Introduction

Recent estimates show that North Carolina is host to nearly 1 million HispanicFootnote 1 residents, representing 10% of the overall North Carolina population. With approximately 1 out of 10 births in 2008 to Latina mothers, the growth of the Latino child population in North Carolina is outpacing that of children in any other racial or ethnic group. Hispanics in North Carolina reside in certain urban counties but also constitute a growing share of the population in many less populated, rural counties such as Duplin, Lee, and Sampson counties (U.S. Census Bureau 2019). Because this ethnic demographic group has the fastest growing population in the state, North Carolina is sometimes characterized as an emerging destination state. On the other hand, North Carolina counties such as Forsyth, the home of Winston-Salem, have a long history of being home to Latinos, many of whom originally arrived as migrants working in the agricultural sector; future waves of workers supported the growth of the tobacco industry. In fact, until the 1990s, when immigration enforcement strategies shifted, Hispanics in Winston-Salem were welcomed and integrated into the community as emerging members of the middle class. Under this unique and evolving backdrop of population diversity and growth, North Carolina—as a state in the racially segregated Southeast—offers a unique context for understanding the experiences of social services and receipt of income assistance benefits for Hispanic families and children.

Income assistance can serve as a support for children’s health and cognitive development; and yet, despite eligibility, Hispanic families are less likely to receive benefits, particularly in comparison with their African American peers (Coleman-Jensen et al. 2016; Duncan et al. 2014; Hoynes et al. 2016; Lee and Mackey-Bilaver 2007; Williams 2013). Connections to social assistance and services, like connections to the formal labor market, may also serve as gateways to other types of support for children (Golden et al. 2013) and foster social inclusion in ways that can yield positive intergenerational returns. Despite these benefits to family and child well-being, little information is available about how state and local safety-net policies and practices are implemented for and experienced by Hispanic families and how these experiences may subsequently affect their uptake of benefits.

With new survey and qualitative data collected in 2019 and 2020, we describe the perspectives of practitioners and administrators serving and assisting Hispanic families and children in North Carolina. While scholars have probed how African Americans have been excluded from and disadvantaged by social policies, Hispanic populations’ unique challenges in accessing safety-net programs have not received comparable attention. Low-income Hispanic families who are eligible for social services and income assistance face challenges related to English language and literacy proficiency, the complexity of relationships of household members typically used to determine eligibility, and citizenship status of family members and other members of the household. Theories of administrative burden suggest that Hispanic families face high learning, compliance, and psychological costs that are compounded by these challenges.

We posit that Hispanic families may be caught in racial division crossfires and that strategies intended to exclude African Americans, resulting from decentralized authority over the implementation of social benefits, have spilled over to Hispanic families under the agenda of immigrant exclusion. Finally, anti-immigrant policies and rhetoric—particularly heightened in the current political landscape—have outsized chilling effects among Hispanics, even among citizen families (Haley et al. 2020). We probe and report on the perspectives and characteristics of practitioners, their work environments, and their perceptions of services and delivery of social safety-net benefits for Hispanic families.

How the Safety Net Benefits Eligible Families and Children

Historically, many of America’s safety-net programs grew out of the New Deal and President Lyndon B. Johnson’s War on Poverty in the 1960s. These include cash assistance (formerly called Aid to Families with Dependent Children and now Temporary Assistance for Needy Families, or TANF) and nutrition assistance programs such as the Special Supplemental Nutrition Program for Women, Infants, and Children (WIC), Supplemental Nutrition Assistance Program (SNAP), and subsidized school meals hosted by the U.S. Department of Agriculture. The safety net is a source of income and a mechanism for poverty reduction for eligible children and families (Hoynes et al. 2016). Additional income enables parents to buy goods and services for their families that support their children’s development. These include higher-quality housing and nutrition; more stimulating and stable home environments; learning opportunities outside of the home; and, because of the ability to reduce or restructure work hours, more parental time available to supervise, monitor, or spend with youth. Additional economic resources may also reduce parents’ own stress and improve their mental health (Masarik and Conger 2017), allowing parents to devote more attention to their children, thus providing a more predictable family life, less conflicted relationships, and warmer and more responsive interactions (Newland et al. 2013; White et al. 2015).

How the safety net supports families may vary by racial/ethnic group. Financial buffers that absorb negative income shocks and support families in smoothing consumption—such as credit, savings, and assets—are more available to white families than to African American or Hispanic families (McKernan et al. 2011; Taylor et al. 2011). Rates of insurance coverage to cover catastrophic events and enable preventive care that would minimize the subsequent likelihood of a health crisis or accumulation of debt also substantially vary by racial/ethnic group (Artiga et al. 2016; Smith and Medalia 2015). Financial management strategies and availability of those who contribute to household finances also vary by racial/ethnic group partly as a function of differing social and family networks (Calzada et al. 2013).

The most direct evidence about the impact of income assistance is from studies of welfare reform that find positive effects of cash assistance, typically combined with work requirements, on preschool- to school-aged children’s schooling outcomes (Chase-Lansdale et al. 2011; Gennetian and Miller 2002; Gennetian and Morris 2003). Welfare waiver experimental studies of the 1990s also show improvements in behavioral problems of preschool-aged children, both internalizing and externalizing, in response to their parents’ increased income and employment, though fewer impacts on adolescent risky behavior (Bitler et al. 2016; Gennetian et al. 2004; Gennetian and Miller 2002; Gennetian and Morris 2003). The welfare reform programs of the 1990s showed different effects for children of different racial and ethnic groups, with African American and Hispanic children benefiting most from programs that had “education-first strategies,” as compared with white children (Yoshikawa et al. 2010). Analyzes of federal programs such as the Earned Income Tax Credit and studies of food subsidy programs such as SNAP and WIC from the U.S. Department of Agriculture Research have the largest poverty reduction impacts for children (Bitler 2015; Schmeiser 2012). The WIC program also has favorable effects on birth and child outcomes, including the nutrition, health, and academic performance of children exposed to the program (Bitler and Currie 2005; Carlson and Neuberger 2017; Chorniy et al. 2020; Jackson 2015; Kowaleski-Jones and Duncan 2002).

Racial and Ethnic Differences in Receipt of Social Programs and Administrative Burden

Figure 1 shows trends in receipt of benefits of children in families from safety-net programs in the United States and North Carolina. At its highest point in 2013, nearly 52% of children in Black families received benefits, compared with 39% of children in Hispanic families and 20% of children in white families. In all years from 2005 to 2018, children in Hispanic families were always more likely to receive benefits than children in non-Hispanic white families and less likely to receive benefits than children in Black families. The trends by race and ethnicity among children in families in North Carolina mimic the trends in the United States. These racial/ethnic discrepancies in utilization rates of means-tested programs are not well understood or researched. Political historians point to the exclusion of African Americans from Social Security through provisions that omitted agricultural and domestic workers, along with exclusion from home ownership and public housing through discriminatory practices (Katznelson 2005; Quadagno 1994; Rothstein 2017). In more recent years, scholars have identified aspects of policy implementation that are “onerous,” discouraging uptake (Burden et al. 2012). For example, Moynihan and colleagues (2015) describe the cost of learning about programs and determining whether one is eligible for them as “administrative burden,” which can also include compliance costs, the costs of providing documentation for application and recertification processes, and the difficulty of following vague program rules. Finally, individuals may experience psychological costs from the stress and stigma of accessing benefits.

Fig. 1
figure 1

Source: National KIDS COUNT, analysis of data from the U.S. Census Bureau, 2005, 2008, 2010, and 2013 to 2018 American Community Survey

Children in families that receive public assistance by race and ethnicity, in North Carolina and the United States. Population of children under age 18 in families that receive Supplemental Security Income (SSI), cash public assistance income, or Food Stamps/SNAP in the previous 12 months (NC vs US average).

While scholars who study administrative burdens have not yet demonstrated how these particular costs shape program uptake among members of racial and ethnic minority groups, research suggests that members of these groups (particularly African Americans) experience less generous benefits and more punitive programs, along with disparate treatment from caseworkers (Barnes and Henly 2018; Gooden 2006; Soss et al. 2011). For example, Hardy et al. (2019) and Michener (2019) connect devolution (i.e., delegation of responsibility for service provision from the federal to state and county level governments) to race-based variation in state spending on programs like TANF and Medicaid, and multiple researchers demonstrate that relative to other racial groups, low-income African American recipients are subject to less generous TANF programs (Hardy et al. 2019; Hero and Levy 2017; Parolin 2019). Further, states with a greater share of African Americans and greater income inequality by race spend less on cash assistance; using state-year panel data from 1997 to 2014, Parolin (2019) finds that states with larger percentages of African Americans spend less on cash assistance, shifting funds away from government assistance with the intent of discouraging single motherhood.

Along with less generous policies, research suggests that federal and state legislators, as well as program administrators, create administrative burdens in the form of punitive rules and complex processes to access and maintain benefits (Moynihan et al. 2013). Scholars suggest that these burdens stem from the ideological leanings and “hidden politics” of state and local governments and the racial makeup of states (Moynihan et al. 2015). Studies on welfare reform demonstrate that African Americans are subject to greater compliance costs in the form of stricter, more punitive policy rules. States with a greater share of African Americans on TANF enacted stricter family cap rules and time limits. As Soss and colleagues (2001) conclude, the “‘devolution revolution’ has created openings for new forms of racial inequality that disadvantage African Americans in the U.S. welfare system. Because states with more black recipients have adopted stricter policy regimes, black families are now more likely to participate under the most punitive program conditions” (p. 390).

These burdens that members of racial minority groups experience are not limited to complicated or stringent program rules but also include the psychological costs or the stigma and stress of program experiences, including interactions with workers that “violate basic psychological needs for autonomy and respectful treatment” (Moynihan et al. 2015, p. 5). Unlike universal programs, means-tested programs may signal exclusion from normalized or typical full citizenship (Campbell 2014). Front-line workers can impose these costs by erecting additional barriers to successful claims (Brodkin and Majmundar 2010; Kogan 2017), subjecting clients to scrutiny (Watkins-Hayes 2011), or leveraging policy rules to sanction or deny access to programs (Kogan 2017). Further, caseworkers offer information about other services, such as child care, work readiness programs, and education and training programs, more often to white recipients than to African American and Hispanic recipients (Hahn et al. 2015; Holzer and Stoll 2001).

Even though children are typically viewed as deserving recipients of public assistance, parents may be less inclined to enroll on their children’s behalf if they may experience stigma, escalation of feelings of failure, and thoughts about their inability to provide for their children. This is particularly problematic when means-tested programs may serve as gateways to other services that are less stigmatizing. Irrespective of a family’s intentions or needs, the design of social assistance and characteristics of the labor market can affect length or receipt of benefits, the quality and stability of benefits, and subsequent utilization. African American and Hispanic recipients are found to be sanctioned—that is, face a reduction or cessation of cash assistance due to noncompliance with work or related requirements—at higher rates than white recipients after controlling for factors like their work history and the ages of their children (Monnat and Bunyan 2008). African Americans have less stable employment, are hired less often, and are more likely to cycle back to TANF (Holzer and Stoll 2001) irrespective of the strength of the labor market.

Hispanic families may be especially disadvantaged by policies and practices that have traditionally excluded African Americans. In particular, elevated burdens like learning and compliance costs, along with the psychological costs of enduring anti-immigrant sentiment, may dissuade program uptake. Factors for Hispanic families such as language barriers, legal and citizenship status requirements, and fear of deportation or other immigration-related concerns may be particularly salient in considering safety-net program participation (Martinez et al. 2015; Rhodes et al. 2015; Sandstrom et al. 2014; Toomey et al. 2014). Nearly 60% of all children born to immigrants are born to Hispanics (Greenberg et al. 2018), and estimates suggest that up to one in four Hispanic children may be residing with an unauthorized immigrant parent (Yoshikawa et al. 2010). Although these characteristics alone might not be a determinant of ineligibility for other members of the household, they may serve as residual deterrents.

For Hispanic families, residency, citizenship, and eligibility determination can be characterized as administrative exclusion under the umbrella of anti-immigrant agendas and administrative burden. These factors can obstruct access assistance, even among those states that might collectively be characterized as providing generous benefits. In a recent review of state policies and practices, North Carolina, in comparison with 12 other highly Hispanic states, lacked an online TANF cash assistance application as of 2019 and requires health screening and child immunization documentation unlike its neighbor Southeastern states included in this review (Gennetian et al. 2020).

Current Study

This study aims to extend insights on implications for Hispanic families residing in a Southeastern state and having system-level features associated with inaccessible and relatively less generous safety nets—in particular, a conservative and resource-constrained state with a high concentration of low-income African Americans. We explore how Hispanic families may be caught in the crossfire in a state where policy administration and practice have typically disadvantaged African Americans.

Have strategies that led to African American exclusion spilled over to Hispanic people under the agenda of immigrant exclusion? To answer this question, we probe characteristics of practitioners, including their work environments and their perceptions of services for social safety-net programs designed to support economic well-being of families. We apply a mixed methods approach to capture (i) appraisals of TANF county administrators’ practices and perceptions across the state through survey items and (ii) perspectives of workers and some administrators, in their own words and voices, across similar agencies charged with TANF, WIC, Medicaid, and SNAP administration. This approach—employing quantitative and qualitative data about on-the-ground practices and perceptions—provides new descriptive information for the field.

Methods

First, 49 in-depth qualitative interviews were conducted with social service administrators and front-line workers in five counties (three rural) across WIC, TANF, SNAP, and Medicaid programs. Staff members were recruited through email. In the recruitment email, staff members were instructed to call a study hotline to schedule and participate in an interview. Interviews ranged from 20 to 90 min. We asked agency workers and administrators to describe their program’s mission and goals, the application and recertification process, and barriers to program participation (see Appendix A). We specifically asked agency workers and administrators to describe the challenges Hispanic families face in accessing assistance programs and resources available for Hispanic families. Staff members were offered $30 as a cash incentive or an electronic gift card. Some workers and administrators declined the incentives based on county-level restrictions on receiving payment.

We entered all transcripts and interview memos into NVivo 12, a qualitative data analysis software package. The research team analyzed the data using an interpretive approach, which prioritizes participants’ own understanding of processes as described in their own words (Haverland and Yanow 2012). We relied on broader codes derived from previous literature on barriers to accessing assistance programs to initially organize the data but also allowed for emergent themes. For example, when agency workers or administrators reported complicated applications as a potential barrier to assistance, we coded this response as a “compliance cost” but coded “language barriers” as an emergent theme when staff commented on Hispanic families’ difficulties in understanding the application process due to language challenges.

The second, quantitative method consisted of a 20-min online survey we designed to elicit perspectives and identify practices of county administrators of agencies delivering TANF across Hispanic majority states. The survey captures six themes: organizational processes (language barriers, application processes, requests for documentation), training and staff development, economic climate and budgetary constraints, demographics, office space, and external political environment.Footnote 2 These sections were designed after an extensive literature review focusing on organizational culture in TANF offices (Gaffney et al. 2018), differences between county- and state-administered TANF programs (Hahn et al. 2015), resources offered by TANF agencies (Justice in Government Project 2020), and outreach and engagement with Hispanic families (Torres et al. 2016), as well as in consultation with leading experts on the TANF program. The survey also included specific questions about the impact of COVID-19 and allowed for open-ended responses on key barriers that Hispanic families face in accessing TANF, approaches that might make it easier for Hispanic families to access TANF, and comparison of accessibility between TANF and other programs like SNAP, WIC, and Medicaid.

In North Carolina, the email and link to the online survey was distributed through a “Dear County Director Letter” sent by the North Carolina’s Health and Human Services Department.Footnote 3 The survey ran from the end of June to the end of September 2020. A total of 45 administrators and 18 front-line workers who are employed in 46 counties replied, including representation from workers in the five counties included in the qualitative study. The 46 counties represented, out of 100 counties total, account for almost 70% of the entire North Carolina population and all regions of the state. In total, the counties surveyed account for almost 75% of the Hispanic population in North Carolina. For this study we report on analyzes from data collected from the 45 county administrators representing 38 counties, weighted at the county level to adjust for five counties that included responses from two administrators and one county that included responses from three administrators. County level administrators who responded vary somewhat in their functions, divisions and titles from county-level Economic Services and TANF directors, supervisors, and managers as self-reported in the survey.

Results

Findings are organized by key themes that emerged from the qualitative study with respect to the 49 administrators and workers’ perceptions as revealed through open ended interviews: 22 (44.8%) reported citizenship concerns, 14 (28.6%) reported transportation barriers, 7 (14%) reported language barriers, 11 (22.4%) reported learning costs, and 3 (6.1%) reported no barriers. Data from closed and open ended questions in the survey of 45 county-level administrators are incorporated as appropriate, with some key findings that correspond to these themes presented in Tables 1 and 2; companion illustrative examples are provided from open-ended responses in the survey.

Table 1 Application process requirements and office characteristics, as reported by North Carolina county administrators
Table 2 Staff retention, recruitment, and training, reported by North Carolina county administrators

Citizenship Concerns and Fear of Deportation

When asked to describe the greatest barriers to applying for assistance, most administrators and workers in the open-ended interviews (n = 22; 44.8%) commented on pervasive fears of deportation and reluctance to apply due to the Trump administration’s public charge rule. When asked whether Hispanic families faced barriers to applying for assistance, one urban WIC staff member explained these concerns in greater detail: Clients are reluctant to keep WIC appointments because they do not want to risk future opportunities for citizenship. She explained,

One, the reason is because it’s a lot of Spanish people that they don’t have papers. We have a community of Spanish-speaking people here and all the time that I ask them why you didn’t come back to the appointments, you miss appointments, “Oh no, I don’t want to go.” I’m like, “Why not? You are allowed to still receive your benefits.” And they always say like, “No, I’m scared they might get all my information. I don’t want to get in trouble or in the future when I get my papers or I’m in the process of papers. So I don’t want that this is going to interfere. So I prefer to stay away from this type of program.”

She added that Hispanic clients fear being taken by immigration enforcement agents outside of the office: “They think that maybe someone is going to be waiting for them outside. That’s what they told me, but that they say.... So, it’s more like scared to come to the program because they [are] going to get my information and so someone might come to my house and get me.” These citizenship and deportation concerns emerged in rural counties as well. This worker, a WIC processing clerk, described the fear of deportation as one of the greatest barriers to program participation for Hispanic families, especially when they are advised by legal advocates to avoid applying for assistance programs. She explained that “one barrier was the Hispanic population was afraid. They were afraid because they were told by people that represent them…to not apply for anything.”

These concerns were not limited to the WIC program; workers across public assistance programs viewed citizenship concerns as a barrier to program participation for Hispanic families. For example, an urban Work First caseworker regarded future eligibility for citizenship as a deterrent to program participation: “I think future eligibility for citizenship or legal standing is impacted by any benefits you receive and I think that would be the drawback or the thing that hinders a lot of people that are Hispanic in Work First.” An urban SNAP and Medicaid intake worker similarly regarded citizenship as a barrier to accessing benefits. While she viewed Hispanic families as having the same opportunity to access benefits as other families, she said they are “nervous” to apply for services due to the threat of deportation. She stated:

I think they have the same window of opportunity that everyone else has. I think sometimes they are a little more nervous to apply for the services. They don’t know that if they are undocumented or here illegally—I hate using that word—that we are not able to report that information, so a lot of times I think they just don’t request the assistance or the help because they are nervous or they’re scared. But they have the same opportunity to apply for services as anyone else.

Fear of deportation also featured extensively in the open-ended questions of the survey. Nine (evenly distributed across urban and rural counties) of the 45 administrators who responded explicitly mentioned deportation and citizenship as a key barrier of access, second only to the language barrier. For example, supervisors and managers across urban and rural counties note the “fear of ICE [U.S. Immigration and Customs Enforcement],” “fear of deportation and law enforcement being contacted,” and fears that “if they are in the process of applying for citizenship, they are afraid that accessing benefits will disqualify them.” When asked about approaches to make it easier for Hispanic families to apply, a rural program administrator responded: “Understanding that our agency is here to help. We are not ICE.”

Surveys of county administrators further suggest the ways in which the types of documents used to determine eligibility could potentially provoke concerns and fears, even among Hispanic citizen families. As shown in Table 1, over 80% of county administrators report that Social Security numbers of the applicant and all household members are collected, as well as birth certificates. Driver’s licenses are the second most common type of document collected, with just over half reporting reliance on bank statements and fewer using passports and vehicle registration. Table 1, however, also suggests little presence of ICE or law enforcement inside agency offices, and the majority of county administrators report that “Know Your Rights” legislation information is provided in the office through posters, signs, or verbal advice.

Transportation Barriers

In addition to citizenship concerns, many of the workers and administrators who were interviewed (n = 14; 28.6%) reported transportation barriers as impeding Hispanic families from participating in programs. Without reliable transportation, program participants could not visit the office to apply for programs or attend WIC or TANF appointments. For example, a WIC staff member attributed low enrollment numbers to a combination of clinic hours that did not fit clients’ work schedules and “car troubles.” WIC participants miss appointments because they cannot take time off from work and cannot get to clinics. She explained,

I hear a lot of people, transportation may be a barrier. Work schedules, they can’t get here and we try to accommodate that now because we went to night clinic to try to accommodate that. But at one point, I think it would just kind of be car problems, transportation issues and they just can’t get off of work, or the time is you know. Which I can understand, working 8:00 to 5:00 by the time you get off, everything is closed up.

Transportation barriers were most pronounced in rural communities that did not have reliable public transit; 12 of 14 workers who reported transportation as a barrier for families were from rural counties. One TANF worker noted the challenges of accessing services—agencies are simply located too far from many clients. Without a bus system, clients have to “find a ride,” the worker explained.

Transportation is a lack here, especially along this north side, where, to get to the DSS [Department of Social Services] office in this town, just from here, it’s probably about—around about 10 min. But if we have families from the [this area] or far as down to the [next town over], that could be as far as 20, 25 min. So if they don’t have transportation, trying to find a ride—and I have heard people to say that people would take them, but people charge lots for rides. Like, a lot. A lot.

While less prevalent, some urban workers reported transportation as a barrier to program participation. An urban TANF and Medicaid worker explained this barrier in greater detail. While they acknowledged that the community had a public transit system, they also shared that public transportation can be difficult to access.

I think a lot of people they can’t get here.... You know transportation is a problem. Yes, we have the transit, but I think they have certain criteria that they have to go through to get that and you have to let them know so much ahead of time and that kind of thing.

Survey findings of county administrators show that approximately 80% of administrators report availability of transportation services, with 88% of urban administrators reporting this service compared with 70% in rural counties (available upon request). Few county administrators mentioned transportation as a barrier in the open-ended responses, and transportation generally ranked low among barriers mentioned. Some exceptions arose. One urban county administrator noted that their agency was unable to provide transportation (and daycare and other services to help applicants become employed) to undocumented Hispanic applicants.

Transportation hurdles can be alleviated or compounded by the location of services (Gaffney et al. 2018; Kirby and Lyskawa 2015). Only 16% of the county administrators responded that serving Hispanic clients factored into the design or location of the TANF offices in their jurisdiction. However, administrators also reported high levels of co-located services, primarily of other welfare agencies like SNAP and WIC. Other relatively high levels of co-located services included child care (67%), substance abuse support (46%), and refugee resettlement services (17%).

Availability and Access: Learning Costs

Many of the workers and administrators interviewed reported concerns that Hispanic families lacked awareness or accurate information about how to access and use assistance programs. One urban frontline worker commented: “I think it may be just a learning barrier within the community to let them know that, look, this is available to you.” Others tied limited information to poor promotion of the program. A rural frontline worker shared this view when describing barriers to the WIC program in the past. She commented that increased promotion had boosted awareness and program participation:

But I think it’s just that—at one time it might have been some are not being promoted, the program was not being promoted like it should maybe, not enough putting out there. But I must say within this last year or two they have really done good. They have really put their stuff, like those signs you see out there say WIC, they put all that out there and I don’t know if it’s... and it just helped boost their clientele up.

Reports from county-level administrators in the survey offer a slightly different picture of overall accessibility. The majority (approximately 80%) of county administrators reported that they perceived the TANF system and program as easy to navigate for English and non-English eligible applicants and that the TANF program ensures that Hispanic families are aware of its services. However, consistent with the reports from workers in the qualitative study, only 18% of county administrators report that targeted outreach through social media is used often or fairly often. Indeed, our independent analysis of tweets from North Carolina agencies confirms very low activity, with a recent burst in 2020. About one in three county administrators reported participation in networks and targeted Hispanic events, and administrators generally reported low use of culturally specific strategies. When asked about the most common ways Hispanic applicants found out about TANF, 60% of county administrators stated “visiting an office (or through a call center)” and/or “referred by other programs (SNAP, Medicaid, WIC, CPS, FNS).” “Word of mouth” was by far the county administrators’ most commonly stated way that they perceive that Hispanic families find out about social programs.

Language Accessibility and Barriers

While language accessibility was a less prominent theme, seven of the workers and administrators interviewed (14.2%) also emphasized the language barriers Hispanic families experience when applying for benefits. Although most workers acknowledged that application materials were available in Spanish, many noted that language barriers between workers and applicants would often confuse clients about what is required for the application process. In this sense, language barriers compounded the learning costs of accessing programs. As an urban Medicaid worker noted, Hispanic applicants may not comprehend complicated application and recertification forms even when they are written in Spanish:

I would think that with the Hispanic families it could be the same thing, language, not fully understanding the program and the qualifications. Even when it comes down to certifying, recertifying again. There could be some barriers there. That’s the only thing I could think of with the Hispanic population.... It’s not fully understanding what is entailed after they apply. So like if they already been approved and then have to recertify, I don’t think they fully understand that. You know they think okay when I apply that was it, you know and I continue to get it, I don’t have to report any changes, why do I have to do that?

A bilingual rural WIC worker similarly commented on the confusion Hispanic clients experience during the application process because of the language barrier. She described an instance where a client tried to apply for assistance and assumed that she was presumptively eligible for WIC because she received Medicaid. When the worker tried to confirm her Medicaid receipt, she realized the applicant wasn’t receiving Medicaid because she had failed to submit income verification paperwork to recertify. She explained, 

So today I did have one client that I could not serve. She had no income. I checked NCTracks, which is a way that we can look up Medicaid. I tried looking up her information. It came up that she did not have Medicaid.... And the client said that she had applied for the baby. So I’m looking up mama’s information, I’m looking up the baby’s information... and I’m like, when did you apply?

When she could not find the applicant’s Medicaid verification, the WIC worker called the Department of Social Services (DSS) and acted as a translator for the applicant. She learned that DSS mistakenly sent out a recertification letter in English. Because the applicant did not understand the letter, she did not know that needed to send additional income verification to maintain her Medicaid benefits. She explained,

So I’m looking up and I’m like, well, let me call DSS and let’s see if there’s a reason why the application’s not showing or they can—maybe there’s a reason why. So I call the caseworker and basically she informed me that the hospital sends in information and lets them know that the client has delivered. But they sent the client a letter saying that she needed to send an income verification for the month of November... and if she didn’t do it, her application was dead. Pretty much it was expired and she would have to apply again. And because it was a language barrier, they mailed out the letter in English and the client does not read English.

The worker rescheduled the applicant’s WIC appointment for a later date when the applicant was able to verify her household income. She added that this kind of confusion is common among Hispanic clients. Hispanic mothers who have just given birth assume they are signed up for Medicaid through the hospital, but they must take additional steps to continue to receive benefits.

It happens a lot with our Hispanic clientele. And I was telling the DSS worker that I’m thinking, because the hospital is filling out general information and sending it to DSS, our Hispanic clientele are thinking, oh, they’ve already applied for the Medicaid. The baby’s already approved for Medicaid. So I’m thinking that’s a little bit of a language barrier that our Hispanic community has and stuff. So I explained to her, please try to see if you can get that information that they’re requesting.

Her client was unable to receive WIC benefits that day because she did not have the right information to determine income eligibility. Had she received a recertification letter in Spanish, she might have been presumptively eligible for WIC.

While front-line workers emphasized the challenges Hispanic families face, many agency staff also noted the resources agencies provided to ease access to programs. No worker or administrator interviewed mentioned any training or professional development that equipped workers to address the unique needs of Hispanic families. Instead, many workers (22; 44.8%) emphasized the presence of bilingual workers or interpreters within agencies as eliminating barriers to assistance for Hispanic families. In short, they regarded the presence of bilingual workers and interpreters as giving Hispanic families the same level of access to programs as other families. For example, one rural WIC staff member regarded the WIC programs as accessible to Hispanic families because the agency has “two bilingual personnel.” For her, “language is not a barrier.” Likewise, an urban WIC staff member bureaucrat described how interpreters in her clinic assist Hispanic applicants in accessing WIC and other public assistance programs. When asked to describe the barriers Spanish speaking families face, she commented:

We have communication with the interpreter at social services. So if our patients are running into an issue with Medicaid or anything we usually call the interpreter there, for the Hispanics now, and say, “I have such and such here what do they need to do?” And usually she’ll say, “Tell them to stop by our office,” because it’s only a couple of blocks.... if you tell us that you’re a US citizen and they were like, “Well, have you applied for Medicaid?” If they say, “No. Because I don’t know where to go,” then we try to guide them, say, “Okay. Social services is down the road. You can go any day.” Because we already know their work hours and then we know the interpreters there, so we communicate with them. Sometimes I will call and say, “Such and such will stop at your building. They’re new to the county. Would you please guide them to what services you offer?”

In North Carolina, the county health departments and departments of social services are required to offer application materials in Spanish and to provide interpreters to clients. Accordingly, many workers (14; 28.6%) commented on the presence of Spanish application materials. An urban DSS caseworker explained that these resources give Hispanic families “the same opportunities” as English-speaking clients to access benefits.

I think it—because we do have—because they can apply online and we do have information that is translated into Spanish. When they come into the agency, they are—we’re required to offer interpreters for them so that it’s not a challenge for them, they get the same opportunities. So I think it’s pretty much the same as far as being able to contact us and apply and to start the process.

Interviewer: Do you have Spanish-speaking families on your caseload?

Respondent: I do. Yeah. And we have interpreters that are on the floor.... If we can’t get ahold of someone, then we do have where we can call other interpreters and they can help us and it can continue to flow so there’s not a delay or they have to wait for someone, they can still have the same service.

This urban DSS administrator described access to programs as “very easy for [Hispanic] clients.” As part of the agency’s efforts to comply with federal antidiscrimination laws, the agency developed strategies to ensure that individuals who did not speak English as a first language could access programs. He described the agency’s efforts to hire bilingual caseworkers to improve customer service for Spanish-speaking clients:

We have a number of bilingual people on board. We’ve gone from needing interpreters and translators to making sure that we have bilingual people in the actual roles. And there’s a significant difference in the type of customer service you receive when you’re speaking with a bilingual caseworker versus speaking to a caseworker and an interpreter. And so, we have that pretty much throughout all of our programs for people who speak Spanish. So, I think that it’s very easy.

Give these efforts, this administrator viewed programs as accessible to Hispanic families. The fear of deportation and “sociopolitical barriers” rather than the language barriers discouraged Hispanic families from applying for benefits. He stated:

I think some of the federal laws, some of the border patrol and ICE enforcement, has added a different level of challenges. But just in terms of general accessibility, particularly because of language, I think those barriers are removed. It’s more sociopolitical barriers that they’re facing now.

A few workers did not share this view; they highlighted the pitfalls of relying on interpreters, bilingual workers, and Spanish application materials. Some workers remarked that there were not enough interpreters to serve the needs of the agency. Further, they stated that the Hispanic families need more one-on-one assistance to fill out applications, even in Spanish. Applications from Hispanic applicants are often incomplete or incorrect. Additional bilingual workers would improve customer service for Hispanic families and ease the application and recertification process.

The vast majority of county administrators in the survey sample agreed that TANF program materials in Spanish translation were readily available in person, online, and during the intake process, with the majority of administrators reporting that staff also have access to interpreters. Just over half of county administrators reported that the availability of materials in Spanish increased in the previous 5 years. However, although access to interpreters was reported as high (across online and on-site formats), administrators report low-to-seldom access to cultural ambassadors or consultants.

Characteristics of Staff and Professional Development

Information about staff characteristics and access to training and professional development was collected as part of the survey; data are presented in Table 2. County administrators report retention rates of approximately 60% for front-line worker and 67% for administrator staff positions, though less than 10% of the positions are filled by Hispanic people or people proficient in Spanish. Even though many county administrators, especially those from urban counties, report targeted recruitment efforts to fill staff positions with Hispanic (69%) or people proficient in Spanish (83%), this is less prevalent for filling senior positions. While just over half (58%) of county administrators report that the agency provides training to support Hispanic families, only 8% report often or frequent review of case practices relevant to Hispanic families and 34% report speaking about cross-cultural barriers. While the front-line staff interviewed in the qualitative study seldom mentioned receiving training to meet the specific needs of Hispanic families, over 95% of county administrators who responded to the survey report that staff receive general training and training specific to documentation at least once a year and 80% or more of administrators report training on immigrant eligibility, prejudice and bias, and cultural competence training at least once a year. Approximately 75% of county administrators report staff receive training specific to serving illiterate applicants, whereas 42% of county administrators report staff receive training specific to serving Hispanic people.

Discussion

Considerations of administrative burden are typically argued for, and documented from, the perspective of individuals and families who are eligible for safety-net and related programs and services. The views of administrators and practitioners is understudied, particularly those who are increasingly charged with serving racial and ethnically diverse populations; yet, these workers have considerable leeway in shaping the experience of eligible recipients and reducing or exacerbating administrative burden. We take a unique approach to filling this gap. We embarked on a coordinated collection of qualitative and quantitative data from administrators and front-line workers to capture their perceptions of implementation and services and the challenges and strategies in serving Hispanic families. This is the first endeavor that we know of to do this in North Carolina, a state with a history of racial division. Our findings uncovered aspects of four domains based on worker reports that shape Hispanic family experiences: fear related to citizenship status, transportation barriers, learning costs, and language accessibility and barriers.

Workers and administrators across both data sources emphasized concerns about citizenship status and the fear of deportation as discouraging program use among Hispanic families, even though county administrators report that offices do not have excessive security measures or access points for ICE; further, information on Know Your Rights legislation is reported as readily available. However, nearly half of the county administrators also reported some presence of private security guards in the office, and approximately one in five reported the presence of metal detectors, with much higher reports of metal detectors (29%) from administrators in rural as compared with administrators from urban counties (11%). The presence of security guards and metal detectors might set a hostile rather than an empowering tone, particularly in the context of fears of deportation. Workers and administrators from both the survey and interviews emphasized the broader psychological costs Hispanic families experience stemming from anti-immigrant rhetoric from national political figures and the public charge rule. Through their own voices and indicators of anti-immigration threat, we show the ways that chilling effects have spread to affect staff responsible for the delivery of services, who find it challenging to offset these perceptions with the resources available to them. Regardless of the agency resources available, these findings may suggest how broader sociopolitical concerns create barriers to participation in safety-net programs. When asked about perceptions of political support, county administrators in the survey data reported political support for Hispanic families across social media and local and federal platforms, but very few reported comparable levels of support for immigrant families.

Administrators and workers from the qualitative and survey data named a number of examples of learning costs imposed on Hispanic families. While some workers interviewed found language barriers to be adequately addressed through the presence of bilingual workers and interpreters, and Spanish-language translation appears prevalent according to county administrators, workers and administrators in both groups of respondents reported ways in which Hispanic families still face challenges in understanding how to navigate complex systems processes to access assistance. While access to language interpreters (in person and online) is high, use of cultural ambassadors, for example, who may serve this supportive role and improve efficiency for workers, is very low. County administrators may not be attuned to the ways that workers face challenges beyond language translation. Specialized training for workers and additional agency resources that can be offered through cultural consultants and ambassadors will increase efficiency and support access for Hispanic families in person or through remote or digital platforms. When asked about the most common way that Hispanic families learn about TANF and perhaps safety-net programs more generally, county administrators and workers agreed that word of mouth is the most typical mechanism of information dissemination. This finding suggests that formal information dissemination through local departments of social services and county health departments may be ineffective in raising awareness about assistance programs for Hispanic families. Instead, agencies may need to pursue creative outlets, such as seeking trusted community ambassadors and employing more grassroots strategies to support and shape peer-to-peer information exchange. More formal avenues to foster cross-state communication may also help: county administrators reported that in 2019 there was very little correspondence with other states about strategies for outreach to Hispanic families or addressing barriers that might affect them (less than 10% indicated correspondence with other states on these topics, not shown).

Our mixed methods approach allows the qualitative evidence to shore up and enrich the quantitative findings, revealing new themes beyond those that emerged in the survey. Transportation emerged in the qualitative interviews as a key barrier for Hispanic families seeking public assistance. This was not a targeted area of inquiry in the survey, although, consistent with the qualitative data, some county administrators described transportation as a barrier in the open-ended responses. Federal and state-level policy makers should consider transportation burden as impeding program access when in-person engagement is necessary for eligibility determination and receipt or recertification. Key shifts in social policy delivery following COVID-19 provide an opportunity to assess how conversion to online and remote interviewing and intake processing might improve access. But even with remote opportunities that might reduce transportation barriers, the challenge of ensuring that Hispanic families understand the complicated application process means that compliance costs beyond language barriers remain a problem.

Investing in cultural competence and related training opportunities might have large returns that can also improve cost efficiency of service delivery. While the majority of county administrators report the availability and receipt of general and specific training (e.g., related to prejudice and bias and eligibility determination for immigrants), review of case practices in serving Hispanic families and provision of training specific to support Hispanic families are less common. Thus, our findings point to the potential benefit of processes for offering supportive feedback and review.

Surprisingly, our quantitative and qualitative evidence provides little indication of racial or ethnic bias in how workers treated Hispanic clients—a departure from the social policy literature that highlights the racialized implications of decentralized social policy delivery. Resources for Hispanic families were comparable across counties and rural and urban contexts. For example, we found only slight differences in the recruitment and retention of Hispanic staff. However, whereas a majority of county administrators report moderate to high efforts to recruit Hispanic and Spanish-speaking individuals for staff positions, a minority of administrators report equivalent efforts to fill administrator positions.

Limitations

The strengths of the mixed methods approach are balanced with limitations on external validity and the ability to derive conclusions that can be generalized across the state. Our sample sizes of 45 county administrators representing the majority of counties with high populations of Hispanic families and 49 worker and administrator respondents are large for endeavors seeking to gather practitioner perspectives (whether by survey or qualitative interviews) but may not represent the full range of practices and perceptions. Second, to preserve confidentiality, our analyzes exclude characteristics of the administrators and front-line workers such as their education, work experience, race, and ethnicity and thus we do not have good information on the administrators who did not respond and why they might not have responded. Third, we examine one state in the Southeast rather than other states in the region. North Carolina has a unique decentralized system as one of the only county-administered systems for TANF, for example, that is not shared by their southeastern state peers (Gainsborough 2003). Nevertheless, our descriptive study offers insights on an important yet understudied area of research, namely barriers to assistance programs for Hispanic families in a system marred by racial exclusion. Our aim was to reveal these barriers from the vantage point of administrators and front-line workers. To that end, the descriptive patterns offered by the survey results and thick description of the qualitative findings are appropriate and point to the need for further research on this aspect of social service delivery.

Implications

Examining the views and practices of administrators and workers is an important piece of understanding families’ administrative burden and access to safety-net programs; this research is intended to complement and accompany the perspectives of families. Because so little research has been conducted from this perspective, mixed methods approaches are essential gateways to uncovering insights and areas that merit further investigation and exploration, particularly related to areas that are not easily conducive to predefined metrics or survey questions, such as training needs or fear and chilling effects. We uncovered many dimensions that are not feasible to capture for example, in companion efforts documenting policy and practice from publicly accessible information (Gennetian et al. 2020). Future quantitative research can examine the relationship between the presence of Hispanic staff members and translators and program uptake, while qualitative research can further probe the kinds of training and professional development needed for all staff to better serve Hispanic families. The discrepancy between administrators’ reports of specialized training and front-line workers’ reported receipt of that training points to gaps in what is prescribed and what is actually happens on the ground, given workers’ time constraints and mounting client demands. Future qualitative research can investigate whether and how training occurs in agencies and how this training translates into practice.

Studies of street-level bureaucracy and administrative burden often point to the role of legislators at the federal and state level in shaping the hurdles families encounter in their attempts to access benefits. Yet our survey and interviews highlight the need to understand Hispanic immigrants as a group that is socially constructed in a negative light both in discourse and in public policy. Workers viewed national anti-immigrant discourse and policy as constraining their capacity to enroll and retain Hispanic families. This negative discourse, rather than the program rules and guidelines, hindered their efforts to serve Hispanic families. This kind of exclusion from assistance programs will undoubtedly have generational consequences on the health and development of Hispanic children. Future research can examine these implications.