While attending the Consumer Electronics Show last week, FCC Chairman Julius Genachowski observed in his speech to the assembled technology glitterati that “… virtually every new product on the CES floor is fueled by broadband Internet—by connectivity and bandwidth, wired and wireless. If you shut off the Internet, virtually nothing on the CES floor would work.” Certainly, the rapid innovation in edge devices is a wonderful thing. But, such innovation may not be traveling alone. That is, economic theory suggests this rapid increase in the number and sophistication of edge devices may be accompanied by an increase in the intensity of price competition among broadband providers and, as a consequence, possibly a reduction in the number broadband service providers. In fact, these changes may bring lower prices and a more concentrated industry structure. You may be thinking: “how can lower prices be the result of higher industry concentration?” This is a sensible question, so let me explain how it works.
While the steady flow of great new inventions make broadband more valuable, the new devices, and/or the consumers using them, tend to be “network agnostic.” That is, modern consumer devices and services function with the Internet connections that can be provided by many providers and technologies, so that consumers do not need a connection from any one particular broadband provider or any one particular broadband technology. Tablet computers, for example, often connect to the Internet using both 3G (provided by a variety of wireless carriers) and wi-fi technologies. In fact, in the case of Amazon’s Kindle, the customer may not even know who the broadband provider is, since the connection is embedded in the device itself. It is Amazon, not the consumer, which chooses the broadband provider. Consequently, as consumers flock to new devices and over-the-top services, attachment to a particular Broadband Service Provider (BSP) is becoming a less important feature of relevance to the consumer. This “commoditization” of broadband services is expected to intensify price competition, thus reducing prices and, in turn, profit margins. The implications of margin shrinkage may be significant, as we detailed in our paper, Competition After Unbundling: Entry, Industry Structure and Convergence—a paper cited frequently by the Commission as an appropriate theoretical tool with which to evaluate the telecommunications industry. This device-driven commoditization is the focus of our recently released paper, Shocks to the Broadband Ecosystem: Implications for Competition and Market Structure.
The mechanism by which device innovation alters industry structure can be described loosely as follows. New innovations in devices and services make consumers more network agnostic. To the economist, this means that the perceived and/or actual differentiation among broadband networks and technologies. Thus, price competition intensifies and profit margins shrink (costs constant). But, providing broadband service requires large capital expenditures, and firms must charge prices in excess of marginal costs in order to pay these fixed expenditures. As margins shrink, the incentive and ability to expend capital and support existing capital shrinks. As a result, falling margins may alter industry structure by reducing the BSPs’ willingness to invest in their networks and possibly reducing the number of the firms that can profitable serve the market. So, while device innovation is welcome, it has consequences, and one such consequence could be fewer BSPs offering the underlying broadband service. This fewness, however, is a consequence of falling prices and, in turn, falling profit margins.
There are a couple interesting and policy-relevant economic concepts at work here. First, while we typically think of prices falling as the number of firms increases (that is, we typically assume Cournot Competition in quantities), in the presence of fixed costs, falling prices (from the intensification of price competition) may lead to a decrease in the number of firms. In other words, market structure and prices are in a feedback loop—more firms may mean lower prices, but lower prices may be fewer firms. A key and policy-relevant corollary of this feedback loop is that the Hirschman Herfindahl Index of industry concentration (or HHI) provides very little information on the extent of price competition in broadband markets (or any market with high fixed costs)—a high HHI could as legitimately indicate intense price competition (via the feedback loop) as it could weak price competition (via the Cournot assumption). As we show in Shocks to the Broadband Ecosystem, device innovation may lead to fewer broadband providers yet lower prices at the same time. Plainly, competition in broadband markets may be more complex than the Merger Guidelines approach can handle.
Second, while we consider the role of innovation on price competition and industry structure, the process at work is not so limited. Increased commoditization of broadband service by any means may render a more concentrated equilibrium. For example, the FCC’s net neutrality rules deliberately seek to commoditize broadband networks and shrink the profits of network providers. As a consequence, such rules may actually lower welfare in the long-run if firms abandon markets and curb investments in the face of shrinking profits. As the Commission itself observed in the National Broadband Plan, if “expected returns to telephone companies do not justify fiber upgrades, then users may face higher prices, fewer choices and less innovation.” Commoditization, whether by regulatory fiat (e.g., the FCC’s Open Internet Rules) or device innovation, should be expected to reduce expected returns and, in turn, lead to the very consequences we predict in our research. With regards to the effect of regulation, we address this issue more fully in our paper Network Neutrality and Industry Structure.
So, as the glamour of Las Vegas begins to fade and the day-to-day reality of the market sets in, we note that the pursuit of one type of benefit may bring with it another type of cost. There are no free lunches. Device innovation may alter industry structure in ways some find undesirable, but this ill feeling may be misplaced. Rising concentration may the result of intensifying price competition. The structural consequence of innovation is not necessarily bad for consumers, since prices may actually fall as industry concentration rises.