Don’t sweat diet?

Summary: Various people (Hurford, Tomasik) have tried to estimate the typical animal welfare cost of a carnivorous diet. These costs have encouraged EAs to become vegetarians or vegans, and EAs particularly focused on animal welfare advocate that others (including EAs) should reduce their consumption of animal products.

Closely following an estimate by Kaufman, I consider the face value of abstaining from dairy or abstaining from all animal products in terms of donations to ACE-recommended animal welfare charities: <1 cent per year given to The Humane League would offset typical dairy consumption, and <$1 year would offset a typical American’s consumption of all animal products. Thus the emphasis on dietary change intra-EA seems misplaced: it is extraordinarily low impact compared to other means to helping animals. 


Industrial agriculture produces vast amounts of animal suffering: not only are billions killed each year, but their lives are often horrible – so much so that informed observers consider their lives to be not worth living. Given this suffering is generated to satiate human’s relatively trivial desire to eat meat, eggs, and dairy, reducing consumption of these products can relieve vast amounts of suffering. Many effective altruists think this (and animal welfare) is the most important thing to work on right now, and encourage others to become vegetarian and vegan. Many other effective altruists, even if they aren’t primarily focused on animal welfare, are vegetarian and vegan due to this reasoning..

How much is being vegan worth?

Effective altruists like quantification and trying to weigh the relative importance of different courses of action. This gives surprising answers in case of veganism.

Working out exactly how much suffering is hard. Happily, much of this work has already done by Brian Tomasik on the ‘suffering per calorie’ estimates of animal food products, and subsequently by Peter Hurford on the suffering typically accrued by a typical American’s diet. Jeff Kaufmann recently used these figures to compare the ‘value’ of abstaining from dairy in terms of a donation to AMF. A precis:

A typical american dairy consumption is 1/45th of a dairy cow’s annual output. Thus consuming dairy for 45 years is approximately a ‘single year of cow life’ (ignoring elasticities). $100 is thought to avert 1 DALY if given to AMF. So if you think a year of cow life averted is about as valuable as one year of human life, then a couple of dollars given to AMF approximately offsets the harm of dairy.

This calculation is necessarily rough, and perhaps the principal problem is comparing ‘years as a dairy cow’ versus ‘years as a human’. Many people will say that human lives are worth more, but others (particularly EAs interested in animal advocacy) might argue that averting years of dairy cow life is more valuable to adding years of human life.

The advent of Animal Charity Evaluators allows a more ‘apples to apples’ comparison, as they offer rough estimates of how effective their charities are in terms of ‘animal lives in agriculture’ averted, which is about the same outcome as we expect from reducing our consumption. These figures are remarkably high: ACE estimates the Humane League as averting 3.4 lives for every dollar spent. It rates Mercy for Animals and Animal Equality International even higher, at 8.8 animals/$ and 10.1 animals/$.

A comparison between donations to THL and abstaining from dairy is instructive. 1/45th of a cow life per year, divided by 3.4 lives per dollar:

Price (in THL donations) of dairy = 1/45 years / 3.4 years/$
= $0.0065

Less than a cent per year to THL (the least cost-effective of ACE’s recommendations) does as much good as abstaining from dairy from the same period, and this might be conservative (we’ve assumed that animal lives only last one year, inter alia).

Dairy is the least harmful of animal products in the typical diet. Consider instead going from a typical american diet to veganism. Peter Hurford helpfully provides the following estimated relative harms of different animal products in the typical american diet:

  • Giving up beef is ~2.1x as important as giving up dairy.
  • Giving up beef is ~1.9x as important as giving up turkey.
  • Giving up pork is ~2.9x as important as giving up beef.
  • Giving up chicken meat is ~11.3x as important as giving up beef.
  • Giving up eggs is ~11.8x as important as giving up beef.
  • Giving up aquacultured fish is ~6.4x as important as giving up beef.

We can convert the relevant foods into ‘dairy equivalents’.

Dairy - 1
Beef - 2.1
Pork - 6.1 (2.1*2.9)
Chicken - 23.7 (2.1*11.3)
Eggs - 24.8 (2.1*11.8)
Fish - 13.4 (2.1*6.4)
TOTAL - 71.1

So the typical american diet is 71.1 times worse than only consuming dairy but abstaining from meat and eggs. We can recalculate the value of turning vegan in terms of yearly THL donations:

Price (in THL donations) of veganism = 1/45 * 71.1 / 3.4 years/$
= $0.46

So being vegan for a year is about as good as giving 46 cents to the humane league.

Sources of uncertainty

This seems remarkably cheap. Should we believe this calculation?

I have made a few short-cuts, and a more careful calculation could be made (perhaps recalculating the animal welfare costs from Hurford rather than taking the bottom line estimates, or dis-aggregating ‘animals’ when looking at the benefit of abstaining from dairy alone). However, it the ‘bottom line’ figure isn’t sensitive enough to these factors to be driven up by several orders of magnitude. Hurford also estimates the typical American diet causes 5.5 years of animal suffering per year. Using this directly gets a factor of 4:

Price (in THL donations of veganism = 5.5 years/3.4 years/$
= $1.62

Further, several elements of the calculation were conservative: I’ve ignored elasticities (which would reduce the amount of animal suffering dietary change would avert), I’ve assumed that all animals saved would live for a year, and I’ve picked THL instead of 2-3x better performing AE or MfA. Scott Alexander has made a similar calculation to mine using similar data, and gets similar ‘bottom line’ figures.

Perhaps weakest component in the calculation is the cost effectiveness of the animal charities. One may think the estimates are too good to be true. Even if one cannot see any obvious feature of over-estimation, regression to the mean means that the estimated-to-be-best charities will be generally overestimated, and this effect can bite particularly hard if the distribution is log-normal or broader.

Is veganism worth it?

Taking these calculations at face value suggests individual dietary change is unimportant, with an equivalent donation value of a handful of dollars each year. Given most people would prefer to keep their current carnivorous diet for more than this amount, this suggests offsetting donations are much better for most than diet change. (Of interest, if milk substitutes are even 1 cent more expensive over the year, then returning to milk and donating the difference seems a better strategy.)

If one is pessimistic about the estimates ACE gives, and anticipates a 3 or 4 order of magnitude correction, then the price of veganism does rise to the level where people might prefer to change diet rather than give moral offsets. However, most animal focused EAs I’ve talked to seem optimistic that ACE’s estimates are not dramatic over-estimates. This optimism seems incongruous with their eagerness to get EAs and EA events to be vegan: persuading them to give their pocket change to THL seems much easier (and better for animal welfare!) than getting them to change their diets. Dietary change looks like a very ineffective intervention.


I foresee two main objections, given the remarks Jeff’s post gathered.

The first is that unlike donation targets, veganism and giving money to THL are not mutually exclusive, so there is no trade-off – just do both! Yet although veganism may not draw directly on our donation budget, it may plausibly draw on our budgets of self-sacrifice and self-control. There are other unpleasant actions we could take which are independent of donations (having cold showers to reduce energy expenditure and thus climate change), which most of us intuit probably aren’t worth it in terms of indirect costs due to their limited impact. I aver that for most people changing their diet, given the extremely low monetary value of an offsetting donation, falls below this threshold. (For those for whom it doesn’t, go vegan!)

The second that this analysis only tries to model ‘first order’ impacts. Being vegetarian or vegan might have a larger impact in terms of persuading others to become vegetarian or vegan, and generally act to signal dismay at animal agriculture. It looks hard for an animal welfare advocate getting any traction if they eat meat. These ‘second order’ effects are hard to quantify, but I struggle to think they would be worth more than 10x or 100x the ‘direct impact’ of not eating animal products. I have been a vegetarian all my life, and I have mentioned it in conversation less than 50 times, and I doubt much more than half the people who know me also know I’m a vegetarian, and much fewer the reasoning why. The amount of ‘second order effects’ of my vegetarianism on animal welfare are wholly trivial, and I doubt I’m wildly away from the population mean in terms of second order impacts.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s