Manifold Markets lives in an interesting halfway point between the for-profit Silicon Valley and non-profit Effective Altruism spaces. We’ve received all kinds of funding: restricted grants, unrestricted grants, angel and VC checks.
At first, it might seem that grants are much better than investment — you get money for nothing! But in practice, I’ve felt that it’s not so clear-cut as that. Some reasons that the nonprofit grants feel less empowering than just investment money.
- Grant money feels restricted and thus meaningfully different/less valuable than investment money
- Think of the difference between “money for a scholarship” vs “cash in your wallet”
- One of EA’s favorite charities is GiveDirectly, premised on the idea that people can make pretty good decisions about spending if they just have the money
- Example: When LTFF gave us $200k we… basically just sat on it. Vaguely thought “oh, maybe we’re supposed to set up a longtermist prediction tournament now like Metaculus’s?”
- Example: Accepting FTX FF’s $500k took a couple extra back and forths with lawyers to draft a contract on how we could spend it (and then a detour into a grants portal issue), vs just taking $1m investment via wire
- Grant money leads to making guesses about how your funder wants you to spend your money, in a low-feedback, slow loop with unclear objectives
- Especially bad the more levels of indirection/regranting are happening
- Example: at Future Forum, we ran a pitch contest to see what projects to fund with $10k. I grabbed Ela Madej and Caleb Prikah (who are wonderful and amazingly talented people) to judge. This lead to:
- Ela and Caleb (judges), guessing how…
- Isaak Freeman (Future Forum president), guessing how…
- FTX Future Fund, guessing how…
- Sam Bankman Fried would want his money to be allocated
- We spent 90+ minutes judging the 20 entries, in large part because we were unsure what exactly the funding criteria was.
- Investors get to say “I think this will make money”
- Consumers get to say “I value this thing at more than the cost of money”
- But when you’re spending funding from other sources, you now have to invent a model of your funder’s preferences. “Does <my funder> think this would be a reasonable use of money? Could I defend that decision?”
- Having different colored dollars makes financing things and accounting a huge pain; beware (more than nontrivial) inconveniences
- For starters: can’t use Stripe Atlas to get incorporated; which also means no access to Stripe Opal, which makes all kinds of getting things paid for much much better
- Sidenote: Stripe Opal is amazing! All your money sits in one place! Issue virtual credit cards to employees or just random with a couple clicks! 1.5% cash back on everything!
- Financial infrastructure (actually moving money around) in the nonprofit space is held together with duct tape and spit
- Then: Have to track how the money we’re spending is allocated, worry about whether it fits the bar
- Especially in an early stage startup: plans change!
- Example: Manifold got $500k to fund charity prediction markets. Our first experiment with that raised $8k; and if we match that with another $8k it’s $16k total, a drop in the bucket compared to what we raised. So…
- We could just keep running it, for a long time?
- We could run other kinds of experiments?
- E.g. donor lotteries, top charity giveaway, buy up impact certs?
- We could return the money?
- But feels like giving up agency/power/ability to shape the world, my hopefully-somewhat-accurate view is that on the margin Manifold could spend $500k more effectively than FTX Foundation
- The US government subsidizes 501c3 grants in the tax code, but it comes out to something like a 70% match at most
- But my guess is that restricted grant money is less than half as useful as actual straight cash money, so 1.7x * 0.5x = 85% as good as actual cash
- Where does 70% come from?
- If you have a $0.01 stock that appreciates to $1000, and then donate it to a 501c3: You get a ~$1000 tax deduction; at a 37% marginal tax rate = $370 less to government. So you keep $370, charity gets $1000 = $1370 of benefits
- If you had sold it instead and then handed it to your friend as cash, you first pay long-term capital gains, e.g. 20%, and then your friend receives 80% = $800 of benefits.
- EA talks about order of magnitude differences between how charities spend money - if so, then a 70% government match should be one of the less important factors into where you donate
- Charity founders and employees sacrifice their earnings and their time for the charity
- Leads to a misalignment between the charity’s mission and personal happiness; leads to burnout
- And the fact that people have to make such sacrifices mean that talented people with a lot of options go work in Silicon Valley or Wall Street instead
So: If you’re rich like SBF, consider not worrying about whether a cause has nonprofit status, and just sending them a check