Debt Collection Pressure and Mental Health: Evidence from a Cohort of U.S. Young Adults

Abstract

The debt collection industry in the United States has grown in tandem with rising indebtedness. Prior research on debt and mental health mainly treats debt as a resource and liability, rather than a power relationship between creditors and debtors. We study the mental health consequences of debt collection pressure using data from the National Longitudinal Survey of Youth-1997 Cohort (N = 7,236). Drawing on stress theory and health power resources theory, we posit collection pressure as a relational stressor that undermines well-being through negative interactions with debt collectors, financial strain, role strain, and stigma. We find that more than one in three young adults in this cohort faced debt collection pressure by around age 40, with higher rates among low-income and Black young adults. Individual fixed-effects and lagged dependent variable regression models indicate that debt collection pressure is associated with increased psychological distress, with more severe consequences among low-income young adults.

Recent cohorts of young adults have come of age during a period of economic crisis, precarity, and uncertainty. The U.S. Millennial cohort, born in the early 1980s, transitioned to adulthood during the Great Recession and the slow recovery that followed, when successfully navigating the precarious economy depended in part on access to financial resources and power. Many young adults in this cohort turned to widely available consumer credit and have taken on historically high levels of unsecured debt to finance their transition to adult independence (Dwyer and DeMarco forthcoming; Dwyer, McCloud, and Hodson 2011; Houle 2014). During the same period, young people faced an unprecedented mental health crisis, with evidence that psychological distress has increased fastest among this age group in recent years (Zheng and Echave 2021). Debt is a well-known stressor in part because taking on debt entails entering into an unequal power relationship with creditors. How, then, has indebtedness impacted the mental health and well-being of recent cohorts of young adults?

Credit and debt have a complex relationship to mental health. On the one hand, credit can be a resource that provides liquidity for consumption and investments. In this sense, access to credit supports mental health, as young adults can use credit to borrow to fund investments and purchases, which reinforces self-concept and financial independence (Dwyer 2018; Dwyer et al. 2011). On the other hand, debt can be a liability that increases financial strain. As a liability, debt places a burden on financial resources that competes with the cost of daily consumption, investments, and medical care (Dwyer 2018; Nepomnyaschy et al. 2021). Debt repayment can therefore cause stress, contributing to worse mental health (Drentea 2000; Drentea and Reynolds 2015; Dunn and Mirzaie 2016; Sun and Houle 2020; Sweet et al. 2013). The mental health consequences of failures to repay debts, measured by delinquency and bankruptcy, are particularly severe (Addo 2017; Gathergood 2012, Sun and Houle 2020; Wolfe et al. 2022).

Both the resources and liabilities of debt derive from the unequal power relationship between creditors and debtors (Dwyer 2018; Krippner 2017). Creditors have legal power over debtors and dictate loan terms (e.g., interest rates), repayment schedules, and have legal authority to levy fines, fees, and other consequences on debtors who are unable to repay their debts. Scholars have thus moved beyond debt levels to focus on the disempowering experiences of indebtedness that can create mental anguish (Addo 2017; Drentea and Reynolds 2015; Sweet, DuBois, and Stanley 2018). A manifestation of that unequal power relationship and a key hypothesized pathway linking indebtedness to psychological distress is pressure from debt collectors (Drentea 2000; Drentea and Lavrakas 2000; Sweet et al. 2018). After a debt has gone unpaid for a few months, creditors typically send the late account to collections. The debt collection industry includes both creditors and third-party agencies and is a multi-billion-dollar sector that has grown in the 2000s (Consumer Financial Protection Bureau 2022). In 2015, 33% of U.S. consumers had at least one account in third-party collections, up from 27% in 2004 (Consumer Financial Protection Bureau 2019). The most common accounts in collections are medical bills, credit cards, personal finance loans, and telecommunications accounts (Consumer Financial Protection Bureau 2019). Debt collectors use a range of aggressive tactics to pressure debtors into repayment, including repeated calls at odd hours; disclosure of debts to family, friends, and employers; use of abusive or obscene language; and use of legal power to initiate wage garnishments and asset seizures (Consumer Financial Protection Bureau 2022). Prior research documents negative effects of debt collection pressure on debtors’ personal lives (Hochschild 2012; Sutton 1991; Sweet et al. 2018; Thorne and Anderson 2006), but no prior study to our knowledge has examined the link between debt collection pressures and mental health in a population-based sample of young adults.¹

We bridge concepts in stress theory (Pearlin 1989) with insights from the sociology of power, debt, and health (Drentea and Reynolds 2015; Dwyer 2018; Reynolds 2021) to argue that debt collection pressure is a relational stressor that undermines the mental well-being of young adults. Relational stressors reflect ongoing situations where relationships with a powerful actor causes stress for the less powerful person or group (in this case between a creditor and a debtor). We hypothesize that debt collection pressure is a relational stressor, exposing debtors to financial stress, and interactions with debt collectors that are both stigmatizing and potentially abusive. We anticipate that debt collection pressure will be more prevalent and stressful among less powerful groups in the financial system.

We test our hypotheses using data on a cohort of Millennial young adults born between 1980 and 1984 from the National Longitudinal Survey of Youth-1997 Cohort, who came of age during a time of growing economic precarity and rising reports of mental health problems. Young adults experience both higher rates of psychological distress (Lee 2023) and a greater risk of having debts in collections relative to older adults (Martinchek, Andre, and Santillo 2022). Our data contain a unique self-reported measure of pressure from debt and other bill collectors and repeated measures of psychological distress symptoms. The data also provide extensive measures of debt, wealth, and other socioeconomic characteristics, as well as demographic characteristics, allowing us to distinguish the effects of debt collection pressure from the effects of debt portfolios and other confounders. We capitalize on the panel nature of the data by including lagged measures of psychological distress and individual and year fixed effects to address concerns about reverse causality and unobserved heterogeneity. Our analyses ultimately contribute new insight on the relational mechanisms linking creditor–debtor power relations to mental well-being and inequality.

BACKGROUND

Research on the social determinants of mental health has long shown that socioeconomic and financial resources are determinants of psychological well-being (Pearlin 1989; Phelan, Link, and Tehranifar 2010; Turner and Lloyd 1999). A common focus in this literature is financial strain—a situation where one’s financial demands exceed available resources (Aneshensel and Avison 2015). Research in this vein has shown that debt is a source of financial strain (Berger, Collins, and Cuesta 2016; Drentea 2000; Drentea and Reynolds 2012, 2015; Dunn and Mirzaie 2016; Dwyer et al. 2016; Sun and Houle 2020; Sweet et al. 2013). Debt is thus increasingly recognized as a social determinant of mental health, joining income, wealth, knowledge, prestige, and other resources as fundamental causes of health disparities (Mendes de Leon and Griggs 2021), and scholars have conceptualized debt as a key element of socioeconomic status (Drentea 2000).

However, research on the socioeconomic determinants of mental health has been criticized for not paying sufficient attention to relational aspects of social location, including power differentials (Muntaner et al. 2015; Navarro 2009; Reynolds 2021). While the social determinants paradigm recognizes complex power relations as the source of unequal resources, socioeconomic status is often operationalized as a characteristic of individuals, in part due to data limitations (Fenwick and Tausig 2007; Muntaner 2013; Muntaner et al. 2015; Parbst and Wheaton 2023; Reynolds 2021; Sweet et al. 2018). Reynolds’ (2021) health power resources theory provides a framework for understanding how power operates on the macro-institutional and micro-interactional levels to shape mental and physical health (Montez, Hayward, and Zajacova 2019; Navarro et al. 2006; Reynolds and Brady 2012; Zheng et al. 2021). Reynolds (2021) argues that power shapes health through four distinct mechanisms. First, stratification shapes the distribution of resources and risks across social groups—determining who has access to socioeconomic resources or is exposed to financial risks, such as debt. The three remaining mechanisms identify interrelated pathways through which power relations affect the “necessity, value, and utility” of socioeconomic status for health (Reynolds 2021:496), including: (1) commodification (exposure to market forces with minimal state support or intervention), (2) devitalization (behavioral responses to exclusion and exploitation), and (3) discrimination (unequal exposure and impacts of risks across social groups).

We draw on health power resource theory to advance understandings of the relationship between indebtedness and mental health. Sociological theories of credit, debt, and inequality posit that creditor–debtor relationships are fundamentally structured by power (Dwyer 2018; Rona-Tas and Guseva 2018). For example, creditors have legal power to compel debtors to repay what they owe, with interest, and can financially extract payments through coercive means if debtors fail to make payments. Indeed, prior research shows that debt repayment problems—including persistently high unsecured debt balances, repayment stress, bankruptcy, and child support arrears—are associated with higher psychological distress (Addo 2017; Drentea, Zhu, and Guo 2021; Dunn and Mirzaie 2016; Gathergood 2012; Robbins et al. 2022; Sun and Houle 2020; Wolfe et al. 2022). And yet, most existing analyses of debt and mental health operationalize debt as an individual-level source of financial strain, rather than a stressor that is shaped by the power relationship between creditors and debtors.

Pressure from Debt Collectors as a Relational Stressor

We posit debt collection pressure as a relational stressor that may undermine the mental health of debtors. We define relational stressors as situations where relationships with a powerful actor causes stress for those who are less powerful. Because such stressors tend to disrupt the day-to-day activities of the less powerful, we view relational stressors as a chronic source of stress (Pearlin 1989). In the language of health power resources theory, debt collection pressure increases psychological distress because it can (1) make the repayment of debt more financially stressful for debtors (commodification) and (2) make indebtedness a more stressful and fearful experience by increasing exposure to unequal treatment and harassment, which can stigmatize debtors in ways that reduce their humanity. This treatment makes it more difficult for debtors to function in their social roles, creating more mental health risks (devitalization).

Debt collection pressures subject debtors to commodification mechanisms—legal factors that shape the “necessity of socioeconomic resources” (Reynolds 2021:496)—that expose them to unscrupulous market actors with minimal state intermediation or regulation. Collection pressure commodifies debtors because debt collectors can leverage legal power to demand repayment. Debt collectors commonly convey an urgent need for repayment, and they have broad authority from the state to pursue debtors through legal action (Sutton 1991). They often threaten to take legal action to persuade debtors to set up a repayment plan. Debt collectors who take legal action typically file against debtors in civil court (Hynes 2012). Civil court cases, which collectors almost always win and rarely provide debtors with legal recourse, create a legal channel for collectors to initiate wage garnishments and the seizure of assets (Fulford and Nagypal 2023; Hynes 2012; Raba 2023). This financial strain, as a chronic stressor, is likely to wear away at debtors’ mental health over time (Selenko and Batinic 2011).

Beyond financial strain, the aggressive tactics used by debt collectors also lead to mechanisms that devitalize—“deprive one of vitality, vigor, or effectiveness” (Reynolds 2021:498)—and in turn create chronic stress. Debt collectors often apply pressure by incessantly contacting debtors (Hunt 2007). The Federal Trade Commission confirms that collection agencies use aggressive tactics to target debtors, including “obscene language, racial and ethnic slurs [and] harassing phone calls at work” (Messac 2023:1623). Such harassment is itself a form of chronic stress that may lead to fear and distress (Houle et al. 2011). These aggressive debt collection tactics also implicate the borrower’s formal and informal social ties. Indeed, prohibited calls to workplaces and illegal communications with family, friends, and neighbors were among the most commonly reported debt collection violations in 2021 (Consumer Financial Protection Bureau 2022). Contact of coworkers might be particularly stressful, as debtors may fear the disclosing of potentially stigmatizing credit information to employers (Kiviat 2019; Maroto 2012). In a similar vein, repeated calls likely conflict with family and personal life. In short, debtors pressured by debt collectors likely feel constantly harassed, which might elevate psychological distress. Debt collectors’ frequent attempts at contacting family members and calls to the workplace likely also make it difficult for debtors to function in their social roles, for example by distracting them from their work or family responsibilities via repeated phone calls, which could in turn create role strain that further undermines mental health.

These interactions with debt collectors are also stigmatizing and demoralizing for debtors. Link and Phelan (2014) argue that stigma can be leveraged by powerful actors to make those with lower status feel cast out and down. Collection pressure can make debtors feel like they have failed to meet cultural ideals of economic self-reliance and other neoliberal standards of self-worth (Lamont 2018:424; Tach and Greene 2014). Indeed, Hochschild’s (2012:139) ethnography of debt collection agency employees revealed that collectors were commonly trained to “deflate the customer’s status … by hinting that the customer is lazy and of low moral character.” Debtors interviewed by Sweet and colleagues (2018) describe powerful feelings of shame and guilt for having accounts in collections, and debtors with feelings of personal failure regarding their debt are more likely to report poor mental and physical health. Interviewees with accounts in collections also describe feeling stigmatized by debt collectors (Sweet et al. 2018; Thorne and Anderson 2006).

In sum, we draw from prior research on indebtedness and mental health and health power resources theory to theorize that debt collection pressure is chronically stressful through the mechanisms of commodification and devitalization and hypothesize the following:

Hypothesis 1: Debt collection pressure is positively associated with psychological distress.

Moderation by Income and Racial-Ethnic Identity

Health power resources theory also points to the importance of discrimination—the systematic institutional and micro-level biases that expose minoritized groups to disproportionate insult and abuse—for shaping health (Reynolds 2021). This suggests that the mental health consequences of debt collection pressures might vary across income and racial-ethnic groups. Socioeconomic status and race shape power in financial markets (Dwyer 2018; Pattillo and Kirk 2021; Seamster 2019). Credit markets empower upper and middle class and white populations to make investments in wealth and status (Fligstein and Goldstein 2015; Seamster 2019). Low income and minoritized racial-ethnic groups, by contrast, more commonly face what Pattillo and Kirk (2021:890) call coercive financialization—an “externally imposed, involuntary, or last-resort entry into financial engagements.” Low-income groups are more likely to take on debt as a defense against economic insecurity (Fligstein and Goldstein 2015), Black and Hispanic populations tend to hold higher-cost unsecured debts associated with wealth extraction (Dwyer and DeMarco forthcoming; Seamster 2019), and Black people experience more stress and anxiety about their credit scores than white people (Norris 2022). Low-income and minoritized populations thus hold less power in financial markets than their advantaged counterparts, which might make collection pressure more stressful for the former groups.

Race- and class-based power differentials permeate interactions with debt collectors for at least three reasons. First, exclusion from mainstream financial services, disproportionate legal entanglements, disparities in wealth, and other factors contribute to inequalities in exposure to collection pressure. Consistent with this view, prior research finds that lower-income and Black and Hispanic populations are more likely to move from a delinquency to collections (Fulford and Nagypal 2023; Hynes 2012; LaVoice and Vamossy 2024; Raba 2024).

Second, debt collectors may discriminate in their targeting of debtors, applying more intense pressure on low-income and Black and Hispanic debtors. There is some evidence that collectors target lower-income populations, who they believe will be more intimidated by legal pressure (Holland 2011), and Hochschild (2012) documents prejudicial views among some of the debt collectors she interviews.²

Finally, high income and white debtors have more access to various forms of support including legal support, this serving as a buffer against collection abuses and associated devitalization. Lower-income defendants are more likely to receive a collections judgement in civil court (Hynes 2012), and Black and Hispanic defendants are less likely to have legal representation (Fulford and Nagypal 2023; Raba 2023).

Taken together, the above literature suggests that collection pressure is more burdensome for the mental health of lower-income and non-white debtors because these groups are more likely to be exposed to collection pressure for longer durations with potentially more intense pressure and less access to legal services that might protect debtors from collection abuses.

Hypothesis 2a: Low-income young adults face more exposure to debt collection pressure in young adulthood than high-income young adults.

Hypothesis 2b: Black and Hispanic young adults face more exposure to debt collection pressure in young adulthood than white young adults.

Hypothesis 3a: Debt collection pressure has a stronger association with psychological distress among low-income young adults relative to high-income young adults.

Hypothesis 3b: Debt collection pressure has a stronger association with psychological distress among Black and Hispanic young adults relative to white young adults.

DATA AND METHODS

Data

Data for this study came from the National Longitudinal Survey of Youth-1997 Cohort (NSLY-97). The NLSY-97 is a national longitudinal survey of 8,984 U.S. young adults born between 1980 and 1984. The NLSY-97 has response rates above 80% and includes an oversample of Black and Hispanic young adults (Bureau of Labor Statistics 2005). The initial survey wave was in 1997, with annual follow-ups surveys from 1998 to 2011 and biannual follow-ups from 2013 to 2019. The NLSY-97 was ideal for this study as it contains a unique measure of debt collection pressure, repeated measures of psychological distress, and comprehensive data on debts, assets, income, and demographic characteristics. The NLSY-97 includes the oldest members of the Millennial birth cohort, who came of age during an unsecured credit expansion in the 2000s and subsequent financial crisis and credit contraction during the 2010s following the Great Recession (Dwyer et al. 2011, 2016; Houle 2014).

We made two restrictions to the NLSY-97 sample. Because measures of debt collection pressure were only available from 2007 to 2019, we first restricted the sample to respondents who participated in at least one survey wave between 2007 and 2019 (N = 8,355 respondents). Second, we restricted the sample to respondents with valid data on all study variables other than income, assets, and debts (N = 7,236 respondents). We reshaped the data into panel form, so person-waves are the unit of observation (N = 37,033 person-waves).

Missing data was minimal on most study variables (below 5%), except for family income (due to non-response) and debts and assets, which were asked about approximately every five years. We used linear interpolation to impute values of family income, debts, and assets between survey waves using the “mipolate” command in Stata 16.³

Measures

Psychological distress.

Our primary outcome measure of psychological distress captured depressive symptoms, using responses to the Mental Health Inventory-5 (MHI-5) (Veit and Ware 1983), which are available in the 2000, 2002, 2004, 2006, 2008, 2010, 2015, 2017, and 2019 survey waves. The MHI-5 has been validated as a measure of anxiety and mood disorders in several populations (Cuijpers et al. 2009; Marques, Pais-Ribeiro, and Lopez 2011). The MHI-5 asks respondents how often in the past month they felt blue, sad, happy, peaceful, or nervous (none of the time, some of the time, most of the time, or all the time). We reverse coded the happy and peaceful items and summed all five responses to generate an additive scale ranging from 8 to 23 that captures the extent of depressive symptoms (Chronbach’s α = 0.792). Following previous research (Ware et al. 1993), we then transformed this to a uniform scale ranging from 1 to 100, with higher scores on the scale corresponding to higher levels of psychological distress.⁴

Debt collection pressure.

The NLSY-97 asked respondents whether in the past 12 months they or their spouse/partner “were pressured to pay bills by stores, creditors, or bill collectors.” We used this measure to construct a dichotomous indicator of debt collection pressure (1 = yes, 0 = no). In our sample, 13.2% reported pressure from debt collectors in a given survey wave.

Because our measure of debt collection pressure has, to our knowledge, not been tested in the literature, we took additional steps to validate it and ensure that it reflects financial difficulties. We estimated an ordinal logistic regressing a five-category measure of current self-reported financial difficulty (1 = very comfortable and secure, 2 = able to make ends meet without much difficulty, 3 = occasionally have some difficulty making ends meet, 4 = tough to make ends meet but keeping your head above water, and 5 = in over your head) on debt collection pressures. Respondents who reported debt collection pressure reported significantly higher financial difficulties than those who did not (see Figure 1). This suggested that debt collection pressure was more common among individuals who were behind on their bills or were otherwise facing financial difficulties, consistent with our expectations.

Figure 1: Predicted Probabilities of Financial Difficulty by Debt Collection Pressure.

Note: Source: 2008–2019 waves of the NLSY-97. Estimates from ordinal logistic regression models predicting current self-reported financial difficulty by debt collection pressure in the last 12 months. Brackets represent 95% confidence intervals.

Control variables.

We controlled for several characteristics that may confound the association between debt collection pressure and psychological distress. Sociodemographic characteristics and life events, including age and family formation/dissolution may influence the relationship between debt repayment and distress (Addo 2017; Dwyer et al. 2016; Sun and Houle 2020; Sweet et al. 2013); thus, we included controls for age, age-squared, education (1 = four-year college degrees), marital status (Categories: Married/partnered, separated/divorced/widowed, never married [reference]); region (Categories: Northeast, North Central [reference], South, and West); urban residence (1 = yes); presence of a work-limiting health condition (1 = yes); and the respondent’s highest ever degree of criminal justice contact (Categories: No contact [reference], arrest but no conviction, conviction but no incarceration, and incarceration). We included several controls for socioeconomic characteristics to ensure that the associations of collection pressure and distress were distinct from the effects of other socioeconomic factors. Socioeconomic controls included the respondent’s family income quartile position, the number of weeks unemployed in the year of the interview, an indicator for homeownership (1 = yes), and an indicator for being late on mortgage or rent in the last 12 months (1 = yes). Additional financial controls included continuous measures of housing debt, vehicle debt, educational debt, credit card debt, other debts to businesses, and overall net worth (total debts minus total assets), inflated to constant 2017 dollars using the Consumer Price Index-All Urban Consumers series. Finally, we included controls for the calendar year (year fixed effects) to account for any remaining period effects.

Analytic Strategy

Our analytic strategy proceeded in three steps. First, we examined trends in exposure to debt collection pressure among the NLSY-97 cohort and documented descriptive differences in average psychological distress symptoms by whether the respondent reported any pressure from debt collectors in the past 12 months, with t-tests for statistical significance in the bivariate analyses. We weighted all trend and descriptive analyses using the NLSY-97 custom weights to adjust for the Black and Hispanic oversample and attrition out of the survey over time.

Second, we tested Hypothesis 1 by estimating linear regression models predicting psychological distress symptoms as a function of debt collection pressure and other factors. We built complexity into the models in a stepwise fashion, controlling for more potential confounders with each step. Our most complex specification is written as follows:

$$Y_{it}={\beta }_0+{\beta }_{1}Pressur{e}_{it}+{\mathit{D}}_{it}^{\prime }+{\mathit{S}}_{it}^{\prime }+{\mathit{F}}_{it}^{\prime }+Y_{it-1}+{\gamma }_t+{\mu }_i+{\epsilon }_{it}$$ (1)

where ${\beta }_1$ captures the association between debt collection pressure and psychological distress, $\mathit{D}$ is a vector of time-varying demographic characteristics, $\mathit{S}$ is a vector of time-varying socioeconomic characteristics, and $\mathit{F}$ is a vector of time-varying financial controls for assets and debts. We included calendar year dummies $\left({\gamma }_t\right)$ and individual fixed effects $\left({\mu }_i\right)$ to control for all time-invariant characteristics of individuals (Halaby 2004; Rabe-Hesketh and Skrondal 2012). Fixed-effects models helped address unobserved selection into debt collection pressure due to personality characteristics and other psychological traits that are stable across time (Letkiewicz and Heckman 2019) and allowed us to examine how within-person changes in debt collection pressures are associated with within-person changes in distress.⁵ ${\beta }_1$ was potentially vulnerable to reverse causality, wherein prior psychological distress affects both current distress and exposure to debt collection pressure (Addo 2017; Finnigan and Meagher 2019; Gathergood 2012; McCloud and Dwyer 2011; Sullivan et al. 2000; Sun and Houle 2020).⁶ We therefore included a one-wave lagged dependent variable to account for potential reverse causality.⁷ Including a lagged dependent variable also helped account for time-varying confounders, such as event stressors, that might be linked to both collection pressure and psychological distress. All models were estimated with robust standard errors clustered on the respondent identifier.

To test Hypotheses 2a to 2b we examined income- and race-specific differences in exposure to debt collection pressure descriptively by estimating the share pressured by debt collectors in the last 12 months and the cumulative share ever pressured since 2007 by income (family income quartile: p1 to p25, p25 to p50, p50 to p75, and p75 to p99) and race-ethnicity (non-Hispanic Black, Hispanic [of any race], and non-Hispanic white). To test Hypotheses 3a to 3b, we re-estimated Equation (1) on samples stratified by income and racial-ethnic group. Stratified sample analyses were desirable for our purposes because they allowed us to assess heterogeneity by time-invariant characteristics (such as race) while using fixed-effects regression. We implemented z-tests to assess statistical differences in the coefficients across models stratified by family income and race.

RESULTS

Descriptive Results

Figure 2 presents trends in exposure to debt collection pressure among the NLSY-97 Cohort from 2007 to 2019. The solid line shows that the share of respondents reporting collection pressure in the last 12 months ranged from 10% to 15% of the sample in each wave, with a slight rise in 2009, likely reflecting the Great Recession, and a gradual decline thereafter, which may reflect a combination of age- and period-related dynamics. These estimates are similar to previous single-year estimates of the share of young adults with at least one account in collections (Martinchek et al. 2022). However, a focus on current exposure to debt collection pressure underestimates cumulative exposure in young adulthood. The dashed line shows the share of the NLSY-97 Cohort that has ever experienced debt collection pressure since the question was first asked of respondents in 2007. We estimate that by around age 40, approximately 38% of this cohort ever experienced debt collection pressure.

Figure 2: Trends in Exposure to Debt Collection Pressure, 2007–2019.

Note: Source: 2007–2019 waves of the NLSY-97. Trends are weighted using the NLSY-97 Custom Weights.

Table 1 presents weighted descriptive statistics by debt collection pressure in the past 12 months. Consistent with Hypothesis 1, those who reported being pressured by debt collectors also reported significantly higher psychological distress symptoms on average than those who did not report being pressured by debtor collectors in the past 12 months. However, Table 1 also suggests that those who face debt collection pressure face several social and economic disadvantages and differ in important ways from those who do not. We therefore turn to multivariable methods to provide a more rigorous test of Hypothesis 1.

Table 1:

Weighted Descriptive Statistics

	Full Sample		Debt Collection Pressure	No Debt Collection Pressure
	Mean	SD	Mean	Mean	t-testa
Mental Health Outcome
Psychological Distress Symptoms in Past Month (1–100)	41.52	29.52	56.61	39.23	***
Debt Collection Pressure
Debt Collection Pressure in Last 12 Months (0,1)	.13		1.00	.00
Time-Varying Demographic Controls
Age	30.67	3.66	30.17	30.75	***
Four-Year College Degree (0,1)	.31		.18	.33	***
Marital Status: Never Married (0,1)	.35		.37	.34	***
Marital Status: Married or Partnered (0,1)	.57		.52	.58	***
Marital Status: Separated/Divorced/Widowed (0,1)	.08		.11	.08	***
Presence of Biological Child (0,1)	.49		.55	.49	***
Census Region: Northeast (0,1)	.16		.17	.15	**
Census Region: North Central (0,1)	.24		.26	.23	**
Census Region: South (0,1)	.38		.36	.39	*
Census Region: West (0,1)	.22		.21	.23	**
Urban Residence (0,1)	.95		.94	.95	*
Work Limiting Health Condition (0,1)	.07		.14	.06	***
Highest Criminal Justice Contact: None (0,1)	.65		.57	.67	***
Highest Criminal Justice Contact: Arrest (0,1)	.14		.17	.14	***
Highest Criminal Justice Contact: Conviction (0,1)	.11		.14	.10	***
Highest Criminal Justice Contact: Incarceration (0,1)	.10		.12	.09	**
Time-Varying Socioeconomic Characteristics
Low Family Income (Quartile: p1–p25)	.22		.27	.21	***
Low-Mid. Family Income (Quartile: p25–p50)	.25		.30	.24	***
High-Mid. Family Income (Quartile: p50–p75)	.26		.25	.26
High Family Income (Quartile: p75–p99)	.28		.18	.29	***
Weeks Unemployed in Year of Interview	2.61	8.33	4.56	2.31	***
Homeownership (0,1)	.33		.18	.35	***
Late Mortgage/Rent in Last 12 Months (0,1)	.02		.10	.01	***
Time-Varying Financial Characteristics
Mortgage Debt (in US $1s)	45,377	80,277	17,753	49,584	***
Vehicle Debt (in US $1s)	6,205	10,669	4,415	6,477	***
Education Debt (in US $1s)	8,236	21,727	10,020	7,965	***
Credit Card Debt (in US $1s)	2,012	5,286	2,883	1,880	***
Other Debt to Businesses (in US $1s)	2,243	16,081	6,256	1,631	***
Net Worth (Assets Less Debts; in US $1s)	70,626	140,285	11,870	79,576	***
Time-Invariant Characteristics Absorbed by the Fixed Effects
Gender: Woman (0,1)	.49		.58	.48	***
Race: Non-Hispanic Black (0,1)	.16		.20	.16	***
Race: Hispanic (Any Race) (0,1)	.13		.11	.13	***
Race: Non-Hispanic Other Race (0,1)	.05		.05	.05
Race: Non-Hispanic White (0,1)	.66		.65	.66
N (person-wave obs.) =	21,041		2,808	18,233

Note: Source: 2008, 2010, 2015, 2017, and 2019 waves of the NLSY-97. SES = Socioeconomic Status.

^aT-tests for statistical difference in means between those who reported any debt collection pressure and those who did not report any debt collection pressure in the past 12 months.

⁺p < .1,

^*p < .05,

^**p < .01,

^***p < .001 (two-tailed)

Psychological Distress Regression Results

Table 2 presents results from the linear regression models predicting psychological distress symptoms in the past month (see Online Supplemental Appendix Table S1 for full model results). Column 1 is a reduced-form model that includes controls for race-ethnicity and gender, time-varying demographic controls, time-varying socioeconomic controls, and period fixed effects. Consistent with Hypothesis 1, results in Column 1 suggest that those who experienced debt collection pressure had psychological distress symptom scores that were 11.7 points higher on average than those who did not report collection pressure (p < 0.001). These findings are robust to controls for time-varying financial characteristics (Column 2).

Table 2:

Linear Regression Models Predicting Psychological Distress Symptoms in Past Month

	(1)	(2)	(3)	(4)
	Psychological Distress	Psychological Distress	Psychological Distress	Psychological Distress
	Coef. (SE)	Coef. (SE)	Coef. (SE)	Coef. (SE)
Debt Collection Pressure in Past 12 Months (0,1)	11.734*** (.664)	11.047*** (.666)	4.168*** (.648)	4.172*** (.642)
Time-Varying Demographic Controls	Yes	Yes	Yes	Yes
Time-Varying Socioeconomic Controls	Yes	Yes	Yes	Yes
Period Fixed Effects	Yes	Yes	Yes	Yes
Time-Varying Financial Controls	No	Yes	Yes	Yes
Individual Fixed Effects	No	No	Yes	Yes
Lagged Dependent Variable	No	No	No	Yes
Constant	30.132 (19.490)	32.436+ (19.536)	94.348*** (26.567)	105.179*** (26.925)
N (person-wave observations) =	21,041	21,041	21,041	21,041

Note: Source: 2008, 2010, 2015, 2017, and 2019 waves of the NLSY-97. Robust standard errors clustered on the respondent ID in parentheses. SE = Standard Error.

⁺p < .1,

^*p < .05,

^**p < .01,

^***p < .001 (two-tailed)

Column 3 adds individual fixed effects. Note that including fixed effects shifts the interpretation to the within-individual change in distress when one does versus does not experience collection pressure (Halaby 2004). Unsurprisingly, including individual fixed effects reduces the size of the debt collection pressure coefficient (b = 4.168, p < 0.001). This is consistent with arguments that selection on time-invariant characteristics plays a role in the associations of collection pressure and distress (Gathergood 2012; Letkiewicz and Heckman 2019). However, collection pressure continues to have a significant and positive relation to distress symptoms, such that within-person exposures to debt collection pressure are associated with a 4.2-point increase in distress. This suggests that the observed effects of collection pressure on distress are not entirely explained by selection into collection pressure on time-invariant characteristics.

Finally, Column 4 adds a control for lagged psychological distress. After controlling for prior distress, the coefficient remains significant and is virtually unchanged from Column 3, suggesting that the observed relationships are not driven by psychological distress in the prior period (Gathergood 2012).^8,9

To help interpret the substantive magnitude of the estimates in Table 2, consider the association between debt collection pressure and distress for the median person in our sample, defined by median values of the sociodemographic variables. This person is a low-income white woman aged 31 with less than a college degree who is currently married or partnered, is living with a biological child, is a renter in an urban area in the South and has $6,415 in total net worth. The baseline predicted psychological distress score for this person is 39.5. Exposure to debt collection pressure is expected to increase psychological distress for this person to 43.7—an increase of 4.2 points or 10.6% from baseline. This translates to one third of a within-person standard deviation in distress scores, is equal to the protective effect of marriage or partnership, and is larger than the distressing effects of one year of unemployment.

Heterogeneity by Income and Racial-Ethnic Group

Figure 3 presents income- and race-specific estimates of exposure to collection pressure from 2007–2019. Consistent with Hypotheses 2a to 2b, low-income and Black young adults are more likely to have been pressured by a debt collector in the last 12 months relative to their high-income and white counterparts, with even larger disparities in cumulative exposure to collection pressure. 55% of low-income young adults and 49% of Black young adults experienced debt collection pressure by around age 40, compared to 28% of high-income young adults, 38% of Hispanic young adults, and 37% of white young adults. These inequalities in collection pressure highlight the specific disadvantages of low-income and Black young adults in the U.S. financial system.

Figure 3: Trends in Exposure to Debt Collection Pressure by Family Income Quartile and Racial-Ethnic Group, 2007–2019.

Note: Source: 2007–2019 waves of the NLSY-97. Trends are weighted using the NLSY-97 Custom Weights.

Does the association between debt collection pressure and distress vary by income and race? Table 3 presents results of the stratified sample analyses testing for heterogeneity by income (Hypothesis 3a) and race-ethnicity (Hypothesis 3b). All models in Table 3 include all time-varying controls, individual fixed effects and lagged dependent variables. Columns 1 to 4 display the coefficients for debt collection pressure by income group. The association between debt collection pressure and psychological distress is stronger for lower-income respondents than for higher-income respondents. For those in the lowest-income group, collection pressure is associated with a 6.3-point higher distress score (p < 0.001). By contrast, collection pressure is not significantly associated with distress for the highest-income group (b = 0.896, p = 0.644). A z-test of coefficients reveals that these effects are significantly different (z = 2.208, p = 0.027). Z-tests for the remaining income contrasts reveal no statistically significant differences. These results are consistent with Hypothesis 3a, as low-income young adults experience a stronger distress response to debt collection pressure than high-income young adults.

Table 3:

Linear Regression Models Predicting Psychological Distress in Past Month Stratified by Income and Racial-Ethnic Group

Subsample:	(1)	(2)	(3)	(4)	(5)	(6)	(7)
	Psychological distress	Psychological distress	Psychological distress	Psychological distress	Psychological distress	Psychological distress	Psychological distress
	Income quartile: p1–p25	Income quartile: p26–p50	Income quartile: p51–p75	Income quartile: p76–p99	Race-Ethnicity: Black	Race-Ethnicity: Hispanic	Race-Ethnicity: White
	Coef. (SE)	Coef. (SE)	Coef. (SE)	Coef. (SE)	Coef. (SE)	Coef. (SE)	Coef. (SE)
Debt collection pressure in last 12 months (0,1)	6.255*** (1.458)	4.354* (1.725)	3.146+ (1.763)	0.896 (1.940)	3.907*** (1.156)	2.769+ (1.444)	4.818*** (0.945)
Time-varying demographic controls	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Time-varying socioeconomic controls	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Period fixed effects	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Time-varying financial controls	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Individual fixed effects	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Lagged dependent variable	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Constant	126.368+ (70.547)	206.828** (79.420)	60.464 (73.741)	17.732 (56.480)	158.404** (54.541)	56.593 (59.005)	118.095** (37.342)
N (person-wave observations) =	5,302	5,251	5,257	5,231	5,730	4,486	10,064

Note: Source: 2008, 2010, 2015, 2017, and 2019 waves of the NLSY-97. Robust standard errors clustered on the respondent ID in parentheses. SE = Standard Error.

⁺p < .1,

^*p < .05,

^**p < .01,

^***p < .001 (two-tailed)

Columns 5 to 7 present results stratified by racial-ethnic group. These results show less heterogeneity than the income results. For each of the racial-ethnic groups examined, debt collection pressure is positively related to psychological distress symptoms, and there are no significant differences in the size of the effects for Black, Hispanic, and white young adults. This result is inconsistent with Hypothesis 3b; there is little evidence that Black and Hispanic young adults experience a stronger effect of debt collection pressure on psychological distress.¹⁰ However, these results need to be interpreted in the context of our finding of significantly higher exposure to debt collection pressure among Black young adults relative to their white and Hispanic counterparts (Figure 3). Even without an interaction effect, given higher exposure, Black young adults face a disparate impact of the increased psychological distress associated with debt collection pressure. Black young adults are also more likely to be low income and thus the added penalty of psychological distress for lower income young adults who face debt collection pressure also falls most heavily on Black young adults.¹¹

Alternative Specifications and Robustness Checks

We estimated additional alternative specifications and conducted several robustness checks to probe the sensitivity of our results to alternative measurement and modeling choices. First, following Porter and DeMarco (2019), we tested an alternative dichotomous outcome capturing major depression, measured as having a 64 or higher score on the psychological distress scale.¹² In linear probability models with individual fixed effects and a lagged dependent variable, debt collection pressure is associated with a 5.3 percentage-point increase in depression (Online Supplemental Appendix Table S2), and all of the results from these models are consistent with the results above.

Second, we tested whether debt collection pressure is more consequential for anxiety symptoms than depression symptoms, as literature on financial strain suggests (Drentea and Reynolds 2015). Following Hodson, Dwyer, and Neilson (2014), we created an anxiety scale using the peaceful and nervous items of the MHI-5 (Range: 3 to 9; Chronbach’s α = 0.591), and a depression scale using the blue, sad, and happy items (Range: 5 to 14; Chronbach’s α = 0.729). We transformed both measures into uniform scales ranging from 1 to 100. In linear models with individual fixed effects and a lagged dependent variable, we find that collection pressure is associated with a 3.3 point within-person increase in the anxiety symptoms scale (Online Supplemental Appendix Table S3), and a 4.2 point within-person increase in the depression symptoms scale (Online Supplemental Appendix Table S4). This suggests that the consequences of collection pressure for psychological distress are similar for both anxiety and depression symptoms.

Might our results be confounded by the effects of bankruptcy and financial strain? Many who face debt collection pressure subsequently file for bankruptcy, and prior studies find that bankruptcy is associated with psychological distress (Addo 2017; Maroto 2012; Wolfe et al. 2022). We also know that debt repayment is stressful in part because it produces financial strain (Drentea and Reynolds 2015). While we view bankruptcy and financial strain as potential mechanisms linking debt collection pressure to psychological distress, we acknowledge that these experiences might also be confounders. We tested models with controls for having ever experienced bankruptcy and current self-reported financial difficulty. While we find quite similar results when we control for bankruptcy, we find evidence that self-reported financial difficulty partially mediates (or confounds) the effects of collection pressure (Online Supplemental Appendix Tables S5 and S6). However, debt collection pressure remains positive and significant in these models, suggesting remaining pathways for negative interactions, role strain, and stigma.

Finally, we tested whether cumulative exposure to debt collection pressure is associated with psychological distress around age 35. We estimated a cross-sectional model predicting distress in 2017, the latest survey wave in which all respondents received the MHI-5, as a function of the total number of waves the respondent experienced debt collection pressure from 2007 to 2017, with a control for baseline psychological distress in 2006. We found that respondents experiencing higher cumulative exposure to collection pressure also reported significantly higher psychological distress in 2017 (Online Supplemental Appendix Figure S1). Moreover, the cumulative effects of exposure to collection pressure from 2007 to 2017 are larger than the effect of point-in-time exposure to collection pressure, as presented in Table 2, Column 1. This is consistent with our theory that debt collection pressure operates as a chronic relational stressor.

DISCUSSION

Recent cohorts of young adults are facing an unprecedented mental health crisis and many have taken on debt to manage the financially intensive transition to adulthood. Prior research on debt and mental health mainly treats credit and debt as a resource and liability, rather than as a power relationship between creditors and debtors. This article contributes a test of theories that power dynamics shape mental health by focusing on the case of debt collection pressure and young adult psychological distress. Leveraging a unique measure in the NLSY-97, we find that exposure to debt collection pressure is both prevalent and highly stratified across income and racial-ethnic groups in this U.S. cohort. Panel regression analyses, including individual fixed effects and lagged dependent variable models, suggest that a consequence of experiencing debt collection pressure in young adulthood is higher psychological distress. In stratified sample regressions, we find that the consequences of collection pressure for psychological distress are especially pronounced among low-income young adults.

The current study makes several contributions. Our core finding that collection pressure is associated with psychological distress supports theories that pressure from debt collectors helps explain why debt repayment problems are distressing (Drentea 2000; Drentea and Lavrakas 2000; Sweet et al. 2018). Debt collectors rely on discursive and legal exercises of power to encourage or force debtors with past-due accounts to repay their debts. Drawing on stress theory and health power resources theory, we characterize collection pressure as a relational stressor that increases financial strain (commodification) and exposes debtors to aggressive actions and targeted harassment by collection agencies, leading to role strain and stigma (devitalization), consequently eroding debtors’ mental well-being. We contribute to the literatures on credit, debt, and mental health by demonstrating the utility of health power resources theory for understanding the mechanisms linking unequal creditor–debtor relations to mental health.

That our findings are most stark for low-income young adults is consistent with arguments that power relations are particularly consequential for disadvantaged groups (Montez et al. 2019; Parbst and Wheaton 2023). Low-income young adults are not only more likely to be exposed to debt collection pressure—they also experience a greater distress penalty from collection pressure. We argue that class-based inequalities in the mental health effects of collection pressure reflect wider creditor–debtor power imbalances among the disadvantaged.

While the mental health consequences of collection pressure are stratified by income, we find no evidence that its distressing effects vary by racial-ethnic group. This null finding might reflect several factors: First, it could be that debt collectors apply similar pressure to Black, white, and Hispanic young adults, net of other characteristics. Thus, conditional on exposure, debt collection pressure has similar psychological consequences across racial-ethnic groups. Second, there is a long literature in the sociology of mental health which suggests that racialized minorities, and especially Black people, have access to an array of coping resources that can protect mental health in the face of chronic racialized stressors (e.g., Louie et al. 2021). Thus, even if debt collection pressure is more distressing for Black young adults, access to coping resources might mitigate some of the negative consequences, potentially explaining why we do not find moderation by race. Third, it could be that race and class intersect to make debt collection pressure most distressing for social groups who hold the least amount of power in the financial system. Indeed, in supplemental analyses testing moderation by race and income, we find evidence that debt collection pressures are most distressing for low-income Black young adults. Nevertheless, our analysis of exposure shows that Black young adults are more likely to be exposed to collection pressure than white and Hispanic young adults. Collection pressure thus places a greater burden on the well-being of racialized populations. Our findings contribute to knowledge of how race and class interact with power to shape mental health by documenting the role of collection pressure in unequal creditor–debtor relationships.

We also make a methodological contribution by leveraging the panel nature of the NLSY-97. Prior research on debt and health has primarily relied on cross-sectional data, or between-person comparisons. Studies of credit, debt, and health are potentially vulnerable to issues of unobserved selection and reverse causality, as health problems often precipitate financial distress (Finnigan and Meagher 2019; Gathergood 2012; McCloud and Dwyer 2011; Sullivan et al. 2000). We address such concerns by using individual fixed effects, to account for unobserved selection into collection pressure on time-invariant characteristics, and by estimating lagged dependent variable models, to account for reverse causality.

It will be valuable for future research to build on the limitations of this study. First, while the NLSY-97 contains a unique measure of debt collection pressure, measures of the mechanisms linking collection pressure to psychological distress are limited. Thus, we are unable to fully test the pathways through which collection pressure increases psychological distress. Second, it is possible that the amount and type of debt in collections moderates the associations of collection pressure and psychological distress. Debt collectors may apply more pressure on debtors who are late on larger balances or those who are late on unsecured credit accounts or past-due bills for which there is no underlying asset to seize. There may also be differences depending on whether the debt collector works for a first- or third-party collections firm or for the state. We are unable to test for such heterogeneities because the NLSY-97 lacks measures of the amount and type of debts in collections. Third, because our analyses focus on U.S. Millennial young adults, we caution that our findings may not generalize to other cohorts, countries, or life course periods. Fourth, the NLSY-97 includes relatively small samples of low-income and Black and Hispanic young adults. It will thus be important for future work to further test for heterogeneity in the mental health consequences of debt collection pressure using larger samples. Finally, because we rely on observational data, our analyses do not account for all potential confounders. However, the consistency of results across models with fixed effects, lagged dependent variables, several alternative specifications, and a battery of sensitivity checks suggest that it is unlikely that our core conclusions are unduly influenced by spurious associations.

In sum, our study suggests that, for debtors, there are psychological costs to the power imbalances that have become common in creditor–debtor relationships. Yet there are conditions under which debtors might gain some power back from creditors. Such conditions can arise from state regulations (Rona-Tas and Guseva 2018). For example, the U.S. federal government regulates the practices of debt collectors under the Fair Debt Collection Practices Act (FDCA, passed in 1977). While the FDCPA provides only minimal restrictions, many states have their own laws governing debt collection and bankruptcy, with 22 states having changed their collection laws since the 1990s (Fedaseyeu 2020; Martin 2022). This suggests that U.S. state policy contexts might moderate the link between collection pressure and mental health, consistent with the commodification concept in health power resources theory (Reynolds 2021). Future research could leverage this institutional variation to better capture power differences between debtors and creditors and test whether policies that constrain the power of debt collectors mitigate the impact of debt collection on the well-being of debtors. As greater attention is given to mental health challenges among young people in the United States, our findings suggest that reducing financial pressures may be a key strategy for enhancing well-being during this crucial life course stage, along with supporting equitable inclusion in financial markets.

Biographies

Alec P. Rhodes is a postdoctoral scholar at the Institute for Research on Poverty at the University of Wisconsin–Madison. He is interested in how power and policy shapes inequality in the United States, with a focus on credit, wealth, debt, labor markets, and health disparities. Ongoing projects examine the unequal returns to worker power across populations and places and the relationships between state safety net policies and high-cost credit use among low-income populations. He has published in Nature Human Behaviour, Social Problems, The Sociological Quarterly, Social Science & Medicine, and Rural Sociology among other outlets.

Rachel E. Dwyer is professor in the Department of Sociology and affiliate of the Institute for Population Research at The Ohio State University. She studies social inequality and economic insecurity in the United States, with a particular focus on credit, debt, and inequality, and growing attention to health disparities. Current work examines debt and well-being among low-income populations. She has published in the American Sociological Review, Social Forces, Social Problems, and Nature Human Behaviour among other outlets

Jason N. Houle is professor in the Department of Sociology at Dartmouth College in Hanover, NH. His research primarily focuses on the social determinants of health and mental health across the life course, and processes of social stratification and mobility. His current work examines the link between rising household indebtedness and population well-being.