Empowerment Through Self-Subordination: Microcredit and Women's Agency

1 Empowerment Through Self‐Subordination?: Microcredit and Women’s Agency Serene J. Khader Forthcoming in Poverty, Agency, and Human Rights Oxford University Press, Diana Meyers, ed. Poverty can undermine people’s agency. The poor often face restricted options and develop limited senses of what they can be and do. According to Deepa Narayan’s famous Voices of the Poor study, the poor themselves characterize their condition as one of acute powerlessness (Narayan 2000, 38‐40). It thus seems logical that successful antipoverty interventions should enhance agency. The idea of a tight correlation between reduced poverty and increased agency has been so widespread in the development community that, until quite recently, traditional poverty indicators often functioned as proxies for agency (Alkire 2007, 10‐11.) Data from the last fifteen years on how antipoverty interventions affect women have called this correlation into question (Alkire 2007, 10‐11). Antipoverty interventions often yield mixed results for women’s agency and empowerment. For example, some studies suggest that income interventions do little to change women’s actual involvement in remunerated economic activity (Goetz and Sen Gupta 1996). Some also show that income interventions do nothing to change women’s sense of inferior entitlement to household resources or decisionmaking authority (Sen 1999; Osmani 1998; Cheston and Kuhn 2002,19). Other studies suggest that increasing women’s access to paid work decreases their political participation (Batliwala and Dhanraj 2007, 24), that increasing women’s access to education can accompany increased support for female genital cutting (Agot 2007, 290)i, and so on. Underlying the puzzlement about this data seems to be a line of reasoning roughly like this: reducing poverty increases agency, and women’s empowerment is the increase in women’s agency, so reducing women’s poverty should empower women. A popular explanation of why antipoverty interventions sometimes seem to fail to empower women is that empowerment is a process (Kabeer 1999; Kabeer 2001; Agot 2007; Nagar and Raju 2003). According to this explanation, the issue is not that antipoverty interventions have failed to empower women; it is that they have not gone far enough to utterly transform gender relations. This explanation is applied to situations where antipoverty interventions seem to leave women’s gender status untouched and ones where they seem to worsen gender inequality. For example, Susy Cheston and Lisa Kuhn (2002) remark that microcredit interventions in Ghana have not changed prevailing views about whether women should be allowed to lead men. However, Cheston and Kuhn express optimism that women’s participation in mixed gender banks will lead to increased comfort with women’s leadership in the wider social world (2002, 38). Naila Kabeer goes farther and argues that what appear to be worsening gender relations may actually be part of empowering processes. According to her, the upsurge in domestic violence the accompanies microcredit is men really resisting their impending loss of power (2001, 65‐66). According to such narratives, antipoverty interventions can constitute steps in the right direction even when they fail to directly challenge gender hierarchy. 2 In this chapter, I call for skepticism of the idea that antipoverty interventions contribute to women’s empowerment by enhancing their agency. My claim is not categorical; I believe antipoverty interventions can empower women. I am more interested in how readily we should apply “steps in the right direction” explanations to cases where antipoverty interventions fail to bring about feminist change. I will argue that such explanations fail to take the following fact seriously: that increases in women’s agency can result from decreases in the egalitarianism of gender relations. This possibility is generated by what I call the “self‐subordination social recognition paradox” or “SSRP.” The paradox is that women can often gain welfare by complying with and internalizing oppressive norms. Access to many goods depends on social recognition, and, under patriarchy, women often have to subordinate themselves to achieve social recognition. The popular understanding of agency ignores this paradox by failing to distinguish what I call “welfare agency”(the ability to enhance one’s welfare) from feminist agency (the ability to identify and change sexist norms). On the popular definition of agency, agency is a single good whose increase tracks increases in a person’s sense of herself as equal. The possibility of women gaining agency by increasing their investment in sexist norms is of more than theoretical interest, as I will show in a discussion of microcredit interventions in South Asia. Microcredit, by materially rewarding female seclusion and the view of women as collateral, may offer women new reasons to act according to, and endorse ,patriarchal expectations. When I say that antipoverty interventions can strengthen the incentives for women to internalize and comply with sexist norms, I do not simply mean that they “reinforce” or fail to change oppressive gender expectations. I will revisit claims that antipoverty interventions can be criticized for “reinforcing” or failing to change gender subordination in my conclusion. My claim about how antipoverty interventions affect women’s agency is stronger in one sense and weaker in another. It is weaker in this sense: I am agnostic about whether to call antipoverty interventions that leave gender inequality intact “failures.” But my claim is stronger in that it suggests that failing to empower women can have worse consequences than simply leaving inegalitarian gender relations intact. Interventions may strengthen patriarchal structures by giving women new incentives to do what they dictate and see them as just. Poverty sometimes gives women reason to question the dictates of patriarchy. Antipoverty interventions can decrease the force of those reasons by increasing the rewards for patriarchally prescribed behavior. The chapter unfolds as follows. I argue in the first section that “right direction” explanations rely on an understanding of women’s empowerment that is counterintuitive, inconsistent with common usage, and that excludes the possibility of women increasing their agency by growing in their acceptance of oppressive norms. Specifically, the operative notion of women’s empowerment conflates empowerment with individual women’s ability to enhance their welfare. In the second section, I argue that, access to benefits often depends on social recognition, and patriarchy creates sexist conditions of social recognition. The upshot is that women often gain social and material benefits from complying with patriarchal norms—and, where social and material rewards for complying with patriarchal norms align, women also gain senses of 3 coherence and self‐esteem from internalizing them. It follows that increases in women’s welfare agency often provide women new incentives to accept patriarchy. In the third section, I examine Kabeer’s now‐famous argument that microcredit empowers women. I argue that her inadvertent conflation of feminist and welfare agency causes her sanguine interpretation of the mixed data on women and microcredit. Kabeer says microcredit increases perceptions of women’s value, but I believe we must acknowledge the fact that women are valued for their increased ability to meet patriarchal expectations. In the conclusion, I ask what general lessons about assessing antipoverty interventions we might draw from my analysis. 1. Defining Women’s Empowerment The claim that antipoverty interventions impel women’s empowerment by enhancing their agency trades on a distorting conflation. This conflation is of enhancements in individual women’s welfare agency with women’s empowerment. An intuitive understanding of the idea of women’s empowerment, as well as common usage of the term, suggests that the concepts are nonidentical. Development theorists tend to agree that empowerment involves the use of conscientizing processes to increase a person’s ability to pursue her welfare (Nagar and Raju 2003; Malhotra, Shuler, and Boender 2002;Kabeer 1999).ii Though many philosophers allow that a person can agentically decide not to value her welfare, development theorists tend to attribute cases where people fail to value their welfare to lacks of agency.iii Call knowledge that one’s welfare is of value, coupled with the ability to pursue it “welfare agency.” It may seem that women’s empowerment occurs when each woman in a group experiences enhanced welfare agency. This idea is conceptually fraught, for reasons I will explain in a moment. First, let us observe that—at least intuitively‐‐ women’s empowerment requires women’s increased ability and desire to agitate for more egalitarian gender relations. When women are empowered, they act—or believe it is worth acting—in ways that challenge women’s subordination. To say this is not to deny that women’s empowerment comes in degrees; one may, for instance, start by questioning women’s subordination in the political arena but not in the home‐‐ or one may have only an inchoate sense that women’s subordination is wrong. My point is that interventions that do nothing to affect the subordination of women, or women’s awareness of their own subordination, do not empower women. Nor do interventions that increase women’s subordination. Though the term “women’s empowerment” is not used consistently in development discourse, both mainstream and radical development actors use it in a way that suggests that it entails opposition to gender hierarchy. Even the language of the third Millenium Development Goal treats the goals of empowering women and eliminating gender inequality as linked (United Nations). Most empowerment theorists explicitly discuss women’s empowerment as involving action and/or conscientization toward the end of decreasing sexist oppression (Nagar and Raju 2003; Mayoux 1998; Swain 2007; Sen 1993 (cited in Malhotra, Shuler, and Boender 2002); Holvoet 2005).iv Call the type of agency that challenges sexist norms‐‐ the type needed for women’s empowerment‐‐ “feminist agency.” 4 The logical points that distinguish feminist and welfare agency are dual. First, an agent’s welfare agency can be enhanced without her feminist agency being enhanced. I may have more food in my stomach than I used to, for instance, and still give my husband the best food—all the while believing this is the right thing to do. Second, an agent’s welfare agency may be enhanced through decreases in her feminist agency. Consider the case of women in South Asia who comply with sexist food distribution norms to keep male relatives happy— and do so in contexts where their access to virtually all goods depends on male guardianship.v Assume that these women are not in a position to alter the “patriarchally risky”vi features of their society and thus decide to stop wasting their time questioning their validity. These women reduce their feminist consciousness and gender‐role violating actions to increase their welfare agency. Their access to food, income, and security depend on their fulfilling subordinate gender roles. If these two points are correct, the logical relationship between feminist and welfare agency is this: they are nonidentical but may overlap. The same action may increase both feminist and welfare agency—as in, say, the choice to leave an abusive husband in a context where this does not expose a woman to further violence or cause economic or social death. The claim that it is sometimes not in women’s welfare interests to increase their feminist agency may seem paradoxical. It may seem to entail the view that sexist norms do not harm women. Here we need to remember that a norm is sexist primarily in virtue of its harming women as a group (Frye 1983; Cudd 2006). To say that women can gain welfare agency from complying with sexist norms is not to say that they gain it from the existence of those norms. It is to say instead that, if the conditions under which the norms obtain are fixed (which it is sometimes reasonable for women to assume), individual women gain welfare agency from complying that they would not get from resisting. It is certainly true that individual women usually experience negative effects from complying with oppressive norms. But we need to remember—and a longer discussion of this fact will be the topic of the next section—that patriarchal societies often make behavior with inherently negative welfare effects a prerequisite for accessing other goods that are also constitutive of welfare. Following Uma Narayan (2002), I term this phenomenon “harm‐benefit bundling.” Take the case of a woman who feeds superior food to her husband and suffers poor nutrition as a result. If insisting on an equal claim tofood will result in the loss of male guardianship‐‐ and thus potential loss of income, shelter, and safety—compromising her nutrition is probably a welfare‐maximizing move. Of course, the real problem here is the structural constraint on women’s options. But this does not change the fact that, if changing the structure is out of a woman’s power, she can advance her welfare by doing what patriarchy prescribes.vii If an action may enhance a woman’s welfare agency while undermining her feminist agency, successful antipoverty interventions may not always be as innocent as they initially seem. We should be skeptical of the idea that successful antipoverty interventions are always—or even usually— a) steps toward women’s empowerment or b) neutral with respect to it. That women can gain welfare agency by trading away feminist agency means that enhancing women’s welfare agency can happen at the 5 expense of women’s empowerment. We need to ask why recent empowerment theorists insist that successful antipoverty interventions are either steps in the right direction or, at worst, unrelated to women’s status. The answer seems to lie in the mechanisms by which empowerment is supposed to occur. Antipoverty interventions are supposed to work on people’s self‐concepts. According to many empowerment theorists, antipoverty interventions empower women by expanding their senses of what they are capable of. The pre‐empowered person characteristically holds two attitudes—one toward herself and the other toward the world. The attitude toward herself involves low self‐worth stemming from general self‐ devaluation. The attitude toward the world is one of uncritical acceptance of the power status quo—a sense that gender inequality is just a part of the natural order of things.viii Antipoverty interventions empower women, because a newfound sense of worth and/ or desire to question causes or manifests skepticism of patriarchal values. Arguments to this effect abound in the microcredit literature. Kabeer argues that microcredit increases women’s experience of being valued by others, and this causes women to perceive themselves as capable of bargaining with their husbands (1998; 2001). Cheston and Kuhn (2002,29‐30) argue that microcredit helps women gain greater senses of self‐ efficacy—even when this is simply increased efficacy at fulfilling traditional roles. They assert that increased self‐efficacy often translates into a desire to challenge oppressive structures. Linda Mayoux argues that women entrepreneurs have increased confidence and skills and that the visibility of such women can increase women’ overall status (2007, 39). Even skeptics seem to share the view that antipoverty interventions would empower women if they successfully changed women’s beliefs about what they are capable of. These theorists tend to argue—not that microcredit has negative impacts on women’s self‐concepts—but rather that microcredit fails to empower women when it fails to work on their self‐concepts. Simeen Mahmud argues that microcredit does not expand women’s sense of what they are capable of because it tracks them into poorly compensated tasks they were already doing anyway (2003,602‐603). I believe this widely‐held view‐‐ that antipoverty interventions that enhance women’s senses of self will increase their desire to change patriarchy—is based on two faulty assumptions about agency. The first is what I call the “cumulative assumption.” This assumption is that agency is an internally undifferentiated good whose quantity necessarily increases when new options appear. When I speak of agency as internally undifferentiated, I do not mean that an agent’s set cannot comprise distinct options. The cumulative assumption, to the extent that it offers a theory of how new options affect existent ones, treats options as distinct from one another. Instead, the idea that agency is internally undifferentiated has two important implications. First, new options do not affect agency primarily by changing its quality; they change only the amount of agency present. Second, increases in agency should, in principle, impel a process that leads to improvements across an agent’s life. An inability to agitate for welfare in some domain of life reflects insufficient agency rather than a lack of something distinct. Theorists who expect self‐worth to eventually permeate all domains of a woman’s life seem beholden to the idea that more agency will become feminist agency. Logically, the problem with the cumulative assumption is that it sees 6 options as self‐sufficient and thus lacks a concept of opportunity cost. It denies that new options and beliefs about the self may eviscerate, or decrease the appeal of, others. As I have mentioned, patriarchy rewards women for complying with it. Part of what this means is that new options for welfare can remove incentives to resist patriarchy. I will make this point more fully in the section on microcredit, but here is one example. In a context that valorizes female seclusion, giving women opportunities to access income without leaving the home decreases the appeal of challenging norms that demand women’s seclusion. The operative notion of agency in development discourse, in addition to suggesting that agency is internally undifferentiated, suggests that it requires specific motivational content. Increases in agency move a person toward the view that she has equal value. I call this assumption “the substantive assumption,” because it sees agency as culminating in a person’s adoption of certain substantive moral beliefs. The idea seems to be that, the more effective a person’s welfare agency becomes, the more she will adjust her self‐concept toward equality. One way this might work is by expanding women’s senses of what they are capable of. When women discover they can engage in activities that patriarchy tells them they are unfit for, they begin to question patriarchy. Consider Cheston and Kuhn’s (2002, 29‐30) claim that women who are effective at meeting their welfare needs will develop increased self‐esteem.This, in turn, is supposed to cause dissatisfaction with sexist limitations on their lives. Consider also Kabeer’s argument that women who bring income into households start believing they have a right to ask for what they need within them (2001, 71). The problem with the substantive assumption, when applied broadly, is that it denies that women can gain self‐esteem and self‐efficacy by believing in the appropriateness of their subordination and fulfilling subordinate roles well. For instance, a woman who is provides superior food to her male relatives without complaint is likely to be regarded a good wife with a particular talent for self‐discipline. Thinking I am getting better at feminine self‐denial is unlikely to lead me to believe that women and men are equal. 2. The Self‐Subordination Social Recognition Paradox The understanding of agency described above excludes the possibility that women’s incentives to accept patriarchy can increase with their welfare agency. Correlations between increased welfare agency and increased collusion with patriarchy are likely to be common in patriarchal societies for reasons I will describe in this section. The subordination‐social recognition paradox occurs because of a structural feature of patriarchal societies. The SSRP is as follows: people’s access to social status, their ideas about what they should become, and their access to material benefits often depend on how well they meet social expectations. In patriarchal societies, women often need to meet sexist social expectations to increase their welfare. For instance, women in the contemporary United States are often penalized professionally for failing to conform to patriarchal beauty standards. This means that they can increase their access to income and social recognition (both objective goods) by wearing makeup and heels, being thin, and so forth. In the development context, the SSRP means interventions may simultaneously increase women’s access to welfare and their incentives to accept patriarchal norms. 7 Interventions may do so by increasing the material or social rewards of patriarchally prescribed behavior. Interventions may also strengthen the links between psychological goods, such as self‐coherence and self‐respect, and acceptance of patriarchal beliefs. Since the attainment of these psychological goods often enhances both objective welfare and subjective happiness, this means that antipoverty interventions may go farther than incentivizing compliance with sexist norms. They may incentivize their internalization. Both complying with and internalizing oppressive norms can increase women’s welfare agency. Women’s incentives to internalize patriarchal norms are strongest when the social and material rewards for acting on the basis of them align. Social benefits include things like status, affiliation, approval, and love. In patriarchal societies, sexist norms dictate the behavior required of “good women.” Women who comply with such norms, especially those who do so particularly well, receive status and praise (or at least protection from certain forms of criticism). The rewards of being a “good woman” often extend beyond the glow of others’ approbation. How a person is perceived by others often affects her ability to meet her material needs. As we have discussed, in poor societies with high levels of patriarchal risk, women’s ability to access food, shelter, and safety depends heavily on their marriageability and ability to keep male family members happy (Agarwal 1997). This means that social benefits are often gateways to material ones. A woman who is seen as a “good woman” will be well‐positioned to meet her material needs. But it is one thing to claim that women can benefit from behaving as “good women” and another to claim that they can benefit from endorsing the ideology surrounding that behavior. Women stand to benefit from internalizing oppressive norms under conditions where they benefit both socially and materially from complying with them. Under such conditions, internalizing oppressive norms facilitates development of a coherent self‐concept. It is theoretically possible for a person to comply with sexist norms for purely instrumental reasons. So, for instance, a woman may give better food to her husband only to keep him happy while simultaneously rejecting the view that this is what good woman should do. But a lifetime of “going through the motions” could also be a lifetime of suffering, compromised self‐efficacy, and difficulty identifying one’s true commitments. Some of the reasons that we can expect difficulty from a life of doing one thing and believing another stem from general psychological mechanisms over which we exert little control. We persuade ourselves that we are personally responsible for positive outcomes of our behavior more than negative ones, we attempt to protect reasons for optimism about our futures, we attempt to avoid cognitive dissonance, and so forth (Cummins and Nistico 2002). Further, and more important to my overall argument, psychological coherence enhances a person’s welfare agentic capacities. A woman who can support her self‐efficacy by internalizing sexist views about herself has self‐interested reasons to do so. As Rawls observes in his explanation of the value of the primary good of (the social basis of) self‐respect, it is difficult to pursue any plan of life without believing oneself or one’s projects are of value. It is also difficult to pursue any plan of life one thinks it is impossible to effectively pursue (1978, 441).ix For Rawls, a person’s sense 8 that her self and projects are of value supplies motivation for pursuing those projects. Rawls also claims that self‐respect has a social basis. He thereby emphasizes how difficult it is for us to remain motivated to pursue our projects absent social affirmation of their (and our) value. An agent who repeatedly hears that her projects are worthless may begin to believe it. In patriarchal societies, women learn that projects that would promote their equality lack value. The woman who refuses to provide superior nutrition to her husband, for example, is likely to be criticized, gossiped about, and worse. She may grow to believe that she is defective (because she fails to meet standards of good womanhood)‐‐ or that her project of criticizing sexist norms is morally suspect (because the people around her treat her moral perspective as alien). Under such circumstances, a woman can preserve self‐worth and motivation by internalizing sexist norms and making her projects consistent with them. In addition to providing motivation to pursue her projects, endorsing oppressive norms can be an agent’s easiest path to maintaining a coherent self‐concept. As Diana Meyers argues, “a reflective commitment to a set of values” can serve to “protect us from others’ scorn” (1989, 212). If we believe our own evaluations to carry special weight or our projects to be objectively valuable, we are not required to refashion our self‐concepts every time others judge us. Though Meyers’ point applies most readily to cases where an agent’s values clash with the dominant ones, it also points to the way scorn from others may threaten an agent’s sense of identity—the sense that she is a stable entity who cares about certain things. Acting against her deeply held beliefs to avoid penalty, as does the agent who complies with oppressive norms without internalizing them, may cause her to question whether she is really committed to what she says she is committed to. Such uncertainty may cause her to act in inconsistent ways that undermine her ability to achieve ends—welfare‐promoting or otherwise.x Third, an agent’s sense of self‐efficacy depends partly on her ability to act successfully in accordance with her principles. Compliance with oppressive norms often demands behavior with painful effects. The life of the woman who has to conform to oppressive norms often involves repeatedly subjecting oneself to suffering. The woman who believes the suffering is unjustified but inevitable is also reminded daily that the world is unfair and unchangeable. Adjusting one’s beliefs about what is justified is a way of protecting oneself against the pain and self‐efficacy losses likely to come with repeated failure.xi My goal in the last few paragraphs has been to reveal an implication of the SSRP that may not be initially apparent‐‐ that, under patriarchal conditions, a woman can increase her welfare agency by internalizing oppressive norms. Part of my argument that microcredit is less empowering than it initially seems will be that it furnishes new reasons to believe in the moral acceptability of patriarchal norms. My point has emphatically not been to show that internalizing oppressive norms is praiseworthy—or even that it will usually yield welfare increases. The number of women who will benefit from internalizing oppressive norms is limited. One reason is simple interpersonal variation. It is clear that a woman who would not feel herself without a feminist identity stands to gain less by internalizing sexist norms than one who has mixed feelings about gender oppression. A second reason is that the force of the incentives to internalize 9 sexist norms varies. This type of variation is a function of differences in social structures rather than individuals. The incentives to internalize patriarchal norms are likely to be particularly weak when the rewards for complying are intermittent, difficult to access, or not uniform. For instance, a woman may improve her shot at marrying into a higher social class by being unusually attractive and dutiful. However, if this outcome is unlikely for a given woman—and she knows this on some level—she has reason to protect her self‐esteem by continuing to believe that beauty and duty are not true measures of a person’s worth. Similarly, and more important for my argument about antipoverty interventions, incentives to internalize patriarchal norms are weaker when long and short‐term rewards, or social and material rewards for complying with oppressive norms diverge. Internalizing sexist norms is a particularly promising path to agentic coherence when an agent is rewarded consistently and uniformly for conforming. Internalization will then produce consistency between the agent’s beliefs about what is acceptable and the external reinforcement she receives. External reinforcement is most uniform when social and material rewards accrue to the same patriarchally prescribed behaviors. Consider the case of a woman who gives superior food to her husband and who is not terribly poor (and thus does not fall below some nutritional threshold). She is held up as an example for women in her community and succeeds in attaining economic security by capturing her husband’s favor. Suppose also there is no other way for her to gain an equal level of income, security, or social recognition. This woman, if she believes sexist food norms are unjustified, will be subject to cognitive dissonance, a lack of self‐efficacy, and a sense of moral failure. She can do away with these threats to her agency and sources of suffering by internalizing sexist norms. (To say this is not to deny that there is an inherent harm in the sense of oneself as unequal; it is only to say that the harm is outweighed by the other benefits.) However, the rewards of internalization would differ if the material and social rewards of complying with sexist food distribution norms diverged. Suppose urbanization and globalization have changed men’s earning power such that women can no longer count on men to provide for them effectively (as Mead Cain, Syeda Rokeya Khanam, and ShamsunNahar (1978) argue they actually have in much of South Asia). Suppose also that it is still widely believed that women should prioritize their husbands’ nutrition. It is far less clear that a woman living under these conditions stands to benefit from internalizing patriarchal norms. Part of the reason is that this may lead her to engage in welfare‐undermining behavior—to depend on her husband when there are better ways of securing food and shelter (goods it might be worth risking social ostracism to attain).xii Another reason internalizing patriarchal norms promises to be less beneficial is that it requires suppressing impulses toward basic survival. It may seem that, by treating women’s beliefs about patriarchy as partly responsive to incentives, I assume that our normative attitudes are constantly shifting or within our cognitive control. To respond to the concern about cognitive control, we can note that changes to our beliefs about ourselves may, but need not, occur consciously. The more troubling concern is about the relative stability most people’s self‐concepts have achieved by adulthood. It may seem that people form self‐concepts 10 early in life, and they are not so mercurial as to change with social tides. I agree that people rarely cast off their self‐concepts in adulthood. However, an agent can alter her self‐concept in response to social conditions without completely rejecting her previous views. Though adults typically have more robust and coherent self‐concepts than children, the struggle for self‐integration is lifelong—and, for most of us, never fully achieved. Adult women may become more or less attached to certain norms without radically altering their self‐concepts. Moreover, if a woman’s self‐concept has always been internally ambivalent despite containing some allegiance to patriarchal values, identifying with previously unendorsed patriarchal norms can increase her self‐ coherence. 3. Does Microcredit Empower Women? I have elaborated the SSRP as follows: women can gain welfare agency from complying with and internalizing sexist norms, and the incentives to internalize sexist norms are highest when women are both socially and materially rewarded for complying with them. I now turn to discussing reasons for worry that microcredit interventions align the social and material rewards of accepting sexist norms. Microcredit is supposed to increase the perception of women’s value (by themselves and others), but , as the SSRP reminds, women can gain increased value by getting better at meeting , and believing in, patriarchal social expectations. In the most comprehensive look at the data on women and microcredit in South Asia to date, NailaKabeer attempts to reconcile contradictory conclusions about its empowerment potential (1998, 2001). According to her “right direction” explanation, the data that is usually adduced to show that microcredit fails to empower women either relies on the wrong indicators or can be reinterpreted as evidence of women’s empowerment. I believe that Kabeer’s inadvertent commitment to the cumulative and substantive assumptions about agency causes her to misassess the cases in her dataset that involve women’s increasing self‐subordination. The mechanisms through which Kabeer sees women’s empowerment as occuring always involve enhanced ability to meet oppressive social expectations. Kabeer underestimates the implications of this for future gender relations. First, however, we should get clear about what Kabeer’s findings are and which feminist critiques of microcredit she challenges. The popular media have, nearly univocally , celebrated microcredit (Poster and Salime 2002, Narayan 2005,Batliwala and Dhanraj 2007). However, a swirling controversy has developed within feminist development studies. One set of feminist criticisms is about the role microcredit plays in perpetuating global inequalities. For instance, Winifred Poster and Zakia Salime (2002) argue that microcredit creates the impression that it is a lack of credit—rather than an unjust global economic order—that causes poverty in the global South. Uma Narayan (2005) argues that microcredit programs serve an ideological function by masking the devastating effects of the International Financial Institutions on the poor. I am sympathetic to these critiques of the structural discursive effects of microcredit, but they are out of the scope of Kabeer’s analysis—and mine. Kabeer is interested in feminist critiques that assert that microcredit fails to empower women. The studies that purport to show this use diverse indicators of 11 empowerment—all of which, however, are associated with the normative goal I called “feminist agency” in the first section. Some studies suggest a positive correlation between a woman’s loan recipient status and the level of domestic violence to which she is exposed (Shuler, Hashemi, and Badal 1998; Aktaruzzaman and Guha‐Khasnobis 2012). Anne‐Marie Goetz and Rina Sen Gupta (1996) find that a large percentage of recipients in Bangladesh transferred loans to their husbands and many had no idea what had happened to their loans. Kabeer (1998) herself finds that some women loan recipients are not involved in economically productive activity at all. Nathalie Holvoet (2005, 86) sand Aminur Rahman (2001) observe that women are frequently “forced” by their husbands to take out loans. Montgomery, Bhattacharya, and Hulme (1986) note that, in most cases, microcredit does not change the gender division of labor within households. Kabeer argues that the data above are consistent with the view that microcredit empowers women. Her argument has two parts. First, she argues that these data demonstrate only that microcredit has failed to completely transform the patriarchal conditions of women’s lives. Her reasoning is that the critical studies rely on end‐state, rather than processual, indicators (Kabeer 2001, 66). Empowerment is a spectrum, and determining whether women are becoming empowered is a matter of asking whether they are moving in the right direction on the spectrum—not asking whether they have arrived. Indicators like domestic violence and women’s control over loans could be evidence that women are more empowered than they were before—even if they have not “arrived.” (Domestic violence could be a move in the right direction because it is signals threats to men’s power (Kabeer 2001, 65‐66)). Second, Kabeer argues that women loanees see themselves as empowered, and that this should urge us to look for indicators of empowerment that can do justice to their experiences (1998,18). The women Kabeer interviewed expressed a particular set of reasons for valuing microcredit. Other community and family members perceived women as having greater worth, and women valued themselves more. Kabeer argues that this is expressed in women’s increased bargaining power within households. Using household bargaining ability as an empowerment indicator allows Kabeer to directly challenge the indicators used by the microcredit critics. Where the critics focus on things like female “control,” of loans, accounting processes, etc., Kabeer argues that focusing on control requires us to ignore incremental growth in women’s involvement in decisionmaking. Kabeer’s idea that microcredit increases women’s sense of self‐worth and the value others attach to them is striking. It suggests an unusual mechanism of empowerment—and one that triggers worries related to the SSRP. This is why: a person’s sense of her own value and her value in the eyes of others are likely a function of her ability to fulfill her prescribed role. If the person’s assigned role is subordinate, increases in her value will probably track increases in her success at complying with and internalizing oppressive norms. To put it starkly, saying that an oppressed person has become empowered through gaining social value could just be another way of saying that a person is (or is perceived as) getting better at accepting her subordination. Of course, this is true only if prescribed roles do not become less hierarchcal as a result of antipoverty interventions. It seems true that a large number of the women in Kabeer’s 12 sample have experienced changed gender expectations with regard to household bargaining power and enterprise management. These are largely women from the economically better off social strata. For some women in Kabeer’s study, however, there is no reason to suppose that microcredit has weakened patriarchal social expectations. There is reason to suspect that the higher value placed on women comes from their increased ability either to meet longstanding sexist expectations or to meet new versions of such expectations that seem to have been generated by microcredit. I begin with examples of the second type of case—women becoming successful at meeting new oppressive expectations. An astonishing fact about many women in Kabeer’s sample (2001, 73) and those in some other studies (Holvoet 2001), is that their value in the eyes of others seems to have increased without their participation in any element of loan use. In other words, some women turn loans over to male family members and do not participate at all in deciding what is done with them. Yet these same women say that their husbands value them more, that they are more socially accepted, and so forth. Though it is a minority of women who report total male control of loans, it is also unclear why many of the women who do participate in loan use are valued more. Specifically, it is not clear that the increased valuation has to do with more egalitarian conceptions of what women can and should do. A number of women appear to experience increased value for reasons that have nothing to do with, or are tenuously related to, being perceived as more competent entrepreneurs, managers, or decisionmakers. What explains the new value placed on women, if it is not a newfound respect for what women’s capabilities? Two alternative explanations suggest themselves—one more pernicious than the other, but neither terribly sound evidence for concluding that microcredit is empowering. The less pernicious mechanism seems to be the reduction of scarcity within the household. According to the women in Kabeer’s study, men’s inabilities to meet their households’ basic needs produces interfamilial tension that often culminates in violence (2001, 70; 1998,39, 49‐51). The potentially more pernicious mechanism involves the valuing of women because they are means of access to loans. One interviewee offers the following analysis of the effect of loans in her community, “Now husbands think, if we beat up our wives, they won’t give us loans, we won’t survive” (Kabeer1998, 51). Though this woman also asserts that her own husband places a higher value on her because she works now, Kabeer’s own analysis of this woman’s testimony includes the idea that men value women because of the fear that loans will dry up. There is evidence that has emerged since Kabeer’s study that men view women as something like collateral since the advent of microcredit (Hoffman and Gnanou 2007, 9). In the researchers’ words, “in places where loans are perceived as being more easily obtained by women, men use women to gain access to loans.” Fauzia Erfhan Ahmed, who conducted a recent study of men in Grameen bank loanee households, found that many husbands whose wives take out loans continue to view the loans as their own property (2008a, 554). Kabeer seems convinced that these new attributions of value to women are steps in the right direction. Part of the reason seems to be an implicit commitment to the substantive assumption about agency. That Kabeer thinks that increased value by 13 others is empowering, and not simply welfare‐enhancing, suggests that she believes increasing women’s perceptions of their own value means challenging women’s acceptance of patriarchy. But there are important reasons to doubt that this is true about the cases at hand. Neither the case of the loan reducing scarcity nor the case of women as collateral suggests an increase in women’s value because of any new beliefs about what women are capable of. Assuming that gender inequality is supported largely by views about differences in men’s and women’s capacities, there is little logical reason the changed perceptions of women should militate against it. Indeed, conditions of reduced scarcity may remove reasons to question women’s exclusion from formal sector work (and I will return to this point in a moment). There is also reason to worry that the perception of women as collateral arises from new incentives to invest in conservative gender ideology. The idea of women as sources of external capital is nothing new in South Asian societies with histories of dowry. Hoffman and Nanou make this connection to dowry explicitly. “Loans may represent a form of dowry. If their wives do not manage to obtain such a loan, this may constitute a source of tension and violence within the household. This is equally the case where the husband who has appropriated the loan for himself no longer has the means to recover the capital” (200 , 9). In other words, microcredit interventions may reward men for manipulating women and reward women for learning to negotiate within this manipulation. Even if microcredit produces short‐term welfare gains for women, it simultaneously gives men reasons to accept an existing sexist view—namely of women as points of access to property. Women also have new reasons to accept this view of themselves; the best way to keep their husbands happy is to help them secure loans. They gain agency to access income and perhaps freedom from violence, but since this agency comes from obsequious behavior, it is potentially welfare agency gained through self‐subordination. This worry that microcredit can strengthen patriarchal views of women as property is exacerbated by certain perceptions of microcredit on the part of rural South Asian populations. Research in Bangladesh describes some men as feeling that microcredit is externally imposed by NGOs and the government whose interest is in corrupting women (Ahmed 2008a, 554). One man in Ahmed’s study, which documents a variety of different male responses to microcredit, says, “We know that the Grameen bank is against Islam. Women go there to show their legs. But we are forced to allow them to stay on as loanees because we are poor. As soon as we are financially stable I will ask my wife to quit the bank” (Ahmed 2008a, 552). Such men’s goal under these circumstances is to use the NGOs for to access income while minimizing women’s corruption. This subsection of men experiences financial dependency on their wives as humiliating—a point which offers further reason to worry about resurgences of gender conservatism. Ahmed even relates the story of a woman who tearfully withdraws from the Grameen bank after ten years because her husband has declared that they have finally earned enough money to stop compromising her honor in this way (2008b). If there is something new that women are capable of after microcredit, it is bringing in money, but this capacity is seen as contingently linked to the goals of the government and NGOs‐‐ rather than revelatory of previously ignored facts about women’s nature. 14 Of course, none of these points conclusively demonstrates that Kabeer is wrong about microcredit moving these women toward empowerment. But it is surprising that she sees men’s continued power as evidence of partially changed patriarchy rather than resurgent patriarchy. Her failure to raise these questions seems symptomatic of a more general difficulty. Kabeer, at least in her interpretations of data, struggles to account for the possibility that interventions might increase women’s incentives to invest in patriarchy. In theoretical discussions of empowerment, Kabeer frequently notes that patriarchal structures create incentives for women and that we cannot determine in advance whether agency increases will produce the desired empowerment effects (Kabeer 1999, 442). But, when interpreting her data, Kabeer seems not to consider the possibility that new incentives to accept patriarchy may arise. We might explain this oversight‐‐and her certainty that the moves in question are moves toward empowerment—with reference to something like the cumulative assumption. Increases in women’s welfare, rather than sometimes having gender‐related opportunity costs, always add to women’s overall agency. Kabeer’s optimism seems to stem from inadvertent commitment to the cumulative assumption about agency—understanding new options as agency‐enhancing, rather than potentially agency‐decreasing in some areas, or as changing the background conditions under which welfare agency can be exercised. We see Kabeer’s implicit commitment to the cumulative assumption more clearly when we consider the theoretical contortions required to make sense of another set of cases. These are cases where women’s increased value comes from increased ability to meet longstanding patriarchal expectations. Some of the poorest women in Kabeer’s sample decided to abandon jobs as fieldworkers or domestic servants. These women valued loans because they allowed them to work from their homes. To understand why they prefer home‐based labor, it is important to remember that their society ties women’s honor to seclusion. One woman says, “Isn’t it better to work in your own house than to work in someone else’s to fill your stomach? You stay at home, you raise some ducks and hens for yourself and you make some profit. Isn’t it bad when people say ‘She goes to work in some people’s houses?” (Kabeer 1998, 66). In their previous employment, hese women faced social penalties for violating purdah constraints. They felt their choice was between working in occupations that would cause them to be seen by other community members and thus losing honor and status‐‐or preserving their honor by starving at home. There were non honor‐related reasons for women to prefer seclusion, most notably the physical difficulty of work in the fields and the desire not to be someone else’s servant. However/ the women themselves consistently cite the desire to gain honor as a major motivation for their workforce withdrawal. Women with the lowest economic status, Kabeer notes, chose home‐based labor after microcredit largely to avoid shame. Drawing on my earlier ideas about the subordination‐social‐recognition paradox, we can understand microcredit as having affected the poorest women’s incentives in two ways. First, it has increased the rewards for engaging in seclusion. Second, by realigning the social and material rewards of complying with patriarchal norms, microcredit has incentivized internalizing the belief that purdah is morally acceptable. 15 Before microcredit, the only way to avoid or mitigate severe material deprivation was to violate purdah norms. This offered self‐interested reasons to question the validity of those norms. That there is greater incentive to violate purdah through one’s actions when one is extremely poor is borne out by research demonstrating that well‐to‐do women tend to adhere to purdah most strictly (Holvoet 2005, 85‐86). One of Kabeer’s informants says that, when she used to engage in agricultural labor with her husband, “if people try to make me feel ashamed of my work, I do not feel ashamed. . . I will work myself, I will feed myself, I will carry whatever load I have to” (Kabeer 1998,67). Describing her life after microcredit, she states that there is no dishonor in work. But she also justifies her exit from the fields, thus, “my value has gone up from before, I feel ashamed, and people say, ‘she has improved so much, how can she still go do this work’” (Kabeer 2001,70)? This particular woman’s narrative strongly suggests that it is not only her conformity with purdah that is changing; her beliefs about its appropriateness are as well. Kabeer seems to want to describe this fact only in terms of a failure to alter existing gender relations. She writes of how “purdah continues to constrain” and is “resilient” (Kabeer 2001, 69). But she seems reluctant to take seriously the possibility that microcredit could have increased the power of purdah or women’s acceptance of it. Kabeer recognizes that these women who withdraw from work outside the home pose the most serious challenge to her positive view about microceredit. Her attempt to deal with the challenge is, I believe, unsatisfactory because of the theoretical contortions it requires. She argues that increased purdah is a step towards women’s empowerment because the women in question value the changes microcredit has allowed (Kabeer 2001,70). The women viewed their previous state as humiliating; they are now less humiliated. Without denying that these women’s lives are better without having to experience degradation and ostracism, we can note that Kabeer’s claim that the purdah is empowering because the women want it is question‐begging. It is question‐begging in two ways. First, as I have argued in my work on adaptive preferences, it is unclear why the judgments of a person who does not see oppressive norms as harmful should be treated as definitive about how to evaluate the retrenchment of those norms (see Khader 2011, 2012, 2013). If empowerment requires increases in feminist agency, those who lack feminist consciousness may not be the most reliable indicators of its presence of absence. Second, if I am correct that feminist agency and welfare agency are separate— and that welfare agency can come at the cost of feminist agency—all these women are saying when they say that microcredit has improved their lives is that they have more welfare agency. Increased welfare agency is not the same thing as increased feminist agency, and saying otherwise denies the SSRP. Kabeer demonstrates awareness that women have to adhere to restrictive gender norms to increase welfare. She writes that the poorest women opt for purdah to increase their social standing (Kabeer 2001, 71). She calls this paradoxical, but to simply call it paradoxical is to undertheorize the implications of women’s increased participation in purdah. Rather than simply saying that women value increased participation in female seclusion, we need to acknowledge that microcredit gives them new welfare‐based reasons to engage in and accept it—and 16 we need to ask serious questions about what the long‐term gender effects of this will be. It may be argued that I am holding Kabeer to a standard of empowerment that she herself does not hold.xiiiKabeer explicitly defines empowerment as an increase in the conscious ability to make strategic life choices (1999; 2001,81). On Kabeer’s stated view, the women’s choice to enter purdah is empowering because the women gain a choice they previously lacked. Where they previously felt compelled to violate purdah because of economic necessity, they now get to decide whether to accept it (2001,81). However, this comment is at odds with her other repeated assertions about empowerment— especially her repeated assumption that having real choices means learning to challenge doxa (1990, 441). Doxa are views that reinforce social inequality by making inequality appear to be the natural order of things. It is not clear how empowerment can simultaneously require challenging doxa and be achieved when a woman endorses patriarchal norms more thoroughly than she used to.xiv More importantly, however, our question is whether microcredit is empowering—not whether it is empowering on Kabeer’s definition. If women’s empowerment were reducible to the expansion of welfare agency or the giving women whatever they happen to value, Kabeer’s claim that increased purdah is empowering could be logically valid. But it would be based on a notion of women’s empowerment that is highly counterintuitive and that does not seem to be motivating the bulk of feminist literature on the topic. As I argued in the first section, most writing on women’s empowerment suggests that it requires a move away from oppressive gender relations. 4. Conclusion We need a theoretical understanding of women’s empowerment that that acknowledges that women’s welfare can increase without resulting in empowerment. This is because of the conditions described in the self‐subordination social recognition paradox. Patriarchal societies structure women’s options so that they can benefit from complying with and internalizing patriarchal norms. One important upshot of this for development ethics is that enhancing women’s ability to achieve welfare can strengthen patriarchal structures. Giving women opportunities to enhance their welfare by getting better at fulfilling and accepting subordinate roles may consolidate patriarchy and women’s relationship to it. There are practical reasons we should think seriously about the long‐term gender‐related opportunity costs that may ensue from just making individual women’s lives better—as is shown by the examples of microcredit producing the view of women as collateral and causing increased acceptance of purdah. What does the fact that women can gain welfare agency by compromising their feminist agency imply about evaluating development interventions? Let us begin with what it does not imply. It certainly does not imply that inspiring feminist consciousness in women is always more important than increasing their welfare. We should be wary of sacrificing women’s basic welfare at the altar of equality—of saying that the lives of the poor women who preferred purdah to social ostracism and backbreaking agricultural labor were better before. Nor does it imply that antipoverty interventions that do not decrease sexism are failures. Development practice involves complicated on‐the‐ground 17 judgments, and incremental changes, and there may be no imaginable intervention in a given case that would reliably decrease incentives to support patriarchy. Though this is not the place to offer a theory of how to measure empowerment, my analysis here asks us to be attentive to two important possibilities. First, as a variety of feminist empowerment theorists have argued, increases in women’s welfare, self‐ esteem, and so forth do not imply a decrease in the power of patriarchal forces. Second, interventions that do not focus on changing gender roles may do more than simply leave gendered power structures intact. There are opportunity costs to interventions that may give women new reasons to accept patriarchy. Rather than assuming that women’s empowerment always increases when they have more opportunities, we need to see that interventions restructure opportunities and incentives. They change the long‐term landscape within which women seek empowerment—and this means we cannot be sanguine about which direction women’s empowerment is moving just because women are better off. The woman who is forced by her husband to withdraw from the Grameen Bank after ten years is arguably better off economically‐‐ but returns to a gender landscape in which many men in her community feel they need to defend their masculinity in the face of humiliation. The woman who enters purdah now because she can participates in creating a gender landscape where poor women are less critical of seclusion. Rather than assuming women who can do more‐‐ or who have greater self‐ esteem‐‐ will change gender relations, we need to ask about gender‐related opportunity costs and imagine strategies for mitigating them. For instance, Ahmed recommends that the Grameen bank should enlist particularly progressive husbands to train others who fear threats to their wives’ honor or their own power (Ahmed 2008b). Holvoet recommends that adding groups that are long‐term and focused on consciousness‐ raising to microcredit packages. She postulates that it will increase the likelihood that women will translate their new household bargaining power into a critique of patriarchy (2005, 97). The SSRP suggests that antipoverty interventions focused on women, because they enhance women’s welfare under oppressive conditions, risk consolidating women’s relationship to patriarchy. To evaluate whether antipoverty interventions are steps in the right direction, we need long‐term vigilance—not just about whether women’s lives are improving, but also about how the gender landscape in which they exert agency is changing. i The female genital cutting case may not, strictly speaking, involve an antipoverty intervention failing to empower women. According to Agot, girls and families opted for cutting because of a lack of economic opportunities. Interventions gave them education but not employment, so marriageability became the best means to income. ii In her influential definition of empowerment, Kabeer (1999; 2001) distinguishes choices from strategic life choices. I believe Kabeer sees strategic life choices as those that would significantly enhance a person’s ability to access welfare. iii A prominent example is Kabeer’s (199) assertion that women’s apparent contentment with injustice is likely caused by an inability to call patriarchal norms into question. 18 iv In an influential study, Anju Malhotra, Sidney Shiller, and Carol Boender argue that diverse theorists of women’s empowerment share a conception of empowerment as the expansion of the ability to make choices important to achieving welfare (2002, 5‐6). It may therefore seem that the operative conception of women’s empowerment focuses on welfare agency rather than feminist agency. However, virtually all of the theorists cited by Malhotra, Shiller, and Boender define empowerment as involving challenges to existing power structures—and they often explicitly mention challenging gendered ones. Perhaps Malhotra, Shiller, and Boender omit the focus on gender because they believe choices about gender are one subset of strategic life choices. I explain why that view is untenable in the next paragraph of the paper’s text. Further, the feminist development theorists who are puzzled by the mixed results of antipoverty interventions seem committed to the idea that women’ s empowerment requires the enhancement of feminist agency. Their evidence that the results of antipoverty interventions are problematic is usually the existence of unchanged or worsened gender relations. They fault antipoverty interventions for failing to change women’s belief that they deserve a lesser share of household goods, for failing to change the gender division of labor, for increasing men’s violence against women, etc. v Bina Agarwal (1997) argues that women’s compliance with sexist food distribution norms under such circumstances is self‐interested. vi Cain et. al (1979) describe societies as having differing levels of “patriarchal risk” depending on the welfare and status losses women stand to incur if they are not attached to male affines. vii I have made this argument more fully elsewhere (see Khader 2013), but collective action problems make the range of cases where women can reasonably believe that changing their behavior is not going to change sexist norms quite large. viii Development discourses frequently characterize the pre‐empowered person as uncritically accepting of the power status quo. Kabeer argues that those who are not yet empowered treat widely held beliefs as though they are beyond question (1999, 440‐ 441). DFID describes the pre‐empowered as seeing themselves as passive objects of the choices of others (Appleyard 2002, 13). Rowlands argues that women’s empowerment involves increases in the perception of the self as “able and entitled to make decisions” (1997, 14). Oxaal and Baaden argue that empowerment requires developing the ability to critically assess one’s own situation in order to transform society (1997, 6). ix Rawls specifically discusses self‐respect rather than self‐esteem. However, those who argue that the two concepts are distinct tend to argue that, since Rawls focuses on the view that one’s self and one’s projects are worthy—rather than whether one has successfully lived up to one’s principles‐‐ he means something closer to self‐esteem. x Boxill (1976) argues that, under oppressive conditions, it can be difficult for people to maintain clarity about their own motives over time. Benson (1999) argues that oppressive socialization can cause women to experience some of their desires as alien and thus experience difficulty knowing which motives they identify with. xi Amartya Sen argues that learning to ignore the injustice of the world can be a way of preserving subjective wellbeing (2002, 634). 19 xii Studies show that the poor themselves tend to rank social ostracism as one of the worst parts of their plight (Narayan 2000; Kabeer 2001). This fact should give us (theorists) pause before assuming that the risks associated with social ostracism are minor or automatically less morally urgent than losses of health, shelter, and so forth. One reason social inclusion may be of high value is that affiliation and social recognition ensure stable access to other goods. This view of the value of affiliation and social recognition is consistent with the sociological theory my subordination social recognition paradox; according to it, relations with others are the gateways to a variety of goods constitutive of welfare. xiii On at least one occasion Kabeer seems to doubt whether empowerment requires increases in feminist agency. See 1998 (12‐13). xiv Kabeer may be able to consistently assert that empowerment is defined as the mere having of choices and that it requires the challenging of doxa. Doing so would require asserting that a person who scrutinizes patriarchy and believes in patriarchy more intensely after the scrutiny has experienced an increase in empowerment. But there are a number of occasions where Kabeer claims that a choice is somehow not real if it reinforces women’s subordination (see 1999, 441 for instance). Further, Kabeer argues that choices whereby women reinforce their subordination are more likely to reveal welfare tradeoffs restrictive societies impose on them than women’s true beliefs (1999; 1998, 28). 20 References Agarawal, Bina. 1997. Bargaining and gender relations: within the household and beyond.” Feminist Economics 3 (1): 1‐51 Agot, Kawango. 2007. Women, culture and HIV/AIDS in Sub‐Saharan Africa: What does the empowerment discourse leave out? In Global empowerment of women: Responses toglobalization and politicized religions. ed. Elliott. New York: Routledge. FauziaErfhan. 2008. Hidden opportunities: Islam, masculinity, and poverty alleviation. International Journal of Feminist Politics 10(4): 542‐562 ‐‐‐‐‐‐‐‐‐.2008. Microcredit, men, and masculinity.NWSA Journal 20 (2): 122‐155. Alkire, Sabina.2007.Concepts and measures of agency. Oxford Poverty and Human Development Initiative Working Paper. Aktaruzzaman,Khondker and BasudebGuha‐Khasnobis. 2012. Microcredit and domestic violence: A regression discontinuity approach. Available at SSRN: http:,,ssrn.com,abstract=2156086 or http:,,dx.doi.org,10.2139,ssrn.2156086. Appleyard, Susan. 2002. A rights‐based approach to development: What the policy documents of the UN, development cooperation agencies, and NGOs say.OHCHR Human Rights Policy Roundtable Background Paper. Batliwala,Srilantha and DeepaDhanraj. 2007. Gender myths that instrumentalize women: A view from the Indian front line. InFeminisms in development: Contradictions, contestations, and challenges. London: Zed. Benson, Paul. 1999. Feeling crazy: Self‐worth and the social character of responsibility. In Relational autonomy. New York: Oxford. Boxill, Bernard. 1976. Self‐Respect and protest. Philosophy and Public Affairs 6 (1): 58‐ 69. Cheston,Susy and Lisa Kuhn. 2002. Empowering women through microfinance. UNIFEM. Cudd Ann. 2006. Analyzing Oppression. New York: Oxford. Cummings , Robert A and Helen Nistico. 2002. Maintaining life‐satisfaction: The role of positive cognitive bias. Journal of Happiness Studies 3: 37‐69. Frye , Marilyn. 1983. The Politics of Reality. Freedom CA: The Crossing Press. Goetz, Anne‐Marie and Rina Sen Gupta. 1996. Who takes the credit? Gender, power, and control over loan use in rural credit programs in Bangladesh. World Development 24(1): 45‐63. Hoffman Elisabeth and Kamala Marius‐Gnanou. 2007. Credit for women: A future for men? ADA Dialogue37: 7‐11 Holvoet, Nathalie. 2005. The impact of microfinance on decision‐making agency: Evidence from South India. Development and Change 36(1): 75–102 Kabeer, Naila. 2001. Conflicts over credit: re‐evaluating the empowerment potential of loans to women in rural Bangladesh. World Development 29(1): 63‐84. —.1998. ‘Money can’t buy me love’? Re‐evaluating gender, credit, and empowerment in rural Bangladesh.IDS Discussion Paper 363. Sussex: Institute for Development Studies. 21 ‐‐‐‐‐‐‐‐.1999. Resources, agency, achievements: Reflections on the measurement of women’s empowerment. Development and Change 30(3): 435‐464. Khader, Serene J 2013 “Can Women’s Compliance With Sexist Norms Be Self‐ Interested?” Unpublished Manuscript on File with Author ‐‐‐‐‐‐‐‐. 2013 Identifying adaptive preferences in Practice: Lessons from Postcolonial Feminisms. Journal of Global Ethics. ‐‐‐‐‐‐‐‐. 2012. Must theorizing about adaptive preferences deny women’s agency? Journal of Applied Philosophy 29 (4): 302‐317. ‐‐‐‐‐‐‐‐. 2011. Adaptive Preferences and Women’s Empowerment. New York: Oxford. Mahmud, Simeen. 2003. How empowering is microcredit? Development and Change 34(4): 577‐605. Malhotra,Anju, Sidney Shuler, and Carol Boender. 2002. Measuring women’s empowerment as a variable in development. Background Paper Prepared for the World Bank Workshop on Poverty and Gender. Mayoux, Linda. 2001. Tackling the downside: Social capital, women’s empowerment, and microfinance in Cameroon. Development and Change 32: 435‐464. Meyers , Diana. 1991.Self , society , and personal choice. New York: Columbia. Montgomery, Richard,Debapriya Bhattacharya, and David Hulme. 1986. Credit for the poor in Bangladesh: the BRAC Rural Development Programme and the Government Thana Resource Development and Employment Programme. In Finance Against Poverty, ed. Hulme and Mosley.London: Routeledge. Narayan, Deepa. 2000. Voices of the poor volume 1: Can anybody hear us? New York: Oxford. Narayan Uma. 2005. Informal sector work, microcredit, and women’s empowerment: A critical overview. Unpublished manuscript on file with author. ‐‐‐‐‐‐‐‐. 2002. Minds of their own: Choices, autonomy, cultural practices, and other women. A mind of one’s own: Feminist essays on reason and objectivity, ed. Louise M. Antony and Charlotte E. Witt. Boulder, Colorado: Westview. Osmani,Lutfun. 1998. The impact of microcredit on the relative well‐being of women. IDS Bulletin29 (4): 31‐38. Oxaal, Zoe and Sally Baden. 1997. Gender and empowerment: Definitions, approaches, and implications for policy. Swedish International Development Agency. Rawls , John. 1978. A theory of justice.Cambridge: Harvard. SenAmartya. 1990. Gender and cooperative conflicts. In Persistent inequalities: Women and world development, ed.Irene Tinker. Oxford: Oxford. ‐‐‐‐‐‐‐‐.2002.Rationality and Freedom. Cambridge: Belknap. Shuler,Sidney Ruth, Syed A. Hashemi, and Shamsul Huda Badal. 1998. Men’s violence against women in rural Bangladesh: Undermined or exacerbated by microcredit programs? Development in Practice 8(2): 148‐57. Swain, Ranjula.2007. Can microfinance empower women? ADA Dialogue 37:61‐82. United Nations.2013. Goal 3: Promote Gender Equality and Empower Women. http:,,www.un.org,millenniumgoals,gender.shtml (accessed February 3, 2013). 22