IP: FT: IT 'has failed to fulfil its promise'

[David Farber <farber () cis upenn edu>]

Freedom fantasy: IT 'has failed to fulfil its promise'


WEDNESDAY AUGUST 13 1997


By Victoria Griffith


Back in the 1960s, computer buffs dreamed of the day technology would
make all of us feel like royalty. Robots could be programmed to bring
us coffee, make our beds and fetch the newspaper. We would all be
freed up to follow worthwhile pursuits like furthering our education
and spending more time with our families.


It didn't happen. Yet 30 years later, according to Stephen Roach,
chief economist at the investment bank Morgan Stanley, corporations
are succumbing to the same sort of fantasy by believing that
investment in computers and multimedia advances will boost the
productivity of their workforces.


"The productivity gains of the Information Age are just a myth," says
Roach. "There's not a shred of evidence to show that people are
putting out more because of investments in technology." As evidence,
he compares the US service industry, which he estimates is responsible
for 80-85 per cent of information technology expenditures over the
past decade, to US manufacturing, which has spent just 15-20 per cent
of the total. If computers boosted productivity, Roach argues,
services should show higher output-per-worker advances than
manufacturing. Yet just the opposite is true. Throughout the 1990s,
productivity in manufacturing has risen more than 3 per cent a year on
average. In services, average annual gains have been less than 1 per
cent.


Another indication is productivity growth in the seven richest nations
over the past three decades. Since these countries have invested most
heavily in information technology, it instinctively follows that their
productivity should be accelerating. Yet productivity growth has
actually declined in these countries, from an average of 4.5 per cent
a year in the 1960s to 1.5 per cent a year in the 1990s. In the US,
where output per worker has grown faster than in Europe, Roach
believes most of the gains can be attributed to the longer hours
Americans are spending at the office.


Roach says a failure to re-design the workplace and educate workers
forms the crux of the problem. For computers to bring sustained gains
in productivity, he explains, they must allow employees to concentrate
more on value-added duties such as product development, customer
relations, and corporate strategy. Yet, according to Roach, that
hasn't happened.


"ATM machines meant bank tellers could be directed to more important
duties, but in general, they weren't," says the Morgan Stanley
economist. "Many were simply fired. Others are just doing less work
than they used to, and still others were siphoned off to other areas
where the banks still haven't quite figured out how to use them."


The substitution of computers for human labour has caused a one-off
productivity lift in some sectors, Roach concedes. Workers in the
telecommunications industry boosted their output by about 7 per cent
per hour between 1973 and 1983, for instance. These gains were
accomplished largely by automating routine tasks, then chopping excess
workers from the payroll.


Yet for sustained growth of employee output, says Roach, that same
workforce must perform in a higher sphere. Most unskilled labour is
not capable of that, he argues. "Human intelligence and poor education
places a ceiling on the productivity growth we can get from
automation," he explains. "If the brain power just isn't there in a
segment of the population, take them from routine tasks, and they
become dead weight."


What if technology were used to free up educated rather than unskilled
workers? Some companies are trying to do this by programming computers
with "management" skills. The Italian clothing group Benetton has just
signed an agreement with the computer company International Business
Machines to allow for automated store inventory control. A computer at
headquarters, for instance, would decide just how many red sweaters
the location should have on stock, and when they should be replenished
- tasks once assigned to human shop managers. The motorcycle group
Harley-Davidson is setting up similar technology.


Roach is sceptical that these programmes will boost corporate output
per worker. "There's been a lot of talk about artificial intelligence
over the years, but it's doubtful that computers can ever be made to
think like humans," says the Morgan Stanley economist. What about the
IBM computer Deep Blue's recent win over chess champion Garry
Kasparov? "That's just data processing," says Roach.


Is it possible that companies have failed to see substantial
productivity gains simply because they are spending money on the wrong
sort of information technology? "That may be part of it," says Roach.
"There is certainly a lot wasted on worthless projects. But a bigger
problem is that companies don't accurately price out technology versus
human capital. It's not just comparing someone's salary over a number
of years to the cost of a computer. Once you invest in the computer,
you have to spend a fortune on maintenance, programming, and
everything else that comes with it. All of that can easily add up to
the cost of the workers you got rid of."


Roach worries about the limited flexibility companies will have in the
next recession to reduce their cost structure.


Roach says he is not advising companies to stop investing in
technology. Some automation is necessary, he remarks, just to keep up
with the competition. Future productivity gains from computers cannot
be ruled out. "The Information Age may end up having a big impact on
productivity further down the road," he concedes. "What I'm saying is
that it hasn't yet, and there's no guarantee that it ever will."






=A9 Copyright the Financial Times Limited 1997 "FT" and "Financial
Times" are trademarks of The Financial Times Limited.