Measuring Community Resilience Equitably

The Community Resilience Estimates provide an easily understood metric for how at-risk every neighborhood in the United States is to the impacts of disasters, including wildfires, flooding, hurricanes, and pandemics such as COVID-19.

Modeled estimates are based on 10 resilience-related risk factors. Current estimates are modeled using 2019 American Community Survey data and display the number and percentage of residents living with 0, 1-2, and 3 or more risk factors for the nation, states, counties, and census tracts.

In addition to providing an easily understood metric of social vulnerability, the CRE program provides data users the information they need to make informed decisions.

The 2019 Community Resilience Estimates Equity Supplement adds to the discussion of equity by pairing data from the 2019 Community Resilience Estimates with 2015-2019 American Community Survey. The end product is a dataset which provides the latest information on social vulnerability and equity in one source.

The CRE program provides data that cannot be replicated with aggregated publicly available data alone because it uses data about individuals and households from the American Community Survey (ACS) to help determine why a community is considered “at risk”.

The CRE for Equity data tool guides users in the interpretation of the uncertainty ingrained in survey estimates. This is done by providing “flag variables” in the CRE for Equity that provide context and help users understand if individual statistics are higher or lower than the national average.

All CRE products have the ability to measure social vulnerability in a more equitable way because the CRE products start by analyzing information about individual people and households.

In contrast, other indices start their analysis by using publicly available information about counties and census tracts — not individuals.

Publicly available Census Bureau data is extremely valuable for community organizers and disaster management. When these data are rolled up into an index, they can sometimes mask the actual risk level of an area because it does not take into account that risks are not spread proportionally throughout a community.

For example, if a community had 10 people and each had a single risk factor such as health status — for a total of 10 risk factors — the CRE would determine that this community was not at risk because the relative risk for each person was low.

But if a community had the same 10 risk factors and all of them were concentrated in three out of 10 individuals, the results would change.

If two of those individuals had three risk factors and one person had four risk factors, the CRE would flag this community because 30% of the population would be high risk while the rest of the community would remain low risk.

This example shows how an inequitable distribution of risk can be hidden in data measured at the community level rather than at the individual level.

On the surface, Coosa County in rural Alabama may not appear to have many concerns about resilience.

The county has poverty and health insurance coverage rates that are on par with the national average. And the county performs above the national average in vehicle availability and English language speaking ability.

While these are encouraging characteristics and the county appears resilient, additional analysis tells a different story.

The CRE shows that 33.8% of residents of Coosa County are considered to be “at-risk” or have three or more risk factors.

This is likely due to the county having an above average percentage of people who have not graduated high school, fewer households with a broadband internet subscription and more people with a disability.

Combining this information using publicly available data can end up muting the outcome of an analysis, making the county seem to be at moderate risk.

In addition, the CRE shows that in Coosa County, these risk factors seem to be concentrated in the Black or African American population.

While the county has a poverty rate that is not statistically different than the national average, the poverty rate for individuals who are White alone is lower than the national average and the rate is higher for individuals who are Black or African American alone. “Alone” refers to people who report only one race.

In Coosa County, individuals who identify as Black or African American alone are also less likely to have internet access in their homes than individuals that are White alone. Finally, Black or African American individuals in the county are less likely than White individuals to be high school graduates.

The concentration of any of these characteristics in a population leads to concerns both about social vulnerability and equity. The CRE’s methodology captures this vulnerability and classifies Coosa County as an area of concern while other indices may miss these subtleties by relying only on aggregated data.

Each community will have its own nuances and the methods used by the CRE attempt to ensure that these concerns are captured by analyzing individuals and their circumstances. By doing this, the Census Bureau is able to determine an individual focused measure of community risk in the face of disasters and other emergencies, including the pandemic.

With the release of the CRE for Equity data product, users can now also make informed decisions about social vulnerability and equity. The CRE will continue to publish updated data.