I partially agree. In my second, high estimate model, cell-based meat arrives in 100 years. However, it more likely arrives sooner, e.g. in 2030. From then on, carbon offsetting starts to count. I agree that we should discount future emission reductions, due to the urgency of the climate problem and the possibility of early threshold values in the climate system being passed. But 10 years is not so long.

I'm surprised by the level of agreement between our assumptions. In your model, 200 M$ funding is required to advance clean meat with 0,7 years, whereas I assumed 100M$ and 1 year. You assume a lower greenhouse gas saving: 50% of the current 7,8 Gton CO2 emissions, whereas I assumed an increase in meat consumption in businass as usual scenario, and a reduction of 1 ton CO2 per vegan year, that means a reduction of around 10 Gton (assuming 10B people), but you assumed a 25% probability of success, whereas I assumed 10%. But with more lognormal error distributions, you arrive at higher $/ton estimates. Here's my guesstimate

https://www.getguesstimate.com/models/16723

I quickly made a guesstimate: https://www.getguesstimate.com/models/16723 (you can also compare it with shaybenmoshe's guesstimate below)

In the scenario where the level of funding F is the same every year, if you make a one time donation x, the outcome gets closer by $\frac{x}{F}$ years, therefore you've produced $U\cdot \frac{x}{F}$ utility.

The main assumption behind this result is that some utility U in the future is worth as much as some utility U right now. Therefore, when judging which of two projects is the best, since C only affects how long it will take to complete each project, it doesn't matter. The only quantities that matter are U of course, and the funding F.

What if you want to take into account the fact that no, some utility in the future is worth less than some utility right now, therefore completing quick projects first is better?

Warning : this part will involve math

One common way to do that is to assume that the utility decreases over time in a geometrical manner : 1 utility unit in 1 year is equivalent to $\tau$ utility units now, 1 unit in two years is equal to ${\tau }^2$ utility units now, with $\tau$ slightly smaller than 1. For example, if $\tau$ is equal to 0.99, then one utility point in one century is worth about one third of one now, and the closer $\tau$ is to 1, the more you adopt a longtermist point of view.

Ok so now we can compute the total utility of a project with cost C, annual funding F, and per year utility U :

$U_{total}=U\cdot \frac{1}{1-\tau }\cdot {\tau }^{\frac{C}{F}}$

Now, making a one-time donation x, is the same as decreasing the total cost C by x, thus the utility gain of this donation will be :

$gain=x\cdot \frac{-\partial U_{total}}{\partial C}=U\cdot \frac{-ln\left(\tau \right)}{1-\tau }\cdot {\tau }^{\frac{C}{F}}\cdot \frac{x}{F}=U\cdot {\tau }^{\frac{C}{F}}\cdot \frac{x}{F}$

So we have three factors : U, of course, $\frac{x}{F}$, which is how many years of funding you'll provide with your donation, and ${\tau }^{\frac{C}{F}}$ which represents how much this future utility is worth, compared to utility right now. Once again, if you think that utility in the future is equal to utility now, which means $\tau =1$, you get $gain=U\cdot \frac{x}{F}$, which is the original formula in the post. C matters only if utility in the future is not equal to utility in the present.

Now, we can re-write this formula with your notion of return on investment :

$gain=x\cdot \frac{U}{C}\cdot \frac{C}{F}{\tau }^{\frac{C}{F}}$

With this version, we see three factors influencing the gains : x, the bigger your donation, the better, $\frac{U}{C}$, the bigger the return on investment, the better, and finally $\frac{C}{F}{\tau }^{\frac{C}{F}}$, which is a function of $\frac{C}{F}$, the number of remaining years. This function is convex, it starts at 0, reaches a maximum, and its limit is 0 again when $\frac{C}{F}$ approches infinity.

#TODO : include a graph of this function, once I figure out how to do that

With this model, when $\frac{C}{F}$ is too small, it means that the project will soon be funded with or without you, thus you shouldn't invest in it. On the other extreme, if $\frac{C}{F}$ is too big, the benefits will take place too far in the future, and because utility points lose value when too far in the future, you shouldn't invest in it either. In the middle are the best projects.

Anyway, the main point is : the cost C matters only if you think that utility right now is worth more than utility in the future, otherwise only the funding F matters

The basic (in my opinion realisitic) assumption is that other people invest in cell-based meat R&D anyway, and that in the business-as-usual scenario (where you do not fund anything) no other strategy (technology, intervention, vegan outreach campaign,...) will be able (even with more funding) to abolish animal farming before cell-based meat enters the market at competitive prices. Suppose cell-based meat arrives within a few decades and eliminates animal farming in say 50 years, whereas another, next best strategy would eliminate animal farming in 100 years. Suppose that this other strategy was less costly, for example requiring only 10 million euro funding per year over a period of 100 years to abolish animal farming, whereas cell-based meat would require 100 million euro funding over 50 years. And suppose that other strategy was more neglected, for example receiving only 10 million euro funding per year, compared to 100 million for cell-based meat. Even then, extra funding for that other strategy would not be effective when it is impossible to speed it up such that it will eliminate animal farming within 50 years. When that other strategy takes more than 50 years anyway, it will become obsolete anyway in the business-as-usual scenario where cell-based meat arrives earlier and eliminates animal farming earlier. A global coordination such that all cell-based meat funding goes to that other, less costly strategy, is not effective (not so feasible). Hence, the most effective thing to do for us, is to accelerate that cell-based meat research, such that it enters the market one year earlier. That saves an extra year of animal suffering and greenhouse gas emissions. If other strategies received more funding, there is a likelihood that they make cell-based meat obsolete, and this consideration is included in the 10% probability of cell-based meat eliminating animal farming.

Thanks

about the 10.000 years assumption: that is only used to calculate a high estimate of clean meat R&D. I'm not so worried if that is an overstimate.

My calculation assumes indeed no diminishing returns for clean meat R&D. I don't expect diminishing returns in the short run, when so much need to be researched. In my model, the decreasing neglectedness accounts for diminishing returns. When funding and investments by others increses to 1 billion dollars, the cost-effectiveness decreases with a factor 10. Anyway, the point is that clean meat R&D is a good opportunity in the short run, for the next 10 or 20 years.

ACE's CEA methodology is different indeed, but Im not convinced that it is really incomparable to mine. A basic assumption is that ACE's top charities who are not involved in clean meat (i.e. the charities except Good Food Institute), are not capable of eliminating animal farming before clean meat can.

The CEA of leafleting could be an overestimate indeed. The study that I did, was not randomized controlled.

About missing the impact of animal advocacy: I'm sceptical about the possibility of attitudinal change: just like the expectations of leafleting were too high (not strong evidence of behavioral change), the expectations about other animal rights advocacy could be too high as well. The case for clean meat is different: in the past we already have striking examples of animals being replaced by more than 90% within 50 years due to new technologies (e.g. horses -> cars, whale oil -> kerosene, messenger pigeons -> telephone/telegraph, sheep wool -> synthetic fibers, animal insulin -> human recombinant DNA insulin, rabbit skin tests for cosmetics -> human skin tissueand perhaps now movie animals -> CGI animals). These transitions were independent from animal rights campaigning.

I do see much room left for attitudinal change, in particular moral circle expansion (see e.g. https://stijnbruers.wordpress.com/2020/03/17/consider-remarkable-animal-capabilities-to-expand-the-moral-circle/), but perhaps after 10 or 20 years, when clean meat is already well on track and lost its opportunity for more funding (and returns diminished). Also, once people automatically decrease their animal meat consumption, they suffer less from cognitive dissonance, which means attitudinal change might become easier.

I'm skeptical about the welfare reforms strategy. For me to be indifferent between the current welfare reforms and an X% reduction of animal farming, I think X is very low, probably lower than 10%. For example cage free eggs: I don't believe that, if all battery cages were abolished and chickens had free range, that count for more than a 10% improvement in welfare, and probably a 0% in animal rights. Given moral uncertainty, I put some probability on a rights-based ethic where animals should not be used as merely a means. Also, some of the future possible welfare reforms could be so difficult, that clean meat (or animal-free eggs) will arrive sooner, making the welfare reforms campaigns obsolete. Also, welfare campaigns are also much less neglected than clean meat R&D.