After the Family Wage: Gender Equity and the Welfare State Author(s): Nancy Fraser Reviewed work(s):

Source: Political Theory, Vol. 22, No. 4 (Nov., 1994), pp. 591-618 Published by: Sage Publications, Inc. Stable URL: http://www.jstor.org/stable/192041 . Accessed: 11/12/2011 11:17
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.

Sage Publications, Inc. is collaborating with JSTOR to digitize, preserve and extend access to Political Theory.

http://www.jstor.org

AFTER THE FAMILY WAGE Gender Equity and the Welfare State

NANCYFRASER NorthwesternUniversity

HE CURRENTCRISISOF THEWELFARE STATE has manyrootsmassive economic movements of trends, refugees and Immigrants, global to the trade of unions and labor parties, weakening popularhostility taxes, the nse of nationaland"racial"-ethnic the decline of solidanstic antagonisms, and the of state socialism. One ideologies, collapse absolutelycrucialfactor, however,is the crumblingof the old genderorder.Existing welfarestates are premisedon assumptionsaboutgenderthatareincreasinglyout of phase with They thereforedo not provide many people's lives and self-understandings. for women social and children. protections,especially adequate The gender order that is now disappearing descends from the industrial era of capitalismand reflects the social world of its origin. It was centered on the ideal of the family wage. In this world, people were supposed to be organized into heterosexual, male-headed nuclear families, which lived princlpally from the man's labor marketearnings. The male head of the household would be paid a family wage, sufficientto supportchildrenand a wife and mother, who performeddomestic labor without pay. Of course, countless lives never fit this pattern.Still, it providedthe normativepicture of a properfamily. The family-wageideal was inscribedin the structure of most industrial-era welfarestates.' Thatstructure hadthreetiers,with social-insurance programs occupying the first rank.Designed to protectpeople from the vagariesof the
AUTHOR'SNOTE:Researchfor this article was supportedby the Centerfor UrbanAffairsand Policy Research, NorthwesternUniversity.For helpful comments,I am indebted to Rebecca Blank Joshua Cohen, Fay Cook, Barbara Hobson, Axel Honneth,Jenny Mansbrdge, Linda Nicholson, Ann Shola Orloff John Roemer,lan Schapiro, TracyStrong, Peter Taylor-Gooby, Eli Zaretsky,and the membersof the FeministPublic Policy WorkGroupof the Judy Wittner, Centerfor UrbanAffairsand Policy Research,NorthwesternUniversity. POLITICAL Vol.22 No.4, November 1994 591-618 THEORY, ? 1994SagePublications, Inc. 591

592

THEORY / November 1994 POLITICAL

labormarket(and to protectthe economy from shortagesof demand),these programsreplaced the breadwinner'swage in case of sickness, disability, unemployment,or old age. Many countriesalso featureda second tier of programs,providing direct supportfor full-time female homemakingand mothering. A third tier served the "reslduum." Largely a holdover from traditional poor relief, these programsprovidedpaltry,stigmatized,meanstested ad to needy people who had no claim to honorablesupportbecause they did not fit the family-wage scenario.2 Today,however,the family-wageassumptionis no longertenable-either We are currentlyexperiencingthe death throes empincally or normatively. of the old, industrialgender orderwith the transitionto new, postindustrial phase of capitalism.The crisis of the welfare state is bound up with these epochalchanges.It is rootedin partin the collapse of the worldof the family wage, and of its centralassumptionsaboutlabormarketsand families. In the labor marketsof postindustral capitalism, few jobs pay wages sufficient to supporta family single-handedly; many,in fact, are temporary or part-timeand do not carry standardbenefits.3Women's employment is increasinglycommon, moreover-although far less well-paid than men's.4 Postlndustrial families, meanwhile,areless conventionaland more diverse.5 Heterosexualsare marryingless and later,and divorcing more and sooner. And gays and lesbians arepioneeringnew kindsof domestic arrangements.6 Gendernormsand family forms are highly contested,finally.Thanksin part to the feminist and gay and lesbian liberationmovements,many people no model. As a resultof homemaker longerpreferthe male breadwinner/female these trends,growing numbersof women, bothdivorcedand never married, are strugglingto supportthemselves and their families without access to a male breadwinner's wage.7 is In short,a new world of economic productionand social reproduction emerging-a world of less stable employmentand more diverse families. Although no one can be certain about its ultimateshape, this much seems clear: the emerging world, no less than the world of the family wage, will If requirea welfarestatethateffectively insurespeople againstuncertainties. the is that It increased. is need for such the clear, too, protection anything, families and old formsof welfarestate,built on assumptionsof male-headed relatively stablejobs, are no longer suited to providingthis protection.We need something new, a postindustrialwelfare state suited to radically new conditionsof employmentand reproduction. welfarestate look like? Conservatives What then should a postindustrial the welfare state,"but their have lately had a lot to say about"restructuring and contradictory; vision is counterhistorical they seek to reinstatethe male homemakerfamily for the middleclass, while demandbreadwinner/female

Fraser/ GENDEREQUITYAND THE WELFARE STATE

593

ing that poor single mothers work. Neoliberal proposals have recently emerged in the United States, but they too are inadequatein the current and obsessed with employmentdespite the context. Punitive, androcentric, absence of good jobs, they are unableto providesecurityin a postindustrial world.8 welfare Both of these approaches ignoreone crucialthing:a postindustrlal must a order. But the state, like its industrial predecessor, support gender only kind of genderorderthatcan be acceptabletoday is one premisedon gender equity. Feminists, therefore,are in a good position to generatean emancipatory the imporvision for the coming period.They, more thananyone,appreciate tance of gender relationsto the currentcnsis of the industrialwelfare state and the centralityof genderequity to any satisfactoryresolution.Feminists also appreciatethe importanceof care work for human well-being and the effects of its social organizationon women's standing. They are attuned, finally, to potentialconflicts of interestwithinfamilies andto the inadequacy definitionsof work. of androcentric To date, however, feminists have tended to shy away from systematic reconstructivethinkingabout the welfare state. Nor have we yet developed a satisfactory account of gender equity that can inform an emancipatory such thinking.We should ask: What new, vision. We need now to undertake order should replace the family wage? And what sort postindustrialgender of welfare state can best supportsuch a new genderorder9Whataccountof And whatvision of social genderequitybest capturesourhighestaspirations? welfare comes closest to embodying it? Two differentlknds of answersarepresentlyconceivable, I think,both of which qualify as feminist. The first I call the universalbreadwinner model. of It is the vision implicitInthe current most U.S. feminists politicalpractice and liberals. It alms to foster genderequity by promotingwomen's employment; the centerpiece of this model is state provision of employmentenabling services such as day care. The second possible answer I call the caregiverparitymodel. It is the vision Implicitin the current politicalpractice of most WesternEuropean feministsandsocial democrats.It alms to promote gender equity chiefly by supportinginformalcare work; the centerpieceof this model is state provision of caregiverallowances. Which of these two approachesshould command our loyalties in the vision of a postindustrial coming perod? Whichexpressesthe most attractive order? Which best embodies the ideal of gender genderequity? In this article, I outline a frameworkfor thinking systematically about these questions.I analyzehighly idealizedversionsof universalbreadwinner and caregiver panty in the manner of a thought experment. I postulate,

594

/ November 1994 THEORY POLITICAL

contraryto fact, a world in which both of these models are feasible in that their economic and political preconditionsare in place. Assuming very favorableconditionsthen, I assess the respectivestrengthsand weaknesses of each. The result is not a standard policy analysis. For neitheruniversalbreadwinnernorcaregiverparitywill in fact be realizedin the nearfuture,and my at policy-makingelites. My intent,rather, discussionis not directedprimarily is theoreticaland political in a broadersense. I aim, first, to clarify some and "difference" dilemmassurrounding by reconsideringwhat is "equality" meant by genderequity.In so doing, I also aim to spurincreasedreflection on feminist strategiesand goals by spelling out some assumptionsthat are implicit in currentpracticeand subjectingthem to criticalscrutiny. These aims converge in the overall logic of my argument. Startingfrom some widely held moral intuitions, I arrive by the end of the thought conclusion:withrespectto social andcontroversial at a surprising experiment welfare, at least, the deconstructionof gender difference is a necessary conditionfor genderequity. My discussion proceeds in four parts. In the first section, I propose an Then, analysis of genderequity thatgeneratesa set of evaluativestandards. to universalbreadin the second and thirdsections, I apply those standards winner and caregiverparity,respectively.I conclude, in the fourthsection, that neitherof those approaches,even in an idealizedform, can deliver full genderequity.To have a shot at that,I contend,we mustdevelop a new vision welfare state, which effectively deconstructsgender difof a postindustrlal ference as we know it.

CONCEPTION A COMPLEX I. GENDEREQUITY. welfare state, we need To evaluatealternativevisions of a postindustrial some normativecriteria.Gender equity, I have said, is one indispensable But of whatprecisely does it consist? standard. Feminists have so far associated gender equity with either equality or difference, where equality means treating women exactly like men, and womendifferentlyinsofaras theydifferfrom wheredifferencemeanstreating men. Theoristshave debatedthe relativemerits of these two approachesas These if they representedtwo antitheticalpoles of an absolutedichotomy.9 have of difference in ended stalemate. have Proponents generally arguments as "the male that shown equalitystrategiestypicallypresuppose successfully on a standard distorted women and imposing norm,"therebydisadvantaging

Fraser/ GENDER EQUITYAND THE WELFARESTATE

595

have argued just as cogently,however,thatdifference everyone. Egalitarians on essentialist notions of femininity,therebyreinrely approachestypically and confining women within existing gender forcing existing stereotypes divisions. Neitherequality nor difference,then, is a workableconception of gender equity. Feministshaverespondedto this stalematein severaldifferentways. Some have tred to resolve the dilemmaby reconceivingone or anotherof its horns; difference or equality in what they consider a more they have relnterpreted defensible form. Othershave concluded"a plague on both your houses"and sought some third,wholly other,normativeprinciple.Still others have tred to embracethe dilemma as an enablingparadox,a resourceto be treasured, not an impasse to be gotten around.Many feminists, finally, have retreated altogether from normativetheorzing-into culturalpositivism, piecemeal reformism,or postmodernantinomianism. Normativetheorizingremainsan None of these responsesis satisfactory. intellectual for enterprise feminism, indeed for all emancipaindispensable We need a vision or pictureof where we are trying social movements. tory to go, and a set of standardsfor evaluatingvariousproposalsas to how we theoreticalimpasseis real,moreover; mightget there.The equality/difference it cannotbe simply sidesteppedor embraced.Nor is thereany "whollyother" thirdtermthatcan magicallycatapultus beyondit. Whatthenshouldfeminist theorsts do9 I propose that we reconceptualizegender equity as a complex, not a simple, idea. This meansbreakingwith the assumptionthatgenderequitycan be identifiedwith any single valueor norm,whetherit be equality,difference, or somethingelse. Insteadwe should treatit as a complex notioncomprising a pluralityof distinct normativeprinciples.The pluralitywill include some notions associated with the equality side of the debate, as well as some associated with the differenceside. It will also encompassstill othernormative ideas that neitherside has accordeddue weight. Assume, for example, that gender equity requiresnot only equal respect for women and men, but also some more substantivekind of equality,such as equality of resourcesor equalityof capabilities.Assume, in addition,that it requires not only parity of participationin socially valued activities, but also the decenteringof androcentrlc measuresof social value. In that case, each of four distinct norms must be respected for gender equity to be achieved. Failureto satisfy any one of them means failureto realize the full meaning of genderequity. This lknd of approach promises several advantages. Treating gender equity as a complex idea lets us spot possible tensions among its component norms. We can see, for example, that some efforts to equalize resources

596

THEORY / November 1994 POLITICAL

between women and men can work at cross-purposeswith some efforts to in socially valued activities,and we can look achieve parityof participation for ways to minimize such conflicts. No longer restricted to the two meganormsof equalityand difference,moreover,we have more conceptual resourcesat our disposal. We can develop more fine-grainedappraisalsof alternative political strategiesand goals. In what follows, I assume thatgenderequity is complex in this way, and I propose an account of it that is designed for the specific purpose of welfarestate.This account evaluatingalternative picturesof a postindustrial other thanwelfare.For such suited to issues be not handling perfectly might issues, it mightbe best to devise a somewhatdifferentpackageof component norms.Nevertheless,I believe thatthe generalidea of treatinggenderequity as a complex conception is widely applicable.The analysis here may serve the usefulnessof this approach. as a paradigmcase demonstrating For this particular thoughtexperiment,in any case, I unpackthe idea of as a compoundof five distinct normativepnnciples. Let me gender equity enumeratethem one by one. Principle Antipoverty The first and most obvious objective of social-welfare provision is to prevent poverty. Preventingpoverty is crucial to achieving gender equity now, after the family wage, given the high rates of poverty in solo-mother families and the vastly increasedlikelihood that U.S. women and children If it accomplishes nothingelse, a welfare state will live in such families.10 should at least relieve suffering by meeting otherwise unmet basic needs. such as those in the UnitedStates,thatleave women,children, Arrangements, and men in poverty,areunacceptable accordingto this criterion.Any postindustrialwelfare state that preventedsuch poverty would constitutea major advance.So far,however,this does not say enough.The antipoverty principle are which of all not of different in a satisfied be ways, variety might and of the as such Some isolating targeted, provision ways, acceptable. stigmatizedpoorrelief for solo-motherfamilies, fail to respectseveralof the following normativeprinciples,which are also essential to genderequity in social welfare. Principle Antiexploitation not only in themselves, but also as a Antipovertymeasuresare important means to another basic objective: preventing exploitation of vulnerable

Fraser/ GENDEREQUITYAND THE WELFARESTATE

597

people.1 This princlple,too, is centralto achieving gender equity after the family wage. Needy women with no otherway to feed themselves and their children, for example, are liable to exploitation-by abusive husbands,by sweatshop foremen, and by pimps. In guaranteeingrelief of poverty then, welfare provision should also aim to mitigateexploitabledependency.'2 The alternative of the of an source income enhances bargaining availability position of subordinatesin unequal relationships.The nonemployed wife who knows she can supportherself and her childrenoutside of her marriage has more leverage within it; her "voice" is enhancedas her possibilities of in "exlt"increase.13 The same holds for the low-paid nursinghome attendant relation to her boss.'4 For welfare measures to have this effect, however, supportmust be providedas a matterof right.When receiptof aid is highly the antiexploitatlonprincipleis not satisfied.'5 stigmatizedor discretionary, At best, the claimantwould tradeexploitabledependenceon a husbandor a boss for exploitabledependenceon a caseworker'swhim.16 The goal should be to prevent at least three kinds of exploitable dependencies:exploitable dependence on an individualfamily member,such as a husbandor an adult andexploitable child; exploitabledependenceon employersandsupervisors; on the whims of state officials. Rather than shuttle personal dependence forth and these back an among exploitable dependencies, adequate people This principle rules out approachmust prevent all three simultaneously."7 that channel a homemaker's benefits throughher husband.It arrangements is likewise incompatible with arrangementsthat provide essential goods, such as health insurance,only in forms linked conditionallyto scarce emwelfare state thatsatisfied the antiexploitation ployment. Any postindustrial a would principle represent majorimprovementover currentU.S. arrangements. But even it might not be satisfactory.Some ways of satisfying this would fail to respectseveral of the following normativeprinclples, prnnclple which are also essential to genderequity in social welfare. EqualityPrinciples A postindustrial welfarestatecould preventwomen's povertyand exploitation and yet still toleratesevere gender inequality.Such a welfare state is not satisfactory.A furtherdimension of genderequity in social provision is redistribution, reducinginequalitybetween women andmen. Equality,as we saw, has been criticizedby some feminists. They have arguedthat it entails treating women exactly like men accordingto male-definedstandards,and thatthis necessarilydisadvantageswomen. Thatargument expresses a legitimate worry,which I will addressunderanotherrubricbelow. But it does not

598

THEORY/ November 1994 POLITICAL

underminethe ideal of equality per se. The worry pertainsonly to certain inadequateways of conceiving equality,which I do not presupposehere. At least threedistinctconceptionsof equalityescape the objection.These three are essential to genderequity in social welfare. Income equality. One form of equality that is crucial to gender equity concerns the distributionof real per capita income. This kind of equality is highly pressingnow, afterthe family wage, when U.S. women'searningsare less than70% of men's, when muchof women's laboris not compensatedat all, and when many women suffer from "hiddenpoverty"due to unequal distribution within families."8As I interpretit, the pnnclple of income equalitydoes not requireabsoluteleveling, but it does rule out arrangements that reduce women's incomes after divorce by nearly half, whereas men's incomes nearlydouble. 9 It likewise rulesout unequalpay for equalworkand of women'slaborandskills.The incomeequality the wholesaleundervaluation prnciple requiresa substantialreductionin the vast discrepancybetween men's and women's incomes. In so doing, it tends, as well, to help equalize the life-chancesof children,becausea majorityof U.S. childrenarecurrently likely to live at some point in solo-motherfamilies.20 Leisure-timeequality.A second kind of equalitythatis crucialto gender of leisuretime. This sortof equalityis highly equityconcernsthe distribution when manywomen,butonly a few men, the after now, family wage, pressing do both paid work and unpaidprimarycare work, and when women suffer One recentBrtish studyfoundthat from"timepoverty."21 disproportionately 52%of women surveyed,comparedto 21%of men, said they "felttiredmost The leisure-timeequalityprinciplerules out welfarearrangeof the time."22 ments that would equalize incomes while requiringa double shift of work from women, but only a single shift from men. It likewise rules out arrangements that would requirewomen, but not men, to do either the "work of of piecing together income claiming" or the time-consuming"patchwork" servicesfromdifferentagencies and fromseveralsourcesandof coordinating
associations.23

Equalityof respect.A thirdkindof equalitythatis crucialto genderequity pertainsto statusandrespect.This kindof equalityis especiallypressingnow, after the family wage, when postindustrialculture routinely represents women as sexual objects for the pleasureof male subjects.The principleof equal respect rules out social arrangementsthat objectify and denigrate women-even if those arrangements preventpoverty and exploitation,and and leisuretime. It is incompatible income even if, in addition,they equalize

Fraser/ GENDEREQUITYAND THE WELFARESTATE

599

thattrivializewomen's activitiesand ignorewomen's with welfareprograms contributions-hence with welfare reformsin the United States thatassume AFDC claimantsdo not "work."Equalityof respectrequiresrecognitionof women's personhoodand recognitionof women's work. A postindustrialwelfare state should promoteall three of these conceptions of equality.Such a state would constitutean enormousadvance over but even it might not go far enough. Some ways of present arrangements, the satisfying equalityprinclpleswouldfail to respectthefollowing princlple, which is also essential to genderequity in social welfare. Antimarginalization Principle A welfare statecould satisfy all the precedingpnnciples and still function to marginalizewomen. By limiting supportto generous mothers'pensions, for example, it could renderwomen independent,well provided for, well rested, and respected,but enclaved in a separatedomestic sphere, removed fromthe life of the largersociety. Such a welfarestatewouldbe unacceptable. on a par with men Social policy should promotewomen's full participation in all areasof social life-in employment,in politics, in the associationallife of civil society. The antimarglnalizatlon princlplerequiresprovision of the conditions for women's necessary participation,including day care, elder and for care, provision breast-feedingin public. It also requiresthe dismanworkculturesandwoman-hostilepoliticalenvironments. of masculinist tling welfare state that providedthese things would represent Any postindustrial a great improvementover currentarrangements.Yet even it might leave something to be desired. Some ways of satisfying the antimarglnalization principle would fail to respect the last principle, which is also essential to gender equity in social welfare. Antiandrocentrism Principle A welfare state that satisfied many of the foregoing principlescould still entrench some obnoxious gender norms. It could assume the androcentrlc view that men's currentlife patternsrepresentthe human norm and that women ought to assimilate to them. (This is the real issue behind the previously noted worryaboutequality.)Such a welfarestate is unacceptable. Social policy should not requirewomen to become like men, nor to fit into institutions designed for men, to enjoy comparable levels of well-being. Policy should aim instead to restructureandrocentricinstitutions so as to welcome humanbeings who can give birthand who often care for relatives

600

THEORY / November 1994 POLITICAL

andfriends,treatingthemnot as exceptions,but as ideal-typicalparticipants. The antiandrocentrlsm principlerequiresdecenteringmasculinlstnorms-in because partby revaluingpracticesand traitsthatare currentlyundervalued as It men well as with women. entails associated are changing changing they women. Herethenis anaccountof genderequityin social welfare.On this account, prnncples, genderequityis a complex ideacomprisingfive distinctnormative one of whlch-equality-is internallycomplex and encompassesthreedistinct subpnnclples.Each of the princlplesis essentialto genderequity.Thus no postindustrialwelfare state can realize gender equity unless it satisfies them all. Some of the five tend usually to How then do the pnnclples interrelate? well work at cross-purposes.Everything, could others one another; support in fact, dependson context. Some institutional arrangements permitsimultaneous satisfactionof several principleswith a minimumof mutualinterferin contrast,set up zero-sum situations,in which ence; other arrangements, one to principle interferewith attemptsto satisfy another. attempts satisfy Promotinggenderequity after the family wage, therefore,means attending to multiple alms that are potentiallyin conflict. The goal should be to find approaches that avoid trade-offs and maximize prospects for satisfying all-or at least most-of the five princlples. to assess two alternativemodels In the next sections, I use this approach welfare state. First, however,I want to flag three sets of of a postindustrial of care work.Precisely relevantissues. One concernsthe social organization how this work is organizedis crucialto humanwell-being in generaland to In the eraof the family wage, care the social standingof women in particular. work was treatedas the privateresponsibilityof individualwomen. Today, however,it can no longerbe treatedin thatway.Some otherway of organizing it is required,but a numberof differentscenariosare conceivable. In evaluwelfare state models then, we must ask: how is responating postindustrial allocatedbetween such institutionsas the family, the work care for sibility market,civil society, and the state?And how is responsibilityfor this work assigned within such institutions:by gender9by class? by "race"-ethniclty9 by age? A second set of issues concernsdifferencesamong women. Genderis the prncipal focus of this article,to be sure,but it cannotbe treateden bloc. The lives of womenandmenarecross-cutby severalothersalientsocial divisions, sexuality,and age. Models of postindustrial includingclass, race-ethnicity, welfare states, then, will not affect all women-nor all men-in the same way; they will generatedifferentoutcomes for differentlysituatedpeople. For example, some policies will affect women who have childrendifferently

Fraser/ GENDEREQUITYAND THE WELFARESTATE

601

from those who do not; some, likewise, will affect women who have access to a second income differentlyfromthose who do not;and some, finally,will affect women employedfull-timedifferentlyfromthose employedpart-time, and differentlyyet again from those who are not employed. For each model then, we must ask: which groupsof women would be advantagedand which groups disadvantaged? A thirdset of Issues concernsdesideratafor postindustral welfare states otherthangenderequity.Genderequity,afterall, is not the only goal of social welfare. Also importantare nonequitygoals, such as efficiency, community, and individual liberty.In addition,there remainother equity goals, such as racial-ethnic equity, generational equity, class equity, and equity among here.Some of them, nations.All of these issues arenecessarilybackgrounded however, such as racial-ethnicequity,could be handledvia parallelthought equity as a complex idea, analoexperiments:one might define racial-ethnic the is treated and then use it, too, to assess to here, way gender equity gous of a state.24 visions welfare postindustrial competing With these considerationsin mind, let us now examine two strikingly different feminist visions of a postindustrialwelfare state, and let us ask: which comes closest to achievinggenderequity in the sense I haveelaborated here?

BREADWINNER MODEL II. UNIVERSAL In one vision of postindustrial society, the age of the family wage would the of to the universal breadwinner. This is the vision implicit age give way in the currentpolitical practiceof most U.S. feminists and liberals. (It was also assumed in the former state-socialist countries!) It alms to achieve gender equity principallyby promotingwomen's employment.The point is to enable women to supportthemselves and theirfamilies throughtheirown role is to be universalized,in sum, so that wage earning. The breadwinner women too can be citizen-workers. Universalbreadwinner is a very ambitiouspostindustrial scenaro, requirnew ing major programs and policies. One crucial element is a set of employment-enablingservices, such as day care and elder care, aimed at freeing women from unpaidresponsibilitiesso that they can take full-time Anotheressential element is a employment on terms comparableto men.25 set of workplace reforms aimed at removing equal-opportunity obstacles, such as sex discriminationandsexual harassment. Reformingthe workplace requiresreformingthe culturehowever-eliminating sexist stereotypesand

602

THEORY POLITICAL / November1994

breakingthe cultural association of breadwinningwith masculinity.Also requiredare policies to help change socialization, so as first, to reorient women's aspirationstowardemploymentand away from domesticity,and second, to reorientmen's expectationstowardacceptanceof women's new role. None of this would work, however,withoutone additionalingredient: macroeconomicpolicies to createfull-time,high paying,permanent jobs for in the primary women.26 These would have to be truebreadwinnerjobs labor force, carryingfull, first-class social-insuranceentitlements.Social InsurThe aim here is to bnng ance, finally, is centralto universalbreadwinner. thathas traditionally women up to paritywith men in an institution disadvantaged them. How would this model organizecarework?The bulkof such work would be shifted from the family to the marketand the state, where it would be Who then are these employees likely to performedby employees for pay.27 be? In the United States today,paid institutionalcare work is poorly remubut such arrangements nerated,largely feminized, and largely raclalized,28 areprecludedin this model. If the model is to succeed in enablingall women it must upgradethe statusandpay attachedto care work to be breadwinners, it too into primarylaborforce work. Universalbreadmaking employment, worth";it winner,then, is necessarilycommittedto a policy of "comparable coded of skills andjobscurrently undervaluation mustredressthe widespread as feminine and/or "non-White,"and it must remuneratesuch jobs with breadwinner-level pay. Universal breadwinnerwould link many benefits to employment and distributethem throughsocial insurance.In some cases, such as pensions, benefit levels would varywith earnings.In this respect,the model resembles The difference is that many more women the industrialera welfare state.29 would be covered on the basis of theirown employmentrecords,and many more women's employment records would look considerably more like men's. Not all adultscan be employed,however.Some will be unableto workfor medical reasons, including some not previously employed. Others will be unable to get jobs. Some, finally, will have care work responsibilitiesthat they are unable or unwilling to shift elsewhere. Most of these last will be must include a women. To provide for these people, universalbreadwinner residualtier of social welfarethatprovides need-based,means-testedwage replacements.30 is far removed from presentrealities. It requires Universal breadwinner massive creation of primarylabor force jobs-jobs sufficient to supporta family single-handedly.That, of course, is wildly askew of currentpostinbutfor "disposable dustrialtrends,which generate jobs not for breadwinners,

Fraser/ GENDEREQUITYAND THE WELFARESTATE

603

Let us assume for the sake of the thoughtexperiment,however, workers."31 thatits conditionsof possibilitycould be met, and let us considerwhetherthe resultingpostindustral welfarestate could claim title to genderequity. Antipoverty would do a We can acknowledge straightoff that universalbreadwinner A of that created secure breadwinnergood job preventingpoverty. policy quality jobs for all employable women and men-while providing the services that would enable women to take such jobs-would keep most families out of poverty,and generous levels of residualsupportwould keep the rest out of povertythroughtransfers. Failingthat,however,severalgroups are especially vulnerableto poverty in this model: those who cannot work, those who cannot get secure, permanent, full-time, good-paying jobsdisproportionatelywomen and/or people of color; and those with heavy, hard-to-shift, unpaidcareworkresponsibilities-disproportionatelywomen. Antiexplottation The model should also succeed in preventingexploitabledependencyfor most women. Womenwith secure breadwinner jobs are able to exit unsatisfactory relations with men, and those who do not have such jobs but know they can get them will also be less vulnerableto exploltation.Failing that, the residual system of income supportprovides back-upprotectionagainst and exploitabledependency-assuming thatit is generous,nondiscretionary, honorable.Failing that,however, the groups mentionedabove remainespecially vulnerableto exploltation-by abusive men, by unfair or predatory employers, by capriciousstate officials. Equality Incomeequality.Universalbreadwinner is only fair,however,at achieving income equality. Granted,secure breadwinner jobs for women-plus the services that would enable women to take them-would narrowthe gender
wage gap.32Reduced inequality in earnings, moreover, translates into reduced

inequality in social-insurancebenefits, and the availability of exit options from marriageshould encouragea more equitabledistributionof resources within it. But the model is not otherwiseegalitarian.It containsa basic social fault line dividing breadwinners from others, to the considerabledisadvanof the others-most of whom would be women. Apartfromcomparable tage

604

THEORY / November 1994 POLITICAL

worth,moreover,it does not reducepay inequalityamong breadwinner jobs. To be sure, the model reducesthe weight of genderin assigning Individuals to unequally compensatedbreadwinner jobs, but it thereby increases the of other variables, class, education,race-ethnicity,and presumably weight are in relationto thosevarables men-who Women-and disadvantaged age. will earnless thanthose who are not. Leisure-time equality. The model is poor, moreover, with respect to It equality of leisure time, although it improves on currentarrangements. assumes thatall of women'scurrentdomesticand care workresponsibilities can be shifted to the marketand/orthe state. But thatassumptionis patently unrealistic.Some things,such as childbearing, attendingto family emergencies, and much parentingwork, cannotbe shifted-short of universalsurroOtherthings, such as arrangements. gacy and otherpresumablyundesirable we were be could and (some) shifted-provided housekeeping, cooking levels of commodior collective to high living arrangements prepared accept fication.Even those tasksthatare shifted,finally,do not disappearwithouta trace, but give rise to burdensomenew tasks of coordination.Women's chances for equal leisure,then,dependon whethermen can be Inducedto do theirfair shareof this work. On this, the model does not inspireconfidence. Not only does It offer no disincentivesto free riding, but in valorizingpaid work, it implicitlydenigratesunpaidwork,therebyfueling the motivationto Women withoutpartnerswould, in any case, be on their own. And shirk.33 those in lower-incomehouseholds would be less able to purchasereplacement services. Employed women would have a second shift on this model one thansome have now; and therewould be then, albeit a less burdensome in sum, is many more women employed full-time. Universalbreadwinner, not likely to deliver equal leisure. Anyone who does not free ride in this world is likely to be harriedand tired. possible postlndustrlal Equality of respect. The model is only fair, moreover, at delivering equality of respect.Because it holds men and women to the single standard of the citizen-worker,its only chance of eliminatingthe genderrespect gap is to admitwomen to thatstatuson the same termsas men. This, however,is unlikely to occur.A more likely outcome is that women would retainmore connection to reproductionand domesticity than men, thus appearingas breadwinnersmanque.In addition,the model is likely to generateanother kind of respect gap. By puttinga high premiumon breadwinnerstatus, it residualsystem invites disrespectfor others.Participants Inthe means-tested will be liable to stigmatization,and most of these will be women. Any

Fraser/ GENDEREQUITYAND THE WELFARESTATE

605

model, even a feministone, has a hardtime constructemployment-centered status for those it defines as nonworkers. honorable an ing Antimarginalization This model is also only fair at combating women's marglnalization. in employment,but its definition Granted,it promoteswomen's participation of participationis narrow.Expecting full-time employment of all who are in politics andcivil society. able, the model may actuallyimpedeparticipation to women's in those arenas. it does nothing promote participation Certainly in a "workenst" It fights women's marglnalization, one-sided, then, way. Antiandrocentrtsm Finally, the model performs poorly in overcoming androcentrism.It valorizes men's traditionalsphere-employment-and simply tries to help female care work, in contrast,is treatedinstruwomen fit in. Traditionally, be It is not it what must is mentally; sloughed off to become a breadwinner. itself accordedsocial value. The ideal typicalcitizen hereis the breadwinner, now nominally gender neutral. But the content of the status is implicitly masculine;it is the male half of the old breadwinner/homemaker couple, now universalizedand requiredof everyone. The female half of the couple has simply disappeared.None of her distinctivevirtues and capacities has been to men.Themodelis androcentnc. for women,let aloneuniversalized preserved in Figure 1. Not We can summarizethe merits of universalbreadwinner to women breadwinner delivers the best outcomes universal surprisingly, whose lives most closely resemblethe male half of the old family-wageideal couple. It is especially good to childless women and to women withoutother major domestic responsibilitiesthat cannot easily be shifted to social services. But for those women, as well as for others,it falls shortof full gender equity.

PARITY MODEL III. CAREGIVER In a second vision of postindustrialsociety, the era of the family wage would give way to the era of caregiverparity.This is the pictureimplicit in the political practiceof most WesternEuropeanfeminists and social democrats. It aims to promote gender equity principallyby supportinginformal

606

POLITICAL THEORY/ November1994

UniversalBreadwinner

Antipoverty Antiexploitation Incomeequality Leisure-time equality of respect Equality Antimarginalization Antiandrocentnsm Figure 1.

Good Good Fair Poor Fair Fair Poor

care work. The point is to enable women with significantdomestic responsibilities to supportthemselves and their families, either throughcare work alone or throughcare work plus part-timeemployment. (Women without significantdomestic responsibilitieswould presumablysupportthemselves throughemployment.)The aim is not to make women's lives the same as men's, but rather to "make difference costless."34Thus childbearing, childrearing,and informaldomestic labor are to be elevated to paritywith formalpaidlabor.The caregiverrole is to be puton a parwiththe breadwinner role-so that women and men can enjoy equivalent levels of dignity and well-being. Caregiverparity is also extremely ambitious.On this model, many (although not all) women will follow the current U.S. female practice of alternating spells of full-timeemployment,spells of full-timetare work,and careworkwith part-time that combine employment.The aim part-time spells is to make such a life patterncostless. To this end, several major new programsare necessary.One is a programof caregiverallowances to compensate childbearing,childraising,housework,and other forms of socially necessarydomesticlabor;the allowancesmustbe sufficientlygenerousat the full-timerateto supporta family-hence equivalentto a breadwinner wage.35 Also requiredis a programof workplacereforms.These must facilitatethe possibility of combining supportedcare work with part-timeemployment between differentlife states. The key here is flexibility. and of transitioning One obvious necessity is a generous programof mandatedpregnancyand family leave so thatcaregiverscan exit andenteremploymentwithoutlosing security or seniority.Anotheris a programof retrainingand job search for those not returningto old jobs. Also essential is mandatedflextime so that theircare workresponsibilicaregiverscan shift theirhoursto accommodate fulland between shifts ties, including employment.Finally,in the part-time to ensurecontinuityof all wake of all this flexibility,theremustbe programs

Fraser/ GENDEREQUITYAND THE WELFARESTATE

607

the basic social-welfarebenefits,includinghealth,unemployment, disability, and retirementinsurance. This model organizescareworkverydifferentlyfromuniversalbreadwinner.Whereasthatapproach shiftedcare work to the marketandthe state,this one keeps the bulk of such work in the householdand supportsit with public funds.36 Caregiverparity's social-insurancesystem also differs sharply.To assure continuous coverage for people alternatingbetween care work and employment,benefits attachedto both must be integratedin a single desertbased system. In this system, part-time jobs and supportedcare work must be covered on the same basis as full-timejobs. Thus a woman finishing a spell of supportedcare work would be eligible for unemploymentinsurance benefits on the same basis as a recently laid off employee In the event she could not find a suitable job, and a supportedcare worker who became disabled would receive disabilitypaymentson the same basis as a disabled employee. Yearsof supportedcare work would count on a parwith years of pensions. Benefit levels would employmenttowardeligibility for retirement be fixed in ways thattreatcare work and employmentequivalently. residualtierof social welfare.Some Caregiverparityalso requiresanother, adults will be unableto do eithercare work or waged work, includingsome withoutpriorworkrecordsof eithertype. Most of these people will probably be men. To provide for them, the model must offer means-testedwage and allowance replacements.37 Caregiverparity'sresidualtier should be smaller than universal breadwinner's, however; nearly all adults should be covered in the integratedbreadwlnner-careglver system of social insurance. far is removed from currentU.S. arrangements. It Caregiverparity,too, of funds to hence requires large outlays public pay caregiver allowances, major structuraltax reform and a sea change in political culture. Let us assume for the sake of the thoughtexperiment,however, that its conditions of possibility could be met. And let us consider whether the resulting postindustrialwelfare state could claim title to genderequity. Antipoverty Careglverparity would do a good job of preventingpoverty-including for those women andchildrenwho arecurrently most vulnerable.Sufficiently allowances would solo-mother families out of povertydunng generous keep full-time of care and a combination of allowances and wages work, spells would do the same during spells of part-time supported care work and part-time employment. (Wages from full-time employment must also be sufficient to supporta family with dignity.) Because each of these options

608

/ November 1994 POLITICAL THEORY

would carry the basic social-insurancepackage, moreover, women with wouldhaveconsiderable feminineworkpatterns security.Adultswithneither care work nor employmentrecordswould be most vulnerableto poverty in this model;most of these wouldbe men. Children,in contrast,would be well protected. Antiexploitation Caregiverparityshould also succeed in preventingexploitationfor most women, includingthosewho aremost vulnerable today.By providingincome it reduces their economic dependence on to wives, directly nonemployed husbands.It also provideseconomic securityto single women with children, reducing their liability to exploitation by employers. Insofar as caregiver allowances are desertbased and nondiscretlonary, finally,recipientsare not with neithercarework it is adults Once whims. to caseworkers' again, subject nor employment records who are most vulnerableto exploitation in this model, and the majorityof them would be men. Income equality.Careglverparityperformsquite poorly, however, with respect to income equality.Although the system of allowances plus wages provides the equivalent of a basic minimum breadwinnerwage, it also institutesa "mommytrack"in employment-a marketin flexible, noncontinuousfull- and/orpart-time jobs. Most of thesejobs will pay considerably less even at the full-time rate than comparablebreadwinner-track jobs. on to one incentive an economic will have families partner keep Two-partner thanto sharespells of care workbetweenthem; trackrather the breadwinner the man will be most given currentlabor markets,makingthe breadwinner and socializaculture current Given heterosexual for couples. advantageous tion, moreover,men aregenerallyunlikelyto choose the mommytrackin the same proportionsas women. So the two employment tracks will carry traditional gender associations. Those associations are likely in turn to track.Caregiver against women in the breadwinner producediscrimination difference make will not it but less cost make difference then, parity may costless. Leisure-timeequality. Caregiverparitydoes somewhat better,however, with respectto equalityof leisuretime. It makes it possible for all women to avoid thedoubleshiftif theychoose, by optingforfull- orpart-time supported this choice is available care work at variousstages in theirlives. (Currently, only to a small percentageof privilegedU.S. women.) Wejust saw, however,

Fraser/ GENDEREQUITYAND THE WELFARESTATE

609

thatthis choice is not trulycostless. Some women with families will not want to forego the benefits of breadwinner-track employment and will try to combine it with care work. Those not partneredwith someone on the caregiver track will be significantly disadvantagedwith respect to leisure time, andprobablyin theiremploymentas well. Men, in contrast,will largely be insulatedfrom this dilemma.On leisuretime, then, the model is only fair. Equalityof respect.Caregiverparityis also only fairat promotingequality of respect.Unlike universalbreadwinner, it offers two differentroutesto that end. Theoretically, citizen-workers and citizen-careglversare statuses of equivalentdignity.But are they really on a parwith one another?Caregiving is certainly treated more respectfully in this model than in current U.S. society, but it remains associated with femininity. Breadwinninglikewise remainsassociatedwith masculinity.Given those traditional genderassociathe two the economic differential between tions, plus lifestyles, caregivingis with In true to attain breadwinning. general, it is hard to unlikely parity but how imagine "separate equal"genderroles could providegenuineequalof ity respect today. Antimarginalization Caregiverparityperformspoorly,moreover,in preventingwomen's marBy supportingwomen's informalcare work, it reinforces the gmnalization. view of such work as women's work andconsolidatesthe genderdivlsion of domestic labor. By consolidating dual labor marketsfor breadwinnersand women withinthe employment caregivers,moreover,the model marglnalizes sector. By reinforcingthe associationof caregiving with femininity,finally, it may also impede women's participation in other spheres of life, such as and civil society. politics Antiandrocentrism Yet caregiver parity is better than universal breadwinnerat combating androcentrsm. It treats caregivlng as intrinsicallyvaluable, not as a mere obstacleto employment,thuschallengingthe view thatonly men's traditional activities are fully human. It also accommodates feminine life patterns, therebyrejecting the demandthat women assimilate to masculine patterns. But the model still leaves something to be desired. Caregiverparty stops shortof affirmingthe universalvalue of activitiesand life patterns associated with women. It does not value caregivingenough to demandthat men do it

610

THEORY / November1994 POLITICAL

too; it does not ask men to change. Thus caregiverparity representsonly one-halfof a full-scalechallengeto androcentnsm. Here,too, its performance is only fair. in Figure2. Caregiverpanty's strengthsand weaknessesare summarized In general, careglverparityperformsbest for women with significant care work responsibilities.But for those women, as well as for others, it fails to deliver full genderequity.

GENDEREQUITY IN A POSTINDUSTRIAL IV CONCLUSION: STATE GENDER WELFARE DECONSTRUCTING REQUIRES and caregiverparityare highly utopian viBoth universalbreadwinner welfare a state. Eitherone of them would representa of sions postindustrial Yet neitheris likely to over current U.S. arrangements. majorimprovement be realized soon. Both models assume backgroundpreconditionsthat are strikinglyabsent today.Both presupposemajorpolitical-economicrestructhe capacityto turng, includingsignificantpubliccontrolover corporations, and the create to investment direct jobs, abilityto tax high-quality permanent at rates sufficient to fund wealth expandedhigh-qualitysocial profits and assume broad popularsupportfor a postindusprograms.Both models also trial welfare state thatis committedto genderequity. If bothmodels areutopianin this sense, neitheris utopianenough.Neither universal breadwinnernor caregiver paritycan actually make good on its promiseof genderequity-even undervery favorableconditions.Although botharegood at preventingwomen's povertyandexploitation,bothareonly holds women fair at redressinginequalityof respect:Universalbreadwinner that prevent to the same standardas men while constructingarrangements them from meeting it fully; caregiver parity,in contrast,sets up a double to accommodategenderdifferencewhile institutionalizing standard policies thatfail to assureequivalentrespectfor feminineactivitiesand life patterns. When we turnto the remainingcomponentsof genderequity,moreover,the two models'strengthsandweaknessesdiverge.Whereasuniversalbreadwinand at reducingincome ner is betterat preventingwomen's marginalizatlon and men women, caregiverparityis betterat redressing inequalitybetween Neither model, at combatingandrocentrism. and time leisure inequalityof with men in politics on a full women's however,promotes par participation and civil society. And neithervalues female-associatedpracticesenough to ask men to do them, too; neitherasks men to change. (The relativemeritsof universal breadwinnerand careglver parity are summarizedin Figure 3.)

Fraser/ GENDER EQUITYAND THE WELFARESTATE

611

CaregiverParity

Antipoverty Antiexploitation Incomeequality Leisure-time equality of respect Equality Antimarginalization Antiandrocentrism Figure 2.

Good Good Poor Fair Fair Poor Fair

Neither model, in sum, provides everythingthat feminists want. Even in a highly idealized form, neitherdelivers full genderequity. If these were the only possibilities, we would face a very difficult set of trade-offs. Suppose, however, we reject this Hobson's choice and try to develop a thirdalternative.The trick is to envision a postlndustrialwelfare state that combines the best of universal breadwinnerwith the best of caregiver parity, while jettisoning the worst features of each. What third alternativeis possible? So far, we have examined-and found wanting-two initially plausible one aimingto makewomenmorelike men arenow, andthe other approaches: leaving men and women pretty much unchanged, while aiming to make women's differencecostless. A thirdpossibility is to induce men to become more like most women are now-that is, people who do primarycare work. Consider the effects of this one change on the models we have just examined. If men were to do their fair share of care work, universalbreadwinner would come much closer to equalizing leisure time and eliminating androcentrism,whereas caregiver parity would do a much better job of Both models, in equalizing income and reducingwomen's marginalization. addition,would tend to promoteequalityof respect. If men were to become more like women are now, in sum, both models would begin to approach gender equity. The key to achieving genderequity in a postindustral welfarestate, then, is to make women's current life patterns the norm. Women today often combine breadwinning andcaregiving,albeitwith greatdifficultyand strain. A postindustrlalwelfare state must ensure that men do the same, while redesigning institutions so as to eliminate the difficulty and strain. Such a welfare state would promote gender equity by dismantling the gendered opposition between breadwinningand caregiving. It would integrateactivities that are currently separatedfrom one another,eliminate their gender coding, and encouragemen to performthem too.

612

POLITICAL THEORY / November 1994

UniversalBreadwinner

CaregiverParity

Antipoverty Antiexploitation Incomeequality Leisure-time equality of respect Equality Antimarginalization Antiandrocentrsm Figure 3.

Good Good Fair Poor Fair Fair Poor

Good Good Poor Fair Fair Poor Fair

to a wholesalerestructuring of the institution This, however,is tantamount of breadwinning andcaregivingas separate of gender.The construction roles, of the coded masculineandfemininerespectively,is a principalunderglrding currentgenderorder.To dismantlethose roles and theirculturalcoding is in effect to overturnthatorder.It meanssubvertingthe existing genderdivision of laborand reducingthe salienceof genderas a structural principleof social At the limit, It suggests deconstructing gender.39 organization.38 PresidentClinton has proclaimedthat his goal is to end welfare as we know it. The presentthoughtexperimenthas led us to a differentgoal: to end gender gender as we know it. Only by embracingthe aim of deconstructing can we mitigatepotentialconflicts among our five componentprnciples of that genderequity,therebyminimizingthe necessity of trade-offs.ReJecting aim, in contrast, makes such conflicts, and hence trade-offs, more likely. Achieving gender equity in a postindustrialwelfare state, then, requires deconstructing gender. A thoughtexpenment, I noted at the outset, is not a policy analysis. But it can nevertheless have political implications. By clarifying that gender gender,the reasoningheresuggests a strategy equity requiresdeconstructing means of radicalreform.This buildingmovementswhose demandsfor equity cannot be satisfied within the presentgenderorder.It means organizingfor of society."40 reformsthat"advancetowarda radicaltransformation a a is Crucialto such strategy thlrd-deconstructive-vislon of a postlnthen What dustral welfarestate. mightsuch a welfarestatelook like? Unlike caregiver parity, its employment sector would not be divided into two differenttracks;all jobs would assume workerswho are caregivers,too; all would have a shorterworkweek thanfull-timejobs have now; and all would have employment-enablingservices. Unlike universal breadwinner,however,employees would notbe assumedto shift all careworkto social services. Some informalcare work would be publicly supportedand integratedon a par with paid work in a single social-insurancesystem. Some would be

Fraser/ GENDER EQUITYAND THE WELFARESTATE

613

performedin householdsby relativesandfriends,butsuch householdswould not necessarily be heterosexualnuclearfamilies. Othersupportedcare work would be located outside of households altogether-in civil society. In state-fundedbut locally organizedinstitutions,childless adults,olderpeople, and others without kin-basedresponsibilitieswouldjoin parentsand others in democratic,self-managedcare work activities. This approachwould not only deconstructthe opposition between breadwinningand caregiving; it would also deconstructthe associated opposition between bureaucratized public institutionalsettings and intimateprivatedomestic settings. Treating civil society as a site for care work offers a wide range of new possibilities for promotingequal participation In social life, now no longer restrictedto formalemployment. Much moreworkneedsto be done to developthis third-deconstructivevision of a postindustnal welfare state. A key is to develop policies that discouragefree riding.Contraconservatives,the realfree nders in thecurrent system are not poor solo motherswho shirk employment.Instead,they are men of all classes who shirk care work and domestic labor,and especially corporationswho free ride on the labor of workingpeople, both underpaid and unpaid. A good statementof the deconstructivevision comes from the Swedish Ministryof Labor:"Tomakeit possible for bothmen andwomen to combine andgainfulemployment, a new view of the male role anda radical parenthood in the of life Thetrickis to imagine organization working arerequired."41 change a social world in which citizens' lives integratewage earning, caregiving, and involvementin the assoclacommunityactivism, political participation, tional life of civil society-while also leaving time for some fun. This world is not likely to come into being in the immediatefuture.But it is the only imaginablepostindustral world thatpromisestruegenderequity,and unless we areguidedby this vision now, we will neverget any closerto achievingIt.

NOTES
1. See Abramowitz(1988), Fraser(1987), Gordon(1988), and Land (1978). An exception is France,which from early on had high numbersof female workers(Jenson 1990). 2. This accountof the tnpartitestructure of the welfarestaterepresentsa modificationof my earlier(1987) view. Heretofore,I followed Nelson (1984, 1990) in positing a two-tierstructure of ideal-typically"masculine"social insuranceprogramsand ideal-typically"ferminne" family supportprograms. Althoughthatview was a relativelyaccuratepictureof the U.S. social-welfare system, I now considerit analyticallymisleading.The UnitedStatesis unusualin thatthe second and third tiers are conflated. The main programof means-testedpoor relief-Aid to Families

614

THEORY/ November 1994 POLITICAL

with Dependent Children(AFDC)-is also the main programthat supportswomen's child as two distincttiersof social welfare.Whensocial thesearebest understood raising.Analytically, insuranceis added,we get a three-tierwelfare state. 3. See Harvey(1989), Lash and Urry(1987), and Reich (1991). 4. Smith (1984). 5. Stacey (1987). 6. Weston(1991). 7. Ellwood (1988). 8. Fraser(1993). 9. Bartlettand Kennedy(1991). 10. Ellwood (1988). 11. Goodin (1988). 12. Not all dependenciesareexploitable.Goodin(1988, 175-6) specifies the following four conditions that must be met if a dependency is to be exploitable:(1) the relationshipmust be supplies; partymust need the resourcethatthe superordinate asymmetncal;(2) the subordinate for the supply of needed (3) the subordinatemust depend on some particularsuperordinate controlover the resourcesthatthe mustenjoy discretionary resources;and (4) the superordinate needs from him or her. subordinate 13. See Hirschman(1970), Okn (1989), and Hobson(1990). 14. Piven and Cloward(1971), Esplng-Andersen (1990). 15. Goodin (1988). 16. Sparer(1970). 17. See Orloff (1993). The antiexploitation objective should not be confused with current U.S. attackson welfare dependency,which are uhghly ideological. These attacksdefine dependency exclusively as receiptof public assistance.They ignore the ways in which such receipt can promoteclaimants'independenceby preventingexploitabledependenceon husbandsand employers(Fraserand Gordon1994). 18. Lister(1990), Sen (1990). 19. Weitzman(1985). 20. Ellwood (1988, 45). 21. Hochschild(1989), Schor (1991). 22. Bradshawand Holmes (1989, as cited by Lister, 1990). 23. Balbo (1987). but I do not have space to consider it 24. A fourthset of issues is also extremelyimportant, here. It concerns the bases of entitlementto provision.Every welfare state assigns its benefits pnnciples,which defines its basic moralquality.That accordingto a specific mix of distributive of threebasic mix, m each case, needs to be scrutimzed.Usually it containsvaryingproportions is the most Need-based and of entitlement: desert, need, provision citizenship. pnnclples but it nsks isolatingandstigmatizlngthe needy;it has been the basis of traditional redistributive, poor relief and of modem public assistance, the least honorableforms of provision.The most honorable, in contrast, is entitlementbased on desert, but it tends to be antiegalitaran and usuallytax payments, exclusionary.Hereone receives benefitsaccordingto one's contributions, work, and service-where tax paymentsmeans wage deductionspaid into a special fund,work of means prmary labor force employment,and service means the military,all interpretations women. Deserthas beenthepnmarybasis of earnngs-linkedsocial those termsthatdisadvantage welfarestate.(Actually,thereis a heavy ideologicalcomponentin the insurancein the industrial Benefit is desert-based. usualview thatpublicassistanceis need-based,whereassocial insurance Moreover,all governmentprolevels in social insurancedo not strictly reflect contributions. grams are financed by contributions,in the form of taxation.Public assistance programsare

Fraser/ GENDEREQUITYAND THE WELFARESTATE

615

financedfromgeneralrevenues,bothfederalandstate.Welfarerecipients,like others,contribute to these funds,forexample, throughpaymentof sales taxes [Fraser andGordon1992].) The third prnclple, citizenship,allocatesprovisionon the basis of membershipin society. It is honorable, egalitaran, andunlversalist,butalso expensive andhence hardto sustainat high levels of quality and generosity;some theonsts worry,too, thatit encouragesfree nding. (The free-rder worry, incidentally,is typically defined androcentncallyas a worry about shirkingpaid employment. Little attentionis paid, in contrast,to a far more widespreadproblem,namely,men's free nding on women's unpaiddomestic labor.A welcome exceptionis a recentunpublished paperby Peter entitlementsare typically found in social-democratic Taylor-Gooby[1993].) Citizenship-based countres, where they may include single-payerumversalhealth insurancesystems, unversal family or child allowances, and umversalflat-rateold-age pensions;they are virtuallyunknown in the UnitedStates-except forpubliceducation.Inexamlinngmodelsofpostindustral welfare states then, one must look closely at the constructionof entitlement. It makes considerable difference to women's and children's well-being, for example, whether day care places are distributedas citizenshipentitlementsor as desert-basedentitlements(i.e., whetheror not they to takeanotherexample, whethercare are conditionalon prioremployment).It likewise matters, work is supportedon the basis of need, in the form of a means-testedbenefit for the poor, or whether it is supportedon the basis of desert, as returnfor work or service, now interpreted nonandrocentncally,or whether, finally, it is supportedon the basis of citizenship under a umnversal basic income scheme. 25. On what basis would these benefits be distributed?In theory, employment-enabling services could be distributedaccording to need, desert, or citizenship, but citizenship accords best with the spirit of the model. Means-testedday care targetedfor the poor cannot help but status, and desert-basedday care sets up a signify a failure to achieve genuine breadwinner catch-22: one must already be employed in order to get what is needed for employment. Citizenship-basedentitlementis best then, but it must make services availableto all. This rules out Swedish-type arrangements,which fail to guaranteesufficient day care places and are plagued by long queues (Hobson 1993). 26. That,incidentally,would be to breakdecisively with U.S. policy,whichtypicallyassumes thatjob creationis for men; Bill Clinton'smuch-touted industnaland infrastructural investment policies are no exception in this regard(Fraser1993). 27. This could be done in several differentways. Governmentcould itself provideday care, and so on, m the formof publicgoods, or it could fundmarketized provisionthrougha system of vouchers.Alternatively, to provideemployment-enabling services employerscould be mandated for theiremployees, either throughvouchersor in-house arrangements. The state option means nevertheless.Mandating highertaxes, of course,butit maybe preferable employerresponsibility creates a disincentiveto hire workerswith dependents,to the likely disadvantageof women. 28. Glenn (1992). 29. It too conditionsentitlementon desertanddefines contribution in traditional androcentnc terms as employmentand wage deductions. 30. Exactly whatelse mustbe providedinside the residualsystem will dependon the balance of entitlementsoutside of it. If health insuranceis providedumversallyas a citizen benefit, for example, then there need be no means-testedhealth system for the nonemployed.If, however, mainstream healthinsuranceis linked to employment,thena residualhealthcare system will be necessary.The same holds for unemployment, retirement,and disabilityinsurance.In general, the more that is providedon the basis of citizenship,instead of on the basis of desert, the less has to be providedon the basis of need. One couldeven say thatdesert-basedentitlementscreate the necessity of need-basedprovision;thus social insurancecreates the need for means-tested public assistance.

616

POLITICAL THEORY / November 1994

31. Kilborn(1993). 32. Exactly how much dependson the government'ssuccess in eliminatingdiscrmination worth. and in implementingcomparable 33. Universal breadwinner apparentlyrelies on persuasionto induce men to do their fair shareof unpaidwork. The chances of thatworkingwould be improvedif the model succeeded in promotingcultural changeandin enhancingwomen'svoice withinmarnage.But it is doubtful thatthis would suffice. 34. Littleton(1991). 35. On whatpnnciple(s) would these benefitsbe distributed? Caregiverallowancescould in on the basis of need, as a means-testedbenefit for the poor-as they have theorybe distributed always been in the United States. But thatwould contravenethe spiritof caregiverparity.One cannotconsistentlyclaim thatthe caregiverlife is equivalentin dignityto the breadwinner life, has always while supportingit only as a last-resort stopgapagainstpoverty.(This contradiction bedeviled mothers' pensions-and later Aid to Dependent Children-in the United States. Although these programswere intendedby some advocates to exalt motherhood,they sent a contradictorymessage by virtue of being means tested and morals tested.) Means-tested allowances, moreover,would impede easy transitionsbetween employment and care work. Because the aim is to makecaregivingas deservingas breadwinnmng, caregiverallowancesmust be based on desert.Treatedas compensationfor socially necessary service or work, they alter of those terms. androcentncmearnngs the standard 36. SusanOkin (1989) has proposedan alternative way to fundcarework. In her scheme the funds would come from what are now consideredto be the earningsof the caregiver'spartner. A man with a nonemployedwife, for example, would receive a paycheck for one-half of his salary;his employerwould cut a second check in the same amountpayabledirectlyto the wife. Intriguingas this idea is, one may wonderwhetherit is really the best way to promotewives' independencefrom husbands,because it ties her income so directlyto his. In addition,Okin's proposal does not provide any care work support for women without employed partners. Caregiverparity,in contrast,providespublic supportfor all who performinformalcare work. Who then are its beneficiares likely to be? With the exception of pregnancyleave, all of the model's benefits are open to everyone;so men as well as women can opt for a "fermnne"life. Women,however,areconsiderablymorelikely to do so. Althoughthe model aims to makesuch a life costless, it includes no positive incentivesfor men to change. Some men, of course, may simply prefer such a life and will choose it when offered the chance; most will not, however, given currentsocialization and culture.We will see, moreover,that caregiverparitycontans some hiddendisincentivesto male caregiving. essential additional model:whatever breadwinner 37. In thisrespect,it resemblesthe umversal goods are normallyoffered on the basis of desertmustbe offered here too on the basis of need. 38. Okin (1989). 39. J. Williams(1991). 40. Gorz (1967, 6). 41. As quotedin Lister(1990, 463).

REFERENCES
Social Welfare Policyfrom Colonial Abramowitz,Miml. 1988. Regulatingthe Lives of Women: limes to the Present. Boston:South End.

Fraser/ GENDEREQUITYAND THE WELFARESTATE

617

and the State, editedby Ann ShowstackSassoon. Balbo, Laura.1987. "CrazyQuilt."In Women London:Hutchinson. Bartlett,KathanneT, and Rosanne Kennedy,eds. 1991. FeministLegal Theory:Readings in Law and Gender Boulder,CO: Westview. Ellwood, David T. 1988. Poor Support: Poverty in the American Family. New York:Basic Books. Gosta. 1990. The ThreeWorlds of Welfare Capitalism.Pnnceton,NJ: PnnceEsplng-Andersen, ton UmversityPress. Fraser,Nancy. 1987 "Women,Welfare,and the Politics of Need Interpretation." Hypatia: A Journal of Feminist Philosophy 2, no. 1: 103-21. (Repnnted in Fraser,UnrulyPractices: Social Theory.Minneapolis:Umversityof Power,Discourse, and GenderIn Contemporary MinnesotaPress, 1989.) 1993. "Clintomsm,Welfare,and the Antisocial Wage:The Emergenceof a Neoliberal Marxism6, no. 1: 9-23. Political Imaginary." Rethinking Fraser,Nancy, and Linda Gordon. 1992. "Contractversus Charity:Why Is There No Social Citizenshipin the United States?"Socialist Review 22, no. 3: 45-68. 1994. "A Genealogy of 'Dependency' Tracinga Keywordof the U.S. WelfareState." in Cultureand Society 19, no. 2: 309-36. Signs: Journal of Women Glenn, Evelyn Nakano. 1992. "FromServitudeto Service Work:Historcal Continuitiesin the Racial Division of Paid ReproductiveLabor."Signs: Journal of Womenin Cultureand Society 18, no. 1: 1-43. ThePolitical Theoryof the Welfare State.Pnnceton, Goodin, Robert.1988. Reasonsfor Welfare: NJ: Pnnceton UniversityPress. Gordon, Linda. 1988. "What Does Welfare Regulate?" Social Research 55, no. 4: 609-30. Gorz, Andre. 1967. Strategy for Labor:A Radical Proposal. Translated by MartinA. Nicolaus and Victona Ortiz. Boston: Beacon. Harvey,David. 1989. The Conditionof Postmoderity: An Inquiryinto the Ongins of Cultural Change. Oxford:Basil Blackwell. Hirschman,AlbertO. 1970. Exit, Voice,and Loyalty:Responsesto Decline in Firms,Organizations, and States. Cambndge, MA. HarvardUniversityPress. Hobson, Barbara. 1990. "No Exit, No Voice:Women'sEconomic Dependencyand the Welfare State."Acta Sociologica 33, no. 3: 235-50. 1993. "Economic Dependency and Women's Social Citizenship:Some Thoughtson Esplng-Andersen'sWelfareState Regimes."Unpublishedtypescnpt. Parentsand the Revolutionat Home. New Hochschild, Arlie. 1989. TheSecond Shift. Working York:Viking. of Gender: Policies to 'Protect' Women Workersand Jenson, Jane. 1990. "Representations Infants in France and the United States before 1914." In Women,the State, and Welfare, edited by LindaGordon, 152-77 Madison:Universityof WisconsinPress. Kilbom, Peter. 1993. March 15. "New Jobs Lack the Old Security in Time of 'Disposable Workers.'" The New York Times,pp. Al, A6. Land, Hilary. 1978. "WhoCares for the Family9"Journal of Social Policy 7, no. 3: 257-84. Lash, Scott, and John Urry. 1987 The End of OrganizedCapitalism.Cambndge:Polity. Journalof Social Policy Lister,Ruth. 1990. "Women,EconomicDependency,andCitizenship." 19, no. 4: 445-67. Littleton, Chrstine A. 1991. "ReconstructingSexual Equality."In Feminist Legal Theory: Readings in Law and Gender,edited by KathanneT. Bartlettand RosanneKennedy,35-56. Boulder,CO: Westview.

618

THEORY / November 1994 POLITICAL

Nelson, Barbara.1984. "Women'sPoverty and Women's Citizenship:Some Political Consein Cultureand Society 10, no. Signs: Journalof Women quences of EconomicMarginality." 2: 209-31. WelfareState:Workmen's and 1990. "TheOrgins of the Two-Channel Compensation theState,and Welfare, editedby LindaGordon,123-51. Madison: Mothers'Aid."In Women, Umversityof WisconsinPress. Susan Moller. 1989. Justice, Gender,and the Family. New York:Basic Books. Okmn, Orloff, Ann Shola. 1993. "Genderand the Social Rights of Citizenship:The Comparative Analysis of GenderRelationsandWelfareStates."AmericanSocilogical Review58, no. 3: 303-28. Piven, FrancesFox, and RichardA. Cloward. 1971. Regulatingthe Poor New York:Random House. Reich, Robert. 1991. The Work of Nations: PreparingOurselvesfor 21st CenturyCapitalism. New York:AlfredA. Knopf. Decline of Leisure.New York: The Unexpected Schor,Juliet. 1991. The OverworkedAmerican: Basic Books. Review of Sen, Amartya.1990. "MoreThan 100 Million WomenAre Missing."TheNew York Books 37, no. 20: 61-6. Smith, Joan. 1984. "The Paradoxof Women'sPoverty:Wage-EarnngWomenand Economic in Cultureand Society 9, no. 2: 291-310. Transformation." Signs: Journalof Women Sparer,Edward V. 1970. "The Right to Welfare."In The Rights of Americans: What They Are-What TheyShouldBe, edited by NormanDorsen,65-93. New York:Pantheon. Stacey, Judith.1987 "Sexism By a SubtlerName? PostindustnalConditionsand Postfemimst Consciousnessin the Silicon Valley."Socialist Review96: 7-28. Peter. 1993. "Scrounging,Moral Hazard,and Unwaged Work:Citizenshipand Taylor-Gooby, HumanNeed." Unpublishedtypescrpt. Weitzman,Lenore. 1985. The Divorce Revolution:The UnexpectedSocial Consequencesfor Women and Childrenin America.New York:Free Press. Weston, Kath. 1991. Families We Choose: Lesbians, Gays, Kinship. New York:Columbia UmversityPress. Gender."In FeministLegal Theory:Readings in Law Williams, Joan. 1991. "Deconstructing and Gender,edited by KatharneT. Bartlettand Rosanne Kennedy,95-123. Boulder,CO: Westview.

Nancy Fraser is Professor of Philosophyand Faculty Fellow of the Centerfor Urban whereshe is also an affiliateof Affairsand Policy Researchat Northwestern University, StudiesProgram.She is the author of UnrulyPractices:Power,Discourse the Women's Social Theory(Universityof MinnesotaPress and Polity and Genderin Contemporary Press, 1989), the co-author of Feminist Contentions: A Philosophical Exchange (Routledge,1994), and the co-editorof RevaluingFrenchFemnusm:CriticalEssays on Difference,Agency, and Culture(IndianaUniversityPress, 1992).