New alternatives for the inclusion of gender components in the multidimensional measurement of

XVIII International Meeting on Gender Statistics, Aguascalientes, Mexico Sabina Alkire, 6 September 2017 Introduction and Motivation: MPI

• The first prominent MPI was the global MPI co-designed by OPHI with UNDP’s Human Development Report Office. • Many countries have official permanent national MPIs and a network of 54 participating countries share lessons. • The dimensions, indicators, and cutoffs of national MPIs vary. Common dimensions are health, , living standards and work, also violence, and environment. • National MPIs are primarily used as governance tools, to shape allocation, policy design, coordination, targeting, monitoring and evaluation. • In recent Voluntary National Reviews, 16 countries so far name MPI as a tool to track poverty. Gendering Multidimensional Poverty A key methodological choice for gendered poverty is the unit of identification:

A household unit of identification identifies each person as poor or non-poor based on a clearly defined assessment. If indicators are individual (work, education, health) or age specific (school attendance) all households’ poverty status reflects these.

An individual unit of identification identifies each person as poor based on their own individual attainments, usually also applying household attainments to each household member. Gendering Multidimensional Poverty: Options Household as Unit of Identification: How to bring gendered components into household measures. 1. Include women specific deprivations as indicators: - lack of access to skilled birth assistance - female malnutrition - FGM - experiences of domestic violence

2. Create and analysed ‘gendered’ deprivation indicators - a household is deprived in adult education unless at least one woman AND one man have completed 6 years of schooling. Gendering Multidimensional Poverty: Options Household as Unit of Identification: How to bring gendered components into household measures. 3. Include women’s and men’s deprivations but deepen the analysis with a gendered component: - 60% of people live in households where no adult has completed 5 years of school; for 50% of cases no woman is educated; in 5% only the man lacks schooling; in 5% both a man and women lack schooling. 4. Include deprivations that often affect women more: - lack of childcare facilities - carrying water - indoor air pollution from solid cooking fuel Global MPI – child specific and related indicators Gendered Poverty Measures: Options Person as Unit of Identification: Individual MPIs reflect deprivations that directly strike each person. They can be disaggregated by age and gender, to inform age-specific and gendered interventions. - May identify if all persons in a given household are poor - May identify the persons most-at-risk in terms of region, gender, age, indicator(s) and so on. - May include household level indicators (water, sanitation) - Most are restricted to one age bracket (children, adult)

Alkire Apablaza and Jung (2014), Vijaya et al (2014), Klasen and Lehoti (2016), Pogge and Wisor (2016) Alkire & Apablaza (2017) Gendered Poverty Measures: Options Person as Unit of Identification: Problem: gendered poverty measures either: 1. Leave out deprivations that do not apply to all - may overlook key issues: care burden, reproduction, VAW - visibly incomplete

2. Include gender-specific deprivations. - In some indicators one gender will be non-deprived - or else indicators are to be defined such that they refer to comparable gendered deprivations (what is comparable?) - Requires normative standards/guidelines (particularly for reproduction and employment/care; intrahousehold sharing) -Requires special datasets (assets, income, autonomy) Example of an Individual Measure:

Women’s Empowerment in Agriculture Index achievements A woman’s or man’s empowerment score shows her or his own Five domains of empowerment What is the WEAI? • An aggregate index in two parts: – Five domains of empowerment (5DE – an MPI -): assesses whether women and men are empowered in the 5 domains of empowerment in agriculture – Index (GPI): reflects the percentage of women who are as empowered as the primary man in their households • It is a survey-based index, not based on aggregate statistics or secondary data, constructed using interviews of the primary male and primary female adults in the same household How is the Index constructed? WEAI is made up of two sub indices 5DE = (1-M ) 0 GPI = (1-P1) Women’s Five domains of Empowerment Gender parity empowerment in Agriculture Index (GPI) (5DE) Index Women’s A direct measure of (WEAI) achievement’s relative women’s empowerment to the primary male in 5 dimensions in hh All range from zero to one; higher values = greater empowerment Who is empowered?

A woman who has achieved ‘adequacy’ in 80% or more of the weighted indicators is empowered Bangladesh Pilot results • 31.9% of women are empowered

• Disempowered women have adequate achievements in 60.7% of domains

• 59.8% of women enjoy gender parity

• Households without gender parity have a Overall, the WEAI 25.2% empowerment gap between the score is 0.749 woman and man Bangladesh: How to increase empowerment?

Contribution of each indicator to disempower- ment of women and men Multidimensional Poverty across Europe by Gender and Year

Alkire & Apablaza 2017 in Marlier et al.

16 Gendered analysis of the global MPI: work in by Robles.

Individual Censored Headcounts Proportion of people who are MPI poor and deprived in each indicator

90% 81.5% 80% 76.2% 76.7% 70.5% 72.0% 70.8%70.6%71.0% 70% 58.7% 59.7% 60% 50% 40% 29.8%31.3%28.5% 30% 20% 10% 0% Years School Child Mortality Nutrition Electricity Sanitation Water Floor Cooking Fuel Assets schooling attendance Education Health Living standards Individual MPI Females MPI Males MPI

Note: in some indicators men’s health achievements may be lower.

17 Some Difficult Issues

• Individual data may be ‘stock’ (educational attainment) • Comparable deprivations in work and care • Comparable deprivations in reproductive health • Comparable deprivations across age cohorts • Intra-household distributions: data gaps • Autonomy and Voice: improving data sources “Measuring Women’s Autonomy in Chad using the Relative Autonomy Index” Vaz, Pratley, & Alkire Feminist 2016 • Validated the Relative Autonomy Index statistically • Demonstrated domains of autonomy are distinct • Women are less autonomous than men in each domain • Correlations with subjective well-being, other agency and decision-making questions, and satisfaction are low. • Regression analysis shows that autonomy information is not replicated by standard proxies for women’s empowerment such as level of education or income poverty status. • Data on autonomy should be gathered as they may unlock new relationships between agency & outcomes. The biggest constraint to gendered poverty measures at present: DATA

Ways forward: - Sample designs and sampling teams - Gendered surveys covering key dimensions

See OPHI’s gendered survey for MPI proposal: http://www.ophi.org.uk/wp-content/uploads/MPPN_SDG- Pov_QuexPost2015_Sept-14a.pdf