| A | B | C | D | E | F | G | H | I | J | K | L | M | N | O | P | Q | R | S | T | U | V | W | X | Y | Z | |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
1 | GWWC Cost-Effectiveness | |||||||||||||||||||||||||
2 | ||||||||||||||||||||||||||
3 | Item | Value | Source | |||||||||||||||||||||||
4 | Pledgers | |||||||||||||||||||||||||
5 | Annual donations per 2025 pledger | 3851 | ||||||||||||||||||||||||
6 | Recording adjustment | 0.78 | ||||||||||||||||||||||||
7 | Counterfactual adjustment: In Dullaghan's (2019) old analysis of RC Forward, it is noted that GiveWell estimates it directed $149 million to its recommended charities in 2017; this is less than 1% of the $49 billion received by the 100 largest charities in the USA in 2018 suggesting only a tiny fraction of donors are diverting funds to EA charities; in the absence of more precise data from donor surveys, Dullaghan assumes RC Forward captured only a small percent of Canadian donors who were planning to donate to non-EA charities and then switched because of RC Forward’s existence, and their best guesses assume 95-100% of donors were planning to give to EA charities anyway. The low baseline fraction of overall philanthropic donations going to EA charities seems to me to have the opposite implication that Dullaghan draws – that is to say, if the baseline rate of donations to EA charities is so low, then chances are in the absence of the EG organization they would have given to non-EA charities. Overall, this seems to me a less reliable estimation method than direct surveys and external reference class. | 0.45 | https://rethinkpriorities.org/publications/cost-effectiveness-of-rc-forward | |||||||||||||||||||||||
8 | Effectiveness adjustment: In Schienle, Schwicker & Schulze (2023), one key difference between their analysis of Effektiv Spenden and GWWC's analysis is that ES treats all donations as having equal impact (i.e. they do not account for potential differences in impact among their recommended charities); this is an important omission, given how substantially even GiveWell charities can differ from each other in cost-effectiveness. | 0.56 | https://docs.google.com/document/d/1hSvK2ZLK7zRXMRq5_RVGJupmc7EQ61M3fzU4Ih55TYI/edit | |||||||||||||||||||||||
9 | Discounted giving lifespan | 8.4 | ||||||||||||||||||||||||
10 | Reporting adjustment | 1.34 | ||||||||||||||||||||||||
11 | Value of a 2025 GWWC pledge in GiveWell dollars: In Townsend & Hoeijmakers (2023), GWWC models the value of a GWWC pledge as a function of a recording coefficient (i.e. the amount we need to multiply recorded donations by to estimate total donations), a counterfactual coefficient (i.e. the fraction of total donations that only happen because of GWWC), an effectiveness coefficient (i.e. the fraction of total donations that go to effective charities), the average giving lifespan of GWWC pledgers, the average annual donation made per GWWC pledger each year, and a discount rate. CEARCH believes that this modelling can be improved upon; in particular, we think that for modelling the GWWC pledge, (a) the recording coefficient should take into account external reference classes; (b) the counterfactual coefficient should take into account external reference classes; (c) the effectiveness coefficient should take into account the relative cost-effectiveness of top and standout charities; (d) the average giving lifespan of GWWC pledgers should be informed by external reference classes; (e) running a regression of dollars donated against pledge batch and year of giving shows that earlier pledgers give more and that pledgers of any given batch given about the same year after year; and (f) GWWC should precisely model the discount rate rather than use the UK Green Book's standard 3.5%, which has a pure time preference element which is philosophically questionable, and which does not take into account other potential considerations relevant to effective giving (e.g. decline in GiveWell's cost-effectiveness over time and inflation,) and in cost-effectiveness more generally (e.g. GCR and uncertainty). Finally, a critical issue that must be addressed is the fact that pledgers who do not report their donation to GWWC may nonetheless be fulfilling their pledge and giving all the same. | 8659 | https://docs.google.com/document/d/1l0RjbMLXjwjy5Pqg4aLe-_YDLXw8T3DT5QP0AQGYHTo/edit & https://docs.google.com/document/d/1EHqCi_ZRdKvsWpB6-V7AOQXj-wDjNFa4BeQVHvsx7Lw/edit | |||||||||||||||||||||||
12 | Expected number of 2025 GWWC pledgers: Uses the annual average increase from 2021 to 2023. We note that GWWC aims to reach approximately this number in 2024 (and is on track to do so). | 1095 | https://docs.google.com/spreadsheets/d/1Yc2VPSmR6zM7Q5k7yWkXxIxsmckAnPHLpNfNaVUN3jE/ & Correspondence with GWWC | |||||||||||||||||||||||
13 | Value of GWWC's 2025 pledge promotion work in GiveWell dollars (Baseline) | 9481114 | ||||||||||||||||||||||||
14 | Adjustment to include largest 10 pledge donors: Baseline estimate based off GWWC data, which is then subject to the following considerations: (a) that these extremely wealthy donors are more likely to be giving counterfactually anyway (n.b. one expert we consulted has a major concern about the GWWC model related to GWWC's failure to model based off a break down by demographic class, such as year of joining, or size of giving, as the more you give, the more likely you are to have been giving anyway); (b) that they are likely more effectiveness oriented; (c) that they are more likely to sustain their giving; and (d) that they are subject to less significant non-reporting bias (given evidence of higher response rates to GWWC's surveys by the top 10 relative to the average donor); overall no net adjustment is made. | 1.22 | https://docs.google.com/spreadsheets/d/1125fOWYq3nv7HxWXNZRTTx6xOdpSqxlU9NKLZ1Igmgk/ | |||||||||||||||||||||||
15 | Value of GWWC's 2025 pledge promotion work in GiveWell dollars (All donors) | 11608833 | ||||||||||||||||||||||||
16 | Trial Pledgers | |||||||||||||||||||||||||
17 | Annual donations per 2025 trial pledger | 1381 | ||||||||||||||||||||||||
18 | Recording adjustment | 0.80 | ||||||||||||||||||||||||
19 | Counterfactual adjustment: Per discussion above | 0.45 | ||||||||||||||||||||||||
20 | Effectiveness adjustment: Per discussion above | 0.51 | ||||||||||||||||||||||||
21 | Discounted giving lifespan | 4.0 | ||||||||||||||||||||||||
22 | Reporting adjustment | 1.10 | ||||||||||||||||||||||||
23 | Value of a 2025 GWWC trial pledge in GiveWell dollars (via trial pledge donations): In Townsend & Hoeijmakers (2023), GWWC models the value of a trial pledge as a function of a recording coefficient, a counterfactual coefficient, an effectiveness coefficient, average total donations made per trial pledger, probability of a trial pledger taking the full pledge, and the value of a GWWC pledge. CEARCH believes that this modelling can be improved upon; in particular, we think that for modelling the GWWC trial pledge, (a) the recording coefficient should take into account external reference classes; (b) the counterfactual coefficient should take into account external reference classes; (c) the effectiveness coefficient should take into account the relative cost-effectiveness of top and standout charities; (d) average total donations made per trial pledge should be modelled as a function of the average giving lifespan of GWWC trial pledgers, the average annual donation made per GWWC trial pledger each year, and a discount rate; and (e) the probability of a trial pledger taking the full pledge and the value of a GWWC pledge need not be taken into account of insofar as these pledges are already included in the aggregate number of pledges when used to calculate pledges obtained per dollar. Finally, a critical issue that must be addressed is the fact that pledgers who do not report their donation to GWWC may nonetheless be fulfilling their pledge and giving all the same. | 1117 | https://docs.google.com/document/d/1l0RjbMLXjwjy5Pqg4aLe-_YDLXw8T3DT5QP0AQGYHTo/edit | |||||||||||||||||||||||
24 | Expected number of 2025 GWWC trial pledgers: Uses the annual average increase from 2021 to 2023 | 670 | https://docs.google.com/spreadsheets/d/1Yc2VPSmR6zM7Q5k7yWkXxIxsmckAnPHLpNfNaVUN3jE/ | |||||||||||||||||||||||
25 | Value of GWWC's 2025 trial pledge promotion work in GiveWell dollars | 748352 | ||||||||||||||||||||||||
26 | Overall | |||||||||||||||||||||||||
27 | Expected benefit of GWWC's 2025 overall pledge work in GiveWell dollars (Baseline): Throughout this analysis, I treat GWWC's 2025 future work as focused on non-EA pledging (as opposed to EA pledges, or non-pledge donations), after having confirmed with multiple GWWC team members that this is the likely future direction for GWWC's: (a) one team member comments that GWWC has always targeted non-EA pledgers and will continue to; (b) another team member affirms that proto-EAs (i.e. non-EAs who can become EAs) will be their main target for the coming few years (maybe 2% of the population would fit that categorization), and that the focus is pledges rather than donations; and (c) a final team member says that GWWC is deprioritizing annual giving; they only care about it as a pipeline to pledging; he says that while the focus will be on EAs in 2024, it has to be non-EAs 2025, as GWWC will exhaust the EA pool quickly, and scaling in the long run requires new audiences; ultimately, he agrees that it makes sense to evaluate GWWC's work on promoting pledging amongst non-EAs. | 12357185 | https://docs.google.com/document/d/1EHqCi_ZRdKvsWpB6-V7AOQXj-wDjNFa4BeQVHvsx7Lw/edit | |||||||||||||||||||||||
28 | Adjustment for prioritization of pledge over donations | 1.10 | https://forum.effectivealtruism.org/posts/a8wijyw45SjwmeLY6/gwwc-is-funding-constrained-and-prefers-broad-base-support | |||||||||||||||||||||||
29 | Adjustment for company pledgers: There are 49 company pledgers compared to 9335 human pledgers as of August 2024, and an adjustment is calculated accordingly, while taking into account the greater income of companies | 1.05 | https://www.givingwhatwecan.org/about-us/members | |||||||||||||||||||||||
30 | Adjustment for post-retirement donations | 1.10 | ||||||||||||||||||||||||
31 | Adjustment for spillover effect on non-pledge donations: Calculated from the ratio of non-pledge to pledge donations in 2020-22, while discounting for future prioritization of pledge over non-pledge work | 1.44 | https://docs.google.com/spreadsheets/d/1MF9bAdISMOMV_aOok9LMyKbxDEpOsvZ9VO8AfwsS6_o/ & https://docs.google.com/document/d/1EHqCi_ZRdKvsWpB6-V7AOQXj-wDjNFa4BeQVHvsx7Lw/edit | |||||||||||||||||||||||
32 | Adjustment for impact on effective giving ecosystem: For example, GWWC might encourage effective giving indirectly via supporting incubation of EG initiatives and hosting the EG summit. | 1.10 | ||||||||||||||||||||||||
33 | Adjustment for talent pipeline | 1.10 | ||||||||||||||||||||||||
34 | Adjustment for GWWC's influence declining over time due to the general growth in EA | 0.9 | ||||||||||||||||||||||||
35 | Adjustment for direct talent cost (i.e. opportunity cost of GWWC team working elsewhere) | 0.9 | ||||||||||||||||||||||||
36 | Adjustment for reduced political donations: There is evidence that charitable donations and political donations are substitutes, and to the extent that EAs are disproportionately liberal and American, this reduces the financial support that the Democrats receive | 0.9 | https://www.nber.org/system/files/working_papers/w26616/w26616.pdf | |||||||||||||||||||||||
37 | Adjustment for diminishing marginal returns in funding: We expect that the typical funding decision will be made in a context where GWWC has already raised - or will likely go on to raise - a meaningful portion of its budget anyway from other donors. Hence, I discount for diminishing marginal returns, based on (a) the assumption that the initial 10% of baseline funding has 10 units of marginal returns, and marginal returns decline by 1 unit for every 10% increase in baseline funding; and (b) the assumption that GWWC's funding situation at the end of 2023 (about 70% of baseline budget expected to be secured) is representative of how much funding will usually be left after institutional and other reliable individual donors have made their contributions or have had their expected contributions factored in. In other words, we are assessing the value specifically of filling the last 30% of GWWC's baseline budget. Note that we have gotten an independent estimate from GWWC as to the likely marginal value of the last 30% of their baseline budget, and they ballpark it at 10%. However, the uncertainty is radical; and both we and GWWC expect better calibration once GWWC trials out different pledge outreach strategies over the next year and measures their returns in terms of pledges secured. | 0.11 | https://forum.effectivealtruism.org/posts/a8wijyw45SjwmeLY6/gwwc-is-funding-constrained-and-prefers-broad-base-support | |||||||||||||||||||||||
38 | Expected benefit of GWWC's 2025 overall pledge work in GiveWell dollars (Fully calibrated) | 2180785 | ||||||||||||||||||||||||
39 | Cost of GWWC's 2025 overall pledge work in GiveWell dollars: Takes GWWC's costs for 2020-2022 and calculates the annual cost. I do not split between GWWC's pledge and non-pledge work, as that is practically impossible - any work to promote pledging (i.e. to give at least 10% of one's income to the organizations that can most effectively use the money to improve the lives of others) or trial pledging (i.e. the same but for >=1% and <10%) will do so by promoting the moral necessity of giving and the logical soundness of doing so effectively, which also promotes non-pledge effective donations, even if one isn't persuaded to pledge; similarly, any work to promote non-pledge effective donations requires promoting giving and effectiveness, which helps promote the pledge as well. Adjustment for marginal cost of remaining 30% of budget, as detailed above. | 209713 | https://docs.google.com/spreadsheets/d/1MF9bAdISMOMV_aOok9LMyKbxDEpOsvZ9VO8AfwsS6_o/edit | |||||||||||||||||||||||
40 | Giving multiplier | 10 | ||||||||||||||||||||||||
41 | ||||||||||||||||||||||||||
42 | ||||||||||||||||||||||||||
43 | ||||||||||||||||||||||||||
44 | ||||||||||||||||||||||||||
45 | ||||||||||||||||||||||||||
46 | ||||||||||||||||||||||||||
47 | ||||||||||||||||||||||||||
48 | ||||||||||||||||||||||||||
49 | ||||||||||||||||||||||||||
50 | ||||||||||||||||||||||||||
51 | ||||||||||||||||||||||||||
52 | ||||||||||||||||||||||||||
53 | ||||||||||||||||||||||||||
54 | ||||||||||||||||||||||||||
55 | ||||||||||||||||||||||||||
56 | ||||||||||||||||||||||||||
57 | ||||||||||||||||||||||||||
58 | ||||||||||||||||||||||||||
59 | ||||||||||||||||||||||||||
60 | ||||||||||||||||||||||||||
61 | ||||||||||||||||||||||||||
62 | ||||||||||||||||||||||||||
63 | ||||||||||||||||||||||||||
64 | ||||||||||||||||||||||||||
65 | ||||||||||||||||||||||||||
66 | ||||||||||||||||||||||||||
67 | ||||||||||||||||||||||||||
68 | ||||||||||||||||||||||||||
69 | ||||||||||||||||||||||||||
70 | ||||||||||||||||||||||||||
71 | ||||||||||||||||||||||||||
72 | ||||||||||||||||||||||||||
73 | ||||||||||||||||||||||||||
74 | ||||||||||||||||||||||||||
75 | ||||||||||||||||||||||||||
76 | ||||||||||||||||||||||||||
77 | ||||||||||||||||||||||||||
78 | ||||||||||||||||||||||||||
79 | ||||||||||||||||||||||||||
80 | ||||||||||||||||||||||||||
81 | ||||||||||||||||||||||||||
82 | ||||||||||||||||||||||||||
83 | ||||||||||||||||||||||||||
84 | ||||||||||||||||||||||||||
85 | ||||||||||||||||||||||||||
86 | ||||||||||||||||||||||||||
87 | ||||||||||||||||||||||||||
88 | ||||||||||||||||||||||||||
89 | ||||||||||||||||||||||||||
90 | ||||||||||||||||||||||||||
91 | ||||||||||||||||||||||||||
92 | ||||||||||||||||||||||||||
93 | ||||||||||||||||||||||||||
94 | ||||||||||||||||||||||||||
95 | ||||||||||||||||||||||||||
96 | ||||||||||||||||||||||||||
97 | ||||||||||||||||||||||||||
98 | ||||||||||||||||||||||||||
99 | ||||||||||||||||||||||||||
100 | ||||||||||||||||||||||||||