Innovation is the primary engine of growth in economies at the technological frontier, and a path to higher profits and growth for individual companies, as the likes of Apple, Alphabet, Microsoft, and Amazon make clear. CEOs play a crucial role in directing and overseeing their firms’ innovation efforts. This, however, creates a tension: The interests of the shareholders and the CEO might not be aligned, opening the door to agency frictions. In turn, such frictions can result in suboptimal investment in innovation, leading to losses in firm value for shareholders and low economic growth and welfare for the broader economy.

Better corporate governance can help align the interests of managers and shareholders, and thus alleviate the negative impact of agency frictions. Recent empirical studies confirm the significance of corporate governance in the growth process. For instance, Aghion, Van Reenen, and Zingales (2013) find that firms with better corporate governance are more likely to come up with patented inventions. A recent report by the OECD summarizes this evidence, stating that “corporate governance exerts a strong influence upon innovative activity and entrepreneurship. Better corporate governance, therefore, should manifest itself in enhanced corporate performance and can lead to higher economic growth.”

Although there is a broad consensus that corporate governance can be relevant for innovation and economic growth, there’s little agreement on how it works. Despite the substantial, and largely separate, strands of literature on corporate governance, agency frictions, and firm innovation, little work has been done to quantify the effects of agency frictions between shareholders and managers on firm innovation, managerial compensation structure, and economic growth. In our efforts to fill this gap, we find that better corporate governance improves firm innovation primarily through the way in which managers are compensated.

One increasingly popular method at U.S. public companies is to increase the share of stock options in manager compensation. We find that firms with higher institutional ownership tend to adopt that method. Stock options reward managers if they succeed in improving the firm’s market value, while isolating them from downside risk. Therefore, stock options provide incentives to engage in risky innovative activities, thereby contributing to firm value and economic growth.

To undertake our empirical analysis, we combine micro-data on patented inventions from the United States Patent and Trademark Office (USPTO) with data on U.S. public firms from Compustat and executive compensation information from Execucomp. Unlike the majority of researchers using similar data, we focus on the quality, not the quantity, of innovations. In other words, we would like to capture the effect of corporate governance on high-impact innovations that provide a significant advantage to a firm over its peers, rather than on what are colloquially called “junk patents.” We employ three scale-independent measures of disruptive innovations: (i) the average number of citations to the patents of a firm, (ii) the fraction of a firm’s patents that make it to the top 10 percent in terms of patent quality, and (iii) the average originality of the patents – a measure which captures the variety of distinct technology classes the new innovation is combining and building upon. We document the following facts:

• Higher institutional ownership is positively associated with disruptive innovations.
• Higher institutional ownership leads to CEO compensation contracts with more stock options and restricted stock grants relative to total compensation.
• Such contracts are associated with more disruptive innovations.
• The positive association between institutional ownership and disruptive innovations is realized largely through the mechanism of tying CEO compensation to performance.

How much do agency frictions between managers and shareholders matter for innovation, firm growth, firm value, and economic growth? To what extent can the inefficiencies caused by agency frictions be alleviated through better corporate governance and tying CEO pay more closely to performance? Are there systematic differences in the degree of this inefficiency across time? Answering these questions in a purely empirical setting is challenging. First, agency frictions are not directly observable. To assess their impact, we must identify what would have happened in a parallel, counterfactual world without agency frictions. Evaluating this counterfactual is difficult, because it is hard to find exogenous shocks that eliminate agency frictions. Even if there were such a shock, it is likely to be limited in scope, raising concerns about external validity. Overall, it is unclear how to quantify the effect of agency frictions without a model.

We overcome these challenges by developing and structurally estimating a new, dynamic general-equilibrium model with firm innovation, and endogenously determined managerial compensation contracts. In the model, the firm’s board determines the CEO’s compensation. Taking the compensation as given, the CEO makes the innovation decisions. The board, however, does not fully represent the preferences of the shareholders. The CEO’s compensation structure is the product of a tug-of-war between the CEO and the shareholders. The CEO has some influence over the board’s final decision. Consequently, the agreed-upon compensation contract deviates from the shareholder-optimal contract that would maximize firm value. Better corporate governance acts to reduce the CEO’s influence, enabling the board to choose compensation contracts with a higher fraction of stock options. This, in turn, motivates the manager to carry out more disruptive innovations. Better corporate governance thus leads to a higher rate of innovation and more knowledge spillovers, thereby increasing long-run productivity and social welfare.

Using simulated method of moments (SMM), the model parameters are estimated to best fit a wide-ranging set of facts about U.S. firms. Using the estimated model, we quantify the impact of agency frictions on firm innovation, economic growth, and social welfare. An experiment in which we remove agency frictions by eliminating CEO influence results in an increase in firm innovation by 26.6 percent of its value. Output growth rate increases by 0.51 percent. This makes the average individual in the economy as well-off as a permanent increase in consumption by 7.3 percent, which is how we measure social welfare. These large gains are attributable to the fact that even a slight increase in economic growth can translate into huge compounded gains in the future. In another quantitative experiment, we attempt to gauge the impact of FAS 123R, a change in accounting standards introduced in December 2004 that discourages firms from compensating employees with stock options. We find that it might have slightly reduced long-run economic growth while concentrating R&D in firms with better corporate governance. The overall effect is a drop in social welfare by 0.87 percent.

A recent strand of literature finds that short-term pressure on CEOs to meet earnings targets can force them to decrease investment in R&D. To check the robustness of our results and to study the interaction between short-termism and CEO influence, we enrich our analysis by incorporating short-term earnings pressure on CEOs into our baseline model. We document new evidence on how short-termism is a more relevant problem for firms with a high level of institutional ownership. In the extended model, CEOs suffer a loss in total compensation if they miss the short-term earnings target. This leads to a reduction in R&D spending during unprofitable years, especially for innovative firms. The cut in R&D spending increases reported profits, thereby allowing the CEOs to avoid the loss. We find that this extension does not change our quantitative results significantly, reducing the estimated social cost of agency frictions from 7.3 percent to 5.8 percent. Removing short-term pressure on its own also leads to gains in growth and welfare, albeit at one quarter of the magnitudes achieved by eliminating CEO influence. Finally, the model predicts amplified gains if both frictions are alleviated simultaneously.

Our findings suggest that institutional shareholders play a crucial role in improving firm innovation. This effect is caused mainly by how managers are compensated. Greater innovation is desirable to shareholders because it increases firm value and growth. At the same time, knowledge spillovers to the rest of the economy help improve long-run economic growth and social welfare. Our quantitative findings suggest that there is significant room for policy intervention to improve social welfare through the alleviation of agency frictions. While no magic recipe exists for improving corporate governance, understanding how and why institutional investors buy stakes in particular firms might be a fruitful area for future research. Another option is changing how various components of CEO compensation are taxed by the government, which can provide more incentives for firms to increase the performance-based parts of CEO compensation such as stock options.

REFERENCE

Aghion, P., J. Van Reenen, and L. Zingales (2013): “Innovation and Institutional Ownership,” American Economic Review, 103(1), 277-304.

This post comes to us from professors Murat Alp Celik and Xu Tian at the University of Toronto. It is based on their recent paper, “Agency Frictions, Managerial Compensation, and Disruptive Innovations,” available here.