The Effects of Severe Childhood Physical and Sexual Abuse on Adult Socioeconomic Prosperity

Abstract

Our study utilizes Adverse Childhood Experience (ACE) scores to estimate the relationship between forced sexual intercourse and physical abuse on socioeconomic outcomes in adulthood. ACEs have been shown to have long-term negative impacts on health, mental health, and cognition. We expand upon the literature that analyzes the effects of ACEs on human capital investment and adult socioeconomic outcomes by focusing on the ACE scores pertaining to repeated physical and forced sexual abuse in childhood. Specifically, we estimate probit models using data from the Behavioral Risk Factor Surveillance System to measure the marginal effects of childhood sexual abuse (CSA) and physical abuse on the probability of high school completion, unemployment, and the likelihood of living in poverty in adulthood. We find adults who suffered physical abuse in childhood are more likely to live in poverty. Adult survivors of CSA are less likely to finish high school and more likely to live in poverty. The likelihood of high school noncompletion increases when the individual suffered both forms of abuse. We also find that only those who suffered both forms of abuse in childhood had a greater likelihood of being unemployed and high school noncompletion. We find the negative socioeconomic impact in adulthood is larger for women than for men, implying gender heterogeneity in outcomes of CSA and physical abuse. Researchers should control for the correlation between sexual abuse and physical abuse in childhood, particularly in women, when estimating their effects on socioeconomic outcomes.

Adverse experiences in childhood can have lasting impacts that continue into adulthood. This is particularly true for child abuse. Child abuse is a societal issue that is prevalent in nations worldwide. According to the U.S. Department of Health and Human Services (2022), over 650,000 children are victims of child abuse and neglect in the United States each year. The World Health Organization (WHO) estimates child maltreatment affects 55 million in the European region, with a 9.6% prevalence rate for sexual abuse and 22.9% for physical abuse (Sethi et al., 2021).

Recent studies highlight the effects of childhood sexual abuse (CSA) and physical abuse on adult economic outcomes such as educational attainment, labor force participation, employment, and income. Understanding the relationship between childhood abuse and adult socioeconomic outcomes is important because child abuse can have direct and indirect costs to society and the economy.

Childhood abuse and neglect tend to be more prevalent in lower socioeconomic status families (Hafton et al., 2017; Matta Oshima et al., 2014; Shaefer et al., 2018). Crouch et al. (2019) show that lower-income households have higher prevalence rates of ACEs. Moreover, children from higher-income households are less likely to experience physical abuse (Shaefer et al., 2018). Eckenrode et al. (2014) show a positive relationship between child maltreatment, child poverty rate, and income inequality.

While most of the research in this area observes health impacts, a growing body of literature investigates the impacts adverse childhood experiences (ACEs) have on socioeconomic prosperity. In this paper, we use the ACE questionnaire to examine the impact of CSA and physical abuse on high school noncompletion, unemployment, and poverty rates using the BRFSS from 2009–2012. We define CSA and physical abuse as having experienced forced sexual intercourse more than once before age 18 and being physically abused by a parent or an adult more than once before age 18, respectively.

Those who experience forced sexual intercourse are likely to also experience physical abuse (Dong et al., 2004). Studies that address the human capital effect of CSA should note its potential upward bias on its effect on human capital because CSA tends to be correlated with physical abuse. We correct for this by grouping victims of child abuse into three categories: those that only experienced physical abuse with no sexual abuse; those that experienced severe sexual abuse but did not experience physical abuse in their youth; and those that experienced both physical abuse and sexual abuse in their youth. By isolating these three categories of individuals, we can measure the impacts each form of abuse has on future outcomes. This method allows us to determine which forms of abuse have the greatest negative impact in adulthood. Understanding these connections is vital for ending the cycle of violence and improving socioeconomic outcomes for future generations.

The purpose of this study is to contribute to the literature in two ways. First, we disaggregate physical and sexual abuse to estimate each form of abuse separately on adult economic outcomes. We focus on repeated physical and forced sexual abuse by estimating economic outcomes for individuals that experienced these forms of abuse in their childhoods. Instead of estimating the effects of the number of ACEs an individual reports, we use the Behavioral Risk Factor Surveillance System (BRFSS) survey to isolate individuals who indicated having experienced forced sexual intercourse and physical abuse either separately or concurrently in childhood. We identify the severity of the abuse by isolating those that experienced physical or severe sexual abuse more than once during their childhoods. Diette et al. (2017) estimate severe child abuse on high school completion and educational attainment. Building upon this work, we estimate the likelihood of being unemployed and living below the Federal Poverty Level (FPL) based on our three measures of abuse. Our dataset includes four years of BRFSS surveys, allowing us to aggregate the data by including thirteen states’ survey responses for the years in which each state included the optional ACE module in the survey data. To our knowledge, this has not been done in previous work.

Long-term Impacts of Child Abuse

Numerous studies highlight the link between the severity of the trauma associated with childhood physical and sexual abuse to mental health and socioeconomic outcomes in adulthood (Almquist, 2015; Almquist & Brännström, 2018; Anda et al., 2006; Goodman et al., 2011; Maniglio, 2009; Rees & Sabia, 2013; Robst & Smith, 2008). However, when controlling for family-level heterogeneity, the severity of CSA and physical abuse on these negative outcomes in adulthood dampens (Boden et al., 2007; Tanaka et al., 2015).

Sufferers of childhood maltreatment are often more likely to report depression symptoms in adulthood and poorer cognition (Goltermann et al., 2021; Zheng et al., 2022). Furthermore, child maltreatment can affect critical periods in human development (Cavanaugh & Nelson, 2022; Li et al., 2022). Currie and Widom (2010) show that individuals with a history of child abuse and neglect have lower IQ test scores than their non-abused, non-neglected cohorts. Adults with a history of abuse and neglect tend to be less educated, less likely to be employed, have lower earnings, and own fewer assets. If employed, these individuals are less likely to be employed in a skilled or professional occupation. They also find that the negative impact of childhood abuse is stronger for women than men.

Ferguson et al. (2013) report that CSA is associated with depression, PTSD, self-esteem, and decreased life satisfaction. CSA exposure, in particular, can have long-term negative consequences that affect individuals well into adulthood, particularly among women. Women’s mental health, academic performance, and educational attainment can be linked to CSA (Rees & Sabia, 2013). The effect of the abuse diminishes when controlling for family-level heterogeneity, however. Robst and Smith (2008) investigate the effects of CSA on women’s income and find female victims of CSA earn less than non-victimized women. Sabia et al. (2013) find that CSA has the strongest effect on women’s labor force participation in addition to lower earnings. They estimate a 6.6% lower probability of labor force participation and a 5.1% lower average wage for women sexually abused in childhood. They attribute the effects of CSA to stress-related adverse psychological and physical consequences. Those who experience CSA are almost three times more likely to be out of the labor force due to sickness and disability. However, there is no effect of experiencing physical abuse and the likelihood of being unemployed (Barrett et al., 2014).

Not only can CSA and physical abuse in childhood impact the socioeconomic outcomes of an individual, but they can also have an intergenerational impact on their offspring. Intergenerational continuity in parenting is often referred to as intergenerational transmission (Assink et al., 2018). Offspring of mothers exposed to childhood abuse and neglect are more likely to have biological, behavioral, and developmental problems in early childhood (Garon-Bissonnette et al., 2022). Seteanu and Giosan (2022) find fathers’ ACE is associated with adult children’s ACE implying that the negative socioeconomic outcomes can persist intergenerationally.

Adverse Childhood Experience (ACE)

ACEs are defined as having any forms of psychological, physical, or sexual abuse and measures of household dysfunction such as substance abuse, mental illness, domestic violence, and history of incarceration. The link between exposure to ACEs in childhood and adolescence and adverse mental and health outcomes in adulthood is well documented (Elmore & Crouch, 2020; Merrick et al., 2017; Schilling et al., 2007; Sonu et al., 2019; Wade et al., 2016). These adverse outcomes in adulthood could potentially be a contributing factor to adverse socioeconomic outcomes. Font and Maguire-Jack (2016) assess the associations of ACEs and adult health outcomes and whether socioeconomic conditions mediate those associations. They find that socioeconomic conditions only explain a small portion of adverse health outcomes in adults who suffered ACEs.

Abbott and Slack (2021), for example, find that an additional ACE is associated with a 1.02 increase in depression using the Epidemiologic Studies Depression Scale-Revised (CESD-R), controlling for demographic covariates. The CESD-R measures 20 items related to depression in the form of loss of interest, appetite, sleep, concentration, feelings of worthlessness, fatigue, agitation, and suicidal ideation. Kerker et al. (2015) show among a sample of a nationally representative survey of children investigated by child welfare, an additional accumulation of an ACE is associated with worsening social development and mental health. Cavanaugh and Nelson (2022) show that nearly all ACEs, except for parental divorce for women and emotional neglect for men, are associated with depression in adulthood. They highlight childhood sexual abuse as a leading association with adult depression for both men and women.

Zielinski (2009) reports that victims of ACEs are more likely to be unemployed and fall into poverty in adulthood. When isolating ACEs by abuse type, he finds no employment effect for CSA victims. However, they are more likely to be in poverty than other ACEs. Those that experience childhood physical abuse are more likely to be unemployed and live in poverty. Metzler et al. (2017) use the BRFSS survey for 2010 and examine the impact that the number of ACEs has on educational attainment, employment, and income potential in adulthood. They find a correlation between the number of ACEs and adverse outcomes in adulthood. Notably, having three or more ACEs increases the probability of unemployment, dropping out of high school, and living in the FPL. Male victims who experienced childhood physical and sexual abuse and the number of ACEs experienced are less likely to be employed than females (Liu et al., 2013).

Suglia et al. (2022) explain that ACE among young adults is associated with lower educational attainment and mental health. Furthermore, they find that ACEs are strongly correlated with childhood socioeconomic status. Inevitably, the negative childhood experience perpetuates into adulthood depression and affects socioeconomic outcomes. Kim et al. (2021) also explain the correlation between childhood adversity resulting in depression in adulthood and low socioeconomic status. Children of low socioeconomic status are more likely to experience ACEs; it may be that parents who themselves experienced childhood abuse may be more likely to abuse and neglect their children, implying an intergenerational transmission (Zielinski, 2009).

Data and Methodology

Behavioral Risk Factor Surveillance System (BRFSS) Survey ACE module

The BRFSS survey, coordinated by the Center for Disease Control and Prevention (CDC), is a nationally representative cross-sectional telephone survey on noninstitutionalized adults in the United States. The survey is conducted by the state health department and captures prevalence data regarding risky behaviors, health conditions, and socioeconomic conditions.

In each survey round, states have optional modules to add to the core questionnaires. The BRFSS survey ACE questionnaires are adaptations from the original ACE Study from Felitti et al. (1998). The BRFSS added the ACE module as an optional module during the 2009–2012 survey rounds. Thirteen states elected to incorporate the ACE module in their survey questionnaires. Arkansas and Louisiana included the ACE module in their 2009 survey rounds. Washington D.C, Hawaii, Nevada, Vermont, and Wisconsin included the ACE module in their 2010 annual surveys. In 2011, Minnesota, Montana, and Washington added the optional ACE module, and Wisconsin and Vermont continued to collect survey data for the ACE module for a second year. 2012 was the last year the ACE module was offered as an optional module. Wisconsin elected to continue to collect survey responses, and Iowa, North Carolina, and Tennessee added the optional module to their annual surveys. Our sample includes all states that opted to add the ACE Module to their annual surveys.1

Although the BRFSS expanded their surveys to include cell phone respondents in 2011, cell phone users’ household size was not captured until 2014. Therefore, we only include individuals that responded via a landline phone and exclude cell phone respondents in our sample because the number of adults in the household is pivotal in our analysis.2

Model and Hypotheses

Childhood trauma and mistreatment have lasting impacts in adulthood. Empirical studies suggest a negative relationship between child abuse and socioeconomic outcomes. Our goal is to examine the impact of childhood physical and sexual abuse on socioeconomic outcomes in adulthood by isolating questions on the ACE questionnaire pertaining to CSA and physical abuse. The BRFSS does not provide information that captures the respondent’s childhood background. Therefore, we utilize the BRFSS ACE variables that capture household dysfunctions to measure the individual's childhood background. Additionally, we explore the relative effect of forced sexual intercourse and physical abuse on our socioeconomic outcomes of interest by gender. We anticipate that recurrent childhood physical and sexual abuse should directly affect high school completion, employment, and income in adulthood.

• H1. Recurrent childhood physical abuse is positively correlated with negative socioeconomic outcomes.

• H2. Recurrent childhood sexual abuse is positively correlated with negative socioeconomic outcomes.

• H3. Recurrent cases of both physical abuse and severe sexual abuse are positively correlated with negative socioeconomic outcomes.

We use an approach similar to Diette et al. (2017), who disaggregate those who experience CSA from physical abuse and create a measure for those who experienced both events. They estimate the severity of the abuse on high school completion and educational attainment and find that experiencing both events results in the greatest negative impact, especially for women. For men, they do not observe significant effects, however. We expand upon this by examining the relationships between frequent physical and severe sexual abuse and adult economic outcomes using the BRFSS survey. Including high school completion in our analysis ensures consistency in the findings between the datasets.

We define physical abuse as those physically abused by a parent or an adult more than once in childhood. We use Physical Abuse (only) to measure those that experienced only physical abuse and no concurrent sexual abuse in Eq. (1):

$$prob\left(y,=,1,|,x\right)=\mathrm{\Phi }({\gamma }_0+{\gamma }_1{PhysicalAbuse(only)}_i+{\gamma }_2^{{}^{\prime }}{HHDysfuntionACEs}_i+{\gamma }_3^{{}^{\prime }}{X}_i+{\gamma }_4^{{}^{\prime }}{\mathit{State}}_i+{\gamma }_5^{{}^{\prime }}{\mathit{Year}}_i)$$ 1

where $y$ represents our socioeconomic variables of interest. ${HHDysfuntion}_i$ is a vector of the ACE control variables intended to capture the household dysfunction characteristics of the individual during childhood. $X_i$ controls for marital status, gender, physical, mental, or emotional disability, and the respondent’s current age. ${\mathit{State}}_i$ and ${\mathit{Year}}_i$ are the state and year fixed effects, respectively.

Next, we focus on severe CSA involving intercourse. We isolate the effect of this form of abuse to measure its impact on socioeconomic outcomes in adulthood. We re-estimate the model in Eq. (2):

$$prob\left(y,=,1,|,x\right)=\mathrm{\Phi }({\beta }_0+{\beta }_1{ForcedIntercourse(only)}_i+{\beta }_2^{{}^{\prime }}{HHDysfuntionACEs}_i+{\beta }_3^{{}^{\prime }}{X}_i+{\beta }_4^{{}^{\prime }}{\mathit{State}}_i+{\beta }_5^{{}^{\prime }}{\mathit{Year}}_i)$$ 2

using Forced Sexual Intercourse (only) as the abuse variable. This measure excludes those who experienced both forced sexual intercourse and physical abuse during childhood. We define forced sexual intercourse by whether the respondent was forced into having sexual intercourse during childhood by someone five or more years older or an adult more than once.

The retrospective nature of the ACE questions naturally leads to recall bias because of imperfect recollection, the mood at the time of the survey, and the impact of experience ACE has on the individual (Colman et al., 2016; Hardt et al., 2006; Robst & Smith, 2008). Widom et al. (2004) also warn researchers about the magnitude placed on findings associated with ACEs on adulthood outcomes. However, several studies note that ACE recollection is relatively accurate (Dube et al., 2004; Hardt & Rutter, 2004; Hardt et al., 2006; Pinto et al., 2014; Yancura & Aldwin, 2009). Experiencing childhood physical and sexual abuse may have further recollection problems, but Hardt et al. (2006) show that the severity of the experience is relatively accurate. As a result, we identify childhood physical and sexual abuse by the severity of having experienced the event more than once. A person may inaccurately report whether the event occurred more than once, but perceived abuse would still have negative outcomes by the self-perception of the event’s trauma. Further, Robst and Smith (2008) find that if the person feels that the event harmed their life, the event’s impact on the individual’s adulthood outcome is larger than those who did not perceive the event as traumatic.

Approximately 14% of those who experienced physical violence also experienced forced sexual intercourse, while approximately 55% of those who experienced forced sexual intercourse were also physically abused.3Forced Sex and Physical Abuse represents those who have experienced both forced sexual intercourse and physical abuse during childhood.

$$prob\left(y,=,1,|,x\right)=\mathrm{\Phi }({\delta }_0+{\delta }_1{ForcedSexandPhysicalAbuse}_i+.{\delta }_2^{{}^{\prime }}{HHDysfuntionACEs}_i+{\delta }_3^{{}^{\prime }}{X}_i+{\delta }_4^{{}^{\prime }}{\mathit{State}}_i+{\delta }_5^{{}^{\prime }}{\mathit{Year}}_i)$$ 3

Table 1 provides descriptions of the variables in our dataset. Our socioeconomic outcome variables are High School Dropout, Unemployed, and Poverty. We define Unemployed as those that are not employed but available to work during the survey year.4 We use an approach similar to Metzler et al. (2017) and Sabik et al. (2018) to estimate whether an individual is living in the FPL. The BRFSS reports ranges of incomes ($0-$10,000, $10,000-$15,000, $15,000–20,000, etc.) based on total household income. We use the midpoint of the household income bracket for each individual based on their reported family size as a matching mechanism. Poverty measures whether the respondent’s family income is in the FPL based on the year in which the respondent was surveyed. We categorize individuals as living in the FPL if their household income falls within the United States Department of Health and Human Services FPL guidelines based on family size (Dickon, 2015).

Table 1.

Variable Definitions

Variable	Definition
Socioeconomic Outcomes
HS Dropout	= 1 if dropped out of high school
Unemployed	= 1 if out of work for less than a year or out of work for more than a year  = 0 if employed or self-employed Excludes those who are homemakers, students, retired, and unable to work
Poverty	= 1 if resides in a household that is in the federal poverty line in the survey year
Physical and Sexual Abuse
Physical Abuse (Only)	= 1 if Physical Abuse (Only)  = 0 otherwise
Forced Intercourse (Only)	= 1 if Forced Sexual Intercourse (Only)  = 0 otherwise
Forced Sex and Physical Abuse	= 1 if Forced Sexual Intercourse and Physical Abuse  = 0 otherwise
Household Dysfunctions ACEs
HH Illegal Drugs	= 1 if lived with someone who used illegal drugs or abused medications
HH Alcoholism	= 1 if lived with someone who was a problem drinker or alcoholic
HH Mental Illness	= 1 if lived with someone who was depressed, mentally ill, or suicidal
HH Incarceration	= 1 if lived with someone who served time in a correctional facility
HH Parental Divorce	= 1 if parents separated or divorced
HH Domestic Violence	= 1 if parents or adults physically abused each other at home
Demographic Measures
College Graduate	= 1 if graduate from college
Disabled	= 1 if limited in activity because of physical, mental, or emotional problems
Female	= 1 if female
Married	= 1 if married during the survey period
White	= 1 if White
Black	= 1 if Black or African American
Asian	= 1 if Asian
Hispanic	= 1 if Hispanic
Age	= Age of respondent in the survey year
Year	= Year of survey
State	= State of residence

All ACE variables correspond to when the respondent was under 18 years old

Our explanatory variables of interest are categorized by sexual and physical abuse ACEs and Household Dysfunction ACEs. Our analysis focused on responses to Question 7 and Question 11 of the ACE module.5 Further, we isolate our sample to only include respondents who answered “more than once” for these questions.

Our household dysfunction ACE explanatory variables are HH Illegal Drugs, HH Alcoholism, HH Mental Illness, HH Incarceration, HH Parental Divorce, and HH Domestic Violence. HH Illegal Drugs defines those who lived with someone who used illegal drugs or abused medications during childhood. HH Alcoholism identifies those who lived with someone who was a problem drinker or alcoholic during their childhood. HH Mental Illness represents those who lived with someone who was depressed, mentally ill, or suicidal during childhood. HH Incarceration identifies those who lived with someone who served time in a correctional facility during childhood. HH Parental Divorce defines those whose parents separated or divorced during their childhood. HH Domestic Violence represents those whose parents or adults in their household physically abused each other during childhood.

Results

Before presenting any estimation results, we provide summary statistics for the variables in Table 2. Column (1) shows the results for the full sample. Column (2) and Column (3) of Table 2 show the summary statistics for females and males in our sample, respectively. For men and women combined, only 5.3% of the respondents dropped out of high school. Males have a slightly higher high school dropout rate than females at 5.7% compared to 5% for females. 9.1% of the overall sample are unemployed, with 8.7% of women and 9.7% of men who are unemployed. Ten percent of the overall sample live in the FPL, but the distribution of this is not equal for men and women. Women have an 11.1% poverty rate, whereas only 8.8% of men in the sample report living in the FPL. Two percent of females and 0.6% of males report Forced Sexual Intercourse (Only).This is consistent with Barrett et al. (2014), who find that males are less likely to report CSA compared to females, and if they report, they are more likely to be educated than males who do not. Males report Physical Abuse (Only) at 11.6%, and for females, it is 10.7%. Forced Sex and Physical Abuse for males is 0.7%, while 2.4% for females.

Table 2.

Unweighted Summary Statistics

	(1)		(2)		(3)
	Full Sample		Female		Male
	Obs	Mean (SD)	Obs	Mean (SD)	Obs	Mean (SD)
HS Dropout	57,465	0.053	34,280	0.050	23,185	0.057
		(0.224)		(0.218)		(0.233)
College Graduate	57,465	0.401	34,280	0.405	23,185	0.395
		(0.490)		(0.491)		(0.489)
Unemployed	43,274	0.091	24,643	0.087	18,631	0.097
		(0.288)		(0.282)		(0.296)
Poverty	52,235	0.102	30,838	0.111	21,397	0.088
		(0.303)		(0.315)		(0.284)
Disabled	57,265	0.245	34,165	0.251	23,100	0.236
		(0.430)		(0.434)		(0.425)
Forced Intercourse (Only)	57,465	0.014	34,280	0.020	23,185	0.006
		(0.119)		(0.139)		(0.080)
Physical Abuse (Only)	57,465	0.111	34,280	0.107	23,185	0.116
		(0.314)		(0.309)		(0.320)
Forced Sex & Physical Abuse	57,465	0.017	34,280	0.024	23,185	0.007
		(0.129)		(0.152)		(0.083)
HH Illegal Drugs	57,465	0.096	34,280	0.097	23,185	0.096
		(0.295)		(0.295)		(0.294)
HH Alcoholism	57,465	0.260	34,280	0.280	23,185	0.231
		(0.439)		(0.449)		(0.421)
HH Mental Illness	57,465	0.183	34,280	0.212	23,185	0.140
		(0.387)		(0.409)		(0.347)
HH Incarceration	57,465	0.054	34,280	0.053	23,185	0.056
		(0.226)		(0.224)		(0.230)
HH Parental Divorce	57,465	0.232	34,280	0.242	23,185	0.218
		(0.422)		(0.428)		(0.413)
HH Domestic Violence	57,465	0.130	34,280	0.139	23,185	0.116
		(0.336)		(0.346)		(0.320)
Female	57,465	0.597
		(0.491)
Married	57,465	0.608	34,280	0.602	23,185	0.618
		(0.488)		(0.490)		(0.486)
White	57,465	0.802	34,280	0.795	23,185	0.811
		(0.399)		(0.403)		(0.392)
Black	57,465	0.062	34,280	0.068	23,185	0.051
		(0.240)		(0.253)		(0.221)
Asian	57,465	0.032	34,280	0.032	23,185	0.032
		(0.175)		(0.175)		(0.175)
Hispanic	57,465	0.037	34,280	0.036	23,185	0.037
		(0.188)		(0.187)		(0.189)
Age	57,465	48.7	34,280	48.7	23,185	48.8
		(11.5)		(11.3)		(11.7)

All variables are binary except for Age which ranges from 18–64. Standard deviations are in parenthesis

Overall, approximately 60% of the respondents were female, 60% of all respondents were married, and 40% obtained a college degree. 25.1% of females and 23.6% of males identify as having some sort of disability, which could be physical, mental, or emotional, limiting their activity. The race demographics are similar to the overall proportions of the U.S. population. The number of Hispanic respondents was only 3.7%, which is significantly lower than the percentage of Hispanics in the overall population.

Table 3 shows the results for Eq. (1) with the primary interest in Physical Abuse (only). In Column (1), we provide the relationships of the covariates with dropping out of high school. We do not observe a significant relationship between Physical Abuse (only) and high school noncompletion. Of all of the HH Dysfunction ACEs, we only find that living with someone incarcerated, parental divorce, or parental domestic violence in the household significantly increases the likelihood of dropping out of high school.

Table 3.

Probit marginal effects of Physical Abuse (Only) on socioeconomic outcomes

	(1)	(2)	(3)
	HS Dropout	Unemployed	Poverty
Physical Abuse (Only)	0.0059	0.0020	0.0169**
	(0.0069)	(0.0081)	(0.0081)
HH Illegal Drugs	0.0008	0.0044	0.0136
	(0.0077)	(0.0100)	(0.0094)
HH Alcoholism	0.0066	0.0081	0.0150**
	(0.0055)	(0.0066)	(0.0064)
HH Mental Illness	-0.0145**	0.0266***	-0.0013
	(0.0063)	(0.0075)	(0.0068)
HH Incarceration	0.0473***	0.0296***	0.0239**
	(0.0088)	(0.0106)	(0.0106)
HH Parental Divorce	0.0278***	0.0153**	-0.0022
	(0.0053)	(0.0062)	(0.0063)
HH Domestic Violence	0.0249***	0.0113	0.0171**
	(0.0066)	(0.0082)	(0.0076)
College Graduate		-0.0511***	-0.1117***
		(0.0064)	(0.0075)
Disabled		0.0710***	0.0858***
		(0.0058)	(0.0054)
Observations	57,465	43,140	52,083

For all regression result tables, the sample is restricted to those ages 18–64. Demographic control variables, and state and year fixed effects are not reported. The standard errors are in parentheses

*** p < 0.01, ** p < 0.05, * p < 0.1

Next, we turn our focus to adult economic outcomes. These models control whether the individual graduated from college or is physically or mentally disabled. In Table 3, Column (2), we focus on the relationship between our covariates and unemployment. As expected, those who graduated from college are less likely to be unemployed, and those who claim to be physically or mentally disabled are more likely to be unemployed. We do not observe a significant relationship between Physical Abuse (only) and unemployment. Living with someone with mental illness, incarceration, and parental divorce are significantly related to being unemployed. We attribute these significant effects to the individual’s household income during adolescence and educational attainment.

Table 3, Column (3), focuses on the relationship between our covariates on living in poverty. Physical Abuse (only) is associated with a 1.7% increased likelihood of being in poverty. The household dysfunction variables HH Alcoholism, HH Incarceration, and HH Domestic Violence are all significantly related to being in poverty.

In Table 4, we show the effect of Forced Intercourse (Only) on high school completion, unemployment, and poverty. Forced Intercourse (Only) is associated with a 2.8% greater likelihood of high school noncompletion. Similar to the results for physical abuse, living with someone incarcerated, parental divorce, or parental domestic violence in the household significantly increases the likelihood of dropping out of high school. We do not find a significant relationship between Forced Intercourse (Only) and unemployment. HH Mental Illness, HH Incarceration, and being disabled do show an increased likelihood of being unemployed, however. Whereas college graduates are 5.1% less likely to be unemployed. Forced Intercourse (Only) impacts the likelihood of being in poverty by 4.4%.

Table 4.

Probit marginal effects of Force Intercourse (Only) on socioeconomic outcomes

	(1)	(2)	(3)
	HS Dropout	Unemployed	Poverty
Forced Intercourse (Only)	0.0283*	-0.0053	0.0442***
	(0.0162)	(0.0191)	(0.0151)
HH Illegal Drugs	0.0009	0.0045	0.0137
	(0.0078)	(0.0100)	(0.0094)
HH Alcoholism	0.0067	0.0083	0.0154**
	(0.0055)	(0.0066)	(0.0064)
HH Mental Illness	-0.0142**	0.0268***	-0.0002
	(0.0062)	(0.0075)	(0.0068)
HH Incarceration	0.0471***	0.0296***	0.0237**
	(0.0088)	(0.0106)	(0.0106)
HH Parental Divorce	0.0277***	0.0153**	-0.0020
	(0.0053)	(0.0062)	(0.0063)
HH Domestic Violence	0.0269***	0.0120	0.0227***
	(0.0062)	(0.0078)	(0.0077)
College Graduate		-0.0512***	-0.1121***
		(0.0064)	(0.0075)
Disabled		0.0710***	0.0856***
		(0.0057)	(0.0054)
Observations	57,465	43,140	52,083

Table 5 reports the relationship between experiencing Forced Sex and Physical Abuse on our socioeconomic measures. Forced Sex and Physical Abuse significantly increases the likelihood of dropping out of high school to 4.3% and unemployment to 3.6%. We find no significant impact on poverty, however. Our results imply that an essential aspect of the relationship between CSA and economic prosperity depends on whether the individual experienced both sexual and physical abuse. The effects of the other household dysfunction variables are still significant and the likelihoods remain similar to the results in the models where each form of abuse is measured separately.

Table 5.

Probit marginal effects of Force Sex and Physical Abuse on socioeconomic outcomes

	(1)	(2)	(3)
	HS Dropout	Unemployed	Poverty
Forced Sex & Physical Abuse	0.0431***	0.0355*	0.0229
	(0.0136)	(0.0188)	(0.0147)
HH Illegal Drugs	0.0001	0.0040	0.0138
	(0.0078)	(0.0100)	(0.0095)
HH Alcoholism	0.0068	0.0082	0.0158**
	(0.0055)	(0.0065)	(0.0064)
HH Mental Illness	-0.0155**	0.0262***	-0.0004
	(0.0062)	(0.0075)	(0.0069)
HH Incarceration	0.0464***	0.0291***	0.0237**
	(0.0088)	(0.0106)	(0.0106)
HH Parental Divorce	0.0277***	0.0149**	-0.0019
	(0.0053)	(0.0062)	(0.0063)
HH Domestic Violence	0.0241***	0.0102	0.0209***
	(0.0063)	(0.0079)	(0.0078)
College Graduate		-0.0511***	-0.1120***
		(0.0064)	(0.0075)
Disabled		0.0707***	0.0857***
		(0.0058)	(0.0054)
Observations	57,465	43,140	52,083

Child Abuse Effects by Gender

To further investigate the impact of child abuse on socioeconomic outcomes, we re-estimate our models by gender. Tables 6, 7 and 8 report the relationship between our ACE and socioeconomic measures among females and males. Table 6 shows the results for Eq. (1) by gender. Columns (1–3) show the relationship of the covariates for females and probabilities of being a high school dropout, unemployed, and poverty, respectively. Columns (4–6) show these relationships for males. Those that experience childhood physical abuse do not seem to have lasting economic impacts in adulthood. We only find a significant relationship between Physical Abuse (Only) and an increased risk of poverty for males.

Table 6.

Probit marginal effects of Physical Abuse (Only) on socioeconomic outcomes by gender

	Females			Males
	(1)	(2)	(3)	(4)	(5)	(6)
	HS Dropout	Unemployed	Poverty	HS Dropout	Unemployed	Poverty
Physical Abuse (Only)	-0.0010	0.0117	0.0083	0.0164	-0.0077	0.0255**
	(0.0080)	(0.0101)	(0.0103)	(0.0113)	(0.0127)	(0.0119)
HH Illegal Drugs	-0.0007	0.0045	0.0041	0.0029	0.0060	0.0212
	(0.0088)	(0.0134)	(0.0105)	(0.0127)	(0.0143)	(0.0151)
HH Alcoholism	0.0080	-0.0068	0.0230***	0.0038	0.0244**	0.0077
	(0.0063)	(0.0076)	(0.0081)	(0.0093)	(0.0105)	(0.0099)
HH Mental Illness	-0.0085	0.0309***	0.0002	-0.0243**	0.0198	-0.0022
	(0.0070)	(0.0088)	(0.0084)	(0.0112)	(0.0126)	(0.0113)
HH Incarceration	0.0231**	0.0109	0.0237*	0.0727***	0.0462***	0.0221
	(0.0103)	(0.0132)	(0.0128)	(0.0137)	(0.0159)	(0.0171)
HH Parental Divorce	0.0317***	0.0260***	-0.0012	0.0226***	0.0052	-0.0035
	(0.0064)	(0.0080)	(0.0074)	(0.0085)	(0.0093)	(0.0102)
HH Domestic Violence	0.0312***	0.0108	0.0176*	0.0137	0.0130	0.0137
	(0.0075)	(0.0098)	(0.0092)	(0.0115)	(0.0133)	(0.0126)
College Graduate		-0.0563***	-0.1268***		-0.0448***	-0.0968***
		(0.0072)	(0.0080)		(0.0103)	(0.0127)
Disabled		0.0669***	0.0783***		0.0744***	0.0923***
		(0.0072)	(0.0066)		(0.0090)	(0.0086)
Observations	34,280	24,566	30,750	23,185	18,574	21,333

Table 7.

Probit marginal effects of Force Intercourse (Only) and covariates on socioeconomic outcomes by gender

	Females			Males
	(1)	(2)	(3)	(4)	(5)	(6)
	HS Dropout	Unemployed	Poverty	HS Dropout	Unemployed	Poverty
Forced Intercourse (Only)	0.0132	0.0109	0.0566***	0.0576	-0.0502	0.0205
	(0.0153)	(0.0205)	(0.0167)	(0.0365)	(0.0353)	(0.0350)
HH Illegal Drugs	-0.0009	0.0044	0.0029	0.0041	0.0056	0.0227
	(0.0088)	(0.0134)	(0.0105)	(0.0129)	(0.0144)	(0.0153)
HH Alcoholism	0.0078	-0.0064	0.0228***	0.0044	0.0241**	0.0089
	(0.0063)	(0.0077)	(0.0081)	(0.0093)	(0.0104)	(0.0099)
HH Mental Illness	-0.0087	0.0316***	0.0003	-0.0232**	0.0195	0.0001
	(0.0069)	(0.0086)	(0.0083)	(0.0111)	(0.0125)	(0.0115)
HH Incarceration	0.0230**	0.0107	0.0226*	0.0730***	0.0463***	0.0230
	(0.0103)	(0.0132)	(0.0127)	(0.0137)	(0.0159)	(0.0171)
HH Parental Divorce	0.0316***	0.0260***	-0.0014	0.0230***	0.0051	-0.0023
	(0.0064)	(0.0080)	(0.0074)	(0.0085)	(0.0093)	(0.0102)
HH Domestic Violence	0.0310***	0.0141	0.0201**	0.0202*	0.0100	0.0237*
	(0.0074)	(0.0092)	(0.0087)	(0.0104)	(0.0126)	(0.0135)
College Graduate		-0.0567***	-0.1271***		-0.0446***	-0.0973***
		(0.0072)	(0.0080)		(0.0103)	(0.0127)
Disabled		0.0672***	0.0774***		0.0741***	0.0930***
		(0.0072)	(0.0066)		(0.0090)	(0.0085)
Observations	34,280	24,566	30,750	23,185	18,574	21,333

Table 8.

Marginal effects of Force Sex and Abuse and covariates on socioeconomic outcomes by gender

	Females			Males
	(1)	(2)	(3)	(4)	(5)	(6)
	HS Dropout	Unemployed	Poverty	HS Dropout	Unemployed	Poverty
Forced Sex & Physical Abuse	0.0426***	0.0380*	0.0424***	0.0214	0.0290	-0.0444
	(0.0142)	(0.0209)	(0.0163)	(0.0291)	(0.0344)	(0.0338)
HH Illegal Drugs	-0.0021	0.0039	0.0029	0.0038	0.0053	0.0231
	(0.0087)	(0.0135)	(0.0106)	(0.0129)	(0.0144)	(0.0153)
HH Alcoholism	0.0078	-0.0063	0.0234***	0.0046	0.0240**	0.0090
	(0.0063)	(0.0076)	(0.0081)	(0.0093)	(0.0104)	(0.0099)
HH Mental Illness	-0.0108	0.0306***	-0.0008	-0.0231**	0.0192	0.0003
	(0.0069)	(0.0087)	(0.0084)	(0.0111)	(0.0125)	(0.0115)
HH Incarceration	0.0211**	0.0095	0.0220*	0.0731***	0.0459***	0.0232
	(0.0103)	(0.0133)	(0.0128)	(0.0137)	(0.0159)	(0.0171)
HH Parental Divorce	0.0312***	0.0253***	-0.0015	0.0234***	0.0048	-0.0021
	(0.0064)	(0.0080)	(0.0074)	(0.0085)	(0.0093)	(0.0102)
HH Domestic Violence	0.0276***	0.0118	0.0162*	0.0191*	0.0091	0.0251*
	(0.0075)	(0.0093)	(0.0087)	(0.0105)	(0.0128)	(0.0137)
College Graduate		-0.0565***	-0.1267***		-0.0446***	-0.0973***
		(0.0072)	(0.0080)		(0.0103)	(0.0127)
Disabled		0.0667***	0.0769***		0.0740***	0.0932***
		(0.0072)	(0.0066)		(0.0090)	(0.0085)
Observations	34,280	24,566	30,750	23,185	18,574	21,333

Results for Eq. (2) by gender are in Table 7. Although we do not find a significant association for high school noncompletion or unemployment, we find that females who experienced Forced Intercourse (Only) are 5.7% more likely to live in poverty. This result is stronger than the results obtained using the combined gender sample. This could indicate that the overall result is skewed upward by the women’s outcomes. Forced Intercourse (Only) does not significantly impact any of the male socioeconomic outcomes, however.

Table 8 displays the results from Eq. (3) for each gender. We only find a negative impact on females’ high school noncompletion, employment, and poverty. Females who experienced Forced Sex and Physical Abuse are 4.3% more likely to drop out of high school, 3.8% more likely to be unemployed, and 4.2% more likely to be in poverty.

Discussion

Sufferers of ACEs have more symptoms of depression and poorer cognition. Socioeconomically, they are expected to complete fewer years of education and earn less. The long-term impact of ACEs varies by the age of experience, gender, family background, type of ACE, and the number of ACEs experienced.

Our primary contribution to the literature is disaggregating those who have experienced sexual abuse from physical abuse from ACEs. In our sample, we find that over 50% of those who experienced sexual abuse are also physically abused. We find, on average, that experiencing physical abuse or CSA is significantly associated with living in poverty. However, experiencing them jointly results in a higher probability of high school noncompletion and unemployment. Furthermore, we disaggregate our sample by gender and find that females are overwhelmingly more affected by experiencing both sexual and physical abuse in childhood compared to males. Our results show that females who experience CSA and physical abuse jointly are more likely to drop out of high school, be unemployed, and live in poverty. For men, we do not find any significant negative association with our socioeconomic measures from experiencing both CSA and physical abuse.

HH Incarceration, HH Divorce, and HH Domestic Violence increase the likelihood of dropping out of high school in all of our models, indicating the impact that negative experiences parents or other adults in the home can have on whether an individual finishes high school. The type of dysfunction in the respondent’s childhood home appears to impact adult employment differently for men and women. HH Incarceration increases the likelihood of men being unemployed in all of our models, whereas HH Mental Illness and HH Divorce increase the probability of women being unemployed. The household dysfunction variables HH Alcoholism, HH Incarceration, HH Domestic Violence, as well as a college degree (which reduces the likelihood), and being disabled all increase the likelihood of living at the FPL regardless of the form of abuse for the combined sample and women. For men, however, Disabled and College Graduate were the only significant variables that impact poverty.

Childhood physical abuse is more prevalent for both men and women in our sample than forced sexual abuse and physical and sexual abuse in childhood. Our results show that women who suffer from repeated sexual abuse, whether sexual only or both sexual and physical, have a greater likelihood of negative socioeconomic outcomes in adulthood.

Why females suffer more than males socioeconomically from experiencing both forced sexual intercourse and physical abuse during adolescence is unclear. However, several factors could explain this heterogeneous outcome for females. One explanation could be gender differences in mental health outcomes after experiencing ACEs. For example, females are shown to internalize their response to child maltreatment and have more symptoms of depression than males (Chung & Chen, 2020; Gallo et al., 2018). Additionally, females who were physically and sexually abused in adolescence have a greater risk of depression than males (Gallo et al., 2018). Other factors that could explain the differential in the socioeconomic outcomes for females are gender discrimination in the labor force participation and market and differences in gender roles. Intergenerational socioeconomic outcomes and age of exposure to ACEs during adolescence can also matter. None of these confounders are testable with our data, however. To our knowledge, no existing empirical studies identify the channels of gender differences in socioeconomic outcomes relating to CSA and physical abuse.

Limitations and Future Directions

The BRFSS is a cross-sectional survey and lacks measures of the respondent’s childhood socioeconomic background. This limits the ability to control for additional channels through which childhood physical and sexual abuse determines adulthood socioeconomic status. Despite these drawbacks, it offers a large sample size and measures the severity of childhood physical and sexual abuse. Most importantly, our results are consistent with Diette et al. (2017), who use data from the National Comorbidity Survey-Replication and the National Survey of American Life.

ACE questionnaires are retrospective, which may lead to inaccuracies. Robst and Smith (2008) recommend including mental health treatment, symptoms, or diagnosis in CSA measures. However, these measures are rarely available to researchers. Another approach is collecting data on documented child abuse reports, but this can result in underreporting (Diette et al., 2017; Schurer et al., 2019). To address the shortcoming of the retrospective aspect of the ACE questionnaire, we rely on the severity of the childhood physical and sexual abuse measured by whether the event occurred more than once.

Another limitation when using ACE questionnaires in the BRFSS to assess the impacts of severe forms of child abuse is that there is no way to control for the magnitude and duration of the respondent’s abuse. Six of the eleven questions on the ACE questionnaire allow for non-binary responses.6 The questions are intended to identify various forms of abuse inflicted or witnessed by the respondent in childhood. These questions, however, limit the respondent to a response of either: “never,” “once,” or “more than once.” Because of this, there is no way to capture aspects of the abuse that could significantly impact their adult outcomes. For example, the age of onset of abuse, duration and frequency of abuse, or whether a parent or another adult inflicted the abuse all could impact how these forms of abuse manifest in adulthood.

There is a growing body of literature on the adulthood outcomes of experiencing ACEs. A few have addressed how childhood sexual abuse affects the victims’ industry choices and the gender wage gap (Robst, 2008; Robst & VanGilder, 2011). Future research can visit the extent to which ACEs impact the gender wage gap as well as the channels through which females who are sexually and physically abused fare worse than males socioeconomically.

Appendix

Table A.1 The BRFSS ACE Questionnaire

The following questions come from the ACE Module in the 2009–2012 BRFSS surveys

All questions refer to the time period before you were 18 years of age

Now, looking back before you were 18 years of age

Declarations

Conflict of Interest

On behalf of all authors, the corresponding author states that there is no conflict of interest.