The QWERTY myth
Economists adore a nice case of market failure. The dogged
persistence of the standard typewriter keyboard, held to be a
technological anachronism, is a great favourite. Yet the charges
against QWERTY were long ago disproved
AT A conference attended the other day by your reporter, a
distinguished academic economist (who had better remain nameless)
cited the "QWERTY" layout of the standard typewriter keyboard as a
clear example of how markets "can make mistakes". It may have been
the millionth such reference. Many a textbook cites this case as
proof of a certain kind of market failure-that associated with the
adoption and locking-in of a bad standard. For years, if you cited an
example of a "pure public good" (another kind of market failure), it
had to be a lighthouse. If you needed a case of "positive
externalities" (yet another), you would very likely go for
beekeeping. In its field, QWERTY has achieved the same iconic
eminence.
Which is only apt, because the tale of QWERTY is a myth-just like
those other two cases. More than 25 years ago, Ronald Coase, a Nobel
laureate, showed that when lighthouses were first built in Britain
they were provided by private enterprise; tolls were collected when
ships reached port. So lighthouses are not pure public goods. At
about the same time Steven Cheung examined beekeeping and
apple-growing in the state of Washington. He found that apple-growers
paid beekeepers for their bees' pollinating endeavours; those
services were not, in fact, an unpriced "externality".
Mr Cheung's article was called "The Fable of the Bees". That is why
the paper on QWERTY published a mere nine years ago by Stan Liebowitz
of the University of Texas at Dallas and Stephen Margolis of the
University of California, Los Angeles, was called "The Fable of the
Keys". It tells you plenty about the history of the typewriter-but
what every economist should have concluded from it was that another
example of "locking in" had better be found. QWERTY simply will not
do.
The myth goes roughly as follows. The QWERTY design (patented by
Christopher Sholes in 1868 and sold to Remington in 1873) aimed to
solve a mechanical problem of early typewriters. When certain
combinations of keys were struck quickly, the type bars often jammed.
To avoid this, the QWERTY layout put the keys most likely to be hit
in rapid succession on opposite sides. This made the keyboard slow,
the story goes, but that was the idea. A different layout, which had
been patented by August Dvorak in 1936, was shown to be much faster.
Yet the Dvorak layout has never been widely adopted, even though
(with electric typewriters and then PCs) the anti-jamming rationale
for QWERTY has been defunct for years.
Why has the bad design endured? Because, the story continues, that
first inefficient standard became locked in. Even though the costs of
new keyboards and retraining for the Dvorak layout would be quickly
recovered, typists won't switch unless others do so as well;
likewise, the keyboard manufacturers refuse to move first. There is a
co-ordination failure-that is, a market failure.
Carriage return
A fine tale, but largely fiction. The paper by Messrs Liebowitz and
Margolis shows, in the first place, that the first evidence
supporting claims of Dvorak's superiority was extremely thin. The
main study was carried out by the United States Navy in 1944
(doubtless a time when every second counted in the typing pools). The
speed of 14 typists retrained on Dvorak was compared with the speed
of 18 given supplementary training on QWERTY. The Dvorak typists did
better-but it is impossible to say from the official report whether
the experiment was properly controlled. There are a variety of
oddities and possible biases: all of them, it so happens, seeming to
favour Dvorak.
But then it turns out-something else the report forgot to
mention-that the experiments were conducted by one
Lieutenant-Commander August Dvorak, the navy's top time-and-motion
man, and owner of the Dvorak layout patent.
In 1956 a carefully designed study by the General Services
Administration found that QWERTY typists were about as fast as Dvorak
typists, or faster. Interest in Dvorak among companies and government
agencies had lately been increasing, but it came to an end with that
finding. Since then, as "The Fable of the Keys" explains, there have
been a variety of other experiments and studies. They find that
neither design of keyboard has a clear advantage over the other.
Ergonomists point out that QWERTY's bad points (such as unbalanced
loads on left and right hand; excess loading on the top row) are
outweighed by presumably accidental benefits (notably, that
alternating hand sequences make for speedier typing).
Which is all very interesting, but the point is this: if you have
learned to type on a QWERTY keyboard, the cost of retraining for
Dvorak (however modest) is not worth paying. This implies, in turn,
that the QWERTY standard is efficient. There is no market failure.
Undaunted by the resilience of the QWERTY myth, Messrs Liebowitz and
Margolis devote the first chapter of a forthcoming book on supposed
market failures, technology and Microsoft to the fable of the keys.
They go on to argue that the fashion for finding new kinds of market
failure (lock in, path dependence, network effects, and so on) in
high-tech industries is misconceived: QWERTY is just one example
among many. Their views on this larger question are for another
article. For now, merely note that the failure illustrated by the
QWERTY myth has more to do with the study of economics than with
markets. For some reason, economists seem to adopt bogus anecdotal
histories and then get locked in.
All the articles were published in The Journal of Law and Economics.
"The Lighthouse in Economics" by Ronald Coase was in the issue of
October 1974, and "The Fable of the Bees: An Economic Investigation"
by Steven Cheung in April 1973. "The Fable of the Keys" by Stan
Liebowitz and Stephen Margolis appeared in April 1990. It can be read
on the Internet at http://wwwpub.utdallas.edu/~liebowit/keys1.html