Virulent Word of Mouse

April 23, 2014

The Fault Lines Along Fast Lanes


highwayUntil recently, a fast lane from a broadband ISP was a remote possibility in the US. ISPs had to give data equal treatment, regardless of the source, and could not offer faster delivery for a higher price while giving slower service as a default.

Although fast lanes were allowed by regulators a few years ago in the wireless networks, the carriers hesitated to offer them. In December 2013, AT&T Wireless broke with the norm and announced just such a program. FCC regulations forbidding fast lanes at landline broadband ISPs had also prevented them, but a January 2014 US appeals courts struck down those regulations.

Is that a good or bad trend? The answer depends on who’s talking. Critics of government regulation despise the rules forbidding fast lanes, whereas net neutrality supporters view the presence of fast lanes as a nightmare come to life.

Legal and political aspects of this topic typically get most of the attention, as do the implications for the variety of speech online. Most reporters find these aspects interesting, and understand them. However, the economics of fast lanes receives less attention. That is a surprise, because the economics is not very difficult, and it’s worth understanding. It illuminates the fault lines between many different points of view.

Mirrors and servers

The public Internet has evolved considerably since the days when the design for packet networks presumed that the message did not have to arrive at an inbox immediately. Users today prefer and expect speedier services. That goes for more than just IP telephony and video chat, where users notice the smallest delay. It also holds true for video, such as YouTube and many online games. Many providers believe it also affects the bottom line—namely, that users switch services if they do not get fast delivery of data.

Long before fast lanes became a real possibility, many participants in the Internet made investments aimed at reducing delays. For example, for some time now, Akamai has sold a well-known approach to improving speed. Their service also defines the first fault line, so this is a good place to start the discussion. Opponents to net neutrality ask why Akamai can operate a business to speed up data delivery but a carrier cannot.

Akamai’s service supports servers inside ISPs, closer to households. Any seriously large Internet content firm must buy these services, and it is considered a cost of doing business online. Many ISPs like working with Akamai, because their customers experience better service without much investment from the ISP.

That is not the only method for speeding up data. For example, Google has bypassed Akamai’s charges in many locations by building its own data network to ISPs. Netflix has recently sought to do the same, though it is not quite done (because it has not successfully negotiated a presence with every US ISP). Any gathering of more than three Internet engineers will generate discussion of even more potential solutions in the cloud. Amazon built a content delivery network with enormous geographic range. Microsoft has similar investments and aspirations, as does IBM. The list goes on.

That leads to the deeper question. The last few years have witnessed robust experimentation among distinct approaches to functional improvement, and these might be either complements to, or substitutes for, each other. Accordingly, carriers have had two roles. They act as a firm whose users benefit from faster delivery, and they act as a supplier that could choose to cooperate—or refuse to cooperate—with solutions offered by others.

When a carrier had no investments in fast lanes, it had every reason to cooperate with solutions offered by others. Will that change if the carrier has its own fast lane?

The answer defines a fault line between points of view. Some observers label this a possibility that might never arise. They want a regulatory response only when a problem emerges, and otherwise they anticipate that a regulator will err. Net neutrality supporters think regulators have an obligation to protect the Internet. Advocates worry that introducing fast lanes messes with a system that already works well. They do not trust carriers to cooperate with solutions that might substitute for a fast lane business or threaten an investment in some way.

Competition and monopoly

The next fault line has to do with the role of money. Defenders of fast lanes expect them to become a cost of doing business for content firms, and forecast that fast lanes will be profitable and generate more investment. Opponents have the same forecast about profitability, but a different interpretation. They worry that fast lanes will lead to an Internet where only rich firms can deliver their content effectively.

This concern tends to get plenty of press, and a few rhetorical questions illuminate the fault line. Will the default speeds offered by ISPs be good enough for startups or for small specialty websites? One side believes that the defaults will be good enough, whereas the other believes that fast lanes will lead ISPs to neglect investing in their slow services.

One’s point of view about the state of competition for ISPs has a big role in interpreting the role of money. Some believe a competitive ISP market would melt away most problems. Others argue that belief about competitive ISP markets is a fantasy and masks many dangers.

The belief in competition is not a belief in magic, so it is worth examining. Rather, this side views competition as a painful process. In competitive markets, customers substitute into alternatives if they do not like what a supplier does. Suppliers hesitate to do things that make their users angry. In other words, ISPs would compete for customers by offering better fast lanes. In this view, users would get angry if they perceived that carriers were slowing down content from firms they cared about, and angry users would find another carrier.

Where is the fault line? Recognize the two key factors that make ideal competitive markets operate well—namely, transparency and the availability of many user options.

Just about everybody is in favor of transparency, but not necessarily more of it if rules require it. Those with faith in competitive processes tend to see the merits in nothing more than a few light-handed requirements, such as programs to facilitate measuring the speed of different ISPs. The other side asks for much more, such as the publication of all fast lane contracts (more on that later).

As for the second concern about options, consider the key open question: Do users have many options available to them, or do they face de facto monopoly ISP markets? Once again, there are different beliefs about the preponderance of competition and monopoly found throughout locales of the US. Those who presume that competition is inadequate lack sympathy for leaving ISPs alone (versus those who presume it is adequate).

That also leads to different interpretation about how lucrative fast lanes will be. Supporters of fast lanes say that ISPs should charge whatever the market will bear, and competition will discipline pricing. Opponents say that the monopolies emerged from granting public franchises and use of public rights of way, and characterize high prices as misuse of utility franchises.

A classic debate about government merger policy also arises. Net neutrality supporters argue that fast lanes give ISPs artificial incentives to consolidate in order to increase their bargaining leverage with content providers, thus concentrating economic power in ISPs. Net neutrality opponents do not see anything wrong with large ISPs. In a competitive market, size is irrelevant.

Mixed incentives

The foregoing leads into the last fault line in discussions about fast lanes—namely, views about mixed incentives at carriers. A mixed incentive arises when a carrier distributes a service that substitutes for one available on the public Internet.

Many broadband ISPs have a thriving broadband service and provide video on demand, and make a pretty good margin on both services. Will most cable firms want to sell a fast lane service to Netflix at a low price? If the carrier did not make money on video on demand, then a carrier’s price for a fast lane for Netflix would be lower, and the same goes for entrepreneurial firms offering video services. That also begins to suggest the intuition behind the concern that cable firms will tilt their other actions against online video to protect their existing businesses.

Mixed incentives also come up in discussions about scrutinizing carrier contracting practices. To put this fault line in perspective, consider a hypothetical scenario: What would happen after a carrier sells a fast lane to, say, ESPN? Can anyone else expect the same terms, even Netflix? Yet again, one side argues that competition will solve these issues, and the other sees a need for regulatory intervention to make terms of fast lane contracts public.

A mixed incentive also can emerge when a carrier has an economic incentive to protect its partner’s business in which it gets a cut. In other words, is it okay if ESPN gets a better deal than Fox Sports because an ISP made a deal with the local team who competes with something done by Fox Sports? The same fault line as just mentioned: should competition solve this question, or should governments intervene to publish fast lane contracts? Should ISPs be required to give the same terms to all takers?

To summarize, the fault lines between perspectives hinge crucially on several beliefs about the economics. Forecasts depend on whether the observer sees a preponderance of competitive or monopoly markets for ISP services. They also depend on whether transparency resolves potential problems.

 

Copyright held by IEEE. To view the original, see here.

mouseonmouse

4 Comments »

  1. Thanks for the post—very timely as there are 3 recent news articles [1,2,3] that are related to BB policy. My comments / questions will probably show my __rudimentary__ understanding of (/confusion about) broadband issues:

    (A)
    It seems to me that there are (at least) two distinct pricing issues and two distinct customer groups: (1) usage-based pricing (UBP) and quality-of-service based pricing (QoSBP), and (2) consumers (households / individuals) receiving traffic over last mile and enterprise customers such as Netflix

    As I understand, wired providers do not (yet) use UBP widely (for consumers) although wireless providers have resorted to UBP plans. On the other hand, carriers have been using speed tiers (QoS) on the consumer side for a while. What is the interplay between usage and QoS? Why is different prices for usage / QoS on the enterprise side a problem while they seem to be OK on the consumer side? (I have read about, but hardly understand, the added complexity of peering)

    (B)
    First, Is there any evidence that price (or contract term) disclosures increase welfare (economic theory, econometric studies in BB or other settings—I’d appreciate any reading references)? Second, AFAIK, contracts with secret terms and prices (w/ NDA) are widely used in other enterprise settings such as enterprise software (for example, IBM probably sells Websphere at a lower price to Chase Manhattan than to a smaller company). Why would we not require disclosure of terms and prices in other highly concentrated markets?
    The following is an excerpt from NYT:
    {
    Under the proposal, broadband providers would have to disclose how they treat all Internet traffic and on what terms they offer more rapid lanes, and would be required to act “in a commercially reasonable manner,” agency officials said. That standard would be fleshed out as the agency seeks public comment.
    }

    Regards,
    Zia Hydari (CMU)

    [1] Welcome to Comcast Country

    http://www.nytimes.com/2014/04/24/opinion/welcome-to-comcast-country.html?hp&rref=opinion

    [2] F.C.C., in a Shift, Backs Fast Lanes for Web Traffic

    http://www.nytimes.com/2014/04/24/technology/fcc-new-net-neutrality-rules.html

    [3] FCC to Propose New ‘Net Neutrality’ Rules

    http://online.wsj.com/news/articles/SB10001424052702304518704579519963416350296

    Comment by Zia Hydari — April 24, 2014 @ 10:31 am | Reply

    • Thanks for the comment.
      (a) I do not quite know how to answer. I do not yet see much interplay (yet) b/w QOS, UBP and fastlanes.
      (b) Transparency in contracts for communications services are not comparable with transparency in enterprise services. In a communications network there are enormous externalities, as well as different competitive conditions. The terms offered by a carrier shapes entrepreneurs and established content providers. For example, ESPN might be able to get a better deal than anybody else if the carrier knows it will be secret, but perhaps not if exposed to public scrutiny. Transparency is also much better for enforcement of antitrust.
      Cheers, Shane

      Comment by Shane Greenstein — April 24, 2014 @ 2:34 pm | Reply

      • Thank your for your reply. It seemed to me that fast lanes are end-to-end QoS guarantees; otherwise, enterprise customers have been able to purchase dedicated internet access for different speed at different prices for a while …

        Comment by Zia Hydari — April 24, 2014 @ 3:16 pm

      • Ah, now I understand you question. The fast lane rules being discussed are about the rules for the “public Internet”, primarily home access, not about access to (large) enterprises, nor about dedicated lines to (large) enterprises. Don’t mix the home and enterprise contracting regulatory environments. Different animals.
        Shane

        Comment by Shane Greenstein — April 24, 2014 @ 3:40 pm


RSS feed for comments on this post.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

The Rubric Theme. Blog at WordPress.com.

Follow

Get every new post delivered to your Inbox.

Join 60 other followers

%d bloggers like this: