Treatment of Married Couples in the SSI Program

by
Issue Paper No. 2003-01 (released December 2003)

The Supplemental Security Income (SSI) program serves as an income source of last resort for elderly or disabled individuals. SSI eligibility is restricted to people with limited income and resources, and recipients' countable income reduces their benefit amount. In establishing eligibility and the benefit amount, SSI also considers the financial resources of certain people associated with the recipient, such as spouses of SSI recipients. Marital status, therefore, can be an important factor in determining eligibility and in calculating the amount of the benefit.

Benefits for a married couple, both of whom receive SSI and have no other income, amount to 25 percent less than the total they would receive if they were living together but not as husband and wife. This analysis identifies how marital status affects benefit rates and the counting of income and resources in determining eligibility. The comparisons made between married couples and two adults living together suggest that the rules provide a financial advantage for a man and a woman who live together but are not married. The paper also presents options for making the program more neutral toward marital status.


This paper was prepared by Richard Balkus and Susan Wilschke, Office of Policy, Office of Disability and Income Assistance Policy. Questions about the analysis should be directed to them at 202-358-6012.

The findings and conclusions presented in this paper are those of the authors and do not necessarily represent the views of the Social Security Administration.

Summary

The treatment of marriage is a frequent consideration in the discussion of government benefit policies. In the Supplemental Security Income (SSI) program, for example, two recipients married to each other receive a benefit that is one-quarter less than if they simply lived together but not as husband and wife. The treatment of marriage has been an issue in other means-tested programs as well. For example, legislation passed in 2001 reduced the marriage penalties identified with the earned income tax credit (EITC), an income supplement for low-income workers. Within that context, this paper examines SSI policy toward marital status.

Although each member of an SSI married couple is guaranteed an income level equal to only 75 percent of the federal benefit rate, they are generally financially better off than SSI individuals living alone. This comparison reflects the economies of scale from sharing living expenses as well as higher incomes. However, members of the opposite sex who cohabitate and do not marry (or are not found to be representing themselves as husband and wife) are each guaranteed an income level equal to 100 percent of the federal benefit rate and generally fare better financially than SSI married couples.

This paper identifies how marital status affects benefits and provides options for making the program more neutral toward marital status. The options include changes to three aspects of the SSI program: the benefit rate, income and resource exclusions, and counting spousal income and resources.

Benefit Rate Options

The first set of options addresses issues associated with the benefits of couples relative to the benefits of two individuals. These options are mutually exclusive and are collectively referred to as the benefit rate options.

The first three options reflect different approaches to setting SSI benefits for married couples relative to other recipients and would make the program more neutral toward marital status. Option 1 would significantly raise program costs and would widen the gap in poverty status between SSI recipients who live alone and those who live with other adults, including other SSI recipients. Options 2 and 3 recognize the economies of scale from sharing living expenses. Option 2 would also simplify the complex rules that now exist regarding living arrangements and in-kind support and maintenance.

Option 4 would limit the FBR for eligible couples to only married couples. Also, for two unmarried persons living together, there would be no consideration of the ineligible person's income and resources in determining the other person's eligibility and benefit amount. Therefore, this option may be viewed as providing a financial gain for couples who do not marry. However, it would result in all unmarried couples being treated in the same way. It would also reduce the amount of information that must be collected and would simplify program rules.

Exclusion and Deeming Options

The paper also examines two sets of rules on what income and resources are counted in determining SSI eligibility and benefits—those governing the income and resources that are excluded from being counted and those for counting the income from the spouse not getting SSI, a process known as "deeming"—and options for changing them. Unlike the benefit rate options, the options for changing the exclusions and deeming are not mutually exclusive.

Five options would change the rules for excluding income and resources. The program costs of each of the options for current beneficiaries would not be significant.

Two options would change the deeming of income from an ineligible spouse. The two options are not mutually exclusive and could be combined. Like the exclusion options, the deeming options would result in more comparable treatment between married couples (or couples representing themselves as husband and wife) and single adults who live together.

Introduction

Do marriage penalties exist in the Supplemental Security Income (SSI) program? Are the benefit rates and the rules for counting income and resources neutral in regard to the treatment of marital status? Are single persons and couples treated fairly relative to each other? Questions about the existence of marriage penalties in social assistance programs have surfaced more frequently in recent years. The Economic Growth and Tax Relief Reconciliation Act of 2001 (P.L. 107-16) reduced the marriage penalties associated with the earned income tax credit (EITC), an income supplement for low-income workers. Proposed legislation for reauthorizing welfare reform would require states' plans for the Temporary Assistance for Needy Families (TANF) program to describe their efforts to "encourage equitable treatment of married, 2-parent families."1

This paper examines the marriage policies of the SSI program, which is a larger cash assistance program than TANF. First, it explains how the program considers marital status in determining eligibility for and the amount of SSI benefits. Second, the paper analyzes the policies to determine whether marriage grants any financial advantages or disadvantages to recipients. For example, how do married couples fare compared with other two-person households? Third, to the extent that differences are identified, it offers options for making the program more neutral toward marital status. It should be noted, however, that the options are not intended to make marital status irrelevant to determining eligibility or benefit amounts. Assuming some financial responsibility between spouses is consistent with SSI's status as a means-tested program.

From the options presented, two approaches emerge for policymakers. One is to raise the payments made to eligible couples so that they are on the same level as those made to individual recipients. The other approach is to extend the argument used to justify paying couples a lower rate to other recipients living with adults and assume that those recipients also benefit from the economies of scale by sharing household expenses.

Marital Status in SSI

The Social Security Act defines the rules for determining marital relationships for SSI recipients. Appropriate state law is applied in determining whether a man and a woman are married, except that if a man and woman have been considered as husband and wife for purposes of Social Security benefits, they are also considered married for purposes of SSI.2

The act also requires that if a man and a woman are found to be "holding out"—that is, presenting themselves to the community as husband and wife—they should be considered married for purposes of the SSI program. An example of such a relationship is one in which the couple are not legally married but consider themselves as being in a common-law marriage. If a member of the couple denies holding out but evidence exists to the contrary, both individuals must complete a questionnaire gathering information about bills, mail, and housing arrangements. Some advocates regard this procedure as administratively burdensome and as infringing on personal privacy. Such concerns have resulted in recommendations to eliminate the concept of holding out and to treat as spouses only those individuals who are legally married (see, for example, Social Security Administration 1992).

Just 24 percent of SSI recipients age 18 or older (1.5 million) are married, compared with 57 percent of all adults in the United States (see Table 1). Approximately 38 percent of married recipients are members of eligible couples (both spouses are entitled to SSI), and the rest have ineligible spouses. The proportion of eligible couples has remained relatively steady over the past quarter century.

Table 1. Marital status of U.S. adult population and SSI adult recipients, by age (in percent)
Marital status 18 or older 18–64 65 or older
U.S. SSI U.S. SSI U.S. SSI
Married 57 24 58 21 55 32
Widowed 7 16 2 6 32 37
Divorced or separated 13 24 13 24 8 21
Never married 23 36 27 47 4 11
SOURCES: For U.S. adult population data: U.S. Census Bureau, Detailed Tables, PCT7, Sex by marital status by age for the population 15 years or older, available at http://factfinder.census.gov/servlet/DTTable?geo_id=01000US&ds_name= DEC_2000_SF3_U&mt_name=DEC_2000_SF3_U_PCT007&_lang=en&_sse=on. For Supplemental Security Income (SSI) recipient data: Social Security Administration, SSI Annual Statistical Report, 2001, Table 33.

Effect of Marital Status on Federal and State Benefit Rates

The amount of a recipient's SSI benefit is based on many factors, including one's marital status. Of the three major income assistance programs, only the SSI program distinguishes marital status. TANF and the Food Stamp program base benefits on household size, not marital status. The distinction made in the SSI program somewhat parallels that of the EITC. (See the appendix for a discussion about the effects of marital status on means-tested programs and the EITC.) The following discussion reviews the different benefit rates for determining the amount of one's benefit and highlights the differences based on marital status.

The Social Security Administration (SSA) uses separate rules for computing benefits for couples and individual recipients. The federal benefit (FBR) rate for a couple ($829 a month for 2003) equals 1.5 times the FBR for an individual. The rationale for paying couples less than the amount that two individual beneficiaries would receive is one of economies of scale. By living together and pooling resources, a couple can live more economically than if each person lived alone.

To determine the amount of SSI benefits a couple is eligible to receive, their combined countable income is deducted from the FBR for a couple. The result is then divided equally and paid to the couple in separate checks. If the couple is determined to be living in someone else's household and receiving in-kind support and maintenance (food and shelter) from within the household, the couple FBR is reduced by one-third.

Forty-four states supplement federal SSI payments, and, in some of them, policies regarding the treatment of marriage for supplemental payments differ from the federal policy (Social Security Administration 2001a). For example, in Colorado, the total payment level available for an eligible couple living independently is equal to two times the total payment level for an individual, thereby making up the difference in federal benefit rates for an individual and a couple. In California, the comparable payment level for an eligible couple is about 1.8 times the total payment level for an individual. In other states, the supplemental payment for an eligible individual with an ineligible spouse is considerably higher than that for an eligible individual with no spouse. For example, in Alaska the adjustment almost equals the federal benefit amount for an individual. In New Jersey, the supplemental payment ensures that an individual with an ineligible spouse is at the same payment level as an eligible couple.

Impact and Policy Implications

Both the poverty threshold and the FBR assume that a couple is better off financially than two individuals with the same total income living alone, but they make different assumptions about the size of the economies of scale. The poverty threshold in 2001 was 28 percent higher for a couple than for an individual—$11,569 versus $9,039—whereas the FBR was 50 percent higher. As a result, the FBR for an individual was 70 percent of the poverty threshold, while the FBR for a couple equaled 83 percent of the poverty threshold for a two-adult family.

Although the benefit rate for couples is based on the assumption that a married couple economizes on living expenses, other recipients, whether or not related, might also choose to live together to economize. This situation raises the question of how married couples should be treated compared with other multirecipient households. Multirecipient households that comprise only married couples represent about 30 percent of all multirecipient households. Table 2 breaks down the number and percentage of multirecipient households by the size of the household. Overall, almost 1.9 million recipients, or 30 percent of all SSI recipients not living in an institution or a board and care facility, live in a multirecipient household: 30 percent of those 65 or older, 28 percent of those aged 18 to 64, and 38 percent of those under 18.3

Table 2. Percentage distribution of SSI recipients in multirecipient households, by age
Household composition Total Under 18 18–64 65 or
older
Married-couple recipients only 30 . . . 22 60
Multirecipient noncouple household 65 98 73 32
Two recipients 50 74 56 27
Three recipients 11 18 12 6
Four to seven recipients 4 5 6 0
Multirecipient household with married couple and other recipients 5 2 4 7
Three recipients 4 1 3 7
Four to five recipients 1 2 1 1
SOURCE: Authors' calculations based on data set used for Koenig and Rupp (2002).
NOTES: Data exclude households identified in the SIPP panel as noninstitutionalized group quarters.
. . . = not applicable.

When several individuals in one family (excluding married couples) receive SSI, each member is eligible for the full FBR minus any countable income. Since SSI benefits are not reduced for each subsequent eligible family member, it is possible for total family income from SSI to exceed the poverty threshold. For example, if both members of a family of two (not a couple) receive the maximum federal benefit, the SSI payments to the family would equal 110 percent of the applicable poverty threshold, compared with the maximum couple benefit, which is 83 percent of the poverty threshold. Comparing the poverty rates for couple versus noncouple families with two SSI recipients shows the advantage that many noncouple multirecipient families have over couples. Chart 1 shows poverty rates for SSI recipients, broken down by size of family and number of recipients in the family. The poverty rate for a married couple receiving SSI is 45.1 percent compared with 9.8 percent for two SSI recipients who are not a married couple (Koenig and Rupp 2002, Table 10).

Chart 1.
Percentage of SSI recipients in poverty, by family size and number of recipients
Two bar charts with text description below.
Show text description

Text description for Chart 1.
Percentage of SSI recipients in poverty, by family size and number of recipients

This bar graph shows two panels, one labeled "one SSI recipient" and the other labeled "two or more SSI recipients." In both panels, the horizontal axis is labeled family size. In the "one SSI recipient panel," family size is one person, two persons, and three persons. In the panel for "two or more SSI recipients," family size is two persons married, two persons unmarried, and three persons. In both panels, the vertical axis is labeled percent, from zero to 90 in increments of 10.

In families with one SSI recipient, about 80 percent of SSI recipients in one-person families are in poverty, about 35 percent of SSI recipients in two-person families are in poverty, and about 28 percent of SSI recipients in three-person families are in poverty.

In families with two or more SSI recipients, about 44 percent of SSI recipients are in two-person married families are in poverty, about 10 percent of SSI recipients are in two-person not married families are in poverty, and about 28 percent of SSI recipients in three-person families are in poverty.

SOURCE: Data in Koenig and Rupp (2002).

The overall poverty rates for the households in which SSI recipients live suggests that the benefits of living with another individual extend beyond households with more than one SSI recipient. Annual poverty measures based on Survey of Income and Program Participation (SIPP) data matched to SSA administrative data indicate that the poverty rate for SSI recipients living alone is 78 percent. For two-person families with one SSI recipient it is 36 percent; for three-person families, 28 percent; and for four-person families, 27 percent.4 Recipients living with a spouse, other relatives, and even nonrelatives benefit from living with others. SSI considers the income and contributions of other household members only in certain situations: spouses and parents (in the case of children) of SSI recipients and when someone in the household provides in-kind support and maintenance to the recipient.

Having a benefit rate for married couples that is lower than that for two singles can provide incentives for beneficiaries to misreport their living arrangements. To receive higher benefits, couples may say they have separated when, in fact, they are still living together. Unmarried persons who are living together may argue that they are not presenting themselves to the community as a couple. Some representatives coach their clients not to give the appearance that they are living as husband and wife.5 In fiscal year 2000, an estimated $26 million in overpayments was attributed to issues of reporting of marital status, confirming the difficulty in determining the marital status of two persons living together.

Options

The following options address policy issues associated with the current couple rate. The first option would make marital status irrelevant to determining eligibility for and the amount of the monthly benefits for two persons living in the same household and applying for benefits. For Options 2 and 3, marital status would not be material in determining benefit amounts for two married SSI recipients but would still be relevant for resource determinations. Marital status would also be relevant for a recipient living with his or her ineligible spouse, since spousal deeming of income and resources would continue to apply. The fourth option would not neutralize the role of marital status, but it would address issues associated with the policy of applying the FBR for couples to unmarried persons living together.

Benefit Rate Option 1: Eliminate the couple rate and treat married SSI recipients as individuals. Under this option, each member of an SSI couple would be entitled to the individual FBR, and income for each spouse would be calculated separately. Past studies of the program have neither recommended eliminating the couple as an eligible unit nor considered doing so a viable option. In addition to the cost factor (approximately $900 million annually for current married couple recipients alone), a major reason for lack of support for such a proposal seems to be the economies-of-scale argument supported by the comparisons between poverty measures for eligible couples and individuals. For example, eliminating the couple rate would increase the maximum benefit rate for a married couple from 83 percent to 110 percent of the poverty threshold, thus creating a wider disparity with the maximum benefit rate for an individual, which represents 70 percent of the poverty threshold. However, that disparity already exists with regard to two recipients who are living together but are not married. The existing wider disparity between married couples and other two-adult recipient households provides a disincentive to marriage and an incentive for two adults not to be found holding out to the community as husband and wife.

Benefit Rate Option 2: Eliminate the current rules for determining living arrangements and in-kind support and maintenance. Reduce the FBR for all individuals living with another adult by a set percentage. Adjust the couple rate to equal two times the reduced FBR. This option would extend the economies-of-scale assumption to all recipients living with another adult. It follows the rationale that all people living together and sharing household expenses can live more efficiently than people living alone. This option was included in the report Simplifying the Supplemental Security Income Program: Challenges and Opportunities (Social Security Administration 2000) as a way to address the complexity of current policies on living arrangements and in-kind support and maintenance. It was also discussed in the 1998 SSI Legislation Workgroup report. Although this change was not specifically intended to address the issue of the couple rate, it would eliminate the differential treatment between couples and other recipients who live together.

Such a change would significantly simplify the current policies on living arrangement and in-kind support and maintenance, which are frequently criticized as among the most complicated policies in the SSI program. SSA would not need to gather information on household expenses, marital status, or whether a couple was holding out. SSA would only need to determine whether the individual was living alone or with another adult.

Under a 5 percent FBR reduction, for example, just over 1 million current recipients would receive higher benefits.6 Couples would be eligible for 190 percent of the FBR instead of 150 percent. Individuals who are subject to the one-third reduction or currently receive in-kind support and maintenance also would receive higher benefits. Nearly half of all recipients (2.7 million) would have their benefits reduced.

Under a 10 percent FBR reduction, approximately 307,000 recipients whose countable income, such as Social Security, brings them close to the income limit would become ineligible. A 10 percent reduction with payment protection—that is, no reduction for the first 3 years for all individuals awarded benefits before the implementation date—would increase program costs by about $400 million over 5 years but would save more than $1 billion over 10 years. A more costly alternative would be permanent payment protection for the approximately 2.7 million current recipients who would be subject to the reduction.

A variation of this option would be to exclude children from the benefit reduction. An argument made for not including children is that considering the parents' income in determining the amount of the child's benefit recognizes that parents have financial responsibility for their children and that their children's benefits are, therefore, already adjusted for the economic support the parent provides. An additional argument is that the economies-of-scale rationale does not work for households that include a disabled child. Disabled children have special financial needs that negate any savings assumed from the economies of scale.

Benefit Rate Option 3: Impose a limit on payments to all SSI recipients who live in multirecipient households. This option would extend the economies-of-scale rationale to all SSI recipients living together. It would be neutral toward marriage by applying the current couple rate, or a similar sliding scale, to all households that have multiple recipients. At least two variations of this option have been discussed in previous studies: applying the sliding scale only to families with multiple child recipients, or to all households with multiple recipients, related or unrelated.

The 1996 report of the Committee on Childhood Disability of the Disability Policy Panel recommended reducing benefits to families with multiple child recipients using a sliding scale. In 2003, the Congressional Budget Office (CBO) updated a sliding scale for this option. For a family with no countable income, the first child would receive $552 (the FBR), the second would receive $333 (40 percent less), and the third would receive $291 (47 percent less). Benefits would continue to decrease for additional children. In making its estimate, CBO assumed that implementation would occur in fiscal year 2005, with estimated savings of $70 million in that year and of $1.345 billion from 2005 to 2013.

Supporters of such a proposal argue that reducing cash benefits for children would account for economies of scale but that each child would still retain full Medicaid coverage, which is sometimes more valuable to families than the cash benefits. Opponents argue that some children with disabilities have unique needs that may not be covered by Medicaid and that families with reduced SSI payments may not be able to meet those needs.

In 1998, the Subcommittee on Human Resources of the House Ways and Means Committee considered legislation that would have reduced benefits for both multiple unrelated recipients living together and multiple child beneficiaries. Under this proposal, two recipients living together would be eligible to receive 150 percent of the FBR (as is currently the case for couples); three recipients, 210 percent; and four recipients, 260 percent. Five or more recipients living together would each receive 60 percent of the FBR. In opposition to the proposal, disability advocates argued that many persons with disabilities, especially mental disabilities, live in group homes offering transitional, supported living situations and would suffer from the loss of benefits, which are used to provide services to residents. The proposal was never submitted to the full committee.

Koenig and Rupp (2002) simulated the effect of providing each recipient in a multirecipient household with a benefit equal to 75 percent of the FBR. They estimated a 19 percent reduction in the number of beneficiaries in noncouple multirecipient households (mostly those individuals whose other income results in a low SSI payment), a 34.4 percent reduction in aggregate benefits for those households, and a 2.2 percent reduction in aggregate benefits for the overall SSI population.

Benefit Rate Option 4: Eliminate the concept of treating as a married couple unmarried persons who represent themselves to the community as husband and wife (the concept of "holding out"). Previous proposals to simplify the SSI program have included this option. The 1992 SSA report Supplemental Security Income Modernization Project: Final Report of the Experts found "the SSI 'holding out' provision as having adverse and disturbing effects both with respect to claimants' personal privacy and to the administrative process." The majority of the experts recommended its elimination. The estimated cost of the proposal was $16 million for the 5-year period 1993–1997. An agency workgroup charged with evaluating proposals for simplifying the program had made the same recommendation 2 years earlier. SSA rejected the recommendation, citing the concern that such a change could be viewed as being antifamily. For example, not being considered as part of a couple could disadvantage some individuals. Those individuals could be considered as living in the household of another and receiving in-kind support and maintenance, resulting in a reduction of benefits. Some people would also view the proposal as creating a more focused and significant marriage penalty by defining married couples only as a man and woman who are legally married. Unmarried persons who are both eligible and holding out as husband and wife would no longer be subject to the couple rate and would be financially better off than if they were married. For unmarried persons with only one member eligible (or potentially eligible), the income and resources of the ineligible person would no longer be considered in determining the other person's eligibility or benefit amount. The elimination of the "holding out" provision, however, would result in all unmarried couples receiving the same treatment. Eliminating it would also reduce the amount of information that must be collected and would simplify program rules.

Effect of Marital Status on Treatment of Income and Resources

The previous section illustrates how marital status determines whether a person receives benefits under the individual or couple rate. Marital status also affects how income and resources are counted in determining a person's SSI eligibility and monthly benefit amount. For eligible couples, some rules for excluding income and resources treat the couple as a unit. Therefore, if two eligible persons married or represented themselves as husband and wife, they would lose the benefit of two separate exclusions. For couples with one member ineligible, the rules require that the income and resources of the ineligible spouse be considered in determining the other spouse's eligibility for and monthly amount of SSI benefits. If an eligible individual lives with another person and they are not married or are not representing themselves as husband or wife, the eligible individual's SSI benefit is determined without considering the income from the other person.

The following analysis reviews the rules for several exclusions (four income and one resource) and the rules for considering the income from an ineligible spouse. The analysis points out how the rules differ in their treatment of married couples or persons representing themselves as two single persons living in the same household. Although the benefit rate options discussed above are mutually exclusive alternatives, the exclusion and deeming options presented below could be implemented in combination with one another and with one of the benefit rate options.

General Income, Earned Income, and Infrequent and Irregular Income Exclusions

The SSI program allows a number of exclusions from income in determining eligibility and one's monthly benefit amount. The rules for the three most frequently used income exclusions are not marriage neutral and result in what may be perceived as marriage penalties.

The most common income exclusion is the general income exclusion. The first $20 of unearned income of an eligible individual or eligible couple, other than unearned income from other federally funded means-tested programs (primarily veterans' pension benefits and TANF family grants), is excluded from countable income. Any portion of the exclusion not used for unearned income is used for earned income. A married couple (eligible individual and eligible or ineligible spouse) is entitled to only one $20 exclusion per month regardless of whether one or both members have income.

The second most common exclusion is the earned income exclusion. The first $65 of earned income plus one-half of the remainder is excluded from the eligible individual's or eligible couple's countable earned income. Like the general income exclusion, a married couple is entitled to only one $65 exclusion per month regardless of whether both members of the couple have earned income. Therefore, two working members of a couple would receive $32.50 less in total income than would two unrelated eligible individuals who have earned income and live in the same household.

The third exclusion—the infrequent and irregular income exclusion—also treats couples as a unit, and the dollar amount does not increase even if both an eligible individual and spouse (eligible or ineligible) have infrequent and irregular income. This exclusion provides for not counting income that is received either infrequently or irregularly and does not exceed $10 per month of earned income or $20 per month of unearned income. The law defines "infrequently" as not receiving the income more than once in a single quarter from a single source and "irregularly" as not reasonably expecting to receive the income.

Impact and Policy Implications. The underlying reasons for the three types of exclusions are consistent with extending the exclusions to both members of the couple. The legislative intent of the general income exclusion was to reward SSI beneficiaries who had previously worked and have monthly income from Social Security benefits. These beneficiaries would receive $20 more in total income than those who had not qualified for a Social Security benefit or other work-related pension. However, the current exclusion does not provide any additional reward for a couple in which both members have worked and are receiving Social Security benefits.

The legislative intent of the earned income exclusion was to encourage beneficiaries to work and obtain economic self-sufficiency. However, if the eligible individual is working, any earned income from the eligible spouse is not subject to a second $65 exclusion. The earnings are subject only to the second part of the exclusion, which disregards one-half of the combined earnings above $65 per month.

Mindful of difficulties in administering a means-tested program, Congress saw the infrequent and irregular exclusion as one way to simplify administration of the SSI program. The legislative intent was to exclude small amounts of income. The current exclusion allows SSA to ignore an eligible individual's receipt of a birthday gift of $20 or earnings of $10 for an odd job. However, the exclusion does not apply to each member of a couple. Consider, for example, an aged couple, each receiving a $20 gift from their son in December. Since the total exceeds $20 and both gifts are from the same source, the full $40 represents countable income. However, two elderly siblings receiving SSI benefits and living in the same household who receive a $20 gift each from their nephew in December would each benefit from the infrequent and irregular exclusion.

Exclusion Options. The following options would extend the above exclusions to both members of an eligible couple and would have a small impact on the guaranteed income level a couple could receive. For example, the guaranteed income level for an SSI eligible couple with each member receiving only unearned income would go from 85 percent to 87 percent of the poverty threshold.

Exclusion Option 1: Give each member of an eligible couple a separate general income exclusion. Both members of a couple would be able to take full advantage of the general income exclusion. In cases in which both members had income, the total benefit for the couple would increase by $20 per month. In June 2002, there were about 90,000 couples with each member having some type of unearned income. Therefore, under this proposal, program costs for current beneficiaries would increase annually by about $20 million. Actual costs could be somewhat higher because more people would be financially eligible for the program.

Exclusion Option 2: Give each member of an eligible couple a separate earned income exclusion. Both members of a couple would be able to take full advantage of the earned income exclusion. If both members had earned income, the total benefits for the couple would increase by $32.50 per month. In June 2002, there were approximately 660 couples in which each member was receiving either wages or self-employment income. The annual cost of this change for current beneficiaries, therefore, would be about $250,000.

Exclusion Option 3: Give each member of an eligible couple a separate infrequent and irregular income exclusion. Eligible couples would have more opportunity to exclude small amounts of income. The annual program cost of such a change would be minimal. For example, fewer than 4,300 members of eligible couples had "other" income in December 2000. That means that a much smaller number of couples would have income that would be excluded under the current definition of infrequent and irregular income if the exclusion were extended to both members of the couple.

Student Earned Income Exclusion

Under current law, a portion of the earned income of a child receiving SSI who SSA determines to be a student is excluded when determining his or her benefit. In 2003, a student can exclude up to $1,340 per month (not to exceed $5,410 per year). The act defines a child as an unmarried individual who is not head of a household and is under age 18, or is under age 22 and a student regularly attending school in preparation for gainful employment. Therefore, students who are married do not qualify for the exclusion.

Impact and Policy Implications. This exclusion is one of several work incentives designed to maximize one's ability to work to achieve economic self-sufficiency. The same purpose would seem applicable to a married student.

Exclusion Option 4: Eliminate the marriage restriction for the student earned income exclusion. This option was included in H.R. 743, The Social Security Protection Act, as reported by the Senate Finance Committee on September 17, 2003. That proposal would allow the student earned income exclusion to apply for any individual who is a student under age 22 by removing the term "child" from the statutory language for the exclusion, thereby eliminating the marriage and head-of-household restrictions.

Life Insurance Exclusion

The SSI program includes several rules for excluding different types of resources in determining eligibility based on countable resources, which cannot exceed $2,000 for an individual and $3,000 for a couple. Some rules apply to each member of a married couple and, therefore, would apply in the same way as for two unmarried eligible individuals living in the same household. For example, each individual is allowed to exclude as countable resources one wedding ring and one engagement ring. Also, each member of a couple is allowed to set aside up to $1,500 in burial funds and to exclude items associated with the burial space exclusion such as the casket, headstone, and gravesite. Other rules for excluding a resource treat a married couple as a unit. For example, an eligible couple, like an eligible individual, can only exclude household goods and personal effects with a total value of up to $2,000. Also, the rules for excluding an automobile treat members of a couple as a unit, and couples with more than one automobile can generally exclude only one automobile.

The rules for excluding life insurance policies treat a married couple as a unit. Life insurance policies are excluded if the total face value of all policies on a single person does not exceed $1,500. If it exceeds $1,500, the cash-surrender value of the policy or policies involved counts as a resource. Life insurance policies owned by one spouse are considered to be an available resource of the other spouse.

Impact and Policy Implications. Although the rule does allow each member of a couple to have more than one life insurance policy, it limits the face value of policies owned by both members on any insured person to $1,500. For example, a married couple is allowed to own two life insurance policies—one with each spouse as the insured, each with a face value not exceeding $1,500. However, two adults who are living in the same household but are neither married nor found to be holding out are each able to own two policies—one for each adult as the insured, each with a face value not exceeding $1,500, for a total of four policies.

Exclusion Option 5: Expand the life insurance exclusion by treating both members of a couple as individuals. Under this option, both members of a couple would be able to take full advantage of the exclusion, and each member would be able to own two life insurance policies—one for each member as the insured. If none of the four policies had a face value exceeding $1,500, then all four policies would be excluded.

Income from an Ineligible Spouse

When an SSI recipient is married to an individual who is not receiving SSI, the eligible individual's income and resources are deemed to include the income and resources of the spouse. Instead of determining the exact value of any support and maintenance supplied by the ineligible spouse, deeming rules recognize some measure of family responsibility and support the premise that SSI should pay benefits only to the extent that needs are not met by other sources. SSA also deems the income and resources of parents of SSI applicants and beneficiaries under age 18 and the income and resources of sponsors to noncitizens. The same rules for excluding certain income and resources for applicants and beneficiaries also apply to deeming situations. Although the Social Security Act requires SSA to consider the income of ineligible spouses and parents, the actual rules for deeming income are determined by regulation.

The SSI applicant or beneficiary must qualify on the basis of his or her own income, before any deeming of the ineligible spouse's income is considered. If the ineligible spouse's income is equal to or less than the difference between the couple and individual FBR, there is no income to deem to the eligible individual. The eligible individual's own income is subtracted from the individual FBR to determine the benefit amount.7

If, however, the ineligible spouse's income is greater than the difference between the two benefit rates, the eligible individual and ineligible spouse are treated as an eligible couple for purposes of counting income. The remainder of the ineligible spouse's unearned income is combined with any unearned income of the eligible individual, and the remainder of the spouse's earned income is combined with any earned income of the eligible individual.

The $20 general income exclusion is applied to the combined unearned income. Any remaining portion of the exclusion may be applied to the earned income. Likewise, $65 of earned income and one-half of the remainder is subtracted from the combined earned income. The remaining countable income is subtracted from the FBR for an eligible couple. The SSI benefit is the lesser of this amount or the amount from subtracting only the eligible individual's countable income from the FBR for an individual. An example of how the rules for spousal deeming apply is given in Box 1.

Box 1.
Example of spousal deeming

An eligible individual receives a Social Security benefit of $200 per month, and the ineligible spouse earns a salary of $800 per month. Since the spouse's income is greater than $277 (the difference between the couple and individual federal benefit rates for 2003), spousal deeming applies. The combined countable income for the eligible and ineligible spouse is less than the federal benefit rate for a couple, so the individual is eligible to receive SSI based on deeming. The calculations are illustrated below (in dollars):

Social Security 200.00
General income exclusion − 20.00
Countable unearned income = 180.00
Spouse's salary 800.00
Earned income exclusion − 65.00
Earned income minus exclusion = 735.00
Countable earned income(735.00 ÷ 2) 367.50
Total countable income(180.00 + 367.50) 547.50
Couple federal benefit rate 829.00
Countable income − 547.50
SSI benefit = 281.50
 

Impact and Policy Implications. Counting the income of an ineligible spouse in many cases results in an SSI benefit that is lower than that of a recipient who lives with an unrelated individual, since more income is counted for the married recipient. However, in some cases—typically when the spouse has fairly small amounts of income—deeming of spousal income has no effect. Examples for both situations are given in Box 2.

Box 2.
Examples comparing treatment of income for a married and unmarried couple

Compare two Supplemental Security Income recipients: one is married and living with an ineligible spouse; the other is unmarried and living with a partner (Situation 1). Both the spouse and the partner earn $500 per month, and neither recipient has income other than SSI. The unmarried recipient will receive the individual federal benefit rate (FBR) of $552. The SSI recipient is entitled to the lesser of the amount from the deeming calculation using the couple FBR and the amount from subtracting only the recipient's countable income from the individual FBR. For this example, the lesser amount is $552.

However, when the example is changed so that both the spouse and the partner earn $1,000 per month, the unmarried individual is better off (Situation 2). After deeming, the married recipient receives $371.50 per month, and the unmarried recipient continues to receive $552.00. The calculations are shown below (in dollars):

  Married couple Unmarried couple
Situation 1.
Ineligible spouse or partner has earned income of $500 per month
Apply the general and earned income exclusions 500.00 − 85.00 = 415.00 ÷ 2 = 207.50 Income of unmarried partner does not count toward recipient's SSI benefit
Subtract countable income from appropriate FBR 829.00 − 207.50 = 621.50 552.00 − 0 = 552.00
SSI benefit 552.00 552.00
Total household income 500.00 + 552.00 = 1,052.00 500.00 + 552.00 = 1,052.00
Situation 2.
Ineligible spouse or partner has earned income of $1,000 per month
Apply the general and earned income exclusions 1,000.00 − 85.00 = 915.00 ÷ 2 = 457.50 Income of unmarried partner does not count toward recipient's SSI benefit
Subtract countable income from appropriate FBR 829.00 − 457.50 = 371.50 552.00 − 0 = 552.00
SSI benefit 371.50 552.00
Total household income 1,000.00 + 371.50 = 1,371.50 1,000.00 + 552.00 = 1,552.00
 

It is also useful to compare the rules for spouse-to-spouse deeming and those for parent-to-child deeming, which differ in several ways. First, spouse-to-spouse deeming uses the couple rate described earlier and subtracts any income from either person from that rate, whereas a parent's income is calculated separately from a child's income. One result is that if both parent and child have income, both can apply the appropriate exclusions in determining countable income, whereas a couple has one exclusion. Therefore, the child may be able to exclude more income than an adult recipient who is subject to spousal deeming.

In determining eligibility for children, a parental living allowance is subtracted from the parent's income. The allowance for one parent equals the FBR for an individual, and the allowance for two parents equals the FBR for a couple. The parental living allowance is subtracted from income after applying the earned and unearned income exclusions. In contrast, no specific deduction is made for the ineligible spouse in the spousal deeming calculation. However, the use of the couple FBR instead of the individual FBR provides an implicit allowance for the spouse equal to the difference between the two benefit rates. When an ineligible spouse and an ineligible parent have the same amount of monthly income, the child recipient in some cases may have more income excluded than the married recipient and thus receive a higher benefit because of the parental allowance.

Deeming Options. The two options for changing the rules for deeming are not mutually exclusive and could be combined. The first option is an extension of the options for excluding income discussed above.

Deeming Option 1: Extend the income exclusion options for couples to deeming situations. A recipient and an ineligible spouse would both be able to use the general income exclusion, earned income exclusion, and infrequent and irregular exclusion if both spouses had income. For example, if both the eligible spouse and the ineligible spouse had income from earnings, both of them would be able to exclude the first $65 of their earnings as well as one-half of the remaining earnings. Currently, only one spouse can apply the $65 exclusion to earned income, although half of the remaining earned income for both of them is excluded.

In June 2002, there were approximately 3,500 cases in which both the recipient and the ineligible spouse received earned income and 36,000 cases in which they both received unearned income. The annual cost of this change on the basis of those numbers of current beneficiaries would be approximately $10 million.

Deeming Option 2: Provide a living allowance for the ineligible spouse that is equivalent to the FBR for an individual. Under this option, spousal deeming would be calculated in the same manner as parent-to-child deeming. Like the parental living allowance, the spousal allowance would be subtracted from income after applying exclusions. Countable income would be deducted from the individual FBR rather than the couple FBR. Box 3 compares this option with the current rules for spousal deeming.

Box 3.
Example comparing rules for spousal deeming under current law and the option to provide a living allowance for the ineligible spouse

The ineligible spouse has $1,500 in earned income and the recipient has no income. Under current rules, the recipient is charged with $707.50 in countable income and receives a monthly benefit of $121.50. The option to provide a living allowance equivalent to the individual federal benefit rate would reduce the countable income to $162.50 and the recipient would receive a monthly benefit of $396.50. The calculations are shown below (in dollars):

  Current
rules
With allowance
for ineligible
spouse
Earned income 1,500.00 1,500.00
General income exclusion − 20.00 − 20.00
Earned income exclusion − 65.00 − 65.00
Earned income minus exclusions = 1,415.00 = 1,415.00
Countable earned income(1,415.00 ÷ 2) 707.50 707.50
Spousal living allowance . . . − 552.00
Total countable income = 707.50 = 155.50
Federal benefit rate 829.00 a 552.00 b
Countable income − 707.50 − 155.50
SSI benefit = 121.50 = 396.50
NOTE: . . . = not applicable.
a. Couple federal benefit rate.
b. Individual federal benefit rate.

This option would eliminate differences that currently exist between different deeming arrangements. In most cases, SSI recipients with spouses would have an advantage over recipients without a spouse, since they would be allowed to receive SSI despite the known existence of other income sources in the household.

Concluding Observations

The higher poverty rate for households consisting of a married couple who are both receiving SSI (45.1 percent) compared with those consisting of two nonmarried recipients (9.8 percent) raises a question of benefit equity. The economies-of-scale argument would seem to apply to both types of households and would not seem to give members of either household an advantage in terms of reduced costs for food and shelter or other necessities of daily living. Steuerle (2001, 2) discusses the shortcomings of using economies of scale as a rationale for justifying marriage penalties:

The problem with using this argument to justify marriage penalties is not that there are no economies of scale from sharing. There are, and indeed, these gains reinforce other natural instincts to engage in mutual support. Economies of scale, however, apply to almost all sharing arrangements—dormitories, retirement homes, cohabitation, and so on. Yet marital vows of allegiance are the only type of arrangement that is taxed.

In addressing the way that married couples are treated in the SSI program, policymakers face two approaches. One is to treat eligible couples like individual recipients. This approach would extend the individual federal benefit rate and income and resource exclusions to each member of the couple, thus raising program costs. The other approach is to extend the argument used to justify paying couples a lower rate to all recipients who enjoy economies of scale by sharing household expenses, lowering benefits for many recipients. This approach may result in a savings or cost depending on the rate of reduction and whether other income and resource changes are included.

Several of the benefit rate options would bridge the gap in poverty status between the two types of households. The three options for changing the benefit rate structure would eliminate the rate differences that now exist between eligible couples and other two-recipient households. Benefit Rate Options 1 and 2 would have additional payoffs by simplifying the SSI program and should result in better payment accuracy.

Benefit Rate Option 1, which calls for eliminating the couple rate and treating all recipients as individuals, would create a wider disparity in the income guarantee as a percentage of the poverty threshold between an individual living alone and a married couple. Program costs would significantly increase as couples were made better off. However, the program would also be made simpler to administer by eliminating the need to determine whether an unmarried eligible couple was holding out to the community as husband and wife and by removing the incentive for persons to misrepresent their status as a couple.

Benefit Rate Option 2 would adjust the couple rate to two times the reduced federal benefit rate for an individual living with another adult and would recognize the economies of scale between one- and two-person households. Like the first option, this option would eliminate the need to determine whether an unmarried man and woman living in the same household and potentially eligible for SSI are holding out as husband and wife. This option would also greatly simplify the current set of complex rules for living arrangements and receipt of in-kind support and maintenance. SSA would no longer need to determine household expenses, payment toward these expenses, and the value of any help from outside the household. SSA would only need information on whether the person was living alone or with another adult. The specific reduction in the individual FBR for this option would dictate whether it would produce a savings or a cost.

Benefit Rate Option 3 is a sliding-scale version of the second option for households with two or more recipients. It would also eliminate the current disincentive to marry. Although this option would create parity between married couples and other recipients who live together, it would most affect households with child recipients and could significantly reduce the SSI income for households with several SSI recipients. (A variation of this option would be to exclude children.) Whether Option 3 would result in a savings or a cost (like Option 2) would depend on the percentages used for the sliding scale.

Benefit Rate Option 4 would maintain the current benefit rate structure but would limit the FBR for couples to married couples. This option would only eliminate the concept of holding out, and, therefore, all unmarried couples would be treated in the same way. Of the four options, this option would do the least to simplify the program and would not change the financial disadvantage for married couples that results from the current rate structure.

The other two sets of options would eliminate the disparities that exist in applying income and resource exclusions and in the deeming of income. None of these options would significantly increase program costs—at least for current beneficiaries. These options are not exclusive. They could be implemented together and would complement any benefit rate option that eliminated the couple as a unit for determining benefit amounts. These options would also increase total income; for the many couples who are married or holding out; it would therefore close the gap between the poverty rate for SSI couples and other two-person households.

Appendix: Marital Status in Other Means-Tested Programs and the EITC

In examining the way the Supplemental Security Income (SSI) program treats marital status, it is also useful to consider how other means-tested programs do so. The Temporary Assistance for Needy Families (TANF) and Food Stamp programs, for example, base benefit amounts on household size rather than marital status. The rate for a veteran receiving a pension based on disability varies depending on the number of dependents, which would include the veteran's spouse. The treatment of marital status for the earned income tax credit (EITC) produces outcomes similar to those of the SSI program.

Temporary Assistance for Needy Families

The 1996 Personal Responsibility and Work Opportunity Reconciliation Act, which established the TANF program, states that the underlying purpose of the act is to promote marriage and the formation and maintenance of two-parent families. Nonetheless, marital status is not a criterion for receiving benefits. That is, TANF does not distinguish between married and unmarried couples, but there are some differences in the rules for two-parent and one-parent families. Two-parent families face higher work participation rates, and 17 states (and the District of Columbia) have retained eligibility rules that make it harder for two-parent families to receive assistance. Some states use TANF monies to fund activities that promote marriage. West Virginia, for example, increases a family's monthly benefit when there is a legal marriage and both members of the couple receive public assistance.

Food Stamps

The Food Stamp program pays benefits based on household size. Marriage may affect the household size, but the amount of benefits paid does not vary by marital status. The income and assets of each household member are considered in determining eligibility and benefit amount. Regulations define a household as comprising an individual living alone; an individual living with others but customarily purchasing food and preparing meals for home consumption separate and apart from others; or a group of individuals who live together and customarily purchase food and prepare meals together for home consumption. The rules further state that spouses who live together but purchase and prepare meals separately are still considered members of the same "household" for the purpose of determining food stamp eligibility. According to definitions in the regulations (7 CFR 271.2), spouses are two individuals who either would be defined as married to each other under applicable state law or are living together and are holding themselves out to the community as husband and wife by representing themselves as such to relatives, friends, neighbors, or tradespeople.

Veterans' Pension Program

The payment rates for means-tested pensions for disabled veterans depend on whether the veteran has dependents, such as a spouse, child, or parent. Veterans receive an incremental increase in their benefit for each additional dependent. Surviving spouses of certain deceased veterans may receive dependency and indemnity compensation. The regulations (38 CFR 3.1(j)) define marriage as a marriage valid under the law of the place where the parties resided at the time of marriage, or the law of the place where the parties resided when the right to benefits accrued.

Earned Income Tax Credit

The EITC provides low- and moderate-income families with a credit that can eliminate any income tax and generate a payment to the family. The size of the credit depends on the level of income and the number of children (that is, no children, one child, or two or more children) in the family. Two wage earners with one child and each with very low income may have some incentive to marry. Consider, for example, a single parent with one child and earnings of $4,000 in 2001. That person would be entitled to an EITC of $1,369. The person's companion also earns $4,000 in 2001 and would be entitled to an EITC of $308. If the two persons were married and had filed a joint federal return, their credit would be $2,428, an increase of $751. However, the EITC penalizes married couples with higher earnings. For example, a single parent with two children has earnings of $25,000 in 2001. The EITC for that person would be $1,494. The person's companion earns $10,000 and would receive an EITC of $52. If the couple had married and filed a joint return, their combined income of $35,000 would have made them ineligible for any credit.

For the 2002 tax year, the maximum tax credit started decreasing for a married couple at an income level $1,000 greater than that of a single person (thereby extending the income range for each filing status for married couples by $1,000). The change, therefore, lessens the impact of the current penalty for some married couples. For example, in the 2001 tax year, a one- or two-parent family with two children would not have received a credit if their income had totaled $32,121 or more. For the 2002 tax year, a single parent with two children earning $33,178 or more would not be eligible for a tax credit, but a two-parent family could earn up to $1,000 more before being ineligible for a tax credit. However, the change for the 2002 tax year does not eliminate the marriage penalty for some couples. The two people in the above example with a combined income of $35,000 would still not be eligible for any tax credit if they had married and filed a joint return for 2002.

Notes

1 See section 103(a), Promotion of Family Formation and Healthy Marriage, in H.R. 4, Personal Responsibility, Work, and Family Promotion Act of 2003, as passed by the House on February 13, 2003.

2 The 1996 Defense of Marriage Act (P.L. 104-199) reinforced the definition of marriage for federal programs as "only a legal union between one man and one woman as husband and wife" and further provided that no state would be required to recognize "a relationship between persons of the same sex that is treated as marriage under the laws" of another state.

3 Koenig and Rupp (2002) find that previous studies using SSA's administrative data sources underestimate multi-recipient households. For example, Table 21 of the SSI Annual Statistical Report, 2000 (which is based on the SSI Quality Assurance Stewardship File) indicates that only 19 percent of all SSI recipients live in a multirecipient household.

4 Koenig and Rupp (2002) using data from the 1996 SIPP panel matched to SSA administrative data for March 1996, January 1998, and December 1998.

5 Neighborhood Legal Services of Erie County, New York, offers the following advice: "Unmarried couples who do not want their SSI to be reduced should do everything they can to appear not married." From The Impact of Marriage on Supplemental Security Income.

6 The estimates for the 5 percent reduction are from the report on simplifying the SSI program (Social Security Administration 2000). The estimates for a 10 percent reduction are also from that report; however, the report notes that those estimates are from 1998 for an option that contained different features, including payment protection for the first 3 years of implementation. The Office of Policy is currently analyzing additional options. All options assume budget neutrality, and the rate of reduction varies depending on the groups of recipients that would be subject to benefit restructuring. Some options would treat eligible couples like other SSI recipients who share the same household.

7 An allocation for each ineligible child equal to the difference between the FBR for a couple and the FBR for an individual is subtracted from the ineligible spouse's income. An allocation for eligible aliens who have been sponsored by the spouse is also deducted.

References

Committee on Childhood Disability of the Disability Policy Panel. 1996. Restructuring the SSI Disability Program for Children and Adolescents. Washington, D.C.: National Academy of Social Insurance.

Congressional Budget Office. 2003. Budget Options. Available at http://www.cbo.gov/showdoc.cfm?index=4066&sequence=13.

Koenig, Melissa, and Kalman Rupp. 2002. The Poverty Status of Different Types of Multirecipient Households: Is SSI Fair to Married Couples? Paper presented at the 2002 Annual Meeting of the Population Association of America, Atlanta, Ga.

Social Security Administration. 1992. Supplemental Security Income Modernization Project: Final Report of the Experts. Washington, D.C. August.

———. 1998. "Report of the SSI Legislation Workgroup: Legislative Options for Addressing High Risk." Baltimore, Md.

———. 2000. Simplifying the Supplemental Security Income Program: Challenges and Opportunities. Office of Policy, Office of Disability and Income Assistance Policy. Washington, D.C. December.

———. 2001a. State Assistance Programs for SSI Recipients. Office of Policy, Office of Research, Evaluation, and Statistics. Washington, D.C.

———. 2001b. SSI Annual Statistical Report, 2000. Office of Policy, Office of Research, Evaluation, and Statistics. Washington, D.C.

———. 2002. SSI Annual Statistical Report, 2001. Office of Policy, Office of Research, Evaluation, and Statistics. Washington, D.C.

Steuerle, Eugene. 2001. Testimony before the Subcommittee on Human Resources of the House Committee on Ways and Means, Hearing on Welfare and Marriage Issues. May 22.

Welch, Karen. The Impact of Marriage on Supplemental Security Income (SSI).