LOADING

Type to search

The Community Eligibility Provision, Free and Reduced-Price Lunch, and Measurement of Poverty: Implications for Education Policy

The Community Eligibility Provision, Free and Reduced-Price Lunch, and Measurement of Poverty: Implications for Education Policy

Cory Koedel
University of Missouri-Columbia/CALDER 

Stacey Preis
PRiME Center, Saint Louis University

CALDER Policy Brief No. 16-0519

 

Highlights

  • Changes to the National School Lunch Program (NSLP) have affected the indicator that has been most commonly used to measure student income status in research and policy: Free and Reduced-Price Lunch (FRL).
  • Existing alternatives such as Direct Certification (DC) have potential, but also limitations.
  • A study comparing FRL and DC data did not show considerable differences in their efficacy in proxying for student disadvantage. Still, DC data have some appealing features, and many states are switching from FRL- to DC-based measures of poverty.
  • Uncertainty around the continued use of FRL data, brought on by the changes to the NSLP, can be leveraged as an opportunity to usher in a more rigorous and thorough approach to measuring student disadvantage in states across the U.S.

Executive Summary

The Community Eligibility Provision (CEP) is a recently-implemented policy change to the federally-administered National School Lunch Program (NSLP). It allows schools and districts serving low-income populations to identify all students as eligible for free lunch, regardless of students’ individual circumstances. The purpose of the CEP is to expand meal access to students who attend low income schools, while at the same time reducing paperwork and streamlining the process of participating in the NSLP. An unintended consequence of the CEP is that it reduces the informational content of NSLP-based measures of student poverty because all students at CEP schools are classified as eligible for free lunch. This has important implications for school accountability and finance policies at all levels of government as these policies have become highly dependent on the use of free and reduced-price lunch (FRL) data to proxy for student disadvantage.

 

What Is the Issue?

Student poverty has long been associated with measurable differences in student outcomes. Federal No Child Left Behind (NCLB) legislation was designed to, among other things, highlight these differences through the focus on accountability for outcomes across multiple student subgroups within schools, including subgroups of high-poverty students.

Motivated by the differences in student outcomes by poverty status, education finance policies are often designed to shift resources toward high poverty students. A prominent national example is the federal Title-I program. Another example at the state level is California’s Local Control Funding Formula (LCFF).[1] These and other substantively similar policies reflect the recognized challenges associated with improving educational outcomes for high-poverty students.

The primary measure of student poverty status used for these types of policies has been participation in the NSLP. In fact, under both NCLB and the Every Student Succeeds Act (ESSA), FRL-eligible students constitute their own subgroup for the purposes of evaluating academic outcomes under federal accountability. Additionally, school funding policies have relied heavily on FRL data to disburse poverty-targeted resources.

With the introduction of the CEP, participating schools and districts identify all students as eligible for free lunch, regardless of students’ individual circumstances. The CEP-induced change to the informational content of students’ FRL designations has important implications for policy design.

How the CEP Influences Accountability & Funding Policies: Missouri as a Case Study

In Missouri’s state accountability system, district accreditation and charter local education agency (LEA) renewal rely on the Department of Elementary and Secondary Education’s (DESE) annual performance report (APR).[2] The APR is divided into five sections. One section captures overall achievement and achievement growth as measured by the state assessment. Another section captures achievement and achievement growth among a “super subgroup” of disadvantaged students. The super subgroup includes FRL-eligible students, racial and ethnic minorities, English language learners, and students with special needs who are served under an individualized education plan (IEP).

For CEP districts and charter LEAs, “super subgroup” accountability outcomes (e.g., achievement) are identical to outcomes for the entire population because all students are identified as FRL eligible. Put another way, the CEP nullifies the super-subgroup section of the APR because the super subgroup can no longer be distinguished from the general population.

Like many states, Missouri’s funding policies have also historically relied on FRL data. State funding is weighted based on factors related to student need, including the number of FRL-eligible students. In legislation passed in 2013, lawmakers anticipated the impact of the CEP on the weighting factors in the state funding formula and put provisions in place to prevent CEP adopters from drawing additional funding based on their 100% FRL eligibility. Specifically, for CEP adopters, funding based on FRL-eligible students is frozen based on the number of eligible students in the year prior to CEP adoption.[3] While there is no sunset date on this provision, it is not a permanent solution. As schools and districts get further and further removed from initial CEP adoption, the applicability of the pre-CEP FRL share declines.

What Is Known?

The Data Effects of the CEP

Given the substantial stakes associated with FRL data, it is unsurprising that the data implications of the CEP have caused much consternation.[4] But can we quantify precisely what the CEP has done to FRL data? The answer is yes. And while the CEP has undoubtedly reduced the ability of FRL data to proxy for student disadvantage, particularly at the school level, its effect on the data is more modest than one might expect.

An analysis of administrative microdata from Missouri shows that the majority of students who attend CEP schools—about 80 percent—would be eligible for free and reduced-price lunches in the absence of the policy.[5] This finding is likely to generalize broadly because it is the result of CEP eligibility rules. Given this, relatively few students change FRL status with the introduction of the CEP. This can be illustrated with a simple back-of-the-envelope calculation. For example, if roughly 20 percent of schools adopt the CEP, and 20 percent of the students in these schools (on average) change FRL-eligibility as a result, then just four percent of students would experience a change in FRL-eligibility due to the CEP. The accuracy of this type of rough calculation is confirmed in the detailed analysis of microdata performed by Koedel and Parsons (2019).

While it is of some comfort that only a small fraction of students experience a change to their FRL-eligibility due to the CEP, this does not nullify the policy concerns raised in the preceding section. Notably, CEP schools still circumvent the spirit of DESE’s subgroup accountability policy in Missouri, and the funding implications of the CEP remain problematic.

What Is Not Known?

Navigating a Shift Toward the use of Direct Certification Data

We have relied on FRL data to inform accountability and funding policies in the U.S. for so long that little is known about how changing the basis for measuring poverty will affect these policies. The most common response of states to the CEP has been to shift their poverty-focused accountability and funding systems to be based on direct certification (DC) data. Students are directly certified for free meals if they participate in other means-tested programs such as the Supplemental Nutrition Assistance Program (SNAP), Temporary Assistance for Needy Families (TANF), and the Food Distribution Program on Indian Reservations. Students can also be directly certified if they are classified as foster, migrant, homeless, or runaway.

Even in the post-CEP era, research from Missouri finds that DC data do not do a superior job of identifying disadvantaged students relative to FRL data.[6] Still, DC data are not less effective in identifying student disadvantage than FRL data, and a case can be made for switching to DC data for several reasons. First, DC data are cheaper to collect—districts and states can essentially plug into data collected by other agencies—and subject to rigorous accountability controls. Second, DC data preserve within-school variability in student status, restoring the functionality of subgroup-based accountability policies like in the example described above for Missouri. Third, DC-based metrics address the funding problem created by the CEP shifting a mass of schools to “100% high poverty.”

Several complications arise with a shift to DC data that states must consider. First, the poverty threshold for direct certification is more stringent than for FRL eligibility (130 percent of the poverty line compared to 185 percent), and correspondingly fewer students are directly certified. An issue that must be resolved is that the simple statistics used in state funding formulas, like the number of disadvantaged students, are affected by switching to DC data because of the more stringent poverty threshold. States moving to DC-based metrics have addressed this issue by multiplying the DC share by a number greater than 1.0 in order to better align schools’ and districts’ DC percentages with their pre-CEP FRL percentages. The post-adjustment DC counts fit better within existing policy frameworks. The typical multiplier is 1.6, although this has been subject to debate and the proper multiplier can vary by context and purpose.[7] An adjustment similar in spirit would be to increase the amount of aid targeted per DC student.

Another issue that merits consideration as states navigate this change is that some student populations are systematically less likely to participate in the broader, non-school-based social safety net programs that lead to direct certification—most notably Hispanic students and undocumented immigrants.[8] States with large Hispanic and immigrant populations have the potential for measured poverty to shift markedly in some schools with a transition from FRL- to DC-based metrics. More research is needed to understand the scope of this problem and potential policy responses. A general solution is to bring in outside data to augment DC data, as discussed in a recent report from the Massachusetts Department of Elementary and Secondary Education.[9]

Policy Levers and Challenges

Opportunities to Develop and Use New Metrics

Given research showing that the CEP effect has been fairly modest thus far, there is no need to rush into a quick fix to address the data problems created by the CEP. In fact, the uncertainty around FRL data brought on by the CEP can be viewed optimistically, in that it can be leveraged as an opportunity to usher in a more rigorous and thorough approach to measuring student disadvantage, whether FRL- or DC-based (and potentially inclusive of other measures of disadvantage, like student mobility). State data systems in place today are far superior to what was available when accountability and funding policies based on FRL data were developed. The longitudinal information in modern systems can allow us to paint a more complete picture of students’ circumstances, both by combining multiple sources of data and by accounting for students’ histories of disadvantaged circumstances. One possibility is to construct a multi-input continuous index of disadvantage for each student that could mitigate some of the limitations of individual poverty metrics in isolation. The specifics of constructing an index are beyond the scope of this policy brief, but the data systems and analytic tools exist to produce and validate such an index. Given that states are already rethinking the data infrastructure upon which their policies in this area are built, it is a good time to push harder on the data to develop more comprehensive measures of student disadvantage.

Policy with Purpose

Policy discussions on measuring poverty during the CEP era should maintain focus on whether the metrics we choose capture the risk factors we aim to address. For example, if we are identifying poverty because of its correlation with academic achievement outcomes, persistence of poverty merits more attention.[10] Concentration of poverty matters as well. In addition to being independently important, schools with concentrations of low-income students often have less experienced teachers and other types of resource limitations compared to schools serving students of higher socioeconomic status.

The CEP has weakened the viability of FRL data as the sole measure of poverty status in public schools. The most common alternative states have turned to, DC data, has promising features but also limitations. The challenge facing policymakers today is how best to pivot from our historical reliance on FRL data. Fortunately, despite limitations with each of the individual measures of poverty that are available, collectively we have more information in state data systems than ever before that can be used to identify disadvantaged students and design and evaluate policies to support them.

Notes 

[1] Johnson, R., & Tanner, S. (2018). Money and Freedom: The Impact of California’s School Finance Reform on Academic Achievement and the Composition of District Spending. Getting Down to Facts II Technical Report. Palo Alto, CA: Policy Analysis for California Education (PACE).

[2] Local education agencies (LEAs) encompasses both traditional public school districts and independent charter schools.

[3]  §163.011(6), RSMo.

[4] Camera, L. (2019). Miscounting Poor Students. U.S. News & World Report. Retrieved 01.09.2019 at: https://www.usnews.com/news/education-news/articles/2019-01-07/why-its-getting-harder-to-count-poor-children-in-the-nations-schools.

Chingos, M.M. (2018). A promising alternative to subsidized lunch receipt as a measure of student poverty. Policy report. Washington DC: Brookings Institute.

Greenberg, E. (2018). New measures of student poverty: Replacing free and reduced-price lunch status based on household forms with direct certification. Education Policy Program policy brief. Washington DC: Urban Institute.

[5] Koedel, C., & Parsons, E. (2019). Using Free Meal and Direct Certification Data to Proxy for Student Disadvantage in the Era of the Community Eligibility Provision. CALDER Working Paper No. 214-0119-1.

[6] See Koedel, C., & Parsons, E. (2019). Using Free Meal and Direct Certification Data to Proxy for Student Disadvantage in the Era of the Community Eligibility Provision. CALDER Working Paper No. 214-0119-1. This finding is consistent with evidence that FRL data are effective in identifying disadvantaged students—see Domina, T., Pharris-Ciurej, N., Penner, A.M., Penner, A.K., Brummet, Q., Porter, S.R., & Sanabria, T. (2018). Is free and reduced-price lunch a valid measure of educational disadvantage? Educational Researcher 47(9), 539-555.

[7] Grich, R. (2019). New Strategies for Measuring Poverty in Schools. Explainer. FutureEd. Content retrieved 02.07.2019 at: https://www.future-ed.org/how-states-measure-poverty-in-schools/

[8] Massachusetts Department of Elementary and Secondary Education. 2017. Low-Income Student Calculation Study. Policy Report from the Massachusetts Department of Elementary and Secondary Education.

Zedlewski, S.R., & Martinez-Schiferl, M. (2010). Low-Income Hispanic Children Need both Private and Public Food Assistance. Policy Brief 2. Washington, DC: Urban Institute

[9] Massachusetts Department of Elementary and Secondary Education. 2017. Low-Income Student Calculation Study. Policy Report from the Massachusetts Department of Elementary and Secondary Education.

[10] Michelmore, K., & Dynarski, S. (2017). The gap within the gap: Using longitudinal data to understand income differences in educational outcomes. AERA Open 3(1), 1-18.

References

 Camera, L. (2019). Miscounting Poor Students. U.S. News & World Report. Retrieved 01.09.2019 at: https://www.usnews.com/news/education-news/articles/2019-01-07/why-its-getting-harder-to-count-poor-children-in-the-nations-schools.

Chingos, M.M. (2018). A promising alternative to subsidized lunch receipt as a measure of student poverty. Policy report. Washington DC: Brookings Institute.

Domina, T., Pharris-Ciurej, N., Penner, A.M., Penner, A.K., Brummet, Q., Porter, S.R., & Sanabria, T. (2018). Is free and reduced-price lunch a valid measure of educational disadvantage? Educational Researcher 47(9), 539-555.

Greenberg, E. (2018). New measures of student poverty: Replacing free and reduced-price lunch status based on household forms with direct certification. Education Policy Program policy brief. Washington DC: Urban Institute.

Grich, R. (2019). New Strategies for Measuring Poverty in Schools. Explainer. FutureEd. Content retrieved 02.07.2019 at: https://www.future-ed.org/how-states-measure-poverty-in-schools/

Johnson, R., & Tanner, S. (2018). Money and Freedom: The Impact of California’s School Finance Reform on Academic Achievement and the Composition of District Spending. Getting Down to Facts II Technical Report. Palo Alto, CA: Policy Analysis for California Education (PACE).

Koedel, C., & Parsons, E. (2019). Using Free Meal and Direct Certification Data to Proxy for Student Disadvantage in the Era of the Community Eligibility Provision. CALDER Working Paper No. 214-0119-1.

Massachusetts Department of Elementary and Secondary Education. (2017). Low-Income Student Calculation Study. Policy Report from the Massachusetts Department of Elementary and Secondary Education.

 Michelmore, K., & Dynarski, S. (2017). The gap within the gap: Using longitudinal data to understand income differences in educational outcomes. AERA Open 3(1), 1-18.

Zedlewski, S.R., & Martinez-Schiferl, M. (2010). Low-Income Hispanic Children Need both Private and Public Food Assistance. Policy Brief 2. Washington, DC: Urban Institute

163.011(6), RSMo.