21 February 2024

Harnessing the Power of Intersectionality Data for Better Development Policies and Impact:

A Concept Note

Prepared by Shaida Badiee and Mayra Buvinic^[1]

Contents

1. Introduction
2. History
3. Four Dimensions of Intersectionality in Development Data
4. Data production & use
5. Data agency
6. Policy analysis
7. Impact & Monitoring
8. Requirements
9. Priorities & way forward

I. INTRODUCTION

Intersectionality is a concept for uncovering and understanding the experiences and challenges faced by individuals who occupy multiple and intersecting group identities. Core components of the concept with data implications include a) ascribed attributes of a group defining identity; b) combined disadvantage from 2 or more group attributes; c) the importance of having data that captures ‘lived experience;’ d) participatory approaches to data collection; e) data ownership issues; and f) multiple disaggregation by gender and other attributes such as age, race, ethnicity, disability and sexual orientation and gender identity (SOGI) status.

Intersectionality has been integrated for many years within development practice.It is also central to global feminist advocacy and identity politics. Intersectionality is receiving increasing development attention as countries and agencies are discovering its importance in the context of achieving development goals, including SDG implementation, particularly for poverty reduction and gender equality, and the ‘leave no one behind’ ambitious agenda.Many development agencies such as UN Women, OECD, and the World Bank are placing intersectionality as a core dimension to consider when addressing gender inequalities.In addition, the concept has become central to global feminist advocacy and identity politics and has gone viral in the international press, resulting in multiple interpretations and political backlash.

The main emphasis of this concept note is to translate aspirational intersectionality concepts into practical guidance for developing countries, particularly with regards to data. This note is motivated by a need to bring together and learn from different strands of thought — development, feminist, and identity theories— to clarify the concept and its political implications and develop a commonly agreed upon approach that applies the core principles of intersectionality to inform development data and policies.

This concept note seeks to motivate discussion and feedback and is a work in progress and will be maintained as a live document.This note starts with a summary of the evolution of the concept of intersectionality in the fields of development, gender issues and development data to help ground the concept of intersectionality.It then outlines a framework with four main dimensions to characterize a successful intersectionality approach to data in development.Third, the note describes each dimension separately and analyzes challenges, risks and opportunities. The last forward-looking section outlines the basic requirements for implementing an intersectionality approach to data production and use in LMICs and lists priorities for the work ahead.

II. HISTORY

The development origins of intersectionality

‘Intersectionality’ was first used to characterize discrimination in the US but the underlying concept is well-known in the field of development. While K. Crenshaw first coined the term in 1989 to describe the double oppression (by gender and race) of women of color in the U.S., the concern with double discrimination has an established tradition in development theory and practice.^[2] In development theory, a primary focus on income inequality branched out in the 1990s to encompass a focus on ‘horizontal inequalities’ affecting socially excluded groups.^[3] Around the same time or earlier, human rights approaches emerged in development work as a counterpoint to structural adjustment programs; development agencies launched community driven development (CDD) programs targeting the poor and excluded; and social inclusion and social cohesion approaches informed development policies and projects.^[4] Fundamental components of these policies and programs include non-discriminatory legal frameworks, participatory development, and identification and documentation of excluded populations as prerequisite for recognizing rights and redressing inequalities of populations with combined (i.e., intersecting) disadvantages.

Decades earlier, leading countries in the global South had identified excluded groups suffering multiple forms of disadvantage and implemented affirmative action policies. Notable examples are reservation quotas in India, documenting race in the census in Brazil, and affirmative action quotas in higher education, also in Brazil.

Intersectionality in gender and development

The gender and development field, since its inception in the 1970s, has recognized the multiple disadvantages arising from the combination of gender and other ascribed group attributes. For instance, the marginalization of indigenous women took center stage at the first UN World Conference on Women in Mexico City in 1975.^[5] Women with disabilities have unfailingly raised their voices in the UN Commission on the Status of Women meetings and in UN Women’s world meetings since the 1970s. A recurrent theme in the field is the need for more data disaggregated by sex and other group identities such as age, class, ethnicity, race, disability and, more recently, gender identity, to be able to accurately document the condition of women and design inclusive policies and programs.^[6]

Intersectionality in development data

The Basic Needs Approach conceptualized in the 1970s changed the prevailing focus on income in traditional measurement frameworks used by development agencies to a broader focus on food, shelter, education, and health indicators. The work of Mahbub ul Haq and Amartya Sen highlighting the importance of considering multiple dimensions of human well-being led this shift towards a people-centered approach to data and measurement. In the 1990s, this approach was used to establish the Human Development Index (HDI) which provided a high-profile tool to help measure and analyze life expectancy, education, and other people focused indicators together in the form of a composite index.

The Millennium Development Goals (MDGs) in 2000 and its successor, the Sustainable Development Goals (SDGs), adopted in 2015, have also helped advance a people-focused approach to data. While the MDGs was still focused on data aggregates, the SDGs, with its sophisticated data framework, is the most explicit recognition of the importance of intersectionality in data at the global level. The SDGs recognizes the importance of the intersection of identities – for instance, for achieving the goals of gender equality (#5), reduced inequalities (#10) and peace, justice, and strong institutions (#16) — and calls for more inclusive and equitable data that Leaves No One Behind (LNOB) and puts people at the center of measurement frameworks.

III. FOUR DIMENSIONS OF THE INTERSECTIONALITY IN DEVELOPMENT DATA FRAMEWORK

What would a successful intersectionality approach to data look like to support development goals? We see four main dimensions, shown in Figure 1 below, that must be developed together by any organization using this approach:

1. Data production and use: A sound data system should be able to collect disaggregated data by attributes important for national development which also are interoperable, open and accessible, and at the same time protect privacy. This describes rather advanced data systems, so it is essential to look for guidance and solutions that enable LMIC countries to build data systems that serve the intersectionality data needs.
2. Data agency:A core tenet of intersectionality is that those left behind should have agency in data collection and that data should capture their lived experiences, that is, the knowledge they have gained through first-hand involvement in everyday events and empower them to bring about social change. Data agency refers to the ability of individuals or groups to have control and ownership over their data, particularly when it comes to data that reflects their intersecting identities. Governance of data should address this link to those left behind as well as the link between data, the populations left behind, and policy.
3. Linkages to policy analysis: Capacity for data usage, including analysis of challenges and needs of populations is a necessary component for more equal outcomes. It is essential that there are strong linkages between data and policymaking to support effective evidence-based policy. Data people need to be in contact with policy analysts to inform collection and policy analysts should have access to data people to inform usage. Therefore, organizational arrangements, skills mix for this work, governance of data and linkages throughout the data value chain are critical.
4. Impact and monitoring: An efficient process must be designed to monitor impact and results achieved along the way and allow for course corrections in any of the data or policy analysis stages. Because of the complexities and risks involved in so many aspects of intersectionality, monitoring results and a dynamic data and policy analysis are of outmost importance.

Figure 1:Intersectionality: Data to Agency to Policy to Impact Framework

Source:Open Data Watch and Data2X

A successful intersectionality approach to development will contribute to ‘leaving no one behind’ and achieving the SDGs (Figure 1). The inner circle in Figure 1 spells out the ascribed attributes that identify groups and are most often associated with disadvantage, marginality and social exclusion. These include standard demographic indicators such as sex and age and attributes that are much less frequently measured such as race, ethnicity, disability, and SOGI (sexual orientation and gender identity). The middle circle contains selected outcomes from exercising agency to use data to affect policy; these outcomes are shown by the red arrow drawn from ‘policy’ and include economic empowerment, education, health, labor market and welfare status outcomes. This middle circle also contains selected factors that mediate the ability to exercise agency over data to influence policy; they are shown by the red arrow drawn from ‘agency’ and include digital transformation, time allocation and time poverty, gender-based violence, biases and discrimination, legal and social norms, and financial inclusion. The outside circle notes the two overarching impacts of a successful intersectionality approach mentioned above.

We now turn to briefly describe each framework dimension (data, agency, policy, and impact) separately, analyze risks and opportunities, and end with a discussion of requirements and issues to consider on the way forward.

IV. DATA PRODUCTION AND USE

Data is placed at the heart of our proposed intersectionality conceptual framework. It is an essential resource to capture the complex and nuanced experiences of individuals and understand the impact of intersecting identities. The goal should be for better data systems along the Data Value Chain (Figure 2), not only in data production but also better dissemination, analysis, and use of intersectional data to design effective policies for reducing inequalities and improving lives.

But how far are we from an agreed approach to data that would meet the needs of an intersectionality in development approach with a gender lens (one of the major attributes that needs intersection)?

Intersectionality and gender data limitations

Traditional data systems cannot fully capture what is needed for an intersectionality in development approach. At production, the traditional data systems such as census and household surveys generally collect information on broad demographic categories such as age, gender, race/ethnicity, and socioeconomic status. While they provide valuable information, they often have several limitations:

• Lack disaggregation and granularity: Traditional data systems lack the details needed to capture the specific experiences of individuals who belong to multiple groups. These traditional systems often collect data on broad categories. However, they rarely collect or report the intersections between these categories and therefore do not fully capture the complexity of people’s identities. What is needed is to improve both data disaggregation (data covering smaller subgroups like age and gender) and data granularity (data providing higher level of details such as small area estimations). Individuals who belong to multiple marginalized groups may face unique challenges that are not adequately reflected in these simplified categories. For example, identities such as a transgender person of color or an immigrant with a disability.
• Underrepresentation of marginalized groups: Traditional data systems may underrepresent marginalized groups due to historical biases, sampling methods, or exclusionary criteria. This underrepresentation can lead to a skewed understanding of the experiences and needs of these groups and bias aggregate results.
• Lack qualitative data: Traditional data systems primarily rely on quantitative measures and may not capture the nuanced experiences and narratives of individuals who belong to multiple marginalized groups. Qualitative data, such as personal stories and lived experiences, can provide valuable insights into the intersecting forms of discrimination and disadvantage faced by individuals for effective policy and program design.
• Limited data on social determinants: Traditional data systems often focus on individual-level characteristics and may not capture quality data on social determinants (of health or educational attainment or labor force participation, for example) that disproportionately affect marginalized populations. Factors such as discrimination, structural inequalities, and access to resources and opportunities are crucial in understanding the experiences of individuals facing intersectional forms of disadvantage.
• Privacy and confidentiality of data: An intersectional in development approach requires granular data with disaggregation by many characteristics to adequately represent groups across various intersectional identities. This calls for data systems to include best practices in applying privacy and confidentiality measures. Compliance with relevant data protection laws and regulations is crucial as well as use of data anonymization techniques to minimize the risk of re-identification.

Building a better data system along the data value chain

To address these limitations, we should aim for improvements in all segments of the data value chain—collection, publication, uptake, and impact. It is important to take a system-wide, comprehensive, and inclusive approach to build better data systems with people at the center. Piecemeal interventions may help with an isolated study or case. However, the goal should be to mainstream intersectionality data requirements within a strong and sustainable national data system. In such system, the ambition should go beyond more inclusive survey questions, diversifying sampling methods, and promoting data disaggregation. Following a system-wide approach would involve not only planning data production improvements, but also building capacity for making collected data open and easily accessible, for analyzing data, and promoting their use and impact.

Use of existing data sources should be maximized. Data based on administrative systems and data from alternative sources such as citizen generated data (CGD) are particularly important for building a better data ecosystem for intersectionality and gender. Lack of data sharing agreements and poor open data practices stand in the way of using such data sources from across government agencies and administrative systems. Based on evidence from the Gender Data Compass, a research project monitoring state of gender data availability and openness in 185 countries, existing data does not always mean that those data are open and accessible to users. In fact, this study finds gender data are among the least open and accessible data sets. Assessment of 185 countries shows the average openness score for all data categories is 53% as compared to only 34.7% average openness score for gender data. Fortunately, there are many best practices since open data issues are not new and in recent years there have been many advancements in the data sharing and open data fields. With better and more systematic use of these practices, access to and use of existing data could improve today’s situation where much of the existing data remains under used.

An intersectionality in development approach to data has transformative potential for improving traditional data systems. This approach calls both for better representation of different population groups and for improvements to data governance along the data value chain. Inclusion of representative groups, in form of voice and participation, is key and this should be done along the data value chain to ensure new data systems are inclusive and safe. This, in fact, could be the most transformative potential of this approach. Such data practices allow us to go beyond data quality improvements, such as better disaggregation to reach for higher goals of improving the power dynamics within data governance practices. This is further explained in the following section on data agency.

V. DATA AGENCY

Agency is the ability to act and affect changes in spheres that are important to the individual or the collective.^[7] Agency is a process that is difficult to measure directly; data usually captures this process indirectly (usually through self-reporting of individual or groups) or measures the outcomes of exercising agency, such as economic achievements, social or psychological gains. Data agency in intersectionality includes the basic tenet that counting the excluded and making them visible in statistics acknowledges their status as citizens with rights and responsibilities. But it goes beyond this basic principle and calls for excluded populations to have a say in, control or ownership over the production and use of data.

Intersectionality questions the traditional processes of data collection where data producers own and control the data collected. It argues that this unequal balance of power between data producers and the subjects of data biases the data collected. Unbiased data captures the ‘lived experiences’ of individuals or groups that have been marginalized for identity reasons through participatory data methodologies. Intersectionality devolves ownership of data to the subjects of data who become empowered to use this data to change policies and redress group inequalities.

There has been some work done to apply these basic principles of intersectionality around data agency in LMICs, which has raised many questions as well as opportunities and risks. Below we briefly discuss, first, the advances in citizen generated data (CGD) and what it offers to translate the theory of data agency into practice. And second, we briefly discuss data ownership issues.

Agency, ‘lived experience’ and citizens generated data (CGD)

Capturing ‘lived experience’ is a special case of citizens’ generated data (CGD), a movement that has emerged in the last decade as the growth of privately generated data has challenged the social contract between people and government regarding their data and its use.^[8] The term “lived experiences” suggests moving away from data systems based on fixed statistical categories collected in a detached way and instead capturing the unique and diverse experiences of individuals using techniques that involve communities and includes those marginalized. The term CGD has been used to define different approaches to data production and initiatives by communities, civil society organizations and citizens. The World Bank, for instance, defines CGD as data that people or their organizations produce to directly monitor, demand, or drive change on issues that affect them.^[9]

A review^[10] of how citizens data have been used in recent years reveals that citizen data initiatives have helped respond to a wide range of demands:

1. Responding to data needs of groups or communities that do not feel represented by the data available in their national or local data ecosystem.
2. Ensuring equitable and effective service delivery in various sectors, such as healthcare, education, social services, etc. This process involves assessing the availability, affordability, and quality of services, and using the gathered data to inform legal and policy initiatives.
3. Assessing and monitoring the extent to which legal rights and international human rights standards are acknowledged and effectively put into practice
   
4. Increasing public understanding and awareness, mobilizing actions and empowering local communities.

CGD can be fully integrated into official statistics, complement them as additional data to be used outside the official statistical boundaries, or challenge official statistics and offer an alternative description of experiences. The latter option is especially welcome by groups that feel misrepresented by the existing approaches to data collection.

Recently, the UN Statistical Commission has endorsed the establishment of a collaborative to define a conceptual framework for citizen data.^[11] The purpose is to encourage responsible production and curation of citizen data and enable CSOs, communities and citizens in countries to use the data they produce more broadly and effectively. The upcoming framework is also expected to leverage the power of citizens’ data for the SDGs, to fulfill the ambition to leave no one behind, and to inform decision-making overall.

Central in this framework is citizens’ agency in defining how data are representing their experiences and addressing the issues that are important to them. On that premise, citizen data are the result of CSOs or communities engaging in the production of data at their own initiative, with control over the process, and the intention to use the data as an instrument to drive change. In addition, the framework offers a set of principles, inspired by existing data principles and a human-rights approach to data (following the principles of human rights, dignity, equality, and non-discrimination in all aspects of data systems), established to ensure that data production and use are responsible, professional, and ethical. These principles serve as a framework for data governance, management, and protection, in the context of organizations that collect citizen data.

Overall, CGD increases data inclusiveness and improves the availability of data needed to address intersectionality but should cover the undocumented and operate under robust governance. To fully fulfill the intersectionality mandate, CGD needs to expand the concept of ‘citizens’ to include people without citizen documents who may be the most marginalized and left behind. In addition, it requires establishing robust data governance. Fundamentals include legal rights to independent data production, funding, and the right incentives to encourage collaboration between government and civil society. Importantly, it requires data literacy, standards for data quality and appropriate level of data quality controls to generate trust in data.

Data ownership issues and opportunities

Data ownership is an important contributor to improving data agency and a key concept that plays a crucial role when building a system for intersectional data. Data ownership refers to the rights and controls a person, institution, or organization has over data they create and collect. Without clarity of data ownership terms and conditions, the value of intersectionality data could be diminished, as we would not know who has the authority to access or use and modify the data.

Data ownership opportunities and risks are greater in the context of intersectionality and as demand is increasing for more and better disaggregated data to capture intersecting identities. The risks are also higher as there is more demand for using new sources of data such as CGD or passively generated digital data, among other sources. These call for better data ownership practices with sound legal, technical, and ethical considerations.

On the legal side there needs to be data governance processes and guidelines that clarify data ownership rights and any regulatory framework associated with specific types of data, such as disability data or SOGI, or other characteristics.

At the technical level, there should be tools to keep data secure as well as safeguard privacy of data and use techniques to anonymize data to protect individuals at risk. At the technical level, data ownership should also ensure that data are safely accessible and usable by a wide group of stakeholders. Data access tools should be welcoming to people with different data literacy skills, including marginalized communities. They should provide opportunities to collect feedback from participating communities, provide chances for comments and correcting mistakes, and increase enthusiasm for and use of this data.

At the ethical level, it is important to raise awareness among users for the ethical use of data and avoid ways that harm individuals and worsen the disadvantages they are facing. This is possible with clear guidelines and procedures and examples of dos/don’ts. There are some inherited biases and discriminations in existing data collections methods and algorithms. It is important to keep looking for these and mitigating them. Building a culture of transparency in all aspects of data governance for intersectionality is critical to clarify many hidden and obscure issues that must be changed for building a better system.

Governments and national statistical offices through efforts to build better data systems in recent years have been considering a new role called Data Stewardship. This role is very promising for overseeing the modernization of the expanded data ecosystem, with increased volume, sources, use, and risks such as those mentioned in this paper. The evolving job description for Data Stewardship includes many key requirements noted to strengthen the legal, technical, and ethical foundation for national data systems. A United Nations Statistical Commission working group is looking into defining the role of Data Stewardship. This is an excellent opportunity for ensuring some of the data agency and data ownerships considerations highlighted above are included in the countries data stewardship discussions going forward.

VI. POLICY ANALYSIS

A basic purpose of having intersectional data and agency over that data is to make policies more inclusive and deliver for the excluded and marginalized. The framework in Figure 1 lists some of the conditions that influence the effectiveness of inclusive policies, such as robust legal and nondiscriminatory frameworks on the positive side and stereotypical norms and biases on the negative side. It also lists some of the expected outcomes of a successful intersectionality approach. While there are many policy implications of this approach to data, below we discuss two related foundational data issues primarily associated with the identification and documentation of different gender identities.

The intersectionality data paradox

The ability to document excluded populations by their ascribed group features is both good and bad. Counting excluded populations makes these populations visible and is required to design and implement affirmative policies that redress inequalities and enforce rights. Identifying as belonging to an excluded group can be a politically meaningful, empowering act. But it can also expose individuals and groups to stigma, discrimination, and violence if this identification takes place in hostile environments, without the protection of the state and a robust human rights legal framework.

The above is especially the case for transgender individuals who do not identify with their sex at birth in official identification (ID) documents. This can be a cause of discrimination when these documents do not match people’s appearance. Possible solutions include to delink gender from the sex attribute recorded in civil registries, collect self-determined gender identity (as Argentina, Colombia and Uruguay do) or at least collect 3 or more categories for gender rather than the binary female/male. India has a third category (T) when it collects data on sex; Nepal does the same and uses O (other). A World Bank paper on ID4D systems and SOGI inclusive design recommends the gradual removal of gender or sex markers from ID documents which reduces the risk of discrimination; it also, however, states that it does not advocate stopping the collection of sex or gender attributes in ID systems (see footnote 5).^[12]

Legal and regulatory human rights frameworks are a first line of defense against possible negative outcomes of counting excluded populations. Anti-discrimination laws provide theoretical protection, but they need to be followed by effective enforcement mechanisms. Government agencies and civil society organizations have complementary roles in ensuring their enforcement.

The need to establish synergies between data objectives for SOGI and ID4D

The right to have an official ID document is a core element of inclusive development yet to be fully realized in all countries. To have an ID and be counted, and therefore have access to the rights, protections from and benefits of the state, are basic rights and also foundational elements of a robust national level data system. Global Findex estimates that 850 million people do not have official IDs and that women living in LICs are 8 pp. less likely to have IDs than men.^[13] These ID gaps add to overall challenges of ID coverage in some countries where ID platforms are not fully developed or where barriers arise from ways ID systems are implemented. There is ongoing work to improve ID platforms and engender civil registration and vital statistics (CRVS) and ID for development (ID4D) programs to close gender gaps in having ID documents.

There is a tension that needs addressing between the objective of closing gender gaps in ID possession and the objective of accurately documenting gender identity in ID documents. Closing gender gaps in ID possession works to ensure that all newborns have birth certificates which record sex at birth, while documenting different gender identities works to delink gender from the sex attribute recorded in civil registration (CR) systems– either recoding the self-determined gender attribute or implementing the gradual removal of the gender or sex marker from ID documents. We need to define practical ways to meet both objectives in a complementary way in national data systems.

VII. IMPACT AND MONITORING

The intersectionality and gender data framework (Figure 1) has a theory of change behind it. The theory of change expects that, through better information on the lived experiences of marginalized groups and their increased participation and voice in governing data about them, there will be a marked improvement in availability and use of relevant data for policymaking. And this data will increase demand and trust for policy analysis in key development areas. Areas such as health, education, labor markets and financial inclusion will have better data to address the needs of marginalized groups. The resulting transparency and equitable processes will show clear outcomes. The agenda for LNOB and achieving SDGs would be where monitoring of results would show progress and impact of this work.

The theory of change presupposes a rather advanced data system functioning in a strong enabling environment. The National Statistical Office and other stakeholders are expected to collaborate in taking advantages of new opportunities with data sources such as CGD and adhering to best practices in data collection, use, and governance. It requires political support and resources for needed improvements along the Data Value Chain and a strong enabling environment to make these hard-earned changes mainstreamed and sustainable. So how could these changes become self-re-enforcing and help create a virtuous cycle?

Defining tangible, measurable results and monitoring impact are critical to build an equitable and safe data system. The resulting changes from an intersectionality approach to data could lead to positive or negative outcomes. Changes in improving data disaggregation by many characteristics, for example, could interfere with the best practices in data privacy agreements and lead to unintended negative outcomes directly or indirectly. Making positive differences would require monitoring if the intended outcomes were in reach and identifying the issues and what needs to change. In fact, many of the techniques highlighted in this note for improving intersectionality data also offer opport-unities to assess outcomes and impacts of intended changes. While improving voice and participation of targeted groups, for example, interviews or focus groups could help with collecting information to ensure changes being made are effective. Case studies could provide important information on if a certain set of data was effective or how it could have been more effective in informing better policies.

Overall, a systematic approach to monitoring impact would identify areas to improve, increasing accountability and transparency, and allow collecting lessons learned for course corrections and providing feedback. The monitoring system for building a better intersectionality data system could also show success of the interventions which will in turn increase support from all stakeholders including groups invested in the new data system as well as users and donors.

VIII. REQUIREMENTS

This paper has outlined several challenges and opportunities in implementing desired changes in each of the four dimensions of the intersectionality data framework. A key requirement is to increase and clarify demand for applying intersectionality in policies and with it increased political support. Without policy demand and political support, national data managers and stewards will not be able to raise the resources and capacities needed. Further work on the topic should provide further information answering some of these questions:

• What are some examples of the political support/lack of it at the international and national levels and what are some of the examples of countries that have succeeded in establishing an intersectionality approach to data?
• There is a need for an agreed international data framework for intersectionality with guidance on priorities of attributes that need to be disaggregated for gender data and other characteristics.
• A question often asked is what does it take/how much does it cost to have a minimum standard data system to respond to intersectionality needs? Could we extend the methodology used for estimating methods in the State of Gender Data Financing project to estimate costs of building up the system for intersectionality needs? This work would also provide us a mapping of where the gaps are in data and capacity of countries as well as tracking financing for this work and areas in need of more and better financing for intersectionality data.
• Building on the work done by the Inclusive Data Charter, what more is needed in terms of getting to a set of operational guidelines for countries to use to adopt better data practices for intersectionality?
• The guidance on building capacity for intersectionality along the Data Value Chain calls for being context sensitive, grounding plans on the starting enabling environment. In this context prioritizing indicators based on policy needs are critical. Should there be an agreed minimum-set of indicators for gathering ‘lived experience’ data with multiple disaggregation? This could also help with estimating what it would cost and enhancing discussions and advocacy around the topic.
• Do we have best practice examples of data governance methods and procedures that cater for what is required in terms of data agency for intersectionality data agenda? This needs to happen both with regular as well as new ways of collecting data such as CGD noted earlier. Should there be further work done to define how to adapt current practices to improve data agency?

IX. PRIORITIES AND WAY FORWARD

The expert views collected, on intersectionality and gender data, through a set of interviews in mid-2023 have provided valuable inputs to this project guiding the articulation of the proposed framework. The priority for going forward is to enhance collaboration and consultation, engaging with key stakeholders to learn from what is being done in research, data, advocacy, and development fields. This should help to motivate open discussions on opportunities and barriers for improving data systems and their enabling environment to better support intersectionality studies. Building a learning collaborative would be an important work stream. The project team could also look further into some areas that need research or preparing case studies. Ideally this work should result in creating a compendium of knowledge material. Some of the priority areas identified so far for further work include:

• Collection of country case studies and practices to ground this work and keep it focused on operations with attentions to different country contexts.
• A more systematic articulation and mapping of data gaps. This would help to unite stakeholders on priorities and if it would be helpful to propose a minimum set of indicators and take steps towards proposing an indicator framework.
• There has been demand for doing a preliminary costing based on the mapping of gaps using similar methodologies used in the State of Gender Data Financing project. This could help with engaging donors to bring focus to what it takes/costs to implement the proposed changes.
• More work is already commissioned on demystifying SOGI data and measurement opportunities so there will be new information to share on this important topic.
• Other high priority policy areas mentioned in this concept note such as SOGI and its link to national ID for development may also require bringing on board experts in this field into roundtable discussions and motivate support and build agreements on shared solutions.
• It is important that data governance and review of legal frameworks for intersectionality be covered by the case studies noted above, but it would be useful to cover this topic in a specific focused paper.
• Since the goal is to mainstream intersectionality along the four dimensions of this framework as part of countries’ operations and development, it would be good to investigate what have we learned from participatory development and social inclusion frameworks and practices. Ideally one of the partners interested to work with us on this topic from the World Bank for example could support this activity.
• Several areas in the data dimension of the framework are showing promising changes such as the movement around the CGD. There is a strong link between the CGD agenda and intersectionality data. It is important that what is needed for the intersectionality data approach stay as high priority while countries and global agencies are adopting the CGD framework start to invest in implementing them.

This concept note will be a live document. It will be used to stimulate discussions among partners at several round table events organized to gather experts to focus on this topic and also used to collect feedback from experts in specific areas such as ID for development, SOGI issues, Citizen Generated Data, and other related work streams. For providing comments or suggestions for improvements please contact:

ShaidaBadiee@OpenDataWatch.com or mayra.buvinic@gmail.com.

^[1] This concept note benefits from useful feedback from Mary Borrowman and rich interviews with José Alvarado, Lee Badgett, Kathleen Beegle, Kate Grantham, Caren Grown, Diva Dhar, Ruth Levine, Francesca Perucci, Papa Seck, Jenna Slotin, and Claudia Wells.

^[2] Crenshaw, K. 1989. Demarginalizing the Intersection of Race and Sex: A Black Feminist Critique of Antidiscrimination Doctrine, Feminist theory and Antiracist Policies. University of Chicago Legal Forum 1: 139-167.

^[3] Stewart, F. 2000. Horizontal Inequalities: A Neglected Dimension of Development. WIDER Annual Lecture 5 UNU/WIDER, Helsinki.

^[4] Dongier, P., Van Domelen J, Ostrom E., Rizvi, A. and Polski, M. 1970. Community Driven Development. Chapter 9. Bhalla, A. and F. Lapeyre. 1997. Social Exclusion: Towards an Analytical and Operational Framework. Development and Change 28: 413-433.

^[5] Domitila Barrios de Chungara, from Bolivia, voiced the plight of indigenous women at the first UN world women’s meeting in Mexico City in 1975. See Adami, Rebecca and Dan Plesch, eds. 2022. Women and the UN: A New History of Women’s international Human Rights. U.K.: Routledge.

^[6] Cite Data2X gender data gap work

^[7] Sen, A. 1999. Development and Freedom. NY: Oxford University Press.

^[8] Trust that people’s data will be protected and that they will share in the value that data can produce.

^[9] World Development Report. 2021. Data for Better Lives. Washington, D.C.: World Bank

^[10] Collaborative on Citizen Data, Citizen data: a conceptual framework, forthcoming.

^[11] United Nations, Statistical Commission Report on the fifty-fourth session (28 February–3 March 2023), E/2023/24-E/CN.3/2023/37, decision 54/102.

^[12] Lebbos, T., Esquivel-Korsiak, V. and Clark, J. 2021. ID Systems and SOGI Inclusive Design. World Bank

^[13] Findex estimates released in 2022.