The U.S. Department of Agriculture’s Community Eligibility Program (CEP) has created an important and exciting opportunity for schools with a relatively large share of low-income students to increase the number of students who receive free meals. This quickly-growing program is reducing the number of students across the Commonwealth at risk of being hungry at school. Expanding CEP is beneficial to all the students in participating schools because it ensures that fewer students are trying to learn while hungry; it reduces possible stigma associated with participating in the school meal programs; and it eliminates some administrative costs for schools.
At the same time, this program identifies and counts eligible low-income students by matching student data with data from other state programs (“direct certification”) to create what is known as the “economically disadvantaged” measure. Prior to CEP, schools had each family fill out a meals application form every year, and used the data from these forms to identify low-income students. This was the vintage “low-income students” measure in place through the 2013-2014 school year. The implementation of CEP brings particular challenges to the schools in areas other than school meals where the data from those forms were used. The expansion of CEP means that state officials and school administrators now need to find new ways to count their low-income students as accurately as possible for school funding formulas (see Proposed Low-Income Student Changes Would Have Varied Chapter 70 Impact) and for other purposes.1
While the Department of Elementary and Secondary Education (DESE) has been making significant progress with the implementation of Community Eligibility, the automated data matching process that directly certifies eligible students currently fails to identify a large number of low-income students. Undercounting these low-income students means that fewer schools are participating in CEP than might be potentially eligible, and schools may be losing out on federal and state funds for which they should be eligible.
This paper outlines some of the problems in the direct certification system as currently operating, and broadly identifies possible solutions. There are systemic and technical changes that could help, and providing supports to schools will assist them in the transition to this new system. By extending the automatic matching system and including all relevant programs, improving the ability of the automated program to identify matches, allowing for the resolution of partial data matches, and making sure that all low-income families are enrolled in the food and health care programs for which they are eligible, the state can provide significant benefit to students, their families, and the schools.
THE POLICY CHANGES
To determine whether a school or district can participate in the Community Eligibility Program, federal law requires that schools use an automated program eligibility matching process known as direct certification. This process links school enrollment records with the state’s enrollment database for the TAFDC (public assistance) and SNAP (formerly known as Food Stamps) programs. The state now performs this match several times a year. School administrators at the school and district level then have regular access to this database to identify their students whose records exactly match the lists from these benefits programs and also a list of records that partially match. (See Appendix for discussion of how this process works.)
Massachusetts is also participating in a special pilot program implemented in the 2014-2015 school year that matches school enrollment lists with the state’s Medicaid (MassHealth) membership. Students whose families have incomes up to 133 percent of the federal poverty level can be directly certified for free school meals if they are covered by any of the MassHealth insurance programs. While schools do not have direct access to the MassHealth database, they receive a listing of income-eligible students whose enrollment records exactly match with current MassHealth records. Schools do not currently receive information about records that partially match with the MassHealth lists.
In addition, schools may certify other categorically eligible students who participate in Head Start or Early Head Start, are foster children under custody of the Dept. of Children and Families (DCF), are on a list of homeless children as identified by the district homeless liaison, are migrants, or are runaway children. Combined, these are the “Identified Students” that can be counted towards eligibility for CEP. At a 40 percent Identified Student Percentage (ISP), a school or district may elect to participate in CEP (see Appendix for more details).
A school’s direct certification count is not necessarily the same as the count of students who are receiving free school meals. In a school participating in CEP, all of the students are receiving free meals. In other schools, there may be students who were not identified via direct certification, but who have completed paper application forms and have met the income criteria and have been determined eligible for free school meals.
Once a school is implementing CEP, it no longer collects paper applications to identify low-income students. Since not all schools or districts are collecting paper applications any more, DESE can no longer rely on data from paper applications to uniformly identify low-income student populations across the Commonwealth.
The state is looking to direct certification as a new way to identify low-income students across the Commonwealth. By strengthening the way the state and districts implement direct certification, and by using this same system to identify additional low-income students, the state should be able to achieve a good count of low-income students to use for the school aid formula and other purposes.
THE MASSACHUSETTS DIRECT CERTIFICATION PROGRAM
Direct certification simplifies and streamlines the school meal application process. Children receiving SNAP or TAFDC benefits, or living in households with other children receiving TAFDC or SNAP benefits have already been determined eligible for programs with financial criteria similar to the school meals programs and can therefore be directly certified for free school meals. Students can also be directly certified for free school meals if they currently receive MassHealth and if they are in a household with an income up to 133 percent of the federal poverty level.
Participating in the national pilot that brings the Medicaid program data into the direct certification process provides an important and exciting opportunity for Massachusetts. Massachusetts has a particularly high rate of health insurance for children, and the MassHealth program provides health insurance for close to 645,000 children, with almost 500,000 of those children school-aged.2 With the fully-functioning inclusion of the MassHealth program in the direct certification program, Massachusetts should be able to identify almost all children in the schools in households under 133 percent of the federal poverty level and in fact also identify a large share of children at incomes up to 185 percent FPL.
Direct Certification and Expanded Program Matching Could Provide an Effective Way to Count Low-Income Students
The direct certification program should be automatically identifying the vast majority of students who are eligible for free meals, but it is not yet doing so. For example, recent caseload estimates from MassHealth show that there are approximately 392,000 school-aged children between the ages of 5 and 19 enrolled on MassHealth with incomes below 133 percent of the federal poverty level. Even though some of these young people might not be in the public schools, an optimally-functioning direct certification system should be able to identify a similar number of students eligible for free meals.
Currently, DESE has directly certified 312,200 students for free meals (see table below). This total is approximately 79,700 less than the low-income MassHealth school-aged caseload, suggesting that there may be as many as 79,700 students who should be directly certified who are not yet getting picked up by the data match. These “lost” students may, of course, not be in the public schools, or they may still be receiving free meals. They may have completed paper applications, or they may be in CEP schools. What is significant is that these students are not included in a school district’s “economically disadvantaged” student count.
With the new policy that bases state educational aid (Chapter 70) on a district’s count of economically disadvantaged students, getting the direct certification process to work accurately becomes all the more imperative. A school district’s funding is now highly-dependent upon the successful implementation of direct certification in the district.
Moreover, because MassHealth has the ability to identify the income levels of its members, DESE could even explore the possibility of using this data matching system to go beyond its economically disadvantaged measure. The state could also identify MassHealth students at incomes from 133-185 percent of the federal poverty level and generate counts of students closer to the vintage low-income student tallies. Not only would this be useful for the purposes of the Chapter 70 funding formula, it would simplify and expand the number of students receiving reduced-price meals in schools that are not participating in the Community Eligibility Program.
In fact, a new U.S. Department of Agriculture (USDA) pilot grant program has just been announced to do exactly this. This new pilot could provide DESE with the opportunity to do MassHealth expanded program matching to 185 percent of the federal poverty level to identify students eligible for reduced-price meals. If Massachusetts were to participate in this pilot, the state might be able to generate data on students to closely approximate or even surpass the totals previously generated by the vintage low-income measure.
School year 2013-2014 was the last year in which Massachusetts tallied students relying on the vintage low-income students category and paper applications for all schools. In 2013-2014, there were approximately 365,900 students identified as low-income (under 185 percent of the federal poverty level). There are approximately 446,000 people ages 5-19 with incomes under 185 percent of the federal poverty level on the MassHealth caseload. Although some of the young people on the MassHealth caseload are likely not in the public schools, the MassHealth totals demonstrate that there are many more low-income school-aged people than the number identified by the vintage low-income count, and successful data matching with MassHealth could capture these students. If an automatic data matching program were able to identify even 80 percent of these young people on the MassHealth caseload, the schools would be identifying as many low-income students through data matching as they were through the old paper application system.
Accordingly, if the state were able to correct the problems with the current data matching system and get approval for the new pilot program to identify students with MassHealth up to 185 percent of the federal poverty level, the state should be able to use data matching to effectively identify even more low-income students than the old paper application system did.
Challenges in the Implementation of Direct Certification
Multiple challenges have emerged with the implementation of the direct certification matching program. Some of these challenges are systemic, some are technical, and some are problems that are likely just the challenges that come with implementing any new system. In order for the direct certification plan to work properly, and to consistently and accurately identify all of the students eligible for free meals, and to also accurately identify every economically disadvantaged student, each of the pieces of the system must be working together properly.
It is particularly important to recognize that the system currently relies on substantial effort at the local level. Massachusetts has more than 400 school districts, with varying capacity to consistently implement these complex administrative data matching tasks. The more that the state can take on direct certification data matching at the state level through systems enhancements, the more likely the entire program will function smoothly and consistently across the Commonwealth, and the more likely the state will be generating reliable student tallies.
Problem: The state has not yet included all programs in the data match.
The entire MassHealth program caseload has not yet been fully incorporated into the matching database. In the fall of 2015 when the MassHealth match began, the database only included a subset of the MassHealth eligibility categories. Although the programs that enroll a majority of children were included, a significant number of programs, including the MassHealth Limited program and its companion Children’s Medical Security Plan (CMSP) were not included in the match. MassHealth Limited provides emergency Medicaid coverage to approximately 16,000 children who are largely immigrants. These children also typically receive CMSP for preventive care. Because of the initial exclusion of these programs from the MassHealth data match, for example, none of these children has yet been directly certified for school meals through the MassHealth match or yet included in a school’s economically disadvantaged count.
Solution: Expand the MassHealth match to include all relevant Medicaid programs.
The state recently announced it would add all of the remaining MassHealth programs into the database in March 2016, at which point children in these remaining programs should be counted by the new system.
Problem: Data matching may not be including all students whose eligibility is due to “extended eligibility.”
The automatic data matching should be able to match and directly certify students when there is another child in the household who receives SNAP or TAFDC benefits. Whereas direct certification via MassHealth requires that the student be a MassHealth member at the time of the data match, for SNAP and TAFDC direct certification, either the student or a minor child in the student’s household can be receiving a SNAP or TAFDC benefit. This is referred to as “extended eligibility.” When the automated data matching system does not capture all students who could be eligible due to extended eligibility, the burden for identifying these additional students is on the administrative staff in each school or school district to document the inclusion of these students in a school’s Identified Student Percentage. These extended eligibility students would be included in a district’s ISP, but would not show up included in direct certification counts.
Solution: Train school personnel in proper documentation of extended eligibility.
There is information about households in the SNAP and TAFDC databases, which increases the number of students automatically directly certified via extended eligibility. However, because extended eligibility criteria can be complicated, school district personnel should be carefully trained in how to identify and document extended eligibility when not automatically identified so as not to be excluding potentially eligible students from direct certification.
Problem: Families that are eligible for SNAP or MassHealth but are not enrolled would not be included in direct certification counts.
The direct certification program works best when every child eligible for SNAP, TAFDC, or MassHealth is enrolled in those programs. For a variety of reasons that is not always the case.
Solution: Support community outreach to make sure all eligible families are enrolled in SNAP and MassHealth.
Making sure eligible families are enrolled in SNAP and MassHealth would reduce food insecurity for low-income families, expand access to health care, and also lead to more accurate identification of low-income students through direct certification.
There are estimates that there are hundreds of thousands of people in Massachusetts potentially eligible for SNAP but not currently enrolled. In the 2013-2014 school year, there were approximately 222,000 school-aged children receiving SNAP in Massachusetts, but close to 100,000 additional students received free school meals in that year.
School staff could provide families with information about programs for which they might be eligible. Also, by coordinating with local community health centers and other community organizations, schools could help uninsured students enroll in MassHealth, and eligible families enroll in SNAP. Other institutional partners, such as local philanthropies, might also consider supporting local community-based outreach and enrollment activities. The state might also consider funding expanded outreach and enrollment for SNAP and MassHealth, recognizing the significant financial benefit of federal funds that would come to the schools and the state.
It is certainly in the best interest of students and schools for families to have access to programs for which they are eligible. Not only are students better off when they have health care and food on the table at home, but the schools themselves benefit by having more eligible students included in their direct certification counts which would increase the number of schools eligible for Community Eligibility, and would also increase the school’s state aid.
Implementing a common application for benefits (discussed below) would also be an important step towards ensuring that people are enrolled in the benefits for which they are eligible.
Problem: The state’s data systems are not aligned.
One of the primary challenges to the direct certification project is that the state’s school data systems, SNAP and TAFDC data systems, and MassHealth data systems are not aligned. Because these are separate databases, connecting and coordinating them is a substantial technical challenge. Inconsistencies in names, spellings, dates, etc., make data matching cumbersome and complicated.
Solution: The state could create a common application for SNAP, TAFDC, MassHealth and similar benefits.
Creating a common and coordinated application for key benefits such as MassHealth, SNAP, and TAFDC, would be an important first step towards coordinating the state’s data systems, as the state could create a common and unique identifier that would be shared across the programs. This might reduce some of the technical challenges associated with adding the MassHealth caseload to the direct certification data base.
Creating a common application would also reduce the number of people missing out on benefits for which they are eligible. State officials have indicated that this is something the Executive Office of Health and Human Services (EOHHS) is working towards, although the timeframe is uncertain.
Problem: The data matching system uses limited criteria for making exact matches.
There are also a variety of technical problems with the implementation of direct certification. The biggest problem surfacing in the direct certification matching program is that the enrollment lists (school enrollment rosters and public benefits program rosters) do not always align properly. This means that for many students, even though they may receive SNAP or TAFDC, or be enrolled in MassHealth, because of a technical misalignment in their data, they will not be matched for direct certification.
There are many reasons why a student’s record might not match the listing in the public benefits databases, including: names might be misspelled in one of the lists, dates might be entered with numbers transposed, a student might have multiple last names that are entered differently, a student’s name might be transliterated in different ways from a non-English alphabet. When records for SNAP or TAFDC match only partially, staff at the local school or school district must devote significant administrative time to “resolving” and documenting partial matches.
Local staff (at the district or school level) can manually review data that do not match, but this process can be very time-consuming. There are school districts (Springfield, for example) that have devoted significant staff time and expertise to manually reviewing student rosters in order to increase their direct certification numbers and Identified Student Percentage. There is significant pay-off for these efforts in terms of expanding participation in the Community Eligibility Program and for drawing down additional state and federal funds, but not all school districts have personnel to devote to these activities.
Solution: Add a common numeric identifier to the databases in order to simplify the matching protocols.
DESE and the EOHHS are taking steps to add DESE’s unique student ID (the SASID) to the DTA and MassHealth databases. It is not yet clear when this process will be fully-implemented, as there appear to be significant technical challenges in connecting these databases.
Once there is a common unique identifier for each student, many of the other technical challenges may be resolved.
Pending the successful addition of the DESE SASID to the other
databases, there are several solutions DESE could consider:
- Improve the matching algorithm to accommodate common data mismatches. Some states (Connecticut, for example) have developed sophisticated probabilistic matching algorithms to increase their direct certification match rates. These changes could include such elements as allowing matches when a student has two last names and reversing the order of the names, adding or subtracting hyphens in last names, substituting “s” for “z” in names such as Rodriguez (Rodrigues).
- Include an additional field for the student’s name as exactly entered on their MassHealth card so that at the very least, there would be no mismatches on a student’s first or last name.
- Providing support and training for district staff to resolve partial direct certification matches.
Problem: The direct certification matching program only allows for an exact match to the MassHealth database.
According to DESE, the current direct certification program can only generate exact matches against the MassHealth database. They have noted that at this point the system is not capable of identifying partial matches, but that this could be a system enhancement in the future.
Solution: Add the capacity to the matching program to generate a list of MassHealth partial matches.
Allowing school administrators to correct inaccurate records identified by partial matches to the MassHealth database could be an important way to increase the number of students directly certified.
The expansion of direct certification and its new-found significance for all school districts has created significant challenges during this period of transition, but there are manageable strategies the state can pursue to address these challenges. If DESE were to provide expanded training at the local level for district and school staff on how to manage and administer the direct certification matching program, districts would continually improve their ability to accurately and consistently generate counts of students directly certified for free school meals. This training could include how to directly certify students through extended eligibility, how to resolve partially-matched data, and what documentation is necessary for resolving the partial matches of Identified Students.
The state could support outreach and enrollment expansion efforts at the local level to ensure that students are enrolled in the programs for which they are eligible. This would also increase the share of students included in direct certification counts.
There are many opportunities for state officials to emphasize the importance of the direct certification program, even for schools not participating in Community Eligibility. First of all, school districts increasing their direct certification rates are potentially eligible for substantial additional funding. Second, it is possible that there are many more schools and school districts that will find they will be eligible to take advantage of Community Eligibility, with all the benefits that come with that important program.
Moreover, if the state uses direct certification counts for administrative purposes, during this time of transition it is important to use up-to-date numbers throughout the state budget process. Particularly with the ongoing inclusion of additional members of the MassHealth caseload to the data match, and with ongoing improvements to the technical systems during this time of transition, it is notable that the direct certification counts are continually improving.
Even so, given the significant and substantial challenges associated with the implementation of this important expansion of direct certification, and the inconsistencies with implementing direct certification across the Commonwealth’s many school districts, the state should consider postponing the use of direct certification counts for such crucial uses as determinations about school funding.
Deferring using direct certification counts for the state’s school funding formula for a year could allow time for the state to put into place the recommended solutions to help generate a more accurate economically disadvantaged count. At the same time, the state should actively pursue the new USDA pilot to use automatic data matching for MassHealth to 185 percent of the federal poverty level. With systemic and technical corrections to the direct certification system, coupled with this expanded data matching, Massachusetts might be able to generate low-income student totals that would closely track or even improve upon the old low-income students measure.
GLOSSARY OF ABBREVIATIONS
CEP—Community Eligibility Program. A federal program that allows schools or districts to provide free school meals to all students once the school or district identifies that 40 percent of student enrollment are categorically eligible for free meals.
CMSP—Children’s Medical Security Plan. Health coverage for primary and preventive care for otherwise uninsured children. Also available to immigrant children qualified for MassHealth Limited.
DCF—Dept. of Children and Families.
DESE—Dept. of Elementary and Secondary Education.
EOHHS—Executive Office of Health and Human Services.
FPL—Federal Poverty Level. A federally-determined income threshold used to determine eligibility for public programs.
ISP—Identified Student Percentage. The share of students enrolled who are categorically eligible for free school meals.
SASID—State Assigned Student Identifier. This unique student identification number is assigned to a student upon enrollment in the Mass. public schools, and remains with that student throughout their time in the schools.
SNAP—Supplementary Nutrition Assistance Program. Formerly known as Food Stamps.
TAFDC—Transitional Assistance for Families with Dependent Children. The Massachusetts transitional public assistance program.
USDA—United States Department of Agriculture. The federal agency administering the school meals program.
HOW COMMUNITY ELIGIBILITY AND DIRECT CERTIFICATION WORK
The school meals programs provide free meals to students in families with incomes under 133 percent of the federal poverty level (FPL) and reduced-price meals to students in families under 185 percent of the FPL. With Community Eligibility (CEP), if a school or district can identify that 40 percent or more of its students are eligible for free school meals (as measured by its Identified Student Percentage, or ISP), that school or district may opt to provide free meals to all its students.
Direct certification takes data from the school district’s roster of students, and matches it against the state’s database for the Transitional Assistance for Families With Dependent Children (TAFDC), the Supplemental Nutrition Assistance Program (SNAP, or “food stamps”) and the state’s databases for the Medicaid (MassHealth) program for members under 133 percent FPL. At least three times a year, districts are required to upload their student roster into the state’s human services online data system known as the Virtual Gateway. School districts have regular access to this matched database, and get a list that will identify each student as either an exact match, a partial match, a household match (for TAFDC and SNAP only), or no match. The system also matches students against the Department of Children and Families caseload. It is up to each school district to then determine whether errors in student data could be re-submitted into the school roster in order to turn a “partial match” or “no match” into an exact match. It is also up to each school district to identify whether there are students not identified as household matches who could indeed be directly certified due to extended eligibility. The share of student data converted from partial matches to exact matches and therefore counted in direct certification counts is highly dependent upon the school district resources put into resolving partial matches.
Schools participating in CEP no longer collect individual program applications and no longer collect meal fees at the point of sale. The federal government reimburses a school participating in CEP based on a formula. Schools are reimbursed for free meals at a rate of 1.6 times the ISP, and all other meals are reimbursed at the lower “paid meal” rate.
When a school reaches 40 percent ISP it could choose to participate in the Community Eligibility Program, but it would cost the school to do so. At a 40 percent ISP, a school would be providing free meals to 100 percent of its students, but would be getting reimbursed by the federal government for 64 percent of its meals at the free meal rate, and 36 percent of its meals at the lower paid meal reimbursement rate. Under this scenario, a school meal program would lose money.
However, once a school reaches an ISP of 62.5 percent, under CEP all meals would be reimbursed at the higher “free meal” reimbursement rate (62.5% x 1.6 = 100%). At an ISP of 62.5 percent, CEP provides free meals to all students at no additional cost to the school, and even at an ISP of approximately 50-55 percent, expanding to CEP may not cost a school anything because of CEP’s reduced administrative costs.
1In particular, schools rely on the identification of individual students from low-income families in order to track academic progress and eliminate achievement gaps. In many schools, identified low-income students might also have access to special targeted programs or might be eligible for fee waivers to participate in field trips or extracurricular activities.
2Data provided by Office of Medicaid, MassHealth Snapshots, November 2015 and by separate request, December 1, 2015.