Sentiment and Sustainability:How the language of U.S. corporate filings reveals divergent paths for management priorities on DEI and Environment
Bibliographic record
Abstract
Corporations are quick to adjust their disclosures when political or regulatory shifts signal changing expectations from investors, regulators, and other stakeholders. This white paper summarizes the preliminary findings from analysing the language of mandatory annual corporate filings (Securities and Exchange Commission 10-Ks) in the United States before and after the recent re-election of U.S. President Donald Trump. Specifically, we examined the changes in sentiment toward two key themes: environment (climate change and environmental sustainability), and DEI (diversity, equity and inclusion). Both are themes that have been central to corporate strategy and stakeholder engagement in recent years. Shifts in priorities of the new administration — including efforts to scale back ESG frameworks, limit DEI programming, and reduce environmental incentives — have created a regulatory atmosphere in which firms are recalibrating how they present these issues in mandatory disclosures. Our results reveal a stark decline in DEI-related disclosure language following the 2024 election, with average DEI keyword mentions across S&P 1500 firms’ filings falling by nearly 25 percent between 2024 and 2025. The decline is observed across industries, firm sizes, and both Republican (red) and Democratic (blue) states, though it is most pronounced among larger firms. In contrast, climate- and environment-related terminology remained largely stable over the same period, registering only a small and statistically insignificant decrease. These findings suggest that usage of DEI-themed language in 10-K filings is more sensitive and responsive to a changing regulatory environment. This contrasts our results related to usage of environment-themed language, which appears resilient to near-term political shifts, possibly indicating acceptance of this language as a baseline across global capital markets. The divergent trajectories highlight how corporate disclosure priorities are impacted not only by federal policy, but also by other considerations possibly to do with risk assessment, public sentiment or investor expectations.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.004 | 0.014 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.002 | 0.002 |
| Science and technology studies | 0.001 | 0.002 |
| Scholarly communication | 0.005 | 0.004 |
| Open science | 0.000 | 0.002 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.003 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".