Covid offsets and carbon offsets
These are my own thoughts entirely and do not necessarily reflect those of my employer.
I am not equating climate change with Covid-19.
Covid-19 has garnered a lot of support from effective altruists, and I notice a strong social resolve to encourage social distancing, and create norms such as wearing a mask in your profile picture.
Over time it’s become clearer that the mortality gradient for Covid-19 is steepest for people who are older, and who have pre-existing health conditions, but this is only an average. The act of mask-wearing, social distancing etc., becomes less about protecting myself and more about an act of goodwill to others. We undertake this act to reduce the probabilities that we might unknowingly spread the dreadful disease to someone who might become very ill or even die from the disease.
Covid-19 spreads throughout populations, through different means including surfaces and through the air. It produces a nebulous harm, an infection in population. At an individual level, a person catches the infection, and may become very ill. So far (July 2020) there have been around 600,000 deaths globally from Covid-19.
It’s an interesting comparison with climate change, which is also a global problem but a very different one. Our actions overall as humanity lead to the production of greenhouse gases, which interact with the earth system, trap more of the sun’s energy on earth, and lead to heatwaves, flooding, and increases in the frequency and severity of storms. This leads to human impacts such as heat stress, famines, disease, and death. Our actions as an individual contribute to a nebulous harm, which then affects other individuals. The UN estimates that climate change causes 150,000 deaths per year, increasing to 250,000 per year from 2030 onwards.
We produce these greenhouse gases when we travel on an airplane, drive a car, power our lives, buy products from around the world, and even likely partially powering the computer I use.
John Broome estimates that a lifetime’s carbon emissions effectively take away half a year of life from someone else in the world, though it may be higher, and possibly higher still if we take into account the small probability of tragedy. And similarly, my decision to hold a party or interact with my friends during a global pandemic creates some fractional, statistical harm spread across society. These seem to me quite similar.
One difference could be that climate change is more likely to harm people outside of my circle of acquaintances—people living along coastlines and in Sub-Saharan Africa, but covid might affect my own older relatives.
I am genuinely interested in what you think about the following questions:
If you are in good health and expect not to be affected by covid-19, then do you think it is right to be lax on social distancing, and ‘offset’ your choices, e.g. through a donation to Johns Hopkins?
Do you think people in good health are morally obliged to carry out social distancing? If so, why?
If you frequently travel, eat meat, buy lots of items etc., then do you think we have an obligation to offset our carbon emissions?
Or do you think we have an obligation to reduce our personal carbon emissions, rather than offsetting them?
What do you think are the differences between covid-19 and climate change that might mean you give different answers to each?
Do you think that we have obligation not to cause harm, or does this all wash out if we take a straightforward consequentialist approach?
Recommended reading
John Broome—How much harm does each of us do?
John Broome—The most important thing about climate change
Friederike Otto and Luke Harrington—Why Africa’s heatwaves are a forgotten impact of climate change
Some thoughts:
There may be more of a personal incentive to avoid COVID-19 than can be assumed from looking at mortality gradients with respect to age and health. SARS-1 has documented long-term effects that range from permanent lung damage to chronic fatigue; there’s compelling reasons not to get involved with our current mass SARS-2 experiment. In general, recovering from a virus is not necessarily smooth, see viral meningitis.
I think it’s interesting to compare offsetting with climate change and this pandemic. I’m no expert on modeling the effects of individual effort to mitigate climate change, but I assume that a pandemic has an inherently exponential nature that makes high compliance to social-distancing guidelines imperative. I would say people in good health are absolutely morally obligated to social distance.
With carbon emissions, I feel a little more comfortable making direct comparisons between offsetting vs decreased personal emissions. Obviously it depends on the nature of the offset, but I think that you can do a straightforward cost-benefit analysis based on how much emissions you prevented/caused.
The pandemic is different. What does a donation to Johns Hopkins pay for? Perhaps better information, better models, better science. That’s not really the same currency as decreasing R0 through social distancing. Better information would be quite useless at a high enough R0.
And say we do make a consequentialist analysis of the situation that donation can make up for social distancing. In that case, I’d be worried about the relationship between variance and chaos—a donation might not always lead to mitigating R0 consistently compared to social distancing. Tweaking parameters in a complicated exponential epidemiological model is a sure way to get unexpected results and invite tail risk.
Relevant EA forum post: “Ethical offsetting is antithetical to EA”