Knowledge spillovers between firms are a key driver of important economic phenomena such as technological progress (Griliches, 1979), growth in international trade (Grossman and Helpman, 1991), and industrial innovation (Bloom, Schankerman, and Van Reenen, 2013). Although prior research has examined how firm-to-firm spillovers can arise through strategic alliances, supply-chain relationships, geographic proximity, or common ownership, very little is known about whether such spillovers can occur through interlocking boards, i.e., direct inter-firm links that arise when an individual serves concurrently on two or more boards. In a new study, we empirically examine whether interlocking boards provide an important channel for scientific knowledge and innovation to flow between firms.

These connections between boards can be expected to lead to knowledge spillovers for two reasons. First, corporate directors are fiduciaries who advise and monitor managers on behalf of shareholders. As such, directors have a duty to provide their firms with information that is relevant to business practices and shareholder value. Second, because directors often communicate with fellow board members and with corporate management, an interlocking director may be privy to firm-specific information at both firms that is neither quantifiable nor widely available to the public.

To examine the effects of interlocking boards on the inter-firm transmission of scientific knowledge, we construct a new dataset that details board linkages and patenting activity for nearly 500,000 pairs of public U.S. firms from  2003 to 2019. We examine  how the sharing of knowledge between a firm that created the patents (the “source firm”) and a firm with a linked board (the “downstream firm”) affects the quantity, quality, and novelty of the downstream-firm’s patents and the extent to which individual downstream-firm patents directly cite or are textually similar to source-firm patents.

A key empirical challenge is that links between boards and downstream firms’ innovation policies are likely to be endogenous responses to one or more characteristics of firm pairs (e.g., policy similarity, geographic proximity, or social connections). To address this challenge, we introduce a novel instrumental variables (IV) approach based on schedule conflicts in annual shareholder meetings. Each year, nearly all U.S. public firms hold shareholder meetings at which director attendance is mandatory. The large majority of these annual meetings are held during the two-month “proxy season” from mid-April to mid-June. Hence, there are many pairs of firms for which it would be infeasible for a director to simultaneously attend both firms’s meetings. Since the timing of a firm’s annual shareholder meeting is usually the same from one year to the next, we argue that schedule conflicts between two firms’ meetings are predictable and avoidable, thus providing a source of exogenous variation in board interlocks between firms. We document that annual-meeting schedule conflicts are, in fact, strong (negative) predictors of board interlocks, and we use these schedule conflicts to construct time-varying instrumental variables at the firm-pair level that help remove the endogeneity in board interlocks.

We find consistent evidence that board interlocks help facilitate knowledge spillovers between firms. First, a board interlock increases the flow of patents from a source firm and the degree to which the downstream firm’s patents are related to the source-firm patents, where relatedness is captured by cross-patent citations or textual similarity between patents. Second, we find that board interlocks significantly increase the sensitivity of a downstream firm’s innovation quantity (patent volume) and quality (the occurrence of highly-cited or exploratory patents) to the flow of patents from the source firm.

We then examine how the board structures of source and downstream firms alter the impact of board interlocks on spillovers. Consistent with theoretical work showing that information flow from management to directors is generally reduced when the board is less “friendly” or has fewer insiders (Raheja, 2005; Adams and Ferreira, 2007; Harris and Raviv, 2008), we find that interlocks have significantly less impact on knowledge spillovers when the source-firm’s board or the downstream-firm’s board is dominated by outsiders. Similarly, we find that interlocks are less conducive to spillovers when either the source firm or the downstream firm has few co-opted directors (i.e., directors who were appointed during the tenure of the current CEO). These results hold both for the relatedness measures of spillover as well as those based on patent quantity and quality.

We also explore how the spillover effects of board interlocks vary with individual characteristics of directors who are the link between firms. Our regressions reveal that directors who hold more than three board seats in a given year contribute less positively to knowledge spillovers between firms. In contrast, directors who are younger have a significantly more positive influence on knowledge spillovers.

Overall, our study provides evidence that board interlocks are an important channel for knowledge spillovers. The results of our analysis contribute to a broader understanding of the mechanisms by which corporate governance and the form of organizations can help transmit scientific knowledge and, ultimately, technology, between public U.S. firms. Our findings also have important implications for practitioners and policymakers. For example, boards that are highly independent or that  monitor their companies closely may weaken information sharing between management and directors, thus reducing the valuable spillover effects of board interlocks. Likewise, regulations or policies that limit directors’ ability to hold multiple board seats may inhibit the beneficial spillover of knowledge between firms.

REFERENCES

Adams, R., Ferreira, D., 2007. A theory of friendly boards. The Journal of Finance 62 (1), 217-250.

Bloom, N., Schankerman, M., 2013. Identifying technology spillovers and product market rivalry. Econometrica 81, 1347-1393.

Griliches, Z., 1979. Issues in assessing the contribution of research and development to productivity growth. The Bell Journal of Economics, 92-116.

Grossman, G., Helpman, E., 1991. Trade, knowledge spillovers, and growth. European Economic Review 35, 517-526.

Harris, M., Raviv, A., 2008. A theory of board control and size. The Review of Financial Studies 21, 1797-1832.

Raheja, C., 2005. Determinants of board size and composition: A theory of corporate boards. Journal of Financial and Quantitative Analysis 40, 283-306.

This post comes to us from professors Mark A. Chen at Georgia State University’s Robinson College of Business, Shuting (Sophia) Hu at Baylor University, Joanna (Xiaoyu) Wang at Georgia State University, and Qinxi Wu at Baylor University’s Hankamer School of Business. It is based on their recent article, “Board Interlocks, Knowledge Spillovers, and Corporate Innovation,” available here.