Commentary: Why The Numbers Miss The PointChristopher Farrell
No matter where you look, computers, software, and telecommunications are transforming the way Americans work. In 1966, the best transatlantic phone cable could carry simultaneously 138 conversations between Europe and North America. Today, fiber-optic cables carry thousands of times more traffic. At least half the workforce now uses computers on the job. Information technologies seem to be paying off big. In the 1990s, productivity has been rising at a 1.9% annual rate, double the pace of the previous two decades.
Yet now, largely because of a chain reaction set off by an adjustment in the method used to calculate computer prices, government statisticians are sharply lowering their estimates of productivity growth to a 1.3% pace in the '90s. After adjusting for business cycle fluctuations, that means productivity grew at a pace little better than that of the '70s and '80s. In other words, despite nearly two decades of restructuring, investing huge sums in high-tech equipment, and the remarkable resurgence in U.S. corporate profits and global market share, the productivity improvement amounts to a big fat zero.
NEW RULES. Hold off on the pessimism, please--and not because these are preliminary numbers that most likely will be revised. The reason is more fundamental. What matters is not how many computer boxes are coming off the assembly lines, but the extent to which information technology is rewriting the rules of business, spurring technological progress, and creating economic wealth, says Timothy F. Bresnahan, economist at Stanford University. As the Industrial Era gives way to the Information Age, new products, new ideas, and new technologies are transforming the economy in ways the statistics aren't capturing.
The signs of a seismic shift are all around us. Supermarkets and bar codes. Banks and ATM machines. Lotus Notes and groupware. In industry after industry, information technologies have inspired innovations, raised efficiencies, lowered prices, and hiked profits. Professor Erik Brynjolfsson and graduate student Lorin M. Hitt of Massachusetts Institute of Technology surveyed almost 400 large companies to measure the effect of investment in information technology on output per employee. They found evidence that the return on investment in information systems could exceed 50% and that such systems have led to improvements in product quality, product variety, and customer service. "Many of the benefits of information technologies are not captured by the official measures and don't show up in GDP," says Brynjolfsson.
Investment is an engine of growth, and the government's numbers miss a big chunk of high-tech investment. For example, business purchases of software aren't included under investment (except for software sold with computer hardware). Instead, they're treated like pencils and paper clips. Yet corporate purchases of package software rose from $7.4 billion in 1983 to $51 billion in 1994, according to Sentry Market Research in Westborough, Mass.
The mismeasurement goes further. With the exception of computers, the latest figures only pick up a fraction of the price and quality changes in most high-tech products. The official data show prices of business and home telecom equipment are up by 4.4% since 1991. Yet prices have come down on everything from central office switching equipment to the equipment needed to run computer networks. For instance, average fax machine prices dropped 28% over the past four years even as quality advanced, according to Personal Technology Research, a consulting firm in Waltham, Mass. The upshot: By missing these price drops, the government understates the widespread use of new technologies and, ultimately, GDP.
OFF THE GAUGES. The productivity improvement that is so evident in Corporate America is not showing up in the numbers because much of it is coming in the information and services sectors of the economy, where productivity is daunting to gauge. In the early post-World War II era, about half the economy was easily measured; by 1990, the fraction of the economy for which the productivity numbers were reasonable was less than one-third, according to Zvi Griliches, economist at Harvard University.
Economic measures capture run-of-the-mill changes in activity. But they have always fallen short in dealing with paradigm shifts. We are living in a new era driven by information technologies. The U.S. is a lot stronger than the government numbers are now telling us.