You may try to argue that natural capital accounting is just about providing governments with better information about the socio-economic utility of nature. But the truth is that, ultimately, this is not what this is about. While the World Bank’s WAVES programme is about integrating natural capital into national accounts, the Natural Capital Declaration envisages the introduction of new ‘risk management tools’, which, seeing as the proposal is coming from financial sector companies, is likely to mean the further development of new financial hedging mechanisms such as ‘biodiversity bonds’ that would take control of ‘natural capital’ away from governments and the people who depend on it and put it at the mercy of highly unstable financial markets. And as for the idea that reducing conservation decisions to a financial cost-benefit analysis will lead to better government decision-making, it is wholly possible that the opposite may occur. For example, we already know that coral reefs can protect coastlines from storm surges – but if we express this in a dollar figure, the implication is that it is acceptable to trash it if the profit opportunities are sufficiently high. And this kind of simplistic utilitarianism ignores the fact that, to further develop your example of the coral reef, the benefits of destroying the reef are likely to accrue to investors rather than the poor, who are often the most dependent on free natural resources. So a government or private company may decide to let the reef die because the overall monetary return of preserving it is less, ignoring the fact that the people impacted by the decision will be the poorest. We’re seeing this already with the REDD+ programme, which is a more advanced example of where ascribing a financial value to nature can lead us. The example of the 20,000 people who were evicted from their homes by the New Forests Company in Uganda as a result of being on the wrong side of the sort of cost-benefit analysis you describe shows what can go wrong when you take this approach.