Public Sentiment Decomposition and Shareholder Actions

Public sentiment regarding corporate practices has become increasingly pronounced, particularly with the rise of social media and the democratization of information. This heightened public engagement encompasses a variety of issues including, for example, a company’s financial performance, products, environmental policies, treatment of employees and corporate governance practices. Traditional media coverage and social media interactions serve as platforms for capturing public sentiment. The sentiment can not only influence a corporation’s management and its board of directors, but it also affects shareholders, including large institutional investors such as mutual funds, pension funds, and asset managers. Given their role as stewards of capital, institutional investors typically monitor public sentiment alongside conducting their own independent research to inform their investment decisions.

The question that arises is whether public opinion influences shareholders’ actions, which we capture through number of shareholder proposals and support rate for director elections. Alternatively, the public sentiment may be irrelevant to shareholders’ actions, as different stakeholders might simply follow their own financial and nonfinancial motives that do not necessarily overlap with the public sentiment. Therefore, understanding the effect of public sentiment on shareholder actions is important, especially with the ever-increasing push for democratization of information. The retailer Target is an example of a firm that received severe backlash in 2023 with customers boycotting its products and it had to immediately make changes to its product offerings. The firm had one shareholder proposal in both 2022 and 2023, but they subsequently had five in 2024.

We study the relationship between public sentiment and shareholder actions and provide two innovative contributions. First, we use unique measures of public sentiment to capture not only a firm’s traditional financial and governance attributes but also newer issues of interest such as climate risk and social issues. Our proxy for sentiment is obtained from LSEG’s MarketPsych Analytics based on sentiment data generated using state-of-the-art textual analysis and machine learning on a large collection of news and social media content.

Our paper also introduces a novel approach to capture shareholder dissatisfaction with firm management. Rather than using the voting outcomes for shareholder proposals to capture shareholder concerns, we use the number of shareholder proposals as our primary measure to capture shareholder dissent. For example, the Amazon proxy statement in 2022 contained fifteen shareholder proposals, the number increased to eighteen in 2023 and then dropped to fourteen in 2024. There were concerns about a wide variety of issues and the proposals varied from requests for management to adopt a policy to consider employees’ pay grades in setting executive compensation targets, establish a board committee on public policy, or provide reports on such issues as retirement plan options related to climate goals, racial and gender pay gaps, warehouse working conditions, packaging materials, and tax transparency among a number of other concerns.

Our approach has the advantage that the number of shareholder proposals measure is a more complete representation of shareholder concerns in any given year. Further, it allows us to examine the association between public sentiment and shareholder dissent for all firms, even those without shareholder proposals. The absence of shareholder proposals is important information in itself and unlike vote totals, our measure allows us to capture this information. Besides being few in number, especially relative to management-sponsored proposals, shareholder proposals are also often withdrawn, usually because management and the sponsoring shareholder reach an agreement before the actual vote. Therefore, our analysis is conducted both with and without withdrawn proposals.

We find that negative public sentiment about a firm on both financial and broad ESG aspects are significantly related to the number of shareholder-sponsored proposals. In addition, the two subcomponents: (1) environment and social (E&S) sentiments, and (2) governance (G) sentiment, also show significant relationships with the number of shareholder proposals that are independent of each other, affecting shareholder proposals on their corresponding category (i.e., the E&S sentiment matters for E&S proposals, while the G sentiment matters for G proposals). Moreover, our data allows us to examine the sentiment reflected in social media sources separately from the sentiment reflected in news sources. We find that both sentiment sources are important in the extent of proposals submitted by shareholders, with the news sources being slightly stronger in affecting the number of shareholder proposals. To ascertain whether the effect of public sentiment on shareholder proposals is causal, we use a creative instrumental variable approach. The use of scandal movies, which expose a firm’s past scandals related to corporate misconduct, as an instrumental variable further strengthens our argument by providing causal evidence that negative ESG sentiment leads to increased shareholder dissent.

To further study the relevance of our metric – the number of shareholder proposals representing shareholder dissent – we examine whether the number of shareholder proposals has any consequences for directors and management of the firm. We find a strong association between the number of shareholder proposals on the ballot and director turnover at the firm. There is also an association between our measure of shareholder dissent and forced turnover of CEOs.  In terms of economic significance, one additional shareholder proposal is associated with a 10.9% increase in director turnover and a 24.8% increase in forced CEO turnover, both relative to the mean.

We also examine the association between director elections and public sentiment, since we know from prior literature that many investors express their concerns through votes against directors. Lower public sentiment leads to a lower support rate for management-sponsored directors. All the components of the public sentiment, from ESG to financial sentiments, appear to matter for the director support rate.

This study not only enhances our understanding of the mechanisms driving shareholder activism but also highlights the importance for firm management to maintain a positive public image to mitigate potential shareholder conflicts.