We recently received a request from a resident in Lincoln County, asking for assistance in locating data related to digital inclusion, in order to help enroll residents who qualify for the FCC Emergency Broadband Benefit.
Specifically, the reader asked us to determine the following:
To answer the inquiry above, we need to compile information about:
All of this is information collected by the U.S. Census Bureau’s American Community Survey (ACS) and the summary tables provide details on household computer availability and internet subscriptions, poverty status at various levels of the federal poverty line (FPL), and SNAP receipt. What the tables don’t provide is the intersection of these characteristics. To answer this, we need the microdata or individual record data.
Here’s how we went about answering this question for Lincoln County, and how you can find the same data for your county.
The ACS microdata is published for geographic areas called Public Use Microdata Areas or PUMAs. Each PUMA is about 100,000 residents, which means some counties are combined with other counties to make up a PUMA. This is the current list of PUMAs in North Carolina.
Lincoln County had just over 78,000 residents in 2010, which means it’s combined with the eastern part of Cleveland County to form PUMA 2700.
IPUMS—the Integrated Public Use Microdata Series—is my favorite data tool. Located at the Minnesota Population Center at the University of Minnesota, the IPUMS projects do the hard work of cleaning and harmonizing data sets over time. They also create some of the most common indicators—such as poverty status based on family size and income—from the underlying data, allowing researchers and journalists to spend less time cleaning data and more time analyzing and evaluating relationships.
It’s free to use (apply here for ACS data), and they have online analytical tools for their most popular databases: IPUMS-USA (Census and ACS data) and IPUMS-CPS (for the March Annual Social and Economic supplement to the Current Population Survey). This means that you can do many data analyses without a statistical software package.
To conduct our analysis, we first need to identify the variables we need. Some of these are household variables, meaning they have the same values for all people in the housing unit. Others are person variables, meaning they can vary by person-to-person within the household.
Here’s what we need:
I want the most current data available, but because I’m trying to look at a very specific subpopulation within one PUMA, I also want to make sure that my sample is large enough for me to get reliable estimates. Because of this, I would typically choose the 5-Year ACS data instead of the 1-Year data for these types of evaluations. Unfortunately, the computer and internet questions changed in 2016, which means that I can’t use the 2015-2019 5-Year ACS data and expect it to contain all the variables I need. Because of this, I’m going to combine data from multiple sample years; this process is called “pooling.” To make the analysis run faster, I’m limiting my selection to just the ACS data. Note: I could also choose my sample when I download the data and read it into the software package of my choice (Stata) or use ipumsr to read data into R, but I’m focusing here on using the online tool.
Often the specific topic that I want doesn’t quite exist in the variables that are available. Instead, I need to combine the data together to create a new variable that captures what I am interested in. Whenever it’s straightforward, I err on the side of providing more detail: we can always add categories together later; disaggregating them has to be done in the variable creation phase.
Here, we are specifically interested in the intersection of computer availability and internet use. To do this, I need the variables CIHISPEED and CILAPTOP described above. First, I click on the “Create Variables” button linked in the top of the online tool.
Next, I use the expression syntax help and information on the variable codes to create a new variable. Specifically, I’m creating a new variable named hhintcomp (Household Internet and Computer) that contains the following four values:
The analysis is a multi-step process, consisting of the following:
If we go back to the initial question, it was about the digital divide in Lincoln County and whether we could say anything about the digital divide in households below 125% of the federal poverty line and receiving SNAP funds. Looking at our tables, we can see that:
Using these numbers, we can calculate that there are:
Because our PUMA contains Lincoln County and part of Cleveland County, I can’t use it to directly estimate the number of households that meet certain conditions in Lincoln County. Instead, I first need to get the number of occupied housing units in Lincoln County. I can do this by looking at the 2020 Census data at https://data.census.gov. According to Table H1, there are 34,306 occupied housing units in Lincoln.
By combining my percentages meeting the conditions of interest (step 7) and total housing units (step 8), I can get an estimate of the number of households meeting the specified conditions:
With those numbers established, we can then estimate the impact of distributing a given number of devices on closing the digital divide.
Need help understanding population change and its impacts on your community or business? Carolina Demography offers demographic research tailored to your needs.
Contact us today at email@example.com for a free initial consultation.Contact Us
Categories: Carolina Demographics, Story Recipe
The Center for Women’s Health Research (CWHR) at the University of North Carolina School of Medicine released the 12th edition of our North Carolina Women’s Health Report Card on May 9, 2022. This document is a progress report on the…
Dr. Krista Perreira is a health economist who studies disparities in health, education, and economic well-being. In collaboration with the Urban Institute, she recently co-led a study funded by the Kate B. Reynolds Foundation to study barriers to access to…
Our material helped the NC Local News Lab Fund better understand and then prioritize their funding to better serve existing and future grant recipients in North Carolina. The North Carolina Local News Lab Fund was established in 2017 to strengthen…
Your support is critical to our mission of measuring, understanding, and predicting population change and its impact. Donate to Carolina Demography today.