This paper documents changes in the tax consequence of marriage over the period 1984 to 1997. Reversing the impact of the 1986 Tax Reform Act, tax acts in 1990 and 1993 are found to increasingly tax marriage. Our decomposition of different components show that, altogether, tax laws explain most (55–60 percent) of the change in the tax cost of marriage between 1984 and 1997. Our decompositions also show that the non–tax changes are almost exclusively driven by the changing labor market attachment of married women (as measured by their share of family earnings) and not by family size or total family income.

This paper examines the impact of changes in labor market conditions on participation in the Aid to Families with Dependent Children (AFDC) program in California. Transitions off welfare and transitions back onto welfare are estimated using discrete duration models that control for local labor market conditions, demographic and neighborhood characteristics, duration effects, county-fixed effects, time effects, and county- speciŽfic time trends. The results show that higher unemployment rates, lower employment growth, lower employment-to-population ratios, and lower wage growth are associated with longer welfare spells and higher recidivism rates. Hispanics, blacks, and two-parent families are the groups that are most sensitive to changes in local labor market conditions.

In this paper, we examine the role played by differential mortality in esti- mates of life cycle wealth profiles. Our study makes three contributions. First, we show that the Survey of Income and Program Participation (SIPP) provides reliable data on mortality as compared to the US life ta- ble data. Second, we provide estimates of the relationship between mortal- ity and wealth and show strong evidence of differential mortality. Lastly, and most importantly, we show that the differences in mortality by wealth-age profiles.

In this paper, I examine the effect of business cycles on the employment, earnings, and income of persons in different demographic groups. I classify individuals by sex, education, and race. The analysis uses data from the Current Population Survey’s Outgoing Rotation Group file, covering the period 1979-1992, and March Annual Demographic files (ADF) covering the period 1975-1997. Many different individual and family outcome measures are considered including: employment to population ratios, weekly earnings, hourly earnings, annual hours, annual earnings, family earnings, family transfer income, and total family income. The regression model is specified such that the key parameters measure how the labor market outcomes of less skilled workers vary with the business cycle relative to the variability for high skill groups. The analysis uses variation across MSAs in the timing and severity of shocks. The results consistently show that individuals with lower education levels, nonwhites, and low skill women experience greater cyclical fluctuation than high skill men. These results are the most striking when examining comprehensive measures of labor force activity such as the likelihood of full-time year around work. Government transfers and the earnings of other family members decrease the differences between groups, as business cycles have more skill-group neutral effects on family income than individual earnings. The paper examines the stability of these results by comparing evidence across the 1982 and1992 recessions. The evidence suggests that the 1992 recession led to more uniform effects across skill groups than earlier cycles.

This paper identifies the impact of "program realignment," a 1991 California state policy that significantly enhanced local governments' financial risk and programmatic authority for public mental health services, on treatment costs per user, and on the mix of inpatient and outpatient service costs. The study employs a natural pre-realignment and post-realignment design using the 59 California local mental health authorities (LMHAs) as the unit of analysis over a seven-year period spanning policy implementation. Total treatment and inpatient cost per user decreases and outpatient cost per user increases after program realignment. Higher levels of contracting with private providers tend to enhance this trend, while risk for institutional services reduces user costs uniformly. Financial and programmatic decentralization can enhance cost efficiency in treatment, while promoting substitution of outpatient services for inpatient services. Local conditions such as risk and contracting determine the extent of the policy response.