E-Commerce Sites Make Great Laboratory For Today's Economists

[R.A. Hettinga]

<http://online.wsj.com/article_print/0,,SB109744444984841427,00.html>

The Wall Street Journal


 October 11, 2004

 PORTALS
 By LEE GOMES


E-Commerce Sites
 Make Great Laboratory
 For Today's Economists
October 11, 2004

Calling them eBayologists doesn't sound quite right; Amazonists is even
worse. There may not be a good name for them, but for a growing number of
academics trying to test ivory-tower theories in the real world, the big
online commerce sites are the new place to be.

Especially for economists, Web sites and their untold millions of
interactions are perfect real-world laboratories. For instance, a
fast-growing specialty in economics is "auction theory," which, as the name
suggests, tries to tease out the basic patterns by which auctions operate.
Before eBay, an aspiring auction theorist didn't have much data to work
with. As a result, many of them spent time sifting and resifting through a
limited set of data on government wireless spectrum auctions of the 1990s
for their auction insights.

But now, with eBay and the rest, there are more data than even the most
dedicated graduate student would know what to do with. EBay itself even
helps with this, supplying anonymized data to university researchers under
special arrangement.

What exactly are economists learning? Frankly, the conclusions aren't the
sort that would cause a layperson's jaw to drop; many are the sort that
only a microeconomist would love.

For instance, one of the cardinal rules of eBay involves the importance of
a seller's reputation, or the rating given the person by those who have
done business with him or her.

Luis M.B. Cabral, an economist at NYU's business school, says that for an
economist, it wouldn't have been immediately obvious that the eBay ranking
would be important -- partly because reputations on eBay, unlike those in
the real world, are essentially anonymous.

But not only is reputation important, Prof. Cabral found, but it affects
the prices sellers can charge and the amount of merchandise they can move.
In one study, he looked at what happens to sellers with perfect ratings who
receive their first piece of negative feedback.

Not only did they begin selling less, but they entered a kind of downward
spiral, with fewer sales and a declining reputation. "It really makes a
huge difference," Prof. Cabral says.

Another academic foray into the online world ended up confirming the power
of brand names. Conventional wisdom among most American economists would be
that prices on Web sites should be essentially the same. After all,
Internet users all have "perfect information" about what everyone is
charging, and they face few, if any, "switching costs" in moving from one
supplier to the next.

But Judith A. Chevalier, an economist at the University of Chicago business
school, looked at price differences between Amazon and Barnes & Noble and
found that Amazon was consistently able to charge more. And in a separate
study, she found that Amazon's customer-written reviews of books do indeed
affect sales, with negative reviews having more of an impact than positive
ones. Prof. Chevalier theorized that prospective shoppers may suspect
positive reviews are some sort of shill for the author but don't bring
comparable skepticism to negative notices.

Even with all the interest by researchers, some of the questions posed by
the e-commerce sites remain unresolved. One of them involves the usefulness
of "sniping," which occurs when a prospective buyer in an auction doesn't
get involved in the bidding until the last minute, and then makes an offer
the others don't have time to top.

The practice works only on eBay-style auctions, which have fixed time
limits. There are many eBay users who swear by the practice. In fact, there
are even a number of services that will snipe for you automatically.

David Reiley, a University of Arizona economist, says that for certain
kinds of auction items -- including standardized electronic gear with many
different bidders -- studies suggest that snipers don't do better than
anyone else, either in their success rate in winning their items or in the
price they end up paying.

But he says there may be other kinds of merchandise, such as those with
fewer bidders or prices that aren't easy to determine, where buyers should,
in fact, snipe. Even that could change, he notes, if everyone sniped. It is
currently unclear how many people on eBay engage in the practice.

Not all of the work being done online by economists is theoretical, and
savvy eBay traders would be wise to acquaint themselves with some of the
literature. Lucking Reiley, for instance, along with a colleague, Rama
Katkar, wanted to find out if it was better to publicly set a minimum bid
for an item or to use an eBay feature called a "secret reserve," which
rejects bids below a certain dollar amount without telling bidders what the
minimum is.

By comparing 50 pairs of identical items, one with the secret minimum and
one without, the economists concluded that the secret minimum dissuaded
buyers from participating in the auction, thus driving down prices. The
study was done using Pokimon cards, but the finding, presumably, could hold
true for anything: Beanie Babies, lawn gnomes, you name it.

-- 
-----------------
R. A. Hettinga <mailto: rah at ibuc.com>
The Internet Bearer Underwriting Corporation <http://www.ibuc.com/>
44 Farquhar Street, Boston, MA 02131 USA
"... however it may deserve respect for its usefulness and antiquity,
[predicting the end of the world] has not been found agreeable to
experience." -- Edward Gibbon, 'Decline and Fall of the Roman Empire'