Abstract
Many studies have attributed the disproportionately high rate of domestic violence in Asian communities to Asian patriarchal “cultural norms” and the psychological and behavioral traits that these norms produce in individuals. This article seeks to expand the scope of domestic violence analysis beyond these individual and cultural frameworks, arguing that Asian domestic violence is also a product of larger scale, social systems of inequality. By examining the funding criteria of the Family Violence Prevention Services Administration (FVPSA) and the Quality-Adjusted Life Year (QALY) standard used by Robin Hood, my research shows how state and private organizations systematically devalue and underfund minority-targeted programs.
At the 2005 release of the World Health Organization’s (WHO) landmark study on domestic violence, WHO Director-General Lee Jong-wook stated that “Women are more at risk from violence at home than in the street, and this has serious repercussions for women’s health.” Worldwide, most women who die of homicide are killed by their partner or ex-partner (WHO, 1997), and domestic violence especially is a leading cause of death during pregnancy (Advocates for Human Rights [AHR], 2003). Even when abuse is not lethal, survivors of domestic violence 1 sustain serious physical injuries, ranging from “bruises and fractures to chronic disabilities” (AHR, 2003). In the United States, violence by an intimate partner is the leading cause of non-fatal injury to women (WHO, 2010). A recent study at Johns Hopkins University found that abused women have neurological, gynecological, and stress-related problems at a rate 50% to 70% higher than never-abused women. Pregnant women who experience abuse are at significantly increased risk for low birth weight, infections, anemia, and other dangerous pregnancy complications (WHO, 1997).
Between 41% and 60% of Asian American women experience abuse at some point during their lifetime (Asian & Pacific Islander Institute on Domestic Violence [APIIDV]; Yoshihama, & Dabby, 2009), as compared with about 25% to 30% of women in the U.S. population overall (American Bar Association, 2010). 2 In recent literature, many studies have attributed this disparity to Asian “cultural norms,” suggesting that because Asian culture is traditionally more patriarchal, Asian women in the United States are more likely than their native-born counterparts to accept abuse as natural and not seek outside help. Especially within the public health community, epidemiological studies of domestic violence have tended to focus on identifying culturally specific risk factors for being a domestic violence victim, usually manifested as individual, psychological, or behavioral traits (e.g., “passivity” or “fatalism”) (Yick & Oomen-Early, 2008). My research, however, seeks to expand the scope of domestic violence analysis beyond these individual and cultural 3 frameworks, positing that domestic violence in Asian communities is also caused and perpetuated by larger scale, social systems of inequality—and, in particular, the United States’ current system of social service provision. Social services are often the determining factor enabling a woman to break free of an abusive relationship, particularly through helping her become financially independent. My research describes how service-providing organizations face significant barriers to making such services available to the Asian community, because the state and private organizations fail to recognize their significance and urgency. As a result, these programs are systematically devalued and underfunded, creating structural, in addition to individual-level, barriers to Asian women seeking to escape domestic violence. 4
Limits of Individual and Cultural Analyses
In a recent literature review of 1,396 publications in leading scholarly journals on gender violence, Salazar and Cook (2002) found that “Most studies focus on an individual level of analysis (p. 410)” and “view violence against women as a problem that resides within individuals (p. 418)” They report that these studies emphasize how individual psychology influences abused women’s attitudes and responses, positing it as a “causal mechanism” that leads to their staying in harmful relationships. Building on this research, Alice Yick and Oomen-Early (2008) conducted a similar literature review of studies specifically concerning domestic violence in Asian or Asian American populations. While they note a general scarcity of studies on this population in general (remarking that they account for only 2.4% of articles published in the five most prominent violence-related journals over the past 16 years), their findings agree with those of Salazar and Cook concerning the focus on individualistic analysis. For example, one study of domestic violence in Korean communities reported that “risk factors” for domestic violence included having low self-reported self-esteem and having been exposed to physical assault in childhood (Lee, 2007). Another study of Korean American immigrant communities found that “the relationship between the attitudes toward partner abuse and the incidence of partner abuse as perpetrator was significantly correlated” and suggested that acceptance of violence by the abused woman played a primary role in allowing and perpetuating the practice (Ahn, 2002).
However, among research that specifically explores the disproportionately high rates of domestic violence among Asians, there is a shift in emphasis from talking about the individual to talking about the community. In particular, many studies tend to attribute the high prevalence of abuse among Asians to the community’s unique and violence-prone culture, with its cultural norms that more readily sanction all forms of physical punishment, including domestic abuse. For example, Lee’s (2007) aforementioned study concluded that “most of the problems that Asian immigrant women experience [are] based on cultural differences.”
The predominant conception of Asian culture 5 in these studies is one in which “the Confucian model is the cornerstone of all social relations” (Yick, 2001). Within this highly patriarchal system, power and authority are transmitted from the father to the eldest son and all are expected to obey the male head (Wong, 1995, in Yick, 2001). Furthermore, this patriarchal family is the model for all social relations and the main source of identity and belonging for all members of that culture. Therefore, these studies claim that women are socialized “from birth” (Yick, 2001) to resign themselves to being dominated within their families. 6 Masaki and Wong (1997, in Yick, 2001) argued that the reason Western social service paradigms seem to fail Asian women is that their focus on individual empowerment is “dissonant with traditional Asian values that emphasize the collective unit, such as the family.”
These studies present a “simplistic,” “flattening and reductive” (Yick & Oomen-Early, 2008) view of traditional culture, assuming that “culture” acts universally and identically across time and space. As Yick and Oomen-Early (2008) write, “culture” in these studies often becomes interchangeable with “ethnicity,” being assumed to act on all members of an immigrant community with a certain shared genetic background. For example, they describe how “a sample of Chinese Americans might be recruited for a study, and then ‘Chinese culture’ would be used to interpret all of the findings, when in fact the level of adherence to Chinese culture is not necessarily fixed.” The authors note that in 46 studies that addressed culture within the review, all 46 used ethnicity unproblematically as a “proxy” for culture, without considering the variable degree of association between culture and ethnicity.
This leads to another flaw in cultural analyses: By failing to engage with culture in all its variability and complexity, these studies can effectively revert to the level of individual analysis. Rather than analyzing Asian culture in terms of its norms, how they are enforced, and how they might be changed, studies tend to use culture as an explanation for individual women’s psychological states and behavioral patterns. Because of this similarity and slippage between them, individual and cultural analyses are often very much alike in the conclusions they reach about necessary points of intervention, as well as in their assumptions about abused women’s needs. Whether these studies find “risk factors” for abuse in individuals’ unstable psychological states or their backward cultural norms, both types of conclusions seem to imply that the suffering of abused women stems from factors within the abused population. Yick (2001) warns that this can lead to unfair victim blaming in which survivors come to be seen as suffering from their own deficiency. Domestic violence in Asian communities is depicted as “‘part of their culture’ or ‘in their nature’—‘their’ designating either the victims of violence, or the perpetrators, or both, as may be expedient’” (Farmer, 2003, p. 48). This kind of scholarship fails to recognize what Salazar and Cook (2002) call “the social, institutional, historical, and political contexts of violence against women.”
Since the publication of Kimberle Crenshaw’s (1991) groundbreaking article, “Mapping the Margins: Intersectionality, Identity Politics, and Violence Against Women of Color,” an increasing number of feminist scholars have treated the structural and historical aspects of domestic violence, offering more nuanced analyses of the social, political, and economic factors (such as class, race, ethnicity, and sexual orientation) that complicate individualistic and cultural models. By paying attention to the specific challenges experienced by minority and/or poor women, these authors not only enrich our understanding of the causes, implications, and modes of prevention of domestic violence but also expand the vision of gender justice itself (Bauer, Rodriguez, Quiroga, & Flores-Ortiz, 2000, Menjivar & Salcido, 2002, Sokoloff & Dupont, 2005). Menjivar and Salcido (2002), for example, argue for a more intersectional and comprehensive approach to domestic violence so that, “[i]nstead of simply adding factors that affect immigrant women” to the existing model, we should aspire toward “a restructuring of domestic violence frameworks” themselves. Sokoloff and Dupont (2005) emphasize the need for addressing “the individual lived experiences of diverse battered women . . . as well as the social structural underpinnings of domestic violence in culturally diverse communities” (pp. 38-39) so that we can implement “public policies that address [the] structural root causes of domestic violence” (p. 39). Furthermore, by retaining the productive tension between the focus on individual experiences and the emphasis on structural factors that define and shape such experiences, these authors offer more comprehensive and responsible ways to theorize “violence” itself. Sokoloff and Dupont (2005) point out that
the lack of adequate institutional support in the form of social services and public housing as well as the intrusions and coercive controls by the state and its agencies . . . is another level of violence experienced by battered women, which occur in ways that are racialized as well as gendered and classed. (p. 44)
This study seeks to contribute to the growing body of feminist scholarship that explores the multiple levels of violence inflicted on women who reside at the intersections of race, class, and gender inequality. In particular, I explore the factors that contribute to the Asian community’s disproportionately high rate of domestic abuse: I examine how structural obstacles currently limit the availability of social services to Asian women, not only by directly impeding individuals’ access to them 7 but also by impeding social service providers themselves from making services available to Asian immigrant populations. In so doing, I challenge the language of “cultural difference”—read: cultural deficiency—that has too long gone unproblematized in speaking about Asian domestic violence. At the same time, I seek to avoid reducing domestic violence survivors to powerless beings whose actions are completely determined by structural, large-scale political, and economic forces beyond their control. Rather, I argue that recognizing how community members’ choices are constrained by these structural forces can lead to more realistic and productive programs for change, according them the agency to name and actively resist their oppression.
Method
Research was conducted at the New York Asian Women’s Center (NYAWC), located in New York City. The Center was founded in 1982 as the first domestic violence organization on the East Coast to serve Asian communities. From June to August, 2009, I gathered data through written surveys among the women clients and structured interviews with NYAWC staff. All respondents were informed about the nature of the study and gave informed consent for the inclusion of their responses in this article. The study received institutional review board (IRB) approval.
Client surveys were based on the NYAWC’s existing client surveys, which were available in English as well as Mandarin, Japanese, Korean, Urdu, Punjabi, and Spanish. These surveys contained questions about the clients’ service needs and whether they were met, whether they had concerns or difficulties contacting the shelter, and what other options they had considered. Approximately 20 surveys were administered in person at the NYAWC’s resident shelters, while 30 were administered verbally over the phone. For the in-person surveys, if clients were not comfortable with the written form, they again had the option to complete the survey verbally with the help of a translator.
I also conducted interviews with a wide variety of NYAWC staff members: counselors, case managers, outreach coordinators, and directors. In addition, I conducted interviews with other social service providers to whom staff members referred me, including social workers and physicians at local hospitals, and local university professors and public health experts. These interviews were largely unstructured and open-ended; each focused on the specific aspect of social service provision on which the interviewee was most expert. I conducted a total of 14 interviews, which were all audio recorded and fully transcribed prior to analysis.
Finally, in addition to these structured surveys and interviews, I attended training sessions for service providers on topics such as immigration, human trafficking, and language access; sat in on New York City Hall hearings on policies to protect and provide for domestic violence victims; and attended directors’ meetings with NYAWC funders. In all these diverse settings, I found new opportunities to gather material about the frustration faced by Asian women victims and service providers who seek to serve the Asian community.
Obstacles to Individuals
In the emerging field of structural analyses of domestic violence, much of the work so far has focused on structural obstacles 8 that can prevent individual women from receiving existing services. Foremost among these obstacles for Asian women are (a) U.S. immigration policies and (b) lack of language capacity.
Immigration status can be a serious barrier to women seeking and accessing services. At the NYAWC in 2009, about 18% of clients were undocumented, “out-of-status” immigrants, meaning that they were subject to deportation at any time. Roughly another third held a temporary visa or green card so that their stay in the United States was contingent on continued sponsorship from their spouses. As a result, case manager Julie Kim describes how clients’ fears of being arrested, deported, or losing their partners’ sponsorship can prevent them from contacting authorities for help, even in times of emergency. Moreover, because of this lack of police support, women are also indirectly denied access to multiple other social services. For example, an official police report is a required part of the application for safe housing from the New York City Housing Association (NYCHA). 9 Julie Kim describes how, because of fears about immigration status, clients who have been dealing with abuse for years still often have no reports that document their abuse and therefore, do not qualify for relocation.
Before 1990, under the 1986 Immigration Act, battered immigrant women who separated from their abusers before 2 years of marriage were categorically denied any social services and were immediately eligible for deportation. In 1990, a provision was added to the act waiving this 2-year requirement in cases of hardship caused by domestic violence (Crenshaw, 1991). Then, in 1994, Congress passed the Violence Against Women Act (VAWA), 10 with additional provisions for immigration relief in cases of domestic violence. In its present form, VAWA “provide[s] avenues for battered spouses and children to leave their abusers without jeopardizing their immigration status” (National Network to End Domestic Violence [NNEDV], 2005). Specifically, it allows battered immigrants to file “self-petitions” for immigration relief, asking to have their status adjusted to that of legal resident without the cooperation of their abusive spouse.
Despite these promising provisions, actually adjusting legal status through VAWA is far from easy. To file a successful application, women must provide a long list of necessary proofs and documentation, including a cover letter, a G-28 Notice of Entry, Form I-360 (Petition for Amerasian, Widow(er), or Special Immigrant), a self-petitioner’s declaration, evidence of a good faith marriage, and evidence of the abuser’s legal status. It is next to impossible for women to collect and assemble all these documents, especially if they are not comfortable reading English. What is more, once all the materials are finally collected, petitioners must pay a US$375 fee to submit the application. This sum is inordinately high for many women who have no assets of their own and are entirely economically dependent on their partner; however, the fee can only be waived with a great deal of additional paperwork, which is also inaccessible for women with limited English.
This requirement of English proficiency is not specific to VAWA discourse; in the United States, the entire system of social service provision appears to be only fully functional in English. Language barriers can be one of the most disempowering and frustrating obstacles on the path to social services, creating a dead end for many immigrant women who do not speak English fluently. Chinese case manager Amy Cheung described long waits when clients called for public assistance, saying, “Someone else might get seen in 2 hours, but to wait for an interpreter would take maybe 3 hours or 4 hours. It’s very discouraging to the clients sometimes” (personal communication, 2009). She also described similar delays often occurring in courtrooms so that clients are forced to wait days or even weeks until a suitable interpreter can be found. By that point, Amy described, many clients are so discouraged that “they just don’t want to do it anymore. It makes them want to give up.” Indeed, in a survey, the NYAWC asked its shelter residents what they would have done if the NYAWC were not available to provide interpretation and translation. One woman responded, “I don’t know. Don’t want to think about it.” Another wrote, “Stayed home or gone to a friend’s house. But I can’t stay at her house for long.”
While federal laws, 11 state laws, 12 and New York City regulations 13 all set out explicit provisions for providing language access services in conjunction with social services so that all services are made available in a client’s “best language,” there appears to be little or no enforcement of these policies. In the summer of 2007, an agency called Legal Services New York (LSNY) conducted a survey of 69 New York City offices of the Human Resources Administration/Department of Social Services (HRA/DSS; HRA, 2010). Because it receives funding from federal, state, and city governments, “HRA is mandated by federal, state, and local laws to provide language services to LEP [limited English-proficient] individuals” 14 (LSNY, 2007). However, the evaluation revealed a “systematic failure to provide legally mandated language access services.” For example, 13% of offices had not posted signs indicating the availability of language services, while another 20% had signs that were in bad condition or difficult to read. Fully two thirds of offices did not provide HRA applications in all 6 required languages, while 15% did not provide any translated applications at all. In addition, an almost ludicrous obstacle arises when translated applications and forms are made available, but no staff members are available who can read the forms and enter them into the system; in such cases, the office effectively discards all forms that are completed in a language other than English. The survey found that overall about 75% of the HRA offices were not in compliance with existing language service policies.
Obstacles to Organizations
Given the above ways that social services are structurally inaccessible to Asian immigrant women, it becomes clear that providing services targeted to that population’s specific needs is crucial to addressing domestic violence in the community. However, in practice, we see the opposite of what is needed: structural—and, specifically, economic—obstacles severely limit organizations’ ability to provide specialized, targeted social services for this population. As soon as I arrived at the NYAWC, multiple staff members described to me their concern about these economic obstacles, posed primarily by funders’ policies concerning program evaluation and resource allocation. They described how funders were increasingly granting or denying support based on the demonstration of standardized “success outcomes,” which were defined within evaluative frameworks that did not reflect their minority clientele’s unique needs.
In this section, I will explore the evaluation policies of two of the NYAWC’s most prominent funders. The Family Violence Prevention Services Administration (FVPSA) is a division of the U.S. Department of Health and Human Services. It has been the major source of government funding for domestic violence service providers since 1984 and has provided US$125 million in support in 2007. Meanwhile, Robin Hood is a New York City–based nonprofit organization founded in 1988, dedicated to reducing urban poverty. I will examine how both of these public and private funders tend to ask service providers to show “success outcomes” that are either unrealistic, or simply irrelevant for domestic violence survivors from Asian immigrant communities. As a result, these evaluations fail to capture the unique merits and successes of minority-targeted programs; they repeatedly devalue them as less successful and award them less funding.
FVPSA
FVPSA’s 2007 publication, Outcome Evaluation Strategies for Domestic Violence Service Programs Receiving FVPSA Funding: A Practical Guide (henceforth Guide) (Lyon & Sullivan, 2007), lays out evaluation criteria for the various kinds of services that FVPSA funds, including emergency shelter, support groups, counseling, and advocacy. From the start, the authors note the difficulty of evaluating domestic violence programs, which have their most meaningful impact in the long-term, 15 using only short-term measures. However, they continue as long-term outcomes can be nearly impossible for domestic violence agencies to measure, especially for anonymous clients or clients who relocate to escape abuse. In view of this difficulty, the Guide presents a review of recent evidence-based research to inform selection of short-term outcomes/metrics that can be used to reliably indicate favorable long-term outcomes. These include receiving higher education, finding employment, finding housing, receiving legal assistance, participating in support groups, developing a safety plan, and forging ties with the community.
After this fairly lengthy review, although, the Guide specifically highlights two short-term outcomes as the most useful surrogate measures for program success: (a) having a safety plan and (b) being aware of community resources. These outcomes are to be measured through two statements on client surveys: (a) “Because of [this service], I feel I know more ways to plan for my safety,” and (b) “Because of [this service], I feel I know more about community resources for domestic violence.” Clients check “yes” or “no” in response to each question, and service providers need to report 65% positive responses for both outcomes to receive FVPSA funding. While the Guide suggests that other survey questions be administered to aid in programs’ own self-evaluations, only these two statements are absolutely required to be administered, and only these come with strict requirements that need to be met to qualify for funding. 16
FVPSA outcomes
The first FVPSA outcome asks whether programs help clients “know more ways to plan for their safety.” To help programs meet this outcome, the FVPSA Guide’s authors recommend providing women with the numbers of police authorities, hotlines, and shelters. The manual also emphasizes the need to offer women “mutual support and understanding” throughout the process of becoming independent. By defining program success in this way, specifically in terms of women’s knowledge, FVPSA’s discourse automatically assumes a subject for whom a functional safety plan can actually be achieved solely through the provision of information (and perhaps empowering counseling).
However, this is far from the reality for immigrant women. Julie Kim noted this failure on the part of FVPSA discourse, saying, “Immigrant women have so many more needs. Basic things, like reading street signs or using the subway system—they need to learn to navigate a whole new world and a whole new country.” So for an immigrant woman, information is not always enough to create a working safety plan; learning the numbers of authorities or crisis hotlines will not mean that she has the English skills to communicate her needs in an emergency. Furthermore, Yi Hui Chang described how physical appearance can be another constraint on Asian women’s mobility, saying,
If she lives in an area where she is the only Asian, she walks down the street and . . . everyone’s looking at [her]. If she enters a shelter, people might make a mental note that she’s entering that house. So a lot of times, [especially] in a small town, you can’t just go into hiding. A regular safety plan would never work.”
In these ways, multiple Asian-specific characteristics, including LEP, unfamiliarity with American public transportation (and general inexperience being in America), and physical appearance, can pose serious barriers to formulating and executing a working safety plan. These barriers stand regardless of how “knowledgeable” a woman is, at least in terms of the phone numbers and shelter locations that FVPSA values. However, FVPSA’s evaluation discourse unquestioningly equates such knowledge with safety. By doing so, it assumes that all domestic violence survivors are English-competent, nonminority, American citizens—and it leaves no discursive space in which to imagine or address the unique barriers minorities face to safety planning. Therefore, FVPSA makes no provisions for minority-targeted services that report lower success rates for this outcome; such services simply get poor evaluations and less funding.
The second FVPSA success outcome is helping women “know more about community resources.” The Guide specifically notes that they chose this metric so as to focus on “building stronger and safer communities,” recognizing how conditions outside the home can exacerbate or ameliorate a domestic violence situation. Using this outcome as a funding criterion is therefore meant to encourage agencies to help individuals reconnect with their communities and find places of refuge within it.
This funding criterion is again problematic because it assumes that women come from a community in which it is feasible and safe to use community resources. Case manager Christina Kim describes how many immigrants live in very intimate communities, such as the Korean community in Flushing, New York, where there is virtually no anonymity. This can make seeking out community resources dangerous and nearly impossible. In these cases, she said, it can be much more prudent to remove a woman from her community—as quickly, completely, and secretly, as possible—rather than try to connect her with community resources.
Furthermore, there is a very real possibility, especially in insular Asian communities, that community resources may not exist at all. The NYAWC’s community outreach coordinator Fronthy Nguyen described how service providers often confront a long-standing tradition of silence surrounding domestic abuse in Asian communities, saying, “In Chinatown, for instance, people just look at us like . . . ‘Oh no! We’re Asian, we don’t talk about that.’ They don’t want to talk about or address it at all” (personal interview, 2009). Therefore, programs that seek to provide domestic violence education as well as more material forms of shelter and aid often have a very hard time establishing themselves in Asian communities, and social service providers cannot help Asian women “know about” resources in their communities that are nonexistent. By assuming that all domestic violence survivors come from a community in which support is available and safe to use, FVPSA holds service providers to a standard of provision that is not designed in light of their clients’ actual options and needs.
This method of evaluation affects minority- and immigrant-targeted organizations, such as the NYAWC, in two ways. First, they fare worse than their competitors in evaluations, because there is no discursive recognition of, or compensation for, the specific obstacles they face in serving minorities. For example, with regard to the Asian client who was too easily recognized by her physical appearance to leave home freely, the staff made special efforts to formulate a working safety plan for her. They needed to arrange her transportation from her home to a shelter to the airport to a different city, formulating a schedule “down to the last hour” to ensure that she would not be in any one place long enough to be found. However, rather than receiving more support because of their clients’ additional service needs, they are ranked less favorably and receive less funding than competitors, because these safety plans are so much more difficult to formulate and implement satisfactorily.
Second, the NYAWC can actually be hindered from providing services that are optimally tailored to their clients’ needs. Services that are specifically targeted to minority populations may want to aim at alternative outcomes than increasing knowledge, recognizing that their clients encounter additional major barriers to escaping abuse. For example, an ESL (English as a second language) program may help women learn the English they need to actually implement a safety plan and use crisis hotlines or safe houses. However, because the FVPSA Guide exclusively values individual knowledge as the key to remedying abuse, it effectively confines the DV program’s role to knowledge provision. There is no discursive space in which to recognize linguistic, economic, and other structural factors that affect women’s experiences of abuse; therefore, services that address these structural factors, but do not directly lead to positive knowledge outcomes, are not legible or deemed valuable.
The FVPSA Guide states that the goal of program evaluation is to “identify and build upon those efforts that are helpful to women with abusive partners; at the same time, we don’t want to continue putting time and resources into efforts that are not helpful or important.” However, at least in its current form, FVPSA discourse exclusively supports a form of service provision that benefits specifically English-competent, citizen subjects. It fails to provide for minorities and immigrants, by neither recognizing their unique needs nor regarding programs targeted to meet those needs as “successful.” As a result, FVPSA can withhold funding from these programs, posing a serious structural obstacle to making essential services available to the Asian community. In this way, uninformed and biased program evaluation leads to the very misguided allocation of time and resources that it seeks to avoid.
Robin Hood
Domestic violence organizations are not only compelled to show positive outcomes to the government to secure state funding but also face pressure to show these outcomes to their private funders. These private, non-governmental organizations each have their own distinct modes of program evaluation, based on various measures of program “success.” Among these NGOs, Robin Hood (one of the NYAWC’s largest funders) has one of the most intricate and quantitative methods of deciding which organizations receive their grants. Robin Hood (2009) allocates funding according to a “complex computation” of the ratio of a program’s budget to the benefits it provides its clients.
While quantifying a program’s budget is fairly straightforward, measuring the benefits provided by the program can be much trickier. For most programs that Robin Hood supports, such as early childhood, education, and job training programs, benefits are calculated for how much the program boosts the future earning of poor families above that which they would have otherwise earned. However, domestic violence programs are classified as “Survival” programs that offer “palliative care” rather than a “cure” for poverty by providing things such as housing, food, or access to medical care. Therefore, benefits provided are calculated differently: Rather than boosted earnings, Robin Hood estimates how much the program improves a family’s standard of living, general well-being, and health. With the help of public health researchers and health economists, they estimate the “monetary value of improved health” using a formulation called Quality-Adjusted Life Years (QALYs).
QALYs
The QALY is a widely used measure of how health and well-being are compromised for various disease states.
17
A year of “life in perfect health” is assigned the value of 1.0 QALY, while death is assigned a value of 0.0. Years of life in poor health are assigned values between 0 and 1, depending on the severity and degree of suffering associated with the illness. Dr. Peter Muennig of the Columbia School of Public Health gives the following example:
Say you’re missing a leg. Now, imagine that an evil genie appeared at the end of your bed, and it said, “I’ll trade you a 10% chance of death for having your leg again.” And the person may say, “No, that’s too high,” and they negotiate back and forth. The risk of death you’re willing to accept, they call it the “equivalent” between living in perfect health and death. So if you were willing to accept a 10% chance of death, living without the leg is worth 90% of a full year of health. (personal interview, 2009)
Thus, a year of life without the leg would be valued at 0.9 QALYs. To determine QALYs originally, individuals with various health conditions were surveyed about how much life they would risk to be restored to full health. Based on this research, QALY equivalents were set for many different health conditions, ostensibly providing a way to compare the degree of suffering for very different ailments. By the 1990s, the QALY framework became “widely accepted as the reference standard” for health outcomes and is now used routinely in cost-effectiveness comparisons of various health interventions. 18 Robin Hood’s funding allocation is entirely based on this kind of comparative cost-effective analyses of different programs, with effectiveness measured in terms of QALYs saved.
However, in the article, “Calculating QALYs,” Franco Sassi (2006) notes that despite this popularity, “the debate on [the QALY’s] theoretical underpinnings and practical implications is still ongoing.” While the QALY can appear to be a value-neutral, objective measure, its definition is based on a set of theories and assumptions that are highly subjective, political, and controversial. One of the most outspoken critics of QALYs was John Rawles (1989), who called the proposal to distribute care and services according to the price of QALYs based on “false premises, faulty reasoning and unjust principles.” In “Castigating QALYs,” Rawles describes several ways in which a QALY-based evaluation system unjustly favors certain kinds of interventions over others.
First, Rawles points out, QALYs favor quick and complete fixes. A hip replacement is shown to be highly cost-effective in terms of QALYs gained per dollar, because it is a one-time operation that restores an individual to full functionality. Meanwhile, a treatment such as regular hemodialyses, which is repeated and ongoing, accumulates to a much higher total cost, while not effecting a complete cure. In fact, the therapy actually appears to prolong the patient’s suffering, resulting in the loss, rather than gain, of QALYs over time. Therefore, hip replacement is considerably favored over hemodialyses in QALY-based cost-effectiveness analyses. However, asks Rawles, if someone were to ask you to choose between surviving lethal renal failure and overcoming arthritis, which would you choose? By distributing resources with the sole view of getting the best return of QALYs per investment, “those whose illnesses happen to be cheaper to treat will be treated in preference to those whose treatments are expensive,” regardless of whether they are actually life-threatening (Rawles, 1989).
This observation leads into Rawles’ second critique, which is that QALYs greatly devalue human life. According to the logic of QALY calculations, a value of 1.0 QALY, representing a completely healthy life, is ideal only because it is a complete “absence of suffering.” As a result, the ultimate goal of health interventions becomes only to eliminate suffering, rather than to preserve life. However, Rawles (1989) writes, “to reduce the value of life to the absence of disability or distress is to undervalue existence very greatly indeed. Life is valued for infinitely more reasons than absence of suffering.” QALYs have no way to assign value to a life involving a great deal of suffering and distress, or to an intervention that preserves such a life.
Structural obstacles for the NYAWC
How do QALY-based evaluations, then, affect funding and support for domestic violence programs? Theoretically, as Dr. Muennig explained, women who receive services are expected to “have a healthier, higher quality of life than women who remain with their abusers . . . and the QALY is gonna capture that. By providing counseling and support, you’re gonna get a lot of QALY gains” (personal interview, 2009). For example, Robin Hood calculates that domestic violence programs save women QALYs by averting injuries, chronic pain conditions, and psychological conditions such as depression and posttraumatic stress disorder (PTSD). It also considers QALYs gained for the women’s children, who avoid psychological damage from witnessing the abuse.
However, as Rawles (1989) suggested, QALY-based evaluations favor one-time fixes over treatments for long-term ailments. Most of the above health conditions associated with domestic violence are not the ones remediable by a quick, one-time intervention. A domestic violence victim must engage in a long and difficult process of receiving social services to build an independent life; during this entire time, she may experience little positive change in her health status and is likely to continue to experience fear, anxiety, and isolation. From the survivor’s point of view, this suffering is still a gain on the path toward ultimate independence and healing; however, QALYs’ valuation of life solely as “the absence of suffering” means that this suffering is seen only as a loss. Because significant time and resources must be invested to treat these conditions, and without immediate effect, Robin Hood’s algorithm does not view domestic violence programs as producing favorable returns of QALYs per dollar. Therefore, Robin Hood may favor other, quicker-fix poverty alleviation programs above domestic violence services, even if those programs address a problem that caused less suffering to begin with.
While QALY-based evaluations can lead to underfunding of all domestic violence programs, my time at the NYAWC also revealed how they can specifically disadvantage interventions targeted to minority populations. In Robin Hood’s algorithm, domestic violence programs are viewed as the most “effective” when they help abused women get psychological counseling, find employment, and get temporary housing. Research exists (despite the questionable applicability of its sample, as noted above) to show that these steps are associated with significant QALY gains. Therefore, domestic violence programs need to demonstrate that they help women achieve these outcomes to be awarded Robin Hood funding.
However, as we saw before in FVPSA’s evaluations, these outcomes again assume a victim who is an English-competent, American citizen. Many clients who are recent immigrants find it very difficult to achieve these outcomes, because they must first acquire basic “survival skills,” such as speaking English and using public transportation. Before attending appointments or filling out job and housing applications, many of the NYAWC’s clients need to learn how to speak, read, and write basic English, how to interpret street signs and use the subway by themselves, and how to open a bank account and manage her own funds. Furthermore, many of the NYAWC’s clients are not legally eligible to work in the United States because of their immigrant status in which case, they must also apply for immigration relief and obtain a government-issued ID and work authorization. NYAWC case managers help clients through all of these essential intermediate steps toward self-sufficiency; however, because these services do not result in immediate material benefits to the women, Robin Hood’s “complex computations” do not recognize them as achieving “positive outcomes” that can be rewarded in evaluations. Meanwhile, the NYAWC is not able to show very positive results for the outcomes that Robin Hood does recognize.
At an annual funding meeting with a Robin Hood representative, NYAWC staff members brought up this issue of how they felt misrepresented and undervalued by Robin Hood’s evaluation. They suggested alternative measures they could use to report their program’s success: a list of what they called “Client Milestones,” or skills (e.g., English, ability to use public transportation, financial planning) that would allow clients to eventually achieve Robin Hood’s original goals. However, when this list of Milestones was presented to the Robin Hood representative, she summarily dismissed the idea. The representative said, “Learning [skills] is not really an outcome at all. What matters is the endgame.” The staff members left the meeting disappointed, frustrated, and entirely uncertain of being awarded funding the next year.
In this way, the QALY-based evaluation discourse of private funding organizations, such as that of FVPSA, also produces and enforces conformity to particular subject positions for domestic violence survivors. This subject position is a distinctly gendered, English-competent citizen—and, specifically, a working citizen, reinforcing the limitation on benefits for unproductive dependents. 19 Because this subject position is entirely irrelevant to the lived experiences of many clients at the NYAWC, it is obvious that Robin Hood does not imagine these women as the beneficiaries of their programs. Therefore, services targeted to such illegible recipients—such as the life skills and language education programs that they most need—are not seen as valuable or fundable.
A final result of valuing domestic violence programs in terms of QALYs gained is that they again constrain interventions to the individual scale. Because QALYs are calculated based on individual health outcomes and expenditures, they assign no value to programs that seek to make an impact on a larger, community-wide scale. In its publication Measuring Success, Robin Hood specifically highlights that their evaluative algorithms only calculate benefits to individuals, with the justification,
Donors do not give Robin Hood money so we can go out and save money for other taxpayers. Instead, donors give money to improve the living conditions of people living in poverty. The benefits that Robin Hood counts are the private benefits that accrue to poor people as a result of our grants. We don’t count public benefits to society at large . . . [Thus,] our metrics keep us focused on our mission. (Robin Hood, 2009)
This focus on measuring benefits to individuals, rather than communities or society, recalls the individualism we saw in the FVPSA discourse. 20 Because they do not value change to the underlying community or social structure that creates and supports these obstacles, Robin Hood, like FVPSA, effectively favors “palliative” programs that treat the effects of domestic violence for specific individuals, rather than proactive programs that address their underlying causes.
For example, one type of “proactive” intervention that is basically disqualified from funding by Robin Hood’s criteria is community outreach aimed at raising awareness of domestic violence. These interventions can be powerful tools for combating domestic violence in all communities but especially in minority communities, where education and outreach about domestic violence have historically been lacking. Attorney Madeline Garcia, a regular collaborator with the NYAWC, expressed the need for “money not just for the women, but for the whole immigrant community. We need money to do outreach to them and find interpreters from among them, so we can empower them to help each other.” However, as community outreach coordinator Fronthy Nguyen described, these interventions can only report “outcomes,” such as the number of people contacted; there is little to no data on how such interventions affect the well-being of individual domestic violence survivors. Therefore, these programs appear close to valueless when plugged into a QALY-based evaluative algorithm. This is especially harmful to minority communities for two reasons: first, because outreach is so conspicuously lacking and needful, and second, because it does not support community engagement and the exercise of community agency in efforts to end and prevent violence. Instead, it supports a mode of social service provision that seems to regard domestic violence as an inherent flaw in minority culture and rely on outside players and service providers for its rectification.
Conclusion
Historically, U.S. feminists have made claims to equality on the basis of (a) citizenship and (b) equal productivity to men. But immigrant women are by definition excluded from the category of “citizen,” or granted it only on such disempowering terms that the very benefits that they originally sought are undermined. At the same time, as Chang (2000) notes, productivity-based claims have always excluded immigrant women who, at least temporarily, lack the social mobility and training to do most kinds of labor. Therefore, these categories of “citizen” and “productive” do not create a discursive space within which minority and immigrant women can be recognized as legible recipients of domestic violence services.
To address the needs of Asian immigrant survivors of domestic violence, then, we must challenge these categories that currently delimit entitlement to social services. On one hand, we must recognize the specificity of Asian women’s social location, which places them outside these categories and creates for them unique service needs. This process of recognizing difference need not reconstitute existing patterns of oppression; recognizing the uniqueness of Asian immigrant women’s experiences need not reinforce their positions as victims of their partners as well as of the structural forces that compound abuse. Rather, this recognition can, and must, shape our efforts to provide useful aid to these women. Not only can it be directly translated into the design of more effective structural interventions to help Asian immigrant women escape abuse; at the same time, it can also put pressure on funding agencies to finally value and support structural interventions targeted to this population. 21
Finally, recognizing these differences between Asian immigrant women and the rest of “mainstream” society—the differences between White and minority, immigrant and citizen, suffering and non-suffering—not only allows us to help women on the other side of various social divides but is also an essential step to bridging those gaps and joining with suffering women in their own cause. Only such a vibrant and inclusive coalition can make real, large-scale change to expand the system of social service provision in our society. In this way, understanding specificity can ultimately allow us to recognize the universality of women’s right to health, independence, and well-being, and demand, for them and with them, the social services that can help them attain it.
Footnotes
Acknowledgements
I would like to thank Melanie Boyd and Laura Whitman for their guidance and encouragement throughout the writing and editing of this manuscript. I would also like to thank Julie Kim and the staff of the New York Asian Women’s Center (NYAWC) for so generously welcoming me into their circle and sharing their experiences. Finally, I would especially like to thank You-me Park for her invaluable insights and critiques, and for always being my greatest teacher, supporter, and inspiration.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the Josephine de Karman Fellowship Trust, the Association of Asian American Yale Alumni, and the Yale College Dean’s Office of Yale University.