It’s up to everyone, not just the Census Bureau, to help avert a looming crisis

Maria Filippelli is the Public Interest Technology 2020 Census Fellow, sitting at the Leadership Conference for Civil and Human Rights. For the past few months, she has been working with national, state, and local organizations to bridge the technology divide that exists with the new online census option.


If you know one thing about the 2020 Census, it’s probably that it will include a citizenship question that multiple lawsuits claim could lead to undercounting populations that are already marginalized and considered hard-to-count. Or you might know about the leaked emails from the Department of Justice (DOJ), stating they are open to “revisiting the sharing of census data,” a direct violation of Title 13 of the Constitution, which states that it is against the law to disclose or publish any personal information.

But what you probably haven’t heard about are the large operational changes planned for the next census. 2020 will be the first time the census questionnaire can be filled out online, a much needed change for the 21st century. This, and other modernizations like the first call-in response feature, have required the buildout of over 50 communication and information systems. If these changes aren’t successfully executed, it is highly likely that the census data will be inaccurate, leading to inequitable redistricting and funding.

So far, the future doesn’t look bright. Several of these systems are still under development and many have been under scrutiny for lack of testing and monitoring. The General Accountability Office (GAO) and the DOC Office of the Inspector General (OIG) have criticized the Census Bureau’s system deployment schedule, thoroughness of system testing, and overall project management. The GAO report identified over 3,000 security weaknesses, IT cost overruns, and lack of contractor management, in addition to labeling the 2020 Census high risk. The OIG report, published in October, found there is lack of continuous security monitoring of some systems, and that Census leadership isn’t fully aware of the risks.

In typical census cycles, new technologies and methodologies are tested and evaluated for several years, culminating in a complete test run referred to as the End-to-End test. The End-to-End test is also normally tested in multiple, varying communities, ensuring a thorough evaluation and the ability to make needed changes before the census goes live. Due largely to staff and funding issues, the End-to-End test ran in only one location this summer: Providence County, RI. Full test results have not been shared yet, though the Bureau claims it was a success.

The critiques around system readiness and increasing concerns about sharing personal information online have census advocates concerned that hard-to-count populations will become increasingly undercounted or not counted in the census. So how do we avert this burgeoning crisis?

A growing network of technologists, policy analysts, researchers, community advocates, and others are working hard to make sure everyone is counted. Advocacy groups across the country have been educating census advocates on the technical aspects of the upcoming census so they are comfortable speaking about census security and can develop accurate messaging for their communities.

Research has shown that when people understand that census data helps fund local schools, roads, and hospitals, they are more likely to complete it. Even so, there is strong evidence that the current administration’s treatment of immigrant and minority population means that a small portion of the population will not fill out the census, regardless of the benefits to their communities.

To counteract this at the state level, Complete Count Commissions have been working with local leaders, community based organizations, and others to raise awareness of the census and importance of responding. Local efforts to engage communities are building off of the success of Get Out the Vote campaigns, rebranding as Get Out the Count (GOTC). GOTC help develop messaging that resonates within their communities and address concerns around the digital divide. What does it mean if a church allows their parishioners to use a community computer to take the census, for example? Will the bureau accept multiple responses from one device? And what about people who want to fill out the census on their phones? Will it be mobile-friendly?

When it comes to understanding the technical complexities that surround the upcoming census, public interest technologists are taking the lead on breaking down the Bureau’s technical documentation into easily understandable resources. For example, knowing the difference between security threats that are internal or external to the Census Bureau is key for understanding proactive steps everyone can take to increase security. Threats internal to the Census Bureau include denial of service and man-in-the-middle attacks on their systems. External threats include phishing scams, fake websites or phone numbers, and other disinformation encountered through traditional or social media. Working with census advocates to understand these differences through stakeholder meetings and webinars, and developing easily accessible resources like short videos or shareable graphics, is key to the security of the census.

Why all the fuss for a 10 question survey? The decennial census is a once-in-a-decade chance to update House of Representatives apportionments and redraw districts to be accurately representative. In addition, the population data collected is used to allocate nearly $800 billion in federal funding annually for schools, hospitals, roads, parks, and other needed resources. Undercounting children then, for example, could mean fewer schools in an area, causing school overcrowding and ultimately leading to poorly performing schools.

The stakes are high for achieving a fair and accurate 2020 census because it is the first time in 10 years that congressional districts can be redrawn, and there won’t be an opportunity for 10 more years. That, and the funding at stake, means now is the time for engagement between all census advocates so that we count everyone only once and in the right place. What has been learned from the census work so far is that it’s up to everyone, not just the Census Bureau, to help avert a looming crisis. We may not be able to solve all of the problems, but at the very least we can help mitigate their impact, and work hard to ensure that everyone can stand up and be counted.