Another American Myth -- Small Businesses Create Most Jobs

I quote from an article by Alan D. Viard and Amy Roden from AEI Online (references excluded):

The most common argument for preferential treatment of small business--its uniquely powerful role in job creation--does not stand up under scrutiny. To begin, the statement that small firms create the majority of jobs does not imply that they play a unique role in job creation. No matter how jobs are distributed across firm sizes, one can always find some threshold size such that firms smaller than that size account for a majority of jobs. 

Careful statistical studies do not assign any special role to small firms. Instead, such studies have largely reaffirmed Gibrat's Law, formulated by Robert Gibrat in 1931, which holds that there is no relationship between a firm's employment size and its growth rate of employment. As Federal Reserve Bank of Cleveland economists Ben R. Craig and James B. Thomson and University of North Carolina professor William E. Jackson III noted in 2004, "economic studies find little evidence to support" the claim that small businesses are an important source of employment growth. In their authoritative book on job creation and destruction, Steven J. Davis of the University of Chicago and AEI, John C. Haltiwanger of the University of Maryland, and Scott Schuh of the Federal Reserve Bank of Boston dismiss what they call the "small business job-creation myth," concluding that "conventional wisdom about the job-creating prowess of small businesses rests on statistical fallacies and misleading interpretations of the data." 

Davis, Haltiwanger, and Schuh, along with other authors, emphasize the "regression fallacy" that arises from temporary changes in firm employment. Most of the studies computing job gains between two dates classify firms as big or small based on their size at the earlier date, a practice that inflates job gains at small firms. When a firm that has temporarily become small due to a recent setback regains its former position, a job gain for a small firm is recorded; when a firm that has temporarily become large due to a recent expansion falls back to its prior position, a job loss at a large firm is recorded. Opposite results are obtained if the firms are classified as large or small based on their employment at the later date.

It is also important to look at net, rather than gross, job gains. For any category of firms and time period, gross job creation is the sum of job gains at those firms that added jobs during the period. Conversely, gross job destruction is the sum of job losses at those firms that reduced jobs during the period. Net job creation is equal to gross job creation minus gross job destruction. As Davis, Haltiwanger, and Schuh--and others--document, smaller firms have proportionately higher rates of gross job creation than larger firms. Unfortunately, small firms' high gross job creation is offset by high gross job destruction. Davis, Haltiwanger, and Schuh conclude that "[i]n a nutshell, net job creation in the U.S. manufacturing sector exhibits no strong or simple relationship to employer size."

Another issue of importance in measuring net job gains is churn. When the mom-and-pop shop in the local strip mall goes out of business, it is often replaced by another mom-and-pop shop with the same number of employees. There is an obvious bias if one only records the job creation in the new company and fails to record the job loss from the failed company. It is great that the number of jobs in the neighborhood does not go down if the empty space in the strip mall is filled, but the employment effect is churn rather than job creation.

A lot of political hyperbole is focused on the merits of small companies and almost none on the merits of large companies. Moreover, many government policies are designed to give small firms breaks, either tax financing or actual government funding.

Some such breaks make sense to me. It makes sense as Justin Lin suggests in his book, New Structural Economics, to provide government incentives to firms that are struggling to create new competitive advantages for a country (with an open, free market economy with a supportive government); some of those companies will be small, but some will be large (e.g. Apple, Google, Amazon). It may also make sense to provide aid for small firms in hard times since they are likely not to have the deep pockets of large firms.

I suspect that much of the answer to our pro small business policies lies in the effectiveness of small business lobbies that espouse very conservative policies. The National Federation of Independent Business (NFIB) and local Chambers of Commerce are very influential in Washington, both very conservative, and can reasonably argue that they represent small business owners who tend to oppose taxes, minimum wage increases and mandatory health care insurance for employees. According to the Boston Review:

While NFIB is relatively small—600,000 members compared to AARP’s 38 million—it is remarkably powerful. Fortune has frequently named it the most powerful business lobby in Washington, and in 2005 Republican members of Congress identified it as the most powerful congressional lobby.

While, based on my last post we can not expect Republicans to read the literature and change their views, perhaps the Democrats and Independents will.