Ch. 3.3. On sustainability metrics: how to learn from previous estimates of environmental benefits to overcome predictable blindspots

An analogy: we now know that the costs of environmental regulation are consistently over-estimated in advance. There is a political reason: the data tend to come from industry, who have an incentive to emphasise costs. And there's an economic reason: costs of pollution abatement are calculated using current technology, but regulation increases the returns to technological improvement, so induces new technologies that cut the cost. The systematic error is only visible in retrospect, looking across studies, but once identified can be taken into account in future estimates.

There might be reasons for thinking that estimates of environmental/ecological benefits are also systamtically biased. The course has explored some of these, including tipping-points and emergent properties - both of which, I would guess, might lead to environmental damage being worse rather than better than expected. In other words, in weighing the cost of pollution or environmental degradation, is it far more likely that there are bad effects than good effects that we haven't thought of?

Could a review of ex-ante and ex-post estimates, and some deeper thinking about why systematic error might occur, shed a similar light to the cost-of-regulation research?