Knowledge that Transforms

This paper considers the effects for offering a “would-not-vote” option in contingent valuation (CV) questions framed using the referendum format. This approach arises from a suggestion made by the National Oceanic and Atmospheric Administration's (NOAA) panel on contingent valuation. The NOAA panel was asked to evaluate the use of this method for estimating the economic value of nonmarketed environmental resources in the context of natural resource damage assessments. This test used the CV questionnaire developed for the study of the Exxon Valdez oil spill conducted by the State of Alaska with in-person interviews. The findings suggest that when those selecting the “would-not-vote” response are treated as having voted “against” the program (a conservative coding), offering this option does not alter (1) the distribution of “for” and “against” responses (2) the estimates of willingness to pay derived from these choices, or (3) the construct validity of the results.

This study tests for the presence of prejudicial or “noneconomic” discrimination on the part of mortgage lenders by evaluating the performance of home mortgage loans. The approach differs from that of previous studies of loan performance in that it is based on the proposition that noneconomic discrimination should be more pronounced in less competitive lending environments, while statistical discrimination should not. Using a rich set of FHA-insured loan records and measures of local market concentration to proxy the competitive environment, we test for the prediction of better loan performance by minority borrowers relative to white borrowers in more concentrated markets. We argue that this approach substantially reduces the potential for omitted-variable bias that has cast a shadow on previous studies of lending discrimination. Results fail to reject the null hypothesis of no noneconomic discrimination.

We assume that the instantaneous riskless rate reverts toward a central tendency which, in turn, is changing stochastically over time. As a result, current short-term rates are not sufficient to predict future short-term rate movements, as it would be the case if the central tendency were constant. However, since longer maturity bond prices incorporate information about the central tendency, longer maturity bond yields can be used to predict future short-term rate movements. We develop a two-factor model of the term structure which implies that a linear combination of any two rates can be used as a proxy for the central tendency. Based on this central-tendency proxy, we estimate a model of the one-month rate that performs better than models which assume the central tendency to be constant.

A feature of U.S. postwar business cycle experience that is by now widely documented is the tendency of the spread between the respective interest rates on commercial paper and Treasury bills to widen shortly before the onset of recessions. By contrast, the paper—bill spread did not anticipate the 1990–1991 recession. Empirical work presented in this paper supports two (not mutually exclusive) explanations for this departure from past experience. First, at least part of the paper—bill spread's predictive content with respect to business cycle fluctuations stems from its role as an indicator of monetary policy, but the 1990–1991 recession was unusual in postwar U.S. experience in not being immediately precipitated by tight monetary policy. Second, movements of the spread during the few years just prior to the 1990–1991 recession were strongly influenced by changes in the relative quantities of commercial paper, bank CDs, and Treasury bills that occurred for reasons unrelated to the business cycle. This latter finding in particular sheds light on the important role of imperfect substitutability of different short-term debt instruments in investors' portfolios, and highlights the burdens associated with using relative interest rate relationships as business cycle indicators.

This paper explores issues that arise in the evaluation of social programs using experimental data in the frequently encountered case where some of the experimental treatment group members drop out of the program prior to receiving treatment. We begin with the standard estimator for this case and the identifying assumption upon which it rests. We then examine the behavior of the estimator when the dropouts receive a partial “dose” of the program treatment prior to dropping out of the program. In the case of partial treatment, the identifying assumption is typically violated, thereby making the estimator inconsistent for the conventional parameter of interest: the impact of full treatment on the fully treated. We develop a test of the identifying assumption underlying the standard estimator and consider whether exclusion restrictions produce identification of the mean impact of the program when this assumption fails to hold. Finally, we discuss alternative parameters of interest in the presence of partial treatment among the dropouts and argue that the conventional parameter is not always the economically interesting one. We apply our methods to data from a recent experimental evaluation of the Job Training Partnership Act (JTPA) program.

This study uses data from the National Longitudinal Survey of Youth to distinguish empirically between mover—stayer, “search good,” and “experience good” models of job mobility. We estimate wage models in which the pattern of overall job mobility affects both the level and tenure slope of the log-wage path. After controlling for the correlation between mobility patterns and time-constant person- and job-specific unobservables, we find that workers who undergo persistent mobility have lower log-wage paths than less mobile workers. This finding is consistent with models in which job mobility is driven by time-varying unobservables, such as “experience good” models, where changes in perceived match quality cause turnover.

Asymmetric behavior has been documented in postwar quar-terly U.S. unemployment rates. This suggests that improvement over conventional linear forecasts may be possible through the use of nonlinear time-series models. In this note an out-of-sample forecasting competition is carried out for a set of leading nonlinear time-series models. It is shown that several nonlinear forecasts do indeed dominate the linear forecast. The results are sensitive, however, to whether a stationarity-inducing transfor-mation is applied to the nonstationary unemployment rate series.

We argue that the effect of wives' earnings can be assessed meaningfully only by comparing the observed distribution of income with a reference distribution. The components of the standard decomposition of the Gini coefficient have no implicit reference distribution and therefore should not be interpreted as a measure of the effect of an income source on inequality. We suggest several intuitive counterfactual reference distributions and illustrate their use with 1979 and 1989 U.S. data. We conclude that wives' earnings reduced inequality in that the income distribution would have been less equal in their absence. Alternative measures of the impact have mixed results.

This paper examines the relationship between productivity, investment, and plant age for over 14,000 plants in the U.S. manufacturing sector for the period of 1972 to 1988. Productivity patterns vary significantly due to plant heterogeneity. Initially productivity increases with respect to plant age, but then it decreases. Productivity and growth in productivity are found to be systematically correlated with plant size and industry. However, there is virtually no observable relationship between investment and productivity or productivity growth. Overall the results indicate that plant heterogeneity and fixed effects are more important determinants of observable productivity patterns than sunk costs or capital reallocation.

During the spring and the fall of 1993, respondents to a national household survey were asked to report expectations of spring 1994 weekly earnings. Elicited in the form of subjective probabilities, these data are potentially much more informative than are typical reports of economic expectations. Subjective probability distributions of future weekly earnings are estimated for each respondent, based on his or her reports of a series of subjective probabilities. This paper analyzes the cross-sectional variation in expectations, revisions of expectations between the spring and the fall of 1993, and the relationship between 1993 expectations and the distribution of spring 1994 earnings realizations. Generally positive findings on the validity of the data bode well for the prospects of eliciting expectations in future surveys.