Interesting take! I found myself having 3 objections when I read it
1. The chart showing meat <> GDP per capita doesn't prove much. We can get similar charts for obesity, or alcohol. Revealed preferences don't reveal everything.
2. The "happiness hit" of a chicken sandwich (or falafel) seems way out of proportion. I'd bet that these things don't replicate, or hold steady, or even be always positive, even for the same person over time. This means the compounding doesn't necessarily happen. i.e., you can't say she was x% more productive today because she was happy, and this compounds for her whole life. More likely being point-happy gives a one time tiny boost, which is nice but decays like diminishing marginal utility.
3. The link of self-proclaimed happiness to productivity is interesting but I'm a tad suspicious of a Said b school article to be honest. Even if its true in a ceteris paribus condition, these things have crazy number of confounders. For instance, if we'd done this same analysis twenty years ago we might've said the same about smoking (makes me happier, happier people are more productive, productivity compounds).
Regarding your comment on carbon offsets, I'd echo Scott Alexander and others in noting that while the existence of an optional ethical offset price can br useful, it is only relevant **if you actually commit to pay it**.
Once the low hanging fruit in GHG offsets are cleared, those prices should start going up and, assuming you are fully committed to lifestyle offsets, the cost tradeoff might encourage your veganism anyway.
Simply knowing that you *could* offset your behaviour does not actually affect your purchasing decisions. It isn't until you internalise the cost that you can credibly claim that the offsets have any relevance. The theological antecedent here being that if you sin all you like because you could buy Indulgences, but then never actually buy Indulgences, you are still going to hell.
On "low hanging fruit":
GHG offsetting has associated costs, whether that is through tree planting, land banking, or Direct Air Capture (DAC). If **everyone** (or even just, say, the wealthiest 80% of people/corporations/governments) were to be offsetting, the supply of the cheapest offsets (third world land banks, forestry restoration) would be rapidly depleted, with low scope for supply response. The marginal carbon offset would be costly DAC, the price of which would be sufficiently high to be worth including in decisions to fly or to eat meat.
Hope that clears up confusion, but if not just let me know what part was hard to understand.
I have a few objections to this but the most obvious one is that it seems totally crazy to think that eating a chicken sandwich rather than a falafel wrap (or whatever) is likely to make a worker 5% more productive, even for someone who loves chicken sandwiches. Is there any evidence at all that adopting a vegetarian/vegan diet reduces productivity?
Yeah so that's the rub, it doesnt really matter what % more productive it makes you, because it compounds. Also, the assumption is that the chicken sandwich is his preference. If the falafel is truly equivalent to his tastes, then by all means he should eat it, but this doesn't square with what we see in global revealed preferences. Meat makes people happy!
I guess I just think the percentages here seem way off, fair enough if you think that has no implications for the point. Like the 10 - 15% claim seems very specific, the paper you link from the Oxford business school is referring to the change in productivity expected given a one unit change in happiness (e.g. moving from saying you're feeling 5/10 to 6/10), and the change in productivity is 12%. But it seems really unlikely to me that switching from a meat sandwich to some lesser sandwich results in a one unit increase in happiness, I wouldn't be surprised if there was zero change tbh.
Haven't looked through this in detail, but here's one study (note: not randomised so take with a huge pinch of salt) showing that workers put on a vegan nutrition plan saw a decrease in productivity impairments relative to the control group:
From a personal perspective, this is the same for me. It really doesn't phase me in terms of satisfaction whether I have the meat or vegan version (in the specific case of falafel vs chicken sandwich I'm falafel all the way). It leads me to question the validity of the revealed preference when:
a. cultural and social factors play a huge role - someone eats meat to signal (mostly to themselves) that they're not poor.
b. meat is a great source of amino acids if you're on the edge of being poor - rice with chicken is better than rice with a tiny bit of chicken, maybe that's why someone eats it. For them it probably really does mean better performance because they've a better diet than the vegan alternative - for me it makes no such difference.
PS. Connor I found your substack by searching for mine. Yours comes up, mine doesn't!
The point about happiness increases for animals being a one shot gain vs happiness levels for humans compounding is an excellent one. I hadn't really considered it before, and it is apparently too subtle for others, but it seems almost certainly true given how productivity is linked to mental state. Given that as the case, how much of a compounding hardly matters, so long as it is greater than zero it will be immense over time and population.
I enjoyed the argument - Connor never fails to find an interesting angle!
One thing I'd add is I don't think plant-based diets are fixed, I think cultures can basically make them better or worse. Vegan options were pretty limited and bland when dining out 20 years ago (and still are in some parts of the country) but today in some places they're plentiful and tasty and that appears to be a youth-driven trend that will keep increasing, but who knows.
You can probably see this with meat too - BBQ options are a lot more varied and enjoyable in parts of the south than say Boston or DC.
(Side note - and maybe I'm alone in this - but I find Potbelly's so bland and disappointing, yet they are everywhere in DC, that I don't think people regardless of veganism are maximizing their utility in the food arena.)
I think you'd also have to bet against plant-based foods never being able to be indistinguishable from animal-based products down the road. I would love to see more blind taste testing.
Interesting take! I found myself having 3 objections when I read it
1. The chart showing meat <> GDP per capita doesn't prove much. We can get similar charts for obesity, or alcohol. Revealed preferences don't reveal everything.
2. The "happiness hit" of a chicken sandwich (or falafel) seems way out of proportion. I'd bet that these things don't replicate, or hold steady, or even be always positive, even for the same person over time. This means the compounding doesn't necessarily happen. i.e., you can't say she was x% more productive today because she was happy, and this compounds for her whole life. More likely being point-happy gives a one time tiny boost, which is nice but decays like diminishing marginal utility.
3. The link of self-proclaimed happiness to productivity is interesting but I'm a tad suspicious of a Said b school article to be honest. Even if its true in a ceteris paribus condition, these things have crazy number of confounders. For instance, if we'd done this same analysis twenty years ago we might've said the same about smoking (makes me happier, happier people are more productive, productivity compounds).
Revealed preferences don't reveal that people enjoy consuming more calories and alcohol as they can afford to?
No. More alcohol does not make you happier in the medium term even though it does in very short term. Hence obv no compounding happiness advantage.
Regarding your comment on carbon offsets, I'd echo Scott Alexander and others in noting that while the existence of an optional ethical offset price can br useful, it is only relevant **if you actually commit to pay it**.
Once the low hanging fruit in GHG offsets are cleared, those prices should start going up and, assuming you are fully committed to lifestyle offsets, the cost tradeoff might encourage your veganism anyway.
Could you elaborate? I didn't understand this
On "sin taxes only work if you pay them":
Simply knowing that you *could* offset your behaviour does not actually affect your purchasing decisions. It isn't until you internalise the cost that you can credibly claim that the offsets have any relevance. The theological antecedent here being that if you sin all you like because you could buy Indulgences, but then never actually buy Indulgences, you are still going to hell.
On "low hanging fruit":
GHG offsetting has associated costs, whether that is through tree planting, land banking, or Direct Air Capture (DAC). If **everyone** (or even just, say, the wealthiest 80% of people/corporations/governments) were to be offsetting, the supply of the cheapest offsets (third world land banks, forestry restoration) would be rapidly depleted, with low scope for supply response. The marginal carbon offset would be costly DAC, the price of which would be sufficiently high to be worth including in decisions to fly or to eat meat.
Hope that clears up confusion, but if not just let me know what part was hard to understand.
I have a few objections to this but the most obvious one is that it seems totally crazy to think that eating a chicken sandwich rather than a falafel wrap (or whatever) is likely to make a worker 5% more productive, even for someone who loves chicken sandwiches. Is there any evidence at all that adopting a vegetarian/vegan diet reduces productivity?
Yeah so that's the rub, it doesnt really matter what % more productive it makes you, because it compounds. Also, the assumption is that the chicken sandwich is his preference. If the falafel is truly equivalent to his tastes, then by all means he should eat it, but this doesn't square with what we see in global revealed preferences. Meat makes people happy!
I guess I just think the percentages here seem way off, fair enough if you think that has no implications for the point. Like the 10 - 15% claim seems very specific, the paper you link from the Oxford business school is referring to the change in productivity expected given a one unit change in happiness (e.g. moving from saying you're feeling 5/10 to 6/10), and the change in productivity is 12%. But it seems really unlikely to me that switching from a meat sandwich to some lesser sandwich results in a one unit increase in happiness, I wouldn't be surprised if there was zero change tbh.
Haven't looked through this in detail, but here's one study (note: not randomised so take with a huge pinch of salt) showing that workers put on a vegan nutrition plan saw a decrease in productivity impairments relative to the control group:
https://pubmed.ncbi.nlm.nih.gov/20389060/. Figure 1 of interest, if true I guess we should say it makes the case for becoming vegan stronger!
From a personal perspective, this is the same for me. It really doesn't phase me in terms of satisfaction whether I have the meat or vegan version (in the specific case of falafel vs chicken sandwich I'm falafel all the way). It leads me to question the validity of the revealed preference when:
a. cultural and social factors play a huge role - someone eats meat to signal (mostly to themselves) that they're not poor.
b. meat is a great source of amino acids if you're on the edge of being poor - rice with chicken is better than rice with a tiny bit of chicken, maybe that's why someone eats it. For them it probably really does mean better performance because they've a better diet than the vegan alternative - for me it makes no such difference.
PS. Connor I found your substack by searching for mine. Yours comes up, mine doesn't!
Thanks for the comment and I hope you enjoy! Sorry to crowd you out in the "All Trades" genre, but I'll be sure to check out your page!
The point about happiness increases for animals being a one shot gain vs happiness levels for humans compounding is an excellent one. I hadn't really considered it before, and it is apparently too subtle for others, but it seems almost certainly true given how productivity is linked to mental state. Given that as the case, how much of a compounding hardly matters, so long as it is greater than zero it will be immense over time and population.
Great essay!
I enjoyed the argument - Connor never fails to find an interesting angle!
One thing I'd add is I don't think plant-based diets are fixed, I think cultures can basically make them better or worse. Vegan options were pretty limited and bland when dining out 20 years ago (and still are in some parts of the country) but today in some places they're plentiful and tasty and that appears to be a youth-driven trend that will keep increasing, but who knows.
You can probably see this with meat too - BBQ options are a lot more varied and enjoyable in parts of the south than say Boston or DC.
(Side note - and maybe I'm alone in this - but I find Potbelly's so bland and disappointing, yet they are everywhere in DC, that I don't think people regardless of veganism are maximizing their utility in the food arena.)
I think you'd also have to bet against plant-based foods never being able to be indistinguishable from animal-based products down the road. I would love to see more blind taste testing.