The Federal Communications Commission (FCC) recently proposed amending its low-income “Lifeline” program—which provides a $9.25 per month credit for consumers of voice services—to permit recipients to apply that same subsidy instead to broadband services. Who could argue against increasing options for low-income Americans?
Before critiquing the FCC’s proposal, it’s important to point out that expanding broadband access is a laudable goal. But financing this expansion through the Lifeline program will eventually lead to the perverse outcome of taxing broadband in order to subsidize it. Better to raise the funds for subsidized broadband from taxes imposed on behavior we want to discourage.
To an economist, a subsidy (or a tax) is warranted only in the presence of a market failure. When the market produces too much a product—think driving—it’s because producers are not internalizing a negative externality (traffic or air pollution). When the market produces too little—think general (as opposed to applied) research and development (R&D)—it’s because producers are not internalizing a positive externality or spillover.
This understanding leads to a simple policy prescription: Tax the industries that produce negative externalities and subsidize those that produce positive spillovers. Yet our politicians won’t support a gas tax to finance our crumbling roads, reflecting their constituents’ myopic desires, even if the result runs counter to economic theory.
Broadband is a classic case of positive spillovers in that every person who joins the network makes the network more valuable for existing users and for application providers. In addition to tapping into those positive spillovers, a broadband subsidy could stimulate more broadband investment: If a broadband provider needs a 30 percent take rate to deploy fiber to a neighborhood, and if a broadband subsidy gives it assurance that that target will be exceeded, the neighborhood has a better chance of being deployed.
Now back to the FCC’s Lifeline proposal. Lifeline is currently funded by a “universal service fee” that shows up on your phone bill for services that are designated as interstate (as opposed to intrastate). The FCC imposes a fee on providers of these voice services, who in turn pass that fee onto their customers. Roughly half of the funds that flow to low-income residential users are raised on the backs of businesses, creating a cross-subsidy of sorts. The FCC proposes to leave the funding alone (for now), but to give Lifeline recipients the option to apply the existing subsidy to broadband instead of voice service.
By my calculations (produced below), to induce non-adopting Americans to share in the costs of broadband, the annual subsidy would cost between $1.1 billion (for a modest addition of 10 million of the 32 million disconnected homes) and $4.3 billion (for 20 million homes, leaving just 12 million disconnected). The immediate problem is that a large chunk of this cost estimate does not fit within the contours of the existing Lifeline budget, which stood at $1.7 billion in 2014.
How did I arrive at these cost estimates? A 2014 study by three FCC economists estimates that up to 10 million disconnected homes would be willing to subscribe to broadband if a subsidy of 15 percent were offered. The annualized cost of connecting the first 10 million disconnected homes would be $1.1 billion (equal to 10M x 15% x $60 per month x 12 months). Because the next tranche of non-adopters are less inclined to adopt, a larger subsidy would be required to reduce the disconnected share further. To the extent that 20 million homes could be induced to adopt broadband in response to a 30 percent subsidy, the subsidy would cost $4.3 billion per year (equal to 20M x 30% x $60 per month x 12 months).
Telling Lifeline-enrolled families that already purchase a bundled voice and broadband service that they can apply their existing $9.25 per month subsidy to broadband rather than voice is not going to reduce the number of disconnected broadband households (nor would it make their lives any better). And Lifeline-enrolled families that didn’t have broadband because they purchase voice on a standalone basis would be forced to lose their voice subsidy if they applied the subsidy to broadband instead (making their lives only slightly better). Tapping the existing base of Lifeline funds just won’t make a big difference when it comes to shrinking the digital divide.
And therein lies the problem. The current base of revenues—interstate voice services—are under siege as consumers increasingly obtain voice service as a free add-on to a wireless broadband data package. To raise the funds to make a real dent in the number of disconnected homes and improve lives, the Lifeline revenue base likely would have to be expanded to include broadband services.
As unelected officials, the FCC Commissioners would be happy to oblige. Some activists are practically begging the FCC to tax broadband to preserve the Universal Service program. And the FCC is not bashful about taxing and spending: In 2014, the FCC expanded the E-Rate program by $1.6 billion to give schools and libraries greater access to broadband.
But for the same reason we would never finance a general R&D subsidy by taxing firms engaged in general R&D, it makes no sense to tax broadband in order to subsidize it. Indeed, for those 10 to 20 million non-adopting households that would come aboard in response to a modest subsidy, there are likely millions of price-sensitive broadband households that would leave the broadband market in response to a modest tax. Unlike interstate voice revenues, which are paid in part by businesses, fixed broadband revenues are overwhelmingly paid by residential consumers. Thus, bringing broadband into the revenue base would cause U.S. households to bear a larger burden of the universal service subsidy.
Even worse, as soon as the FCC imposes a federal universal service fee on broadband to meet the surging demand for broadband among low-income Americans, the states are free to get in on the action. By reclassifying broadband as a “telecommunications service” in its February Open Internet Order, the FCC activated a series of dormant state and local telecom-based fees that had never been extended to broadband; Internet service was previously designated as an “information service,” and thereby immunized from this form of state taxation.
Recognizing this risk, the FCC preempted states from moving forward with their own universal service fees for broadband until the FCC adopted fees at the federal level: “[W]e preempt any state from imposing any new state USF contributions on broadband—at least until the Commission rules on whether to provide for such contributions.” The combined universal service fees from the FCC and the states would perversely contribute to the digital divide by driving even more price-sensitive adopters out of the broadband market.
To avoid this spiral, we need to look elsewhere for the financing of a broadband subsidy. Were it designed by an economist, the subsidy would be financed through the general treasury so as to reduce any distortions in the broadband marketplace. And the source of the funds would be the elimination of existing subsidies for sugar, corn, coal, or oil—all of which generate negative externalities.
Raising taxes on broadband users in order to subsidize broadband makes no economic sense.
This is cross posted from Forbes.