On Thursday, November 16, the Federal Communications Commission (FCC) is set to vote on a proposal that would adopt several reforms to the Universal Service Lifeline program, which subsidizes broadband and telephone service for low-income consumers. The program, expanded to include broadband Internet in 2016, distributes upwards of $1.5 billion annually in the form of subsidies.
Major reforms on the table include:
- Adopting a “self-enforcing” budget that makes it more difficult to spend more than the rules allow;
- Limiting subsidy payments to facilities-based providers;
- Reauthorizing State Commissions to delegate Lifeline ETCs; and
- Limiting lifetime Lifeline benefits.
The FCC contends these proposed reforms will “more effectively and efficiently help close the digital divide by directing Lifeline funds to the areas where they are most needed.” Opponents, however, believe the proposed changes “will gut the program and continue to widen the digital divide.”
The likely outcome, if the proposal is enacted as currently written, will be somewhere in between. Some of these proposed reforms are important, positive steps that will improve the Lifeline program’s efficiency. Some are unlikely to lead to any benefits and may yield net costs. Most importantly, the proposed reforms fail to acknowledge key problems with the Lifeline program: First, we still do not truly understand why many low-income people do not subscribe to broadband, as the Commission’s own experiments demonstrated, although studies point to lack of interest or perceived relevance followed by price as key determinants. Second, the program provides no mechanism for evaluating whether it is actually helping to close the digital divide.
Here is my take on these proposals:
The Commission takes an important step forward by proposing a hard (or “self-enforcing”) budget on Lifeline, noting that it is important to keep spending “at a responsible level.” Keeping spending in check is important, but a budget also encourages other beneficial behavior. A strict budget encourages efficiency in ways beyond strict money management. Resource constraints create pressure to think creatively about solutions, and to get the biggest bang for the buck in a way that does not exist if more money is always available. Furthermore, it creates incentives to reduce waste, fraud, and abuse, which is an ongoing challenge. Keeping spending in check is particularly important given that the Lifeline program is funded through a regressive tax that falls heavily on the poor.
This proposal does not get an “A” because it leaves the budget at $2.25 billion per year, which is much higher than the $1.5 billion spent in 2016, and therefore not a binding constraint. As such, it will not create the benefits that it should.
Limiting payments only to facilities-based providers: C
Where a strict budget is likely to yield net benefits, the FCC’s proposal to limit subsidy payments only to facilities-based providers is unlikely to increase network investment, as the Commission hopes it would. A long debate has focused on the effects of “unbundling” regulations—rules that require facilities-based providers to give wholesale network access to resellers. Much of the research in this area, including my own, finds that such rules reduce network investment. It therefore seems logical to conclude that directing Lifeline to facilities-based carriers would likewise encourage investment on the margin.
That conclusion is unlikely to hold in the case of Lifeline. The aspect of unbundling that reduces investment incentives is not selling wholesale services to other companies, per se. The problem is that the government mandates such behavior at regulated rates. Because the government has no reliable way to set regulated rates equal to market-based rates, the rules end up distorting and, generally, reducing investment.
On their own, however, resellers offering Lifeline service have reached market-based commercial agreements with facilities-based providers. In principle, some of these resellers are able to reach certain populations—through marketing or specialized services, for example—that aren’t worth the effort for facilities-based providers. Given that resellers are market driven, conditional on the existence of Lifeline, restricting subsidies to facilities-based providers is unlikely to have much, if any, effect on investment incentives. In fact, it is conceivable that this could worsen the digital divide by removing from the program some of the providers who explicitly seek out those who would not otherwise have service.
Returning responsibility to the states: Legal: NA | Economics: C+
The FCC presents a legal argument as to why it should reverse the 2016 creation of a new category of federally-designated Lifeline providers, which effectively preempted state authority to designate eligible providers of Lifeline service. As an economist, I have no insight into the legality of the prior arrangement aside from general principles of federalism. However, program administration is the kind of activity that benefits from economies of scale, and state boundaries are generally unrelated to the underlying economics of telecommunications. In other words, it would seem to make more sense to reduce the number of entities with certification or other management responsibilities than to increase it. While there are reasons why not all programs should be centrally administered, the proposal presents no economic argument as to why that should be the case with Lifeline.
Limit lifetime benefits: Incomplete
The FCC proposes imposing lifetime limits on the total benefits any given household can receive, and asks for input into many potential aspects of such a limit. As with any proposal related to Lifeline, the key question at hand is what effect such a rule would have on the digital divide. The FCC reports that households remain in the Lifeline program for about 1.75 years, on average, but we know little about why they subscribed, why they left the program, and what happened when they left. More broadly, how likely are recipients to continue broadband service once they are no longer receiving the subsidy? Without further research, the answer is unclear.
Two pieces of evidence suggest that most customers would continue service and that limiting lifetime Lifeline benefits would not increase the digital divide. First, recent research has found that Lifeline has historically not done a good job at targeting people who will sign up for broadband service only if they receive a subsidy. It appears that many Lifeline recipients would have broadband service even without the program. If that is true, discontinuing the subsidy at a certain point should not cause these recipients to disconnect. Second, survey research finds that people are more likely to value broadband service once they have experience with it. Again, to the extent that this is true, it reduces the chances that a Lifeline recipient would disconnect service even if the subsidy induced them to sign up in the first place.
On the other hand, some Lifeline recipients cannot afford unsubsidized broadband service or do not value it sufficiently to purchase it at market prices. These people would drop their service without the subsidy.
Before it moves forward with this reform, the FCC should conduct experiments to study how many people are likely to cut their broadband connections under a limited Lifeline benefit. Ideally, this would involve experiments exploring this question. Such experiments might reveal that subsidies are required for only a short period of time and that, once exposed to broadband, a household develops a strong willingness to pay for the service. In that situation, imposing a benefit limit could significantly reduce the cost of the Lifeline program with little negative effect on the digital divide. On the other hand, experiments might reveal that a large share of participating households do drop their service without the subsidy. In that case, the benefits of lower costs must be balanced with the social cost of reducing connectivity.
Real Reforms: Waiting for Godot
The use of experiments brings us to the bigger question of what real Universal Service reforms would look like.
First, they would include the Connect America Fund (CAF) in addition to Lifeline. CAF suffers from the same problems and more—it is three times the size of Lifeline, has had little to no effect on telecommunications adoption, and subsidies flow to wealthy and poor rural residents alike.
With respect to Lifeline, though, in short, we don’t really know what changes will better target the digital divide. We know that the program is not doing a good job of increasing the number of low-income people who are connected, and the program makes no effort beyond eligibility checks to ensure that recipients would subscribe only if they receive a subsidy.
Research by Olga Ukhaneva at Georgetown found that a large share of telephone Lifeline recipients would have had phone service even without the subsidy, and the broadband reforms did nothing to try to target those people who truly need a subsidy in order to subscribe. In other words, much of the support is going to people who would have service anyway.
In addition, we still know very little about getting people online. The FCC conducted 14 experiments in 2012 to explore the effects of different subsidy models on subscribership by low-income people who never had broadband service. Overall, participation in these subsidy programs was only about ten percent as much as providers had expected. More specifically, the results indicated that it was difficult to encourage those low-income people who remain who do not subscribe to connect. These experiments confirmed that we do not yet have a good understanding of what keeps the remaining people offline.
Answers to two crucial research questions might radically improve Lifeline’s efficiency. First, how many people have service specifically because of Lifeline? Second, how do we get the remaining unconnected low-income people online? Until the FCC takes these questions seriously then it will not be possible to determine whether Lifeline has much, if any, effect on the digital divide. Until then, reforms are just window dressing.