When a company issues a green bond, the proceeds are earmarked for green projects. But are those projects actually green in an economically meaningful sense — projects that would not have been funded without the label? Or are they the same projects the company would have funded anyway, just wearing a green label for investor relations reasons?
This is the additionality question. It sits beneath every green bond transaction, is usually politely ignored in the deal pitch, and is now the subject of a serious and growing empirical literature. The four papers below represent the most honest engagement with the question the academy has produced: two of them find the answer uncomfortable, one offers partial redemption, and one reframes the question entirely.
This is not an abstract debate. Regulators have noticed. Litigants are beginning to notice. The EU Green Bond Standard’s external review requirement is a direct response to it. Practitioners who engage with this literature now will not be surprised by it later.
1. Lam & Wurgler (2024) — the NBER answer: same projects
Lam and Wurgler examined what US firms actually do with green bond proceeds versus what they say. They matched green bond issuers against comparable non-issuers and tested whether the issuers’ capital expenditure patterns, R&D, and environmental investment changed meaningfully after issuance. They did not. The green label is applied to projects that would have been funded regardless. The proceeds are fungible; the earmarking is cosmetic.
The paper is careful in its framing. It does not say green bonds are fraudulent. It says that the use-of-proceeds mechanism — the market’s central additionality device — does not, on its own, produce additional green investment. The label shifts which projects get called green within a fixed capital budget; it does not expand the budget.
Why this matters for your work. If you write impact reports or post-issuance SPO reviews, this paper changes the question you should be asking issuers. The current practice is to verify that proceeds were allocated to the nominated projects. Lam and Wurgler show that this verification does not get at the underlying question: would those projects have happened anyway? You are not required to answer that question — current market standards do not ask you to — but you should know it is the question that will define the next phase of labelled bond market regulation.
2. Curtis, Weidemaier & Gulati (2023) — the legal structure: no teeth
Where Lam and Wurgler ask whether green bonds fund different projects, Curtis et al. ask whether the legal documentation requires them to. They conducted a systematic review of green bond indentures and found that use-of-proceeds covenants are almost universally aspirational rather than enforceable. Issuers can misallocate proceeds to non-green uses with limited legal consequence; bondholders have little recourse. The market’s integrity rests almost entirely on reputational discipline, which is real but fragile.
This is a law review paper written for lawyers but readable by non-lawyers. The core finding can be stated simply: the standard green bond covenant says roughly “we intend to allocate proceeds to green projects,” not “we are legally obligated to allocate proceeds to green projects, and here is your remedy if we don’t.” These are different sentences with very different investment implications.
Why this matters for your work. If you review green bond documentation for an investment committee, this paper gives you the right question to ask: is the use-of-proceeds covenant enforceable, or is it a statement of intent? The authors identify jurisdictions and structures where stronger covenants exist and note that the EU GBS’s external reviewer obligation is a meaningful step toward accountability even if it does not create a bondholder cause of action. For DCM lawyers, the paper is also a prompt: client exposure to ESG litigation risk is higher for aspirational covenants in jurisdictions where ESG misrepresentation claims are being actively tested.
3. Sangiorgi & Schopohl (2023) — why issuers issue: legitimacy, not greenium
This paper takes an unusual approach: instead of measuring what green bonds do, it asks corporate issuers why they did it. The survey responses are illuminating. Most issuers cite non-financial motivations — investor relations quality, signalling commitment to sustainability, broadening the investor base — rather than cost of capital advantages. The greenium is secondary or irrelevant for most issuers. They are not issuing to save basis points; they are issuing to be seen as aligned with a sustainability narrative, and to attract investors who value that alignment.
The paper does not condemn this. Legitimacy and investor access are real benefits of issuance, and they are often what clients actually want when they come to a sustainable finance desk. But the mismatch between the market’s official rationale (proceeds funding additional green investment) and the issuers’ actual motivation (institutional signalling) is documented here in unusually direct terms.
Why this matters for your work. This paper is the most useful for understanding why the additionality critique does not kill the market. Issuers are not issuing because they expect additionality; they are issuing because they want what green bond issuance provides socially and institutionally. This does not resolve the additionality question — it repositions it. The honest case for green bonds is not “we are funding things that would not otherwise be funded.” It is “we are signalling commitment and building a constituency of aligned investors.” For DCM bankers, this paper tells you what you are actually selling. For RI analysts, it tells you what your portfolio companies actually bought.
4. Liu, Huang & Mbanyele (2024) — the behavioral nudge: Chinese evidence
The Chinese evidence offers a more optimistic reading of the additionality question. Liu et al. examine whether Chinese firms that issue green bonds subsequently engage in more environmental responsibility activities — not just use-of-proceeds allocation, but broader behavioral change in how the firm manages its environmental footprint. They find that green bond issuance does nudge environmental engagement, with effects that persist for several years after issuance and are stronger for firms under higher institutional monitoring pressure.
The paper is careful to situate its findings in China’s institutional context, where regulatory enforcement accompanies the green bond label and issuers face more direct pressure to demonstrate environmental performance. The nudge effect does not operate through the financing mechanism; it operates through the reputational and regulatory accountability that labelling creates.
Why this matters for your work. This paper shows that the additionality problem is not irreducible. It is a function of institutional design. In a market where green bond labels carry enforceable obligations and external accountability — as the EU GBS is intended to create — the behavioral effects that Lam and Wurgler find absent in current US practice could emerge. The paper does not prove additionality in Western markets; it shows what conditions would need to exist for additionality to emerge. That is a different and more useful claim for a practitioner thinking about market evolution than for one writing an impact report this quarter.
How to read these four together
Read in this order:
1. Lam & Wurgler (2024) — to calibrate the problem: use-of-proceeds earmarking does not produce additional green investment in current US practice. 2. Curtis et al. (2023) — to understand why: the documentation does not require additionality, and weak covenants give issuers no incentive to pursue it. 3. Sangiorgi & Schopohl (2023) — to understand why issuers issue anyway: legitimacy and investor access are the actual benefits, and they are sufficient to sustain the market even without additionality. 4. Liu et al. (2024) — to understand what conditions would produce the additionality the market promises: institutional pressure, enforceable accountability, active monitoring.
What these four papers do not say is that the labelled bond market is fraudulent or worthless. What they say is that the market currently lacks the mechanisms to ensure additionality, and that most participants — issuers, bankers, investors — have implicitly accepted this without saying it out loud. The EU Green Bond Standard’s mandatory external review is a direct response to exactly this problem. Whether it will be enough is a question the next decade will answer.
The additionality literature should also be read alongside Part 2 of this series (“Does Green Mean Green?”), which examines the environmental impact of use-of-proceeds separately from the question of whether the projects would have happened anyway. Additionality and impact are related but distinct claims.
What you can do with this on Monday
- For SPO analysts: before your next post-issuance review, read Lam & Wurgler and consider asking the issuer what would not have been funded without the green bond. You are not required to verify the answer — current standards don’t ask for it — but the question is increasingly on regulators’ minds, and being ahead of it is more defensible than being surprised by it. - For RI analysts: read Curtis et al. as a covenant checklist. For green bonds in your portfolio, establish whether the use-of-proceeds obligation has teeth. If not, engagement on covenant structure is a legitimate stewardship ask, and increasingly one that EU investors can support with regulatory language. - For DCM bankers: read Sangiorgi & Schopohl to understand what your clients actually value. You are selling legitimacy, investor access, and ESG narrative alignment — and this is a real and sustainable value proposition — but stop selling additionality you cannot demonstrate. The honest pitch survives academic scrutiny; the additionality pitch does not. - For all four roles: each paper has a permanent XFID. The findings will not move; the URL will not break. You can cite them in IC papers, engagement letters, and policy responses knowing the references are durable.
Part 8 of 8. Previous: Nature and Biodiversity Bonds. Browse all RESEARCH papers at xframework.id/registry?type=RESEARCH.