There’s been a lot of handwringing about people’s obsession with getting the relatively few jobs at the relatively few explicitly EA-branded organizations. The discussions have been interesting, but they tend to miss the essential reason for this phenomenon in my experience: when you’re an EA applicant, EA orgs may like you more than non-EA orgs do. A lot more.
Personally, I never felt much pressure, or even necessarily desire, to work only at explicitly EA organizations. I want to work as an analyst or researcher in an EA or EA-adjacent cause area, but that hardly restricts me to EA organizations! Outside the EA-sphere, there are think tanks, philanthropic foundations, consultancies that work in the public interest, and departments of government, among others, that I would be delighted to work for. Over the past ~year, while looking for my first job out of grad school, I have submitted 32 applications to non-EA organizations, alongside 6 applications to EA orgs.
And how is that working out for me?
Of the 6 applications to EA orgs, 5 got back to me asking for at least one interview or test task, and 4 asked me to do multiple rounds of interviews and/or test tasks. Typically, they expressed enthusiasm about me as an applicant, seemed genuinely sorry to be unable to hire me right then, and encouraged me to try again in the future. This is sweet of them, but it’s a disheartening experience overall.
Meanwhile, of 32 applications to non-EA orgs, 3 got back to me asking for at least one interview or test task, and exactly 1 asked me to do multiple rounds of interviews and/or test tasks.* Typically, I got no response of any kind. Sometimes form rejections drift in, pursuant to applications I submitted months ago. This is disheartening in a totally different way.
(*From my perspective, even this exception proves the rule: the one non-EA application that got me multiple rounds of consideration was when I applied specifically to the Charity Navigator subteam that used to be ImpactMatters; a previous application to Charity Navigator as a whole got no response. But from a more neutral, preregistration-demanding perspective, you should probably ignore this line of argument.)
Obviously, I have limited access to the reasoning behind the hiring decisions here. But for what it’s worth, here’s my personal speculation as to what’s going on:
At this time in my life, my CV consists of (a) academic accomplishment in formal, abstract fields, and (b) some student jobs, a couple of which reflect my longstanding interest in effective altruism. From the perspective of an EA org, this shows a reasonable amount of competence and commitment, enough that they’re happy to toss me a test task and see how I do. But to a mainstream org in a similar cause area, I just seem like kind of a weird fit. Why is someone who studies “category theory” and “formal semantics” applying for this job in policy/development/climate/etc? They don’t have a test task culture, and they do have a stack of candidates whose degrees are in policy/development/climate/etc, so they simply go with one of them.
I’m not saying it’s easy to get a job at an EA org; it’s definitely not, and I haven’t. But for some of us, getting a job anywhere else can feel even harder.
Thank you for writing this. I think this points to an important point/risk/trade-off for people who take an EA path in their careers. EA can be really interdisciplinary, in a way that may not be legible outside EA. This is tricky for career planning.
Once upon a time, I found myself with a bunch of unconditional $25 charity gift cards from an every.org promotion. This seemed like a great opportunity to encourage the people in my life to pick charities to donate to, without the awkwardness of talking directly about how they should spend their own money. So I sent four gift card links and an explanation to a group chat with my four closest friends from college.
The first thing that happened was that one friend expressed enthusiasm, claimed a gift card, and donated it to the Florence Project, an organization that gives legal aid to detained migrants in Arizona. The second thing that happened was that the other three friends said nothing, and never claimed or used the gift cards.
I felt disappointed with this response rate. I mean, why wouldn’t someone want a free $25 for charity?? I know on an intellectual level that I’m much more excited about picking charities to donate to than most people are, but do I really have to internalize that knowledge in order to make accurate predictions? Darn.
I didn’t want to pressure anyone, so I didn’t say any more about it. As time passed and the gift cards drew close to their expiration date, I figured I should just use them myself. But here I ran into an issue. If I just directed this unclaimed money to the organizations that I would have chosen in the first place, wasn’t I kind of retroactively giving myself reason to prefer that my friends not participate? I didn’t want that! At the time I had genuinely wanted my friends to participate, and I didn’t want to ruin that after the fact!
Of course, I also didn’t want to waste $75 of donations for no reason. Or for dubious decision-theoretic reasons that definitely had not been relevant to the actual behavior of any of the actual people involved.
And so, this has been the convoluted story how I came to donate a particular $75 to GiveDirectly. Usually, I don’t donate to GiveDirectly, because I’m on board with GiveWell’s view that donations to AMF and other top charities are something like ten times as cost-effective as unconditional cash transfers. (See e.g. this funding report.) But I’m a big fan of cash benchmarking—the idea that all global development interventions should be compared to the baseline option of just giving the beneficiaries unconditional cash transfers instead. Similarly, it seemed to me that if money was earmarked for charity, and no one stepped in to make a case for anything else, cash was a sensible default.
I think this approach is potentially useful in a much more common situation: the matching campaign. Matching campaigns are popular, but misleading (see e.g. discussion here, here, and here). The misleadingness could be fixed if the matching donor agreed to set fire to any money that wasn’t used to match other donations, but they typically won’t agree to that (and it wouldn’t be a great look). The idea here is, instead of fire, the matching donor picks a “floor” donation option that they’re okay with their money going to, but not maximally excited about. For best results, the audience of potential donors should largely agree with this assessment, and also the floor option should have some kind of ontological basicness to it, such that it feels reasonable and not insulting to designate it as the “floor”. Obviously I’m describing cash transfers here, but I think other interventions with these properties would work for the same reasons; direct air capture comes to mind.
I’d be interested to hear if something like this has been tried.
Interesting suggestion. I’m not familiar with anyone doing a donation match like this.
It seems like having a default charity for matching money to go to could be counterproductive to the matcher’s goals. E.g. Every.org wanted to get more people to use their platform to donate. But I think many people don’t really find it more valuable for money to get directed to one charity over another. EAs are different in that regard. While we’re certainly not unique in caring which charities money goes to, I think many people might think “Why should I donate when the money is already going to go to charity?” and decide not to participate.
While generally I wouldn’t advise people to do donation matches, would it be good for organizations already running them to make cash transfers the default use of the money if matching donors don’t direct it elsewhere? Maybe. One benefit might be that it just gets people to think more about the value of directing money to one organization versus another, instead of merely thinking that they’re raising more money for a charity of their choice.
Why I Apply to EA Orgs
There’s been a lot of handwringing about people’s obsession with getting the relatively few jobs at the relatively few explicitly EA-branded organizations. The discussions have been interesting, but they tend to miss the essential reason for this phenomenon in my experience: when you’re an EA applicant, EA orgs may like you more than non-EA orgs do. A lot more.
Personally, I never felt much pressure, or even necessarily desire, to work only at explicitly EA organizations. I want to work as an analyst or researcher in an EA or EA-adjacent cause area, but that hardly restricts me to EA organizations! Outside the EA-sphere, there are think tanks, philanthropic foundations, consultancies that work in the public interest, and departments of government, among others, that I would be delighted to work for. Over the past ~year, while looking for my first job out of grad school, I have submitted 32 applications to non-EA organizations, alongside 6 applications to EA orgs.
And how is that working out for me?
Of the 6 applications to EA orgs, 5 got back to me asking for at least one interview or test task, and 4 asked me to do multiple rounds of interviews and/or test tasks. Typically, they expressed enthusiasm about me as an applicant, seemed genuinely sorry to be unable to hire me right then, and encouraged me to try again in the future. This is sweet of them, but it’s a disheartening experience overall.
Meanwhile, of 32 applications to non-EA orgs, 3 got back to me asking for at least one interview or test task, and exactly 1 asked me to do multiple rounds of interviews and/or test tasks.* Typically, I got no response of any kind. Sometimes form rejections drift in, pursuant to applications I submitted months ago. This is disheartening in a totally different way.
(*From my perspective, even this exception proves the rule: the one non-EA application that got me multiple rounds of consideration was when I applied specifically to the Charity Navigator subteam that used to be ImpactMatters; a previous application to Charity Navigator as a whole got no response. But from a more neutral, preregistration-demanding perspective, you should probably ignore this line of argument.)
Obviously, I have limited access to the reasoning behind the hiring decisions here. But for what it’s worth, here’s my personal speculation as to what’s going on:
At this time in my life, my CV consists of (a) academic accomplishment in formal, abstract fields, and (b) some student jobs, a couple of which reflect my longstanding interest in effective altruism. From the perspective of an EA org, this shows a reasonable amount of competence and commitment, enough that they’re happy to toss me a test task and see how I do. But to a mainstream org in a similar cause area, I just seem like kind of a weird fit. Why is someone who studies “category theory” and “formal semantics” applying for this job in policy/development/climate/etc? They don’t have a test task culture, and they do have a stack of candidates whose degrees are in policy/development/climate/etc, so they simply go with one of them.
I’m not saying it’s easy to get a job at an EA org; it’s definitely not, and I haven’t. But for some of us, getting a job anywhere else can feel even harder.
Thank you for writing this. I think this points to an important point/risk/trade-off for people who take an EA path in their careers. EA can be really interdisciplinary, in a way that may not be legible outside EA. This is tricky for career planning.
Cash by Default
Once upon a time, I found myself with a bunch of unconditional $25 charity gift cards from an every.org promotion. This seemed like a great opportunity to encourage the people in my life to pick charities to donate to, without the awkwardness of talking directly about how they should spend their own money. So I sent four gift card links and an explanation to a group chat with my four closest friends from college.
The first thing that happened was that one friend expressed enthusiasm, claimed a gift card, and donated it to the Florence Project, an organization that gives legal aid to detained migrants in Arizona. The second thing that happened was that the other three friends said nothing, and never claimed or used the gift cards.
I felt disappointed with this response rate. I mean, why wouldn’t someone want a free $25 for charity?? I know on an intellectual level that I’m much more excited about picking charities to donate to than most people are, but do I really have to internalize that knowledge in order to make accurate predictions? Darn.
I didn’t want to pressure anyone, so I didn’t say any more about it. As time passed and the gift cards drew close to their expiration date, I figured I should just use them myself. But here I ran into an issue. If I just directed this unclaimed money to the organizations that I would have chosen in the first place, wasn’t I kind of retroactively giving myself reason to prefer that my friends not participate? I didn’t want that! At the time I had genuinely wanted my friends to participate, and I didn’t want to ruin that after the fact!
Of course, I also didn’t want to waste $75 of donations for no reason. Or for dubious decision-theoretic reasons that definitely had not been relevant to the actual behavior of any of the actual people involved.
And so, this has been the convoluted story how I came to donate a particular $75 to GiveDirectly. Usually, I don’t donate to GiveDirectly, because I’m on board with GiveWell’s view that donations to AMF and other top charities are something like ten times as cost-effective as unconditional cash transfers. (See e.g. this funding report.) But I’m a big fan of cash benchmarking—the idea that all global development interventions should be compared to the baseline option of just giving the beneficiaries unconditional cash transfers instead. Similarly, it seemed to me that if money was earmarked for charity, and no one stepped in to make a case for anything else, cash was a sensible default.
I think this approach is potentially useful in a much more common situation: the matching campaign. Matching campaigns are popular, but misleading (see e.g. discussion here, here, and here). The misleadingness could be fixed if the matching donor agreed to set fire to any money that wasn’t used to match other donations, but they typically won’t agree to that (and it wouldn’t be a great look). The idea here is, instead of fire, the matching donor picks a “floor” donation option that they’re okay with their money going to, but not maximally excited about. For best results, the audience of potential donors should largely agree with this assessment, and also the floor option should have some kind of ontological basicness to it, such that it feels reasonable and not insulting to designate it as the “floor”. Obviously I’m describing cash transfers here, but I think other interventions with these properties would work for the same reasons; direct air capture comes to mind.
I’d be interested to hear if something like this has been tried.
Interesting suggestion. I’m not familiar with anyone doing a donation match like this.
It seems like having a default charity for matching money to go to could be counterproductive to the matcher’s goals. E.g. Every.org wanted to get more people to use their platform to donate. But I think many people don’t really find it more valuable for money to get directed to one charity over another. EAs are different in that regard. While we’re certainly not unique in caring which charities money goes to, I think many people might think “Why should I donate when the money is already going to go to charity?” and decide not to participate.
While generally I wouldn’t advise people to do donation matches, would it be good for organizations already running them to make cash transfers the default use of the money if matching donors don’t direct it elsewhere? Maybe. One benefit might be that it just gets people to think more about the value of directing money to one organization versus another, instead of merely thinking that they’re raising more money for a charity of their choice.