The Gender Division of Labor: "Keeping House" and Occupational Segregation in the United States Author(s): Philip N.

Cohen Source: Gender and Society, Vol. 18, No. 2 (Apr., 2004), pp. 239-252 Published by: Sage Publications, Inc. Stable URL: http://www.jstor.org/stable/4149435 Accessed: 25/08/2009 12:36
Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in the JSTOR archive only for your personal, non-commercial use. Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at http://www.jstor.org/action/showPublisher?publisherCode=sage. Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed page of such transmission. JSTOR is a not-for-profit organization founded in 1995 to build trusted digital archives for scholarship. We work with the scholarly community to preserve their work and the materials they rely upon, and to build a common research platform that promotes the discovery and use of these resources. For more information about JSTOR, please contact support@jstor.org.

Sage Publications, Inc. is collaborating with JSTOR to digitize, preserve and extend access to Gender and Society.

http://www.jstor.org

THE GENDER DIVISION OF LABOR "KeepingHouse" and Occupational Segregation in the United States
PHILIPN. COHEN Universityof California,Irvine

Thisarticle exploresthe effect of women'smovementinto the labor marketon the gender segregationof work, using the CurrentPopulationSurveyfrom 1972 to 1993. The author includes as working those respondentswho were "keepinghouse" and codes keeping house as an occupation. The results show higher estimatesof gender segregation,and slightly steeper declines over time, thanwere seen inprevious studies.Analysis of one-year longitudinalchanges reveals less movementout offemale-dominated occupationswhen keepinghouse is includedas an occupation. Finally, a decompositionof the segregation trendshows thatthe movementof womenawayfrom keepinghouse contributedas muchto the overall decline in gendersegregationas did thedesegregationofpaid occupations. Theauthorconcludes that the movementof women's workfrom the household to the labor markethas been a drivingforce in the changing natureof gender inequality. Keywords: occupational segregation; housework; gender inequality

Gender segregationin the labormarketis high, fueled by genderedand discriminatory practices and assumptions (Baunach 2002; Nelson and Bridges 1999; an institutionalizedgenderdivision of labor,and Reskin 1993). But for reproducing labor marketis still no matchfor the "genderfactory" women's the work, devaluing of the married-couplefamily (Berk 1985). This article explores the effect of women's labor moving into the paid marketon the overall gender segregationof work and thereforeon the changing natureof gender inequality. The genderdivision of laboris a centralfeatureof gender inequality,both in its economic aspects and in the social constructionof gender identities (Huber 1991; all systems of gender stratification Lorber1994). As Chafetzwrote, "undergirding is a gender-baseddivision of labor,by which women arechiefly responsiblefor difon the genferenttasksthanaremen"(1991, 77). However,the empiricalliterature der division of labor is uncomfortablydivided between those who examine the
AUTHOR'SNOTE:I would like to thankChristineBose, Nancy Folbre,MattHuffman,and the Gender & Society reviewers for their commentsand suggestions and Liana Sayerfor her contributions. REPRINTREQUESTS:Philip N. Cohen,DepartmentofSociology, UniversityofCalifornia, Irvine, CA 92697-5100.
GENDER & SOCIETY,Vol. 18 No. 2, April 2004 239-252 DOI: 10.1177/0891243203262037 0 2004 Sociologists for Women in Society

239

240

GENDER& SOCIETY / April2004

division of household labor and those who study gender segregationin the paid labor market. The overall gender division of labor has not been considered in empirical studies of the United States (Miller and Garrison1982). The problem with separatestudies of houseworkversus occupationalsegregation is thatthey cannotshow the dynamicrelationbetweenthe two. In this article,I bring together the division of household and labor marketwork in one, partial attemptto fashion a unifiedmeasureof the division of laborfor one periodof recent history.The resultsunderscorethe importanceof the movementof women's labor from the household to the labor marketin reducinggender inequality. HOUSEWORK AND OCCUPATIONS Researchconsistentlyhas shown thatwomen do the lion's shareof unpaidlabor within households (Coltrane 2000). Although this inequality has decreased in recent decades, the household division of laborremainshighly gendered(Bianchi et al. 2000). At the same time, gendersegregationin the labormarketremainshigh, althoughaftera half centuryof apparentstability,therewere declines in the 1970s and 1980s (Blau, Simpson, and Anderson 1998; Cotteret al. 1995; Reskin 1993; Wells 1999). Of course, change in these two arenasis linked, but thatconnection is rarelythe subjectof directexamination.The entryof greaternumbersof women into thelabor force occurredas household services, products,and technology reducedwomen's housework obligations and increased the demand for female labor in the market (Cohen 1998; Cohen and Bianchi 1999; Cotter, Hermsen, and Vanneman2001; Cowan 1983; Presser 1999; Strasser 1982; Uttal 2002). The movement of care work from within the family to the marketrepresentsa fundamentalshift (Folbre and Nelson 2000). We have not reachedthe point at which we might "eliminate the home as a place of work and housewives as a functionalgroupof the population" (Durand 1946, 222), but we have unquestionablymoved in that direction (Stacey 1993). With women more likely to be employed, the segregation of paid work has increasedin importanceas a componentof gender inequality.Chang (2000, 1658) has been thatoccupationsarethebackarguedthat"thelong-standingpresumption bone of the class stratification system, butas women enterinto the formaleconomy in ever-increasingnumbers,the occupationalstructurebecomes the main locus of gender stratificationas well." This echoes an earlierbody of researchon the shift from home to market,which stressed the continuity of gender segregationin the new context of the labor market: The sexualdivisionof laborreappears in the labormarket, wherewomenworkat
women'sjobs, often the veryjobs they used to do only at home .... As these jobs are

low-status andlow-paying, relations remain theirmaterial intact, patriarchal though

Cohen / GENDER DIVISION OF LABOR

241

baseshiftssomewhat fromthefamilyto thewagedifferential, fromfamily-based to 1981,25) (Hartmann industrially-based patriarchy. Thereclearly is a connectionbetween the work women do at home and the occupations that are female dominatedin the labor market.However, it is misleading to collapse the two entirely because there are very few paid occupations that are as female dominatedas "women's work"in the home. Manypeople who studythe segregationof paid occupationsareconcernedtheoretically with the overallgenderdivision of labor.For example, in his international comparison,Jacobs(1989) revieweddataon the genderdivision of labor,including paid and unpaid work. But in the data analysis, he examined paid work only. Because women tend to move between occupations that are more or less female dominatedduring the course of their careers, through "revolvingdoors,"he suggested that gender segregation is continually reproduced through processes of social controlthat define male- and female-dominatedwork, ratherthan women's "taste"for certainjobs or theirhuman capital assets. The question he examinedhow gender segregationis reproduced-is importantfor all kinds of work, but the labor force data he used restricthis analysis to paid employment. Similarly,in ValerieOppenheimer's(1970) landmarkstudy,she acknowledged that "census statistics best reflect . .. paid employment outside the home, rather than the trends in all kinds of productive work carried out by women." But she added, "Thislimitationshould not, however,be a serious drawbackin the analysis of the labor force of an industrialsociety" (p. 10). From the perspective of these laborforce studies, then, a defining characteristic of industrialsociety-of modernity, in fact-is that the paid labor marketreplaces the home as the centralsite of gender inequality. Thus, many researchersconcernedwith gender inequalityhave moved to focus primarilyon occupationalinequality,even as most feminists stressthe continuityin the division of laborbetween home and market(Cohen and Huffman2003; Cotter et al. 1997). But the transitionfromunpaidlaborat home to paid laborin the market is itself a source of change in the gender division of labor.Direct comparisonsare difficult to find, but consider the examples of cooking and cleaning. In 1995, women did 74 percentof all unpaidcooking at home, butin the market,only 45 percent of all cooks were women (this category excludes those working in private households, a tiny fractionof the total). Similarly,women did 80 percentof unpaid housecleaning at home, but only 35 percentof janitors and cleanerswere women.1 Insofaras the division of labor is a cornerstoneof gender inequality,then, women leaving home and going to work may itself reduce gender inequality.In fact, the market's ability to pull women from the household (Cotter, Hermsen, and Vanneman 2001) has been a leading factorin the partialredivisionof houseworkin the past few decades. This may be seen in the many studies that show less gender inequalityin couples' houseworkwhen women areemployed (BatalovaandCohen 2002; Bianchi et al. 2000; Coltrane2000).

242

GENDER & SOCIETY / April 2004

Existing studies of trends in occupational segregation implicitly treat women who enterthe labormarketas if they arejust beginningto work:The level of gender segregationis assessed only in their paid work capacity.2In this analysis, I treat women who leave home for the labormarketas if they were exchangingonejob for another.This is possible because the CurrentPopulationSurvey (CPS) until 1993 identified respondents who were "keeping house." Although this method raises definiteproblems,as discussed below, it also offers a uniqueopportunityto see the in the overall gender division role of women's increasinglaborforce participation of labor.

KEEPING HOUSE: COUNTING HOUSEWORK AS WORK Feminists argue that women's work is devalued by the failure of mainstream economics to account for unpaidwork, both internationally (Waring1999) and in the United States. As Folbreand Abel (1989, 547) reported,in the nineteenthcentury,"thecensus institutionalizeda definitionof 'work' as 'marketwork'thatliterin ally devaluedwomen's unpaidwork"-and it still does. Women'sparticipation the paidlaborforce did not reach50 percentuntilthe late 1970s, even amongBlack women, whose participationrate was historically higher than white women's (Goldin 1990, 17). Thus, neithernationaleconomic accountsnorlaborforce statistics take into accountthe work thatfor much of Americanhistorywas the focus of most women. Similarly,contemporaryanalyses of occupational segregationdo not include the disproportionate shareof unpaidlaborwomen perform.Reskin and Hartmann (1986, 7) acknowledgedthis problemin theirstudyof sex segregation,notingas an is one aside that"theoccupationof most women not in the laborforce, homemaker, for studying of the most segregatedoccupations."That selectivity is appropriate some labor market dynamics, but it precludes us from evaluating the overall division of labor. The U.S. Census Bureauintroducedthe term "keepinghouse" in 1870. In the the instructionsto assistant marshals for the 1870 census, under "occupation," Bureauwrote the following: distinct willbereserved forsuchpersons asreceive Theterm wagesor "housekeeper" fortheservice. Women housefortheir ownfamilies orforthemselves, keeping salary will be entered house." without as "keeping Grown any othergainfuloccupation, themwill be reported without occupation.3 daughters assisting Those women coded as "keepinghouse"were not includedin tabulationsof the laborforce. British census takers,on the otherhand, counted keeping house as an from 1875 to 1905 (Folbreand occupationfrom 1851 to 1881, as did Massachusetts Abel 1989). In 1910, the U.S. Census Bureau instructedenumeratorsto count women doing unpaid labor on family farms as "farm laborers,"resulting in an

Cohen / GENDER DIVISION OF LABOR

243

upwardspike in the historical trend for women's labor force participation.They subsequently produced estimates to eliminate this anomaly, and the offending instructionto enumeratorswas not repeated (Oppenheimer 1970, 2-6). Durand (1946, 221) defendedthe exclusion of farmwives from the laborforce, as "strictly of farm accuratefigures would greatlyoverstatethe relativedegree of participation women in the labor force because a large percentageof them do very little gainful work,"but he providedno evidence to supportthis conclusion. Some researchersof nationalincome in the nineteenthandearly twentiethcenturies attempted to estimate the value of this work, before interest apparently waned. Folbre and Wagman(1993, 279) concludedthat "whatvariedacross states and over time was women's participationin marketwork, probablynot their productivework in general."If thatis the case, we should recognize thatthe exclusion of houseworkersfrom existing occupational segregation studies is an artifactof patriarchal assumptionsin the conception and collection of labor force data.With the work of Waring(1988, 1999), analysts once again startedto factor women's unpaidwork into nationaleconomic estimates, but this newfound intereststill has not reachedoccupationalsegregationstudies. METHOD The analysiscomprisesthreesections. First,I recode laborforce datafor respondentsin the CPS to includethose respondentswho were recordedas keeping house, andI code keeping house as an occupation.ThenI calculategendersegregationlevels with and without the keeping house occupationacross the period from 1972 to 1993 (when the question was dropped).Second, using the one-year longitudinal propertyof the CPS, I broadenJacobs's (1989) "revolvingdoor"-which reflects the movement of women into and out of female-dominatedwork-to include the houseworkoccupation.This will enable us to see the gendercomposition of occupations that women enter when they shift between keeping house and paid work. Finally,I decompose the trendto show the relative contributionsto desegregation of women enteringpaid work versus paid occupations becoming less segregated. The data are from the CPS Annual Demographic Files (March).The CPS is a large,monthly,nationallyrepresentative surveyconductedby the CensusBureauto measureattributesof the laborforce. Duringthis period, the sample for the March survey consisted of approximately 50,000 households per year. My sample civilian adultsages 25 to 54, the ages most commonly includesnoninstitutionalized used in labor force studies (except in the longitudinal analysis, where I include women ages 18 to 64 to increasethe samplesize). Although these samplesarelarge, I pool several years of data at each point to decrease fluctuations due to random variation.I use the person weights providedby the CPS. Until 1993, the CPS asked the "majoractivity" question of each household member: "What was X doing most of last week?" The categories offered were "working,""looking for work," "keeping house," "going to school," "unableto

244

GENDER & SOCIETY / April 2004

I code all those who answered "keepinghouse" as working work,"and "retired." and include "keepinghouse" in the new analysis of occupationalsegregation. Slightly morethan 10 percentof people listed as keeping house in the CPS were also listed as "inthe laborforce"on the CPS employmentstatusvariable.Presumably,these people were keepinghouse as theirmajoractivitybutalso gave information that led them to be coded as in the labor force. I include all those who chose "keepinghouse"in this category,regardlessof theiremploymentstatus,as this represents their self-described majoractivity. To measurethe division of labor,I estimate occupationalsegregationusing the index of dissimilarity(Blau, Simpson, and Anderson 1998). The numberreflects the percentageof either men or women who would have to change occupationsto of men andwomen acrossoccupations.Jacobs(1989) achieve an equaldistribution and Baunach(2002) made a persuasivecase for using more elaboratemeasuresof and isolation). However,my objectiveis less segregation(indexes of concentration to arriveat the most precise measureof segregationthanit is to show the difference once houseworkersare included, which is unlikely to be affected by the choice of index. Because the CPS has smaller samples than the decennial censuses used by others, and small occupations are more likely to be segregated by chance (Cotter et al. 1997; Wells 1999), I limit my analysis to the largest 100 paid occupationsat each time point. However,the resultswere nearlyidentical when I used 200 occupations instead. To analyze how the movement of women from housework into paid occupations, and vice versa, affects the gender division of labor, I take advantageof the one-year longitudinal propertyof the CPS. Each household in the CPS is interviewed for four consecutive months,then misses eight months,and finally is interviewed againfor four more months(U.S. Census Bureau2000). As a result,households interviewed in March are supposed to be interviewed in the following March's survey. The matching across the two years is not perfect, however.For example, household composition may change, and people may move or die between Marchsurveys;thereare also nonresponseand recordingerrors.Furthermore, it is impossible to know for sure that matches are perfect (for example, if a personis divorcedand then marriesanotherspouse of the same age in the intervening year). However,using fairlyrestrictivecriteria,it is possible to achieve most of the eligible matcheswith a high degreeof confidence(MadrianandLefgren 1999). I conductthe longitudinalanalysisusing the 1991 to 1993 MarchCPS, combining three years of transitionsto increase sample size. Individualsare matchedby householdnumberand individualline numberandincludedif they arethe same sex and race/ethnicityand have aged between <1 and 3 years in the interveningyear. These criteria,developed by Madrianand Lefgren (1999), representan attemptto balance the tradeoffs of false matches and false rejections.4This method yields 2,353 women who were keeping house in one year and were employed the next year. Finally, I do a simple decompositionof the trendin desegregation.First,I estimate the trend in segregationlevels holding constant the proportionof men and

Cohen / GENDER DIVISION OF LABOR

245

women keeping house, while letting the distributionof men and women over the largest 100 paid occupationschange throughthe period. Second, I repeatthe exerof men and women in the 100 largestoccupacise holding constantthe distribution tions and letting only the proportionkeeping house change over time. Severalcaveatsarein order.The most obvious problemis the strictassumptions I impose. For everyone to be in one and only one occupation,no one may be both keeping house and in a paid occupation.Thus, this method cannotcapturethe division of laborwithin householdsbeyond a simple indicatorof whethera respondent is keeping house. Of course, the fact that women shouldermost of the housework even when they do paid workis itself a sourceof genderinequality(Hochschildand Machung 1989). Women's higher rates of paid employment have in fact contributedto an increasein theirshareof all work.Forexample, Sayer (2001) showed that in paid and unpaidwork combined, Americanwomen worked 11 minutes less per day than men did in 1975, but they worked 26 minutes more in 1998. However, women's houseworkhours still droppedby 42 percentfrom 1965 to 1995, as total time spent on housework declined 21 percent. Furthermore, men who work prefor contribute some and their housework time also housework, dominantly pay doubled from 1965 to 1995 (Bianchi et al. 2000, 208). Some occupational segregation studies (e.g., Blau, Simpson, and Anderson 1998) makethe opposite mistake,includingthe paid occupationsof even part-time workers,most of whom spendmore time on houseworkthanon paidwork.According to data from the National Survey of Families and Households, women in the labor force part-time(less than 20 hours per week) in the early 1990s averaged more than 35 hours per week of unpaid household work (Liana Sayer, personal communication).In fact, a significantportion(6 percentin 1992-1993) of currently employedworkerslisted "keepinghouse"as theirmajoractivity.Thus,the assumption in existing occupationalsegregationstudiesthatlabormarketoccupationoverrides houseworkerstatusis questionableas well. The problemof overlapbetween keeping house and paid occupationsis not differentin principlefrom the problemof people working in multiplepaidjobs, who have been coded into one occupation in previous segregation studies. I conclude that identifying some houseworkersis better than excluding all of them from the analysis. While imperfect, this method may still serve as a correctiveto existing occupationalsegregationstudies-moving us in the directionof an estimateof the overall gender division of labor. An additionalproblemcould result from social desirabilityin the identification with keeping house. Folbre and Abel (1989, 548), in a review of databefore 1940, noted that some women may have self-identified as housewives because "traditionalpatriarchal normsattachedsome stigmato marriedwomen who relinquished theirprimaryidentity as housewife."However,they did not presentevidence conhousewife status,so it is impossible to evalufirmingthis suspicion of overreported ate. Furthermore, it is reasonableto expect thatthe stigma of not being a housewife diminishedin subsequentdecades as women's employment became more normative. One womaninterviewedin 2002 for a separatestudy on women's work/family

246

GENDER & SOCIETY / April 2004

histories said of her decision to enterthe laborforce years earlier,"Idid feel thatit thatyou stay home with your kids. But thereis the outsidepreswas very important sure to be productive.You'renot productiveif you are a housewife." RESULTS Table 1 shows the percentageof working women and men who were keeping house from 1972 to 1993, at approximately10-year intervals.Because the CPS changed occupationcoding schemes in 1983, and I include pairsof years thatuse the same schemes, I had to use 1981-1982 insteadof 1982-1983. For comparison, FolbreandNelson (2000, 125-26), underthe assumptionthat85 percentof women age 16 or older withoutpaidjobs are full-time homemakers,arrivedat an estimate for 1990 thatis within 5 percentagepoints of the CPS figure (33 percentversus38 percent)when I use the same age range (16 and older). Table 2 shows the 15 occupationswith the largest numberof women in 19921993, rankedby percentagefemale in the occupation.Clearly,keeping house was the largestoccupationfor women as late as the early 1990s, and it was also among the most segregated,at 96.5 percentfemale. Note thateven when women in thepaid labor market work in "reproductivelabor"occupations (Glenn 1992)-such as waitresses, nurses'aides, cooks, or teachersof young children-their occupations are less segregatedthan household labor. The basic results arepresentedin Table3, which shows the index of dissimilarity, or gendersegregation,with and withoutincludingkeeping house as an occupation.6First, segregationis more pronouncedwhen keeping house is includedas an occupation, droppingonly to 62.6 by the end of the period. Second, because the proportionof women keepinghouse declines duringthe period,its inclusion in the calculationmakesless differencein lateryears.Finally,the decline in segregationis somewhat steeper-19.7 points versus 16.1-when keeping house is included. RevolvingDoors Wherearethe formerhouseworkers going when they enterpaid work?Women's out of into and female-dominated measureof mobility occupationsis an important barriersupholding segregationover time. Including houseworkersin an occupational analysis broadensthe pictureof women's mobility with regardto the segregation of work. Researchersfrom the 1940s (Kyrk 1947) to the present(Budig and England 2001) have analyzed the effect of women's spells outside the paid labor force, but the data used for such studies count houseworkingwomen as out of the laborforce. We can addressthis questionby using the longitudinalpropertyof the CPS to trackwomen from one year to the next. For all women workers,and those who changed occupationsfrom one year to the next, I show the distributionacross male-dominated,balanced, and femaledominatedoccupations in Table4.7 For occupation changers,the table shows the

Cohen / GENDER DIVISION OF LABOR

247

TABLE1: Percentage of Women and Men Keeping House, 1972 to 1993 Year 1972-1973 1981-1982 1992-1993 Women 52.8 38.9 26.2 Men 0.2 0.4 1.0

NOTE: Civilian adults ages 25 to 54 (weighted),eitheremployedor "keepinghouse." TABLE 2: Occupations with 1 Percent or More of All Women, 1992/1993 Occupation Secretaries Receptionists Keeping house Registered nurses Bookkeepers,accounting,and auditingclerks Nursingaides, orderlies,and attendants Teachers, elementaryschool Cashiers Waitersand waitresses Administrative supportoccupations,n.e.c. Teachers,secondary school Cooks (not privatehousehold) Accountantsand auditors sales occupations Supervisorsand proprietors, n.e.c. Managersand administrators, Percentage Female 99.0 97.9 96.5 94.5 91.7 91.1 86.7 79.8 79.4 74.8 57.2 51.1 50.6 34.4 31.2 Percentage of All Women Working 5.2 1.0 26.2 2.6 2.3 2.0 2.4 1.7 1.1 1.5 1.1 1.1 1.1 2.1 3.0

NOTE: Women,ages 25 to 54 weighted.n.e.c. = not elsewhere classified. TABLE 3: Occupational Gender Segregation, with and without Keeping House Year 1972-1973 1981-1982 1992-1993 Totalchange Without KeepingHouse 70.2 63.1 54.1 -16.1 WithKeeping House 82.3 73.5 62.6 -19.7 Difference 12.1 10.4 8.5 -3.6

NOTE: Civilians calculatedfromthe 100 largestpaidoccupations. ages 25 to 54 (weighted),

compositionof theirdestinationoccupations.As expected, the female workforceis more skewed toward female-dominated occupations when houseworkers are included. Whether houseworkersare included or not, the table also shows that women who change occupations end up in less segregated occupations than the

248

GENDER & SOCIETY / April 2004

TABLE 4: Percentage Female in CurrentOccupation, 1991 to 1993 Male Female Dominated, Dominated, <30% 30%-69% 70+% Allwomen Excludinghouseworkers houseworkers Including Occupationchangers Nonhouseworkers and nonhouseworkers Houseworkers Formerhouseworkers NOTE: Civiliansages 18 to 64 (weighted).

Mean

9.9 7.2 11.7 10.2 11.4

37.4 27.1 42.0 35.3 31.7

52.7 65.7 46.3 62.7 56.9

66.4 74.7 63.3 69.5 68.1

general populationof women workers.Thatis, excluding houseworkers,52.7 percent of all women are in female-dominatedoccupations,comparedto 46.3 percent of those who changedoccupations.Whenhouseworkersareincluded,65.7 percent of all women are in female-dominatedoccupations, comparedto 62.7 percentof those who changed occupations. The most importantrow of Table 4, however, shows the distributionof former houseworkers in their new occupations. Although less heavily skewed toward female-dominatedoccupations than all women workers, they are more concentratedin segregatedoccupations than women who moved between paid occupations (56.9 percentversus46.3 percent),andtheiroccupationsaverage68.1 percent female, comparedto 63.3 percentfor women who changed paid occupations.The formerhouseworkersare also somewhatmore segregatedthanthe totalpopulation of nonhouseworkers. Thus, women leaving the keeping house occupationaremore concentratedin female-dominatedoccupationsthanotherwomen in the paidlabor force. When houseworkersare included,therearealso a substantialnumberof women who moved from paid occupations back to keeping house (for example, women who have marriedor had children). In previousanalyses of occupationalsegregation, they arecountedas having left the workforce.I find thatwhen only paidwork is counted, 25.3 percentof women who left male-dominatedoccupations moved into female-dominatedoccupations the next year-what Jacobs (1989) called the "revolvingdoor."However, when houseworkersare counted, the numberrises to 36.4 percent, a significant increase in the rate at which women move back into segregatedwork (not shown). Anotheruseful indicatorof mobility is the correlationbetween the gendercomposition of women's former versus currentoccupations after they change jobs. Jacobs (1989) arguedthata low correlationis one indicationof potentialopenness in the labormarket,showingthatwomen may move in andout of female-dominated occupations.I find thatwithoutkeepinghouse includedas an occupation,the correlation between the percentage female in women's previous occupation and their

Cohen / GENDER DIVISION OF LABOR

249

TABLE 5: Decomposition of Occupational Segregation Trends

1972-1973 1981-1982 1992-1993

Observedlabormarkettrend Observedtrendincluding keeping house Change in paidworkonly Change in keeping house only NA= not applicable. NOTE:

70.2 82.3 NA NA

63.1 73.5 78.3 78.0

54.1 62.6 74.5 74.7

currentoccupationis .21. Once keeping house is included, the correlationrises to .28. Thus,the inclusion of women as unpaidhouseworkersin the laborforce reveals greaterbarriersto women's mobility.8 TwoPathsto Desegregation If enteringthe paid labormarketand moving out of female-dominatedoccupations both contribute to declines in the overallgenderdivision of labor,what arethe relative contributions of these two paths to desegregation? The results of the decomposition analysis are shown in Table 5. Table5 shows thatfor each 10-yearinterval,women leaving the keeping house occupation and the desegregationof paid occupations contributedequally to the overallerosion of the genderdivision of labor.Thus, althoughthe level of segregation remainshigh at the end of the period, by these measures,women achieved as much desegregationof work from enteringthe paid laborforce as they did from the has not been captured changingcompositionof paidoccupations.This contribution in previous studies of occupationalsegregation. CONCLUSION I have arguedthat we need to take housework into account when trackingthe overallgenderdivision of labor-including both paid andunpaidwork. This analysis is a first attemptto do that.The results show thatthe inclusion of houseworkers as an occupationaffects measuresof gendersegregationin two ways. First,because the houseworkeroccupationis large and predominantlyfemale, estimates of gender segregation are higher when these women are included. Second, because women enteringthe formallaborforce areon averageenteringoccupationsthatare less segregatedthanhousework,the rate of decline in gender segregationis somewhat steeper over the period once houseworkersare included. However, the outcome is not a simple one, because formerhouseworkersmove into more segregated occupations than women who are already in the labor market and because some women reenterthe houseworkeroccupation when they leave the paid labor force. Finally, a decomposition of the trends shows that, over time, women leaving

250

GENDER & SOCIETY / April 2004

housework as an occupationcontributedas much to the overall decline in gender segregationas did the desegregationof paid occupations. Although we have many studies of trendsin the gender segregationof occupations (e.g., Blau, Simpson, and Anderson 1998; Cotter et al. 1995; Reskin 1993; Wells 1999), we do not have empiricalassessmentsof the overallgenderdivisionof laborin the United Statesor otherindustrialsocieties. This is disappointingconsidering the importance of the division of labor to theories of gender inequality (Chafetz 1991). Not only do existing occupationalsegregationstudiesof paidwork understatethe overall gender division of labor;they also cannot capturehow the shifting location of women's work contributesto the trendin the division of labor. If, as Reskin argued, segregationis "a fundamentalprocess in social inequality" (1993, 241), then some of its decline shouldbe creditedto the movementof women into paid work. NOTES
1. Houseworkfigures are calculatedfrom 1995, as reportedby Bianchi et al. (2000, Table 1); labor marketfigures are my calculationsfrom the 1995 MarchCurrentPopulationSurwy. 2. A similarsituationexists in studiesof occupationalprestige,which generallyexclude housework. In one exception, Bose andRossi (1983) showed thatwhen "housewife"was includedamong 110occupations in a 1972 survey, housewives' prestige scores (51) were lower than some female-dominated occupationsnot associated with housework,such as practicalnurse (56.4) and privatesecretary(60.9), but much higherthansome female-dominated occupationsdoing similarwork, such as waitress(24.4), short ordercook (21.5), and babysitter(18.3). 3. The instructionsto enumerators for each census have been compiled by the Integrated PublicUse MicrodataSeries (http://www.ipums.org). 4. Jacobs(1989) used a simplermethod,which relies on a retrospectivequestionaboutoccupationin the previous year, which he comparedwith respondents'occupation in the previous week. However, because the "keepinghouse"occupationis derivedfromthe current"majoractivity"question,thereare no retrospectivedata on this occupationin the cross-sectional data. 5. The CurrentPopulationSurveychangedthe occupationcoding scheme in 1992. I recode a small numberof occupationsto allow occupationmatchingacross the years 1991 to 1993. 6. Forcomparison,using detailedoccupationsin the decennialcensus, Blau,Simpson,andAnderson (1998) found segregationindexesof 67.7 in 1970, 59.3 in 1980, and53.0 in 1990;Baunach(2002) found indexes of 60.8 in 1980 and 50.0 in 1990. 7. For a discussion of these cut-pointsin occupationpercentagefemale, see Jacobs (1989). 8. The correlationJacobs(1989) foundusing 1981 Current PopulationSurveydatais only . 11.But in additionto the time period difference,his method is not comparable(see note 4).

REFERENCES
Batalova, Jeanne A., and Philip N. Cohen. 2002. Premaritalcohabitationand housework:Couples in cross-nationalperspective.Journal of Marriage and the Family 64 (3): 743-55. Baunach,Dawn M. 2002. Trendsin occupationalsex segregationand inequality,1950 to 1990. Social Science Research 31 (1): 77-98.

Cohen / GENDER DIVISION OF LABOR

251

Berk, Sarah F 1985. The genderfactory: The apportionmentof work in American households. New York:Plenum. Bianchi, SuzanneM., Melissa A. Milkie, LianaC. Sayer, and John P. Robinson. 2000. Is anyone doing the housework?Trendsin the gender division of household labor.Social Forces 79 (1): 191-228. Blau, Francine D., Patricia Simpson, and Deborah Anderson. 1998. Continuingprogress? Trends in occupationalsegregationin the United States over the 1970s and 1980s. FeministEconomics 4 (3): 29-71. Bose, ChristineE., and Peter H. Rossi. 1983. Gender and jobs: Prestige standings of occupations as affected by gender.AmericanSociological Review 48 (3): 316-30. AmericanSociological Budig, Michelle J., andPaulaEngland.2001. The wage penaltyfor motherhood. Review 66 (2): 204-25. Chafetz, Janet S. 1991. The gender division of labor and the reproductionof female disadvantage: Towardan integratedtheory.In Gender, family, and economy: The triple overlap, edited by Rae L. Blumberg.NewburyPark,CA: Sage. Chang,MarikoL. 2000. The evolution of sex segregationregimes.AmericanJournal of Sociology 105 (6): 1658-701. Cohen, Philip N. 1998. Replacinghouseworkin the service economy: Gender,class, and race-ethnicity in service spending.Gender & Society 12 (2): 219-31. Cohen, Philip N., and SuzanneM. Bianchi. 1999. Marriage,children,and women's employment:What do we know? MonthlyLabor Review 122 (12): 22-31. Cohen, Philip N., and MattL. Huffman.2003. Individuals,jobs, and labormarkets:The devaluationof women's work.AmericanSociological Review 68 (3): 443-63. Coltrane,Scott. 2000. Researchon household labor:Modeling andmeasuringthe social embeddedness of routinefamily work. Journal of Marriage and the Family62 (4): 1208-33. Cotter,David A., JoAnn DeFiore, Joan M. Hermsen, Brenda M. Kowalewski, and Reeve Vanneman. 1997. All women benefit: The macro-level effect of occupationalintegrationon gender earnings equality.AmericanSociological Review 62 (5): 714-34. and Occupations22 (1): 3-21. . 1995. Occupationalgender desegregationin the 1980s. Work Cotter,David A., JoanM. Hermsen,and Reeve Vanneman.2001. Women'swork and working women: The demandfor female labor.Gender & Society 15 (3): 429-52. Cowan,RuthS. 1983. More workfor mother:Theironies of householdtechnologyfromthe open hearth to the microwave.New York:Basic Books. women in the laborforce.AmericanJournalof Sociology 52 (3): 217-23. Durand,JohnD. 1946. Married Folbre, Nancy, and MarjorieAbel. 1989. Women's work and women's households: Genderbias in the U.S. census. Social Research56 (3): 545-69. Folbre,Nancy, and J. A. Nelson. 2000. For love or money-or both?Journalof Economic Perspectives 14 (4): 123-40. Folbre,Nancy,andB. Wagman.1993. Countinghousework:New estimatesof realproductin the United States, 1800-1860. Journal of Economic History 53 (2): 275-88. Glenn, Evelyn Nakano. 1992. From servitudeto service work:Historicalcontinuitiesin the racial divilabor.Signs: Journal of Womenin Cultureand Society 18 (1): 1-43. sion of paid reproductive the gender gap: An economic history of Americanwomen. New Goldin, Claudia. 1990. Understanding York:Oxford UniversityPress. Heidi. 1981. The unhappymarriageof Marxismand feminism:Towardsa more progressive Hartmann, and revolution:A discussion of the unhappymarriage of Marxismand Feminism, union. In Women edited by Lydia Sargent.Boston: South End. Hochschild, Arlie R., and Anne Machung. 1989. The second shift. New York:Avon Books. Huber,Joan. 1991. A theoryof family,economy, andgender.In Gender, family, and economy: Thetriple overlap, edited by Rae L. Blumberg.Newbury Park,CA: Sage. Jacobs,JerryA. 1989. Revolvingdoors: Sex segregationand women's careers. Palo Alto, CA: Stanford University Press.

252

GENDER & SOCIETY / April 2004

Kyrk,Hazel. 1947. Who works and why. Annals of the AmericanAcademyof Political and Social Science 251:44-52. Lorber,Judith. 1994. Paradoxes of gender.New Haven, CT: Yale UniversityPress. Madrian,Brigitte C., and LarsJ. Lefgren. 1999. A note on longitudinallymatchingCurrentPopulation Survey respondents.Technicalworking paper247. National Bureauof Economic Research,Cambridge, MA. Miller,Joanne,andHowardH. Garrison.1982. Sex-roles:The division of laborat home andin theworkplace. Annual Review of Sociology 8:237-62. Nelson, RobertL., and William P. Bridges. 1999. Legalizinggender inequality: Courts, markets,and unequalpay for women in America. Cambridge,UK: CambridgeUniversityPress. ValerieK. 1970. Thefemale laborforce in the UnitedStates:Demographicand economic Oppenheimer, Studies, factors governingits growthand changingcomposition.Berkeley:Instituteof International Universityof California. Presser,HarrietB. 1999. Demographics:Towarda 24-hour economy. Science 284 (5421): 1778-79. Reskin, Barbara.1993. Sex segregationin the workplace.Annual Reviewof Sociology 19:241-70. eds. 1986. Women's work,men'swork:Sex segregationon the Reskin,Barbara F., andHeidi I. Hartmann, job. Washington,DC: National Academy Press. Sayer, LianaC. 2001. Trendsand genderdifferencesin men's and women's time use: The relationship workandtime for leisure.Paperpresentedat the annualmeetbetweentime in marketandnonmarket ings of the PopulationAssociation of America, Washington,DC. Stacey,Judith.1993. Good riddanceto the family:Response.Journalof Marriageand the Family55 (3): 545-47. Strasser,Susan. 1982. Never done: A history of Americanhousework.New York:Pantheon. U.S. Census Bureau.2000. CurrentPopulationSurveytechnical paper 62: Design and methodology. Washington,DC: U.S. Census Bureau. Uttal, Lynet.2002. Makingcare work:Employedmothersin the new childcaremarket.New Brunswick, NJ: RutgersUniversityPress. Waring,Marilyn. 1988. If women counted:A newfeminist economics. San Francisco:Harper& Row. Uni1999. Countingfornothing: Whatmenvalueand whatwomenare worth.Toronto,Canada: .. versity of TorontoPress. Wells, Thomas. 1999. Changesin occupationalsex segregationduringthe 1980s and 1990s. Social Science Quarterly80 (2): 370-80.

Philip N. Cohenis an assistantprofessor of sociology at the Universityof California,Irvine.His researchinvolves the relationshipbetweenfamily structureand inequalitywithinand between families. He also studiesmicro-macrolinkagesin social inequality,includingthe effectsof labor marketracial/ethnic compositionand inequalitywithinand betweenjobs.