Divestment and Engagement: The Effect of Green Investors on Corporate Carbon Emissions

Green investors, frustrated with what they see as inaction by governments, increasingly seek to use capital markets to pressure companies to combat climate change. These efforts have brought to the surface an important debate about what is the most effective strategy – should investors divest from fossil companies in order to deprive them of capital and free up resources for clean energy, or should they acquire fossil fuel stocks and use their ownership rights to press for emission cuts?

In a new empirical study we examine the competing arguments in this debate. We estimate how corporations adjusted their carbon emissions in response to a change in the composition of their shareholders: did they reduce emissions when green investors divested or when they invested? The evidence points to a clear conclusion: engagement is better than divestment for investors that want companies to reduce carbon emissions. Green investors make companies greener.

In 2021, Maine became the first state to require its public pension funds to divest from fossil fuel businesses; the huge New York State and New York City pensions have announced their intention to stop investing in fossil fuel companies; and California lawmakers are advancing legislation to force the state’s two massive pension funds, CalPERS and CalSTRS, to do the same. By one estimate, almost $40 trillion in assets has been committed to divestment. Yet in a statement opposing California’s divestment bill, CalPERS argued that “divestment has little – if any – impact on a company’s operations and therefore does nothing to reduce greenhouse gas emissions. . . . The companies in question can easily replace CalPERS with new investors, ones who are unlikely to speak up as loudly or as consistently as we have about the urgent need to move toward a low-carbon economy.” And CalSTRS added that “it is important that long-term investors, such as CalSTRS, actively engage fossil fuel companies . . . to transition their business models to cleaner forms of energy,” and that divestment would “severely hinder” such collaboration.

To assess these competing views, we examine how facility-level corporate greenhouse gas emissions changed in response to changes in ownership by green investors over the period 2010-2021. We focus on stock ownership by an important class of investors – public pension funds –  which control $5.6 trillion in assets. We classify a public pension fund as green if it was controlled by Democrats and not green if it was controlled by Republicans. We define control based on the party of the governor (because governors can influence pension fund investment through legislative and regulatory actions and by appointing pension board trustees) and by the majority party on a fund’s board of trustees (because trustees can influence pension funds’ operations directly).

To study this issue, we need to quantify the effect of changes in green ownership on a company’s carbon emissions. The challenge is that green ownership of a company might change in reaction to a company’s actions, and not necessarily cause those actions. We address the issue of causality by focusing on two sources of variation in green ownership that are arguably exogenous with respect to company emissions. The first source is changes in control of a pension due to changes in the political party of the governor and the trustees. The second source is changes in stock holdings by a given fund due to variation in the fund’s returns on the non-public-equity components of its portfolio. Since public pension funds typically maintain target ratios for their investment in public equities versus other asset classes such as private equity, real estate, and commodities, changes in the value of non-public-equities holdings require a fund to acquire more public equities to restore its target ratio. We show that this rebalancing, which is also unconnected to emissions in portfolio companies, is a strong predictor of changes in a fund’s stock holdings.

Our key finding is that an increase in the fraction of shares held by green public pension funds caused companies to reduce their carbon emissions. In our baseline estimate, a 1 percentage point increase in a company’s shares held by green pension funds was associated with an approximately 3 percent reduction in plant emissions over four years. In comparison, we find no association between non-green ownership and changes in corporate carbon emissions. We conduct a battery of tests and find that the basic pattern is highly robust. Engagement reduced emissions; divestment did not. This suggests that divestment strategies were counterproductive – causing increased emissions compared to the alternative of holding the stock.

We also investigate potential mechanisms by which ownership of a company’s stock might lead to emission cuts. First, the mere ownership of a company’s stock by green shareholders might cause corporate managers to alter company policies, as would be the case if managers sought to maximize shareholder utility. Second, ownership might allow investors to engage management. Engagement can be collaborative, involving expressions of preferences, sharing of knowledge, and attempted persuasion, or it can be confrontational, such as voting against incumbent directors and sponsoring shareholder proposals. The evidence we assemble is largely suggestive but points to a primary role for collaborative engagement compared to confrontation. We find that emissions reductions were more strongly associated with ownership by pensions known for actively engaging management than more passive funds, and we do not find that green ownership attracted more shareholder proposals or that green pension funds had markedly more confrontational voting records.

In addition to environmental issues, our findings speak to the broader issue of stock divestment as a strategy to achieve social goals. Activists have tried to advance social causes through divestment going back at least to the 1980s, when divestment from apartheid South Africa was common. California pension funds have divested from Iran, Sudan, thermal coal, tobacco companies, and gun manufacturers. Previous evidence on divestment has focused on its effect on financial markets and asset prices (often finding minimal effects) or on the return earned by the divesting funds, but little is known about whether financial markets can cause companies to change their real behavior. Our paper provides some of the first direct evidence on the real effects of divestment, suggesting that it may be ineffective and possibly counterproductive compared to engagement.