Effective Altruism and Everyday Decisions
Ask for your drink without a straw.
Unplug your microwave when not in use.
Bring a water bottle to events.
Stop using air conditioning.
Choose products that minimize packaging.
I’ve recently heard people advocate for all of these, generally in the form of “here are small things you can be doing to help the planet.” In the EA Facebook group someone asked why we haven’t tried to make estimates so we can prioritize among these. Is it more important to reuse containers, or to buy locally made soap?
I think the main reason we haven’t put a lot of work into quantifying the impacts of these everyday choices is that they’re minor compared to questions like “what should I work on?”, “if I’m donating where should the money go?”, “how can we figure out the impact of our choices at all reliably?” etc. Quantification, even at a very rough level, is really hard and so we should focus on the most important questions first.
A second reason, however, is that these sorts of activities are often shockingly poor tradeoffs. Perhaps you give up AC to save electricity, but then you get less done during the day, sleep poorly at night, and only save $3/day in electricity and ~$0.75/day in CO2  Or you buy zero-waste laundry paste which you dilute at home, putting money and effort into avoiding a very small amount of plastic packaging. Or you take cold showers and enjoy them dramatically less while slightly reducing your use of heating fuel. Advocacy often explicitly or implicitly treats actions as free, while a full evaluation needs to also consider the cost to yourself.
I’m not saying our personal choices don’t matter and that we should give up, but a small number of our choices matter far more than others, and we should put our efforts there.
 This is figuring 600W average usage for a window unit, which is 14.4kWh/day. Our marginal cost for power is $0.21/kWh, on the high side nationally, so $3/day. Figuring 1T CO2 per kWh, this is ~0.007T CO2/day. Using the same 95th percentile EPA social cost estimate I used in this post, $105/T, that’s $0.76/day. These are very rough numbers, but they’re enough to see that the costs are low.