Net Neutrality? There's No Such Thing.

Part of the new Dutch petascale national supercomputer, “Cartesius.” Flickr/Dennis van Zuijlekom

Its supporters have fallen victim to sloppy thinking.

Considering the internet a global common good has become fashionable. Calling for net neutrality is even more in vogue. Most importantly, lawmakers and regulators worldwide are subscribing to the idea of an unalienable right to a nondiscriminating, and possibly free, internet. Well, all this is bad news. The consequences of governments forcing net neutrality might be dire. Not only is net neutrality based on bad economics, but it also takes its toll—on internet users.

It is undeniable: the internet is an important part of many people’s everyday life, of the global economy and of many essential services or facilities. It is equally hard to make a case against the falling costs of accessing the web. Free hotspots, flat rates and immense data packages are witnesses of this development. It might be concluded, then, that the internet is a reality and a necessary tool for modern life. Therefore, the net should be neutral. That is why the web is often compared with the highway system, and its maintenance and management to broadcasting.

But not only are both these comparisons problematic, but the very idea that there ought to be a nondiscriminating, neutral and possibly free internet is false at best. Turning its implementation into a further task of government is even worse. It puts the very fabric of the internet at risk, delaying investment, lowering its service levels and creating more potential for discrimination.

Let’s review this: first, the mental imagery used to make the case for net neutrality, and second, the economic arguments that support—or do not support—the claim.

Getting Metaphors Right

The internet is often compared to a highway system. In the U.S. legal system, an analogy was made to broadcasting licenses. Both these images are wrong because they focus on some features but by far not on all—and not on all the important—facets of public roads and broadcasting. Especially, they fail to mention the less successful aspects of their subjects. If one is searching for an image, it is much more realistic to compare the internet to a shopping mall. Here is why.

The highway metaphor: Data moves through cable and signal waves. Where there are no wires, information cannot pass. Where there are no signal transmitters, there is no net. The picture seems clear enough.

Many people conclude, then, that if there is enough cable and transmitter capacity, there is no need for managing whatever moves through them. In other words: if there is enough capacity, the surplus of capacity guarantees neutrality in its use. (And many would even claim that if there is no capacity, the government should simply build some.)

However, there is more to a highway system than the actual traffic. First, highways are not privately owned; the cables and transmitters of the web are. Second, the road system has been built with public money and maintained with taxes; internet infrastructure providers are financed with private money and cannot impose or collect taxes. And third, highway systems compete with other public roads, pay-per-use private expressways and so on. Therefore, there is room for user discrimination according to their preferences and profiles.

By the way, there is further non-neutrality, such as in the case of traffic jams, where the police often direct the flow of vehicles, giving preference to some. And there are those places with mobility pricing, which is the most non-neutral a road system can get. This shows that if you want net neutrality, highways are the wrong way to think about the web.

The broadcasting license analogy: this image is more serious, since the FCC and U.S. courts seem to be fond of it. The Radio Act of 1927 established that the radio frequency spectrum belongs to the public, and licensees have no property rights to use it. The analogy to the internet roughly assumes that broadband is a similar case of public domain, and that therefore whoever is licensed to operate it must provide public service, in this case nondiscriminatory traffic. What is wrong with that analogy?

For starters, among economists, even those of a more institutional persuasion, the idea of licensing is considered inefficient. The economist Ronald Coase proposed that, as for other resources (such as land and metal), the market should regulate the use of the radio spectrum: with well-defined property rights, the free market will allocate resources to their most efficient use. Then, we all know how public-service broadcasting turned out: its programs usually consist of vintage features, and its commercial breaks disrupt every program. No wonder consumers are willing to pay more for nonpublic channels like HBO, or switch altogether to disruptors of broadcasting like Netflix.

Here, again, the point becomes obvious: the very regulation aiming at neutrality lowered the level of service by neutral providers while creating the potential for discrimination. There is, however, a third argument against this analogy. While the radio spectrum is somewhat limited, broadband is expandable. Those firms whose business model is running the net create broadband. Their return on investment is on the provision of infrastructure and less on the usage of existing systems. So, here again, the analogy misses the important points—points that would deeply affect the web.

The mall comparison: There is more to this image than the apparent irony of the web driving shopping malls out of business. Usually, a developer-manager sets up a shopping mall envisaging renting space to individual shops, restaurants and other services. While it is in the interest of the developer-manager to offer a diversity of goods, attractions and price categories, the mall is not a public service. And not every business can set up shop there. Indeed, businesses compete for space and pay a price for it.

Most developers-managers also use other criteria in adjudicating space than just price; for example, there are reserved spaces for seating, for childcare or for parking—all of them free and nondiscriminatory. A shopping mall is therefore a mix of discriminatory and nondiscriminatory spaces, and even in the non-neutral spaces there is not only a race for price, but for diversity. All these factors make the mall attractive. Most important of all: the mall developer-manager is a private entity that set up the mall in order to generate economic benefits. And selecting who sets up shop is a very important aspect of the mall’s success.

The same principles apply to the companies that build and manage the internet’s infrastructure. They are private entities investing in a space they want others to use. Naturally, they also want to generate revenues from it. It is in their interest to make their space attractive, by allowing a diverse mix of offerings, but also by sorting out how these offerings operate. Yes, there is a considerable portion of management of that infrastructure—be it DSL, broadband, point-to-point or something else—and management is always, to some degree, discretionary. Good management, however, is rule-based and eventually, the rule of discretion becomes known to the stakeholders. It is then up to the stakeholders to decide whether to use that one mall (or broadband provider) or to switch to the competition. Yes: competition between internet infrastructures (DSL, broadband, point-to-point and things to come) is reality nowadays. And it will become even fiercer as the costs of building broadband, as well as the costs of its alternatives like direct, are rapidly decreasing.

What do these metaphors show us? Getting the right comparison is important. If there is need for an image of the web, this image cannot be drawn from the aspects one might like while at the same time disregarding all the others. A good metaphor tries to reproduce the dynamics of the system. The shopping-mall comparison is superior to the highway metaphor and the broadcasting analogy because it gets the dynamic of the internet right. This dynamic is business-driven, because it is set up by for-profit private businesses. And the dynamic is not neutral. But it is exactly the rule-based non-neutrality of that dynamic that makes shopping malls and the internet attractive to investors and users.

Getting Economics Right

The most important mistake of the highway metaphor and the broadcast analogy is that they consider the internet to be a public good. But it is not. It is private, developed by private entities, maintained by private firms and especially paid by private agents, be they content generators or end users. Treating the net as a public good is a severe misunderstanding of its economics. More than it, it leads to even worse policy proposals. That is why it is important to get the economics of public goods—and of the internet as a private good—right.

The public versus private dilemma: The term “public good” is based on the much older “common good.” It comes from the traditional English legal term for common land. It refers to resources accessible to all members of a society; these resources are held in common, not owned privately. Therefore, they are also free. If a member of a society does use the resource, he or she pays nothing. Examples of commons apart from common land are air, international waters and, according to some people, habitable earth.

The term “public good” is a nuance of “commons.” Their supply generates costs, but their consumption is (somewhat) free. Typical examples are public parks, defense or, yes, broadcasting. The difference between a common and a public good is that the common good does not cost anything to produce, whereas the public good does. Their similarities are that their use is generally free of cost and that they do not belong to any private entity. Common- and public-good economics, based on this, makes a semi-normative claim about these goods: they are both non-excludable and non-rivalrous, in that individuals cannot be effectively excluded from use, and use by one individual does not reduce availability to others.

Economics has a rather unfavorable view of common and public goods. Since they are supposed to belong collectively to the community and are free of charge to consume, commons and publics are usually abused—and there is also a great deal of free riding. Why is this? Because leaving the cost of maintaining something to the collective is tantamount to leaving it to nobody. Collective responsibility is contrary to individual responsibility. And being free of charge, commons and publics are a boon for free riders. Keep in mind: just because something is free, it does not mean that its creation or maintenance does not come with costs. That is the problem of common and public goods: their maintenance still bears costs and, more often than not, the community cannot summon enough means for compensating use and abuse. Generally, publics and commons ends in a state economists prefer to call a tragedy.

Naturally, there are possibilities for the social management of common and public goods. But usually, social management is pretty coercive. Most of the time, it means that every participant of common  or public goods has to put in some work for it. And try to apply it to the net: does that mean that everyone should have a duty to expand broadband and manage it? This is hard to believe and much harder to implement.

Of course, government is the last resort in trying to avoid the tragedy of common and public goods. But government will use taxes to provide for common goods and at the same time to provide for its own institutions. These taxes are usually higher than the private production of a similar private good would cost. And it is more effective, too. A tax has to be paid by everyone, regardless of how much the public good is used. Private production follows individual interests, as users want to use the private good.

Again, applied to the internet: would this mean that the government should expropriate all private investors that have built the web so far? And then what? Will taxes for social programs or defense be used to build broadband? Or will there be tax increases? The theoretical solutions to all public-good problems are usually more complicated than the problem itself.

Is it then a mistake to imagine the internet as a common or public good (remember, this is the intellectual cornerstone of net-neutrality advocates and of the FCC’s ruling)? Yes, it is. In reality, there are two faults within the same claim. The first is a fundamental error in economic reasoning. The second is a more preoccupying normative claim disguised as a factual statement.

Bad—and good—economics: The free-of-charge use of some of its applications doesn’t make the whole internet a public good. Granted, there are many free places in the web, and even in its infrastructure: think of forums, newspapers, magazines and even games—as well as free hotspots, free broadband, free data packages. But usually, the nonpaying consumer is paying for the used service in another way, whether by willingly becoming the target of advertising, or by making a particular page more attractive in accessing it, or by sharing data with the service’s provider. For the infrastructure provider it is even simpler. Diversity of users brings traffic, and traffic is good news. Also, most web users pay flat rates, and other prices that are not directly linked to the traffic they generate. So, there is already a price for using the web, which makes it a private good. The internet is like the proverbial free lunch: there is almost never one.

Now to the more important error in the net-neutrality argument: it disguises a normative claim as a factual one. Normativity means saying what something ought to be; factuality, on the other hand, is an assessment of things as they are. Why is this second error so troublesome? Saying that the internet is a public good because it should be one is a normative claim, even a political program. It takes a desired end state while disregarding all factors that might influence it—in this case, the private nature of the web, the individual incentives to invest and, most importantly, the business cases that maintain service levels and quality. Think of it this way: in physics, the earth’s gravity generates friction and that might be problem. Many physical experiments and applications would work much better if there was no friction. But just assuming that there is none is not a formula for success. Good physics has to deal with gravity, in the same way that good economics has to deal with the nature of private property.

But in this case, the web being a system of private goods is the good news. Private interests built the net in the first place, and they will continue to expand and make it better. In the case of the internet, private property is the motor of innovation. There is no net without private investment; understanding this also means getting the economics of the internet right. There are no websites without agents spending money on them. There is no content without its development envisaging some sort of monetary return. And there is no infrastructure without a for-profit business model at its base.

From the private nature of the internet, it necessarily follows that the infrastructure provider can use those principles for managing the web as they see fit. Indeed, the net is as valuable as its management, and management is never neutral. All businesses are by necessity non-neutral. How would restaurants manage if they were neutral? Instead, they choose their customers by the cuisine they offer, by their price range, by allowing for reservations and setting aside tables for preferred customers. Financial-services providers have specialized products according to clients’ profiles—for example age, risk propensity, financial goals or experience. Non-neutrality in these and other business models increases profits and customers’ welfare. In fact, managerial non-neutrality is very often the core of value added.

Why should it be different with the web? In order for the system to work, internet providers will have to discriminate according to traffic, data usage, data package and, yes, special agreements, reciprocity contracts and the like. This is what makes the net usable in the first place. If there were no active management, the internet would be congested with pornography and drug traffic taking up most of its space. Thanks to non-neutral management, important (messages, health, financial) and entertaining (music, games) data packages have a measure of priority. Keep in mind: management is always discretionary, but management of goods is rules-based discretion. While the web’s providers have a right to non-neutrality as they deem fit, they will bear users’ response to their non-neutral policies. The clearer they are, the better they will be rewarded by customers.

Getting It Right

“Net neutrality” is not even a romantic dream. It is a series of mistakes flanked by misconceptions. Putting it into practice jeopardizes the web, making it slow, lowering its service levels and finally delaying all investment cycles. The internet is a bundle of private goods. Business models are behind every offer in the web, and especially at the base of its infrastructure. And business models are there for profit. In order to make profits, infrastructure providers must be able to manage the traffic. And management is by necessity non-neutral. But this is good news, since it is the rule-based discretionary management of the web that makes it usable in the first place. Accepting non-neutrality is simply good economics. And it is win-win for all involved, especially users.

The internet is a bundle of privately owned and privately run businesses. And it is only successful and helpful to all people if it continues so.

Henrique Schneider is chief economist of the Swiss Federation of Small and Medium Enterprises.

Image: Part of the new Dutch petascale national supercomputer, “Cartesius.” Flickr/Dennis van Zuijlekom