Sonakshi (Sona) Agrawal

The data centers behind artificial intelligence are raising U.S. electricity demand after a decade of flat growth. Several states have responded by reviving Construction Work in Progress (CWIP), a cost-recovery rule that lets a utility begin collecting from customers while an asset is still being built rather than only after it enters service. Using the historical wave of state-level CWIP adoptions in a staggered difference-in-differences design, I find that utilities build more and finish construction faster once their state allows CWIP. I find less evidence for an eased financing constraint and more for a higher payoff to investing. Looking beyond investment, the data do not suggest customers are better served. In the current wave, CWIP states are already planning far more customer-funded capacity than states without the rule, where new building is left to independent producers. These results contribute to the emerging literature on financing the infrastructure demands of artificial intelligence.

We explore the role of ESG raters’ business models in the production of their ratings, noting that increasingly ESG raters not only produce ESG ratings but also construct and sell ESG indices. We examine whether index licensing incentives are associated with how ESG ratings relate to stock returns. By comparing ratings across raters for the same firm, we control for fundamental ESG performance. We find that raters with strong index licensing incentives assign higher ESG ratings to firms with better stock performance compared to raters with weaker licensing incentives, and that this association is concentrated among index-eligible firms. We further show that index construction is significantly associated with stock returns, and that the high-incentive rater adjusts its ratings following index composition changes, while the low-incentive rater does not. Our findings highlight the need for greater transparency in the incentives of ESG rating producers.

This paper integrates two earlier working papers: “ESG Ratings of ESG Index Providers” by Agrawal, Liu, and Rajgopal, and “ESG Rating Agency Incentives” by Sridharan, Yan, and Yohn.

Climate disclosure standards, including the ISSB’s IFRS S1 and S2, primarily target publicly traded firms. We test whether this focus aligns with where climate risk actually falls. Using establishment-level data covering 8.9 million private and 13,513 public parent firms matched to over two decades of federal disaster declarations, we find a striking divergence. Public firms experience no significant operating performance effects from climate disasters, while private firms experience sales declines of 0.7% on average at their local operations in disaster-affected counties. The declines are concentrated in manufacturing, retail, chemicals, and healthcare, and persist for several years in capital-intensive industries. We examine geographic diversification and access to credit as candidate mechanisms. Both partially explain the public-private differential, but neither mechanism protects firms in capital-intensive industries that cannot easily relocate production. The firms most exposed to climate disasters are private firms in specific industries, precisely those for which investors lack standardized climate risk information. Our findings inform the international debate about extending sustainability disclosure beyond publicly traded firms.

Corporations increasingly use investor-state dispute settlement (ISDS) to challenge foreign regulations through international arbitration, raising concerns about corporate influence over government policymaking. Using comprehensive data on 1,023 ISDS cases, we document three findings. First, among firms that file ISDS cases, public firms are more likely than private firms to challenge industry-wide regulations, while showing no differential propensity to challenge firm-specific rules. This pattern is consistent with public firms having larger industry stakes that reduce free-riding incentives. Second, industry-wide challenges filed by public firms take significantly longer to resolve. This difference persists under identical investment treaties, a pattern more consistent with regulatory delays than case complexity. Third, cross-sectional tests reveal these patterns are strongest in countries with strong shareholder protections and for regulations protecting foreign stakeholders. Stock markets respond positively to ISDS filings, particularly for challenges targeting stakeholder-protective regulations. Our findings suggest public firms use ISDS in ways that prioritize domestic shareholder interests over foreign stakeholder welfare.

The paper examines how mandatory disclosure of voting rationales reshapes institutional voting and firm governance, using a 2014 Indian mandate that requires mutual funds to justify every vote. We find that abstentions fall by about a quarter, with dissent concentrated in auditor approvals, executive compensation, and director appointments. Disclosure quality is overwhelmingly a fund-level decision rather than a firm characteristic. Treated firms reduce CEO duality, busy directorships, and dividends, while raising board financial expertise.