Report on the PDP-Feb06 from Sao Paulo

[Harold Feld]

At 06:06 PM 12/5/2006, Milton Mueller wrote:
> >>> Mawaki Chango <ki_chango at YAHOO.COM> 12/5/2006 3:20 PM >>>
> >TERM OF REFERENCE 3
> >Policy for price controls for registry services
> >Policy Recommendation N (Option 2):
> >
> >The NCUC has argued that it is premature to formulate policy in the
> >area of pricing without having had the benefit of an intensely
> >focused study on this topic. They believe that a new PDP is required
> >to address the specific issue of price controls. ("We believe that
> >existing price caps should be left in place for the short term, and
> >another, separate PDP inaugurated on methods and criteria for
> >changing, raising or eliminating price caps in the future.")
> >
> >Thus, another option is to keep the status by encouraging ICANN to
> >continue with existing pricing provisions and initiating a targeted
> >PDP on this issue alone taking into account the upcoming economist's
> >report (http://www.icann.org/minutes/resolutions-18oct06.htm).
> >
> >NCUC: N
>
>Yeah! Good stuff. Also, the "market dominance" analysis in the Option 1
>is fallacious from an economic point of view. To an individual
>registrant with an established domain, opportunistic pricing hurts them
>whether or not the registry is dominant in the total  market.

Allow me to expand on this a bit, because it is important.

The theory of competitive markets holds that competition between
providers serves consumers (some would say better than
regulation).  When consumers can freely choose closely substitutable
products from a variety of vendors, the vendors must work hard to
attract customers.  Thus, in competitive markets where consumers can
freely switch, vendors will generally seek to hold prices to the
lowest profitable level (lest a competitor attract price sensitive
customers) or offer services that justify a provider premium (e.g.,
better customer service).

There is a difference between whether competitive alternatives are
available and whether consumers have the ability to take advantage of
the competitive alternatives.  In economic terms, it is possible that
the cost of switching from one alternative to another ("switching
cost") is so high that a consumer cannot switch to a theoretically
available alternative.  This is called "consumer lock-in."  Lock in
may also occur if a theoretical alternative is not, in fact available
to the specific consumer, if the theoretical alternative is not a
close substitute, or for other reasons.

Finally, where there are relatively few choices, and providers of the
choices all operate under the same general conditions and with
considerable knowledge of each other's practices, parallel behavior
with regard to pricing is likely to occur even in markets that are
considered "competitive" under most indicia (e.g., have four or more
equal size firms).

Apply this to the TLD market. There are now five truly generic TLDs
(.com, .org, .net, .biz, .info), a number of "sponsored" TLDs (e.g.,
.travel) and a number of ccTLDs open to large numbers of registrants
(but not necessarily all registrants) (e.g., .us (must have U.S. nexus)).

.com and .net are run by the same registry, and are thus unlikely to
genuinely compete against one another.  Verisign has no incentive to
try to drive customers from one TLD to the other.  This leaves only
three competing registries that can be regarded as possibly close substitutes.

But a consumer switching from one TLD to another faces very
significant costs.  Leaving aside the costs of changing addresses and
informing everyone else to change the address, a name holder has
usually invested considerable goodwill in a name.  Even switching
from Example.TLD1 to Example.TLD2 is likely to create customer
confusion and dissipate good will by creating inconvenience for
customers/website visitors.

This gets worse if "Example" is not available in TLD2, since it
requires not merely a change in TLD, but a substantive change in the
name.   Finally, the holder of Example.TLD1 that abandons that name
due to concerns about pricing runs the risk that someone else will
register Example.TLD1 and exploit the goodwill created by the
original registrant or create confusion (either innocently or deliberately).

There is also the question of whether the registrant regards TLD1 and
TLD2 as "substitutes" or "close substitutes."  The difference is
important.  I regard something as "close substitutes" if I am
indifferent or care only a little about which one I use.  By
contrast, if I will use a different product when I must but really
don't regard it as the same, it is merely a "substitute."  For
example, Diet Coke and Diet Pepsi are both diet colas.  Many people
are equally happy drinking Diet Coke or Diet Pepsi, and therefore
regard them as close substitutes.  I have a strong preference for
Diet Coke, and regard Diet Pepsi as merely a substitute to use when
no Diet Coke is available.  Further, while water is also a drink with
no calories, it is clearly different from a diet cola.  Again, some
people may regard water as a close substitute, while others merely
regard it a substitute and will pay a premium for diet cola if that
is what they genuinely want.

For all these reasons, a name holder may not regard an available
competitive alternative as actually available, or may endure onerous
terms the name holder would have rejected when initially selecting a
TLD in which to register.

Worse from a consumer perspective, the relatively small number of
providers makes it less likely that the consumer will have a real
choice.  Even if there is no market power in the usual sense, the
limited number of competitors ensures that all competitors will
quickly switch to the more profitable pricing strategy unless there
is an immediate, strong consumer rejection of the change in pricing
policy that drives so much business to the few competitors that it is
worth maintaining the old policy.  But lock in factors discussed
above make such a mass exodus/market correction unlikely.

In short, "market power" or the presence of possible competitive
alternatives is not the only measure of whether consumers have
sufficient real choices to rely on market forces in place of
regulation.  A better public interest analysis will look not merely
to the number of possible competitors or even only to market chare,
but to the nature of consumer lock-in and other factors that may
limit the ability of competitive market forces to protect consumer interests.

Harold