Bridging the Gaps Method and Data
The Bridging the Gaps project is a unique, collaborative effort. State-level partners collected eligibility rules for six public work supports in their states: child care assistance; EITC (Earned Income Tax Credit, federal and state); Food Stamps; housing assistance (Section 8 and public housing only); Medicaid /SCHIP (State Children’s Health Insurance Program); and TANF (Temporary Assistance to Needy Families). In five of the states, they also conducted focus groups. Our partners participated in each step of the research process: discussing methods to estimate the gaps, learning how to conduct focus groups, and discussing the results. This is the first project to use survey data to estimate eligibility and coverage of work supports in these ten states. More detail on our data and methods can be found in, Bridging the Gaps: Technical Report on Data and Methods (forthcoming October 2007).
A family falls into the hardships gap when their income remains below a basic standard of need, even after the value of work supports is incorporated into their family budget calculation. To estimate the hardships gap, we make adjustments to the family budgets and total reported income. The amount of a family’s hardships gap is the difference between their income, including all work supports, compared to costs of goods and services in their local area.
If a family reports receiving Food Stamps, the EITC, or TANF, then we add the cash value of these benefits to their total income.
If a family reports receiving childcare assistance, housing assistance, or Medicaid/SCHIP, we use the self-reported values for these items instead of the family budget cost estimates for these expenses.
The eligibility gap is the share of people living below a basic family budget who are eligible for each work support program. We estimate who is eligible for each work support by mapping the work support eligibility rules onto survey data. This share is then multiplied by the state population from the U.S. Census Bureau for the same year as the administrative data, except in the case of the EITC where the population estimate is based on tax filers as reported by the Internal Revenue Service. The relevant Census state population is based on the unit of analysis for each program: child care is based on the number of children under age 13; EITC is based on the number of tax filers; housing assistance is based on the number of households; and Food Stamps, Medicaid/SCHIP, and TANF are based on number of people.
The coverage gap is the share of people within each state who receive each of the six work supports compared to the share of the state population eligible for these programs. The number of units that receive the work support comes from state administrative data for the latest year available. We do not use reported receipt of programs from the survey data because it is under-reported and would therefore overestimate the eligibility gap.
For some programs, we refer to the eligibility gap and “take-up,” while for others, we talk of “effective coverage.” The EITC, Food Stamps and parts of the Medicaid program are entitlements, meaning that a person or family who is eligible for the program can access it so long as they apply, so we refer to the coverage gap as a problem of take up. Child care assistance, housing assistance, and TANF work supports, however, are often not available to all who are eligible. In many BTG states, there are waiting lists and in some cases, the list have been closed to new applicants. Here, we refer to the eligibility gap as a problem of effective coverage.
Besides the EPI/CEPR family budgets (explained in Chapter 2 of Bridging the Gaps: Technical Report on Data and Methods), we use four kinds of data to estimate the hardships, eligibility, and coverage gaps:
Survey data. We use the Survey of Income and Program Participation (SIPP) for all nine states and the Current Population Survey’s Annual Social and Economic Supplement (ASEC) for the District of Columbia. These surveys are representative of the U.S. non-institutionalized population, including the elderly, children, disabled people, and any others not in the military or an institutional setting. The SIPP is a three-year panel, beginning in 2001 and the ASEC is conducted annually each March; we use the data that cover calendar years 2001 through 2005 (survey years 2002 through 2006).
Administrative data. We use administrative counts of the average number of units served per month by each work support. CEPR, in conjunction with BTG partners, gathered this data from available sources online, usually the government agency responsible for administering the program. The unit of analysis differs across administrative programs: for child care, the unit of analysis is the number of children under age 13 served; for EITC, the unit of analysis is tax filers; for housing assistance, the unit of analysis is the number of households served; and for Food Stamps, Medicaid/SCHIP and TANF, the unit of analysis is the number of people served.
Work support eligibility rules. The BTG state partners, working with CEPR, gathered the work support eligibility rules. The rules were compiled from government sources as well as non-governmental sources, including information from advocates and providers. We collected all eligibility rules possible, including those related to income thresholds, asset limitations, citizen status, and work requirements. The exception is the EITC where we used the National Bureau of Economic Research’s online TAXSIM model to estimate eligibility.
The work support program eligibility rules are from the latest year available, and, unless noted, match the year of the administrative data. For the District of Columbia, Iowa, Illinois, New York, and North Carolina, the eligibility rules are from 2006; for Minnesota, Ohio, and Washington, they are from 2005; and for Massachusetts and Texas, they are from 2004. For each work support program in BTG states and the District of Columbia, we include an appendix that describes the eligibility rules.
To ensure accuracy, the eligibility rules were reviewed by those most familiar with the programs in each state. The eligibility rules for the six major work supports programs rarely (if ever) could be found in any central source. Further, some are extremely complicated; thus, working closely with partners in each state was key to ensuring that the rules are as accurate as possible. As a part of this process, CEPR and BTG partners held meetings with advocates and policy experts to discuss the accuracy of the rules that we put together.
Focus Groups. Five of the BTG partners (District of Columbia, Illinois, Massachusetts, Minnesota, and North Carolina) conducted a total 22 focus groups with over 100 parents whose incomes ranged from 75 percent of the federal poverty threshold to 325 percent. In all of the states, except Illinois, separate focus groups were conducted for those with higher family incomes (225-325 percent of poverty threshold) and those with lower incomes (75-250 percent of poverty threshold). Using the same protocols across the BTG locations, focus group participants discussed the ways they combine earnings and public work supports; the reasons so many families who are eligible for public work supports do not use them; and the strategies families employ to make ends meet.