Lessons From Early Implementation of Pandemic-EBT

Before the COVID-19 pandemic hit, nearly 30 million low-income children were approved to receive free or reduced-price meals at school each day. When schools closed, parents, administrators, and policymakers worried about how families would provide these children with ten extra meals each week even as workers’ earnings and jobs were disappearing. One key way federal policymakers responded was by promptly enacting Pandemic EBT (P-EBT) in March 2020, an entirely new program that allowed states to provide approximately $250 to $450 per child in grocery benefits (depending on the average number of days schools closed in the state) to make up for the meals missed in the spring of 2020.^[2]

The goal of P-EBT was simple: while schools are closed, to provide families whose children qualified for free or reduced-price meals with the funds that otherwise would have gone to schools to provide them with breakfast and lunch. But this was an entirely new program, requiring collaboration across the Supplemental Nutrition Assistance Program (SNAP, formerly food stamps) and school meal programs, data matching, and the mailing of millions of benefit cards.

Though P-EBT was optional for states and the Department of Agriculture (USDA) required them to bear half of the administrative costs, all 50 states, the District of Columbia, and the Virgin Islands rose to the challenge — even as they were dealing with countless other pandemic-related demands, changes, uncertainties, and costs. As a result, millions of families had additional resources to buy food for their children to provide the meals they normally would eat at school. The program’s success is a testament to the dedication and ingenuity of countless elected officials, public servants, and other stakeholders across the country.^[3]

With millions of children still not getting enough to eat^[4] and preliminary evidence showing that P-EBT substantially reduced food insecurity,^[5] policymakers in October extended P-EBT through fiscal year 2021. They also added flexibility to address the more complicated ways schools are combining virtual and in-person instruction, allow states to replace meals missed at child care for children in households receiving SNAP benefits, extend P-EBT to schoolchildren in Puerto Rico, and provide federal funding for all state administrative costs.^[6] States now have an important opportunity to enhance and expand the P-EBT program they built over the past few months to help ease the ongoing hardship low-income families face over the coming months.^[7]

To document how states operationalized P-EBT, the Center on Budget and Policy Priorities and the Food Research & Action Center conducted a rapid assessment project over the summer of 2020, primarily through a survey and interviews of state officials and other stakeholders.^[8] (See box, “CBPP/FRAC P-EBT Documentation Project.”) A longer report and other materials accompanying this report detail how state officials made decisions about how to implement P-EBT, operationalized the program, and worked to ensure that families received benefits.^[9] (See box, “Pandemic EBT Implementation Documentation Project: Full Report.”) This report reviews early findings regarding state implementation of P-EBT in the spring and summer of 2020 in order to support improved implementation for the current school year.

To deliver benefits, states had to design and staff a new program infrastructure, as well as create new policy to govern the program. They had to determine how to deliver benefits to eligible children, including assessing whether the state had sufficient information to issue benefits directly to eligible households or instead had to require and design a new application. They also had to establish a new communications plan to support the program and develop a mechanism through which families could ask questions about the new benefit and resolve issues. State agencies and school districts organized and managed this significant undertaking for millions of children despite operating in a new remote work environment, with limited federal guidance, amidst the fear and concern surrounding COVID-19, and while their agencies were transforming services for their students or program participants to a remote service delivery model.

Implementation took tremendous effort, creativity, and agility by state officials and other stakeholders. Our materials describing their work provide a rich source of information that state officials can draw on as they consider how to strengthen and streamline their P-EBT program, restructure it to accommodate new instructional models, and reach younger children.

This report, which summarizes key findings from the longer report and our own observations from working with state officials and anti-hunger advocacy organizations during implementation, describes the groups of children that states sought to reach, the data they drew on to issue benefits, and issues they might wish to consider as they plan their P-EBT program for the 2020-2021 school year.

Reaching Eligible Children

Schoolchildren were eligible for P-EBT meal replacement benefits if they had been able to receive free or reduced-price school meals prior to the pandemic and their school was closed for at least five days due to the pandemic. The eligible group last spring included an estimated 30 million schoolchildren. State officials typically break down the eligible groups based on the way the state could identify them and deliver benefits:

• Children in households already receiving SNAP benefits. Under federal law, all children in households receiving SNAP automatically qualify for free school meals. These children represent roughly half of P-EBT-eligible children nationwide and are the easiest for states to reach because they are known to the SNAP eligibility system and someone in their family already has an EBT card, since that is how SNAP benefits are delivered. All but two states (Louisiana and Wyoming) issued P-EBT benefits directly to these households without requiring any action by families, which is referred to as “direct issuance” in this report, and all states but California added P-EBT benefits to the household’s SNAP EBT card.
• Children in households receiving other assistance. Thirty states were able to use information about children assisted by programs other than SNAP that also confer eligibility for free school meals to issue P-EBT benefits directly, without requiring any action by families. These programs include Temporary Assistance for Needy Families cash assistance (25 states), Medicaid (16 states), and foster care (15 states). Although these children are known to a state’s eligibility system, their families typically do not have EBT cards, so states generally mailed new P-EBT cards to these households.
• Additional children receiving free or reduced-price school meals. This group includes children approved for free or reduced-price school meals because of their household income or because their school offers meals at no charge to all students (such as high-poverty schools operating under the Community Eligibility Provision).^[10] Most states did not have statewide lists with the data needed to issue benefits to these households, so they gathered information using one of the following approaches.
  • Direct issuance: When states have sufficient information about eligible children, they can issue benefits directly to that household. Thirty-one states used information from state education agencies, school districts, and schools to mail a new P-EBT card directly to the family at the address on file with the school, without requesting additional information from the family. In these states, families did not have to take any action to request P-EBT benefits, unless there was missing or inaccurate data.
• Application (or other information collection): When states did not have enough information to directly issue benefits, they gathered it through an application process. Twenty-five states required at least some parents to apply for P-EBT or submit a data-collection form, which typically required the name of the head of the household, the current mailing address, the student’s name, and the student’s date of birth.^[11]
• Newly eligible children. Thirty-four states reported extending P-EBT benefits to children in households that lost income due to the pandemic and thus became newly eligible for free or reduced-price school meals. At least 12 states added children whose families were enrolled in SNAP after the pandemic began. Likewise, 21 states provided P-EBT benefits to children whose family submitted a free or reduced-price meal application to their local school district, and at least one of those states (Oregon) set up a statewide school meal application for newly eligible families.^[12] But not all school districts routinely accepted and processed school meal applications, and not all districts explained to families that there was a reason to apply for free or reduced-price school meals when schools were closed.

How States Obtained Necessary Information to Issue Benefits

Figure 1 shows, for families not receiving SNAP benefits or another form of assistance that allowed for direct issuance, the states that issued benefits directly to eligible children because they had sufficient information to do so, those that used an application or other form, and those that used a combination of both approaches. As P-EBT implementation has evolved, the distinctions between these models have blurred somewhat. For example, several states (including Arizona and Montana) that initially issued P-EBT directly based on information from school districts later developed online forms, similar to an application, to collect information from families that needed to update their addresses or other inaccurate data. Other states (including Georgia, Tennessee, and Wyoming) that had required some families to submit an application shifted to direct issuance in the fall out of concern that many families had not applied.

States Can Strengthen, Streamline P-EBT for 2020-2021 School Year

The P-EBT programs that replaced school meals missed during the spring of 2020 reached a remarkable number of children quickly at a time of great need. But states had to make decisions rapidly, often without full information, and under difficult circumstances. State officials now have the benefit of experience, more familiarity with how their agencies operate during a pandemic, and more time to refine their P-EBT programs. During our interviews, state officials shared their reflections on what worked well and where they’d like to make improvements. Those observations are summarized in the full report on our project, “Pandemic EBT Implementation Documentation Project.”

This section draws on key findings from the documentation project and our own observations from working with state officials and anti-hunger advocacy organizations during implementation to offer several key recommendations for states to consider as they design their P-EBT programs for the 2020-2021 school year. While many of these recommendations would apply to providing P-EBT benefits to younger children as well, they focus on providing benefits to school-age children and do not cover the full range of issues states will need to consider when designed P-EBT programs for younger children. By taking these steps, states can provide benefits to more eligible children with less burden on families and program administrators.

Directly Issue Benefits to Children Whenever Possible

Issuing benefits based on data the state already has, without requiring families to apply, can reach more eligible families with less burden on struggling families and less work for state administrators. To be sure, gathering the necessary data from other programs (such as TANF or Medicaid) and from the state education agency, school districts, and schools can be difficult, and the data itself may have limitations. But the alternative — developing a new application, informing families about it, and responding to inquiries from families — can make benefit issuance slower and more complicated.

In states that can issue benefits directly to children, several features would improve the reach and ease of issuing benefits:

• Collaborate early. Close collaboration between the SNAP and Child Nutrition agencies led to smoother implementation and made it more likely that state officials anticipated and addressed data challenges when developing an implementation plan.
• Improve data matching. For more than a decade, school districts have been required to automatically enroll children in households receiving SNAP benefits for free school meals using a data-matching process known as direct certification. Over the years, states have substantially improved their data-matching processes, bringing down the share of children inadvertently missed from 32 percent to 8 percent.^[13] With a little more time to develop implementation plans, states can apply some of the successful data-matching processes used for direct certification to their data matching for P-EBT benefits.
• Do not require redundant school matches for children who are categorically eligible for free school meals. For districts in which all schools are offering only virtual instruction, states can reduce the number of children missed by data matching by issuing benefits to all children in households participating in SNAP, TANF, the Food Distribution Program on Indian Reservations (FDPIR), or Medicaid^[14] or who are in foster care, homeless, or migrant to the maximum extent allowable, without confirming which school the child attends.^[15] The cost of potentially including in P-EBT a very small group of low-income children who attend schools that do not participate in the federal child nutrition programs or who are homeschooled isn’t worth the burden of conducting this match and the consequences of missing eligible children.
• Issue benefits to eligible children rather than their caretakers. About half the states issued spring 2020 benefits to the head of household even when this forced the state to create a household P-EBT application. But states that instead issued the P-EBT card to children routinely, or when parent information was not available, found that this worked smoothly, eliminated the need to group children by household, and made direct issuance possible in situations where information about the parent or guardian was not available from education data. It made it easier to resolve situations where a card was inadvertently sent to a non-custodial parent.
• Improve address accuracy. Ideally, states or school districts could provide a mechanism for families to update their mailing address even before issuing benefits. For example, the School District of Philadelphia used robocalls to alert families that they would receive P-EBT benefits; the call included the address currently on file with instructions about how to update it. Providing such automated calls or texts and establishing a simple, secure mechanism for providing updated data could slow the provision of benefits, but lowering the number of cards mailed to the wrong address might be worth this delay.
• Provide a resolution process. Providing a clear and secure process for families inadvertently left out of direct issuance can reduce their frustration and reduce the number of inquiries states have to field. For example, 11 states developed an online inquiry form that families could submit. The most useful system would provide real-time information on whether a household’s benefits had already been issued (and if not, when they would be issued), the address on file, and a mechanism for updating it.

Make Applications Accessible

States that cannot obtain complete or current data to issue benefits directly to all eligible children will have to provide an application to families to collect the needed information. While more burdensome for states and families, applications have the advantage of collecting current address information so that fewer P-EBT cards are mailed to outdated addresses. If an application is unavoidable, the following features will make it more accessible to the full range of eligible families.

• Keep the application simple and short. If the form is simple and short, eligible families are likelier to complete it and do so correctly. It is also important for the application to work well on a mobile phone because that is the main source of internet access for many low-income families. Another way to simplify the application is to ask only for the information necessary to issue benefits. Some states, for example, asked for the Social Security number of the head of household or the child but (in keeping with USDA guidance) did not require families to submit it.^[16] By eliminating such optional questions, states can reduce the likelihood that families decline to submit an application because the household includes someone who doesn’t have a Social Security number or has privacy concerns.
• Make the online application public. Several states made their application available only to families approved for free or reduced-price meals, via a private application link sent by their school district. In these states, not all eligible families received the application link, so families reached out to the state and community partners for assistance, creating backlogs and confusion. In two states, the application links were eventually published by the state.
• Provide the application in multiple languages. Federal law requires that communications with families relating to the school meals programs must be, to the maximum extent practicable, in a language that they can understand.^[17] Ideally a state would provide the P-EBT application in the languages in which it offers school meal applications. Minnesota, for example, made its application available in five languages in addition to English (Hmong, Russian, Somali, Spanish, and Vietnamese).
• Provide an alternative to an online application. It is important to provide an alternative for families without internet access, especially since public computers may not be available due to the pandemic. For benefits covering the spring of 2020, of the 25 states that used an application, at least ten states offered a paper application and at least 16 states offered a way to complete the application by phone.

Develop a Robust Communications and Outreach Plan

USDA required states to develop a public information campaign about P-EBT, but state plans varied considerably. By developing clear communication plans that explain to families how P-EBT benefits work, the state, local SNAP agencies, school districts, and community-based organizations can reduce confusion and the number of inquiries to which states must respond. The state officials we interviewed said they would like to approach this area more deliberately in the future.

• Communicate key messages to the public. Responses to our survey indicated that the most common inquiries from families related to eligibility, benefits going to some but not all school-age children in the household, and benefit status. States can reduce the number of inquiries and reduce confusion for families by providing information about when benefits will be issued, when cards will be mailed, how to activate cards, and what to do if a family does not receive benefits or has difficulty using their card; sharing it broadly; and updating it as needed. Now that P-EBT has been extended, it is also important to advise families to keep their cards because they might receive additional benefits in the future.

• Equip partners. In addition to sharing key information publicly, states can share it with organizations that might receive inquiries from families, such as local SNAP offices, school districts, schools, advocacy organizations, and community-based service providers. It also would be helpful to inform these stakeholders how they can be most helpful in disseminating the information, give them easy-to-share materials, explain the circumstances under which they should refer an inquiry to the state, and provide points of contact for them and families.

  In partnership with anti-hunger advocates, many states, such as Colorado, developed a P-EBT communications toolkit that included these pieces and shared it with interested partners for easy dissemination.^[18] Many states also hosted a webinar to share information about the program and provide a venue to answer questions from partners to ensure they were equipped to help families. School districts can be important envoys since they are in regular contact with families about other school-related matters. By requiring school districts to inform families about P-EBT, as Texas did, states could ensure that eligible families receive information from a trusted source.

• Communicate directly with families. By communicating directly with eligible families, states can provide tailored information, such as how to activate a P-EBT card. It is important to communicate both with families receiving SNAP benefits and with those who are not, using the following approaches:
  • Send families receiving SNAP benefits a notice, text, or automated call when P-EBT benefits are issued, explaining what they are for and how much the family will receive for each child and in total.
  • Clarify that P-EBT benefits are distinct from SNAP benefits, especially when communicating with families not already receiving SNAP, who might otherwise think they are receiving unsolicited SNAP benefits.
  • Work closely with the EBT vendor to ensure communications to families explain clearly when and how P-EBT cards will be sent and how to activate them.
  • Design identifiable envelopes for mailing cards (which also protects privacy) and show families what to look out for so they won’t discard the cards because they do not understand the cards’ purpose or suspect they are a scam.
• Collaborate to reach certain children. State officials and stakeholders we interviewed described difficulty in reaching children in foster care, children who are homeless, and children in immigrant families. By partnering with service providers and community groups that routinely work with these children and families and have earned their trust, states can more effectively deliver benefits to them. For example, partner organizations might be able to develop messaging that addresses families’ specific concerns. In the case of children experiencing homelessness, the school district’s homeless education liaison can help with communication and distributing EBT cards.
• Include newly eligible families. To reach children who become eligible when families lose income, it is important to provide a straightforward way to access P-EBT benefits without having to apply for SNAP. For example, a state could provide a statewide online combined application for free or reduced-price school meals and P-EBT benefits. If a combined application is not possible, then a statewide online application for free or reduced-price school meals would be helpful. It is also important to publicize the mechanism widely, including through school districts’ communications channels, and explain why families should apply for free or reduced-price school meals even if their child is receiving virtual instruction.

Simplify Card Activation

After a family receives a new EBT card, it needs to activate the card and set a Personal Identification Number (PIN) to use at the grocery store. Given the need for speedy implementation in the spring, many states (and their EBT vendors) did not adapt their SNAP process for families to activate their new P-EBT cards or update their call center prompts to better accommodate P-EBT. This caused confusion and may have led some families to not complete the activation process. States can take steps to minimize these problems.

• Use information other than Social Security number (SSN) for card activation. SNAP’s typical approach of using the head of household’s SSN and date of birth to activate cards does not work well for P-EBT, which does not require families to report their SSN in order to receive benefits. Instead, many states successfully used student date of birth, zip code, or another number for card activation.
• Adapt the SNAP call scripts for P-EBT card activation. Some states reported that many families were confused about the automated telephone prompts for P-EBT. One of the EBT vendors did not make any changes to its automated telephone prompts to reflect the different process for activating a P-EBT card. By adapting the script, states and EBT vendors can reduce confusion for households and extra work for state staff.

Adequately Staff P-EBT

To deliver benefits for the spring of 2020, states invested in staffing the new P-EBT program in a variety of ways. The extent to which they could dedicate staff or invest in technology to support the program appears to have affected how quickly and effectively they issued benefits and addressed problems. The variation in state investment in part reflects USDA’s requirement that states contribute half the of the administrative costs of designing and implementing the program.

The legislation extending P-EBT through fiscal year 2021, however, provides federal funding to cover all administrative costs associated with operating P-EBT; this change should allow states to assess how much staffing is needed to run an effective program and reassign or hire as needed. When developing staffing plans, it might help to consider the following areas in addition to the resources needed to obtain data, conduct data matches as needed, issue benefits, and mail P-EBT cards.

• Anticipate troubleshooting. Troubleshooting was required for some families regardless of whether benefits were issued directly were based on an application, or a combination of those approaches. To manage troubleshooting for benefits for the spring of 2020, nearly every state utilized a new or existing call center and supplemented it with an email address, online inquiry form, or online benefit status portal. Nonetheless, backlogs of inquiries developed in some states we interviewed. Some states created a troubleshooting process after many families began experiencing issues because of how quickly things moved.

  Developing a clear process from the beginning would have reduced frustration for families and school districts, as well as the workload for states from addressing the same inquiry multiple times. A strong troubleshooting approach would have clear points of entry, give families immediate feedback on when to expect a response, consolidate inquiries regardless of how they are submitted, have clear protocols for how to respond to routine inquiries, and include training staff on how to respond to special but predictable situations, such as children in foster care, children who are homeless or in unstable housing, divorced or separated parents, and custody disputes. Once a plan is in place, it can be adequately staffed and the process can be clearly communicated to other agencies, school districts, community partners, and families.

• Consider agency roles. P-EBT required state SNAP agencies and Child Nutrition agencies, which are usually within state education departments, to collaborate. State officials we interviewed valued the ways in which launching P-EBT strengthened existing relationships and led to new ones. Now that these relationships are in place, states can consider the most sensible role for each agency for this school year and how best to share information across agencies. For example, in some states the SNAP and Child Nutrition agencies managed different aspects of the program; states could consider consolidating information and program management into a single agency or team to reduce the time spent sharing information across agencies and ensure that consistent information goes to partners and the public.
• Engage community partners. Anti-hunger organizations, community-based service providers, and other community partners can be important allies in delivering benefits to eligible families. Such organizations could be especially helpful in conducting targeted outreach, helping families complete an application, informing a broad array of community-based organizations about P-EBT, and troubleshooting when eligible families do not receive benefits. During the spring and summer of 2020, some states worked with community partners to conduct targeted outreach or manage family inquiries. As states plan their staffing and budgets, it is worthwhile to consider whether conducting outreach through contracts with community partners could more effectively reach and help certain families and alleviate burdens on state staff.

Assess Progress

Setting goals with timelines and measuring progress are important tools for building on successes, identifying challenges, and modifying the P-EBT program as needed. Goals could include, for example, reaching all eligible children. States could measure their progress by assessing the share of eligible children who receive benefits, the share of new P-EBT cards that are activated, and the share of benefits on new cards that are redeemed. Another goal might be to keep the response time to inquiries low. By closely monitoring data and assessing progress toward identified goals, states can work with partners and address issues as they emerge. This could be especially helpful in the following areas.

• Monitor applications or card activations. By regularly collecting data on applications or card activations and comparing it to the potential numbers of activations, states can assess their progress in providing benefits to eligible children. The more local the data, the more useful it is in guiding targeted outreach, technical assistance, and troubleshooting.
• Make adjustments. Monitoring data in real time can allow program administrators to adjust plans and operations to address shortcomings. For example, many states tracked application rates, and in response to low rates, Georgia, Tennessee, and Wyoming developed a mechanism to directly issue P-EBT cards, while other states extended their application deadline and conducted additional outreach. When officials in Ohio realized that many directly issued P-EBT cards had been mailed to incorrect addresses, they developed a simple online form for families to update their address.^[19]
• Share data. Sharing data publicly or with community partners can help them support state efforts. For example, the Massachusetts SNAP agency regularly shared zip code and city-level data with advocates, schools, and other outreach partners showing the percentage of P-EBT cards that had been activated; outreach partners then used this localized data to focus resources in communities with the lowest card activation rates. The Minnesota SNAP agency shared county-level data on the share of eligible children who had applied for P-EBT, which allowed an anti-hunger organization to run targeted social media and radio ads. Likewise, the Texas SNAP agency regularly provided application data at the county and zip-code level to education and advocacy partners, which used the information to target paid radio, television, and social media ads.

Conclusion

In the spring of 2020, state officials acted quickly to launch new P-EBT programs to provide grocery benefits to replace the breakfasts and lunches that low-income schoolchildren were missing because their schools had closed. Their tenacity and creativity helped millions of struggling families feed their children.

P-EBT has now been extended for the 2020-2021 school year and adapted to meet the variety of school closure and hybrid approaches that are now in place and may develop over the school year. As states assess potential changes in their programs going forward, an understanding of how states implemented spring 2020 P-EBT programs can serve as a starting point for state officials and other stakeholders. By making deliberate decisions about the key implementation considerations described in this report, state officials can strengthen and streamline their P-EBT programs.