Investors are increasingly interested in whether firms implement environmental, social, and governance (ESG) policies that, for example, reduce the firms’ carbon footprints, diversify their workplaces, or better protect customers’ private information. Some socially beneficial policies (e.g., improving energy-efficiency) may also improve shareholder wealth or would do so with appropriate government actions (e.g., given a carbon tax). However, political realities, and the exhaustion of opportunities that are not costly to the firm, will increasingly present firms with a social dilemma: whether to adopt a policy that benefits society but does not benefit the firm enough to cover its cost. For example, the firm might invest in an expensive tracking system that enables it (and helps enable other firms) to eliminate suppliers that use unsustainable tropical hardwoods or child or forced labor. The overall benefit – reducing tropical deforestation or the abuse of vulnerable populations – might be large but the benefits to the firm’s reputation might not cover the costs of the system and the extra expense of supplies.

The traditional corporate objective of maximizing shareholder wealth would discourage such costly but socially beneficial choices. Hart and Zingales (2022) argue that a more appropriate corporate objective is maximizing shareholder utility, taking into account that shareholders may care about benefits to society. A problem with that objective is that some shareholders care more than others or not at all. Since shares can change hands in the secondary market, it is not obvious which shareholders’ preferences are relevant, and so it is not obvious how to apply the shareholder utility objective.

In a new paper, we examine the firms’ dilemma by building a coherent theoretical model that incorporates shareholder utility and testing the model in a laboratory experiment. Our active agents are investors who are aware of a proposed but costly ESG policy by a given firm. Some investors weigh the public good benefits heavily while others assign little or no weight to those benefits.  In our model, investors with such heterogeneous preferences initially all own shares of the firm’s stock. The investors then play a two-stage game. In the first stage, they participate in a secondary market where they can either sell their shares, buy more, or stand pat. Owning shares conveys the right to vote for or against the ESG policy, so those who desire or oppose the proposed policy need to hold shares for their voices to be heard. After shareholder positions have been established in the first stage of the game, the ESG policy is put to a vote in the second stage. Each share provides one vote, and a majority of shares determines the outcome.

Using a standard rational actor approach, we show that both rejection and approval equilibria are possible in our model.  Whether the ESG policy is approved or rejected depends on the costs borne by the firm and also on the distribution of the weights investors place on the public benefits of the ESG policy. Given that distribution, approval equilibria exist for all policy costs below a certain upper bound, and rejection equilibria exist for policy costs above a certain lower bound.  In many cases, e.g., with evenly dispersed preference weights, there is a substantial overlap between the bounds, so the policy can be either approved or rejected in equilibrium.  In other cases, e.g., with highly polarized preference weights, the bounds coincide and so the equilibrium type is unique. The model also predicts  how the share price in the secondary market depends on the anticipated rejection or approval of the policy.

We test those theoretical predictions in a laboratory experiment with human subjects who take the role of investors. Their social benefit preference weights are induced via the usual financial incentives and thereby tightly controlled. (We chose not to rely on subjects’ true feelings about ESG policies since those are hard to measure and may have distributions that are not useful for testing predictions.)  As predicted, we find that the ESG policy is generally approved when the costs of the policy are low, while it is generally rejected when those costs are high. We also find that the distribution of preferences regarding ESG policy benefits matters.  For policy costs in the range that permits both types of equilibria under dispersed preferences, we find that the approval rate is lower when cost is higher, while under polarized preferences the unique equilibrium predictions find strong support.  We further find that observed prices are generally in line with equilibrium predictions when the policy is rejected.

Observed share prices in the experiment’s secondary market, however, tend to be substantially higher than predicted when the policy is approved. Indeed, observed prices in this case are usually close to the fundamental value of shares when the firm rejects the costly ESG policy. This price anomaly seems to be driven by investors who care least about the social benefits of the policy. The model predicts that in the approval equilibrium those investors will sell their shares at a relatively low price that reflects the firm’s cost of implementing the ESG policy. In the experiment, however, those investors hold out for higher prices. To a large extent they are accommodated by the investors who care most about the policy’s social benefits and who acquire shares to ensure that the policy is approved. It seems that there is a struggle for control when the vote ultimately is for approving the ESG policy, and so the share price includes a substantial voting premium not predicted by the model.

The upshot is that, in our experiment, the prospect of policy adoption generally increases demand for the firm’s shares, and the resulting higher share prices largely compensate for the ESG policy costs. That is, ESG policy adoption in our experiment is less costly to the original shareholders than predicted. To the extent that similar forces are at work in the wider world, this observation may help explain the recent shift in investors’ consideration of ESG policies in the valuation of firms.

REFERENCES

Hart, Oliver D and Luigi Zingales, “The New Corporate Governance,” Technical Report, National Bureau of Economic Research 2022.

Paper: Duffy, John, Friedman, Daniel, Rabanal, Jean Paul and Rud, Olga, A., “Trade, Voting, and ESG Policies: Theory and Evidence” https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4306156.

This post comes to us from professors Daniel Friedman at the University of California at Santa Cruz, John Duffy at the University of California at Irvine, and Jean Paul Rabanal and Olga Rud at the University of Stavanger. It is based on their recent paper, “Trade, Voting, and ESG Policies: Theory and Evidence,” available here.