Download PDF
Perkins, Deja, Dillon Mahmoudi, Erica Henry, and Caren Cooper. 2026. “Geography of race and income shape spatial data gaps in two national participatory science projects.” People and Nature, 8(5), 1485–1498. doi.org/10.1002/pan3.70307.
Summary
This paper shows how the spatial logics of race and income quietly structure where “participatory science” data actually get made, and where they are conspicuously absent. Drawing on more than 90 million eBird and iNaturalist observations across the United States, the paper demonstrates that these marquee biodiversity platforms are not neutral infrastructures but deeply entangled with residential segregation, wealth stratification, and the long geographic afterlives of white supremacy. In doing so, the paper reframes contributory science as part of the broader machinery that can either reproduce or contest environmental injustice, depending on whose places count as data-rich and whose are systematically ignored.
Contributory projects may inadvertently be creating data-rich and data-poor areas, amplifying their perceived value in whiter, more affluent spaces while diminishing attention to less privileged places.
Methodologically, the paper uses hurdle models to assess both the presence and abundance of observations relative to census tract racial composition and median household income, disaggregating results for urban and rural contexts. This design moves beyond participation headcounts to ask whether the geography of people aligns with the geography of data, and how that alignment shifts across different settlement types. The decision to hold race and income together, but analytically distinguish their spatial effects, makes visible how different socio-spatial processes (segregation, deindustrialization, incarceration, and uneven development) converge into patterned data gaps.
Empirically, the paper finds that urban eBird data are significantly less likely to occur, and are less abundant, in tracts that are predominantly Black, Indigenous, and people of color, while both urban and rural iNaturalist data are significantly underrepresented in lower-income tracts. These patterns show that the much-celebrated democratization of biodiversity science rests on a profoundly uneven substrate: contributory projects help produce data-rich and data-poor geographies that map uncomfortably well onto racialized and classed landscapes. Because environmental data operate as a form of social capital—mobilized to justify investments, shape regulations, or support “evidence-based” planning—these gaps risk amplifying existing inequalities in environmental governance.
If contributory science projects are not collectively creating an equitable distribution of data, then they are exacerbating environmental injustice rather than democratizing knowledge production.
Ultimately, we argue in this paper that such data gaps are not incidental noise but the cumulative outcome of white supremacy culture, colonial land-ordering, and project designs that privilege whiter, more affluent spaces as default sites of scientific value. We call for disaggregating participation data by specific minoritized communities, rejecting the flattening effects of “BIPOC” as a catch-all category, and insists that any attempt to “fix” these gaps must be community-specific and historically grounded. In this way, the paper situates eBird and iNaturalist as exemplars of a broader field of contributory projects whose compounded spatial biases constitute an environmental justice problem in their own right.