Source, Form, and Amount of In-kind Support and Maintenance Received by Supplemental Security Income Applicants and Recipients

Introduction

In January 1974, Congress created the Supplemental Security Income (SSI) program, which provides income of last resort to aged, blind, and disabled persons to help them meet their basic food, clothing, and shelter needs. During December 2012, about 8.3 million persons received SSI payments (SSA 2014a). Because SSI is means tested, the Social Security Administration (SSA) must count all income and support received by an individual, including “in-kind support and maintenance” (ISM), to determine his or her monthly payment amount. SSA defines ISM as unearned income received by SSI applicants and recipients¹ in the form of food and/or shelter from anyone living within or outside their households. To determine the monetary value of ISM, SSA requires applicants and recipients to answer detailed questions about their household members and expenses: how they divide household expenses and what help they get from others within or outside of their households. About 9 percent of SSI recipients have their benefit rates reduced because of ISM during any given year (SSA 2013, Table 8).

ISM policies have several equity, incentive, and administrative issues. Many experts in this field believe that certain ISM policies place some SSI recipients at an economic advantage, while other ISM policies may discourage families from assisting low-income relatives on SSI because such contributions can result in dollar-for-dollar reductions in recipient payment amounts (SSA 2000b; Balkus and others 2009). Those authors and many others have identified ISM policy as one of the leading policies that make the SSI program difficult, time-consuming, and costly to administer (Kennedy 1983; GAO 2002b; SSA 2000a, 2012b). In fact, the Government Accountability Office (GAO) and SSA's Office of the Inspector General have repeatedly identified ISM policy as a leading cause of SSI payment errors (GAO 2002a, 2002b, 2012; SSA 2000a, 2000b, 2012a, 2012b; SSAB 1999, 5). Subsequently, SSA has made ISM simplification a leading priority.

Until recently, limited information has existed to inform ISM policy and its simplification. Available ISM publications have left the following four perennial ISM questions unanswered:

• How many SSI applicants and recipients alleged receiving ISM from within and/or outside of their households?
• What proportion of SSI recipients alleged receiving ISM in the form of food and/or shelter?
• What proportion of SSI recipients alleged obtaining ISM exceeding the amount deducted from their federal benefit rate?
• To what extent did the total ISM alleged vary by ISM source and form, as well as by age group of SSI recipients?

The existing literature does not quantify several facets of ISM because of the absence of detailed ISM research data.

For over 20 years, SSA has used the Modernized Supplemental Security Income Claims System (MSSICS) to administer the SSI program. I used administrative data from the MSSICS to quantify the source, form, and amount of ISM received by SSI recipients. This article provides basic statistics about the ISM that SSI recipients acquire from others. More importantly, it answers the four noted questions so policy and decision makers can make data-driven decisions as they strive to simplify ISM policy and minimize SSI payment errors. Subsequent articles will describe the household composition of recipients who receive ISM and may explore ISM policy options for simplifying the SSI program.

Program Background

The SSI program provides a basic monthly national income guarantee, called the federal benefit rate (FBR), to children and adults with disabilities (including blind persons) as well as the aged (persons 65 or older).

SSI Program Eligibility

To be eligible for SSI, all applicants must meet income and resource requirements.² In addition to the federal SSI payment, some states provide supplemental benefits to their residents (SSA 2013).³

Financial eligibility requirements. The law requires SSA to reduce the monthly SSI FBR dollar for dollar by the amount of the individual's “countable” income—that is, income minus all applicable exclusions. SSI financial eligibility rules require that the countable income (after any applicable exclusions) of applicants and recipients be less than the current FBR plus any available state supplement. In certain situations, SSA also considers the income of other individuals who live with the applicant when determining eligibility for the program. This includes spouses who are not eligible for SSI and parents if the applicant is younger than age 18.

Disability requirements. The disability test for children requires the applicant to have a medically determinable impairment (or a combination of impairments) resulting in “marked and severe functional limitations.” However, the disability test for nonaged adults is the same as that used for the Disability Insurance (DI) program covered under Social Security and requires that the applicant be blind or have a physical or mental impairment that prevents him or her from engaging in any substantial gainful activity (SGA).⁴ The identified impairment must also have lasted or be expected to last for a continuous period of at least 12 months or to result in death. For 2014, the SGA standard is $1,070 per month for nonblind persons and $1,800 per month for blind persons (SSA 2014c). However, the SGA rate for blind individuals is not applicable to SSI recipients, but rather for blind participants in the DI program.

Payment Amounts

SSA uses a “couple”-FBR unit for recipients who live with an eligible spouse and an “individual”-FBR unit for all other recipients to determine the recipient's eligibility and payment amount. For 2014, the monthly individual FBR is $721, while the monthly couple FBR is $1,082 (SSA 2014b). SSA generally adjusts the individual and couple FBRs annually for inflation. Many states augment the federal SSI payment by offering some SSI recipients a state supplemental payment. In sum, a recipient's monthly SSI payment is equal to the applicable FBR plus any applicable state supplement, minus any countable income (after any applicable exclusions).

How does SSA apply ISM to determine payment amounts? The agency must reduce payments if an applicant or a recipient has countable earnings or other income, including ISM.⁵ The rationale for reducing benefits by the value of the ISM received is that persons receiving food and/or shelter assistance need less help fulfilling their basic needs than those without such support.

Current ISM rules require applicants and recipients to answer detailed questions about household composition and expenses as well as the contributions toward household expenses made by themselves and/or those living in the household. SSA collects ISM-related information from recipients during their initial application interview and after a change of address, household composition, or household expenses.

Two phases compose the ISM evaluation process. During the first phase, claims representatives identify which of the four (A, B, C, or D) federal living arrangement (FLA) categories the recipient belongs to (Box 1). SSA categorizes a person as living in FLA-B if he or she resides in another person's household throughout a month and receives both food and shelter from other people living in that household. The agency identifies a recipient as belonging to FLA-C if he or she is either an eligible child younger than age 18 who lives in his or her parent's household or younger than age 22 and is a full-time student. An SSI recipient belongs to FLA-D if he or she resides in a public or private medical institution throughout a month and Medicaid is paying more than 50 percent of the cost of care or, effective December 1996, if he or she is a child younger than age 18 who resides in a medical care facility in which private insurance (or a combination of private insurance and Medicaid) pays more than 50 percent of the cost of care. If an individual belongs to FLA-D, SSA does not count food or shelter from the medical treatment facility as ISM, but the agency may count ISM from other sources. The FLA-A category includes all persons for whom FLA-B, FLA-C, or FLA-D categories do not apply. SSI recipients belonging to FLA-A include the homeless; transients; persons who earmark their contributions for food or shelter; persons who live with others, but separately consume or purchase their food; and those who live in public assistance households. Although all SSI recipients may receive outside ISM, claims representatives must identify an SSI recipient's living arrangement to determine whether he or she lives with others and could receive inside ISM.

The second phase of the ISM evaluation process involves one of two ISM counting methods (Box 2).SSA applies the value of the one-third reduction (VTR) rule to recipients who live in another person's household (FLA-B) throughout a month and receive both food and shelter from within the household. Claims representatives reduce the FLA-B recipient's FBR by one-third rather than counting the actual value of support received. If an individual or a couple receives ISM but is not subject to the VTR, SSA applies the presumed maximum value (PMV) rule. The agency applies this rule to an individual in FLA-A (living in own home) or FLA-C (child living with parents) generally because the recipient lives in another person's household, but does not receive both food and shelter from that person or the recipient lives in his or her own household and receives inside and/or outside ISM. The PMV is equal to one-third of the FBR plus $20 (the general income exclusion) and caps the amount of ISM that SSA counts. The agency may apply an amount less than the PMV to calculate a person's payment if the individual can show that the actual value of the ISM received is lower than the PMV.

ISM Literature

Today, a limited body of ISM literature exists. The content of available ISM research falls into one of the following topical areas: ISM policy, past simplification attempts, and prevalence of ISM—discussed in the next three sections.

ISM Policy

An SSA report on SSI, released in 2000, identifies ISM policy as being exceptional because no other federal program counts in-kind support when determining benefit eligibility (SSA 2000b). Balkus and others (2009) explain that Congress' reasoning for including ISM as countable income was to direct SSI payments to persons with the least amount of income and support. However, a substantial portion of the ISM literature criticizes ISM policies for being inequitable, complex, intrusive, and burdensome (Balkus and others 2009; GAO 2002a, 2002b; Kennedy 1983; SSA 2000a, 2000b, 2012a, 2012b; SSAB 1999, 5). Numerous publications identify ISM policy as one of the leading policies that make the administration of the SSI program difficult, time-consuming, and costly, although only 9 percent of SSI recipients have their benefit rates reduced by ISM each year (Balkus and others 2009; SSA 2000a; 2013, Table 8). Furthermore, GAO and SSA's Office of the Inspector General have repeatedly declared ISM policy as one of the leading causes of SSI improper payments (GAO 2002a, 2002b, 2012; SSA 2012a, 2012b). Balkus and others (2009) indicate that ISM policy does not treat recipients equally. For instance, recipients with higher household expenses who receive support with a higher monetary value have a lower percentage of their total ISM offset by benefit reductions than those who have lower household expenses and need less support to fulfill their needs. Balkus and others (2009) and SSA (2000a) note that ISM policies also create disincentive issues by deterring families who would like to assist low-income relatives on SSI. Finally, the Social Security Advisory Board indicates that SSA must often base ISM determinations on what is alleged by SSI applicants and recipients rather than on verifiable information, such as expense receipts (SSAB 2005).

Past Simplification Attempts

Various articles and reports highlight SSA's numerous attempts to reduce the administrative burden and errors spurred by counting ISM. For example, the agency presents several ISM options and acknowledges that implementing alternative ISM policy might simplify the SSI program, but could create other dilemmas (SSA 2000a). Several other SSA documents discuss past attempts made by the agency's managers, researchers, and legislative workgroups to develop, study, and propose new approaches for simplifying ISM policy (Balkus and others 2009; SSA 2000a, 2012b). Repeatedly, GAO (2000a, 2000b, 2012) has reported limited progress on simplifying ISM complexities and addressing the persistence of ISM-related challenges. Several sources have acknowledged that a lack of detailed and comprehensive ISM data has impeded past ISM simplification efforts and opportunities to better inform ISM policy decisions and options (SSA 2000a, 2000b; SSAB 1999, 5; 2005).

Prevalence of ISM

Although existing SSI literature provides summary ISM statistics, it does not detail the characteristics or amounts of ISM received by SSI recipients. SSA (2000a) and Balkus and others (2009) report that nearly 9 percent of SSI recipients have their benefit rates reduced because of ISM during any given year. The SSI Annual Statistical Report, 2012 identifies ISM as the second most common source of unearned income received by SSI recipients (SSA 2013, Table 8).

Available ISM literature provides insight into the qualitative aspects of ISM policy and evaluation efforts, but falls short on providing detailed quantitative information. The current body of ISM literature leaves many questions unanswered about the source, form, and amount of ISM received by SSI recipients. This article alleviates the ISM literature gap by quantifying the ISM received by SSI participants, as of October 3, 2009, using MSSICS data.

Research Data and Methodology

For over 20 years, SSA has used the Modernized SSI Claims System to support the administration of SSI claims. I have deciphered and manipulated the content of MSSICS administrative data to support ISM research, despite that huge undertaking. My original MSSICS file (pulled on October 3, 2009) contained records for 1,120,817 SSI units identified as having positive ISM amounts recorded in certain MSSICS fields.⁶ Later, I matched that original MSSICS file to the Characteristic Extract Record (CER) format of the Supplemental Security Record (SSR) to identify which SSI cases were in current-pay status within a week of the MSSICS pull date.⁷ Then, I reconfigured couple-unit records to allow person-level comparisons. As a final step, I applied five selection criteria and limited my original MSSICS sample to the 53 percent of SSI recipients who were in current-pay status and alive during the week leading up to the MSSICS pull date. Appendixes A and B detail the structure, limitations, and uses of the MSSICS; my final sample selection criteria; and the methods I used for computing recipients' source, form, and amount of ISM support.

Overview of the Final Study Sample

My final MSSICS research sample consists of 611,192 recipients,⁸ of which about 91 percent received SSI payments as an SSI-individual unit and 9 percent as a member of an SSI-couple unit. I focused on persons who received an individual FBR because they represent the majority of persons receiving ISM. Among my final individual-FBR sample, 35.0 percent belonged to FLA-A, an additional 57.8 percent resided in FLA-B, and the remaining 7.2 percent were in FLA-C (Chart 1).⁹ From a different perspective, almost 45 percent of individuals younger than age 18 lived in FLA-C (Chart 2). The overwhelming majority (82.3 percent) of young adults aged 18 to 24 were in FLA-B, and 51.7 percent of individuals aged 25 or older were also in the FLA-B category.¹⁰ Young adult SSI recipients comprised 18.3 percent of all individual SSI recipients who received ISM (Chart 3). I separated young adults (aged 18–24) from all other working-age (25–64) recipients to examine how those young adults fared during the years following their transition from childhood to adulthood.

Findings

My new use of MSSICS data for research purposes reveals noteworthy differences between recipients who receive support and reflects the different ISM assumptions and counting rules applicable among those individuals. As mentioned earlier, SSA determines that recipients who live in the home of another person and receive support in the form of both food and shelter are in the FLA-B category. The agency applies the VTR rule to FLA-B members rather than counting ISM. Furthermore, SSA applies the PMV rule to all other SSI recipients receiving ISM who are not subject to the VTR rule, such as those having rental liability or ownership of their home or those paying at least their pro rata share of the household food and shelter expenses. Because SSA does not count ISM for FLA-B members, this section details the self-reported ISM of SSI recipients in categories FLA-A and FLA-C as of October 3, 2009.

Research Question 1: How Many ISM Applicants and Recipients Alleged Receiving Support from Within and/or Outside of Their Households?

SSA's ability to simplify SSI policy and reduce improper SSI payments has been contingent on its knowledge of recipients' sources of in-kind support. FLA-A and FLA-C subsample members comprised 43.3 percent of my final study sample (not shown), the majority (56.6 percent) of whom received support from only outside of their homes (Table 1). The remaining 56.7 percent of my final study sample belonged to FLA-B, and SSA applied the VTR rule rather than counting ISM (not shown).

An estimated 40.1 percent of FLA-A and FLA-C subsample members alleged receiving ISM from only inside of their homes (Table 1).¹¹ However, the FLA-A members were far more likely to receive ISM from exclusively within their homes than those in FLA-C (42.9 percent versus 24.0 percent).

The data in Table 1 suggest that one option for simplifying ISM policy and reducing improper SSI payment amounts may be to increase the efficiency of ISM evaluation efforts among FLA-A and FLA-C group members receiving ISM from outside the home.

Research Question 2: What Proportion of SSI Recipients Alleged Receiving ISM in the Form of Food and/or Shelter?

It is important to know whether individuals in FLA-A and FLA-C received support in the form of food and/or shelter because one can anticipate that the value of ISM would be greater if a person received assistance in the form of shelter rather than food. Chart 4 indicates that persons in FLA-A and FLA-C were most likely to allege receiving only shelter assistance if they received outside ISM, and they were most likely to allege receiving food and shelter assistance if they received inside ISM (86.6 percent versus 42.8 percent).¹² For persons receiving ISM from inside the household, the type of support received is not readily identifiable;¹³ an estimated 31.4 percent of FLA-A members and all FLA-C members with inside ISM had an undefined-ISM type.¹⁴

This analysis suggests that persons receiving ISM from the outside are more likely to have ISM totals exceeding the ISM amount deducted from their FBR because they are more likely to receive shelter assistance than those receiving only inside ISM. The Research Question 4 section discusses this hypothesis.

Research Question 3: What Proportion of SSI Recipients Alleged Obtaining ISM Exceeding the Amount Deducted from Their FBR?

My MSSICS data present an opportunity to study the actual dollar amount of ISM allegedly received by SSI recipients and identify how many of those recipients alleged an ISM total exceeding the amount deducted from their FBR. Recall that SSA applies the VTR rule to FLA-B group members and does not count any ISM within or outside of the household. Because we do not need to calculate the actual value of support received in most cases, it is not surprising that no FLA-B MSSICS sample members had ISM totals greater than the VTR. However, FLA-A and FLA-C members may have an alleged ISM amount exceeding the PMV cap (or one-third of the FBR plus $20), even though SSA does not deduct amounts in excess of the PMV from payments.¹⁵ As mentioned previously, the PMV cap policy has been the source of many equity issues. More explicitly, recipients with higher household expenses who receive support with a higher monetary value have a lower percentage of their total ISM offset by benefit reductions than those who have lower household expenses and need less support to fulfill their needs (Balkus and others 2009).

The current ISM cap rules create an advantage for the 31.4 percent of all FLA-A and FLA-C SSI recipients who alleged having an individual (not household) support total exceeding the PMV and did not have their FBR reduced by every dollar of ISM recorded in the MSSICS (Table 2). In fact, an estimated 13.0 percent of all FLA-A and FLA-C group members receiving ISM allegedly declared an ISM total equal to or greater than 200 percent of the PMV and had their FBR reduced by no more than 50 percent of the value of the support they had received (not shown). Meanwhile, almost half (47.1 percent) of FLA-A and FLA-C members had a “small” ISM total, less than 50 percent of the PMV (not shown) and had their FBR reduced by every dollar of ISM recorded for them.

Research Question 4: To What Extent Did the Total ISM Alleged Vary by ISM Source and Form and by Age Group of SSI Recipients?

This section sheds light on the characteristics of FLA-A and FLA-C subsample members receiving ISM who were most likely to benefit from the PMV cap. The data in Table 2 reveal that persons in FLA-A were more likely to allege support exceeding the PMV if they lived in their own home (34.2 percent) and/or alleged outside ISM (41.5 percent). The table also shows that FLA-A members receiving inside and outside (dual) support were most likely to allege ISM totals exceeding the PMV (67.6 percent), but those individuals represented no more than 4 percent of persons with ISM totals above the PMV cap (not shown). Over half (55 percent) of those benefiting from the PMV cap were FLA-A, individual SSI recipients alleging outside support (not shown). I also examined FLA-A, individual-FBR SSI recipients receiving ISM by age group and their source of ISM support.

Of the 98.8 percent of FLA-A, individual-FBR SSI recipients who alleged outside and/or inside ISM (Table 1), I found that the aged (65 or older) subset was most likely to allege ISM totals greater than the PMV, while the young adult (18–24) subset was least likely to do so (40.8 percent versus 23.6 percent), as shown in Table 3. The subset for children (younger than age 18) was the most likely to allege assistance exceeding the PMV cap among those alleging only outside ISM, while the aged subset (65 or older) was most likely to do so among those alleging only inside ISM (61.6 percent versus 27.5 percent). The latter findings are most likely the result of at least 90 percent of child SSI recipients who belonged to FLA-A and were receiving outside ISM and at least 70 percent of aged SSI recipients who belonged to FLA-A and were receiving inside-ISM assistance in the form of shelter (not shown).

These analyses suggest that aged, FLA-A SSI recipients were most likely to benefit from the current PMV cap provision because at least 80 percent of them received outside ISM and/or assistance in the form of shelter (not shown). These findings reaffirm my earlier hypothesis that SSI recipients receiving outside ISM are more likely to have ISM totals greater than the PMV if they receive assistance in the form of shelter.

Conclusions

This groundbreaking MSSICS research has provided valuable insight on the incentive, equity, and administrative issues associated with current ISM policies. First, this study reveals that over half of FLA-A and FLA-C SSI recipients receive ISM support from persons living outside of their homes. Second, most recipients receiving outside-ISM support obtain that support more often in the form of shelter rather than food (Chart 4). Third, an estimated 47.1 percent of FLA-A and FLA-C subsample members receiving ISM alleged individual ISM totals less than 50 percent of the PMV cap (not shown), while an additional 31.5 percent alleged ISM totals greater than the PMV cap and did not incur a dollar-for-dollar reduction of benefits because of ISM received (Table 2). Also, my MSSICS data show that SSI recipients receiving ISM were more likely to receive support exceeding the PMV cap if they were aged members of the FLA-A category—recipients who received support from outside of their households and/or in the form of shelter (Table 3). In conclusion, this research presents new insights on the source, form, and amount of ISM received by SSI applicants and recipients and how SSA policy and decision makers can use MSSICS data to understand complex policy issues when considering alternative options.

Appendix A: Research Data for the MSSICS and ISM Evaluation Process

For most SSI cases, SSA claims representatives use the MSSICS to gather, record, and update SSI claims information and to support SSI administrative efforts, such as the ISM evaluation.¹⁶ During the initial interview process, those representatives navigate through several MSSICS computer screens, while recording information provided by applicants or third parties. A person's application type or posteligibility event determines which MSSICS screens or paths a representative must navigate. SSA has programmed over 20 MSSICS screens to support the FLA and ISM evaluation process, but not all screens apply to every applicant or recipient. Therefore, some MSSICS fields have skip patterns. The MSSICS confirms eligibility and calculates an applicant's or a recipient's benefit rate after claims representatives record sufficient information about the claim to make a determination or enter information about a posteligibility event, such as a change in address, household composition, or household expenses (SSA 2001). Essentially, the MSSICS is a dynamic system because it is a constantly changing pending file that serves as a repository for claims representatives to use in creating permanent SSI records.

The MSSICS captures only information gathered during the ISM evaluation process that is necessary to determine benefit eligibility and payments. Recall that SSA's ISM policy involves identifying a recipient's FLA and applying one of two ISM counting methods. Claims representatives first use the MSSICS to determine a recipient's FLA during the ISM evaluation process (Box 1). (Thirteen MSSICS screens directly support the determination of one's living arrangement.) Thereafter, representatives use the MSSICS to gather the information needed to determine the amount of chargeable ISM (Box 2). For FLA-A and FLA-C group members, claims representatives use the MSSICS to determine the specific amount or type of ISM (that is—food, shelter, or both) received by those recipients along with their contribution to household expenses. SSA needs this information to determine a recipient's pro rata share of household food and shelter expenses and whether a representative should reduce that recipient's FBR by the PMV or a lesser amount. The agency determines that recipients who live in another person's household and receive both food and shelter assistance from within the home are in the FLA-B payment category. SSA reduces the applicable FBR by the VTR. Claims representatives do not complete any ISM development for FLA-B members, but instead reduce their benefits by the VTR.

Limitations of MSSICS Data

The MSSICS provides the most comprehensive ISM data and maintains a database of information with more detail than the CER format of the SSR. In addition to collecting information about individuals who apply for SSI as well as information about the parents if the applicant is a minor (younger than age 18), SSA uses the MSSICS to collect information on persons who are ineligible to receive payments, but who are part of a family in which someone does receive SSI payments (for example, ineligible spouses, children, others). The agency retains information on the value of resources in the MSSICS to determine eligibility. In the context of ISM evaluation efforts, the MSSICS indicates the amount of support received by most individuals in FLA-A and FLA-C, identifies whether support originated inside or outside of a recipient's home, and specifies for the majority of persons receiving ISM whether they received assistance in the form of food and/or shelter. MSSICS records also reflect what segments of the ISM evaluation process an SSI recipient receiving ISM participated in during an initial or posteligibility interview. Furthermore, those records show a household member's relationship to a recipient (that is—parent, child, spouse, and so forth). In contrast, the SSR provides monthly SSI records, but only flags a record if an SSI recipient had a positive ISM amount recorded and if he or she had the monthly payment reduced under the VTR or PMV rule (Panis and others 2002, chap. 6).

Like most data sources, the MSSICS has limitations. The MSSICS is a dynamic system that only reflects the characteristics of pending cases on the day that SSA staff members pull a set of records. The MSSICS does not provide data for past SSI recipients or reflect any subsequent updates made to incomplete cases. The system does not provide data on the current-pay status of recipients or whether they were alive on the pull date of their records. Furthermore, one can only acquire ISM data through the MSSICS for a point in time because extract files are unavailable; this access restriction greatly impedes researchers who wish to examine ISM trends. For the most part, MSSICS data do not capture all of the information that recipients or third parties provide to claims representatives during the ISM evaluation process, unless SSA needs that information to determine eligibility and payment amounts. For instance, SSA applies the VTR rule and does not count ISM if an individual belongs to FLA-B during an entire month and receives both food and shelter from the household. The system does not record uniform information among SSI recipients with ISM because of the different MSSICS screens or paths applicable to different groups. The data are unverified and largely based on the anecdotal evidence supplied by recipients or third parties. Despite there not being a corresponding MSSICS record for every SSI case, MSSICS data best equip SSA to quantify the ISM received by SSI recipients and to inform policymakers who wish to simplify SSI administration and avoid SSI payment errors.

ISM Study Data

My original MSSICS file provides a snapshot of the source, form, and amount of ISM received by SSI recipients, as of October 3, 2009. That file contained records for 1,120,817 SSI units. I matched my MSSICS file with the CER format of the SSR¹⁷ to verify which cases involved persons who were alive, had positive ISM records, and were in current-pay status at approximately the same time as the October 2009 MSSICS pull date.¹⁸ Matching MSSICS and CER/SSR records revealed that nearly 47 percent of my original MSSICS sample were not in current-pay status, alive, and/or receiving ISM during the week agency staff pulled my MSSICS file.¹⁹ Surprisingly, one out of eight SSI units (140,281 in all) had matching CER/SSR files indicating that the unit head was deceased before the MSSICS file's October 3, 2009, pull date. In addition, less than 2 percent of my original sample had matching MSSICS and CER/SSR ISM files equal to zero. Linking MSSICS with CER/SSR data enabled me to ensure that all final study individual-unit and couple-unit sample members were alive and eligible for SSI during the week leading up to October 3, 2009.

MSSICS Sample Selection Criteria

I selected my final study sample using five criteria:

1. FLA status;
2. Death date;
3. Payment status;
4. Presence of a positive MSSICS or CER/SSR ISM amount; and
5. Receipt of an individual or a couple FBR.

First, I retained MSSICS cases for individuals who belonged to FLA-A, FLA-B, or FLA-C and had matching CER/SSR records indicating that they were alive and were in current-payment status during the week immediately preceding October 3, 2009. Second, I omitted the portion of my original MSSICS file that involved FLA-A or FLA-C SSI units with matching MSSICS and CER/SSR ISM records equal to zero because SSA did not reduce those persons' FBR using the VTR rule, as was the case for their FLA-B counterparts. Third, I differentiated remaining SSI units by whether their members had received SSI payments as individuals or as eligible spouses. Next, I duplicated SSI couple-unit records and divided their recorded values in half to establish a record for each eligible spouse, so I could make person-level FBR comparisons among individual-unit and couple-unit SSI recipients and compare my estimates with those appearing in the SSI Annual Statistical Report (SSA 2013). Last, I verified that my couple (or eligible-spouse) subsample comprised only SSI recipients who were both alive and receiving SSI payments under the couple FBR during the week of October 3, 2009. Subsequently, my final study sample comprised 556,472 individual units and 54,720 couple units, which totaled 611,192 SSI recipients.

Appendix B: Research Methodology Used in Identifying Sources and Type of ISM and Calculating ISM Amounts

In this section, I describe my methods for computing recipients' source, form, and amount of ISM.

Outside ISM is food and/or shelter provided to an SSI recipient by at least one person living outside of the recipient's household. Outside ISM may also include rent-free shelter or a rental subsidy (reduced rent). Inside ISM, on the other hand, is food and/or shelter provided to an SSI recipient by persons within the household. If the household does not cover all of the recipient's shelter and food costs, the ISM is treated as outside ISM.

Determining Outside ISM

Determining receipt of outside ISM is relatively straightforward. The MSSICS records the form (that is, food and/or shelter) and amount of support received from someone outside the household. I calculated outside-ISM amounts by summing the six outside-ISM values recorded for FLA-A and FLA-C individual and couple SSI units on the MSSICS outside-ISM screen. Then, I divided each eligible spouse's outside-ISM total in half to compute the amount received by each eligible spouse.

Determining Inside ISM

The value of inside ISM received by an SSI recipient is dependent on his or her household shelter and food expenditures, separate-food-consumption and purchasing status, and his or her contribution to household expenses. Recipients can receive inside ISM provided they do not live alone, reside only with an ineligible parent or spouse, and/or belong to a public assistance household. Therefore, all noted inside-ISM numbers pertain to the members of my final sample who belonged to FLA-A or FLA-C and resided with at least one ineligible household member.²⁰ I initiated my inside-ISM calculations by dividing FLA-A and FLA-C sample members living with at least one ineligible household member into two groups:

1. Recipients who shared their food expenses with all members of their household (59.4 percent); or
2. Recipients who either consumed all of their meals outside of their household or purchased their food separately from their household members (40.6 percent).

For FLA-A and FLA-C sample members who shared food expenses equally with their household members, I based their inside-ISM amounts on their personal share of total household food and shelter expenses minus their personal contribution toward household expenses. As for FLA-A and FLA-C sample members who separately consumed their meals or purchased their food separately from their household members, I based their inside-ISM amounts on their personal share of total shelter expenses minus their personal contribution toward household expenses.²¹

Identifying Outside-ISM Type

It is also straightforward to determine the type of outside ISM received by persons not living in the household of another. I simply referred to the six fields of the MSSICS outside-ISM screen that specify the type of outside ISM received. I identified SSI recipients as acquiring only food outside ISM if their positive outside-ISM flags were only equal to “1” (denoting food), and I classified SSI recipients with outside-ISM flags all equal to “2” (denoting shelter) as having only shelter outside ISM. I then categorized all MSSICS observations as receiving food and shelter outside ISM if they had one of the following:

1. Outside-ISM flags equal to 1 (food) and 2 (shelter); or
2. At least one outside-ISM flag equal to “4,” indicating they had received food and shelter outside ISM because of a nonhousehold situation.²²

Identifying Inside-ISM Type

Next, I investigated whether inside-ISM recipients received support in the form of food, shelter, or both. Assessing recipients' food consumption activities is essential for computing what, if any, inside-ISM amounts they received; identifying if they had received assistance in the form of food and/or shelter; and determining their FLA. I immediately classified FLA-A and FLA-C recipients who lived with at least one ineligible household member as receiving inside-ISM shelter if their MSSICS records indicated they had either consumed all of their meals outside of the home or purchased their food separately from their household members. However, distinguishing inside-ISM food and shelter was more involved among the majority of recipients in FLA-A and FLA-C who shared food expenses with all of their household members. The MSSICS only distinguishes the type of inside ISM received by recipients without rental or home ownership liability because the system only allows claims representatives to record earmarked contributions for FLA-A and FLA-C sample members. Almost half (47.3 percent) of the recipients in FLA-A and FLA-C who shared food with others had neither rental nor ownership liability for their households; 5 percent of that subset had earmarked contribution fields specifying positive amounts.

I identified the few sample members in FLA-A and FLA-C without rental or home ownership liability and without missing records as receiving only inside-ISM shelter if a claims representative had earmarked those individuals' personal contributions toward food, and vice versa. For members of FLA-A and FLA-C with contributions earmarked for food and shelter, I based their inside-ISM type on whether their food and/or shelter contributions fell below their pro rata share of total household food and shelter expenses. For example, I identified sample members with food and shelter earmarked contribution records as receiving inside-ISM shelter if their food contribution was equal to or greater than their pro rata share of total household food expenses and their shelter contribution was less than their pro rata share of total household shelter expenses. Because one cannot identify the type of inside ISM received by renters or homeowners, I classified that type of ISM as undefined. It is possible that the subgroup with undefined-ISM types received both food and shelter assistance from within their homes.

Computing ISM Amounts

One of my research objectives was to identify the proportion of October 2009 SSI recipients who received ISM from inside and/or outside their households.

I based my MSSICS outside-ISM counts on the number of recipients who had at least one positive outside-ISM field.²³ Nearly 27 percent (n=162,202) of my final study sample had outside-ISM amounts greater than zero; almost 80 percent (n=127,018) of that group belonged to the FLA-A category.

Calculating the proportion of SSI recipients who received inside ISM was more involved. Those calculations required examining SSI recipients' household member counts, household shelter and food expenditures, separate-food-consumption and purchasing status, and their personal contribution toward household expenses. I initiated my inside-ISM computations by dividing SSI recipients into two groups: (1) those who ate all of their meals in their household and shared food purchases with their household members, and (2) those who either ate all meals outside of the household or separately purchased their food. For the first group (n=541,359), I set their inside-ISM amount to their pro rata share of total household expenditures minus their personal contribution toward household expenses. For individuals separately consuming or purchasing food (n=74,933), I calculated their inside-ISM amount as being equal to their pro rata share of their total household shelter expenses minus their personal contribution toward household expenses. I found that 25.9 percent (n=159,831) of my final study sample had MSSICS records indicating that they had received positive inside ISM.²⁴