It's widely believed that small businesses provide most of the nation's job growth.

But how true is that common belief?

According to the Small Business Administration, small business "establishments" created about three-quarters of all of the net new jobs between 1994 and 1995, when the boundary between small and large is set at 500 employees.

To get this result, SBA subtracts job losses at shrinking or shuttered businesses from gains at growing ones, using a comprehensive database of U.S. companies.

That's the kind of evidence small-business advocates use to make their case.

But there are other ways of counting the job numbers that show that larger businesses deserve more credit than they customarily receive.

Sorting out this issue seems to require an advanced degree in statistics.

For some time, David Hirschberg, a onetime SBA statistician, has been dueling with his former employer over the small-versus-large question. Hirschberg, author of "The Job Generation Controversy: The Economic Myth of Small Business," argues that the SBA's studies are biased to exaggerate the contribution small firms make to job growth. Because so many people believe what he calls the "small-business mantra," proposals to increase the minimum wage or require that small businesses meet environmental or workplace regulations imposed on bigger companies are undermined, he says.

Hirschberg notes that the SBA data customarily focus on business "establishments," meaning offices with employees. A business "firm," in SBA terms, is more inclusive, covering parent companies that could be operating many establishments at different locations.

The distinction turns out to be significant. When data on firms instead of establishments are used, big-business job gains look a good deal more impressive.

According to the SBA, business firms with fewer than 500 employees in 1994 created about 60 percent of the job growth between 1994 and 1995. Firms with 500 or more employees produced the rest of the net job growth. That's still a noteworthy performance by small firms -- but clearly they must share the applause with larger firms.

A city mayor trying to boost the local economy might care more about establishment growth. A legislator wrestling with business regulations might well focus on what's happening to firms.

And there's another fine point in the argument.

One approach traditionally used by the SBA is to sort companies by size based on the beginning year of a study. An establishment with 100 employees in 1994 was considered small, so if it gained 500 jobs during the year, all of that growth was chalked up in the small-business column, even though by the end of 1995 the company had grown into the large-business category. That's misleading, Hirschberg contends.

To get around the problem, the SBA has begun analyzing job growth a second way, using an establishment's average size during a growth period, notes SBA analyst Alicia Robb. In the case above, the establishment's average size over 1994-95 was 300 employees, putting it in the small camp at the beginning and end of the period.

When the average-size approach is used, only 56 percent of net job growth in 1994-95 was produced by "small" firms, not three-quarters, Robb said.

Again, with this approach, small-firm employment grew by 4.4 percent in 1994-95, compared with 3 percent for larger firms. The small firms still led the way, but not by an overwhelming margin.

The SBA data also show that smaller businesses experience much higher birth and death rates than larger ones.

During the 1994-95 period, successful establishments with 1 to 99 employees added 10.9 million workers overall. But other firms in that category, ones that were shrinking or were forced out of business, eliminated 7.9 million jobs. That's a high churn rate.

Which may suggest a reason for extending a helping hand to small business: They have a much harder life than their larger cousins.

Information on how to obtain the SBA data is available at www.sba.gov/advo/stats/leem_avail.html.

Big Numbers

The percentage of net new jobs created between 1994 and 1995 by small business* equaled 78 percent:

Job growth at smaller* firms in 1994-95 equaled 5.1%

Job growth at bigger** firms in 1994-95 equaled 2.1%

* 499 employees or fewer at the start of the period in 1994.

** 500 employees or more at the start of the period in 1994.

SOURCE: Small Business Administration