A Dynamic Analysis of Broadband Competition: What Concentration Numbers Fail to Capture
The 117th Congress is considering whether to devote significant federal resources toward promoting broadband access in underserved communities. Legislative proposals to do so include President Joe Biden’s draft American Jobs Plan—a $2.3 trillion budget-reconciliation package that sets aside $100 billion for broadband infrastructure. They also include the Accessible, Affordable Internet for All Act, which would create a $79.5 billion federal program.
The instinct to promote network buildout is understandable, particularly in the wake of the COVID-19 pandemic and the various socioeconomic disparities it highlighted. But precisely how that infrastructure funding is deployed will determine whether such proposals succeed or fail.
In fact, the U.S. broadband market is already healthy, and in most cases, competitive outcomes are close to optimal. Charges that broadband markets are dominated by monopolies or oligopolies and that they are therefore stagnant, over-priced, and of low quality do not comport with the empirical and economic realities. To take but one example, even with the overall rise of prices across the economy, and in the face of surging demand during the COVID-19 pandemic, U.S. broadband prices fell.
Concentration is a poor predictor of competitiveness, and broadband markets with even a small number of competitors can be—and are—quite healthy. Indeed, the multi-year, multi-billion-dollar investment plans broadband firms execute—amid constant pressure from alternative modes of Internet access like 5G, fixed wireless, and satellite—tell the story of a highly competitive, dynamic market.
To be sure, there are a few areas where there has been no meaningful wireline broadband buildout: Approximately 4.4 percent of the U.S. population does not have access to at least 25/3 Mbps fixed service. Even then, however, many of those areas are served by wireless Internet service providers (WISPs), cellular broadband, and/or satellite service.
But while the digital divide—both rural and urban—may be real, that fact alone does not justify wholesale intervention into broadband markets. Instead, the actual scope of the problem should be assessed, and policies tailored to remedy specific needs. The policies required to reach that stubborn 4.4 percent tail of broadband rollout are likely to be very different than those that facilitated the buildout of the first 95.6 percent.
Policies designed to close the digital divide should have two broad features: they should reach consumers where they are, and they should not disrupt the otherwise healthy broadband market. Reaching consumers where they are means targeting subsidies directly to consumers to make it more viable for existing providers to build out into new areas. Such policies should be technology-neutral and designed to stimulate demand to jumpstart markets that have otherwise proven too costly for any provider to enter. Avoiding disruption of healthy markets entails refraining from interventions that artificially introduce new competitors, skew investment planning by broadband providers, or dictate how and where providers should build networks.
There is much that can be done to encourage better and timelier broadband rollout, but not all solutions are equally effective. As we detail below, policymakers must choose carefully among competing options to realize the best possible result.
This paper aims to address common misconceptions associated with broadband competition that, in turn, undercut practical solutions for connecting the unconnected. It first details some of those misconceptions and contrasts them with the realities of current broadband markets. It then provides an overview of how to properly understand healthy competition in local broadband markets. It then provides a critique of commonly advanced proposals that are based on fundamental misunderstandings of how broadband markets work. And finally, it offers an approach to policy that incorporates a variety of solutions for connecting the unconnected.
Read the full white paper here.