At the framework level
The basic framework you envision seems to have a certain fixed level of charitable resources (potential "sacrifice"), to be stewarded and spent in a utility-optimizing way over time. Once spent, these resources are gone. In this framework, asking "is the amount I would have to donate to the best charity to accomplish the same benefits as donating a kidney less than the price for which I would sell my kidney?" is a coherent and useful question, because you plan to offset your charitable contributions by that amount if you decide to donate a kidney, in order to maintain a stable level of personal welfare.
This sense of a broad scarcity of "charitable resources" (meant to include money, time, organs, etc.) is well-expressed by Julia's comment:
People think about doing their good deed for the day. I bet if you asked someone who'd just undergone major surgery to save someone's life if they now wanted to donate money as well, they'd say, "Look, I've sacrificed enough."
My framework is different. I think of moral sacrifice as something closer to a muscle: it can be used up, but it can also be worked out, flexed, strengthened. I think of donating a kidney as something more like a workout that will train the moral muscle than like a marathon that will exhaust it. If you think that donating a kidney would exhaust your moral capacity, I agree that you probably shouldn't do it.
Julia pointed to (a LessWrong writeup of) a study that purports to show that people behave less ethically after purchasing "green" products because they have "licensed" some immoral behavior with their positive actions. I think this dynamic makes a lot of sense with my experience in small actions but, big, life-changing decisions like donating a kidney or giving away a large portion of your income seem more likely to have the opposite effect: reinforcing an identity as a "charitable person" who is committed to the welfare of others. A couple weeks after I donated my kidney, I made my annual charitable donation; I'm nowhere near you and Julia, but I tithe. I don't think this is especially idiosyncratic: there's a psychology literature on "moral identity" that shows more or less the exact opposite of the literature on "moral licensing."
Of course, in these psych studies, the sample sizes are small and the conclusions are based on sample averages. The psych studies showing that any mention of statistics or more than one victim reduces donations clearly aren't characterizing GiveWell donors, and in the same vein I don't see much reason to think that these studies, even the one I cited, are characterizing people like me. So even if the "moral licensing" phenomenon is the dominant one in a broad swath of social activity or individuals, I think that gives me little reason to believe it dominates in my case or yours.
Details
Does the $100K account for the costs of the transplant operation and followup costs such as anti-rejection drugs?
Yep.
There are also economic benefits of fewer people getting malaria. I don't know the relative magnitudes of these.
Agreed, though the far greater productivity of workers in the developed world (and the social support that they draw on when not working) means that the dollar magnitude of those economic gains will be much higher. Conversely, the higher baseline also means that the welfare returns to the developed-world economic gains are lower than the developing world returns for similar dollar values.
Because I estimated DALYs just from an overall $/life-saved and an estimate at life expectancy then we the question here is whether I'm right to treat one life saved via a malarial net as four times more valuable than ten additional years via a kidney donation. We don't have to bring DALYs into it.
Agree that we don't literally have to use DALYs, but I don't think this is responsive. You're saying that you value saving the child's life at 4x the amount of extending the adult's life by 10 years; my point is that I value additional life for mid-life people much more than I do for infants, so I don't share your value here. I'm not sure what my multiple is, but it tilts the other direction (i.e. for your value of 4x, I have a value of less than one).
As far as I can tell, I care about happiness and avoiding suffering, where two years of at joy-minus-suffering level X matter twice as much as one year at level X.
How do you deal with the repugnant conclusion?
But neither the AMF's nor kidney donation's benefit comes much from an effect on newborns, so that doesn't apply here.
AMF's benefits don't come from newborns, but the $1600/life saved estimate assumes that the entire benefits of bednets come from protecting children under 5.
That's unfortunate. I would really like it if GiveWell could give me $/life-saved that corrected for these issues and was as accurate enough to take literally. How far off do you think this number is? 2x? 100x?
I don't know the right answer to this question. I agree with what Holden wrote about why what you're asking for is too hard to do in a reasonable way.
Isn't medical leave generally unpaid, especially for elective surgeries? If you miss two weeks of work and take home $50K/year then that's about $2K in missed pay.
I don't know of the general case here. I know a guy who got paid medical leave from his law firm, and I don't explicitly know of any cases where people had to take unpaid leave. The federal government offers 30 days of paid leave for organ donation for federal employees, as do many state governments (PDF). I would be mildly surprised if Google wasn't willing to give paid leave for altruistic kidney donation. My general guess is that this is going to track the overall generosity of the benefits of your employer. I do agree that this consideration could easily be decisive.
"Because of the very general framing of "sacrifice," donating a kidney--if it were "worth doing"--would imply that you thereby reduce your charitable obligations." Yes. I would do this.
I tried to address this in the framework section, above. You've shown that you can live happily and healthily on the amount of income that you and Julia save for your own consumption. I don't really understand why would you decide to keep more money in response to another, totally non-fiscal decision.
To put it in a different context, let's say that your job becomes far more pleasant because you got new coworkers who you liked more. Should you thereby give a higher portion of your salary to charity, to offset the increase in pleasure? If something negative happened in your personal life, would you begin withholding a higher portion of your salary for personal consumption? I have a feeling that I'm probably not being fair with these examples, but I'm genuinely not sure why.
Anyway, as much as a I don't understand this approach, I do agree that the bar for donating a kidney should be high if it leads you to reduce your fiscal charitable contributions. If you would reduce your charitable contributions by $10K, I'm not sure if it's worth donating a kidney; if you would reduce your contributions by $50K, I'm pretty sure it's not (with the exception of the hard-to-quantify outreach benefits).
"Envisioning people as having a fixed amount of suffering they're willing to undergo does not strike me as a very accurate psychological model." It doesn't? As long as you account for things like "giving up meat is hard at first but gets easier" it seems about right to me.
I addressed this in the framework section, but I wanted to add that I think "giving up meat is hard at first, but gets easier" dynamics are pervasive. I think giving a big portion of your income is hard at first but probably gets easier over time. If you weren't giving at all, giving 50% of one's income might seem inconceivable; after giving 25% for a long time, it might seem far more doable (even though the next 25% is harder). Most of the studies I've read on hedonic adaptation suggest that humans are quite good at adapting (over time) to new circumstances and new forms of suffering.
Not sure what to say about this. I don't talk about it much with strangers. My friends and family seemed to think that donating a kidney was a bigger deal than giving away or foregoing a lot of income; I don't really agree. Broadly, I think that donating a kidney has convinced people I have more "skin in the game" than just working for GiveWell has, and I can think of one or two cases where this has made them more likely to take my advice on charitable donations. I think that if I were more comfortable talking about organ donation, it would be possible to use that as an entree into talking about charity: "... but you don't have to donate a kidney to make a huge difference! Just donating money can also be transformative."
On the other hand, I think it's at least possible that the net effect could be harmful. Kidney donation is still widely perceived (incorrectly, I've argued) as a big deal, so if you're trying to convince people that donating a portion of their income is easy, it could be a disadvantage.
Just to be clear, I think that in jkaufman's argument, the $400 is not the opportunity cost of undergoing surgery, but rather an estimate of the amount of money AMF would require to achieve comparable benefit as donating a kidney.
I agree that if you had to forego income during surgery or incur other costs, then it would be reasonable to add those onto the other side of the ledger as you propose. As I mentioned above, though, I think that's typically not the case; medical leave policies are often quite generous in these cases, and public funds are available for incurred costs.
A couple disclosures right off the bat: I donated a kidney to a stranger last year and I work for GiveWell (but this post represents my views, not GiveWell's).
I think this is a really interesting and informative analysis (you might also be interested in David Barry's post along the same lines). However, I think these both miss a few important points:
Living kidney donation operations involve significant savings for the health system, on the order of $100,000 (coming from averting the need for dialysis, which is extremely costly) (Matas and Schnitzler 2003). Giving $100K to medicare or a private insurer is obviously not the best conceivable use of money, but it's hardly a 100% waste. Similarly, most people on dialysis are not able to work, but most who receive transplants are able to do so; the economic benefits from recipients returning to work are not included in the $100K figure.
I don't think it's the right decision to assume away chains. For most altruistic donors in the U.S., chains are possible, and if you opt against a chain, you can choose a recipient who is demographically likely to achieve greater than average benefit from a donated kidney (someone young and healthy) through a website like matchingdonors.com. I've talked to some fairly senior transplant surgeons and people who helped write the algorithms for matching kidney exchanges, and they don't think chains have been saturated in, e.g., the National Kidney Registry, so adding another altruistic donor is expected to increase the number of kidneys passed along. That said, the right multiplier to use is unclear; the average chain length has been declining over time, giving some credence to the worry that more donors are just shortening the length of the average chain. My personal guess is that the chain multiplier is something like 2x; that's meant to be conservative, if haphazard, but it certainly could turn out to be too aggressive. That is significantly less than the average chain length: I've never heard of a chain that included fewer than 3 kidneys; mine had 8.
I think DALYs are the wrong ethical framework and that adult lives matter more than the DALY framework makes them seem. Undiscounted DALYs assume that saving a newborn is twice as good as saving someone midway through their life, whereas I believe something roughly like the converse (see this discussion at the GiveWell blog ).
I don't think the AMF cost-effectiveness figures and the kidney-donation figures are comparable in the way that you're reporting. Holden's posts about Bayesian charity evaluation give some reasons for my view here, but the basic story is that I don't think the expected value of a $1600 donation to AMF is actually anywhere near one life saved. The reason for this has nothing to do with how AMF works and is more a feature of its place in the total distribution of charity cost-effectiveness. I think there are a variety of practices in cost-effectiveness estimation that push in favor of a difficult-to-estimate positive bias (e.g. using evidence from RCTs, which are generally conducted in the most promising circumstances), that the most extreme cost-effectiveness estimates are more likely to be biased, and that the benefit of a marginal contribution is almost always less than the benefit of an average contribution. All of these conspire to make me think that the estimate that GiveWell provides for the "cost-per-life saved" for AMF is not the correct number for estimating the expected value of a contribution to AMF. (I do think it is extremely useful for deciding between charities. I just don't think these numbers mean what they appear to mean. I haven't discussed this thoroughly with other GiveWell staff and don't know whether they agree; GiveWell hasn't published a clear view about how these cost-effectiveness estimates should be used.)
As a practical matter, most people I've heard of who have been altruistic donors have been able to take medical leave of some kind from their job (many states have a law requiring this), which does't entail using up vacation. This doesn't answer the point about being willing to pay $400 to avoid the suffering, but it does answer the "just work and donate the money" objection.
Separately from the fact that I don't think $400 gets as much benefit as you seem to, I don't think the idea that you would be willing to pay $400 to avoid a nephrectomy is a good reason not to donate a kidney, at least without the additional stipulation that you're not willing to donate $400 more. Imagine a case where an individual believes that donating a kidney is worthwhile for benefits to others, but that they would happily pay $400 to not undergo the surgery (this may be your case, I think it would be mine, too). He pays the $400, and then face the question again. He still has both kidneys, can still benefit someone else by the same amount as before, but he has $400 less, so then he decides again whether to donate a kidney or donate $400. (...induction...) Eventually, he either donates a kidney or decides that enough has been asked of him. Another, perhaps more straightforward, way to put this is to say that the fact that there's some amount of money that achieves equivalent benefit to donating a kidney only tells you not to donate a kidney if you're sure you're not willing to give more money; the implication could just as easily be that you need to give a lot more money. I'm actually not sure what the amount of money (i.e. the $400) has to do with it in your framework, since for any given amount of money the repeated question means that you eventually reach your limit in willingness to pay to avoid surgery.
The model I have of what you're trying to say here is something along these lines:
As rational altruists, we can achieve great outcomes for others at relatively little harm to ourselves. In order to maximize the return to others for the sacrifices we're willing to accept, we should sort potential actions by their expected return and work our way down the list until we've used up our willingness to sacrifice.
In this framework, if you thought the return to donating a kidney was high enough that it fell above your marginal charitable contribution on your ranked list, it would offset that contribution below the limit for sacrifice you're willing to endure. Because of the very general framing of "sacrifice," donating a kidney--if it were "worth doing"--would imply that you thereby reduce your charitable obligations.
I find this an unlikely conclusion, and I don't share your very general conception of "sacrifice." I have a policy on how much money I give to charity, and I decided to donate a kidney, but both decisions depended on the specific circumstances of what would be asked of me. Envisioning people as having a fixed amount of suffering they're willing to undergo does not strike me as a very accurate psychological model, and living on the margin of the maximum suffering you're willing to endure to benefit others doesn't strike me as a good life choice (or at least self-conception).
A final consideration is that I think we relate differently to our bodies than we do our cash. I think that donating a kidney is fundamentally different from donating money in terms of impact on a donor, and that may or may not be something that matters to you. Relatedly, donating a kidney may also have different implications for a professional giving advocate than simply giving more money does. I suspect that it opens up publicity avenues that wouldn't otherwise be available, and changes the tenor of conversations with people who might be more typically emotional in their giving.
I tried to respond to this issue above: http://lesswrong.com/lw/d4v/altruistic_kidney_donation/6w9j