Measuring women’s progress in aglobal era

Sylvia Walby

Introduction

The 1995 UN Conference on Women in Beijingagreed a Platform for Action to improve the livesof the world’s women (United Nations 1995). Thisconstituted a challenge to many conventionalconceptions of progress, of equality, and of justice.It remains a challenge to see the full implementa-tion of this programme, agreed by all the world’scountries. In 2005 there was a review of theprogress made in implement-ing the Platform, the Beijing110 process. A series of meet-ings and reports were preparedfor this process, including at aregional level (LuxembourgPresidency 2005, United Na-tions Economic Commissionfor Europe 2004).

The Platform for Actionis a challenge to the notionthat progress can be mea-sured by the extent of eco-nomic growth. In challengingthis traditional indicator ofsuccess, it is not alone, sincethis is contested by the United Nations Devel-opment Programme (UNDP 1990). But whatexactly should replace this dominant paradigmis highly contested. There are various alternativebases on which assessments of progress andclaims for justice can be made, including justice,equality, human rights, and capabilities (Nuss-baum 2000, Sandel 1998, Sen 1984, 1987, 1992,1999, Walby 2001). Further, in the act ofagreeing the Platform for Action, the UNconference challenged the view that differences

between cultures made standards of justiceincommensurable, but nevertheless there remainimportant issues in how these standards areoperationalised in different locations.

But what is gender equality? The conceptand content of ‘‘gender equality’’ is highly con-tested. ‘‘Gender equality’’ is a ‘‘signifier’’ thatactors attempt to fill with their own preferences.How is the tension between models of genderequality that are based on sameness, the equal

valuation of difference,and the transformation ofexisting gender practicesand standards to be re-solved (Bacchi 1999, Rees1998, Verloo 2001, Walby2005)? There are impor-tant differences betweenthe routes by which genderequality might be achieved.

While the sources ofchange in gender relationsare many, three main typescan be identified. Some arecommon to processes ofeconomic and human de-

velopment; some are specific to path-dependentforms of change that vary between countries;and others are related to global political wavesand the development of international regimes.The UN initiatives associated with the Platformfor Action take place within these wider sourcesof change.

The UN Platform for Action requiresattention to the specificities of gender disadvan-tage, which were not to be subsumed within ageneric concept of inequality, and, by further

Sylvia Walby is Professor in the Depart-ment of Sociology, Lancaster University.Previously Professor of Sociology at theUniversities of Leeds and Bristol. Authorof Gender Transformations (Routledge,1997) and Theorizing Patriarchy (Black-well, 1990); editor of special issues ofSocial Politics (2005) and InternationalFeminist Journal of Politics (2005) ongender mainstreaming. Her next book isComplex Social Systems: Theorisationsand Comparisons in a Global Era (Sage,2006).Email: S.Walby@lancaster.ac.uk

ISSJ 184rUNESCO2005. Published byBlackwell PublishingLtd., 9600GarsingtonRoad,OxfordOX42DQ,UKand350Main Street,Malden,MA02148,USA.

differentiating this into 12 critical areas foraction, it challenges any simple treatment ofgender inequality as reducible to a single dimen-sion. One of the critical areas identified by thePlatform for Action was that of the ‘‘institu-tional mechanisms for the advancement ofwomen’’. The capacity for effective policydevelopments to achieve gender equality needsto be established in institutional mechanisms, ofwhich three kinds were identified: nationalmachinery and other governmental bodies, theintegration of gender perspectives in policymaking, and the generation of gender-disaggregated data for policy evaluation. It isimportant to be able to measure the extent towhich progress has been made on the 12 criticalareas of the Platform for Action. In order to dothis, it is necessary to operationalise the conceptsbehind these critical areas and to developsummary indicators supported by data. Thismatters for both social science analysis andpolicy development. The movement from ab-stract concepts to measurable quantifiable in-dicators supported by valid and reliable data iscomplex. In this context it is important to haveindicators and data that are comparable be-tween countries and over time.

The Beijing 110 process is not the only onethat is demanding the development of indicatorsof gender equality. A similar policy dynamic canbe seen within the EU strategy for genderequality. In the context of the development ofgender mainstreaming, there is an ever-presentconcern with the relationship between gender-specific measures and those of the mainstream,so that gender is no longer a marginalisedconcern. While the development of gender equa-lity indicators in the EU has often been framedby the needs of economic policy, there is anadditional impetus, which comes from theirattention to the Beijing 110 process. Whileseveral UN bodies, such as UNDP and UNI-FEM, and objectives, such as the MillenniumDevelopment Goals, have focused primarily onissues for the South (Kabeer 2003), the concernin this article is the EU context for translatingthe Beijing 110 process into one relevant for themore developed world.

There has been some ambivalence infeminist theory about the process and implica-tions of abstraction, since it entails the produc-tion of representations of the gendered world

that are somewhat removed from women’sdirect experiences. Some of the early forms offeminist challenges to patriarchal representa-tions of knowledge prioritised the use ofwomen’s experiences to confront and replacemistaken conceptions. However, as the exclu-sion of women from science and formalisedknowledge institutions has diminished, there hasbeen an increased capacity to generate forms ofknowledge that are more consistent with theworld as experienced by women. Statistics, oncetreated with suspicion, on the basis that theywere most likely to reflect a patriarchal view ofthe world order, can now be viewed as a newlyimportant domain of contestation over therepresentation of gender inequality. Genderstatistics are a new and key site of activity ofthe global feminist epistemic community.

Conceptualising progress,equality, justice, humanrights, and capabilities

In order to situate the challenge constituted bythe Platform for Action, I shall begin with areview of some contemporary debates about theconcept and measurement of progress, equality,and justice. First, is the challenge to the use ofGDP per capita as the key measure of progressby the capabilities approach (Nussbaum, 2000,2003, Sen, 1984, 1987, 1992, 1999). Second, aconsideration of the concepts of equality, justice,and human rights as diverse frames within whichgender progress might be conceptualised. Third,a consideration of the implications of differentmodels of gender equality for determining theappropriate standards of gender equality, inparticular, whether they are the same ones formen and women, promote equal valuation ofdifferent contributions, or propose new trans-formed gender standards for all. Fourth, ad-dressing the tension in developing concepts andindicators of progress and equality for genderthat are either separate from or close to themainstream (Jahan 1995, Walby 2005).

GDP per capita

The most traditional indicators of economicprogress are the level of economic development

372 Sylvia Walby

r UNESCO 2005.

and the rate of economic growth, used by theWorld Bank, InternationalMonetary Fund, andmany national government Finance Ministries.The measure of progress embodied in mostglobal institutions of financial governance, suchas the World Bank and International MonetaryFund, as well as adopted by many othergovernmental bodies from the EU to the UK,is that of the level of economic development, asmeasured by the size of the Gross DomesticProduct (GDP) per capita, and its rate ofgrowth. This is defended as the prime indicatorof progress on the basis that this constitutes anindication of the average standard of living ofpeople in a country and that governments arerepeatedly democratically elected on such amandate (HM Treasury 2004).

It is well known that this has been subject tomany challenges. While it is widely noted as ameasure of the performance of the monetarysector of an economy, the extent to which itcorrelates with human, as well as economic,development is highly disputed, not least by theUnited Nations Development Project. The cri-tique of the use of GDP per capita as the keyperformance indicator of economic developmentis most developed in the context of developmentpolitics and theory. It is contested on the groundsthat the average income per person hides in-equalities such as those associated with class andgender, and that income is not an ade-quate measure of human development in thesense of the capacity to function. Thus it is notan adequate measure of human progress, whichis the proper goal of economic policy. The debatehas broadened into questions of how develop-ment is best conceptualised as well as measured,engaging in philosophical as well as substan-tive questions. It addresses the question ofwhat values should be incorporated into globaleconomic policy, in particular, what it means,both abstractly and substantively, to value humandevelopment as the end goal of economic devel-opment.

Capacities,capabilities, and theUNDP

The contestation between these rival approachesis most clearly articulated in the work of Sen(1984, 1987, 1992, 1999) and Nussbaum (2000,2003). Sen argues for a re-focusing of theultimate goal of economic policy away from

the level of monetary income per person to thatof human capabilities. Human capabilities areunderstood as what people are actually able todo and to be. Human capabilities can beunderstood in relationship to the notion ofwell-being. In this approach monetary income ismerely one of the means by which the outcomeof human capabilities is generated. It impliesthat merely raising the level of average incomemay not be an adequate goal of economic policy,if the pattern of its distribution and use are notoptimal for the achievement of human capabil-ities. High levels of inequality may preclude thedevelopment of human capabilities in all thepeople of a country, and may be hidden by theuse of average income as the key indicator. Thechange in the key performance indicator fromincome per person to some measure of humancapabilities is likely to change the nature ofeconomic policy.

Sen rejects approaches that rely on people’sarticulated subjective preferences on the groundsthat oppressed people may have adjusted totheir subordinated status. Adaptive preferencesare not always the best guide to best policy.Rather it is necessary tomakemore fundamentalclaims about what is needed to enhancecapabilities. But Sen leaves the question asto exactly what are human capabilities ratheropen. So how is a more concrete list to bedeveloped?

Nussbaum (2000, 2003) argues for thedevelopment of a concrete list of central humancapabilities, though not in quantified form. Shelists ten: life and not dying prematurely; bodilyhealth, including good health, reproductivehealth and shelter; bodily integrity, includingfreedom to be mobile, to be secure from violentassault including domestic violence, and oppor-tunities for sexual satisfaction; being able to usethe senses, imagination and thought in a wayinformed and cultivated by education, as well asfreedom of expression and religion; being able tohave emotional attachments; being able toengage in practical reason and reflection; abilityto affiliate with others and to receive social res-pect, whatever one’s race or sex; live with otherspecies successfully; ability to play; and theability to participate politically, to hold property,and to enter decent employment. Nussbaumconsiders that these are valid universally, thoughalways subject to revision. The basis of her list

Measuring women’s progress in a global era 373

r UNESCO 2005.

appears to be her own reading of the literatureand engagement with the issues.

Nussbaum (2003), while agreeing with Senonmany matters, is critical of Sen for leaving hisargument at too abstract a level. She considershis strong stance on many issues of social justiceand his role in the development of the UNDPmeasures as puzzlingly inconsistent with hisrefusal to endorse any given list of capabilities.She suggests that the most likely explanation forthis is the priority that he gives to processes ofdemocratic deliberation, which he feels would beinhibited by the endorsement of a set ofcapabilities at the international level. Sen(2004) confirmed this as part of the explanation.So who and how are the key lists determined?What balance of democracy, power, and ex-pertise?

An attempt to operationalise the concept ofhuman capabilities in measurable forms in amanner informed by the work of Sen has beenconducted by the UNDP (1990). The advocacyby the UNDP of its human development indicesconstitutes one of the most significant policychallenges to the use of GDP per capita. Thiscontestation over the goals of global economicpolicy is led at the policy level by the UnitedNations Development Project (UNDP), whichproduces an annual publication with a set ofalternative summary and detailed performanceindicators for almost all countries in the world.This is considered in detail below.

Varieties of models of genderequality

There has been more than one way of con-ceptualising the nature of and route to genderequality. At least three major types of approachcan be identified: equality through sameness(equal opportunities or equal treatment),through equal valuation of difference (specialprogrammes), and the transformation of gen-dered practices and standards of evaluation. Thefirst model is one in which equality based onsameness is fostered, especially where womenenter previously male domains, and the existingmale norm remains the standard. The second isone in which there is a move towards the equalvaluation of existing and different contributionsof women and men in a gender segregatedsociety. The third is one where there is a new

standard for both men and women, that is, thetransformation of gender relations.

There is question as to whether the first twomodels actually constitute gender mainstream-ing, because they retain the gender standards ofthe status quo in some form. For Rees (1998),only the third strategy constitutes gender main-streaming and has the potential to deliver genderjustice because this is the only strategy thatinvolves the transformation of the institutionsand the standards necessary for effective equal-ity, while Booth and Bennett (2002) argue thatall three are gender mainstreaming approaches.

While the elimination of gender inequalityis the goal of the gender mainstreaming strategy,the extent to which this can mean accepting andvaluing existing gendered differences is a keysource of disagreement within gender main-streaming theory and practice. This has been adebate within gender theory more generally.While all the definitions of gender equalityinclude equality within each social domain, theyvary as to whether a change in the balance of thedomains, and the equalisation of any differentialrepresentation of women and men in eachdomain, constitute legitimate areas for interven-tion or not.

Underlying these discussions is the‘‘sameness/difference’’ debate that has takenplace within feminist theory (Felski 1997, Folbre2001, Fraser 1997, Lorber 2000). This is a multi-faceted debate, which is simultaneously norma-tive, philosophical, theoretical, substantive, em-pirical, and policy-relevant. Thus within ananalysis of gender mainstreaming are the classicarguments within feminist theory about differ-ence, universalism, and particularism. In parti-cular, there are dilemmas in how to recognisedifference, while avoiding the trap of essential-ism (Ferree andGamson 2003, Fraser 1997), andtaking account of the global horizon. There is anissue as to whether traditional equal opportu-nities policies are limited because they mean thatwomen can only gain equality with men if theyare able to perform to the standards set by men(Guerrina 2002, Rossilli 1997). However, there isa question as to whether there can be an effectiveroute to gender justice in which existing separategender norms/standards are retained, in that it isnot possible to be ‘‘different but equal’’ becausedifferences are too entwined with power andresources.

374 Sylvia Walby

r UNESCO 2005.

The Council of Europe definition of genderequality includes diverse elements:

Gender equality means an equal visibility, empowerment

and participation of both sexes in all spheres of public and

private life . . . Gender equality is not synonymous with

sameness, with establishing men, their life style and

conditions as the norm. . . . Gender equality means

accepting and valuing equally the differences between

women and men and the diverse roles they play in society

(Council of Europe 1998: 7–8).

The Council of Europe (1998) specifies the needfor the ‘‘equal participation of women and menin political and public life’’ and for ‘‘theindividual’s economic independence’’, and that‘‘education is a key target for gender equality’’.This defines equal participation in political andpublic life, in education and the achievement ofeconomic independence, as universal goals,while other spheres (notably the family andcare-work) remain sites of difference. An under-lying question is that of the assumed degreeof connection among the gender practicesin different domains. If they are coupled tightly,it may not be possible to have equality throughsameness in one domain and equality withdifference in another. If the links are looser,this may be theoretically and practically possi-ble. This debate depends on an implied theoryof gender relations that needs to be made explicitin order to understand the nature and degreeof the postulated connections between differentgendered domains and the implications ofchanges in one of them for the others (Walby2004).

These theoretical debates about models ofgender equality underpin the empirically baseddebates as to preferred indicators of genderedprogress. In particular, they are relevant to theissue of whether equal participation in employ-ment should be taken as an indicator of genderequality, or whether unequal participation maybe consistent with gender equality since theremight be equal valuation of women’s unpaidwork with paid work. This is one of theimportant issues underlying the debates on thedevelopment of indicators of the advancementof women.

The debates on the conceptualisation ofgender equality are informed by debates on thenature of the processes of change in genderrelations. In particular, there is the question of

the extent to which progress for women is closelyassociated with economic development as con-trasted with a democratically inspired humandevelopment that involves state or communityprovision of welfare services.

Development of institutionalmechanisms for theadvancement of women

Gender and the mainstream

Gender mainstreaming was identified by thePlatform for Action as a key process in thedevelopment of policies for the advancement ofwomen. Gender mainstreaming is an essentiallycontested process (Bacchi 1999, Elgstrom 2000,Walby 2005). This is because it inevitablyinvolves tension between the ‘‘mainstream’’and the ‘‘gender equality’’ positions. Never-theless, there is often a focus on those areaswhere there is potential overlap between the twoagendas. In practice, there is often a difficultstrategic decision as to how close to or distantfrom the mainstream should be the positionadopted by the advocates of gender equality. If itis too close, then it is likely to end in the seriousdilution and even loss of the gender equalityperspective as it becomes integrated into themainstream (Jahan 1995). If it is too distant,then it may be rejected as ‘‘extremist’’ and havelittle impact.

These general considerations concerninggender mainstreaming apply to the process ofthe conceptualisation and operationalisation ofgender equality. On the one hand, there is a needto identify separately the gender issue at stakeand to capture its specificity. On the other, if theindicator that is developed is too far from theexisting repertoire, then there is a danger that itwill be marginalised and that the data needed tosupport it will not be forthcoming.

Institutional mechanisms

The Platform for Action identified the develop-ment of institutional mechanisms for the ad-vancement of women as one of the 12 criticalareas. This involved three strategic objectives:H1, create or strengthen national machineries

Measuring women’s progress in a global era 375

r UNESCO 2005.

and other governmental bodies; H2, integrategender perspectives in legislation, public poli-cies, programmes, and projects; and H3, gen-erate and disseminate gender-disaggregated dataand information for planning and evaluation(UN 1995). These three developments are highlyinter-related. The development of gender per-spectives in policy making requires informationabout the potential gender impact of policies.This requires the development of indicators inorder to benchmark what would be importantgender impacts, as well as the data necessary tosupport these. It is unlikely that this informationwould be produced without an institutionalinfrastructure, so it is necessary to have specia-lised national gender machinery and othergovernmental bodies and to have forms ofindependent scrutiny and democratic input.

The development of the institutional me-chanisms has involved the creation of specialisedgender units within government. Although theyhave been created in most countries, the formsthey take are highly varied. These have beenlocated at various levels, from the highest levelwith a remit to overview all governmentactivities, to whole or major parts of ministries,to small units tucked away within anotherministry. The unit may be headed politically bya Cabinet or more junior minister. The unit andits political head may be dedicated to genderissues, or have many other duties. They mayhave many or few resources; authority todetermine the practices of other ministries ornot (Luxembourg Presidency 2005). They mayoperate with much public scrutiny from andengagement with civil society and NGOs, or not(Clavero et al. 2004, Rai 2003, Verloo 2001).They may be affected by the presence of womenin decision-making bodies, for example theextent to which women are elected as parlia-mentary representatives, promoted as govern-mental ministers, and involved in othergovernance institutions (Huber and Stephens2000, Manza and Brooks 1998). The relation-ship between the gender machinery, electedwomen political representatives and civil societymay be key to understanding the impact of thegender machinery (Halsaa 1998, Mazur 2002),for example in the development of a ‘velvettriangle’ (Woodward 2004). The nature of therelationship with transnational and interna-tional governance bodies, for example, EU,

ILO, World Bank, IMF, UN may also be ofimportance (Keck and Sikkink 1998, Pietila1996, Walby 2002).

The implementation of gender mainstream-ing and gender equality policies more generallyrequires the development of specific tools andprocesses. Major tools include ‘‘gender impactassessment’’ and ‘‘gender budgeting’’. These inturn require further tools including ‘‘indica-tors’’, ‘‘benchmarking’’, ‘‘targets’’ and ‘‘genderdisaggregated statistics’’. These tools are utilisedby the normal policy actors in engagement withthose representing women’s interests. Genderimpact assessment is an example of a new formof gender mainstreaming practice that uses botha new gender toolkit and new forms of inclusionof women in decision-making processes. Genderbudgeting is a specific form of gender impactassessment that is applied to financial decision-making.

Gender impact assessment involves theanalysis of the gender implications of policies.It is best performed during the early stages ofpolicy development, so that revisions can bemade to the proposed policy if its impactappears to be detrimental to gender equality. Itis recommended in the Platform for Action.

Gender budgeting challenges the tradi-tional notion that financial governance is agender-free zone. Gender budgeting is a gendermainstreaming tool that includes a genderequality perspective in financial decision-makingat the highest levels. It is a process of disag-gregation of budgets by gender in order todiscover the extent to which policies that havegendered implications are differentially funded.It is not about a separate budget for women. Itinvolves the introduction of a gender equalityperspective in forms of policy making which hadbeen beyond the reach of more traditional equalopportunities approaches. The purpose is tomake financial decision-making at governmen-tal level transparent in relation to gender(increased budget transparency is recommendedby the Organisation for Economic Co-operationand Development (OECD 2001) as a goodgovernance practice) (Balmori 2003, Budlenderet al. 1998, 2002; European Commission 2003a,European Parliament 2003, Sen 1999, Sharp andBroomhill 2002, Villagomez 2004).

The development of both gender impactassessment and gender budgeting depend

376 Sylvia Walby

r UNESCO 2005.

upon the availability of gender-disaggregatedstatistics, indicators and benchmarks. It isimportant to have available the necessary data,as well as knowledge as to the preferred levels ofsummary indicators, and to be able to constructrelevant benchmarks to assist in the monitoringchange. For example, gender budget projectsthat examine the implications of tax and benefitpolicies require the generation of data disaggre-gated by gendered individuals, not merelyconstituted at the traditional household level.

Both processes require not only the main-streaming of gender perspectives into the workof the normal policy actors, but also theinclusion of women’s voices and interests inareas of decision-making in which they are stillunder-represented, often requiring the engage-ment of expert civil society groups and NGOs.This is especially important for the process ofmonitoring, where such independent expertise isimportant.

Development of genderedindicators

Analysis of gender inequality requires the devel-opment of gendered indicators supported bygender disaggregated statistics. This is noted as acomponent of the UN critical area on Institu-tional Mechanisms for the Advancement ofWomen in the Platform for Action. StrategicObjective H.3 is to ‘‘Generate and disseminategender-disaggregated data and information forplanning and evaluation’’. Paragraph 206 lists aseries of agreed actions to be taken by ‘‘national,regional and international statistical services andrelevant governmental and United Nations agen-cies’’ detailing the steps that are to be taken. Theadoption of a twin strategy of gendermainstream-ing alongside specific actions for gender equality(UN 1995) has increased the priority attached togender disaggregating data in a wide range offields previously thought to be ungendered.

The development of indicators and the dataneeded to support them constitute a critical linkbetween policy aspirations and knowingwhether they are working. Robust, valid andreliable indicators of changes over time andcomparatively between countries are key toevaluating the effectiveness of innovative poli-cies. Without indicators and the data to support

them, it is hard to assess the extent to which thedevelopments in policy, such at those noted inthe country reports to the UN Beijing 110process, have positive outcomes for genderequality. Indicators are important as challengesto rhetoric. In the absence of commonly agreedindicators, there is a potential for considerableslippage between rhetoric and policy practice.

There is always a distinction between theconcept and a quantitative measure that isintended to represent the concept. There arelong-standing debates as to how economic andother indicators should best be developed. Beyondthe economic, there is increasing interest in anddevelopment of quantitative indicators of variousdomains of social life that are cross-nationallycomparative (Berger-Schmitt and Jankowitsch1999, UNDP 1990). There are three majorattempts to create gendered indicators appropri-ate for comparative cross-national analysis overtime: a suite of indicators from the UNDP(discussed below); the Millennium DevelopmentGoals, for developing countries (not discussedhere because of the focus of this paper ondeveloped countries); and indicators from the EU.

United Nations Development Projectindicators

The United Nations Development Project hasdeveloped various indices to capture progress inhuman development. The UNDP (1999) pro-duces a Human Development Index that in-cludes not only income per person (GDP percapita) but also longevity (as an indicator ofhealth) and education. It produces severalversions of the indicator, which sharpen thefocus on different inequalities, including genderand poverty. Supporting this is a series of tableson many diverse aspects of human well-being.The basic UNDP indicator is the HumanDevelopment Indicator, which is composed ofthree elements: life expectancy; education (bothenrolment and achieved literacy); and incomeper capita (UNDP). However, since this mea-sure is an average for each country it necessarilycannot reveal variations in human developmentwithin a country, which is essential if a genderanalysis is to be conducted.

The UNDP provides two further indices,the Gender-related Development Index (GDI)and the Gender Empowerment Measure

Measuring women’s progress in a global era 377

r UNESCO 2005.

(GEM), which are gendered and thus addressthis issue. The GDI applies gender ratios to eachof these three dimensions: relative life expec-tancy; relative education; and the proportion ofearned income. The GEM attempts to measurefemale empowerment and is made up of threeindicators: the proportion of women in parlia-ment, the proportion of women in top jobs(professional, technical, administrative, andmanagerial), and the female share of earnedincome (UNDP). These indices have beenimportant and influential; however, they havealso been subject to several criticisms.

The first disadvantage of these indices istheir conflation of absolute levels of develop-ment (such as absolute GDP per capita) with agender relationship in a single index. This isappropriate if the intention is to measure theabsolute development of women, but if what isbeing sought is a measure of gender inequality,then this practice does not achieve this goal(Dijkstra 2002, Dijkstra and Hanmer 2002).Second, in practice the GDI is overwhelmingly(over 90% in most countries) driven by onecomponent, that of income share, which isinappropriate for an index that purports to havethree components. This is for technical reasonsassociated with the greater variation in thiscomponent than the others (Bardhan andKlasen 1999). Third, the income share compo-nent is predominantly driven by the rate ofparticipation in employment, for technicalreasons associated with the paucity of data onthe gender wage gap (Bardhan and Klasen1999). Fourth, there is concern that the genderincome share may be a poor proxy for genderedstandard of living and women’s well-being. Thisis because there may be household sharing (withwomen benefiting from resources despite no orlittle earned income) or there may be anuncertain correspondence between a woman’searned income and her access to resources(Bardhan and Klasen 1999, Dijkstra andHanmer 2000). This issue can be seen as goingbeyond a technical problem to one that isassociated with the choice of model of genderequality, which might be seen as culturallyspecific. In particular there is the question as towhether equal participation in employment ismore likely to deliver gender equality than agendered division of labour in the household.However, this may still be seen as an empirical

question, that is, the extent to which a woman’searned income is associated with improvementsin her well-being (Anand and Sen 1995), is notone of incommensurable cultural difference. Afifth problem is that neither the GDI nor GEMencompasses all 12 critical areas. In particular,violence against women is missing. The UNDPgendered indicators of progress for women,while flawed, nevertheless constituted an inno-vative and important step in the process ofdeveloping more adequate indicators of humandevelopment.

Millennium Development Goals

The launch of the Millennium DevelopmentGoals in 2000 constituted a further ambitiousattempt to produce indicators and targets, whichwere supported by robust statistical data, tomeasure progress towards development. Likethe UNDP and capabilities approaches, theyfocused on outputs for people, rather thaninputs such as the size of income. The 18 targets,including the eradication of extreme poverty andhunger, the achievement of universal primaryeducation, and many associated with basichealth were associated with eight goals: eradi-cate extreme poverty and hunger; achieveuniversal primary education; promote genderequality and empower women; reduce childmortality; improve maternal health; combatHIV/AIDS, malaria, and other diseases; ensureenvironmental sustainability; develop a globalpartnership for development. However, onlyone quantitative target named gender, that ofthe elimination of gender disparity in education(Millennium Development Goals 2004).

While the Millennium Goals constituted aserious attempt to operationalise the concept ofdevelopment for less developed countries, mostof its goals and targets have already been met bymore developed countries. Hence for countriessuch as those in the EU they do not constitute anappropriate set of indicators and targets forfurther development.

Development of EU indicators for thePlatform for Action

The European Union committed itself to thedevelopment of a simple suite of indicators at theMadrid Council in 1998, in order to be able to

378 Sylvia Walby

r UNESCO 2005.

Workers in a Samsung Electronics factory in Suwon, South Korea, October 1999. AFP/Choo Youn-Kong

Measuring women’s progress in a global era 379

r UNESCO 2005.

effectively carry out the annual review of theimplementation of the Platform for Action towhich it committed itself in 1995. The EuropeanCommission has energetically engaged in theprocess of developing gender equality indicators,especially those relevant to the more developedworld, establishing working groups (EuropeanCommission 2001a), advisory committees (Eur-opean Commission 2001b), commissioning ex-pert reports (Plantenga et al. 2003, Rubery et al.2002), engaging in its own research (EuropeanCommission, 2003f), discussing the issue at theHigh Level Group on Gender Mainstreaming,engaging with Eurostat (Eurostat 2004), gainingthe support and approval of the EuropeanCouncil and various EU Presidencies (Luxem-bourg Presidency 2005) for this programme ofwork, and presenting reports on gender equalityto the Council (European Commission 2004).

The development of indicators, bench-marks and targets is a key part of the EuropeanUnion process of policy development, especiallywith the development of the new Open Methodof policy Co-ordination (OMC) since 1997.The OMC has been especially important fornew policy developments in the EuropeanEmployment Strategy (EES), based on agree-ment of hard targets at meetings of the SpringEuropean Council, implemented by policiesdeveloped to suit national frameworks, thoughwith exchange and learning between differentcountries, and monitored through annualreports to Council, using agreed StructuralIndicators, with data overseen by Eurostat.Insofar as gender equality issues have beenmainstreamed within the EES, then these receivethe full attention of Council and Commissioncombined with data support from Eurostat. TheEuropean Employment Strategy was given a keyplace in European Union policy development bythe LisbonCouncil in 2000 which established theaim for the EU to be by 2010 ‘‘the mostcompetitive and dynamic knowledge basedeconomy in the world, capable of sustainableeconomic growth with more and better jobs andgreater social cohesion’’. The achievement ofthis aim is considered to require an effectivegender dimension, in particular the reductionof gender gaps in employment, unemployment,and pay. A key target is to increase the employ-ment rate for women to 57% by 2005 (Stock-holm Council) and to 60% by 2010 (Lisbon

Council) in the context of raising overallemployment rates to 67% in 2005 and 70% by2010 (European Commission 2003b, 2003c,2003d).

In the early stages of the European Employ-ment Strategy (EES), when Equal Opportunitieswas one of four pillars, the Commission devel-oped seven equal opportunities indicators,which were reported annually in the JointEmployment Report. These were: EO1, Absolutegender unemployment gap; EO2, Absolutegender employment gap; EO3, Index of gendersegregation in occupations; EO4, Index ofgender segregation in sectors; EO5, Genderpay gap; EO6, Gender gap in the employmentimpact of parenthood; and EO7, Employmentimpact of parenthood (European Commissionand European Council 2002a, 2002b).

In the revised and current practice, genderequality is one of ten guidelines in the EES andthere is a separate annual report on equalitybetween women and men from the Commission(European Commission 2004). This reportprovides a general overview of progress andchallenges and provides supporting data (pro-vided by Eurostat) on five dimensions: paidwork (gender gaps in employment and unem-ployment, gender share of part-time working),income and pay (gender pay gap and risk ofpoverty), decision-making power (proportion ofwomen in parliament andmanagerial positions),knowledge (educational attainment at uppersecondary level and percentage of professorswho are female), and time (the gap between thehours worked by women and men who havechildren) (European Commission 2004).

A key set of indicators for policy makers inthe EU are the newly developed StructuralIndicators, some of which are provided in agender-disaggregated form: employment growth;employment rate and of older workers; effectiveaverage exit age; gender pay gap; life-longlearning (adult participation in education andtraining); unemployment rate; science and tech-nology graduates; at-risk-of-poverty (before andafter social transfers); dispersion of regionalemployment rates; long-term unemploymentrate; jobless households.

The EU has begun the process of develop-ing indicators and supporting data in relation toeach of the 12 critical areas of the Platform forAction, but this is not yet complete. The extent

380 Sylvia Walby

r UNESCO 2005.

of the development varies between the 12 criticaldifferent areas, as discussed below.

Women and poverty

There are some gender disaggregated structuralindicators developed within the agenda on socialinclusion and social cohesion, which are reportedannually to the Spring European Council, how-ever, the Presidencies and Councils have notagreed an indicator for Women and Povertyspecifically in relation to the Platform for Action.The most important indicator here is ‘‘at-risk-of-poverty (before and after social transfers)’’ anddata are adequately and appropriately providedby EUROSTAT. The inclusion of data by ageenables the especially disproportionate poverty ofwomen older than working age to be revealed.However, this indicator is predicated on theassumption of equal sharing of resources withinhouseholds. It is thus likely to understate thepoverty of women. The challenge is to develop anindicator and collect data that would captureresource distribution within as well as betweenhouseholds. This is especially important for thedevelopment of social protection policies and tax-benefit systems that should provide economic andfinancial autonomy for women as well as men.

Education and training

There are three structural indicators relevant towomen and education and training, which arereported annually to the Spring Council,although the EU has not agreed them asindicators in relation to the Platform for Action.These are ‘‘life-long learning’’ (adult participa-tion in education and training); ‘‘science andtechnology graduates’’; and ‘‘the educationalattainment of women and men’’. The data forthese are available from EUROSTAT. How-ever, while these indicators capture the degree ofparticipation of women in these forms ofeducation and training, they omit the extent ofthe segregation of women in areas of educationthat lead to less well-paid jobs.

Women and health

There are some EU-wide data on ‘‘healthy lifeyears’’ by gender which is used in EU analysis ofsocial inclusion (though not all countries pro-

vide data every year). The UNDP uses lifeexpectancy (adjusted by the five years estimatedto be women’s biological advantage), which is areasonable outcome measure. However, there isno agreed indicator on women and health,although this has been under discussion by theHigh Level Group on Gender Mainstreaming.This topic is not included within the EuropeanEmployment Strategy, nor is it in the list ofStructural Indicators.

Violence against women

Several EU Presidencies (Spanish, Danish, Irish,Greek, Dutch, 2002–2004) have made progresson the development of indicators on violenceagainst women (Presidencia de la Union Eur-opea 2002a, 2002b). Three indicators on domes-tic violence have been adopted: the number offemale victims; types of victim support; andmeasures to end violence. Further, there areproposals from the Netherlands for indicatorson sexual harassment in the workplace. In the2002 Joint Report on Social Inclusion, 10 out of15 member states cited domestic violence as oneof the major challenges in relation to socialinclusion concerning gender equality (EuropeanCommission and European Council 2002b), butthere is no structural indicator on this topic.There are three major outstanding challenges.First, to develop and adopt indicators on theremaining forms of violence against women, forexample, rape and other forms of sexual assault,so-called ‘‘honour’’ crimes, and trafficking inwomen. Second, to refine the definition ofdomestic violence so that it can be appropriatelyoperationalised, for example to include not onlyprevalence but also the number of incidents andextent of injury. This is important both toprovide an indication of severity of the violenceand also to ensure that it can be translated intoand thus mainstreamed into crime statistics andthe criminal justice system. Third, to collect datathat are valid, reliable, and meaningful on anannual basis that is comparable between coun-tries. Despite the progress in the development ofindicators, the data to support them does notexist in any member state (although there havebeen several ad hoc one-off studies), let alonecomparable data at the level of EU, nor do thereappear to be any plans to collect it. Thechallenge is to complete the development of

Measuring women’s progress in a global era 381

r UNESCO 2005.

indicators concerning violence against women,and to collect the data necessary to use them.

Women and armed conflict

While there have been discussions about theimportance of the presence of women in peace-keeping forces and armies and of the training ofall personnel in conflict zones in the specialissues that affect women civilians in suchlocations, there is neither an agreed indicatornor relevant statistical data collected on womenand armed conflict.

Women and the economy

The EU has made most progress in the devel-opment of indicators in the area of women andthe economy. This is strongest whether there areoverlaps between the recommendations of Pre-sidencies on Platform for Action indicators withthe requirements European Employment Strat-egy, where there are Structural Indicators thatare supported with data collection directly orindirectly under the auspices of Eurostat. Never-theless, there remain several challenges.

The gender pay gap may appear a straight-forward indicator in that it was agreed as anIndicator by the Belgian Presidency and is also aStructural Indicator. Further, the 2003 Councilrecommended a target of a significant closing ofthe gender pay gap. While there are somereasonable data to support this, there remainsa challenge to collect data that are fullycomparable, inclusive and annual in order tosupport it. For example, some countries providedata for full-time workers only, even though theoperational definition of the Structural Indica-tor includes those working more than 15 hours,thereby in some countries, but not all, excludinga particularly low paying set of workers. There iscurrently a gap in the main source of datacollection, as one survey, the European Com-munity Household Panel (ECHP), stops and itsreplacement, the Survey on Income and LivingConditions (SILC), has not yet started, produ-cing a challenge to provide statistics to supportthe indicator at present.

The EU made progress on indicators of theprovision of care for children and other depen-dants under the French Presidency in 2000 andthe Barcelona Council in 2002 set hard targets,

within the European Employment Strategy, onchildcare, so that member states should provideby 2010 childcare for at least 90% of childrenbetween 3 years and school age by 2010 and forat least 33% of children under 3 years of age.However, there is no structural indicator in thisarea and, currently, an absence of annuallycollected cross-nationally comparable data. Thechallenge is to provide the statistics to supportthe Presidency’s indicators (and indeed theBarcelona targets). Eurostat has announcedplans to include questions about care in twofuture surveys (Eurostat 2004). In the LabourForce Survey for 2005, an ad hoc module willcontain a set of questions on how child care andother care responsibilities are dealt with and willalso ask about the take up of parental leave. Thenew Statistics on Income and Living Conditions(SILC) survey will, in a couple of years, includequestions on childcare. The challenge will be toensure that the plans are delivered, the ques-tioning sufficient, and that the repeats aresufficiently frequent to enable adequate mon-itoring (for example, there appears to be nocommitment to repeat the LFS questions).

Statistical information on two furthertopics relevant to gender equality in the econo-my is managed by Eurostat: the gender gap inemployment, and the gender gap in unemploy-ment. Under the early version of the EuropeanEmployment Strategy, these were both indica-tors of equal opportunities. Raising the employ-ment rate and reducing the unemployment ratehave the status of EU structural indicators.Targets to narrow these gender gaps were set bythe Lisbon Council in 2000, which were qualifiedby age by the Stockholm Council.

Women in power and decision-making

Indicators on women in political decision-making were agreed by the Finnish Presidencyin 1999 and on women and men in economicdecision-making by the Italian Presidency in2003. Statistical data are available on: theproportion of women in the single or lowerhouse of Parliament; the proportion of womenwho are members of national/federal govern-ments; the proportion of women and menamong the members of executive boards of the

382 Sylvia Walby

r UNESCO 2005.

top 50 firms publicly quoted on national stockexchanges.

Institutional mechanisms for theadvancement of women

Although EU member states have providedqualitative information about their institutionalmechanisms for the advancement of women tothe UN as part of the Beijing 110 process, thereis no agreed indicator and no statistical orcomparable information.

Human rights of women, women andthe media, women and theenvironment, the girl-child

There are no agreed indicators for the humanrights of women, women and the media, womenand the environment and the girl-child, althoughthere have been discussions as to the kind ofissues that should be included.

Discussion of EU indicators for thePlatform for Action

In 1998, the EU pledged to develop a simplesuite of indicators to monitor progress on thePlatform for Action to which its member statessigned up in 1995. While there has been someprogress in developing such indicators and thestatistical information needed to support them,this programme of work is as yet incomplete. Insome of the 12 critical areas the EU has agreedindicators, but in some of these the data tosupport them are not available. The develop-ment of indicators has progressed most in areasassociated with the economy, though even hereare there are major gaps, especially in relation tothe adequacy of data. In several areas, there areno agreed indicators. In others, such as violenceagainst women, while there is agreement on thebroad nature of the indicators, there is anabsence of data to support them. The EU hasmost developed indicators in the area of theeconomy, broadly defined to encompass educa-tion, social inclusion and poverty, hence extend-ing across several of the 12 critical areas, notonly ‘‘women in the economy’’. However, in theremaining areas, where the policy lead lies withthe member states rather than the EU, the

development of indicators and supporting com-parable data is much less developed.

The initial appearance of the model ofgender equality implied by the EU indicators isthat of equality through sameness. The target onnarrowing the gender employment gap makes itclear that the aim is to increase the employmentrate of women. However, equality throughsameness is not unequivocally to a malestandard, since the priority and nature of thecommitment to the reconciliation of workingand family life has often meant the regulation ofthe work-place so as to enable the combinationof employment and caring. Thus there is, to alimited extent, a strategy of transformationunderlying the EU model of gender equality.The tension between these two strategies under-lies many of the debates about the nature ofgender equality in the EU and the targets andindicators that are selected for priority attention.

Conclusions

In order to be able to assess the progress of theworld’s women it is necessary to producerelevant indicators and gender-disaggregatedstatistics. These are a key part of the institutionalmachineries agreed in the 1995 UN Platform forAction. While the Millennium DevelopmentGoals attempted to produce indicators andsome (limited) targets appropriate for develop-ing countries, there is much less development inrelation to the more developed countries, despitethe commitment of the EU in 1998 to do this. Inmost of the 12 critical areas there are as yet nowidely agreed indicators appropriate for themore developed countries, and in even fewer arethere appropriate data to support this. Insofar asthere are data it is often uneven, collected on anad hoc basis and inappropriate for comparisonsbetween countries and over time.

There are two underlying issues that affectthe process of development of such indicators.First there are technical difficulties in creatingrobust information that is comparable acrossdifferent countries. The overcoming of thesetechnical difficulties requires resources and com-mitment. Second, there are theoretical questionsas to the nature of the model of gender equality.There is more than one preferred vision of genderequality. There are three types of approach to

Measuring women’s progress in a global era 383

r UNESCO 2005.

this, varying as to the extent to which thestandards are based on an assumption of same-ness, of the equal valuation of different contribu-tions, or the transformation of gender relations.

The EU has demonstrated priorities in thedevelopment of some indicators before others.The EU has a three-fold dynamic in the policyprocess for the development of gender-sensitiveindicators and supporting data. First is thecommitment to the UN Platform for Action;second, is the commitment to gender equalityand gender mainstreaming; third, are the re-quirements of the EES to develop the Europeaneconomy. As a consequence of these develop-ments, in some areas the EU has been a globalleader alongside the UNDP in the production ofindicators to monitor the progress of women.However, progress in the EU has been in veryspecific fields, notably those associated with theeconomy, broadly defined, and primarily as partof the Lisbon Agenda and the EuropeanEmployment Strategy, and in some other areasthe commitment to develop the suite of indica-tors to monitor the development of the Platformfor Action remains unfulfilled.Where the gendermainstreaming agenda overlaps with otheragendas, such as the expansion of employmentin order to boost GDP, then there has beenconsiderable development of institutional ma-chinery, indicators and statistics. Where the

gender mainstreaming agenda is more specific togender equality, justice and women’s humanrights, then the development of institutionalmachinery, indicators, and statistics has beenconsiderably less.

The more fundamental issue is that of themodel of gender equality that underpins thedevelopment of the indicators. A key issue hereis whether it is assumed that paid employmentconstitutes a universal or contingent routetowards the advancement of women. Some ofthe indices that were first developed whichintegrated several domains, such as the UNDPGDI, were heavily weighted towards women’spaid employment andwere criticised by some forthis weighting. The lesson from this has been todevelop and report indicators separately fordifferent areas, rather than to integrate theminto a single index. Such an approach allows forthe testing of various theories as to the nature ofthe relationship between gender relations indifferent domains.

The development of gendered indicatorsoutside employment and its related fields, suchas violence against women, has been slow. Onlywhen there is the full development of indicatorsand the quantitative data to support them in allthe major areas will it be possible to test fullytheories of gender inequality and to measureuneven progress in the advancement of women.

References

ANAND, S. AND SEN, A. 1995.‘‘Computer gender-equity-sensitiveindicators’’, In Human DevelopmentReport 1995, UNDP, ed., NewYork: Oxford University Press.

BACCHI, C. L. 1999. Women, policyand politics: the construction ofpolicy problems. London: Sage.

BALMORI, H. H. 2003. Gender andbudgets: overview report. Falmer,UK: Institute of DevelopmentStudies, University of Sussex.

BARDHAN, K. AND KLASEN, S. 1999.‘‘UNDP’s Gender-related indices: Acritical review’’, WorldDevelopment, 2(6), 985–1010.

BERGER-SCHMITT, R. AND

JANKOWITSCH, B. 1999. Systems ofsocial indicators and social reporting:the state of the art. EuReportingWorking Paper No. 1. Mannheim:Centre for Survey Research andMethodology.

BUDLENDER, D. ELSON, D., HEWITT,G., AND MUKHOPADHYAY, T. 2002.Gender budgets make cents:understanding gender responsivebudgets. London: CommonwealthSecretariat.

BUDLENDER, D., SHARP, R., AND

ALLEN, K. 1998.How to do a gender-sensitive budget analysis:contemporary research and practice.

London: Aus Aid andCommonwealth Secretariat.

CLAVERO, S., AND DALY, M. WITH

BRAITHWAITE, M. 2004.EQUAPOL: Gender sensitive andwomen friendly public policies: acomparative analysis of theirprogress and impact. Report toconference.

COUNCIL OF EUROPE 1998. Gendermainstreaming: conceptualframework, methodology andpresentation of good practices.Council of Europe. EG-S-MS(98)2 rev.

384 Sylvia Walby

r UNESCO 2005.

DIJKSTRA, A. G. 2002. ‘‘RevisitingUNDP’s GDI and GEM: Towardsan alternative’’, Social IndicatorsResearch, 57, 301–338.

DIJKSTRA, A. G. AND HANMER, L. C.2000. ‘‘Measuring socio-economicgender inequality: Toward analternative to the UNDP gender-related development index’’,Feminist Economics, 6(2), 41–75.

ELGSTROM, O. 2000. ‘‘Normnegotiations. The construction ofnew norms regarding gender anddevelopment in EU foreign aidpolicy’’, Journal of European PublicPolicy, 7(3), 457–476.

EUROPEAN COMMISSION 2001a. Noteon gender indicators. Prepared by theWorking Group on GenderIndicators of the AdvisoryCommittee on Gender Equality.DOC-EQOP 60-2001. Brussels:European Commission.

EUROPEAN COMMISSION 2001b.Opinion of the Advisory Committeeon Gender Equality on GenderStatistics and Indicators. EQOP62-2001. Brussels: EuropeanCommission.

EUROPEAN COMMISSION 2002.Employment in Europe 2002: recenttrends and prospects. Luxembourg:European Commission Directorate-General for Employment and SocialAffairs.

EUROPEAN COMMISSION 2003a.Opinion on gender budgeting.Advisory Committee on equalopportunities for women and men,May.

EUROPEAN COMMISSION 2003b.Proposal for a Council Decision onGuidelines for the EmploymentPolicies of theMember States. COM(2003) 176, 26.05.03.

EUROPEAN COMMISSION 2003c.Communication from theCommission to the Council, theEuropean Parliament, theEconomic and Social Committeeand the Committee of the Regions.Strengthening the social dimensionof the Lisbon strategy: Streamliningopen coordination in the field ofsocial protection. Brussels,07.06.2003, COM (2003) 335 final.

EUROPEAN COMMISSION 2003d.Council Decision of 22 July 2003(2003/578/EC) on guidelines for theemployment policies of the MemberStates. Official Journal of theEuropean Union L 197/13,05.08.2003.

EUROPEAN COMMISSION 2003e.Gender Pay Gaps in EuropeanLabour Markets: Measurement,Analysis and Policy Implications.Commission Staff Working Paper.Brussels, 4.9.2003. SEC(2003) 937.Brussels: Commission of theEuropean Communities.

EUROPEAN COMMISSION 2004. Reporton Equality Between Women andMen, 2004. Report from theCommission to the Council, theEuropean Parliament, the EuropeanEconomic and Social Committeeand the Committee of the Regions.Brussels, 19.2.2004. CON (2004) 115final. Brussels: EuropeanCommission.

EUROPEAN COMMISSION AND

EUROPEAN COUNCIL 2002a. JointEmployment Report. Brussels:European Commission.

EUROPEAN COMMISSION AND

EUROPEAN COUNCIL 2002b. Jointreport on social inclusion.Luxembourg: Office for OfficialPublications of the EuropeanCommunities.

EUROPEAN COMMISSION AND PLSRAMBOLL MANAGEMENT 2002.Feasibility study for a EuropeanGender Institute. Brussels: EuropeanCommission.

EUROPEAN PARLIAMENT 2003. Publichearing on ‘‘gender budgeting.European Parliament Committee onWomen’s Rights and EqualOpportunities, 23 January 2003.

EUROSTAT 2004. ‘‘Gender statisticsin Eurostat: overview of projectsand publications’’, Paper submittedby Eurostat and prepared by KarinWinqvist for Conference ofEuropean Statisticians. UNECEWork Session on Gender Statistics.Working Paper 21. Geneva 18–20October 2004.

FELSKI, R. 1997. ‘‘The doxa ofdifference’’, Signs, 23(1), 1–22.

FERREE, M. M. AND GAMSON, W. A.2003. ‘‘The gendering of governanceand the governance of gender’’, InHobson, B., ed., Recognitionstruggles and social movements:contested identities, agency andpower. Cambridge: CambridgeUniversity Press.

FOLBRE, N. 2001. The invisible heart:economics and family values. NewPress.

FRASER, N. 1997. Justice interruptus:critical reflections on the‘‘postsocialist’’ condition. London:Routledge.

GUERRINA, R. 2002. ‘‘Mothering inEurope: Feminist critique ofEuropean policies on motherhoodand employment’’, EuropeanJournal of Women’s Studies, 9(1),49–68.

HALSAA, B. 1998. ‘‘A strategicpartnership for women’s policies inNorway’’, In Lycklama, G.,Nijeholt, A., Vargas, V., andWieringa, S., eds., Women’smovements and public policy inEurope, Latin America and theCaribbean. New York: Garland.

HUBER, E. AND STEPHENS, J. D. 2000.‘‘Partisan governance, women’semployment, and the socialdemocratic service state’’, AmericanSociological Review, 65, 323–342.

JAHAN, R. 1995. The elusive agenda:mainstreaming women indevelopment. London: Zed Books.

KABEER, N. 2003. Gendermainstreaming in poverty eradicationand the Millennium DevelopmentGoals. London: CommonwealthSecretariat.

KECK, M. E. AND SIKKINK, K. 1998.Activists beyond borders: advocacynetworks in international politics.Ithaca: Cornell University Press.

LORBER, J. 2000. ‘‘Using gender toundo gender: a feminist degenderingmovement’’, Feminist Theory, 1(1),79–95.

LUXEMBOURG PRESIDENCY 2005.Report of the LuxembourgPresidency on the Progress made bythe Enlarged EU following the

Measuring women’s progress in a global era 385

r UNESCO 2005.

Beijing Platform for Action (Beijing110). Prepared by: A. M. THEISEN,N. SPODEN, M. VERLOO and S.WALBY. Luxembourg: LuxembourgPresidency.

MANZA, J. AND BROOKS, C. 1998.‘‘The gender gap in US Presidentialelections: When? Why?Implications?’’, American Journal ofSociology, 103(5), 1235–1266.

MAZUR, A. 2002. Theorizing feministpolicy. Oxford: Oxford UniversityPress.

MILLENNIUM DEVELOPMENT GOALS

2004. Millennium DevelopmentGoals. Accessed at: www.developmentgoals.org

NUSSBAUM, M. C. 2000. Sex andjustice. New York: OxfordUniversity Press.

NUSSBAUM, M. C. 2003.‘‘Capabilities as fundamentalentitlements: Sen and social justice’’,Feminist Economics, 9(2–3), 33–50.

OECD-DAC (ORGANISATION FOR

ECONOMIC CO-OPERATION AND

DEVELOPMENT – DEVELOPMENT

ASSISTANCE COMMITTEE) 1998.Sourcebook on Concepts andApproaches linked to GenderEquality. Paris: OECD.

OECD (ORGANISATION FOR

ECONOMIC CO-OPERATION AND

DEVELOPMENT) 2001. OECD BestPractices for Budget Transparency.PUMA/SBO(2000)6/FINAL.Public Management Service, PublicManagement Committee, 15 May2001.

PETERS, J. AND WOLPER, A., eds,1995. Women’s rights, human rights:international feminist perspectives.London: Routledge.

PLANTENGA, J., FIGUEIREDO, H.,REMERY, CH., AND SMITH, M. 2003.Towards an EU Gender EqualityIndex. Feasibility studycommissioned by and presented tothe European Commission. Utrecht,Netherlands: Utrecht School ofEconomics.

PRESIDENCIA DE LA UNION EUROPEA

2002a. Good practice guide tomitigate the effects of and eradicate

violence against women. Prepared bythe Institute of Women,Autonomous Organisation assignedto the Spanish Ministry of Labourand Social Affairs and by MariaJose Diaz Aguado and RosarioMartınez Arias of the UniversidadComplutense de Madrid. Madrid:European Presidency.

PRESIDENCIA DE LA UNION EUROPEA

2002b. Study on measures adopted bythe member states of the EuropeanUnion to combat violence againstwomen. Prepared by the Institute ofWomen, Autonomous Organisationassigned to the Spanish Ministry ofLabour and Social Affairs and byMaria Jose Diaz Aguado andRosario Martınez Arias of theUniversidad Complutense deMadrid. Madrid: EuropeanPresidency.

RAI, S. M. ed. 2003. Mainstreaminggender, democratising the state?Institutional mechanisms for theadvancement of women. Manchester:Manchester University Press.

REES, T. 1998. Mainstreamingequality in the European Union,education, training and labourmarket policies. London: Routledge.

ROSSILLI, M. 1997. ‘‘The EuropeanCommunity’s policy on the equalityof women: From the Treaty ofRome to the present’’, EuropeanJournal of Women’s Studies, 4(1),63–82.

RUBERY, J., FAGAN, C., GRIMSHAW,D., FIGUEIREDO, H., AND SMITH, M.,eds, 2002. Indicators on genderequality in the European employmentstrategy. Prepared by the EuropeanCommission Expert Group onGender and Employment. Brussels:Equal Opportunities Unit,European Commission.

SANDEL, M. J. 1998. Liberalism andthe limits of justice. 2nd edition.Cambridge: Cambridge UniversityPress.

SEN, A. 1984. Resources, values anddevelopment. Oxford: Blackwell.

SEN, A. 1987. On ethics andresources. Oxford: Blackwell.

SEN, A. 1992. Inequalityre-examined. Oxford: ClarendonPress.

SEN, A. 1999. Development asfreedom. Oxford: Oxford UniversityPress.

SEN, A. 2004. Comment at IAFFEconference, Oxford.

SEN, G. 1999. Gender mainstreamingin finance: a reference manual forgovernments and other stakeholders.London: CommonwealthSecretariat.

SHARP, R. 2003. Budgeting forequity: gender budget initiativeswithin a framework of performanceoriented budgeting. New York:United Nations Development Fundfor Women (UNIFEM).

SHARP, R. AND BROOMHILL, R. 2002.‘‘Budgeting for equality: TheAustralian experience’’, FeministEconomics, 8(1), 25–47.

SQUIRES, J. 1999. ‘‘Re-thinking theboundaries of politicalrepresentation’’, In Walby, S., ed.,New agendas for women. London:Macmillan.

UNITED NATIONS 1995. Fourth worldconference on women: platform foraction. New York: United Nations.

UNITED NATIONS DEVELOPMENT

PROJECT 1990. Human DevelopmentReport. New York: UNDP.

UNITED NATIONS ECONOMIC

COMMISSION FOR EUROPE 2004.Regional Overview of Progress:Review of the Implementation of theBeijing Platform for Action and itsRegional Outcomes. Paper for ECERegional Preparatory Meeting forthe 1-year Review ofImplementation of the BeijingPlatform for Action.

UNITED NATIONS ECONOMIC

COMMISSION FOR EUROPE

STATISTICAL DIVISION 2004. DraftReport on the Status of OfficialStatistics related to Gender Equalityin Western Europe and NorthAmerica. Paper for Conference ofEuropean Statisticians. Paperprepared with the contributions ofE. BISOGNO, A. ME and F.

386 Sylvia Walby

r UNESCO 2005.

COMINCINI. UNECE Work Sessionon Gender Statistics. Geneva18–20 October 2004. WorkingPaper No. 4.

VERLOO, M. 2001. Another VelvetRevolution? Gender Mainstreamingand the Politics of Implementation.IWM working Paper No. 5/2001.Institute for Human Sciences,Vienna.

VILLAGOMEZ, E. 2004. ‘‘Genderresponsive budgets: Issues, goodpractices and policy options’’,Background Paper V for theRegional Symposium on

Mainstreaming Gender intoEconomic Policies of the UnitedNations Economic Commission forEurope, 28–30 January 2004, Palaisdes Nations, Geneva.

WALBY, S. 2001. ‘‘From communityto coalition: the politics ofrecognition as the handmaiden ofthe politics of redistribution’’,Theory, Culture and Society,18(2–3), 113–135.

WALBY, S. 2002. ‘‘Feminism in aglobal era’’, Economy and Society,31(4), 533–557.

WALBY, S. 2004. ‘‘The EuropeanUnion and gender equality:Emergent varieties of genderregime’’, Social Politics, 11(1), 4–29.

WALBY, S. 2005. ‘‘Gendermainstreaming: Productive tensionsin theory and practice’’, SocialPolitics, 12, 3.

WOODWARD, A. 2004. ‘‘Buildingvelvet triangles: Gender andinformal governance’’, InChristiansen, T., and Piattoni, S.,eds. Informal governance in theEuropean Union. Cheltenham:Edward Elgar.

Measuring women’s progress in a global era 387

r UNESCO 2005.