For the large numbers of families now essentially confined to their homes, having a reliable broadband connection is more important than ever. A large income-based “digital divide” still exists, with more than 40 percent of adults with annual incomes below $30,000 lacking a home broadband connection.
Senator Ed Markey introduced a bill with real potential to mitigate the digital divide. Most proposals simply call for more money for existing programs or for new programs without evidence they will help. Real-world experience, however, has demonstrated how little we truly understand about why many low-income people do not subscribe.
The Markey bill tackles this underlying issue. It sets the stage to run experiments, gather data and undertake rigorous analysis to learn why these gaps exist, what expectations are realistic, how to make real progress in closing the divide, and how to get the biggest bang from subsidy bucks.
Well-meaning policymakers and advocates typically say that the biggest issue keeping low-income people from subscribing is the cost. Surveys consistently list that as one of the top two reasons. If that were so, the solution would be obvious: increase subsidies. And, to be sure, lower prices for low-income households would encourage additional adoption. As economists, far be it from us to argue that demand curves do not typically slope downwards.
But evidence from people’s behavior, as opposed to survey responses, has shown that the price of broadband service is not the primary factor that keeps many low-income households from subscribing.
In 2011, Comcast launched Internet Essentials. This program offered home broadband at $9.95 per month initially to low-income households with school-age kids who did not already have broadband. It was thus targeted at a group likely to value broadband but did not yet have it—a direct attempt to mitigate the “homework gap,” as FCC Commissioner Rosenworcel has called it. Since then, the program has expanded to other groups and other internet service providers have introduced similar programs.
In recent research, we found that Internet Essentials successfully signed up large numbers of households who likely would not have purchased service without the program. But we also found that by 2015, a large share of eligible households in Comcast’s territory still did not have service. Reducing the price to $9.95 per month was not able to narrow the gap as much as society would have hoped.
One response is that perhaps $9.95 per month is still too expensive for many households, either because they cannot afford it or because they do not place that high a value on broadband.
But it turns out that lowering the price further does not encourage much additional subscribership.
In 2013, the FCC conducted 14 experiments with wireline and wireless broadband companies to learn what strategies might encourage low-income people to subscribe. The experiments tested several combinations of price, speed, access to digital literacy classes, and subsidized hardware. The program came up with a surprising result: Only about ten percent of the expected number of households signed up, even with the price of one plan set at $1.99 per month.
These experiences show why Senator Markey’s focus on analysis is so important.
We need to know more in order to create programs likely to encourage low-income households to subscribe and now is the time to learn it.
First, the FCC should be collecting data about various approaches being tried to bridge the gap during the pandemic and make that information available to researchers. For example, schools and communities across the country are rolling out different approaches to help kids access distance learning, and ISPs are making their WiFi signals available for free. The thousands of ongoing experiments will provide invaluable data to help us learn which approaches are effective.
Second, Congress should make money available to plan and conduct more experiments of the type the FCC did in 2013 once the pandemic has subsided. Such experiments would complement the data collected about actions during the pandemic because demand is different in more normal times. We need information from non-pandemic life in order to develop programs that remain effective over time. Such experiments are not expensive. In 2013, the FCC spent only $25 million on these experiments, which is tiny compared to the billions of dollars we already spend, and the billions more we may spend, on bridging the digital divide.
Third, we must be willing to learn from the results. This may sound like a no-brainer, but the FCC’s 2015 Lifeline reform order did not once mention the experiments conducted in 2013. Ignoring those results was a lost opportunity to make Lifeline more effective. Let’s not lose that opportunity again.
Bridging the digital divide is important. Achieving that goal requires more than additional subsidies. It means taking a clear-eyed look at what does and does not work and being willing to accept that certain beliefs about how to close the divide may be wrong while simultaneously being willing to consider ideas that have not yet been tried.
