To split or not to split – part 1

Should we put all our dollars into one charity, or does it make more sense to diversify our charitable investment? The answer to this question depends on both the personal goals that motivate one’s donations, as well as on the specifics of the charities in question.

1. The goal: helping others

Let’s first consider the straightforward case where we want our donations to help as many people as much as possible. Given this goal, there is an argument to only pick one charity, the one we consider to have the highest impact in expectation. The “in expectation” part is important: It might be, although unlikely, that our chosen charity ends up being ineffective, or perhaps even harmful in some unanticipated way. This would be very unfortunate, but assuming that we made the best out of all the information available to us and selected a charity for good reasons, then such an outcome would simply be bad luck – there is nothing we could’ve done better! Sometimes the cards turn against us, even when we make the best decision. By splitting donations across several charities, including some with less than maximal expected impact, we decrease the risk of a worst-case scenario where our entire donation ends up ineffective or harmful. However, at the same time, splitting also decreases the likelihood of the best-case scenario and, more importantly, it decreases the expected impact of our donation:  If we are interested in maximizing our expected positive impact, then splitting is irrationally risk-averse, similar to a poker player shying away from putting their entire stack on the line when they’re slightly ahead in equity.

But what about bankroll management, is there an equivalent concept when it comes to the impact of donations? The situation seems significantly different. There is an important difference between diversification in finance and splitting donations to charities: When the goal is to help others, helping twice as many people is twice as important. That’s all it is about. If what we care about is positive impact, then it would be irrational for us to be risk averse in regard to it.

In finance on the other hand, the goal-unit of the game is money. People usually don’t value money in itself; rather, they value it instrumentally: they value what money can buy them, or the comfort and status that comes with having a lot of it. So this means there is a further step to getting to your actual goal. The amount of useful things you can do with more money doesn’t increase linearly: there are threshold effects and generally there is diminishing marginal utility. For someone with a net worth of one million, doubling this amount is not twice as good as losing everything would be bad. For these reasons, it can be perfectly rational for investors to sacrifice some monetary expected value in order to diversify their risks. It is not the expected value of money they are interested in, but rather the expected utility of the things they can do with the money.

To summarize, “risk aversion” can be rational when instrumental goods such as money are concerned, but being risk averse directly in regard to the unit you care about, e.g. helping other people, cannot. Thus, when it comes to selecting charities with the intention of helping others effectively, splitting is, all else being equal, a bad idea.

However, all else is rarely exactly equal. There are many circumstances under which splitting makes sense, but there needs to be a reason behind it each time. So far, we have been assuming that the first dollar donated to a charity will be used equally effectively as the last dollar. This is probably correct for small donations or for very large charities, but when it comes to donations in the range of five, six or seven figures, it starts to matter a lot whether a charity has enough room for more funding to quickly make effective use of the entire donation. If a charity is short on funding and therefore needs to hold back money for a financial emergency, unable to pursue its most effective interventions with maximum effort, then donations to this charity have a very large impact up to the point where the funding gap is closed. If the charity becomes less effective the more money it receives, because e.g. they have more than enough money to pick the low-hanging fruits and would have to expand and take risks to continue to do what they do, there will come a point where giving further dollars to the second best guess is the better choice.

To conclude, as a default, it makes sense to only pick one charity, but intelligent splitting is the best choice when a charity’s room for more funding is affected. For a more detailed analysis, see this blogpost.

Coming soon: Part 2 on splitting, focused on getting personal feel-good value out of donations.