As a global funder, we often hear from, and share with, our peers the challenges of collecting grantee demographic data internationally. This includes questions of how to navigate different legal landscapes in terms of what data we are allowed to collect; how to approach sensitive data; and what is even meaningful to collect given the importance of cultural context.
Requirements around compliance, privacy, and security of data varies from place to place. And, identification with certain demographic categories (e.g. LGBTQ+) may actually place respondents at risk in some places. All of this makes collecting demographic data internationally more complicated. So while we have seen some significant progress in demographic data collection among those funding in the United States in the last few years, there has been less movement for those funding outside of the U.S. to identify useful ways to capture diversity data.
In 2023, the Ford Foundation made the decision to begin collecting more demographic data for organizations based in the ten regions outside of the US in which we fund. Five actions came of these efforts:
- We continue to collect data on gender as we have been.
- We began collecting data on disability.
- For some of the regions where we fund, we began collecting data on race/ethnicity in a way that would be meaningful locally.
- For those regions where race and ethnicity data was less meaningful, we began collecting data on more locally relevant categories of identity.
- Taking into account local privacy protection laws, we decided to limit international data collection to board and executive team, avoiding data collection of any specific, identifiable individual.
In this piece, we share a little bit about our process, lessons learned, and decisions around which data to collect.
Where We Were: The Challenge of Collecting Demographic Data Abroad
In 2018, the Ford Foundation revised its approach to DEI data collection for grantees with the aim of focusing less on counting people and more on embedding equity and belonging into organizations. And yet, we also recognized that some amount of data collection was important, and part of pursuing that goal of equity. In the U.S., tracking race and ethnicity was particularly salient and we relied on the following categories:
- Multiracial or Multiethnic
- Arab or Arab American
- Asian or Asian American
- Black or African American
- Hispanic or Latino/a/x
- Native American, American Indian, or Alaska Native
- Native Hawaiian or Other Pacific Islander
- White
- Other
- Unknown or decline to state
Compared to gender and disability, race/ethnicity is more complicated since it’s not a universal lens through which oppression can happen. For example, it would not be meaningful to track BIPOC-led organizations in countries/regions where there are no BIPOC communities as we understand it here in the U.S. And this is the case for the other regions in which we work: Mexico/Central America; Brazil; the Andean region; India; China; Indonesia; East Africa; Southern Africa; West Africa; and Middle East/North Africa.
But inequality based on identity exists globally. What we needed was to find a way to track what those meaningful categories were — and do so while respecting legal requirements in each country about data collection on identity. We also recognize that because we wanted to collect this information with specific categories for specific contexts, it will be more difficult to talk in aggregate about diversity overall for all of our grantees.
Applying Demographic Data Collection in Localized Contexts
We began with our regional offices around the world and the staff within them, asking what identity categories were most meaningful to them. Responses varied across our regional offices, but they could generally be categorized into one of three groups:
Group 1: Offices that believed race and ethnicity were not only relevant to track, but crucial to their local context. For those, we adapted the categories to ones that would make sense for the regions. For example:
For the Andean region, we use the following categories for race and ethnicity:
- Perteneciente a una comunidad Negro/a rural (Belonging to a Black rural community)
- Afrodescendiente urbano/a (Urban Afro-descendant)
- Moreno
- Perteneciente a un pueblo Indígena rural (Belonging to an Indigenous rural community)
- Perteneciente a un pueblo Indígena urbano/a (Belonging to an Indigenous urban community)
- Persona de raza mixta o mestizo (Mixed race)
- Persona romaní o gitano/a (Roma)
- Blanco/a (White)
- Unknown or decline to state
For South Africa, we use the following categories for race and ethnicity:
- Black
- White
- Colored
- Indian/Asian
- Other
- Unknown or decline to state
Group 2: Offices that did not view race and ethnicity as relevant to their local context, but where there were other drivers of inequality for which tracking data would be important. For example:
For India, we use the following categories for caste or tribal group:
- Adivasi
- Bahujan
- Dalit
- Other
- Unknown or decline to state
For Indonesia, we used locations within the country in order to see distribution between the power center of Java and other regions of the country.
- Bali
- Java
- Kalimantan
- Papua
- Sulawesi
- Sumatra
- Other region/island or not listed
- Unknown or decline to state
Group 3: Offices where there was concern that tracking data on race and ethnicity or other categories might further divides and could potentially exacerbate tensions rather than advance our efforts at equity. Examples of this include Eastern Africa and the Middle East/North Africa. In order to respond to these local concerns, we decided to limit our data collection to just tracking disability and gender.
Finally, in order to abide by local legal restrictions around collecting and storing personal data (many countries outside the U.S. have much stronger data protection laws similar to the more well-known General Data Protection Regulation EU law), we decided to limit data collection to anonymized aggregate data that is not tied to a specific individual. In practice, this means we ask for aggregate data about an organization’s board and executive leadership teams, rather than specific data about an executive director or board chair. For example, we ask how many women are on an organization’s board, not who on their board identifies as a woman.
For a complete list of all the categories we landed on, please click here.
What We’re Learning from the Data
This effort to expand our demographic data collection to our non-U.S. grantees was put into action at the end of 2022. After a full year of data collection under this new framework, we are now able to begin analyzing Ford’s expanded data. With any rollout of new processes, there will be a learning curve in the first year, especially with something this complex. One aim for us this year will be to create tools to help our teams interrogate this data better and make meaning of it within the contexts they are funding.
We also recognize that this entire effort has been happening against the backdrop of a changed environment here in the United States around diversity, equity, and inclusion (DEI). The Supreme Court ruling last year against affirmative action in college admissions and a rising backlash against DEI efforts in general have many companies and organizations either pulling back on their DEI commitments or at least unsure of how to navigate what has become an increasingly uncertain space. And many have wondered whether they should be collecting demographic data at all as part of grant applications.
For us, our answer has been a resounding “yes, and…” Yes, it is increasingly important for us to understand who we are funding, how they are informed and accountable to those on the margins, and how their work and institutions have changed over time to respond to the communities they serve. And we want to be able to understand that outside of the U.S. as well.
As we mentioned, with our new international data being collected with specific categories for specific contexts, it will be challenging, if not impossible, to talk in aggregate about the diversity of our grantees. For instance, we will never be able to say that a certain percentage of our grant dollars is going to People of Color-led organizations globally, because outside of the U.S. that concept either looks very different or, depending on the country, is not applicable at all.
And that is okay. We intentionally recognize that this data is not meant to be aggregated or analyzed as a whole. The alternative — applying U.S.-based race and ethnicity categories globally — would have meant that the data we are collecting would be mostly meaningless for those sitting outside of the U.S. Rather, for us what is important is making sure that we are intentionally thinking about DEI and applying that lens to our grantmaking and data collection in culturally-specific and meaningful ways.
We hope these reflections can be useful for other global funders who are struggling with similar questions, and would appreciate continuing to be in conversation as our practices and thinking continues to evolve.
Eric Li is data governance manager at the Ford Foundation, Luc Athayde-Rizzaro is grantmaking effectiveness officer at the Ford Foundation, and Bess Rothenberg is senior director of strategy and learning at the Ford Foundation.
