�������

It’s a problem that U.S. policymakers have grappled with since the dawn of the telephone. As modern communications transform the way we live, not everyone benefits.

Take the internet. You can go almost anywhere in the country and get online thanks to cable, satellite, DSL and other services. And yet, according to the Pew Research Center, one in 10 Americans didn’t have service at home last year.

Just like they did with the telephone, policymakers have zeroed in on pricing as the answer to narrowing the digital divide — not just as a matter of equity, but also to help lift Americans out of poverty. If rates are low enough, the thinking goes, cash-strapped consumers will subscribe and tap into job training and other potentially poverty-reducing services now available to them.

As simple as the fix sounds, it raises some important questions that Greg Rosston, the Gordon Cain Senior Fellow at the Stanford Institute for Economic Policy Research (SIEPR), is answering: Are low-cost internet services enough to drive demand among low-income subscribers? Do they spur enrollment in online courses or applications for jobs? For providers, how do low-rate plans impact their bottom line? And when they offer cheaper service to one set of customers, do they raise prices for others to offset their costs?

To date, evidence that would shed light on these issues has been scant. That’s largely because it wasn’t until 2011 that an internet provider first offered cheaper broadband plans to low-income households and only a small number of companies do so today.

In a new working paper, Rosston finds that offering cheap internet service is a powerful tool for encouraging low-income customers to subscribe. At the same time, he and his co-author Scott Wallsten — president of the Technology Policy Institute — conclude that pricing as a strategy for closing the digital divide isn’t enough.

“The glass half-full view is that low-cost service can make a big difference,” says Rosston, who is also the director of Stanford’s Public Policy program. “But the number of low-income households with internet access still lags far behind higher-income households with service. If the goal is to bridge that gap, pricing alone won’t do it and we don’t yet know what will.”

Rosston also analyzed whether these new subscribers then took advantage of online education and other poverty-alleviating services.

“People don’t see the benefits of these programs,” Rosston says. Wallsten has shown in  that people are willing to pay money not to take job training courses online. According to Wallsten’s study, it’s possible that people don’t have the time for these classes or are skeptical about their usefulness. His research also found that some people don’t subscribe to internet service even with significant discounts on price.

Subscriptions rise, but…

In their study, Rosston and Wallsten looked at the first five years of a groundbreaking program that Comcast launched in 2011 as a regulatory condition of its merger with NBC Universal. Because other internet service providers didn’t roll out similar plans until 2016, the researchers analyzed data on Comcast’s program between 2011 and 2015 in order to conduct a clean experiment.

Under the program, called Internet Essentials, households with school-age children eligible for free or reduced-price school lunches were offered $10-a-month broadband service, computers for $150 and training on core functions like web browsing and email.

The scholars, using data from the U.S. Census and the Federal Communications Commission, calculate that 64 percent of low-income households in areas served by Comcast had internet service in 2011. Five years later, 74 percent of them did.

While the 10 percent jump is significant, says Rosston, not all of it can be attributed to Internet Essentials. That’s because some of the new customers would likely have subscribed even without the promotion, while others already had service through another provider but switched to Comcast because of the low rate.

Rosston and Wallsten conclude that the program netted slightly more than 292,000 new internet subscribers, or about 7 percent of the 10 percent increase in internet-connected households.

Rosston says the fact that a significant number of eligible households didn’t subscribe points to a larger challenge U.S. policymakers face in making home internet access ubiquitous: It may not be possible.

“Lowering the monthly cost to $10 still didn’t get everybody online,” Rosston says. “We don’t yet know how to get these people online — and it’s not even clear that they want it or would benefit from it.”

Mobile phones as gamechanger

Rosston cautions that the study has some limitations. Comcast does not provide service throughout the country and its markets tend to be mostly urban and suburban. It’s not known, either, how similar plans that AT&T and Charter Communications made available after the 2015 study period are impacting subscriptions rates or pricing for non-program customers.

What’s more, Rosston points out that market trends could ultimately bring about the solution policymakers have been looking for.

“Since 2015 there have been huge changes in the way people use wireless phones to access the internet and I expect that will continue over the next five years,” Rosston says. “Since most low-income households have wireless phones, these changes will hopefully make internet accessibility even easier for them.”

Krysten Crawford is a freelance writer.