> Regulation works when experts really do have the answer. I think there are other significant factors to consider as well. In the US at least I think regulat

> Regulation works when experts really do have the answer. I think there are other significant factors to consider as well. In the US at least I think regulation, in our current day, has failed in many areas due to 1) the slowing down of decision makers to become informed and make decisions (see congress for stagnation in passing legal regulations) 2) industries or large industry players have waged effective influence campaigns in mainstream and scientific communities to push a friendly agenda or inject uncertainty which has nock on affects when it comes to funding (see pesticides like neonics and glyphosates, oil and gas companies with respect to global warming and plastic recycling, tobaco companies and smoking causing cancer, …), and 3) a certain degree of capture in regulatory agencies by industry (industry players later work for the regulatory agency or the reverse, as well as other tactics). 4) the politicization of regulation (not sure how much this can be helped though). It’s a mess. > We could switch from "allow by default" to "deny by default", but I'm not even sure that would help. There may be a middle ground here. As a chemic becomes more widely used that could trigger required enhanced testing by a regulatory agency as well as research grants looking into their safety. And as a chemical becomes ever more widely used rigorous long-term studies could be required. I’m not sure what the right long term answer is but I can’t help but think there is some fundamental mismatch of top down regulation by regulatory bodies in a market based system where companies are innovating and also creating these harms / negative externalities. In the former regulators are slow, reactionary, and often substantial harm must have already occurred for a regulation to be passed. The latter is a dynamic and fast moving system where companies are ruled by a fitness function which is often myopic and locally greedy. It seems introducing another market for regulation could offer a solution by using one dynamic system to regulate another. Carbon credits is an example of this. However two or more interacting systems will make things much harder to reason about and may have their own significant flaws if not designed well (for example trees are planted but then die earlier than expected or are later harvested after the carbon credits have been sold).