I have a PhD in finance and am the strategist at Affinity Impact, the impact initiative of a Singapore-based family office that makes both grants and impact investments.
Wayne_Chang
Thanks for posting this, Jonathan! I was going to share it on the EA Forum too but just haven’t gotten around to it.
I think GIF’s impact methodology is not comparable to GiveWell’s. My (limited) understanding is that their Practical Impact approach is quite similar to USAID’s Development Innovation Ventures’ impact methodology. DIV’s approach was co-authored by Michael Kremer so it has solid academic credentials. But importantly, the method takes credit for the funded NGO’s impact over the next 10 years, without sharing that impact with subsequent funders. The idea is that the innovation would fail without their support so they can claim all future impact if the NGO survives (the total sum of counterfactual impact need not add to 100%). This is not what GiveWell does. GiveWell takes credit for the long-term impact of the beneficiaries it helps but not for the NGOs themselves. So this is comparing apples to oranges. It’s true that GiveWell Top Charities are much more likely to survive without GiveWell’s help but this leads to my next point.
GiveWell also provides innovation grants through their All Grants Fund (formerly called Incubation Grants). They’ve been funding a range of interventions that aren’t Top Charities and in many cases, are very early, with GiveWell support being critical to the NGO’s survival. According to GiveWell’s All Grants Fund page, “As of July 2022, we expect to direct about three-quarters of our grants to top charity programs and one-quarter to other programs, so there’s a high likelihood that donations to the All Grants Fund will support a top charity grant.” This suggests that in GiveWell’s own calculus, innovation grants as a whole cannot be overwhelmingly better than Top Charities. Otherwise, Top Charities wouldn’t account for the majority of the fund.
When thinking about counterfactual impact, the credit one gets for funding innovation should depend on the type of future donors the NGO ends up attracting. If these future donors would have given with low cost-effectiveness otherwise (or not at all), then you deserve much credit. But if they would have given to equally (or even more) cost-effective projects, then you deserve zero (or even negative) credit. So if GIF is funding NGOs that draw money from outside EA (whereas GiveWell isn’t), it’s plausible their innovations have more impact and thus are more ‘cost-effective’. But we are talking about leverage now, so again, I don’t think the methodologies are directly comparable.
Finally, I do think GIF should be more transparent about their impact calculations when making such a claim. It would very much benefit other donors and the broader ecosystem if they can make public their 3x calculation (just share the spreadsheet please!). Without such transparency, we should be skeptical and not take their claim too seriously. Extraordinary claims require extraordinary evidence.
A 7% real investment return over the long-term is in my opinion, highly aggressive. World real GDP growth from 1960 through 2019 is 3.5%. Since the proposed fund expects to invest over “centuries or millennia,” any growth rate faster than GDP eventually takes over the world. Piketty’s r > g can’t work if wealth remains concentrated in a fund with no regular distributions.
Even in the shorter run, it’s unrealistic to expect the fund to implement a leveraged equity-only strategy (or analogous VC strategy):
1) A leveraged approach may not survive (e.g. will experience −100% returns). Even if the chance is small over a given year, this will be increasingly likely over a longer horizon. Dynamic leverage strategies can be implemented to reduce this risk but this likely reduce returns too.
2) A high-risk strategy will result in extremely painful drawdowns. In bad times, any fiduciary running the fund will face enormous pressure to shift to a more conservative strategy. During the Great Depression, US equities declined by nearly 90% during the course of just 3 years, even without leverage. Sticking to the same approach in the face of a potentially worse decline is nearly unimaginable.
3) A consistently leveraged portfolio approach has never been done before over long investment periods. Foundation/university endowments are probably in the most analogous position and few apply leverage. Harvard tried a modest 5% leverage during the 2000’s, and it blew up during the Financial Crisis.
4) Any successful strategy will be mimicked and thus face increasing competition and declining returns. If the fund grows to any significant size, it will start facing competition from itself. For example, Yale’s legendary endowment has seen declining returns from a ~9.5% real rate over the past 20 years to a ~5.5% one over the past decade. Similarly, given Berkshire Hathaway’s large size, it’s now increasingly difficult for Warren Buffet to beat the stock market.
Indeed, the proposed fund may actually have to be quite conservative for it to survive over time (through broad diversification even into low-return assets) and be accepted by the world (to avoid scrutiny or excess taxation). In my opinion, when investing over centuries with an unprecedented strategy, I would characterize a 2-4% real return (broad asset class diversification that keeps up with world GDP) as reasonable, and a 5%+ real return (all equity with or without leverage) as aggressive.
Hauke’s calculation simply determines a standard Benefit/Cost ratio. If it costs $10 to avert a tonne of CO2 that provides benefits of $417 (in damages averted), this Benefit/Cost ratio equals 41.7. This ratio should be directly comparable to Copenhagen Consensus ‘Social, economic, and environmental benefit per $1 spent.’ For the Post-2015 Consensus, ‘Climate Change Adaption’ is listed as providing a Benefit/Cost ratio of 2 while climate-related ‘Energy Research’ has a ratio of 11. I would weight these results from meta-level research must more strongly than that from a single study. But even if we believed Hauke’s study, a benefit/cost ratio of 41.7 still lags ‘Reduce Child Malnutrition’ (ratio of 45) or ‘Expanded Immunization’ (ratio of 60). This hardly suggests that “we should consider prioritizing climate change over global development interventions.” The unconditional cash transfer benchmark that Hauke uses is a minimum and not representative of highly cost-effective interventions in global development. Using GiveWell’s estimates, deworming and malaria nets are more than 10x more cost-effective than cash. Before rushing to replace well-established priorities and interventions that are based on decades of research, we need to have substantial confidence in the new priority/intervention. This study is far from it.
Note that the Copenhagen Consensus and GiveWell results do not apply utility adjustments. If this new climate change study does so, its Benefit/Cost ratio would be distorted by improperly inflating Benefits, which make the ratio larger than it actually is.
Thanks for your response, kbog!
Animal welfare issues are plausibly getting worse and not better so I’d be less confident to assume it will not be an issue in the future. As the world develops and eats more meat, Compassion in World Farming estimates that annual factory farm land animals killed could increase by 50% over the next 30 years. Assuming people’s expanding moral circle will reverse this trend is dangerous when the animal welfare movement has progressed little over the past few decades (number of vegetarians in US have been flat; there are some animal welfare legislative victories but also setbacks like ag-gag rules). Innovations like clean meat could help but it is still early, and there are also ways technology can make things even worse. Assuming animal welfare issues remain as they currently are (neither deteriorating nor improving) seems to me a plausible and more responsible projection.
If so, for the Long Term Future EA Fund, let’s assume the Animal Welfare EA Fund “offset ratio” (to account for the meat eater problem) is the same for future generations as it is for the current generation. Based on your blog’s estimate of a nickel a day, it costs a person ~$1000 to offset a lifetime of meat consumption ($0.05/day x 365 days/year x 50 years). It seems your estimate is for people living in rich countries though, so maybe 30% of that or ~$300 is more applicable to the average human. This can be compared to the Long Term Future Fund’s expected cost effectiveness of saving a human life (for just the current generation). I’ve seen one estimate that assumes a reduction in x-risk of 1% for $70 billion dollars spent (again for the current generation only). This leads to ~$1000 per human life saved ($70 billion / 7 billion humans / 1%). If so, the meat eater problem offset ratio for the Long Term Future Fund is very roughly ~30% (~$300 offset per life saved / ~$1000 to save a life).
Let’s apply a similar logic to the Global Health EA Fund. Instead of ~$1000 to offset a lifetime of meat consumption, let’s assume 10% of that for someone living in extreme poverty, or ~$100. GiveWell estimates that AMF can save a life for ~$3000, leading to an offset ratio of ~3% (~$100 offset per life saved / ~$3000 to save a life). This is two orders of magnitude larger than your comment response (of 0.008% ~ 0.04% from $0.08 ~ $0.4 / $1000). One reason might be because you’re only accounting for one year of the meat eater problem when I’ve accounted for a lifetime’s worth of impact (which I believe is the more complete counterfactual comparison). However, I’ve not had a chance to dive into your spreadsheet so I could be mis-using your results. Any corrections or reactions are much appreciated!
Finally, I’m curious as to why you think offsetting makes little sense under utilitarianism. I’m thinking it would actually be required if one were uncertain about the conversion ratio between human and animal welfare. If we were certain about the conversion, we should just do the one intervention that’s most cost effective, in whatever domain it happens to be in (human or animal). But if we were uncertain about the conversion, we will need to ensure that one domain’s actions doesn’t inadvertently produce overall negative utility when the other domain’s consequences are summed together. In the case of saving a human life, we wouldn’t want to lower overall utility because of our underestimation of the meat eater problem. On the other hand, we wouldn’t want to just focus on animal welfare if it turns out human welfare is especially significant. Offsetting cross-domain spillover effects avoids this dilemma (I teach finance, where analogies include hedging different FX risks or asset-liability matching). For the meat eater problem, it ensures saving a human life does not lead to negative utility even if we find out that animal welfare is unexpectedly important. The offset trades one animal life for another animal life, ensuring neutral utility impact within the animal domain.
Sorry for the long reply but I’ve been worrying about the meat eater problem so found your post to be especially interesting and informative. Any response you might have would be very appreciated!
Thanks so much for such a thorough and great summary of all the various considerations! This will be my go-to source now for a topic that I’ve been thinking about and wrestling with for many years.
I wanted to add a consideration that I don’t think you explicitly discussed. Most investment decisions done by philanthropists (including the optimal equity/bond split) are outsourced to someone else (financial intermediary, advisor, or board). These advisors face career risk (i.e. being fired) when making such decisions. If the advisor recommends something that deviates too far from consensus practice, they have to worry about how they can justify this decision if things go sour. If you are recommending 100% equities and the market tanks (like it did last year), it’s hard to say ‘But that’s what the theory says,’ when the reflective response by the principal is that you are a bad advisor because you don’t understand risk. Many advisors have been fired this way, and no one wants to be in that position. This means tilting toward consensus is likely the rational thing to recommend as financial advisors. There are real principal-agent issues at play, and this is something acutely felt by practitioners even if it’s less discussed among academics.
I suspect the EA community is subject to this dynamic too. It’s rarely the asset owners themselves who decide the equity mix. Asset allocation decisions are recommended by OpenPhil, Effective Giving, EA financial advisors, etc. to their principals, and it’s dangerous to recommend anything that deviates too far from practice. This is especially so when EA’s philanthropy advice is already so unconventional and is arguably the more important battle to fight. It can be impact-optimal over the long term to tilt toward asset allocation consensus when not doing so risks losing the chance to make future grant recommendations. The ability to survive as an advisor and continue to recommend over many periods can matter more than a slightly more optimal equity tilt in the short term.
Keynes comes to mind: “Worldly wisdom teaches that it is better for reputation to fail conventionally than to succeed unconventionally.”
I highly recommend the Founder’s Pledge report on Investing to Give. It goes through and models the various factors in the giving-now vs giving-later decision, including the ones you describe. Interestingly, the case for giving-later is strongest for longtermist priorities, driven largely by the possibility that significantly more cost-effective grants may be available in the future. This suggests that the optimal giving rate today could very well be 0%.
Hi Joel, I skimmed your report really quickly (sorry) but suspect that you did not account for soda taxes being eventually passed anyway. So the modeled impact of any intervention shouldn’t be going to 2100 or beyond but out only a few years (I’d think <10 years) when soda taxes would eventually be passed without any active intervention. You are trying to measure the impact of a counterfactual donated dollar in the presence of all the forces already at play that are pushing for soda taxes (how some countries already have them). This makes for a more plausible model, and I believe is how LEEP or OpenPhil model policy intervention cost-effectiveness (I could be wrong though).
Minor suggestion: in your title and summary, please just write out “10 k” as 10,000. No need to abbreviate when people may be unsure that it’s actually 10,000 (given that it’s such a large difference).
I agree with Michael that concrete examples would be very helpful, even for researchers. A post should be informative and persuasive, and examples almost always help with that. In this case, examples can also make clear the underlying logic, and where the explanation can be confusing.
For example, let’s think about investing in alternative protein companies as a way to tackle animal welfare. Assume that in a future state where lots more people eat real meat (bad world state), the returns for alt-proteins in that state are low but cost-effectiveness is high. This could be because alt proteins have faced lower rates of adoption (low returns) but it’s now easier to persuade meat eaters to switch (search costs are now low since more willing-switchers can be efficiently targetted). The opposite situation is true too. In a good future state with few meat-eaters, alt protein returns are high but cost-effectiveness is low. So this scenario should put us in your table’s upper left quadrant (negative correlation btw/ World State and Cost-Effectiveness + negative correlation btw/ Return and Cost-Effectiveness).
This example illustrates how some of your quadrant descriptions may be confusing or even inappropriate:
“Underweight investment”: I agree with this one since to have a greater EV, you want investments with a positive correlation between returns and cost-effectiveness. This isn’t true for alt proteins here, so you should avoid them.
“Divest from evil to do good”: I don’t think this makes sense because alt proteins are not “evil” (but you should avoid them given the scenario).
“Mission leveraging”: I was quite confused initially because I was assuming that the comparison is to no investment at all. If so, then investing in alt proteins can lead to an ambiguous impact on volatility (depending on the relative magnitude of return changes versus cost-effectiveness changes). It could in fact be mission hedging (with an improvement in the bad state) if the low returns end up producing more total good because of the state’s high cost-effectiveness. However, I eventually realized that the comparison is to a fixed grant within the animal welfare space (although this was never made explicit in the post and may not be what most people would assume). If so, then indeed this is always mission leveraging since a positive correlation between the world state and returns does ensure lower volatility.
So as you can see, an example makes clear where table descriptions may be inappropriate and where a clearer description can be helpful. It also makes more concrete what various correlation signs mean and how to think about them.
This post (and the series it summarizes) draws on the scientific literature to assess different ways of considering and classifying animal sentience. It persuasively takes the conversation beyond an all-or-nothing view and is a significant advancement for thinking about wild animal suffering as well farm animal welfare beyond just cows, pigs, and chickens.
Thanks for the clarification, Owen! I had mis-understood ‘investment-like’ as simply having return compounding characteristics. To truly preserve optionality though, these grants would need to remain flexible (can change cause areas if necessary; so grants to a specific cause area like AI safety wouldn’t necessarily count) and liquid (can be immediately called upon; so Founder’s Pledge future pledges wouldn’t necessarily count). So yes, your example of grants that result “in more (expected) dollars held in a future year (say a decade from now) by careful thinking people who will be roughly aligned with our values” certainly qualifies, but I suspect that’s about it. Still, as long as such grants exist today, I now understand why you say that the optimal giving rate is implausibly (exactly) 0%.
Here’s the math on moral/financial fungibility:
...
You’re probably better off eating cow beef and donating the $6.03/kg to the Good Food Institute
Is refraining from killing really morally fungible to killing + offsetting? Would it be morally permissible for someone to engage in murder if they agreed to offset that life by donating $5,000 to Malaria Consortium? I don’t mean to be offensive with this analogy, but if we are to take seriously the pain/suffering that factory farming inflicts on animals, we should morally regard it in a similar lens to inflicting pain/suffering on humans.
So, no, moral acts are not necessarily fungible. It is better to not eat meat in the first place than to eat meat and donate the savings to farm animal charities (even if you could save more animals). This is obvious from a rights moral framework but even consequentialists would consider financial offsetting dangerous and unpalatable. The consequences of allowing people to engage in immoral acts + offsetting would be a treacherous and ultimately inferior world.
So your calculations are not the cost of eating meat but rather, the cost of saving animals. You have not estimated the cost of chicken/cow suffering (which would require estimating utility functions and animal preferences), but rather, the cost of alleviating suffering. Your low-cost numbers don’t imply that eating meat is inconsequential, but rather, that it’s very cost-effective to help chickens and cows. GiveWell’s $5,000 per human life doesn’t make human life cheap or murder trivial, it means we have an extraordinary opportunity to help others at a very low cost to ourselves.
I don’t think it makes sense to compound the model distributions (e.g. from 1 year to 10 years). Doing so leads to non-intuitive results that are difficult to justify.
1) Compounded model results (e.g. 10x impact in 10 years) are highly sensitive to the arbitrarily assumed shape, range, and skewness parameters of the variable distributions. Also, these results will vary wildly from simulation to simulation depending on the sequence of random draws. This points to the model’s fragility and leads to unnecessary confusion.
2) The parameter estimates may use annualized growth rates, but they need not correspond to an annual time frame. Indeed, it is more realistic to make estimates for longer horizons because short-term noise averages out (i.e. Law of Large Numbers). In other words, it is far easier to estimate a variable’s expected mean than its underlying distribution. Estimates for the expected mean will already be highly uncertain. I don’t think it’s possible to reasonably defend distribution assumptions of the variables themselves.
The exercise is to compare giving-today vs. investing-to-give-later. The post usefully identifies key variables in this consideration. I think the most it can do is propose useful estimates of these variables’ expectations over the long run (i.e. their averages over time) and their key uncertainties (i.e. Knighting uncertainty and not quantifiable distribution parameters). If the expectations’ net sum is above 1, it makes sense to give later. If it falls below 1, it makes sense to give now. Reasonable areas of uncertainty can be further discussed and debated. Already, there will be much irreconcilable (rational) disagreement. Compounding returns using arbitrary distribution parameters won’t (and shouldn’t) reconcile any differences and likely confuses the matter.
Thanks for posting this, kbog! I would be interested in your recommendation for someone donating to the EA funds. The Long Term Future and Global Development funds focus on humans and thus potentially runs into the meat eater problem. For every dollar donated to the above funds, what would be an appropriate amount to donate to the Animal Welfare Fund that is enough to offset this issue? Thanks!
Thanks for your response, Joel!
Stepping back, CEARCH’s goal is to identify cause areas that have been missed by EA. But to be successful, you need to compare apples with apples. If you’re benchmarking everything to GiveWell Top Charities, readers expect your methodology to be broadly consistent with GiveWell’s and their conservative approach (and for other cause areas, consistent with best-practice EA approaches). The cause areas that are standing out for CEARCH should be because they are actually more cost-effective, not because you’re using a more lax measuring method.
Coming back to the soda tax intervention, CEARCH’s finding that it’s 1000x GiveWell Top Charities raised a red flag for me so it seemed that you must somehow be measuring things differently. LEEP seems comparable since they also work to pass laws that limit a bad thing (lead paint), but they’re at most ~10x GiveWell Top Charities. So where’s the additional 100x coming from? I was skeptical that soda taxes would have greater scale, tractability, or neglectedness since LEEP already scores insanely high on each of these dimensions.
So I hope CEARCH can ensure cost-effectiveness comparability and if you’re picking up giant differences w/ existing EA interventions, you should be able to explain the main drivers of these differences (and it shouldn’t be because you’re using a different yardstick). Thanks!
Hi Huwelium, thanks so much for your post! I’m also advising someone on highly cost-effective interventions, so I found your thoughtful analysis to be very interesting. My question relates to your cost effectiveness estimates vs GiveWell’s. Based on GiveWell’s spreadsheet, their modeling of DDK (2017) places that program’s cost effectiveness at 0.5x – 2.5x GiveDirectly’s. Their modeling of Bettinger et al (2017) places that program’s at 0.2x – 1.4x GiveDirectly’s. Both of these estimates are for consumption effects only and excludes non-pecuniary benefits like reduced teenage pregnancy. This seems most comparable with your document’s cost-effectiveness estimates, which are based on income effects only. However, for Pratham, you conclude its cost effectiveness is 20x − 200x GiveDirectly’s.
I’m having trouble undertanding how your estimates are one to two orders of magnitude different from GiveWell’s. I’m probably missing something important so I was wondering if you’ve attempted a reconciliation. Any clarification on assumption differences and their relative importance would be very much appreciated. Thanks so much!
New phrasing works well!
Got it. But I think the phrasing for the number of animals that die is confusing then. Since you say “100 other human [sic] would probably die with me in that minute,” the reference is to how many animals would also do during that minute. I think what you want to say is for every human death, how many animals would die, but that’s not the current phrasing (and by that logic, the number of humans that would die per human death would be 1, not 100).
I’d suggest making everything consistent on a per-second basis as smaller numbers are more relatable. So 1 other human would die with you that second, along with 10 cows, etc.
Thanks for writing this! The very last sentence seems off. Did you mean to say every second (instead of minute)? Also, the number of farm animals that die every second should be 1⁄60 (not 1⁄120) of that in the “minute” table above.
This last sentence was quite shocking for me to read. It’s sad…but very powerful.
This criticism seems unfair to me:
It seems applicable to any type of advocacy. Those who promote global health and poverty are likely biased toward foreign people. Those who promote longtermism are likely biased toward future people. Those who advocate for effective philanthropy are likely biased toward effectiveness and/or philanthropy.
There’s no effective counter-argument since, almost by definition, any engagement is possibly biased. If one responds with, “I don’t think I’m biased because I didn’t have these views to begin with,” the response can always be, “Well, you engaged in this topic and had a positive response, so surely, you must be biased somehow because most people don’t engage at all.” It seems then that only criticisms of the field are valid.
This is reminiscent of an ad hominem attack. Instead of engaging in the merits of the argument, the critique tars the person instead.
Even if the criticism is valid, what is to be done? Likely nothing as it’s unclear what the extent of the bias would be anyway. Surely, we wouldn’t want to silence discussion of the topic. So just as we support free speech regardless of people’s intentions and biases, we should support any valid arguments within the EA community. If one is unhappy with the arguments, the response should be to engage with them and make valid counterarguments, not speculate on people’s initial intuitions or motivations.