Census 2020 Outreach Efforts Hampered by COVID, But Continue

Boston Indicators wrote this brief with input and guidance from the Massachusetts Census Equity Fund, which supports community organizations and grassroots leaders who are organizing within their community to ensure the most accurate count in the Commonwealth.

The United States Census is the oldest regularly recurring data-gathering exercise in the world. It provides invaluable information about who makes up our large, diverse country, and it helps us analyze how we’re changing over time. Since it won’t happen again until 2030, results from the 2020 count will have decade-long implications for how particular communities are represented in government and for how hundreds of millions of federal dollars are distributed across Massachusetts cities and towns. For more detail, see Boston Indicators’ report Census 2020 Explained: How It Works and What’s at Stake for Massachusetts.

Online self-response for the 2020 census went live on March 12, two days after Governor Baker declared a state of emergency. Having the Census emphasize online responses for the first time this year has provided a useful mechanism for households to fill out their forms even while socially isolating at home. But even though the first wave of the 2020 count is happening online, a core strategy for getting a complete statewide count has always involved proactive door-to-door outreach to support households that nonetheless do not fill out their form online.

While Census self-response rates grew steadily in March and April, since around mid-May both Massachusetts and U.S. self-response rates have plateaued. As a result of the pandemic, many local organizations that would typically lead get-out-the-count efforts were shuttered during the first months of the count. Even though libraries, community centers, Councils on Aging and more are slowly reopening, they remain largely unable to offer their spaces or services to assist in self-response. So while Massachusetts is on track for a higher response rate than the U.S. overall, we are still below our self-response rate for the 2010 Census, at 68 percent.

Looking at statewide response rates, however, obscures large differences within Massachusetts. Lower-income communities, especially those with large immigrant populations that speak a variety of different languages, have tended to have lower census response rates. A range of socio-economic challenges contribute to these lower response rates, but this year may be especially challenging in wake of the Trump administration’s attempt (albeit unsuccessful) to add a citizenship question. While this question was not ultimately included, it’s quite possible that the effort itself is still having a chilling effect leading to depressed response rates.

The two interactive features below allow users to analyze which communities have the lowest self-response rates to date and also compare that with important demographic information. The map in the first feature shows self-response rates for every city and town statewide, and the interactive table allows for detailed comparisons with measures of population change since the last decennial census in 2010. The most recent population estimates we have are from the 2018 American Community Survey, so these tools compare roughly eight years of change (note: these estimates are especially rough for smaller cities and towns). Places that have changed most since 2010 are of particular concern, because without an accurate 2020 count these new population centers risk being underrepresented for years to come.

Perhaps most troubling from this data is the fact that Chelsea has both the state’s single-highest coronavirus infection rate and also the third-lowest self-response rate (among the largest 20 percent of cities and towns). Chelsea’s response rate of 49.3 percent is 15 percentage points below the current state average.

The vast majority of cities and towns with low response rates are places with large immigrant populations and large communities of color, and are often particularly hard-hit by COVID-19. These include places like Lawrence, Everett and Worcester. Falmouth, Fall River and Barnstable are three outliers in this regard. They also have lower self-response rates but are each more than 75 percent White.

Below is an interactive scatterplot that uses some of the same data to facilitate comparisons between current self-response rates and individual measures of population change from 2010 to 2018. We’ve set the default graph to look at foreign born residents, because they are among the groups that most often go undercounted. We’ve also set the default to look at total population change, but for each of these measures you can also toggle to percent change from 2010 in order to adjust for different base population sizes.

Some additional key findings from this work:

• The cities and towns with the greatest absolute increases in foreign born, Black and Latino populations also have some of the lowest self-response rates statewide. Absent strong get-out-the-count efforts in coming weeks and months, these communities are at the greatest risk of being further underrepresented by the 2020 Census.
• Boston’s foreign born population increased by about 27,000 between 2010 and 2018, and troublingly, the city’s current self-response rate is only 52 percent (the sixth-lowest among the state’s largest cities and towns).
• Brockton had the largest absolute increase in its Black population statewide, larger even than Boston’s Black population increase (+17,713 vs +17,600), even though Boston’s total population is more than seven times as large. No other cities have gained more than 5,000 Black residents.
• By contrast, the state’s Latino population is growing across a larger number of different cities. Boston still has the largest growth of Latinos in Massachusetts, but 11 cities have grown their Latino populations by around 5,000 or more. Asian population growth has also been more evenly spread.
• Many towns in far western Massachusetts and on Cape Cod also have very low self-response rates. The Census Bureau uses a different, more labor-intensive process for encouraging responses in these more remote areas, and this process has been slowed significantly due to the coronavirus.
• Among largest 20 percent of communities statewide, Chelsea, Boston and Cambridge are the three fastest-growing—each of them have grown roughly 13 percent since 2010. While Cambridge is doing a bit better, all three have self-response rates well below the state’s 2010 level.

Reacting to the unique challenges posed by COVID-19, the Census Bureau originally extended the response period by 3 months to October 31. The Bureau has since shortened that timeline, scheduling an end to self-response and non-response followup by September 30, potentially impacting the success of the count. In addition, every operational timeline has been extended. Originally scheduled to open in March, field offices across the United States only began opening in May, with Massachusetts’ offices opening later than many others on May 25th. The count of residents in group quarters and other enumeration processes have also been extended. Likewise, the homeless count has been officially rescheduled to September 22-24 after originally being scheduled for early April. Processing of data for political redistricting purposes has been pushed back to May 1, from January 1, 2021.

Without doing effective real-time organizing work over these coming months, there’s real risk that federal funding and political representation for these lower-income communities will suffer over the coming years.