Download the PDF in English and Spanish.

CMAP analysis of U.S. Census Bureau data and federal COVID-19 relief funding found that:

• Communities of color are filling out the census at much lower rates than communities with mostly white residents. Compared to the last census, some communities have self-response rates more than 20 percentage points below rates from 2010.

• Hispanic communities (in both Chicago and the suburbs) have the biggest drops in participation since 2010, with average declines of more than seven percentage points. This mirrors trends in other large urban areas in the country.

• CMAP estimated how a census undercount could impact federal resource allocation by examining the CARES Act formula. If the state’s 2019 population estimate had been 5 percent less — and the population of other states stayed the same — Illinois would have lost $235.5 million in COVID-19 relief funding.

• To increase a complete count by September 30, 2020, community leaders and groups should share the effects of an undercount and encourage their networks to fill out the census. They can use this social media toolkit to help get the word out.
   

Fewer people in northeastern Illinois are filling out the census, putting the region at risk of a costly undercount. A new analysis of 2020 census data by the Chicago Metropolitan Agency for Planning found that the region’s reduction in self-responses, compared to 2010, is concentrated in communities with large Hispanic and Black populations. The COVID-19 pandemic has made achieving a complete count more difficult at a time when maintaining government support is more important than ever.

The implications of an undercount for the region overall are substantial. All seven counties have communities, including rural, suburban, and urban areas, with low response rates. More than $675 billion in federal funding is distributed in the U.S. every year based on census numbers. An undercount would mean billions of lost federal dollars for the region, creating further barriers to equitable reinvestment.

The recent announcement by the U.S. Census Bureau that field data collection is ending on September 30, 2020 — one month earlier than planned — increases the urgency of getting a complete count of all residents in northeastern Illinois. During outreach activities over the next few weeks, community leaders and groups must use every resource to reach everyone and urge them to complete the census.

The health and economic crises brought on by COVID-19, compounded with the effect of efforts to include a “citizenship question” in the 2020 census, make a complete count exceptionally difficult to achieve this year. This was also the first primarily digital U.S. Census, presenting a new challenge for households with limited internet access or computer literacy.

To better understand the challenge, CMAP analyzed data on the current rates of “self-response” across the region, compared to 2010. These data capture individual households’ voluntary completion of the census and are distinct from responses gathered by Census Bureau staff on the ground. The map below shows that low self-response rates are a shared challenge. Each of metropolitan Chicago’s seven counties has communities — urban, suburban, and rural — with below average response rates, including some areas with rates lower than 40 percent.

Response rates often vary across the region, but disparities in response rates are worse. Net census responses have historically been lower in communities of color. The regional self-response rate has fallen from 70.8 percent in 2010 to 69.1 percent in 2020 as of August 19, and the data suggest that areas with more people of color are filling out the census at lower rates than similar areas did in 2010. This small regional gap obscures bigger disparities in some tracts that could significantly impact future resources in these communities. The map below shows census tracts across the region where current self-response levels are substantially below 2010 levels. Compared to the last census, some areas have rates that are lower by more than 20 percentage points.

CMAP identified these disparities by analyzing the relationship between self-response rates and the share of nonwhite residents in a given census tract. On average, tracts that have at least 50 percent white residents have the same response rates as tracts with a similar racial composition in 2010. However, tracts with predominantly Black, Hispanic, and Asian residents have much lower rates than they did in 2010. (The Census Bureau uses the term “Hispanic” to refer to residents of any race who have self-identified as being of Hispanic ethnicity.)

Black and Hispanic communities have average self-response rates of 55 percent — 14 percentage points below the region’s current rate of approximately 69 percent. The decline is sharpest in majority Hispanic communities, where rates have fallen by more than seven percentage points since 2010. As shown in the map below, this pattern is playing out in communities across the region. Majority Black communities (as well as majority Asian communities) have seen smaller, but still concerning, average declines of two to three percentage points versus 2010.

Even during the pandemic, some communities, including Aurora and Elgin, are improving results through recent awareness-raising activities, including parade-like census caravans that encourage residents to respond. While some tracts in these cities still have self-response rates below 2010, local census leaders and the Metropolitan Mayors Caucus have reported that responses in the targeted areas accelerated immediately following the caravans. More ideas on how to encourage census participation are available from the Mayors Caucus’ Peer Advisory Group.

Many of these disparities are also present in the response rates for peer regions. Self-response rates have fallen compared to 2010 in 16 of the country’s 20 largest metropolitan statistical areas. Lower self-response rates in Hispanic areas across almost all peer regions are driving a disproportionate share of that decline. While metropolitan Chicago is doing better than peers like Philadelphia, Boston, and New York City, the region is underperforming relative to others like Detroit and Washington, D.C.

The recently enacted federal Coronavirus Aid, Relief, and Economic Security (CARES) Act illustrates the importance of accurate census data for Illinois communities. One CARES Act provision divided $139 billion among states according to population. Based on 2019 Census Bureau estimates, built on the 2010 census count, Illinois received $4.91 billion to help respond to COVID-19.

To show how the census influences federal resource allocation, CMAP  estimated the impact of a significant undercount on the State of Illinois’ CARES Act funding. If the state’s 2019 population estimate had been 5 percent less — and every other state’s population remained unchanged — Illinois would have lost $235.5 million in COVID-19 relief funding.

Another way to understand the importance of counting all residents is to look at the impact at the individual level. Assuming other factors held constant, under the same scenario of a 5 percent lower count, the state would have lost $371 for every uncounted person. That would have resulted in less funding for the state and local governments to buy personal protective equipment, clean facilities, and address other critical needs related to COVID-19. Local governments still need to support uncounted people but must do so with fewer resources.

Northeastern Illinois was already facing uneven population growth and decline. That longer-term decline, coupled with 2020’s unique challenges, make a complete count in Illinois all the more important.

Complete Count Committees throughout the region are working to promote census participation. These groups are supported by the Illinois Department of Human Services’ Census Office, which was created by Governor J.B. Pritzker in 2019 with the largest investment for census activities in the nation. To distribute these funds, Illinois tasked Regional Intermediaries with awarding grants and partnered with the Metropolitan Mayors Caucus (MMC) to help local governments and their Complete Count Committees. One of many organizations committed to census outreach, MMC has a program, Complete the Count 2020, which helps municipalities coordinate census efforts and enable peer-to-peer sharing of best practices. CMAP’s regional analysis of the geographic and demographic trends in census response rates are helping these committees and other partners target outreach activities to areas with the greatest need.

In the face of a daunting economic recovery and the continued effects of the pandemic, a strong last push is vital to secure deeply needed resources over the next 10 years. Filling out the census is one of the quickest ways to make sure Illinois has the resources it needs over the next decade. As we enter the final month of the census period, local outreach efforts must receive strong support to ensure a complete count.

CMAP is working with partners and elected officials to re-extend the census completion date. Last week, CMAP sent a letter to the northeastern Illinois congressional delegation regarding the Census Bureau’s decision to shorten the outreach deadline and its impact on the region. CMAP also signed a letter sent to congressional leadership this week by the National Association of Regional Councils, Association of Metropolitan Planning Organizations, and National Association of Development Organizations. The letter calls for additional data collection time, raises concerns about the use of differential privacy (which may undermine resource distribution in high population areas), and urges an accurate count in communities with colleges and universities while students are unable to live on campus.

As shown by the work in Aurora, in Elgin, and through local Complete Count Committees, local leaders can play a key role in supporting increased participation. The Metropolitan Mayors Caucus has resources and updates to share with local governments and can connect municipalities to nonprofits for more direct engagement with residents. Contact LaTiyfa Fields at lfields@mayorscaucus.org for information.

Between now and September 30, CMAP urges local leaders and our partner organizations to highlight with their networks the costly impact of an undercount, particularly for Black and Hispanic communities. Here’s a social media toolkit to help get the word out about filling out the census.

How ready is your community or county to help improve the census count?

• How are you engaging with Black and Hispanic residents? What strategies have you developed to encourage more residents of color to fill out the census?

• Who are the trusted voices you can use to reach undercounted residents, explain to them why the census is important, and answer questions?

• What community benefits would be jeopardized if you had fewer funding and resources over the next 10 years? Make this part of the story as you talk about the importance of a complete census count.
   

About the data

The U.S. Census Bureau releases self-response data through its publicly accessible data portal. For the 2020 census, this includes households that have responded to the Census Bureau’s invitation to complete the census, both over the internet and by mail. The Census Bureau releases the share of households that have responded as the “self-response” rate and continues to update the data throughout the census period. This data does not include any responses gathered by Census Bureau’ enumerators: the staff who visit households that have not yet responded online or by mail.

CMAP analyzed tract-level, self-response data for both the 2010 and 2020 census. Longitudinal comparisons leveraged 2006-2010 5-year ACS data to establish the racial composition of each tract in 2010 and the 2014-2018 ACS data (the most recent available at the tract level) to establish the racial composition of each tract in 2020. Average response rates by group were determined by performing a weighted average of response rates on the relevant tracts, weighted by the number of housing units in a given tract. Because these demographic data are associated with 2010 tract definitions, they were interpolated into the planned 2020 tracts using a correspondence file provided by the Census Bureau. Averages based on tract-level data are similar but not identical to averages obtained using higher-level geographies (such as counties), which may be due to discrepancies introduced when relating data between 2010 and 2020 tract geographies. For reference, average responses for the region using county-level figures were 69.8 percent in 2010 and 68.9 percent as of August 19, 2020.

To work with this data, CMAP used APIs to directly download tract-level demographics and then used R, a free open source statistical package, to analyze the data. The R script used to perform this analysis is available on CMAP’s Github page. Partners are encouraged to use and build on this analysis.