The Social Vulnerability Index and American Community Housing Survey data
One of the most common visualizations constructed by organizations working with city data is demographics. Especially for jurisdictions looking to adequately convey the challenges involved in providing service to their community, demographic data such as census data often aids presentation tremendously. Unfortunately, getting to that data typically requires someone well versed in database management to import it locally and someone experienced with GIS to display it. Additionally, the most commonly used resource for this type of demographic data is the American Community Survey (ACS), and processing its fields to provide views of the data relevant to first responders can be a time-consuming and costly process.
Fortunately, the Centers for Disease Control (CDC) has already done this work on ACS data in order to create what they call the Social Vulnerability Index (SVI). From the CDC’s SVI Fact Sheet:
What is social vulnerability? Every community must prepare for and respond to hazardous events, whether a natural disaster like a tornado or disease outbreak, or a human-made event such as a harmful chemical spill. A number of factors, including poverty, lack of access to transportation, and crowded housing may weaken a community’s ability to prevent human suffering and financial loss in a disaster. These factors are known as social vulnerability
What is CDC’s Social Vulnerability Index? ATSDR’s Geospatial Research, Analysis & Services Program (GRASP) created databases to help emergency response planners and public health officials identify and map communities that will most likely need support before, during, and after a hazardous event. CDC’s SVI uses U.S. Census data to determine the social vulnerability of every census tract. Census tracts are subdivisions of counties for which the Census collects statistical data. The SVI ranks each tract on 15 social factors, including poverty, lack of vehicle access, and crowded housing, and groups them into four related themes. Maps of the four themes are shown in the figure below. Each tract receives a separate ranking for each of the four themes, as well as an overall ranking.
So the data available from the SVI specifically targets those factors that put communities at risk for hazardous events, and two of the most common hazardous events in communities, as fire departments can well attest to, is EMS events and structure fires.
Thus, NFORS has recently incorporated this data at the census tract level and will be adding it to incoming incidents so that fire departments will be able to easily explore the impact of demographics on their service demand and convey these insights to decision makers!
Visualizing Community Vulnerability and Call Volume
The most basic graphics used to convey intuition about how demographics affect call volume are maps, so let’s make some maps in NFORS!
First the standard response map conveying call volume is available by default
Next, we may want to view call volume on a per census tract basis since that most readily maps to our demographics. Note that we need to use the 2016 SVI FIPS code field for this, NOT the 2010 census tract data. It is a little known fact that the Census bureau slightly alters census tracts (usually by producing new ones) annually. Make sure your geography is the same vintage (same year) as your data!
Finally, for this example, we want to visually compare the number of incidents in these census tracts with their actual vulnerability, available from the social_vulnerability_2016.summary.overall field.
SVI Themes
Above, we looked at the overall SVI. SVI itself is a summary metric calculated from four themes, which are groupings of underlying ACS variables that have been identified in literature as contributing to community vulnerability.
SVI is measured as the percentile ranking of a census tract with respect to these themes. Thus, it is on a scale of 0 to 1 with 0 representing the least vulnerable (demographically) census tract and 1 representing the most vulnerable census tract among all census tracts in the United States.
Each of the underlying 15 variables are available either as raw data, such as number of persons, or number of households, or percentages, such as percentage of population in this census tract that is 65 or older, etc.
Examining Calls and Responses using SVI
Suppose we want to find out how many of our department’s calls are originating from areas in our city that could be classified as highly vulnerable. We could graph our call volume against the overall SVI for tracts in our city, as shown below:
This graph shows us that for this jurisdiction, a large portion of their call volume is coming from the areas of their city with highly vulnerable populations. Keep in mind that this metric is slightly different from high risk, insofar as it is capturing an area’s ability to absorb the impact of disastrous events rather how often and how severely it is subject to disastrous events. Thus, by drilling down from an examination of overall SVI into the various themes, it may be possible to identify areas of the city where particular forms of community risk reduction programs, such as smoke alarm installation in homes, may help harden the population against, for example, residential fires.
Additionally, it may be fruitful for a jurisdiction to examine response performance relative to SVI. Slow response times to an area hurt fire and EMS outcomes there, but a vulnerable population will be especially affected, so it may be beneficial to confirm that service to highly vulnerable populations is not deficient.
As we can see in the above graph, this jurisdiction’s response performance to highly vulnerable areas, value 0.9 on the graph, is comparable to its response elsewhere, indicating that this department’s coverage is evenly distributed geographically. However, should a department be examining its coverage in light of resource deployment or perhaps a station study, it may be worth considering that if the area is less able to absorb the impact of disastrous events, improvements to response times there may have a stronger positive impact on incident outcomes.
Heat maps for SVI variables
In addition to the general examinations above, we can use heat maps to examine how certain combinations of ACS variables underlying the SVI are associated with call volume and response. For example, we might build the following graph examining call volume as a function of income and minority status in order to better understand what is driving call volume:
From this we can see that a strong majority of the calls in this jurisdiction are from areas with a very low per capita income and that are mostly non-white. Exactly which areas these are could be ascertained by clicking on the high volume boxes and applying the filters in order to see geographically where these calls are coming from.
Conclusion
The SVI and its underlying ACS variables at a census tract level have now been incorporated into NFORS as an enrichment to incoming incident data. These variables are very useful for helping to tie community demographics into the fire problem, especially when building stories to present to communities and to city or county officials to help them understand some of the demographics associated with the fire and EMS calls in their city. Additionally, these can be used to help identify certain areas within the community that may benefit from community outreach efforts and aid in correctly targeting those efforts.
Want to start exploring? A sample SVI dashboard, with all the visualizations discussed above, is now included in every new workspace! To start exploring with our starter SVI dashboard, create a new workspace and view the SVI Introduction.
