How I’m finding my place on the spectrum of altruism

Research notes on Effective Altruism

Table of Contents

1. How I got here, and why this matters to me
2. What does “doing good” mean?
   - Promoting welfare, and how you measure it
   - Defining moral impartiality & expanding our moral circles
   - Long term view
3. What should we focus on in order to do the most good?
   - Prioritizing how to do good
   - Using the framework: Some high and low priority causes
   - Walking the walk
4. How much should we help / how much should we give?
   - Giving by donating
   - Giving consistently, starting now, rather than waiting till life is over
   - Other ways to give
5. Conclusion and call to action
   - A call to action
6. Appendix
   - Criticism of Effective Altruism
   - Further research on EA

How I got here, and why this matters to me

Like many of my friends and family, I grew up in an environment that emphasized the value of helping others. I learned the values of charity, kindness to others, and fairness from my parents, my teachers, and activities like boy scouts. I always thought that I would use my career to have a positive social impact on the world. I’m really thankful for the environment I grew up in, but it left the door open on what exactly constitutes “positive social impact”. I didn’t have a great framework for comparing career paths, potential employers, or charitable causes that I could donate to. Of course, I knew I would never work in the oil industry or in tobacco, but that still leaves a lot of options for “socially impactful” careers. So, over the past 5 years of my career, I’ve found myself joining organizations and donating to charities that have any kind of positive social impact on basically any cause that feels good to help out with, whether that’s cancer research, supporting the arts, or a food bank. Recently, that’s started to change.

A few months ago, I discovered Effective Altruism (EA). EA is a social movement that was started by the philosopher Peter Singer, and has grown into a philanthropic community of a handful of institutions and maybe 10–100k individuals around the world. Reading about EA is starting to convince me of two things:

1. Blindly joining any seemingly good cause before me might make me feel good, but likely isn’t having the most impact on the world in the ways that I might truly care about. And it may not be a small difference in impact — not just a matter of 1.5x or 2x impact. Picking a career or making donations in line with causes that are most aligned with my moral theory might mean that the same contribution has a 100–1000x difference in impact compared to whatever socially impactful cause randomly pops up in front of me.
2. On an even deeper level, understanding the philosophical arguments for what constitutes good is important in order to define positive impact on the world. Otherwise, it is very easy to think anything that helps anyone is worthy, whether it helps one person or a billion people for the same amount of effort.

This post compiles my research and conclusions on a range of questions that arose as I learned about the EA framework for defining how to have the most positive social impact possible.

What does “doing good” mean?

Photo by Tim Marshall on Unsplash

The first question I researched was EA’s definition of the meaning of doing good. A basic assumption in EA is that the purpose of altruism is to do good for the world that we live in, not to make oneself feel good (the latter concept is called affective altruism). That aligns with how I’ve thought about altruism, though everyone might not agree. In any case, with that assumption in mind, let’s take a look at EA’s definition of doing good:

“Doing good” or “making a difference” is about promoting welfare, considered impartially, over the long term — without sacrificing anything that might be of comparable moral importance.

There were three key parts of this definition that I researched in depth: “promoting welfare”, “considered impartially”, and “over the long term”.

There were two things that I intentionally left out of my research:

1. I didn’t research the last part of the EA definition of doing good, “sacrificing anything that might be of comparable moral importance”. This part of the definition can lead to some interesting debates, but in the rest of my research it didn’t really have practical significance in any tradeoffs that had to be made about doing good or choosing causes.
2. I didn’t look into the welfare of animals and the environment, which are actually significant components of EA. I decided to start by diving deep into human welfare.

Promoting welfare, and how you measure it

EA’s utilitarian definition of promoting welfare seemed compelling to me at face value: “promoting happiness, health, and the ability for people to live the life they want”. But as an engineer and product manager, I naturally wanted to know how the philanthropic community measures welfare. In my experience a lot of the nuance of what counts as positive or negative is hidden in how something is measured. Here’s what I found:

Macro and micro metrics for welfare

At the macro level (large populations over long periods of time), the World Bank and other international institutions often use GDP as a proxy for welfare. It’s a fair start — high productivity should mean higher quality of life (see “does money make you happy?”). The UN has tried to create a more holistic macro metric called the Human Development Index (HDI). It includes life expectancy and expected education in addition to productivity (gross national index per capita). One criticism of HDI is that it assigns arbitrary weights to education, GDP, life expectancy.

At the micro level (assessing the impact of a specific intervention), a metric used by some NGOs and welfare economists to assess utility is quality-adjusted life years (QALYs). An economist might ask “how many QALYs did this intervention create or save per $ of cost?”. QALYs is an ideal metric from an economist’s perspective because it tries to capture how much individuals value their quality of life with or without an intervention. However, QALYs can be hard to measure across heterogeneous populations who have different types of problems. Often, the measures for the value of interventions are specific to the region and problem being solved. For instance, measures like earnings, food security, height or weight of children at various ages compared to benchmarks, are more objective measures of poverty reduction.

Measuring happiness, and does money make you happy?

Photo by Sharon McCutcheon on Unsplash

As I read about measuring welfare, a key question I had was whether money makes people happy. I have consistently heard things like “the poor adjust to their lower income and wealth and can have lives just as happily as the rich”, and “it’s perfectly likely for a rich person to be depressed”. So, I explored the research on whether money makes people happy.

At a micro level, the two main ways of measuring happiness are 1) asking individuals about their “general life satisfaction” or 2) asking individuals about their “current emotional state”. In the studies I read, 1) correlated pretty well with income (notably the correlation was logarithmic), whereas 2) did not correlate with income much at all. This is the most recent and thorough work on the topic (by two Nobel laureates), and this EA article explains further, with links to additional relevant papers. I found it interesting that on average, money does make people happier, tapering off after ~$75k in the US.

At a macro level, the World Happiness Report is one of the leading measures of 1). Though it measures happiness in line with somewhat Western values, there’s a notable correlation between GDP and “happiness score” based on their findings.

My takeaways: The philanthropic community uses pretty sensible methods to measure welfare. They may not be perfect, but they seem to be good enough to understand the general picture of welfare in different countries and populations. Secondly, while additional money does not significantly increase the happiness of middle and high-income individuals, it has a big impact for the poorest members of society. For the poor, money does buy happiness.

Defining moral impartiality & expanding our moral circles

The next important clause in the EA definition of promoting welfare is to consider all lives “impartially”. Here’s what the EA community means by this:

“[We should] strive to treat equal effects on different beings’ welfare as equally morally important, no matter who they are — including people who live far away or in the future, and including non-humans”

I found that moral circle expansion was the analogy that resonated for me. The image below (from this paper), illustrates this concept well, describing how the values of justice, equality, and fairness that many of us have grown up with can help us empathize with larger and larger groups of individuals. Each layer of the moral circle is more challenging to expand to, because it may mean we are making personal sacrifices for other beings whom we can relate to less and less.

Centripetal and centrifugal forces in the moral circle

But as I read about moral circle expansion, I had to ask myself some challenging questions. Do I really believe that I should try to empathize with all people equally, even if they live in a country that I will never see myself? Do I have an obligation to help people far away, or should I focus on my community?

To me, the ethical answer is yes, all people matter equally. It’s hard to come up with a counterargument without contradicting the central tenants of fairness and equality. The veil of ignorance analogy helped me reason about this — if I could have been randomly assigned any human life on Earth (and therefore could very well be poor), I would certainly hope for privileged people to help the poor, at least a little. Now this isn’t to say I have lived my life in the most ethical way possible, or that I even think it’s possible for me to be 100% ethical. But more on this in “how much should we give” later on.

My takeaway: I should strive to expand my moral circle to all humans.

Long term view

The third piece of EA’s definition of promoting welfare is “over the long term”. This is simply an expansion on the idea that we should leave the world in a better place for our children and their children.

The basic philosophical idea behind EA’s proposal is that humanity is likely to exist for many thousands of years, and at the end of it all, what will matter most is the sum total of welfare that existed during humanity’s time on the planet. I found the work of Hillary Greaves (1, 2) pretty convincing on this topic.

As convincing as the theoretical argument is, I think it’s pretty challenging to put into practice, especially the farther we look out into the future. Sure, there are clear risks like pandemics, nuclear war, and global warming that we know can affect humanity negatively in the next century. But we really don’t know what the world will look like in 1000 years, and that makes it difficult to know how we can promote the welfare of that future world.

The one exception to all this is if you are a nihilist and don’t think there’s a point to anything humans do in our blip of the universe. Nihilism is interesting, but doesn’t change that there are hungry people out there!

My takeaway: It’s important to prioritize the welfare of future generations, but it’s hard to predict the future and not worth prematurely optimizing for the good of humanity 100 generations from now.

What should we focus on in order to do the most good?

Once I was more or less aligned with EA on how to define doing good, I explored the question of how to do the most good. This question is at the crux of “effective” in Effective Altruism. Here are EA’s basic points on this topic:

1. Prioritize how to do good: In order to find the optimal focus areas, we should use a prioritization framework that assesses how much welfare we can create across all people in the world equally, without preference given to any cause.
2. The best causes may not be what you thought: If we judge what we could work on using our equal prioritization framework, we’ll find some causes actually have 100x the impact of others! Moreover, we should be open minded about this ordering changing over time.
3. If we believe the talk, we should walk the walk: We should align our philanthropic actions with our moral theory as much as we can.

Prioritizing how to do good

EA uses a 4-part prioritization framework to assess the impact an individual could have working on a class of problems in the world. The framework includes the following components, in descending order of impact on the final priority.

• Scale (0–16): how much welfare can be gained in terms of productivity or health outcomes? How many lives can be saved or improved?
• Neglectedness (0–12): “Is there any reason to expect this problem not to be solved by: (i) the free market (ii) government (iii) other individuals looking to have a social impact?”
• Solvability (0–8): What is the capacity to solve this problem given more resources? Do we have good technical solutions? Would doubling efforts would have a big impact on the amount of good done?
• Personal fit (0–4): This applies more to careers directly related to the cause area. Skill set match and personal interest in a problem can be a 10–100x multiplier of impact, especially in a domain like research which has long investment times and uncertain rewards.

I find this image to be really illustrative of why this prioritization framework is important:

My takeaways: This framework makes sense to me for prioritizing how to do good with donations. I think this framework fits even if one’s values are different from EA’s — it’s still a good way to prioritize how to spend scarce resources.

Using the framework: Some high and low priority causes

Some high priority EA causes

• Eliminating poverty fits squarely into having a large positive impact, because GDP and happiness are logarithmically correlated [1] [2]. Effective international philanthropy has a much higher impact than domestic philanthropy, because $1 buys much more welfare in developing countries than in the US (and because the baseline welfare is lower in developing countries).
• Reducing extinction risks ranks as a high impact problem under the EA framework, because the # of potential future lives is very large. One issue with choosing this type of cause is how hard it might be to estimate the probability that any particular threat will be an extinction risk, and further estimate how much progress can be meaningfully made against it with my effort. This class of causes is interesting, but often receives a lot of eye rolls from EA critics. I think it’s an important theme to address, but I’m personally not passionate about most of these causes right now.
• Global Priorities Research is the economic and policy research required to figure out what is important to spend time on. This cause is arguably underfunded but very high impact. The fact that most US giving goes to domestic causes when the same impact in developing countries can be made for 10–100x cheaper is an indicator that research and advocacy into the right causes to focus on could be impactful.

Some lower impact causes that I thought were really important

• Cancer has a significant scale, but very low neglectedness (hundreds of billions of research and spending annually), and low solvability (researchers are making progress, but quite slowly). This makes additional funding or efforts to cure or treat cancer a low impact cause. On a personal note, this is what I’ve worked on for the past 5 years, and it’s interesting to learn how EA ranks this cause.
• Education in the developed world is a problem with very low neglectedness and solvability, yet it gets a lot of attention. Read the linked article to learn more.

Walking the walk

EA doesn’t actually say a lot about this topic, but I think this is challenging. It’s one thing to learn about Effective Altruism and to agree to statements like “all people are equal” in theory. But actually changing our donation strategy, our consumption habits, or our careers can be big steps. I started thinking about all the donations that I’ve been making under social pressure over the past 5 years (most of them, as it turns out):

• I donated to my alma mater, Cooper Union, because I felt lucky to have attended the institution while it was tuition free and funded by philanthropy.
• After Trump was elected I rage-donated to Planned Parenthood, environmental groups, and the ACLU.
• Any time a family member or friend was doing a fundraiser for pretty much anything at all, I donated some amount of money to support them.
• I’ve given lots of small change to panhandlers.

After reading about EA, I started wondering about the thousands of dollars I’ve given to these causes… should I be saying no to these institutions, to friends and family, and to panhandlers? The same money could have 10x or 100x the impact for the poorest people on the planet, even if I never meet those people myself.

My takeaway: I should strive to spend most of my philanthropic effort on causes using the EA framework, prioritizing all humans impartially. But these donations I’ve been making to causes that come before me on a day to day basis are not meaningless! They bring value to me, my loved ones, and my community, and so I should prioritize this second bucket of causes for myself. I’ll continue to make small donations to Planned Parenthood and boy scout troops, knowing those $’s may only have 1/100th the impact they could have on human lives, but that that smaller impact will be closer to home for me and my community, and therefore meaningful.

How much should we help / how much should we give?

At this point in my research, I had started to feel aligned with spending the majority of my philanthropic energy on EA causes. But the big question that remained was how much I should give. I talked to several people about their philosophy on this topic, and did some independent research to come up with an answer that works for me. What I came up with was that giving level depends on your personal situation. Giving everything is obviously infeasible, but in my opinion giving nothing is also the wrong answer for most of my friends and family. In this section, I lay out some of the reasoning for how I came to a giving level for myself.

Giving by donating

The first question I thought about was how much happiness I might be sacrificing with every dollar that I donate. After reading Kahneman and Angus Deaton’s paper, I wondered whether I could live happily on a $75k income (to be transparent, I am lucky enough to work in tech and make more than that). I thought about how I’ve spent money over the past several years, and what kind of expenses lie in my future. After doing some rudimentary analysis, I decided that at my current income level, I could probably donate at least 10% of my income without having a noticeable negative impact on my happiness. From that 10%, each dollar I donate to causes supporting the lowest income people in the world will have a wayyy bigger impact than spending the dollar on myself for a nicer car, house, or another ski trip.

My takeaway: Aim to give something like 10% of my income every year. I might even take the giving what we can pledge after a year or so of seeing how it goes.

Giving consistently, starting now, rather than waiting till life is over

The second question that came up was whether I should start giving now or wait until my major life expenses were satisfied (buying a house, raising kids, saving for retirement) and donate what was left over. Admittedly, I haven’t done a ton of research on this topic, but my intuition says that if I wait till the end of life, I will end up spending much more money than I may need to be happy. Just like work fills the time you allot to a task, I think spending can fill the budget you allot. Secondly, donating consistently over time is a much better way to get tax breaks. Speaking of tax advantages, donor advised funds (DAFs) can help your donations appreciate over time tax free if you aren’t ready to commit to a cause yet. Note, there’s no good evidence on whether investing to donate 2–3x more money in the future is better or worse than donating now. It’s hard to say, so I’m considering donating half my money directly to causes and half to a DAF to allocate later in life.

Finally, the EA community has recognized that a major barrier to making effective donations can be identifying the causes and charities that are actually most effective. The community has invested a lot of time into institutions like GiveWell, which conduct deep research to prioritize the most effective major charities that have capacity to deploy more funding.

My takeaway: Giving consistently over time is a better way to hold myself to my philanthropic goals.

Other ways to give

Of course, donating money isn’t the only way to do philanthropy. I could work full time at a charity, volunteer my time, or raise money from others. I’m still thinking about these options, and weighing how my skill set overlaps with what the most effective causes could use in terms of support. 80000 Hours, an EA affiliated organization, has a great post on how to choose a career to have a lot of impact. It’s going to take some time for me to think through this, and my career / donation balance may change over time.

Conclusion and call to action

So after all that research and thinking, here are the key things I’m coming away with:

• Since “doing good” is an important part of my values and life, I should think rigorously about how to align my philanthropic actions with my philosophy for who to help and how to help them.
• If I truly believe that all people are equal and should have an equal opportunity at a happy life, then certain causes, such as international health and poverty alleviation, can help me have a 100x impact with my donations. As a start, I’m going to use GiveWell to help prioritize most of these donations.
• It seems quite likely that I can live a very happy life even if I give 10% of my income to charity every year. Most of these donations should go to the most effective causes for global human welfare in alignment with my philosophy, but that doesn’t mean that a small amount of donations can’t go to causes that are in my community or that I’m personally going to benefit from.
• There is so much more to learn about Effective Altruism and how to have an impact on the world!

A call to action

If you’ve made it this far, then I want to leave you with a call to action. I’m really interested in the opinions of my friends and family about the research I’ve done here, and really anyone else who comes across this.

• If you disagree with what I’ve found, I really want to know why and hear your opinion.
• If you agree with what I’ve found, I really want to know if / how you want to act on it.
• If you are on the fence, please follow the links I’ve shared to read more, and then tell me what convinces you one way or the other!

Some difficult questions that I want you to ponder:

• Will learning about EA change your pattern of giving?
• How do you decide how much of your time / money to give?
• How important is it for you to have tangible causes to donate to vs “optimal” causes?

Appendix

Criticism of Effective Altruism

I took a look at the main opposition and criticism to Effective Altruism, and here’s what I found. There are some valid points here, but on the whole they haven’t convinced me that EA is a poor way to think about altruism. A more extensive list of criticism and responses can be found on EA’s website itself.

The most famous criticism of EA is an article by Ken Berger & Robert M. Penna, of the charity rating organization called Charity Navigator. The authors call out EA for being superficial and marginalizing lots of good meaning organizations. They argue that the utilitarian attitude to philanthropy can kill altruistic spirit. I think this criticism is valid if you believe the purpose of altruism is for your own happiness. If you want to do good for the world at large, then I don’t understand why you wouldn’t want to take a good hard look at the impact of different charities on your definition of “good”. Sure, some of those might resonate with you more than others, but when the impact is 100–1000x, taking an initial look might be worthwhile!

Similarly, Simon O’Regan has a nice short post which calls out positives as well as some negatives of EA. The first negative that stands out to me is how subscribing to EA can feel less tangible than contributing to causes you might be close to. This is definitely true, and could take a toll on enthusiasm over the years. One thing I want to learn more about is how the EA community mitigates this risk. Secondly, O’Regan points out that EA can neglect to fund smaller organizations with less measurable outcomes. Also valid, and something I’d want to learn more about.

Further research on EA

As I wrote and got feedback on this post, a few areas stood out as needing more research and thinking. As next steps, I plan to:

• Understand how to compare a bigger personal career shifts to continuing to work at interesting tech jobs and donating a chunk of my salary.
• Doing more detailed financial planning to figure out how different levels of annual donations may affect my long term happiness and ability to meet financial goals. For instance, if I want to buy a house in a particular neighborhood, how much money do I need to save. And based on things like that, how much can I give without impacting my happiness?