What's the best marginal aid dollar?

[Being]

But you said so right here...

In fact, given demographic pressures in much of the developed world (and some middle income countries), it would make sense for those countries to be much more welcoming of migrants and refugees to improve their demographic balance while simultaneously reducing population pressures in the developing world. Perhaps that is a rather more ethical and effective solution rather than 'preventing pregnancies'.

[Steely Glint]

I think the reality is that no matter how hard we try to fix 'root causes', there are always going to be people who, through bad circumstances or a ruinous context (e.g. wars, natural disasters, climate, etc.) are going to need immediate life saving assistance. So, sure, root cause effects are important, but at times it's hard to understand where the line should be drawn between meeting the immediate and urgent needs of people and trying to limit the numbers of those who need such assistance in the future.

As an addendum, I think there's different criteria when there's a specific emergency caused by a natural/manmade disaster.

[wiggin]

Since I posted this thread, I've been doing a lot more reading on this topic. There's a lot of discussion of this in the effective altruism literature (think Peter Singer) as well as the measurement side of things with e.g. RCTs for poverty alleviation from Esther Duflo et al. In that context, there's a few general trends that this approach tends to highlight:

- Measurable endpoints and interventions, with frequent checks on efficacy
- Scalability (will more $$ at the margin be as impactful as the initial $$)
- Focusing on areas of greatest 'suffering' irrespective of the identity of the sufferer

This is the basis for organizations like GiveWell, which claim to have a hard nosed empirical based approach to identifying top performing charities. They currently argue that a single life can be saved with ~$3-5k, and focus on generally unobjectionable charities addressing mainly health issues in the poor world - malaria, parasites, childhood vaccination, nutritional deficiencies, etc. There is no doubt that these are worthy and deserving charities, and that the basic framework described above will yield positive results, but I have to say that the more I read the less I was convinced that it was a defensible approach on its own.

First off, I am not sold that you can fully quantify 'suffering', let alone quantify the effect of an intervention. There are a lot of long term, nebulous, difficult to measure interventions (and endpoints) that don't lend themselves to good RCTs or even basic impact reporting. These interventions might be incredibly impactful, but with such a empirically driven approach they will always get overlooked in favor of more malaria nets. This isn't a free pass for charities to avoid attempts at quantifying their impact, but it does mean we should be cautious about measurement bias (it reminds me of the architectural bias in archeology).

Secondly, I think the effective altruism folks sometimes get too expansive with their agnosticism wrt cause. I am unabashed in broadly prioritizing the welfare of humans over the welfare of animals. A big component of that community spends time focusing on animal welfare based on equivalence arguments about 'suffering' but I am not convinced. On the flip side, you get people who use reductio ad absurdum arguments to say that all of our money should go towards fighting theoretical but large existential risks. In particular, there's a branch of this that is deathly afraid of an unfriendly general AI, and thinks this is the only truly important effort going on today (at the expense of other more prosaic efforts to limit suffering).

Lastly, I take exception to the specific claim that one should not favor local communal needs over those elsewhere in the world (ostensibly because the degree of suffering that can be mitigated per dollar spent is much higher in other parts of the world). I already alluded to this earlier, but I think that there is a unique responsibility that each of us has to our local community in addition to our general responsibility to the global community. A few examples:

1. My wife and I financially support refugee resettlement efforts in our region. While it's true that our dollars would go further towards supporting refugee resettlement efforts in cheaper countries, we choose to support local ones. Why? Well, firstly, we think that resettlement of refugees in the United States is (or at least should be) a central theme of what it is to be American - nearly everyone in the US has come from somewhere else, often fleeing persecution, war, or seeking a better/freer life for themselves and their families. As a country we have frequently failed to meet this ideal, but it's one we want to support; thus, we find it meaningful and important to support a continued (and stronger) welcome for those seeking America's shores. More personally, both of us descend from people just a few generations back who were refugees (either legally or effectively, depending on the era), and we are well aware of how challenging it can be. Lastly, a close colleague and friend of mine is himself a refugee from Iraq who was resettled here with his family while he was in high school. It would be a real disservice to our country if we didn't welcome bright, hardworking people of diverse cultural backgrounds to join in building and improving our society.

2. My wife and I financially support water projects in the Levant. While there are many reasons for the conflicts in that portion of the world, water scarcity is definitely one of them. As most of you no doubt know, my wife and I have family members living in the region (and close cultural ties to one country there), and we think this is one relatively non-partisan and important way to help reduce the chances of a conflict over water. Are there places that need water projects more urgently? Sure! But our specific interest in the stability of this region (and, more broadly, our belief that this region has a sufficiently well developed technology and bureaucratic base to manage large scale water projects) means that we believe we have a responsibility to engage with efforts to enhance regional stability more than improving water access elsewhere.

3. We financially support (and volunteer) with a local charity providing social services, particularly related to food insecurity, to a broad swath of poor people in the region. While our dollars (but not time) could go farther in, say, Ethiopia or another famine-stricken region, we feel that there is an additional responsibility not to ignore the poor who are our own neighbors (literally in my case - at least one of the clients of our organization live two doors down in my building; they're elderly immigrants from the FSU living on limited fixed incomes).

All of these would be big no-nos under the 'rules' of effective altruism. Peter Singer would probably call it immoral! But I think that these community ties matter. Of course, that isn't to say that we shouldn't (or don't) support more global efforts to reduce suffering/fight poverty, generally along the rubrics described by e.g. GiveWell. But we don't prioritize those efforts to the exclusion of meaningful efforts that affect our own communities (however you want to define it). I haven't quite figured out how to square this circle - how to know what's the right mix of communal vs. global involvement, and how to articular an internally consistent moral position to defend it. But it definitely feels more right than the theoretically 'objective' approach taken by this new breed of altruist.

[Loki]

I think the rules you mentioned make more sense in the context of a large philanthropic organization. You should focus on things that make most sense to you. You should give to organizations that are better run and more effective, but that doesn't mean opting for causes you feel no connection to. And if you do give enough to have some leverage, it might be a good idea to use that leverage to get the organization to do a study to make sure there's no room for improvement.

[Steely Glint]

Trying to min-max altruism like you're doing any% speed run of of compassion on twitch dot com is stupid. If you want to min-max charity, it has to be done on the collective level or else you can't answer basic min-maxer questions like 'how many does this need vs how much is already allocated'. If everyone used altruism science to identify the most optimal organisation and cause to donate to in order to maximise their altruism points to dollars spent ratio, then we'd be vastly over-allocating resources to that one cause and neglecting everything else.