Populist Industrial Policy

The myth that small business creates most jobs serves both liberals and conservatives.


Just about everybody likes small business—liberals and conservatives, Democrats and Republicans, even aging Chinese Communists. Bill Clinton frequently says good things about small business to prove he's a "different kind of Democrat."

He even appealed to small companies in his big health-care speech: "These rising [health-care] costs are a special nightmare for our small businesses—the engine of our entrepreneurship and our job creation in America today."

Clinton's opponents, of course, argue that the president's policies are the nightmare. Responding to the State of the Union address, House Minority Leader Bob Michel said, "We agree with the president that we have to put more people to work, but remember this: 80 to 85 percent of the new jobs in this country are created by small business. So the climate for starting and expanding businesses must be enhanced with tax incentives and deregulation, rather than imposing higher taxes and more governmental mandates."

Small business is, apparently, the opposite of the weather: Everybody praises it, and everybody does something about it. But all this posturing is based on bad economics and worse politics. Contrary to endlessly repeated conventional wisdom, small companies do not account for the vast majority of new jobs.

That notion stems from the work of David Birch, a former MIT researcher who now runs a consulting firm called Cognetics. In the 1980s, Birch claimed to show that most new jobs came from small companies. His findings were trumpeted by small-business advocates, notably the Small Business Administration and my former employer, Inc. magazine. It seemed impolite to subject Birch's research to normal scientific checking.

But Birch has now recanted. He says, "The relative role of smaller firms in generating jobs varies enormously from time to time and place to place….Most small-firm job creation occurs within a relatively few firms—the Gazelles." These "Gazelles" are, quite simply, high-growth companies. That growing companies hire more people than non-growing companies is hardly surprising. The "Gazelles," says Birch, represent every sector of the economy and are extremely unstable.

As a celebration of a dynamic, entrepreneurial economy, Birch's vision is appealing. It holds up on anecdotal grounds. Birch cites such Gazelle successes as AST Research and Federal Express. But his research has absolutely no predictive value. You identify a Gazelle by looking at its past growth, not predicting its future prospects. The implication of Birch's research is that no one, including David Birch, knows where new jobs will come from.

That is also the conclusion of a trio of more-academic researchers, economists Steven J. Davis of the University of Chicago, John Haltiwanger of the University of Maryland, and Scott Schuh of the Federal Reserve Board. They identify the analytical fallacies that lead people to see job-creation patterns where none exist. And they have done careful empirical research on manufacturing jobs. The upshot: "In a nutshell, net job creation behavior in the U.S. manufacturing sector exhibits no strong or simple relationship to employer size."

Big companies account for the largest number of newly created (and newly destroyed) manufacturing jobs, because they have the most jobs to begin with. Small companies create and destroy jobs at a higher rate, since they start from a smaller base. Some small firms grow, and some shrink. Some big firms grow, and some shrink. Size alone tells you nothing.

None of this is surprising when you think about it. If David Birch or the SBA knew which companies would prosper and add workers, they'd be making a killing in venture capital. And even allowing for some inefficiency in the capital markets, we'd see a lot more unsubsidized investment pouring into little companies.

Plus, we've all known for years that high-growth startups have little in common with mom-and-pop dry cleaners. Talking about "small business" as an analytical category is silly on its face.

But it makes powerful sense politically, which is why the job-creation myth has persisted and why it has pervaded our political rhetoric. Both liberals and conservatives need the myth, though for radically different reasons.

Liberals, among them many of my friends and former Inc. colleagues, need a way to come to terms with capitalism. Repulsed by corporate bureaucracy, they can't defend General Motors, or even Microsoft. But they can get behind the Birchean vision of little companies struggling mightily to make up for all the layoffs from the aging behemoths. The myth neatly divides the world into two camps—big business, which they already knew was evil and now discover is useless as well, and small business, which is virtuous and productive.

The problem then arises of what to do about small businesses that succeed: What about Microsoft and MCI and Wal-Mart? What about those Gazelles? This isn't just a matter of cultural attitudes. A lot of government policy is friendly toward small companies as long as they stay small. If you don't hire more than 49 employees, you don't have to give 60-days notice of plant closings; you avoid mandatory family and medical leave; you can do business with the government without filing affirmative-action plans; and the Clinton health-care plan will give you subsidies. Government policy punishes companies that grow.

It does so because of the way conservatives use small business in the policy debate. It is hard to stand up and say that some nice-sounding regulation is a bad idea or none of the government's business. Phil Gramm isn't going to stick up for large corporations. Instead, he tells stories about his constituent Dickie Flatt, who owns a small printing business and gets ink under his fingernails. Flatt's company is populist. General Electric isn't.

There's nothing inherently wrong with this rhetorical strategy. Small companies are a lot easier to picture. Their budgets operate on a scale closer to the household than the nation-state. Lots of people can identify with Dickie Flatt; not many can imagine running General Electric.

But most populist politicians don't use small companies as examples. They use them as special cases. They don't say, "Look what this regulation does to the corner dry cleaner. Imagine that effect multiplied over all of IBM." Instead, they suggest that small companies are better, more precious, more deserving of government favor than big companies.

By treating small business as special, they discredit the notion that government should be neutral. They make enterprises unequal before the law. Of course, as Steven Davis observed at a recent economics conference, it is not better to apply a bad policy more thoroughly. Small-business exemptions do reduce the drag on the economy created by various wealth-consuming regulations.

But over the long term, the political dynamic created by special exemptions only encourages more regulation. It buys off the constituencies that could make that regulation politically costly.

The politics of small-business favoritism has sacrificed truth to myth. And it has encouraged the hubris of would-be planners. Telling government officials that you can identify what size companies will create jobs doesn't encourage entrepreneurship. It supports the old political game of picking winners. It is industrial policy with a populist face.