The 2020 Census and the Dashboard

Aug. 17, 2022

Taylor Lampe & Alex Chen

While exploring some of our new 2020 data derived from the American Community Survey (ACS) – metrics such as Children in Poverty, Excessive Housing Cost, and others – you may notice that your city’s maps look a little different between 2020 and older years. For example, take a look at maps below for Fayetteville, AR. You can see that not only has the distribution of Broadband Connection changed between 2019 and 2020, but some of the census tracts (i.e. neighborhoods) seem to have changed shape, and there are more census tracts in 2020 than in 2019 (for example on the west side of Fayetteville).

Broadband - 2020 Census

What’s going on here? You are seeing the impact of the 2020 Census!

What is the US Census?

The U.S. Census occurs every ten years, and its goal is to provide as complete of a count as possible of all people and households across the United States, as well as various important characteristics of those households. Census data are used to allocate funding for government programs and re-draw congressional boundaries, among others.

The decennial Census is also fundamental for calculating key metrics on the City Health Dashboard and updating the maps of those metrics. Census tract boundaries (see our blog about them here) are updated by the Census every 10 years to more accurately reflect population distributions. City boundaries can change, too – generally to reflect local changes like annexation, consolidation, and more. All of our data sources will eventually start using the 2020 Census geographic boundaries as the foundation for their metric estimates; our ACS metrics (using the US Census as the data source), are just the first to do so.

How does this impact the Dashboard?

You may notice differences between tract maps when you look at 2020 ACS data vs other metrics or years. The most visible changes will be tracts that split into two (or more) because of population growth. In some cities, you’ll likely notice these changes on the maps! In Bend, OR, changes are especially visible in the city's center neighborhoods.

Diversity - 2020 Census

You can see both Bend and Fayetteville below in the top 10 Dashboard cities with the highest percent of census tracts splitting in 2020. The impact on these maps will be much more visible. These tend to be either very small cities, or more suburban areas that have experienced huge population growth in the last decade. While larger, more urban areas also experienced population growth (and therefore, many tracts splitting in 2020), those cities are so big that the map impacts may not be as visible. On average, Dashboard cities had about 20% of census tracts split in 2020, with some not changing at all, and some cities with all tracts splitting! Check out your city’s maps to see your local impact.

Tract Splitting

You may also notice changes to city boundaries. Unlike census tracts, cities boundaries are dictated by the local government and don’t necessarily change when the population grows. While these are mostly more subtle changes, there may be a few city boundaries, like Sugar Land, TX, that change substantially. Sugar Land’s boundary expanded to the west in 2020.

Segregation - 2020 Census

We also added 42 new cities to the Dashboard in the July 2022 update – thanks to population growth. These cities surpassed the Dashboard’s minimum population threshold of 50,000 residents in the 2020 Census. Explore which cities we added here.

How to assess trends over time

If you are comparing maps for 2019 (or earlier) with 2020 ACS data, and the city or tract boundaries change significantly, we recommend taking a zoomed out approach to identify general visual trends of darker and lighter areas between years, since you can’t always compare values in the exact same tract across time. We also recommend reviewing our tips on working with multi-year data here!

For more advanced users interested in accessing our data via the API or downloadable data, we have added a variable called census_parent_shape_year that distinguishes which Census (2010 or 2020) was used to calculate the metric. You can access advanced resources for standardizing data across 2010 and 2020 tracts from our friends at NHGIS.

Feel free to email [email protected] with any questions. Enjoy the new data and maps!

Explore More