Census 2020

Main page content

Census 2020

CAUR partnered with the Greater Tacoma Community Foundation to identify the uses of census data ahead of the 2020 US Census. The goal was to identify the different ways that census data is used and then to understand the impact of an undercount in the Census. We specifically created data models for the Pierce County area to demonstrate the local impact of an undercount or inaccurate census data.
 

Download the report here:

CAUR explored four use-cases for census data: Planning bus routes, Identifying expansions for healthcare clinics, allocating federal grant funds and creating district boundaries for state and federal elections.

 

Bus Route Planning

Bus Route Planning

Transit operators identify potential areas of expansion by finding locations where there is a high likelihood of finding residents who will frequently ride transit. Common census tables used to identify these areas are:

  • Areas with high senior population
  • Areas with high youth population
  • Areas with high disabled population
  • Areas with low household vehicle access
  • Areas with high/low transit usage
  • Other internal data sources*

Transit Fast-Facts

Details Value Data Source
Total occupied housing units in Pierce County 312,839 ACS17 5-yr B25003
Homes with no vehicle available 17,975 ACS17 5-yr B25044
Residents using transit to get to work 13,595 ACS17 5-yr B08141
Avg (mean) commute time to work (driving a vehicle) 31.2 Minutes ACS17 5-yr S0801
Avg (mean) commute time to work (taking transit) 53.9 Minutes ACS17 5-yr S0804

Modelling the Data

Using census data to reflect these categories, we hypothesized where Pierce Transit might begin to consider expanding service areas. We identified census block groups that fit each of the critera, and then found areas where all of the shapes overlap. We then modified the census data to reflect an 8% undercount in household access for vehicles and ran the model again. 

The result was the elimination of a candidate region for transit expansion. This undercount represents only 88 homes with an inaccurate census return and we can see that the result is dramatic.                      

Actual Census Data      

     Modified Census Data

We can see that in this model an 8% error completely eliminates a census block group from contention for expansion. It is imperative that census data be completed as accurately as possible to ensure that residents can be considered for access to transit. 

 


Explore the data

Use the map below to view the census data we used for our model. Use the legend on the right to toggle layers on and off to observe patterns in the data. 

 

* Transit providers also use internal datasets with ridership details, ride times, boarding/de-boarding locations and other data to inform their decisions. The details here are for demonstration purposes only.

Business & Healthcare Locations

Healthcare Clinic Locations

Businesses rely on census data to understand the demographic characteristics of their region. Businesses want to ensure that there are enough customers in the are to support the business. Additionally, specialized businesses need to ensure that there are enough skilled employees in the area to support opening a location. For example, opening a new healthcare clinic requires that the local population have enough nurses, medical specialists, and support staff to work there.

CAUR used literature from the Urgent Care Association of America to identify potential Pierce County locations that a health clinic might consider for expanding service. As with identifying bus service areas, we identify areas that meet a specific set of criteria so that we can find areas to investigate. Health Clinics might check for:

  • Areas with high population age 21-49 (prime working age)
  • Areas with high percentage of families with children
  • Areas with high employment rates (assumed to have healthcare benefits)
  • Areas with high percentage of home ownership
  • Areas with greater than average median income

Healthcare Fast-Facts

Details Value Data Source
Total non-institutionalized population 822,539 ACS17 5-yr S2701
Total with health insurance 755,573 (91.9%) ACS17 5-yr S2701
Total uninsured 66,966 ACS17 5-yr S2701
Uninsured people with jobs 36,259 (59.6% of the uninsured population has a job) ACS17 5-yr S2702
Male/Female Breakdown 49.7% Male / 51.3% Female ACS17 5-yr DP05
Median Age 36.0 ACS17 5-yr DP05
Median Income $63,881 ACS17 5-yr S1901

Modelling the Data

Similarly to how we modeled bus route planning, we developed a hypothetical model to identify areas for health clinic expansion. We identified individual areas where each of the selection critera was met, and then used computer software to identify places where all the criteria were met. The results highlighted a possible location in the Bonney Lake area. We evaluated an undercount of 54 children in a single block group and the results were the elimination of an entire block from contention in the model. 

Actual Census Data 

Modified Census Data

A 54 child inaccuracy represents only a 6.6% undercount for that specific census table. Again, it is imperative that the census is completed completely and accurately. 

 

*Details presented here are for demonstration purposes only. 

Federal Grant Funding

Federal Grant Funding

CAUR investigated the Community Development Block Grant (CDBG), distributed by the US Department of Housing and Urban Development. This grant provides funds to cities and jurisdictions to provide a wire range of services for their communities including, but not limited to: providing housing opportunities to low-income residents, and to expand economic opportunities for low and moderate-income persons.

CDBG funds are determined through a formula to determine the financial need of a community. The inputs of this formula are provided by the US Census bureau. CAUR generated sample data to demonstrate how census data can influence funding decisions. Using an example formula and hypothetical values, we determined that Pierce County might see a significant reduction in CDBG funding with a 10% census undercount.

The table below represents Formula A from the CDBG allocation template. We hypothesized funding amounts based on 2010 Census and 2019 CDBG allocation values. These are for demonstration purposes only, actual CDBG funding may differ.

A 10% undercount in Pierce County could represent a difference of nearly $750,000 in CDBG funding.

Congressional Redistricting

Congressional Redistricting

Census data is used extensively in creating federal and state legislative districts. The Washington State Redistricting Commission is entrusted to delineate maps for our elections. In doing so they abide by the following criteria:

Districts must be drawn to:

  • Encompass, as nearly as can be done (or is 'practicable') equal numbers of people
  • Comply with the Voting Rights Act to ensure that minorities have an equal opportunity to elect representatives of their choice
  • Make sure that parts of a district are not physically separated
  • Make sure that, to the extent possible, boundaries of cities, counties, neighborhoods and communities that have common interests are respected, and their division minimized
  • Make sure they do not favor or discriminate against any incumbent, candidate, or political party

The 10 congressional districts in Washington State were drawn to be nearly identical in population. In 2011 at the time of redistricting, each district contained 672,454 people (+/- 10). Meaning that all districts had a population within 10. This incredibly precise redistricting means that there are cases where small segments of roads, neighborhoods, or towns might be shifted back and forth to equalize populations. Using the map below, you can see this phenomenon by toggling between congressional districts to see both large and small scale changes in district boundaries.


Redistricting Fast-Facts

Details Value Notes
Pierce County Council Districts 7  
Population/County Council District 110,752-116,475 Districts are within 5% population of one another
District Undercount 14% An undercount of 14% is like losing an entire district's population
   
State Legislative Districts 49  
Population/State Legislative District 137,192-137,285 Districts are within 0.07% population of one another
District Undercount 17% An undercount of 17% is like losing an entire district's population
   
Federal Congressional Districts 10  
Population/Congressional District 672,454 Districts are within +/- 10 people of one another
District Undercount 10% An undercount of 10% is like losing an entire district's population