The Digital Equity Lab is pleased to release our report "Preparing For the Digital Decennial Census" which includes an analysis of risk, safety, and trust for the first-ever digital decennial census. As advocates for structural equity, we believe the digital transition of the census is understudied and requires both attention and investment at national scale in order to achieve a level of success sufficient for our emerging era of digital governance.

In 2020, the United States will hold its first digital census. The Constitution mandates that every ten years the US government must count every household in the country. The demographic and economic data collected is then used to decide $700-$900bn in federal funding for federal and state programs and to redraw electoral districts at the federal and state levels. Despite the persistence of a national digital divide — 35% of US adults still do not have internet at home —the Census Bureau is moving forward with its plan to ask 80% of households to complete the 2020 survey over the internet.

This strategy prioritizes counting those who have internet access at home, for whom the process will be simple and quick. Yet because of the demographics of the digital divide — 53% of Latinx and 43% of Black adults, more than half those earning under median income, 42% of rural residents, and half of elders age 65+ — optimizing the count for the best-connected among us could lead to an overcount of affluent White populations and a systemic undercount of immigrants, people of color, and children, mirroring existing structural inequities. This in turn could lead to underrepresentation in government for unconnected populations, the loss of funding for critical social programs, and a skewed idea of who we are as a nation.

To offset the effect of the digital divide, the Bureau suggests that those without internet access at home fill out the online survey by going to a neighbor, a local organization, or finding public internet access points; yet neither the federal government nor, in most cases, local governments are resourcing public-facing digital infrastructure at scale for census self-response. This lack of core investment may further undermine the digital method for attaining a fair and accurate count. In the absence of coherent proactive messaging about the digital process and sufficient access and literacy support for digitally marginal and challenged populations, the Digital Equity Laboratory (DEL) and its partners are deeply concerned about the success of the digital transition of this massive public mobilization.

In response, DEL with the NY Counts 2020 campaign’s Tech & Tools Committee initiated a holistic risk assessment process to:

1) identify potential vulnerabilities of the digital census for individual participants, at-risk communities, and organizations and institutions acting as census sites
2) suggest mitigation strategies to protect the count while also protecting communities already disproportionately impacted by the digital divide and by predictive and surveillant technologies.

Our goal is to provide digital tactics and techniques to help prevent possible harms, and enable communities and agencies to better prepare against the uncertainties of a digital census and the likelihood of a resulting undercount. Our focus is at the user-level, and aims to address holistic safety concerns, not solely cybersecurity.

The report that follows provides the best possible recommendations given a number of uncertainties about the Census Bureau’s plans and systems, and takes a “power not paranoia” approach to building capacity and awareness among community stakeholders, rather than ignoring or glossing over uncertainties and threats.

Full Report: Preparing for the Digital Decennial Census: Building Consent, Equity, and Safety into Digital Transition
(*accessible interactive PDF available here)

This report was co-authored by Greta Byrum and Meghan McDermott based on risk assessments and digital safety recommendations compiled by Nasma Ahmed, Sarah Aoun, David Huerta, Sam Lavigne, Dhruv Mehrotra, Rebecca Ricks, and Norman Shamas. This work was inspired by the critical work of the New York Counts 2020 campaign, which has built a powerful network of community advocates and organizers dedicated to ensuring the most robust possible count in 2020 while supporting their communities. We are grateful for the New York Counts 2020 Steering Community’s valuable feedback in the process of compiling these recommendations, as well as to the New York Counts Tech & Tools Committee, which has advised and shaped this work every step of the way; and to the Open Society Foundations for their support. We are particularly grateful to Lauren Moore for her perspective and partnership, to Holly Dowell for her support and skill, and to Maya Wiley for her leadership.

This work is licensed under a Creative Commons Attribution-NonCommercial 3.0 Unported License.

The original reports by security consultants are available:
- Census 2020 Community Risk Assessment and Recommendations | David Huerta, Sam Lavigne, and Dhruv Mehrotra
- Risk Assessment of the 2020 U.S. Census: Recommendations for Action | Nasma Ahmed, Sarah Aoun, Rebecca Ricks, and Norman Shamas