The above estimates, although alarmingly high, are underestimates for the several reasons, including memory error, small samples, reluctance to recall traumatic memories, fear of reprisal, and embarrassment (DeKeseredy, 1995; Kennedy & Dutton, 1989; Schwartz, 2000; Smith, 1987, 1994). Sizeable portions of female public housing residents are also recent immigrants from war-torn countries, dictatorships or police states. For example, about 20 percent of the women who answered the QNLS stated that they were refugees or recent immigrants to Canada, and the research team strongly suspects that this group constituted a much higher percentage of those women who did not complete the questionnaire (DeKeseredy et al., 2003). Hence, many women probably did not believe assurances of confidentiality because of prior traumatic experiences with authority figures (e.g., police) (Schwartz, 2000). It is also possible that language barriers precluded many ethnic minority women from filling out the QNLS (Koss, 1993).

          It is fair to argue, then, that had Ireland et al. (2003) included measures of male-to-female violence, significant differences in rates of violence reported by Rochester public housing residents and those who do not live in public housing would have emerged. Moreover, the rates of violence reported by their adolescent Pittsburgh public housing residents would have also been even higher than they were compared to those gleaned from people not living in public housing. However, this is an empirical issue that can only be addressed empirically. Hopefully, future surveys on violence in public housing communities will pay more careful attention to the gendered nature of criminal victimization in and outside these neighborhoods.

THE VALUE OF GATHERING QUALITATIVE DATA

          As Lab (2003, p. 42) reminds us in his response to Ireland et al.’s (2003) article, many criminologists avoid, for one reason or another, gathering qualitative data, which contributes to “the lack of contextual depth in much research.” Hence, in addition to measuring violence against women and other forms of woman abuse (e.g., verbal harassment in public places), researchers should conduct rich in-depth interviews and use observational techniques. These are methods that feminist and other scholars have used for years and they yield valuable data that can contribute to the development of even better surveys, as well as theoretical perspectives. Consider Bourgois’ (1995) path-breaking ethnographic study of crack dealing in East Harlem. He found that violence against women in public housing areas is fostered by a “crisis in patriarchy” generated by men’s inability to find legitimate employment. Indeed, as others have argued (DeKeseredy, Alvi, & Schwartz, 2003; DeKeseredy & Schwartz, 2002; Websdale, 2001), psychological stress spawned by the failure to fulfill the role of “bread winner” in a current political economic order characterized by the rapid disappearance of working-class jobs motivates economically and socially disenfranchised men to beat or sexually assault their female partners “to reassert their grandfathers’ lost autocratic control over the household….” (Bourgois, 1995, p. 85).

          It would be very difficult, if not impossible, to uncover such data using only survey research techniques. Similarly, it is highly unlikely that a survey would reveal what Websdale (2001) found in his ethnographic study of African-American public housing residents in Nashville, Tennessee. In the community he examined, violence against women occurred in the context of “the frantic competition between men in the projects over every conceivable issue, including sexual access to women” (2001, p. 134). His data, those gathered by Bourgois (1995), and in-depth interviews conducted by Renzetti and Maier (2002) also call into question the value of reducing violence rates through changing the physical structure of public housing. For example, if poor men are primarily motivated to beat, rape, and psychologically abuse women because of perceived threats to their masculine identity, transferring them and women into smaller housing developments will do little, if anything, to eliminate the alarming amount of violence-induced pain and suffering that occurs behind closed doors. To develop truly effective means of curbing “intimate intrusions” (Stanko, 1985), the major causes must first be identified. It is to this issue that we turn to next.

THE NEED TO THEORIZE VIOLENCE AGAINST WOMEN IN PUBLIC HOUSING

          Commonly espoused theories of crime in public housing communities, such as defensible space, social disorganization, and “broken windows,” may account for adolescent street crime in these areas (see, for example, Ireland et al., 2003), but they cannot explain why so many female public housing residents are physically and sexually victimized by male partners and acquaintances. Even feminist scholars have devoted little theoretical attention to the abuse of economically marginalized women in North American public housing. Perhaps the paucity of theoretical work on this topic is due in large part to the fact that research on crime in public housing is in a state of infancy. Regardless of what accounts for the lack of theory testing and construction, ethnographic studies reviewed here and elsewhere (see Renzetti & Maier, 2002) make it clear that economic and cultural variables need to be incorporated into the “current theoretical mix” (Raphael, 2001, p. 454).

          To the best of our knowledge, only one theoretical perspective on violence against women in public housing was constructed accordingly (see DeKeseredy & Schwartz, 2002). Heavily influenced by theoretical work done by Sernau (2001), DeKeseredy and Schwartz (1993) and Young (1999), this economic exclusion/male peer support model combines both macro- and micro-level factors, such as formal labor market exclusion and patriarchal male subcultural dynamics. Still, at this point in time, the model lacks strong empirical support because it has not been tested, and future testing will depend to some extent on the availability of funding. As several studies reveal (Savelsberg et al., 2002; Schacht & Eitzen, 1990), state funded research is more likely to focus on micro- than on macro-level analysis, which is one of the key reasons why virtually all of the public housing studies done so far have “failed to theorize the state or its role constructing social life” (Venkatesh, 2003, p. 53). Moreover, it is doubtful that federal agencies driven by an intense neo-conservative agenda will fund research on how the current political economic order contributes to violence against women.

          Neglecting to address the state and broader political economic forces is more evident in work done by consultants and other nonacademic researchers because their livelihoods depend primarily on state-sponsored grants and contracts. This is not to say, however, that all nonacademic researchers are driven by the principles of “administrative criminology.” For example, Popkin et al. (2000) sharply critique the Chicago Public Housing Authority, and their book was co-authored by several researchers who are not based in institutions of higher learning. Social scientists affiliated with the Center for Impact Research (CIR) in Chicago also devote much