Measurement of economic abuse among women not seeking ...

Measurement of economic abuse among women not seeking social or support services and

dwelling in the community

Rachel J. Voth Schrag, PhD, LCSW

University of Texas-Arlington School of Social Work

Kristen Ravi, LMSW

University of Texas-Arlington School of Social Work

Corresponding Author Information:

Rachel J. Voth Schrag, PhD LCSW

University of Texas-Arlington

School of Social Work

211 S. Cooper

Arlington TX 76019

[email protected]

1

Abstract

Scholars have defined economic abuse (EA) as tactics used by abusive partners to undermine the

self-sufficiency and economic self-efficacy of survivors of intimate partner violence (IPV).

However, no measures of EA have been tested in non-IPV-service seeking samples. The current

study assesses the psychometric properties of the Scale of Economic Abuse-12 (Postmus,

Plummer, & Stylianou, 2016) in a non-service seeking sample of adult females attending

community college. A quantitative web-based survey was administered to a simple random

sample of female community college students (n=435). Analyses included confirmatory factor

analysis (CFA) and exploratory factor analysis (EFA). CFA indicated a poor fit for the three-

factor model of the SEA-12 in this sample. The results of the EFA found a single factor model

retaining four items (the Scale of Economic Abuse-Short, or SEAS). Women are experiencing

EA outside of IPV service-seeking populations, and that tactics of economic control seem to be

central to EA in this sample.

Key Words: Economic Abuse, Intimate Partner Violence, Economic Stability, Measurement

Main Text

Intimate partner violence (IPV) is pervasive worldwide and affects at least one in four

women in the United States their lifetime (Black et al., 2011). It comprises a range of repetitive

behaviors including physical, sexual, economic, and psychological abuse and/or threats, which,

when combined, create a relational context of coercive control that limits the freedoms and

opportunities available to survivors (Stark, 2007). One domain of IPV is economic abuse (EA).

Scholars have defined EA as tactics used by abusers that undermine another person’s self-

sufficiency and economic self-efficacy, which include financial exploitation, economic control,

and employment sabotage (Adams, Sulivan, Bybee, & Greeson, 2008; Postmus, Plummer,

2

McMahon, Murshid, & Kim, 2012; Weaver, Sanders, Campbell, & Schnabel, 2009). Examples

of these include preventing or limiting work or school hours, harassment at the woman’s place of

employment or school, adversely affecting her credit history, and controlling shared finances by

demanding receipts, limiting accessibility to funds, or making unilateral financial decisions

(Adams et al., 2008; Authors, 2017a; Authors 2017b; Ericksson & Ulmestig, 2017; Postmus et

al., 2012; Weaver et al., 2009). Adams et al. (2008) and Postmus et al. (2012) found that 99% of

sheltered IPV survivors and 98% of service-seeking women reported experiencing tactics of EA

in their abusive relationship.

EA can be conceptualized as a subset of emotional abuse that includes domains such as

work sabotage, economic control, and economic exploitation (Postmus et al., 2012). Consistent

with coercive control theory (Stark, 2007), economic tactics are used by an abusive partner to

methodically, and often subtly, establish and maintain power and control. The financial impact of

EA on survivors is a point of important scholarship. Studies document that abusive intimate

partners will often interfere with employment, steal or spend women’s money, and sabotage

women’s credit (Adams, Greeson, Kennedy, & Tolman, 2013; Adams et al., 2008; Riger,

Ahrens, & Blickenstaff, 2001; Sanders, 2015; Tolman & Rosen, 2001). These behaviors are

linked to higher levels of depression, employment and housing instability, increased use of

public assistance, greater material hardship, and increased economic dependence on their abusive

partners for financial stability (Adams et al., 2013; Authors, 2015; Goodman, Smyth, Borges, &

Singer, 2009; Stylianou, 2018).

When paired with gendered dynamics related to financial control and economic

opportunity found in the United States and abroad, the impacts of EA for women survivors of

IPV become even starker. Data demonstrate that women face a greater prevalence and a greater

3

severity of poverty than their male counterparts, with forces including structural inequality,

workplace inequity, child-rearing and family responsibilities, and social institutions combining to

limit women’s economic security (Broussard, Joseph, & Thompson, 2012; Reid & LeDrew,

2013; U.S. Census Bureau, 2011). When these structural inequalities already at play are

combined with the tactics of coercive control and economic abuse, they further entrench the

economic and psychological impact of EA (Authors, 2015; Pyles, Katie, Mariame, Suzette, &

DeChiro, 2012; Stark, 2007). These compounding layers of economic dependence create an

environment which is ripe for an abusive partner to increase and maintain their control over

survivor’s economic present and future. In the context of EA, abusive partners actively limit the

ability of survivors to become self-sufficient by “making all financial decisions, reducing her

ability to acquire, use, and maintain money, and/or forcing her to rely on him for all of her

financial needs” (Postmus et al, p. 413, 2012).

Understanding the extent, impact, and causes of EA in the non-service seeking population

of survivors of IPV is an essential step in developing effective interventions and preventions for

IPV survivors. Many IPV survivors do not access domestic violence shelters or IPV specific

counseling agencies, yet they still experience a wide range of adverse impacts from violence. In

a nationally representative survey of survivors, nearly half reported having specific service needs

(such as medical, legal, and advocacy help) that were not met in the wake of victimization

experiences (Breiding, Chen, & Black, 2014). These survivors may have a unique set of

experiences and subsequent service needs due partly to differences in the extent, type, and

impact of economically abusive behaviors they have experienced. However, no extant measures

of EA have been used or tested in non-service seeking samples, leaving the field at a significant

deficit in our measurement of these dynamics. To address this gap, the current study is the first

4

to test the psychometric properties and results of the Scale of Economic Abuse-12 in a non-

service seeking sample of adult females.

The adverse impact of IPV on survivor’s financial stability has been well documented in

the literature. Historically, scholars have aggregated EA with emotional, psychological, or non-

physical violence (Stylianou, Postmus, & McMahon, 2013. Only recently have researchers

shifted their focus to identifying specific abusive behaviors that are present within EA (Adams,

Greeson, Kennedy, & Tolman, 2013) These behaviors seek to sabotage the economic efforts of

survivors and maintain economic power and control (Adams et al., 2013).

While there is an emerging consensus related to the gendered impact of economic

coercion and its link to other forms of IPV, there remains a lack of consistency in

conceptualization and measurement of EA (Postmus, Hoge, Breckenridge, Sharp-Jeffs, &

Chung, 2018). To this point, specific measures of EA have generally been used within service-

seeking populations of women (i.e., women in IPV shelters, seeking IPV counseling, or seeking

economic services related to IPV) (Adams et al., 2008; Postmus et al., 2016). Studies that have

evaluated economic control among IPV survivors in the general population have usually

depended on one or two items from larger studies such as the National Violence Against Women

Survey or the Fragile Families and Child Well-Being Survey (Outlaw, 2009; Postmus, Huang, &

Stylianou, 2012; Authors, 2015). A number of IPV measures include items that tap

economically abusive tactics within broader psychological, non-physical, or emotional abuse

subscales, while some newer measures include independent subscales addressing EA along with

other forms of violence. Table 1 provides examples of measures with items or sub-scales

addressing tactics of economic abuse (Campbell, Campbell, Parker & Ryan, 1994; Lehmann,

Simmons, & Pillai, 2012; Postmus et al., 2016; Shepard & Campbell, 1992; Tolman, 1999;

5

Weaver et al, 2009). The Scale of Economic Abuse and Scale of Economic Abuse-12 provide

the most commonly used standalone measures for these behaviors (Adams et al, 2008; Postmus

et al., 2016; Stylianou et al., 2013).

<Insert Table 1 about here>

The Scale of Economic Abuse

The Scale of Economic Abuse (SEA; Adams et al., 2008) is the first scale to measure EA

independently of other forms of abuse. The measure was created based on empirical research as

well as interviews with female IPV survivors and advocates. Adams et al. (2008) identified

several concepts related to EA including 1) preventing women’s resource acquisition, 2)

preventing women’s resource use, and 3) exploiting women’s resources. Initially, the authors

constructed a 120-item scale with Likert-type answers that ranged from 1 (Never) – 5 (Very

Often). The original SEA queries respondents about the frequency of behaviors “since the

relationship began.” After testing the measure with 103 service-seeking survivors, the scale was

reduced to 28 items that included two subscales Economic Exploitation (11 items) and Economic

Control (17 items). The full SEA had a reliability coefficient of .93, and the two subscales had

alpha coefficients of .89 (Economic Exploitation) to .91 (Economic Control; Adams et al., 2008)

indicating good internal consistency.

Postmus et al. (2016) further explored to psychometric properties of the SEA. They

initially conducted a CFA to test the two- factor structure of the SEA, but found a poor fit and

subsequently conducted an EFA of the SEA using the data collected from the 120 female

survivors from 15 IPV organizations across the United States. Results indicated the presence of

a three-factor structure that included 12 items of the original 28. The first factor, economic

control, consisted of five items. Examples of these items include “make you ask for money” and

6

“make important financial decisions without talking to you first.” The second factor,

employment sabotage, contained four items. Examples include “threaten you to make you leave

work” and “do things to keep you from going to your job.” The third factor, Economic

Exploitation, consisted of three items which included “spend the money you need for rent or

other bills” and “built up debt under your name.” The total SEA-12 had a reliability coefficient

of .89 indicating good internal consistency. The economic control (α=87), employment sabotage

(α=.86), and economic exploitation (α=.89) subscales also had good internal consistency.

Study Aims

Existing measures of EA have only been evaluated within IPV service-seeking samples

of women. The potential impact of EA on women living in the community is substantial, yet we

lack measures tested in non-service-seeking populations for these dynamics. It is possible that,

among the broader population, tactics of EA may be more varied, or that a few fundamental

dynamics may stand out as the most frequently observed forms of economic coercive control.

Having a greater understanding of these tactics outside of service-seeking samples could help

interventionists and preventionists tailor outreach and awareness campaigns to fit the experiences

of a broader array of survivors. Thus, the aims of the current study are to 1) test the factor

structure of the SEA-12 in a non-service seeking community college sample and 2) if possible,

reduce the number of items in the scale to make it more useable alongside measures for other

domains of IPV and associated constructs for population level surveys conducted among non-

service-seeking individuals. Compared to previous studies conducted among those living in a

shelter or explicitly seeking help for the economic impacts of abuse, it is expected that the

current study will observe less extreme EA, and potentially observe decreased levels of

behaviors such as those tapped in the SEA economic exploitation and work sabotage sub-scales

7

which imply fairly substantial financial entanglement between intimate partners. Given a

framework which identifies EA as a form of emotional abuse, we expect that a measure of EA

should be more strongly correlated with emotional abuse than with physical abuse. Similarly,

given the demonstrated link between EA and economic hardship, a valid measure of EA should

be significantly linked with economic hardship.

Methods

Data Collection

A quantitative web-based survey was administered to a simple random sample of

community college students from four campuses of a Midwestern college system. Using student

e-mail addresses provided by the system’s Office of Institutional Research, potential respondents

were contacted until a final sample of 435 respondents was obtained. The survey began with

screening questions for inclusion and exclusion criterial. Eligible participants identified as

female, were at least 18 years of age, and had been in an intimate relationship in the past 12

months. Because the larger study focused on academic and economic outcomes, all participants

were college students (see Authors, 2018 for more information on the larger study). The

institutional review boards (ethics panel) of the sponsoring University and the four community

colleges approved the study protocol before the beginning of the data collection, and participants

provided informed consent on a web-form before taking the survey.

The survey’s sampling frame was for-credit (i.e., credit earning) students enrolled during

Fall 2015 (N= 19,238). Using a random number generator, a simple random sample of potential

respondents was selected contacted via their school e-mail addresses. An initial screening

question was used to identify female respondents. Fifty-nine percent of students enrolled in the

8

system identified as female in Fall 2015, so approximately 11,350 of the 19,238 students in the

sampling frame were eligible to participate.

Selected students (n = 9053) received an e-mail inviting their participation and a follow-

up reminder e-mail approximately 12 days later. The study was framed as an investigation of the

economic, personal, and educational experiences of community college women, without

reference to IPV in the recruitment literature in order to protect potential survivors in the sample.

The survey included demographic measures along with standardized measures for exposure to

forms of violence. The survey took approximately 30 minutes to complete, and a $20 gift card

was provided to thank respondents for their time.

Response Rate. 5341 of the 9053 potential participants e-mailed would be expected to be

female. Of these, 15% (n=1,358) opened a recruitment e-mail. Fifty-six percent of them opened

the survey link, and 620 provided consented to participate. These 620 consented participants

represent 11.6% of the females initially e-mailed and 45.7% of those who opened a recruitment

e-mail. From consenting participants, 171 were screened because they had not been in an

intimate relationship in the past 12 months. Initially, 450 students screened into the study;

however, 14 were removed from the analysis because they dropped out of the survey before

completing the demographic questioner, leaving a final sample of 435 participants.

Description of the Sample. There were no statistical differences between the resulting sample

and the institutional demographics overall on any available demographic variable (see Table 2).

Respondents were 27 years old on average and less than half were full-time students. Fifty-eight

percent identified as White, while 27% identified as Black. Almost 95% reported that their

partner identified as male.

<Insert Table 2 about here>

9

Measurement

Scale of Economic Abuse SEA (12). In the current sample of community college

women, the alpha of the overall SEA was .86 and subscale alphas were .81 for ‘economic

control’, .68 for ‘work sabotage’, and .68 for ‘economic exploitation,’ suggesting issues with

reliability for the second two subscales in this sample. Table 3 reports the mean and frequency of

responses to the SEA-12 by subscale. Among study participants, 43.5% reported experiencing at

least one tactic of EA. The most frequently endorsed items were “made financial decisions

without you” and “kept financial information from you,” while “beat you up if you said you

needed to go to work”, “built up debt under your name”, and “threatened to make you leave

work” were all reported by less than four percent of respondents.

<Insert table 3 about here>

Measures of IPV and economic hardship which have previously been used with the SEA were

employed to assess the construct validity of the SEA-12 in this sample.

Intimate Partner Violence. Experiences of other forms of IPV is measured using the

Abusive Behavior Inventory (Revised) (Postmus et al., 2016). The ABI(R) includes three factors

covering domains of physical, emotional, and sexual abuse, and has a response set on a five-

point Likert scale (1=never to 5=very often) assessing the frequency of specific tactics over the

past 12 months. It was normed on a sample of over 400 female IPV survivors and has been used

previously with the Scale of Economic Abuse (Postmus et al., 2016; Stylianou et al., 2013). In

the current sample, subscale alphas were .89 for physical IPV, .71 for sexual IPV, and .89 for

emotional IPV.

Economic Hardship. Economic Hardship Index (EHI)- has been previously used with

the SEA-30 to examine relationships between EA and material hardship among IPV survivors

10

seeking services (Adams et al., 2008). The EHI is a checklist of 13 types of material hardship

which include food insecurity, difficulty obtaining stable housing, eviction, and utility

disconnection. Participants were asked to identify if they experienced any type of hardships over

the past one year and a total summed score is calculated. Adams et al. (2008) determined a

reliability coefficient of .86 in a sample of IPV survivors. Among the current sample of

community college women, the alpha was .88.

Monthly Individual Income. Participants were asked to report their individual monthly

income from all sources, including from work, public assistance, spousal or child support, and

any other sources. Students reported an average monthly individual income of $1,115.98

(SD=$1,064.65). Comparatively, the Bureau of Labor Statistics reported the 2018 median

personal income was $900 weekly, or approximately $3,783 monthly, for all full-time workers

(BLS, 2019).

Data Analysis

The goal of this study is to assess the factor structure and validity of the SEA-12 in non-

service seeking sample of women and assess if it is possible to reduce the items for use in

surveys of this population. For the purpose of the current analysis, the sample was split in half,

with 217 observations comprising the half on which the CFA was performed, and 218 in the half

on which the EFAs were performed. In Phase 1, a confirmatory factor analysis (CFA) was

conducted to determine if the three-factor solution for the SEA-12 found by Postmus et al.

(2016) fit the data for the current sample (Brown, 2007). Robust Maximum Likelihood

estimation (Satorra & Bentler, 1994) was used to address the substantial kurtosis and skewness

of individual scale items (scale mean skewness = 2.8, scale mean kurtosis = 11.78). All items

are interval level. STATA 12.1 was used to test the confirmatory model, and chi-square

11

statistics, comparative fit index (CFI), and the root mean square error of approximation

(RMSEA) were used as fit indices. A non-significant chi-square test, a CFI value higher than .9,

and a RMSEA below .06 are indices of a good fit (Brown, 2006; Bryne, 2012).

Phase 2 was to conduct an exploratory factor analysis (EFA) to assess the factor structure

with the current population and identify indicators to keep for non-service seeking population

surveys. A principal factor EFA was used, as it is free of distributional assumptions and less

prone to improper solutions than maximum likelihood techniques (Fabrigar, Wegener,

MacCallum, & Strahan, 1999). This is preferable in instances with substantial non-normality is

present in indicators (Brown, 2006). We performed the EFA through a series of principal axis

factor analysis with no rotation. We then conducted a series of principal axis analyses with

varimax rotation (Wood, Tatryn, & Gorsuch, 1996). Finally, we conducted the EFA using

principal axis functioning with direct oblimin rotation. We examined the scree plots and

eigenvalues greater than one to guide our model selection. A total of three models were

compared.

Phase 3 consisted of the validation of the new scale through the analysis of coefficient

alphas for reliability, and correlations between the new scale and the ABI total scale as well as

the physical, sexual, and emotional abuse subscales. To assess the relationship between EA and

economic hardship in the non-service seeking population of women, an additional bivariate

correlation was conducted to examine the relationship between the new scale, the EHI, and an

individual’s total monthly income.

Of the 435 respondents, only 14 observations had any missing data across the variables of

interest (3% missing). Missing data were at random, and maximum likelihood estimation was

used to address missing items (Allison, 2001).

12

Results

Phase 1: CFA

A CFA was conducted to test the fit between the current data and Postmus et al.'s (2016)

three-factor solution for the SEA-12. In our sample, the results indicated a poor fit X2 (40)

=152.58, p<.001, CFI=.82, RMSEA=.12 (90% CI=.10, .43). Therefore, we conducted an EFA to

examine the factor structure within the community sample of female college students.

Phase 2: EFA

Bartlett’s test of sphericity (X2 = 406.35, p<.001) and Kaiser-Meyer-Olkin [KMO]

(KM0=.81) were acceptable, indicating we could proceed to conduct an EFA. The results of the

EFA indicated the presence of a one-factor structure that utilized four of the original items which

had an eigenvalue of 2.50 and explained 62.61% of the variance. The factor loadings ranged

from .75-.86. This scale, comprised of 4 items, was named the Scale of Economic Abuse-Short

(SEA-S) (M=1.40, SD= .70). Retained items were “made financial decisions without you”,

“demanded to know how money was spent”, “kept financial information from you,” “made you

ask for money” (see Table 4).

<Insert Table 4 about here>

Phase 3: Validation

We assessed the internal consistency of the SEAS by examining the Cronbach’s alpha

coefficient and item-total correlations of the scale. The SEAS had a reliability coefficient of .87

indicating good internal consistency. The item correlations ranged from .56-.72 indicating a

moderate to strong relationship. We examined the convergent and discriminant validity of the

SEA-S by analyzing the correlations between the SEAS and the ABI total scale and its subscales

as well as the EHI and total monthly income (see Table 5). The SEAS was positively correlated

13

with the ABI-R total scale (r=.56, p<.001), ABI-R Physical IPV subscale (r=.38, p<.001), ABI-R

psychological subscale (r=.61, p<.001), ABI-R sexual abuse subscale (r=.50, p<.001), indicating

that higher levels of EA are associated with higher levels of other forms of abuse. Further, the

SEAS is significantly correlated with the EHI (r=.23, p=.001), suggesting that increased severity

of EA is linked to increased economic hardship in the current sample. There was no significant

correlation between the SEAS and respondent’s total individual monthly income. Although all

forms of abuse are correlated, they were not correlated highly enough to suggest

multicollinearity.

<Insert Table 5 about here>

Discussion

The SEAS appears to provide a brief, reliable, and valid tool for measuring experiences

of EA among non-service seeking women. It could be a useful and low-burden tool for assessing

EA as a distinct form of IPV in non-service seeking population prevalence studies, or for

scholars seeking to include EA among a range of IPV domains to be studied in non-service

seeking. The initial psychometric evaluation of the new scale found that it possesses strong

internal consistency reliability, and that the items are moderately correlated with each other. The

SEAS score was found to be significantly correlated with a participant’s extent of economic

hardship, pointing to initial evidence for the scale’s validity. The fact that the SEAS was more

strongly correlated with psychological/emotional abuse, of which EA is considered a sub-

category, than physical violence, provided initial evidence for convergent and discriminate

validity. The fact that EA was found to be a salient domain for many in this population of

community college students, who may have more transient relationships and thus less financial

entanglement, calls for future investigation into how EA functions in less committed or long

14

term relationships. The diversity of relationship types in the current sample provides a fresh look

at the role of EA across populations, pointing to a need for greater population-level attention to

this issue. Future work should validate the SEAS within other populations, including men, older

individuals, and teenage dating partners, as well as assess its test-retest reliability and association

with other theoretically linked constructs including health and well-being outcomes.

These findings also demonstrate that women are experiencing the tactics of EA outside of

IPV service seeking populations. Among study participants, 43.5% reported experiencing at least

one tactic of EA in the past 12 months, with about 25% reporting experiencing more than one

tactic. This suggests that coercive control related to economic and financial well-being is a

salient issue for many women outside of service seeking populations who are in or have recently

been in an intimate relationship. With such rates, education about the tactics of EA and

strategies for addressing them should be part of not only IPV services but other financial and

economic support services for women and families. Moving forward, it will be critical to

highlight the central role of economic control in EA for many survivors, and implement

strategies to rebuild women’s economic autonomy. This can include facilitating access to

economic resources, financial information, and financial services. IPV advocates, as well as

others working on issues of financial empowerment, should be educated about tools to support

survivors, including economic education, individual development accounts, credit counseling,

and individual economic advocacy (Postmus et al., 2015; Sanders, 2015; Von Delinde, 2016).

EA was distinct from, but moderately correlated with, other forms of IPV (physical,

psychological, and sexual violence). In the current sample, 77% of those who reported any other

experiences of IPV (i.e., physical, sexual, or psychological abuse) also reported at least one

instance of EA. In this population, EA is a common but not universal experience among

15

survivors of IPV. The evidence also suggests that tactics which have previously been labeled as

‘economic control’ may be uniquely salient when seeking to identify the core of EA in the

broader population, as the three most frequently endorsed items were from that SEA-12 subscale.

In a sample of women who were seeking economic services from IPV agencies, Postmus and

colleagues (2012) found only slightly higher rates of tactics of ‘economic control’ compared to

behaviors linked to ‘economic exploitation’ or ‘work sabotage.’ However, among this sample of

women, tactics of economic control emerged as the most frequently experienced examples of

EA. This finding may be unique to the context of community college students, who may be

vulnerable to economic control if they make choices to trade off remunerative work for

additional time and effort for school. Alternately, this could signal that economically controlling

tactics are a hallmark of EA in a non-service seeking population, comprising of behaviors that

are more likely to appear in less-violent or extreme relationships. Economic control may be

uniquely difficult to assess, as its tactics may be more covert that other forms of EA (Stylianou et

al., 2013). As argued by Stylianou et al. (2013), “they may often be perceived as innocuous and

can be more easily blend in as ‘normal' financial behavior that occurs between individuals in a

relationship” (p. 3200). Such behaviors are also concerning because they can be perpetrated

remotely and could continue post-breakup, especially in relationships where there is continued

financial entanglement (e.g., child or spousal support). Future research should determine the

extent to which economic forms of abuse persist post-breakup and how that relates to

revictimization risk and other negative outcomes for survivors (Stylianou et al., 2013). As these

tactics are identified, they should appear prominently in prevention education for young people

and public health style campaigns about intimate partner violence. Additional research should

16

focus on determining how EA manifests in additional non-service seeking population samples,

and on identifying warning signs for future victimization and perpetration.

The SEAS provides a useful tool for expanding EA research into these and other

populations because of its comparative parsimony. While versions of the SEA with more items

can provide a deeper and more comprehensive look into the varying dynamics faced by survivors

of IPV, their length may cause scholars and advocates to shy away from using them, especially

in population level surveillance studies or in clinical practice. This is especially true given the

fact that, to capture IPV more broadly, measure of EA should be paired with measures of other

forms of coercive controlling tactics. Short screening tools such as the SEAS could help scholars

and service providers to identify potential risk markers and develop tailored interventions.

Researchers could also use such a tool to understand the trajectory of EA tactics within

relationships, potentially helping to identify early warning signs of future controlling behavior.

It is also noteworthy that, in our sample, EA is correlated with economic hardship and not

correlated with individual income. This may suggest that experiencing high levels of economic

control within intimate relationships can result in economic hardship for survivors across the

income spectrum. It also underscores the importance of considering alternative metrics for

economic well-being when studying and working with survivors of intimate partner violence. A

survivor may have ‘sufficient’ income without any ability to access or leverage that income due

to the economic control of their partner (Authors, 2015). Alternately, this finding could reflect

specific dynamics within this collegiate population. For community college students, increased

monthly individual income may be associated with increased hours worked and fewer credits

taken per semester. Community college students are doing school on top of many other

commitments including work and parenting and may not have the financial security to go to

17

school fulltime. As such, increased monthly income may not function as a useful marker of

economic security in this population (Authors, 2018 a & b). Increased income also opens new

potential venues for economic exploitation or coercion by an abusive partner, who may steal or

divert substantial amounts of money. Given the unique dynamics of collegiate income earners,

whose income may actually be a sign of having less overall economic security, such tactics could

create increased vulnerability.

Limitations

These findings should be viewed in light of a number of limitations to the study. First,

these data come from a cross-sectional study, and without repeated measures, we have no ability

to assess the test-retest reliability of the SEAS. Further, the argument for the convergent validity

of the SEAS would be stronger if some of the other extant measures of EA (for example, the EA

subscale of the DV-FI) were also administered to the sample so that the correlation between

measures could be established. Future work should seek to assess these important dimensions of

validity. Further, there is significant evidence to suggest that surveys in non-service seeking

population samples may fail to capture the extent of forms of IPV, as survivors may be less

likely to respond, may fear retributions for honest answers, and may doubt that their answers will

lead to real change (Fincher et al., 2015). Study participants are all current community college

students, who may be different in important ways from other non-service seeking samples. For

example, they may have access to key resources due to their academic affiliation, and may have

certain social advantages associated with those attending higher education, such as prestige,

access to internships or college associated job search help (Belfield & Bailey, 2011).

Comparatively, community college students are often less economically advantaged than their

18

‘traditional college’ peers. They work more hours a week for pay, and often have many financial

obligations for which they are individually responsible (Ma & Baum, 2016).

The current study applies measures in a new context, and the use of electronic

recruitment may have decreased the likelihood of accessing respondents who have partners who

use electronic surveillance tactics. Next, there is a chance that study participants vary in

systematic ways from study non-participants, threatening the generalizability of findings.

Although a simple random sampling approach decreases this threat to generalizability, the use of

a web-based survey means that potential respondents are more likely to participate if they are

comfortable with the electronic interface, motivated to complete the survey, and frequently check

their e-mail.

Mirroring previous psychometric work done on the SEA, the current study only captures

female community college students. Future work should evaluate the SEAS, and EA more

broadly, among men, gender non-conforming individuals, and those in the workforce, among

other key groups. Additional research that looks across socio-economic levels, wealth

indicators, countries, and cultures would also be beneficial, as differences in labor practices,

social welfare institutions, and gendered expectations related to finances could all impact the

conceptualization and measurement of EA. Finally, the ABI, SEAS, and EHI were all used with

a 12-month recall time frame; thus they fail to capture historical experiences of abuse. They also

do not account for the intensity or motivation behind the acts of IPV.

Conclusions

The current study provides a glimpse at the dynamics of EA in a non-service seeking

sample of women and documents a reliable and valid short measure for EA which provides a

19

low-burden option for non-service seeking population studies of violence against women. Future

work should build on these findings in other samples, and look at the association of EA with

potentially important covariates. This work will move the field forward in supporting survivors

and establishing effective prevention strategies. Because of the vital role of economic stability in

safety from abuse, furthering this work is critical. Especially for those with children, economic

control may further entrap survivors by making them more dependent on their partners to

provide their basic needs, and limiting their access to resources that they might otherwise be able

to mobilize (Sanders, 2015). As argued by Brush (2004): “for women, the consequences of

poverty include not only hardships such as homelessness and hunger but also additional

vulnerability to being trapped in relationships with abusive men” (pg. 24).

20

References

Adams, A. E., Greeson, M. R., Kennedy, A. C., & Tolman, R. M. (2013). The effects of

adolescent intimate partner violence on womenʼs educational attainment and earnings.

Journal of interpersonal violence, 28(17), 283-300.

Adams, A. E., Sullivan, C. M., Bybee, D., & Greeso n, M. R. (2008). Development of the scale

of economic abuse. Violence Against Women, 14, 563-588.

Allison, P., (2001). Missing Data. Thousand Oaks, CA: Sage,

Authors, 2015

Authors, 2017a

Authors, 2017b

Authors, 2018a

Authors, 2018b

Belfield, C., & Bailey, T. (2011). The benefits of attending community college: A review of the

evidence. Community College Review, 31(1), 46-68.

Black, M.C., Basile, K.C., Breiding, M.J., Smith, S.G., Walters, M.L., Merrick, M.T., Chen, J.,

& Stevens, M.R. (2011). The National Intimate Partner and Sexual Violence Survey

(NISVS): 2010 Summary Report. Atlanta, GA: National Center for Injury Prevention and

Control, Centers for Disease Control and Prevention.

BLS. (2019). Usual weekly earnings of wage and salary workers: Fourth quarter 2018.

Washington D.C.: Bureau of Labor Statistics. Accessed from

https://www.bls.gov/news.release/pdf/wkyeng.pdf

21

Breiding, M.J., Chen J., & Black, M.C. (2014). Intimate Partner Violence in the United States —

2010. Atlanta, GA: National Center for Injury Prevention and Control, Centers for

Disease Control and Prevention.

Broussard, C., Joseph, A., & Thompson, M. (2012). Stressors and coping strategies used by

single mothers living in poverty. Affilia, 27, 190–204.

Brown, T., (2006). Confirmatory Factor Analysis for Applied Research. New York: Guilford.

Brush, L. D. (2004). Battering and the Poverty Trap. Journal of Poverty 8(3), 23-43.

Byrne, B., (2012). Structural Equation Modeling with Mplus: Basic Concepts, Applications, and

Programming. New York: Taylor & Francis Group.

Campbell, D., Campbell, J., King, C., Parker, B., & Ryan, J., (1994). The reliability and factor

structure of the index of spouse abuse with African-American women. Violence and

Victims, 9(3), 259-274.

Dutton, M. A., & Goodman, L., (2005). Coercion in intimate partner violence: Toward a new

conceptualization. Sex Roles. 11(12). 743-756.

Eriksson, M., & Ulmestig, R. (2017). “It’s not all about money”: Toward a more

comprehensive understanding of financial abuse in the context of

VAW. Journal of Interpersonal Violence. Online First. DOI:

10.1177/08860517743547

Fabrigar, L. R., Wegener, D. T., MacCallum, R. C., & Strahan, E. J. (1999). Evaluating the use

of exploratory factor analysis in psychological research. Psychological Methods, 4(3),

272-299. doi:10.1037/1082-989X.4.3.272

22

Fincher, D., VanderEnde, K., Colbert, K., Houry, D., Smith, L., & Yount, K., (2015). Effect of

face-to-face interview versus computer-assisted self-interview on disclosure of intimate

partner violence among African American women in WIC clinics. Journal of

Interpersonal Violence, 30(5), 818-838.

Goodman, L., Smyth, K. F., Borges, A. M., & Singer, R. (2009). When crises collide: how

intimate partner violence and poverty intersect to shape women’s mental health and

coping? Trauma, Violence & Abuse, 10(4), 306–29.

Lehmann, P., Simmons, C., & Pillai, V., (2012). The validation of the checklist of controlling

behaviors (CCB): Assessing coercive control in abusive relationships. Violence Against

Women, 18(8), 913-933.

Ma, J., & Baum, S., (2016). Trends in community colleges: Enrollment, Prices, Student Debt,

and Completion. College Board Research Brief, accessed from

https://trends.collegeboard.org/sites/default/files/trends-in-community-colleges-research-

brief.pdf

Outlaw, M. (2009). No one type of intimate partner abuse: exploring physical and non-physical

abuse among intimate partners. Journal of Family Violence, 24(4), 263-272.

Postmus, J., Hetling, A., & Hoge, G., (2015). Evaluating a financial education curriculum as an

intervention to improve financial behaviors and financial well-being of survivors of

domestic violence: Results from a longitudinal randomized controlled study. Journal of

Consumer Affairs, 49 (1), 250-266.

23

Postmus, J. L., Huang, C., & Stylianou, A. (2012). The impact of physical and economic abuse

on maternal mental health and parenting. Children and Youth Services Review 34(1),

1922-1928.

Postmus, J. L., Hoge, G., Breckenridge, J., Sharp-Jeffs, N., & Sharp-Jeffs, N. (2018). Economic

abuse as an invisible form of domestic violence: A multicountry review. Trauma,

Violence, & Abuse, online first, DOI: 10.1177/1524838018764160

Postmus, J. L., Plummer, S., & Stylianou, A. M. (2016). Measuring economic abuse in the lives

of survivors: Revising the scale of economic abuse. Violence Against Women, 22(6), 692-

703. doi:10.1177/1077801215610012

Postmus, J., Plummer, S., McMahon, S., Murshid, N., & Kim, M., (2012). Understanding

economic abuse in the lives of survivors. Journal of Interpersonal Violence, 27(3), 411-

430.

Postmus, J., Stylianou, A., & McMahon, S. (2016). The Abusive Behavior Inventory-Revised

(ABI-R). Journal of Interpersonal Violence, 31(17), 2867-2888.

Pyles, L., Katie, M., Mariame, B., Suzette, G., & DeChiro, J. (2012). Building bridges to safety

and justice: Stories of survival and resistance. Affilia, 27, 84–94.

Reid, C., & LeDrew, R. (2013). The burden of being ‘employable:’ Underpaid and unpaid work

and women’s health. Affilia, 28, 79–93.

Riger, S., Aherns, C, & Blickenstaff, A. (2001). Measuring Interference with Employment and

Education Reported by Women With Abusive Partners: Preliminary Data. In K. O’Leary

& R. Mairuo (Eds.) Psychological Abuse in Violent Domestic Relations. New York:

Springer.

24

Rubin, D.B. (1987), Multiple imputation for nonresponse in surveys, New York: John Wiley &

Sons, Inc.

Sanders, C. K. (2015). Economic Abuse in the Lives of Women Abused by an Intimate

Partner: A Qualitative Study. Violence Against Women, 21 (1), 3 – 29.

Shepard, M., & Campbell, J. A. (1992). The Abusive Behavior Inventory: A measure of

psychological and physical abuse. Journal of Interpersonal Violence, 7, 291-305.

Stark, E., (2007). Coercive control: How men entrap women in personal life. New York: Oxford

University Press.

Stata Corp (2012). Stata: Release 12. Statistical Software Guide. College Station, TX: StataCorp

LP.

Stylianou, A., (2018). Economic abuse experiences and depressive symptoms among victims of

intimate partner violence. Journal of Family Violence, online first, DOI: 10.1007/s10986-

018-9973-4.

Stylianou, A., Postmus, J., & McMahon, S., (2013). Measuring abusive behavior: Is economic

abuse a unique form of abuse? Journal of Interpersonal Violence, 28(16), 3186-3204.

Tolman, R., (1999). The validation of the psychological maltreatment of women inventory.

Violence and Victims, 14(1), 25-37.

Tolman, R. M., & Rosen, D. (2001). Domestic violence in the lives of women receiving

welfare: Mental health, substance dependence, and economic well-being.

Violence Against Women, 7(2), 141-158.

U.S. Census Bureau. (2011). Current population survey—2011, Annual social and economic

supplement. Retrieved from http://www.census.gov/

25

VonDeLinde, K. (2016). Survivor Centered Economic Advocacy: A Multilevel Approach to

Enhancing Survivors’ Economic Security. Washington, D.C.: The Center for Survivor

Agency & Justice.

Weaver, T. L., Sanders, C. K., Campbell, C. L., & Schnabel, M. (2009). Development and

preliminary psychometric evaluation of the domestic violence-related financial issues

scale (DV-FI). Journal of Interpersonal Violence, 24(4), 569–585.

Wood, J., Tataryn, D., & Gorsuch, R., (1996). Effects of under-and-over extraction on principle

factor analysis with varimax rotation. Psychological Methods, 1(4), 354-365.

Table 1

Example IPV Measures with Subscales or Items Tapping EA Domains

Measure (Author, year) Individual item/Subscale Example items

Abusive Behavior Inventory

(ABI; Shepard & Campbell,

1992)

Two items “Prevented you from

having money for your own

use”, “Put you on an

allowance”

Index of Spouse Abuse

(Campbell et al., 1994)

One item “Tried to control your

money”

The Psychological

Maltreatment of Women

Inventory (Tolman, 1999)

Long form: five items

Short form: one item

“My partner used our

money or made important

financial decisions without

talking to me about it”

Domestic Violence-Related

Financial Issues Scale (DV-FI;

Weaver et al., 2009)

5 Subscales covering:

Economic abuse, financial

self-efficacy, financial

security and future safety,

perceived financial role in

partner abuse, financial

distress and relationship

decisions

“My partner used our

money or made important

financial decisions without

talking to me about it”,

“My partner negatively

affected my credit rating”.

Checklist of Controlling

Behaviors (CCB; Lehmann et

al., 2012)

Seven-item subscale “Did not allow me equal

access to the family

money,” “Used my fear of

not having access to money

to control my behavior”

26

Table 2

Participant Demographics (n=435)

Sample

Mean (SD) or

% (N)

CC

Overall a

Female 100% (435) 59%

Age (years) 27.1 (9.9) 27n/s

Full Time Student 43.2% (179) 40%n/s

Race

White 58.1% (252) 56%n/s

Black/AA 27.4% (119) 31%

Asian 5.3% (23) 4%

Other 9.6% (40) 5%

Relationship Status

Single 31.3% (136)

Dating, not living together 27.0% (117)

Married 21.2% (92)

Dating, living together 16.8% (73)

Separated/Divorced/Widowed 3.7% (16)

Most recent partner identifies as

Male 94.5% (411)

Female 4.4% (19)

Transgender 1.2% (5) a Fall 2015

27

Table 3

Means and percentages for the Scale of Economic Abuse-12 in a non-IPV service seeking sample

of community college women (n=421)

SEA-12 Subscale Mean (SD) %a

Economic Control

Made financial decisions without you 1.50 (.95) 28.24

Demanded to know how money was spent 1.27 (.70) 16.27

Kept financial information from you 1.41 (.96) 20.00

Made you ask for money 1.23 (.69) 12.71

Demanded receipts or change 1.09 (.46) 5.88

Economic Exploitation

Spent money needed for rent/other bills 1.19 (.61) 11.06

Paid bills late, not at all 1.27 (.73) 15.60

Built up debt under your name 1.05 (.29) 3.06

Employment Sabotage

Did things to keep you from going to your job 1.10 (.42) 6.35

Demanded that you quit your job 1.08 (.42) 4.24

Threatened to make you leave work 1.04 (.40) 2.12

Beat you up if you said you needed to work 1.01 (.12) 0.71

a Ever Occurred

28

Table 4

Rotated Pattern Matrix and Indicators of the Scale of Economic Abuse-Short (SEAS)

Item SEAS

Made financial decisions without you 0.80

Demanded to know how money was spent 0.75

Kept financial information from you 0.86

Made you ask for money 0.75

% of variance 62.61

Note: Percentage of variance is post-rotation.

Table 5

Correlations between Scale of Economic Abuse-Short (SEAS) and Key Indicators

Mean (SD) Eco

nom

ic A

bu

se

(SE

AS

)

AB

I T

ota

l

Ph

ysi

cal

IPV

Psy

cholo

gic

al

IPV

Sex

ual

IPV

Eco

nom

ic H

ard

ship

Economic Abuse (SEAS) 1.34 (.63) --

ABI total score 6.17 (8.77) .56

***

--

Physical IPV 1.08 (.26) .38

***

.72

***

--

Psychological IPV 1.41 (.53) .61

***

.97

***

.59

***

--

Sexual IPV 1.14 (.49) .50

***

.58

***

.36

***

.56

***

--

Economic Hardship 2.35 (2.96) .23

***

.17

**

.20

**

.19

***

.18

***

--

Individual Monthly Income $1116 ($1065) .04 .04 .05 .02 -.05 .25

***

***p>.001 **p>.01