(I’m assuming this open thread is for random EA questions. Yes?)
A non-EA friend said to me today, that she believes/fears that EA is a small echo chambery group, and so its conclusions are not to be trusted? (And was also concerned that the conclusions of GiveWell are not peer reviewed).
My response was that while you can’t eliminate echo chamber possibilities, EAs are very self-reflective and consider critiques carefully, and that GiveWell is very transparent, and makes recommendations based on scientific evidence from universities and other organisations. I also said that the goals and methodology of EA make a lot of sense to me.
I don’t think I did justice to EA there, and I can’t say that my response was very convincing to her. The conversation made me very uneasy—partially because I always feel a bit upset by friends criticising what I hold dear. Perhaps she thinks I have joined a cult. Perhaps I have joined a cult!
I would welcome any comments on how you feel about the smallness and reliability of EA, and whether anyone has some different ways of thinking about it.
(I’m assuming this open thread is for random EA questions. Yes?)
Yes, all posts of the EA newsletter to the Forum also double as open threads.
A non-EA friend said to me today, that she believes/fears that EA is a small echo chambery group, and so its conclusions are not to be trusted? (And was also concerned that the conclusions of GiveWell are not peer reviewed).
It depends how wide you want to cast a net in defining ‘effective altruism’. Effective altruism is now a social movement of several thousand people. This isn’t simply based on how the public Facebook group, arguably the most front-facing presence for EA anywhere, has ~10k members. While there are many community members who are students, or aren’t necessarily donating tons of money right now, or working for an effective organization full-time, by attendance to EA conferences and membership in local EA clubs and chapters worldwide, there are at least a few thousand people who take EA very seriously.
Effective altruism is largely divided among four cause areas: poverty alleviation (in practice, primarily evidence-based global health interventions, and cash-transfers); animal advocacy (in practice, mostly focused on factory farming, with a wide spread of positions from mere animal welfare reform to full-out animal liberation); existential risk mitigation (largely focused on potential future risks from artificial/machine superintelligence); and metacharity (fundraising or raising awareness of effective altruism or its composite causes). From years of personal experience, I can attest there is much informal criticism within and between causes, by a variety of advocates within EA, of different approaches and styles of evaluation within EA.
However, most of this isn’t front-facing, and, from the outside, it more or less appears effective altruism is somewhat monolithic. For the record, it seems the community consensus that, other things being equal, Givewell’s research is considered top-notch. Within a focus on poverty alleviation itself, there isn’t much deviation by other organizations or camps from Givewell’s recommendations.
One of the first EA organizations was Giving What We Can (GWWC), which, among its other roles, acts as a charity evaluator. It uses a different methodology than Givewell, but largely reaches the same conclusions in its charity recommendations.
AidGrade is another charity evaluator which is often associated with effective altruism, and evaluates charities on similar empirical principles to GWWC and Givewell. How they differ is that AidGrade uses statistical meta-analyses, based on its own inclinations, and at the behest of philanthropists with an evidence-based approach, in a way more similar to more in-depth analyses reminiscent of fields like health or development economics. AidGrade works by providing data on various interventions on international aid, but eschews making specific recommendations and prefers to let consumers of their meta-analyses reach their own conclusions.
Much of Givewell’s work is focused on determining which organizations are the most competent and efficient at delivering a given intervention. These interventions are usually determined beforehand, by academic researchers, as being the most cost-effective and some of the best ways to increase quality-adjusted life years (QALYs), disability-adjusted life years (DALYs), or some other metric of well-being. Examples include:
While there are criticisms of effective altruism as a narrow and new phenomenon in philanthropy and doing good in general. However, the foundations for all of Givewell’s recommended charities, and the interventions they pursue, all have a basis in peer review.
(I’m assuming this open thread is for random EA questions. Yes?)
A non-EA friend said to me today, that she believes/fears that EA is a small echo chambery group, and so its conclusions are not to be trusted? (And was also concerned that the conclusions of GiveWell are not peer reviewed).
My response was that while you can’t eliminate echo chamber possibilities, EAs are very self-reflective and consider critiques carefully, and that GiveWell is very transparent, and makes recommendations based on scientific evidence from universities and other organisations. I also said that the goals and methodology of EA make a lot of sense to me.
I don’t think I did justice to EA there, and I can’t say that my response was very convincing to her. The conversation made me very uneasy—partially because I always feel a bit upset by friends criticising what I hold dear. Perhaps she thinks I have joined a cult. Perhaps I have joined a cult!
I would welcome any comments on how you feel about the smallness and reliability of EA, and whether anyone has some different ways of thinking about it.
Yes, all posts of the EA newsletter to the Forum also double as open threads.
It depends how wide you want to cast a net in defining ‘effective altruism’. Effective altruism is now a social movement of several thousand people. This isn’t simply based on how the public Facebook group, arguably the most front-facing presence for EA anywhere, has ~10k members. While there are many community members who are students, or aren’t necessarily donating tons of money right now, or working for an effective organization full-time, by attendance to EA conferences and membership in local EA clubs and chapters worldwide, there are at least a few thousand people who take EA very seriously.
Effective altruism is largely divided among four cause areas: poverty alleviation (in practice, primarily evidence-based global health interventions, and cash-transfers); animal advocacy (in practice, mostly focused on factory farming, with a wide spread of positions from mere animal welfare reform to full-out animal liberation); existential risk mitigation (largely focused on potential future risks from artificial/machine superintelligence); and metacharity (fundraising or raising awareness of effective altruism or its composite causes). From years of personal experience, I can attest there is much informal criticism within and between causes, by a variety of advocates within EA, of different approaches and styles of evaluation within EA.
However, most of this isn’t front-facing, and, from the outside, it more or less appears effective altruism is somewhat monolithic. For the record, it seems the community consensus that, other things being equal, Givewell’s research is considered top-notch. Within a focus on poverty alleviation itself, there isn’t much deviation by other organizations or camps from Givewell’s recommendations.
One of the first EA organizations was Giving What We Can (GWWC), which, among its other roles, acts as a charity evaluator. It uses a different methodology than Givewell, but largely reaches the same conclusions in its charity recommendations.
AidGrade is another charity evaluator which is often associated with effective altruism, and evaluates charities on similar empirical principles to GWWC and Givewell. How they differ is that AidGrade uses statistical meta-analyses, based on its own inclinations, and at the behest of philanthropists with an evidence-based approach, in a way more similar to more in-depth analyses reminiscent of fields like health or development economics. AidGrade works by providing data on various interventions on international aid, but eschews making specific recommendations and prefers to let consumers of their meta-analyses reach their own conclusions.
Much of Givewell’s work is focused on determining which organizations are the most competent and efficient at delivering a given intervention. These interventions are usually determined beforehand, by academic researchers, as being the most cost-effective and some of the best ways to increase quality-adjusted life years (QALYs), disability-adjusted life years (DALYs), or some other metric of well-being. Examples include:
The positive reception of cash transfers by economists, in particular by the quality of experiments conducted by GiveDirectly, Givewell’s #2 top-rated charity, who founded their charity after doing research on their own on what sorts of interventions are considered top-notch.
The review of deworming by the World Bank concluding its an effective intervention, one pursued by two of Givewell’s top-rated charities.
The World Health Organization’s conclusion long-lasting insecticidal nets (LLINs) are the most effective malaria prevention program available, and the one pursued by the Against Malaria Foundation, Givewell’s top-rated charity for several years now.
While there are criticisms of effective altruism as a narrow and new phenomenon in philanthropy and doing good in general. However, the foundations for all of Givewell’s recommended charities, and the interventions they pursue, all have a basis in peer review.
Thanks Evan, that was a very thorough and helpful reply. I am now better equipped to answer that question, and feel less uneasy too.