The fiction of a 23% decline

The plural of anecdote, social scientists are fond of saying, is data. For

that reason, it's a positive sign that the focus of the debate about the

impact of procurement reform on small business has shifted from anecdotes

to examining the numbers.

Looking at those numbers, I'm reminded of the old Holiday Inn ad slogan,

"The best surprise is no surprise." For all the sound and fury, the overall

numbers show fewer bumps than the kiddy roller coaster at an amusement park.

In 1997, small-business prime contractors received 22.7 percent of federal

prime contracting dollars. Last year, they received 23.1 percent. The percentage

of dollars going to small firms is down 0.8 percent from its 1995 high of

23.9 percent, but it's still comfortably higher than at the beginning of

the Clinton administration in 1993, when it was 21.9 percent.

Those figures might surprise you, because a recent report by the House

Small Business Committee trumpeted that the number of contracts awarded

to small businesses — subsequent news stories said nothing about dollars — had declined a dramatic 23 percent between 1997 and 1999. Some reports

even stated that the small business "share" had declined 23 percent and

that the culprit for this catastrophe was the rise in contract "bundling,"

the practice of consolidating smaller contracts into larger ones.

The presentation and interpretation of those numbers demonstrates that

no matter how valuable it is to replace anecdotes with numbers, it is also

true that a warning needs to be flashed into any statistics, such as those

in the Federal Procurement Data System (FPDS) that served as a basis for

the Small Business Committee report.

Members of the committee who are concerned about a 23 percent drop can

breathe easier: FPDS data doesn't include credit card transactions.

The number of federal credit card transactions has increased each year.

Therefore, FPDS data, which excludes those actions, has become a poorer

measure of the actual number of government contracts. In all, the number

of contract actions reported into FPDS has shrunk by half since 1993, from

19.6 million to 10.4 million. If every contracting officer in the government

had spent their time doing nothing but "bundle" contracts since 1993, they

couldn't have taken 9 million transactions out of the system.

The transactions missing in the reported data haven't disappeared — they're just not being reported. And if you look at the small-business share

of contract actions of more than $25,000 — the numerator of small-business

actions over the denominator of total actions — it is not down 23 percent

since 1997. It has been stable, going from 45 percent to 44 percent.

Actions of greater than $25,000 are where one should look for any impact

of bundling. When people talk about bundled contracts, they typically mean

jobs worth $100,000 or $1 million or $3 million being consolidated into

jobs worth $100 million.

Surely, there are issues worth discussing. No data exists on the small-business

share of credit card micropurchases less than $2,500, but that share surely

is down somewhat from before credit cards were used. Previously, those purchases,

by law, had to be made from small businesses.

There has been a decline in the small-business share of contract actions

of less than $25,000 — from 53 percent to 47 percent since 1997 — probably

because some small-dollar orders previously going to local small firms now

go to large national firms through the General Services Administration schedule.

And the percentage of dollars going to 8(a) firms is down a bit since 1997,

though still higher than 1993. At the same time, the overall dollar figures

imply that small firms are making up for any losses in these two areas elsewhere

in the system.

So discussion, yes; hysteria, no. Unfortunately, the apparent dramatic

"23 percent decline" is being used to justify several taxpayer rip-off

bills Congress is considering. More on that in my next column.

—Kelman, administrator of the Office of Federal Procurement Policy from 1993

to 1997, is Weatherhead Professor of Public Management at Harvard's John

F. Kennedy School of Government.