Job creation is currently the Holy Grail for Washington policymakers. In order to craft better job policies, it is valuable to understand when, where, and by whom jobs are created. Rigorous data analysis tells us that start-up firms are disproportionate job creators and that new firms tend to appear in cities with smaller incumbent firms.

Policymakers should keep future job-creating start-ups in mind whenever they hear an argument to protect the economic status quo.

Job Growth Through Churning

In 2002, the U.S. Census compiled data on businesses going back to 1976 into the Longitudinal Business Database.[1] By virtue of size, firms with more than 500 employees provide a large share of the jobs in the private nonfarm economy: 45 percent.[2] Generally, larger sectors and industries see more jobs created as well as more jobs destroyed. The churning is part of a healthy economy, as long as gross jobs created exceeds gross jobs destroyed. From 1975 to 2005, economists John Haltiwanger, Ron S. Jarmin, and Javier Miranda report that gross annual job creation was 17.6 percent of all jobs and gross annual job destruction was 15.4 percent, resulting in net job creation of 2.2 percent annually.[3]

The firms that systematically create more jobs than they destroy are young ones. Sixty percent of the jobs created by start-ups still exist five years later. Haltiwanger, Jarmin, and Miranda note that “startups are a critical component of the experimentation process that contributes to restructuring and growth in the U.S. on an ongoing basis.”[4]

Small Firms, Jobs, and Young Workers

Young firms are vital to growth. Two other papers show that concentrations of young or small firms are associated with higher subsequent growth across U.S. cities[5] and across countries.[6]

Haltiwanger, Jarmin, and Miranda emphasize that young firms are the most important net job creators. Since young firms are almost always small, a casual look at the data shows that small firms create many net jobs. However, deeper econometric analysis shows that the key feature is firm age, not firm size. Thus, we should expect roughly the same rate of net job creation from firms of equal age, regardless of their size.

Economists Edward Glaeser, William Kerr, and Giacomo Ponzetto go a step further. Knowing that entrepreneurial start-ups cause job growth, they ask the data where entrepreneurship occurs. Controlling for industry and region effects, they find that cities where the average establishment had fewer employees in 1992 had higher start-up rates there from 1992 to 1999. Thus, even if small businesses do not directly create jobs, they are a feature of a business environment that encourages entrepreneurship, and entrepreneurs do create jobs.

In a follow-up paper, Glaeser, Kerr, and Sari Pekkala Kerr dig deeper into the past and find that cities built near coal and iron mines tended to develop larger establishments and, therefore, to be less entrepreneurial.[7]

Economists Ejaz Ghani, Kerr, and Stephen O’Connell confirm the same pattern in India. Using an Indian dataset that includes both formal and informal manufacturing firms, they find that young firms create jobs in India, too. Likewise, clusters of small firms are associated with more start-ups in the future.

Economists Paige Ouimet and Rebecca Zarutskie find that young workers are the group most likely to find jobs at start-up firms.[8] In fact, they find that younger firms tend to offer better wages for young workers than older firms. And young firms seem to benefit from hiring young workers. Ouimet and Zarutskie suggest that some characteristics of young workers—such as their willingness to take risks—makes young workers a good match for young firms.

How It Works

Job creation and destruction occur in a highly complex, intertwined system. Glaeser, Kerr, and Ponzetto work to eliminate some possible explanations for the geographic correlation between average establishment size and subsequent start-ups. The link is not due to a city’s industrial composition, nor to general features of the city or region. They also rule out geographic differences in productivity and average age of incumbent firms, although those independently affect start-up rates.

The data leave two explanations standing: differences in fixed costs of starting a new business and geographically differing supplies of entrepreneurs. Regulatory hurdles, paperwork, and environmental restrictions, for example, can all raise fixed costs and discourage start-ups.

Policy Implications

New jobs come from new firms. New firms tend to appear alongside existing small firms. Given that these patterns persist in economies as different as the U.S. and India, it seems likely that the patterns of job creation represent fundamental economics, not a particular institutional framework.

Government can encourage job creation by making it easy to start a new business. Although large firms have lawyers working full time to help them navigate regulatory mazes, small businesses generally do not. The risk involved in entrepreneurship is ample; government should seek to remove barriers to entering business instead of erecting them.

Because incumbent firms and large corporations have the resources to lobby their representatives for favorable treatment, Washington policymakers often see persuasive arguments in favor of protecting the economic status quo. Where job creation is concerned, however, future start-ups are the champions, and they have no lobbyists.

The success and failure of startups are crucial mechanisms by which the economy organically encourages good ideas and discourages ones that do not meet consumers’ needs. Thus, although start-ups as a group create many net new jobs, individual start-ups frequently fail. Government should stay out of the business of “picking winners.” When a start-up based on a bad idea goes out of business, its capital and labor are freed up to better serve consumers.