Last week, I joined “Revolutionist’s Night” — a cabal aiming to radically improve animal welfare, with some flavor of 1800s slavery abolitionism or the 1700s American revolution. Our topic for the evening was inspired by Aaron Boddy’s post “The current market price for animal welfare is zero”: can we create a market where anyone can directly buy credits for animal welfare?

I’m very sympathetic to this idea, and similar ideas like Paul Christiano’s for offsetting factory-farmed eggs. (A year ago, I spent some time brainstorming how one might actually implement the egg-offset-thing.)

Some reasons I’d be excited:

• Markets are great at optimizing for particular objectives, if you shape them correctly. The world abounds in cheap high-quality consumer goods, thanks to markets.
• There are some precedents for using markets for public goods, from carbon credit markets and renewable offsets (eg electric vehicle or solar credits)
• Welfare credits let you buy directly what you want, vs labels like “organic” which bundle that together with food quality
• Typical animal welfare interventions (going vegan, protests, corporate commitments) rely more on persuasion or coercion; markets rely on voluntary participation
• Markets scale up well: they facilitate trillions of dollars of activity, a level of coordination only (perhaps) matched by major governments and religions
• Corporations are better at interfacing with markets and price signals; an animal welfare market would let existing corporations participate, and might induce new corporations
• On a personal level, it sure would be nice to just pay some $ every month and then be able to eat whatever I like, rather than thinking a lot about. (FarmKind offers a service approximating this).

We threw around a bunch of ideas during Revolutionist’s Night, though we didn’t end up with a concrete proposal or specific next steps. In lieu of that, here are some lingering questions I have around the whole concept:

• What is the total addressable market (TAM) of animal welfare?
  • We broke this down into “altruism” (doing good for animals) and “consumer” (offsetting harms caused by yourself)
  • From an “altruism” perspective, we estimated ~$300M each year being spent by funders on farmed and wild animal welfare
  • From a “consumer” perspective, we ballparked:
    • 300M US citizens * 5% vegetarianism rate * 1k meals/y * $2/meal welfare offset ⇒ $30B/y market
    • For an early adopter/MVP market, imagine: 10k EAs * 1k meals/y * $2/meal offset ⇒ $20M initial market.
  • I think there might be an additional “corporate” perspective; something like: “how much are companies already spending beyond the bare minimum, to make animals happy?”
    • (I’m not sure how to estimate this)
• What does the track record of carbon credits imply about animal welfare markets?
  • Positives:
    • Stripe Frontier is an awesome example of using advance market commitments to kickstart a robust market; see this writeup by the founder Nan Ransohoff
    • Many companies have adopted “net zero” commitments, buying carbon credits to offset their electricity & operational usage.
    • Lincoln recent installed solar panels for his group house; ~1/3rd of the projected revenue for these panels comes from govt tax credits which incentivize renewables (vs 2/3rds from savings on electricity)
  • Negatives:
    • Very few people I know actually offset their carbon usage. I haven’t done this myself; I vaguely feel like I should but haven’t yet
      • My friend Barak Gila wrote about spending $10k offsetting plane & car miles, in Cars and Carbon
    • Carbon credits famously suffer from verification and goodhearting problems.
      • Unclear how much of these problems are theoretical/nerdsnipe-y, vs fundamentally killing the entire concept
        • Xander Balwit mentions that many people who used to be excited by carbon credits have now moved on to other greener pastures
• Why isn’t there already “human welfare market”? Should we aim to set up the human welfare market first?
  • In one sense, Givewell is already a place to go to spend money to buy human welfare units (often QALYs). By analogy, Animal Charity Evaluators is already an “animal welfare market” — but this seems unsatisfying.
    • Verifying impact is hard! There are some criticisms of ACE as an evaluator specifically; and some broader concerns about epistemics in animal welfare-land.
    • Lincoln: “When I buy carbon credits, I believe it’s going into a marketplace where people are competing to deliver the product”. And this is less true for charity evaluators.
      • There’s some competition for impact among the charities being evaluated, but we’re mostly placing trust into fairly-subjective calls made by the evaluators.
• What social technology might get people on board with buying these credits?
  • Awareness: In some sense, the low absolute TAM of animal welfare is downstream of the fact that still, not that many people care about animal welfare.
    • What makes certain causes much more popular than others? Eg why is climate change a much bigger issue than animal welfare?
      • Constance suggested that people have more selfish reasons to care about climate (they’re concerned about themselves and their kids dying); and climate impacts are more visible (people see weird weather events & news reports, but not much of factory farms).
      • I suspect that’s part of the story, but there’s also some amount of contingency — eg Al Gore choosing to do “An Inconvenient Truth”
  • Concrete pledges: The Giving What We Can pledge is a cool piece of social technology/coordination that successfully convinces many people to give a certain % of their income to charity. The Giving Pledge and Founder’s Pledge are similar examples.
    • Notably, these are all examples of pledging a certain % or $ amount (an input), rather than pledging to buy a certain amount of welfare (an output)
  • Incepting with status: I like the story of how diamond monopolist De Beers manufactured the tradition of proposing with a diamond ring, by getting famous people on board and showcasing them in movies. Can we do this for animal welfare units?
    • Eg manufacture eg $500 tungsten jewelry or $50 3D-printed trinkets; attach a $10k “animal welfare” impact certificate; and convince all our EA friends to gift these during major life milestones (marriage, childbirth, etc?)
• Could market-based credits work for other public goods we care about?
  • It’s easy to brainstorm ways to commit dollars:
    • AI xrisk — safety-conscious lab employees could pledge some fraction of their income to safety work
    • GHD/human welfare — people in rich countries could pledge some fraction of the differential between their personal income and the poverty line, to GiveDirectly or other welfare causes
    • Abundance — people who own property could commit some fraction of land appreciation prices to a YIMBY fund; “voluntary Georgism”
  • But dollars are an input, not an output. It’s harder to operationalize how to buy the thing we actually care about (units of welfare, basis point reduction in xrisk)
    • Carbon credits are reasonably well-operationalized (tons of CO2 removed); Covid vaccines for Operation Warp Speed are also well operationalized
    • Markets work best with fungible standardized units (think oil or corn commodities)
  • Also: does this kind of moral offsetting lead to very high demandingness, to have to think through the downstream moral externalities of every action you take?
    • Perhaps it’d be ideal to aggregate these decisions higher up the stack than “individual consumers” — perhaps corporations, perhaps government
      • But aggregation also glosses over differences in values — eg some people might consider cows or pigs worthy of consideration, but not shrimp or bees
• How can Mox (our coworking space, where Revolutionist Night was held) help push for animal welfare markets?
  • I’ve been excited for Mox to be an early adopter of prosocial technologies — eg we’re the first coworking space to deploy far-UV (I think); we’re currently demo-ing lamps from three different providers
  • For animal welfare, we currently exert some influence when we cater food, opting to serve more vegetarian meals than our members/event clients might otherwise choose. We also don’t stock meat snacks.
  • Some market-based proposals:
    • We could tax all meat eaten on premises, eg charging people $5 for any meat meal
      • This is the most targeted and most internalizes externalities, but it seems high friction (so hard to actually implement) and seems like it’d be unpopular (has high social cost)
    • We could subsidize vegetarians, eg providing a 20% off discount code for memberships for pledged vegetarians, or a $40/mo rebate for vegetarians
    • We could commit some % of our general catering budget to buying animal welfare offsets, or estimate the number of meat meals each month. (To use a term coined by Lincoln, Mox could aim to be “welfare-neutral”)
  • Similar proposals might also work well in other contexts where someone is aggregating lots of demand for food, eg at Manifest or EAGs

Thanks to Lincoln Quirk, Constance Li, and many others for discussion

Further Reading

• Aaron Boddy on shrimp stunning credits
• Paul Christiano on humane egg offsets
• Scott Alexander on the ethics of moral offsets