In our forthcoming American Economic Review paper, Innovation and Institutional Ownership, we examine the incentives to innovate at the firm level by studying the relationship between innovation and institutional ownership. Innovation is the main engine of growth. But what determines a firm’s ability to innovate? Innovating requires taking risk and forgoing current returns in the hope of future ones. Furthermore, while any type of financing is plagued by moral hazard and adverse selection, the financing of innovation is probably the most vulnerable to these problems (Arrow, 1962) since the information that needs to be conveyed is hard to communicate to outsiders. This paper is an attempt at analyzing the corporate governance of innovation and more specifically the role of institutional owners in fostering (or hindering) innovation.
While the ability to diversify risk across a large mass of investors makes publicly traded companies the ideal locus for innovation, managerial agency problems might undermine the innovation effort of these companies. In publicly traded companies, the pressure for quarterly results may induce a short-term focus (Porter, 1992). And the increased risk of managerial turnover (Kaplan and Minton, 2008) might dissuade risk-averse senior managers from this activity. Finally, innovation requires effort and “lazy” managers might not exert enough of it. Hence, it is especially important to study the governance of innovation in publicly traded companies, which account for a large share of the private investments in research and development (R&D).
Probably the most important phenomenon in corporate governance in the last thirty years has been the remarkable rise in institutional ownership. While in 1970 institutions owned only 10% of publicly traded equity, by the start of 2006 they owned more than 60%. Thus, in this paper we focus on the role of institutional ownership on the innovation activity of publicly traded companies. Did the rise in institutional ownership increase short-termism, undermining the innovation effort? Or did it reassure managers, making them more willing to strike for the fence? To answer these questions we assemble a rich and original panel dataset of over 800 major US firms over the 1990s containing time-varying information on patent citations, ownership, R&D and governance. We show that there is a robust positive association between innovation and institutional ownership even after controlling for fixed effects and other confounding influences. Institutions have a small and positive impact on R&D, but a larger positive effect on the productivity of R&D (as measured by future cite-weighted patents per R&D dollar).
To uncover the source of this relationship we build a model that nests the two main reasons for this positive effect. The simplest explanation is managerial slack: managers may prefer a quiet life but institutional investors force them to innovate. An alternative explanation is based on career concerns. Innovation carries a risk for the CEO: if things go wrong for purely stochastic reasons, the board will start to think he is a bad manager and may fire him. This generates a natural aversion to innovation. If incentive contracts cannot fully overcome this problem, increased monitoring can improve incentives to innovate by “insulating” the manager against the reputational consequences of bad income realizations. According to this view, institutional owners have better incentives (they typically own a large share of the firm) and abilities (they typically own stock in many firms so they benefit from economies of scope in monitoring) to monitor. This more effective monitoring will therefore encourage innovation. The lazy manager hypothesis predicts that product market competition and institutional ownership are substitutes: if competition is high then there is no need for intensive monitoring as the manager is disciplined by the threat of bankruptcy to work hard. In contrast, our career concern model predicts that more intense competition reinforces the positive effect of institutional investment on managerial incentives.
We find that the positive relationship between innovation and institutional ownership is stronger when product market competition is more intense (or when CEOs are less “entrenched” due to protection from hostile takeovers), which is consistent with the career concerns hypothesis and inconsistent with the “lazy manager” one. Another implication of the career concern model is that the decision to fire the CEO is less affected by a decline in profitability when institutional investment is high. We find this is indeed the case. While in the absence of a large institutional investor a decline in profit leads to a high probability the CEO is dismissed, this probability drops when institutional investors own a substantial amount of stock.
Next, we try to uncover which institutions have the biggest impact on innovation by using Bushee (1998) classification. We find that quasi-indexed institutions (to use Bushee’s classification) have no association with innovation, while other forms of institutional owners have an equally positive association with innovation. We also show that the effect of these non-indexed institutional owners on innovation appeared to grow stronger after the 1992 change in the American Proxy Rules that increased their influence.
Finally, we argue that the correlation between institutional ownership and innovation is not primarily due to a selection mechanism whereby institutions are simply better at selecting the companies who will innovate more in the future. We also show that the exogenous increase in institutional ownership that follows the addition of a stock to the S&P500, has a positive effect on innovation, even for the non-indexed investors.
While there is a large literature on the effect of financing constraints on R&D (for surveys see Bond and Van Reenen (2007) and Hall(2002)), there is very little on the relation between institutional ownership and innovation. Notable exceptions are Francis and Smith (1995), who find a positive correlation between ownership concentration (which includes institutions) and R&D expenditures and Eng and Shackell (2001), who find a positive correlation of institutions with R&D. In a similar vain, Bushee (1998) finds that cuts in R&D following poor earnings performance are less likely the greater is the degree of institutional ownership. Unlike all these papers, we focus on the actual productivity of the innovation process, rather than only on the quantity of innovative inputs (R&D expenses). In addition, our use of an instrument and policy changes allows us to examine the possibility that this relationship is due to institutions’ ability to select the most innovative firms. Finally, our model allows us to probe deeper into the fundamental agency problem underlying this relationship.
Our model is complementary to Edmans. He shows how institutional ability to sell improves the information embedded into prices and in so doing improve the incentives to innovate. We take that effect on prices as a given and derive its impact on a career concern model. The most related paper is Sapra et al. (2008), who independently look at the effect of takeover pressure on innovation, using both R&D expenditures and patent citations as measures of corporate governance.
The full paper is available for download here.