gender differences in workplace authority: discrimination and the role of organizational leaders

16
http://asj.sagepub.com/ Acta Sociologica http://asj.sagepub.com/content/41/2-3/99 The online version of this article can be found at: DOI: 10.1177/000169939804100201 1998 41: 99 Acta Sociologica Mia Hultin Organizational Leaders Gender Differences in Workplace Authority: Discrimination and the Role of Published by: http://www.sagepublications.com On behalf of: Nordic Sociological Association can be found at: Acta Sociologica Additional services and information for http://asj.sagepub.com/cgi/alerts Email Alerts: http://asj.sagepub.com/subscriptions Subscriptions: http://www.sagepub.com/journalsReprints.nav Reprints: http://www.sagepub.com/journalsPermissions.nav Permissions: http://asj.sagepub.com/content/41/2-3/99.refs.html Citations: What is This? - Jan 1, 1998 Version of Record >> at COLUMBIA UNIV on November 5, 2014 asj.sagepub.com Downloaded from at COLUMBIA UNIV on November 5, 2014 asj.sagepub.com Downloaded from

Upload: m

Post on 10-Mar-2017

213 views

Category:

Documents


1 download

TRANSCRIPT

Page 1: Gender Differences in Workplace Authority: Discrimination and the Role of Organizational Leaders

http://asj.sagepub.com/Acta Sociologica

http://asj.sagepub.com/content/41/2-3/99The online version of this article can be found at:

 DOI: 10.1177/000169939804100201

1998 41: 99Acta SociologicaMia Hultin

Organizational LeadersGender Differences in Workplace Authority: Discrimination and the Role of

  

Published by:

http://www.sagepublications.com

On behalf of: 

  Nordic Sociological Association

can be found at:Acta SociologicaAdditional services and information for    

  http://asj.sagepub.com/cgi/alertsEmail Alerts:

 

http://asj.sagepub.com/subscriptionsSubscriptions:  

http://www.sagepub.com/journalsReprints.navReprints:  

http://www.sagepub.com/journalsPermissions.navPermissions:  

http://asj.sagepub.com/content/41/2-3/99.refs.htmlCitations:  

What is This? 

- Jan 1, 1998Version of Record >>

at COLUMBIA UNIV on November 5, 2014asj.sagepub.comDownloaded from at COLUMBIA UNIV on November 5, 2014asj.sagepub.comDownloaded from

Page 2: Gender Differences in Workplace Authority: Discrimination and the Role of Organizational Leaders

Gender Differences in Workplace Authority:Discrimination and the Role of OrganizationalLeaders

Mia HultinSwedish Institute for Social Research, Stockholm

ABSTRACTThis paper examines to what extent discrimination accounts for inequalities in authority exertionbetween women and men in the Swedish labour-market. Processes governing authorityattainment are studied in terms of human capital and family responsibilities as well as of thehorizontal sex segregation in the labour-market. The empirical results strongly indicate thatwomen are being unduly restricted from attaining supervisory positions at work, primarilywithin the private sector of the economy. The assumption that discrimination is brought aboutby decision-makers within work organizations was tentatively tested and proved not to hold,since it was determined that neither women’s nor men’s chances to reach higher supervisorypositions are affected by the sex of the highest workplace manager. The analyses are based upondata from the 1991 Swedish Level of Living Survey and the 1991 Swedish Establishment Surveyon a sample of 2017 employees.

Mia Hultin, Swedish Institute for Social Research, Stockholm University, S- 106 91Stockholm, Sweden© Scandinavian Sociological Association 1998

1. Introduction

The patterns of women’s work-related beha-viour have changed, during recent decades, ina number of ways in Sweden, as well as in mostother Western societies. Women have come tomake up approximately half of the Swedishlabour force, and there has been an increase, ifa modest one, in the number of women in male-dominated occupations and educational fields.However, it is well established that despitechanges of this kind, women still tend to be

disadvantaged regarding various job rewards.Swedish women earn less than men, net of a

range of individual and structural factorsrelevant for wages (e.g. le Grand 1991,1994), and they also occupy fewer highpositions within authority hierarchies in workorganizations (Mueller et al. 1994; Roman1994; Statistics Sweden 1992; Wright et al.

1995). ’

Results from a recent comparative studyindicate that the authority differential betweenSwedish women and men (net of, for example,

formal qualifications) is relatively large com-pared to the gender gaps found in Australia,Canada, the United States and the UnitedKingdom (Wright et al. 1995). This pattern isthe reverse of that which has generally beenfound in gender wage differentials in Swedenrelative to, for example, Canada and the UnitedStates (Rosenfeld & Kalleberg 1990).1

The present paper examines to what extentSwedish women and men’s unequal chances toreach authority positions can be attributed todiscrimination. Since gender discrimination isdifficult to measure directly by means of large-scale quantitative data sets, conclusions on theincidence of discrimination almost inevitablymust be drawn from inference. Explanations ofdifferences in authority exertion between menand women which emphasize the role ofdiscrimination are supported to the extent

that the authority gap cannot be accountedfor by gender differences in the effects ofrelevant career-related characteristics (seeHalaby 1979; Jaffee 1989; Wright et al. 1995for conclusions in line with this argument) . 2

at COLUMBIA UNIV on November 5, 2014asj.sagepub.comDownloaded from

Page 3: Gender Differences in Workplace Authority: Discrimination and the Role of Organizational Leaders

100

The data material used in the presentstudy permits a simple yet direct test of the

assumption that discrimination is broughtabout by decision-makers within work organi-zations. If gender differences in authorityattainment are generated partly by unjusttreatment of female employees in the work-place, it seems reasonable to assume that thesex of the highest workplace manager affectsemployees’ chances to reach authority posi-tions. The empirical analyses will thereforeaddress the question of whether women’schances to exert supervisory authority are

relatively more depressed in workplaces wherethe highest workplace manager is a man. Theanalyses are based upon individual-level datafrom the 1991 Swedish Level of Living Survey,and upon organizational-level data from the1991 Swedish Establishment Survey.

2. Workplace authority - defmition andrelevance

To have authority is to have some degree oflegitimate control over other people’s work.This relation to other employees’ work is afeature that distinguishes authority from

autonomy, which is a control dimension con-cerning the worker’s relation to his or her ownwork (Wolf & Fligstein 1979a). Supervisoryauthority refers to the extent to which anindividual has the work organization’s mandateto guide and control the work of others. Hence,if women face relatively more restrictions in theauthority attainment process and thereforehave limited access to resources which reachbeyond their own work, women’s abilities toaffect the organization of work can be seriouslylimited.

Studies of gender differences in authorityhave relevance for the problem of genderinequality at work in several ways. Authoritycan be considered a valued job attribute in thatit confers financial rewards and status, and alsobecause it can be intrinsically rewarding.3 3Furthermore, positions in authority hierarchiesare associated with power, which implies thatgender differences in authority are not onlyoutcomes of gender inequality at work, but thatthey can also be regarded as a significant causeof inequalities and their persistence (Wright etal. 1995). The idea that&dquo; women’s limited accessto positions of authority should be treated as animportant mechanism behind discriminationhas been empirically substantiated in a recent

Swedish study on gender wage discrimination(Hultin & Szulkin 1997). In that study,women’s wages were found to be lower in

workplaces in which comparatively more of themanagers are men than in workplaces in whichthe male representation in management is

relatively poor.Moreover, it can be argued on meritocratic

or efficiency-related grounds that valuable

competence and resources go to waste in

society to the extent that women with similarqualifications and career aspirations to those ofmen are unduly barred from workplace author-ity. In a meritocratic society, individuals are tobe judged by their achievements, not by theirascribed characteristics. In line with such

reasoning, gender differentials in authoritythat are not attributable to gender disparitiesconcerning relevant qualifications and aspira-tions can have negative consequences for thefunctioning of society as a whole.

3. Explanations of the gender gap inauthority

Below, different plausible explanations of

authority differentials between women andmen are discussed. The factors emphasized inthe explanations often put forward in theliterature can be broadly categorized as indivi-dual-level factors and organizational-level fac-tors. Generally, individual-level explanationsare stated within a non-discriminatory theore-tical framework. Such approaches are normallybased upon a firm confidence in individuals’

voluntary choices and prospects to shape theirown opportunities. Perhaps the most promi-nent individual-level approach to inequality injob rewards is offered by human capital theory.Explanations of gender inequality at workderived from human capital theory will beelaborated in the first section. Researcherswithin an organizational or a structural fieldhave gone beyond the individual level, focusinginstead on what restrictions and opportunitiesare posed to the individual employee by thesurrounding organizational milieu. It is, for

example, primarily agents on the organiza-tional level who have the capacity to makedecisions which potentially bring about discri-mination against female employees. Hence,researchers within the organizational field

primarily stress the influence on reward attain-ment processes of factors beyond employees’own reach. Discriminatory processes will there-

at COLUMBIA UNIV on November 5, 2014asj.sagepub.comDownloaded from

Page 4: Gender Differences in Workplace Authority: Discrimination and the Role of Organizational Leaders

101

fore be discussed primarily within the frame oforganizational-level explanations, in the secondsection.

Individual-level explanations of the gendergap in authorityGender differences in individual qualifications andfamily responsibilities. One of the implications ofhuman capital theory is that there existdifferences between women and men in accu-mulated qualifications and other individualtraits relevant for labour-market . outcomes(Becker 1964). Women’s limited representationin authority hierarchies is therefore primarilydue to their disproportionately low numbersamong the suitable candidates for supervisorypositions. Compositional gender differences in,for instance, education, labour-market experi-ence and tenure are attributed primarily to thedivision of labour in the household (Becker19 8 5; Mincer & Polachek 19 74). Since womenhave had primary responsibility for the familyand the household, they have a weakerattachment to the labour-market and feweropportunities to invest in human capital withpotential value in the labour-market.

In addition to influencing acquired quali-fications, family responsibilities also affectcareer chances more directly. Managerial posi-tions often require working at least full-timeand being flexible in terms of business travellingand working overtime. Interviews with employ-ers have confirmed the importance of managersnot letting family responsibilities intrude uponwork demands (Edlund et al. 1990). Sincewomen still have the main responsibility for thefamily, it is expected that being married,cohabiting, or having children present in thehousehold reduces women’s chances of reach-ing supervisory positions.4 For men, researchhas, in general, indicated that the effects onauthority of having a family are either absent orpositive. Gender-differentiated findings of thiskind could indicate both that women withfamilies are less inclined than correspondingmen to strive for supervisory positions, and thatemployers consider women with family obliga-tions to be unsuitable for managerial positions.It may be that marriage and children signifystability, maturity and commitment in assess-ments of men’s suitability for exerting author-ity, while the same factors signal weakercommitment for female employees (Baron &

Bielby 1985; Hill 1980).5If women, compared to men, in general

have fewer qualifications in terms of work

experience and education, it seems reasonableto expect that men will attain higher levels ofauthority 6 However, the empirical merit of thisargument is far from impressive. Given thatdiscrepancies between women’s and men’seducational levels and labour-market experi-ence have diminished over time in Sweden aswell as in other Western societies, it has becomethat much more difficult to explain the gendergap in authority by such human capital factors.Empirical results from several studies indicatethat women are disadvantaged with respect toauthority even when they are compared to menwith similar qualifications and family responsi-bilities (Mueller et al. 1994; Wolf & Fligstein1979a; Wright et al. 1995). Such researchfindings are certainly compatible with the ideathat employers and other organizational deci-sion-makers tend to treat women unfairly whenassigning individuals to supervisory positions.

Also quite contrary to the predictions ofhuman capital theory, men have been found toacquire greater authority returns than womento various human capital investments (Edlundet al. 1990; Hill 1980; McGuire & Reskin 1993;Wolf & Fligstein 1979b). Consequently, menhave relatively more to win in authority termsfor each additional unit of human capitalinvestment they make. The finding thatwomen tend to possess less authority thanmen in spite of similar career-related attributesimplies that discrimination is at work. However,proponents of human capital theory have putforward alternative hypotheses to account forgender inequality in job rewards. These expla-nations are based upon assumed differencesbetween women and men regarding lessobservable career-related traits, such as workeffort and preferences vis-d-vis the working life.

Gender differences in effort and preferences. Becker(1985) has stipulated that women put lesseffort into their labour-market work and theirwork careers than do equally qualified men,mainly due to extensive family responsibilities. , IHowever, Bielby and Bielby (1988) defy Beck- (

er’s assumption of unobserved gender differ-ences in market productivity as rooted inwomen’s lower levels of work effort and

preferences for less demanding work. Theycome to the conclusion that women’s workeffort seems to be stronger compared to that ofmen with similar family situations and levels ofhuman capital Jacobs (1992) also empiricallychallenges the notion of women’s relatively lowlevels of work effort. In his study, no systematic

at COLUMBIA UNIV on November 5, 2014asj.sagepub.comDownloaded from

Page 5: Gender Differences in Workplace Authority: Discrimination and the Role of Organizational Leaders

102

gender differences in work commitment werefound. Obviously, there are good reasons to

question the empirical merit of human capitaltheory’s assumption of significant gender dif-ferences in less observable career-related traits.

However, even if women work as hard andas capably as men, there could neverthelessexist gender differences in preferences or

aspirations related to career outcomes such asauthority exertion. For example, Kanter (1977)found, in an American case study, that menindeed reported somewhat higher motivationfor promotion, but that they also had consider-ably greater opportunity to get promoted.Women, on the other hand, judged a numberof promotion outcomes as highly desirable, butat the same time as less likely to be attained.Kanter’s findings point to the importance ofinterpreting aspirations and preferences inrelation to perceived opportunities. Other

empirical studies also refute the assumptionthat women in general do not aspire to higherpositions within work hierarchies. Many menhave preconceptions about women’s unwilling-ness to hold authority positions. Womenthemselves often state that they are equallycompetent and ambitious to supervise as theirmale counterparts, but that they as women seemore severe structural and interpersonal obsta-cles to fulfilling these aspirations (SOU 1994:3;Wahl 1992)..

In the present study, it is possible partly toaccount for the individuals’ aspirations vis-d-visworking life through measures of educationalchoices, both in terms of their chosen fields ofeducation and of the total number of educa-tional years. In addition, the respondents’family situation can be assumed to indicategender differences in aspirations and earlier-made choices regarding working life. Inclusionof indicators of family situation in analyses ofauthority attainment can therefore improvecomparability between career orientations offemale and male employees, which, in turn,makes conclusions about potential discrimina-tion against women more reliable.

Organizational-level explanations of thegender gap in authoritySex segregation. The Swedish labour-market ishighly sex segregated in that women and men,to a great extent, work in different occupationsand sectors (Charles 1992; Jacobs & Lim 1992;Nermo 1996). Occupational sex segregationresults in depressed opportunities for women toreach supervisory positions (Jaffee 1989). In

addition, sex segregation implies limited accessto internal labour-markets and to their benefitsfor women (Baron & Bielby 1985; Bielby &Baron 1986). According to Baron and Bielby,occupational sex segregation entails not onlydifferences across occupations, but also divi-sions within occupations, so that men andwomen with similar qualifications performdifferent roles in their work. Such a ’hidden’system of segregation implies that women

occupy positions of lower rank within occupa-tions, and consequently are assigned to author-ity positions less often than men. The genderpenalty in the labour-market is thus double.First, women are disadvantaged since manytraditionally female occupational areas offer

relatively few opportunities for career advance-ment. Second, the opportunities for upwardmobility which women do have are oftenlimited compared to those of men with similarqualifications in the same occupation.

As mentioned above, traditionally femaleoccupations are often characterized by limitedcareer opportunities. However, when occupa-tional sex composition is accounted for, the sexcomposition in the workplaces has even furtherimplications for women’s chances to attain

positions of authority. Reskin and Padavic(1994) found that women in general havebetter chances of promotion in female-domi-nated workplaces. Such workplaces have

greater experience with female employees,which in turn implies less gender-stereotypingand improved possibilities of spotting talentedand suitable individuals. Correspondingly, Kan-ter (1977) claims that a workplace sex compo-sition skewed in men’s favour seriously reduceswomen’s authority and career chances. Whenwomen are in a minority, their access to

necessary resources and opportunities at work- such as the opportunity to benefit profession-ally from informal networks - are reduced. It

has also been found that men in male-dominated work environments are reluctantto have women in authority positions (Cock-burn 1991).

The Swedish labour-market is character-ized by a pronounced prevalence of internalcareer ladders, which, in a non-discriminatorycontext, would entail greater chances forwomen to reach supervisory positions in workenvironments in which women constitute a

relatively large share of the employees. Anotherfeature of the sex-segregated Swedish labour-market is that women and men differ greatlyregarding which employment sector they work

at COLUMBIA UNIV on November 5, 2014asj.sagepub.comDownloaded from

Page 6: Gender Differences in Workplace Authority: Discrimination and the Role of Organizational Leaders

103

in. Women make up a considerable share of thelabour force in the public sector, whereas mento a larger extent work in the private sector.Accounted for their overall distribution in thetwo sectors, women have fewer chances to

reach authority positions in the private as

opposed to the public sector (Statistics Sweden1992). Such findings, together with researchindicating that wage differences betweenwomen and men are more articulated in theprivate than in the public sector (e.g. Andersen& Tomaskovic-Devey 1995), lead us to expect alarger gender gap in authority chances in theprivate sector than in the public sector, evenafter accounting for sex composition at the

occupational and organizational levels.

Allocative discrimination. Sex segregation in thelabour-market can be a result both of self-selection processes on the part of employeesand of discrimination carried out by employersand managers in allocation processes (Jacobs1989; Reskin 1993; Tomaskovic-Devey 1993).However, labour supply mechanisms, such asindividual choices and preferences, ought to berelatively more salient in explanations ofhorizontal sex segregation, that is, segregationinto different occupations and sectors. When itcomes to explaining vertical sex segregation,that is, the greater concentration of women inlower levels of work hierarchies within occupa-tions and sectors, the individuals’ preferencesare of lesser importance since both women andmen in general value the autonomy and thefinancial rewards that higher positions withinwork hierarchies entail. Hence, one should

expect that the authority differential betweenwomen and men who have similar career-

related characteristics and who are situated insimilar labour-market segments reflects discri-minatory practices in sorting processes in thelabour-market. The discriminatory element inallocative processes pertains to women’s limitedaccess to positions within a work organizationeither at the time of entry into it, or in terms ofcareer advancement once hired.

What is the rationale behind discrimina-

tory practices among employers? One powerfulapproach to this problem is the theory ofstatistical discrimination (Phelps 1972). Phelpsexemplifies statistical discrimination as follows:

[T]he employer who seeks to maximize expectedprofit will discriminate against [...] women if hebelieves them to be less qualified, reliable, long-term, etc. on the average than [...] men, [...], andif the cost of gaining information about the

individual applicants is excessive. [S]ex is takenas a proxy for relevant data not sampled. (Phelps1972: p. 659)

Phelps stresses that this mechanism is even

applicable to employers and workers who

actually have no problem hiring and workingalongside (in this case) women. According tothe argument of statistical discrimination, awoman may be cut off from authority positionsirrespective of her qualifications insofar as

employers judge her on the basis of assump-tions about women generally being unsuitablecandidates for exerting authority. An employercan, for instance, act on the premise thatwomen’s quit rates are higher than men’s,which would imply the risk of augmented costsin finding and training new employees. Prob-abilistic reasoning of this kind thus reflectsrational considerations (however bounded) tothe extent that women are perceived as having,on average, lower marginal productivity for agiven line of work (Bielby & Baron 1986).However, since women and men have over-

lapping distributions for practically all assets ofimportance to organizational goals, treatingindividuals only as carriers of a certain

category’s average characteristics inevitablyleads to unequal treatment of individuals

(England & Farkas 1986).Theories of discrimination in general, and

the theory of statistical discrimination in

particular, are problematic in that they arevery difficult to test directly by means of

existing large-scale quantitative data sets.

However, if discrimination is perceived as a

more direct and active process involvingindividuals’ use of organizational power, the

assumption that women are discriminated

against in the authority attainment processcan be (at least tentatively) examined. The datamaterial used in the present study permits ananalysis of whether the sex of the highestmanager in the workplace influences employ-ees’ chances to reach supervisory positions. Ifdiscrimination is a result of an active processgoverned by powerful agents who have themotivation and capacity to treat other peoplewithin the work organization unfairly, womenmay face especially low chances to exert

authority in those workplaces in which the

highest manager is a man. The probability thata female workplace manager uses her organi-zational power to debar other women from

legitimate rewards in recruitment and promo-tion processes should reasonably be lower than

at COLUMBIA UNIV on November 5, 2014asj.sagepub.comDownloaded from

Page 7: Gender Differences in Workplace Authority: Discrimination and the Role of Organizational Leaders

104

the probability that a male workplace managerdiscriminates against women. A female organi-zational leader can be expected not to holdstrong prejudices against women occupyinghigher positions in hierarchies, since she herselfis a female leader. Furthermore, the presence ofa female workplace manager can be treated asan indicator of the decision-makers within theorganization being overall relatively favourablydisposed to assigning women to authoritypositions.

The above organizational features offer

explanations that imply the influence on theauthority attainment process of mechanisms

beyond the scope of individual employees’ ownactions. According to the reasoning offered byresearchers like Baron and Bielby (e.g. 1985),one such important mechanism is that womenend up in labour-market segments which entaillimited opportunities for advancement to

higher positions within organizational hierar-chies. However, as suggested, for example, byKanter (1977), the structural constraints

implied by sex segregation can partly becounteracted by another mechanism operatingon the organizational level. In those workenvironments within certain occupations in

which women are numerous, discriminatorytendencies are weaker than in workplaces inwhich women are in a minority The applic-ability of both stipulations will be evaluated inthe analyses which follow.

4. Data, variables and analytical strategy

The empirical analyses in this paper are basedon two interrelated sets of data. The source ofthe individual-level data is the 1991 SwedishLevel of Living Survey (LNU). The total sampleof that survey consists of approximately 6000individuals representative of the Swedish popu-lation, aged 18-75. The overall response ratefor the LNU was 79 per cent.8 8 The workestablishments that employed the LNU respon-dents, and which had a total of ten or moreemployees, are included in a second sample,which makes up the Swedish Establishment

Survey (APU). The APU contains data onvarious organizational features, obtained frominterviews with workplace managers. Theresponse rate of the APU was 93 per cent.9

9

The analyses below are conducted for 926women and 1091 men who were employed theweek before the interview and who fulfil theabove-discussed criteria.

Authority Supervisory authority is determinedby how many persons the respondent super-vises in his or her work. This variable is treatedmainly as a logarithmic continuous variableioin the analyses, but is, in some instances, usedas a dummy variable with the value 1,signifying that the respondent supervises at

least 20 per cent of the employees at the

workplace. Analyses with a number of categor-izations of the dependent variable have beenconducted. However, 20 per cent seemed to bean adequate breaking point, since there exists.areasonable representation of both women andmen above this level. Among those respondentswho are supervisors at even higher levels (i.e.those who supervise, for example, 30, 40 or 50per cent of the employees at the workplace), therepresentation of women is too small to

generate a target category that would be

meaningful to analyse by means -of multivariateregression techniques.

Individual qualifications. The employee’s educa-tion is included in the analyses with the help ofboth metric and nominal variables. In metricterms, education is assessed as the number ofeducational years. Furthermore, field of study iscontrolled for by means of a nominal variablewith ten different categories.ll Labour-marketexperience is measured directly by a variableindicating how many years the individual has. been working. 12 Furthermore, labour-marketexperience squared is included in the equationsto account for diminishing marginal returns.Tenure with current employer is measured innumber of years.13 3

Family responsibilities. Marital status is indexedwith a categorical variable on two levels,namely married/cohabiting and single, withthe latter as baseline category. The number ofchildren of the respondent or respondent’sspouse living in the household is indicated bya metric variable.

Horizontal sex segregation and the sex of thehighest workplace manager. Occupational sectoris assessed by a dummy variable which is coded1 if the respondent is employed in the privatesector and 0 otherwise. The share of women

among the employees at the establishment andthe share of women in the respondent’soccupation are both metric variables rangingfrom 0 to 1. The latter of these variables is

based upon a three-digit occupational distribu-tion by sex in the 1990 census. A dummy

at COLUMBIA UNIV on November 5, 2014asj.sagepub.comDownloaded from

Page 8: Gender Differences in Workplace Authority: Discrimination and the Role of Organizational Leaders

105

variable indicates whether or not the highestmanager at the workplace is a man. 14

Analytical strategy. The multivariate analyses ofwomen’s and men’s authority attainment areconducted by means of ordinary least squaresregression and logistic regression. The ordinaryleast squares regression analyses presented inthe first empirical part of the paper address twoimportant issues: (i) the magnitude of the

gender gap in authority attainment and (ii)gender-specific mechanisms in the authorityattainment process. The analyses of the gendergap in supervisory authority, and the potentialexplanation of the gap by individual qualifica-tions, family responsibilities and the horizontalsex segregation in the labour-market, make itpossible to assess the extent to which womenwith career-related characteristics and labour-market location similar to those of men are

unduly barred from authority positions withinorganizational hierarchies. Hence, these ana-lyses offer an indication of the potential scopefor discrimination against women in the

authority attainment process. ~~ sTechnically, the magnitude of the gender

gap is assessed through comparisons ofwomen’s and men’s adjusted intercepts in themultivariate models. These intercepts reflectthe predicted (logged) number of subordinatesfor a woman or a man with average values onall included predictors in the given model. 16

The gender gap in the number of subordinatesexpresses how many times more subordinatesmen with average characteristics have in

comparison to women with average character-istics

The next step in the empirical part of thispaper is to examine whether there exist gender-specific elements in the authority attainmentprocess, that is, to what extent women and menare differently rewarded in authority terms forsimilar career-related characteristics. The ana-

lyses are conducted for women and men

separately so that the coefficients, i.e. thereturns or pay-offs, can be compared forwomen and men.l8 Sector differences in theprocesses of women’s and men’s authorityattainment are studied by means of separateregressions for publicly- and privately-employedindividuals. This strategy is adopted becauseearlier research has shown that women’s

opportunities and job rewards in general arebetter in the public than in the private sector.

In the final part of the empirical section, itis examined whether women’s chances to exert

supervisory authority are relatively more

depressed in work organizations where the

highest workplace manager is a man. To detecta relationship between the sex of the highestmanager at the workplace and women’schances to exert authority, the analyticalprocedure seeks to maximize the applicabilityof the assumption that the incumbent of the

Table 1. Descriptive statistics (means and percentages) for variables indicating workplace authority, individual qualifications,family situation, horizontal sex segregation and the sex of the highest workplace manager.

* p:5 0.05 for the difference between women and men to be 0.

at COLUMBIA UNIV on November 5, 2014asj.sagepub.comDownloaded from

Page 9: Gender Differences in Workplace Authority: Discrimination and the Role of Organizational Leaders

106

Table 2. Gender-specific influences on authority exertion of individual qualifications, fainily situation (Model A), horizontalsex segregation (Model B) and the sex of the highest workplace manager (Model C). ’Ordinary least squaresregression of the naturallogarit1l111 of the number of subordinates for women and men separately. Unstandardizedregression coefficients and adjusted intercepts.

a Educational sector is controlled for in all models.b For calculations of the adjusted intercepts, see footnote 16.’For calculations of the gender gaps, see footnote 17.* p :5 0.05 for the estimate to be equal to 0.**p < 0.01 for the estimate to be equal to 0.t p ~ 0.05 for the estimates to be equal for women and men.

highest supervisory position has some substan-tial say over recruitment processes within theorganization. It seems reasonable to expect thatthe highest workplace manager has primaryinfluence over the recruitment of supervisors athigh levels within the organization. Therefore,it will be examined to what extent the sex of thehighest workplace manager has a bearing onthe employees’ chances to supervise at leastone-fifth of all employees in the workplace.

5. Empirical results

As shown in Table 1, men have, in general,more than twice as many subordinates as

women have. This disadvantaged situation forwomen does not seem to stand in proportion tothe generally rather small gender differencesbetween the employees’ individual qualifica-tions and family situations. As shown in thetable, women’s and men’s average numbers ofeducational years are the same. Some statisti-cally significant differences between womenand men in labour-market experience and

seniority exist, but are rather small. Thereare, however, some clear-cut features of the sex-segregated Swedish labour-market. Womentend to be employed in occupations in whicharound 75 per cent of the employees are

women, whereas men in general work in

occupations in which 75 per cent of the

employees are men. The sex segregation is alittle less pervasive at the workplace levelcompared to the occupational level, but isnevertheless substantial. Both women andmen work in organizations where, on average,only a third of the employees are of the oppositesex. The same pattern applies to the distribu-tion of women and men over the public andprivate sectors of the labour-market. Womentend to work in the public sector, whereas menare more often employed in the private sector.The sex segregation also pertains to the highestmanagerial level. More than 75 per cent of thefemale respondents are employed in workplaceswhich are led by a man. Men’s dominanceamong the highest workplace managers is evenmore salient in workplaces in which the malerespondents work.

at COLUMBIA UNIV on November 5, 2014asj.sagepub.comDownloaded from

Page 10: Gender Differences in Workplace Authority: Discrimination and the Role of Organizational Leaders

107

...: ~...~:: u.:~

The questions addressed in the analysespresented in Table 2 are (i) to what extent thechances to reach positions of authority differbetween similar women and men and (ii)whether the processes governing authorityattainment operate in different ways forwomen and men. It can be seen from _-modelA in Table 2 that women and men with similarindividual qualifications and family situationsscore very differently as regards authorityexertion. The gender gap calculated on thebasis of the adjusted intercepts (model A, inTable 2) indicates that men with averagecharacteristics have more than three times

more subordinates than have women withsimilar qualifications and family situations.Women’s disadvantages are somewhat mutedin model B, in which effects of the horizontalsex segregation are studied, but they are still

quite substantial. Still, men have more thantwo-and-a-half times more subordinates thanhave comparable women. Hence, women’slimited access to positions of authority is not,to any great extent, explained by the fact thatwomen tend to be situated in female-dominatedlabour-market segments. Neither is the genderdifferential in authority exertion influenced bythe sex of the highest workplace manager(model C, Table 2). In sum, women have

significantly fewer chances to reach positionsof authority than men who have similar

average qualifications and family situations,and who are located in similar labour-marketstructures. Accordingly, the fact that womenare left behind in the authority attainmentprocess cannot be explained by inferior quali-fications, more demanding family situations orby the condition that women tend to be

employed in .female-dominated occupationsand work organizations.

The results of the analyses presented inTable 2 also give information about to theextent to which the authority attainment

process operates differently for women andmen. By and large, individual qualificationsseem to influence employees’ chances to attainpositions of authority in the ways predicted byhuman capital theory. Women’s and men’seducational qualifications, labour-market

experience and tenure with their current

employer are rewarded in the authority attain-ment process. There is a tendency, although nota statistically significant one, pointing to thecondition that women are rewarded a little lessthan men for their labour-market experience.The somewhat weaker relationship between

experience and authority exertion in the case ofwomen may indicate that women in youngercohorts tend to be more attached to the labour-market and more career-oriented than are

women in older cohorts. Both women andmen have diminishing marginal returns fromlabour-market experience, which means thateach additional year of experience is worth

subsequently less in authority terms. In sum,the differences between women and men asregards how well they are rewarded for theirindividual qualifications are not especiallyprominent.

In models A, B and C in Table 2, all resultsare indicative of family responsibilities havingsomewhat different implications for women andmen in the authority attainment process.Married or cohabiting men have better oppor-tunities to exert authority than single menhave, whereas women’s - but not men’s -authority chances are limited to the extent

that they have children in the household. Thesetendencies are quite difficult to interpret in thatthey could reflect both the existence of self-selection on the part of employees and thecondition that employers treat women and menwith similar family situations in different ways.

In models B and C in Table 2, organiza-tional-level variables are included in theanalyses. It was hypothesized in the theoreticalparts of this paper that the allocation of womenand men into different labour-market struc-tures can influence the gender-differentiatedauthority attainment process in two partlycounteracting ways. The first assumption wasthat women face depressed opportunities toreach supervisory positions as a consequence oftheir concentration in female-dominated occu-

pations. The second assumption was thatwomen, once they are allocated to certain

occupations, are less subjected to discrimina-tion if they work in organizations in which

many other women are employed than if theywork alongside comparatively more men.

The first assumption is supported by theresults presented in Table 2. Women tend tohave fewer opportunities to reach supervisorypositions the more female incumbents there arein their occupation. This finding is compatiblewith the reasoning put forward by Baron andBielby (e.g. 1985), who stress that occupationalsex segregation implies limited access forwomen to internal labour-markets and theirbenefits. Obviously, occupations with a largerepresentation of female incumbents are asso-ciated with inferior opportunities for women to

at COLUMBIA UNIV on November 5, 2014asj.sagepub.comDownloaded from

Page 11: Gender Differences in Workplace Authority: Discrimination and the Role of Organizational Leaders

108

~ .8.~ u

15~#,-M)

~-~GJ tl~ U

~ .~.....::is

% 1 It::: .S~s.- E~ c~;s U £~’£

c s.9 E ba 3~ a

15~l# GJ

x ~ a~ ~Q t-

~n#b 1NO Ob b9 g,W

~Tl £- 1

~ ~ i

b W j

m9_B° ’

O_b £$91# .i~bt~ ~’ 15~ eE ~ -~]1’5,#%GJ .....

-~4t:::&dquo;&dquo;’’’8’^· O .

1 .~ &dquo;~ S -2

Uz m.3i$ .#i £.5 ~ ’-’O ~ .1-hb GJ#jfi ~ 15W~ lieN ~ ’~ ~ .~#%fl’5?~g -~~s

~~£ l

M ~ .

~$i~i Wll~#U ~ <:.:>

=U m,

~ <u .9 _ O CS ,I £-,t4 GJ ~ U07 ~% £ , g m u

s O3 0M42

IP

,6 d~ s~ bp cL&dquo; ca

. -1 r1 0-0 v 0

~t7 O r,V~ .G O0 0

00~~..90 I-<CO 8 - 0 c...

CO..9 I-< .S .... ê’ en ff ::I V N’b &dquo;’G b0 N¿S2~..c..c..c~ -~ aj p yp 00 -&dquo;

§ ~ ~ ~ S ~u ~ (u r< B c.s -3 ~ B ~ B0 0 v CI) (1) v t;I)s g S.S~.s&dquo;’00&dquo;&dquo; ....ëØ:C::¡::I-<~1-<cd ’ c~ ‘~’ c~ ~ o O 0 0uuuoooco 0 0 u vi C:) VI C:5 VIw w w e~ a s::..«1.00*:&dquo;-

at COLUMBIA UNIV on November 5, 2014asj.sagepub.comDownloaded from

Page 12: Gender Differences in Workplace Authority: Discrimination and the Role of Organizational Leaders

109

reach positions of supervisory authority,whereas men’s authority chances are not

significantly affected by the occupational sexcomposition.

The second assumption must be rejectedsince it appears to be men, and not women,who benefit in authority terms from working inestablishments in which the share of femaleemployees is relatively large. This result indi-cates discriminatory practices on the organiza-tional level. If equally qualified women andmen were perceived by employers and otherorganizational decision-makers as equally cap-able of exerting authority, a female-dominatedworkplace would provide better chances forwomen to reach higher positions within thework hierarchy. The fact that men, but notwomen, seem to benefit from a female-domi-nated work environment certainly lends sup-port to the hypothesis that discriminatorypractices influence the authority attainmentprocess.

In sum, it seems as if sex segregationprocesses imply twofold disadvantages as far aswomen’s authority attainment is concerned.First, the occupations that women tend to workin provide fewer chances for women to attainauthority. Second, the fact that many womenwork in the establishment is not an advantageto women, whereas men in female-dominatedwork environments appear to be relatively wellrepresented in supervisory positions.

The results from models B and C in Table 2also indicate that employment sector is relatedto women’s and men’s authority chances inquite different ways. Men have better chancesto reach positions of authority in the privatesector than in the public sector, whereaswomen’s chances appear to be lower, althoughnot significantly so, if they work in the privateas opposed to in the public sector. Thesefindings, together with earlier research indicat-ing that women and men generally facedifferent opportunities considering rewardattainment in the public and private sectors,provide reasons to study women’s and men’sauthority attainment separately for each sector.The results from these analyses are presented inTable 3. .

Table 3 also shows the differences between .women’s and men’s authority attainment interms of the gender gaps calculated on the basisof the intercepts expressing the (logged) num-ber of subordinates for women and men withthe same average qualifications, family situa-tions and labour-market locations. Women’s

disadvantaged situation in the authority attain-ment process is especially accentuated in theprivate sector, despite a tendency indicatingthat differences also exist between publiclyemployed women and men. The gap betweenwomen and men is, in general, more than threetimes larger in the private sector than in thepublic sector. Men who have average charac-teristics and who work in the private sectorhave more than seven times as many sub-ordinates as corresponding women in the

private sector, whereas men working in thepublic sector have one-and-a-half to two timesmore subordinates than have similar women inthe public sector. A comparison betweenmodels A and B in Table 3 shows that thegender differences in the public sector are moreexplicable in terms of the occupational andorganizational sex composition than are thegender differences in the private sector. Hence,the fact that publicly employed women haveless authority than publicly employed men can,in part, be assigned to the inferior opportunitiesthat publicly employed women in female-dominated occupations and workplaces face.Women in the private sector have restrictedopportunities to reach supervisory positionsirrespective of the sex composition of their

occupations and workplaces.Table 3 also shows some sector-specific

results worth mentioning within the context ofgender inequality in authority attainment.

Regarding individual qualifications, it seems

as if women in the private sector are rewardedneither for their education nor for their

seniority In the public sector, on the otherhand, women and men seem to be more

similarly rewarded for their human capital.Men’s positive authority returns from beingmarried or cohabiting seem to be restricted tothe private sector of the economy, whereas thehindering effect for women of having childrenin the household appears to exist only in thepublic sector. Since there is ample reason tobelieve that women are less discriminated

against in the public than in the private sector,it seems reasonable to posit that the negativeeffect for women of having children reflectssome kind of self-selection rather than discri-minatory behaviour on the part of employeesand other organizational decision-makers.

Evidently, some of the results in the

analyses presented in Table 2 can be explainedthrough the sector-specific analyses of models Band C presented in Table 3. It seems as if it isprimarily women in the private sector who are

at COLUMBIA UNIV on November 5, 2014asj.sagepub.comDownloaded from

Page 13: Gender Differences in Workplace Authority: Discrimination and the Role of Organizational Leaders

110

Table 4. The influence of the sex of the highest workplace manager on women’s and men’s chances to exert authority athigher levels in the organization, net of individual qualifications, family situation and horizontal sex segregation.aLogistic regression of the probability for supervising at least 20 per cent of the employees at the workplace. Oddsratios.

a Educational sector controlled for.

* p ~ 0.05 for the estimate to be equal to 1.** p ~ 0.01 for the estimate to be equal to 1. - ’

t p S 0.05 for the estimates to be equal for women and men.

more disadvantaged in the authority attain-ment process the more female incumbentsthere are in their occupation. Furthermore,the finding that men face better possibilities toreach supervisory positions the more womenwho are employed at the establishment seemsto pertain to the private sector only. In sum,men seem to have more to win by working inthe private sector, whereas women face stron-ger obstacles to reaching supervisory positionsin the private as opposed to the public sector.

Finally, model C in Table 3 shows that thechances for both publicly- and privately-employed women and men to attain authoritypositions seem unaffected by whether thehighest workplace manager is a man or a

woman. However, it seems reasonable toassume that the sex of the highest workplacemanager affects women’s and men’s chances inthe process of allocation to higher-level super-visory positions within organizational author-ity hierarchies. Thus, the analyses presented inTable 4 pertain to women’s and men’s chancesto reach authority positions which entailsupervision of at least one-fifth of the employeesin the workplace. 19

The main problem addressed in the ana-lyses presented in Table 4 is whether the sex ofthe highest workplace manager implies differ-ent chances for women and men in assign-

ments to higher positions within organizationalauthority hierarchies. We posited that femaleemployees face more obstacles in the authorityattainment process at workplaces where thehighest supervisor is a man. However, this

assumption is not supported by the empiricalresults presented in Table 4. Neither women’snor men’s chances to reach higher supervisorypositions are affected by the sex of the highestworkplace manager. Obviously, the processwhich results in women facing inferior oppor-tunities to reach positions of authority shouldbe described in terms other than activediscrimination brought about by top-levelmanagers within work organizations . 20

6. Conclusions

z

In the present paper, gender differences in

authority attainment at the establishment levelhave been examined. The analyses indicate thatwomen with qualifications, family situations

and horizontal labour-market positions similarto those of men have severely limited chances ofreaching supervisory positions. It seems rea-sonable to make at least some inference of theexistence of discrimination on the basis of theconsiderable gender gaps that persist even afterextensive controls for different career-related

at COLUMBIA UNIV on November 5, 2014asj.sagepub.comDownloaded from

Page 14: Gender Differences in Workplace Authority: Discrimination and the Role of Organizational Leaders

111

characteristics on both individual and organi-zational levels. Primarily within private sectororganizations, the underrepresentation ofwomen in supervisory positions is both sub-stantial and pervasive. This finding certainlylends support to the assumption that discrimi-nation continues against women in allocationprocesses in the labour-market. Moreover, theanalyses resulted in other findings which arecompatible with interpretations of the genderauthority gap in terms of discrimination.Women seem to be disadvantaged the morefemale incumbents there are in the occupation,whereas men benefit the more women whowork in the establishment. Furthermore, pri-vately employed women tend not to berewarded in the authority assignment processfor additional education and seniority.

Attempts have been made to examine towhat extent women’s relatively low chances toreach supervisory positions are influenced bythe sex of the highest workplace manager. Theunderlying assumption in these analyses wasthat if discrimination is an active process

brought about by organizational decision-makers, female employees with a male work-place manager have fewer authority chancesthan female employees working under a femaleworkplace manager. This hypothesis was

rejected, since the sex of the highest supervisorin the work organization had no effects at all onwomen’s and men’s authority attainment.

However, the whole discrimination hypothesiscannot be discarded due to the fact that the sexof the highest workplace manager does notinfluence employees’ chances to reach super-visory positions. In general, the analyses haveresulted in consistent and substantial evidencefor the assumption that female employees aredisadvantaged in the authority attainment

process. The fact that women with similarcareer-related individual characteristics andsimilar structural locations to those of menhave considerably lower chances to reach

positions of authority certainly indicates theinfluence of factors beyond women’s own reach.Accordingly, the discrimination hypothesisreceives support, albeit indirectly.

Since women’s possibilities to reach super-visory positions are unaffected by whether aman or a woman is at the top of the organiza-tional power hierarchy, it seems reasonable todiscuss the gender authority gap in other termsthan discrimination brought about more

actively and purposefully by organizationaldecision-makers. Statistical discrimination is

one mechanism applicable to male as well asfemale employers and other organizationaldecision-makers who actually have no distastefor hiring or promoting women to positions ofauthority. Statistical discrimination is based

upon rational considerations of how to max-imize the productivity at the workplace, in partby minimizing employees’ absence and quitrates. The outcome of such considerations could

very well be exclusion of women who aspire tosupervisory positions, in that women tend to beviewed as less reliable and less stable employeescompared to their male counterparts.

But why is the authority gap betweenwomen and men so wide in a country likeSweden? Sweden is a society known for a

relatively high degree of equality betweenwomen and men in a number of life spheres.However, policies concerning job rewards andtheir distribution between women and menhave primarily been focused on wage inequal-ity If wages are considered the reward for work,authority can be viewed as the reward for, or anexpression of, an employee’s merit in a broadersense, as conceived by the work organization(Halaby 1979). Merit is, of course, a vagueconcept, rather elastic and open to interpreta-tion and even manipulation. Due to the

vagueness of merit definitions, as well as thelink between perceived merit and rank, dis-

criminatory practices in authority allocationare likely to persist, or, as Halaby expresses it:

’ [F] ailure to achieve high rank is interpretableas ipso fact9 evidence of lack of merit’ (Halaby1979:124).

First version received August 1996Final version accepted January 1998

AcknowledgementsI thank colleagues at the Swedish Institute for Social Researchand the Department of Sociology at Stockholm Uniyersity,especially Johan Fritzell, Carl le Grand, Ryszard Szulkin andMichael Tahlin, for valuable comments on earlier versions ofthis paper. I also gratefully acknowledge the suggestions fromthe reviewers of this article. The Swedish Council for Social

Research has partly financed this study (grant number: 95-0205 :lc).

Notes ,

1 It is worth mentioning that the overall wage distributionis more compressed in Sweden than in countries like Canadaand the United States. A compressed wage structure quitenaturally implies a somewhat limited scope for wage differ-entials between various groups in society.

2 Approaches in which residual gender differences (i.e.

at COLUMBIA UNIV on November 5, 2014asj.sagepub.comDownloaded from

Page 15: Gender Differences in Workplace Authority: Discrimination and the Role of Organizational Leaders

112

differences that persist when relevant characteristics are

accounted for) are treated as indicators of discrimination haveoften been used in studies of gender inequality in job rewards.3 A position of authority is not necessarily only a positive

thing for the individual, given the responsibilities and workdemands that tend to be connected to authority. Thus,authority can be viewed partly as a burden or a cost forwhich the individual has to be compensated. Nevertheless, it

seems reasonable to assume that authority, in the end, haspositive net effects, considering various job rewards forindividuals who hold such positions.

4 Marriage can, in part, be treated as an indicator of futureplans to have a family. Moreover, it seems reasonable to assumethat women who are highly committed to their work haveweaker economic incentives to marry (Bernhardt 1993). Boththese assumptions indicate that feedback mechanisms are atwork.5 Men have also been found to benefit from marriage in

terms of wages (Richardson 1997).6 Of course, the processes generating such compositional

differences between women and men’s qualifications may havebeen influenced by gender discrimination at earlier stages. Thisinfluence is, however, more indirect compared to discriminationtaking place in the labour-market. Moreover, effects of familyfactors and qualifications on authority attainment are almostinevitably contaminated in the process of selection because offeedback between choices and outcomes related to familysituation and labour-market behaviour. For example, women’spositions at work can influence the division of labour in thehousehold (Nermo 1994), and women’s investments in educa-tion and labour-market experience can reflect earlier prefer-ences and choices concerning future work and family plans(Lehrer & Nerlove 1986; Waite & Stolzenberg 1976). However,by comparing men and women who have made similar choicesand investments regarding qualifications and family situation, itbecomes possible to infer to what extent the authorityattainment process is influenced by discriminatory practices.7 Bielby and Bielby’s (1988) analyses are based upon

women’s and men’s self-reported efforts, meaning that the

conclusions may be contaminated by some gender-related bias.However, earlier findings have pointed to the condition thatwomen tend to undervalue their work, which implies thatwomen’s work effort may be underestimated when established

through subjective ratings.8 For a more thorough description of the Level of Living

Survey, see Fritzell and Lundberg (1994).9 For a more thorough description of the Establishment

Survey, see le Grand et al. (1994).10 Before the logarithmic transformation of the authority

variable was made, the value 1 was added to the number ofsubordinates.

11 This variable’s effects are not presented in the tables.The variable for field of study is coded as follows: the first

category is the baseline, referring to public primary education,the second to aesthetics, humanities and religion, the third topedagogy, the fourth to administration, trade, economics, socialscience and behavioural science, the fifth to industry, crafts,engineering and natural science, the sixth to transportationand communication, the seventh to health care, the eighth toagriculture, forestry and fishing, the ninth to service and

military education and the tenth category to the respondentswho either have no education or who cannot be classified

according to the above categories (cf. Erikson & Jonsson 1996).12

Respondents who have more than 50 years of labour-market experience are given the value 50. Since labour-marketexperience is highly correlated with age, age is not included inthe analyses. However, it is, of course, an important componentin the experience variable.

13 Respondents who have worked more than 30 years fortheir current employer are given the value 30.

14 Those respondents who themselves are highest work-place managers are excluded from the analyses.

15 The main results obtained by the ordinary least squaresregressions were reproduced by means of logit regressions(tables not shown).

16 The adjusted intercepts are calculated for models withcentred values on the predictors, that is, after all individuals areassigned predictor values expressing the difference between theoriginal variable value for the individual and the averagevariable value calculated for women and men jointly. Theoriginal estimates for the various predictors’ effects on

authority are not altered after this procedure is applied. Hence,

where Y is the dependent variable, xj is a vector of independentvariables, xj is the overall means for these variables, bjcoefficients to be estimated and a* adjusted intercepts. Indicesm and f stand for males and females (for further elaboration, seele Grand 1989).

17 Strictly, the value 1 is subtracted from the exponen-tiated intercepts for women and men before the differences arecalculated. This procedure is applied because the authorityvariable is constructed as the logarithm of the number ofsubordinates plus 1. Hence:

where G is the gender gap in authority, and &alpha;m and &alpha;f are the

adjusted intercept for males and females, respectively.18 Differences between women’s and men’s estimates will

be discussed in the text insofar as they appear to be substantialand stable throughout the statistical models. Some genderdifferences seem important even though (due to the size of thestandard errors) they do not fully fulfil the formal requirementsof tests of significance.

.

19 The analyses are conducted jointly for employees inpublic and private sector organizations, since the number ofcases is too small to allow separate analyses for each sector.More specifically, there are very few women in the private sectorwho supervise at least 20 per cent of the employees in theworkplace.

20 The remaining results shown in Table 4 are also

enlightening. The human capital factors show predicted effectson the chances to reach higher supervisory positions, withsome exceptions. The employee’s education does not seem toaffect the assignment to authority positions at higher levelswithin the work organization. Presumably, potential incum-bents of higher supervisory positions are evaluated more

according to informal criteria and less according to formal

qualifications, such as educational credentials. Furthermore,only women obtain significant positive returns from their

seniority. Again, men’s authority returns from being married orcohabiting turn out to be significant.

ReferencesAndersen, C. D. & Tomaskovic-Devey, D. 1995. Patriarchal

Pressures: An Exploration of Organizational Processes thatExacerbate and Erode Gender Earnings Inequality. Work andOccupations, 22, 328-356.

Baron, J. N. & Bielby, W T. 1985. Organizational Barriers toGender Equality: Sex Segregation of Jobs and Opportunities.In A. S. Rossi (ed.), Gender and the Life Course, pp. 233-252.New York: Aldine de Gruyter.

at COLUMBIA UNIV on November 5, 2014asj.sagepub.comDownloaded from

Page 16: Gender Differences in Workplace Authority: Discrimination and the Role of Organizational Leaders

11

Becker, G. S. 1964. Human Capital. New York: Columbia

University Press.Becker, G. S. 1985. Human Capital, Effort, and the Sexual

Division of Labor. Journal of Labor Economics, Suppl. 3, 33-58.Bernhardt, E. 1993. Fertility and Employment. European

Sociological Review, 9, 25-42.Bielby, D. D. & Bielby, W. T. 1988. She Works Hard for the

Money: Household Responsibilities and the Allocation of. Work Effort. American Journal of Sociology, 93, 1031-1059.Bielby, W T. & Baron, J. N. 1986. Men and Women at Work: Sex

Segregation and Statistical Discrimination. American Journalof Sociology, 91, 759-797.

Charles, M. 1992. Cross-national Variation in Occupational SexSegregation. American Sociological Review, 57, 483-502.

Cockburn, C. 1991. In the Way of Women. London: Macmillan.Edlund, C., Ahltorp, B., Andersson, G. & Kleppest&ouml;, S. 1990.

Karri&auml;rer i kl&auml;m. Om chefen, familjen och f&ouml;retaget (JammedCareers: the Boss, the Family and the Firm). Stockholm:Norstedts.

England, P. & Farkas, G. 1986. Households, Employment, andGender: A Social, Economic and Demographic View. New York:Aldine de Gruyter.

Erikson, R. & Jonsson, J. O. (eds). 1996. Can Education be

Equalized? The Swedish Case in Comparative Perspective.Boulder, CO: Westview Press.

Fritzell, J. & Lundberg, O. (eds). 1994. Vardagens Villkor.

Levnadsf&ouml;rh&aring;llanden i Sverige under tre decennier (EverydayLife: Living Conditions in Sweden during Three Decades).Stockholm: Brombergs f&ouml;rlag.

le Grand, C. 1989. Interna Arbetsmarknader, ekonomisk segmen-tering och social skiktning (Internal Labour-markets, EconomicSegmentation and Social Stratification). Stockholm: Institu-tet f&ouml;r Social forskning.

le Grand, C. 1991. Explaining the Male-Female Wage Gap: JobSegregation and Solidarity Wage Bargaining in Sweden. Acta

Sociologica, 34, 261-278.le Grand, C. 1994. L&ouml;neskillnaderna i Sverige: F&ouml;r&auml;ndring och

nuvarande struktur (Wage Differentials in Sweden: Changesand Current Structure). In J. Fritzell & O. Lundberg (eds.),Vardagens villkor. Levnadsf&ouml;rh&aring;llanden i Sverige under tre

decennier (Everyday Life: Living Conditions in Sweden duringThree Decades), pp. 117-160. Stockholm: Brombergs f&ouml;rlag.

le Grand, C., Szulkin, R. & T&aring;hlin. M. 1994. OrganizationalStructures and Job Rewards in Sweden. Acta Sociologica, 37,231-252.

Halaby, C. N. 1979. Job-Specific Sex Differences in Organiza-tional Reward Attainment: Wage Discrimination vs. RankSegregation. Social Forces, 58, 108-127.

Hill, M. S. 1980. Authority at Work: How Men and WomenDiffer. In G. J. Duncan & J. N. Morgan (eds.), Five ThousandAmerican Families : Patterns of Econornic Progress, Vol. 8, pp.107-146. Ann Arbor, MI: University of Michigan: Institutefor Social Research.

Hultin, M. & Szulkin, R. 1997. The Sex of Workplace Managersand the Wages of Employees - An Empirical Test of ActiveDiscrimination. Mimeo (Stockholm: Swedish Institute forSocial Research/Department of Sociology).

Jacobs, J. 1989. Revolving Doors: Sex Segregation and Women’sCareers. Stanford, CA: Stanford University Press.

Jacobs, J. 1992. Women’s Entry into Management: Trends inEarnings, Authority, and Values among Salaried Managers.Administrative Science Quarterly, 37, 282-301.

Jacobs, J. & Lim, S. 1992. Trends in Occupational and IndustrialSex Segregation in 56 Countries 1960-1980. Work and

Occupations, 19, 450-486.

Jaffee, D. 1989. Gender Inequality in Workplace Autonomy andAuthority. Social Science Quarterly 70, 375-390.

Kanter, R. M. 1977. Men and Women of the Corporation. NewYork: Basic Books.

Lehrer, E. & Nerlove, M. 1986. Female Labour Force Behaviourand Fertility in the United States. Annual Review of Sociology,12, 181-204.

McGuire, G. M. & Reskin, B. F. 1993. Authority Hierarchies atWork: The Impacts of Race and Sex. Gender and Society, 7,487-506.

Mincer, J. & Polachek, S. 1974. Family Investment in HumanCapital: Earnings of Men and Women. Journal of PoliticalEconomy 82, 76-108.

Mueller, C. W, Kuruvilla, S. & Iverson, R. D. 1994. SwedishProfessionals and Gender Inequalities. Social Forces, 73,555-573.

Nermo, M. 1994. Den ofullbordade j&auml;mst&auml;lldheten (TheIncomplete Equality). In J. Fritzell & O. Lundberg (eds.),Vardagens Villkor. Levnadsf&ouml;rh&aring;llanden i Sverige under tre

decennier (Everyday Life: Living Conditions in Sweden duringThree Decades), pp. 161-183. Stockholm: Brombergs F&ouml;rlag.

Nermo, M. 1996. Occupational Sex Segregation in Sweden1968-1991. Work and Occupations, 23, 319-332.

Phelps, E. S. 1972. A Statistical Theory of Racism and Sexism.American Economic Review, 62, 659-661.

Reskin, B. 1993. Sex Segregation in the Workplace. AnnualReview of Sociology, 19, 241-2 70.

Reskin, B. & Padavic, I. 1994. Women and Men at Work.

Thousand Oaks, CA: Pine Forge Press.Richardson, K. 1997. Essays on Family and Labour Economics.

Doctoral dissertation no. 28. Stockholm: Swedish Institutefor Social Research.

Roman, C. 1994. Lika pa olika villkor. K&ouml;nssegregering i

kunskapsf&ouml;retag (Equal on unequal terms). Stockholm:

Symposion Graduale.Rosenfeld, R. A. & Kalleberg, A. L. 1990. A Cross-national

Comparison of the Gender Gap in Income. American Journal ofSociology, 96, 69-106.

SOU 1994 :3, M&auml;ns f&ouml;rest&auml;llningar om kvinnor och chefskap (Men’sconceptions of women and leadership). Stockholm: Fritzes.

Statistics Sweden (SCB). 1992. Man &auml;r chef. En studie av kvinnoroch m&auml;n i ledande st&auml;llning i privat och offentlig sektor (Men arebosses. A Study of Women and Men in Supervisory Positionsin the Private and Public Sector). Stockholm: Statistiska

Centralbyr&aring;n.Tomaskovic-Devey, D. 1993. Gender and Racial Inequality at

Work. The Sources & Consequences of Job Segregation. Ithaca:ILR Press.

Wahl, A. 1992. K&ouml;nsstrukturer i organizationer (Gender Struc-tures in Organizations). Stockholm: Ekonomiska Forskning-sinstitutet.

Waite, R. M. & Stolzenberg, L. J. 1976. Intended Childbearingand Labor Force Participation of Young Women: Insightsfrom Non-Recursive Models. American Sociological Review, 41,235-252.

Wolf, W. C. & Fligstein, N. D. 1979a. Sexual Stratification:Differences in Power in the Work Setting. Social Forces, 58,94-107.

Wolf, W. C. & Fligstein, N. D. 1979b. Sex and Authority in theWorkplace: The Causes of Sexual Inequality. American

Sociological Review, 44, 235-252.Wright, E. O., Baxter, J. & Birkelund, G. E. 1995. The Gender

Gap in Workplace Authority in Seven Nations. American

Sociological Review, 60, 407-435.

at COLUMBIA UNIV on November 5, 2014asj.sagepub.comDownloaded from