Politicians, and many economists, have often pushed for programs that support small and medium enterprises. One justification is that SMEs employ a large portion of the labor force, so policies that help them should generate more jobs.

But evidence from the US and Latin America suggests that it is new firms, not small firms, which have a disproportionate impact on net employment. Moreover, new firms, rather than established firms that remain small, are a natural conduit for introducing innovative, high productivity ideas to the marketplace.

The process of converting good ideas into good new businesses, however, is rife with market failures. A new firm that introduces a novel idea could be imitated. If so, the entrepreneur – who took all the risks — may not appropriate all the benefits of introducing her idea, thus discouraging entrepreneurship.

Entrepreneurs may have a great ideas but not have the complementary skills — such as managerial capacity or marketing knowledge — to start and grow a viable organization, or to attract those who do. Furthermore, new innovative firms typically lack access to finance, since even good ideas are hard to collateralize.

Not all new firms have the same potential impact on job creation or innovation. In fact, the majority of new business owners — such as construction contractors, auto mechanics, dentists or insurance brokers — do not plan for their firms to innovate or grow substantially. But market failures tend to be more severe for innovative firms.

Efforts to encourage the creation of new firms should not, therefore, consist of blanket policies favoring entrepreneurship. Rather, they should focus on dynamic entrepreneurship, that is, on encouraging the creation of innovative new firms with high growth potential.

But if policies need to target the most promising new firms, how can countries identify the “right” ones? Indeed, there is no need to pick them beforehand. Rather, countries need to adopt policies that attract high productivity firms, or implement policies that leverage private sector capabilities to screen and identify the most promising ones.

Supporting business incubators and accelerators may have an impact on a relatively small number of firms, but at least may be targeting the right ones. Incubators provide entrepreneurs with access to networks of mentors and financing, and may help promising firms emerge and grow. But for incubators to select the right firms and provide valuable services, incentives have to be right. Incubators should be rewarded according to firm performance, not just based on the number of incubated firms, as often done in the region.

Encouraging the development of a venture capital industry is another complementary way to stimulate dynamic entrepreneurship. Unlike banks, VC funds can accept innovative ideas as collateral because they participate in the financial upside of firms. While VC is not possible everywhere (it requires a critical mass of good projects), it was the cornerstone of the recent development strategy in countries like Taiwan and Israel. But, with the possible exception of Brazil, the VC industry is underdeveloped in Latin America.

Encouraging dynamic start-ups, however, is not enough. To make a dent, high productivity start-ups need to scale up and absorb workers and other factors employed in less productive activities. But high productivity firms in Latin America do not scale-up well, and tend to remain too small. The reasons range from distortions in labor markets and tax codes favoring small firms, to family ownership. Understanding the constraints that prevent firms scaling up — and addressing them — is a key. Without it, policies encouraging start-ups will not deliver on their promise. LF

Ernesto Stein is principal economist at the research department of the Inter-American Development Bank, and co-author of Rethinking Productive Development, published by the bank in September