Psychology of Violence

© 2019 American Psychological Association 2020, Vol. 10, No. 3, 268 –278
2152-0828/20/$12.00 http://dx.doi.org/10.1037/vio0000244

The Revised Scale of Economic Abuse (SEA2): Development and Initial

Psychometric Testing of an Updated Measure of Economic Abuse in
Intimate Relationships

Adrienne E. Adams Megan R. Greeson

Michigan State University DePaul University

Angela K. Littwin McKenzie Javorka

This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

University of Texas—Austin Michigan State University

This document is copyrighted by the American Psychological Association or one of its allied publishers.

Objective: The purpose of this study was to advance the measurement of economic abuse by developing
an updated version of the Scale of Economic Abuse that addresses key limitations of existing instruments.
Building on the original Scale of Economic Abuse, we constructed a 2-dimensional Revised Scale of
Economic Abuse (SEA2) to measure abusers’ use of economic restriction and economic exploitation to
exert control over the economic domain of their partners’ lives. Method: Using data collected through
a survey of 248 women seeking services for intimate partner violence (IPV), we examined the factor
structure of the 14-item SEA2 to test the psychometric soundness of the 2-dimensional conceptualization.
We also performed an initial test of the instrument’s construct validity by examining its relationship with
closely associated constructs, material dependence on the abuser and outstanding debt. Results: Confir-
matory factor analysis provided support for the 2-factor structure of the SEA2. Regression analysis
results suggested that the SEA2 measures an economic dimension of IPV as intended and provided initial
evidence that the 2 subscales measure distinct forms of economic abuse. Conclusion: The SEA2 appears
to be a psychometrically sound instrument for measuring the economic abuse construct. Researchers can
use this instrument to further our understanding of the correlates and consequences of this distinct form
of IPV. Practitioners could use the SEA2 to assess the types and extent of economic abuse their clients
experienced. The substantive findings of the study also have implications for practice and policy.

Keywords: economic abuse, measurement, intimate partner violence, coercive control, domestic violence

Intimate partner violence (IPV) committed against women is recently, researchers have turned their attention to understand-
a widespread problem and significant public health concern ing and addressing economic abuse. Economic abuse involves
(Black et al., 2011). IPV is “a pattern of abusive behavior used behaviors that control a person’s ability to acquire, use, or
by one partner to gain and maintain power and control over maintain economic resources, thus threatening their economic
another intimate partner” (U.S. Department of Justice, 2011). security and potential for self-sufficiency (Adams, Sullivan,
The behaviors include physical, sexual, psychological, or eco- Bybee, & Greeson, 2008). These behaviors began to surface in
nomic threats or actions. Decades of research have produced a the violence against women literature in the 1970s (Walker,
substantial body of literature on the correlates and conse- 1979). The term “economic abuse” first appeared in the late
quences of physical and psychological forms of abuse. More
1980s (Pence & Paymar, 1986). In 2008, Adams and colleagues
produced the first instrument to measure economic abuse, the
Scale of Economic Abuse (SEA). That study and others since
have demonstrated that economic abuse is a distinct form of
This article was published Online First May 30, 2019.
IPV that has severe negative effects on women’s lives (Adams,
Adrienne E. Adams, Department of Psychology, Michigan State Uni-
versity; Megan R. Greeson, Department of Psychology, DePaul University; Beeble, & Gregory, 2015; Adams et al., 2008; Stylianou, Post-
Angela K. Littwin, School of Law, University of Texas—Austin; McKen- mus, & McMahon, 2013). In 2016, Postmus and colleagues
zie Javorka, Department of Psychology, Michigan State University. published a revised version of the SEA, called the SEA-12, with
We are tremendously grateful to the survivors and service organizations fewer items and a different factor structure than the original
that participated in this study, as well as our research assistants, Briann instrument (Postmus, Plummer, & Stylianou, 2016). Both the
Wilson and Elizabeth Dennis. This project was only possible because of
original SEA and the SEA-12 have limitations that need to be
your contributions.
Correspondence concerning this article should be addressed to Adrienne addressed to move the field forward. The purpose of this study
E. Adams, Department of Psychology, Michigan State University, East was to advance the measurement of economic abuse by devel-
Lansing, MI 48824. E-mail: [email protected] oping and validating an updated version of the SEA.

268
 SEA2 269

Literature Review mands is more likely since there are fewer resources to combat the
pressure to comply” (p. 749).
Research on economic abuse has burgeoned in the past decade. Economic abuse is a component of coercive control (Dutton et
We now know that economic abuse is distinct from and occurs as al., 2005; Hamberger, Larsen, & Lehrner, 2017; Pence & Paymar,
frequently in help-seeking samples as physical and psychological 1986; Stark, 2007). Abusers force their partners to make economic
abuse (Adams et al., 2008, 2015; Stylianou et al., 2013). We also decisions that they might not have otherwise. For instance, abusers
know that economic abuse has detrimental consequences for vic- direct how their partners spend money, make them quit or not take
tims’ economic and psychological well-being. Studies show that a job, and demand that they put the household bills or credit in
economic abuse is associated with reduced economic self- their name (Adams et al., 2008; Adams, Littwin, & Javorka, 2019;
sufficiency, increased financial strain, increased material hardship, Chowbey, 2017; Haj-Yahia, 2000; Sanders, 2015). Abusers also
and reduced access to financial resources (Adams et al., 2008, control their partners by making economic decisions that affect
2015; Hetling, Stylianou, & Postmus, 2015; Postmus, Plummer, their victims without their knowledge or consent. For instance,
McMahon, Murshid, & Kim, 2012). Recently, economic abuse has abusers refuse to include their partners’ names on assets, take out
been linked to increased depression and anxiety symptoms and
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

credit in their partners’ names, steal their partners’ money and

reduced quality of life (Adams & Beeble, 2018; Haj-Yahia, 2000;
This document is copyrighted by the American Psychological Association or one of its allied publishers.

property, and withhold financial information (Adams et al., 2008,

Postmus, Huang, & Mathisen-Stylianou, 2012; Voth Schrag, in press; Littwin, 2012; Sanders, 2015; Sharp, 2008). Within the
2015). broader context of coercive control, abusers create environments in
While the knowledge base is growing, there remains much to which refusing their demands or questioning their behavior is
learn, and measurement is vital to that endeavor. In 2008, Adams dangerous. If told to provide receipts or change after making a
and colleagues developed the first comprehensive measure of purchase or “asked” to assume sole responsibility for a lease or
economic abuse, the SEA (Adams et al., 2008). They used the IPV utility service, the victim does so or risks harm. If the abuser hides
literature and conversations with researchers, advocates, and sur- bills that come in the mail or uses his partner’s personal informa-
vivors to generate an initial pool of 120 items capturing econom- tion to take out credit in her name without her knowledge and she
ically abusive behaviors. With data from a sample of 103 service- suspects or discovers the bills or debt, confronting him—as one
seeking IPV survivors, they used an iterative conceptual and might do in a nonabusive relationship—means risking harm (Lit-
statistical process to reduce the initial pool of items to a 28-item twin, 2012; Sanders, 2015; Shoener, 2016). Ultimately, economic
scale. Exploratory factor analysis indicated that a two-factor solu- abuse is an effective control tactic because it limits women’s
tion was the best fit for the data. One factor, consisting of 17 items, economic resources, thereby restricting their autonomy/space for
was labeled “economic control.” The second factor, consisting of action and ability to combat the abuse (Dutton & Goodman, 2005;
11 items, was labeled “economic exploitation.” In 2016, Postmus Sharp-Jeffs et al., 2018).
and colleagues published a revised version of the SEA (Postmus et The function of economic abuse as a mechanism of control is
al., 2016). They collected data from 120 women participating in a reflected in the definition of economic abuse that Adams and
financial literacy program for IPV survivors. After a confirmatory colleagues (2008) constructed from the literature—“behaviors that
factor analysis (CFA) of the factor structure of the SEA showed control a woman’s ability to acquire, use, or maintain economic
poor model fit, they conducted an exploratory factor analysis and resources, thus threatening her economic security and potential for
arrived at a 12-item scale with three factors: economic control, self-sufficiency” (p. 565, emphasis added). Put another way, eco-
economic exploitation, and employment sabotage. The current nomic abuse involves exerting control over the economic domain
approaches to conceptualizing and measuring economic abuse of another’s life. Despite the centrality of control to economic
have two key limitations that warrant attention. First, they do not abuse, this is not well reflected in current conceptualizations. Both
sufficiently reflect the function of economic abuse as a mechanism the SEA and SEA-12 are conceptualized as multidimensional
of control. Second, they lack adequate conceptual coverage of scales with a distinct “economic control” dimension. This concep-
abusers’ use of the consumer credit system. tualization suggests that the element of control applies to a subset
of economically abusive behaviors when in fact it underlies all
Economic Abuse as a Mechanism of Control
economically abusive behavior. Further, the item wording in the
Coercive control theory posits a type of IPV that is rooted in SEA and SEA-12 does not adequately capture the controlling
abusers’ desire to control their partners (Crossman & Hardesty, nature of economically abusive tactics. Some items may inadver-
2018; Dutton & Goodman, 2005; Stark, 2007). Abusers enact tently tap poor financial management behavior instead of abuse.
control by making demands and tacitly or explicitly threatening For example, the measures ask respondents to report how fre-
harm for failure to comply (Dutton, Goodman, & Schmidt, 2005). quently their partners would “spend the money you needed for rent
Threats deemed credible based on the abusers’ past behavior or other bills.” This behavior could reflect imprudent spending
compel victims to comply, subordinating their own interests, de- habits rather than an attempt to control one’s partner. To address
sires, and values. In this way, coercive control effectively limits these limitations, we are proposing a two-dimensional conceptu-
women’s autonomy or “space for action” (Sharp-Jeffs, Kelly, & alization of economic abuse that more clearly articulates how both
Klein, 2018; Stark, 2007). Dutton and Goodman (2005) explained dimensions involve exercising control over the economic domain
that resisting abusers’ control requires tangible, social, and per- of another’s life.
sonal resources. Abusers wear down resistance by limiting and Economic restriction. The original economic control dimen-
depleting these resources. This “wearing down of resistance” is sion of economic abuse captured the ways abusers keep their
central to maintaining control over time. As Dutton and Goodman partners from accessing and using economic resources (Adams et
stated, “When resistance is lower, compliance with coercive de- al., 2008; Postmus et al., 2016). “Economic restriction” is a more
 270 ADAMS, GREESON, LITTWIN, AND JAVORKA

precise term for this set of tactics. It better distinguishes these the SEA and SEA-12, respondents are asked to report how often
tactics from the exploitive tactics abusers use to control their their partners “pay bills late or not pay bills that were in your name
partners. Research shows that abusers restrict their partners’ access or in both of your names” and “spend the money you needed for
to and use of a range of economic resources including income, rent or other bills.” Such behaviors are potentially damaging, and
financial information, and property. For instance, abusers limit in the context of abuse, the victim may not be able to intervene out
their partners’ access to and use of income by keeping them from of fear of reprisal (Sanders, 2015). However, absent further con-
working or taking their paychecks, hiding money, denying access text, these items may not distinguish abusive from nonabusive
to accounts, and dictating and monitoring spending (Adams et al., behavior. It is possible that such items tap poor financial manage-
2008; Brewster, 2003; Chowbey, 2017; Moe & Bell, 2004; Post- ment practices rather than tactics to exercise control over the
mus et al., 2016; Sanders, 2015; Sharp, 2008). Abusers limit their economic domain of a partner’s life. To capture the controlling
partners’ access to financial information about income, debt, and nature of economically abusive behavior, the wording of scale
investments by doing things like intercepting and hiding the mail items should reflect the victim’s lack of voice in the economic
(Adams et al., 2008; Brewster, 2003; Littwin, 2012). Abusers decision or freedom to choose or influence an economic course of
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

impose limits on the use of property by doing things like hiding car action (Dutton & Goodman, 2005; Katz, 1997; Stark, 2007).
This document is copyrighted by the American Psychological Association or one of its allied publishers.

keys or damaging their vehicle to prevent its use (Adams et al.,

2008; Sanders, 2015). Economic restriction is an effective mech-
Abusers’ Use of the Consumer Credit System
anism of control because it limits the victim’s economic resources
and forces her into a position of supplication. By restricting their The SEA was developed over 10 years ago (Adams et al., 2008).
partners’ access to and use of income, financial information, and At the time, there was empirical and anecdotal evidence that
property, the abuser forcibly establishes an arrangement in which abusers generate debt in their partners’ names. This behavior was
the victim is dependent on the abuser for financial and material captured in the SEA and SEA-12 by an item assessing the fre-
resources. quency with which an abusive partner would “build up debt under
Economic exploitation. In addition to restricting economic your name by doing things like use your credit card or run up the
resources, abusers also control their partners through economic phone bill” and in the SEA with an item assessing the frequency
exploitation. Economic exploitation involves forcibly, coercively, with which an abusive partner would “force you to give him
or fraudulently using a partner’s economic resources for one’s own money or let him use your checkbook, ATM card, or credit card.”
advantage (Adams et al., 2008, 2019). There are a number of ways We now know much more about this form of economic exploita-
abusers take advantage of their partners economically. Abusers tion and the measurement should reflect this learning.
“freeload” by demanding that the victim solely pay for household In her pioneering article “Coerced debt: The role of consumer
necessities, put their money in a joint bank account so the abuser credit in domestic violence,” Littwin (2012) coined the term “co-
can spend it freely, or use their money to buy the abuser goods or erced debt” to refer to all nonconsensual, credit-related transac-
pay their bills (Adams et al., 2008; Anderson et al., 2003; Chow- tions that occur in an intimate relationship where one partner uses
bey, 2017; Littwin, 2012; Sanders, 2015; Sharp, 2008). Another coercive control to dominate the other partner. Through interviews
way abusers economically exploit their partners is by stealing their with lawyers, researchers, and advocates, Littwin showed that
money or property. This can involve the abuser taking money from abusers generate debt through fraudulent and coercive transac-
their partner’s purse or wallet or a joint bank account without tions. Abusers use their partners’ personal information to fraudu-
consent or taking the victim’s property for their own use or lently open credit cards, take out loans, or put leases, utilities, or
liquidation (Adams et al., 2008; Anderson et al., 2003; Brewster, phone services in their partners’ name without their knowledge.
2003; Chowbey, 2017; Sanders, 2015). In addition to free-loading They also use coercion to force their partners to use credit cards,
and stealing, IPV victims also report that their partners exploit take out loans, or put services in their own names when they would
them by generating debt in their names through fraud and/or not have otherwise done so. Of the 55 professionals interviewed,
coercion (Adams et al., 2019; Littwin, 2012). Economic exploita- 93% knew IPV victims who experienced coerced debt. For exam-
tion is an effective mechanism of control because it depletes the ple, participants shared stories of victims whose partners forged
victim’s economic resources. By forcing the victim to use their their names on credit card offers or loan documents, coerced them
money or credit for the abuser’s benefit or by taking the victim’s to purchase items on credit, and required that household debts be
assets or property for their own use, the abuser coercively estab- in their names.
lishes an arrangement in which the victim’s economic resources Since Littwin’s original work, two other studies have provided
are drained to their benefit, diminishing the victim’s resources and evidence of coerced debt. In a qualitative study with 30 low-
compromising their economic well-being. income IPV survivors participating in an economic development
Capturing control with item wording. Along with clearly program, Sanders (2015) found a common theme among partici-
articulating the element of control in the conceptualization of the pants of accumulating debt and damaged credit because of abuse.
dimensions of economic abuse, grounding economic abuse in She reported that, on average, women were US$9,132 in debt, and
coercive control theory requires attending to how control is re- in many cases, the abuse played a significant role. In another study,
flected in the measurement of the construct. In economically Adams and colleagues (2019) surveyed 1,863 women who called
abusive relationships, abusers make economic decisions unilater- the National Domestic Violence Hotline in response to IPV. They
ally and impose them on the victim through fraud, force, coercion, found that 52% of the callers reported that their partner had put
or manipulation (Anderson et al., 2003; Brewster, 2003; Littwin, debt in their name via a coercive and/or a fraudulent transaction.
2012; Sanders, 2015; Stark, 2007). Items used to measure eco- Forty-three percent reported a coercive transaction; that is, an
nomic abuse need to better reflect this dynamic. For instance, in intimate partner coerced them into borrowing money or buying
 SEA2 271

something on credit when they did not want to. Twenty-two property.” We added “pawn your property or shared property when
percent reported a fraudulent transaction; that is, they found out you didn’t want him to” to tap the controlling nature of the act.
about debt or bills that their intimate partner had put in their name Also, the SEA included a number of work interference items that
without their knowledge. would only be applicable to women who were working, such as
Recent research has expanded our understanding of this form of “do things to keep you from going to your job” and “demand that
economic exploitation. We now know that the two items used in you quit your job.” Including items that are not applicable to a
the SEA and SEA-12 do not adequately capture the concept. The subgroup of participants detrimentally affects the measure’s con-
item “build up debt under your name by doing things like use your struct validity (DeVellis, 2017). We added “keep you from having
credit card or run up the phone bill” limits attention to specific a job or going to work” to capture work interference in a way that
types of behavior, namely, using an existing credit card and applies regardless of whether the participant had a job. In addition
running up a phone bill. “Force you to give him money or let him to adding revised versions of original items, we wrote 10 new
use your checkbook, ATM card, or credit card” conflates a debt- items based on the literature and conversations with survivors and
generating action with other forms of economic restriction and/or advocates. The new items largely tapped exploitive tactics not well
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

exploitation—taking one’s money. Research suggests that to en- assessed with the SEA, such as “Make you use your money to buy
This document is copyrighted by the American Psychological Association or one of its allied publishers.

sure adequate conceptual coverage, the SEA needs to more fully him things or pay his bills when you didn’t want to” and “Take out
capture abusers’ use of fraud and coercion to take out loans, buy a loan or buy something on credit in your name without your
items on credit, or put bills in their partners’ names. permission.” Each of the 46 items was categorized as either eco-
nomic restriction or economic exploitation. The initial pool in-
Current Study cluded 22 economic restriction items and 24 economic exploitation
items rated on a 5-point Likert-type scale with response options of
The purpose of the current study is to advance the conceptual- 0 ⫽ never, 1 ⫽ hardly ever, 2 ⫽ sometimes, 3 ⫽ often, and 4 ⫽
ization and measurement of economic abuse by addressing key very often, indicating the frequency with which an intimate partner
limitations of existing instruments. Building on the original SEA, engaged in the behavior during the relationship.
we constructed a two-dimensional Scale of Economic Abuse
(SEA2), capturing economic restriction and economic exploita-
tion. Our research questions and hypotheses are as follows: Other Measures
In addition to the initial pool of items for the SEA2, we used
Research Question 1: Is the factor structure of the two-
measures of psychological abuse, physical abuse, economic well-
dimensional measure psychometrically sound?
being, and demographic characteristics to validate the new mea-
Research Question 2: Do the SEA2 and its subscales demon- sure.
strate internal consistency reliability? Psychological Maltreatment of Women Inventory. We
used the 14-item Psychological Maltreatment of Women Inventory
Research Question 3: Are the SEA2 and its subscales valid (PMWI; Tolman, 1999) to assess the degree of psychological
measures of the constructs they are designed to measure? abuse participants experienced in their relationship. Items were
rated on a scale ranging from 0 (never) to 4 (very frequently).
Hypothesis 1: The SEA2 will be significantly positively Examples of the items included “My partner called me names” and
associated with material dependence on the abuser after “My partner told me my feelings were irrational or crazy.” We
controlling for physical and psychological abuse. computed a mean score for analyses (M ⫽ 2.97, SD ⫽ .76). The
Hypothesis 2: After controlling for economic exploitation, PMWI is a valid and reliable measure of psychological abuse used
economic restriction will have a positive association with extensively in IPV research (Tolman, 1999). In this study, the
material dependence on the abuser. instrument had an internal consistency coefficient of .90. Tolman
(1999) showed that the PMWI differentiated between women with
Hypothesis 3: The SEA2 will be significantly positively and without abusive partners, providing evidence of validity.
associated with outstanding debt, an indicator of resource Modified Conflict Tactics Scale. The Conflict Tactics Scale
depletion. (CTS) as modified by Sullivan, Basta, Tan, and Davidson (1992)
was used to assess the frequency of physical violence experienced
Hypothesis 4: After controlling for economic restriction, during the relationship. The modified CTS (Straus, 1979; Sullivan
economic exploitation will have a significant positive rela- et al., 1992) contains 23 items rated on a scale ranging from 0
tionship with outstanding debt. (never) to 6 (more than four times per week). Example items
include “Pushed or shoved you,” “Choked or strangled you,” and
Method “Forced sexual activity.” We computed a mean score for analyses
(M ⫽ 1.38, SD ⫽ 1.14). This scale has demonstrated good internal
SEA2 Item Pool Generation consistency in other studies with service-seeking IPV victims (e.g.,
␣ ⫽ .92, Goodkind, Sullivan, & Bybee, 2004; ␣ ⫽ .90 to .93,
An initial pool of 46 items for the SEA2 was generated from Adams et al., 2015). In this sample, the reliability coefficient was
several sources. We started with the original 28 items from the .95. The CTS is one of the most widely used measures to assess
SEA and added eight new items with revised wording to better IPV and has been shown to be valid and reliable across diverse
reflect the element of control and ensure universal applicability. samples (Lucente, Fals-Stewart, Richards, & Goscha, 2001; Straus
For example, the SEA included “pawn your property or shared & Mickey, 2012).
 272 ADAMS, GREESON, LITTWIN, AND JAVORKA

Material dependence. Material dependence on the abuser race/ethnicity was not among the options provided. Most partici-
was assessed with the item “During your relationship, to what pants (83%) had at least a high school education. Thirty-one
extent did you rely on the financial resources from the person who percent had completed trade school, college, or an advanced de-
hurt you in order to have basic necessities (for example, housing, gree. Almost half (48%) of the sample was employed at the time
food, transportation)?” Response options were “not at all,” “a of data collection. Twenty-six percent worked full-time, 16%
little,” “somewhat,” “quite a bit,” and “completely,” coded 0 to 4, worked part-time, and 6% worked sporadically. Over half (58%)
respectively (M ⫽ 2.39, SD ⫽ 1.4). This item was written for this of the sample had a yearly net household income of $15,000 or
study. Therefore, its construct validity has not been tested. less. Another 18% reported an annual household income of
Outstanding debt. To assess outstanding debt, participants $15,001–$30,000, and 13% had an annual household income of
were asked if they currently owed money (yes/no) for the follow- $30,001–$50,000. Eleven percent reported a net household income
ing: (a) unpaid rent or mortgage, (b) an unpaid utility bill, (c) an of $50,000 or more. The majority (93%) of women surveyed
unpaid medical bill, (d) a student loan, (e) a credit card bill, (f) a reported that their abusive partner was male, 6% said their partner
payday loan, and (g) a loan from family or a friend. We opera- was female, and 1% reported that their partner was transgender.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

tionalized outstanding debt as the number of types of debt the Their relationship length ranged from 1 month to 45 years (average
This document is copyrighted by the American Psychological Association or one of its allied publishers.

participant owed. “Yes” responses were summed to arrive at a of 9 years), and almost all (94%) participants had lived with their
score ranging from 0 to 7 (M ⫽ 2.73, SD ⫽ 1.9). This measure was partners at some point. Almost all of the participants experienced
created for this study. Therefore, its construct validity has not been physical abuse (98%), and 100% experienced psychological abuse.
tested. Based on the number of latent factors, number of indicators, and
Control variables. Four demographic variables were in- the size of the factor loadings, 248 is a sufficient sample size to
cluded as controls owing to potential associations with the vari- evaluate overall model fit and individual model parameters in the
ables of interest. These included the participant’s age in years, CFA (based on power of 80% or greater for all parameters; see
whether the participant identified their racial background as Monte Carlo work by Wolf, Harrington, Clark, & Miller, 2013).
“White” (coded 0 ⫽ no, 1 ⫽ yes), the number of children they had
under age 18, and whether the participant ever lived with their Procedures
abusive partner (coded 0 ⫽ no, 1 ⫽ yes).
Staff provided eligible clients with private space in which to
complete a self-administered, paper questionnaire. The consent
Participants form appeared on the face-page of the questionnaire. The form
Participants were recruited from domestic violence (DV) service instructed women who did not want to participate to place their
organizations in a Midwestern state. The state-wide DV coalition blank questionnaire in the envelope, seal it, and return it to the staff
provided us with a de-identified list of 26 organizations. The list person. The form directed women who were willing to participate
included the region of the state in which the organization was to proceed to the first page of the questionnaire. The end of the
located, the types of services offered, the number of adult clients questionnaire instructed participants to put the completed ques-
served annually, and the monthly income of clients served. Select- tionnaire in the envelope provided, seal it, and return it to the staff
ing potential sites from the list, we sought diversity in geographic person. When the participant returned the questionnaire, the staff
location, types of services provided, and client income. We ex- person gave her a $5 gift card to thank her for her time. The lead
cluded sites that primarily served non-English-speaking clients author’s university institutional review board approved the study
owing to resource constraints that prohibited translation of study procedures.
materials. In total, we identified 14 potential sites. The state
coalition sent an e-mail to the executive directors of those orga- Results
nizations to introduce the study. We made contact with those who
expressed interest to provide details about the study procedures. In Final Scale Construction
the end, 11 DV service organizations agreed to partner with us on
this study. We used an iterative statistical and conceptual process to con-
Using an instruction sheet that we provided, the staff of the DV struct the final SEA2. Our aim was to produce a brief measure that
service organizations told clients about the study, screened for maintained conceptual coverage. We also sought to ensure that
eligibility, and invited them to participate. To be eligible for the items were widely applicable. Five items were excluded because
study, clients had to be female, at least 18 years old, English- they did not apply to all participants, resulting in a high degree of
speaking, and in an intimate relationship with someone who was missingness. For example, “steal your car keys or take your car so
abusive in the past 6 months. The 6-month time frame was chosen you couldn’t go look for a job or go to a job interview” was
to maximize recall of events that took place during the relation- excluded because almost 20% of the sample marked “not applica-
ship. ble” or left the question blank. Fifteen items were excluded owing
A total of 248 women participated in the study. They ranged in to a lack of clear conceptual fit or weak statistical fit with one of
age from 19 to 79 years (M ⫽ 36, SD ⫽ 11.28). Over half (61%) the subscales. For example, “take your paycheck, financial aid
of the women surveyed had children. The number of children check, tax refund check, disability payment or other support pay-
ranged from zero to six (M ⫽ 1.43, SD ⫽ 1.52). Forty-three ments from you” and “take money from you without your permis-
percent reported their race/ethnicity as White, 28% as Black, 18% sion” were excluded because these behaviors could be classified as
as Hispanic/Latina, 9% as biracial or multiracial, 1% as Native economic restriction or economic exploitation. Twelve items were
American, and 0.4% as Asian. Another 0.4% indicated that their excluded owing to conceptual redundancy with better performing
 SEA2 273

items. For example, “demand to know how money was spent” was Factor structure (Research Question 1). After reducing the
dropped because it was conceptually similar to “demand that you item pool to the 14-item SEA2, we conducted a CFA to test the
give him/her receipts or change when you spent money,” which proposed two-factor structure. A two-factor CFA model was tested
had a higher item-total correlation. Through this process, we in Amos Version 23 using maximum likelihood estimation. Eco-
reduced the initial pool of 46 items to the 14-item SEA2. nomic restriction and economic exploitation were modeled as
correlated, latent constructs, each with seven indicators (see Table
Participants’ Experiences of Economic Abuse as 1 for the final 14 items and the subscale they were assigned to).
Measured by the SEA2 Model fit was assessed by chi-square, root mean square error of
approximation (RMSEA), and standardized root mean squared
Of the 248 women surveyed, 96% had an abusive partner who
perpetrated at least one tactic of economic abuse as measured by residual (SRMR) statistics. The chi-square likelihood ratio test was
the 14-item SEA2. Ninety-one percent experienced economic re- used to compare relative fit of alternative nested models.
striction. The most commonly used economic restriction tactics The CFA provided support for the proposed two-factor structure
of the SEA2. We tested an initial two-factor model, ␹2(76) ⫽
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

were “decide how you could spend money rather than letting you
215.52, p ⫽ .00, RMSEA ⫽ .086, SRMR ⫽ .05, and then con-
This document is copyrighted by the American Psychological Association or one of its allied publishers.

spend it how you saw fit” (74%), “make you ask him/her for
money” (73%), and “keep financial information from you” (67%). ducted post hoc modifications to improve model fit. Specifically,
Eighty-three percent experienced economic exploitation. The three covariances were added between four pairs of residuals when there
most common exploitive tactics were “spend his/her money how- was solid conceptual justification for doing so, owing to shared
ever he/she wanted while your money went to pay for necessities” item wording (e.g., multiple items about loans, shared wording
(71%), “make you use your money to buy him/her things or pay applied to three pairs of residuals) or shared tactics across the
his/her bills when you didn’t want to” (54%), and “steal your items (multiple items involving hiding something, one pair of
property” (54%). The descriptive statistics for the SEA2 items are residuals). The final model demonstrated strong model fit,
presented in Table 1. ␹2(72) ⫽ 125.84, p ⫽ .00, RMSEA ⫽ .055, SRMR ⫽ .044. See
Table 1 for CFA results. This model represented a statistically
Psychometric Assessment of the SEA2 significant improvement in model fit over and above a one-factor
model, suggesting the two-factor model is preferable (change in
To prepare the data for the psychometric analysis, we performed
missing data analysis and imputed missing data. The amount of ␹2(1) ⫽ 153.46, p ⫽ .00).
missing data was low. Only 1.50% of values were missing for the Reliability (Research Question 2). Internal consistency was
46 potential SEA2 items and 1.45% of values were missing for the assessed by Cronbach’s ␣ and by standardized loadings from the
full dataset. Missing data were imputed using expectation maxi- CFA. The SEA2 and both subscales showed strong internal con-
mization (Schafer & Graham, 2002). Other variables that were not sistency. Cronbach’s ␣ was .89 for the seven-item Economic
analyzed in the study were also included in the matrix to assist in Exploitation subscale, .91 for the seven-item Economic Restriction
imputation. Little’s missing completely at random test was not subscale, and .93 for the full scale. Standardized regression load-
significant, ␹2(9396) ⫽ 9386.76, p ⫽ .53, suggesting that the data ings in the CFA were all high, indicating strong, positive relation-
could be treated as missing completely at random. ships between each item and its respective factor. Standardized

Table 1
Descriptive Statistics and Standardized Factor Loadings From CFA of the SEA2

Item % M SD ␤ SE

Economic Restriction 91 1.82 1.30 NA NA

Decide how you could spend money rather than letting you spend it how you saw fitb 74 2.10 1.55 .782 .084
Make you ask him or her for moneyb 73 2.13 1.58 .674 .091
Keep financial information from youa 67 2.03 1.69 .720 .095
Keep you from having the money you needed to buy food, clothing, or other necessities 63 1.72 1.58 .889 .080
Hide money so that you could not find ita 60 1.74 1.66 .809 .089
Demand that you give him/her receipts or change when you spent money 60 1.59 1.60 .816 .085
Keep you from having a job or going to work 52 1.41 1.57 .696 .089
Economic Exploitation 83 1.26 1.15 NA NA
Spend his/her money however he/she wanted while your money went to pay for necessitiesc 71 2.13 1.63 .672 .097
Make you use your money to buy him/her things or pay his/her bills when you didn’t want toc 54 1.41 1.54 .787 .085
Steal your property 54 1.36 1.50 .655 .089
Put bills in your name, leaving you to pay them 44 1.13 1.48 .730 .084
Force or pressure you to give him/her your savings or other assets 42 1.13 1.53 .739 .087
Make you take out a loan or buy something on credit when you didn’t want tod 37 .93 1.39 .688 .081
Take out a loan or buy something on credit in your name without your permissiond 29 .73 1.31 .778 .073
Note. SEA2 ⫽ Revised Scale of Economic Abuse; CFA ⫽ confirmatory factor analysis. Items rated on a 5-point Likert-type scale with response options
of 0 ⫽ never, 1 ⫽ hardly ever, 2 ⫽ sometimes, 3 ⫽ often, and 4 ⫽ very often. The percentage (%) reflects the percent who responded 1 (hardly ever) to
4 (very often). Descriptive statistics were calculated with raw data. Factor loadings and error terms were calculated with imputed data. All factor loadings
were statistically significant at p ⬍ .01. Shared superscripts indicate that the residuals for those items were allowed to covary in the final CFA model.
 274 ADAMS, GREESON, LITTWIN, AND JAVORKA

loadings ranged from .67 to .89 for Economic Restriction items associated with material dependence after controlling for eco-
and .66 to .79 for Economic Exploitation items (Table 1). nomic exploitation. Data were inspected for outliers, linearity,
Validity (Research Question 3). Construct validity was as- heteroscedasticity, normality of residuals, multicollinearity, and
sessed via the CFA, as well as additional correlation and regression influential cases. Data violated the assumptions of heteroscedas-
analyses. The CFA provides evidence of construct validity by ticity and normality of residuals. Therefore, we conducted boot-
demonstrating that the hypothesized two-factor structure was strapped regression using bias-corrected confidence intervals (CIs;
sound. The CFA also showed a moderately high positive correla- Wright, London, & Field, 2011). Contrary to our hypothesis,
tion between the economic restriction and economic exploitation higher scores on the SEA2 were not significantly associated with
latent variables, r ⫽ .76, p ⫽ .00, suggesting that they represent higher ratings of material dependence on the abuser (Regression 1;
distinct, but strongly related, constructs. Furthermore, bivariate b ⫽ .221, 95% CI [.003, .451], p ⫽ .05). However, consistent with
correlations that examined relationships between measures of eco- our hypothesis, we found that economic restriction was signifi-
nomic abuse (the SEA2 and the subscales) and other forms of cantly positively associated with material dependence after con-
abuse (i.e., physical and psychological abuse) provide evidence of trolling for economic exploitation (Regression 2; b ⫽ .545, 95%
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

discriminant validity. The moderate correlations (rs range from .48 CI [.335, .746], p ⫽ .00). We also found that after controlling for
This document is copyrighted by the American Psychological Association or one of its allied publishers.

to .68) suggest the SEA2 and its subscales are related to, but economic restriction, the unique variance left over in economic
distinct from, other forms of abuse (Table 2). exploitation was negatively associated with material dependence
Regression analyses were also conducted to assess convergent on the abuser (Regression 2; b ⫽ ⫺.362, 95% CI [⫺.597, ⫺.140],
validity by examining whether the full SEA2 and the Economic p ⫽ .00; Tables 3 and 4).
Restriction and Economic Exploitation subscales were associated Full SEA2 and Economic Exploitation subscale predicting
with financial outcomes with which we would expect them to outstanding debt. We expected that the full SEA2 would be
correlate: outstanding debt and material dependence on the abuser. significantly positively associated with outstanding debt (Hypoth-
We examined bivariate correlations between the outcome variables esis 3), and that economic exploitation would be significantly
and demographic variables (ever lived with the abuser, participant positively associated with outstanding debt after accounting for the
age, race/ethnicity, and number of children under 18). Demo- effects of economic restriction (Hypothesis 4). Therefore, the third
graphic variables that were significantly related to the outcome model (Regression 3) examined whether the SEA2 was associated
variables were included as control variables in the respective with outstanding debt, and the fourth model (Regression 4) exam-
analyses. All regressions also controlled for physical and psycho- ined whether economic exploitation was associated with outstand-
logical abuse. ing debt after controlling for economic restriction. Data were
Regressions were conducted in SPSS on the imputed data set. inspected for outliers, linearity, heteroscedasticity, normality of
Correlations between all variables used in the regressions are residuals, multicollinearity, and influential cases. Both regressions
presented in Table 2. Findings provide support for the convergent had cases with high leverage values, and therefore, models were
validity of the subscales and the full SEA2. rerun without high leverage cases. The pattern of statistically
Full SEA2 and Economic Restriction subscale predicting ma- significant results did not change, and therefore, all cases were
terial dependence. We expected that the SEA2 would be signif- retained in the analyses. All other assumptions were met. Results
icantly positively associated with material dependence (Hypothe- supported our hypotheses, providing evidence of the construct
sis 1), and that economic restriction would be significantly validity of the SEA2 and the Economic Exploitation subscale.
positively associated with material dependence after accounting Specifically, the SEA2 was significantly positively associated with
for the effects of economic exploitation (Hypothesis 2). Therefore, outstanding debt (Regression 3; b ⫽ .359, SE ⫽ .15, p ⫽ .01), and
the first regression model (Regression 1) examined whether the economic exploitation was significantly positively associated with
SEA2 was associated with material dependence, and the second outstanding debt after controlling for economic restriction (Re-
model (Regression 2) examined whether economic restriction was gression 4; b ⫽ .554, SE ⫽ .15, p ⫽ .00; Tables 5 and 6).

Table 2
Bivariate Correlations Among Variables Used to Assess the Construct Validity of the SEA2 and the Subscales

Variables 1 2 3 4 5 6 7 8 9 10

1. SEA2
2. SEA2-Restriction .928ⴱⴱ
3. SEA2-Exploitation .906ⴱⴱ .683ⴱⴱ
4. Outstanding debt .169ⴱⴱ .094 .224ⴱⴱ
5. Material dependence .187ⴱⴱ .290ⴱⴱ .038 .139ⴱ
6. Physical abuse .554ⴱⴱ .480ⴱⴱ .540ⴱⴱ .047 .050
7. Psychological abuse .620ⴱⴱ .632ⴱⴱ .499ⴱⴱ .086 .145ⴱ .516ⴱⴱ
8. Participant age .064 .119 ⫺.008 ⫺.007 .090 ⫺.160ⴱ ⫺.018
9. Number of children ⬍18 ⫺.094 ⫺.098 ⫺.073 .037 .007 ⫺.108 ⫺.097 ⫺.112
10. Race .006 .018 ⫺.009 .021 .143ⴱ ⫺.223ⴱⴱ .005 .121 ⫺.104
11. Ever lived with abuser .080 .071 .077 .044 .070 .118 .069 ⫺.116 .065 .084
Note. SEA2 ⫽ Revised Scale of Economic Abuse. Race was coded as 0 ⫽ racial/ethnic minority, 1 ⫽ White. Ever lived with abuser was coded as 0 ⫽
never lived with the abuser, 1 ⫽ did live with the abuser.
ⴱ
p ⬍ .05. ⴱⴱ p ⬍ .01.
 SEA2 275

Table 3 Table 5
Results of Bootstrapped Regression Analysis Examining SEA2 as Results of Regression Analysis Examining SEA2 as Predictor of
Predictor of Material Dependence on the Abuser Outstanding Debt

Variables b SE 95% bias-corrected CI p Variables b SE ␤ p

SEA2 .221 .113 .003, .451 .050 SEA2 .359 .145 .212 .014
Physical abuse ⫺.061 .097 ⫺.253, .133 .534 Physical abuse ⫺.107 .131 ⫺.064 .414
Psychological abuse .111 .164 ⫺.215, .404 .499 Psychological abuse ⫺.030 .207 ⫺.012 .885
Race .369 .189 ⫺.005, .742 .052
Note. SEA2 ⫽ Revised Scale of Economic Abuse. R2 ⫽ .032.
Note. SEA2 ⫽ Revised Scale of Economic Abuse; CI ⫽ confidence
interval. Race was coded as 0 ⫽ racial/ethnic minority, 1 ⫽ White. R2 ⫽
.058.
expected a significant relationship between the full SEA2 and
material dependence on the abuser (Hypothesis 1). The finding
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

showed a trend-level relationship (p ⫽ .050) between the two

This document is copyrighted by the American Psychological Association or one of its allied publishers.

Discussion variables. Examination of the relationships between the two sub-

The purpose of this study was to develop the SEA2, an updated scales and material dependence sheds some light on this finding.
version of the SEA (Adams et al., 2008). We developed an initial Bivariate correlations indicated that economic restriction was pos-
pool of items from the original SEA, the literature, and conversa- itively related to material dependence but showed no relationship
tions with survivors, advocates, and researchers. With data from between economic exploitation and material dependence. As hy-
248 women seeking services for IPV, we used an iterative con- pothesized, with both subscales in the regression model, we found
ceptual and statistical process to arrive at a 14-item scale with two that economic restriction was still significantly positively related
dimensions: economic restriction and economic exploitation. The to material dependence on the abuser after accounting for the
Economic Restriction subscale included seven items that capture effects of economic exploitation (Hypothesis 2). However, after
control tactics abusers use to impose limits on their partners’ removing the shared variance with economic restriction, economic
economic resources. The Economic Exploitation subscale con- exploitation was significantly negatively associated with material
sisted of seven items that capture ways that abusers exert control dependence, with more exploitation related to less material depen-
by using their partners’ economic resources for their own advan- dence on the abuser. Although not hypothesized, this finding is
tage. CFA supported the proposed two-factor structure (Research logical. If an abuser is exploiting their partner’s resources, the
Question 1). This means economic abuse tactics can be meaning- direction of the dependence is reversed—the abuser is relying on
fully categorized into two distinct dimensions, economic restric- the victim’s resources rather than the other way around. This
tion and economic exploitation, which belong to the higher order dynamic may explain why the full SEA2 is not significantly
economic abuse construct. The measures of internal consistency related to material dependence.
indicated that the full SEA2 and its subscales are reliable measures We also tested the construct validity of the SEA2 by examining
of economic abuse and its dimensions (Research Question 2). It is its relationship with outstanding debt as an indicator of resource
also important to note that, consistent with prior studies, this depletion. We expected that the full SEA2 would be significantly
research shows that economic abuse is associated with yet distinct positively related to outstanding debt after controlling for the
from physical and psychological forms of IPV (Adams et al., 2008; effects of physical and psychological abuse (Hypothesis 3). The
Stylianou et al., 2013). findings supported this hypothesis. The more economic abuse
Prior research shows that economic abuse is related to negative women experienced, the more types of debt they owed. We further
economic outcomes for victims (Adams et al., 2008, 2015; Hetling hypothesized that economic exploitation would be significantly
et al., 2015; Postmus et al., 2012). If the SEA2 is a measure of positively related to outstanding debt after controlling for eco-
economic abuse, then it should be related to economic variables. nomic restriction (Hypothesis 4). The findings supported this hy-
The current study provided initial evidence of the construct valid- pothesis. The more abusers used their partners’ economic re-
ity of the SEA2 and the two subscales (Research Question 3). We sources for their own advantage, the more types of debt their
participants owed.
Overall, these findings suggest that the SEA2 measures an
Table 4 economic dimension of IPV as intended. Further, these results
Results of Bootstrapped Regression Analysis Examining
Economic Restriction as Predictor of Material Dependence on
the Abuser Table 6
Results of Regression Analysis Examining Economic
Variables B SE 95% bias-corrected CI p Exploitation as Predictor of Outstanding Debt
Economic restriction .545 .097 .335, .746 .000 Variables b SE ␤ p
Economic exploitation ⫺.362 .115 ⫺.597, ⫺.140 .002
Physical abuse .020 .101 ⫺.183, .228 .841 Economic exploitation .554 .149 .334 .000
Psychological abuse ⫺.063 .163 ⫺.386, .228 .695 Economic restriction ⫺.165 .140 ⫺.113 .240
Race .381 .178 .032, .725 .032 Physical abuse ⫺.169 .131 ⫺.102 .196
Psychological abuse .108 .210 .043 .608
Note. Race was coded as 0 ⫽ racial/ethnic minority, 1 ⫽ White. R2 ⫽
.150. CI ⫽ confidence interval. Note. R2 ⫽ .063.
 276 ADAMS, GREESON, LITTWIN, AND JAVORKA

provide initial evidence that the two SEA2 subscales measure instrument to further our understanding of the correlates and
distinct forms of economic abuse, one capturing restriction of consequences of this distinct form of IPV. At this stage, lon-
economic resources and one capturing exploitation of economic gitudinal research on the cumulative, lasting, and/or rippling
resources. effects of economic abuse is of particular importance. The field
would also benefit from further psychometric work on the
SEA2. A limitation of the current study is that the CFA was
Limitations
conducted on the same sample that was used in item reduction.
The study findings need to be considered in light of its The CFA should be replicated on an independent sample to
limitations. First, the CFA was conducted on the same sample strengthen conclusions that can be drawn about the instrument
of data that was used to reduce the item pool. Future research and its factor structure. It would also be beneficial to use a
should continue to examine the factor structure and validity of longitudinal design to examine the predictive validity of the
the scale, particularly by attempting to replicate the factor scale and test whether the scale is biased for particular groups
structure within a new sample. Second, a number of sampling by examining measurement invariance based on demographic
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

issues affect the generalizability of the study findings. The variables such as gender, age, and ethnicity. In addition, re-
This document is copyrighted by the American Psychological Association or one of its allied publishers.

study participants were all seeking services for IPV in one searchers might take a closer look at the instrument’s two
Midwestern U.S. state. It is possible that the nature and effects subscales. Are there differences in who perpetrates or experi-
of economic abuse differ for women who have not sought ences one or both forms of economic abuse? What are the
formal help, or those who seek help from organizations in other differential effects of these two forms of economic abuse in
parts of the United States or elsewhere in the world. Also, all of victims’ lives? Any such research would advance the study of
the study participants were women and almost all were in this distinct form of IPV.
heterosexual relationships. We do not know how the SEA2
performs for IPV victims who do not identify as women or
Implications for Practice and Policy
whose abusive partners do not identify as men. Further, while
participants’ household incomes ranged from $0 to $50,000 or The methodological and substantive findings of this study have
more, a large proportion lived in households with a net annual important practice and policy implications. This research provides
income of $15,000 or less. Additional research is needed to further evidence that economic abuse is as common as physical
understand economic abuse perpetrated in households with and psychological abuse among women seeking help for IPV. It
greater economic resources. Finally, owing to resource con- also shows that abusive partners use a range of economically
straints that made translation services prohibitive, the sample abusive tactics to control their partners and that the SEA2 is a
was entirely English-speaking women. To be used with ethni- reliable and valid instrument to detect IPV involving economic
cally diverse samples, the SEA2 needs to be validated with IPV abuse. Practitioners in DV service organizations, legal services
victims who speak languages other than English. Although the agencies, financial counseling programs, and other service settings
sample was limited in these ways, it is important to note that the where people with abusive partners seek help could use the SEA2
demographics of this sample were similar to the demographics to assess the tactics and extent of economic abuse their clients
reported in a study with a national sample of help-seeking IPV experienced. This information could then be used to guide safety
survivors (Lyon, Bradshaw, & Menard, 2011). This suggests planning to reduce the risk of further economic abuse. As with all
that the SEA2 may be more widely generalizable to women safety planning, the process should be collaborative, ongoing, and
seeking help for IPV. focused on strategies that meet the victim’s self-defined needs and
minimize the risk of retaliation or other harms (Davies & Lyon,
2014). An economic abuse assessment could also help frame a
Implications for Research
conversation about ways to address damage stemming from past
With the development of SEA2, there are now three versions and ongoing abuse. This study indicates that material dependence
of the SEA. The SEA2 reflects conceptual and methodological on the abuser and owing debt are two potential consequences of
enhancements that have important implications for the construct economic abuse. For victims currently in a relationship with an
validity of the instrument. We improved the conceptual clarity abusive partner, the conversation might center on strategies for
of the subscales by recasting the “Economic Control” subscale increasing resources for independence from the abuser and safely
as “Economic Restriction.” Now, the subscales are linked more disputing and/or managing the debt when abuse is ongoing. The
clearly to the definition of economic abuse as a mechanism of conversation would be similar for victims no longer in a relation-
control and capture methods by which control is exerted: eco- ship with their abusive partner. Practitioners could help victims
nomic restriction and economic exploitation. In addition, to identify and address the lasting and rippling effects of material
ensure the instrument captures abusive tactics as opposed to dependence and safely dispute and/or manage their debt going
financial management behaviors, we included items with word- forward.
ing that directly reflect the victim’s lack of choice. We also Practitioners’ efforts would be aided by policy targeting eco-
included items that more accurately and discretely capture the nomic abuse. A first step is to recognize economic abuse as a form
use of coercion and fraud to generate debt in another person’s of IPV in legislation, such as the Violence Against Women Act in
name and excluded items that systematically did not apply to the United States. Definitions of the problem ought to have three
unemployed respondents. elements. They should reflect the function of economic abuse as a
The SEA2 appears to be a psychometrically sound instrument mechanism of control. They should capture key ways abusers exert
for measuring economic abuse. Researchers can also use this control, including through the restriction and exploitation of eco-
 SEA2 277

nomic resources. And they should identify common targets of that Dutton, M. A., Goodman, L. A., & Schmidt, R. J. (2005). Development and
control, including income, credit, assets, expenditures, and finan- validation of a coercive control measure for intimate partner violence:
cial information. A second step is to allocate funds specifically for Final technical report: National Institute of Justice, Office of Justice
services to prevent and help victims recover from economic abuse. Programs, US Department of Justice.
Funds could be directed to prepare generalist advocates to address Goodkind, J. R., Sullivan, C. M., & Bybee, D. I. (2004). A contextual
analysis of battered women’s safety planning. Violence Against Women,
economic abuse effectively, as well as support the work of advo-
10, 514 –533.
cates who specialize in economic issues. A third step is to ensure
Haj-Yahia, M. M. (2000). Implications of wife abuse and battering for
that legal remedies exist to combat economic abuse. For instance, self-esteem, depression, and anxiety as revealed by the Second Pales-
state/provincial personal protection order laws and local imple- tinian National Survey on Violence Against Women. Journal of Family
mentation of those laws ought to offer an avenue for financial Issues, 21, 435– 463. http://dx.doi.org/10.1177/019251300021004002
relief from economic abuse. States/provinces could adopt DV- Hamberger, L. K., Larsen, S. E., & Lehrner, A. (2017). Coercive control in
specific tort laws through which victims can get restitution for the intimate partner violence. Aggression and Violent Behavior, 37, 1–11.
abuse. Law-reform organizations could draft a uniform law to http://dx.doi.org/10.1016/j.avb.2017.08.003
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

address coerced debt for adoption by states/provinces. Targeted Hetling, A., Stylianou, A. M., & Postmus, J. L. (2015). Measuring financial
This document is copyrighted by the American Psychological Association or one of its allied publishers.

practice and policy responses such as these are warranted given the strain in the lives of survivors of intimate partner violence. Journal of
remarkably high rate at which economic abuse appears to occur Interpersonal Violence, 30, 1046 –1064. http://dx.doi.org/10.1177/
among service-seeking IPV victims and its detrimental effects on 0886260514539758
victims’ lives. Katz, E. (1997). The intra-household economics of voice and exit. Feminist
Economics, 3, 25– 46. http://dx.doi.org/10.1080/135457097338645
Littwin, A. (2012). Coerced debt: The role of consumer credit in domestic
References violence. California Law Review, 100, 1–74.
Lucente, S. W., Fals-Stewart, W., Richards, H. J., & Goscha, J. (2001).
Adams, A. E., & Beeble, M. A. (2018). Intimate partner violence and Factor structure and reliability of the Revised Conflict Tactics Scales for
psychological well-being: Examining the effect of economic abuse on incarcerated female substance abusers. Journal of Family Violence, 16,
women’s quality of life. Psychology of Violence. Advance online pub- 437– 450.
lication. http://dx.doi.org/10.1037/vio0000174 Lyon, E., Bradshaw, J., & Menard, A. (2011). Meeting survivors’ needs
Adams, A. E., Beeble, M. L., & Gregory, K. A. (2015). Evidence of the through non-residential domestic services and supports: Results of a
construct validity of the Scale of Economic Abuse. Violence and Vic- multi-state survey. Washington, DC: National Institute of Justice.
tims, 30, 363–376. http://dx.doi.org/10.1891/0886-6708.VV-D-13- Moe, A. M., & Bell, M. P. (2004). Abject economics: The effects of
00133 battering and violence on women’s work and employability. Violence
Adams, A. E., Littwin, A., & Javorka, M. (2019). The frequency, nature, Against Women, 10, 29 –55.
and effects of coerced debt among a national sample of women seeking Pence, E., & Paymar, M. (1986). Power and control: Tactics of men who
help for domestic violence. Violence Against Women. Advance online batter. Duluth, MN: Minnesota Program Development Inc.
publication. http://dx.doi.org/10.1177/1077801219841445 Postmus, J. L., Huang, C.-C., & Mathisen-Stylianou, A. (2012). The impact
Adams, A. E., Sullivan, C. M., Bybee, D., & Greeson, M. R. (2008).
of physical and economic abuse on maternal mental health and parent-
Development of the Scale of Economic Abuse. Violence Against
ing. Children and Youth Services Review, 34, 1922–1928.
Women, 14, 563–588. http://dx.doi.org/10.1177/1077801208315529
Postmus, J. L., Plummer, S. B., McMahon, S., Murshid, N. S., & Kim,
Anderson, M. A., Gillig, P. M., Sitaker, M., McCloskey, K., Malloy, K., &
M. S. (2012). Understanding economic abuse in the lives of survivors.
Grigsby, N. (2003). “Why doesn’t she just leave?” A descriptive study
Journal of Interpersonal Violence, 27, 411– 430. http://dx.doi.org/10
of victim reported impediments to her safety. Journal of Family Vio-
.1177/0886260511421669
lence, 18, 151–155. http://dx.doi.org/10.1023/A:1023564404773
Postmus, J. L., Plummer, S.-B., & Stylianou, A. M. (2016). Measuring
Black, M. C., Basile, K. C., Breiding, M. J., Smith, S. G., Walters, M. L.,
economic abuse in the lives of survivors: Revising the Scale of Eco-
Merrick, M. T., . . . Stevens, M. R. (2011). The National Intimate
nomic Abuse. Violence Against Women, 22, 692–703. http://dx.doi.org/
Partner and Sexual Violence Survey (NISVS): 2010 Summary Report.
10.1177/1077801215610012
Atlanta, GA: National Center for Injury Prevention and Control, Centers
for Disease Control and Prevention. Sanders, C. K. (2015). Economic abuse in the lives of women abused by an
Brewster, M. P. (2003). Power and control dynamics in prestalking and intimate partner: A qualitative study. Violence Against Women, 21,
stalking situations. Journal of Family Violence, 18, 207–217. http://dx 3–29. http://dx.doi.org/10.1177/1077801214564167
.doi.org/10.1023/A:1024064214054 Schafer, J. L., & Graham, J. W. (2002). Missing data: Our view of the state
Chowbey, P. (2017). Women’s narratives of economic abuse and financial of the art. Psychological Methods, 7, 147–177. http://dx.doi.org/10
strategies in Britain and South Asia. Psychology of Violence, 7, 459 – .1037/1082-989X.7.2.147
468. http://dx.doi.org/10.1037/vio0000110 Sharp, N. (2008). What’s yours is mine: The different forms of economic
Crossman, K. A., & Hardesty, J. L. (2018). Placing coercive control at the abuse and its impact on women and children experiencing domestic
center: What are the processes of coercive control and what makes violence. London, United Kingdom: Refuge.
control coercive? Psychology of Violence, 8, 196 –206. http://dx.doi.org/ Sharp-Jeffs, N., Kelly, L., & Klein, R. (2018). Long journeys toward
10.1037/vio0000094 freedom: The relationship between coercive control and space for
Davies, J. M., & Lyon, E. (2014). Domestic violence advocacy: Complex action-measurement and emerging evidence. Violence Against Women,
lives/difficult choices. Thousand Oaks, CA: Sage. http://dx.doi.org/10 24, 163–185. http://dx.doi.org/10.1177/1077801216686199
.4135/9781483352916 Shoener, S. (2016). The price of safety: Hidden costs and unintended
DeVellis, R. F. (2017). Scale development. New York, NY: Sage. consequences for women in the domestic violence service system. Nash-
Dutton, M. A., & Goodman, L. A. (2005). Coercion in intimate partner ville, KY: Vanderbilt University Press.
violence: Toward a new conceptualization. Sex Roles, 52, 743–756. Stark, E. (2007). Coercive control: The entrapment of women in personal
http://dx.doi.org/10.1007/s11199-005-4196-6 life. New York, NY: Oxford University Press.
 278 ADAMS, GREESON, LITTWIN, AND JAVORKA

Straus, M. A. (1979). Measuring intrafamily conflict and aggression: The U.S. Department of Justice. (2011). Office of Justice Programs fact sheet:
Conflict Tactics Scale (CTS). Journal of Marriage and the Family, 41, Domestic violence. Retrieved from https://ojp.gov/newsroom/factsheets/
75– 88. http://dx.doi.org/10.2307/351733 ojpfs_domesticviolence.html
Straus, M. A., & Mickey, E. L. (2012). Reliability, validity, and prevalence Voth Schrag, R. J. (2015). Economic abuse and later material hardship: Is
of partner violence measured by the conflict tactics scales in male- depression a mediator? Affilia, 30, 341–351. http://dx.doi.org/10.1177/
dominant nations. Aggression and violent behavior, 17, 463– 474. 0886109914541118
Stylianou, A. M., Postmus, J. L., & McMahon, S. (2013). Measuring Walker, L. E. (1979). The battered woman. New York, NY: Harper &
abusive behaviors: Is economic abuse a unique form of abuse? Journal Row.
of Interpersonal Violence, 28, 3186 –3204. http://dx.doi.org/10.1177/ Wolf, E. J., Harrington, K. M., Clark, S. L., & Miller, M. W. (2013). Sample
0886260513496904 size requirements for structural equation models: An evaluation of power,
Sullivan, C. M., Basta, J., Tan, C., & Davidson, W. S., II. (1992). After the crisis: bias, and solution propriety. Educational and Psychological Measurement,
A needs assessment of women leaving a domestic violence shelter. Violence 73, 913–934. http://dx.doi.org/10.1177/0013164413495237
and Victims, 7, 267–275. http://dx.doi.org/10.1891/0886-6708.7.3.267 Wright, D. B., London, K., & Field, A. P. (2011). Using bootstrap esti-
Tolman, R. M. (1999). The validation of the Psychological Maltreatment of mation and the plug-in principle for clinical psychology data. Journal of
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

Women Inventory. Violence and Victims, 14, 25–37. http://dx.doi.org/ Experimental Psychopathology, 2, 252–270. http://dx.doi.org/10.5127/
10.1891/0886-6708.14.1.25 jep.013611
This document is copyrighted by the American Psychological Association or one of its allied publishers.

Appendix
The Revised Scale of Economic Abuse (SEA2)

0 1 2 3 4
Using the 0–4 scale below, during your relationship, how often did your Hardly Very
partner do the following: Never ever Sometimes Often often

1. Keep you from having the money you needed to buy food, clothes, or other necessities e e e e e
2. Keep financial information from you e e e e e
3. Decide how you could spend money rather than letting you spend it how you saw fit e e e e e
4. Make you ask him/her for money e e e e e
5. Hide money so that you could not find it e e e e e
6. Demand that you give him/her receipts or change when you spent money e e e e e
7. Keep you from having a job or going to work e e e e e
8. Make you use your money to buy him/her things or pay his/her bills when you didn’t want to e e e e e
9. Spend his/her money however he/she wanted while your money went to pay for necessities e e e e e
10. Take out a loan or buy something on credit in your name without your permission e e e e e
11. Make you take out a loan or buy something on credit when you didn’t want to e e e e e
12. Put bills in your name, leaving you to pay them e e e e e
13. Force or pressure you to give him/her your savings or other assets e e e e e
14. Steal your property e e e e e
Note. Researchers and practitioners should contact the lead author for permission to use the SEA2.

Received September 24, 2018

Revision received April 12, 2019
Accepted April 14, 2019 䡲