Gender and Social Rights of Citizenship PDF

Gender and the Social Rights of Citizenship: The Comparative Analysis of Gender Relations and Welfare States Author(s): Ann Shola Orloff Reviewed work(s): Source: American Sociological Review, Vol. 58, No. 3 (Jun., 1993), pp. 303-328 Published by: American Sociological Association Stable URL: http:/...

Gender and and Welfare States Author(s): Ann Shola Orloff Reviewed work(s): Source: American Sociological Review, Published by: American Sociological Stable URL: http://www.jstor.org/stable/20959 Accessed: 04/07/2012 07:39 Your use of the JSTOR archive indicates http://www.jstor.org/page/info/about/policies JSTOR is a not-for-profit service that content in a trusted digital archive. of scholarship. For more information American Sociological American Sociological Review. http://www.jstor.org GENDER AND TIE SOCIAL RIGHTS THE COMPARATIVEANALYSIS OF GENDER RELATIONSAND WELFARE State social provision affects structures political conflict bilization of identities and interests. gender, while most feminist research comparative. I develop a conceptual cial provision that draws on feminist tive variation suggested by power gender: (1) the state-market relations tries organize the provision of markets; it is then termed the tion dimension is expanded to gender relations, especially the zenship rights/decommodification the sexual division of caring on men and women of benefits that proposed to capture the effects paidworkand capacityto formandmaintainan No one who has listened to debates about the welfare state in the United States or in other advanced capitalist and democratic countries- about "welfaremothers"or child- care support- could doubt the importanceof *Direct all correspondenceto Ann Shola Departmentof Sociology, University of 1180 ObservatoryDrive, Madison,WI 53706. researchwas partiallysupportedby a grantfrom National Science Foundation(SES 8822352) being carried out jointly with a project Gender Regimes of Liberal Welfare States" Dr. Sheila Shaver and Dr. Julia O'Connor). versions of this paper were presentedat workshops,and seminarsin the United den, Germany, and Australia. For their on draftsof this paper,I thank:JuliaAdams, Baxter, Jane Collins, Bob Connell, Linda Gordon, Alex Hicks, BarbaraHobson, David James, Jane Jenson, TrudieKnijn, Walter Korpi, Marilyn Lake, Christiane Lemke, Leslie McCall, Eileen McDonagh, Margit Meyer, Pavla Miller, Deborah Mitchell, Renee Monson, Julia O'Connor, Joakim Palme, Wendy Sarvasy, Sheila Shaver, Birte Siim, Theda Skocpol, BarbaraSullivan, Pamela Walters, Dorothy Watson, and three ASR reviewers. Leslie McCall and HeatherHartley provided researchas- sistance. I Using the term "welfarestate"to AmericanSociological 304 lationships, structures political conflict and participation,and contributesto the formation and mobilization of specific identities and in- terests. Many would also argue- and I would agree - that, as a result of varied political struggles, the state may also offer important political resourcesto women and to other sub- ordinate groups. Moreover, variation across countriesand over time in the characterand ef- fects of social provisionon genderrelationshas been considerableand significant. Recent feminist work on social provision is concerned with the extent to which welfare states have the potential to be or already are "woman-friendly"(Hernes 1987), or, to pose the proposition somewhat differently,the ex- tent to which - and how - they reproduce male dominance.2 Some analysts have lighted the role of women's political securing social rights based simply ship. The recognition of the genderedcharac- ters of the welfare state and social of the agency of women, are importantcor- rectives to the "mainstream"literatureon welfare state, which is all too often gender- blind in its conceptions of class, citizenship and the economy.3 Yet if little of the ployment, industrialaccident,retirement,disability,social ill health,deathof a family breadwinner,or extreme poverty; some analysts also include provision of education and housing (Flora and Heidenheimer 1981). Despite these difficulties, I maintain continuity with social-scientific torical studies of the causes and effects social provision. I define the welfare social provision, as interventions by civil society to alter social and marketforces. ever, I do not judge a priori that all terventions are aimed at, or actually produce, greaterequality among citizens; I discuss furtherbelow. 2 I preferto use the term "male-dominant," than "patriarchal,"to describe Western states in the late nineteenth centuries. Historically, "patriarchy"refers form of male dominance in which fathers control families and families are the units nomic power," not to "a universal, terministic social structurewhich denies women" (Gordon 1988b, p. vi; see also Cockburn 1990). 3 Using terms like "mainstream"and to describebodies of researchthatcontainstrikingly divergent conceptual frameworksis an fication. My terminologyconveys an importantdif- ference in the two literatures:Mainstreamwork GENDERAND SOCIALRIGHTS cial policies reflect and reinforce relations of dominance and exploitation, thus arguing for understandingthe welfare state as functional for patriarchy and capitalism (Wilson 1977; McIntosh 1978; Abromovitz 1988).6 They were concerned with the qualitativeeffects of modern social provision, but assumed invari- ance in the regulatory function of welfare states. Meanwhile, most comparativestudies of the welfare state have focused on expendituredata unsuitedto examiningpower relations,making it easier for feminist researchersto dismiss this work as irrelevant to their concerns. Indeed, mainstream scholars simply assume that the welfare state is a mechanismfor making soci- ety moreegalitarian;they routinelyreferto sys- tems of social provision in Westerncapitalist democraciesas "welfarestates,"taking at face value the claims of stateelites aboutthe charac- ter of social programs. The extent to which states actually promotecitizens' well-being or equality beyond income security is rarely in- vestigated (Esping-Andersen 1990, chap. 1; Cates 1983, chap. 1). Mainstreamscholarsmay have argued about the extent of equality pro- moted by social programs,but saw variationin linear terms - a state's "welfareef sured by social expendituresas a proportionof GNP, resulted in more or less equality,usually conceptualizedin class or income terms. Until recently,then, the predominanttheoreticaland methodological approachesw-ithinthe two lit- eraturestended to neglect qualitativevariation in the effects of state social provisionover time and across nations. Recent methodological and theoretical changes within the two groups may increase the chances of a fruitfulconversation.Feminist research on the welfare state is taking a com- parative turn, focusing on the variationin the gender content of systems of social provision (Ruggie 1984; Jenson 1986, 1991; Michel and Koven 1990; Lewis 1992; Skocpol and Ritter 1991). Meanwhile, several mainstreamcom- parative welfare state researchers, especially those associated with the "power resources" school of analysis, are focusing on variationin 6 These feminist functionalist approaches are similar to radical or Marxist functionalist proaches,which never dominatedmainstreamcom- parative scholarship. However, the Marxists ig- nored genderjust as mainstreamscholars did (Offe 1984; O'Connor 1973; Piven and Cloward 1971). 306 work for evaluating the content of social pro- vision, the "dependentvariable"if you will.7 The concern of the power resourcesanalysts with qualitative ratherthan quantitativechar- acteristics of the welfare state is useful for in- vestigating power relations, of paramountim- portance in understandingthe relationshipbe- tween state social provision and gender rela- tions. Over the last decade, scholars at the Swedish Institute for Social Research, under the direction of Korpi and Esping-Andersen, have assembled an impressive data set on the quality of social rights and how these rightsaf- fect differentcitizens in eighteen OECD coun- tries for the period 1930 through 1985 (Korpi 1989; Esping-Andersen1990). These countries include most of the rich capitalistcountriesthat have been democraticsince WorldWarII. This research group has formulated a systematic scheme for comparativeanalysis of state sys- tems of social provision that focuses on three key dimensions: (1) state-marketrelations, (2) stratification,and (3) social citizenship rights, including, in Esping-Andersen's (1990) scheme, how this affects the "decommodifica- tion" of labor. The power resources analysts' framework provides a good startingpoint for analyzingthe gender content of state social provision. But serious conceptual work must be done before applyingit to the interrelationbetweenstate so- cial provision and gender relations. I do not mean that researchshould simply look at what the welfare state does for or to women, al- though that is clearly part of the task. Rather, gender must incorporatedinto the core con- cepts of researchon the welfare state - "citi- zen," "social rights," and the analytic dimensions used to evaluate inputs, content, and effects. "Gendering"the analytic framework means two things. First, because power resourcesanalysisdoes not con- sider genderrelations,its conceptualapparatus must be reworkedto incorporategender. Sec- ond, the key issue for a feminist analysisof the welfare state - the effects of state social pro- 7Thus, to the question, "Why own perspective, state-centered or institutionalist analysis?,"I would say thatfew dealt with its conceptualization of welfare states. Rather, their attention on the role of characteristicsof the developments. GENDERAND SOCIALRIGHTS power resources explains variationin the out- comes of political struggles for social rights (Korpi 1989). Capitalists have greater re- sources in the market,while workers (because of their numbers)have greaterresourcesin the polity. Wage earners,they argue, will use their political resources to modify marketprocesses and extend social rights. Conversely, capital- ists will fight to let market-basedprocesses de- termine welfare outcomes and to limit social rights. This is the theoretical context for the claim that workers' political struggles can se- cure social rights that "pushback the frontiers of capitalistpower"(Esping-Andersen1990, p. 16). This is accomplished by empowering workers vis-a-vis the market- providing so- cial benefits that "weakenthe whip of the mar- ket" and promote working-class political soli- darity (Palme 1990, p. 8). Thus, the analytic scheme for describing the content of the wel- fare state is related to their theory of working- class interests and the dynamics of social policy development - the scheme was devel- oped to answer the question of how states af- fect class relations. What of gender in the power resources framework?To put it bluntly, it is simply absent. Its concepts are explicitly gender-neutral- but the categories of work- ers, state-marketrelations, stratification,citi- zenship, and decommodificationare based on a male standard;moreover, gender relations and their effects are ignored. Feminists are interested in gendering the questions and categories of the power re- sources analysts. However, feminists are also asking a different question: Can the welfare state alter gender relations?This question sug- gests a differentresearchagendaand new con- ceptual categories. Feminist analysts are less unified aboutthe factorsthatunderliewomen's oppression than are mainstreamwelfare state analystson the factorsunderlyingclass oppres- sion.9 This lack of agreement makes it more difficult to specify just what the state would have to do to push back the frontiersof male dominance.10But despite importanttheoretical 9 The problem of specifying gender women's interests has troubled enced literature in political sociology specifying class interests has troubled political sociologists) (Jonasdottir1988; 1985). 10One body of feminist theory highlights portance of men's control of women's 308 CITIZENSHIPAND SOCIALPOLICYOR "POLITICSAGAINSTMARKETS"13 AND MALE DOMINANCE? WhoIs a Citizen? The political strugglesof citizens are criticalto the power resourcesanalysts' understandingof policy developments. But just as the indepen- dent male householderserves as the ideal-typi- cal citizen in classical liberal and democratic theory, the male worker serves as the ideal- typical citizen in the literatureon social rights (in Korpi 1989 or Esping-Andersen 1990 as much as in Marshall 1950). As Hernes notes, "The social-democratic citizen is the citizen worker, a male family provider, a working- class hero. His rights, identities and participa- tion patternswere determinedby his ties to the labourmarket,and by the web of associations and corporatestructureswhich had grown up aroundthese ties" (1988, p. 190, emphases in original). Indeed, "in the 'democra state... employment ratherthan military ser- vice is the key to citizenship"because it be- stows the independencethatis the "centralcri- terion for citizenship" and is associated with men (Pateman 1988a, pp. 238-39). Power resources analysis begins with eco- nomically independent citizens (i.e., wage earners) and considers the cross-national,his- torical, and class variationsin the ways social rights affect them. Analysts focus on those as- pects of state social provisionthatare most rel- evant for male wage earnersand breadwinners, that is, programsthat compensate workersfor losses incurredin the paid labor market,such as old-age pensions and unemploymentinsur- ance. They then examine the basis on which people make claims for state help -need, fi- nancial contributionor citizenship and the associated variationin outcomes. The use of a male standardis not explicit, and power re- sources analysis - like most mainstreamcom- parative researchon the welfare state - uses gender-neutrallanguageand categories.Myles (1989, pp. 135-36) acknowledgedthis problem its own difficulties as currentdata state usually address different issues. I do not ex- pect a one-to-one relation between specific dimen- sions and a particularprogramof the welfare state. Rather,these dimensions describe propertiesof en- tire systems of social provision that vary. 13 I here borrow the title of Esping-Andersen's (1985) book. GENDERAND SOCIALRIGHTS reference to gender inequalities in his discus- sion of social rights (1990, pp. 21-23). Feminist analyses of citizenship highlighted sexuality, reproduction,and physical bodies: "Citizenshipis defined as a practiceof embod- ied subjects whose sex/gendered identity af- fects fundamentallytheirmembershipandpar- ticipation in public life" (Jones 1990, p. 786). Women face gender-specific threats to their bodily integrity both inside and outside the family.As Shaver (1990) argued,"rightsin the control of one's body and sexual person, as in marriage, consent to sexual activity, and the control of fertility and reproduction"have a taken-for-grantedcharacter for men, but are contested issues for women. Feminist theory points to the subjugationof women in the pri- vate sphere of the family, which accordingto liberal theory ought to be (and in practiceusu- ally has been) free from state interference (Pateman 1988a, 1988b, 1989; Eisenstein 1981).15 For example, because of the inviolate nature of "family privacy" in Western coun- tries, husbandshave been allowed to rape and batter their wives. Only recently have there been some tentative legal reforms limiting these "rights"and endowing women with the right to be free from such attacks(Breines and Gordon 1983; Yllo and Bograd 1988; Pateman 1988a, pp. 238-39; MacKinnon1989, chap. 9; Smart 1989, p. 32; Russell 1982; Hanmer, Radford,and Stanko 1989). The "femalefear" of rape also curtails women's access to public spaces, sexual harassmentis an importantcom- ponent of men's power in the workplace, and the state also threatensreproductivefreedoms (Connell 1987; MacKinnon 1989; Cockburn 1991; Petchesky 1984). In sum, relations of dominationbased on control of women's ies in the family, the workplace, and public spaces underminewomen's abilities to partici- 15 Indeed, gendered analyses point out the illib- eral aspects of capitalist and democratic societies which are usually also considered liberal. (Some analyses investigating race and ethnicity, critical race theory, make similarobservation e.g., Crenshaw 1989; Williams 1991; Matsuda 1989]). "Feminists have pointed to the contradic- tion between the rules of the public sphere built on consent and voluntaryassociations, and the rules of the private sphere built on oppression and natural subjugation,and they have arguedthat this division prevents women from realizing a full democratic citizenship" (Siim 1988, p. 163). 310 between the state and the marketin providing welfare, the effects of the welfare state on so- cial stratification, and the characterof social rights (which in Esping-Andersen's scheme includes how these affect the decom- modification of labor). Clustering of systems along these three dimensions defines regime types. Esping-Andersenand Korpi(1987) built on the work of analysts, such as Titmuss (1958), who distinguished between "residual" and "institutional" welfare states (see also Baldwin 1990). "Residual"welfare states only react to market or family "failures"and limit assistance to marginal or especially "deserv- ing" social groups; "institutional" states are pro-active and are committed to the welfare needs of all strata of the population. Esping-Andersen (1990) constructed a typol- ogy of regimes representing"three worlds of welfare capitalism" - liberal, conservative- corporatistand social-democratic- by char- acterizingsystems of social provisionalong the threedimensions. Liberalregimes roughlycor- respond to the "residual"states, while social- democratic and conservative-corporatist re- gimes may be distinguished within the group of "institutional"states. Social-democraticre- gimes are universalistic and egalitarian,while the conservative-corporatistregimes preserve status and class differentials.Despite the fact that "there is no single pure case," Andersenclassified the United States, Canada, Australia, and (probably)GreatBritain as lib- eral regimes; the Nordic countries are identi- fied as social-democraticregimes; andAustria, France, Germany, Italy, and the Netherlands are conservative-corporatistregimes. The State-MarketRelations Dimension A fundamental dimension that varies across welfare states concerns the "range,or domain, of human needs that are satisfied by social policy" instead of by the market (Esping- Andersenand Korpi 1987, p. 41), thatis, "how state activities are interlocked with the market'sand the family's role in social provi- sion" (Esping-Andersen 1990, p. 21). Korpi (1989) contendedthat as public provisions are put in place, this "decreasesthe scope of mar- kets and changes the basis of distributionin these areas from marketpower to political re- sources. In the areas involved in the equal sta- tus of citizenship, the criteria for distribution GENDERAND SOCIALRIGHTS strata,such as distinctprogramsfor blue-collar and white-collar workers.Liberal regimes en- courage social dualisms between the desperate minority,that relies on limited forms of social assistance, and the majority,that relies princi- pally on the market for welfare (e.g., private pensions and health plans). These regimes do not greatlymodify market-generatedstratifica- tion or social mobility - any reductionsof so- cial inequalitiesoccur over the life-span rather than across classes. The three types of regimes (social-demo- cratic, conservative-corporatist, and liberal) also institutionalize distinctive patterns of policy intereststhat help to shape the political alliances and enmities that affect subsequent political struggles over policy (Esping-Ander- sen 1985;Baldwin 1990). Powerresourcesana- lysts have focused almost exclusively on the ways in which policies affect class Others highlighted the ways in which specific features of social provisions "feed back" into politics by encouragingcertainalliances while discouragingothers, defining the terms of de- bate and developing certain state capacities (Weir,Orloff, and Skocpol 1988; Stryker1990; Heclo 1974; Jenson 1986, 1991). The Social CitizenshipRights! DecommodificationDimension The third dimension of the welfare state con- cerns the characterof the social rights of citi- zenship. Some benefits are universal, that is, they are available to all citizens or to all citi- zens of a certain age or condition (e.g., sick- ness, unemployment,parenthood);some ben- efits dependon labor marketparticipationand financial contribution;and some benefits are income-tested,thatis, they areavailableonly to those with incomes and assets below a certain level. These distinctionsregulateaccess to ben- efits, andalong with benefitlevels andthe range of entitlements,they determinethe "degreeto which the individual's typical life situation is freed from dependence on the labor market" (Esping-Andersenand Korpi 1987, p. 40). Esping-Andersen(1990) linked social to what he termed the decommodificationof labor: It is as marketsbecomeuniversalandhegemon thatthe welfareof individualscomesto depend entirelyonthecashnexus.Stripping societyof the institutionallayersthatguaranteedsocialrepro- 312 lytic coherence to diverse comparativestudies and have revealed distinctive clusters of coun- tries based on theirsystems of social provision. Yet, the dimensions are clearly flawed because of their inattentionto gender. Genderingthe State-Market-FamilyRelations Dimension Power resourcesanalysts generallyhave given more attention to the "division of labor" be- tween states and marketsin providingwelfare than to relations among states, markets, and families. Indeed, the distinction between pub- lic and privateis seen as a distinctionbetween politics and the market- families are ignored as "private"providers of welfare goods and services (Esping-Andersenand Korpi 1987, p.. 41). Provision of welfare "counts"only when it occurs throughthe state or the market,while women's unpaid work in the home is ignored. Furthermore, the sexual division of labor within states, markets, and families also goes unnoticed. This dimension should be recon- structedbased on the recognitionof the impor- tance of families and women's unpaidwork to the provision of social welfare, in addition to considering gendered patternsof work. State provision that helps to shift the burdenof wel- fare from the family to the state, or from women to men within the family furthers women's gender interests. Recently,Esping-Andersen(1990) has noted some importantdifferences across regimes in relations between states and families, that is, "how state activities are interlocked with the market'sand the family's role in social provi- sion" (p. 21). He focused on services that re- spond to "family needs... [and] also allow women to choose work ratherthan the house- hold"(p. 28). Conservative-corporatist respectthe principleof subsidiarity- the "state will only interferewhen the family's capacity to service its members is exhausted"(Esping- Andersen 1990, p. 27), but will not provideser- vices that enable mothers (or other primary caretakers)to enter the paid labor force. Thus, these regimes reinforcetraditionalfamily rela- tions. In liberal regimes, "concernsof gender matter less than the sanctity of the market" (Esping-Andersen1990, p. 28), whereassocial- democratic regimes attempt to "preemptively socialize the costs of familyhood,"for example, by assuming partialresponsibility for care of GENDERAND SOCIALRIGHTS women are heavily concentratedin part-time employment. Of the three countries, Sweden has the highest level of sex segregationin oc- cupations,and althoughGermanyhas a some- what lower level of segregation, many fewer women work outside the home in the first place. In Sweden, occupational upgrading among women was accompaniedby continued segregation, while sex desegregationof occu- pations is strongest in the United States and weakest in Germany (Esping-Andersen1990, p. 210). Thus, while the conservative-corporat- ist regime of Germany would be expected to preserve traditionaleconomic dependence for women, the decline in the sex segregation of occupationsin the liberalregime of the United States is unexpected. Nor would progressive Sweden be expected to have high levels of sex segregation of occupations,part-timeemploy- ment and women doing the bulk of unpaiddo- mestic work. (Working wives in Sweden do about72 percentof houseworkcomparedto 74 percent in the United States [Wright, Shire, Hwang, Dolan, and Baxter 1992, p. 262; Ruggie 1988; Lewis 1992; Baxter 1993]). These analytic inadequacies are related to some of Esping-Andersen's premises, which neglect gender relations and feminist scholar- ship. He sees women as choosing between "workand the household,"with work possible for most women only if state services are widely available. Yet women in Scandinavia and elsewhere do not choose between paid work and unpaidhousewifery(includingmoth- ering) as exclusive activities (Hobson 1991a) - they can choose to be stay-at-homewives and mothers only or combine paid work with their domestic work. Nowhere in the industri- alized West can marriedwomen and mothers choose not to engage in caring and domestic labor (unless they are wealthy enough to pur- chase the services of others). Land and Rose (1985, p. 93) call this situation "compulsory altruism"for women (Taylor-Gooby1991, p. 102). The core aspects of the sexual division of labor remain:Womenperformmost domes- tic work whether or not they work for pay, while men do very little domestic work. Understandinggenderrelations,particularly the sexual division of labor, helps explain women's employment patterns.Many women work part-timebecause this arrangement"al- lows" them to do their domestic work (Beechey and Perkins 1987). Thus, the dimen- 314 der power, such as sex segregationof occupa- tions, the division of householdwork, or men's control of women's bodies. Men as a gender have power - a power re- source, to use the terminology of Korpi and Esping-Andersen- flowing from theircontrol of women's domestic and caring labor and of their bodies. Of course, women are not power- less in families, any more than workers are powerless on the job because capitalistshave greatermarketresources.But the power imbal- ance between the gendersin families is impor- tant politically. Because of the power relations in families, shifting decision-makingaboutthe distribution of resources or the provision of services from families to polities is parallelto shifting decision-making from markets to states, for it is a shift from an arenain which resources are disproportionatelycontrolledby men to one in which power may be more equally distributedbetween men and women. Althoughthe (electoral)numbersare not as de- cisively favorable to women as they are to wage earners,polities' decision rules and guid- ing ideologies are more likely to equalize out- comes than to leave these decisions to indi- viduals within families.18The failureto recog- nize gender relations and power within the family and outside the family blinds the power resources analysts to aspects of social policy regimes that affect genderrelations. The conceptualizationof a "division of la- bor" among states and markets must also in- clude families as significant providersof wel- fare, and the unpaidcaring and domestic work of women must be explicitly recognized.19The state is woman-friendlyto the extent thatpoli- 18 Of course, one should not assume that all women will "vote feminist" anymore than one should assume socialist affiliations for workers. However, large numbersmay be a necessary condi- tion for the enactmentof welfare policies favorable to disadvantagedgroups. 19Researchersoutside the power resourceschool have highlighted the importanceof the "public-pri- vate split" (in a nonfeminist sense), vision of labor"between state and voluntaryor cor- porate organizations in providing welfare (Katz 1983, 1986; Flora and Alber 1981; Flora and Hei- denheimer 1981). Voluntary or private social wel- fare bodies have always been importantproviders of welfare services and assistance. Historically and currently,women have predominatedamong char- ity and volunteer workers (Michel and Koven 1990). GENDERAND SOCIALRIGHTS based or universal entitlements,men's claims are usually made because of loss of paid em- ployment. In contrast,most women's claims in most Westernwelfare statesarebasedonfamil- ial or marital roles (i.e., on the basis of unpaid domestic and caring work) although the pro- portion has been declining in the last few de- cades (Fraser1989; Gordon1990;Nelson 1984, 1990). (Of course, all such claims do not re- ceive equal treatment, as factors like race, ethnicity,or maritalstatusalso have effects.) In all systems of social provision,claims basedon motherhood or marriage to a covered wage earner,which often have more stringenteligi- bility requirements,are associated with lower benefit levels than are direct, work-based claims. A number of American scholars have mapped the differences in treatmentbetween caring/unpaidlaborand paid laboronto the du- alistic structureof the U.S. welfare state. They have identified a "two-tier"system in which social assistance programs serve a predomi- nantly female clientele, while contributoryso- cial insurancetargetsa predominantlymale cli- entele (Pearce 1978, 1983, 1986; Nelson 1984, 1990; Acker 1988; Fraser 1989).21 Stark in- equalities exist between the two types of pro- grams. Social assistance programs,on which many single mothers rely for income protec- tion (e.g., "welfare" or Aid to Families with DependentChildren)are politically less legiti- mate, less generously funded, and more ori- ented to monitoring clients' behavior and in- come than are social insuranceprograms(e.g., "social security,"or Old Age, Disability Insurance) on which most unem- ployed and retiredwage-earningmen rely.22 211I am not sure where the "two-tier&q originated, but countless authorsnow indeed exists. Nelson (1984, pp. 221-23) an early and often-cited discussion in which she recognizes the gender differences the clienteles of social assistance surance programs in the United States, the distinction between the types of claims made within these differenttiers. While Nelson is careful not to conflate the two, many of those who cite her work have been less careful. 22 American men who do not qualify for social insurance programs - disproportionately ethnic and racial minorities - must rely on social assis- tance programs(usually called general assistance), which are not available nationwide, receive no fed- eral funding, and are less generous than Aid to 316 This complicates the argumentthat two-tiered systems generate distinctive interests for all women - some women are tied to the welfare state as mothers, while others are tied to the stateas wives (althoughdivorce- or the threat of it - may weaken that bond for some women). This distinctioncan be significantpo- litically because it reinforces differences be- tween two-parent families and single-parent families, as has occurredin the United States (Weir,Orloff, and Skocpol 1988). Althoughthe "two-tiered"formulationhas sometimes been generalized to other nation's systems of social provision, this is problematic. For example, most of Australia'ssocial programsare means- tested, while the Nordic countriesrely on uni- versal programssupplementedby contributory insurance(Shaver 1983; Hernes 1987). Social insurance programs may not treat men's and women's work-based claims equally,either.Gainingeligibility for social in- surance programs is often more difficult for working women thanfor workingmen. For ex- ample, until recently,marriedwomen had to be unable to performhouseworkand paid work to claim work-related disability benefits under Britain's social insurance system (Pateman 1988a, pp. 247-50; Pascall 1986). Under U.S. unemploymentinsuranceprograms,claimants may be declared ineligible because they are unable to work at any time or place because of child care responsibilities or spouses' work commitments(Pearce 1986). In a few welfare states - primarilyin the Nordic countries - men and women receive benefits solely on the basis of citizenship.This arrangementis most common in health insur- ance or medical care and flat-rateold-age pen- sions. Thus, many analysts consider benefits based on universalcitizenshipto be most likely to further gender equality. Although citizen- ship-basedbenefits may be more conducive to equalitythan work- or need-basedbenefits, the range of needs covered by such benefits often betrays a gender bias. For example, benefits claimed on the basis of paid workreceive fund- ing priority while the public services that women depend on are not funded sufficiently to serve all those eligible (Hobson 1990, p. 247; Ruggie 1984, chap. 6). The Scandinavianstates also tend to privi- lege those claiming benefits based on labor marketparticipation.Since the 1960s, entitle- ments based solely on citizenship have de- GENDERAND SOCIALRIGHTS out access to market-basedwelfare benefits, and thus aremore dependenton public benefits and services than are men (Shaver 1983; Piven 1985; U.S. House of RepresentativesCommit- tee on Ways and Means 1988, pp. 34-35; Tay- lor-Gooby 1991; Nelson 1984). Thus, dualism promotedby the welfare state has a genderdi- mension as well as a class dimension.As a re- sult, some analystshave arguedthatwomen are more likely than men to give political support to public social provision. Social-democratic regimes are alleged to producea universalistsocial politics. However, an analysis of four Nordic welfare states differences between men's and women's links to and attitudestowardthe state (Hernes 1987, 1988). Men are politicized by their participa- tion in the labormarket,thatis, they to the state by theirparticipationin organizations- employers' associations and labor unions - that bargaindirectly with the state over social benefits.Althoughwomen are often workers, they are usually under- represented in corporatist decision-making bodies. More important,women's statusas cli- ents and employees of the welfare state cizes and mobilizes them. In Scandinavia, these gendered patterns of state-citizen rela- tionshipshave genderedthe debateover public versus private provision of welfare such that women are strongersupportersof state welfare benefits and services. Systems of social protectionproducegender differencesas well as class differencesin ests and coalitions. Indeed, comparisonsof the relative salience of gender, class, and other bases of identity and mobilizationin different welfare states should be revealing (Michel and Koven 1990; Skocpol and Ritter 1991; Jenson 1991; Skocpol 1992). Genderingthe Social CitizenshipRights/ DecommodificationDimension In Esping-Andersen'sversion of sources scheme, social rights that decom- modify labor are essential to realizing tentialof the welfare statefor emancipatingthe working class from the capitalist market individual employers.24 Decommodification 24 Some class analystsare dissatisfiedwith notionthatstatesocial provisionundercapitalism coulddecommodifylabor.Clearly,Esping-Ander- 318 CorcoranandDuncan 1979). The implicitmale standardfor "worker"obscurespowerrelations in the family and the conditions under which social reproductiontakes place. Yet domestic work must be done and care providedfor chil- dren, the elderly, and the disabled. The social organizationof domestic and caretakingwork must be examined, as well as the extent to which access to services is a right of citizen- ship or is conditionedon laborforce participa- tion, maritalor family status,or financialneed. TWO NEW DIMENSIONSOF WELFARE STATES Access to Paid Work How does an analysis based on the situationof commodified male workers deal with women working in the home? For many women and others excluded from paid labor, commodifi- cation - that is, obtaining a position in the paid labor force - is in fact potentially emancipatory.Contemporaryand historicalre- search has found that many women want paid work because it provides independence and enhancedleverage within marriageand the pa- triarchal family (Blumstein and Schwartz 1983; England and Kilbourne1990; Benenson 1991). Equal access to paid employment and equal pay has been a consistent - and con- tested - demandof women's movementsover the past century(Hobson 1991b). In marriages (or other family relationships)in which power relationsarebased largely on economic depen- dence, access to paid work and to the that make employment a viable option for mothers(or othercaretakers)is as importantas - perhaps even more important than insulation from marketpressuresprovided decommodification.25 Thus, the decommodifi- cation dimensionmust be supplementedwith new analytic dimension that taps into tent to which states promote or discourage 25 Wives' economic dependence "both reflects labor marketrealities and reinforceswomen's position in the labormarket"(Hobson 1990, p. 236; Sorenson and McLanahan1987). Economic depen- dence is associated with less power within the fam- ily because decision-making in marriageis largely based on spouses' contributionsto family income (Blumstein and Schwartz1983; EnglandandFarkas 1986; England and Kilbourne 1990). By affecting decisions about investment in "humancapital"and GENDERAND SOCIALRIGHTS working-classmen couldfulfill the role offam- ily provider This was achieved by improving male workers'marketposition, by supplement- ing men's wages, and by providing honorable public benefits when work was unavailable. This policy strategywas linked with a "family wage" strategyin the marketand was premised on a traditionalgender division of labor, that is, women were responsible for domestic and caring work (even if also engaged in paid la- bor) and men were responsible for providing the bulk of the family income. The dominant goal of post-World War II workers' movements has been to extend social rights. Yet labor movements vary in the extent to which they continue to defend the family wage and male family headship. Some work- ing-class movementshave changedin response to the increasing numbers of women in their constituencies and have supported anti-dis- criminationand comparableworth legislation and services such as day care (Ruggie 1984, 1988; Jenson, Hagen, and Reddy 1988; Milk- man 1990). Yet in few (if any) instances have such movementsembracedan explicit feminist goal of economic independence for women. Rather,their goal usually is to allow married women to combine paid work with family re- sponsibilities- to be secondaryearnerswhile continuing to service their husbands (Hernes 1988; Lewis 1992). (Single women may fare somewhat better.) The Capacity to Form and Maintainan AutonomousHousehold The concept of decommodificationoriginated in analyses of class relationsand class politics. New categories are needed to deal with the ef- fects of state social provision on gender rela- tions. If individuals who carry out caring and domestic work do not enter the labor market, or enter it only as secondary workers, the re- sulting distributionof income within the fam- ily and the availabilityof otherincome sources affects their own and their children'swell-be- ing. Over the last century or so, the wage" supplementedby social rights vided unevenly for wives and children.Femi- nists often say that women are "a husband away from poverty";if you've got a husband and he shareshis income with you, you're pro- tected, but if not, you're likely to suffer eco- nomically.Most men simply do not sharetheir 320 The decommodificationdimension could be subsumed under a more generic dimension measuringindependenceor autonomy,that is, it would indicate individuals' freedom from compulsionto enterinto potentiallyoppressive relationshipsin a numberof spheres. This di- mension would reflect the growing individual- ization flowing from processes of moderniza- tion and state-building,processes that have re- placed the networks of mutual duties and re- sponsibilities - and hierarchies tional corporate bodies with direct links be- tween citizens and states. States now offer re- sources to the differentparties in relationships of domination, accommodation, and conflict (e.g., markets, families, and interracial rela- tions). These state-providedresourcesalterthe balance of power in these relationsand within the polity. Individuals typically participatein many such relationships.Thus, the role of the state cannot be understoodin referenceto only one relationship- decommodificationvis-a- vis the market cannot ignore gender relations in the family or race relationsin communities. The total package of resources available from both public and private sources across social locations must be considered.Attentionwould shift from dimensions tied to only one set of potentially unequal or oppressive relations to an examinationof the combined effects of all programson individualsin specific politically and socially significantgroups. This solution would meld the concepts of decommodificationand access to an indepen- dent income (outside of marriage)into a uni- tary concept of individual independence, or better yet, a concept of self-determination are conditioned jointly on the incomes of both spouses in the case of marriedcouples, which ef- fectively disqualifies the second earner,usually the woman, from benefits when her income rupted (Shaver 1983, 1990; Bernier and LaJoie 1986, pp. 105-06). Women's claims to benefits are also undercutby the "cohabitationrule"(presentin many countries' social assistance programs),which presumes that living or sleeping with a man indi- cates that he is financially supportinghis Whether a family allowance offers an independent income to mothers depends on which parent re- ceives the benefit - feminists had to fight for women's right to be designated beneficiaries (Pascall 1986, p. 220). Regulations allowing women independentaccess to benefits and services are more woman-friendlythan those that force de- pendence on household qualification. GENDERAND SOCIALRIGHTS sources analysts on "socializing the costs of familyhood" and "allowing women to choose work." The capacity to form and maintainan autonomoushousehold relieves women of the compulsion to enter or stay in a marriagebe- cause of economic vulnerability(thus parallel- ing the effects of the citizen's wage for work- ers vis-A-visthe market).Following Hirschman (1970), the right of exit - in this case, to be able to choose not to enter or stay in a mar- riage - alters the power relationswithin mar- riages (Englandand Kilbourne1990). The state is woman-friendlyto the extentthatit enhances the Social Rights of Citizenship: The Comparative Analysis of Gender Relations Vol. 58, No. 3 (Jun., 1993), pp. 303-328 Association 03. your acceptance of the Terms & Conditions of Use, available at. /terms.jsp. helps scholars, researchers, and students discover, use, and build upon a wide range of We use information technology and tools to increase productivity and facilitate new forms about JSTOR, please contact [email protected].. Association is collaborating with JSTOR to digitize, preserve and extend access to OF CITIZENSHIP: STATES* ANN SHOLA ORLOFF Universityof Wisconsin women's material situations, shapes gender relationships, and participation, and contributes to the formation and mo- Mainstream comparative research has neglected on the welfare state has not been systematically framework for analyzing the gender content of so- and mainstream work. Three dimensions of qualita- resources analysts are reconstructed to incorporate dimension is extended to consider the ways coun- welfare through families as well as through states and state-market-family relations dimension; (2) the stratifica- consider the effects of social provision by the state on treatment of paid and unpaid labor; (3) the social citi- dimension is criticizedfor implicit assumptions about and domestic labor and for ignoring the differential effects decommodify labor Two additional dimensions are of state social provision on gender relations: access to autonomoushousehold. gender relations to social provision by the state. Many recent analyses have recognized that states regulate gender relations in the la- bor market,polity, family, and elsewhere (Wil- son 1977; Peattieand Rein 1983; Shaver 1983; Ruggie 1984; Piven 1985; Pascall 1986; Sapiro Orloff, Wisconsin, 1986; Connell 1987; Sassoon 1987; Gordon This 1988a, 1988b, 1990; Pateman 1988a; Abro- the movitz 1988; Laslett and Brenner 1989; Mink and is 1990; Walby 1990; Orloff 1991; Lewis 1992; on "The Skocpol 1992). Theorists may disagree about with the causes of gender inequality and women's Earlier subordination, but few would deny that the meetings, character of public social provision affects States, Swe- women's materialsituations,shapes gender re- comments Janeen ern state social provision may be misleading be- cause it assumes what must be proved - that states promotethe welfare of their citizens throughsocial policy - and also because it assumes that a com- mitment to public social provision, once estab- lished, is irreversible.Generally, the welfare state is conceptualized as a state committed to modify- ing the play of social or market forces in order to achieve greaterequality (Ruggie 1984, p. 11). The welfare state is often operationalizedas the collec- tion of social insurance and assistance programs describe mod- that offer income protection to victims of unem- Review, 1993, Vol. 58 (June:303-328) 303 AMERICANSOCIOLOGICAL REVIEW stream comparative research on the welfare state has considered gender relations, most feminist researchon the welfare state has not been comparative.Thus, not enough is known about how and to what extent systems of so- cial provision actually do vary in their gender content, how social provision and other state institutionsaffect genderrelations,andhow the state's impact on gender relations is related to its effects on other social relations. Conversationsbetween mainstreamwelfare state researchers and feminist researchers would benefit both groups,yet therehave been far too few structuredconfrontationsof con- ceptual frameworks and empirical findings across the two bodies of scholarship.4Mutual appreciationhas been hamperedby the differ- high- ent analytic strategiespursuedby feminist and agency in mainstreamscholars.Until recently,most femi- on citizen- nist empirical analyses of the welfare state have been case studies and have not engaged politics, and the conceptualframeworksand empiricalfind- ings of the mainstreamliterature.5Moreover, the they often emphasized the ways in which so- usually ignored gender differences and inequalities main- in social politics and policies, while feminist work has been premised on the importanceof gender in and political life. 4 Quadagno (1989) has confronted the assump- tions of power resources analysis. Some other works also span feminist and mainstreamresearch: use the term to Gordon (1988a) discussed Piven and Cloward's and his-(1971) influentialwork from a feminist perspective; of modernHobson (1990) examined women's economic de- pendence in OECD countries and how this is re- state, or state the state in lated to efforts to decrease class inequality (the fo- cus of much of the mainstream comparative re- How- state social in- search); and Skocpol (1992) offered a gendered analysis of American social provision of the late nineteenthand early twentiethcenturies in contrast this issue to mainstreaminterpretations. rather5 Feminist theory on the welfare state has come the gender order in from two camps, neitherof which has been particu- and twentiethlarly engaged in scholarly debate with researchers to "a carryingout empiricalinvestigations of the welfare state. First, a socialist-feminist group has debated Marxists about the characterof the system that the of social and eco- unchangingde- welfare state allegedly reinforces - a debate that agency to is essentially an extension at an abstractlevel of the debates about capitalism and patriarchy(McIntosh 1978; for criticism of this approachand an exem- "feminist" plary comparativeanalysis, see Jenson 1986). Sec- ond, women working in the area of democraticand oversimpli- liberal theory have critiquedthe "masters"for their nongendered analyses of citizenship and political has participation(for a review, see Jones 1990). 305 the social rights of citizenship (Korpi 1989; Esping-Andersen and Korpi 1987; Esping- Andersen 1985, 1989, 1990; Kangas 1991; Palme 1990; Myles 1989; Kolberg 1992). This work providesa bridgeto recentfeminist work on the welfare statethatexaminesthe gendered characterof social rights or claims on the state andthe possibilitythatthese rightscan enhance the relative position of women (Piven 1985; Hernes 1987, 1988; Siim 1988; Gordon 1990; Orloff 1991; Skocpol 1992; Shaver 1990; Sarvasy 1992; O'Connorforthcoming).These social citizenship perspectives emphasize the potential of social provision in democratic states, securedat least partiallythroughthe po- litical struggles of citizens and others, to counterdominationeven as they acknowledge that this potential is often far from being real- ized. The programsof the modernwelfare state differentiallyadvantagevarious social groups, and there is importantvariation across coun- tries and programs,as well as over time, in the extent to which the interests of dominant and subordinategroups are enhanced.In short, so- cial citizenship analysts envision social policy as having an emancipatoryas well as a regula- tory potential.Even where emancipationis not fort,"mea- a manifestobjective,social programsmay have unintended"independenceeffects." To understandthe mutualeffects of state so- cial provision and gender relations requires a conceptualscheme that can be used in system- atic comparativeresearch.Rather than devel- oping such a scheme anew, I would argue that it will be more fruitful to directly engage the conceptual frameworksof mainstreamlitera- ture and propose amendmentsthat will reflect what is already known about gender relations and the state. Feministresearchcan therebyin- corporateadvancesin the mainstreamliterature while transformingit to incorporategender re- lations. I offer some critical reflections on the ana- lytic categoriesof mainstreamcomparativeand historical research on the welfare state, espe- cially those employedby the influential"power resources"school of analysis. The power re- sources school has demonstratedin some care- ful studies that "politics matter,"in contrastto ap- those thatcontendthat social policy simply re- flects the systemic "needs"of capitalist,indus- trialized societies. Rather than enter into this debate, I concentrateon the power resources group because they have developed a frame- AMERICANSOCIOLOGICALREVIEW vision on genderrelations- requiresnew con- ceptualcategories and analyticdimensions. Power resourcesresearchersbegin from the premise that workersare oppressedby capital- ism, which transformslaborpower into a com- modity.However,politicalrightsin democratic polities enable workers to mobilize to further their interests. These scholars build on Mar- shall's (1950) distinctionbetween types of citi- zenship rights - civil, political, and social (Korpi 1989, p. 312; see also Hasenfeld, Rafferty,and Zald 1987).8The developmentof social rights reversed the nineteenth-century separation of social protection from citizen- ship. Along with Marshall,they maintainthat this critical social transformationtook institu- tional form in a move away from poor relief to modern social policies, like social insurance and universal benefits based on citizenship. They link social rights based on citizenship to an accountof political mobilizationthat draws on Marxist and Weberiananalysis and the so- cial-democratictraditionsof parliamentaryso- cialism. They identify two major power re- sources in Westernsocieties: capital, an inher- ently unequally distributedmarket-basedre- source, and the right to vote and organize for collective political action, a right that is pre- sumedto be equally distributedin democracies (Korpi 1985). The class-relateddistributionof 8 Marshall(1950) defined citizenship as "a status bestowed on those who are full membersof a com- munity,"and he saw citizens as "equalwith respect to the rights and duties with which the status is en- dowed." However, the content of citizenship rights varies because "no universal principle.. deter- mines what those rights and duties shall be" (pp. "claims," "welfare" 28-29). Analyzing the experience of British work- ing men over the last three centuries, Marshallpre- sented an evolutionary argument about the devel- opment of civil, political, and social citizenship rights (Barbalet 1988). He argued that in the eigh- teenth century,civil rights graduallyattachedto the status of freedom already enjoyed by male mem- bers of the community.Political rights - primarily the franchise- were first grantedto propertyown- ers, but were extended to all citizens, including women, over the course of the nineteenthand early twentieth centuries (Marshall 1950, p. 20). Social not pick on my rights - "the whole range from a modicum of eco- nomic welfare and security to the right to share to commentatorshave the full in the social heritageand to live the life of a the content of civilised being according to the standardsprevail- has focused ing in the society" - which is associated with the state in policy welfare state, are the productof twentieth-century political changes (Marshall 1950, p. I 1). 307 differences, I think it is possible to focus on the processes and institutions that most femi- nist analysts agree are importantto gender re- lations and that are affected by state policies. Fundamental to full social participation and self-determinationare control over one's body and bodily capacities (including sexuality and reproduction)and the right to political partici- pation. These are not central to the welfare state, but are part of the relevant context for evaluatingthe effects of the state on gender re- lations. Many institutions and processes con- stitutegenderrelationsand are directlyaffected by state social provision:the sexual division of labor(includingthe treatmentof care work and caseworkers),access to paid work (as a central role in our societies and as a means of sur- vival), and marriageand family relations.11In the following pages, I discuss and critique the power resources analysts' understanding of citizenship and their analytic scheme for de- scribingsocial policy regimes, then propose an alternativescheme for evaluating and catego- rizing state social provision that can capture both class and gender effects.12 sexuality and the concomitant compulsion for women to enter heterosexual relationships to ser- vice men's personal and sexual needs. Another school of thoughtfocuses on genderdifferences and the control of biological reproduction.Many femi- nist political theoristshave been concerned with le- gal, political, and organizationalbarriersto gender equality, including women's subjugation to male family heads and exclusion from politics - both participatorydecision-makingarenasand control of the means of administration and coercion. Still other theorists focus on men's control of women's labor throughthe sexual division of labor, exploit- ing women's economic dependence, and sex segre- gation in occupations;these theorists are especially concernedwith the relationshipbetween gender and class power. Of course, many analysts recognize that more than one dynamic is involved in produc- ing gender relations. (For overviews, see Walby 1990; Connell 1987; Tong 1989; Collins 1991.) 1 I focus on elements of social provision that have a significant impact on gender relations, but these elements do not necessarily affect all women interests or or all men in the same ways (Spelman 1988; the feminist-influ- Crenshaw 1989; Harris 1990). These dimensions (much as will need rethinkingand supplementationalso from mainstream the perspective of incorporatingracial and ethnic Molyneux relations, which I do not here attempt. 12 I focus on the conceptualization of these di- the im- mensions, ratherthan on their operationalizationin bodies and empirical studies. The operationalizationpresents AMERICANSOCIOLOGICALREVIEW in his own past work and called for gendering these categories in future analyses (also see Pateman 1988a, p. 232). An illustrationof the implicit male standard appearedin recentcomparativeanalyses by the Korpi-Esping-Andersenproject that focused on the ways in which social rights affect "typi- cal cases" because "legislative statements are difficult to compare"(Palme 1990, p. 27). The "typical cases" used in the project's analyses included "an 'average' production worker in manufacturingindustry"and the same worker with dependent spouse and two children (Palme 1990; Korpi 1989, p. 315; Esping- Andersen1990). Of course, because of prevail- ing sex segregationin occupations and house- hold composition, both these "average"work- ers "happened"to be men. The analysts then assessed the qualityof benefitsthatreplacelost income for these average citizens based on their degree of labor marketparticipation,us- tic'welfare ing average wages as a baseline.14 The social rightsof citizens who are economically depen- dent, the vast majority of whom are women, were not considered.Analysts should not rest with an adumbrationof social rights for "typi- cal" worker-citizens- such engendered citi- zens do not exist. Men make claims as worker- citizens to compensatefor failures in the labor market;women make claims as workers, but also as members of families, and they need programs especially to compensate for mar- riage failures and/orthe need to raise children alone. Genderand CitizenshipRights The power resourcesschool assumes that civil and political rights are equally available to all citizens to use in mobilizing to secure greater social rights. Thus, it ignores gender differ- ences in access to civil and political rights, in- cluding the legal rights of personhood. Korpi (1983) noted variationin the extent of formal equality of citizenship rights but not with ref- erence to gender; Esping-Andersenmade no on the welfare 14 The average wage was calculatedfrom data for adult men and women in manufacturing;it assumed full-time, all-year work. Thus, the typical male standardof living was the benchmarkfor pension adequacy,but deviations from that standardare not understoodas having genderimplications.Analysts see inequalitiesin benefits solely in class terms. 309 pate as "independentindividuals"- citizens - in the polity, which in turn affect their ca- pacitiesto demandandutilize social rights.The ways that states intervene - or refuse to- are critical to women's situation. Political rights are also a problem for women. Women'sgender was once considered reason enough for exclusion from the suf- frage.16After the vote was won, the extent to which actual equality of rights was achieved has varied.However, this variationhas not en- tered the analytic frameworksof mainstream researchers.In all forms of formalpolitical ac- tivity save voting, women participateat a lesser ratethanmen (Randall1987; Nelson 1984); for power resourcesanalysts such participationis, of course, the basis for enrichingsocial citizen- ship in the first place. A complete analysisof states'effects on gen- der relations should not rest with "social rights"as they are defined by mainstreamre- searchers. Rather, analyses of social rights should include an examination of family law and the legal frameworksand social programs dealing with legal personhoodand the control of one's bodily capacities and functions. Fur- thermore,the analysis must examine issues of political rights and participation (Hernes 1988). An accuratepicture of the content and effects of state social provision should not be- gin from the premise of a gender-neutralciti- zenship. Rather,one must take account of the very real gender differences in productiveand reproductivelabor and access to civil and po- litical rights and how these differences influ- ence the ways in which men and women struggle for and claim benefits from the state as citizens. bod- THE DIMENSIONSOF THE WELFARE STATE Esping-Andersen(1990) and Korpi(1989) pro- pose three dimensions that characterize the content of the welfare state: the relationship such as s[see, 16 Skocpol (1992) described women's unusual political capacities in the Progressive Era in the United States. Despite being denied the franchise, women waged successful campaigns in most states for mothers' pensions and other legislation. This episode underlines the difficulties of applying Marshall'sevolutionaryanalysis of citizen rights to women. AMERICANSOCIOLOGICALREVIEW thus shifts from buying power on the market towardpolitically based considerationsof jus- tice" (p. 313). Thus, there will be class-influ- enced debates over the content of social policy and over the relative roles of marketsand poli- tics in determiningwelfare outcomes. States in a given regime-typeact similarly in regardto the market.In countrieswith a liberal social policy regime, the market,ratherthanthe state, guaranteesmost welfare needs of most citizens. For instance, in Canada,Britain, and the United States public pension benefits make up a smaller proportionof the incomes of the elderly than they do in Scandinaviancountries welfare or in Europe(Myles 1989, pp. 123-24). More- over, liberal states tend to respond to societal "failures"ratherthaninterveneto preventprob- lems from occurring.Thus, programsin liberal regimes avoid undercuttingthe market by of- fering only stigmatizing subsistence-level grantsto those unableto participatein the mar- ket. In contrast,social-democraticand conser- vative regimes arepro-activeandretaina larger range of welfare activities, effectively crowd- ing out the market.For example, in both these types of regimes, privatepension schemes for better-offworkershave been forestalledby the expansion and elaborationof state programsto cover all strataof the population(Palme 1990; Esping-Andersen1990, chap. 4). Esping- The StratificationDimension A second dimensionof policy regimes is strati- fication: "The welfare state... is, in its own right, a system of stratification.It is an active force in the ordering of social relations" (Esping-Andersen1990, p. 23; Esping-Ander- sen and Korpi 1987, p. 40). Thus, against the common view that "welfare states" promote greater equality, power resources analysts ar- gue thatsystems of social provision have strati- fying effects: Some policies may promote equality, cross-class solidarity, or minimize economic differences, while others may pro- mote social dualism or maintainor strengthen class, status, or occupational differentiation (Esping-Andersen 1985, 1990). Social-demo- craticregimes foster solidarityby including all citizens in common programs,and they reduce class differences through income redistribu- tion. Conservative-corporatist regimes rein- force statusand occupationaldifferentiationby offering separateprovision for differentsocial 311 duction outside the labor contract meant that people were commodified. In turn, the introduc- tion of modernsocial rightsimplies a loosening of pure commodity status. De-commodification oc- curs when a service is rendered as a matter of right,andwhen a personcan maintaina livelihood without reliance on the market.(pp. 21-22) For Esping-Andersen (1990), decommodifi- cation is the heart of the welfare state's emancipatory potential: As commodities, people are captive to powers beyond their control... o If workers actually do behave as discretecommodities,they will by defi- nition compete; and the fiercer the competition, the cheaper the price. As commodities, workers are replaceable,easily redundant,and atomized. De-commodificiation is..., as Polanyi argued, necessary for system survival. It is also a precon- dition for a tolerable level of individual welfare and security. Finally, without de-commodifica- coalitions. tion, workersare incapableof collective action;it is, accordingly,the alpha and omega of the unity and solidarity required for labor-movement de- velopment. (p. 37) Esping-Andersen argued that the extent to which the rights embodied in social programs promote or circumscribe decommodification of labor is a critical dimension that varies across welfare states. The most decommodifying sys- tems offer many generous benefits simply on the basis of citizenship, whereas the least decommodifying have a relatively circum- scribed range of social rights and most assis- tance is means-tested, which severely limits the emancipatory potential of benefits. Social- democratic regimes are the most decommodi- fying since provision is generous, many ben- efits are universal, and access is relatively easy for workers. Liberal regimes limit decommod- ification of labor by conditioning limited ben- efits on means tests or contributions based on work. Conservative-corporatist regimes, which have strong citizenship rights to social benefits, do not promote decommodification of labor, because the conditions for benefits reinforce reliance on work and the market - typically benefits are linked to contributions. rights GENDERING THE DIMENSIONS OF WELFARE STATES ic The three dimensions proposed by Esping- Andersen and Korpi (1987; see also Korpi 1989; Esping-Andersen 1990) have given ana- AMERICANSOCIOLOGICALREVIEW the aged (Esping-Andersen1990, P. 28). So- cial-democraticregimes also encouragemoth- ers to work in the paid work force by providing day care and parentalleaves. AlthoughEsping-Andersenis rightto recog- nize the effects of services on women's abili- ties to enterthe paid labor force, his classifica- tion scheme does not reflect differencesin how careis provided.While liberalregimes, like the United States and Britain, lag in government provision of welfare services, like child care or elder care, and allow the market to provide them, the social-democraticSwedish state of- fers extensive services (Ruggie 1984; Bergmann1986). Yet among social-democratic states, services are not the same: Women in Sweden are likely to work outside the home, whereasNorway's day-careprovision is much less developed than Sweden's and relatively more Norwegian mothers stay at home (Leira 1992). Within the conservative-corporatist group,Franceprovidesmanyservices for work- ing mothers while Germanypromotes house- wifery by offering few services. In the Nether- lands, despite a strong social-democraticpres- ence that has helped to establish extensive so- cial rights for wage earners(Esping-Andersen 1990, pp. 52-53), families have access to few services, and women have high rates of house- wifery and economic dependence(Knijn 1991; Jenson 1991; Hobson 1990, 1991a; Kamerman and Kahn 1981; Lewis 1992). Esping-Andersen'sregime types do not fully predictwomen's employmentpatterns.Esping- Andersen (1990, chaps. 6, 8) analyzed these patternsin Sweden, Germany,and the United States (as "representatives" of the threeregime types). Provisionof services is importantto the Swedish welfare state, and this "provides a phenomenal multiplier-effect for female em- regimes ployment: Social services both allow women to work, and createa largelabor-marketwithin which they can find employment" (Esping- Andersen 1990, p. 159). The lack of services in Germanyhas retardedthe growth of female employment,whereaswomen's employmentin the United States has increasedsharply,driven by marketforces, in spite of the dearthof pub- lic services.Yet otheraspectsof women's work (e.g., sex segregationof occupations and part- time versus full-time work) are not accurately predicted by the regime types (O'Connor 1992). In Germany(a conservative-corporatist state) and Sweden (a social-democraticstate), 313 sion of state-marketrelationsas formulatedby Esping-Andersensimply ignores the tremen- dous amount of caring labor and housework provided by women - housewives and wage earnersalike. Feminist researchon the "labor of caring"shows thatin all industrializedWest- ern countries, welfare- tending to children, the elderly, the sick and disabled - is largely provided in private households by women withoutpay, ratherthanby states, marketsand voluntarynonprofitorganizations;all Western welfare states depend upon this care to a great extent (Finch and Groves 1983; Land 1983; Land and Rose 1985; Waerness 1984; Taylor- Gooby 1991, p.101). The sexual division of labor in caretaking and domestic work within institutions other than the family must also be considered. In- deed, women carryout a disproportionateshare of welfare work, whetherit is provided by the state,privateorganizations,corporations,or the family. To the extent that this work is under- valued in terms of benefits and political re- spect, women suffer disproportionately. Power resources analysts recognize that the "divisionof labor"between states and markets in providingwelfareis a political question,that is, it is a question of which decision rules ap- ply and which actorscontrolthe distributionof valued resources.Clearly,the division of labor between marketsand polities is based on rela- tions of power. There is no similarrecognition of the division of labor between families and states in providingdomestic and caretakingla- bor, welfare services, and goods. Power re- sources analysts simply do not discuss power relations within the family. For example, they ignore the fact thatthe distributionsof income, resources,and work within the family are con- ditioned on power as well as benevolence and shared interests (Hobson 1990; Pahl 1983, 1988; Land 1983; England and Kilbourne 1990; Acker 1988).17To the extent that any as- pect of men's power within the family is ac- knowledged, it is that based on women's eco- nomic dependence (e.g., Esping-Andersen 1990, p. 28). However,it is not namedas men's power, nor is it relatedto other sources of gen- 17 I should note that I am hardlythe first to point out that liberal or leftist men, even those with good intentions, seem to wear conceptual blinders when the issue is men's and women's unequal power in the family (e.g., see the essays in Sargent 1981). AMERICANSOCIOLOGICAL REVIEW cies reduce the sexual division of labor by shifting the burdenof domestic work to public services and to men. Of course, whether it is men or public servants(especially if the public servants are women) who take on the private and unpaid caring and domestic burdens of housewives and caretakersalso affects gender relationsand the shapeof distinctivegenderre- gimes. Genderingthe StratificationDimension Power resources analysts have focused on the effects of state social provision on class hierar- chies, but have ignored its effects on gender hierarchies.Their existing scheme could cap- ture gender differences in benefit levels that depend on labor market processes. Women's inferior status in the work force means that women are disproportionatelydisadvantaged when benefitsreflect work-relatedinequality.20 To date, however, researchershave not done even this minimal gender-sensitive analysis (but see Quadagno 1988; Pearce 1978, 1983, 1986). They have not addressedtwo significant ways thatstatesreinforcethe genderhierarchy: (1) privileging full-time paid workers over workerswho do unpaidwork or who combine part-timepaid work with domestic and caring labor, and (2) reinforcing the sexual division of laborin which women do the bulk of unpaid work. In most systems of social provision, men's claims are based on paid work,while far fewer women make such claims. Contributionsfrom wages to social insurancefunds bring entitle- ment to benefits, and even in the case of needs- 20 Using the Korpi-Esping-Andersendata set, an analystcould comparethe minimumpensions avail- able to people lacking labor market experience to pensions available to people with lifetime labor marketparticipation(personalcommunicationwith Walter Korpi; see also Palme 1990, pp. 33-34). Housewives and women who have been secondary earnerspredominateamong personsreceiving mini- that is, the "di- mum pensions, while male workers predominate among those receiving the highest pensions (Burkhauserand Holden 1983; Quadagno 1988a; Roos 1985; Matthei 1982; Bradley 1989). Although the proximatecause of these unequalbenefits is dif- ferent patterns of labor force participation, these work patternsin turndepend on relationsof depen- dency and power in the breadwinner-housewife household and men's power within the workplace, the polity, and public spaces. 315 Although this "two-tier" formulation cap- tures some of the ways that careworkers are undervalued,it is misleading in some respects. It focuses only on the direct claims made by men and women - male workers' contribu- tions entitle them to social insurancebenefits, while needy mothers claim benefits based on an income test and their family/maritalstatus. In fact, although women are overrepresented among the clients of social assistanceschemes, they are a majority of clients in most social welfare programs,including the old-age pro- gramsof Social Security(unemploymentinsur- ance is one prominent exception). Indeed, many more women are indirectlyincorporated in the welfare state on the basis of their hus- bands' contributions than claim benefits as needy carers. Unlike women who receive so- cial assistance, wives or widows with or with- out children who receive social security are treatedas "rights-bearers" ratherthanas clients - their marital tie to a covered breadwinner entitles them to the same standardizedtreat- ment and nationally-determined,inflation-in- dexed benefits accorded to men who receive social security.These women arethus betteroff than women who depend on welfare, but they are also worse off relative to men within the same program because dependents' benefits are only 50 percent of the main beneficiary's entitlement (although a survivor gets the full amountafterthe deathof the main beneficiary) (Burkhauserand Holden 1983).23 In the United States, the difference between the two tiers of social insuranceand social as- Survivors' and sistance - often understoodas the difference in treatmentbetween men and women - is better conceptualizedas a difference between membersof families that are, or were, headed uot;formulation by a male breadwinnerwith an economically repeat that it dependent wife (and children), and families provided maintainedby women who are not in the paid of this concept, in labor force, or work on its fringes, who must versus social in- make claims based on their status as mothers. as well as Families with Dependent Children. Also, childless women are also ineligible for AFDC. 23 Access to indirectclaims differs for women in different classes and racial and ethnic groups: A woman must be marriedto a man in covered em- ploymentto receive dependents'benefits. However, men who are blue-collarworkersor membersof ra- cial or ethnic minorities are more likely to be un- employed, which then affects the women who de- pend on them for financial support. AMERICANSOCIOLOGICAL REVIEW dined because the universalflat-ratepensions have been supersededby earnings-relatedpen- sions (Esping-Andersenand Korpi 1987, pp. 50-53; Kangasand Palme 1992). Although the povertyratein Sweden is much lower thanthat in the United States, women and men outside the paid labor force in both countries fare worse than those who have or had a secure at- tachment to it (McLanahan, Casper, and Sorenson 1992). Indeed,Hernes(1988) argued that "there is the underlying assumption [in Sweden and throughoutEurope],almost never expressed openly, that universal welfare state services and transfersmust not exceed those earned in the labour market.... There is a po- litical struggle between men and women to count also certain types of unpaid work as a legitimatebasis for welfare rights"(p. 194). States perpetuatethe gender division of la- bor in a variety of ways; for instance, gaining entitlementto social assistance sometimes re- quires women to demonstrate homemaking skills (Abromovitz 1988; Pascall 1986; Fraser 1989). Both the U.S. and British social secu- rity systems offer greater benefits to "house- wife-maintainingfamilies"(usuallyby offering dependents' or survivors' benefits under the old age program)than to single individuals or dual-earnercouples (in which the woman can receive her own work-relatedbenefits or the dependent'sbenefits, but not both) (Bergmann 1986, p. 258; McIntosh 1978; Pascall 1986; Acker 1988). Other public mechanisms from tax systems to the absence of services to alleviate domestic responsibilities - also maintaintraditionaldivisions of labor. Given the differential treatment accorded unpaidcaring and domestic labor comparedto paid labor and the ways in which programre- quirementsreinforcethe sexual division of la- bor in householdsand the workplace,analyses of states' effects on stratification should in- clude genderrelations.The concept of stratifi- cation - if amendedto account for these fac- tors - remains a useful one. Gendereffects are also apparentin the ways in which systems of social provision affect so- cial politics (i.e., "politicalfeedback").Power resources analysts have noted the dualism of liberalregimes:Benefits and services are mea- ger and availableonly to the poorest individu- als, forcingothersto turnto the marketfor ben- efits and services. In liberal regimes, women make up a disproportionateshareof those with- 317 providesworkerswith income from outside the market,therebystrengtheningtheir leverage in the market.Decommodification,which is tied to the political power inherent in citizenship, influences the political fortunes of working- class movements, and hence, the possibilities for social and political transformation.By ig- noring gender differences in the situations of men andwomen workers,particularlywith ref- erence to domestic and caringlabor,and in ac- cess to the paid laborforce, Esping-Andersen's concept is inadequatefor understandingthe ef- fects of state social provision on all workers. found Decommodification,as a dimension of policy regimes, must be understoodin the context of gender relations and also must be supple- mented by a new analytic dimension: the ex- are linked tent to which states guaranteewomen access corporatistto paid employment and services that enable them to balance home and work responsibili- ties, and the mechanisms and institutionsthat implementthese guarantees. Power resources analysts implicitly begin with the situationof male workers and ignore the genderdivision of laborthatmakes the situ- politi- ations of men and women in the paid work force different.Benefits that decommodify la- bor give male workers greatercapacity to re- sist capital and enter the marketon their own terms, but unpaid services provided by wives, mothers,daughtersalso enhancemale workers' capacities.Thus, to focus only on decommodi- fication is misleading about male workers' inter-situation. What of women workers? Again, power resourcesanalysis startsfrom an implic- itly male premise:Women have "chosen"be- tween housewiferyandpaid work, so that once they enter the paid labor force, their domestic responsibilities disappear from the analysis, and they become indistinguishablefrom male workers.Social benefits that decommodify la- bor affect women and men in different ways because their patternsof participationin paid and unpaid labor differ. For instance, taking the power re- parental leave, an example of a benefit that decommodifies labor, may reduce a working the po- woman's earningcapacity because continuous service with an employer often pays off in in- and creased wages (Bergmann 1986, pp. 77-80; sen's analysis follows the traditionof reformist so- the cial democracy in which significant alleviation of the problems flowing from capitalist relations is believed to be possible. AMERICANSOCIOLOGICALREVIEW women's paid employment - the right to be commodified, if you will. I call this fourth di- mension of welfare-state regimes access to paid work. In some countries, men's rights to jobs are promoted through full employment and active labor marketpolicies. Thus, I con- tend that the extent to which the state ensures access to paid work for different groups and the mechanismsthat guaranteejobs (e.g., reli- ance on private employment, creation of tax incentives, legal regulationof privateemploy- ers, or public jobs programs) are dimensions of all policy regimes.26The key issue in inves- tigatingstates'effects on genderrelationsis the extent to which women (or subgroups of women) can claim this right. Of course, paid work exchanges one form of dependence familial - for another - the dependence on an employer for employment.Classical Marx- ism arguedthat women had to be proletarian- ized as a preludeto their emancipation,which would come to them in their statusas workers. However, I am more concerned with the po- tential of paid work to provide women with some autonomy vis-a-vis marriage(or depen- dence on parents). The historical development of decommodi- fication indicates the importanceof access to paid work and social benefits. Expanding so- cial rights was an historically-specificstrategy of some male-dominatedlabormovementsand their elite (male) allies, as most early social programs - old-age pensions, workmen's compensation, unemployment insurance, and healthinsurance- aimed at securingthe posi- services tion of male workers as breadwinnerswhen they were unable to supporttheir families due to loss of jobs or wage-earning capacities - the (Skocpol and Ritter 1991; Skocpol 1992; by Orloff 1993, chaps. 5-9; Jenson 1986; Hernes 1988, pp. 198, 203). Although some labor a movements have preferred a "voluntarist" the ex- strategy to a strategy based on public provi- sion, the broadly-sharedgoal of male-domi- natedcross-class alliances (at least throughthe first half of the century) was to ensure that weak participationin paid labor, the marital balance of power, in turn, has implications for both spouses' currentand future earning power, economic well- being, and entitlement to social benefits (Hobson 1990; Quadagno1988a). 26 Esping-Andersen (1990, p. 22) did note the importanceof full employmentto social-democratic regimes. 319 income with theirchildrenafterthe dissolution of marriage,and states do not make up the dif- ference fully (althoughthere is some cross-na- tional variationin this) (Garfinkeland McLan- ahan 1986; Kamerman 1986; Kahn and Ka- merman 1988). Single mothers, who have lower earning capacities relative to men and more responsibility for their children's well- being, exemplify the economic vulnerabilities of all women - vulnerabilitiesthat are hidden when women have a secure tie to breadwin- ners. Indeed, the deprived circumstances of single mothersare sometimes an incentive for women to marry(or to not divorce). Moreover, family income is not always sharedequally in marriages (Pahl 1983, 1988), and women's economic dependencyis a significant basis for men's power advantagein families. If decommodificationis importantbecause it frees wage earnersfrom the compulsionof par- ticipatingin the market,a paralleldimensionis needed to indicate the ability of those who do most of the domestic and caring work - al- most all women - to form and maintain au- tonomous households, that is, to survive and supporttheir childrenwithout having to marry to gain access to breadwinners'income. I see two ways to conceptualizea dimension of social provision that characterizesdegrees of family supportand the exigencies of marry- ing. First,a generaldimension of self-determi- nation could be developed that would include independence from markets and marriages. Second, proceeding inductively, a dimension based on the demandsof women's movements could be developed,just as decommodification developed from the aims of male-dominated workers' movements. (This inductive strategy is not innocentof theory,of course, but allows the use of feminists' suggestions about what may be necessaryto emancipatewomen to fur- therclarify the significance of the demandsfor which social movements have struggled.)I ar- gue that the appropriatedimension is the ca- pacity to form and maintain an autonomous householdd.7 "family 27 An importantaspect of social benefits as they has pro- affect the capacity to form and maintainan autono- mous household is the extent to which they indi- vidualizeor "familize"recipients(Fraser1989). For example, the use of household means tests under- mines women's abilities to claim benefits as indi- viduals. Australia's and Canada's means-tested as- sistance programsfor the unemployed and the sick AMERICANSOCIOLOGICAL REVIEW within webs of interdependencies (complete individual autonomy does not exist).28In the end, I preferthatseparatedimensionsdeal with different social relations, but a single dimen- sion that explicitly considers gender as well as class relations would, I think, be an improve- ment over the decommodification/socialrights dimension alone. The problem can also also be dealt with in- ductivelyandhistoricallyby linkingthe dimen- - of tradi- sions of the welfare state to social politics and historicalagency.A focus on women's agency in social politics would supplement decom- modificationwith a dimension that taps a goal of women's movements - the capacity to form and maintain an autonomoushousehold (which can be secured in a numberof ways). Indeed,women's movementshave pursuedtwo principalstrategiesto gain economic indepen- dence: (1) establishing secure incomes for women who engage in full-time domestic work and caring for their children;and (2) improv- ing access to paid work and establishing ser- vices that reduce the burdenof caring on indi- vidual households (Pateman1988a). (This fur- ther supportsthe claim that access to paid em- ployment should be considereda dimension of policy regimes.) These strategies have over- lappedhistorically,but the second has emerged as the more importantin the "secondwave" of feminism. Both strategies would provide women with incomes sufficient to support themselves and their children apart from any claims on breadwinners' income. Indeed, if successful, these strategies would extend to women rights that are implicitly or explicitly now guaranteedto men (of the dominantrace/ ethnic group), as in the Italiancase where, ac- cordingto Saraceno(1992, p. 8), men have the right to a family -- that is, a man has the right to a job or income that allows him to maintain is inter- a wife and children. The focus on women's independentincome for supportinga household and their choices about (at least potentially oppressive) mar- riages goes beyond the focus of the power re- partner. 28 Esping-Andersen(1990) noted that"thesocial- democratic regime's policy of emancipation ad- dresses both the market and the traditional fam- ily... the ideal is... to maximize... capacities for individual independence"(p. 28). However, he did not pursue the implications of independence vis-a'-vis the traditionalfamily and how indepen- dence might differ from decommodification. 321 Ritter 1991; Pedersen 1989). They allowed women and their children to survive without husbands, but in relatively deprived circum- stances. Nowhere did a maternalist strategy achieve parity between benefits for stay-at- home mothersand wage earners'benefits or a standardof living for single mothers compa- rable to their marriedcounterparts.Feminists and others still debate the extent to which a maternaliststrategy- paying women to stay home with theirchildren- is a viable strategy, and this seems to vary cross-nationally. Increasing work opportunities and shiftin

Gender and Social Rights of Citizenship PDF

Document Details

Tags

Related

Summary

Full Transcript

Upgrade to continue