Critics say the U.S. high-technology industry is threatened by too much entrepreneurship. Statistics suggest there's not enough

There is a stereotype of high-tech companies -- particularly the venture capital-backed, glamorously "entrepreneurial" sort.

In the stereotypical characterization, the average high-tech business is just one of a thousand small companies that spin off from universities or larger companies to capitalize on some sort of breakthrough -- a new drug or a new chip design, say. It's a risky business, the stereotype suggests. A few high-tech companies make it, and make it big. The rest quickly fade away, most dying at an early age. The companies that do survive go through steep ups and downs as they gain (or lose) their competitive edge. Or so goes the stereotype, anyway.

Some have argued recently that the description is accurate and the process corrosive. They say that the thousands of start-ups steal vital talent from larger, more viable companies, making it virtually impossible for the large companies to compete effectively with their massive Japanese counterparts.

We cannot, it is claimed, compete effectively in world markets if we permit our technology base to fracture into thousands of inefficient operations. We must institute a way to stamp out this fracturing by sharply increasing the capital-gains tax -- thereby erasing the investment appeal of speculative and irresponsible venture formation.

What are the facts about high-tech companies?

Well, one aspect of the stereotype holds up. High-tech businesses do tend to get bigger faster than most other companies. By the end of their first five years, high-tech companies are nearly twice as likely as other manufacturers (14.4% versus 7.7%) to have more than 50 employees. Only 51.5% of five-year-old high-tech companies employ fewer than 10 people; 65.4% of other manufacturers do. It's little wonder that venture capitalists find them attractive.

But that's where the stereotype fails. There are other reasons the venture capitalists love high tech: it's neither particularly risky nor particularly volatile. With the exception of the very youngest, high-tech companies are systematically the manufacturing enterprises least likely to go out of business (see chart, "Odds of Closing by Age and Type of Company," page 2). And even the youngest ones are only slightly more vulnerable.

Employees in high-tech companies are only marginally more likely to be bounced around than are people working for other manufacturers. We read so often about high-tech expansions and layoffs that it's not difficult to conclude it's a boom-or-bust part of the economy. Not particularly. The whole economy experiences big ups and downs, and high tech is not particularly unusual in this regard.

Most surprising to the uninitiated is the prevalence of big companies in high technology. Only 4% of the people who work in high tech work for the small, struggling companies so often romanticized (versus 10% for other manufacturers -- see chart, "Employment by Size and Type of Company," page 2). Conversely, almost three-quarters of all high-tech employees work for large companies with more than 500 employees (versus 51% for other manufacturers). High tech is one of the most concentrated sectors of the U.S. economy -- only slightly less concentrated in large companies than are the makers of cars and planes (78%, as opposed to 73% for high tech), and significantly more concentrated than the steel business (66% versus 73%). Virtually no one works for the smaller spin-offs struggling in garages.

In sum, high tech looks much more like a mature sector than an embryonic one. It has relatively low failure rates, average volatility, and employs an extremely high percentage of its workers in a fairly few large companies.

In this context, it is difficult -- if not impossible -- to understand the argument that entrepreneurship is killing our technology base by draining vast amounts of talent into thousands of smaller, inefficient enterprises. Very few industry sectors in our economy have a larger share of their workers in large companies and a smaller share in small ones.

The discovery that high technology is mature and ordinary rather than boisterous and extraordinary does not, and should not, diminish its importance to the economy. High tech is to today's economy what the steam engine was to the nineteenth century's. Very few jobs were created by manufacturing steam engines. Very few small companies manufactured them. Their importance flowed not from their construction, but from what they made possible after they were constructed.

Similarly, the huge surge of job growth we have experienced in this decade, and are likely to experience in the decades ahead, is based in large measure on what can be done with technology. It is technology's application, not its construction, that fuels the New Economy.

If anything, we should be concerned with how few people work for smaller, innovative technology inventors, not how many. It may be that our technology sector is maturing too fast for its own good, squeezing out the very kind of entrepreneur that developed the technology giants of today from infant companies started as recently as a decade or two ago.

David L. Birch is president of Cognetics Inc., in Cambridge, Mass., and director of MIT's Program on Neighborhood and Regional Change.

Odds of Closing by Age and Type of Company

As the selected examples show, high tech outsurvives other manufacturers. Contrary to stereotypes, high-tech failure rates are low.

% of companies that failed between 1983ñ1987

Age of companies in 1983 (years) High Tech Other manufacturers
0ñ4 31.7% 28.8%
10ñ14 22.9 24.1
20ñ24 16.1 17.5
40ñ49 16.3 18.2
75ñplus 4.2 11.4

Employment by Size and Type of Company

Almost three-quarters of high-tech employees work for large companies. High tech is one of the most concentrated sectors in the U.S. economy.

Company size

(Number of employees in 1987) % of total high-tech work force employed % of total other manufacturing work force employed
1ñ19 3.9% 10.3%
20ñ99 9.4 18.1
100ñ499 13.8 20.8
500ñplus 73.1 50.7