Controlling Externalities: Ownership Structure and Cross-Firm Externalities

Our new working paper on “Controlling Externalities: Ownership Structure and Cross-Firm Externalities” develops a general conceptual framework for understanding how firms’ ownership structure and corporate law affect the internalization of cross-firm externalities and proposes a new metric (called “Controller Wealth Concentration”) designed to provide a simple characterization of the incentives of controllers in this regard. We use this metric—and other sources of evidence—to argue that the prevalence of controlling shareholders around the world (including at many important US firms) poses a significant challenge to the internalization of cross-firm externalities through the influence of diversified owners such as index funds. Further, this suggests that working toward better regulation and liability regimes may be inescapable, even though these have their own challenges.

In recent years, debates over the social purpose of corporations have taken center stage in both public discourse and in corporate law scholarship. This development has been spurred by rising concern about externalities generated by the activities of corporations, such as those associated with climate change and harmful speech. A central underlying premise of these debates is that government regulation and liability regimes appear not to be functioning sufficiently well to force firms to internalize these externalities. There is thus rising interest in exploring alternative mechanisms. In particular, a rapidly growing body of scholarship argues that index funds increasingly approximate diversified “universal owners” with incentives to maximize portfolio value (and thus to internalize cross-firm externalities).

However, much of this analysis has focused on diffusely held US firms. In contrast, most firms in the world—including many important firms in the US—have controlling shareholders. These include many of the firms that are thought to be large contributors to the aforementioned externalities. Our paper examines whether index funds can influence such firms to internalize externalities, and explores other potential mechanisms to achieve this aim.

First, we provide novel empirical evidence suggesting that index funds are not well positioned to force controlled firms to internalize their cross-firm externalities. In particular, we use a dataset of engagements by the “Big Three” index funds with firms on climate-related environmental issues (constructed by Azar et al.). We combine this with recent data on the ownership structure of firms around the world, identifying those countries with relatively dispersed ownership structures. We show that index funds’ environmental engagements are strongly concentrated among firms in countries with relatively dispersed ownership structures, controlling for a variety of potentially relevant factors. This suggests an important limitation of the “universal owner” theory—namely, that the ability of index funds to affect firm behavior (via engagements or other means) may be quite limited in the presence of a controlling blockholder.

Second, we document that controlling shareholders are common among the largest firms in the energy, automobile, and technology sectors. In particular, we collect data from various publicly available sources to document the prevalence of controlling shareholders among the largest firms in these sectors (which are widely thought to be associated with important externalities). Across these sectors, the majority of large firms have controlled ownership structures. In the technology sector, virtually all the largest firms are controlled by a founder, the state, or family. Founding families are prominent among large automobile firms, while state ownership is common among large energy firms.

Third, we propose a simple characterization of the incentives of controllers by introducing the concept of “controller wealth concentration” (CWC). This refers to the fraction of a controller’s aggregate personal wealth that consists of stock in the firm that she controls. The lower CWC is the more scope there is for the controller to hold investments in other firms (that are potentially affected by the externalities created by the controlled firm). A low CWC is a necessary condition for controllers to have a pecuniary incentive to take cross-firm externalities into account (indeed, controllers with low CWC may be more effective than index funds in getting controlled firms to internalize their externalities because of their status as controllers). It is not, however, a sufficient condition for controller diversification (for instance, a controller with low CWC may have substantial investments in a handful of other firms that are similar to the controlled firm).

We construct measures of CWC for the controlling shareholders of a global sample of large technology-focused firms. For this sample, CWC is very high relative to that of a diversified portfolio, typically varying from about 50% to close to 100%. This is the case despite the existence of controlling minority ownership structures (CMS)—such as dual class stock—that permit controlling shareholders to exert control while holding modest cash flow rights. Thus, we conclude that undiversified controlling shareholders constitute a significant obstacle to the internalization of cross-firm externalities, limiting the ability of universal owners to encourage their portfolio firms to internalize such externalities.

The divergence between cash flow and control rights associated with CMS has generally been viewed in corporate law scholarship as creating incentives for controllers to extract private benefits of control. Our framework suggests that, in principle, dual class structures (and other CMS) have the hitherto ignored advantage of allowing controllers to diversify their personal wealth (thereby potentially mitigating cross-firm externalities). We show that a legal regime that permits firms to issue dual class stock can potentially make controllers and all other shareholders better off (relative to one that mandates one-share-one-vote) when cross-firm externalities are sufficiently large in relation to the incremental private benefits of control that controllers may extract under the dual class regime.

While this is true in principle, we find that controllers do not in fact typically diversify and lower their CWC even when they maintain control through dual class structures or other CMS. We discuss possible reasons why controllers fail to diversify. These include founders’ over-optimism about their firms, the need to incentivize founders’ ongoing effort, and founders’ incentives to defer capital gains taxes. Finally, we discuss various possible measures that might encourage controllers to diversify. These include minimum float requirements to list on stock exchanges, along with increased taxation of the returns from large blocks of stock in controlled firms. Ultimately, however, we conclude that such measures—even if they could be implemented—are unlikely to have large effects.

The premise of the recent scholarly discussion of mechanisms to internalize corporate externalities is that the traditional solutions—government regulation and liability regimes—have failed or proven to be insufficient. In this context, the “universal owner” theory appears to provide a promising path forward. However, the facts we document about the ownership structure of important firms and about index funds’ environmental engagements, along with the conceptual framework we develop, suggest that this promise is largely limited (at best) to externalities generated by widely-held firms. For the large fraction of corporations around the world (including many large and important US firms) that have controlled ownership structures, it is difficult to identify feasible mechanisms to internalize corporate externalities. Ultimately, this suggests that redoubling efforts towards increased regulation and enhanced liability may be inescapable (even though these traditional solutions present their own significant challenges).

The complete paper is available for download here.