Measuring Community Resilience

Ohio Census

Given the somewhat recent emphasis on social determinants of health (SDOH), the last few years have seen a sharp increase in the number of indicators and data-sets that tap one or more SDOH measures. One such series, Census-generated, is the community resilience data, with reslience defined as “the capacity of individuals and households to absorb, endure, and recover from the health, social, and economic impacts of a disaster such as a hurricane or pandemic.” In this post I explore these data to better understand what they may tell us.

Ani Ruhil true

Photo by Alex Shute on Unsplash

In 2018 the Census Bureau released experimental estimates of community resilience – “the capacity of individuals and households within a community to absorb the external stresses of a disaster.” These estimates were subsequently updated and released in August of 2021, and have drawn the attention of policymakers and stakeholders because of the ongoing COVID-19 pandemic.1

In this post I take a closer look at them, with the goal of understanding how Ohio and Ohio’s counties compare to the rest of the nation, and of course zeroing in on our Census tracts as well. In the long-run this may lead to other explorations via the social vulnerability and area deprivation indices, respectively. Two estimates are of primary interest to me, (i) the percent of the population with one to two risk factors, and (ii) the percent of the population with 3 or more risk factors.

How are these risk factors estimated? The technical documentation points out that individual and household data are combined from the following metrics:

Each individual who meets a given risk criterion is assigned a binary value of 1 (0 if they are not deemed to be at risk for that criterion). These series of 0 and 1 values are then summed for each individual. Then the total number of individuals with 0 risk factors is calculated, followed by the total number with 1 or 2 risk factors, and then the total with 3 or more risk factors. Dividing by the population in the geography (state, county, etc) generates the percent of the population with 0, 1 to 2, or 3 or more risk factors. The 90% margin of error accompanies each estimate, and would be a good indicator of the uncertainty surrounding these estimates. As a benchmark, note that more than 65% of Americans have one or more risk factors (see Table 1). Note also that the smaller the percent of the population with 1 or more risk factors, the more resilient the community to fallout from a natural disaster, health pandemic, and so on.

How do the states fare?

Figure 1 shows the percent of each state’s population with 1 or more risk factors. Minnesota tops the list, with only 56.1% of its population with one or more risk factors, while New Mexico tails with 73.3% of its population with one or more risk factors. Ohio is 13th (60.4%), doing much better than most of the states.2

The Counties?

Surprise, surprise, Delaware (ranking 6th) is in the top-10 list of most resilient counties, while Warren also features in this marquee list (ranking 7th). Alaskan ‘counties’ feature very prominently in the bottom of the distribution (with six of the 10 least resilient locations).

Table 2: The 10 Most versus Least Resilient Counties
Most Resilient
Least Resilient
County Percent County Percent
Carver County, Minnesota 44.31 Kusilvak Census Area, Alaska 95.25
Hamilton County, Indiana 44.78 Northwest Arctic Borough, Alaska 92.81
Scott County, Minnesota 44.87 Dillingham Census Area, Alaska 92.39
Hanson County, South Dakota 44.95 Todd County, South Dakota 91.45
Douglas County, Colorado 45.41 Oglala Lakota County, South Dakota 91.20
Delaware County, Ohio 46.12 Nome Census Area, Alaska 90.71
Warren County, Ohio 46.39 McKinley County, New Mexico 90.69
Dallas County, Iowa 47.08 Bethel Census Area, Alaska 90.45
Loudoun County, Virginia 47.46 Yukon-Koyukuk Census Area, Alaska 90.29
Forsyth County, Georgia 47.63 Apache County, Arizona 89.56

What about the rest of our counties? How do they fare? To explore this question we turn to a static map that displays the percent of the county population with one or more risk factors.

As is quite evident from the map in Figure 2, there is quite a bit of variability across the counties. Compared to Delaware and Warren, both of whom have less than 50% of their population with one or more risk factors, all other counties have more than 50% of their population in this risk category. Monroe (70.3%) and Holmes (71.6%) have the most vulnerable populations, and so do Southeast Ohio counties.

Last but not the least – the Census Tracts!

In my experience, one of the problems with diving below the county level is that you start seeing blobs rather than anything too clear, particular in places with tightly clustered tracts. The map (see Figure, for example) illustrates this problem well; look at what is going on in Franklin, Hamilton, Cuyahoga, and so on. The one takeaway I would argue we could have faith in from this map is the heightened vulnerability of census tracts in Southeast Ohio.

To delve into variation within a county, let us look at the tracts in a few counties. I will start with Athens, just to see what is going on in my backyard. Lighter fill-colors indicate higher resiliency; please keep this in mind. Now, if you zoom in a bit and click on the map you should see the tract names and resiliency estimate popup. Look at Census Tracts 9730; is the east side of town where most of your faculty, staff, and other well to-do residents live. Note the reddish tract, Tract 9739.02; these are the less well to-do parts of town. The larger tracts are out in the country, as they say, and sparsely populated, with a mix of everything from mobile homes to large several acre residential properties.

Okay, now that we have had a feel how these maps operate, why not look at tracts in each of the six largest counties in our state. If you are familiar with some of the places and neighborhoods in these counties, zoom in and explore the distribution of vulnerable tracts. As you see some pattern, no doubt you will wonder how race and ethnicity factor into this story. That is, are the more vulnerable tracts also the ones more likely to be home to larger proportions of minority Ohioans?







  1. These are small-area estimates derived from the 2019 American Community Survey amd the Population Estimates Program. See here for details↩︎

  2. Ohio has 39.6% of its population with no risk factors, 39.3% with one or two risk factors, and 21.1% with three or more risk factors.↩︎


Text and figures are licensed under Creative Commons Attribution CC BY-SA 4.0. The figures that have been reused from other sources don't fall under this license and can be recognized by a note in their caption: "Figure from ...".


For attribution, please cite this work as

Ruhil (2022, Feb. 12). From an Attican Hollow: Measuring Community Resilience. Retrieved from

BibTeX citation

  author = {Ruhil, Ani},
  title = {From an Attican Hollow: Measuring Community Resilience},
  url = {},
  year = {2022}