By Sarah Watson —

India’s government is faced with the task of not just creating new jobs but of upgrading existing jobs: converting them from low-productivity single-man businesses to high-productivity establishments that create exportable goods. Prime Minister Narendra Modi’s ‘Start Up India, Stand Up India’ initiative is a well-intentioned effort to encourage entrepreneurship. But recent research on job creation suggests that start-ups’ major impact on job growth comes not when they are first founded but over time, as they scale and grow. In light of the low dynamism of Indian firms, the Modi government should place an equal emphasis on helping older start-ups to survive as on creating new ones.

India’s relatively low unemployment rate, around 8 percent, according to a real-time index produced by CMIE, belies its poverty. Most Indians are employed; the problem is that they work low-paying and highly unproductive jobs. As Manas Chakravorty recently noted, 25 years after the first major steps towards economic liberalization India’s labor landscape has become more fragmented and less ‘modern’ than it was in the late 1980s. A closer look at data from the Economic Census (EC), which has been published at roughly 10-year intervals since 1977, shows while India has been adding non-farm workers and enterprises at a rapid pace it has failed to move these workers into productive formal sector jobs.

The most recent EC data show that the percentage of enterprises that are “own account” (do not employ any hired workers) has returned to 1990 levels after dipping sharply in 2005 (Figure 1). Even more remarkably, the percent of all workers who work for themselves and have no employees (i.e. in own account enterprises) has risen from 33 percent in 1990 to over 44 percent in 2015, indicating that formal establishments have fewer employees than they did in 1990 (Figure 2).

As these figures suggests, the number of employees per enterprise has steadily decreased, from 2.88 in 1990 to 2.24 in 2015 (Figure 3).

Perhaps the most dramatic change is in the percentage of workers who are employees of larger enterprises. Between 1990 and 2005 this dropped from 37.1 percent to 25.5 percent as the percentage of enterprises employing more than 10 people more than halved, to 1.5 percent (Figure 4). Equivalent data was not available for the 6^th EC in 2015, but the percentage of all establishments that employed 10 or more people dropped to 1.4 percent that year. In 2015, less than one percent of all business employed more than 14 people. In 1990, 55.8 percent of Indian workers worked for someone else; in 2015 that portion had dropped to 43.5 percent.

While India has lost ground in terms of the average size of its businesses, it has added them at a rapid clip: the number of enterprises increased by 234 percent between 1990 and 2015. But it does not appear to be succeeding in converting these new enterprises into globally competitive businesses that produce exportable goods and that are in a position to hire large numbers of employees. As a result, India has 15 million one-employee retail shops and more than 10 million one-employee manufacturing enterprises.

This picture exemplifies a divide in recent economic research on job creation. Although economists still debate whether large or small businesses create the most jobs, recent research on developed economies seems to establish that it is new firms, rather than specifically large or small ones, that fuel job creation. From this perspective, India is doing well — it is generating millions of new enterprises. But studies of developing economies present a more complex picture: since small enterprises are so likely to go under or to stop at a single self-employed worker, net job creation only takes place when a small firm survives and grows.

The government’s start-up policy, however, explicitly excludes older firms, whether or not they have yet found their sea-legs. The Department of Industrial Policy and Promotion’s definition of a start-up limits the category to businesses that are less than five years old and have never achieved an annual turnover of more than about $3.8 million dollars. Start-ups can have a turnover far higher than that sum without being profitable. Consider that Amazon did not turn a profit until 2003 despite having revenues of $1.6 billion in 1999. And many of today’s highest profile and most innovative companies struggled for at least five years before finding their footing. The policy also states that to qualify as start-ups businesses must also offer new products “driven by technology or intellectual property,” despite the debate over whether or not high-tech start-ups are really drivers of job growth.

Once a start-up has aged out of ‘Start Up India, Stand Up India,’ it will exchange the initiative’s protections — such as self-certification of compliance with labor and environmental regulations — for a general business environment ranked 130^th in the World Bank’s Ease of Doing Business Rankings. Middle-aged Indian firms with more than 10 employees face inefficient markets and a harsh regulatory environment. In the absence of a long-term plan to improve the environment for all businesses — not just start-ups and small enterprises, but establishments at all stages of the growth cycle — many of India’s new start-ups will likely fail to thrive, and struggle to produce the jobs that India so desperately needs.

Ms. Sarah Watson is an associate fellow with the Wadhwani Chair in U.S.-India Policy Studies at CSIS. Follow her on twitter @SWatson_CSIS.