<h1 class="heading">What the airline industry can teach us about broadband caps</h1>
                <h2 class="standalone-deck">The free market can't do it alone; good government policy can promote competition.</h2>
                <div class="article-content clearfix">
                         <p class="byline">
        by <a href="http://arstechnica.com/author/timothy-b-lee/">Timothy B. Lee</a> - <span class="date">May 11 2012, 1:36am RDT</span>
</p><p>Writing for <em>Wired</em> (which shares a parent company with Ars Technica), Susan Crawford, currently a visiting professor at Harvard, <a href="http://www.wired.com/epicenter/2012/04/opinion-crawford-cableization/">complained</a>
about a recent travel mishap. An airline agent told her she wouldn't be
allowed to bring her viola on the airplane despite the fact that the
flight's overhead compartments had plenty of room for it. Violas are too
fragile to check, Crawford says, so she was forced to spend an extra
night in DC in order to take a different flight—and deal with a
different, more accommodating gate agent—the next morning.</p>
<p>She used her story as an extended metaphor for the current debate
over Comcast's broadband cap. Netflix is the hapless passenger in this
analogy, while Comcast is the erratic gate agent who demands it pay
extra to allow its <span style="text-decoration:line-through">viola</span> high-bandwidth videos on the <span style="text-decoration:line-through">airplane</span> network.</p>
<p>In a response to Crawford, Hance Haney of the Discovery Institute <a href="http://techliberation.com/2012/05/03/nothing-to-fear-from-pricing-freedom-for-broadband-providers/">pounced on this example</a> and argued that it demonstrated precisely the opposite point:</p>
<blockquote><p>Most people don't own a viola, nor do they want to
subsidize viola travel. They want to pay the lowest fare. Differential
pricing (prices set according to the differing costs of supplying
products and services) has democratized air travel since Congress
deregulated the airlines in 1978. First class helps make it possible for
airlines to offer both lower economy ticket prices and more frequent
service. Which is probably why Crawford's column isn't about airlines.</p></blockquote>
<p>The comparison of broadband to airlines <em>is</em> instructive, but
not for the reasons that either Haney or Crawford suggest. Haney is
right that airlines' ability to offer a first-class option is good for
everyone because it induces those first-class passengers to pay a
disproportionate share of the costs of air travel. But airlines'
discriminatory policies are not a matter of public concern precisely
because good public policy has made the airline industry relatively
competitive. If one airline makes its coach passengers miserable to
induce them to upgrade to first class, many will switch to a competing
airline.</p>
<p>That's often not a realistic option for broadband customers, who
rarely have more than one serious alternative to their current broadband
provider. Thus, Internet data caps have generated so much more outrage
than price discrimination by airlines. But the success of airline
deregulation has important lessons for fixing our dysfunctional
broadband markets.</p>
<h2>Airlines don't own airports</h2>
<p>The deregulation of air travel was an important achievement of the
Carter administration. Before 1978, the airline industry was a cozy
cartel insulated from competition by government regulations. The
elimination of these regulations allowed a number of airlines to enter
the market or expand service to new cities. Fares plummeted, bringing
air travel within the financial reach of many more travelers.</p>
<p>Paradoxically, a key factor in the success of airline deregulation
was that airports remained in government hands. It's a useful thought
experiment to imagine what would have happened if the same policymakers
who deregulated the airlines had also privatized the nation's airports
by selling them to incumbent airlines. Superficially, that might look
like a further step in a free-market direction, but appearances can be
deceiving.</p>
<p>Deregulation succeeded because it opened the market up to new firms.
But new airlines require access to existing airports, and incumbents
would have little incentive to rent gates on good terms to would-be
competitors.</p>
<p>Theoretically, new firms could just build their own network of rival
airports. But this would not only be prohibitively expensive, it would
also require demolishing thousands of homes and subjecting thousands of
households to the noise and other nuisances of living near these
redundant airports. Indeed, constructing a new major airport is so
disruptive to nearby residents that in practice it only occurs with
government assistance. That's one reason that most metropolitan areas
have only one major airport, and even the largest metropolitan areas
only have two or three. City officials are understandably reluctant to
have more airports than absolutely necessary.</p>
<p>Fortunately, incumbent airlines don't control the nation's airports;
as a result, the airline industry has become highly competitive. New
firms like US Airways, Frontier, JetBlue, and Virgin enter the market at
regular intervals.</p>
<h2>Vertical integration can harm competition</h2>
<p>We can illustrate the problems of the current broadband market by
extending our airline analogy one step further. Suppose that in addition
to owning the nation's airports, the incumbent airlines also acquired
hotel chains. These "vertically integrated" airlines could then boost
their profits by pursuing the following strategy. First, hike the cost
of an ordinary airplane ticket. Second, offer customers "double play"
deals if they stay at airline-owned hotels. This would make it easier
for airlines to identify their richest customers based on which hotels
they stayed at, and to charge those customers correspondingly higher
prices.</p>
<p>This kind of bundling would be good for the bottom lines of incumbent
airlines. But it would be a terrible development for independent hotels
and tour companies, which would be starved of customers and struggle to
stay in business.</p>
<p>The broadband market seems to be undergoing a similar transformation.
Comcast already offers "triple play" deals of data, phone, and video
services. And it recently acquired NBC, paving the way for even greater
vertical integration down the road. This makes independent firms like
Netflix and Sony <a href="http://arstechnica.com/tech-policy/news/2012/05/sony-warns-comcast-cap-will-hamper-video-competition.ars">understandably nervous</a>,
because it means that Comcast is both a key source of customers—Comcast
controls more than 20 percent of the US residential broadband
market—and a major competitor.</p>
<p>This isn't just bad for Netflix; it's bad for consumers too.
Competitive markets produce a wide variety of innovative content that a
handful of vertically integrated telecom-entertainment conglomerates
can't match. If incumbent firms destroy their competitors in an effort
to increase their profits, consumers will suffer from reduced
competition and perhaps higher prices as well.</p>
<h2>Fixing broadband is hard</h2>
<p>By separating the airports from the airlines, regulators ensured that
airport monopolies didn't undermine competition in the airline market.
In principle, this is a good model for the telecommunications market,
and it's a model some countries around the world have pursued.</p>
<p>In the UK, for example, the incumbent telecom company, BT, was
required to create a separate subsidiary, called Openreach, that
provided wholesale access to the "last mile" of its network. A variety
of companies, including another division of BT, offer retail broadband
access on top of the bare infrastructure provided by Openreach. This has
enabled <a href="http://www.ispreview.co.uk/story/2011/09/06/uk-home-and-business-broadband-isp-connections-top-19-9-million-in-q1-2011.html">serious competition</a> among wired broadband providers.</p>
<p>The United States tried a similar regulatory approach starting with
the 1996 Telecommunication Act, requiring the local phone incumbents to
"unbundle" their lines for use by competitors. But the incumbents were
not required to fully separate their wholesale and retail divisions, and
that made it almost impossible for the Federal Communications
Commission to guarantee that third parties were being treated fairly.
The incumbents wore their would-be competitors down with years of
foot-dragging and litigation. The unbundling experiment finally ended
when the Supreme Court <a href="http://en.wikipedia.org/wiki/Brand_X">allowed</a> the FCC to abandon it in 2005.</p>
<p>At this point, resurrecting unbundling would be a tough sell, and the
National Broadband Plan doesn't call for it. Most of the independent
firms that took advantage of the first unbundling program wound up in
bankruptcy. If a future FCC tried to revive the policy, it would face a
tough sell with potential entrants.</p>
<p>Still, the success of competition in the airline industry—and a <a href="http://www.nationalaffairs.com/publications/detail/keeping-the-internet-competitive">similar success in trucking</a>—suggests
that the long-term goal should be to achieve a clean separation between
the monopolistic parts of the broadband industry and the competitive
parts. In recent years, we've been moving in the opposite direction,
with telephone and cable incumbents merging with each other and
acquiring "upstream" businesses like NBC. Just as we would have been
concerned about a monopolistic TWA acquiring the Hilton hotel chain, so
regulators should be concerned of mergers between broadband incumbents
and content companies.</p>
<p>Regulators missed the boat with the Comcast-NBC merger; they should be more skeptical next time such a merger is proposed.</p>
                </div>