MEASURING GENDER INEQUALITY IN EDUCATION IN · PDF fileMeasuring Gender Inequality in...

MEASURING GENDERINEQUALITY IN EDUCATIONIN SOUTH ASIA

Elaine Unterhalter

Titles in this Series

The Move to Programme-Based Approaches – An Effective Partnership for Girls' Education?The Experience of Recent EvaluationsTed Freeman

Educating Girls in South Asia: Promising ApproachesBarbara Herz

Reaching the Girls in South Asia: Differentiated Needs and Responses in EmergenciesAlexandra Mathieu

Mainstreaming Gender for Better Girls' Education: Policy and Institutional IssuesRamya Subrahmanian

Measuring Gender Inequality in Education in South AsiaElaine Unterhalter

Addressing Social and Gender Disparity in South Asia Through SWAps and PBAsin Education: How Can We Use World Experience?Amanda Seel

MEASURING GENDERINEQUALITY IN EDUCATIONIN SOUTH ASIA

Elaine UnterhalterSenior Lecturer, Institute of Education Universityof London, UK

© The United Nations Children's Fund (UNICEF)

Regional Office for South Asia; and

United Nations Girls' Education Initiative (UNGEI)

October, 2006

Short excerpts from this paper may be reproduced for non-profit purposes without authorization

on condition that the source is acknowledged. For longer extracts, permission in advance must be

obtained from the copyright holders via email at [email protected].

The opinions expressed in this paper are those of the author(s) and publication does not

necessarily constitute an endorsement by UNICEF or UNGEI.

Chair, Steering Committee: Susan Durston

UNGEI: Raka Rashid

Series Editor: John Evans

Feedback and correspondence to:

[email protected]

Websites:

www.unicef.org

www.ungei.org

Cover photo: © UNICEF/Bangladesh/SHEHZAD NOORANI

Design and Colour Separations: DigiScan Pre-press, Kathmandu, Nepal

Printing: Format Printing Press, Kathmandu, Nepal.

CONTENTS

Series Foreword ................................................................................................ vAcknowledgements ......................................................................................... viiSummary ....................................................................................................... viii

1. Introduction ............................................................................................... 1

2. Measures of Gender Equality in Schooling ................................................. 3Three Current Measures ............................................................................. 3

Gross and Net Enrolment Ratios (GER and NER) ................................... 3Gender-related EFA Index (GEI) ............................................................. 5Attendance and progression .................................................................. 7

UNESCO Education Development Index (EDI) ............................................. 8

3. Increases in Enrolment and Gender Parity in South Asia –Further Dimensions .................................................................................. 10

4. Gender Equality in Education Index (GEEI) ............................................... 14Towards an Alternative Measure ............................................................... 14The Beyond Access GEEI ......................................................................... 16What Does The GEEI Tell Us? .................................................................. 18

5. Conclusion .............................................................................................. 20

iv

Measuring Gender Inequality in Education in South Asia

References .................................................................................................... 22

Annex ............................................................................................................ 25

About the Author ............................................................................................ 27

v


SERIES FOREWORD

There is a growing sense of momentum around education in South Asia.Governments are engaged and a lot has been done. The Millennium DevelopmentGoals have added an additional spur to action as indeed have greater awarenesson gender disparity and the need for educated workers. There is though a longway to go if the rights of all children are to be realised.

Providing access to education is only part of the story. Once children are enrolledand attending, the quality of their education must make it a worthwhileexperience. The special needs of girls in the social and cultural context of SouthAsia call for special measures, as do the needs of all children in situations ofconflict and emergency. South Asia has many rich, positive examples of successin advancing basic education. It is important that these are shared and built on ifthere is to be an overall improvement throughout the region.

This series of papers aimed at promoting better education in South Asia grew outof collaboration between the UNICEF Regional Office for South Asia and thenewly formed UN Girls' Education Initiative, and had its genesis at a RegionalMeeting on Accelerating Girls' Education in South Asia in February 2005.

Essentially the series is intended to be a forum that allows debate, exchange ofideas and to break new ground. It will aim to capture the momentum and extolgood practice to all engaged in educational policy and implementation.

v

vi


The series does not seek to represent a specific viewpoint, but rather is intendedto enable specialist contributors to present issues in greater depth and breadththan is often the case in official documents.

Initially the series will focus on girls' education but it is hoped that eventually itwill broaden into a platform for more general education issues related to SouthAsia, with a particular emphasis on social inclusion. Contributions and feedbackare invited from academics and practitioners from throughout the South Asiaregion and beyond. The series editors are particularly interested in submissionswhich offer new ideas and strategies that can assist those needing answers, andwhich can add impetus to the ongoing efforts in the region to provide qualityeducation for all.

Come, join the debate!

vii


ACKNOWLEDGEMENTS

This paper is based on work commissioned from the Beyond Access project in2004–05 by the Commonwealth Secretariat, UNESCO Bangkok and UNICEFRegional Office for South Asia. It develops and adapts ideas set out inUnterhalter, Rajagopalan and Challender (2005) and Unterhalter and McCowan(2005). I am grateful to all three commissioning organizations for permission topublish this paper, which draws on reports prepared for them.

Thanks are due to Joseph Crawford and Jacob Steel for help in developing theformula for calculating the GEEI and to Tristan McCowan for research assistance.

viii


SUMMARY

This paper assesses current measures for gender equality in education. Whilegender disaggregating enrolment and progression data has been an importantstep forward, measures that are concerned only with the presence or absence ofgirls in school provide little insight into aspects of gender inequalities withinschools or the gendered environment beyond school. Nevertheless, suchmeasures do generally show improvement for nearly all the countries in SouthAsia for which there is data.

Current measures of education quality underplay the significance of genderequality as an aspect of quality. In reviewing economic and gender indicatorstogether with education indicators for South Asia, no consistent pattern emerges.

The paper presents the Gender Equality in Education Index (GEEI) – a measuredeveloped to draw on existing data sources to consider gender equality ineducation in more dimensions than simply enrolments. The GEEI for thecountries in South Asia for which there is data shows that gains were made overthe 1990s in all countries except Pakistan. However, to reach a GEEI scoreconsonant with achieving the gender and education MDGs will require sustainingthe huge mobilization of the previous decade and in many countries increasingthis. The assessment of GEEI gives some indication of the size of the task andthe levels of mobilization needed.

1

MEASURING GENDER INEQUALITY IN EDUCATION IN SOUTH ASIA

INTRODUCTION

1

Millennium Development Goal 3 forgender equality and the empowerment ofwomen is the goal that was set with theearliest date for achievement – 2005.The thinking behind this was that thepersistence of gender inequalities wouldundermine all the other MillenniumDevelopment Goals (MDGs). Gender –unlike class, race, ethnicity or disability– is one form of social division commonto all societies across the world.Addressing gender inequality is anindicator of commitment to theprinciples that underlie all the MDGs,that is that each and every person on theglobe is the subject and object of moralconcern for the whole population andthat we accord each other reciprocalrespect so that each of us may live a lifeof dignity, without suffering or harm,particularly with regard to poverty,hunger, ill health, lack of education,

discrimination, or the consequences ofenvironmental degradation.

The target associated with MDG3 was anend to gender disparities in primary andsecondary education. We know now thatthis target has been missed in 94 out of149 countries for which data isavailable (UNESCO, 2005). This is amatter of grave concern. It is shamefulin a world where huge wealth isamassed, where enormous talents exist,together with great eagerness foreducation, that in five years we have notbeen able to enrol equal numbers ofgirls and boys in school and ensure thatequal numbers pass just five grades. Ifwe miss the target for this first MDG arewe not calling into question the wholeproject? I fervently hope we are not. Itrust that the fact that we have missedthe first MDG is a wake-up call to us all

2


to think better, discuss deeper, and putthe insights we distil into action.

This paper analyses the progress thathas been made in relation to MDG3 andgives some pointers as to whatprocesses have been associated withthat progress. It also shows how muchfurther countries in South Asia need togo in order to fulfil the educationaspirations associated with MDG3 by2015. Reflection on these patternsindicates where we need to allocateresources and how we can use theglobal failure in relation to the first MDGto galvanize specific actions to helpachieve all the other seven.

Developing an understanding of how toimprove gender equality in education isnot a simple matter. One of the manychallenges that it poses is the nature ofthe information base. Qualitative workhas provided insights into how social

relations inside and outside schools shapegender inequalities and can contribute tochange; but generally this work is basedon small-scale, in-depth research.Qualitative work can deepen knowledgeabout the nature of gender inequalities,but it cannot provide an overview of theirextent, or an indication of whereadditional resources to address theirconsequences would best be deployed.For this purpose, quantitative work isneeded – but there are major problemswith the methods by which genderinequality in education is currentlymeasured. This paper critically reviewsthe existing measures of gender equalityin education used by internationalagencies and governments. It goes on topropose alternative forms of measurementwhich seem better able to capture theaspirations of Education for All and theunderstanding that gender equalityrequires forms of counting that themselvesgo beyond mere statistics about access.

3


more punctilious collection of gender-disaggregated data on primary grossenrolment ratios (GER) and netenrolment ratios (NER). GER is thenumber of children enrolled in school,expressed as a proportion of the childrenof a specific age cohort (say 5–11) whoshould be enrolled in school. GER cansometimes be more than 100 per cent ifthere are large numbers of under-ageand over-age pupils in school. NER isthe number of children in the appropriateage group enrolled in school, expressedas a proportion of the official age grouprequired to be in school. GER and NER,even when disaggregated by gender, onlygive us a picture of the number ofchildren on the school register. They cantell us nothing about whether childrenattend regularly, once registered;whether they complete gradessuccessfully; or whether passing a grade

Three Current Measures

There are currently three measures usedfor assessing gender equality in school.Enrolment rates are assessed throughgender-disaggregated gross and netenrolment rates. A more complexmeasure that seeks to measure genderparity, that is whether or not there areequal numbers of boys and girls at alllevels of the education system, is thegender related EFA index (GEI). A thirdapproach is to look at gender-disaggregated data on retention andprogression.

Gross and Net Enrolment Ratios (GER andNER)The mobilization of resources forEducation for All (EFA) after the Jomtienconference in 1990, and the follow-upmeeting at Dakar in 2000, resulted in

MEASURES OF GENDER EQUALITY IN SCHOOLING

2

4


means that children have acquiredknowledge that they can use outside aschool context. In addition, because inmany countries children's births are notregistered, NER is often based onestimations.

Generally GER and NER data are basedon the Education ManagementInformation System (EMIS) of educationministries, and then passed to theUNESCO Institute for Statistics for thecompilation of international datasets.EMIS is only as good as the relations oftrust, truthfulness and accuracy thatunderpin that system. In some contextslocal officials do not know the reasonswhy they collect data for EMIS. Theymay have difficulties in reaching areasthat are socially or geographicallydistant to collect information; they maybelieve that underestimating oroverestimating children on the schoolregister may bring additional facilities toa locality. Carr-Hill et al. (1999)emphasize the fragility of the data onwhich many national and internationalconclusions are based. Whenparticipatory activities are held invillages to identify children who are notat school, more robust data isassembled. However, there aredifficulties in translating localmobilizing actions into official data onGER and NER, although in somecountries this form of micro-planning isused by governments as well as NGOs.

Throughout the 1990s gender-

disaggregated GER and NER in manycountries showed a gap between girls'and boys' levels of enrolment. Thisgender gap came to be seen as a majorsource of concern, demonstrating that inmany countries fewer girls than boyswere enrolled in primary and secondaryschools (although it should be noted thatthere can be gender parity, that is equalproportions of boys and girls in school,when there is low GER or NER). So greatwas the concern with reaching genderparity – that is, equal numbers of girlsand boys in school – that this becamethe chief indicator for the MillenniumDevelopment Goal 3 on theempowerment of women (MillenniumCommission, 2000).

What changes in girls' NER have beenevident in South Asia since the early1990s? Table 1 lists countries for whichwe have data for 1990 and 2001. Thereis, however, some debate about howaccurate the official data for Pakistanand Afghanistan is. The Afghanistan datareflects the situation in 2003 when therewas a huge drive to enrol girls in school.The extent to which this has beensustained is the matter of critical review(Beyond Access, 2006) Unfortunatelythere is not a full set of data forMaldives for this period.

The table shows that in three countriesin the region by around 2000 nearly 90per cent or more girls of the appropriateage were in primary school. In anotherfour between two-thirds and three-

5


quarters of all girls in this age groupwere in school, indicating considerabledifficulties in enrolling all girls inschool. In Pakistan there appeared tohave been a fall in NER with only 50 percent of girls in the age group enrolled.

Table 1 also shows that through the1990s all countries in the region forwhich there is data, with the exceptionof Pakistan, made percentage gains inthe levels of girls' enrolment. For somecountries (Nepal and Afghanistan)despite the decade being marked byconflict, these gains were enormous. ForIndia and Bangladesh the percentagegain was sizeable. Only in Pakistan isthere a large percentage fall. From thedata held by UN bodies we cannotdetermine whether this is because ofincomplete data or data that has beenwrongly processed. With the exception ofPakistan, using only girls' NER thepicture for the region would be one of

steady growth and reasonable optimism.However, the problems with NERoutlined above entail some doubts aboutwhether this is an adequate enoughpicture of levels of gender equality ineducation.

Gender-related EFA Index (GEI)In contrast to the measurement ofprimary enrolment through gender-disaggregated GER and NER, thegender-related EFA index (GEI),developed by UNESCO for use in itsGlobal Monitoring Reports, is an attemptto indicate the extent to which boys andgirls are equally present at differentlevels in the education system (primary,secondary and adult education).However, a country can have a GEI of 1,indicating complete equality betweenboys and girls, but still have low rates ofaccess, retention and achievement forgirls and boys. For example, in 2003Myanmar had a GEI of 0.949, with only

TABLE 1TABLE 1TABLE 1TABLE 1TABLE 1 Percentage gain in girls’ NER, South Asia: c.1990–c.2003Percentage gain in girls’ NER, South Asia: c.1990–c.2003Percentage gain in girls’ NER, South Asia: c.1990–c.2003Percentage gain in girls’ NER, South Asia: c.1990–c.2003Percentage gain in girls’ NER, South Asia: c.1990–c.2003

Country Girls' NER c.1990 Girls' NER c.2001 Percentage Improvementof Girls' NER c.1990–c.2001

Sri Lanka 90 100 11%Bangladesh 66 87.5 33%India 61 75.7 24%Nepal 41 66 61%Bhutan 50 60 20%Pakistan 62 50 -19.5%Maldives n.a. 92Afghanistan 27 67* 148%

Source: Derived from Unterhalter, Rajagopalan and Challender, 2005; UNDP, 2004; Maldives, 2000;World Bank, 2005* 2003 figure

6


84 per cent primary NER for girls, andKuwait had a GEI of 0.966 with aprimary NER of 83 per cent for girls(UNESCO, 2003: 288-9). Gender parityon its own cannot tell us much aboutgender equality in relation to accessingeducation, progressing through school,and living in a gender-equitable societyafter school.

In Table 2 the improvements in GEI forcountries in South Asia are presented.Unfortunately no calculations wereavailable for Bhutan, Maldives andAfghanistan.

The table shows that in all the countriesof the region there has been animprovement in GEI. In Sri Lanka thereis gender parity at all levels, with asmall bias against boys. The othercountries of the region for which there isdata had roughly similar GEI levels

around 2001, and had improved at moreor less the same level, although Nepalhad made enormous gains from a muchlower base in 1990.

How can we explain the increasing levelof GEI in Pakistan and the falling levelof NER (assuming the figures in Table 1are roughly accurate)? It may be thecase that richer families in Pakistanhave supported their daughters intosecondary school and this hasmaintained reasonable levels of adulteducation over the decade resulting ingender parity at all levels. Thus we see arise in GEI over the period in Pakistanhappening at the same time as poorerchildren have been denied access toschool. The increase may also indicateprogrammes to bring certain girls intosecondary school and set up adultliteracy without concomitant attention tothe expansion of provision for girls in

TABLE 2 Percentage gain in GEI, South Asia: c.1990–c.2001

Country GEI c.1990 GEI c.2001 Percentage Improvement of GEIc.1990–c.2001

Nepal 0.42 0.68 62%Bangladesh 0.62 0.76 23%Pakistan 0.52 0.645 23%India 0.64 0.74 16%Sri Lanka 0.98 0.99 3%Bhutan,Maldives,Afghanistan n.d. n.d.

The shaded figure denotes gender disparity at the expense of boys/men; that is, in Sri Lanka there aremore girls than boys in school

For the purposes of this analysis, movement towards disparity in favour of girls/women is included in thecalculation for improvement

Sources: UNESCO, 2004, 1995; UNICEF, 1994

7


primary school. There may also be somestrong regional variations resulting inthis unevenness.

This raises questions about theaggregated data at country level.Increases in GEI may reflect advances incertain regions and not others. All that isrequired for a high gender parity score isequal numbers of enrolments of girls andboys, not a high overall enrolment as aproportion of the age group which NERmeasures.

The case of Pakistan points tohow important it is to not rely on asingle measure, be it NER or GEI, inorder to measure gender equality inschooling.

Attendance and progressionA third measure of gender equality in

education has focused not just onenrolments, but on attendance andprogression. From the late 1990s,gender-disaggregated data has becomeavailable on progression through school(that is, the completion of primary andsecondary school), with data oftenavailable by district. This gives richerinsight than mere enrolment figures canprovide on whether or not an educationsystem is delivering gender equity inprogression. The picture for South Asiais presented in Table 3.

Table 3 shows a different pattern again.While for all the countries for whichthere is data the percentage of girls whoattend school regularly is much lowerthan the percentage of boys, generallygirls who remain in school pass 5 gradesof primary school at a slightly higherrate than boys.

TABLE 3TABLE 3TABLE 3TABLE 3TABLE 3 Girls' (and boys') primary school attendance and progression over five years, South Asia:Girls' (and boys') primary school attendance and progression over five years, South Asia:Girls' (and boys') primary school attendance and progression over five years, South Asia:Girls' (and boys') primary school attendance and progression over five years, South Asia:Girls' (and boys') primary school attendance and progression over five years, South Asia:

c.1990-c.2001c.1990-c.2001c.1990-c.2001c.1990-c.2001c.1990-c.2001

Country Girls' (boys') net primary school Girls (boys) completing fiveattendance 1992-2002 % years primary school c.2001 %

Afghanistan 14 (58) n.a.Bangladesh 78 (77) 68 (63)Bhutan n.a. 93 (89)India 73 (79) 63 (60)Maldives n.a. n.a.Nepal 66 (79) 81 (75)Pakistan 51 (62) n.a.Sri Lanka n.a. n.a.

Source: UNICEF, 2004; UNESCO, 2004

8


In all these three measures of enrolment, parity and progression, 'gender' is viewed asmerely the numbers of boys and girls entering and progressing through a schoolsystem. These forms of measurement give no indication of gendered relations ofpower in schooling, which may compromise the numbers who progress. This approachto measuring gender does not provide information on the ways in which genderequality or inequality links with other dimensions of human flourishing, for examplehealth, access to decision making, the labour market, or income.

In fact, these figures can give an impression quite at odds with the literature basedon qualitative research in a country. In South Africa, for example, quantitative datashows high levels of gender equality in access and progression, but qualitative datahighlights danger at school from sexual harassment and violence, girls' anxiety abouttheir futures, and considerable discrimination against many women teachers(Unterhalter, 2005a).

UNESCO Education DevelopmentIndex (EDI)

The UNESCO Global Monitoring Reporttried to develop a definition of quality inschools, linking it analytically withequality. Koïchiro Matsuura, Director-General of UNESCO, in his Foreword tothe 2005 volume, commented that'Quality must pass the test of equity: aneducation system characterized bydiscrimination against any particulargroup is not fulfilling its mission'(UNESCO, 2004).

However, the 2004 Report itself payslittle attention to gender dimensions ofquality and does not suggest what the'test of equity' might be. The GlobalMonitoring Report focuses instead onfour proxy measures for quality, only one

of which has data with a genderdimension. The measures used arepupil:teacher ratios, teachers'qualifications, expenditure on education,and learning achievements. Only thislast has been measured with respect togirls and boys. The failure of the GMR tolink quality substantively with equity is aproblem of both measurement andanalysis. The assumption of the report isthat girls and boys enter schools whichare unmarked by gender with regard toquality. The gender-neutral ways inwhich quality has been assessed offer noopportunity to understand the similar ordifferent achievements of children.

In an attempt to bring togetherinformation on access, quality, and thegender gap, UNESCO developed theEducation Development Index (EDI). The

9


EDI constituents and related indicatorsare:

universal primary education: netenrolment ratio;

adult literacy: literacy rate of the groupaged 15 and over;

gender: gender-specific EFA index(GEI, the arithmetical mean of theGender Parity Indices for the primaryand secondary gross enrolment ratiosand the adult literacy rate);

progression: survival rate to grade 5.

The problem with the EDI with regard togender is threefold. Firstly its maingender component, the GEI, isconcerned with parity, which, asdiscussed above, gives insufficientinsight into context. Men and women, orgirls and boys, may have gender parity inliteracy or access to schooling but havelow levels of participation.

Secondly the EDI does not take accountof gender in children's survival inschooling. It primarily considers genderin relation to access and notachievement.

Thirdly the EDI weights each of its fourcomponents equally. Thus enrolments,and gender parity in enrolments, areweighted equally with achievements.However, research in many countriesshows that enrolling children in schoolis only the first hurdle. Ensuringattendance and completion are muchharder tasks, and this is particularly thecase for girls, whose progress isconstrained by many factors linked tosafety, hygiene, nutrition and familyresponsibilities (Watkins, 2000;Tomasevski, 2003). Weighting accessas equivalent to achievementunderestimates the EFA challenge thatconfronts governments, but it isparticularly serious because of itsfailure to adequately assess gender-related aspects of school achievement.

Table 1 painted a particular picture ofefforts in relation to increases inenrolments in South Asia. But can thesignificance of this be interpretedwithout further information? We nowhighlight how considering additionalinformation on the region allows us tosee some further dimensions of thecontext in which increases inenrolment and gender parity haveoccurred.

10


INCREASES IN ENROLMENT AND GENDER PARITYIN SOUTH ASIA – FURTHER DIMENSIONS

Table 4 shows that the level of GDPimprovement between 1990 and 2000does not correlate with the levels ofimprovement in NER. Thus there is aspectacular percentage gain in girls'primary NER in Nepal, with more than60 per cent improvement but only 20per cent growth in GDP per capita. InBangladesh the growth in NER is alsoconsiderably less than the growth inGDP per capita. By contrast, the smallerpercentage gain in girls' NER in Indiaseems out of step with the huge growthin GDP per capita of 132%. The size ofthe fall in girls' NER in Pakistan seemsenormous compared to the level ofreduction in GDP per capita.

Table 4 also shows that considerableincreases in girls' primary NER havebeen made in Bangladesh and Nepaleven with more than one-third of the

population living on extremely lowincomes, but with a smaller proportionof the population living in extremepoverty in Pakistan gains on a similarscale were not made.

Some further insight can be gained inrelation to the pattern of improvementand decline in girls' NER if we comparethe information in Tables 1 and 2 withthat in Table 5, which charts rises andfalls in countries' Human DevelopmentIndex (HDI).

Table 5 shows that there have been risesin HDI in all the countries for which wehave data, raising questions about thepattern seen in Table 1 concerning NERand Table 4 concerning income. Whatwe can see from looking at levels ofgirls' NER, income, wealth and humandevelopment is interesting, but not

3

11


TABLE 4TABLE 4TABLE 4TABLE 4TABLE 4 GDP (PPPGDP (PPPGDP (PPPGDP (PPPGDP (PPP, US$) per capita, South Asia: 1990–2000; and % population living below $1 per day 1991-, US$) per capita, South Asia: 1990–2000; and % population living below $1 per day 1991-, US$) per capita, South Asia: 1990–2000; and % population living below $1 per day 1991-, US$) per capita, South Asia: 1990–2000; and % population living below $1 per day 1991-, US$) per capita, South Asia: 1990–2000; and % population living below $1 per day 1991-

20022002200220022002

Country 1991 2002 Percentage Change Percentage Populationin GDP per Capita Living Below $1 per Day

Afghanistan 700 n.a.Bangladesh 1160 1700 47% 36%Bhutan 620 n.a.India 1150 2670 132% 34%Nepal 1130 1370 21% 37%Maldives n.a. n.a.Pakistan 1970 1940 -2% 13%Sri Lanka 2650 3570 35% 7%

Source: UNDP, 2004

decisive with regard to assessing whatprogress has been made towards theMDG and what processes facilitate this.Enrolment data does not tell us anythingabout gender equality in attendance,participation or achievement as a resultof schooling. GDP per capita does notgive us the multidimensional picture ofhuman flourishing painted by the HDI,while the HDI itself is not gendersensitive enough.

The Gender Development Index (GDI)and the Gender Empowerment Measure(GEM) provide additional perspectives.The GDI is an index made up of the levelof gender inequalities in life expectancyat birth, adult literacy rate, the grossenrolment ratio at primary, secondaryand tertiary levels of education, andearned income. While the GDI measuresconditions for health, income andeducation, the GEM measures the extentto which women are able to use these

TABLE 5TABLE 5TABLE 5TABLE 5TABLE 5 Rise in HDI, South Asia: 1990–2002Rise in HDI, South Asia: 1990–2002Rise in HDI, South Asia: 1990–2002Rise in HDI, South Asia: 1990–2002Rise in HDI, South Asia: 1990–2002

Country HDI 1990 HDI 2002 PercentageImprovement

of HDI1990–2002

Afghanistan n.a. n.a.Bangladesh 0.417 0.509 22%Bhutan 0.511India 0.519 0.595 15%Maldives 0.751Nepal 0.418 0.504 21%Pakistan 0.444 0.497 12%Sri Lanka 0.698 0.740 6%

Source: UNDP, 2004

resources to play an equal role in thepolitical, economic and social life oftheir countries. The GEM is an index ofthe percentage of seats in parliamentheld by women, female legislators,senior officials and managers, femaleprofessional and technical workers, andthe ratio of estimated female to maleearned income.

12


TABLE 6 Improvers' Gender Development Index, South Asia: 1993–c.2003

Country Gender Development Gender Development Index Percentage ImprovementIndex (GDI) 1993 (GDI) c.2003 in GDI 1993–c.2003

Afghanistan n.a. n.a.BangladeshBhutanIndia 0.410 0.574 40%Maldives 0.599 0.735 23%Nepal 0.308 0.479 56%Pakistan 0.383 0.469 22%Sri Lanka 0.679 0.726 7%

Source: Derived from Unterhalter, Rajagopalan and Challender, 2004

TTTTTABLE 7 Improvers' Gender Empowerment Measure, South Asia: c.1995–2000

Country GEM 1995 GEM 2000 Percentage Change in GEM 1990–2000Sri Lanka 0.409 0.272 -33%Bangladesh 0.309 0.218 -28%Pakistan 0.414

Source: UNDP, 2003; UNDP, 2001

Table 6 shows that for Nepal and Indiasignificant increases in girls' NER andGEI from the 1990s to 2000 have beenaccompanied by high levels ofimprovement in the GDI. Indeed in bothcountries the improvement in GDI issignificantly greater than theimprovement in HDI. Pakistan, wherethere was a fall in NER but an increasein GEI, shows increase in GDI. SriLanka, with consistently high levels ofNER and GEI, has a relatively low levelof increase in GDI and a spectacular fallin GEM, possibly due to the high level ofwomen's migration, reducing theirnumbers in political processes andobscuring the amounts of their earnedincome in national data.

It can be seen that differentinformation for the region paintsdifferent pictures. In summarizing theoverall picture Janet Raynor hasdeveloped the overview tablereproduced as Table 8. The 'best' ineach category has a heavy outline andfigures shown in bold; the 'worst' areshown in shaded boxes. She points outthat there are many gaps, which canmake comparison difficult ormeaningless, and in many countriesthe accuracy of the data isquestionable. Missing or inaccuratedata is common in the countries withsections of the population living indistricts with little communicationwith the capital or experiencing war.

13


The spread of outlined or shaded boxesshows that there is no clear pattern. Thatis, there appear to be no consistent linksbetween overall human development,wealth, gender or education. In eachcountry, there is a complex relationamong factors that impact on enrolment,progression and the interplay ofeducation with other aspects of socialdevelopment.

What do these trends tell us? Theinformation suggests that evaluatingprogress on MDG3 requires not onlymeasures of enrolment and genderparity, even though these are the targets.

To give expression to the spirit of theMDG we need ways of understanding notonly how girls and women are able toaccess education in equal numbers toboys and men, but also how they areable to put that education to good use.We need an indicator that will show howgirls and women are able to use primaryschooling to secure better health,access to more education and higherincomes. The disjunctures betweenimprovements in girls' NER, GEI andGDI in Pakistan, outlined above,suggest the need for a different way ofmeasuring achievements towards theMDG.

TABLE 8 Summary indicators for the region

HD

I ra

nk(o

ut o

f 1

77

cou

ntri

es)

GE

M(o

ut o

f 8

0 c

ount

ries

)

PP

P U

S$

,2

00

3

Pub

lic

expe

ndit

ure

on e

duca

tion

(% G

DP,

20

00

–02

)

Adu

lt l

iter

acy

rate

%F:

M,

20

02

–03

Pri

mar

y F:

M r

atio

,2

00

2–0

3

Sec

onda

ry F

:M r

atio

,2

00

2–0

3

Tert

iary

F:M

rat

io,

20

02

–03

Com

bine

d G

ER,

prim

,se

c, t

erti

ary

% F

:M,

20

02

–03

1 2 3 4 5 6 7 8 9

Afghanistan - - - - - - - - -Bangladesh 139 79 1770 2.4 31/50 1.04 1.11 0.50 54/52Bhutan 134 - 1960 5.2 - - - - 14/16India 127 - 2892 4.1 48/73 0.94 - 0.68 56/64Maldives 96 - - - 97/97 1.00 1.15 - 75/74Nepal 136 - 1420 3.4 35/63 0.88 - 0.34 55/66Pakistan 135 71 2097 1.8 35/62 0.74 - 0.81 31/43Sri Lanka 93 72 3778 - 89/92 - - - 69/67

Source: Raynor, 2006

14


GENDER EQUALITY IN EDUCATION INDEX (GEEI)

Towards an Alternative Measure

There are three major problems with theexisting measures of gender equality andinequality in education. Gender parityand the gender gap are inadequatemeasures of gender equality, becausethey do not acknowledge context.Existing measures of quality obscure thegender question. The EDI fails to takefull account of the significance ofgender inequality in achievement. Theseproblems have led the Beyond Accessproject to develop a new measure whichbetter expresses the aspiration forgender equality in education.

The approach has been developed as acontribution to the debate about apublicly accountable criterion of justicein terms of gender equality in education.Thus the approach is offered partly in the

4

hope that it will elicit useful criticaldiscussion. The approach tomeasurement draws on the workundertaken by Amartya Sen and MarthaNussbaum, who distilled a generalapproach to human flourishing based oncapabilities and human rights (Sen,1999; Nussbaum, 2000). These ideashave been operationalized in the UNDP'sHuman Development Reports, whichhave developed the Human DevelopmentIndex (HDI) and the Gender DevelopmentIndex (GDI) (Fukuda Parr and Kumar,2003; UNDP, 1995-2004). A number ofwriters explore capabilities in relation toaspects of education (Alkire, 2002;Unterhalter, 2003; Unterhalter andBrighouse, 2003; Terzi, 2004; Walker,2004; Unterhalter, 2005b).

A key idea when measuring capabilities,valued doings and beings, is that they

15


are multi-dimensional. Measuringgender equality in education is not onlyabout recording the gender gap inenrolments of girls and boys in school(gender parity), but about measuringsome of the other cross-sectoral aspectsof gender equality and equity in relationto health, wealth and decision-makingwhich all have a bearing on genderequality in school. A second aspect ofcapabilities is that, while the concepthas particular strengths with regard toother measures of equality – forexample, people's expressed desires, oraggregated utility (the greatest good forthe greatest number) – the more oneneeds to draw comparison at a cross-country level, the less fine-grained arethe capabilities that can be measured,and the more one has to rely either onmeasures of resources (like access toschool) or on other routinely collecteddata that can act as some kind of proxyfor capabilities (Unterhalter andBrighouse, 2003; Unterhalter, 2005a).

There are many problems withdeveloping a quantitative measure ofgender equality in education. Itrepresents the interrelationship betweencountries or regions as competitive –creating a culture of winners and losers– when in fact they are deeplyinterlinked and in need of each other'ssupport. It sets up an arbitrary board ofscorers, who usually have littleexperience of delivery, to judgeperformance. And it tends to extinguishthe processes of working towards

achievement. These are compellingreasons not to proceed down this path ofanalysis, relying either on scorecards oron quantitative measures of genderequity or equality. However, alongsidethese arguments must be considered theconfusion that results from not knowingwhich countries or districts areimproving gender equity or equality ineducation; which areas need resources,and why we deem this to be the case;and in what areas countries can learnfrom each other. These reasons, basedon harnessing available resources towork together on developing amethodology for measurement of aproblem of global significance, seem tomitigate to some degree the negativedimensions described above(Unterhalter, 2005c).

However, it should be stressed that theutilization of this or any other version ofmeasurement of gender equality ineducation should not be a substitute fordetailed quantitative and qualitativeresearch. A key dimension that requiresconsideration in any form ofmeasurement is an analysis of socialand cultural relations and theopportunity for dialogue, debate and theexploration of differences, particularlywith regard to the public–privateinterface. Such work must be conductedrigorously to provide a corrective to thesimplifications and crude assumptionsof any approach based on scorecards orleague tables. Only in-depth analysiswill furnish the detailed knowledge of

16


local contexts and actions necessary totake forward any of the very generaldirections that measurements of genderequality in education might point to.

The Beyond Access GEEI

Bearing these issues in mind, theBeyond Access Project scorecard forgender equality in education wasdeveloped. This has since been renamedthe Gender Equality in Education Index(GEEI). The GEEI puts together datagathered by UNICEF on girls' attendanceat school, by UNESCO on girls'achievement in primary school andaccess to secondary school, and byUNDP on the gender development index(GDI). The GDI is a measure thatconsists of the distribution of female tomale life-expectancy in a country,literacy and enrolment in school, andestimated earned income. Each of thethree indicators (life-expectancy,education and income) is equallyweighted when compiling the index,although the education index gives two-thirds weight to the adult literacy indexand one-third weight to the grossenrolment rate (UNDP, 2003: 343-4).

The Beyond Access GEEI has beendeveloped to assess both access andretention in broader ways than hitherto.It includes not only the numbers of girlswho attend and remain in primaryschool, but also an assessment ofwhether those girls are able to translatethat attendance and retention into future

secondary schooling, healthy lives andreasonable incomes. Four widely usedmeasures have been used to develop theGEEI for girls' access to and retention inschool:

Girls' net attendance rate at primaryschool

Girls' survival rate over five years inprimary schooling

Girls' secondary Net Enrolment Ratio(NER)

A country's Gender DevelopmentIndex (GDI).

These measures were selected becausethey indicate access to primaryschooling (net attendance rate), derivedfrom household surveys; retention inprimary schooling (survival rate);potential of the education system togenerate teachers and managers withsome concerns with gender equality(girls' secondary NER); and thepossibilities for these women to surviveand flourish as adults (GDI).

The Beyond Access GEEI is not anunweighted index. In compiling theindex, girls' survival over five years inprimary school and the capacity ofwomen to survive into adulthood, retainliteracy and earn a decent livelihood(signalled by the GDI) were weighted astwice as important as attendance inprimary schools. Girls' enrolment in

17


secondary school, which it is believedpoints to the emergence of a cadre ofwomen who will work in socialdevelopment with some orientationtowards gender equality and equity, wasweighted as 50 per cent more importantthan attendance. (See Annex for anexplanation of how the GEEI iscalculated.)

There are a number of critiques of theGDI as a measure of gender equality.Charmes and Wieringa point out that theGDI measures general welfare, ratherthan gender inequality. In fact, thevalues for the GDI are very similar tothose for the HDI, particularly forcountries with high human development(Charmes and Wieringa, 2003). Theyalso point out that the choice ofmeasuring health by using life-expectancy, a very long-termmeasurement, is not likely to offer aprecise indication of women's health forthe current time period – unlike, forinstance, infant and child mortalityrates. Other critics point out thatstatistics of earned income do notinclude the work that women do in thesubsistence economy (Elson, 1999).Pogge (2002) argues that thecalculations for the HDI and GDI areproblematic, because it is implied thatgains in one dimension, for example inschooling or income, can be traded offagainst losses in another, for examplelongevity. These are substantialcriticisms, but in our view they do notnegate the usefulness of the GDI as an

easily accessible measure of someaspects of gender in relation to humanwell-being, which is why it has beenused in the GEEI.

The UNDP's Gender EmpowermentMeasure (GEM) might have been a bettermeasure of how girls and women areable to translate their education intoearning and political decision-making.The GEM is an average of three indices:women's share of parliamentaryrepresentation; their economicparticipation through share of positionsas legislators, senior officials, managersand professionals; and their share ofearned income. There are, however, alsocriticisms of the GEM. For example, thenumber of seats in parliament occupiedby women does not fully indicate howmuch power women actually have.Emphasizing women's earned income inthe formal sector undervalues women'searnings in the informal sector or careeconomy, where a great many exchangesthat are of value to women and theirsocieties take place. Nevertheless,despite these limitations, the GEM doesprovide a proxy measure of the level towhich women are visible in key politicalposts, earn equivalent amounts to menin the formal sector, and haveprofessional employment. Unfortunately,however, the GEM has not beencalculated for many countries, and if ithas been calculated recently there areno time-series data, so comparisonscannot be made. Because GDI hasgenerally been calculated for the early

18


1990s and 2000s for most countries, ithas therefore been used in GEEI insteadof GEM.

Table 9 presents the GEEI for South Asiabetween c.1993 and c.2001. This isbased on work commissioned from theBeyond Access project by theCommonwealth Secretariat, UNESCOBangkok and UNICEF Regional Officefor South Asia (Unterhalter et al., 2004;Unterhalter, Rajagopalan andChallender, 2005; Unterhalter andMcCowan, 2005). Unfortunately,because of incomplete data in oneof the fields the GEEI measures nocalculations have been possiblefor Afghanistan, Bhutan andMaldives.

The table shows that Sri Lanka madespectacular gains in GEEI in a decade oflow economic growth. Bangladesh andNepal made huge gains in GEEI, butremain under a score of 50. India,despite its large growth in GDP percapita, did not make gains in GEEIequivalent to Nepal or Bangladesh but

TABLE 9TABLE 9TABLE 9TABLE 9TABLE 9 GEEI for South Asia: c.1993–c.2001GEEI for South Asia: c.1993–c.2001GEEI for South Asia: c.1993–c.2001GEEI for South Asia: c.1993–c.2001GEEI for South Asia: c.1993–c.2001

Country GEEI c.1993 GEEI c.2001 Percentage Increase/DecreaseSri Lanka 68 94 38India 28 41 46Bangladesh 23 48 109Pakistan 23 20 -13Nepal 20 36 80Afghanistan, Bhutan, Maldives n.a. n.a.

Source: Unterhalter, Rajagopalan and Challender 2005; Unterhalter and McCowan, 2005

nevertheless saw a large rise from a lowbase. Only in Pakistan is there a fall inGEEI over the decade, leaving it with thelowest level GEEI of the countries in theregion for which there is data.

What Does The GEEI Tell Us?

The trends in GEEI in South Asia aresimilar to those in other regions(Unterhalter, Rajagoplan and Challender,2005). Countries with long anddevastating histories of war or repressivegovernment have low GEEI (Pakistanand Nepal), but Sri Lanka is theexception to this. Countries with highlevels of women's mobilization orpolitical participation score higher thancountries where there has been minimalor only 'top-down' mobilization on theseissues; Bangladesh and India comparewith Nepal in this respect. Countrieswith vast regional inequalities scoreconsiderably lower than countries whereregional inequalities are not an issue onthis scale. India and Pakistan comparestrikingly with Bangladesh and Sri Lankain this regard.

19


What does the GEEI tell us about how farthe world needs to go to meet the twoMillennium Development Goals thatrelate to education: MDG 2 and MDG 3?Table 10 compares the rate ofimprovement in GEEI for countries in theSouth Asia region between c.1993 andc.2001 and estimates what level offurther improvement would be needed toreach a GEEI score of 95. A GEEI scoreof 95 would indicate net girls' primaryattendance of 90 per cent and above,girls' primary survival rate of 90 per centand above, girls' secondary NER of 60per cent and above, and GDI of 0.800and above (equivalent to gender equalitylevels in life expectancy, education andincome of Korea, Singapore and Japanin 2003).

Table 10 shows that despite theconsiderable increases in GEEI forvirtually all countries in South Asia forwhich there is data, the percentage

TABLE 10TABLE 10TABLE 10TABLE 10TABLE 10 GEEI scores South Asia c.1993–c.2001 and improvements needed to reach GEEI 95 by 2015GEEI scores South Asia c.1993–c.2001 and improvements needed to reach GEEI 95 by 2015GEEI scores South Asia c.1993–c.2001 and improvements needed to reach GEEI 95 by 2015GEEI scores South Asia c.1993–c.2001 and improvements needed to reach GEEI 95 by 2015GEEI scores South Asia c.1993–c.2001 and improvements needed to reach GEEI 95 by 2015

Country GEEI GEEI Percentage Improvement Percentage Improvement neededc.1993 (%) c.2001 (%) c.1993–c.2001 to reach GEEI of 95 by 2015

Bangladesh 23 48 109 98India 28 41 46 132Nepal 20 36 80 164Pakistan 23 20 -13 375Sri Lanka 68 94 38 1Maldives 85 n.d.Afghanistan,Bhutan n.d. n.d.

Source: Derived from Table 9

improvement required to reach a GEEIscore of 95 per cent by 2015 is stillenormous. Sri Lanka would be requiredto sustain the gains made between 1993and 2001. It already scores highly in theeducation areas of the GEEI, and it isonly in the GDI that its score is notexcellent. Bangladesh would have tomaintain the levels of improvement ingirls' education and women's lives put inplace between 1993 and 2001. Thiswould entail work in all the threeeducation dimensions of the GEEI,where it currently is at a mid level, andconsiderable improvement of its GDI,which points to quality in and equalitythrough education. Nepal would have todouble its rate of improvement in GEEIand India would need to treble this.Pakistan, which saw a fall in GEEIbetween 1993 and 2001, would need togive maximum momentum to thisimprovement with nearly 400 per centimprovement over the next decade.

20


CONCLUSION

This paper has shown that while theremay be questions about the way data iscollected at local and national level, ifwe work with the data that UN agencieshave put together we have a remarkablyconsistent picture. There have beengains in gender equality in and throughschooling in many countries in SouthAsia, whether we use an indicator ofaccess like NER, of gender parity likeGEI, or of more comprehensive aspectsof human flourishing like GEEI. In somecountries – Bangladesh, Nepal, India –the gains over a decade have beenconsiderable. In others, such as SriLanka, good levels of gender equality atthe beginning of the decade have beenmaintained. While increases in incomeand wealth are associated with thesegains, there is not a simple causalrelationship, pointing to other factorssuch as ambitious government

programmes and the political will to putthem into effect. However, there arecountries where the indicators suggestthe decade has not seen anenhancement of gender equality. Thenegative signs with regard to growth ingender equality indicators for Pakistanpossibly bear out a finding from GEEIwork in Africa, that conflict and wideregional divisions are associated withlower GEEI scores (Unterhalter et al.,2004). For Pakistan this fall is alsoassociated with falling GDP, but this isnot the whole story and much more workis needed both on the quality of the datathe indicators rest on and on how we canread these conclusions in relation toqualitative analysis.

The GEEI was initially developed forCommonwealth countries in Africa. Thepicture in South Asia is, however, much

5

21


more encouraging with regard to thelevels of improvement over the lastdecade. If gains on this scale could bemade before the resources linked to theMDGs became available, there arepossibilities that in many countries inAsia the aspirations for gender equalityand the MDGs may be within reach.However, in South Asia particularly, hugeeffort is needed. For some countries thisentails maintaining the effort that hasresulted in the considerable gains madeover the past decade, but for othercountries double or triple the level ofimprovement is required. Pakistan facesthe greatest challenges, coming fromsuch low GEEI levels.

The GEEI figures tell a compelling story.But they are not nuanced enough toshow us the local picture and the fine-grained image of where success is notthe only story and more resources ofmany different kinds are needed. For thiswe need indicators at local level andlocal ownership of the data to holdgovernments, researchers andinternational agencies accountable forfull realization of the MDG vision.

Improving girls' NER, the GEI and GEEIrepresent formidable tasks for the next

decade. In countries where there hasbeen political commitment andambitious programming significant gainshave been made. To measure up to theMDG commitments this needs tocontinue and expand. Achieving genderequality is generally linked with multi-sectoral development, and often withpopular mobilization in support ofpolitical, economic and social demands,particularly with regard to education,health and sustainability. There is muchconsensus on what needs to be done.This paper has tried to show with whatintensity and urgency we need to act.

The GEEI presents an alternative meansof measuring gender-equality gains andlosses in and through education. Hugechallenges remain to increase GEEI inSouth Asia. Reaching the targets forMDG 2 and 3 by 2015 is not impossible,given the talent and wealth of the world.This assessment of GEEI gives someindication of the size of the task and thelevels of mobilization needed. This taskfalls not only to the people who live inthe countries with low GEEI. The MDGsare challenges to global collaborationand resource mobilization. The numberspoint to the heightened levels at whichwe need to work together.

22


REFERENCES

Alkire, S. (2002). Valuing Freedom. Sen's Capability Approach and Poverty Reduction.Oxford: Oxford University Press.

Beyond Access (2006). Girls' Education in South Asia. Education and Gender EqualityProgramme Insights. Oxford: Oxfam online at http://www.oxfam.org.uk/what_we_do/issues/education/downloads/edpaper9.pdf.

Carr-Hill, R., Hopkins, M., Riddell, A. and Lintott, J. (1999). Monitoring the Performance ofEducational Programmes in Developing Countries. London: Department for InternationalDevelopment.

Charmes, J. and Wieringa, S. (2003). Measuring Women's Empowerment: An assessment ofthe Gender-related Development Index and the Gender Empowerment Measure. Journal ofHuman Development, 4/3: 430.

Elson, D. (1999). Labour Markets as Gendered Institutions: Equality, efficiency andempowerment issues. World Development, 27/3: 611-27.

Fukuda Parr, S. and Kumar, A.K.S. (eds) (2003). Readings in Human Development. NewYork: Oxford University Press.

Maldives National EFA Assessment Group (2000). Maldives EFA Assessment: Country reports.Paris: UNESCO. Online at http://www2.unesco.org/wef/countryreports/maldives/rapport_1.html(consulted March 2005).

23


Millennium Commission (2000). The Millennium Development Goals. Online at www.un.org/millenniumgoals.

Nussbaum, M. (2000). Women and Human Development. Cambridge: Cambridge UniversityPress.

Pogge. T. (2002). Can the Capability Approach be Justified? In Nussbaum, M. and Flanders,C. (eds): Global Inequalities, special issue 30:2 (appeared February 2004) of PhilosophicalTopics, 167-228. Online at http://mora.rente.nhh.no/projects/EqualityExchange/Portals/0/articles/pogge1.pdf.

Raynor, J. (2006). Girls' education in South Asia. Paper commissioned by the Beyond Accessproject, London: Institute of Education, University of London.

Sen, A. (1999). Development As Freedom. Oxford: Oxford University Press.

Terzi, L. (2004). On Education on a Basic Capability. Paper presented at the 4th InternationalConference on the Capability Approach, University of Pavia, Italy, 2-7 September.

Tomasevski, K. (2003). Education Denied: Costs and Remedies. New York: Zed Books.

UNDP (1995-2004). Human Development Reports. New York: UNDP.

UNDP (2001). Human Development Report 2001. New York: UNDP.



UNESCO (1995). World Education Report 1995. Paris: UNESCO.

UNESCO (2000). World Education Report 2000. Paris: UNESCO.

UNESCO (2003). Education for All Global Monitoring Report 2003-4: Gender and Educationfor All. Paris: UNESCO.

UNESCO (2004). Education for All Global Monitoring Report 2005: The Quality Imperative.Paris: UNESCO.

UNESCO (2005). Global Monitoring Report 2006. Literacy for Life. Paris: UNESCO.

UNICEF (2004). The State of the World's Children 2004. New York: UNICEF.

24


Unterhalter, E. (2003). The Capabilities Approach and Gendered Education: An examinationof South African complexities. Theory and Research in Education, Vol. 1, No. 1: 7-22.

Unterhalter, E. (2005a). Gender Equality and Education in South Africa: Measurements,scores and strategies. In Chisholm, L. and September, J. (eds) Gender Equity in SouthAfrican Education, 1994-2004. Conference Proceedings. Pretoria: HSRC.

Unterhalter, E. (2005b). Global Inequality, Capabilities, Social Justice: The MillenniumDevelopment Goal for gender equality in education. International Journal of EducationalDevelopment, Vol. 25: 111-222.

Unterhalter, E. (2005c). Mobilisation, Meanings and Measures: Reflections on girls'education. Development, Vol. 48, No. 1.

Unterhalter, E. and Brighouse, H. (2003). Distribution of What? How will we know if we havereached EFA by 2015? Paper presented at the 3rd International Conference on the CapabilityApproach, University of Pavia, Italy, September.

Unterhalter, E. and McCowan, T. (2005). Girls' Education and the Millennium DevelopmentGoals: What do the indicators show us? Paper presented at Meeting on 2005 and Beyond:Accelerating Girls' Education in South Asia, Bangkok, February 2005. UNICEF: RegionalOffice for South Asia.

Unterhalter, E. et al. (2004). Scaling Up Girls' Education: Towards a scorecard on girls'education in the Commonwealth. Online at www.ungei.org (Beyond Access link).

Unterhalter, E., Rajagopalan, R. and Challender, C. (2005). A Scorecard on Gender Equalityand Girls' Education in Asia, 1990-2000. UNESCO Bangkok. Online athttp://www2.unescobkk.org/elib/publications/gender_equality_asia/index.htm.

Unterhalter, E., Rajagopalan, R. and Challender, C. (forthcoming). A Scorecard for Girls'Schooling in Latin America.

Walker, M. (2004). South African Girls' Narratives on Learning: Insights from the capabilityapproach. Paper presented at the 4th International Conference on the Capability Approach,University of Pavia, Italy, 2-7 September.

Watkins, K. (2000). The Oxfam Education Report. Oxford: Oxfam.

World Bank (2005). Strategy for Education Development in South Asia. Washington: WorldBank. Online at http://www2.unesco.org/wef/countryreports/maldives/rapport_1.html(consulted January 2006).

25


ANNEX

Calculating the GEEI

The GEEI was constructed by using four measures deemed useful as indicators of girls'access to and retention in school, and women's health and levels of income after school. Theindicators selected were:

Girls' net attendance rate at primary school

Girls' survival rate over five years in primary schooling

Girls' secondary Net Enrolment Ratio (NER)

A country's Gender Development Index (GDI).

Data from the EFA Monitoring Reports, UNICEF's State of the World's Children, the HumanDevelopment Reports, World Bank reports, and countries' own EFA assessments were used.Occasionally, where there were no figures available, secondary literature was consulted.

Net primary attendance rates, survival at school, secondary NER, and GDI levels were given avalue based on the following assessments shown in Table A.1.

26


TABLE A.1 Criteria for scoring achievements with regard to access and achievement in girls'

education

Score Criteria to achieve the score5 Excellent conditions. Already at or extremely well positioned to achieve gender

equity in 2015 and fulfil the aspirations of the Beijing Declaration4 Very good conditions. Substantial achievement with regard to gender equity and

well on the path to achieving 2015 goal with regard to access. Some gainsneeded in order to improve retention

3 Good conditions. Progress towards 2015 evident, but further work necessary onaccess and retention

2 Poor conditions. Progress towards 2015 slow. Considerable and intensive workneeded on access and retention

1 Very poor conditions. 2015 goals unlikely to be reached without massivemobilization on all fronts to secure access and achievement

TABLE A.2 GEEI scores and indicators

Score Net Girls' Primary Girls' Primary Girls' Secondary GDIAttendance Survival Rate NER

5 90% and above 90% and above 60% and above 0.800 and above4 80-89% 80-89% 50-59% 0.700-0.7993 70-79% 70-79% 40-49% 0.600-0.6992 60-69% 60-69% 30-39% 0.500-0.5991 59% and below 59% and below 29% and below Below 0.499

Net Girls' Primary Girls' Primary Girls' Secondary GDI GEEIAttendance Survival Rate NER

Raw score x 1.25 Raw score x 2.5 Raw score x 1.75 Raw score x 2.5 Sum of weightedmeasures divided

by 4

Using these criteria, the scoring system illustrated in Table A.2 was developed withregard to the indicators.

Detailed calculations of the GEEI for Africa and Asia are to be found in earlier papers(Unterhalter et al., 2004; Unterhalter, Rajagopalan and Challender, 2005).

27


ABOUT THE AUTHOR

Elaine Unterhalter is a Senior Lecturer in Education and International Developmentat the Institute of Education, University of London. She is joint co-ordinator of theBeyond Access: Gender, Education and Development project which has publishedbooks, policy papers and newsletters linked to achieving the gender and educationMDGs. She has written a number of journal articles on gender and education in Indiaand Bangladesh. Her book Gender, Schooling and Global Social Justice: An analysisof global initiatives to achieve gender equality in education will appear in 2007.

Published by

United Nations Children’s FundRegional Office for South AsiaP. O. Box 5815Lekhnath MargKathmandu, Nepal

Telephone: 977-1-4417082Facsmile: 977-1-4418466 / 4419479www.unicef.org

MEASURING GENDER INEQUALITY IN EDUCATION IN · PDF fileMeasuring Gender Inequality in...

Documents

Transcript of MEASURING GENDER INEQUALITY IN EDUCATION IN · PDF fileMeasuring Gender Inequality in...