Gas station without pumps

2019 July 4

How much to give to charity

Filed under: Uncategorized — gasstationwithoutpumps @ 14:56
Tags: ,

I have been doing a short series of posts about charitable giving:

  • Thinking about charity started the series by giving a series of links to articles my son recommended on effective altruism.
  • Why charitable giving gave some of my initial thoughts on why people (and more particularly, why I) give to charity.
  • Readings about charity gave my reactions to each of the articles my son had recommended to me.

Today, I want to think through the amount that I will give and maybe a rough allocation for what sorts of charities I’ll give to.

The effective altruism movement has selected 10% of gross income as a good target for people to hit for feeling good about themselves—it is neither so high that everyone will feel bad because they can’t do it, nor so low that giving is meaningless symbolic gesture.  There is a pretty good explanation of why this arbitrary goal was chosen in https://www.givingwhatwecan.org/about-us/frequently-asked-questions/, so I won’t repeat those arguments here.

The effective altruism movement seems to consistent mainly of young people, for whom income is the primary measure of what they can afford.  Net financial worth (assets minus liabilities) is low or even negative for many of them, so plays little or no role in their calculations.  For people like me, who have a moderately good income but have been frugal and invested much of their income for decades, the 10%-of-income rule of thumb underestimates what is reasonably affordable.  (For people with high debt load, it probably over-estimates what is affordable.)

There are several ways to handle invested wealth consistent with the 10%-of-income guideline:

  • Wait until the investment is converted to income and give away 10% of the amount withdrawn as income.  This is simple to compute, but doesn’t seem consistent with the effective-altruism principle of giving now, rather than delaying giving until later.  It also makes for some difficulties in figuring out how much to contribute if withdrawals from the investment are being made only for the purpose charitable giving.
  • Compute the change in investment value (minus contributions from income to the investment) each year and give away 10% of the growth in the investment.  This is a little messy to compute, as it requires keeping track not just of the current value of the investment, but previous values and how much was contributed to the investment from income.  Also, taking differences is essentially a high-pass filter on the investment value, and high-pass filters accentuate noise in the estimates of investment value (particularly a problem for illiquid investments like real estate, whose value is hard to estimate accurately).
  • Estimate the income that would be obtained by converting the investment into an annuity, and use 10% of that estimate as the amount to give.  This approach has the advantage of asking relatively little from the wealth of young people, but scaling up the distribution of wealth as a person ages.  Because annuity rates are set to be safely less than reasonable expectations of investment income, this method on average results in lower income estimates than growth of investment for young people, but higher estimates for old people (as principal is depleted over the duration of the annuity).

I favor the third approach, as it only requires estimating net financial worth (excluding any assets that have already been converted to annuities, which get treated as income). The resulting formula suggests giving about 0.5% of net worth each year for someone my age (0.4% for someone much younger and up to 1% for someone much older), in addition to 10% of income.

This year, because I’m forgoing 1/3 of my salary in order to take leave in the fall, 0.5% of net worth would make up a bigger share of my giving than 10% of household income—more than doubling the 10%-of-income guideline.  This means that I expect to give about five times as much this year as in previous years (when I fell short of even the 10% guideline based on income alone).

I don’t plan to simply multiply my checks by a factor of five to the same charities I’ve been giving to—if I’m giving that much more, I want to more certain that the giving is effective at meeting my goals.  I do not have a linear scale of “utilons” and the different sorts of charity are for incomparable goals, so I’ll be setting levels of giving for each type of charity based on what I think is sufficient to meet my needs for contributing to that sort of charity, not trying to maximize the value of my donations in a single dimension.

  • Global effective charities in development and health—possibly the Against Malaria Foundation that is one of Givewell’s top charities.
  • Local health and nutrition services (Planned Parenthood, Second Harvest Food Bank, …)
  • Local education (student projects at UCSC, Cabrillo College Foundation, WEST Performing Arts, … )
  • Environmental organizations (particularly ones involved in more sustainable transportation at the local, state, and national level)
  • Local cultural organizations (Museum of Art and History, Santa Cruz Shakespeare, …)
  • Social justice organizations mainly at the national level (Southern Poverty Law Center, ACLU, …)
  • Political organizations mainly at the national level (Verified Voting, PAC for a Change, …).  I probably won’t give much directly to political campaigns, as that seems to be a rather inefficient use of money (most of it seems to go towards raising more money).  There is no tax benefit to giving to politics, but tax advantage is not the primary reason for giving (and I may end up taking the standard deduction anyway, so charitable giving would have no effect on my tax deductions).

I have not decided how to balance the different sorts of giving, nor exactly what organizations to give to in each category.  This will require some calculation to figure out exactly how much we will give, then some probably long discussions as we allocate the money to different pools and different charities.

The effective-altruism approach would argue for picking whichever charity is most valuable (maximum utilons/dollar) and putting all the donations there, but that approach does not work for me—the different sorts of giving are not comparable and I’d rather give a little in several directions than a lot in one.

2019 June 24

Readings about charity

Filed under: Uncategorized — gasstationwithoutpumps @ 19:33
Tags: ,
In Thinking about charity, I listed some references that my son sent that influenced his thoughts on charitable giving.  Now that my grading is done, I finally have time to read them.
This one is a fairly simple plea for doing expected-value calculations, but it does not provide any insight into the relative values of different outcomes, assuming that everything is trivial to quantify.

This argues that money is the unit of caring, and that professionals should give money rather than time to organizations (under the assumption that the division of labor with specialists doing what they do best is most efficient).  The post does not preclude volunteering time, but sees it as a pleasure for the donor that should be bought with hard currency.

There is some validity to the thought that a highly paid professional will do more good by working an extra hour at their profession and donating the extra income, rather than volunteering to do something less skilled that someone could be hired to do more cheaply. The author misses a few points though: 1) Salaried professionals do not generally earn more by working longer hours, so volunteering time may provide working hours that would otherwise be wasted on unpaid effort for some corporation. 2) For many people in semi-skilled trades and low-level office work, the extra income from working longer (after taxes and other deductions) is less valuable than their direct labor (thanks in part to underpayment in those fields). 3) Working continuously on the same things can lead to burnout and a shortened career, hence to less total money to give—a change of pace can be valuable for maximizing the total value of a person’s labor.

https://www.lesswrong.com/posts/3p3CYauiX8oLjmwRF/purchase-fuzzies-and-utilons-separately

This post suggests “I recommend that you purchase warm fuzzies and utilons separately.  Not at the same time.  Trying to do both at the same time just means that neither ends up done well.  If status matters to you, purchase status separately too!”

I think that this post had the main point that my son was trying to make—that it is more efficient to buy the different benefits of charitable giving separately.  The post identifies three benefits: warm fuzzies, status, and utility (though again it gives no real guidance in determining utility).  I identified a few other reasons for giving in Why charitable giving, but the basic notion that the different benefits are separable and may be most efficiently purchased separately is probably a good one.

This post makes a good argument for giving 10% of income more or less continuously, rather than delaying donations to the “wealth-distributing years” as university development officers refer to the post-retirement time of life.  They make a good argument for 10% being a reasonable target.

They do point out reasonable exceptions to the rule, but do not take into consideration reserves people should set aside for emergencies and for retirement (they are more focused on student debt and mortgages).  For people like me, whose salaries are much larger than living expenses and who have good insurance and liquid investments for unexpected expenses, the 10% guideline is a good one, but for people with little in savings, it may be better for them to delay charitable giving until they have adequate reserves—it is not efficient to give money and then need to have others give to you in order to survive.

Their guideline for retired people is also unclear: “People who have retired or partially retired (which we roughly define as having started to draw a pension) can join Giving What We Can and remain members for as long as they continue to donate at least 10% of their spending money.”  Their definition of spending money assumes an externally determined income (like a pension or annuity) rather than drawing down invested money.  Once one reaches 70½, the required minimum distributions from tax-deferred retirement investments may provide a reasonable definition of income, but there should probably be also a wealth-based, and not just an income-based component.

Overall, I felt that this post made a better case for effective altruism than the “lesswrong” posts.

This one argues for charitable giving over political action as being much more efficient use of time and money, and provides a rationale for the rather arbitrary 10% of income giving level recommended in the effective altruism community.

This post presents (and rejects) a utilitarian argument that animal rights are more worth funding than alleviating human suffering (because of the very much larger number of animals killed than humans).  It presents a bit of a reductio ad absurdum for utilitarian arguments about charitable giving, and those utilitarian arguments are the main basis of the effective altruism movement.

Some of the discussion in the comments of the post is also interesting.

This article discusses some rather silly arguments about “existential risk”—the risk of human beings going extinct—and points out the absurdity of calculating expected values of extremely high value events with extremely low probabilities.  Humans are extremely bad at dealing with very large and very small numbers, and our ability to estimate either large or small numbers is terrible.  As a result, the expected value (the product the badly estimated high value and the badly estimated low probability) can be almost anything.  Using this essentially arbitrary number as an input to a decision-making process is absurd.

The comments were again interesting, and I particularly liked one by Murphy:

Keep in mind that the “become a hedge fund manager” advice is based on the current situation where many charities have volunteers lining up to help while being strapped for cash.

In a world that already has many hedge fund managers channeling their money to the most effective charities but a lack of volunteers the most effective choice might be to volunteer.

If you imagine it as a market where the options all have some expected return, the rational thing is to invest in the cheapest item. This moves the price and the guy walking in the door behind you might look at the same market and see that the most effective choice is to invest in the next option which is now the cheapest.

I’m going to agree that I don’t buy arguments based on numbers of future people, I don’t believe in there being a moral incentive to maximize numbers of future people though future-suffering of people who have yet to exist I would attach some weight to. I don’t care if there could be a trillion people on earth, that’s not a goal in my mind but I do care if the grandkids of the current generation might live in misery.

This post argues against investing in politics, particularly for large-scale systemic change (man-vs-man instead of man-vs-nature problems).  The basic argument is that people, even intelligent people, have been historically very bad at making or attempting large-scale changes (the disastrous communist states resulting from Marxism are brought up as an example, as is the eugenics movement).  Focusing on smaller-scale incremental changes, where results are more predictable, is more likely to result in the desired net positive effect.

One comment on the post that I liked is

In some cases, I think it’s pretty obvious that there’s a market inefficiency in morality. The general reason for this is that I’m utilitarian and most people aren’t. Most humans have revealed preferences to care more about people close to them in time and space and mind-shape, all of which I reject. So it’s totally plausible that given my values, the most important interventions are massively underfunded.

There’s good evidence for this. Most Americans happily (or unknowingly) torture and kill dozens of animals every year. The total spending on existential risk reduction is a tiny, tiny proportion of what it would be if the world cared about humans a thousand years in the future as much as we cared about humans now. And the ratio of spending on local versus international charities speaks for itself.

That’s the more polite reason that I believe that there’s an inefficient market in morality. The less polite version is that most people aren’t strategic or selfless, and their implicit main motivation in morality is demonstrating that they’re trustworthy and generous. I think it’s pretty obvious that very few people are actually trying to help others in a cause-neutral way. So I think it’s very likely that I can “beat the market”, as it were.

I’m not a pure utilitarian, and I do care more about people close to me physically, genetically, or in terms of what they value.  I could probably cast my beliefs into the expected-value computations preferred by the effective altruism movement by assigning numeric values to different benefits for different people, but I’m not going to do that, because I don’t really accept the underlying utilitarian ethics.  (Also, the numbers would argue against almost all the local charitable giving I do, unless you consider the life of child at risk of malaria to be much less than a week of theater summer camp for a kid whose parents can’t quite afford it.)

Like Buck, I find So it’s totally plausible that given my values, the most important interventions are massively underfunded.”    I end up, however with a different set of charitable investments, because I have different values and a different basis for figuring out how to spend my money.  (Note: I still haven’t made up my mind what charities I’ll support this year, much less on an on-going basis, nor am I at complete odds with Buck—I’ll probably assign some portion of my giving to the sorts of charities Buck would support, but I’ll also be giving locally and to organizations that support culture and not just life.)

I noticed (after the fact) that the author of the post called out the same comment by Buck as particularly interesting.

This post argues for the value of having a community set a standard for how good one has to be to be considered a good person, and the economic argument for setting this level to try to maximize the goodness actually achieved.  The basic theory is that if most people are being good in order to meet society’s expectations, then setting the expectation too low results in little goodness, but setting the level too high results in many (most?) people not giving because they see the goal as unobtainable.  Setting a threshold somewhere where people are willing to strive to achieve it maximizes the goodness obtained.  Different thresholds (perhaps self-imposed) may apply to different people.
The post seems to be about letting people feel good about being “less bad” without requiring them to be saints in order to feel good.  It seems reasonable in principle, though the examples given did not resonate with me.  The discussion in the comments seem to wander off into random topics with little connection to the post itself.
This post presents two arguments: one in favor of diversifying charitable contributions (at least to some extent) and one that presents a simplistic utilitarian argument in favor of individuals finding the most effective charity for their values and donating everything that they are going to donate to that organization.
I don’t have a constant and consistent way of valuing charities—some I value for one thing some for another.  There is no one-dimensional scale (“utilons” is the term used in the effective altruism posts) that makes them comparable.  Multidimensional spaces are inherently unorderable.  Of course, if I am trying to decide on dollar amounts to give, the dollar scale is a simple one-dimensional scale, and so I need to make decisions on mapping incomparable benefits onto a linear scale.  But I don’t have a linear scale of “utilons” to work with (some effective altruists like to use quality-adjusted life years), and the optimal assignment of dollars is quite likely to be a mixture of different charities, each of which is aiming at a different dimension of the space.
This describes some of the groups at an effective altruism conference.  Some of them seem pretty weird to me (which is a strong statement from someone who has lived in Santa Cruz for 33 years).
This post explains why Giving What We Can recommends giving money to Against Malaria Foundation as the most cost-effective charity for reducing death of children and preventing suffering. Givewell also list Against Malaria Foundation as one of their top charities.
My post today is not intended to reach any conclusions—merely to provide some knee-jerk reactions to the effective altruism posts that my son sent me.  I don’t think I accept the premises of the effective altruism movement, and I still plan to give to a diverse range of charities, but I will take advantage of research done by organizations like Givewell and Giving What We Can to send some of my charitable contributions towards organizations they see as effective.

%d bloggers like this: