How Indian Boards Should Read ESG Rating Divergence

An Indian listed firm can sit in the top quartile of one ESG rating universe and the third quartile of another in the same quarter. Management then arrives at the board with a slide that cherry-picks the flattering number. A lender arrives with the other number. Both rooms treat the gap as a communications problem. It is usually a measurement problem. This briefing shows how boards, CFOs and credit committees in India should read ESG rating divergence as a diagnostic, not as a contest to be won with a press release.

The setting is India after BRSR, BRSR Core, reasonable-assurance pilots, and a crowded market of domestic and global ESG scores. The discipline travels to other emerging markets where disclosure is catching up faster than the underlying data systems.

1. Why two ESG scores for the same firm can both be honest

ESG ratings are not credit ratings with a second coat of paint. A credit rating is anchored to a relatively narrow question: will contractual cash flows be paid. An ESG rating is a composite of contested constructs—climate exposure, labour practice, board quality, product harm, community impact—weighted by a vendor’s own theory of what “matters.” Two vendors can look at the same annual report and produce different ranks without either vendor fabricating a number.

Academic work on rating disagreement has been consistent on this point: correlations across major vendors are often modest, sometimes closer to 0.4–0.6 than to the near-unity correlations boards expect from credit ratings. The disagreement is largest on the “S” and on forward-looking environmental metrics, and smaller on simple governance observables such as board size or auditor change. Indian firms add a further layer: vendors differ in how they treat group structures, related-party intensity, contractor labour, and Scope 3 that is still estimated rather than measured.

Boards go wrong in three predictable ways. First, they average the scores and call the average “our ESG position.” Averaging hides the issue that actually moved. Second, they shop the vendor that ranks them best and put that logo on the sustainability page. Third, they commission a rebuttal letter to the low vendor and ignore the data gap the letter accidentally reveals.

2. The four engines of rating divergence

Before a board argues with a vendor, it should classify the disagreement. Most Indian cases fall into four engines. They require different responses.

A fifth, quieter engine matters in India: language and document access. Vendors that read only English PDFs will miss plant-level consent conditions, state-pollution-board orders, and labour-inspector reports that sit in regional-language annexes. A “low social score” can be a retrieval problem. It can also be a real problem that the English sustainability report never mentioned.

3. How BRSR, ratings and internal dashboards talk past each other

India now runs three parallel ESG information systems that boards often treat as one. They are not one.

BRSR and BRSR Core are issuer disclosures under a regulator’s template. They are comparable across listed firms only to the extent the firm applied the same definitions, the same organisational boundary, and—where required—the same assurance standard. They are not a rating. A complete BRSR can still describe a high-emission, high-attrition, related-party-heavy business.

Vendor ESG ratings are third-party compressions. They mix BRSR-like disclosures with media, NGO, and modelled data. They exist to sort a universe for an asset manager who will not read 220 pages.

Internal ESG dashboards are management’s own KPIs: specific energy, water intensity, lost-time injury frequency, gender mix in the officer cadre, supplier audit closure rates. These are the numbers that can actually change a plant manager’s bonus. They are also the numbers most likely to be missing from the vendor file.

The board’s job is translation. If Vendor A penalises the firm for Scope 3 and the internal dashboard still reports only Scope 1 and 2 for owned plants, the disagreement is not mysterious. If Vendor B rewards a diversity policy while the officer pipeline has not moved in four years, the disagreement is a weighting choice plus a measurement choice. Write both sentences in the board pack. Do not write “the market does not understand our journey.”

Assurance changes the conversation only when the assured line items are the same line items the vendor uses. Assuring a community-spend rupee figure will not move a climate-weighted score. Boards that buy assurance as a reputation product, rather than as a measurement product, are routinely surprised by this.

4. A board pack that treats divergence as information

Replace the single-score slide with a one-page divergence memo. Keep it to six blocks. The company secretary can own the calendar; the CFO should own the financial mapping; the sustainability head should own the issue-level reconciling items.

4.1 The scorecard of scores

List every vendor the firm is scored by, the date of the last refresh, the peer set, the overall rank or score, and the three weakest issue pillars. Add the BRSR Core indicators that map to those pillars. If a mapping does not exist, say so. Silence is how weak pillars survive two more years.

4.2 The engine label

For each material gap versus last year or versus the median vendor, name the engine from Section 2. “Peer-set change after the overseas acquisition” is a different action item from “unassured fugitive emissions at the new mine.”

4.3 The financial hook

Boards act when a number touches capital. State whether the weak pillar sits inside a loan covenant, an export-customer questionnaire, a transition-finance eligibility screen, or a key-managerial-personnel scorecard. If it sits in none of those, say that too—and then ask whether it should.

4.4 The data gap list

Five lines, maximum. Missing contractor hours. Scope 3 category 1 estimated with an industry average. Water data for two leased warehouses. Gender data that stops at the listed parent. A pollution-board direction not yet reflected in the English pack. Each line needs an owner and a month.

4.5 What management will not chase

Some vendor flags are not worth a project. A global methodology that punishes concentrated promoter ownership as if it were a New York dual-class structure may be a fact of Indian listed life, not a defect to “fix” with a consultant. Record the non-chase so the next meeting does not reopen it as a branding issue.

4.6 What will be true in twelve months

One quantitative commitment per material pillar. Not “improve the score.” “Publish an assured Category 1 and Category 4 inventory for the top twenty suppliers by spend.” Scores may still diverge. The firm will have a number it can defend.

5. What lenders and investors should do with the spread

Indian banks and NBFCs are being asked to bring ESG into credit files without being given a single official score. That is fortunate. A single official score would hide the same disagreement under a public seal. The useful credit practice is to treat the spread itself as a risk indicator.

High agreement across vendors on a weak environmental pillar is more informative than a high average with a wide spread. High agreement on governance red flags—auditor resignation, repeated related-party spikes, delayed BRSR—should move the file even if one boutique vendor is still complimentary. Wide spread with no controversy news often means the firm is mid-transition: new plants, new geographies, incomplete Scope 3. That is a monitoring trigger, not an automatic downgrade.

For relationship managers, three questions beat a dashboard traffic light. What organisational boundary did each vendor use. Which of the borrower’s material credit risks—water for a textile mill, thermal coal offtake for a generator, contractor safety for an EPC firm—actually enter the score. What would have to be true in the next audited year for the low score to be wrong. If the borrower cannot answer the third question, the low score is the working assumption.

A credit committee does not need a perfect ESG rating. It needs to know whether the rating disagreement is about methodology, about missing data, or about a plant that is already in the newspaper.

Investors running India dedicated strategies should stop using a single vendor as a screen and then discovering the portfolio’s “ESG quality” flips when the consultant changes house. Document the primary vendor, the challenger vendor, and the issues on which you will override. Overrides need a memo. Overrides without a memo are just taste.

6. When disagreement is a greenwashing alarm

Not every spread is innocent methodology. Some patterns should send the board into a different mode.

Watch for a marketing-led high score paired with a low score from a vendor that weights controversies and physical assets. Watch for a sudden jump after a reporting-boundary change that dropped a dirty subsidiary. Watch for social scores that celebrate policies while attrition, contract-labour share, or lost-time injuries move the wrong way in the BRSR tables. Watch for climate narrative that talks about net zero 2070 while capex still lengthens the life of the high-emission line.

Greenwashing, in this narrower operational sense, is the gap between the story sold to the high vendor and the quantities in the firm’s own schedules. BRSR makes that gap easier to see than it was five years ago, provided someone on the board reads the schedules and not only the chairman’s letter.

Assurance is useful here as a brake, not as a trophy. Reasonable assurance on a defined Core indicator set reduces the room for a vendor to invent a worse number from thin news. It does not prevent a vendor from weighting that number differently. Boards that understand the difference stop buying assurance as if it were a rating upgrade coupon.

7. A research and teaching note for Indian classrooms

For doctoral students and faculty in finance, accounting and sustainability, rating divergence is not only a governance nuisance. It is an identification and measurement topic. If your dependent variable is “ESG score,” you have chosen a vendor’s theory of ESG. Results can flip when you change house. That is not a footnote. It is often the paper.

Better practice in Indian samples is to report at least two vendors or to unpack the score into pillars that map onto BRSR line items you can defend. Better still is to take the underlying quantity—emissions intensity, water, board independence, related-party intensity, gender mix—and use the rating only as a market-perception variable when that is the actual question. Papers that treat “ESG” as a single latent trait measured without error are increasingly hard to desk-accept in serious journals, and they are poor training for students who will sit on credit committees.

In the MBA and commerce classroom, a two-session exercise works. Session one: give groups the same firm’s BRSR tables and two vendor snapshots with the methodologies redacted. Ask them to predict where the scores will diverge and why. Session two: reveal the methodologies and grade the groups on diagnosis, not on whether they guessed the number. Students leave with a habit boards still lack.

8. Field checklist and closing brief

Use this list before the next ESG committee or credit review.

• Can we name every vendor that scores us, the peer set, and the date of the last refresh?
• Have we labelled each material gap as scope, weighting, vintage, or controversy?
• Do the weak pillars map to BRSR / BRSR Core line items we actually disclose and, where required, assure?
• Is there a financial hook—covenant, customer questionnaire, transition-finance screen, or KMP scorecard—for each pillar we claim to care about?
• Is the data-gap list owned, dated, and shorter than it was two quarters ago?
• Have we recorded which vendor flags we will not chase, and why?
• Would an independent director rather defend a quantity than a letter grade in a public meeting?
• If we are lenders: does the file treat spread as information, or do we still keep only the borrower’s favourite logo?

ESG rating divergence will not disappear as Indian disclosure improves. Better disclosure often increases disagreement in the short run, because vendors finally have different quantities to weight. That is progress. The firm that treats the spread as a puzzle to diagnose will make better capex, better supplier, and better credit decisions than the firm that treats the spread as an insult.

The next briefing in this series can go deeper on assurance scope, on climate risk in bank books, or on board-level ESG oversight charters. The rating question comes first because it is the number most rooms already argue about—and the number few rooms have learned to read.