EXAMINING ECONOMIC ABUSE AND RELEVANT PSYCHOSOCIAL FACTORS AMONG UNSTABLY HOUSED DOMESTIC VIOLENCE SURVIVORS By Jasmine Engleton A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of Psychology – Master of Arts 2020 EXAMINING ECONOMIC ABUSE AND RELEVANT PSYCHOSOCIAL FACTORS AMONG UNSTABLY HOUSED DOMESTIC VIOLENCE SURVIVORS ABSTRACT By Jasmine Engleton Objective: To assess if economic abuse, race, citizenship status, and criminal record are associated with severe housing instability among unstably housed domestic violence (DV) survivors. Data Source: Secondary data from the Domestic Violence Housing First (DVHF) Demonstration Evaluation, a longitudinal evaluation that assessed how mobile advocacy and flexible funding leads to desired outcomes for DV survivors and their children. Participants: Data from 392 unstably housed adult, female DV survivors. Methods: A cumulative ordinal logistic regression model series was conducted to determine if economic abuse, race, citizenship status, and criminal record were associated with severe housing instability. Results: DV survivors who had a criminal record were more likely to experience severe housing instability than were DV survivors without a criminal record. Contrary to my hypotheses, DV survivors who were U.S citizens were more likely to experience severe housing instability compared to non-citizens. Economic abuse was not associated with severe housing instability and racial differences were not evident. Conclusion: Overall, this study supported only one hypothesis – having a criminal record was associated with experiencing severe housing instability. Future research should focus on types of criminal records that contribute to severe housing instability. In addition, implications for policy suggest strengthening anti-discriminatory housing laws and conducting a holistic assessment of DV survivors with a criminal record who need housing. Furthermore, this study highlighted the need for a more sensitive measure of housing instability. ii This thesis is dedicated to my late mother, Deidre Engleton. She is my motivation to see an end to gender-based violence. iii ACKNOWLEDGEMENTS I would like to express my deepest gratitude to everyone who has contributed to making this thesis a final product. First, I would like to thank my partner and close friends – Michael, Roxana, and Erica– for your constant support, constructive feedback, and encouragement. I would also like to thank my thesis committee, Dr. Cris Sullivan, Dr. NiCole Buchanan, and Dr. Amy Drahota for your guidance, insight, and expertise throughout this process. Cris, I would like to recognize the invaluable assistance that you provided while making this thesis. Thank you for your professional guidance, patience, and being an exceptionally supportive mentor throughout this process. Finally, I would also like to thank my community of friends and peers at MSU and University of Maryland Baltimore County (UMBC) who have supported and guided me throughout my graduate experience. Thank you for listening and supporting my ideas, growth, and rallying behind every one of my accomplishments throughout the process of completing my thesis. iv TABLE OF CONTENTS LIST OF TABLES ......................................................................................................................... vi INTRODUCTION ...........................................................................................................................1 LITERATURE REVIEW .................................................................................................................4 Housing Instability and DV Survivors ........................................................................................5 The relationship between domestic violence and housing instability ...................................5 Economic abuse and housing instability………………..…………………………………7 Marginalized Identities and Housing Instability: Intersectional Perspective……..………………………………………………………………………….9 Race and housing instability .....................................................................................10 Citizenship status and housing instability .................................................................11 Criminal record and housing instability ....................................................................13 Housing instability & overlapping marginalizing identities ......................................14 CURRENT STUDY .......................................................................................................................15 Research Hypotheses ................................................................................................................15 METHODS ....................................................................................................................................17 Recruitment Process..................................................................................................................17 Data Collection Process ...........................................................................................................17 Analytic Sample .......................................................................................................................18 Measures ...................................................................................................................................18 Analytic Approach ....................................................................................................................21 RESULTS .................................................................................................................................................. 24 Demographic Characteristics .............................................................................................................. 24 Control Variables ................................................................................................................................. 24 Hypothesis Testing .............................................................................................................................. 26 DISCUSSION ................................................................................................................................34 Limitations ................................................................................................................................38 Implications for Future Research ..............................................................................................39 Implications for Practice and Policy .........................................................................................41 Conclusion ................................................................................................................................42 APPENDIX ....................................................................................................................................43 REFERENCES ..............................................................................................................................46 v LIST OF TABLES Table 1 Descriptive characteristics of housing instability index ………………………….……23 Table 2 Inter-item correlation matrix of the 3-item housing instability index……………...…...23 Table 3 Sample socio-demographics (N=392) ……….….………………….….…………........25 Table 4 Correlation matrix of dependent and independent variables…………………………. 26 Table 5 Collinearity diagnostic of dependent and independent variables……………...……....28 Table 6 Cumulative logistic regression of severe housing instability by economic abuse……..27 Table 7 Cumulative logistic regression of severe housing instability by race (People of Color)………………….………………………………………………………………………...29 Table 8 Cumulative logistic regression of severe housing instability by race (Black)……….. .29 Table 9 Cumulative logistic regression of severe housing instability by race (Latinx)………...30 Table 10 Cumulative logistic regression of severe housing instability by citizenship status…...31 Table 11 Cross-tabulation of housing instability by citizenship status (N=389)…..…….......….31 Table 12 Cumulative logistic regression of severe housing instability by criminal record……. 32 Table 13 Moderated cumulative logistic regression of severe housing instability by race and criminal record………………….…………………………………………………………….…33 vi INTRODUCTION Housing instability is a pervasive issue affecting millions of people in the U.S (Joint Center for Housing Studies of Harvard University, 2019). Housing instability can be broadly defined as the extent to which one can maintain or obtain safe and stable housing. Experiencing housing instability encompasses varied forms, including being unable to afford rent, spending more than 50 percent of one’s income on housing, having to move frequently, living in substandard or dangerous housing, experiencing overcrowding, homelessness, risk of eviction, and facing severe landlord disputes (Adams et al., 2018; Brisson & Covert, 2015; Gilroy et al., 2016; Dichter et al., 2017; Breiding et al., 2017; Montgomery et al., 2018). People experience numerous detrimental outcomes as a result of housing instability (Desmond & Gershenson, 2016; Desmond & Kimbro, 2015; Niccolai et al., 2019; Pollack et al., 2010; Reid et al., 2008). Housing instability increases the likelihood that one will experience job loss and other financial consequences (Desmond & Gershenson, 2016). People experiencing housing instability may not be able to afford food, transportation, or healthcare-related costs (Joint Center for Housing Studies of Harvard University, 2019; Reid et al., 2008). Along with financial consequences, unstably housed people experience physical health problems such as hypertension, arthritis, and increased risk for sexually-transmitted infections at higher rates than do individuals who are stably housed (Niccolai et. al., 2019; Pollack et al.,2010). DV survivors are especially vulnerable to experiencing housing instability (Pavao et al., 2008). Previous research has shown that homelessness is particularly high among female DV survivors, the most severe form of housing instability (Baker et al., 2003; United States Conference of Mayors, 2005; Jasinski et al., 2009). 1 DV is the leading predictor of housing instability for DV survivors (Jasinski et al., 2005; Levin et al., 2004; Tessler et al., 2001). One form of DV, economic abuse, may be relevant for understanding housing instability among DV survivors; however, to date, economic abuse has been understudied (Adams et al., 2018; Breiding et al., 2017; Dichter et al., 2017; Pavao et al., 2007; Ponic et al., 2011). Economic abuse occurs when the abuser intentionally exploits, restricts, and controls survivors’ financial resources (Adams, 2008). Economic abuse may diminish survivors’ job stability and financial security, reducing their ability to secure safe and stable housing (Adams, 2008; Adams et al., 2019; Crowne et al., 2011; Kimberling et al., 2009). However, the link between economic abuse and housing instability has not been extensively explored. Moreover, housing instability may not be experienced by all domestic violence survivors equally. Scant research has attended to how survivors with multiple marginalized identities - People of Color, non-citizens, and having a criminal record - may experience housing instability differently (Adams et al., 2018; Baker et al., 2010; Pavao et al., 2007: Gezinski et al., 2019). Studies consistently show that, in the general population, disparities of housing instability exist among People of Color, non-citizens, and people with a criminal record (Chang, 2019; Desmond, 2016; Evans & Porter, 2014; Evans et al., 2018; Lyubansky et al., 2013; Malone, 2009; Maykovichl et al., 2018; Phinney et al., 2007). Further, People of Color who have a criminal record show a disparate experience of housing instability (Lucius, 2018; McConnell, 2013; Olivet et al., 2018). These patterns may be similar among DV survivors; yet these relationships have not been explored. This study addressed these critical gaps by investigating: 1) whether economic abuse contributes to increased risk of housing instability over and above other forms of intimate partner violence, 2) whether housing instability differs by race, citizenship, and criminal record, and 3) 2 whether housing instability differs based on having overlapping marginalized identities (e.g., having a criminal record and being African American or Latinx) among a population of unstably housed DV survivors. It is essential for DV agencies and advocates to understand the contribution of economic abuse to housing instability and the additional factors that increase survivors’ vulnerability to housing instability. This understanding can aid them in tailoring intervention efforts to support survivors in attaining safe and stable housing. 3 LITERATURE REVIEW Housing instability is a widespread issue impacting millions of people in the U.S (Joint Center for Housing Studies of Harvard University, 2019; Mahathey et al., 2018). Housing instability can be broadly defined as the inability to maintain or obtain housing and may consist of multiple forms, such as the inability to afford rent, spending a large percentage of income on housing, having to move frequently, living in substandard or dangerous housing, experiencing overcrowding, risk of homelessness and eviction, and severe landlord disputes (Adams et al., 2018; Brisson & Covert, 2015; Gilroy et al., 2016; Breiding et al., 2017; Rollins et al., 2016; Routhier, 2019). Research points to increasing housing costs and lack of affordable housing as major contributors to housing instability (Joint Center for Housing Studies of Harvard University, 2019; Ellen & Karfunkel, 2016; Aurand et al., 2020). In the U.S, current housing trends place lower-income earners at a disadvantage. Specifically, newly constructed rental properties are priced at or above $2050 in urban areas and $1300 in midwestern areas (Joint Center for Housing Studies of Harvard University, 2019). Although individuals with a high income can afford this increase, lower-income individuals become overwhelmed with substantial housing cost burden, affecting over 37 million people (Joint Center for Housing Studies of Harvard University, 2019). Some government-funded options have addressed this concern -- such as creating public housing which requires recipients to only pay 30% or less of their income for housing (Mahathey et al., 2018). However, a national report indicates only 7.3 million affordable housing units are currently available, leaving over 30 million people in need of housing (Aurund et al.., 2020). Housing cost burden and lack of affordable housing places individuals at increased risk for eviction and homelessness (Desmond, 2016; Mahathey et al., 2018). For example, 2.3 million 4 households are evicted every year due to non-payment or late payment of rent (Desmond, 2016). Additionally, slightly over half a million people are currently experiencing homelessness (Mahathey et al., 2018). Housing Instability and DV Survivors A significant population experiencing housing instability in the U.S. are female DV survivors (Baker et al., 2003; United States Conference of Mayors, 2009; Jasinski et al., 2005; Pavao et al., 2007). DV is defined as a pattern of abuse that consists of psychological, physical, or sexual abuse, controlling behaviors, and economic actions done by one partner to maintain power and control over the other (National Coalition Against Domestic Violence, 2015). DV is a serious public health issue that impacts women at higher rates than men in the United States (Smith et al., 2018). DV negatively affects women's physical and psychological wellbeing, and economic stability (Breiding et al., 2014; Peterson et al., 2018; Truman & Morgan, 2014; WHO, 2013). Along with a host of other negative consequences, abuse has been recognized as a leading contributor to housing instability for DV survivors (Tessler et al., 2001; Levin et al., 2004; Jasinski et al., 2005; Pavao et al., 2007). One national report on homelessness among women in four states found that 20-50% were homeless due to DV (Jasinski et al., 2005). Another study on housing instability found that women who experienced violence were four times as likely to be unstably housed compared to women who had not experienced violence in their relationships (Pavao et al., 2007). Considering this, it is important to explore how DV contributes to housing instability, and whether its impact differs by the type of violence they experience as well as the social location of survivors. The relationship between domestic violence and housing instability. DV can both directly and indirectly lead to survivors being unable to maintain or obtain safe and stable housing. For instance, many survivors report losing their homes or becoming homeless after 5 leaving an abusive partner (Baker et al., 2003; Janiski et al., 2005; O’Campo et al., 2016; Thomas et al., 2015). In a mixed-methods study about safety-related tradeoffs for 301 DV survivors, more than half of the sample reported having to give up “a lot” when trying to leave their partner, and about 20 percent reported losing their homes as a consequence of leaving their partners (Thomas et al., 2015). In addition to losing one’s home and experiencing bouts of homelessness, survivors may have to constantly move to evade the abuser, who may engage in stalking, harassment, violence, or even potentially lethal actions (Smith et al., 2014; Logan et al., 2008). Indirectly, DV can contribute to housing instability through the abuse’s impact on the survivor’s employment (Adams et al., 2012; Crowne et al., 2011; Jategaonkar & Ponic, 2011; Kimerling et al., 2009; Moe & Bell, 2004). Many survivors report that DV resulted in a reduced number of hours worked and job loss (Crowne et al., 2011; Moe & Bell, 2004; Showalter, 2016). A longitudinal study on low-income women’s job stability, experience with DV, and economic wellbeing found that DV contributed to women working about three fewer months at their jobs compared to women who had not experienced DV in the last two years (Adams et al., 2012). In Showalter's (2016)’s systematic literature review of employment and DV, seven studies showed that DV was associated with job loss and nine studies showed that DV reduced employment stability over time. Two longitudinal studies that focused on the relationship between employment and DV showed that DV was associated with job loss, working fewer hours, and unstable employment overtime (Crowne et al, 2011; Meisel et al., 2003). In addition, DV can affect survivors’ mental and physical wellbeing, which can then impact their employment opportunities, which can lead to housing stability(Kimerling et al., 2009; National Center for Injury Prevention, 2003). In a qualitative study of 19 women dealing with DV, women reported leaving their jobs due to the physical injuries and emotional health 6 problems stemming from the abuse (Moe & Bell, 2004). Employment instability and detrimental health consequences stemming from DV may make it increasingly hard for survivors to maintain their current housing or find new housing (Jategaonkar & Ponic, 2011). DV can also indirectly lead to housing instability by triggering eviction and housing denial (Arnold, 2019; Desmond, 2016, 191-192; Stern et al., 2007). For instance, nuisance ordinance laws function to hold landlords responsible for excessive 911 calls to rental properties (Arnold, 2019). Nuisance laws are used to curb crime, but often come with devastating consequences for survivors if the police are called in response to suspected abuse (Arnold, 2019). Studies have found that survivors living in rental or public housing end up being threatened with eviction or become evicted due to the abuser engaging in criminal activity, including domestic violence (Arnold, 2019; Desmond, 2016, p.290). For instance, survivors who call the police or have the police called on them are sometimes evicted from their homes due to the violence (Arnold, 2019; Desmond, 2016, p.290). In one national report on survivors' experience of housing denial and evictions in public and subsidized housing, 11% of evictions and 28% of housing denials occurred due to the violence they had experienced (Stern et al., 2007). Survivors are thus denied housing and even removed from their current housing due to experiencing violence in their relationships. Economic abuse and housing instability. Though there is growing understanding that DV can, directly and indirectly, contribute to housing instability, less is known about whether certain forms of DV, such as economic abuse, may be more or less likely to impact housing instability. A significant portion of the literature on housing instability and DV has focused solely on physical, psychological, emotional, and sexual abuse, but have excluded economic abuse (Adams et al., 2018; Breiding et al., 2017; Dichter et al., 2017; Pavao et al., 2007; Ponic et al., 2011). 7 Economic abuse may be related to increased housing instability among DV survivors due to its direct and negative impact on survivors’ financial stability (Postmus et al., 2012). Economic abuse occurs when abusers intentionally seek to control survivors' ability to obtain and maintain financial resources (Adams, 2008). Abusers may control DV survivors through coerced debt, restricting their use of financial resources, and employment sabotage (Adams et al., 2019; Voth Schrag, 2015). One longitudinal study on economic abuse and later material hardship found that economic abuse was associated with 66% increased odds of experiencing material hardship later in life (Voth Schrag, 2015). Resulting financial instability can make it increasingly difficult for survivors to maintain or obtain safe and stable housing (O'Campo et al., 2016; Voth Schrag, 2015). Abusers may create debt in the survivors’ names either fraudulently or by force, referred to as coerced debt (Littwin, 2012). In Adams and colleagues’ (2019) study of DV survivors calling the national helpline, over half of the women in the research had experienced coerced debt, and those who had experienced it were six times more likely to have their credit damaged. These women were also more than two times as likely to stay with abusers longer due to having more financial obstacles (Adams et al., 2019). Damaged credit and the accumulation of more financial hurdles can make it difficult for survivors to obtain and maintain safe, stable housing. In one qualitative study of 45 women experiencing DV, coerced debt was reported as a reason for experiencing housing instability (O'Campo et al., 2016). Abusers can also restrict survivors' use of money (Adams et al., 2008; Sanders, 2015; Voth Schrag, 2015). Postmus and colleagues’ (2012) study on the experience of economic abuse among survivors showed that a majority of the women had experienced abusers controlling how they used their money and monitored their usage. In a qualitative study of 30 survivors' experiences of economic abuse, women reported that they had to get permission from 8 their partners to pay for necessities and had to give their money to their partners (Sanders, 2015). Abusers' ability to control how survivors use their money may, then, ultimately impact their ability to find new housing or maintain their housing. Additionally, economic abuse can contribute to housing instability through employment sabotage (Adams et al., 2013; Crowne et al., 2011). Abusers may keep women from working, or stalk or harass them at work, resulting in their working fewer hours, becoming unemployed, or quitting (Adams et al., 2012; Meisel et al., 2003; Moe & Bell, 2004; Riger et al., 2004 ). In a qualitative study of 19 DV survivors living in a shelter, many reported being forced not to work (Moe & Bell, 2004). Four studies in one systematic literature review showed that women experiencing harassment at their job or workplace disruption lost time from work (Showalter, 2016). Without employment and consistent income, survivors may find it hard to afford to pay for their housing and this can result in their staying with their abusive partners or becoming homeless (Adams et al., 2019; Moe & Bell, 2004; O’Campo et al., 2016). Prior literature supports the relationship between economic abuse and decreased financial stability, but there are few studies on the relationship between economic abuse and housing instability. Those that have examined this connection to date have been qualitative (Clough et al., 2014; O’Campo et al., 2016); clearly, more research is needed that examines if economic abuse contributes to a greater extent of housing instability above other forms of violence. Marginalized Identities and Housing Instability: Intersectional Perspective When assessing how DV contributes to housing instability, it is important to understand that domestic violence survivors are not a monolithic group. Differences across race, ethnicity, gender, and other factors all impact their experience of housing instability (Crenshaw, 1991; West, 2004). It is therefore important to take an intersectional approach that attends to these 9 differences by considering how inequities -- in this case, housing instability -- may manifest differently due to forces housed within historical, political, material, and social contexts (Cole, 2009). The literature on general homelessness and risk of eviction has found that People of Color, non-citizens, and people with criminal records experience greater instances of housing instability than do White people, US citizens, and those without criminal records (Chang, 2019; Desmond, 2012; Evans & Porter, 2014; Evans et al., 2019; Lyubansky et al., 2013; Malone, 2009; Maykovichl et al., 2018; Phinney et al., 2007; Pavao et al., 2007). One might expect, then, that similar patterns may be evident among DV survivors with similar marginalized backgrounds; however, this has yet to be examined among a population of DV survivors (Barata & Stewart, 2010; Clough et al., 2014; Gezinski et al 2019). Furthermore, these marginalized identities often overlap with one another and may increase a survivor’s risk of experiencing housing instability. The following sections briefly describe what prior research has shown to be the interrelationships among race, citizenship, criminal record, and housing instability among the general population. Race and housing instability. In the U.S, People of Color tend to experience higher rates of housing cost burden, eviction, and homelessness compared to their White counterparts (Brisson & Covert, 2015; Desmond, 2012; Desmond & Shollenberger, 2015; Jones, 2016; Maykovichl et al., 2018;Olivet et al., 2018). In a nationally representative report on housing instability in the U.S., approximately 55% of African Americans and 54% of Latina households reported experiencing housing-cost burden in comparison to 43% of White households with similar income (Joint Center for Housing Studies of Harvard University, 2019). Therefore, People of Color may be increasingly vulnerable to experiencing housing instability because of these disparities in housing cost burden. Not only are People of Color cost-burdened more than 10 their White counterparts, they also experience more instances of lease violations that can lead to eviction and homelessness (Brisson & Covert, 2015; Desmond, 2012; Desmond & Shollenberger, 2015; Jones, 2016; Maykovichl et al., 2018; Olivet et al., 2018). One study found that subsidized housing recipients identifying as Black, Latina, or “other” racial identities were more likely to experience lease violations of late or nonpayment of rent than were White recipients (Brisson & Covert, 2015). Just as important, Latina and African American populations are evicted more often than their White counterparts (Desmond, 2012). Specifically, Black women were more than two times as likely to be evicted than Black men and three times more likely than White women (Desmond, 2012). Similarly, a report in Seattle found that Women of Color were more likely to be evicted than were White women even though they made up a smaller percentage of tenants (Maykovichl et al., 2018). Similar racial trends also occur among the homeless population. A national report by the Center for Social Innovation found that African Americans and Native Americans were overrepresented in the homeless population compared to White and Latina populations (Olivet et al., 2018), even after controlling for income. These studies indicate drastic racial differences among those who are stably housed, unstably housed, or homeless. Citizenship status and housing instability. Rates of housing instability also differ by citizenship status. Specifically, non-U.S citizens experience higher housing cost burden, are more likely to live in poor quality neighborhoods, and experience homelessness more than U.S citizens (Chavez, 2012; Cort et al., 2014; McConnell, 2013; Hall & Greenman, 2013). Further, undocumented individuals experience more housing instability than do U.S citizen and their documented counterparts (Chavez, 2012; Cort et al., 2014; McConnell, 2013; Hall & Greenman, 2013). In one study examining housing affordability, race, and legal status, undocumented Latinas were almost two times as likely to be burdened by housing costs in comparison to U.S 11 citizens (McConnell, 2013). Another study found that undocumented immigrants were less likely to own homes and were more likely to live in overcrowded conditions and live in poor-quality neighborhoods than were documented immigrants (Hall & Greenman, 2013). A similar study found that undocumented immigrants were more likely to be evicted due to the inability to pay rent and to become homeless than were legal residents and citizens (Chavez, 2012). Studies on housing instability and homelessness among undocumented populations demonstrate how poverty, immigration policies, and housing discrimination contribute to these disproportionate rates (Chavez, 2012; Chaudry et al., 2010; Hall et al., 2010; Lyubansky et al., 2013). For example, one longitudinal study on legal status and wage disparities found undocumented Latina immigrants made significantly less income than their documented counterparts (Hall et al., 2010). Specifically, undocumented women made 9% less than documented women, and undocumented men made 17% less than documented men (Hall et al., 2010). Thus, undocumented immigrants have disproportionately less income than their documented counterparts, which can contribute to their housing instability. Immigration policies also contribute to these differences in housing instability, specifically impacting undocumented immigrants who have been targeted and arrested or put in detention centers (Chaudry et al., 2010). A report on the impact of immigration enforcement found that undocumented immigrants who have been arrested are more likely to lose their homes, experience income hardship, and even become homeless than those who are documented (Chaudry et al., 2010). Housing discrimination is also a contributor to housing instability, as undocumented immigrants experience more discrimination when trying to rent an apartment or buy a home (Lyubansky et al., 2013). Thus, there are various structural dynamics such as poverty, immigration policies, and housing discrimination, that disparately affect undocumented people and place them at greater risk of experiencing housing instability. 12 Though the documented population tends to fare better than undocumented counterparts, individuals with lawful status still experience greater instances of housing instability than do U.S citizens (Hernandez et al., 2016). In one study on rental burden and energy insecurity among native-born and immigrant low-income families, rental burden was consistently higher among all immigrants in comparison to their native-born counterparts despite racial differences (e.g. White, Latina and Black; Hernandez et al., 2016). Criminal record and housing instability. In addition to race and citizenship status, housing instability differs by the presence of having a criminal record. Those with a criminal record are denied housing more frequently than are people without a criminal record (Evans, 2016; Evans, 2019; Human Rights Watch, 2014; Malone, 2009). One national report found that people with a criminal record and who were formerly incarcerated were 10 times more likely to be homeless than were people in the general public (Lucius, 2018). In addition, formerly incarcerated women were found to be more likely to become homeless than were formerly incarcerated men (Lucius, 2018). One study found that having any criminal record, regardless of charge, was associated with rental agents being less likely to assist clients with rental properties (Evan et al., 2019). Evan and colleagues (2016) found that having a criminal record of molestation was associated with rental agents being less likely to consider an applicant for an apartment. One national report on different public housing authorities found that people with 15-year-old criminal charges were still being denied housing (Human Rights Watch, 2014). Though having a criminal record may impact housing outcomes, the pathway through which having a criminal record leads to housing instability appears to be primarily through housing discrimination and poverty (Olivet et al., 2018; Lucius, 2018). For instance, people with a criminal record may be denied vouchers or be denied certain types of public housing (Olivet et al., 2018; Lucius, 2018). For people with a 13 criminal record, poverty also contributes to their experience of housing instability where they may have limited financial resources to afford housing (Valles & Dietrich, 2014). Housing instability & overlapping marginalizing identities. Not only do people with these marginalized identities experience disparities in housing instability but having more than one marginalized identity may exacerbate their experience of housing instability (McConnell, 2013; Olivet et al., 2018). People of Color who have a criminal record experience higher rates of housing instability in comparison to People of Color without a criminal record (Lucius, 2018). In a report on homelessness among formerly incarcerated individuals, Black and Hispanic people experience more homelessness than their White counterparts (Lucius, 2018). In another mixed methods study on homelessness among people in the general population, qualitative data revealed that being involved with the criminal justice system was a contributor to disparities in homeless rates among People of Color, who are often overrepresented in the criminal justice system (Carson, 2015; Olivet et al., 2018). Given that prior studies have found that People of Color and people with a criminal record have increased vulnerability to experience housing instability, as well as the overlapping effects of race and having a criminal record on housing instability among the general population, it is reasonable to expect that these relationships may also be important to examine with DV survivors. 14 CURRENT STUDY While there has been limited research examining DV survivors’ experiences of housing instability, many areas continue to be unaddressed. For example, the relationship between housing instability and commonly measured forms of DV (i.e. physical, sexual, psychological, and harassment) has been established (Adams et al., 2018; Breiding et al., 2017; Dichter et al., 2017; Pavao et al., 2007; Ponic et al., 2011), yet the relationship between economic abuse and housing instability has not. Understanding whether economic abuse contributes to survivors’ experience of housing instability is important for researchers, practitioners, and policy makers. In addition, numerous psychosocial factors are understudied in their relation to housing instability among DV survivors. While the connection among race, citizenship, criminal record and housing instability have been studied within the general population (Chang, 2019; Desmond, 2012; Evans & Porter, 2014; Evans et al., 2019; Lyubansky et al., 2013; Malone, 2009; Maykovichl et al., 2018; Phinney et al., 2007), there are no studies on how housing instability may look different for non-citizens who are DV survivors, DV survivors with a criminal record, and how People of Color with a criminal record may have increased risk of experiencing housing instability (Olivet et al., 2018). Although, there was one study that assessed if People of Color who are DV survivors experienced more housing instability than their White Counterparts (Adams et al., 2018), limitations were noted. Housing instability may be differentially impacted for survivors from these marginalized backgrounds and this needs to be understood. Research Hypotheses The current study assessed the extent to which severe housing instability was associated with economic abuse and specific marginalized identities among a sample of unstably housed survivors. The following hypotheses were addressed in the study after controlling for financial 15 difficulties and other types of domestic violence (e.g. physical, sexual, emotional/psychological abuse, and stalking/harassment): H1: Higher economic abuse will be associated with greater housing instability. H2: Survivors of Color (specifically Black and Latina survivors) will report greater housing instability than will their White counterparts. H3: Survivors who are non-citizens will report greater housing instability than will survivors with citizenship. H4: Survivors with a criminal record will report greater housing instability than will survivors without a criminal record. H5: Having a criminal record will significantly moderate the association between race and housing instability. Specifically, the association will be stronger for Black and Latina survivors who have a criminal record and attenuated for Black and Latina survivors who do not have a criminal record. . 16 METHODS Data for this study came from the DV Housing First (DVHF) Demonstration Evaluation, a longitudinal evaluation conducted by Dr. Cris Sullivan. The study was designed to extensively evaluate how mobile advocacy and flexible funding lead to desired outcomes for DV survivors and their children. The current study used data gathered at the baseline time point when homeless and unstably housed DV survivors reached out for services. IRB approval was received at MSU before recruitment and data collection. Further, IRB approval for this thesis was received before any secondary analyses were conducted. Recruitment Process Survivors were recruited from five DV agencies in King County and South Central Washington. Site coordinators and advocates worked together to determine the survivors’ eligibility for the study. Eligibility criteria were that survivors: 1) must recently have sought services at one of the five DV agencies, 2) had to be 18 or older, and 3) had to be homeless or experiencing housing instability. Staff provided coordinators with information about potentially eligible clients after receiving permission from the client to do so. Site coordinators then reached out to survivors to determine eligibility and interest in the study. Data Collection Process Baseline data collection of survivors consisted of face-to-face interviews conducted by trained interviewers. The baseline interview consisted of over 100 questions focused on experiences of violence, housing instability, history of homelessness, housing barriers, participants’ characteristics, and needs. Participants were given $50 to thank them for their time and participation. Baseline interviews lasted from 39 minutes to 3 hours and 13 minutes, averaging 1 hour and 14 minutes, and were recorded to check for accuracy. 17 Analytic Sample At baseline, 409 participants enrolled in the study. For the current study, 392 cis- gendered women (95.8%) were included in the final analytic sample. Although the overall study included men and transgendered individuals, there was not a large enough sample to conduct gender group differences.. Measures The measures included in the current study were housing instability, experiences of violence (physical, sexual, emotional, psychological, stalking/harassment, and economic abuse), financial difficulties, race, citizenship status, and criminal record. Housing Instability. To assess housing instability, a composite index consisting of three items was created, following Trochim & Donnelly’s (2001) steps on creating an index. An index was utilized instead of single item indicators that only capture some aspect of housing instability (Brisson & Covert, 2015; Gilroy et al., 2016; Dichter et al., 2017; Breiding et al., 2017; Montgomery et al., 2018; Adams et al., 2018). The three-item index assessed the frequency of moves in a six-month period, the need for housing in the prior six months, and current housing status. All three items were consistent with the previous literature’s conceptualization of housing instability (Brisson & Covert, 2015; Gilroy et al., 2016; Dichter et al., 2017; Breiding et al., 2017; Montgomery et al., 2018; Adams et al., 2018). The frequency of moves in a six-month period was originally a count variable. It was then converted into a dichotomous variable: zero and one moves (0) and two or more moves (1). Need for housing in the prior six months was originally a three-category variable: “did not need housing,” “needed housing and looked” and “needed housing and did not look.” This variable was collapsed into two categories for “did not need housing” (0) and “needed housing” (1). Current housing status was originally measured by asking people where they currently lived, and 18 there were 10 housing categories (homeless, shelter, transitional housing, substance abuse program, living with friend/family and not paying rent, living with friend/family and paying partial rent, renting by yourself and owning a home). This variable was collapsed into two categories: people who did own their homes or pay rent, even partially (0) and people not paying any rent or living in shelter/transitional housing/homeless (1). The three recoded items were then summed to give each participant a score from 0-3. The housing instability index was assessed for internal consistency in the sample (α= 0.59). According to Hinton and colleagues (2004), a reliability coefficient between 0.5-0.75 is indicative of a moderately reliable index (See Table 1). The different scores were treated as four ordinal categories with higher values indicating greater housing instability. Economic Abuse. Economic abuse was assessed using the updated version of the Scale of Economic Abuse (Adams et al., 2019). The scale consists of 14-items that capture abusive tactics that reduce survivors’ economic stability. Examples of items included in the scale: "How frequently has [ABUSER] done the following [do things to keep you from going to work….threatened you to make you leave work] in the last six months?” Each item was measured on a six-point Likert scale ranged from Never (0) to Quite often (5). Each item on the scale was averaged to give a final score that ranged from 0-5. The reliability coefficient for the updated version of SEA is α= 0.93. For the analytic sample, the updated version of SEA was internally consistent in measuring economic abuse (α= 0.91). DV victimization was measured by the Composite Abuse Scale (Loxton et al., 2013). The scale consists of four subscales with a total of 31items that capture experiences of stalking/harassment, sexual, physical, and emotional/psychological abuse. Examples of items included in the scale: “How often, if at all, did [ABUSER’S NAME] do any of the following [follow you, slap you…..tell you that you were ugly] in the last six months?” Each item was on a 19 six-point Likert scale ranging from (0) Never to (5) Daily. The overall four subscales were then summed and averaged by the total number of items to give a to a final score that ranged from 0- 5. The reliability coefficient for the Composite Abuse Scale is (α= .85) and for the analytic sample, the scale was internally consistent in measuring DV victimization (α= .95). Race. Race was captured as a question "what is your race/ethnicity?" where participants could choose as many race/ethnicity options as they felt applied to them. For this study, racial and ethnic categories were collapsed into four categories: Black (1), Latina (2), White (3), and Other (4). Race was collapsed into these four categories for two reasons. First, the prior literature has generally focused on differences among people who are Black, Latinx, and White (Desmond, 2012; Desmond & Shollenberger, 2015; Maykovichl et al., 2018; Olivet et al., 2018). Second, the sample did not include large enough numbers of other races to create other meaningful categories. Citizenship status. Citizenship was captured through seven items. Participants were asked whether they 1) were a US citizen, 2) had their documentation status tied to another person, 3) were a permanent resident or had a green card, 4) had work authorization, 5) were a U-visa holder, 6) were a T-visa holder, and 7) were a refugee. Given how few participants identified as having no documentation, being refugees, or holding T- or U-visas, for this study I created a single dichotomous item differentiating U.S. citizens from non-citizens. Categories were (1) for U.S citizens and (0) for non-citizens. Criminal record. Criminal record was measured by one dichotomous item: “Do you have a criminal charge that would show up on a background check?” Categories were (1) do not have a criminal record and (0) have a criminal record. Financial Difficulties. Financial difficulties were measured with the Adequacy of Financial Support scale (Mowbray et al., 1999). The scale consisted of 10 items that assessed 20 how difficult it was for survivors to pay for: food, rent/mortgage, utilities, medical expenses, transportation, social activities and to pay debts and childcare. Responses to the items were reported using a four-point Likert scale indicating how difficult it was paying for these basic needs ranging from Not Difficult (0) to Very Difficult (3). The 10 items were then summed and averaged by the total number of items to give a to a final score that ranged from 0-3. The reliability coefficient for the 10-item measure was α=.87 and for the analytic sample, the reliability coefficient for the 10-item measure was assessed (α= 0.82). Analytic Approach For the purpose of this study, participants were group into different categories of housing instability based on how many risk indicators they answered in the affirmative – no indicators of housing instability, one indicator of housing instability, two indicators of housing instability and three indicators of housing instability. Almost half of the participants (46.1%) fell into the category of having three indicators of housing instability. The next most highly endorsed category was having two indicators of housing instability (26% of the sample). Lastly, 18.6% of the sample had one indicator of housing instability, and only 8.4% of the sample had no indicators of housing instability. As previously stated, the three indicators were the number of moves, housing status, and the need for housing. For the number of moves, 59.5% of the sample had moved two or more times in the last six months and 40.5% moved zero to one time in the last six months. In addition, 67.2% were homeless. Lastly, 84.2% indicated that they needed housing. All the analyses first controlled for domestic violence and financial difficulties, as these have been shown in prior studies to impact housing stability. Financial difficulties were measured rather than income or employment status, as it was a more reliable proxy for poverty than either of these other variables in this particular study. For instance, participants were asked 21 about their household income in the prior year, but this does not measure the participant’s access to this money (Correia & Rubin, 2001). Employment does not indicate whether the job provides enough income to pay for housing. Measuring “financial difficulties” gave a more complete picture of participants’ economic circumstances that may play a role in their housing instability. All data analyses were conducted in SPSS 25 software. In order to compare scores across economic abuse, DV victimization, and financial difficulties scales, scores were standardized. To answer the five hypotheses, five regression analysis models were conducted. Missing data was less than 10 percent; therefore, no missing data analysis was conducted. Collinearity diagnostics were conducted to address multicollinearity among all relevant variables and to test assumptions of ordinal logistic regression. After testing for multicollinearity and regression assumptions, an ordinal logistic regression was conducted. In all models, the two control variables (domestic violence and financial difficulty) were entered in the first step. To address the first hypothesis, economic abuse was entered in the second step to see if was a relevant predictor of housing instability. To address hypothesis two, another model was conducted to assess if racial differences in experiencing more housing instability were evident. In the second step, a variable observing only difference among People of Color and White survivors was entered into the model. In the third step, a variable only assessing Black survivors was entered. In the fourth step, a variable assessing only Latina survivors was entered into the model. To address hypotheses three and four, two more models were conducted. In each model, citizenship status and criminal records were entered in the second step. To assess hypothesis five, another model was conducted. First, race and criminal history were added to the model. Next, an interaction term (criminal record X race) was added to the second step as a moderator to assess the relationship between race and housing instability. The 22 moderator assessed whether criminal record affects the relationship between race and housing instability such that Black and Latina survivors who have a criminal record will experience more housing instability than Black, Latina, and White survivors who do not have a criminal record. Table 1 Descriptive characteristics of housing instability index Housing Instability 0 indicators 1 indicator 2 indicators 3 indicators Number of Moves (in 6 months) 0-1 moves 2 or more moves Housing Status Homeless Not Homeless Housing Need Yes No Table 2 n 33 73 102 181 158 232 264 129 331 61 Sample % 8.4 18.6 26.0 46.1 40.5 59.5 67.2 32.8 84.2 15.5 Inter-item Correlation Matrix of the 3-item Housing Instability Index Housing Status Housing Status Housing Need Number of Moves 1 - - Housing Need 0.243 1 - Number of Moves 0.396 0.342 1 23 In this section, the overall descriptives of the sample are provided, followed by results RESULTS from hypothesis testing. Demographic Characteristics The sample included a total of 392 unstably housed or homeless female DV survivors. The majority of the sample was White (35.6 %) or Latinx (32.8%), heterosexual (86.4%), and U.S citizens (81.7%). Participants ages ranged from 19 to 62 years old (M = 34.5, SD = 9.06). In addition, over half of the sample was employed in the last six months (58%), had one to two minor children (54.7%) and had no criminal record (67.1%). Educational attainment varied across the sample, with 28.8% having less than high school education, 22.4% have a high school education or GED, 29.2% having some college or vocational education, 15.5% having a college degree, and 4.4% having an advanced degree. See Table 3 for the full sample socio- demographics. Control Variables Overall, participants reported that it was somewhat difficult to pay for basic necessities in the prior six months (M=1.94, SD=.743). Regarding domestic violence, on average, participants reported experiencing stalking, physical, sexual, or emotional/psychological abuse ‘several times’ in the last six months (M =1.73; SD=1.07). 24 Table 3 Sample socio-demographics (N=392) Race White Latinx Black Asian Middle Eastern Native American/Pacific Islander Multiracial (2 or more) Citizenship Status U.S Citizen Non-U.S Citizen Criminal Record Yes No Sexual Orientation Heterosexual Gay, Lesbian, Bisexual Unsure Education Less than High school High school GED Some college/Vocational College degree (2-4 years) Advanced college degree (4+) Employed (last 6 months) Yes No Number of Children (under 18) 0 1-2 2+ Age Financial Difficulties Domestic Violence (physical, sexual, emotional, stalking/harassment) Economic Abuse n 140 129 67 13 4 25 14 321 72 129 263 339 49 4 113 49 39 115 61 16 228 165 101 215 77 M (S.D.) 34.5 (9.06) 1.94 (.743) 1.73 (1.07) 1.96 (1.15) 25 Sample % 35.6 32.8 17.0 3.3 1.0 6.3 3.5 81.7 18.3 32.9 67.1 86.4 12.5 1.1 28.8 12.5 9.9 29.2 15.5 4.1 58.0 42.0 25.7 54.7 19.6 Range 19-62 0-3 0-5 0-5 Hypothesis Testing Before hypothesis testing, bivariate analyses, collinearity diagnostics and proportional odds testing were conducted (See Table 4). Collinearity diagnostics indicated that multicollinearity was not a concern (See Table 5). In addition, the assumption of proportional odds was met, as assessed by a full likelihood ratio test comparing the fit of the proportional odds location model to a model with varying location parameters, χ2(14) =15.19, p = .365. As shown in Tables 6-13, all hypotheses were tested using a cumulative odds ordinal logistic regression with proportional odds. Table 4 Correlation matrix of dependent and independent variables (N=392) Criminal Record -.170** Race Citizenship Status .159* .116* Economic Abuse .087 DV Victim .187* Finan Diff .034 1 -.093 -.290** -.054 .171** .048 1 .285** .134** .045 .065 .099 .096 .053 1 - - .308** .259** 1 - .071 1 1 - - - - - - - 26 Housing Instability Housing Instability Criminal Record Race Citizenship Status Economic Abuse DV Victim. Financial Difficulties 1 - - - - - - - - - - - Note. *p < .05; **p < .01; ***p <.001 Table 5 Collinearity diagnostics of dependent and independent variables (N=392) DV Victimization Economic Abuse Financial Difficulties Race Criminal Record Citizenship Status Tolerance .87 .84 .92 .91 .89 .85 VIF 1.14 1.20 1.08 1.10 1.13 1.18 For each hypothesis, the control variables of domestic violence victimization and financial difficulties were entered into the first step of the model. Only domestic violence victimization was associated with an increase in the odds of experiencing severe housing instability, with an odds ratio of 1.50, 95% CI [1.23, 1.83], χ2(2) = 16.96, p = .000) for every one unit increase (See Table 6-13). Hypothesis 1: higher economic abuse will be associated with having more housing instability over and above other forms of domestic violence. This hypothesis was not supported. As shown in Table 6, after controlling for other forms of domestic violence victimization and financial difficulties, economic abuse was not a significant predictor of housing instability (p =.875). 27 Table 6 Cumulative logistic regression of severe housing instability by economic abuse Domestic Violence (e.g., physical, emotional, sexual) Financial Difficulties Economic Abuse Note. *p < .05; **p < .01; ***p <.001 Block 1 AOR 95% CI 1.50*** 1.23, 1.83 1.04 .862, 1.26 Block 2 AOR 1.49*** 1.04 1.02 95% CI 1.22, 1.83 .853, 1.26 .831., 1.24 Hypothesis 2: Survivors of Color, specifically Black and Latina survivors, will experience more housing instability than will their White counterparts. First, the difference between People of Color versus White survivors was assessed to test if racism overall was a factor in DV survivors’ experience of housing instability. This hypothesis was not supported. After putting in the control variables (domestic violence victimization and financial difficulties), there was no significant difference between People of Color and White survivors (p = 0.51; See Table 7). In the model assessing whether Black survivors experienced greater housing instability compared to the other racial groups, no significant difference was found after controlling for domestic violence victimization and financial difficulties (p = 0.170; See Table 8). In the model assessing Latinx survivors against all other racial groups, Latinx survivors did not have a significant difference in comparison to White, Black, and Other racial counterparts survivors (p = 0.243; See Table 9 ). 28 Table 7 Cumulative logistic regression of severe housing instability by race (People of Color) Domestic Violence (e.g., physical, emotional, sexual) Financial Difficulties Race White (ref.) People of Color Block 1 AOR 95% CI 1.50*** 1.23, 1.83 1.04 .862, 1.26 Block 2 AOR 1.51 95% CI 1.23, 1.84 1.03 .853,1.25 .671 .450, 1.00 Note. *p < .05; **p < .01; ***p <.001 Table 8 Cumulative logistic regression of severe housing instability by race (Black) Domestic Violence (e.g., physical, emotional, sexual) Financial Difficulties Race Other racial groups (White, Latinx, Other) Black Note. *p < .05; **p < .01; ***p <.001 Block 1 AOR 95% CI 1.50*** 1.23, 1.83 1.04 .862, 1.26 Block 2 AOR 1.45 1.03 1 .689 95% CI 1.17, 1.81 .825, 1.27 .404, 1.17 29 Table 9 Cumulative logistic regression of severe housing instability by race (Latinx) Domestic Violence (e.g., physical, emotional, sexual) Financial Difficulties Race Other racial groups (White, Black, Other) Latinx Note. *p < .05; **p < .01; ***p <.001 Block 1 AOR 95% CI 1.50*** 1.23, 1.83 1.04 .862, 1.26 Block 2 AOR 1.49 .990 .777 95% CI 1.20, 1.86 .795, 1.23 .508, 1.19 Hypothesis 3: survivors who were non-citizens will experience more housing instability than will U.S citizens. This hypothesis was not supported. After controlling for domestic violence victimization and financial difficulties, citizenship status was a relevant predictor but not in the hypothesized direction. Non-citizens had 51% lower odds of experiencing more housing instability than U.S citizens (p=.004; see Table 10). To better understand this finding, chi-square analysis was done to examine the number of housing instability indicators by citizenship status. There was a substantial difference in the number of people having all three indicators of housing instability, with over half of the U.S citizens endorsing all three indicators (51.6%, and only 23.9% of the non-citizens endorsing all three. Table 11 presents the chi-square analysis. 30 Table 10 Cumulative logistic regression of severe housing instability by citizenship status Domestic Violence (e.g., physical, emotional, sexual) Financial Difficulties Citizenship Status Citizen (ref.) Non-citizen Note. *p < .05; **p < .01; ***p <.001 Table 11 Block 1 AOR 95% CI 1.50*** 1.23, 1.83 1.04 .862, 1.26 Block 2 AOR 95% CI 1.46*** 1.19, 1.78 1.03 .848, 1.24 1 .493** .306, .794 Cross tabulations of housing instability by citizenship status (N=389) 0 1 2 indicators indicator indicators 7 (9.8%) 26 (8.2%) 19 (26.8%) 54 (17.0%) 28 (39.4%) 74 (23.3%) 3 indicators 17 (23.9%) 164 (51.6%) Total 71 (100%) 318 (100%) Non- Citizens U.S Citizen Hypothesis 4: survivors who have a criminal record will have more indicators of housing instability than will survivors without a criminal record. This hypothesis was supported. After controlling for domestic violence and financial difficulties, having a criminal record was a significant predictor of more housing instability, χ2(3) = 25.86, p = .000. Specifically, survivors with a criminal record had about twice the odds of experiencing more severe housing instability compared to survivors without a criminal record (AOR=1.83, 95% CI [1.21, 2.79], p=.005; See Table 12). 31 Table 12 Cumulative logistic regression of severe housing instability by criminal record Domestic Violence (e.g., physical, emotional, sexual) Financial Difficulties Criminal Record No (ref.) Yes Note. *p < .05; **p < .01; ***p <.001 Block 1 AOR 95% CI 1.50*** 1.23, 1.83 1.04 .862, 1.26 Block 2 AOR 95% CI 1.44*** 1.18, 1.76 1.06 .876, 1.28 1 1.83** 1.21, 2.79 Hypothesis 5: having a criminal record will significantly moderate the relationship between race and housing instability, such that racial minorities (Black and Latinx) will report more indicators of housing instability than White survivors without a criminal record. This hypothesis was not supported. After control variables were entered into the first block of the model, racial differences among Black and Latinx were assessed. To be able to test whether criminal record moderated the relationship between race, specifically that Black and Latinx survivors with a criminal record would differ from Black and Latinx and White survivors without a criminal record, an interaction term was entered into the model. When the interaction term was added to the model there were no significant differences between Black and Latinx survivors with a criminal record and White, Black, and Latinx survivors without a criminal record (p =.759) (see Table 13). 32 Table 13 Moderated cumulative logistic regression of severe housing instability by race and criminal record Domestic Violence (e.g., physical, emotional, sexual) Financial Difficulties Race White x No Criminal Record (ref.) White x Criminal Record (ref.) Black x Criminal Record Latinx x Criminal Record Black x No Criminal Record Latinx x No Criminal Record Note. *p < .05; **p < .01; ***p <.001 Block 1 AOR 95% CI 1.50*** 1.23, 1.83 1.04 .862, 1.26 Block 2 AOR 95% CI 1.42** 1.00 1 1 1.22 1.59 1 1 1.14, 1.78 .803, 1.25 .352, 4.21 .559, 4.54 . 33 DISCUSSION Unstably housed and homeless DV survivors are in a vulnerable position; therefore, it is of utmost importance to understand the factors that contribute to their experience of severe housing instability. This is the first study to assess whether economic abuse, race, citizenship status, and criminal record are associated with greater housing instability among a sample of unstably housed DV survivors. This section considers the study findings within the context of prior literature and the study limitations. Implications for research, policy, and practice are then discussed. This study confirmed the hypothesis that survivors with a criminal record were more likely to experience more severe housing instability than those who did not have a criminal record. These findings aligned with studies from the general population that found having a criminal record was associated with experiences of housing instability (Evans, 2016; Evans, 2019; Human Rights Watch, 2014; Malone, 2009). Prior studies on the relationship between criminal records and housing instability have found that housing discrimination may be a contributing factor to this relationship (Evans et al., 2019; Human Rights Watch, 2014). Although having a criminal record was associated with experiencing more housing instability, the conclusion that this association was through the pathway of discrimination can only be inferred since this thesis did not include measuring discrimination. Furthermore, the current study did not differentiate whether the criminal record was a felony or a misdemeanor. Prior research has found that types of criminal records have a differential impact on the ability to obtain housing (Evans, 2016; Evans & Porter, 2015). In one experimental study on criminal record and landlord rental decisions, people who had a misdemeanor were more likely to obtain housing than someone with a felony (Leasure & Martin, 2017; Carey, 2005). Considering the 34 findings of prior studies, it may be worth exploring if the type of criminal record has an impact on housing instability among DV survivors. Contrary to expectations, no other hypotheses were supported in this study. After controlling for domestic violence victimization and financial difficulties, economic abuse was not associated with experiencing more housing instability. One possible explanation for this finding was that the measure of housing instability created for this study did not include financial indicators of housing instability such as the inability to pay rent and housing cost burden. Previous research has shown that economic abuse can diminish survivors’ financial independence, suggesting that it can lead to becoming housing unstable (Postmus et al., 2015; Voth Schrag, 2015). Though financial difficulties were controlled for in the current study, they were not associated with experiencing more housing instability. In addition, the sample only included homeless or unstably housed DV survivors which may have suppressed this association. In one study that had both unstably housed and stably housed individuals, having a low income was highly associated with housing instability (Pavao et al., 2007). No racial differences were found with regard to severity of housing instability. Overall, these findings contradict some previous studies from the general population that have found that People of Color, specifically Black and Latinx, do experience more housing instability than their White counterparts (Brisson & Covert, 2015; Desmond, 2012; Desmond & Shollenberger, 2015; Jones, 2016; Maykovichl et al., 2018;Olivet et al., 2018); however, one prior study on domestic violence survivors also found no racial differences in housing instability -- operationalized as moving more than twice in two years (Adams et al., 2018). Two factors may help to explain why these findings diverged from most of the previous literature. First, the non-significant findings could be due to problems with the study’s measure of housing instability. The housing instability measure was created for this thesis, and only 35 included three indicators. The measure lacked a number of aspects of housing instability that may have been salient for racial minorities such as housing cost burden, lease violations, and experience of evictions (Brisson & Covert, 2015; Desmond, 2012; Joint Center for Housing Studies of Harvard University, 2019). Second, the study did not capture how participants were perceived by others with regard to race or ethnicity. Race is often used in studies to assess differences that can highlight racial inequities around health, housing, and other outcomes, or in the case of this study, housing instability (Garcia et al., 2015; Lopez et al., 2017). Using someone’s self-identified race to suggest racial disparities that may be influenced by racism and discrimination, however, is problematic. People may self-identify with a particular racial and/or ethnic group but be perceived differently by others. In other words, someone may be perceived as White when they are not, or they may be perceived to be a Person of Color when they are not. The perception that others have about their racial identity could be a more accurate indicator of racial disparities and discrimination, which the current measure could not do. Furthermore, having a criminal record was not a significant moderator between race and experiencing more housing instability, which was anticipated given prior literature indicating having a criminal record impacts the relationship between race and housing instability (Carson, 2015; Lucius, 2018; Olivet et al., 2018). One rationale for this finding was that there was not a large enough sample to detect a moderating effect. According to Faul et al. (2008), a minimum sample size of 550 is needed to detect even a small effect. The current study only had a sample of 392 DV survivors which was not enough to detect even the smallest effect of the moderator in this relationship. Lastly, a surprising finding from this study was that being a US citizen was positively associated with experiencing more housing instability -- non-citizens experienced less housing 36 instability than U.S citizens. This finding contradicts previous studies that have found that non- U.S citizens, specifically undocumented individuals, experience more housing instability than do U.S citizens (Chavez, 2012; Cort et al., 2014; McConnell, 2013; Hall & Greenman, 2013). This finding was likely due to measurement problems. As noted earlier, the housing instability measure lacked important indicators that may have been relevant to non-citizens such as housing cost burden and poor quality housing (Chavez, 2012; Cort et al., 2014; McConnell, 2013; Hall & Greenman, 2013). When examining the three indicators of housing instability by citizenship status, I noted that 52% of the US citizens endorsed all three indicators, whereas only 24% of non-citizens endorsed all three. This difference may be due to the type of living arrangements that are more common among non-citizens than U.S citizens (Van Hook & Glick, 2007). In one study that examined the living arrangements of Latinx immigrants in the U.S, findings showed that recent Latinx immigrants were more likely to stay with extended family and live with relatives than U.S citizens (Van Hook & Glick, 2007). Though only one study observed these trends in living arrangements, this could be a plausible explanation for why non-citizens were less likely to endorse items in the current study’s housing instability index. In addition, due to small subgroup sample sizes of those in different non-citizenship categories (e.g., permanent resident, T-visa, U -visa holder, undocumented), it was not possible to conduct analyses by subgroup. Previous literature supports that undocumented and documented individuals living in the U.S have vastly different experiences with access to housing (Chaudry et al., 2010). Undocumented people tend to deal with more discrimination and restrictive immigration policies than their documented counterparts, placing them at increased risk of experiencing housing instability (Lyubansky et al., 2013). For instance, undocumented individuals are unable to use public housing or access public assistance that could help with maintaining housing (Chaudry et al., 2010; McCarty & Siskin, 2015). Furthermore, non-citizens 37 who are permanent residents or who are authorized to live in the U.S have more access to housing resources than their undocumented counterparts (McCarty & Siskin, 2015). Considering that people who were undocumented were included in the non-citizen category, these differences, if present, could not be detected. Limitations The study findings need to be considered in light of limitations. First, a significant limitation was the measure of housing instability. It should be noted that the field lacks an adequate measure of housing instability (Fredrick et al., 2014). Prior studies examining housing instability among DV survivors have tended to rely on single indicators such as inability to pay rent or being concerned about losing housing, or having moved more than twice in a two year period (Adams et al., 2018; Breiding et al., 2017; Dichter et al., 2017). One measure, the Housing Instability Index (Rollins et al., 2012), does include multiple indicators, which enhances its ability to accurately measure the construct. Despite this, four of the ten items are reflective of people with landlords and it does not measure homelessness (Lewis-Beck et al., 2011; Rollins et al., 2012). The current study attempted to improve upon prior studies by using multiple indicators, but only included three items and was not validated. The three indicators of housing instability were limited to current housing status, frequency of moves, and housing need. The measure lacked other aspects of housing instability such as landlord issues, evictions, poor housing quality, inability to pay rent, and other financial indicators that may have been relevant among a population of unstably housed DV survivors. As noted earlier, the measure of race/ethnicity had several issues. The measure used self- identified racial categories which only captures one aspect of the construct. Studies have shown that self-identified racial categories are not as important as other measures of race (e.g. ascribed 38 race, skin color) when assessing racial discrimination (MacIntosh et al., 2013; Garcia et al, 2015). The study sample size and composition also introduced limitations to the project. The entire sample was comprised of homeless or unstably housed DV survivors. Therefore, I was unable to compare unstably housed with stably housed survivors, and the findings cannot be generalizable to all DV survivors. If the sample comprised both unstably housed and stably housed DV survivors, possible protective factors could be identified, and it would increase the generalizability of the sample. The study’s sample also comprised only cis-gendered female DV survivors hence these findings are not representative of survivors who identify as lesbian, gay, bisexual, transgender or queer. In addition, all DV survivors in this sample were from a help-seeking population. DV survivors who seek out support differ from DV survivors who do not seek out support in terms of the severity of violence, wellbeing and needs (Macy et al., 2005). Finally, due to the small sample size, undocumented survivors could not be separated from non-citizens. As previously stated, undocumented people tend to experience more housing instability compared to their documented counterparts (Chavez, 2012; Cort et al., 2014; McConnell, 2013; Hall & Greenman, 2013). By not being able to separate survivors who were documented and undocumented, this study assumed that their experiences were the same when this may not be the case. Therefore, separating the two groups would have been preferable but there were not enough undocumented individuals in the sample to conduct analyses on this group. Implications for Future Research The current study revealed the need for a more comprehensive, validated measure of housing instability for DV survivors. The field would benefit from having a measure that 39 captures all the ways in which DV survivors could be experiencing severe housing instability (Fredrick et al., 2014; Routhier, 2019). In the current study, all DV survivors were experiencing housing instability in some form (e.g., frequent moves, homelessness), limiting variability on this construct. For instance, a survivor could be experiencing severe housing instability when they have a multitude of housing issues such as simultaneously dealing with being homeless and not having enough money to pay rent which could be quite different from someone who simply cannot afford to pay rent. In addition to considering the severity of housing instability, there should be a distinction made between difficulties in getting housing versus difficulties in maintaining one's housing. Some DV survivors may be currently struggling with issues that pertain only to obtaining housing such as having been evicted, discrimination by landlords, or owing back rent. Those trying not to lose their housing, on the other hand, may have other concerns such as financial concerns and landlord disputes. Measuring these distinctions could provide insight into specific indicators that may be linked to these two aspects of housing instability. Possible directions for future studies on housing instability could include conducting more qualitative studies to understand the aspects of housing instability among DV survivors in more depth. Additionally, conducting exploratory cluster analysis could help identify types of indicators of housing instability that are linked to certain characteristics or circumstances (e.g. discrimination, DV). Overall, having a validated measure that can address these aspects would be integral for the development of effective approaches to reducing housing instability among DV survivors. Future research should also consider improving measurements of race by including other dimensions such as ascribed race, skin color, and experiences of racial discrimination than only 40 using someone’s self-identified race to access experiences of racism, enhancing its explanatory power (Garcia et al., 2015). Furthermore, future research should examine what types of criminal records among DV survivors are associated with experiencing housing instability. This effort could help to reveal what types of criminal records may contribute to an increased risk of severe housing instability. Implications for Practice and Policy This study provides important implications for advocacy and policy. Having a criminal record was associated with experiencing more housing instability among female DV survivors. It is important for policymakers to understand that some DV survivors may have acquired a criminal record due to the abuse they have experienced (Dichter, 2013; Hirsh, 2001). Policymakers need to enforce anti-discriminatory laws that bar survivors with a criminal record from obtaining housing. One solution could be continue reauthorizing the Violence Against Women Act (VAWA) that has clear guidelines about DV survivors' access to housing and not excluding survivors from housing if someone in the household has a criminal record (Sacco, 2019). In addition, policymakers should take notice of what anti-discriminatory laws that have been put in place and examine their effectiveness in meeting the needs of DV survivors (Keefe & Hahn, 2020). For DV survivors with a criminal record, conducting individualized reviews of their situations should be also enforced (Carey, 2005). An individualized review would allow for a complete assessment of the survivor's past experience with violence and their current ability to obtain housing (e.g. being able to pay rent, etc.) without automatic exclusion based on having a criminal record. In addition, having an exhaustive measure of housing instability that addresses different issues related to obtaining housing versus maintaining housing could provide DV agencies and 41 advocates with a tool for assessing survivors’ experiences of housing instability. This tool could help with providing tailored support for DV survivors. Conclusion Housing instability is a critical concern for many DV survivors. Providing some understanding of what contributes to severe housing instability can aid in the effort to support unstably housed and homeless DV survivors. As this study demonstrated, having a criminal record can influence their experience of housing instability. The field would greatly benefit from the creation of an accurate and exhaustive measure of housing instability. It can aid researchers, DV agencies, and advocates to have a better understanding of all the ways in which DV survivors are experiencing housing instability and could lead to crafting effective interventions for this population. Similar to the need for a housing instability measure, sociodemographic measures need to be more expansive with the purpose that there is an increased understanding of how survivors’ social location factors into their experience of housing instability. 42 APPENDIX 43 APPENDIX A: IRB APPROVAL LETTER 44 REFERENCES 45 REFERENCES Adams, A. E., Greeson, M. R., Littwin, A. K., & Javorka, M. (2019). The Revised Scale of Economic Abuse (SEA2): Development and initial psychometric testing of an updated measure of economic abuse in intimate relationships. Psychology of Violence. 10(3), 268–278. Adams, A. E., Littwin, A. K., & Javorka, M. (2019). The frequency, nature, and effects of coerced debt among a national sample of women seeking help for intimate partner violence. Violence Against Women. Adams, A. E., Sullivan, C. M., Bybee, D., & Greeson, M. R. (2008). Development of the scale of economic abuse. Violence Against Women, 14(5), 563-588. Adams, A. E., Tolman, R. M., Bybee, D., Sullivan, C. M., & Kennedy, A. C. (2012). The impact of intimate partner violence on low-income women’s economic well-being. Violence Against Women, 18(12), 1345-1367. Adams, E. N., Clark, H. M., Galano, M. M., Stein, S. F., Grogan-Kaylor, A., & Graham- Bermann, S. (2018). Predictors of housing instability in women who have experienced intimate partner violence. Journal of Interpersonal Violence. Arnold, G. W. (2016). From victim to offender: How nuisance property laws affect battered women. Journal of Interpersonal Violence, 34(6), 1103-1126. Aurand, A., Emmanuel, D., Errico, E., Pinsky, D., & Yentel, D. (2020) The gap: a shortage of affordable homes. Washington (DC): National Low Income Housing Coalition. https://reports.nlihc.org/sites/default/files/gap/Gap-Report_2020.pdf Baker, C. K., Billhardt, K. A., Warren, J., Rollins, C., & Glass, N. E. (2010). Domestic violence, housing instability, and homelessness: A review of housing policies and program practices for meeting the needs of survivors. Aggression and Violent Behavior, 15(6), 430-439. Baker, C. K., Cook, S. L., & Norris, F. H. (2003). Domestic violence and housing problems. Violence Against Women, 9(7), 754-783. Barata, P. C., & Stewart, D. E. (2010). Searching for housing as a battered woman: Does discrimination affect reported availability of a rental unit? Psychology of Women Quarterly, 34(1), 43-55. Breiding, M. J., Basile, K. C., Klevens, J., & Smith, S. G. (2017). Economic insecurity and intimate partner and sexual violence victimization. American Journal of Preventive Medicine, 53(4), 457-464. 46 Breiding, M. J., Chen, J., & Black, M. C. (2014). Intimate partner violence in the United States-- 2010. Centers for Disease Control and Prevention. Brisson, D., & Covert, J. (2015). Housing instability risk among subsidized housing recipients: Characteristics associated with late or nonpayment of rent. Social Work Research, 39(2), 119-128. Carey, C. A. (2004). No second chance: People with criminal records denied access to public housing. U. Tol. L. Rev., 36, 545. Chang, C. D. (2019). Social determinants of health and health disparities among immigrants and their children. Current problems in pediatric and adolescent health care, 49(1), 23-30. Chaudry, A., Capps, R., Pedroza, J. M., Castaneda, R. M., Santos, R., & Scott, M. M. (2010). Facing our future: Children in the aftermath of immigration enforcement. PsycEXTRA Dataset. Chavez, L. R. (2012). Undocumented immigrants and their use of medical services in Orange County, California. Social Science & Medicine, 74(6), 887-893. Clough, A., Draughon, J. E., Njie-Carr, V., Rollins, C., & Glass, N. (2013). ‘Having housing made everything else possible’: Affordable, safe and stable housing for women survivors of violence. Qualitative Social Work: Research and Practice, 13(5), 671-688. Correia, A., & Rubin, J. (2001). Housing and battered women. Violence Against Women. http://www.vawnet.org Cort, D. A., Lin, K., & Stevenson, G. (2014). Residential hierarchy in Los Angeles: An examination of ethnic and documentation status differences. Social Science Research, 45, 170-183. Crowne, S. S., Juon, H., Ensminger, M., Burrell, L., McFarlane, E., & Duggan, A. (2010). Concurrent and long-term impact of intimate partner violence on employment stability. Journal of Interpersonal Violence, 26(6), 1282-1304. Desmond, M. (2012). Eviction and the reproduction of urban poverty. American Journal of Sociology, 118(1), 88-133. Desmond, M. (2016). Evicted: Poverty and profit in the American city. Broadway Books. 192- 290. Desmond, M., & Gershenson, C. (2016). Housing and employment insecurity among the working poor. Social Problems, 63(1), 46-67. Desmond, M., & Shollenberger, T. (2015). Forced displacement from rental housing: Prevalence and neighborhood consequences. Demography, 52(5), 1751-1772. 47 Dichter, M. E. (2013). “They arrested me—And I was the victim”: Women’s experiences with getting arrested in the context of domestic violence. Women & Criminal Justice, 23(2), 81-98. Dichter, M. E., Wagner, C., Borrero, S., Broyles, L., & Montgomery, A. E. (2017). Intimate partner violence, unhealthy alcohol use, and housing instability among women veterans in the Veterans Health Administration. Psychological services, 14(2), 246. Ellen, I. G., & Karfunkel, B. (2016). Renting In America’s Largest Metropolitan Areas Renting In America’s Largest Metropolitan Areas Authors. Evans, D. N. (2016). The effect of criminal convictions on real estate agent decisions in New York City. Journal of Crime and Justice, 39(3), 363-379. Evans, D. N., & Porter, J. R. (2014). Criminal record and landlord rental decisions: A New York quasi-experimental study. Journal of Experimental Criminology, 11(1), 21-42. Evans, D. N., Blount-Hill, K., & Cubellis, M. A. (2018). Examining housing discrimination across race, gender, and felony history. Housing Studies, 34(5), 761-778. Faul, F., Erdfelder, E., Buchner, A., & Lang, A. G. (2008). G* Power Version 3.1. 2 [computer software]. Universitat Kiel, Germany. Retrieved from http://www.psycho.uni- duesseldorf.de/abteilungen/aap/gpower3/download-and-register. Ford-Gilboe, M., Wuest, J., Varcoe, C., Davies, L., Merritt-Gray, M., Campbell, J., & Wilk, P. (2009). Modeling the effects of intimate partner violence and access to resources on women's health in the early years after leaving an abusive partner. Social Science & Medicine, 68(6), 1021-1029. Frederick, T. J., Chwalek, M., Hughes, J., Karabanow, J., & Kidd, S. (2014). How stable is stable? Defining and measuring housing stability. Journal of Community Psychology, 42(8), 964-979. Garcia, J. A., Sanchez, G. R., Sanchez-Youngman, S., Vargas, E. D., & Ybarra, V. D. (2015). Race as lived experience: the impact of multi-dimensional measures of race/ethnicity on the self-reported health status of Latinas. Du Bois Review: social science research on race, 12(2), 349. Gezinski, L. B., & Gonzalez-Pons, K. M. (2019). Unlocking the door to safety and stability: Housing barriers for survivors of intimate partner violence. Journal of Interpersonal Violence, 088626051985179. Gilroy, H., McFarlane, J., Maddoux, J., & Sullivan, C. (2016). Homelessness, housing instability, intimate partner violence, mental health, and functioning: A multi-year cohort study of IPV survivors and their children. Journal of Social Distress and the Homeless, 25(2), 86-94. 48 Hall, M., & Greenman, E. (2013). Housing and neighborhood quality among undocumented Mexican and Central American immigrants. Social Science Research, 42(6), 1712-1725. Hall, M., Greenman, E., & Farkas, G. (2010). Legal status and wage disparities for Mexican immigrants. Social Forces, 89(2), 491-513. Hinton, P. R., Brownlow, C., McMurray, I., & Cozens, B. (2004). Using SPSS to analyse questionnaires: Reliability. SPSS explained, 356-366. Hirsch, A. E. (2001). “The World was Never a Safe Place for Them” Abuse, Welfare Reform, and Women with Drug Convictions. Violence Against Women, 7(2), 159-175. Human Rights Watch. (2014). No second chance: People with criminal records denied access to public housing. https://www.hrw.org/reports/2004/usa1104/usa1104.pdf Jasinski, J. L., Wesely, J., Mustaine, E., & Wright, J. D. (2005). The experience of violence in the lives of homeless women: A research report. Washington, DC: National Institute of Justice.https://www.ncjrs.gov/pdffiles1/nij/grants/211976.pdf?origin=publication_detail &ev=pub_ext_prw_xdl Jategaonkar, N., & Ponic, P. (2011). Unsafe and Unacceptable Housing: Health & Policy Implications for Women Leaving Violent Relationships. Women's Health & Urban Life, 10(1), 32-58. Joint Center for Housing Studies of Harvard University. (2019). The State of the Nation’s Housing in 2019. https://www.jchs.harvard.edu/state-nations-housing-2019 Jones, M. M. (2016). Does race matter in addressing homelessness? A review of the literature. World Medical & Health Policy, 8(2), 139-156. Keefe R., & Hahn, S. A. (2020). Policy roles in promoting affordable housing for survivors of intimate partner violence. Violence against women, 1077801220913629. Kim, J., & Gray, K. A. (2008). Leave or stay? Battered women's decisions after intimate partner violence. Journal of Interpersonal Violence, 23(10), 1465-1482. Kimerling, R., Alvarez, J., Pavao, J., Mack, K. P., Smith, M. W., & Baumrind, N. (2008). Unemployment among women: examining the relationship between physical and psychological intimate partner violence and posttraumatic stress disorder. Journal of Interpersonal Violence, 24(3), 450-463. Leasure, P., & Martin, T. (2017). Criminal records and housing: An experimental study. Journal of Experimental Criminology, 13(4), 527-535. Levin, R., McKean, L., & Raphael, J. (2004). Pathways to and from homelessness: Women and children in Chicago shelters. Center for Impact Research. Pl Chicago, Il. 49 Lewis-Beck-Beck, M., Bryman, A. E., & Liao, T. F. (2003). The Sage encyclopedia of social science research methods. Sage Publications. Littwin, A. (2012). Coerced debt: The role of consumer credit in domestic violence. Calif. L. Rev., 100(951). López, M. M., Bevans, M., Wehrlen, L., Yang, L., & Wallen, G. R. (2017). Discrepancies in race and ethnicity documentation: a potential barrier in identifying racial and ethnic disparities. Journal of racial and ethnic health disparities, 4(5), 812-818. Loxton, D., Powers, J., Fitzgerald, D., Forder, P., Anderson, A., Taft, A., & Hegarty, K. (2013). The Community Composite Abuse Scale: reliability and validity of a measure of intimate partner violence in a community survey from the ALSWH. Journal of Women's Health Issues Care, 2(10.4172), 2325-9795. Lucius, C. (2018). Nowhere to go: Homelessness among formerly incarcerated people. Prison Policy Initiative. https://www.prisonpolicy.org/reports/housing.html Lyubansky, M., Harris, P. A., Baker, W. E., & Lippard, C. D. (2013). “One day on the red hills of Georgia”: The effects of immigration status on Latina migrants’ experience of discrimination, utilization of public services, and attitudes toward acculturation. Norteamérica, 8 (Número Especial), 85-117. MacIntosh, T., Desai, M. M., Lewis, T. T., Jones, B. A., & Nunez-Smith, M. (2013). Socially- assigned race, healthcare discrimination and preventive healthcare services. PloS one, 8(5), e64522. Macy, R. J., Nurius, P. S., Kernic, M. A., & Holt, V. L. (2005). Battered women's profiles associated with service help-seeking efforts: Illuminating opportunities for intervention. Social Work Research, 29(3), 137-150. Malone, D. K. (2009). Assessing criminal record as a predictor of future housing success for homeless adults with behavioral health disorders. Psychiatric Services, 60(2), 224-230. Mahathey, H. M., Morrill, A., Robinson, T., Shivji, A., & Watt, R. (2018). The 2018 Annual Homeless Assessment Report (AHAR) to Congress. Part 1: Point-in-Time Estimates of Homelessness. December 2018. Maykovichl, X., Witter, E., Cookson, T., & Diddams, M. (2018). Losing Home: The Human Cost of Eviction in Seattle. Seattle Women's Commission. McCarty, M., & Siskin, A. (2012). Immigration: Noncitizen eligibility for needs-based housing programs. Congressional Research Service. McConnell, E. D. (2013). Who has housing affordability problems? Disparities in housing cost burden by race, nativity, and legal status in Los Angeles. Race and Social Problems, 5(3), 173-190. 50 Meisel, J., Chandler, D., & Rienzi, B. M. (2003). Domestic violence prevalence and effects on employment in two California TANF populations. Violence Against Women, 9(10), 1191- 1212. Moe, A. M., & Bell, M. P. (2004). Abject economic: The effects of battering and violence on women’s work and employability. Violence Against Women, 10(1), 29-55. Mowbray,C ., Jordan,L ., Kewalramani, A., Luke, D., Ribisl, K., Herman, S., & Bybee, D. (1999). Analysis of post-discharge change in a dual diagnosis population. Health and Social Work, 24, 91-101. National Center for Injury Prevention and Control. (2003). Costs of intimate partner violence against women in the United States. Center for Disease Control Prevention, 1-64. https://www.cdc.gov/violenceprevention/pdf/ipvbook-a.pdf National Coalition Against Domestic Violence (NCADV). (2015). Domestic violence national statistics. Niccolai, L. M., Blankenship, K. M., & Keene, D. E. (2019). Eviction from renter-occupied households and rates of sexually transmitted infections. Sexually Transmitted Diseases, 46(1), 63-68. O’Campo, P., Daoud, N., Hamilton-Wright, S., & Dunn, J. (2015). Conceptualizing housing instability: Experiences with material and psychological instability among women living with partner violence. Housing Studies, 31(1), 1-19. Olivet, J., Dones, M., Richard, M., Wilkey, C., Yampolskaya, S., Beit-Arie, M., & Joseph, L. (2018). Supporting Partnerships for Anti-Racist Communities (SPARC): Phase I Study Findings. Center for Social Innovation. http://center4si. com/wp- content/uploads/2018/03/SPARC-Phase-1-Find ings-March-20181. pdf. Pavao, J., Alvarez, J., Baumrind, N., Induni, M., & Kimerling, R. (2007). Intimate partner violence and housing instability. American Journal of Preventive Medicine, 32(2), 143- 146. Peterson, C., Kearns, M. C., McIntosh, W. L., Estefan, L. F., Nicolaidis, C., McCollister, K. E., Gordon, A., & Florence, C. (2018). Lifetime economic burden of intimate partner violence among U.S. adults. American Journal of Preventive Medicine, 55(4), 433-444. Phinney, R., Danziger, S., Pollack, H. A., & Seefeldt, K. (2007). Housing instability among current and former welfare recipients. American Journal of Public Health, 97(5), 832- 837. Pollack, C. E., Griffin, B. A., & Lynch, J. (2010). Housing affordability and health among homeowners and renters. American Journal of Preventive Medicine, 39(6), 515-521. 51 Ponic, P., Varcoe, C., Davies, L., Ford-Gilboe, M., Wuest, J., & Hammerton, J. (2011). Leaving ≠ moving. Violence Against Women, 17(12), 1576-1600. Postmus, J. L., Plummer, S. B., McMahon, S., Murshid, N. S., & Kim, M. S. (2012). Understanding economic abuse in the lives of survivors. Journal of interpersonal violence, 27(3), 411-430. Riger, S., Staggs, S. L., & Schewe, P. (2004). Intimate partner violence as an obstacle to employment among mothers affected by welfare reform. Journal of Social Issues, 60(4), 801-818. Rollins, C., Glass, N. E., Perrin, N. A., Billhardt, K. A., Clough, A., Barnes, J., Hanson, G. C., & Bloom, T. L. (2011). Housing instability is as strong a predictor of poor health outcomes as the level of danger in an abusive relationship. Journal of Interpersonal Violence, 27(4), 623-643. Routhier, G. (2019). Beyond worst-case needs: Measuring the breadth and severity of housing insecurity among urban renters. Housing Policy Debate, 29(2), 235-249. Sacco, L. N. (2019). The violence against women act (VAWA): historical overview, funding, and reauthorization. The Violence Against Women Act (VAWA): historical overview, funding, and reauthorization. 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. Smith, S.G., Zhang, X., Basile, K.C., Merrick, M.T., Wang, J., Kresnow, M., Chen, J. (2018). The National Intimate Partner and Sexual Violence Survey (NISVS): 2015 Data Brief – Updated Release. Atlanta, GA: National Center for Injury Prevention and Control. https://www.cdc.gov/violenceprevention/pdf/2015data-brief508.pdf Stern, N., Valles, J., & Randal, A. (2007). Lost Housing, Lost Safety: Survivors of Domestic violence Experience Housing Denials and Evictions Across the Country. National Law Center on Homelessness and Poverty. https://nlchp.org/wp- content/uploads/2018/10/Lost-Housing-Lost-Safety.pdf Showalter, K. (2016). Women's employment and domestic violence: A review of the literature. Aggression and violent behavior, 31, 37-47. Tessler, R., Rosenheck, R., & Gamache, G. (2001). Gender differences in self-reported reasons for homelessness. Journal of Social Distress and the Homeless, 10(3), 243-254. Thomas, K. A., Goodman, L., & Putnins, S. (2015). “I have lost everything”: Trade-offs of seeking safety from intimate partner violence. American Journal of Orthopsychiatry, 85(2), 170-180. Truman, J. L., & Morgan, R. E. (2014). Nonfatal Domestic Violence, 2003–2012 (NCJ 244697). 52 Trochim, W. M., & Donnelly, J. P. (2001). Research methods knowledge base. Chapter 5. Scales and Indexes. Cengage Learning. p.126-129. U.S. Census Bureau. (2018). American Community Survey Data. http://www.census.gov/acs United States Conference of Mayors. (2009). United States Conference of Mayors 2008 Hunger and Homelessness Report. Retrieved from http://www.usmayors.org/publications. Vallas, R., & Dietrich, S. (2014). One Strike and You're Out: How We Can Eliminate Barriers to Economic Security and Mobility for People with Criminal Records. Center for American Progress.https://cdn.americanprogress.org/wpcontent/uploads/2014/12/VallasCriminalRe cordsReport.pdf Van Hook, J., & Glick, J. E. (2007). Immigration and living arrangements: Moving beyond economic need versus acculturation. Demography, 44(2), 225-249. VAWA: Udall celebrates Senate reauthorization of VAWA. (2012). Lanham, MD: Federal Information & News Dispatch. Voth Schrag, R. J. (2015). Economic abuse and later material hardship: Is depression a mediator?. Affiliate, 30(3), 341-351. 53