HDG #007: Why you must look deeper than county level
Read time: 5 minutes
Today I’m talking about why we need to analyze wayyyyyyyy…yyyy beyond the county level if we have any hope of moving the needle on any health, healthcare, or community initiatives—plus a list of resources to get you started!
What do I mean? Deeper than county-level data, baby!
Last week I wrote about how to do a community assessment in minutes and listed 16 different data websites to use. And while those were great resources, today we’re going to go deeper—and today that is to the census tract level (in this case).
Why is this now a non-negotiable?
Aside from the fact that we’ve been looking at the county level for… forever? Analyzing county-level data is extremely valuable—don’t get me wrong. But IMHO, we have all but exhausted the utility of county-level analysis.
The differences between neighborhoods within counties can be like night and day. And that is the only way you’re going to fish out disparities so you can address them.
New strategies that have any hope of changing the status quo must be as unique as the neighborhoods, communities, and regions that they are meant to serve.
They must be hyper-local, targeted, and specific.
And for that, we need more than county-level data.
. . .
Let’s consider my city of Albuquerque, NM.
In general, we have 4 quadrants divided by the 2 major highways that run through the city. Albuquerque is sorta known for having pockets of wealth or poverty sprinkled throughout the city, sometimes within close proximity to one another. Anecdotally, locals talk about areas of town like this:
Northeast (upper upper right): most affluent, more Caucasian
Southeast (lower right): near the airport, industrial, pockets of affluence mixed with poverty
Northwest (upper left): upper middle-class
Southwest (lower left): the most generally lower-household income areas of town, more Hispanic
I used PolicyMap to visualize this on the map using census tract data (roughly equates to a neighborhood). The darker the purple, the more money people tend to make in that neighborhood.
The map generally mirrors my colloquialisms. However, what I did not know until I looked at this in detail was exactly where all of these random pockets were sprinkled specifically throughout the city, I only knew of some. And—in even more cases than I realized—they’re a mere one block away from one another (or even “divided” by a single street).
But one area particularly stood out to me (where the cursor hovers in the image above).
A noticeable, randomly placed dark spot (more affluent) surrounded by a bunch of lighter spots (average-to-lower-income neighborhoods). It made me think about how some kids might fare at their local schools if those schools tended to have more affluent kids, or vice versa.
. . .
Then I looked at something related to health: PolicyMap has its own COVID risk score, which represents the relative risk for residents in a given area to develop serious health complications from COVID-19 because of underlying health conditions identified by the CDC as contributing to a person's risk of developing severe symptoms from the virus.
The darker the purple, the higher the risk of severe COVID complications for residents.
See if you notice anything interesting:
Luckily, MOST of Albuquerque was in the “below average” category, with only 3 neighborhoods in all of metro Albuquerque standing out as noticeable outliers.
But that same neighborhood that looked well-off among a sea of not-as-well-off households also has a noticeably higher risk of severe COVID complications than most of Albuquerque. Your eye is drawn right to them because of the difference, in fact.
What gives?
I don’t actually know, but I’m certainly going to keep digging. I thought it could be a pocket of elderly homeowners who had more money but also more chronic conditions and COVID risk, but another layer showing Median age by tract rebuffed that hypothesis.
But something unique to that neighborhood might be going on there. And data at this level can give us targeted directional pointers for things like, and so many other things from which we can determine where to focus outreach, policy, research, funding, and so much more.
We would not see anything remotely like this by looking at the county alone.
This is just one reason of many why looking at the data beyond just the county level is no longer an option.
. . .
So why haven’t we been doing it?
It’s hard to find reliable data at this level.
It can be potentially misleading (there could actually not be anything unique or perplexing about that neighborhood I pointed out; it could have just been circumstantial evidence based on where the range limits were drawn for coloring the heatmap).
It takes a lot longer.
It reveals SO MUCH data that it can get overwhelming very quickly.
The sheer volume of data and micro-indicators can get unwieldy pretty quickly.
We don’t know what to do with it or we’re not at a place we can target at this level, and county suffices for now.
This list goes on and on.
These are all legitimate reasons!
But if we really want to start moving the needle, they are going to be reasons that we have to overcome.
. . .
Actionable Idea of the Week:
Push the geographical limits of your next analysis.
Here are some data websites that have data at the zip code or census tract level that allow you to compare different neighborhoods like I did today:
These all work best for non-rural areas, as you can imagine. More on that another week.
. . .
How can we get more data at this level? Hit reply and let me know your thoughts.
See you next week!
-Stefany
2 more ways I can help you:
1. If you want to learn more about health data quickly so you can market yourself, your company, or just plain level up your health data game, I'd recommend checking out my free Guides. Courses and more resources are coming soon, so check back often.
2. Book some consulting time to chat with me 1:1 about your data questions, project or product ideas, and to get expert technical data advisory.