, , , , , , , , , , , , , , , , , ,

The FCC recently voted to reverse its earlier actions on so-called net neutrality, which would have treated internet service providers (ISPs) as “common carriers” and subjected them to detailed federal regulation of their services, pricing, and profits. Many believe net neutrality would ensure a sort of fairness and nondiscrimination on the internet, but it is actually a destructive regulatory regime under which certain firms are allowed to extract economic rents from the efforts of others. Warren Meyer has a nice take on this at Coyote Blog:

Net Neutrality is one of those Orwellian words that mean exactly the opposite of what they sound like…. What [it] actually means is that certain people … want to tip the balance in this negotiation towards the content creators ….  Netflix, for example, takes a huge amount of bandwidth that costs ISP’s a lot of money to provide. But Netflix doesn’t want the ISP’s to be be able to charge for this extra bandwidth Netflix uses – Netflix wants to get all the benefit of taking up the lion’s share of ISP bandwidth investments without having to pay for it. Net Neutrality is corporate welfare for content creators.

I made the same point almost three years ago in “The Non-Neutrality of Network Hogs“. Meyer emphasizes that in the net-neutrality fight, the primary tension is between content creators and ISPs (and transport providers), but it is like any other battle to capture the gains from a vertical supply chain. Think of suppliers of goods versus shippers, for example, or traditional publishers versus delivery services, or oil extraction versus refining. Ultimately, all of the various parties must cover their costs in order to survive, and obviously each would like to capture a larger share of the value from its stage of the production process. In a series of arms-length transactions, one might assume that their shares would correspond roughly to the value they add to the final product, but things are more complicated than that. Much depends on the competitive state of the market and on the cost structures faced by different parties.

While the ISPs are often said to exercise monopoly power, there are few if any local markets in which that is actually the case, even in rural areas. Almost everywhere in the U.S., local internet markets could be better described as oligopolistic: there are at least a couple of rival firms (and alternatives for consumers), even if the technologies are sometimes radically different, so some competition exists. The same is true of the internet backbone.

Obviously, content providers compete with one another in a large sense, but many popular forms of content are unique and consumers demand access to them through their ISPs. Therefore, some content providers exercise a degree of monopoly power. And they might also require a lot of bandwidth.

The nature of the costs faced by ISPs and content providers is quite different. The latter have a much lower proportion of fixed costs than ISPs, who must invest in network capacity. Ultimately, the costs of providing that capacity must be priced. At first blush, it seems natural for users of capacity to be billed proportionately, but allocating those costs over customers and over time is a complex undertaking. Like all problems in economics, however, network usage involves a scarce resource. A large increment to demand can lead to network congestion and higher costs, not only directly to the ISPs but to users experiencing a degradation in the speed and quality of their service. ISPs have traditionally had the flexibility to negotiate with large content providers, reaching mutually agreeable terms. That’s what brought us to the state of today’s internet, and most observers would say that it’s pretty damn good!

It is the network that makes all of these wonderful services possible. The ISPs provide and maintain that network, and they must provide for expansion of that network as traffic grows. It is important that ISPs have adequate incentives to do so. However, the form of regulation to which so-called common carriers are subjected is known historically for its failure to provide good incentives. That history goes back as far as 130 years in transportation and about 80 years in telecommunications. This is why many analysts, and FCC Chairman Ajit Pai, contend that common carrier status for ISPs, and “net neutrality”, would lead to shortfalls in network capacity and a deterioration in the quality of service. It would also reward large content providers (think Netflix) in the short term at the expense of ISPs, essentially giving the former access to the existing network at less than cost. That’s the whole idea for industry advocates of net netrality, of course. But in the end, net neutrality is a shortsighted goal, even for the content providers.

The content providers have made every effort to propagandize the public, stoking fears that the ISPs are treating certain kinds of traffic unfairly. Without net neutrality, would ISPs unfairly discriminate against certain kinds of content? Or against certain types of users? Price discrimination is one of the primary criticisms of the presumed behavior of ISPs in the absence of net neutrality. Economist Bronwyn Howell points out that price discrimination is not unusual, however, and is not necessarily undesirable. Indeed, consumers of internet, telephone, mobile, and cable TV services seem to prefer certain forms of price discrimination! Consumers with heavy usage who purchase flat rate monthly internet access pay a lower charge per Gb than light users. Consumers who purchase “bundles” of internet and voice service may benefit from price discrimination relative to those who choose not to bundle their services. Strictly usage-based pricing would prevent price discrimination on this basis, but few would advocate the abolition of bundled offers, which provide benefits in terms of flexibility of use and predictability of cost, yielding net welfare gains for many consumers at no incremental cost to others. Like all voluntary trade, these are positive sum transactions: consumers capture more  “surplus” value while ISPs earn a greater contribution to the fixed costs of the network.

When ISPs charge a data rate based on usage, consumers face a positive marginal cost on incremental data. As usage increases, its marginal value to the consumer declines; the consumer will not use data beyond the point at which its value equals the data rate they pay. That places a cap on consumer surplus (the area above the price and below the consumer’s demand curve). When the consumer faces a zero marginal cost (an unlimited data plan), their usage rises to the point at which its marginal value is zero. The total amount of “surplus” in that scenario is larger, and it is possible for an ISP to split the gain with the consumer by offering a price for unlimited usage. Thus, as long as the network capacity is in place, both parties are made better off! If not, the practice can lead to congestion, but competition for users often dictates that such packages be offered.

Especially in the presence of positive network externalities, it makes no sense for the ISPs, as a group, to price users or traffic out of the market, unless they are punished for doing otherwise at below cost. As always, pricing is an exercise in balancing costs with the benefits to potential buyers. It should remain a private and unfettered exercise ending only in trades that are mutually beneficial.

And what of network capacity and the big content providers? At the “price discrimination” link above, Howell says:

… available bandwidth allowed Netflix to happen, not the other way around. But now, as Netflix comes to dominate existing bandwidth, leading to higher costs, it is causing externalities (delays) and higher costs (ISP fees are now rising in real terms in some markets) to pay for new capacity.

Should the ISPs charge all customers higher rates in order to manage growth in traffic and fund new capacity? How can they allocate costs to the cost-causers? Usage-based data rates are one simple alternative. Tiered rates would act to minimize the extent to which light users are penalized. ISPs have also negotiated with individual content providers directly, reaching agreements to compensate ISPs for access to their customers. Tim Wu, the Columbia Law professor credited with coining the term “net neutrality”, was quoted at the last link bemoaning these types of deals:

‘I think it is going to be bad for consumers,’ he added, because such costs are often passed through to the customer.

Well, yes! Netflix charges its customers, and it will attempt to recover these payments for network capacity. Streaming is an integral component of the service they offer, and they cannot do it without the ISPs. Would Wu propose that the pipes be provided at less than cost?

Some have said that it is more economically efficient for ISPs to charge users directly for incremental short-run network “externalities” caused by large data demands. (Conceptually, it is better to think of these costs as long-run marginal costs of network expansion.) It may be that a tiered rate structure can approximate the optimal solution, and packages are often tiered by download speed. Nevertheless, passing costs along to large content providers is a viable approach to allocating costs as well.

Another argument is that small content providers cannot afford these payments. However, if they don’t generate a significant amount of traffic, they probably won’t have to negotiate special deals. If they grow to require a large share of the “pipe”, it would indicate that they have passed a market test. Ultimately, their customers should pay the costs of providing the capacity in one way or another.

Net neutrality and regulation of ISPs is the wrong approach to encouraging the growth and value delivered by the internet. It would stifle incentives to provide the needed capacity and to develop new network technologies. We certainly didn’t get here by treating the ISPs like public utilities. Rather, the process was facilitated by the freedom to experiment technologically and contractually. ISPs are well aware that the value of their networks are enhanced by ubiquity. Affordable access to a broad share of the population is in their best interest. In the end, consumers are sovereign and should be the sole arbiters of the value offered by ISPs and content providers. Regulators will promise to protect us, but the inevitable result will be a market hampered by rules that degrade the network, leading to substandard service and a less vibrant internet.