Abstract
Objectives
Quantitative studies of institutional (non)engagement are limited by available measures and temporal ordering concerns. Using unique and detailed qualitative data, we examine (non)engagement behaviors and perceptions among formerly incarcerated adults.
Methods
We conducted around 100 face-to-face interviews to elicit rich information about an array of behaviors and perceptions for three key formal institutions: financial, medical, and employment.
Results
Almost everyone in our sample was at least partially engaged post-release and experienced bounded engagement as they navigated the option set available to them. When people were not engaged, common reasons included insufficient funds to meet bank requirements or institutional or logistical barriers to access, which align with an administrative burden perspective. Concerns about wage garnishment, a system avoidance explanation, were also raised; however, avoidance was rare and contained to the relevant institution(s). We also observed institutional interdependency, where barriers encountered with one institution could prevent engagement in another.
Conclusions
Reducing or removing barriers to healthcare insurance and access, minimum balance requirements, and burdens placed on job applicants with criminal records can improve institutional engagement. Alleviating burdens in one domain (e.g., formal identification or schedule restrictions) may also improve access in other institutions (such as formal employment or bank accounts).
Criminal records are no longer rare negative credentials. Among men born in the early 1980s—now in their “prime-age” for employment (US BLS 2020)—close to half of active job seekers and a quarter of those employed had an adult misdemeanor or felony conviction (Bushway et al. 2022). Researchers argue that successfully integrating into society post-conviction involves access to formal social, economic, and political institutions (Travis, Solomon, and Waul 2001), aided by social support as people “settle into communities permanently” (James 2014, p. 1). Yet formerly incarcerated individuals have disproportionately lower engagement levels with formal 1 institutions relative to their non-incarcerated peers, including levels of opening and owning savings or checking accounts (Bryan and Collins 2024), medical insurance or stable coverage (Zhao et al. 2023), and formal employment (Holzer 2007; Looney and Turner 2018; Western 2002).
Various reasons have been proposed for system nonengagement among individuals with criminal records. One straightforward reason is that some people do not want (or do not believe they need) certain services. They could be considered voluntarily opting out, and exploring the reasoning underlying these decisions is important. When people determine there is a need and seek formal institutions, they may encounter laws and policies that restrict eligibility (e.g., licensing barriers). We can consider those individuals formally excluded.
However, there may also be more coercive forms of opting out when people determine they need services, are not officially excluded, and still experience nonengagement. People attempting to engage with governmental agencies routinely encounter administrative burdens, or bureaucratic processes and complicated regulations that lead to frustration and limited access to services (Herd and Moynihan 2018; Pahlka 2023). Barriers to information access and stigma are also considered administrative burdens (Herd et al. 2023). In these cases, people are unable to engage not because they are ineligible or formally prohibited, but because they are dissuaded through the process. While not commonly studied in criminology, Bryan's (2023) recent test of administrative burdens using public benefit programs finds evidence that this framework also applies to individuals with criminal histories.
A more commonly discussed possibility in criminology is system avoidance, where people with prior criminal justice system involvement strategically avoid formal institutions that can easily link their personal information across agency databases (Brayne 2014). There is existing support for the system avoidance perspective (e.g., Brayne 2014; Goffman 2009; see also Haney 2018), but researchers also point to limitations with existing national survey data, including unclear temporal ordering and restricted measures (e.g., Brayne 2022; Remster and Kramer 2018). Whether facing formal exclusion, administrative burdens, or system avoidance, people may replace formal institutions with informal alternatives, such as seeking payday lenders, medical care from an unlicensed friend, or under-the-table employment. However, nonengagement reasons also offer different theoretical and policy implications.
The goal of the current study is to examine how formerly incarcerated adults navigate institutional nonengagement decisions across three major formal institutions—financial, medical, and employment—and investigate the complexities of institutional engagement. By examining transcripts derived from in-depth interviews with formerly incarcerated individuals, we were able to systematically investigate details and context surrounding such decisions, which can “illuminate the mechanisms” between criminal justice system exposure and nonengagement decisions (Remster and Kramer 2018, p. 14). Our study participants reported widespread interest and engagement, with almost everyone engaging with at least one formal institution post-release. When people were not engaged, commonly reported reasons involved institutional or logistical barriers to access, which align with formal exclusion and administrative burdens perspectives. System avoidance explanations (such as wage garnishment) were rare and appeared relevant to only the impacted formal institution(s).
We describe this concept of engaging with available institutions when feasible as bounded engagement. In line with prior observations that marginalized populations appear to circumvent problematic formal institutions when barriers are present, rather than all formal institutions or the problematic institution all the time (Ewald 2024; Fong 2019), we found that participants navigated the option set available to them. We define bounded engagement as generally engaging with formal institutions when there is a need and ability to, but opting out or finding informal replacements—typically as temporary solutions—when accessibility to formal institutions is restricted (or structurally bounded). This concept can apply to a single institution or barriers in one formal institution can prevent engagement in another; we use the term institutional interdependency to describe the latter. Given the reintegrative importance of formal institutional engagement and strong participant interest in these institutions, we conclude the paper with a set of specific barrier-reduction strategies to improve accessibility.
Nonengagement Reasons for Individuals with Criminal Records
Institutions can be defined and categorized by a range of criteria, such as how much they encompass a person's time and interest (Goffman 1961), the level of surveillance or formal record keeping involved (Brayne 2014), or whether they are formal or informal (Garland 2001). While there are numerous interesting institutions we could explore, to be able to speak to the system avoidance perspective, if applicable in our data, we selected the three key institutions in Brayne (2014). 2 We describe long-standing and well-established financial, medical, and labor market institutions as “formal,” and alternatives (such as working “off-the-books”) as “informal” to align with the terminology in labor market research for disadvantaged populations (e.g., Edin, Lein, and Nelson 2002) and to incorporate the role of record keeping (Brayne 2014).
Formal Exclusion
A commonly cited reason for a lack of engagement involves formal exclusion: formerly incarcerated individuals are ineligible for and excluded from certain economic and social opportunities (USCCR 2019). These prohibitions vary dramatically across U.S. states; depending on where a person lives and their criminal record, they may encounter legal restrictions pertaining to employment, holding public office, occupational licensure, voting, jury service, securing public housing, receiving public assistance, eligibility for financial aid, owning a firearm, and obtaining a driver's license (USCCR 2019). Structural exclusion and associated alienation can have profound impacts on how people navigate their social worlds (Bell 2017), and researchers have extensively documented how blocked access can lead people to seek alternative options, including illegal and extra-legal opportunities. As Trimbur (2009, p. 275, emphasis in original) described, people in this predicament may “take reentry upon themselves” and “draw upon resources outside of the system.”
Opting out
Another non-engagement reason involves voluntarily opting out of involvement with formal institutions. For example, employment exclusion policies push people out of the formal labor market, but a sizable proportion of those incarcerated were already nonengaged prior to prison (Sykes et al. 2022). Furthermore, individuals with criminal records, and formerly incarcerated people in particular, face a host of reintegration challenges, including declining mental and physical health, housing and employment instability, and substance abuse (Lattimore, Steffey, and Visher 2010; Mears and Barnes 2010; Petersilia 2003; Travis 2005). As Bryan and Collins (2024) described, the factors correlated with being unbanked—including low socioeconomic status and limited education and employment histories—are similar to the challenges formerly incarcerated people encounter. Some people may feel certain services are not needed, such as not needing a bank account if they do not have income, and others may be actively involved in crime and not seeking services. Reengaging in formal institutions may seem like distant goals for those struggling to reenter society, leading people to opt out.
Reasons rooted in the concept of administrative burdens describe a more coerced form of opting out, where bureaucratic demands and complex rules bring about frustration and withdrawal. Administrative burdens include psychological, compliance, and learning costs—including the time and energy required to gather information about a program or service, eligibility, and access (Herd and Moynihan 2018; Herd et al. 2023). Researchers have found evidence that the psychological costs from criminal record stigma can lead people to retract from public life (e.g., Lageson 2016; Soss 2005; Weaver and Lerman 2010). Even when institutions do not pose a financial cost to engage in services, the costs to learn about and comply with eligibility and application processes can be daunting. For example, respondents in a southern U.S. state reported encountering frustrating application requirements when trying to establish a bank account, such as needing state-issued identification that they did not have (Mielitz et al. 2019). Halushka's (2020, p. 247) ethnographic and interview-based research similarly portrayed a group of formerly incarcerated people, often in desperate need of social safety net programs, attempting to seek services but finding the process to be “physically and emotionally stressful.” His study participants spent extraordinary amounts of time simply waiting and used nontrivial portions of their limited incomes traveling back and forth to appointments. Herd and Moynihan (2018, p. 8) argued that such burdens are the result of “administrative and political choices” (see also Lipsky 1984), and they can apply to services in the public or private sector (Niskanen Center 2023). 3 Administrative burdens have only become amplified with advances in technology and online services; as Pahlka (2023) illustrated, the disconnect between “elite” policy crafters and those responsible for implementation requires cultural change.
The system avoidance perspective implies a different form of coercive opting out. It posits that people understand that they could access formal institutions but purposefully avoid engaging with those institutions because they fear that doing so will result in data sharing and surveillance. Active or strategic system avoidance is a reaction 4 to a concern that interacting with one formal entity will trigger surveillance across others, including public benefits or the criminal justice system (Brayne 2014). Brayne's (2014) influential study incorporated several key institutions—financial, education/employment, and medical—in a test of system avoidance. Based primarily on the National Longitudinal Study of Adolescent to Adult Health (Add Health) data, she found that individuals with prior criminal justice system involvement are less likely to interact with formal institutions than their counterparts without such involvement. She did not find evidence of reduced involvement when examining non-surveilling institutions (religious activities or volunteering), which lends plausible support for the system avoidance perspective. System avoidance is also intuitive; research finds connections between criminal justice system interactions and reduced trust, particularly among Black residents (Bell 2017; Brunson 2007) and members of hyper-surveilled communities (Fader 2021; Giuffre and Huebner 2023; Goffman 2009; Stuart 2016).
However, data limitations make the underlying mechanism(s) undetectable, and differentiating system avoidance from other reasons is challenging. The practice of avoiding surveilling institutions could also be limited to certain situations, such as concerns about wage garnishment. For example, O’Brien (2008) found that even when people understand the value of saving and feel comfortable using retail banks, they may avoid these institutions for fear they will become ineligible for public assistance services by virtue of becoming more visible to state agencies responsible for collecting court-mandated child support payments. In Holzer, Offner, and Sorensen’s (2005) study, young Black men with low levels of education avoided the labor market in response to child support laws that garnished wages, as did a “small yet vociferous” subset of Haney's (2018, p. 34) formerly incarcerated fathers attempting to evade child support payment.
Alec Ewald (2024, p. 2) similarly suggested cynical perceptions may be “domain-specific” and bounded. He describes his interviewees, a group of adults receiving reentry services in New York City, as having “intensely negative, distrustful ideas about certain actors and policies [that] were juxtaposed alongside a strong desire to participate in public life” (p. 2). In her study of low-income mothers, Fong (2019, p. 1786) pointed to what she calls selective or constrained visibility. She describes how child protective services concerns “rarely prompted mothers to avoid systems wholesale, but within their participation in healthcare, educational, and social service systems, they engaged in a selective visibility, presenting themselves but not their full selves to authorities.” People may also react in ways that are counterintuitive from a system avoidance perspective. The Black and Latino “hypercriminalized” participants in Rios’ (2011) Oakland, CA study, for example, developed an increased interest in activism and political engagement even though these activities could lead to heightened surveillance (see also Walker 2020). While Bryan (2023) focused on public safety net programs and did not test engagement in medical, financial, or employment institutions, her study also casts doubt about system avoidance as a primary mechanism for nonengagement.
In sum, there are contrasting potential reasons for institutional nonengagement. The key contribution of the current study is the ability to provide a detailed and nuanced examination of why formerly incarcerated individuals are institutionally (non)engaged with a set of common formal institutions, which enables us to consider the full set of potential reasons and craft relevant policy recommendations. In addition, our geographic context presents a unique opportunity to examine these different perspectives. Formal medical institutions routinely collect and manage sensitive individual-level data, which could discourage engagement (per system avoidance), but in Massachusetts, administrative burdens related to insurance have been shifted from individuals to the state, a practice that has been found to increase engagement (Herd et al. 2013). If formerly incarcerated people avoid surveilling institutions, increased accessibility should not theoretically drive engagement decisions, but if individual-level administrative burdens are a key barrier, access should matter. Taken together, we aim to provide new insights into how, when, and why formerly incarcerated individuals engage with formal institutions, and how policymakers and service providers may be able to strengthen engagement for those seeking services.
Current Study
We were interested in examining whether and why nonengagement occurs among returning citizens 5 in the community more broadly (i.e., a sample of the formerly incarcerated population), rather than higher-risk populations recently released from prison (e.g., following a cohort of people exiting prison) or only those living in hyper-surveilled neighborhood contexts. As a result, we used a population-based sampling strategy rather than cohort-based recruitment (Bushway and Denver 2025; Kalra et al. 2022; Rhodes et al. 2016) with a combination of purposive and snowball sampling techniques (see, e.g., Ewald 2024; Fader 2021; Lerman and Weaver 2014). First, we developed a collaborative agreement with the Office of Returning Citizens (ORC) in Boston, which is housed within the Mayor's Office. The ORC provides individualized case management services to returning citizens in the Boston area, including assistance with housing, employment, and education. Staff members also engage in advocacy work to identify resource gaps in distressed neighborhoods, and at the time of data collection, partnered with over 200 organizations throughout New England to assist people with reentry and reintegration. The ORC had 575 clients when our study began in the summer of 2021 and continued to enroll new clients on a weekly basis.
While participation in the ORC is voluntary, it is also a city-funded program, and clients need to take initiative to engage with this resource. As a result, this approach could produce a sample that is likely to have higher levels of engagement with formal institutions than the broader population of formerly incarcerated adults in Massachusetts. Therefore, we also recruited through less formal methods. Specifically, we asked participants to identify other people who were released from prison but not connected with the ORC or similar city services. We developed and disseminated a one-page flyer to recruit study participants, appeared on a local radio station broadcast, and conducted extensive outreach with community-based organizations directly and by attending local reentry events. Interested individuals could inquire about eligibility criteria and schedule interviews in a few ways: call the number listed on the flyer, which connected to a shared OpenPhone account; contact the research team through a shared email account; or scan a QR code on the recruitment flyer to complete a short, online scheduling form. We interviewed 42 people through the ORC and 55 using other methods.
To be eligible for the study, participants needed to be formerly incarcerated adults (age 18 and older) who resided in Massachusetts. We defined “incarcerated” as spending at least one consecutive year in a state prison or a House of Correction (HOC) as an adult. 6 We also aimed to recruit a large proportion of people who were out of prison or an HOC for at least six months. This was due to our interest in institutional engagement; while we wanted to include people who were recently released, we also wanted to ensure that most people in our sample had time to make various choices we asked about in the interview, rather than capturing the often chaotic first few weeks upon release. Current probation/parole status did not impact eligibility, and no one was excluded based on offense type (including sex offenses). We conducted interviews in English or Spanish, depending on the participant's preference; ultimately, five were in Spanish.
The research team developed semi-structured interview guides in the Summer of 2021 based on an interdisciplinary review of prior survey and qualitative instruments, and interviews were conducted between July 2021 and October 2022. Our IRB approved returning to in-person data collection activities by the summer of 2021, and all interviews were conducted in person and in accordance with university-approved COVID safety protocols. Interviews typically ranged from 1.5–3 hours, depending on the level of detail participants provided and the length of the accounts and examples they wanted to share. The interview team was diverse in terms of race, gender, and career positions. The interview guide contained prompts and probes throughout to help remind the interviewers to ask for clarification or elicit additional details.
We conducted all interviews in a private space on our campus, which is located between two subway station lines and is connected to the commuter rail. We compensated participants for transportation costs by providing either a public transit day pass entitling riders to unlimited subway/bus rides for a 24-hour period or by paying for campus parking. We also gave each participant a $50 Visa gift card at the end of the interview. All interviews were audio recorded, and the files were transcribed verbatim.
Sample Characteristics
Table 1 displays sample descriptives by three groups: the full sample, the “ORC recruitment” subsample, and the “other recruitment methods” subsample.
Descriptive Statistics.
Note. Categories may not sum to 100% due to rounding. Time since release is missing seven observations.
Abbreviations. For the formal engagement categories, E, F, and M refer to employment, financial, and medical, respectively.
We asked participants to self-report their gender, age, and race/ethnicity at the start of the interview. We interviewed 89 men and 8 women, including two people who identified as transgender, and the median age was 45 years old (range: 22–69). Around a third of our sample self-identified as Black, 32% as White, and 23% as Hispanic. Our sample had higher levels of education than anticipated; while about 21% had less than a high school diploma and 41% had a high school degree or General Educational Development (GED) (often acquired during incarceration), 4% had an associate degree, and over a third had some college or higher. Almost 38% of adults in the broader U.S. population had a bachelor's degree in 2022, making this sample proportion strikingly similar (U.S. Census Bureau 2023). In terms of criminal histories, our respondents self-reported a wide range of time since release from last incarceration (less than six months to 27 years), with an average of around three years. For the crime connected to their most recent eligible incarceration stay, over half of the sample reported a violent conviction crime type, followed by drug (25%), other (18%), and property (7%). Criminal histories were also extensive, with around 42% of the sample reporting five or more total convictions.
Basic demographics (male, age) and highest education level are comparable across recruitment methods, although there was a higher proportion of Black participants in the ORC recruitment pool and more White participants recruited in other ways. The ORC group also has a notably higher proportion of violent (67% vs. 38%) and fewer drug (12% vs. 35%) convictions for the most serious recent incarceration offense, and a lower average time since release (2.0 vs. 3.6 years). Over half of both groups were under supervision at the time of the interview.
Data Analysis
First, we developed codes to determine whether people were, or were not, engaged in each of our three formal institutions. This led to a summary profile for each study participant (e.g., Fader 2021, 2023) that provided an overview of engagement with each institution prior to and after the most recent incarceration. The researchers then filled in an entry for each person to summarize whether the person was engaged with all, some, or none of the formal institutions.
We made a set of decision rules for what “counted” as formal employment and medical engagement based on prior studies (e.g., Brayne 2014) and other sources. For employment, active involvement in part-time or full-time paystub or taxable income employment is counted as formal. Alternatives include illegal work (which was sometimes reported but something we did not directly inquire about) or informal work, such as working “under-the-table” or “off the books.” Informal work refers to jobs or side gigs that people are paid in cash to do, but that do not involve illegal activities (Nguyen, Loughran, and Topalli 2023). It is often considered precarious work because it is “uncertain, unstable, and insecure” and “employees bear the risks” (Kalleberg 2018, p. 3, emphases in original). Formal medical engagement included having a primary care physician and engaging in preventative care; accessing medical, dental, or mental health care as needed; and/or seeking care in community-based clinics (all of which are typically connected to insurance). Receiving medical care only through emergency rooms (ERs) or informally (e.g., a friend prescribing or sharing medication) did not count. 7
The most challenging decision rules involved financial engagement; given the rapidly emerging use of less formal banking methods, we had to pause to conduct research on neobanks to determine what “counted” as formal engagement. Following the Federal Deposit Insurance Corporation's definition of being “banked” (FDIC 2021), our base definition included having a bank account (checking or savings) at a formal bank or credit union. Other options now include online (or digital) banks, like Varo or Chime, and money transfer systems (e.g., Cash App) that are considered neobanks. 8 Digital banks and neobanks are routinely described as informal options; they typically do not have a minimum balance requirement, have fewer fees, and they have lower eligibility thresholds, accepting people with poor credit histories and scores. As Chime (n.d.) described, their company provides a “second chance bank account.” Entry requirements are also often lower; for example, Cash App requires limited personal information (phone number or an email address and a zip code) to become a user, and there are multiple ways to deposit money into the app, including paper money at a participating retailer. Cash App also offers a debit card option, and Chime requires a checking account, which makes the distinction between formal banks and neobanks blurrier. These enhanced services require additional personal information (including full name, date of birth, the last 4 digits of the applicant's social security number, and mailing address) and enable users to “deposit paychecks, tax returns, unemployment benefits, government stimulus payouts, and more directly” (Cash App, n.d.) Cash App also describes complying with government authorities (e.g., garnishing wages), a feature more typical of traditional banks. Recognizing this is a gray area, we considered online banks and neobanks as informal options but also paid close attention to how our participants define these entities. We reviewed summaries together to ensure we were capturing the same information, and we made several adjustments to capture additional details.
To identify (non)engagement explanations, we adopted an iterative coding approach. The researchers independently conducted line-by-line coding in relevant sections of the transcripts within each key institution (medical, financial, and employment). We were aware of a preliminary set of theoretical reasons for nonengagement drawn from the existing literature—such as child support obligations, lack of knowledge, or structural exclusions (i.e., collateral consequences)—but employed an inductive approach that allowed for the emergence of new themes (Saldaña 2021). For example, initial codes within the employment domain included the SSI dilemma, under-the-table hustling, criminal record barriers, identification/license, logistical barriers, and medical issues; we placed the first two in an “Alternatives” theme, criminal record barriers into its own top-level category, and the remaining into “Another barrier to employment.”
Initial categorizations and identified themes were compared through iterative discussions where codes were refined based on each round of feedback. 9 Discrepancies most commonly arose when coding formal or informal statuses due to the unexpected complexities in what “counts.” This directed us to co-create clearer and more comprehensive definitions. 10 While we did not disagree on any explanation codes, we had a few initial discrepancies in how to cluster subcodes within broader categories (e.g., whether medical issues should be within employment barriers or separate); when this occurred, we reached a consensus through discussion. Finally, one author drafted a summary of the key findings, which the co-authors then reviewed, including spot-checking coded segments in the transcripts for accuracy.
Results
Levels of Engagement
While we initially set out to examine nonengagement decisions, our first finding is that the sample overwhelmingly reported formal institutional engagement at the time of the interview. As displayed in the first graph in Figure 1, over a third of study participants were engaged with one (34%) or two (38%) institutions, and a quarter reported full engagement (i.e., actively involved in all three). Only two study participants reported complete nonengagement.
Engagement by number of institutions and institution type. Note: Everyone engaging with two institutions was engaged with medical care.
Another notable finding, displayed in the second graph in Figure 1, is that engagement varies widely by institution type. While only 31% of the sample was engaged in formal employment at the time of the interview, 59% percent engaged with formal financial institutions and 98% with medical institutions. As shown in Table 1, the full (non)engagement patterns look similar between individuals recruited by the ORC and other methods, although the non-ORC group is disproportionately engaged with just medical (40% vs. 26%), and the ORC group is more commonly engaged with the other institutions.
It is also worth emphasizing the prominent role of medical institutions in driving high overall engagement levels. Although this population is typically financially eligible for Medicaid, most lose coverage while incarcerated due to the Medicaid Inmate Exclusion Policy and have trouble restarting coverage post-release (Albertson et al. 2020). However, due to an ongoing collaboration between MassHealth (the state Medicaid program) and the Department of Correction (DOC) since 2015, eligible incarcerated individuals on pre-release status (within 30 days of release) are (re)enrolled. As a likely result, all but two respondents reported having insurance, and most had MassHealth. As our participant Sean, 11 a White man in his 30s aptly summarized, medical care is not “ever an issue because we’re in Massachusetts … they can’t deny you medical.” Around a third of our respondents were only involved with formal medical institutions; if we were to remove this institution from the analysis, financial institutions would be the most popular form of formal engagement at 59% (which is similar to Bryan and Collins’ [2024] national estimate for formerly incarcerated adults).
However, there are multiple ways to measure medical engagement, and another option is to focus on just those engaged with primary care physicians (PCP) for preventative or ongoing care. Even here, the engagement levels are high: 88% of our sample reported having a PCP. Around a tenth of our sample referenced Barbara McInnis, a Boston Health Care for the Homeless Program, as a current or past provider. We grappled with how to categorize this organization; it could be considered informal because staff can provide care anywhere, including on the streets, in shelters, or in clinics. However, the nonprofit organization receives funding from the federal Health and Human Services and provides comprehensive care (including primary care, mental health, dental care, recovery services, and specialized services) that is connected to retrieving and maintaining medical records (McInnis Policies & Procedures 2021). While one of our participants did not seem excited about being referred to Barbara McInnis, most reported positive experiences. As Terrance, a Black man in his late 50s, explained, “it's easy to get to and they look out for you.” Curtis, a 60-year-old Black man, described becoming more deeply engaged with medical care because of the organization. As he explained, “They did everything for me, everything. That's how I got my benefits. That's how I got my doctors, my medical appointments, my transportation, the ride. They did everything.” For those not ready or able to fully engage with more traditional formal services while experiencing homelessness, this type of quasi-formal institution was often viewed as valuable. Next, we examine the fully nonengaged individuals in detail to understand widespread nonengagement, followed by nonengagement explanations for formal financial and employment institutions.
Widespread Nonengagement
We refer to the two fully non-engaged participants by their assigned pseudonyms, Aaron and Raymond. Aaron, a non-Hispanic White man in his late 30s, had been out in the community for several years at the time of the interview. He had a history of drug addiction, an estimated dozen convictions, and was unemployed. Aaron did not have a primary doctor and said he would go to the emergency room if he needed care. However, rather than not engaging due to system avoidance, he seemed to engage where and when he could. He had bank accounts before—but not after—his prison sentence, and attributes his current detachment to his financial choices leading up to incarceration. After taking out a payday loan, I started opening bank accounts everywhere just so I could get a line of credit… So I racked up a considerable amount of debt in a very, very short time with no intentions of ever paying any of it back… And because that was never paid back, even though it was over 14 years ago, they will not let me open a bank account. (emphasis added) …just trying to get a fucking Mass ID was … a nightmare. I had to make three separate trips to the RMV [Registry of Motor Vehicles] for them to tell me twice I didn’t have all the proper documentation. That was probably one of the most stressful things about getting out, was just getting that shit in order.
Likewise, Raymond, a Cape Verdean man in his mid-30s, also seemed to work within what was accessible to him. Raymond was aware of Barbara McInnis, the clinic that provides free health services to the homeless without requiring identification or formal paperwork. Although he needed dental work and was seeking mental health services, he was reluctant to set up an appointment because he didn’t “like to frequent that [geographic] area too much … it has everything that signifies incarceration.” He also described feeling conflicted about financial institutions; when asked if members of his community trust banks, he replied, I’d go by the model of what my grandmother would say … my cash is safe[r] under my pillow than in a bank … and I feel like in the inner city predominantly that people don’t trust banks because they feel like they can’t access their money when they actually want it. …I also would rather have my cash on me. But I do understand that it's important … to have credit to be able to have savings and put your money away, you know, for a rainy day.… I’m kind of torn about it… But I do know that it's necessary, whether I trust it or not…
Rather than attempting to exist entirely off the grid—which may be a reasonable assumption if we only had access to a handful of survey questions for Aaron and Raymond—even those reporting the most detachment from formal institutions were open to, and even actively seeking, formal engagement. However, unlike Aaron and Raymond, most of our participants engaged with at least some institutions post-release. We turn to their experiences next to better understand formal institutional nonengagement and related concepts.
Formal Financial Nonengagement
Among the 40 participants who were not engaged with formal financial institutions at the time of the interview, there was a range of alternative sources for money transactions. Western Union was a popular choice when people needed to transfer money to (or receive from) family or friends. Direct express debit cards (which enable access to federal benefits), gift cards, prepaid cards (e.g., Netspend or RushCard), and cash were also common. The use of these alternative sources was often connected to at least one of three explanations: limited funds, access issues, and avoiding wage garnishment. This section discusses each explanation in turn, followed by a primary alternative option our participants identified: online banks and neobanks.
Limited Funds
The most common issue was a lack of money, or not having enough money to make having an account possible (i.e., not meeting minimum account balance requirements) or worthwhile. As Isaiah, a Black man entering his 60s who was starting a job training program but not currently employed bluntly put it, “I don’t have money like that.” Salvador, a White man in his mid-60s, expanded on this idea: Money is so tight that it's hard to just take a lump sum and put it in the bank … to have for the checking account. I’m going to, what, put it in today and take it out tomorrow, when writing a check? …I trust traditional banks [but] it's sometimes the fees. That's why I like [my] myWisely [mobile banking app] account because it's no monthly [fee] … like with those [formal] banks, I went somewhere and you got to pay almost $30 a month, but you got to keep a minimum balance. If you… keep over $1,500 in your account, you don’t have to pay fees. So sometimes I feel like banks charge too much in fees. Because it's online banking, and that allows me … to get a debit card without having to put nothin’ in the bank, like my balance is zeroes. It's been zero since I started the account… They didn’t penalize me for that.
Access Issues
Another set of explanations involved access issues. A small subset of respondents indicated that they were looking to open an account soon but were waiting for something—such as paying off old accounts in collections (i.e., the debt was sent to a third-party collection agency) and/or getting formal identification. Angel, a Puerto Rican man in his 40s, described how a previously frozen bank account created challenges; even after the issue was resolved, he declined to reengage with banks. Although he “could use it now, cause [he's] legal,” Angel referenced needing to find documentation and consult with family members, making reengagement “too much trouble—they give you too much trouble for that.”
Some also encountered information access barriers, such as not knowing how to obtain an account or only understanding certain services. Gustavo, a Puerto Rican man in his 50s, stumbled through an explanation for why he did not engage with traditional banks. He initially reported, “Because I can’t,” and the interviewer encouraged him to elaborate. Gustavo continued, “That's something—I don’t have credit for … that's [why I] only use that Western Union card because … for credit, I don’t got anything.” Luis, a Puerto Rican man in his early 40s, described not trusting formal banks because “anything can happen.” He mentioned hearing negative stories from others, including a situation where a bank “couldn’t refund, like give back the money.” Luis concluded that “the best way to go is to have a stash … in cash.” When the interviewer asked if he would rather keep his cash at home, he replied, I don’t even know how that works. You got money in the bank, and when times goes by, you—you owe right?… All my life … it's been about, selling drugs and doing drugs… That's all I know. That's the life I know.
Herd and Moynihan (2018) would categorize these different access issues as administrative burdens, even in cases of logistical barriers that can potentially provide a benefit (e.g., requiring identification to protect against fraud) or when barriers in one domain (such as employment) are potentially connected to other administrative burdens (such as financial access). Institutional interdependence, or the idea that experiencing a challenge in one institution led to difficulties in another, was also a recurring theme across institutions, and something we further explore in a later section.
Avoiding Garnishment
Finally, a handful of participants pointed to concerns about bank garnishment, which aligns with the system avoidance perspective.
12
Ramiro, a Cape Verdean man in his 30s, reported having three children under the age of 18 and owing tens of thousands of dollars in child support. He reasoned that if he had a bank account now, “child support would take all my money anyway.” Similarly, Thomas, a White 50-year-old father of two, said he used only prepaid debit cards. He explained that he “won’t get bank accounts, [because] I still owe an outstanding child support from when I was in jail… I’m afraid they’re just going to confiscate all my money.” For Thomas, his concerns over garnishment were validated when one of his COVID-19 stimulus checks was directly routed to child support: “everybody got the one [check], then child support takes the other one.” Although Christopher—a White man in his late 20s who did not report having any children—had never experienced wage garnishment, he reported a similar explanation for not setting up a bank account: I have an issue with owing the government a little bit of money. So I have [the bank account] under my mother's name.
Roberto, a Hispanic man in his 30s, and Christopher were the only two participants not involved with formal financial institutions who suggested avoidant behavior also affected their employment activities: Christopher was reluctant to get a job because of wage garnishment and Roberto drifted in and out of employment, sometimes selling drugs and cigarettes as a side hustle or working for employers for gift card payments. However, Roberto was earning a stipend through an education program and was planning to seek more formal employment after securing identification. In this sense, the avoidant behavior appeared specific to the affected institution, rather than widespread across institutions.
Still, the three overarching financial nonengagement explanations (limited funds, access issues, and avoidance) led some of our participants to seek out alternative banking options. Interestingly, however, this did not entail payday lenders or elusive companies. Instead, participants engaged with prominent online or neobank services, as described next.
Second Chance Banking as an Alternative Option
A tenth of our sample (including the two fully formally nonengaged participants, Aaron and Raymond), used online or neobank services such as Chime or Cash App as their primary form of banking. Marvin, a Hispanic man in his early 20s, explained that Chime is a useful service for “simple bank users.” He withdrew money from his formal checking and savings accounts and switched to Chime because “I wasn’t employed or there wasn’t much. It wasn’t much needed.” Christopher, who we introduced earlier and is a few years older than Marvin, also used digital services in limited ways, although, as he explained, that was because “I really don’t know how to use banking too much.”
Digital banks also presented an option for those with restricted choice sets. Diego, who was 40 at the time of the interview, “defaulted in one bank or another in my past,” noting that formal banks “wouldn’t take me.” He searched online and discovered online banking services. When asked if Chime is his main account, he replied, Yeah, and I love it. ‘Cause I’ll tell you this, I don’t know if it was beginner's luck, but ... there's certain places, like mostly direct deposit, you get it [the paycheck] two days early, Wednesday… Tuesday morning it was there. I’m like, whoa, wait a minute. I have to do an online bank account. I can’t get a real bank account. I can’t get a credit card, ‘cause everything went into debt, in collections… I was gone (incarcerated) for a year and a half. …obviously, I just have to pay off the bank thing no matter what because if I pay that off, I can open up a normal bank account. Which, I mean, online banking's fine, it's just, I’d rather have a bank I can go to. ‘Cause when you use the card … it's like a $3 fee from one ATM and then a $3 fee for using the ATM and then a $5 fee. I don’t understand where the $5 comes in, but it's $8, it's over $8 … you get three charges. It's crazy.
Overall, there were mixed reviews about these alternative digital financial accounts. A few of the banking services our participants mentioned have since shut down, and the fintech industry is under scrutiny. In September of 2024, U.S. Senators Elizabeth Warren (D-Mass.) and Chris Van Hollen (D-Md.) raised concerns about the lack of consumer protection regulations when fintech companies cease to exist or in other situations that are not covered by FDIC insurance, including frozen or locked accounts rather than quick payouts (Warren 2024). Yet this style of banking opened access for those unable to access formal options, including removing the minimum balance restriction and other forms of inaccessibility. These services reduce administrative burdens and provide the “second chance banking” that Chime (n.d.) advertises—but at a potential cost.
Formal Employment Nonengagement
Although our participants were the least engaged with formal employment across the three key institutions (Figure 1), not being employed was not typically due to a lack of perceived need or lack of trying. Of the 67 people not formally employed at the time of the interview, more than two-thirds reported they were actively looking for jobs, preparing to start a new job soon, or were engaging in educational workforce development programming to position themselves for the job market. However, most encountered barriers or difficult choices that influenced their formal and informal work experiences. The next subsection discusses shared barriers (most notably, criminal record and supplemental security income challenges), individualized barriers, and other exceptions to the shared barriers theme. Following this, we describe our participants’ experiences with and perceptions of informal jobs that served as alternative options.
Shared and Individualized Barriers and Delays
Criminal Record Barriers
Several participants expressed frustration with continuously applying for opportunities, and the most common shared challenge was related to criminal record (known in Massachusetts as Criminal Offending Record Information, or CORI) barriers that people experienced or perceived. As Omar, a 45-year-old Black man with a GED diploma, described: …it's been hard, nobody really want to take a chance and I just almost had a job with Amazon, right. I had everything taken care of, passed the drug test, everything, it was now just the CORI shit, the background—they did that and right there just ‘x-ed’ the job, like they never got back to me. Yeah, I applied to a warehouse in [Massachusetts city] and I didn’t get the job and I asked a ZipRecruiter guy and he was like … he was trying to tell me, “No, it didn’t have anything to do with my CORI” but I know it did because these are all happening after the [electronic] bracelet incident. Now he [the recruiter] don’t even want to work with me. …I guess that's the biggest thing.… I don’t want to find that I’m qualified for a job, get excited about a job, and then have it [the CORI] come back—“Oh, we can’t give you this job because of your criminal [record]”—I could not.
The Supplemental Security Income (SSI) Dilemma
A subset of our respondents identified a different shared barrier: what one person referred to as the Supplemental Security Income (SSI) dilemma. As John, a White man in his early 40s, explained, after completing a set of classes, he would be eligible for a job but faced a dilemma: “I … was at a standstill and I’m like what the fuck should I do?” John went on to explain that he spoke to an SSI lawyer, who indicated it was “not good to work while you’re applying.” As a result, he was trying to find under-the-table work or connect with a friend to find work at a temp agency to secure the benefits.
While receiving SSI (or less commonly for our participants, Social Security Disability Insurance [SSDI] for those with qualifying work histories), there are work limitations to remain eligible. In total, a tenth of our sample reported strategically seeking part-time work, side gigs, or under-the-table jobs to retain benefits with income eligibility restrictions. As Herbert, a Black man in his early 50s explained, “even on the books I can work up to 15 hours and still not affect my [SSI].” After identifying an employer who was looking for workers on the weekends, Herbert was “waiting to hear something from them in the next day or so” and “keep[ing] my fingers, toes, everything crossed to hope at least somebody call me.” Brenda, a Cape Verdean woman around the same age as John, was similarly knowledgeable about the employment restrictions. After being offered a part-time job at a convenience store, she noted, I’ll get $13.50 an hour but I can’t work more than 20 hours a week and I can’t make more than a certain amount of money anyway… I can’t make $20 an hour … and stay on [SSI].
Individualized Barriers and Exceptions
In addition to these two key shared barriers, participants reported other challenges that were varied and individualized. For one person a medical issue could be the driving factor, while it could be a lack of car access or a license for another. Each additional barrier can cut off types of jobs, industries, or geographic areas with available work. While targeted applying can be a valuable job search strategy (Vuolo, Schneider, and LaPlant 2022), each added restriction further minimized limited options.
In addition, some of the non-employed participants were exceptions to the barriers theme; this subset typically did not view the CORI as a barrier (at least in the types of work they were pursuing), did not report SSI or other additional barriers, and/or otherwise appeared optimistic. As Diego, who attended high school until 9th grade but had past work experience in food service, maintenance, grounds keeping, and warehouse positions, expressed, “I’m hoping it doesn’t take that long to catch a job. There's a lot of people hiring.” Ramiro, the participant who described child support issues above, was also open to a variety of job types and was confident about securing one: Every time I try to find a job in Boston, I get one… I can’t make … no complaints. It's easy, this is Boston man, [if] you can’t find a job in Boston man, you just … ain’t looking. I mean, I won’t settle… [but] I’m not going to just wait for [a] certain job. I’m [going to] grab what I can get, and maybe if something better comes along, then I’ll switch it up…
Precarious Work: Side Gigs, Hustles, and Temp Jobs as Alternative Options
Informal work, which includes side gigs, side hustles, and temporary (temp) jobs, is common among marginalized populations (e.g., Edin and Lein 1997; Venkatesh 2006), including formerly incarcerated individuals (Harding et al. 2014; Sugie 2018; Western and Sirois 2019). Around 20% of our participants not employed at the time of the interview reported engaging in informal income options. Among those not working formal or informal jobs, around half indicated they would work a side gig if they found an opportunity. Reported side gigs 13 included cutting hair, detailing cars or mechanic work, selling bowties, landscaping, shoveling snow, cleaning, handyman repairs, and “boosters” (buying or shoplifting products and reselling at an upcharge). Some considered these gigs part of self-run underground businesses (what Edin, Lein, and Nelson [2002] called “informal-sector entrepreneurs”) and people sometimes had other self-directed hustles like selling food stamps or blood/bone marrow.
Perhaps unsurprisingly, the top-cited reason for taking informal jobs was to make money. Salvador, who was one of our oldest participants at age 65, told us that “the bottom line to it all, is the money.” He went on to explain, That's what we’re all working for, right? That's why we work, to get money, to survive. So whichever is more beneficial, that's the way I’d like to go… I’ll go back to the age thing. It's not like I’ve got a big plan. I’m already over the hump… I’m not thinking about a career and a life ahead of me and a family and so on. I’ve already passed that… The best I can do now is just maintain some kind of happiness… …I have not made more than like maybe 150–200 dollars at the end of the work week … but the pros of having a side job is it helps supplement accidentals so to speak, something you need breaks on the car… And working on the books is good too… I don’t mind working on the books, but see working on the books is going to change my social security [benefit].
Kalleberg (2018) described precarious work as having three key features: it is volatile, without benefits, and lacking protections. Our participants’ descriptions aligned. In addition to earning minimal money, key disadvantages included inconsistency, issues related to time (e.g., finding side gigs to be a waste of time or unpredictable for scheduling), and the lack of protection. As Kevin, a White man in his late 30s explained, [A side gig is] not something that's going to last. You know, it's just maybe like a day or two that they need you.
Garrett enjoyed his informal job but noted it could pose challenges for formal employers. A Haitian American in his mid-30s, Garrett secured a job within two months after release at a grocery store. He then moved to another grocery store chain, but the start of the pandemic upended that job in March of 2020. We spoke to him almost a year later, as he was about to start a job at U-Haul and was working a security job under the table for $20 an hour. Garrett thought the security job pay was “good” and the role was “in my lane,” but described how having an informal job could “mess up your schedule” when you are also seeking or holding formal employment. He explained: A lot of employers, a lot of employers, they don’t like that. Like, they’ll say “Oh, keep doing your thing,” but a lot of employers really want people to be there whenever they want them to be there.
When considering the lack of protections, concerns included not having retirement savings, health benefits, or protections if injured on the job. As Joseph, a Black man in his 30s who reported starting a job “like, yesterday,” and was open to side gigs, put it, “It's not steady work … No insurance … no benefits, really. It's a side job.” Eugene, a White man in his mid-50s, told us he would “definitely consider” a side gig, but noted the lack of benefits as a major disadvantage. He explained, “…you don’t get benefits, you can’t buy health insurance. So yeah. Unless it's serious money…”
The distinctions between formal and informal work—and the accompanying benefits and downsides—became blurry when discussing temp agencies. Chelsea, who was in her early 30s and identified as female (but in the early stages of transitioning to male), was seeking employment through a temp agency. As she described, the pay arrangement: …depends on the company, because they have plenty of different wages that they can pay, some pay under-the-table… I was working at a security job, and he paid me through his personal checks, you know the whole team, but I can’t do that because I pay taxes, I mean it was nice, but I gotta worry about my taxes so I can collect retirement. It's just a birth certificate, Social Security card and ID. [The agency will] match you with a job, you do, like a one hour, 30-minute training and boom, you’re working. It's not like applying for Target, waiting three weeks just for your application to get seen and for them to say no. And you finding all these jobs for them saying no, no, no, no, and you’re … left blind, basically.
Given the typical features of informal work—including sporadic hours and unstable income—under-the-table work is typically considered a “bad job” relative to formal work. As Sugie (2018, p. 1481) notes, “the characteristics of foraging are directly at odds with the reasons we value work and employment … foraging for work may actually exacerbate poor outcomes.” Informal work has also been found to be positively correlated with the probability of engaging in crime—regardless of whether the individuals are also engaged in formal employment or not (Nguyen, Kamada, and Ramakers 2022). However, when facing dire employment opportunities, informal jobs can also serve as an “economic safety net” (Ravenelle 2023, p. 2) and provide a temporary solution.
Bounded Engagement and Institutional Interdependence
Examining multiple institutions in depth enabled us to observe both bounded engagement and institutional interdependence. The notion of bounded engagement draws from Ewald's (2024) original description of bounded legal cynicism, which describes how people may have deeply negative views about a particular institution or set of institutions (in his context, the police and employers) but are interested in engaging with other institutions (such as juries or voting). In that sense, negative perceptions and subsequent opting out might be “limited or domain-specific” (p. 9). Similarly, our respondents seemed to experience bounded engagement, where they were not engaged with the formal institutions actively presenting structural barriers, but were engaged with other formal institutions and often planned to re-engage with the restrictive institutions when they could. In the meantime, many temporarily sought out informal alternative institutions to meet their needs.
The “need-seek-access” diagram displayed in Figure 2 can help illustrate how different choices and barriers to seeking formal institutions can lead to (non)engagement. Among those both needing and seeking formal services, people may fluctuate between statuses—successfully engaged, ineligible (formal system exclusion), and pushed out (administrative burdens)—within an institution over time, or across institutions at a given point in time. Although the interconnected nature of institutions is particularly visible with the system avoidance nonengagement explanation, as people avoid one institution to prevent surveillance or information-sharing with another, system avoidance was bound to the institution(s) directly impacted in our study (see also Ewald 2024; Fong 2019). Instead, we more commonly observed institutional interdependence, or barriers in one formal institution leading to barriers in another.
Formal nonengagement displayed through a need-seek-access framework. Note: This figure aims to organize major types of nonengagement pathways, but we recognize it may not be exhaustive. For example, there may be other reasons people do not seek formal services in addition to active avoidance; the diagram is condensed here for clarity and to map overarching explanations proposed in past research. An additional possible extension is to engage with alternative institutions; that is also not illustrated here.
The connections between institutional barriers were most noticeable when considering formal employment as an obstacle. Respondents reported waiting for various events to occur or conclude before fully engaging in a job search or being able to secure employment. The identified problem domains included education (finishing a program, certificate, or licensing requirements), identification (acquiring a driver's license or other state-issued card), transportation (completing a Cinderella license [i.e., restricted driving hours] or having transportation access to jobs), medical (future and often distant appointments to resolve injuries or illnesses), and halfway house rules (restricting hours out in the community, and therefore, work schedules).
Scott, a 29-year-old White man, illustrated the potential complexity of institutional dependencies. He had been in recovery for five months when we spoke to him and described an unstable housing situation, where he was “living off of people's couches” until he recently started staying in a halfway house. However, he was interested in building his credit score and applied for a Capital One secured credit card. He was unable to receive the card because he lacked proper identification, and he could not obtain a state ID due to outstanding debt and his housing situation. He explained, …I owe thousands of dollars to get my license back and to get a state ID, you need to have two forms [of] ID plus proof of address, which I don’t have. …when I was offered a couple of jobs like on the spot, like, “Hey, I’ll hire you right now, you got a ID.” “No, I don’t. You got to hire me under-the-table or not me at all.” And that doesn’t go over well when you're talking to an employer who wants to hire you right now. But you need to do it on the books. You can’t hire me on the books because I don’t have an ID.
While barriers in other institutional domains (in Scott's case, debt, not having a permanent address, and lacking a state ID) could block employment access, not having a job also served as an institutional barrier to other needed services. For example, we asked Carl, a Black man in his early 50s who had resided in the community for about 15 months, what keeps him away from banks. He pointed to needing a solution to paying back a thousand dollars he overdrew on a previous bank account: “Well, I’m trying to get another job, so I can get some money to pay the bank back.” Working under-the-table was not in his option set—he said he had “Never thought about it” when we asked—leaving food stamps and SSI as his main sources of income. More broadly, our participants appeared to weave through institutional barriers and administrative burdens to find the best opportunities within their available option sets at a given time.
The Limitations of Engagement
For the sizable subgroup of participants who engaged with formal institutions, they discussed needs, easy accessibility, and positive past experiences as motivations to engage. Yet even for the group fully engaged in formal institutions post-release, this was not always synonymous with perceived financial or employment stability, and they often had complicated relationships with formal institutions. As an example, our participants differentiated between bank accounts and available credit. Edward, a 40-year-old White man, had bank accounts and credit cards beforehand, but because of his incarceration, his credit card debt “went unpaid” and his credit “is absolutely garbage right now.” While he currently has a checking account, he has not applied for new credit cards because he is unsure if he will be approved. When asked about his current financial situation he noted, “…my head's barely above water but I can breathe.”
Other participants were similarly reluctant or cynical about these institutions even as they were actively engaged. For example, Breanna, a Black woman in her 50s, describes banks as “us[ing] our money” while the interest earned is “hogwash.” Sean, our participant who was quoted earlier when explaining widespread access to health insurance in the Commonwealth, similarly expressed distrust toward banks. His wariness stemmed from a bad personal experience; as he described, “There was one time I used my debit card and I got charged for buying a bottle of alcohol like seven times.” Still, he currently uses them because “I have to” and acknowledged “I guess my [current] bank isn’t so bad.”
Our engaged participants routinely described their current financial situation as “okay” but needing improvement and often expressed suspicion as they struggled to reach their goals. However, removing barriers to access, improving services, and increasing available information on (and transparency about) services can all work to improve experiences among those currently engaged. While formal employment has increasingly moved towards a reduced worker protection model, including replacing traditional job roles with gig workers (Ravenelle 2023), there may be new incentives for the formal financial sector to reduce burdens. As the Vice President and Product Manager at JP Morgan & Chase explained, “Traditional banks, long revered for their stability and trust, are now leveraging fintech innovations to reshape their operations and customer interactions” (Komandla and Perumalla 2017, p. 1).
Discussion
Formerly incarcerated adults are more disconnected from institutions relative to their non-system involved peers, but a key empirical question concerns what drives (non)engagement decisions. Our analysis of in-depth interviews with formerly incarcerated adults in the Greater Boston area enabled us to examine formal institutional engagement and led to two key findings.
First, almost everyone in our sample (98%) engaged with formal institutions at the time of the interview. This high level of engagement was driven in part by the availability of medical insurance for formerly incarcerated adults in Massachusetts (Western and Simes 2019). While medical institutions are often considered “surveilling” due to the type of personal data collected and stored (Brayne 2014; Goffman 2009), the DOC/MassHealth pre-prison release enrollment program shifts the administrative burdens from individuals to the state, increasing accessibility (Herd et al. 2013). Even when excluding medical institutions, engagement was high; around two-thirds of our sample engaged with at least one of the other major institutions we considered.
Our second key finding is that formal institutional engagement was often bounded. Our participants’ option sets were shaped by barriers, but they typically were engaging or attempting to engage whenever feasible (see also Ewald 2024; Fong 2019). In other words, there were restrictions or structural bounds influencing whether (and how) our participants could engage with formal institutions, but inaccessibility in one domain did not dissuade formal engagement in other institutions or create permanent nonengagement. For example, even those actively accessing informal alternatives such as neobanks or side gigs often expressed concerns about those options and were hoping to establish a “real” bank account or secure a formal job soon. The people we spoke with rarely seemed disinterested or adverse to formal institutions; instead, these opportunities currently were (or were perceived to be) closed off to them, largely due to administrative burdens. Our respondents described learning, psychological, and compliance costs with both financial institutions (e.g., not meeting minimum balance requirements or having the knowledge to set up an account) and in the employment context (e.g., stigma) (Moynihan, Herd, and Harvey 2015). The delays in engagement that were associated with compliance costs also revealed the institutional interdependence our participants encountered. Logistical barriers or delays in one institutional domain (such as obtaining state identification, unfreezing a bank account, following a halfway house's schedule restrictions, or waiting for a medical appointment) could influence a person's option set in another domain (such as the labor market). Our participants’ persistence with bounded formal engagement was particularly striking given the interconnected nature of administrative burdens.
Theoretical Contributions and Extensions
By examining (non)engagement across multiple institutions using rich qualitative data, we are able to contribute evidence and extensions to three existing explanations. First, we build off Bryan's (2023) and others' recent advances towards understanding administrative burdens for formerly incarcerated individuals. In addition to formal exclusions, which were experienced and/or perceived in the formal employment context for some of our participants, administrative burdens can be a concealed but still impactful form of coercive opting-out. Second, our findings suggest that system avoidance, a common nonengagement explanation, appeared to be bounded. When avoidance occurred, it was typically related to specific events, including garnishment for child support payments, and in a few cases, when people were actively engaged in crime (although instances of the latter were all reported to have occurred in the past). This can be useful for both theoretical and policy purposes: if formerly incarcerated people refuse to engage with institutions altogether due to fear of further monitoring from the criminal justice system, some policymakers may feel powerless to respond. However, if system avoidance is bounded and other explanations predominantly drive nonengagement, the ability to improve access to institutions and services lies directly with policymakers and practitioners. This does not negate returning citizens’ concerns about state surveillance but rather highlights the importance of structural inclusion policies (Bell 2017) and reducing administrative burdens.
Third, we build on existing sociolegal frameworks for understanding alternative engagement strategies. In Augustine's (2019) study of job seekers with felony records on community supervision, she illustrated how her interviewees worked “around” the law to comply with the mandate to secure a job when they encounter legal blockades to doing so—what she terms a legal double bind. Our respondents similarly encountered legal limitations for working while retaining federal benefits and worked around what our participant John termed the “SSI dilemma.” In addition to within institutional barriers (such as conflicting legal mandates related to formal employment), we also detected restrictive interdependencies across institutions.
Policy Implications
Administrative burdens have been described as a purposeful “policy instrument” (Moynihan, Herd, and Harvey 2015, p. 52). People reporting conservative beliefs and higher levels of racial resentment tend to support administrative burdens (Haeder and Moynihan 2025), and when policymakers are unable to exclude people legally or formally from accessing state programs, those decision makers may instead rely on burdens to dissuade eligible participants (Moynihan, Herd, and Ribgy 2016). However, increased accessibility and reducing learning, psychological/stigma, and compliance costs can improve how people experience reentry and (re)integration after prison. As such, monitoring which burdens exist within and across entities can help identify burdens that can be mitigated or fully removed (Moynihan, Herd, and Harvey 2015).
Tangible policy solutions to reduce identified administrative burdens include reducing paperwork and streamlining access (Hunt et al. 2018) and using simple and non-bureaucratic language to communicate eligibility and other criteria (Baekgaard, Döring, and Thomsen 2025). While certain information strategies may be more effective (such as video clips improving understanding about program eligibility over flyers), “providing information, even in a basic manner, has important consequences for comprehension” (Walsh, Porumbescu, and Hetling 2024, p. 8). Medicaid policies provide a model for the large-scale shifting of burdens from people to the state; in addition to Massachusetts, Wisconsin has also reduced Medicaid administrative burdens for the broader population and individuals released from prison (Burns et al. 2021; Dorn, Hill, and Hogan 2009; Herd et al. 2013). At the national level, recent guidance from the Centers for Medicare & Medicaid Services (2023) also enabled correctional facilities to obtain “qualified entity” status to make Medicaid eligibility determinations. This allows facilities to enroll incarcerated people pre-release to have coverage as soon as they return home. 14 However, recent proposed changes to Medicaid requirements would reverse some of these changes and exacerbate administrative burdens through requirements to reverify eligibility every six months, added work requirements, and restrictions based on immigration status (Center on Budget & Policy Priorities 2025).
Informal alternatives to formal institutions can also pose a conundrum. “Second chance banking” and side gigs or under-the-table work can be attractive alternatives for marginalized populations because they alleviate administrative burdens and remove barriers to entry. As such, they can provide an opportunity not otherwise available. However, these alternatives can also be risky. Digital banking services have notable consumer protection concerns (Warren 2024) and precarious work is known for its instability, lack of benefits, and lack of worker protections (Kalleberg 2018; Ravenelle 2023). One way to strengthen protections includes legislation to shape the boundaries of ethical institutional behaviors. Another option is having policymakers encourage formal institutions to voluntarily improve services. In the employment context, this could include extending benefits, time off, and insurance to part-time employees and paying workers that are not truly “independent” as W-2 employees rather than 1099 contractors (Ravenelle 2023); in the financial sector this might involve reduced bank fees and penalties. Rather than framing these changes as altruistic, employers benefit from loyal and productive workforces and banks profit from retaining customers that are leaving for fintech alternatives (Ravenelle 2023)—and research indicates strategic issue framing can influence policy adoption (e.g., Chen, Lageson, and Adams 2025). Similarly, formal institutions should reduce financial and logistical barriers to entry. In the employment space, for example, there have been attempts to remove processes that create administrative burdens (e.g., expungement applications and fees) and instead shift these burdens to the state (e.g., automatic expungement). As employers move towards individualized assessments for the large portion of job applicants with criminal records who are ineligible for sealing or expungement (e.g., Denver 2020), a key question is how applicants can provide positive information without taking on excessive burdens.
Finally, given the problematic nature of institutional interdependence described by our participants, alleviating burdens that are not directly connected to banking or employment could have notable benefits to those seeking new accounts or jobs. Easier access to obtaining state-issued identification, for example, can expand access across a variety of essential domains, such as employment, banking, housing, healthcare, and public benefits. As our respondent Lisa, who used a mobile banking app because she encountered formal banking barriers, suggested, There needs to be like a thing at the RMV [Registry of Motor Vehicles] that you can walk in with your prison [ID] and be like, I need an ID, and they fucking print it out and give it to you right there… There needs to be easy, like, user interface services.
Study Limitations
We note several limitations in the current study. First, it is possible that this older sample of returning citizens—many of whom have been released from prison for several years and have high levels of formal education—is in the process of desistance, which may not reflect the experiences of higher-risk, less-educated, and younger samples. However, our multi-method recruitment strategy enabled us to interview people with incarceration records already living in the community rather than following people as they are released from prison, which presents a more diverse set of formerly incarcerated experiences. A second related consideration is that part of our sample is connected to the ORC. Those likely to engage in general system avoidance may not have chosen to seek services from a city-run agency. We sought to overcome this shortcoming by employing a range of recruitment strategies, including posting flyers in transitional housing facilities and snowball sampling. Participants recruited through the ORC did not appear notably distinct in their levels of engagement.
Third, our sample was primarily composed of men. While returning citizens are disproportionately men, we hoped to have women represent a larger portion of our sample. Formerly incarcerated women often have heightened needs related to substance abuse, mental health, and prior victimization that create additional barriers to successful reintegration (Barrick, Lattimore, and Visher 2014; Spjeldnes and Goodkind 2009), and researchers have called for gender-specific reentry programs to address these differential needs (Cobbina 2010). The intersection of race and gender is also an important consideration in reentry research (Leverentz 2006), with prior studies demonstrating experiential differences when centering upon race (Garcia-Hallett and Kovacs 2020) and gender (Crewe, Hulley, and Wright 2017). Although we used targeted recruitment efforts at several local women's reentry facilities, fewer women contacted us. Among those who did, many were ineligible (i.e., had short-term jail stays). Among the few eligible women that we did not interview, most encountered scheduling or travel issues related to their community supervision restrictions. Future research extending the analyses here to a broader sample of women would be useful.
Conclusion
While formerly incarcerated individuals may avoid surveilling institutions in certain situations (e.g., when actively involved in crime and/or attempting to evade wage garnishment), formal institutional nonengagement appears to be more complex. Given the array of exclusions, administrative burdens, and interdependent agency compliance surrounding accessibility, many of our participants experienced what we consider bounded engagement. Most respondents engaged with at least one (if not multiple) formal institutions when they could; when they could not, substituting with informal options (such as online banking or side jobs) was often appealing when that was available to them. Still, while informal institutions had certain benefits, they were rarely the preferred choice given the instability and risks they carried.
Our recommendations include identifying, assessing, and alleviating barriers and burdens in formal institutions—while retaining and enhancing worker and consumer protections in informal sectors—to enhance accessibility and promote engagement opportunities. Given the interconnected nature of burdens (or institutional interdependency) that our participants faced, burden reductions in one domain could also have benefits for accessing other institutions. Finally, the complications we encountered with distinguishing formal and informal engagement across institutions—including primary care medical organizations that meet marginalized populations where they are, jobs and hustles that dip in and out of standard categories, and fintech banking options that blur the lines—are a useful prompt to reconsider what “counts” as formal institutional engagement. By moving beyond existing measures of formal engagement and encouraging a discussion about reconceptualizing these measures, we hope to shift the conversation towards minimizing structural and administrative barriers to reintegrative success.
Footnotes
Acknowledgements
This manuscript is based upon the work supported by the National Science Foundation (NSF) under Grant #2043437. We are indebted to the men and women who spoke with us and candidly shared their experiences and perceptions. Ayanna Miller-Smith and Cassidy Pereira also conducted many of the interviews and supported the broader project in important ways, and we appreciate the research assistance from several undergraduate and master's students. We would like to thank our partners at the Office of Returning Citizens, with a special thanks to Kevin Sibley for his early support of the project, Robert Turner and Rupert Saunders for their time, assistance, and invaluable advice throughout the recruitment period, and Ashley Montgomery for the ORC's ongoing support. We deeply appreciate feedback and insights from the 2023 Duke Law Criminology Roundtable, Alec Ewald, and the Northeastern faculty writing group. Finally, we would like to thank Isabel Wanner from the Mass Law Reform Institute (MLRI) for providing feedback on our summary of Medicaid policy reforms in the discussion section.
Declaration of Conflicting Interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the National Science Foundation (grant number 2043437).
