---
title: "Does doing good pay? Four gates for ESG claims"
description: "Larry Fink's 2019 letter says purpose and profit are linked. Four questions on measurement, causation, aggregates and incentives, read with Vogel and Friede."
url: "https://staging.khoi.co.uk/en/blog/lam-tot-co-lai-bon-cau-hoi"
lang: en-GB
author: "Lê Quốc Khôi"
date: 2026-11-03
updated: 2026-11-03
topics: ["Ethics & corporate governance"]
tags: ["ESG", "Corporate social responsibility", "Correlation and causation", "Reverse causality", "Shareholder primacy"]
---

# Does doing good pay? Four questions to ask before you believe it

## Key takeaways

- Larry Fink's 2019 letter says profits and purpose are "inextricably linked" and that companies ignoring stakeholders will stumble (Fink, 2019); I use it as an example of the "doing good pays" claim.
- Four gates: can it be measured (M), has reverse causality been ruled out (C), does the evidence go beyond aggregate correlation (G), and is management's incentive limited to the level that still pays (I).
- Friede and colleagues (2015) find about 90 per cent of studies non-negative, but as correlation, and weaker for portfolios (Friede et al., 2015); Vogel (2005) argues the business case holds for some firms in some circumstances (Vogel, 2005, p. 3).
- I do not conclude whether ESG investment is worth it; the post shows where Fink's claim stops.

In 2019 Larry Fink, head of BlackRock, wrote to CEOs that purpose is not the sole pursuit of profit but the force that drives it, and that profits and purpose are "inextricably linked" (Fink, 2019). It sounds reasonable and is often repeated. When an ESG proposal reaches your desk with the same justification, is there a way to test it before you spend the money?

## What exactly does Fink say?

Three statements in the 2019 letter matter here. First, purpose is not the sole pursuit of profit but the animating force for achieving it. Second, companies that fulfil their purpose and responsibilities to stakeholders reap rewards over the long term, while those that ignore them stumble and fail. Third, as wealth shifts and investing preferences change, ESG issues will be increasingly material to corporate valuations (Fink, 2019). The letter (a web page, so no page numbers) also lists BlackRock's 2019 engagement priorities: governance (including board diversity), strategy and capital allocation, long-term compensation, environmental risks and opportunities, and human capital (Fink, 2019). See the case file on [Larry Fink's 2019 letter](https://staging.khoi.co.uk/en/case/blackrock-2019-ceo-letter).

The letter cites no study of financial results in support of the "stumble and fail" clause. That matters because, in my reading, the statement sounds like a law, while it is written as a leader's view.

## Gate M: what does the claim measure?

The first question: how are "doing good" and "profit" measured, and over what period? Fink gives no indicator; the letter speaks of purpose and stakeholders, so it cannot be checked against data. That is not his fault alone: measurement is a problem for the whole field. Vogel notes there is no consensus on what counts as virtuous corporate behaviour, and that CSR's many dimensions make firms hard to rate (a company can be good on the environment and poor on labour) (Vogel, 2005, pp. 4-5).

Norman and MacDonald criticise a related form of the problem in the "triple bottom line" idea (economic, social, environmental). They argue that, as far as they know, nobody has proposed a way to use social data to compute a net social "profit or loss", and that the values involved have no common unit (Norman and MacDonald, 2004). The post relies on the March 2003 "In Press" draft, not the journal version, so it gives no page numbers. If it cannot be measured, the claim cannot be tested; it has not been refuted, only left with nothing to test.

## Gate C: did profit come first, or ESG?

Suppose there is a positive correlation between ESG scores and profit. There are two readings: ESG creates profit, or profitable firms have money to spend on ESG. The second is [reverse causality](https://staging.khoi.co.uk/en/glossary#reverse-causality), and telling the readings apart needs a different study design from counting how many studies come out positive. See also [correlation and causation](https://staging.khoi.co.uk/en/glossary#correlation-vs-causation).

Friede, Busch and Bassen call their result a correlation: they pooled 60 review studies (35 vote-count, 25 meta-analyses) covering over 2,200 individual studies to the end of 2014 and found about 90 per cent non-negative, with an average correlation of about 0.15 (Friede et al., 2015). In the paper they use both "positively correlated" and "investing in ESG pays financially" (Friede et al., 2015). In my reading the second phrase goes beyond what the data allow, since the paper does not analyse causal direction. That is my remark, not the authors' conclusion.

## Gate G: does the evidence go beyond aggregate correlation?

An average over thousands of studies does not say whether it holds for any particular company. Friede and colleagues show results differ by method: in the vote-count reviews, only 15.5 per cent of portfolio studies found a positive relation, against 56.7 per cent of non-portfolio studies; in the meta-analyses the differences are smaller but still significant (Friede et al., 2015). They explain the gap by noise, mixed ESG screens and fees of about 2.5 per cent a year (Friede et al., 2015). They also do not analyse cost of capital, only corporate financial performance (Friede et al., 2015).

Vogel stands at the other end. He concludes that the business case for CSR "has little empirical basis" (Vogel, 2005, p. 14) and that "CSR is best understood as a niche rather than a generic strategy: it makes business sense for some firms in some areas under some circumstances" (Vogel, 2005, p. 3). He also writes that more responsible firms are not necessarily more profitable, but neither are they less so (Vogel, 2005, p. 13). Every Vogel passage here comes from the book's opening chapter; the empirical analysis in chapter 2 has not been checked.

The two sources are ten years apart and use different methods, so I do not treat either as having "won". What they share: neither supports a law that doing good certainly pays for every company. See also [the market for virtue](https://staging.khoi.co.uk/en/glossary#market-for-virtue).

## Gate I: how much cost is still reasonable?

The last gate is the one most often forgotten. Vogel writes that in most cases CSR only makes business sense if the costs of more virtuous behaviour remain modest, which limits the resources a company can devote to it (Vogel, 2005, p. 4). He also holds that CSR is sustainable only if virtue pays off (Vogel, 2005, pp. 2-3). For a small company the gate means: before starting, set a spending ceiling and a stopping condition, instead of measuring success by "we are still doing it".

Fink's statement states no such limit: "stumble and fail" is stated for companies that ignore stakeholders, with no condition on cost (Fink, 2019). Vogel says the case is strongest in two groups: firms that have made CSR part of their strategy to attract customers, employees and investors, and highly visible global firms targeted by activists; in his view most firms fall into neither (Vogel, 2005, pp. 14-15).

> **Figure:** [Evidence check for 'doing good pays' claims](https://staging.khoi.co.uk/en/framework/esg-claim-evidence-check) Four gates for the claim "doing good pays" (redrawn from the site's library framework; origin credited to Vogel, 2005, and Norman and MacDonald, 2004).

## How does Fink's claim fare at the four gates?

The table below is my reading, not any author's verdict, and it does not say Fink's letter is wrong.

| Gate | Question | Result for Fink's statement |
|:--|:--|:--|
| M | Is there a clear method of measurement? | The letter names no indicator; untestable as it stands |
| C | Has reverse causality been ruled out? | The letter does not discuss it; the reviews cited above report correlation |
| G | Does the evidence go beyond aggregate correlation? | A non-negative tendency in aggregate, weak for portfolios; Vogel disagrees |
| I | Is the incentive limited to the level that still pays? | The letter states no cost limit |

A conditional reading: as a law-like statement for all companies, the letter stops at gate M. For a narrower claim, say about governance or capital allocation (the priorities BlackRock listed), you could look for specific evidence at each gate. The post on [when profit is legitimate](https://staging.khoi.co.uk/en/blog/khi-nao-loi-nhuan-chinh-dang) asks a different question: when is pursuing profit legitimate at all.

## How does a small business owner use the four gates?

A hypothetical example, every detail mine. A garment workshop that exports is advised to spend VND 600m on wastewater treatment and a social report, because "European customers will pay more". The four gates turn that sentence into four tasks: write down specific indicators (price, orders, contract renewals) and a time frame; check whether the workshop was already selling at higher prices before it invested (reverse causality); look for evidence from workshops of similar size and customers rather than industry averages; and set a VND 600m ceiling and a review date.

This suits you when the spend is large against annual profit or hard to reverse. It does not suit you if you act because the law requires it or because of your own values: then the question is not "does it pay?" but "how much is enough?". The posts on [sustainability marketing is not CSR](https://staging.khoi.co.uk/en/blog/marketing-ben-vung-khong-phai-csr) and [four kinds of responsible investing](https://staging.khoi.co.uk/en/blog/bon-kieu-dau-tu-trach-nhiem-dws) give two other angles.

## Going deeper: how were the sources used?

Fink (2019) is a web page on BlackRock's site, so it has no page numbers. Friede, Busch and Bassen (2015) is used in its full open-access text. Vogel (2005) is used only for the opening chapter (pages 1 to 15), taken from the sample on Brookings' site; the lines on "little empirical basis" are therefore Vogel's own summary on page 14, not an analysis that has been checked. Norman and MacDonald (2004) is used in the March 2003 draft on the authors' site and has not been compared with the printed version in Business Ethics Quarterly.

The four-gate table applied to Fink's letter is this post's own reading. The garment workshop is hypothetical. The post does not examine evidence about BlackRock or about ESG funds, and does not say whether ESG investment is a good idea.

## Frequently asked questions

### Does research prove that doing good pays?

Not in a causal sense. Friede, Busch and Bassen (2015) pooled over 2,200 individual studies and found about 90 per cent non-negative, with an average correlation of about 0.15, but they describe it as correlation and the portfolio studies are much weaker (Friede et al., 2015). Vogel (2005) concludes that the business case for CSR 'has little empirical basis' (Vogel, 2005, p. 14).

### What is reverse causality and why does it matter here?

It is the opposite direction of the hypothesis: profitable companies have money to spend on ESG, rather than ESG producing profit. In the sources used in this post, Friede and colleagues report correlations and do not discuss causal direction (Friede et al., 2015). So a positive correlation is not enough to say ESG spending produces profit; see [reverse causality](https://staging.khoi.co.uk/en/glossary#reverse-causality).

### So should a company invest in ESG?

This post does not answer that. It says the claim 'doing good pays' needs to pass four gates before it is used as the main reason. Vogel argues CSR makes business sense for some firms, in some areas, under some circumstances (Vogel, 2005, p. 3), and in most cases only while costs stay modest (Vogel, 2005, p. 4).

### How can a small business use the four gates?

Write the proposal down and answer each gate in one sentence: what is measured and over what period, why profit did not come first, whether the evidence comes from firms like yours, and what spending ceiling would make you stop. Any gate left blank is where to ask more.

## References

- Fink, L. (2019) *Larry Fink's 2019 letter to CEOs: purpose & profit*. Available at: https://www.blackrock.com/corporate/investor-relations/2019-larry-fink-ceo-letter.
- Friede, G., Busch, T. and Bassen, A. (2015) ‘ESG and financial performance: aggregated evidence from more than 2000 empirical studies’, *Journal of Sustainable Finance & Investment*, 5(4), pp. 210–233. https://doi.org/10.1080/20430795.2015.1118917
- Norman, W. and MacDonald, C. (2004) ‘Getting to the bottom of "triple bottom line"’, *Business Ethics Quarterly*, 14(2), pp. 243–262. https://doi.org/10.5840/beq200414211
- Vogel, D. (2005) *The market for virtue: the potential and limits of corporate social responsibility*. Washington, DC: Brookings Institution Press. Available at: https://www.brookings.edu/books/the-market-for-virtue/.

---

Canonical HTML version: https://staging.khoi.co.uk/en/blog/lam-tot-co-lai-bon-cau-hoi