Child welfare allocations: state-level estimates of effective guarantees

American Journal of Economics and Sociology, The,  Oct, 1995  by Howard Gensler

• E-mail
• Print
• Link

I

Introduction

Welfare is awarded primarily to single female household heads, and since women generally obtain custody of children, the amount of welfare which such families receive depends principally on the number of children in the family. Welfare program statistics are widely quoted and discussed, but the most interesting statistic receives scant attention: the amount of welfare support per child. An examination of welfare on a per child basis demonstrates that the level of support per child in America does not keep pace with the corresponding increase in the poverty level of income.

The welfare guarantee level is the amount of welfare which a family with no other income would be allocated by the income maintenance system. The guarantee level increases with family size. Adults are treated quite differently from children by the American welfare system. Accordingly, the welfare guarantee level can be decomposed into a base amount for the parent, and then a supplemental amount for each child. From this total family guarantee level, benefits are forfeited as income is earned. Although welfare is funded primarily by the federal government, the benefit levels are set by the states, and, therefore, vary widely. Benefit levels also change from year to year. Finally, nominal welfare statistics are inherently deceptive, and overstate the actual amount of welfare which a class member can expect to receive. For instance, suppose that two states both have nominal guarantee levels of $6,000 per year for a single mother with two children. However, in the first state, all such families receive such welfare. In the second state, through problems of access or qualification, only one-half of all such families receive such welfare. The expected welfare guarantee level in the second state would be $3,000 after taking into account the probability of receiving welfare. Expected, not nominal, welfare guarantee levels provide an unbiased description of a state's true support level.

Accordingly, here the grant allocation features of the entire cash or cash-equivalent welfare system on a state by state basis are estimated for each year from 1979 to 1990. Both the parent and per child grant levels are estimated for the main class of welfare beneficiaries under several programs in America for single female household heads, controlling for income. Welfare in this study refers to need-based income maintenance programs which do not depend on prior contributions, retirement status, disability, or veteran status. The programs which are excluded from this definition of welfare include the Old Age and Survivors Disability Income and the Supplemental Security Income Programs of the Social Security Act, military benefits, unemployment insurance, public and private pensions, Medicare, Medicaid, Workers' Compensation, educational programs, scholarships and the Head Start Program. Also excluded are housing programs due to valuation problems and the non-cash-equivalent nature of such programs. Because less than 7% of the sample received any form of housing subsidy, the magnitude of the understatement on the estimate of the expected benefit is small. Programs which are included in this definition of welfare are Aid to Families with Dependent Children (AFDC), the Food Stamp Program, Emergency Assistance, General Assistance, and the Low Income Home Energy Assistance Program.

After the data are described, a model of welfare determination is developed. The model is estimated by means of a tobit procedure to account for the probability of not receiving welfare when family income levels would otherwise warrant welfare program participation. Every year from 1979 to 1990 is estimated, and the results for the most recent year are presented in Figure 1. Substantial interstate variation is found in every year. The vast majority of the estimates indicate that for each additional child which a family has, the family is plunged deeper into poverty. Moreover, across years, national average real benefit levels suffered a 15% erosion.

II

The Data

The Data are derived from the United States Department of Commerce, Bureau of the Census' Current Population Surveys: Annual Demographic Files (March), 1980-1991.(1) This cross-sectional data set provides previous year economic information on civilian residents of the United States older than fourteen, living in non-institutional circumstances and men in the Armed Forces residing in non-military housing. This study uses a sample of low income, non-military, non-farm households. Total household income was limited to twice the official poverty level. Since this study analyzes welfare, subjects with higher incomes would not provide additional relevant variation. The main class of welfare beneficiaries in the United States analyzed herein, single female household heads were further limited to range in age from 18 to 65. Since the economic data is on the previous year, information for the years 1979 through 1990 was collected.

For this study of single female household heads, 20,281 observations are analyzed. The average age of the single female is 31.5, with an average of 11.1 years of education. The sample is 33% black, and 56% receive some amount of welfare. Earned income averaged $3,728 per year. Unearned, non-welfare income averaged $2,268 per year. Finally, 59% participated in the labor force. These women received an average of $2,135 of welfare per year. If the subject actually received welfare, the average annual grant was $3,810.