I'd like to hear from people about a process they use to decide how much to give to charity. Personally, I have very high income, and while we donate significant money in absolute terms, in relative terms the amount is <1% of our post-tax income. It seems to me that it's too little, but I have no moral intuition as to what the right amount is.

I have a good intuition on how to allocate the money, so that's not a problem.

Background: I have a wife and two kids, one with significant health issues (i.e. medical bills - possibly for life), most money we spend goes to private school tuition x 2, the above mentioned medical bills, mortgage, and miscellaneous life expenses. And we max out retirement savings.

If you have some sort of quantitative system where you figure out how much to spend on charity, please share. If you just use vague feelings, and you think there can be no reasonable quantitative system, please tell me that as well.

Update: as suggested in the comments, I'll make it more explicit: please also share how you determine how much to give.

New Comment
72 comments, sorted by Click to highlight new comments since:
Some comments are truncated due to high volume. (⌘F to expand all)Change truncation settings

The important thing is not to burn out, which would substantially reduce future giving. Right now I'm aiming only to max out Google's $6K/year matching limit. I probably won't increase that until I have a lot of savings, like on the order of several months of salary. Then if my partner is OK with it I'll head towards giving 10%, and after that I'd hope to give away 50% of future pay rises; whether we go beyond that will depend on how our income and outgoings compare at that point.

I wouldn't pay any attention to any comments that don't discuss the commenter's giving, by the way! You would probably get more informative answers if the question was "How did you decide how much of your income to give to charity?"

You should err in both directions. If you aren't at risk of burning out, you aren't donating enough. Even if you have a 90% chance of burning out at 10% of your income, that's better than donating less than 1%.
3Paul Crowley
This needs a little adjustment because income goes up with age, but yes, do an expected utility calculation!
The tithe is a well-established Schelling point, and one that I follow. 10% Jaibot and Bride of Jaibot of income goes to GiveWell's top charities. I Use Google's one-a-day app for non-optimal warm fuzzies, and I pitch in to explicitly rationalist causes like MIRI or helping a dying person afford cryonics. (here is where I notice Ambien kicking in. Post becomes less reliably representative of Jai from here on out. The important thing is to measure yourself against what you are: You are a human. Humans are not very good at world optimization, when left to their own devices. They knock things over and cheat and steal - and some goof comes of it, sometimes...but if you can take actions to make life better for people, at all, you're already winning. You made the choice that almost no one else made, the choice which you amplified and echoes after you made it: "I want to help. I realy want to help. I don't mean "look busy" or "be complimented", I mean "alter the state of the world such that everyine gets the things they would want if they were better at being the sort of person I want them to gtow ::falls down::
Note from a week later: I have no memory of writing this and find it hilarious. I still endorse the general message, if not the descent into semi-madness.
I see that people have very different attitudes to money. To me, several months of salary is not a lot of savings, but a very precarious amount. I would be uncomfortable without savings of several years of salary, and the ideal minimum amount would be enough to live on indefinitely without drawing on the capital. I don't miss trains either.
I was hoping that answering "How did you decide how much of your income to give to charity?" is obviously one way of answering my original question, and so some people would answer that. But you may be right that it's too ambiguous.

Meta: please consider crossposting this (and future topical posts) to the Effective Altruism Forum.

The logical thing to do seems to be: figure out how much more you value yourself than a random African. Figure out how much money you have to have so that the marginal dollar is that many times more valuable to them than to you. Donate everything beyond that.

I just plan on trying to spend as little as I can, and just donating the rest.

5Paul Crowley
Don't tell us what you plan on doing, tell us what you are already doing.
Spending as little as I can and going to college.
It's really difficult to 'shut up and multiply' in some cases. I mean, I'm going to get personal here because it feels like the best way to articulate my problems with mathematical utilitarianism. But right now, I don't produce anything like what I cost my society (in terms of socialized medicine, and support I receive from my parents). I feel very strongly that I shouldn't value myself more than a random African. But there are charities that claim I could save at least one life with what I spend on prescription fees every month. In terms of pure utilitarianism, unless I'm certain that I'm going to produce a lot more in the future and give some of that away, I probably ought to persuade my parents to give the help they give me with the rent to effective charities, borrow a bunch of money and give that to effective charities, then give the money I spend on my meds to effective charities until I basically kill myself. That doesn't feel right, but it's what I get from shutting up and multiplying.
That is sort of what I meant by it being the "logical thing". I don't expect many people to actually manage it. You have to work around your emotions. But it's still a useful first approximation for what you should be doing.
Maybe this means your feelings are wrong, and you should value yourself more than a random African. If you go outside LW and EA, the idea that anyone values themselves only as much as a random other person would be considered bizarre.
I don't think that's true? I think that, in practice, people value themselves more. But I think that it's a fairly common tenet of normal peoples' moralities that people are equal in value, and that if you asked random people, most of them would not say that they consider themselves to be more valuable or important than everyone else. Which, yes, means that there's a discrepancy between what people say they believe and what their actions say they believe, but that's pretty normal too.
Most people don't analyze things much at all. It's possible to ask a random person and be told he values everyone equally, and that's in some literal sense not saying what he believes. But if you just rephrased the question as "do you care more about yourself than someone else--would you pay my mortgage as readily as your own", he would answer "oh, if that's what you mean, then of course I care about myself more". Technically he's inconsistent, but it's a very shallow sort of inconsistency based mostly on the fact that he doesn't analyze things much; it isn't some kind of hypocrisy or denial.
Maybe I am amoral, but I don't value myself the same as a random person even in a theoretical sense. What I do is I recognize that in some sense I am no more valuable to humanity than any other person. But I am way more valuable to me - if I die, that brings utility to 0, and while it can be negative in some circumstances (aka Life is not worth living), some random person's death clearly cannot do so, people are constantly dying in huge numbers all the time, and the cost of each death is non-zero to me, but must be relatively small, else I would easily be in the negative territory, and I am not.
The problem with this is that you have to be extremely honest about just how much more you value yourself than a random African. In absolute terms, a dollar can do so much good in Africa that you'd go rather broke even if you "only" considered yourself as valuable as a hundred Africans. If you lie to yourself about your preferences, as people tend to do, you'll quickly see your true preferences reveal themselves.
This is probably not a bad idea for a heuristic, but reality is more complicated. For instance, they don't share your values. You might not eat meat from animals who feel pain during the production, but they might. You might not circumcize babies without anaesthesia, but they might. You might not give money to harmful religious organizations, but they might. You can factor all this is, but it is a lot more complicated than just figuring out how much marginal value a dollar would have for them. You have to figure out how much you value their having the marginal dollar, which requies modelling all the indirect consequences as well.
True, but my point was that you figure out at what point it's better to donate the marginal dollar, and you donate everything beyond that.
That's tautological, if you absorb all relevant factors into "better to donate". My point was you can't just project your current mind into the recipient's situation and then project how much they would value the dollar. In addition to this, personal factors such as lack of long-term commitment devices, probability of motivation-breaking also have to be integrated. On the upside, each additional dollar strenghtenes the market for effective charities, which incentivizes future people to create better and more attractive charities. (This is similar to the way in which marginal veg*anism incentivizes R&D of better and cheaper non-animal products.) I personally find it near-impossible to factor all of this into one number.
I do not know how much the recipient should keep, but I still think it would be helpful for him to know that it's independent of his current salary. For one thing, it's awfully suspicious for the amount he should keep to be that close to the amount he earns. Things like motivation-breaking also have to be integrated, but it's not enough to justify keeping 99% of his money. At that point, almost nothing is lost of he loses motivation, so the benefits of greatly increasing donation would seem to easily outweigh the risks.
By the way, you were implying that giving money to poor Africans is optimal altruism. This is clearly false. By doing so, you merely make up for other people's policy failures and create perverse incentives and moral hazards for the causes of poverty. In addition, other people are more likely than you to give money to the poor rather than to abstract causes, which means that you have an intellectual comparative advantage in focussing on those other causes. I think all of the following are better candidates for hed.utils. than poverty relief: * lobbyism (only very narrow topic range) * technological and scientific research (again, only a small topic range) * liberal eugenics * research and advocacy against nonhuman suffering (not many good candidates, but some are better than all povery relief charities) * utilitarian munchkin ideas like hedonium or artificial utility monsters * singularity and x-risk stuff (again, not many good candidates, but some are better than all poverty relief)

Giving What We Can recommends over 10% of income. I currently donate what I can spare when I don't need the money, and have precommitted to 50% of my post-tax income in the event that I acquire a job that pays over $30,000 a year (read: once I graduate college). The problem with that is that you already have a steady income and have arranged your life around it- it's much easier to not raise expenses in response to income than it is to lower expenses from a set income.

Like EStokes said, however, the important thing isn't to get caught up in how much you should be donating in order to meet some moral requirement. It's to actually give in a way that you, yourself, can give. We all do what we can :)

Its good to give factual numerical values. But I looked up GWWCs explanation for the suspiciously round number of 10% and it is: So this 10% appears to be arbitrary from the point of view of the OPs question. There seems to be no ethical reasoning behind the 10%. At least it looks more like charity-optimization. This may sound harsh, but this is what it looks in this light.
Could you explain what you mean by charity-optimization? I'm confused why you characterize labeling something as charity-optimization as sounding harsh.

My temporary solution is to max out my employer's annual match. That the maximum match is somewhere between 10 and 15 percent of my income is very convenient, as that makes me feel like I'm contributing the "expected" amount for an EA (this feeling is only important for fuzzies) but still leaves me with what seems to be a good amount to save and spend. It also allows me to avoid committing an answer to the question of whether to donate now or invest and donate later. The guaranteed, almost-immediate, soon-to-expire 100% return provided by the match wins pretty clearly over the EV of investing and donating later, and since I feel like I'm donating enough for now, I can evaluate what to do with what's being invested later on, based on my wants and needs.

Another interesting kink in your utility function: you can deduct up to half of your income from your taxes through charitable giving. As far as I know, this represents an opportunity cost: each year I have the option to donate half of my income for that year to charity and only pay taxes on the remaining half, and that opportunity goes away if I don't take it.
That's a very useful point. I do have employer match and it is likely to be an inflection point for effectiveness of any money I give.

Historical Reasons

Growing up, I had a pastor who advocated setting an upper limit on your income and donating everything you earned beyond that. I admired the principle of that stance. (It helped that he followed his own advice, living simply even though his books sold many copies.) But when I started making money after college, I learned how easy it was to increase your spending as your income grows.

A few years ago, when I was in the process of converting from being an Evangelical Christian to an agnostic/atheist, I decided that I didn't want a financial ... (read more)


I see a lot of “rules of thumb” suggestions in the comments (% of income, match Google/employer, etc.). I disagree with these approaches.

I recommend that you spend the requisite time to create a detailed projection of your cash flow. Try to determine how you want to use every dollar you will ever have.

A probabilistic projection will return an optimal giving range. The range may be wide, but at the very least I think this exercise will help you identify upper and lower boundaries.

Added bonus - It’s a great way to take account of your true priorities.

[This comment is no longer endorsed by its author]Reply

I like Julia Wise's thoughts in "Cheerfully".

(Personally I aim for donate 20%, invest 20%.)

"What's the right way to think about how much to give to charity?" My answer: how much better a place do you want to make the world? Do you want to make it a little better, a medium amount better, or a lot better? And how does this goal trade off against your other goals? How much would you sacrifice by making the world a lot better? Keep in mind that money is not a good way to buy happiness. So you sacrifice surprisingly little. But giving up things that you already have is painful due to the endowment effect. Try to fight status quo bia... (read more)

Yes it is. Furthermore, while technically paying for school tuition and medical expenses are "buying happiness" in the sense that if you don't pay for them you won't be very happy, the arguments against using money for happiness don't apply well to alleviating specific needs.
Note that both graphs have a log scale as their x-axis. It's pretty standard for economists, psychologists, etc. to suggest that humans have a logarithmic utility for money (i.e. your happiness is proportionate to the number of digits in your bank balance, so giving away 90% of your capital and reducing that number of digits by 1 has only a marginal impact on your happiness level). I think the statement "money is not a good way to buy happiness" captures the intuition behind logarithmic utility for money fairly well. (Also note that the article does not dispute my claim that money is not a good way to buy happiness. It just notes the lack of an asymptote in the utility curve.)
What does it mean for humans to have logarithmic utility for money? Do we have a measurable quantitative concept of utility that's natural enough that it would be silly to pull stuff like "utility2= log utility1, now humans have linear utility2 for money"!
The main ways to get a handle on this are to use subjective well-being scores (which is what those graphs do, and is somewhat questionable as to whether it's a natural unit), or to ask people about trade-offs or gambles they'd make (to elicit preferences as in a vN-M utility function). Both approaches lead to data saying it's approximately logarithmic, and there are also some theoretical reasons to think this is roughly right.

My current rationalisation for my level of charitable giving is "if, say, the wealthiest top billion humans gave as much as me, most of the worlds current problems that can be solved by charity would be solved in short order".

I use this as a labor-saving angst prevention device.

Me: "Am I a good person ? Am I giving too little ? How should I figure out how much to give ? What does my giving reveal about my true preferences ? What would people I admire think of me if they knew ?"

Me: "Extra trillions thing. Get back to work."

That's interesting, but how much money is needed to solve "most of the world's current problems"?
Also, do the wealthiest top billion humans share your values? If you asked them what the "world's current problems" are, would they give the same answer as you? For instance, there are hundreds of millions, if not billions, of people who want me to burn in hell forever for not believing in a god. I don't mean "who think I will burn", I actually do mean "want me to burn". Needless to say, their values are opposed to my values.

If you are looking to build up savings while donating in the third world, a good middle ground is to make 100 dollars a week of short term (6months To a year) loans on Kiva.

After not too long you build up a decent roll of money each month and in the meantime you are helping third world business growth

I have a good intuition on how to allocate the money

Isn't that a far more formidable problem than just deciding how much to give? Maybe you should tell us your allocation method.

I don't mean that I have one that's superior to anyone else's, but there are tools to deal with this problem, various numbers that indicate risk, waste level, impact, etc. I can also decide what areas to give in based on personal preferences/biases.

I think that something is better than nothing.

Sometimes setting an amount to give that is high can be discouraging, and can lead people to not give at all, or spend time feeling bad about not meeting their standards, rather than focus on what they can do to help.

If you give more than 0.7% of your income, you are doing better than most governments around the world, many of which have promised to spend 0.7% of GDP on foreign aid, but do not meet that threshold.

I don't think "How much money shall I give to charity?" is as important a question as &... (read more)

Also, what money goverments do spend on foreign aid is definitely spent in a much much less efficient way than GiveWell's top charities, and probably in certain cases even worse than not giving anything at all.

This post demonstrates a common failure of LessWrong thinking, where it is assumed that there is one right answer to something, when in fact this might not be the case. There may be many "right ways" for a single person to think about how much to give to charity. There may be different "right ways" for different people, especially if those people have different utility functions.

I think you probably know this, I am just picking on the wording, because I think that this wording nudges us towards thinking about these kinds of questions in an unhelpful way.

But it asks about “the right way to think about how much to give to charity”, not “the right amount to give to charity”. It is well possible (depending on what one means by “way to think about”) that there is one right way to think about how much to give to charity but it returns different outputs given different inputs.

I'd like to chime in to counter the common intuitions seen in the other comments.

First, that the more you give, the better it is.
It's known that money has a powerful effect on human psyche, and I strongly doubt that pouring say a million dollar to a charity that is used to manage ~10K$ donations will suddenly multiply its efficency hundredfolds. More likely, that money will be wasted. Or given to local warlords. Or used to pay higher salaries, etc.
The unreasonable assumption here is that there's a direct path from money to utility for other people, as if c... (read more)

It's known that money has a powerful effect on human psyche, and I strongly doubt that pouring say a million dollar to a charity that is used to manage ~10K$ donations will suddenly multiply its efficency hundredfolds. More likely, that money will be wasted. Or given to local warlords. Or used to pay higher salaries, etc.

Why would you give to charities past the point where they have room for more funding?

The unreasonable assumption here is that there's a direct path from money to utility for other people, as if charities were engines of perfect efficiency that burns money and churns out utility for poor people. Extremely unlikely, even for charities listed top tier on GiveWell.

Life is full of uncertainty, and there's nothing wrong with that. If you're saving lives on expectation, it's worth doing.

I agree that investing in R&D can be a better way to improve the world than donating to traditional charities.

Unless you are unhappy with "saving lives" as a proxy for ethical utility. A discussion of which you are surely aware.
Because it's a very uncertain point Also: Yes, unless you're neglecting a better way readily available. In which case you're paradoxically killing people on expectation.

Reading over the other comments, I think a lot of this is about finding the right schelling point.

This past summer, I put a bunch of reminders spaced out by a month or two into my google calendar that say "make effective altruist plan" - the idea being to make some sort of contract with myself before I graduate and get a regular income again, and sit down and think about what goes into that contract many different times before actually "signing" (which will probably be showing it to a trusted friend or two and asking them to help hold m... (read more)

I was recently thinking about the same thing. I recommend starting small and giving to every cause you like, including causes that don't have 501c3s. Start with a small amount, maybe $1k or so, and the same amount for everyone. They'll all want to treat it as annual, so don't overdo. Then once you know them better, decide whether you're happy with it, want to increase it, or want to cut it before they ask for an annual contribution.

I also pick one month of the year for all of it so you don't make 3 "annual" contributions a year. My parents ... (read more)

I agree with Mr Mind that there are a number of questionable hidden assumptions here. I would suggest that if you truly care about improving the world, you should seek to invest, not to give to charity. Ideally, this would be an area in which you already have some expertise, so you can pick wise investments. Most problems that charities work on are also business opportunities for entrepreneurs to solve; by being part of that, your money will do far more good in the long run. Market feedback makes sure that for-profit businesses really are solving the probl... (read more)

My rebuttal: ... Typically when LWers talk about charitable giving, we are talking about effective altruism.
I didn't say anything about trying to maximise your investment. I strongly suspect that the people behind Endaga, for example, could get more money by operating in a more conventional business environment. Entrepreneurship in the third world is emphatically not thoroughly picked over, but it is just not as profitable as selling to rich Americans. If you are willing to pay that opportunity cost, you can have a big effect. Secondly, effective altruism is a result, not a process. If the most effective way of helping the world is to give nothing to charity and set up businesses instead, then effective altruism would say set up the businesses. I am precisely questioning the assumption that effective altruism == charitable giving.

I am precisely questioning the assumption that effective altruism == charitable giving.

In that case your example of bike rides is pretty bad. It's a strawman. The comparison is bed nets or deworming.

This thread is interesting, but off-topic. There is lots of useful discussion on the most effective ways to give, but that wasn't my question.
To forestall an objection: I think investing with a goal of improving the world as opposed to maximizing income, is basically the same as giving, so that comes into the category of how to spend, not how much money to allocate for it. If you were investing rather than giving, and had income from it, you'd simply allocate it back into the category.
It's not a strawman, I linked to an actual website. No, those people don't call themselves Effective Altruists, but they are engaged in altruism and are trying to be effective. EA is an outcome, not a process, and the EA movement has no patent on it. Yes, it's a weakman, in that I deliberately chose an obviously ineffective charity. But my opinion of the rest of the EA movement is not much higher. The comparison is neither bed nets nor deworming - according to GiveWell's top ranked charity, it's sending money unmonitored, and hoping against 60 years of experience that this actually improves things rather than just being a leaky bucket.
There not much experience with sending money directly. Most of aid spending traditional went to big organisations and not to individual people in form of money.
You might try reading up on microfinance, it seems like it's in the same reference class as your idea. Givewell writes about it here (and other places). If I want to invest in companies being started in Africa, I'm going to have to do one of two things: make investments myself or give my money to someone else to invest it for me. If I make investments myself, I'll either do it remotely (which seems like a bad way to get to know investment opportunities, do due diligence, etc.) or travel to Africa and effectively work as an angel investor myself. Giving the money to someone else to invest seems like it might run in to the same sort of problems that giving the money to someone else to do good deeds with would, but it could be mitigated by my money manager's incentive to get a good return for me in order to attract additional clients. So this idea seems more or less equivalent to: find a venture capital firm operating overseas and invest your money with them. My impression is that venture capital firms typically have institutions rather than individuals as their clients; I'm not sure if there's a good reason for this. Investing in the stock markets of developing countries is another idea. My friend who's an expert on economics has told me that it's common for a country's economy to grow well while its stock market performs poorly, I guess due to a high amount of turnover among the companies listed? I'm not sure how this would effect the EA implications of investing in a developing country's stock market.

There is no fixed tithe... Donate slightly more than required to not feel guilty. If in doubt, increase the amount until you are sure.

This advice seems to assume that either your reason for giving is simply to assuage guilt, or your guilt-feeling machinery is accurately calibrated to the moral and non-moral ends you seek to achieve. The first case seems unlikely to apply to anyone who's seriously trying to answer questions along the lines of "how much should I give?" by rational means. (But if it does, simply training oneself to feel less guilt would seem more efficient. Alas.) The second seems very optimistic.
The only reason that medical bills and supporting a family are even questions is that he values his family's well being more than he does that of someone who might be helped by charity. Phrasing things in utilitarian terms, he puts a much higher utility on his family than on a random person. As such, he should only donate to charity once he has spent enough money on his family that the utility from doing is, per dollar, so much less that that factor overwhelms the large multiplier that he gives for his family's welfare compared to a random person's. Given his description, he clearly hasn't reached that point yet, so he should spend nothing on charity. This leaves guilt (or warm fuzzies, which are the flip side) as the only reason to spend anything at all. The rational amount (ignoring questions like "what utility to I get from assuaging my guilt") is zero. Take care of your family, who needs it, and stop feeling guilty for not donating.
I'm failing to follow your reasoning at this point. How does his description indicate that? ("A random person" isn't quite the right reference point, by the way. It seems reasonable to hope that a well selected charity will direct help to people who are in distinctly more need than average.)
There are two factors here: how much he values the utility of a family member compared to the utility of someone else, and how much utility a given amount of money can produce for his family compared to someone else. The person helped by the charity is non-random with respect to the second factor, but random (in the not-special sense) with respect to the first.
Oh, I see. ... Although, actually, it might be non-random w.r.t. the first factor too. (But probably in the "wrong" direction: that is, I guess that most people care more about a given person's utility when that person is (a) near them and (b) like them, and when the person in question is an affluent Westerner these probably both anticorrelate with being among the world's neediest.)
I apologize for being unclear in my description. At the moment, after all my bills I have money left over. This implicitly goes toward retirement. So it wouldn't be slighting my family to give some more to charity. I also have enough saved to semi-retire today (e.g. if I chose to move to a cheap area I could live like a lower-middle class person on my savings alone), and my regular 401K contributions (assuming I don't retire) would mean that I'll have plenty of income if I retire at 65 or so.
Not sure why the above got so many downvotes... seems pretty reasonable to me.

The more is donated, the better, so figure out how much you expect to want for your own spending/saving and donate the rest. Don't give so much that it takes a toll on you; it must remain something achievable that you want to do.