Geographic diversity in EA

post by AmAristizábal · 2020-06-15T06:13:40.968Z · score: 122 (77 votes) · EA · GW · 19 comments

Hi, I´m from Colombia (first post) and I want to share my thoughts on lack of geographic diversity in EA.

I suspect that due to lack of diversity, questions that could be relevant to EA have not been considered enough and here I share some of the ones that I deal with the most (although I don't have a strong position about most of these things and probably I just have not been aware if they DO HAVE been considered, in that case I would appreciate a lot if you could send links or recommendations):

-Whether giving locally could be better (or not) for donors in low and middle income countries:

Countries with weak currencies such as mine face high exchange rates (especially in hard times such as this pandemic). I have the intuition that with a volatile dollar price it doesn't always make sense to donate to EA recommended charities and perhaps donors could allocate better their donations by donating locally. In my case I just switch to save and donate later (because I'm young and my salary is low haha) but what if I still want to donate a little bit to keep motivation? Or what if I want to convince my friend's uncle to donate?I still want to have an informed opinion.

-Spot regional differences within countries when answering different types of questions: Even if my country's GDP is higher than many countries where effective donations according to EA are allocated, there are many regions within my country where poverty is extremely high, even higher than in richer cities from poorer countries. Those differences are hard to spot if EA spots “poverty” as a whole without zooming in geographical zones.

-Addressing the real potential of going into policy in LMICs: EA recommends policy careers but I suspect that it's an even more important path in LMICs, where policies are weaker, policymakers are even less evidence based and where institutions have a lot more potential to improve.

-Whether there is a chance to adapt EA to other cultural values:

Individualism vs collectivism: I feel that EA was born in cultures that value individualistic goals (even if the focus is on the world as a whole). For example, I see EA deeply linked to “western”´s understanding of freedom, independence of thought, skepticism, mistrust for authority and social norms, etc… However, other cultures with more collectivist mindsets can struggle to link altruism to those specific values. In many cultures altruism is deeply linked to religion or family bonds and giving is prioritized when you help those that surround you. Even if there is no rational argument to value more a life in my country vs a life in sub-saharan Africa, what if EA is losing an opportunity to take advantage of these cultural drives towards giving by, for example, strengthening local networks of charities.

Nationalism: Even if I'm not fond of nationalism I do recognize it as a huge drive for altruism in my country (probably in many others as well). I won't convince my friend's uncle to donate to Against Malaria but I could convince him to donate to a colombian charity. Could we use those emotional bonds to promote doing good in an effective way at the same time?

-I wonder if there is a bias when EA talks about problems not being “neglected” enough when dismissing some cause areas or focus topics: an example that comes to my mind is gender inequality in governments or in the workplace. In EA there is a whole focus area on improving institutional decision making, which is great (actually there is where I want to focus); but what if there are easier and more urgent steps to be taken towards IIDM in LMICs such as focusing on women's access to governments (something that in high income countries is not that neglected and has been widely addressed, or at least a lot more than in other countries). So would donating to organizations that are promoting women participation in governments in LMICS be a good cause to donate, taking into account that institutions in these countries face a huge problem that comes even before cognitive biases or poor decision making processes and it's that there is not enough representation of 50% of the population?

I would love to know what has been said about these topics and feel free to reach out.

Update: Due to this post I addressed the topic during the EA Unconference, here is the link to the video:https://www.youtube.com/watch?v=byWYr2oH1y8

19 comments

Comments sorted by top scores.

comment by clairey · 2020-06-15T13:45:01.064Z · score: 34 (19 votes) · EA(p) · GW(p)

Thank you for writing this post! These are really good points. I think EA Singapore is looking to do more research on poverty and development in Southeast Asia, which can unlock more funding from funders who want to donate to a particular country or region. But I'm not the best person to speak about this.

I also thought of another point: if you're outside of the US and Western Europe,[1] it's just harder to stay involved in EA as a human being. This feels obvious to me and I almost didn't post this comment because it seemed like 'duh', but I remembered that I've found simple, clear explanations of 'obvious' things helpful in the past, and at least they reminded me to consider them.

  • EA in general: unlike some other communities or ideologies, being involved in EA requires a good amount of effort. You're expected to take action -- donating, planning your career -- and stay updated on current thinking, which changes very quickly -- earn to give vs. direct impact? chicken eggs...no wait, fish! no wait, insects![2]
  • But it's not easy to visit or live in an EA hub city like London or San Francisco, for most of the global population (financially, legally, for family reasons)
  • In less rich countries, professional altruism & charity aren't as well-regarded and it can be difficult to get support from your family or friends
  • Fewer like-minded people around you means you have to put in a lot more effort to stay engaged and informed
  • Donating 10% (or any %) is more difficult. Your salary isn't as high. Maybe you're already giving 10% to your parents. Maybe you've grown up in a strong culture of saving. And you're very aware that the 10% of your salary is the equivalent of 1% or 0.1% for someone in a richer country
  • You have to create your own opportunities, e.g. organise your own EA group, find impactful policy jobs, set up your own charity, get people to translate EA materials from English [EA · GW]
  • On top of all this, you probably feel consistently excluded from narratives, like your post explains

There are some good posts on the EA forum about value drift, but a bunch of the actions that people can take are less accessible for people outside of the US and Western Europe.

A Qualitative Analysis of Value Drift in EA [EA · GW]

Concrete Ways to Reduce Risks of Value Drift and Lifestyle Drift [EA · GW]

At some level, this is basically 'how to set up EA communities in other countries and the challenges in this'. But I think there is value in considering people's varying circumstances and the effort needed to be 'EA', even if the effect is just that some people feel a little more encouraged and validated.


  1. Not sure about the exact geographies. It's a spectrum (e.g. Australia is somewhere in between) and there's variation within countries as well. ↩︎

  2. Just to be clear, I think taking action and changing your mind frequently based on the best available evidence are good things. ↩︎

comment by AmAristizábal · 2020-06-16T00:36:13.250Z · score: 4 (4 votes) · EA(p) · GW(p)

Thanks for this reply, I agree with everything you say and those are very good points, it´s definitely harder to be involved. Thanks for pointing out the example from Singapore, I will check it out.

Another issue that comes to my mind is the discussion around AI safety, it would be great to have more visions from low and middle income countries since a very important aspect of AI safety includes how AI is unequally distributed in the world and within countries (in some countries we are already seeing how AI can magnify existing inequalities, who knows how that will turn out in the future...).

Thanks again and I´m making a list of all of these issues to keep them in mind :)

comment by vaidehi_agarwalla · 2020-06-16T02:58:23.055Z · score: 23 (13 votes) · EA(p) · GW(p)

Agreed with a lot of the points you've made here. I really liked the point about different geographic regions where there may be extremely high poverty. Some of the poorest Indian states have population sizes and poverty levels comparable or worse than that of some of the poorest African countries. Of course, there are many EA organisations in India, but there may be similar regions in other countries where EA is less active.

One of the potential challenges of doing projects in different countries is a lack of EA resources and networks in those regions, so initial costs may be high. Also, entering policy roles could be challenging if you're the only one promoting an EA-type mindset, and it could lead to burnout or frustration if you're unable to make change

One concrete idea to try to reduce some of these uncertainties could be to research opportunities in a country, which could both help you determine the best opportunities in your country, but also help others in a similar spot.

  • Doing high-level cause prioritisation on how your country may have a comparative advantage making traction on a particular cause area. Some thoughts on an MVP for cause area research.
  • Conducting expert interviews on those promising cause areas to understand the ways you could contribute to this cause area. These could include people working in industry, policy, or NGOs. Questions you could ask experts.
  • Reading up existing research on your country, if any exists. Organisations like JPAL and CGD have done research on the Global Health & Development/IIDM side for a number of countries.

The biggest challenge with this type of reserach is evaluating the impact such roles have, but this is broadly true for many EA roles.

comment by RomeoStevens · 2020-06-26T23:44:00.273Z · score: 5 (2 votes) · EA(p) · GW(p)

Great points. I feel like there's a rule of thumb somewhere in here like 'marginal dollars tend to be low information dollars' that feels helpful.

comment by AmAristizábal · 2020-06-18T04:58:31.190Z · score: 3 (3 votes) · EA(p) · GW(p)

Hi Vaidehi, your reply got me thinking and I have a couple of questions (I´ll put it here in case somebody else has an answer). So... for people in middle or low income countries that want to pick a career to do the most good, should we focus on global problems or is it a good idea to find country-level or regional-level focus areas in terms of their importance, neglectedness and tractability? I suspect that there are some considerations such as our opportunity to reach higher-level positions in our countries vs somewhere else. For example, I imagine that if someone from Brazil wants to improve institutional decision making it would be easier for him/her to improve Brazil's executive branch than finding a path where he/she could improve US institutions (even if US politics could affect more people due to its higher influence).

Do you think that 1. zooming in to country-level or region-level cause areas could help EA identify more accurately priority paths worldwide? and 2.do you think more impact could be achieved if people from different countries find their comparative advantage to focus on their own regions?

comment by BrianTan · 2020-06-18T12:44:45.184Z · score: 10 (7 votes) · EA(p) · GW(p)

Hello, I'm Brian, co-founder of EA Philippines. I think it's a good idea for EAs from non-US / UK countries to do some cause prioritization and career advice research to help determine what are the high-impact career paths for people from your country. And for people doing this, I think it's good to list both global and more regional/local problems, as well as find local roles to solve these local/global problems.

So yes, I think zooming in to country-level or region-level cause areas could help EA identify more accurately priority paths worldwide. I also think more impact could be achieved if people from different countries find their comparative advantage to focus on their own regions.

EA Philippines' cause prioritization and career advice research is currently housed in this Google Sheet. It's still very much a work in progress, and I'm unsure of whether to show this publicly at this early stage, but I think even just the framework and structure of the Google Sheet would help people in figuring out how to get started on this work themselves and how to list various causes, roles, and organizations.

comment by AmAristizábal · 2020-06-20T06:37:53.770Z · score: 4 (2 votes) · EA(p) · GW(p)

Thanks a lot Brian! I´m impressed by EA Philippines haha your achievements are great and I love to see how this is being done in different countries. I wonder if global priorities would be different (and how) if many countries did something like this. Good luck with that research but so far it looks very interesting and replicable in other countries haha thanks a lot.

comment by dpiepgrass · 2020-08-21T17:26:40.359Z · score: 2 (2 votes) · EA(p) · GW(p)

I'm thinking that it would be relatively smarter for EAs in low-income countries to work in local nonprofits, compared to those in high-income countries who are relatively more effective by earning-to-give. Does that sound right to you?

However this does require that a suitable nonprofit job be available in your country! I just checked the 80000 Hours job board and found that the total number of jobs in the "biggest impact" category in Low-Mid Income Countries was 15, versus 336 jobs in the (less populated!) first world. It could well be that there are fewer EAs in LMICs, but probably not 22 times fewer.

comment by vaidehi_agarwalla · 2020-06-18T16:41:51.927Z · score: 6 (3 votes) · EA(p) · GW(p)

Agreed with much of Brian's comment, and I've also PMed you to discuss more. (If others are interested, please reach out!)

comment by AmAristizábal · 2020-06-16T19:26:25.278Z · score: 3 (3 votes) · EA(p) · GW(p)

Wow! Thanks a lot Vaidehi, this is so helpful. The prioritisation part looks like a great starting point. I´ll check it out carefully and think more about this. Probably discussing this with people from similar countries could bring up great new ideas and cause areas where we could have that comparative advantage. Thanks for those links and for your insightful reply!

comment by Ramiro · 2020-06-16T14:24:09.956Z · score: 12 (7 votes) · EA(p) · GW(p)

Thanks for the post. My comments:

I have the intuition that with a volatile dollar price it doesn't always make sense to donate to EA recommended charities and perhaps donors could allocate better their donations by donating locally

1. Actually, if you're from a poor country and use the current TLYCS calculator, you likely have to be rich for them to recommend you to donate a significant portion of your income.

2. I have mixed intuitions here, maybe someone could better disentangle them: a) if my currency vs. U$dollar exchange rate goes from 1:2 to 1:4, my donations apparently lose half of their value; b) however, if this movement is global (because exchange rates markets overvalue U$ dollar, due to the uncertainties caused by the pandemic), then probably the currency in the countries receiving aid will drop, too - so, on average, everything remains the same; c) due to recession, people donate less, thus saving money to donate later may have a cyclical effect.

EA recommends policy careers but I suspect that it's an even more important path in LMICs, where policies are weaker, policymakers are even less evidence based and where institutions have a lot more potential to improve.

I totally agree with that. But LMICS have their own peculiarities and serious governance issues; for instance, I haven't found 80kh advice on public policy that is applicable to someone beggining a civil service career in Brazil. It'd be probably impactful to find organizations with more local expertise.

I won't convince my friend's uncle to donate to Against Malaria but I could convince him to donate to a colombian charity

I don't know how much it scales, but in Brazil, Doebem offers to tranfer donations to GiveWell charities (AMF, GD and SCI), and also to Brazilian charities recognizedly transparent and that have had their impact previously evaluated by international researchers (though not with the same rigor of GW). Besides, they have experimented with direct transfers during the pandemic.

On the other hand, in LMICS, I think many people are often suspicious of local charities they don't have direct contact with, and might be more trustful of foreign recognized charities - with established reputations and rigorous evaluation. For example, when I talk about GD, people usually say "great idea"; but when I mention doedireto, I face all kinds of questions: "how can you ensure the money gets to the right person? or that they won't spend in drinks? etc." This is not unjustified, considering the bad rep the charity sector may have in some circles.

I wonder if there is a bias when EA talks about problems not being “neglected” enough when dismissing some cause areas or focus topics

1. I think "neglectedness" is actually a proxy to assess the expected marginal impact of and additional contribution to a cause - . So, it might not be applicable to causes advocating for systemic change, where you should perform some sort of tipping point analysis instead. On the other hand, the true problem here is: how do you evaluate charities / projects aiming for systemic change?

2. This might lead to a selection bias - we'll end up focusing on projects that might be easier to evaluate; this is often compared to that joke where an economist searches for her keys under the lightpost because that's the only place she can see. I think most people working with charity evaluation in EA are aware of that; on the other hand, requiring no evidence would likely lead to bad incentives, and you still need some evidence to assess the opportunity costs of a project.

3. I actually think improving women participation in LMIC governments (and leading positions in general) would be a good cause precisely because (epistemic status: guess based on anecdotal experiences and some light readings on organizations and management) it would improve institutional decision-making (besides, of course, mitigating discrimination). It would be interesting to see a more profound assessment of this area.

comment by dpiepgrass · 2020-08-21T17:40:56.806Z · score: 1 (1 votes) · EA(p) · GW(p)

I think the calculator you mentioned is kinda... broken. I notice that the local cost of living is ignored and no recommendation is given for incomes under $40,000 USD (or rather the recommendation is "we recommend giving whatever you feel you can afford without undue hardship"). A "well-paying" job in a LMIC is usually below $40,000/year. My highest gross income ever was about $100,000 CAD, and for this they recommend a 1% donation. Nah, I'll stick with 10%+ thanks. You have to make over $83,000 USD for the recommendation to inch past 1%.

comment by Ramiro · 2020-06-16T12:49:36.177Z · score: 11 (9 votes) · EA(p) · GW(p)

I wonder what you'd think about having a network connecting South American EAs. In Brazil, we have considered many of the matters you are now posing; I believe it would be, overall, mutually beneficial, possibly even fun.

comment by AmAristizábal · 2020-06-16T14:45:27.070Z · score: 6 (4 votes) · EA(p) · GW(p)

I´m in, sounds fun! Thanks for all of your comments, definitely they add important aspects that I had not considered. Let´s discuss them further and thanks for reaching out :)

comment by DavidNash · 2020-06-16T14:34:48.647Z · score: 9 (7 votes) · EA(p) · GW(p)

There is a group here that it might be worth posting this to - Altruismo Eficaz América Latina

comment by AmAristizábal · 2020-06-17T14:02:54.314Z · score: 1 (1 votes) · EA(p) · GW(p)

Thanks David!

comment by AmAristizábal · 2020-06-21T17:24:19.777Z · score: 1 (1 votes) · EA(p) · GW(p)

Due to this post I addressed this topic during the EA Unconference, here is the link to the video: https://www.youtube.com/watch?v=byWYr2oH1y8

comment by devon_fritz · 2020-06-26T15:27:40.738Z · score: 2 (2 votes) · EA(p) · GW(p)

Thanks for this! The video doesn't have the Google Doc linked in it - where can I find that?

comment by AmAristizábal · 2020-06-26T23:31:30.205Z · score: 2 (2 votes) · EA(p) · GW(p)

Hi Devon, thanks for watching. Here is the Google doc: https://docs.google.com/document/d/1wsAivVBb2yj3tXjG6_3dBMK5BSJy0SoOUT1P2ek0Ddk/edit

But some of the best and more informative replies are on the comments of this forum post, I´ll organize them on the document one of these days(still thinking how haha). Thanks!