We develop a model of impact activism where activist investors contribute to a firm's green transition through engagement. Two intertwined free-rider problems hinder activists' impact. First, an internal free-rider problem, where insiders and activists free-ride on each others' transition-related efforts, can render activism ineffective or even counter-productive. Second, an external free-rider problem, where activism-driven gains are reflected in the stock price and accrue to passive investors, prevents activists from investing or tilts investments towards firms that can transition easily without activist engagement. Carbon taxation strengthens these frictions, while sustainability preferences help overcome them.

We study how firms choose to allocate decision powers to shareholders or management in a historical context where control allocation was not mandated by corporate law but ownership structures were constrained by differences in regional wealth distributions. Firms with numerous small -and plausibly less informed- owners tend to disempower shareholders and increase managerial authority. Moreover, shareholder empowerment is accompanied by provisions promoting information acquisition by owners, suggesting that control and information are complements. Our findings highlight how the debate on shareholder empowerment should account for the critical role of ownership structure.

We construct a unique dataset on mandatory retirement policies for independent directors at U.S. public firms from 1994 to 2020 by combining machine learning techniques with manual inspections. We investigate factors that drive firms to adopt mandatory retirement policies and assess the value implications of such decisions. We document an increasing trend in policy adoption, particularly among S&P 1500 firms. Our results indicate that firms with greater monitoring needs are more likely to implement mandatory retirement policies, while those with greater advising needs are less likely to do so. Moreover, mandatory retirement policies are associated with an increase in firm value when monitoring benefits are high but a decrease in value when advising benefits are high. This study is the first to examine both the determinants and outcomes of mandatory retirement policies, suggesting that their value effect depends on firms’ specific governance needs.