The reviewed record of science sign in
Pith

arxiv: 2606.31469 · v1 · pith:CHOR5A5S · submitted 2026-06-30 · q-fin.GN

Same Firms, Different Verdicts: ESG Rating Choice and the Measurement of Greenwashing

Praveen Kumar Ashok Kumar , Rafa{\l} Sieradzki This is my paper

Reviewed by Pith T0 review T1 audit T2 compute T3 formal T4 kernel 2026-07-01 02:37 UTCgrok-4.3pith:CHOR5A5Srecord.jsonopen to challenge →

classification q-fin.GN
keywords greenwashingESG ratingsdisclosure performance gapindex membershipenvironmental disclosureEuropean firmsaggregate confusion hypothesis
0
0 comments X

The pith

The measured gap between environmental disclosure and performance varies by ESG rating, with flagship index membership predicting larger gaps only under the CDP score.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

This paper measures the Disclosure-Performance Gap as the standardized difference between firms' voluntary environmental disclosures and their actual emissions outcomes for 200 large European companies. It tests what predicts this gap and finds that membership in flagship stock indexes is the strongest factor widening it, pointing to ceremonial conformity in reporting. Renewable energy use and environmental spending narrow the gap, while governance variables do not. Replacing the CDP Climate Score with the LSEG Environmental Pillar Score removes the index effect, demonstrating that conclusions about greenwashing depend on the chosen rating lens.

Core claim

The paper establishes that the Disclosure-Performance Gap is significantly wider for firms in flagship indexes, with a coefficient of 0.78, consistent with institutional pressures for symbolic compliance rather than substantive performance. This effect disappears when using an alternative rating source, while actual investments in renewables and capital expenditure reduce the gap, supporting signalling over monitoring explanations.

What carries the argument

The Disclosure-Performance Gap (DPG), defined as the standardized divergence between voluntary disclosure and realized emissions performance, compared across CDP and LSEG ratings to identify rating-dependent predictors.

If this is right

  • Flagship index membership correlates with larger gaps under CDP ratings.
  • Renewable energy use and environmental capex narrow the measured gap.
  • TCFD endorsement shows a positive but directional association with wider gaps.
  • Governance and monitoring variables have no explanatory power for the gap.
  • Detected greenwashing is conditional on the ESG rating applied.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • Rating agencies' differing methodologies can lead to inconsistent greenwashing identifications across the same firms.
  • Investors relying on single ratings may misjudge firm environmental commitment.
  • Future studies could test if similar patterns hold after mandatory reporting under CSRD begins.
  • The ceremonial conformity effect may apply to other ESG dimensions like social or governance scores.

Load-bearing premise

The standardized divergence between disclosure and emissions performance isolates intentional greenwashing rather than differences in reporting incentives or measurement errors.

What would settle it

Observing no difference in the gap for index members when using a third independent rating source or when emissions data is verified by a single standard would challenge the rating-dependence claim.

Figures

Figures reproduced from arXiv: 2606.31469 by Praveen Kumar Ashok Kumar, Rafa{\l} Sieradzki.

Figure 1
Figure 1. Figure 1: Unified conceptual framework Substantive (costly) signals Renewable energy Environmental capex Symbolic (visibility) signals Index membership TCFD support Moderating controls Board diversity, financial slack SBTi, analyst coverage Signalling theory Costly, irretrievable outlay → separating equilibrium Institutional theory Coercive isomorphism; policy-practice decoupling Agency & criti￾cal mass theory Monit… view at source ↗
Figure 2
Figure 2. Figure 2: Stepwise AICc progression across the forward-selection search. [PITH_FULL_IMAGE:figures/full_fig_p010_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Information-criterion comparison across the five variable pools. [PITH_FULL_IMAGE:figures/full_fig_p010_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: Variance Inflation Factors across candidate pools; all values are well below the threshold of 5.0. [PITH_FULL_IMAGE:figures/full_fig_p011_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: Distribution of the Disclosure-Performance Gap by sector [PITH_FULL_IMAGE:figures/full_fig_p012_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: Disclosure-Performance Gap by sector Note: Medians are pinned near zero by within-sector standardisation; the boxes and whiskers display dispersion and tail behaviour, not a cross-sector greenwashing ranking. Source: Author’s own elaboration Bivariate Pearson correlations track the theoretical predictions. Among substantive signals, renewable energy has the strongest negative correlation with the DPG (r = … view at source ↗
Figure 7
Figure 7. Figure 7: Renewable energy coefficient under the two disclosure proxies [PITH_FULL_IMAGE:figures/full_fig_p016_7.png] view at source ↗
Figure 8
Figure 8. Figure 8: Coefficient estimates under the CDP and LSEG disclosure proxies on the identical 200 firms [PITH_FULL_IMAGE:figures/full_fig_p017_8.png] view at source ↗
Figure 9
Figure 9. Figure 9: Diagnostic plots for the full sample (n=200), before outlier removal [PITH_FULL_IMAGE:figures/full_fig_p022_9.png] view at source ↗
Figure 10
Figure 10. Figure 10: Diagnostic plots for the cleaned-sample specification (n=184), after outlier removal [PITH_FULL_IMAGE:figures/full_fig_p023_10.png] view at source ↗
Figure 11
Figure 11. Figure 11: Coefficient estimates with HC3 95 per cent confidence intervals, full sample (n=200) versus cleaned sample [PITH_FULL_IMAGE:figures/full_fig_p024_11.png] view at source ↗
Figure 12
Figure 12. Figure 12: Correlation matrix of the dependent variable and candidate predictors. [PITH_FULL_IMAGE:figures/full_fig_p025_12.png] view at source ↗
Figure 13
Figure 13. Figure 13: Pairwise relationships among the key variables. [PITH_FULL_IMAGE:figures/full_fig_p025_13.png] view at source ↗
read the original abstract

This paper investigates the Aggregate Confusion hypothesis (Berg, Kolbel, and Rigobon, 2022) at the firm level by measuring the Disclosure-Performance Gap (DPG), the standardised divergence between a firm's voluntary environmental disclosure ("Talk") and its realised emissions performance ("Walk"). The sample comprises 200 large European firms from the Energy, Materials, Industrials, and Utilities sectors of the STOXX Europe 600 in fiscal year 2023, the final cross-section of the voluntary reporting era before the Corporate Sustainability Reporting Directive. The model is selected through a six-stage process, candidate assembly, correlation screening, VIF based multicollinearity filtering, stepwise forward search under the corrected Akaike Information Criterion, Cook's distance screening, and HC3 re-estimation across 421 candidate specifications, estimated by ordinary least squares with HC3 robust standard errors on the full sample. Flagship index membership is the strongest predictor of a wider gap ($\beta$ = +0.78, p < 0.01), consistent with institutional ceremonial conformity. TCFD endorsement is also positive ($\beta$ = +0.86, p < 0.05) but identified off a small group of non-supporting firms, so it is read as directional, not a precise magnitude. Renewable energy use ($\beta$ = -0.31, p < 0.01) and environmental capital expenditure ($\beta$ = -0.22, p < 0.05) significantly narrow the gap, consistent with signalling theory, while governance and monitoring variables carry no explanatory power. Results are robust to influence trimming, rank-based recoding of the disclosure score, and removal of the TCFD variable. Replacing the CDP Climate Score with the LSEG Environmental Pillar Score eliminates the index-membership effect while the renewable-energy effect survives, showing that detected greenwashing is conditional on the rating lens applied.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

3 major / 2 minor

Summary. The paper investigates the Aggregate Confusion hypothesis at the firm level by constructing a Disclosure-Performance Gap (DPG) as the standardized divergence between voluntary environmental disclosure (CDP Climate Score) and realized emissions performance for 200 large European firms from Energy, Materials, Industrials, and Utilities sectors of the STOXX Europe 600 in 2023. Using a six-stage model selection process (candidate assembly, correlation screening, VIF filtering, stepwise forward AIC search, Cook's distance, HC3 re-estimation) on 421 specifications estimated by OLS with HC3 robust errors, it reports flagship index membership as the strongest predictor of wider DPG (β = +0.78, p < 0.01), consistent with ceremonial conformity; renewable energy use and environmental capex narrow the gap, while results are robust to trimming and rank recoding but the index effect vanishes when CDP is replaced by the LSEG Environmental Pillar Score.

Significance. If the DPG construction isolates greenwashing, the finding that detected greenwashing is conditional on the rating lens contributes to the Aggregate Confusion literature by documenting firm-level heterogeneity in measurement. The transparency in reporting the vanishing index effect under the alternative rating, the six-stage specification search with multiple robustness checks, and the use of HC3 errors are strengths that enhance credibility of the estimation approach.

major comments (3)
  1. [DPG construction] DPG construction (abstract and variable definition): The performance component relies on absolute (unnormalized) emissions. Flagship index members are systematically larger firms within the STOXX 600 sample, which can mechanically widen the standardized divergence even absent any disclosure-performance mismatch. The six-stage search and reported robustness checks do not indicate tests of emissions intensity, log(assets), or revenue as controls or alternative DPG definitions, leaving open a scale confound that directly affects the central interpretation of the flagship coefficient as ceremonial conformity.
  2. [Model selection process] Model selection (six-stage process described in abstract): Stepwise forward selection under corrected AIC across 421 candidates is known to inflate Type I error rates and produce overfitted models; while VIF and Cook's distance are applied, this procedure remains load-bearing for the reported flagship index coefficient and its p-value.
  3. [Results on TCFD] TCFD result (abstract): The positive coefficient (β = +0.86, p < 0.05) is identified off a small non-supporting subgroup, so the estimate cannot support a precise magnitude claim even if the directional sign is retained.
minor comments (2)
  1. [Methods] A summary table listing the number of specifications retained after each of the six stages would improve transparency of the selection process.
  2. [Abstract] The abstract states the sample comprises firms from four sectors but does not report the exact sector breakdown or any sector fixed effects; adding this detail would clarify generalizability.

Simulated Author's Rebuttal

3 responses · 0 unresolved

We thank the referee for their thoughtful comments, which help strengthen the paper. We address each major comment below, indicating where revisions will be made.

read point-by-point responses

  1. Referee: [DPG construction] DPG construction (abstract and variable definition): The performance component relies on absolute (unnormalized) emissions. Flagship index members are systematically larger firms within the STOXX 600 sample, which can mechanically widen the standardized divergence even absent any disclosure-performance mismatch. The six-stage search and reported robustness checks do not indicate tests of emissions intensity, log(assets), or revenue as controls or alternative DPG definitions, leaving open a scale confound that directly affects the central interpretation of the flagship coefficient as ceremonial conformity.

    Authors: We acknowledge this potential scale confound. Although the DPG is constructed as a standardized measure, larger firms may have higher absolute emissions. To address this, we will incorporate log(assets) as an additional control variable in the model selection process and test an alternative DPG definition using emissions intensity (emissions per unit of revenue). These additional specifications will be reported in a revised robustness section. revision: yes

  2. Referee: [Model selection process] Model selection (six-stage process described in abstract): Stepwise forward selection under corrected AIC across 421 candidates is known to inflate Type I error rates and produce overfitted models; while VIF and Cook's distance are applied, this procedure remains load-bearing for the reported flagship index coefficient and its p-value.

    Authors: We recognize the limitations of stepwise selection procedures, including potential inflation of Type I errors. However, our six-stage process includes correlation screening, VIF filtering to address multicollinearity, Cook's distance for influence, and HC3 robust errors. Moreover, the flagship index result is robust to trimming, rank recoding, and alternative rating measures. In revision, we will add a discussion of these limitations and report results from a pre-specified model that includes the key variables of interest without relying solely on stepwise selection. revision: partial

  3. Referee: [Results on TCFD] TCFD result (abstract): The positive coefficient (β = +0.86, p < 0.05) is identified off a small non-supporting subgroup, so the estimate cannot support a precise magnitude claim even if the directional sign is retained.

    Authors: We agree with this assessment. The manuscript already notes that the TCFD coefficient is identified off a small group of non-supporting firms and should be interpreted as directional rather than providing a precise magnitude. No further change is required on this point. revision: no

Circularity Check

0 steps flagged

No significant circularity detected in the empirical derivation

full rationale

The paper's central results are obtained via OLS estimation (with HC3 standard errors) of a linear model for the Disclosure-Performance Gap on a cross-section of 200 firms. The reported coefficients, including the flagship-index effect, are direct statistical outputs of that regression and do not reduce by construction to any fitted parameter or self-referential definition. No self-definitional constructions, fitted-input predictions, load-bearing self-citations, or ansatz smuggling appear in the six-stage specification search or robustness checks. The analysis is therefore self-contained against external benchmarks.

Axiom & Free-Parameter Ledger

1 free parameters · 1 axioms · 1 invented entities

The central claim rests on the validity of the newly constructed DPG measure, standard OLS assumptions after filtering, and the absence of selection bias from the multi-stage model search.

free parameters (1)
  • OLS regression coefficients = 0.78 for flagship index
    Coefficients including the index-membership beta of 0.78 are estimated from the 200-firm sample.
axioms (1)
  • domain assumption OLS assumptions (linearity, homoscedasticity after HC3, no multicollinearity after VIF) hold for the selected specification.
    Invoked for all reported estimates and robustness checks.
invented entities (1)
  • Disclosure-Performance Gap (DPG) no independent evidence
    purpose: To quantify firm-level greenwashing as the standardised divergence between talk and walk.
    New construct introduced for this study; no external validation provided.

pith-pipeline@v0.9.1-grok · 5887 in / 1345 out tokens · 74792 ms · 2026-07-01T02:37:23.630400+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

7 extracted references · 7 canonical work pages

  1. [1]

    Aggregate Confusion

    INTRODUCTION The integration of Environmental, Social, and Governance (ESG) criteria into capital allocation is one of the most consequential structural changes in modern finance. ESG-related assets under management have surpassed USD 30 trillion and are projected to approach USD 40 trillion by 2030 (Bloomberg Intelligence, 2024). This accumulation of cap...

    work page 2030

  2. [2]

    brownwashing

    LITERATURE REVIEW AND THEORETICAL FRAMEWORK The divergence between corporate environmental disclosure and environmental performance has generated a substantial but theoretically fragmented literature. This section proceeds in four steps. It first traces the shift from binary, scandal-based measures of greenwashing to the continuous, gap-based measure adop...

    work page 2011

  3. [3]

    It first describes the sample and the rationale for restricting the analysis to four carbon-intensive sectors in a single fiscal year

    DATA AND METHODOLOGY This section sets out the empirical design in three steps. It first describes the sample and the rationale for restricting the analysis to four carbon-intensive sectors in a single fiscal year. It then defines the dependent variable and the candidate independent variables, including the standardisation that places disclosure and perfo...

    work page 2023

  4. [4]

    RESULTS This section reports the empirical findings in the order in which a reader needs them. It first profiles the sample and the distribution of the dependent variable, then establishes the bivariate associations that anticipate the multivariate results, verifies the ordinary least squares assumptions, and reports the primary hybrid regression with its...

    work page arXiv

  5. [5]

    CONCLUSIONS This paper provides evidence that the pre-CSRD voluntary reporting regime in European capital markets failed to align corporate environmental disclosure with operational performance. Rather than functioning as an accountability mechanism, the voluntary framework created a bifurcated market in which symbolic visibility was rewarded over substan...

    work page 1977

  6. [6]

    critical mass

    REFERENCES Becker, B. and Milbourn, T. (2011) 'How did increased competition affect credit ratings?', Journal of Financial Economics, 101(3), pp. 493–514. Berg, F., Kölbel, J.F. and Rigobon, R. (2022) 'Aggregate confusion: The divergence of ESG ratings', Review of Finance, 26(6), pp. 1315–1344. Bingler, J.A., Kraus, M., Leippold, M. and Webersinke, N. (20...

    work page 2011

  7. [7]

    APPENDICES Appendix A. Regression diagnostics and coefficient robustness Figure 9: Diagnostic plots for the full sample (n=200), before outlier removal Note: The normal Q-Q plot and residual histogram show the departure from normality (Jarque-Bera = 49.66, p < 0.001) driven by the 16 influential observations. Source: Author’s own elaboration 22 Figure 10:...

    work page