Innovation Investment and Labor Mobility: Employee Entrepreneurship from Corporate R&D

How does corporate research and development (R&D) investment affect employees? This paper examines the effect of R&D on employee retention and moves out of the firm. We are motivated by the importance of R&D to economic growth. In the process of generating new knowledge, technology, and skills, R&D is also human capital-intensive. Human capital is inalienable and portable. Employee mobility is therefore a channel for R&D outputs to leave the firm.

We consider four possible effects, not all of which are unambiguously beneficial for employees. First, R&D could increase worker retention if it leads to more amenities or internal growth opportunities (Rosen 1986). Second, R&D may lead to automation or other structural changes that make labor redundant or skills obsolete, leading to layoffs and unemployment (Autor 2015, Acemoglu & Restrepo 2018). Third, R&D could create new skills or ideas that increase employees’ outside options at other incumbent firms (Herkenhoff, Lise, Menzio & Phillips 2018). Finally, R&D could generate skills or ideas that are specifically valuable for the founding teams of new firms.

Testing the effect of R&D on labor reallocation requires matching R&D-performing employers to employees and following the employees’ subsequent career paths. We accomplish this with U.S. Census data between 1990 and 2008. We consider firm R&D – a measure of innovation input – rather than patents because while patents are a useful measure of innovation output, they capture only the subset of innovation outputs that is contractible and over which the firm has chosen to establish property rights (Sampat 2018). We consider all employees rather than only inventors because the latter comprise a small fraction of the human capital involved in performing R&D and commercializing its outputs. Also, this permits considering broader mechanisms for how R&D might affect labor, including skill obsolescence.

We use two strategies to estimate the effect of R&D increases within publicly traded firms. The first is a regression model with firm, state-year, and granular industry-year fixed effects, as well as time- varying firm characteristics. We find no effects on exits from the employment sample (which mainly corresponds to unemployment and departures out of the labor force), no effects on employee retention, and no effects of R&D on employee departures to other incumbent firms. We do find weak positive effects on employee departures to young firms, which points us in the direction of entrepreneurial outcomes.

Indeed, we find robust effects of within-firm R&D increases on employee departures to entrepreneurship, an intuitive and interesting result because startups are known to be important conduits for commercializing new ideas. Specifically, a 100 percent increase in R&D (about one standard deviation) predicts an 8.4 percent increase in the mean rate of employee departures to entrepreneurship. We use the term “entrepreneurship” in a broad sense to mean the founding team of a new firm; the group most likely to contribute ideas and crucial skills to the startup.2 As we would expect, the effect is higher among high-tech establishments of parent firms. It is not driven by recent hires (who might, for example, have been hired because of the R&D project).

Despite fine controls, the estimate may be biased if, for example, an unobserved new technological opportunity increases both the R&D of parents and employee-founded startups. The second approach is therefore to instrument for R&D using changes in state and federal R&D tax credits, which affect the firm’s user cost of R&D. We follow Bloom, Schankerman & Van Reenen (2013) and provide new, exhaustive detail on the sources of within-firm variation for both instruments. The instrumental variables (IV) analysis also finds that R&D has no effect on other types of employee departures or on employee retention. The IV effect on employee-founded startups is about five times larger than the ordinary least squares (OLS) estimate. This could reflect downward bias in the OLS result. Alternatively, the IV strategy estimates the marginal effect of R&D (the effect of an additional “last” dollar), while OLS gives the average effect (the effect of increasing the optimal amount of R&D by one dollar). The causal effect may be higher for the last dollar if it is spent on projects that are further from the firm’s core focus or have less crucial outputs, and thus are more often rejected. It is also possible that adjustable R&D, the type sensitive to tax credit changes, has a larger causal effect. The OLS yields our preferred estimate because we are most interested in the effect of average increases in R&D.

While the data do not permit us to pin down a precise channel for R&D-induced entrepreneurial departures, we find cross-sectional support for a mechanism where employees take ideas, skills, or technologies created though the R&D process that are of relatively low value to the parent firm to startups. The starting point for motivating this mechanism is that the innovation process is serendipitous, producing unforeseen outputs. The second premise is that innovation effort is hard to contract ex-ante, and hard to verify ex-post. These information, agency, and contracting frictions create benefits to allocating control rights to the individuals who are performing innovation (Grossman & Hart 1986, Aghion & Tirole 1994). Contending with these frictions, the firm may opt not to pursue all good innovations, enabling employees to take some outside the firm.

Innovation frictions are magnified when an idea is riskier, making high-risk, high-reward growth options more often best located outside the firm boundary (Gromb & Scharfstein 2002, Robinson 2008, Frésard, Hoberg & Phillips 2017). Many risky ventures benefit from the high-powered incentives that exist in small, focused firms financed through external capital markets or in entities such as joint ventures (Rhodes-Kropf & Robinson 2008, Phillips & Zhdanov 2012). Consistent with this, we find that higher parent R&D is strongly associated with venture capital (VC) backing among employee-founded startups. Also, R&D-induced startups are more likely to be incorporated, more likely to be in high-tech sectors, pay higher wages on average, and are more likely to exit (fail or be acquired). Therefore, the effect appears to be driven by risky, new-to-the-world ideas, rather than Main Street-type businesses. We expect this mechanism to be particularly salient when the idea is less valuable to the parent firm, which may more often be the case when the idea is far from the firm’s core focus and would impose diversification costs. Indeed, we find that more parent R&D is negatively associated with the employee-founded startup being in same broad industry as the parent.

Firms may permit employees to depart with R&D outputs as a way to induce retention and effort, especially if the lost ideas are peripheral. The long-term labor contract with employees may include some degree of permissiveness towards R&D-induced entrepreneurship. Such a policy could help the firm to hire the best talent and induce optimal effort. A permissive approach to employee entrepreneurship is likely most feasible if the lost innovations are not ones the firm intends to sell or develop. Consistent with this, R&D outputs over which the firm does establish explicit property rights do not yield employee-founded startups. There are no effects of patents or patent citations on employee entrepreneurship.

While there are many studies of knowledge diffusion, to our knowledge there is no work on how innovation investment affects labor mobility. The relationship between innovation and human capital reallocation is important to understand, as these forces are central to productivity growth (Davis & Haltiwanger 1999, Giroud & Mueller 2015, Decker, Haltiwanger, Jarmin & Miranda 2018). The positive effect of R&D on entrepreneurship that we document is one mechanism for why high-growth startup founders are often former employees of large incumbent firms (Gompers, Lerner & Scharfstein 2005, Klepper 2009). We offer corporate R&D as a new source for where ideas for high-growth startups come from, a topic of considerable recent interest (Aghion & Jaravel 2015, Babina 2017).

Tania Babina is an Assistant Professor of Finance at Columbia’s Graduate School of Business. Sabrina Howell is an Assistant Professor of Finance at the NYU Stern School of Business.

References

Acemoglu, D. & Restrepo, P. (2018), ‘The race between man and machine: Implications of technology for growth, factor shares, and employment’, American Economic Review 108(6), 1488–1542.

Aghion, P. & Jaravel, X. (2015), ‘Knowledge spillovers, innovation and growth’, The Economic Journal 125(583), 533–573.

Aghion, P. & Tirole, J. (1994), ‘The management of innovation’, The Quarterly Journal of Economics 109(4), 1185–1209.

Autor, D. (2015), ‘Why are there still so many jobs? The history and future of workplace automation’, Journal of economic perspectives 29(3), 3–30.

Babina, T. (2017), ‘Destructive creation at work: How financial distress spurs entrepreneurship’. Working Paper.

Bloom, N., Schankerman, M. & Van Reenen, J. (2013), ‘Identifying technology spillovers and product market rivalry’, Econometrica 81(4), 1347–1393.

Davis, S. J. & Haltiwanger, J. (1999), ‘On the driving forces behind cyclical movements in employment and job reallocation’, American Economic Review 89(5), 1234–1258.

Decker, R. A., Haltiwanger, J. C., Jarmin, R. S. & Miranda, J. (2018), Changing business dynamism and productivity: Shocks vs. responsiveness, Technical report, National Bureau of Economic Research.

Frésard, L., Hoberg, G. & Phillips, G. (2017), ‘Vertical acquisitions, integration and the boundaries of the firm’.

Giroud, X. & Mueller, H. M. (2015), ‘Capital and labor reallocation within firms’, The Journal of Finance 70(4), 1767–1804.

Gompers, P., Lerner, J. & Scharfstein, D. (2005), ‘Entrepreneurial spawning: Public corporations and the genesis of new ventures, 1986 to 1999’, The Journal of Finance 60(2), 577–614.

Gromb, D. & Scharfstein, D. (2002), Entrepreneurship in equilibrium, Technical report, National bureau of economic research.

Grossman, S. J. & Hart, O. D. (1986), ‘The costs and benefits of ownership: A theory of vertical and lateral integration’, Journal of Political Economy 94(691), 693–95.

Herkenhoff, K., Lise, J., Menzio, G. & Phillips, G. (2018), ‘Worker mobility and the diffusion of knowledge’.

Klepper, S. (2009), ‘Spinoffs: A review and synthesis’, European Management Review 6(3), 159–171.

Phillips, G. M. & Zhdanov, A. (2012), ‘R&D and the incentives from merger and acquisition activity’, The Review of Financial Studies 26(1), 34–78.

Rhodes-Kropf, M. & Robinson, D. T. (2008), ‘The market for mergers and the boundaries of the firm’, The Journal of Finance 63(3), 1169–1211.

Rosen, S. (1986), ‘The theory of equalizing differences’, Handbook of labor economics 1, 641–692.

Sampat, B. N. (2018), ‘A survey of empirical evidence on patents and innovation’, NBER Working Paper (w25383).

Download the complete
article PDF