• IRP Podcasts
• IRP Webinars
• IRP on Facebook
• IRP on YouTube

The Institute for Research on Poverty (IRP) at the University of Wisconsin–Madison is a center for interdisciplinary research into the causes and consequences of poverty and inequality and the impact of related policies and programs. As the National Poverty Research Center sponsored by the U.S. Department of Health and Human Services, IRP coordinates the U.S. Collaborative of Poverty Centers in an integrated set of activities with the ultimate goal of improving the effectiveness of public policies to reduce poverty and inequality and their impacts on the well-being of the American people.

IRP Discussion Paper Abstracts - 1998

Full Text: DP 1183-98

This paper uses an historical perspective to examine the labor market prospects and the macroeconomic setting facing mothers with dependent children who were (or would have been) enrolled in the old AFDC program, now that their welfare status will be handled by the new state programs in the wake of the Personal Responsibility and Work Opportunity Reconciliation Act, passed by Congress in 1996. The new law mandates the end to entitlements to cash payments for these women and their families and requires that they become self-supporting after the time limits for the cash payments are reached-a maximum of five years. Effectively, job holding is to replace welfare assistance as the main means of self-support. This paper documents the historical record of three trends that, given the new welfare laws, will largely determine the future poverty status of the affected women: wage growth, women's labor force participation, and single-parent families (which reflect trends in marital breakups and in out-of-wedlock births). Since 1959, the first year for the modern series of poverty statistics, both women's labor force participation and female headship of families have increased, the latter increasing poverty rates and the former, by itself, reducing poverty rates. The paper argues that wage growth is central to reducing poverty, especially now that government income support programs have been drastically reduced. The favorable economic record in the United States from 1959 to 1973, when wages and family incomes grew, is contrasted with the period from 1973 to 1997, when wages stopped growing and the growth in family incomes was slow. Given the difficulty in reversing demographic trends, macroeconomic economic growth appears necessary and effective to reduce poverty.

Full Text: DP 1182-98

This paper provides recent national estimates of the short-term economic outcomes of marital dissolution for mothers, fathers, and children. In addition, the paper estimates the current and potential impact of private child support transfers on the economic well-being of the various parties involved. Data are from the Survey of Income and Program Participation. Mothers and children fare dramatically worse than fathers; however, these differences would be much more pronounced in the absence of private child support. Substantial increases in custodial family income are possible within the structure of the existing child support system, with minimal impact on poverty among nonresident fathers.

Full Text: DP 1181-98

This paper investigates whether employer sanctions for hiring undocumented workers introduced by the 1986 Immigration Reform and Control Act (IRCA) adversely affected the hourly earnings of Latino workers in the southwestern United States. We exploit the staggering of the sanctions and employee verification requirements across sectors to estimate this effect. In particular, IRCA's employer-sanctions provisions were not extended to agricultural employers until 2 years after their imposition on nonagricultural employers. Hence, Latino agricultural workers provide a control group against which to compare changes in the wages of Latinos in nonagricultural employment. We find substantial pre-post IRCA declines in the hourly earnings of Latino nonagricultural workers relative to Latinos in agriculture. This pattern, however, is considerably stronger for Latino men than Latina women. We do not observe similar intersectoral shifts in relative wages among non-Latino white workers. In fact, the relative wage changes for non-Latino white workers are of the opposite sign. Finally, the pre-post IRCA relative decline in Latino nonagricultural wages reverses the pre-IRCA trend in which the relative earnings of Latino nonagricultural workers had been increasing.

Full Text: DP 1180-98

We investigate the effects of participation in the Chicago Child-Parent Center and Expansion Program from ages 3 to 9 on early school dropout at age 17. The Child-Parent Centers offer a government-funded educational intervention program in preschool through second or third grade in 20 locations in Chicago's poorest neighborhoods. Using data from the Chicago Longitudinal Study, we address two major questions: (1) Is participation in the Child-Parent Centers program associated with a lower rate of high school dropout at age 17? (2) Which nonintervention variables predict high school dropout? After comparing children in 20 intervention sites with similar children who attended schools in similarly poor neighborhoods in which the intervention program was not offered, we find that participation in the intervention offered by the Child-Parent Centers is associated with a 7 or 8 percentage point reduction in the probability of dropout. Our findings also indicate that parental involvement in schooling and avoidance of frequent school mobility are important predictors of high school completion.

Full Text: DP 1179-98

This paper describes how a public-private partnership paved the way for an evaluation of Medicaid managed care with previously inaccessible data. Also included are (1) a brief description of Wisconsin's Medicaid expansion program, (2) research questions and reflections on implications generated by key informant interviews, (3) some answers to these questions and their policy implications gleaned from a Medicaid enrollee survey, and (4) an assessment of the benefits and policy implications of this partnership. The partnership developed in Wisconsin is relevant for other states with large rural populations who will be facing both welfare reform and managed care.

Full Text: DP 1178-98

This report presents the findings from a telephone survey of 313 respondents who have family members enrolled in Medicaid managed care in a multicounty region encompassing both rural and urban counties in Wisconsin. Consumer perspectives on access to, utilization of, and satisfaction with health care provided by a managed care organization are presented and discussed. Differences between the rural and urban counties are noted and implications for policy are explored.

Full Text: DP 1177-98

In this paper we present data from a new survey of 900 employers in Michigan that was designed to gauge employer demand for welfare recipients. The results show that, given the current tightness of labor markets in Michigan, prospective demand for recipients is fairly high. On the other hand, prospective employment is quite highly correlated with measures of unmet labor demand at the establishment level, implying that much of this employment could disappear during the next recession. Many of the prospective jobs are also found in establishments to which inner-city minorities might have limited access, such as small/suburban establishments that receive few black applicants or that recruit informally. Absenteeism and basic skill readiness are potential problems for welfare recipients seeking employment, based on jobs filled by recipients to date or those that are prospectively available. The effects of a variety of potential policy responses targeted at private employers (such as job placement efforts, tax credits for employment or training, etc.) are also considered.

Full Text: DP 1176-98

This report analyzes different approaches to measuring food problems among impoverished households. Researchers investigating what public policy analysts refer to as hunger have sketched out alternative conceptual spaces within which these food problems can be measured. The narrower conceptual space may be termed food insufficiency and is distinguished by restricted household food stores, too little food intake among adults or children in the household, and direct reports or perceptions of hunger among household members. The broader conceptual space may be termed food insecurity. This term subsumes food insufficiency and extends to include resource insufficiency, the inability to acquire enough nutritious food through culturally normalized means, and anxiety about this inability, along with various attempts to augment or stretch the food supply. Since the late 1980s these two definitions of food problems in impoverished households have been understood as hunger, insofar as hunger is a measurable phenomenon for policy purposes in an advanced industrial nation such as the United States. These definitions are now central in the development of survey research items used to estimate the population prevalence of hunger, along with its predisposing socioeconomic conditions and resultant health and developmental consequences. Drawing on a data set containing survey responses from more than 5200 low income households with children in 11 sites from around the nation, we conduct an empirical inquiry of questionnaire items tapping phenomena from each conception defined above. Specifically, the study examines 34 distinct questionnaire items, and it addresses four research questions: (1) To what aspect of food insecurity or food insufficiency does each indicator point? (2) Can particular combinations of items be scaled? (3) When scaled, do the items demonstrate content validity? (4) How do the alternative measures perform in an operationalized model of the antecedents and consequences of household food problems? We test models that include variables such as household income, household food and shelter expenditures, and bills in arrears, along with the health status of a randomly chosen child from each household.

Full Text: DP 1175-98

Past studies have concluded that Social Security retirement benefits and private pension plan survivor benefits are important sources of income for widows. Using data from the New Beneficiary Surveys of the Social Security Administration, we examine the effect of the 1974 Employee Retirement Income Security Act on the choice of a joint-and-survivor pension, which provides pension benefits to the surviving spouse of a pensioner. We find that ERISA has substantially increased the percentage of widows who receive a survivor pension. We then explore some of the factors associated with choice of a joint-and-survivor pension.

Full Text: DP 1174-98

While overall employment in the United States has risen in the last 30 years, the employment and earnings prospects for youths have fallen relative to those for older workers. This deterioration in youth labor market conditions has been most pronounced for low-skilled youths, high school dropouts, and those with low IQs. Using data from national longitudinal studies of young men, young women, and youths, this paper examines a number of aspects of the labor market outcomes of youths entering the labor market at two different times. The first group entered the robust labor market of the late 1960s, while the second group entered the deteriorated labor market of the mid-1980s. Consistent with previous research, this paper finds an improvement over the two periods in levels of employment and earnings for high-skilled youths, with a corresponding deterioration for lower-skilled youths. The paper presents a unique analysis of the growth trajectories of earnings and employment for high- and low-skilled youths in the two cohorts. We find substantial within-cohort growth for high-skilled youths in both cohorts (as well an improvement in household economic circumstances), with a corresponding deterioration in earnings, employment, and household economic circumstances for lower-skilled youths, especially those in the later cohort.

Full Text: DP 1173-98

Wisconsin's welfare reform program, Wisconsin Works (W-2), is among the most ambitious and comprehensive state reforms supported by the U.S. government's Temporary Assistance for Needy Families (TANF) block grant. This paper describes the W-2 program in Wisconsin, compares distinctive features of the program to TANF programs in selected other states, discusses how Wisconsin came to offer such a program, and describes early trends in W-2 program implementation. The paper also makes suggestions for evaluating distinctive features of W-2.

Full Text: DP 1172-98

The proportion of the population residing in high-poverty urban areas grew in the 1970s and 1980s (Wilson 1987; Jargowsky 1997). This paper examines why the number of high-poverty neighborhoods increased by using data from the Panel Study of Income Dynamics PSID) matched with data on tracts from the decennial census. The main findings are that (1) African Americans are moving into white neighborhoods at a high rate, but the white population is declining in areas with substantial black populations quickly enough that the proportion black in white areas is not increasing and (2) there is no systematic tendency for poverty rates among stayers in poor neighborhoods to increase over time relative to poverty rates of other neighborhood types, although there is some evidence of a larger increase in the poverty rate of moderately poor black neighborhoods than other neighborhood types during the early 1980s recession. Implications of the findings for theories of high-poverty neighborhoods and racial segregation are discussed.

Full Text: DP 1171-98

The trend in national policy over the past two decades has emphasized self-reliance and a reduced role of government in society. Given this ideological shift, the official poverty measure, which is based on the premise that all families should have sufficient income from either their own efforts or government support to boost them above a family-size-specific threshold, appears now to have less policy relevance than in prior years. In this paper we present a new concept of poverty, the inability to be self-reliant, which is based on the ability of a family, using its own resources, to support a level of consumption in excess of needs. This concept closely parallels the "capability poverty" measure that has been proposed by Amartya Sen. We use this measure to examine the size and composition of the poor population from 1975 to 1995.  We find that poverty in terms of self-reliance increased more rapidly over the 1975-95 period than did official poverty. We find that families commonly thought to be the most impoverished-those headed by minorities, single women with children, and individuals with low levels of education-have the highest levels of self-reliance poverty. However, these groups have also experienced the smallest increases in this poverty measure. Families largely thought to be economically secure, specifically those headed by whites, men, married couples, and highly educated individuals, while having the lowest levels of self-reliance poverty, have also experienced the largest increases in that measure. We speculate that the trends in self-reliance poverty stem largely from underlying trends in the U.S. economy, in particular the relative decline of wage rates among whites and men, and the rapidly expanding college-educated group.

Full Text: DP 1170-98

Most discussions of welfare and work have focused on how demographic characteristics, schooling, training, and work experience limit welfare mothers' employment and wages, but they have largely ignored factors such as inappropriate workplace behaviors, expectations of discrimination and harassment, depression, alcoholism, and domestic violence, all of which may affect welfare mothers and make employment difficult. In this paper we review the prevalence of these individual-level barriers and argue that they, in combination with an economy which does not pay low-skill workers well, are likely to impede employment and self-sufficiency for a large proportion of welfare mothers. At the end of the review, we summarize the current state of knowledge about barriers to the employment of welfare recipients and suggest several ways in which welfare-to-work programs might address these barriers.

Full Text: DP 1169-98

We use data from a survey of employers to investigate how Affirmative Action in recruiting and hiring influences hiring practices, personnel policies, and ultimately employment out- comes. Our results show that Affirmative Action increases the number of recruitment and screening practices used by employers, raises their willingness to hire stigmatized applicants, increases the number of minority and female applicants as well as employees, and increases employers' tendencies to provide training and to formally evaluate employees.

When Affirmative Action is used in recruiting, it does not lead to lower credentials or performance of women and minorities hired. When it is also used in hiring, it yields female and minority employees whose credentials are somewhat weaker, though performance generally is not. Overall, then, the more intensive search, evaluation, and training that accompany Affirmative Action appear to offset any tendencies of the policy to lead to hiring of less-qualified or less-productive women and minorities.

Full Text: DP 1168-98

We use data from the National Education Longitudinal Survey to examine the effects of family structure and school changing on attendance during high school and educational continuation through college entry. We find that both family structure and changing schools are associated with more attendance problems during high school and with school continuation decisions after high school. The results also show that family structure, changing schools, and attendance patterns play important roles in shaping the educational attainment of individuals, including their postsecondary educational experiences.

Full Text: DP 1167-98

This study extends Reynolds' (1992) investigation of the social- psychological influences on grade retention and school adjustment in early childhood by tracing the predictors and consequences of grade retention for school achievement, perceived competence, and delinquency in early adolescence (age 14). The study sample included 1,164 (93 percent of the sample from the original study) low-income, mostly black children in the Chicago Longitudinal Study. Twenty-eight percent of the study sample were retained-in-grade by age 14 (first grade to eighth grade). The strongest predictors of retention were early school performance (test scores and grades), sex (boys were more likely to be retained), parent participation in school, and school mobility. Overall, grade retention was significantly associated with lower reading and math achievement at age 14 above and beyond a comprehensive set of explanatory variables. Results based on same-age comparison groups yielded larger effects of retention on school achievement than those based on same-grade comparisons, but both approaches indicated that grade retention was associated with significantly lower reading achievement. In the full model, grade retention was unrelated to perceived school competence at age 12 and to delinquency infractions at age 14. With the exception of reading achievement, retention during the primary grades and retention during grades 4 to 7 yielded a similar pattern of effects. Findings were largely consistent with the earlier study and suggest that intervention approaches other than grade retention are needed to better promote school achievement and adjustment.

Full Text: DP 1166-98

When the twentieth century is viewed as a whole, no clear trend in income inequality emerges. Inequality was high and rising during the first three decades and peaked during the Depression. It fell sharply during World War II and remained at the lower level in the 1950s and 1960s. From the 1970s through the mid-1990s inequality steadily increased to levels not seen since World War II, though well below those during the first three decades.

The rate of poverty exhibited a long-run downward trend from about 60-70 percent in the earlier years of the century to the 12-14 percent range in recent years, with considerable fluctuation around this secular trend.

Changes in inequality were produced largely by demographic and technological changes, the growth and decline of various industries, changes in patterns of international trade, cyclical unemployment, and World War II. The primary drivers of the rate of poverty were economic growth and factors that produced changes in income inequality, particularly demographic change and unemployment.

Public policy has reduced the market-generated level of inequality, but since 1950 has had little effect on the trend in inequality. Prior to 1950, the growth of government, and particularly the introduction of a broadly based income tax during World War II, coincided with and partly produced the sharp downward shift in inequality of that era. Government had little effect on poverty rates until 1950. Public income transfer programs have reduced poverty rates appreciably in recent decades. Since World War II, when they have been on a large enough scale to matter, changes in tax and transfer policy have tended to reinforce market-generated trends in inequality and poverty rather than offset them.

Full Text: DP 1165-98

In this paper, we examine the effects of family structure on age at first sexual intercourse before marriage for recent cohorts of women. Previous research on the linkage between family structure and sexual initiation has employed relatively crude measures of family structure-typically a snapshot of the respondent's family structure at age 14. We use retrospective parent histories from the 1979-87 National Longitudinal Survey of Youth to construct dynamic measures of family structure, using information on the number and types of parents in the respondent's household between birth and age 18. We use these measures to test the effects of prolonged exposure to a single-mother family, prolonged absence of a biological father, parental presence during adolescence, and instability in family structure. For white women, age-specific rates of first sexual intercourse are significantly and positively associated with time-varying measures for the number of family transitions; for black women, age-specific rates are significantly and positively associated with time-varying variables for having resided in a mother-only or father-only family during adolescence. Net of other effects of family structure, we find no significant effects for white or black women of being born out of wedlock, prolonged exposure to a single-mother family, or prolonged absence of a biological father. We interpret our results for white women as consistent with a turbulence and instability hypothesis, but as providing little support for socialization or parental-control hypotheses; for black women, our results are consistent with the parental-control hypothesis, but provide little support for the socialization and turbulence hypotheses. Overall, these findings suggest that the processes influencing the transition to sexual activity may vary quite markedly for white and black women.

Full Text: DP 1164-98

The recent decline in the national welfare rolls suggests that mandatory work programs can reduce dependency by more than evaluations suggest. The nonexperimental literature does not test that possibility well. This study uses field interviewing and program data more fully than previously to portray the forces shaping caseload decline. It focuses on Wisconsin, the state with the most dramatic caseload fall.

A time series analysis of the state caseload trend over 1986-94 casts doubt on the view that good economic conditions and benefit cuts alone account for the caseload decline. Cross-sectional analyses comparing counties find strong evidence that both a good economy and demanding work requirements helped drive the caseload down. However, the consequences for recipients are unclear, and to reduce dependency this way makes heavy political and administrative demands on government.

Full Text: DP 1163-98

This paper uses nationally representative data to describe monthly cycles in food expenditure and food intake by food stamp recipients. Food expenditure peaks sharply in the first 3 days after food stamps are received. The corresponding cycle in food intake differs for various categories of food stamp recipients. Food stamp recipients who also receive AFDC appear to maintain steady food intake across the whole month, while AFDC nonrecipients experience a significant drop in intake at the end of the month. Children appear to maintain steady food intake, while adults appear to experience a significant drop. Households that conduct major grocery shopping trips more frequently than once per month maintain steady food intake, while households that shop less frequently experience a significant drop. The food stamp cycle has implications for two areas of research: the measurement of hunger and food insecurity in the United States and the measurement of the impact of the U.S. Food Stamp Program. Intramonthly patterns in food expenditure and food intake have potential implications for policy decisions about the frequency of food stamp benefit delivery, the evaluation of new electronic benefit transfer systems that are replacing traditional food stamp coupons, and nutrition education efforts.

Full Text: DP 1162-98

In this paper, I use data from a new survey of employers in four large metropolitan areas to analyze the flow of black applicants to different kinds of employers and the extent to which these applicants are hired. The results show that less-educated black workers apply less frequently for jobs in the suburbs than in central cities, especially at smaller establishments. Their lower tendency to apply for suburban jobs is mostly accounted for by the higher costs to central-city black residents of applying there, and by lower information flows as well. Black applicants, especially less-educated black males, are also less likely to be hired at suburban establishments, particularly where they must deal with white customers. These results suggest the need for policies to reduce the costs of applying for suburban jobs and to improve the flow of information about suburban employment prospects to less-educated blacks, and perhaps a need to complement such policies with more effective enforcement of antidiscrimination laws in small suburban establishments.

Full Text: DP 1161-98

We develop and estimate a model of social-psychological determinants of entry to the Aid to Families with Dependent Children (AFDC) program, the primary cash welfare program in the United States for 60 years until replaced in 1996. The structural model holds that attitudes and personality characteristics influence a woman's likelihood of becoming demographically and financially eligible for welfare and her willingness to bear the stigma of receiving benefits. These factors, in turn, affect the likelihood of actually going on welfare. We test for a relationship between social-psychological variables and welfare participation using data from the youngest cohorts of women in the National Longitudinal Survey of Youth. We estimate logit models of the probability of ever participating in AFDC up to age 25 and hazard models of the timing until first use of AFDC. The attitudes and personality characteristics in the empirical model are self-esteem, locus of control, attitudes toward school, attitudes toward women's work and family roles, commitment to work, and aversion to accepting public assistance. We find strong associations between welfare use and several attitudes and personality characteristics, but most of the associations are not robust to the inclusion of exogenous personal and family background characteristics. Consistent, strong evidence suggests that more positive attitudes toward school lower the likelihood of using welfare and increase duration until first receipt.

Full Text: DP 1160-98

This paper presents a test of the hypothesis that employers in suburban locations are more likely to discriminate against African Americans than are employers located in central cities. Using a difference-in-difference framework, we compare central-city/suburban differences in racial hiring outcomes for firms where a white person is in charge of hiring (white employers, for short) to similar geographic differences in outcomes for firms where a black person is in charge of hiring (black employers). We find that both suburban black and white employers hire fewer blacks than their central-city counterparts. Moreover, the central-city/suburban hiring gap among black employers is as large as, or larger than, that of white employers. Suburban black employers, however, receive many more applications from blacks and hire more blacks than do white firms in either location.

Full Text: DP 1159-98

The 1996 welfare reform legislation establishing the Temporary Assistance for Needy Families (TANF) program marks a significant change in U.S. social and economic policy. This legislation represents the ascendance of the view that individuals and families need to be self-reliant and that collective support for individual well-being should be minimized. We first describe the major provisions of TANF, providing some background on its differences from prior policy targeted at needy families. Then we catalogue the wide variety of economic changes that are implicit in the new law, stressing those related to changed property rights, fiscal relations among jurisdictions, and economic incentives facing families. Third, we illustrate the form of state reforms that are likely to develop in response to the federal policy change by describing the actions of the state of Wisconsin, which has taken the lead in implementing the new policy. We conclude with a list of yet unanswered questions that will ultimately determine just how far this policy change will slide the nation along the efficiency-equity tradeoff function, away from the equity axis. The answer to these questions will influence the attraction the U.S. reform might hold as a model for other nations concerned with their own safety net programs for poor people.

Full Text: DP 1158-98

This paper reviews trends in poverty since the late 1960s. Poverty is as prevalent now as it was then. A good deal about the nature of poverty and our efforts to deal with it, however, has changed. The paper has three major themes. First, we note that urban poverty is no longer a predominantly black issue; the composition of the urban poor has changed considerably since the late 1960s. Second, poverty outside the central city continues to be a problem and should not be ignored. Third, we argue that the current approach to poverty that emphasizes personal responsibility is clearly ineffective and should be replaced with a focus on poverty as a public health issue. Viewing poverty as a public health issue points to the crucial role of both government and individuals in efforts to overcome it.

Full Text: DP 1157-98

We use data from surveys of employers and households in four metropolitan areas to predict the degree of job availability for various types of disadvantaged workers, such as minorities, high school dropouts, and welfare recipients. We conduct simulations in which we "match" workers to jobs on the basis of skill, spatial, and racial characteristics of each. Our results show that roughly 9 to 17 percent of actual or potential jobseekers are likely to have difficulty finding work, even in tight labor markets. Simulated mismatch rates for disadvantaged workers are considerably higher. We conclude that disadvantaged workers face quite limited job availability, at least in the short run. The wages and benefits for jobs available to these workers are also quite low.

Full Text: DP 1156-98

By 1989, the Michigan Panel Study on Income Dynamics (PSID) had experienced approximately 50 percent sample loss from its initial 1968 membership due to cumulative attrition. We study the effect of this attrition on the unconditional distributions of several socioeconomic variables and on the estimates of several sets of regression coefficients. We provide a statistical framework for conducting tests for attrition bias that draws a sharp distinction between selection on unobservables and on observables and that shows that weighted least squares can generate consistent parameter estimates when selection is based on observables, even when they are endogenous. Our empirical analysis shows that attrition is highly selective and is concentrated among individuals of lower socioeconomic status. We also show that attrition is concentrated among those with more unstable earnings, marriage, and migration histories. Nevertheless, we find that these variables explain very little of the attrition in the sample and that the selection that occurs is moderated by regression-to-the-mean effects from selection on transitory components that fade over time. Consequently, despite the large amount of attrition, we find no strong evidence that attrition has seriously distorted the representativeness of the PSID through 1989, and considerable evidence that its cross-sectional representativeness has remained roughly intact.

Full Text: DP 1155-98

This paper lays out the advantages of using a model for developing research questions and methodologies aimed at evaluating how managed care arrangements for rural Medicaid moms and their children might affect their access to health care and their related health status. The PRECEDE (predisposing, reinforcing, and enabling causes in educational diagnosis and evaluation) health-education program planning model is refined and applied to a population of Medicaid-eligible moms and their children to help organize and clarify the research questions and to identify the types of variables we need to consider for this exercise. An explanation of those variables, why they are important, and how they can be obtained is presented.

Full Text: DP 1154-98

This paper explores the possibility that states respond asymmetrically to increases versus decreases in their neighboring states' welfare benefit levels. We present a theoretical model suggesting that states respond more to decreases than to increases in their neighbors' benefit levels. To test this proposition empirically, we use a panel of annual state-level data from 1983 to 1994 for each of the contiguous United States and the District of Columbia, and we observe changes in state demographic and economic characteristics as well as changes in state welfare benefits. We find substantial empirical evidence that uniformly supports our argument. State responses to neighbor benefit decreases tend to be at least twice as large as their responses to neighbor benefit increases. Our empirical results are robust to modeling neighbor benefits as endogenous. Our results, therefore, have substantial implications for public policy in the wake of the increased decentralization of welfare policy associated with the welfare reforms of 1996.