Encouraging start-ups is real key

Conventional wisdom argues that small businesses are the engines of job growth. Thus, Republican and Democratic politicians and other policy makers frequently propose incentives to assist small businesses.

While elements of truth lie within this view, the dynamics of the labor markets prove more complex. And recommending policy options requires an understanding of these dynamics.

The bottom line: A few new small businesses will be the engines of job growth; older, established small businesses will not.

Changes in employment mask two underlying dynamics -- job gains and job losses. Employment increases when job gains exceed job losses; employment decreases in the reverse case. In addition, the size of job gains and losses greatly exceed the net change in employment. That is, businesses not only create lots of jobs every month but also destroy lots of jobs as well. The net change in jobs is a much smaller number relative to the number of jobs created and destroyed.

In an August 2010 research report that relies on Business Dynamics Statistics from the Census Bureau, Haltiwanger, Jarmin and Miranda conclude that new, start-up businesses are important for net job creation. This observation supports the conventional wisdom, since most start-up firms are also small firms. But, this observation also counters conventional wisdom, since most small firms are not start-up firms.

Imagine two businesses -- a small, 20-year-old independent clothing store and a brand new small information technology business. The independent clothing store is unlikely to add new employees, whereas the new information technology business with an innovative idea is more likely to grow, and grow quickly. That is, older, established, small firms are less likely to grow than brand new start-up firms. In fact, we expect that a few of these start-up businesses will grow quickly into much larger firms.

The Census Bureau posts interesting on-line data on job creation and job destruction. I consider the 2004 to 2009 data, the latest data available, which I average for the 2004-2006 and 2007-2009 periods. The 2004-2006 period captures a robust U.S. economy with significant job gains whereas the 2007-2009 period captures the financial crisis and Great Recession with significant job losses.

When we categorize firms by size and age, interesting observations emerge. First, consider the 2004-2006 period. Relatively "large" firms (100 employees or more), irrespective of age, generally produced net new jobs -- jobs created exceed jobs destroyed.

Relatively "small" firms (fewer than 100 employees), irrespective of age, generally generated net job losses. The important exception: start-up businesses, regardless of size, (by definition) created net new jobs. But, more importantly, the number of net new jobs produced by start-up firms exceeded the net job losses generated by all other small firms.

Second, consider the 2007-2009 period. Now, all classes of firms, irrespective of age or size, saw lower (higher) net job creation (destruction) than during the 2004-2006 period, not a surprise.

Start-up firms by definition created net new jobs. Now, "smaller" firms (fewer than 500 employees) generally produced net job losses. In addition, these job losses exceeded the net jobs created by start-up firms of the same size, reversing this outcome from the 2004-2006 period.

Only "moderately large" firms (500 to 5,000 employees) showed net job gains, although a lot fewer than during the 2004-2006 period.

In sum, net job gains, on average, came from start-up businesses, even though many start-up businesses failed after their first few years of operation. To a lesser extent, net job gains came from large firms.

At first glance, the policy implications seem reasonably straight forward. If policy makers want to stimulate net job creation, then targeting small businesses with incentives may prove wrongheaded, since existing small businesses generally produce net job losses. Rather, the incentives should focus on start-up businesses, not small businesses as a group. That is, targeting all small businesses with policy incentives will likely lead to an unexpected gift to older, established, small businesses that does not create net new jobs.

Providing incentives to start-up businesses, however, runs the risk of altering the dynamics of the labor market. That is, an ill-conceived, poorly designed incentive program may attract exactly the wrong types of start-up businesses, diminishing the efficiency of the start-up business job machine. Alternatively, stripping away regulatory burdens that impose disincentives on start-up businesses may represent an alternative desirable policy option.

What about Nevada? I collected the same information on net job creation for Nevada. Like the national numbers, the same pattern of reductions in net job creation from the 2004-2006 to the 2007-2009 period also occurred in Nevada.

Consider the 2004-2006 period. "Very small" firms (fewer than 20 employees) saw net job losses that exceeded the net jobs created by start-up firms of the same size. Most "medium" and "large" firms (20 employees or more) generated net job creation, suggesting better job creation in Nevada than in the United States.

Now, consider the 2007-2009 period. "Small and "medium" sized firms (fewer than 1,000 employees) produced net job losses that more than offset the net job creation of start-up firms. Only "very large" firms (1,000 employees or more) saw net job creation, but that job creation was anemic.

In sum, Nevada proved a more vigorous net job creator during the 2004-2006 period than the country overall, but proved a much weaker net job creator during 2007-2009.

Stephen M. Miller is professor and chair of the Department of Economics, College of Business at the University of Nevada, Las Vegas.