April 14, 1999
 

          Computer Age Gains Respect of
          Economists

          By STEVE LOHR

            In a nation of technophiles, where Internet millionaires are minted
             daily, it seems heresy to question the economic payoff from
          information technology -- the billions upon billions spent each year by
          companies and households on everything from computers to software to
          cell phones.

          But for more than a decade, most of the nation's leading economists have
          been heretics. They have not been much impressed by the high-tech
          dogma -- embraced by corporate executives, business school professors
          and Wall Street alike -- that regards the transformation of the economy
          through the magic of information technology as a self-evident truth.

          "You can see the Computer Age everywhere," Robert Solow, a Nobel
          prizewinner from the Massachusetts Institute of Technology wrote a few
          years ago, "but in the productivity statistics."

          For years, even as the computer revolutionized the workplace,
          productivity -- the output of goods and services per worker -- stagnated,
          barely advancing 1 percent a year. So it is easy to see how Solow's pithy
          comment became the favorite punchline of the economic naysayers.

          Yet today, even renowned skeptics on the subject of technology's
          contribution to the economy, like Solow, are having second thoughts.
          Productivity growth has picked up, starting in 1996, capped by a surge in
          the second half of last year, after eight years of economic expansion. That
          has drawn attention because past upward swings in productivity typically
          occurred early in a recovery as economic activity rebounded. Once
          companies increased hiring, it slowed again.

                                     But something seems fundamentally
                                     different this time, something
                                     apparently having a lot to do with the
                                     increased speed and efficiency that
                                     the Internet and other pervasive
                                     information-technology advances are
                                     bringing to the mundane day-to-day
                                     tasks of millions of businesses.

                                     The question, posed by economists,
                                     is whether the higher productivity
                                     growth, averaging about 2 percent in
          the last three years, roughly double the pace from 1973 to 1995, is the
          long-awaited confirmation that the nation's steadily rising investment in
          computers and communications is finally paying off. The evidence is
          starting to point in that direction.

          "My beliefs are shifting on this subject," said Solow. "I am still far from
          certain. But the story always was that it took a long time for people to
          use information technology and truly become more efficient. That story
          sounds a lot more convincing today than it did a year or two ago."

          Another pillar in the pessimist camp was Daniel Sichel, an economist at
          the Federal Reserve. His work, along with another Fed economist,
          Stephen Oliner, in 1994, and on his own in 1997, found that computers
          contributed little to productivity growth. But recently, Sichel ran similar
          calculations for the last few years and came to a different conclusion.

          In a paper that has just been published in the quarterly "Business
          Economics," Sichel wrote that his new work points to "a striking step up
          in the contribution of computers to output growth." And the nation's
          improved productivity performance, he noted, is "raising the possibility
          that businesses are finally reaping the benefits of information technology."

          The impact of information technology on the economy is more than an
          academic debate. If, as some experts assert, the technology dividend is a
          key reason for the nation's extraordinary run of high growth, rising wages
          and low inflation, there are significant policy implications. If the recent
          gains are not just a temporary blip, it suggests that the Federal Reserve
          can be less fearful of inflation and keep interest rates stable rather than be
          forced to raise them to cool off what would otherwise be considered an
          overheated economy.

          Indeed, the Fed chairman, Alan Greenspan, and the other Fed governors
          are scheduled to hear presentations on information technology's effect on
          the economy from several academics during a private meeting in
          Washington on Thursday.

          Greenspan, for one, seems to believe a fundamental change is under way.
          He told Congress early this year that the economy was enjoying "higher,
          technology-driven productivity growth."

          The Fed governors will hear a forceful case for technological optimism
          from Erik Brynjolfsson, an associate professor at the MIT Sloan School
          of Management.

          Brynjolfsson asserts that the economic value of speed, quality
          improvements, customer service and new products are often not
          captured by government statistics. "These are the competitive advantages
          of information technology," he said. "We need a broader definition of
          output in this new economy, which goes beyond the industrial-era
          concept of widgets coming off the assembly line."

          The government, after years of defending its figures, conceded two
          weeks ago that productivity growth may be understated. The core of the
          problem, government economists say, is the increasingly complex
          challenge of defining and measuring output in much of the economy's
          fast-growing service sector, which includes the vast reaches of banking,
          finance, health care and education.

          According to the official statistics, a bank today is only about 80 percent
          as productive as a bank in 1977. Yet that seems to take scant account
          of, say, 24-hour automated teller machines, which clearly benefit
          customers who no longer have to wait in lines to be served by human
          tellers during regular "bankers' hours."

          Edwin Dean, chief of the productivity division of the Bureau of Labor
          Statistics, wrote in a new research paper that the agency was increasingly
          concerned that its measurements did not "fully reflect changes in the
          quality of goods and services" or "capture the full impact of new
          technology on economic performance."

          Still, the government's methods of measurement will not be overhauled
          anytime soon. "These are tough, tough questions and we are not going to
          get instant solutions," Dean explained in an interview.

          American corporations long ago made up their minds, voting for
          technology with their dollars. Investment in information technology --
          computing and telecommunications gear -- has quadrupled over the last
          decade, rising as a share of all business spending on equipment from 29
          percent to 53 percent, according to the Commerce Department. And
          that is only the hardware. There have been similar surges in corporate
          spending on software, consulting, technical support and training related to
          the field.

          "The payoff from information technology is unquestionably there with
          individual companies and we're seeing it over and over again," said
          Chuck Rieger, a senior consultant at IBM's services division.

          Of course, anecdotal evidence from individual companies is no proof of
          broad-based benefits in an $8.5-trillion economy. But what many experts
          find encouraging is that the rapid introduction of low-cost Internet
          technology means most companies can now afford to set up electronic
          links with customers and suppliers. For example, a recent survey of
          2,500 manufacturing companies, conducted by
          PricewaterhouseCoopers, found that the number of factories with
          Internet links to customers and suppliers doubled last year.

          At more and more companies, these Internet-based networks are
          already streamlining the mundane chores of business life like invoicing,
          purchasing and inventory control. This is not the glamorous side of
          Internet commerce, occupied by Amazon.com and others selling
          consumer products. Yet if a technology dividend in productivity is at
          hand, the place to look is in the back offices of business. "That is where it
          will be," Solow, the MIT economists, said, "in the wholesale automation
          of corporate transactions."

          This business-to-business commerce over the Internet is projected to
          jump from $48 billion in 1998 to $1.5 trillion by 2003, according to
          Forrester Research Inc. During the same period, the research firm
          estimates that consumer sales over Internet will rise from $3.9 billion to
          $108 billion.

          The service sector of the economy is where productivity gains appear to
          have been especially sluggish and where experts are looking most closely
          for evidence of an efficiency payoff from technology.

          In Chicago, Michael Rushmore, a banker, speaks of how Internet
          computing has "fundamentally changed the way we do business" over the
          last three or four years. Take the way corporate loans are syndicated
          among many banks, notes Rushmore, a managing director of
          Nationsbanc Montgomery Securities, the securities arm of BankAmerica
          Corp.

          Until about two years ago, syndicating a large corporate bank loan meant
          distributing a lengthy offering document, often running more than 200
          pages, to 50 to 100 banks. It was, Rushmore recalled, a nightmarish,
          inefficient process that involved waves of overnight mail, constant faxing
          and armies of messengers.

          Today, much of that process is handled over the Internet on bank Web
          sites that other banks tap into to read and download the offering
          document, ask questions and exchange views. Rushmore estimates that
          the Internet-based system trims 25 percent from the time it takes to close
          a deal, not just improving the ease of the transaction but also saving an
          immense amount of hours of work.

          About a year ago Booz Allen & Hamilton began using the Internet to bill
          several federal agencies that are its clients. Booz Allen estimates that it
          has saved $150,000 a year by eliminating the paper handling on its $10
          million in monthly billings to the government. The greater speed and
          efficiency of the electronic billing also means that the consulting firm is
          being paid 30 percent, or six days, faster than before.

          "Getting that money into the bank much more quickly is probably the
          biggest benefit," said Mark Arnsberger, an assistant controller for Booz
          Allen & Hamilton.

          The rapid spread of Internet-based computing, experts say, promises to
          compress the time it takes for any new technology to enhance economic
          welfare in general. The classic study of the phenomenon, "The Dynamo
          and the Computer: An Historical Perspective on the Modern Productivity
          Paradox," by Paul David, an economic historian at Stanford University,
          was published in 1990.

          The electric motor, David noted, was introduced in the early 1880s but
          did not generate discernible productivity gains until the 1920s. It took
          that long, he wrote, not only for the technology to be widely distributed
          but also for businesses to reorganize work around the industrial
          productionline, the efficiency breakthrough of its day.

          "The process takes longer than people think, but I still believe that we will
          get a revival of productivity growth led by the spread of computing,"
          David said.

          His is a misplaced faith, according to the dwindling band of
          techno-pessimists whose own beliefs remain unshaken. Sure, they
          concede, there has been surprisingly strong productivity growth for the
          last three years. Could this represent a break in the trend? Possibly, they
          grudgingly admit, but only a tiny shift at best, they insist.

          The real problem, they explain, lies in the composition of the nation's vast
          service economy. More than half of all white-collar workers are what
          they term "knowledge workers" -- managers, executives and
          professionals like doctors, lawyers, teachers, even economists.

          "The work they do does not lend itself to technology-driven
          improvements in productivity, and any gains are really difficult to eke out
          and are glacial," said Stephen Roach, chief economist at Morgan Stanley
          Dean Witter. "Paul David's electrical motor has nothing to do with the
          knowledge-intensive process of work in a service economy."

          The real technology cynic at the Fed meeting on Thursday will be Paul
          Strassmann, a former chief information officer of Xerox and the
          Pentagon. Strassmann, author of "The Squandered Computer," published
          in 1997, believes that corporate America's spending spree on information
          technology amounts to an "economic arms race," fueled by misguided
          management theories.

          The recent improvement in productivity, according to Strassmann, is
          mainly attributable to the lower cost of capital because of low interest
          rates. His summary view, though at odds with those of technology
          optimists like Brynjolfsson of MIT, may also be received warmly by the
          Fed.

          "The explanation for the productivity improvement is interest rates, not
          information technology," Strassmann said. "The hero here is not Bill
          Gates. It's Alan Greenspan."

          Yet even Strassmann finds the technology undeniably useful, if not a
          productivity elixir. When asked a detailed question, he replied, "Just look
          it up on my Web site. It's a lot more efficient that way."