I strongly dislike claiming that “there’s a role for small EA donors in campaign finance” in a post which makes essentially no argument for the intervention’s effectiveness.
Maybe there’s a role, but assuming there is seems like wishful thinking. GiveDirectly still has a plausible funding gap in the hundreds of billions or trillions of dollars, so one should at least make the case that it might be better than that.
I agree more nuance in the headline would have been better (eg., if it included the word “potentially” to say “There’s potentially a role for small EA donors in campaign finance”), but note that’s effectively what the body of the piece says, such as here: “consider that election campaign contributions might be a way in which you can have a substantial impact as a small donor” (emphasis added).
If the concern is that donations don’t have any impact on electoral outcomes, there is a good bit of high-qualitysocial science research indicating that television advertising, at least, does, particularly (as OP notes) in down-ballot races. If the concern is that it nonetheless isn’t worth its cost, that’s plausible, but I don’t think OP said anything to suggest strong grounds to believe campaign donations beat GiveWell’s Maximum Impact Fund, nor (I assume) would most readers leap to that conclusion, given the unique depth and rigor of GiveWell’s research process and the far greater difficulty of modeling cost-effectiveness in politics. The thrust of this post seems to be more that this is something worth considering, which seems like a fair assessment, particularly given the extentof preexistingEAactivity in this area (and the reasonable argument that there are decreasing returns to scale).
I think this is a disingenuous motte-and-bailey argument.
The OPs suggestions aren’t to ‘look into whether this might be effective, build some models of cost effectiveness, and compare against existing opportunities’.
They are ‘donate to some of the candidates Elizabeth Edwards-Appell recommends’, ‘form lists of good candidates’, ‘set up an EA funding bloc for candidates’ and ‘devote resources to training EA candidates’.
Answering the question of whether a candidate is “good,” might well (at least on certain EA world views) be sufficient to answer the question of whether donating to the candidate would be (sufficiently) cost-effective (given evidence that 1) donations matter for getting elected, and 2) getting elected allows one to influence policy). Consider the case of a candidate running on a longtermist platform. My impression is that when longtermist grantmakers evaluate giving opportunities in existential risk mitigation, their decision process is much closer to “determine whether the opportunity in question has a reasonable chance of improving humanity’s longterm trajectory within a range of broadly acceptable costs” than to “conduct a thorough, systematic, GiveWell-style cost-effectiveness analysis.” I would think that roughly the same principles that apply to donations to organizations that lobby Congress for better biosecurity policy apply to donations to candidates for Congress who strongly favor better biosecurity policy. This seems to be the thinking behind OP’s post. The back-of-the-envelope intuition here is pretty straightforward; insisting on a GiveWell-style CEA in its place reads like an isolated demand for rigor.
If you can point out where I asked for “a Givewell style CEA” I might agree that it was an isolated demand for rigor.
I didn’t do that, however. Instead, I asked for an attempt to make the case that it could be better than GiveDirectly—I didn’t specify how one might make the case or any level of rigor at all.
What I was imagining was a basic back of the envelope sketch of how this intervention might be cost effective, which I don’t think OP provided.
The supposed motivation for the post was EA having a funding overhang—in that context asking how it compares to another intervention which can potentially absorb near limitless amounts of money without diminishing returns seems totally reasonable to me.
I strongly dislike claiming that “there’s a role for small EA donors in campaign finance” in a post which makes essentially no argument for the intervention’s effectiveness.
Maybe there’s a role, but assuming there is seems like wishful thinking. GiveDirectly still has a plausible funding gap in the hundreds of billions or trillions of dollars, so one should at least make the case that it might be better than that.
I agree more nuance in the headline would have been better (eg., if it included the word “potentially” to say “There’s potentially a role for small EA donors in campaign finance”), but note that’s effectively what the body of the piece says, such as here: “consider that election campaign contributions might be a way in which you can have a substantial impact as a small donor” (emphasis added).
See my reply to HStencil above.
If the concern is that donations don’t have any impact on electoral outcomes, there is a good bit of high-quality social science research indicating that television advertising, at least, does, particularly (as OP notes) in down-ballot races. If the concern is that it nonetheless isn’t worth its cost, that’s plausible, but I don’t think OP said anything to suggest strong grounds to believe campaign donations beat GiveWell’s Maximum Impact Fund, nor (I assume) would most readers leap to that conclusion, given the unique depth and rigor of GiveWell’s research process and the far greater difficulty of modeling cost-effectiveness in politics. The thrust of this post seems to be more that this is something worth considering, which seems like a fair assessment, particularly given the extent of preexisting EA activity in this area (and the reasonable argument that there are decreasing returns to scale).
I think this is a disingenuous motte-and-bailey argument.
The OPs suggestions aren’t to ‘look into whether this might be effective, build some models of cost effectiveness, and compare against existing opportunities’.
They are ‘donate to some of the candidates Elizabeth Edwards-Appell recommends’, ‘form lists of good candidates’, ‘set up an EA funding bloc for candidates’ and ‘devote resources to training EA candidates’.
Answering the question of whether a candidate is “good,” might well (at least on certain EA world views) be sufficient to answer the question of whether donating to the candidate would be (sufficiently) cost-effective (given evidence that 1) donations matter for getting elected, and 2) getting elected allows one to influence policy). Consider the case of a candidate running on a longtermist platform. My impression is that when longtermist grantmakers evaluate giving opportunities in existential risk mitigation, their decision process is much closer to “determine whether the opportunity in question has a reasonable chance of improving humanity’s longterm trajectory within a range of broadly acceptable costs” than to “conduct a thorough, systematic, GiveWell-style cost-effectiveness analysis.” I would think that roughly the same principles that apply to donations to organizations that lobby Congress for better biosecurity policy apply to donations to candidates for Congress who strongly favor better biosecurity policy. This seems to be the thinking behind OP’s post. The back-of-the-envelope intuition here is pretty straightforward; insisting on a GiveWell-style CEA in its place reads like an isolated demand for rigor.
If you can point out where I asked for “a Givewell style CEA” I might agree that it was an isolated demand for rigor.
I didn’t do that, however. Instead, I asked for an attempt to make the case that it could be better than GiveDirectly—I didn’t specify how one might make the case or any level of rigor at all.
What I was imagining was a basic back of the envelope sketch of how this intervention might be cost effective, which I don’t think OP provided.
The supposed motivation for the post was EA having a funding overhang—in that context asking how it compares to another intervention which can potentially absorb near limitless amounts of money without diminishing returns seems totally reasonable to me.