Of late, I have been quite interested in understanding the motivations and strategies of different donors. My goal is to better understand the entire dynamics of the ecosystem (of donors and donees affecting each other). Ultimately, I hope this will help with better identifying the true marginal impact of donations.

In this post, I describe different donor strategies on the extent to which they separate the questions of "how much" to donate from "where" to donate. In other words, does the donor first decide that they'll donate a specific amount, and then decide how to allocate that, or does the donor first determine whom to donate to, then figure out the amount to donate?

I describe:

• Three broad strategies that donors use (setting aside now versus later considerations): predetermined rule for amount, target selection first, and joint optimization

• The role that saving plays in adding nuance to the strategies

• Philosophical assumptions and beliefs that might justify different strategies. In particular, I am interested in exploring assumptions and justifications for the predetermined rule strategy.

• Questions for readers

Three broad strategies that donors use

Predetermined rule strategy: Use predetermined rule for "how much" and decide the "where" within that budget

A number of donors use prespecified rules for figuring out how much to donate. For instance:

• Jeff Kaufman and Julia Wise: goal is 50% of pre-tax income; their donations are consistent with this goal
• Ben Kuhn: goal is 50% of pre-tax income; donations for 2017 and 2018 have been consistent with this goal
• Haseeb Qureshi: goal is 33% of pre-tax income; donation amounts since 2015 are consistent with this goal
• Aaron Gertler: goal is 10% of pre-tax income

The charities to which they donate may vary quite a bit between years. For instance, Haseeb Qureshi switched from donating to a mix of global health and effective altruism charities, to donating to a mix of AI safety and effective altruism charities.

Beyond individual donors, this strategy also seems encouraged by various donor pledges. For instance:

• The One for the World pledge sets a target of at least 1% of income to donate
• The Giving What We Can pledge sets a target of at least 10% of income to donate
• The Life You Can Save pledge allows the pledge-taker to determine a percentage, but still pegs the percentage to income
• The Giving Pledge sets a target of at least 50% of lifetime wealth to donate

Target selection first strategy: Identify the "where" first, then decide "how much"

I don't know of too many donors who explicitly declare that they are following this strategy, but it's probably an implicit one for many donors (I believe my own donations have largely followed this pattern). This is a more opportunistic approach to donation: the donor finds a charity that seems promising to donate to, then figures out how much the donor can afford to donate to that charity and how much the charity needs. The donor decides the amount based on that.

Donors who follow this strategy will differ from donors who follow predetermined rules in a few ways:

• Their giving is expected to fluctuate more year-over-year
• The total amount that they give will react more to new information

Joint optimization strategy: Decide the "how much" and "where" together

I don't know of donors who explicitly declare that they are following this strategy, but it's probably an implicit one for many donors. I believe Peter Hurford and Patrick Brinich-Langlois fit this strategy.

For instance, this could be a donor who, periodically, looks at their finances to identify the general range of money that person is willing to donate. This could, however, be a pretty wide range (potentially with zero as the lower end). The donor then looks at donation opportunities. Based on how compelling the donation opportunities are, and the room for funding of those opportunities, the donor decides which ones to donate and how much.

In other words, the amount the donor donates can vary based on the strength of the opportunities (the stronger the opportunities, the more the donor will donate). The opportunities also compete with one another; particularly good opportunities may displace others.

The role that saving plays in adding nuance to the strategies

Charity, consumption, and saving

If it were not possible to save from one year to the next, the problem of allocation would simplistically boil down to deciding how much of one's money to allocate to charity versus how much to allocate to consumption. Donors following the predetermined rule strategy would be determining this allocation beforehand, and only after that allocating within the charity and consumption buckets.

However, the ability to save now makes it possible to defer both charity and consumption to a future year.

With saving introduced into the picture, we could adopt either of two positions:

• Allocate the saving into "saving for charity" and "saving for consumption" at the time of saving (i.e., within the year itself).

• Allocate the saving to a general, fungible pool of saved money, that could be used later for either charity or consumption.

Allowing saving in the predetermined rule strategy

Some people who pledge a percentage of their income (i.e., use a predetermined rule strategy) may prefer not to donate right now, because they didn't find good donation targets, or want to investigate or wait longer. To meet their goal of donating, they may donate into a donor-advised fund or other vehicle that can hold the funds till a future donation. Ben Kuhn did this in 2017 and 2018.

Donor-advised funds present tax benefits if they are restricted to only donate for charitable purposes. Some donors who want the flexibility to donate to things not recognized as charitable in the tax code may therefore use another vehicle (or perhaps just store the money in personal savings), thereby not being able to take immediate tax benefits, while still clearly demarcating the money as "only to be used" for charity.

The role of saving for other strategies

For people not using a predetermined rule to decide how much to donate, they can:

• allocate some of their savings explicitly to a donor-advised fund, to avail of tax benefits,

• keep the rest of their savings personal, so that the money could opportunistically be used in a later year to finance donations, but doesn't have to be used this way

Many of the wealthy people who ultimately signed the Giving Pledge followed the second strategy, and ultimately decided, after accumulating a fair amount of wealth, that they could afford to donate a lot of it.

Philosophical assumptions and beliefs that might justify different strategies

The strategy that a donor follows is a deeply personal decision based on a number of personal financial factors, values, and preferences. So, we cannot draw definite conclusions about the donor's broader philosophy from whether the donor follows a predetermined rule strategy. Nonetheless, some philosophical assumptions make some strategies more justifiable than others.

For people who trust themselves and want to retain maximum flexibility and don't care about simplifying their lives, a joint optimization strategy seems best

Joint optimization, where the donor decides "how much" and "where" together, seems to be the strategy that allows the maximum use of information in making decisions. A donor can take into account developments both on the charity side (e.g., whether the charities they are interested in donating to continue to be worthwhile, and whether they still have room for more funding) and on the consumption side (e.g., whether the donor had unexpected personal expenses, or loss of income).

A predetermined rule strategy may make sense for couples because it avoids haggling and may promote harmony

Disputes over allocation of money can affect a couple's relationship. If the two members of a couple have different feelings over how much to donate, it may be most effective to sit down and figure out a rule beforehand that represents a reasonable compromise for both sides, rather than try to negotiate the boundary every year. The couple may also decide how much each member controls of the charity allocation (just as they may have similar rules on the consumption side).

[Although I can't find a link to where I read this, I believe the 50% figure that Jeff Kaufman and Julia Wise came up with is an example of this sort of compromise in action.]

A predetermined rule strategy may be viewed as beneficial in dealing with time-inconsistency issues (people who don't trust their future selves)

The header explains it. Further, saving for charity into a donor-advised fund (that requires the money to be spent on charity) is a similar precommitment to hedge against a future, less-charitable self.

A predetermined rule strategy might be good for norm-setting and community-level predictability

When donors precommit to donating a specific percentage of their income, this sets a norm for other donors. We can think of the pledges, such as the One for the World Pledge, GWWC Pledge, and TLYCS Pledge, as strengthening these norms by giving them more explicit labels. This norm may encourage other donors to donate more than they otherwise would.

Also, when there's a significant amount of money flowing into charities' coffers from the predetermined rule, charities, and the community as a whole, know that some amount of money can be counted on. This is true even if individual charities don't know for sure if the money will go to them -- at least they know that there's enough of a pool of money that's committed to charity that they might hope to get. If the amount of money that's committed is more than enough to meet the budgets of the charities that the donors are considering, each charity can be somewhat confident that they'll get at least some of it (because the other charities' room for more funding would be exhausted before the donors had spent out all the money they committed).

Predetermined rule strategies make sense for people who believe that their opinion of the "right" amount to donate to charity isn't likely to change with new evidence or developments

Having a predetermined rule doesn't seem to hurt much if the donor believes things in this general cluster about giving opportunities:

• The donor's opinion of what cause area matters the most won't change much, or the right amount to donate won't change much even if the donor's opinion changes.

• Cost-effectiveness estimates in the area won't change much, or the right amount to donate is not sensitive to cost-effectiveness estimates (see below).

• The entry or exit of other major donors to the cause areas or charities of interest is unlikely, or such entry or exit won't affect the right amount to donate

Also, it makes sense if the donor believes things in this general cluster about personal finances:

• The structure of the predetermined rule (whether in terms of income and wealth) makes sense across different possibilities for the person's wealth, for instance, if the person has to leave a job, or has some other unexpected accident or other expense (such as a wedding, or kid). [From this viewpoint, donating a percentage of income may not make sense; donating a percentage of income-above-a-threshold may be more sustainable]

Of course, there are degrees of commitment, and "predetermined rule" doesn't necessarily mean "predetermined rule for life". So, a more accurate way of framing this is that predetermined rule strategies make sense if the donor doesn't expect changes to the "right" amount to donate within the timeframe that the donor has committed to. In other words, the right amount to donate shouldn't be too fluid. (However, it's different for people who take lifetime pledges, which might be why there is some criticism of lifelong pledging from Buck Shlegeris and Andrew Critch).

My past puzzlement at insensitivity to cost-effectiveness estimates

A few years ago, I had trouble understanding this belief that the "right" amount to donate doesn't change much with cost-effectiveness estimates. I even wrote a long comment on the GiveWell blog about this. Later, I asked the question on the Effective Altruism Facebook group, where there were a number of interesting replies that resolved much of my puzzlement.

Questions for readers

My goal with posting this to the Effective Altruism Forum is partly to provide a concrete public record of my thoughts. I can link to this post and build upon it as I further explore the dynamics of donations.

However, I'm also interested in hearing thoughts from people on the Forum, many of whom are donors of the sort I try to understand in the post.

Some questions I'm particularly interested in hearing thoughts on:

• What's your strategy as a donor? Does it fit within the three broad types of strategies I describe in the post?

• Are there other aspects of these strategies that you consider important, that I missed? For instance, are there some other pros and cons of predetermined rule strategies?

• My impression is that the effective altruism (EA) community uses predetermined rules more than other donors, who come closer to "target selection first" or "joint optimization" in their strategy (i.e., they are more likely to be moved by specific events or developments in their decision of how much to donate). Am I right, and if so, is this good?