Hard to Count 2020



Visualizing Hard to Count communities for the US 2020 census

See The Visualization

Why have we done this?

The census only occurs once every 10 years and is critically important

Census data is used to:

  • Determine the apportionment of seats in the U.S. House of Representatives
  • Define legislative districts and school districts
  • Distribute over $675 billion funds to local, state and tribal governments
  • Inform redistricting efforts

The current administration seeks to undermine the census

  • At the end of March 2018, Secretary of Commerce Wilbur Ross approved a last-minute request by the Department of Justice for the addition for inclusion of citizenship question in the 2020 census
  • Underfunding of the census has led to only one end to end test and a lawsuit from the NAACP
  • The citizenship question was added too late to be part of the sole end to end test, so there will be no testing on the effect of adding the question
  • Immigrant communities are already fearful and distrustful of the current administration due to anti-immigrant and anti-Muslim rhetoric and policies
    • The inclusion of a citizenship question will have a chilling effect and may lead to a failed census
  • As of mid-April 2018, four lawsuits have been filed on the constitutionality of the inclusion of a citizenship question
    • These suits represent 17 states and 7 cities
    • The Brennan Center is tracking current cases here

Our target audience

This visualization is intended for funders, advocates, elected officials, and non-profit organizations who want to get out the count for at-risk populations.

What is hard to count?

The census bureau has done research over the years to determine which populations are missed at higher rates in the census. Their research shows that children, homeless, lower income, lower education, English language learners, undocumented immigrations, and racial/ethnic minorities are least likely to be enumerated properly.

Since the 2020 census will be conducted mostly online, households with poor internet access will likely also be undercounted.

Which metrics were included and why?

The metrics that we visualize are proxies for whether a community might be harder to count. In our visualization, darker green indicates harder to count.




Race & Ethnicity

We included percentage households of different ethnic backgrounds (percentage Latinx, percentage Asian, and percentage Black)

Income

We included percentage of households with renters as a proxy for income

Return Rate

We included failure to return rates from the 2010 census as a proxy for overall failure to return rate for the 2020 census

Internet Access

We included percentage of households with poor internet connectivity to show communities most likely to be affected by the census moving online

Children

We included the percentage of households with children under 5, since young children are most likely to be missed

2020 Census Hard to Count Visualization

Who we are

We are a team of UC Berkeley Master of Information and Data Science (MIDS) students who think the 2020 US census is a big deal and this is our W209 final project!




Robert Deng

Krista Mar

Sombiri Enwemeka

Our data sources

  • For our hard to count metrics we used data from several sources that we linked by using the census bureau’s geographic identifiers at the census tract level. We used FIPS codes for county level visualizations
  • We used the census bureau’s API to access the 5 year estimates available from the American Community Survey (2010-2015) for the race, ethnicity, and socioeconomic variables
  • We used the census bureau’s planning data for 2010 census return rates and the census bureau’s official low response score
  • We used the Federal Communication Commissions’ internet access services data to determine internet connectivity