Invest now, give later?

The power of exponential growth seems to make a compelling case for effective altruists to delay their donations. An average 5% return on investment (ROI) would turn one dollar into ten in 50 years time. If saving a life costs $2000 now and similar opportunities will exist in the future it would cost just $200 to save a life in 2062 – a relative bargain! Sadly things aren’t so simple. Whether we really should delay depends on specifics of the activities we are funding and difficult predictions about the future. Here I’ll summarise the most important uncertainties as a roadmap for future posts.

Our goal can be summarised as choosing the time t which maximises

(1 + Return on investment)^t × Cost effectiveness of donation_t 

× Probability of donation actually being made_t.

Unless you are a multimillionaire, the relevant expected ROI is the highest one available without regard to risk. Giving $2m will do about twice as much good for the world as $1m, so to maximise your expected impact you should just maximise your expected donation. Note that if your favourite charity would be able to use money now to attract donations at a rate faster than you expect your investments could return profits then donating would have to be better.

The second and more challenging issue is how cost effective your donation will be in the future relative to now. If you thought basic health would be the optimal cause this would involve anticipating things like

• the extent of poverty
• the cost of delivering health services
• how much other donors will be funding the low hanging fruit.

The last point is especially relevant for those like me thinking of funding existential risk reduction because a few billion from governments or philanthropists could make a big impact on the value of further funding in that area.

In evaluating cost effectiveness we must factor in that any good charity will have impacts that propagate through time and so offer its own ROI. For instance, combatting contagious diseases now rather than in 2062 should lead to fewer people becoming infected in the meantime and so result in a richer and healthier population in 2062. Similarly, spending on existential risk reduction draws attention, money and researchers to that issue. Giving now leaves your donations more time to have this snowball effect during the window of greatest extinction risk.

On the other hand delaying leaves you more time to identify cost-effective targets for donations. Personally, I am investing rather than giving mostly because I expect groups like 80,000 Hours to give me a much better idea of how to best reduce existential risk within the next decade.

Finally you must assess the risk of your donation never being made, for example due to a catastrophe which eliminates your savings. If you can’t bind yourself through a trust fund, you must also worry about changes to you or your life which result in you deciding not to give.