Why some entrepreneurs start more firms

Why do some entrepreneurs start multiple firms while most start just one? And what does this tell us about the environment facing new businesses in developing economies? A large body of research shows that the creation of new productive firms is an important engine of growth, especially in emerging economies (Hsieh and Klenow 2009, Brandt et al. 2012). However, as Baumol (1990) famously argued, whether entrepreneurial activity translates into economic growth depends critically on the institutional ‘rules of the game’. If institutions reward productive innovation, economies thrive; if they reward unproductive rent-seeking or favouritism, growth suffers (Acs and Szerb 2009, Naudé 2010). In recent research, we examine these forces through the lens of serial entrepreneurs in China, using data covering the universe of Chinese firms from 1995 to 2015 (Brandt, Dai, Kambourov, Storesletten, and Zhang forthcoming).

In many countries, serial entrepreneurs – those who establish more than one firm – act as engines of growth. But are they successful because they are genuinely more skilled at identifying business opportunities? Or do they succeed because they have connections that help them secure financing and overcome barriers to entry?

These questions matter for understanding not just entrepreneurship, but the broader forces shaping economic development. If serial entrepreneurship is driven by skill, it's a sign of dynamism; if it's driven by connections, it signals distortions that misallocate resources.

In principle, serial entrepreneurs could be selected because of highly persistent skills – they are individuals good at identifying new business opportunities (Holmes and Schmitz 1995). Alternatively, their advantage may lie in better access to scarce factors or connections, which help them secure finance and overcome barriers to doing business (Evans and Jovanovic 1989, Song et al. 2011).

We document that serial entrepreneurs are a quantitatively important part of the economy – they represent about one-third of all firms and hold nearly half of all registered capital. Their performance, however, reveals a more complicated picture than previous research has suggested.

The puzzle of serial entrepreneurs

On average, serial entrepreneurs in China look impressive. Their first firms are 11% more productive than those of non-serial entrepreneurs, and their second firms are 21% more productive. This is consistent with recent evidence from other countries finding that serial entrepreneurs outperform non-serial entrepreneurs (Lafontaine and Shaw 2016, Felix et al. 2021). They also operate larger firms, with 43% and 74% more assets, respectively.

This seems to support the view that serial entrepreneurs succeed because of persistent skills – they are simply better at identifying and running businesses. These averages, however, mask striking variation. When we dig deeper, we find that most serial entrepreneurs underperform relative to those who start only one firm. How can this be?

The answer lies in what happens when serial entrepreneurs switch industries.

Learning or favouritism?

About two-thirds of serial entrepreneurs switch industries when starting their second firm. In contrast, a majority remain in the same prefecture. Table 1 shows just how common this behaviour is.

Table 1: Geographical and industry location, registry data, 1995–2015

Notes: Authors' calculations from the Registry Data. Stayers are serial entrepreneurs whose first and second firms are in the same I-O industry. Switchers are those whose first and second firms are in different I-O industries.

This pattern sets up the central puzzle. Why do so many entrepreneurs switch industries? And what does their choice reveal about why they become serial entrepreneurs in the first place?

To answer these questions, we develop a model where entrepreneurs face two key forces. First, skills may be persistent across firms started by the same person – success in one business predicts success in another, especially in the same industry (Jovanovic 1979). Second, some entrepreneurs may have preferential access to finance, allowing them to start firms even when their skills are modest.

These forces have opposite implications. If serial entrepreneurship is driven by persistent skills, serial entrepreneurs should be more productive than non-serial entrepreneurs. If it's driven by favouritism, they should be less productive but operate with more capital.

The data reveals that both forces are at work. Serial entrepreneurs who stay in the same industry are dramatically more productive – their first firms are 51% more productive than non-serial firms, and their second firms are 75% more productive. These are the truly skilled entrepreneurs.

But serial entrepreneurs who switch industries tell a different story. Their firms are 7% less productive than non-serial firms. Yet despite lower productivity, they operate with more capital than their more productive peers who stay in the same industry.

This combination – lower productivity but more capital – is exactly what our model predicts for entrepreneurs with preferential access to finance. These ‘favoured’ entrepreneurs can start firms even when their skills are modest because they face looser borrowing constraints. The patterns in our data – with capital increasing in TFP up to a point, and debt-equity ratios varying systematically with equity – are consistent with a model where entrepreneurs face collateral constraints on borrowing (Evans and Jovanovic 1989, Midrigan and Xu 2014).

Learning from industry choice

Why would skilled entrepreneurs mostly stay in the same industry while favoured entrepreneurs switch? Our model suggests a learning mechanism. Entrepreneurs discover their industry-specific ability by operating their first firm (Jovanovic 1979). If that firm performs well, they learn they have a comparative advantage in that industry and stay. If it performs poorly, they learn they don't and switch to try something else.

But this creates a puzzle: if switchers are learning from poor performance, why do they start second firms at all? The answer, our evidence suggests, is that many switchers are favoured entrepreneurs whose access to finance lowers the bar for entry. They can afford to try again even when their first attempt didn't go well.

Consistent with this interpretation, switchers have substantially lower productivity than stayers but operate with more capital – the signature of preferential treatment.

Concurrent versus non-concurrent firms

We also examine whether serial entrepreneurs operate their firms concurrently or close the first when starting the second. About 60% operate concurrently, while 40% close the first firm.

The patterns here are revealing. First firms that are eventually closed have 43% lower productivity and 4% less equity than those that continue operating alongside the second firm. And second firms that replace first firms have 13% higher productivity but 29% less equity than those run concurrently.

These patterns match our model's predictions about financial constraints. Entrepreneurs with less equity and a much more productive opportunity for a second firm face higher opportunity costs of keeping the first firm operating, making closure more likely.

Diversification and business linkages

Beyond learning, we explore other motives for industry choice. Risk diversification matters: entrepreneurs are more likely to choose second industries whose returns are negatively correlated with their first industry, and they require a substantial productivity premium – 31% – to choose positively correlated industries.

Business linkages also matter. Serial entrepreneurs are more likely to start second firms in industries that are upstream or downstream from their first, or that share similar input or output patterns. This suggests that some serial entrepreneurs build integrated business groups, potentially capturing gains from coordination.

What this means for development

Our findings paint a complex picture of entrepreneurship in China – one that combines genuine dynamism with significant distortions.

On one hand, highly skilled serial entrepreneurs who stay in their industry of strength drive productivity growth. On the other hand, favoured entrepreneurs with access to finance but modest skills start multiple firms, misallocating capital and lowering average productivity.

This mix of dynamism and distortion may characterise many developing economies. The presence of highly productive serial entrepreneurs shows that entrepreneurial talent is abundant. But when connections matter more than skill for accessing finance, that talent cannot fully realise its potential.

Our findings contribute to a large evidence base on how distortions and misallocations affect economic development (Restuccia and Rogerson 2008, Hsieh and Klenow 2009). In the Chinese context, previous work has emphasised differences between the state and private sectors (Song et al. 2011, Hsieh and Song 2015). Our results show that even within the private sector, there are stark differences among entrepreneurs in their productivity and size – differences we link to preferential access to finance.

The costs are likely significant. Given that serial entrepreneurs account for one-third of all firms and half of all registered capital in China, the aggregate effects of these distortions – favouring some entrepreneurs while constraining others – are substantial.

Policy implications

Our results suggest that policies aimed at levelling the playing field for entrepreneurs could have large payoffs. When access to finance depends on connections rather than merit, resources flow to less productive uses and talented entrepreneurs are held back.

Reforms that reduce favouritism and broaden access to credit could improve both static efficiency – better allocation of capital across existing firms – and dynamic efficiency – encouraging more talented individuals to become entrepreneurs in the first place.

The fact that many serial entrepreneurs succeed through genuine skill is encouraging. It suggests that policies that reduce distortions could unleash substantial productivity gains by allowing the most talented to expand and by ensuring that new firm creation reflects ability rather than access.

Serial entrepreneurs in China are, on average, more productive than those who start only one firm. This average, however, conceals a more troubling reality: a substantial fraction of serial entrepreneurs succeeds not because of skill but because of preferential access to finance. These favoured entrepreneurs have lower productivity yet operate with more capital, misallocating resources and dampening aggregate productivity.

Understanding this heterogeneity – and the distortions that create it – is essential for designing policies that foster genuine entrepreneurial dynamism rather than simply rewarding connections (Baumol 1990, Holmes and Schmitz 1995).

References

Acs, Z J, and L Szerb (2009), "The global entrepreneurship index (GEINDEX)," Foundations and Trends in Entrepreneurship, 5(5): 341–435.

Baumol, W J (1990), "Entrepreneurship: productive, unproductive, and destructive," Journal of Political Economy, 98(5): 893–921.

Brandt, L, R Dai, G Kambourov, K Storesletten, and X Zhang (forthcoming), "Serial entrepreneurship in China," American Economic Journal: Macroeconomics.

Brandt, L, J Van Biesebroeck, and Y Zhang (2012), "Creative accounting or creative destruction? Firm-level productivity growth in Chinese manufacturing," Journal of Development Economics, 97: 339–351.

Evans, D, and B Jovanovic (1989), "An estimated model of entrepreneurial choice under liquidity constraints," Journal of Political Economy, 97: 808–827.

Felix, S, S Karmakar, and P Sedlacek (2021), "Serial entrepreneurs and the macroeconomy," Unpublished manuscript.

Holmes, T J, and J A Schmitz (1995), "On the turnover of business firms and business managers," Journal of Political Economy, 103: 1005–1038.

Hsieh, C-T, and P J Klenow (2009), "Misallocation and manufacturing TFP in China and India," Quarterly Journal of Economics, 124: 1403–1448.

Hsieh, C-T, and Z Song (2015), "Grasp the large, let go of the small: The transformation of the state sector in China," Brookings Papers on Economic Activity: 295–346.

Jovanovic, B (1979), "Job matching and the theory of turnover," Journal of Political Economy, 87: 972–990.

Lafontaine, F, and K Shaw (2016), "Serial entrepreneurship: Learning by doing?" Journal of Labor Economics, 34: S217–S254.

Midrigan, V, and D Y Xu (2014), "Finance and misallocation: Evidence from plant-level data," American Economic Review, 104: 422–458.

Naudé, W (2010), "Entrepreneurship, developing countries, and development economics: New approaches and insights," Small Business Economics, 34: 1–12.

Restuccia, D, and R Rogerson (2008), "Policy distortions and aggregate productivity with heterogeneous plants," Review of Economic Dynamics, 11: 707–720.

Song, Z, K Storesletten, and F Zilibotti (2011), "Growing like China," American Economic Review, 101: 196–233.