Skip to main content
SearchLoginLogin or Signup

Gender Data Gap

Published onDec 28, 2019
Gender Data Gap


  • Gender data is considered important because it provides meaningful insight into structural differences in distribution of opportunities, services, and infrastructure.

  • There exists a “sexist data crisis” in national statistical systems, with 60 percent countries sex-disaggregating data in the fields of education and health - with a much lower percentage for sensitive fields such as crime.

  • Intersectional data, or that which is disaggregates and correlates gender with other marginalities, such as geographical location, income, race, ethnicity etc. is even more scarce.

  • Concepts, methods, and assumptions of statistical systems often do not capture the reality of women’s lives, producing data that is exclusionary and of poor quality.

  • The exclusion of women from civil registration systems has a severe impact on the protection of their fundamental rights, especially among at risk-groups such as migrants, poor, minority groups, widows, and single mothers.


Opening up gender data is crucial to monitoring progress on gender-based indicators across Asia. There is a significant gap when it comes to openness and availability of such data, sometimes resulting from an overall resistance to data - especially gendered data.1 Closed data cultures can result from absence of informational disclosure laws and complex requirements in policy to access information.2 Such challenges can hamper efforts to transparently track performance on gender indicators.3 Above and beyond overall failures in transparency, there exists a gender data gap in data systems. In addition to simply opening up data, it is critical to ensure that open data practices employ a gender lens while collecting as well as using data.4 This article discusses the state of gender data across two largest public data systems, official statistics and identification systems.

As per the United Nations Statistical Division, gender data is considered important because it provides meaningful insight into structural differences in distribution of opportunities, services, and infrastructure across economic, health, political, and social dimensions.5 According to Data2x, gender data is data that is collected using sex as a primary classification; reflects gender issues; is based on concepts that adequately capture gender differences; or involves collection methods that account for sociocultural biases and other factors related to gender.6 Gender statistics refer to “Data concepts and definitions that adequately reflect the diversity of women and men and capture all aspects of their lives; collection methods that take into account stereotypes and social and cultural factors that may induce gender bias in the data”.7

The Open Data Barometer, “a global measure of how governments are publishing and using open data for accountability, innovation and social impact”8, identifies a “sexist data crisis” in open data globally.9 In 2017, of the countries that had provided national statistics, sixty percent had sex-disaggregated data for the fields of education and health, with a radically smaller percentage for more “sensitive” areas such as crime statistics.10 While data on women and girls is lacking globally, the dearth is more severe in Asia and other parts of the Global South.11

According to Mayra Buvinic,12 “the dearth of data makes it difficult to set policies and gauge progress, preventing governments and organisations from taking measurable steps to empower women and improve lives. Not having data on a certain area, behavior or society means that you cannot track progress, you cannot evaluate. You are basically not accountable.”13

Sex-disaggregated Data

The Statistics Division of the United Nations Economic and Social Commission for Asia and the Pacific defines sex-disaggregated data as “data that reflects the combined effects of gender roles, gender-based norms and stereotypes, as well as women’s reproductive roles”.14 According to Data2x, sex-disaggregated can be used to achieve mutiple goals - quantify women’s marginalisation and impact on their needs and capabilities, structural challenges in accessing opportunities, as well as resource allocation.15

Quantitative data can also be used to track progress across time and geographical locations, and impact, if any, of development interventions.16 For instance, data on land ownership can measure the efficacy of any redistribution efforts17, that on political participation can measure the progress of affirmative action to bring more women into politics18, and so on. With the efforts of the United Nations Statistical Commission, a minimum set of gender indicators has been established.19 Despite this, there is very little data in a variety of critical areas, such as participation in governance and the public sector, or data on the nature of positions held by women in the private sector.20

Equal Measures 2030 has published a list of issues that have a scarcity of sex-disaggregated data - such as social protection, land tenure, food security, mental health, among others; and issues that impact the lives of women and are data scarce - such as intra-household income, age at first pregnancy, women’s household decision-making power, etc.21 A report on the gender data gap by Data2x discusses data gaps by domain, in the fields of health, education, economic opportunities, political participation and human security.22 They point out critical gaps in data on maternal health, disability, violence against women, unemployment, participation in civic and political life, conflict-related violence, etc.23

Intersectional Gaps in Data

Intersectional data, or data that disaggregates by gender and correlates it with other variables, can identify critical differences between women across different levels of income, education, literacy, disability, geographical location, ethnicity, race and caste.24 Data that is disaggregated by gender alone does not capture differences between women, reducing usability.25 In Asia and the Pacific, the importance of such data in policy planning is not yet widely understood.26

Women from marginalised groups are doubly or triply disadvantaged, requiring policy that is specifically targeted at them. It is then critical to collect intersectional data to design such policy that can grapple with disadvantage on multiple fronts - social, economic, cultural, and political.

Information on rural women is an area with critical gaps - including data on agriculture and rural development and women’s contribution to these.27 Data2x further points out a critical gap in data on girls that are “socially excluded” from education systems, as a result of race, religion, ethnicity, geographical location, and disability.28 Women from specific groups, such as refugees, migrants, widows, single mothers, older women, etc. are more vulnerable to human rights violations, and require targeted data to be produced. Data on displacement, land rights, climate change and environmental data are examples of specific fields that need intersectional disaggregation.29

These can be made cross-comparable across regions by mapping them onto existing indicators such as the SDGs. For example, UN Women has found that 40.6% of rural Sindhi women (aged 18 to 49) living in rural areas of Pakistan are undernourished compared to 2.4% of the richest urban Punjabi women.30 To be useful, data must be disaggregated along variables that are appropriate to the context it is being used in.31

Exclusion from Official Statistics

Gender data is often absent or of poor quality across national statistics globally, impeding the measurement of policy outcomes and development goals - ranging from access to resources, land ownership, education and literacy, credit and financial inclusion, ICTs, health, and labour, as well as data on crime and gender-based violence.32

Women’s labour is a domain with numerous examples of underreporting, and poor quality or incomplete data. For instance, a report by Data2x illustrates the underreporting of women’s involvement in subsistence and other casual occupations.33 Exclusion happens at the stage of definition itself - the definition of employment as provided by the 19th International Conference of Labour Statisticians (ICLS) is “work performed for pay or profit”.34 For the purposes of compiling national accounts, the scope of ‘production’ is restricted and therefore excludes unpaid household labour.35 The new ICLS standards therefore have significant bearing on women in low-income contexts, who are heavily involved in unpaid but economically productive activities such as subsistence agriculture, self-employment and entrepreneurship, and contributing to family work in small enterprises.36 Accurately measuring the type and scale of women’s labour requires data collection that is able to capture undervalued and invisible labour in the informal economy.37

Additionally, consistent guidelines for sex-disaggregated data are not available. Often data is not comparable across countries due to inconsistent definitions and standards of data collection.38

Methods of data collection and analysis that are used most often are often designed in ways that fail to take into account the reality of women’s lives. For example, apart from gender-specific data such as that on women’s reproductive health, surveys collect data from proxy-respondents which are usually male heads of households, administered by male interviewers.39 Further, official surveys very often take households as the unit of analysis, which masks differences within households.40 A survey collecting household-level information of access to mobile phones or the internet will not indicate the difference between access by men, women, and children in the household.41

Innovation with respect to new methods and approaches at data collection are fundamental to pave the way to better gender data.42 Individual-level data, which is dissagregated by sex, is essential to prevent flawed or incomplete datasets that can misrepresent and overreport progress on the indicator it is studying.

The integration of a gender perspective throughout the statistical system is a methodological requirement for improving both data coverage and data quality. This will require the integration of gender and participation of women at different stages of the process.43 A gender perspective is necessary in determining the breadth of issues covered by statistical analysis, as well as the process of arriving at those issues.44 It is also useful in shaping the structure of the analytical system, including “concepts, definitions and classification systems”.45 A gendered perspective can then influence each stage of the data collection process: in determining which issues to cover, methodology for data collection and analysis, and outreach.46 This will ensure that marginalised gender groups are not merely consumers of data, but are actively represented at every stage of the project design.

Data on the Gender Wage Gap

Globally, there exists a gender wage gap across different sections of the economy.47 As per the Eurostat, “Unadjusted gender pay gap is defined as the difference between the average gross hourly earnings of men and women expressed as a percentage of the average gross hourly earnings of men”.48 This gap exists between the income of men and women due to factors such as occupational segregation49, direct pay discrimination due to biases against women workers, higher burden of unpaid care work on women, and penalisation for maternity breaks.50 In Central Asia and East Asia, the wage gap is higher than anywhere in the world.51

Research conducted in Denmark has demonstrated the positive effect of pay transparency on closing the gender gap.52 Pay transparency will allow organisations to be held publicly accountable, and enable the achievement of the goal of fair pay, as it forces organisations to ensure equitable compensation across gender and other intersections.53

Low levels of transparency in pay across public and private sectors, as well as “patchiness” of the available information has created barriers in making correlations between different factors and identifying cause and effect.54 Calderon and O’Donnell critique the popular approach taken by governments to open data within sectors rather than companies and organisation, which then makes it harder to direct advocacy efforts.55

Disaggregating data along other social characteristics reveals greater inequality. Black, Latino, and Asian women in the United States, for example, face greater barriers to equal pay than Caucasian women.56 Biases can also be exacerbated by ethnic occupational segregation - in India, for example, the gender wage gap in the informal sector is larger, which includes a large proportion of Dalit and indigenous women in the labour force.57 Data disaggregated by other factors, such as education and literacy levels, age, and disability, is close to none.58 In Pakistan, the estimated wage loss for rural women is more than double the loss of wage loss for urban women.59 Transparency reports on pay, as well as official labour data, very rarely disaggregate data along multiple axes which creates barriers for understanding intersectional inequality in pay.

Apart from official statistics, it has been difficult to collect data on the pay gap in the informal economy, and subsequently to strengthen advocacy efforts. This is concerning because a large proportion of women in the Global South work in the informal economy - in South Asia, that number is ninety-five percent.60 Similarly, data on unpaid labour, which is overwhelmingly performed by women, is also scarce.61 This particularly impacts rural women in Asia and the Pacific, where the distribution of unpaid is among the lowest in the world.62

When it comes to policy making around gendered wage data, policy frameworks should ensure that data about pay is collected and publicly available for public and private enterprises.63 For instance, companies in Great Britain with more than 250 employees are mandatorily required to report their gender pay gap figures at the end of every financial year.64 Governments globally should thus introduce and implement policy to support open intersectionally disaggregated wage data across different sectors.

Data Gap in Civic Registration

Civil Registration and Vital Statistics (CRVS) is “a system for recording vital statistics such as birth, death, marriage, etc.” It provides a mechanism to assess national progress along global indicators such as the SDGs, and provide data for effective policy making.65 CVRS systems are thus aimed at collecting “continuous and up-to-date” demographic information, including birth and death rates, thereby providing policy makers with reliable information on which to design policy frameworks.66

Crucially, they also serve as a “gateway for exercising individual rights and protections”.67 A number of international conventions related to fundamental human rights call for regulation of the registration of civil status - the International Covenant on Civil and Political Rights (1966) in Article 24 and the Convention on the Right of a Child (1989) in Article 7 “provides the right of any child, without discrimination as to, among others, race, color, sex, national or social origin, to be registered immediately after birth and to have a name.”68

The rights that are guaranteed in these and other similar conventions necessarily require the registration of civil status or presupposes the registration of civil status - they “cannot be safeguarded if a person is denied the right to register civil status”.69 Civil registration provides easier access to services like health and education, and sex-disaggregated vital statistics provide essential demographic data on important issues thus aiding in the formulation of gender sensitive policies.70 It is crucial for women to have access to civil registration, to address issues such as child marriage and protection of fair rights during divorce or widowhood.71

As per the SDGs Report 2019, the average birth registration rate globally is just 73 percent, and less than half the children under 5 have had their births registered.72 Overall coverage in several countries in the Asia Pacific is quite low - 20 percent in Bangladesh and 34 percent in Pakistan.73 Failure to register birth can impede access to subsequent documentation and access to essential services and rights.74

It is critical for countries to achieve 100 percent registration of birth, which then provides access to subsequent adult documentation. However, it is even more crucial for women, especially those who are migrants, poor, belong to a minority group, widows, and single mothers.75 This is because these groups are at greater risk of human rights violations such as trafficking and early marriage, and identification documents provide the basis for protections against such violations.

The operation of CRVS is not merely a function of statistical or technological systems, but also the contexts in which it is embedded - national, political, economic, and social.76 Adapting CRVS systems to these contexts is essential to actively include marginalised groups in the registration process.77 Inclusion can be achieved by identifying and adapting to local norms that disincentivise registration, as well as contextual needs and challenges. Knowles and Koolwal (2016) identifies social customs around marriage and common law, complex registration requirements, payment of fees, and poor levels of information regarding registration as significant barriers to registration.78 There exists a rural-urban divide in registration across Asia, due to poorer access to registration sites and constraints on infrastructure in rural areas.79 The impact of physical constraints is higher for women as compared to men, as women face greater challenges to mobility and may have higher constraints on time.80

Good CRVS systems are “open, accessible and usable”.81 There are several notable initiatives for better implementation of the CRVS systems. The Centre for Excellence for CRVS Systems, under the International Development Research Centre, has launched a knowledge brief series that aims to document produce knowledge on the relationship between gender and CRVS systems.82 The World Bank and the World Health Organisation have developed a Global Civil Registration and Vital Statistics (CRVS) Scaling Up Investment Plan, which aims to map activities across a decade-long period, from 2015 to 2024, “with the goal of universal civil registration of births, deaths, marriages, and other vital events, including reporting cause of death, and access to legal proof of registration for all individuals by 2030.”83 Data2x has also produced vital work assessing the gendered impact of CRVS systems.84


This article aims to map the gender gap in data-driven systems across Asia and globally. We note two crucial systems, official statistical systems and civil registration and identity systems that suffer from a gendered gap. These lead to severe consequences for women, particularly those belonging to marginalised groups. There is risk of exclusion from human and civil rights and protections, from policy making plans and agendas, and public and private services including healthcare. Some data-driven systems report low levels of registration across all demographic groups, which could have a greater impact on rights for vulnerable groups. Others report structural gaps in sex-disaggregated and gender-relevant data. Addressing these failures is crucial to close the gap in gender open data.


No comments here

Why not start the discussion?