December 2005

Does Supplemental Security Income Enduringly Reduce Child Poverty?

Does Supplemental Security Income Enduringly Reduce Child Poverty?

A recent paper by Mark Duggan and Melissa Kearney found that participation in the Supplemental Security Income (SSI) program, which offers a monthly stipend to individuals who prove that they cannot work, leads to a significant and persistent reduction in the probability that a child lives in poverty. The disability benefits from this program have grown from covering 260,000 children in 1989 to covering 1.03 million in 2005. Duggan and Kearney found that SSI enrollment does not appear to crowd out work, affect the structure of the family, or impact health insurance coverage, but does translate to a direct increase in total household income. Is SSI a welfare benefit that does not degrade the recipient, or is it a program with no strings attached that fails to help the needy? At a December 12 AEI panel discussion, Mark Duggan of the University of Maryland and Melissa Kearney of the Brookings Institution presented their findings.

Mark Duggan
University of Maryland

Since 1972, SSI has provided assistance to low-income, aged, blind and disabled Americans. Since 1990, the eligibility criteria has been liberalized and become more inclusive. As a result of the U.S Supreme Court’s ruling in Sullivan v. Zebley, eligibility has increased by 260 percent.
 
In 2004, 7 million received benefits in a typical month, costing $37 billion--compared with $9.5 billion for the Temporary Assistance for Needy Families (TANF) program. Although there has been much research on the effects of TANF and the Aid to Families with Dependent Children (AFDC), there has been little research on the effects of SSI on families.

The percent of children covered by SSI rose from 0.5 percent in 1990 to 1.3 percent in 2003. By contrast, 12.1 percent were covered by AFDC in 1990 and only 5.5 percent by TANF in 2003. Between 1990 and 2003, the number of children in poverty remained roughly constant. Over the past twenty years, TANF enrollment has declined by 4 percent per year, and SSI enrollment has increased by 4 percent per year

With the U.S Census Bureau’s Survey of Income and Program Participation (SIPP) statistics, we were able to use reliable and nationally representative longitudinal data. Identifying causal relationships is difficult since SSI is a means-tested program, and so eligibility criteria will pre-screen according to particular activities. We also had to be careful, since changes to the Earned Income Tax Credit (EITC) and adult SSI eligibility would also affect our study of SSI child enrollment.

To be eligible for SSI, one must be below federal income and asset limits and be aged, disabled, or blind. Sullivan v. Zebley liberalized standards for getting children on SSI--since the same standards for adults’ willingness to work could not justifiably be held for children. As a result, the number of boys qualifying because of “mental disorders” has increased by a factor of fourteen and the number of girls by a factor of ten.

The choice facing families is: does the expected gain from a program exceed the cost of qualifying and applying for it? The Zebley decision changed the cost of applying for the program. Until 1989, only 0.2 percent of children applied for SSI; that quadrupled in the next four years, before falling back slightly.

After Zebley, states had an incentive to shift families from AFDC to SSI, which was federally funded. Whether SSI or welfare was more attractive for families depended on individual state rules. AFDC/TANF benefits varied more widely between states than SSI. After the late 1990s and Zebley, the increase in adoption in SSI in Mississippi and Alabama soared above the rate in states with higher welfare provisions, such as California.

Melissa S. Kearney
Brookings Institution

Households on SSI see their income rise by $500 per month, even though there is some offset of TANF and food-stamp income. For every dollar transfer, 91 cents is not offset by a fall in other income. There is also no evidence that this leads to reduced labor supply. It is therefore a very effective program. Overall, there is a 20 percent increase in income when a family qualifies for SSI.

Does this increase poverty? Since 38 percent lived in poverty ex ante, there is an 11 percent chance that SSI moves the family above the poverty line. For every 100 children eligible for SSI, twenty-five fewer are lifted out of poverty. For every 100 children, forty-five are lifted out of severe poverty.

The level of enrollment in health care does increase, but since many were already enrolled in Medicaid, this increase is not substantial. Nor is there any change in family structure. For households previously living on welfare, the probability of living in poverty reduced as they were shifted to the SSI program.

From 1990-2003, roughly 650,000 children were added to SSI. As a result 161,000 fewer children were in poverty, and 197,000 fewer were in severe poverty. SSI increases total household income by around $5,500 per year.

While welfare was never generous enough to lift families above the poverty line, SSI is more generous; and by federalizing the issue, SSI equalizes the benefits to families in poorer states. But, the potential of SSI is restricted to families with disabled children.

Further research could investigate what families do with this extra money, and whether it improves actual social outcomes.

Richard V. Burkhauser
Cornell University

The U.S census does not track employment and the well-being of people with disabilities. To my knowledge, this is the first paper to focus on the economic well-being of children with disabilities.

Their main finding is that SSI enduringly reduces poverty. I would rather ask whether the SSI enduringly helps children. We should ask: what effects are we looking for over the life cycle of kids, and what are the social expectations of what people with disabilities should grow up to be?

There has been a dramatic change of how we treat people--particularly those with disabilities who have a work limitation. A disability is an impairment that effects the ability to do something (such as hear or see), but may not necessarily prevent someone from working. The social environment, which makes it possible to work or not, essentially determines the extent to which disabled people can be integrated into the workforce and the broader society. I am not sure whether the expansion of the SSI program achieves this.

To understand the SSI program, you must understand where it came from. In the 1960s and 1970s, economists as different as Milton Friedman and James Tobin agreed on the proposal for a Negative Income Tax (NIT). But the measure was defeated, since Congress felt that the nation was not ready for a program that did not distinguish between the deserving and undeserving poor. The “deserving poor” were the aged and disabled. Regarding single mothers with children and adults with disabled children, society was not sure whether they wanted them to work or not.

We then got a program for parents with disabled kids--SSI for those not expected to work. For disabled adults, we limited eligibility to the poor. For children, there were only medical criteria. Over the nearly twenty years since the 1989 court decision, there has been a major change in expectations regarding single women who have been moved from deserving to undeserving poor.

The EITC was the alternative to the NIT and was contingent on work. This had a major impact on single mothers with kids, integrating them more into the workforce. We had a major increase in the expectation that parents of disabled kids were not expected to work. If we can stretch the definition of a disability, this could become a disability. SSI offers money to states to shift people from TANF to SSI. This provides higher benefits to parents who can get their kids certified as disabled.

Over the age of sixty-five, for people under the poverty line, take-up rates for SSI have fallen from 78.5 percent in 1974 to 56 percent in 2003. For young adults aged eighteen to sixty-four, it has risen from 14.8 percent to 20.3 percent, and for those aged up to seventeen, it has increased 0.7 percent to 7.5 percent.

In 1989, 34 percent of single mothers with kids did not work, but in 2003 it was 23.2 percent. This increase in employment of single mothers is the result of welfare reform. Poverty rates for single mothers working are around 20 percent, whereas for those not working it is around 70 to 80 percent.

It is more interesting to look at people who were on welfare and move to be expected to work, rather than to focus on those who were on benefits before as well as after. The median income and relative poverty of those with disabilities has increased. We have made it possible for those with disabilities to get some money, but we have pulled them out of the stream that benefits from economic growth--which is the employed.

Did poverty rates of single mothers with kids fall more than for single mothers with disabilities? In the long run we are concerned about the kids. What do these transfers do to encourage or discourage families to make it in the real world after school? What happens when these kids reach the age of eighteen? Do they simply go on to the SSI rolls for the rest of their lives?

There are significant data limitations. We cannot tell what happens to families that apply but do not receive payments. The idea may be that if you are on SSI, you are there forever. We should see whether this is true.

Mark V. Nadel
Georgetown University

Duggan and Kearney find that SSI does not reduce work effort but increases income and reduces poverty. Other studies find that work effort has already been reduced because of the disability of the child. Some see this reduction as $80 per month of earnings foregone--considerably less than the SSI benefit.

What happens at eighteen when these benefits are terminated? RAND finds an increase in income and work after a reduction in child benefits. SSI is a generous grant program relative to TANF in most states, since it is indexed.

Only the most severely disabled get SSI, which has reduced the poverty rate for kids qualifying. But what about other poor families?

Poor families are very likely to have a member with a disability, although this is often not of the sort serious enough to qualify for SSI. This may be more defensible if there is a rational distinction between the deserving and undeserving poor. There was no lasting impact of congressional attempts to tighten standards.

There is a strong correlation between having a college graduate parent and receiving SSI, which highlights the importance of awareness in gaining receipt. Getting on SSI and disability benefits is not as clear cut as it should be, and it raises substantial equity issues.

SSI kids in general have poor outcomes. They have problems in school (31 percent have been suspended in the past twelve months), many drop out of high schools, and 26 percent have been arrested. Is this due to poverty or disability? Can we disentangle the problems of poverty from the problems of disability?

We have a problem of income support which helps today but does not improve the lifetime prospects of SSI kids. It also begs the question: why not try to reduce poverty for all kids, not just those who qualify for SSI?

AEI research assistant Chris Pope prepared this summary.