Here are some economists tying themselves in mental knots on how you would do cost-benefit analysis on complete annihilation of humanity.
…estimating these benefits means that we need to determine the value of a reduction in preventing a possible future catastrophic risk. This is a thorny task. Martin Weitzman, an economist at Harvard University, argues that the expected loss to society because of catastrophic climate change is so large that it cannot be reliably estimated. A cost-benefit analysis—economists’ standard tool for assessing policies—cannot be applied here as reducing an infinite loss is infinitely profitable. Other economists, including Kenneth Arrow of Stanford University and William Nordhaus of Yale University, have examined the technical limits of Mr Weitzman’s argument. As the interpretation of infinity in economic climate models is essentially a debate about how to deal with the threat of extinction, Mr Weitzman’s argument depends heavily on a judgement about the value of life.
Economists estimate this value based on people’s personal choices: we purchase bicycle helmets, pay more for a safer car, and receive compensation for risky occupations. The observed trade-offs between safety and money tell us about society’s willingness to pay for a reduction in mortality risk. Hundreds of studies indicate that people in developed countries are collectively willing to pay a few million dollars to avoid an additional statistical death. For example, America’s Environmental Protection Agency recommends using a value of around $8m per fatality avoided. Similar values are used to evaluate vaccination programmes and prevention of traffic accidents or airborne diseases…
The value of life as a concept is a natural candidate for a tentative estimation of the benefit of reducing extinction risk. Yet the approach seems somewhat awkward in this context. The extinction risk here is completely different from the individual risk we face in our everyday lives. Human extinction is a risk we all share—and it would be an unprecedented event that can happen only once.
I’m not sure we want to turn over the keys to civilization’s future to these guys, who insist that their science must be values-free. In other words, they try to discern people’s values through their actions and statements, but try to make no ethical judgments independent of those observations. I think there is room in this world for ethical principles of right and wrong that are not economic in nature, and more of us need to be actively thinking every day about what those might be. Even though all 6 billion of us would certainly not agree on the details, we could certainly come to a consensus on the broad outlines. Couple this with better mental tools for understanding the complex nested systems we are embedded in, and it could really guide our choices as a civilization in a better direction.