Wednesday, May 23, 2012


Or so Paul David and Brian Arthur might think after the results of this experiment in Hong Kong become better known.  The researchers are Tanjim Hossain and John Morgan, who set out to test the QWERTY theory; that an inferior product or technology can lock out a superior one due to network effects.

It's has been known for over 20 years that the story of the supposed inferiority of the standard typewriter keyboard has absolutely no basis in fact (history does matter), and no other examples have been demonstrated to be valid either, thanks to intrepid UCLA PhDs Stan Liebowitz and Steve Margolis.  Now we can add that there isn't any controlled laboratory experiment supporting it either.

Hossain and Morgan report on their 'platform competitions';

When pooled, [experimental] sessions 1–4 constitute 60 iterations of dynamic platform competition. In every instance, the market tipped to a single platform, but in none of these instances was that platform the inferior one. This was true despite giving the inferior platform an initial monopoly advantage that persisted with a full 33 percent of the lifespan of the market. It was true regardless of whether the inferior platform had the higher or lower access fee. It was also true when the inferior platform enjoyed the possibility of a novelty effect from being introduced later. In short, the quest for QWERTY in the lab proved utterly fruitless.
It’s been almost 25 years since the economics of QWERTY first appeared in the American Economic Review. The intellectual impact of David’s observation is undeniable. His seminal article has been cited over 788 times. The risk of QWERTY-type outcomes in platform competition is now accepted as conventional wisdom rather than something counterintuitive. While the QWERTY effect is certainly an interesting theoretical possibility, the dearth of examples of the phenomenon, both in the field and now in the lab, leads us to conclude that the danger lies more in the minds of theorists than in the reality of the marketplace.
Yes, you've read that correctly, one of the most cited journal articles in economics is almost certainly full of it.  Bunk.

No comments:

Post a Comment