I recently attended a presentation on economic modeling, climate change, and the social cost of carbon. Initially, this was presented as a process where we took our best scientific information, fed it into our best economic models, and ended up with our best projections about how much harm climate change would do, and thus what the ‘social cost’ of carbon really is. The point I raised was that this approach is in no way divorced from ethical assumptions. Indeed, they are deeply ingrained in the economic models and have profound effects on how they turn out.
Here are a few of the most important aspects of that:
1) The discount rate: Nicholas Stern took a lot of flak for setting this value so low. Basically, it pertains to how much we value the welfare of future generations. The lower you set it, the more the welfare of future generations will affect your calculations. In financial planning, discount rates are often in the neighbourhood of 8%. That means we would be indifferent to having $X today or $X + 8% in one year. The trouble is, with a value that high the welfare of distant future generations becomes almost completely unimportant in your calculations. If we knew that climate change would instantly kill everyone alive in 100 years, using an 8% discount rate would make this fact largely unimportant in terms of working out what the ‘social cost’ of one tonne of carbon is today.
Of course, there are problems with using a very low discount rate as well. If we care as much about all future generations as about our own, we are compelled to put all of our wealth towards investments for them. After all, current spending only benefits us, whereas investment could increase the welfare of a potentially infinite chain of future generations.
In any case, the discount rate selected has a massive effect on what social price for carbon you end up with. Stern worked it out as about $85 per tonne. William Nordhaus, another economist, came up with a figure of $7, largely because he used a higher discount rate. This one choice has the power to massively affect any economic analysis of climate change.
2) The marginal utility of income Take $100 per year from Bill Gates and he will never notice. Take it from everyone living in Sub-Saharran Africa, and you would probably kill millions. Despite this, most economic models assume that a dollar is a dollar is a dollar. If melting permafrost makes us abandon a northern community, at a cost of $20 million, but the cost to Canadian industry of avoiding the emissions that caused it would have been $21 million, the economically optimal outcome would be to allow the community to be destroyed.
To some extent, this can be built into economic models. We can create a mathematical function for how useful each extra dollar a person gets is. If we use that ‘utility’ measure in place of a dollars measure, the impact of different choices on the least well off becomes more important. Actually doing so on climate change would almost certainly hugely increase the social cost of carbon, since the welfare of those threatened by sea level rise in Bangladesh and drought in Sudan would be considered on more equal terms to wealthy Floridians with property threatened by hurricanes and oil company employees hoping to exploit new fields in the Arctic.
In addition to having economic importance, this has massive ethical importance. An approach based on potential Pareto optimality supports any move that improves overall welfare. It doesn’t matter if the people gaining are residents of suburban Toronto while those losing live in villages in Ghana. In everyday life, we recognize that we cannot go around harming people just because we gain more from doing so than they lose.
3) Valuing catastrophic risks If we manage to turn the world’s carbon sinks into net sources, we will have created self-sustaining climate change. If that occurs at an accelerating rate, we will be facing runaway climate change, which threatens to cause enormous physical changes and mass extinctions – possibly including humanity itself. Integrating such possibilities into economic models requires a number of ethical assumptions. Even a very small possibility of such an outcome can have a giant influence on certain kinds of models; likewise, choosing to ignore such outcomes has highly ethically relevant effects.
In short, we cannot combine scientific and economic models and produce a technocratic answer about how much climate change should be permitted. We need to acknowledge and consider the ethical implications built into and arising from those models, we need to choose what kind of world we want to hand over to future generations, we need to consider how important we think responsibility for the problem is when allocating costs, we need to consider the special circumstances of the very poor, and we need to consider how big a risk of catastrophe we should really tolerate.
I think an honest examination of those issues, alongside the best climatic science we have, creates a powerful and immediate ethical and economic argument for change. It is virtually certain that – if they could speak to us – people fifty or one hundred years in the future would be screaming at us to do dramatically more than we are doing now.