Quality Control: How to Spot a Weak Carbon Credit

Transcript

SPEAKER_1: Alright, last time we established that in voluntary markets, value flows from trust in project quality. Today I want to stress-test that trust. SPEAKER_2: Good framing. The key idea here is that not all credits are equal. A weak credit fails to reliably represent a real, additional, measurable, permanent, and independently verified emission reduction. SPEAKER_1: So there are four core tests. Walk me through them quickly. SPEAKER_2: Sure. Additionality—would the action have happened without carbon finance? Permanence—will the stored carbon stay stored? Leakage—did the reduction just push emissions elsewhere? And double counting—is the same ton being claimed by more than one party? SPEAKER_1: Additionality seems like the easiest one to game. How does that actually fail in practice? SPEAKER_2: Think of a wind farm already profitable from electricity prices alone. If it sells credits claiming it only exists because of carbon finance, that's a failed additionality test. The climate benefit was happening anyway—the credit is claiming credit for something it didn't cause. SPEAKER_1: And that connects to baseline setting, right? If you inflate the baseline, the reduction looks bigger than it is. SPEAKER_2: Exactly. Overstated baselines make credits appear far more valuable than the real climate benefit warrants. Weak methodologies can systematically overcredit entire project categories—not just individual projects. SPEAKER_1: Now, permanence—that one seems especially fragile for land-based projects. Suppose a forest burns down after credits are issued. SPEAKER_2: That's the core risk. Carbon stored in forests can be reversed by fire, disease, or land-use change. Industrial projects—like capturing methane from a landfill—tend to be more durable because the reduction is a one-time physical event. Forestry storage is inherently more fragile. SPEAKER_1: So leakage is essentially a boundary problem—the reduction happens inside a fence, but the emissions jump the fence. SPEAKER_2: That's a good way to put it. For example, a protected forest stops logging locally, but timber demand doesn't disappear—logging shifts to an unprotected area nearby. Net global emissions barely move, but the credit claims a full reduction. SPEAKER_1: And double counting—how does that actually happen mechanically? SPEAKER_2: The same emission reduction gets claimed by both the project developer and the host country toward its national climate targets. Now two parties are reporting the same ton. The accounting simply doesn't hold—it's a fundamental integrity failure. SPEAKER_1: So what's the practical due diligence process for screening these weaknesses before buying? SPEAKER_2: Registry data is the starting point—it shows project type, issuance history, retirement records, and ownership chain. Then check the methodology and the credit's vintage, because older credits may reflect earlier, looser rules. Confirm independent third-party verification actually happened. SPEAKER_1: There's also the co-benefits angle. A project might have a compelling story—biodiversity, community development—but that narrative doesn't automatically mean the climate accounting is sound. SPEAKER_2: Critical point. Co-benefits can be genuinely valuable, but they don't substitute for climate integrity. A project can have an attractive social narrative while still producing low-integrity credits. Those two dimensions need separate evaluation—not bundled together. SPEAKER_1: And the financial risk is real, not just reputational. If a company builds a climate claim on weak credits and that claim gets challenged... SPEAKER_2: The exposure is significant. Low-quality credits create reputational, financial, and regulatory risk when underlying claims don't hold up. Remember, market participants increasingly distinguish credits by integrity criteria—not price alone. For Wynton and everyone following along, the takeaway is this: the four tests—additionality, permanence, leakage, and double counting—are the actual quality filter. Price is a signal, but it's not the signal.