by identifying them as atomistic, interchangeable beings. Indeed, distributive theories of justice treat as irrelevant such politically significant features of persons as their gender, race, and class. In contrast, Young's social justice conception recognizes the importance of attending to these kinds of considerations in order to correct the effects of oppressive practices that arise from the biases inherent in sexism, racism, classism, and so on. Treating people as if they are equally situated with respect to power and privilege when they are not is a way of maintaining existing power structures. Understanding justice as social justice, defined as including concern for matters of domination and oppression, invites us to respond to the crippling biases and differing access to power that underlie systematic patterns of oppression. Even when our focus is purely distributive, it is important to raise questions about the social context in which distributions are carried out.
Thus, when we are trying to decide on a matter of social policy with respect to an oppressed group, this conception of social justice requires us to investigate how the proposed practice is likely to affect current patterns of domination and oppression in society: i.e., is it likely to worsen or improve the existing levels of oppression in society?3 Among other ethical questions, we need to ask of each practice or policy reviewed, whose interests does it serve and whose does it harm? In the current context, where the task is to decide about policy regarding the inclusion of women in clinical studies, the ethical questions raised should include questions about how the proposed policy will affect the oppressed status of all women and, more specifically, how it will affect the status of women who are multiply oppressed by virtue of the intersection of gender with their race, class, age, and so on. By this measure, we can identify several specific areas of feminist concern, including: (1) some or all women may be unjustly excluded from some studies and suffer as a result; (2) women may be unjustly enrolled in studies that expose them to risk without offering appropriate benefits; (3) the research agenda may be unresponsive to the interests of oppressed groups; and (4) most generally, the process by which research decisions are made and carried out may maintain and promote oppressive practices. I will expand on each of these concerns in turn.
The exclusion of women from important clinical studies is the best known of the problems of injustice identified as falling within the scope of the topic of women's role in clinical studies. Historically, many studies of diseases that are common to both sexes have systematically excluded women from participation, so the necessary data for guiding treatment decisions for women are unavailable.4 Women's health care must often be based on untested inferences from data collected about men, but because there are important physiological differences between women and men, such inferences cannot always be presumed to be reliable; and, even when some data are collected about women's responses to the treatment in question, we may lack information about how a proposed treatment will affect specific groups of women (e.g., those who