of winners from economic reform and the probability of its adoption. An increase in the number of winners may lower, rather than increase, the likelihood of adoption of economic reform. In particular, even though reforms that benefit a minority or an overwhelming majority are adopted, reforms that benefit a smaller majority are not adopted.

We assemble bank-level and other data for Fed member banks to model determinants of bank failure. Fundamentals explain bank failure risk well. The first two Friedman-Schwartz crises are not associated with positive unexplained residual failure risk, or increased importance of bank illiquidity for forecasting failure. The third Friedman-Schwartz crisis is more ambiguous, but increased residual failure risk is small in the aggregate. The final crisis (early 1933) saw a large unexplained increase in bank failure risk. Local contagion and illiquidity may have played a role in pre-1933 bank failures, even though those effects were not large in their aggregate impact.

The consequences of bank distress for the economy during the Depression remain an area of unresolved controversy. Since John M. Keynes (1931) and Irving Fisher (1933), macroeconomists have argued that bank distress magnified the extent of the economic decline during the Depression. As the intermediaries controlling money and credit, banks were in a special position to transmit their distress to other sectors. But the mechanism through which banking distress mattered for the economy has been hotly contested.

Some Evidence on Race, Welfare Reform, and Household Income by Marianne P. Bitler, Jonah B. Gelbach and Hilary W. Hoynes. Published in volume 93, issue 2, pages 293-298 of American Economic Review, May 2003