Hide table of contents

(crossposted to https://astralcodexten.substack.com/p/impact-markets-the-annoying-details)

Impact markets (aka impact certificates, retroactive public goods funding) are a frequently-mooted EA idea to create "a VC ecosystem for charity".  Instead of funders (eg large foundations) prospectively paying teams to do charitable work, they would retrospectively "buy" moral impact from successful projects. Then investors would "buy equity" in charitable proposals, funding their work in exchange for a portion of the eventual profits. For more information, read Paul Christiano’s or Vitalik Buterin’s in-depth explanations.

I'm interested in using impact markets for the Winter 2022 ACX Grants round, which would require turning this sketch of an idea into a concrete proposal. Below I'm listing some of the open questions about details of an impact marketplace, and my preliminary thoughts on each.

Thanks to everyone who reviewed and made suggestions on this draft. I'm defaulting to keeping reviewers anonymous, but if they prefer to be named I'll add them here later.

1: What Is The Basic Format Of The Market?

A: Simple Retroactive Funding

Some charitable funder announces they will give money to good things that they like. If somebody does a good thing, the funder might give them money later. There isn’t necessarily any investing, certificates, or tokens.

Advantage: This is very simple, and can be done right now: in fact, a team is already doing this for good Effective Altruism Forum posts.

Disadvantage: This is a good start, but really just equivalent to giving prizes to good EA Forum posts. Absent someone else setting up an auxiliary structure, it doesn’t help useful projects get funded. At best, it encourages a few marginal people to do cheap things they might have done anyway.

B: Block Impact Certificates

A charity offers a single impact certificate representing a project. For example, if they need $1 million to cure malaria in Senegal, they sell a single certificate representing that project for $1 million. Then, when an final oracular funder decides the project is worth $5 million, they give $5 million to the holder of the certificate.

Advantage: Again, it's simple.

Disadvantage: Someone needs to have the entire $1 million in order to fund the project. There’s no way to split the costs, which makes it worse than eg stocks. When Tesla needs $1 billion in new funding, it doesn’t just ask billionaires. It sells (eg) one million shares of stock for $1000 each, and then people with much less than $1 billion can bet on their success.

C: Fractionalized Impact Shares

A charity offers splits its offering into many shares or tokens. For example, if they need $1 million to cure malaria in Senegal, they could sell 10,000 tokens for $100 each. Then, when a final oracular funder decides the project is worth $5 million, they compensate each token holder with 1/10,000th of the final reward, so $500.

Advantage: this allows people with much less than $1 million to invest, and creates a secondary market in tokens which people could probably do interesting things with.

Disadvantage: Suppose the charity wants $1 million because it will take $500,000 to build a medicine factory, and then they want to make $500,000 worth of medicine. If only $400,000 in tokens are sold, they can’t even build the factory, and the whole project is worthless. So an investor wouldn’t want to invest $400,000 unless they were sure that someone would produce the other $600,000. But since nobody can know that, people might not invest. Or they might invest, then get very angry when their investment turns ends up useless.

D: Fractionalized Impact Shares With Assurance Contract

As above, except instead of buying a token, investors commit to buy a certain amount of tokens if all tokens get bought. For example, Alice might say “I commit to buying $400,000 of tokens, if eventually we get to $1 million”. Bob might say “I commit to buying $300,000 of tokens, if eventually we get to $1 million”. Carol might say “I commit to buying the final $300,000 in tokens”, and then everyone’s commitment “activates” and they all have to buy the tokens. You might be familiar with this as the mechanism behind Kickstarter campaigns.

Advantage: This solves the problem discussed above.

Disadvantage: It’s more complicated - but I think most people capable of buying certificates at all should be able to understand this, especially if it’s explained well on the marketplace’s website.

My thoughts: The fractionalized shares with assurance contracts seem clearly the best. Some reviewers noted that in real venture capitalism, VCs are able to talk to each other to coordinate funding without a formal assurance contract. I think our market will involve smaller and less savvy players (at least at first), which would make the contract more useful.

2: What Is Being Sold?

A: Full Credit For The Impact Of The Project

This is the traditional solution. If I cured malaria, and you previously bought an impact certificate equal to 100% of my “equity” in the project, then in theory you have the right to say you cured malaria, and I don’t.

Advantage: Elegance, everyone knows what's going on.

Disadvantage: It’s hard to enforce from a common sense point of view. “The credit for” doing something partly takes the form of accurate costly signals about my personal characteristics, which can't be transferred. 

For example, suppose I cure malaria by inventing a novel antimalarial drug. Part of “getting the credit” for this is people understanding that I am a brilliant biologist. They might use me as a role model for young scientists, ask my opinion on important science policy questions, or make me chair of the biology department at Harvard. There's no credible way to sell these benefits - just because I sell you the impact certificate for curing malaria, doesn’t mean Harvard would be willing to make you the department chair instead.

This is also going to make a lot of people mad when some photogenic kid from Kenya sweats and toils to builds a water pump for his community or something, and then some billionaire says no, I built that water pump.

B: Credit For Funding The Project

I think this is what Ben Hoffman writes about here.

Both the people who do the project (ie the founder who gets people to build the bednet factory, the scientist who discovers the malaria cure) and the people who fund the project deserve some share of credit in the success. In this model, impact markets would distribute the funding-related credit only. That is, I might pay $1 million for the impact of a project to cure malaria; when they succeed, I can sell it to a big foundation for $5 million. So long as I have the shares, I can claim that I funded the project; once the foundation has the shares, they can claim this (even though causally they were not the ones who give the project the funds). In either case, the founders and scientists retain the credit for founding the project and doing the science.

C: Rules Around Equity

We might (for example) say that in an average project, the actual team involved deserves half the credit, and the funders also deserve half the credit (or some other distribution). Then we might have a norm that the team can sell half the equity (representing the funding credit) but not the other half (representing a sort of inalienable moral credit that attaches to their hard work). 

Advantage: Matches common sense

Disadvantage: More complicated, less elegant, founders can't sell as much equity.

My thoughts: There are good arguments for all of these. See below for more.

3: How Should The Market Handle Projects With Variable Funding Needs?

Suppose a charity says “For $1 million, we can cure malaria in Senegal. But if we had $10 million, we could cure malaria in all of West Africa!”

This is a bad match for the token impact certificates with assurance contract discussed above; do we active the assurance contract at $1 million, or at $10 million?

A: Activate The Contract At $1 Million, Then Keep Selling More Tokens

This has the following problem: suppose that you are an investor who believes that the plan to cure malaria in Senegal will succeed (produce $5 million in value), but the grander plan to cure it elsewhere in Africa will fail (produce $0 in value). You buy the $1 million in original tokens, activating the contract. Then other, less savvy investors buy the $9 million in remaining tokens.

You turn out to be right about everything. The first project succeeds and produces $5 million in value; the second project fails and produces $0. A final oracular funder pays $5 million, distributed evenly among all token-holders. Your 1/10th of the total token supply gets you $500,000. So even though you were right about everything, you have turned a $1 million investment into only $500,000, and lost half your money.

Since this was your original prediction, you will never invest your original $1 million, even though you correctly predict that will be good!

In other words, under this method, proposing a grander secondary plan might prevent people from investing in your primary plan, even if your primary plan is good, which seems like a bad feature.

B: Activate The Contract At $1 Million, Then Sell A Different Token For The Other Plan

IE the charity sells $1 million worth of “cure malaria in Senegal” tokens. Then, if that contract activates, they offer another $9 million of “expand the plan to all of West Africa” tokens. Everyone knows which kind of token they are buying.

Advantage: This avoids the problem mentioned above.

Disadvantage: More complicated.

One extra advantage: it can handle different levels of needing assurance contracts. For example, if the plan is “$1 million to start off, and then for each $1 we get after that we can produce one additional bed net”, then you don’t need any assurance contract on the secondary tokens. If it’s $1 million to start off, and then for each additional $500,000 we can open a factory in another African country”, then maybe there should be an assurance contract on each extra $500,000, or - if this would get too complicated - on the full $10 million.

My thoughts: Reviewers brought up that real venture capitalists solve this problem by having multiple funding rounds where they sell off different parts of their equity. For example, you might sell off 10% in the first round, achieve some amount, and then sell off another 10% for your next big expansion. I am nervous about this because I want to be able to pitch impact certificates to non-savvy people who may not be able to make good decisions about equity.

One possible solution reviewers brought up would be to offer a “standard deal” like Y Combinator does. If you don’t want to think about impact certificates, you take the standard deal, convert some pre-arranged percent of your equity to certificates, and never think about it again. Then if you want to do something more complicated later, you can look into your options further.

4: Should Founders Be Getting Rich Off Impact Certificates?

A: Yes

Jeff Bezos founded Amazon, sold off some of his equity for seed funding, and kept some other equity for himself. As Amazon grew, the equity appreciated in value, making him very rich.

We could imagine impact certificates working the same way. I start a charitable project, sell 50% of the impact equity for seed funding, and then succeed. My investors can sell their impact certificates to a final oracular funder for more than they paid, making a profit. But I can also sell off my remaining equity to the same funders, making me a profit, potentially a very big one. For example, if I have a project that uses $1 million in funding to produces $5 million of value in the form of malaria cures, I can sell 50% of my equity to investors to get the $1 million I need. When it succeeds, a final oracular funder can pay $2.5 million to my investors, and an additional $2.5 million to me. Now I’m a multimillionaire for coming up with such a great charitable scheme.

Advantage: This encourages people to start charities; the more good they could do, the more it encourages them.

Disadvantage: This redirects at least some of “the surplus” from charities and beneficiaries to founders. That is, right now if a charitable project that costs $1 million produces $100 billion in value, then the founder gets some normal salary, and the funding charity and its beneficiaries have gotten an amazingly good deal. This is good for charities and beneficiaries. But if we allowed founders to capture more of the surplus, then maybe the founder would end up with $50 billion, and the charity and its beneficiaries would only get $50 billion, which seems much worse for them. Given that our goal here is to help beneficiaries (and by implication charities, since any money charities save goes to the beneficiaries eventually), this seems pretty bad.

Or to put it another way, if someone makes $50 billion off charity and uses it to buy a yacht, then some of your charitable donations are going to yachts, which seems worse than them going to feeding the hungry or curing the sick.

Or to put it yet another way, if founders generally kept 50% of their equity, all charitable projects would be twice as expensive - the normal $X to fund the project, and then an additional $X to pay off the founder. If founders generally kept 75% of their equity, all projects would be 4x as expensive, and so on.

And to bring things closer to home: a reviewer mentions an certain excellent AI safety researcher. Given how much money is in this field, and how few excellent researchers there are, the big foundations might plausibly consider this person to be worth $10 million per year (that is, they would be happy to fund a program that spent $10 million per year to create another researcher of the same quality). But right now this person just earns an ordinary six figure AI engineer salary. If impact markets were “a success”, they might have to more than dectuple the amount they pay this person, with no change to their output.

B. No

Impact certificates could be used only to fill projects; need for funding, the same way grants are used now. Social norms would mandate that founders only ask for as much money as they needed for their project. This might include paying themselves a fair salary, but they wouldn’t be getting rich.

Advantage: Avoiding the problems mentioned above.

Disadvantage: There might be some situation where an especially brilliant person considers for-profit vs. charitable work and goes with for-profit because they won’t be able to internalize the value produced by their charity.

My thoughts: Extremely split on this.

Part of the point of impact markets is to spend some of the surplus on incentivizing more charity to exist.

There are some innocuous ways to do this. For example, suppose that there are an unlimited number of malaria charities that want $1 million, and 25% of those will succeed and produce $5 million in benefits, with the other 75% producing nothing. So one way to produce $5 million in expectation is to spend $4 million funding four of these charities. Another way is to buy an impact certificate for a successful project from an investor. So suppose a savvy investor is twice as good at picking projects as the final oracular funder; she can pick a winner 50% of the time. So she pays $2 million to fund two projects, producing in expectation one successful impact certificate representing $5 million in benefits. Then it would seem that the investor and final oracular funder should agree to a price somewhere between $2 million and $4 million for the certificate; whatever number in this range they choose, both will make a profit over their next-best option. They can use normal negotiating tactics to figure out where in that range they fall, but nobody should lose money relative to the no-impact-certificates world.

But if there are projects that everyone recognizes as good, then an impact marketplace will shift the surplus from final oracular funders (who are forced to pay prices more commensurate with their benefits) to someone else. If we let founders get rich, it shifts the surplus to founders. If we don’t let founders get rich, it shifts the surplus to fast investors who may not have added any value - that is, to the first people who snapped up the obviously-underpriced certificates. See Section 7 for more on this problem.

I don’t have a good solution to this latter issue other than either not using impact markets, or allowing founders to get rich. I sketch a hack-ish solution in Section 11C.

5: How Do We Kickstart The Existence Of An Impact Market?

How do we get an impact market to exist? Investors will only buy impact certificates if they expect that a final oracular funder will buy them, but we'll have a hard time convincing them when this has never happened before.

A. Retroactive Prizes First, Then Impact Certificates

Someone who wants to be a final oracular funder gets in the habit of giving out retroactive prizes. For example, out of the blue they find someone who helped fight malaria, and award them $1 million. They make it clear that if someone sells an impact certificate, they will give the prize to the certificate-holder rather than the original charity. Eventually this happens enough that people grow to expect and predict it, and they can sell impact certificates around it.

Advantage: It's simple and would probably work.

Disadvantage: It takes a long time, and a lot of granting prizes to projects that don’t need them (ie they would have been done anyway even without the prizes). Also, it kind of elides over the part where the impact certificate market gets set up. Is someone else doing this behind the scenes? What if they don’t?

B: Committed Pot Of Money (aka The Original ACX Grants Plan)

In the original ACX Grants plan, I solicit grant proposals, and probably get ~500 like last year. I promise to be Final Oracular Funder and give out prizes in one year’s time, and I commit a specific sum to this purpose (let’s say $1 million). I link everyone who’s interested to a market where people can buy shares in the project, and they do this, competing for my $1 million.

Advantage: This sure does set up a marketplace.

Disadvantage: Investors will have to predict (and maybe be sabotaged by) the behavior of other investors. If people only buy a total $1 million in tokens, then on average every investor will break even. Not every investor will in fact break even, because the whole point is that money will flow towards the good investors and away from the worse investors. But on average they’ll break even.

But suppose you invest when there are $1 million of tokens sold, thinking that you are a slightly-better-than-average investor and so can probably do better than break even. Then later some other people come in and buy $9 million more worth of tokens. Now there are $10 million in tokens chasing $1 million in prizes, and on average each investor will only get 10% of their money back. Although some great investors could still come out ahead, it would be an unpleasant surprise to start out expecting to break even, only to learn later that you should actually only expect to get 10% of your money back.

C: Pot Of Money Plus Cap

As above, except that the amount of tokens that can be bought is some multiple of the amount of money in the pot - for example, only $1 million in tokens can be bought, so everyone knows they can on average expect to break even.

Advantage: Solves the problem above.

Disadvantage: It’s pretty limiting. Also, investors don’t have a god-given right to break even. Also, if every project produces massive value, some of them will necessarily be under-rewarded - and knowing that makes them necessarily under-invested-in.

D: Market Kickstart Plus Semi-Unlimited Pot Of Money

Here we would kickstart the supply side of the market with a set of grant proposals as above, but the final oracular funder would be a very large charitable organization (eg Open Phil or the Future Fund), operating at a scale so far beyond the size of the market that they could commit to funding every project that was good enough to deserve it.

Advantage: Solves the problems above.

Disadvantage: It places a lot of the buy/don’t-buy decision in the hands of a single oracular funder. For example, suppose that OpenPhil agreed to do this, but a few years later they got a new CEO who was more conservative than the previous CEO, and only funded about 2/3 as many projects. Now the chance of getting a return on your investment has gone down by 2/3 for a reason unrelated to its fundamentals. Or suppose someone invested in a malaria project, but OpenPhil pivoted away from tropical diseases and now there was no hope of getting any return. Seems like the sort of uncertainty that would discourage people from investing.

E. Market Kickstart Plus Multiple Unlimited Pots Of Money

This is more of the ideal. Have so many people making open-ended commitments to buy impact that changes in one or two of them don’t matter as much. This is how the stock market works: you can buy a stock with good fundamentals because there are so many people who buy stocks that you can abstract out concerns like “Warren Buffett is pivoting away from that industry”; there will always be someone around to offer you a fair price.

Disadvantage: This seems less like a way of starting an impact market, and more like a final desired end state.

My thoughts: I think the lower-down options are clearly better, but they’ll depend on how many people are willing to commit giant pots of money to this project.

6: Should The Market Use Cryptocurrency?

A: Centralized impact certificates and fiat currencies

We could have a standard online marketplace where people buy impact certificates (block or fractionalized) from the site with dollars (or other fiat currencies). The site would track all purchases, and in the very unlikely event that the site ever got erased, everyone would remember it anyway (eg the charity would remember who funded them).

Advantage: Easy to set up and use.

Disadvantage: We don’t get the “coolness” boost of using crypto.

B: Impact NFTs/token and cryptocurrencies

Or we could have a crypto marketplace, like eg OpenSea’s NFT marketplace. The impact certificates could be NFTs (if block) or cryptographic tokens (if fractionalized), and people could buy or sell them using cryptocurrency.

Advantage: We're asking people to spend thousands of dollars on meaningless tokens with no direct personal use value, this is usually a hard sell, but people get excited about this kind of deal when you call it an "NFT". Also, many cryptocurrency users have lots of money and like investing it in experimental projects. Also, EA is friendly with FTX and they might be able to help set this up.

Disadvantage: This is harder for the technical people to set up (unless they’re FTX), harder for users to use unless they’re already familiar with crypto, and has a reputation for including scams.

I’m not concerned about environmental issues because we can just use a blockchain without those problems, but other people who don’t understand crypto might not understand that, and it might be bad PR for a charitable project.

My thoughts: Crypto-literate reviewers say that crypto exchanges take fiat and handle transactions within themselves in a non-crypto way - but then the crypto exists in case someone wants to bring it to a different exchange. This seems like a best of both worlds scenario.

7: How Should The Market Price IPOs?

A: Sell Certificates/Shares for Exactly The Amount Needed To Fund The Project

For example, if a project needed $1 million, it might sell 10,000 shares and price them at $100 each. Whoever was the first person to pay $100 for a share would get it. If all the shares sold out quickly, and someone else wanted one later, they would have to buy it from an existing owner (probably at a higher price).

Advantage: It’s easy to price shares, and project founders will get exactly the amount of money they want.

Disadvantage: All surplus goes to fast people. For example, suppose there is a $1 million project that will create $5 million in value, and everyone knows this. The first few people to see the shares will buy up all them at $100, then immediately sell them for $499 (or whatever), making a huge profit without doing any work.

Note that this isn’t just a question of “the project should have priced their shares higher”. It might be common knowledge that this project with $1 million in costs will create $5 million in value, but if we're asking project founders to honestly report their funding needs, then they will honestly say they need $1 million. In this model, there's no way to get the cost to $5 million without the founders lying.

B: Auction Off The Certificates/Shares

Certificates/shares would get auctioned off for some period of time, kind of like how NFTs are sold now.

I don't know how to use an auction process with fungible shares, although someone has probably solved this problem. A very easy version would be that the share price starts at $100, and if I try to buy 25 shares for $2500, instead I begin an auction with that as the starting price, which lasts some amount of time.

Advantage: Founders and projects deserve the surplus more than lucky/fast investors.

Disadvantage: Might be technically harder. If the founders get the surplus, then they might get rich, which has advantages and disadvantages (see Section 4). The project getting the surplus would be less morally fraught, but not all projects have room for arbitrary amounts of extra funding.

My thoughts: Split on this one. I think the answer will depend on how we resolve the earlier question, “should founders be getting rich off impact certificates?” If yes, then clearly auctions are the way to go (or implied auctions where founders sell off impact at a high price). If no, I can’t think of a good solution to the “fast people get rich” problem.

8: How Should The Oracular Funder Buy Successful Projects?

A. By Buying A Certain Number Of Shares

The same way as in the usual stock market. An investor who owns (for example) 10% of a project can negotiate a sale to the funder.

Advantage: Everyone understands stocks, and this is very simple.

Disadvantage: oracular funders might be semi-monopsonists who can play investors off against each other. This sounds like the kind of process that gets an impact certificate-related fight featured on Money Stuff.

B. By Creating A Funding Floor

Vitalik Buterin’s idea. A final oracular funder states that they value the project at (let’s say) $5 million, and as long as they continue to do so, they will buy any share in the project for $5 million/percent of the project.

Under this model, somewhat might “apply” to have a funder value a certain project. The funder will come up with a valuation, and then anyone (including the original applicant, or other shareholders) can sell their shares at that price.

Advantage: Makes people less concerned about monopsony.

Disadvantage: Might be slightly more complicated.

9: What Should The Final Oracular Funder’s Decision Process Be?

How does a final oracular funder decide how much money a successful project is worth?

A: Fund Based On Regular Charitable Decision-Making Procedure

Presumably final oracular funders already have grantmaking procedures, and can ask themselves questions like “How much would we have paid to prospectively fund a project like this one, knowing that it would succeed as much as it did?”

Advantage: Easy (?)

Disadvantage: Opaque to everyone except the funder. If you’re paying $1 million for a project to cure malaria in Senegal that you think has a 50% chance of success, it matters a lot to you whether the oracular funder would pay $500,000 vs. $5 million upon successful completion. But there might not be enough of a published history of grants for you to know this, and it might not even be a question with a fixed answer: it might depend on which of several different grantmakers got it, or on whether they were in a good or bad mood that day, or any of several other confounders.

Although in theory oracular funders and investors are aligned insofar as the funders want to encourage future projects by paying fair prices for current ones, the alignment might break down - for example, the particular employee charged with evaluating how much money to pay for a successful project might be rewarded for “keeping costs down”, but not be punished for this having a vague long-run negative effect on impact markets. Even if this isn’t true, investors might worry that it’s true, and so under-invest.

One reviewer suggested that final oracular funders go through a long list of past grants (for example, the 2021 ACX Grants) and publish how much money they would have paid for each; future investors can check the list and get a rough idea of what gets how much money.

B: Fund Based On Some Sort Of Clear, Quantified Schedule

For example, “we will pay $1000 for every life saved”.

Advantage: this is clear, objective, and removes investor uncertainty.

Disadvantage: Only works for highly legible projects. How do you quantify the lives saved by lobbying to prevent future pandemics (especially if the lobbying is successful, and a counterfactual pandemic quietly fails to happen)? How do you quantify the success of a CFAR-style educational intervention to "raise the sanity waterline"?

These aren't an unsolveable problem: final oracular funders could easily say "We will pay $500 for each person who has been successfully educated about rationality". But unless they specified every possible outcome, this would shift incentives in favor of commonly-considered, easily-specified outcomes. And outcomes they do specify risk Goodharting - a self-interested founder with these incentives will "educate" people the smallest amount possible to still qualify for the award.

C: Fund Based On Some Kind Of Very Vague Schedule

For example, “we will pay $500,000 for what we think of as an important insight into AI alignment, similar in magnitude to X or Y previous insights. We will pay $100,000 for what we think of as a minor but still useful insight, similar to Z previous insight.”

Or “we will pay $500,000 for what seems like a major decrease in malaria prevalence in this country. We will pay $100,000 for what seems like a smaller but still measurable decrease.”

Advantage: Strikes a balance between the previous two methods. 

Disadvantage: The oracular funder won't be able to predict all the different good things that might come out of charitable projects, and so it will either have to commit to making some things up post facto (as in A) or subtly discourage people from doing unusual things that can’t be measured on the usual metric (as in B).

D: Offer Consultations On Funding Schedules

The final oracular funder could have an officer who takes calls from potential investors and explains the amount of funding they would get for different levels of success. For example, you might say “I’m funding a project to save the endangered blue leopard, how much is that worth to you?” and they could answer “$500 per leopard saved, plus an extra 200% bonus if you save more than 1,000 of them”.

Advantage: this makes investment payoffs clear without preemptively demanding too much legibility.

Disadvantage: requires more work from final oracular funders, and is especially vulnerable to Goodharting ("But you promised you would pay us $500 per leopard!")

E: Have So Many Potential Funders That It Converges To A Market Price

This whole system is partly predicated on the idea that funders will like knowing that they funded useful things - both because they are genuinely altruistic, and because they want to gain status and credibility by showcasing their charitable successes. If there are many oracular funders willing to buy impact certificates, then (absent collusion) they could potentially converge on a “market price”. 

That is, if the impact certificate cost an absurdly low amount, like $1, then people could buy “curing malaria in an entire country” for $1, and I (who would love to be able to brag about saving an entire country at parties) would outbid them and buy it for $2. But if the price were absurdly high - let’s say $100 billion - then nobody would buy it, because you could do more good some other way with that much money. So buyers and sellers will converge on a market price for impact certificates, people could mostly predict what that would be, and it would be fair in some sense.

Advantage: this is how we do everything else under capitalism, and it usually sort of works.

Disadvantage: This will require many independent funders and high liquidity, which might not be realistic while kickstarting a market (or ever). The EA ecosystem is tightly-knit, and even if oracular funders didn't explicitly collude, they would be starting from such similar values and such a cooperative stance towards each other that the equilibrium would probably feel collusive. 

My thoughts: Guess this one is up to the final oracular funders.

10: Who Are We Expecting To Have As Investors?

A: Institutional Types And VC Firms

These are who we would eventually be hoping to attract, given that we are trying to encourage the creation of institution-level expertise in charity-picking. Also, they have the most money.

Disadvantage: We probably couldn't get them right away, and there aren’t that many who really understand effective altruism yet.

B: Ordinary People

Not too ordinary - we still need for them to make more than $200,000/year and so qualify as accredited investors (see I-A below) - but random Google engineers who are interested in EA, that kind of thing.

Advantage: there are lots of them, including in EA. Some of them probably have a hidden talent in charity selection and it would be very useful to learn who those are.

Disadvantage: They would bring less intellectual firepower to the problem per entity than large firms. They would have less money, meaning that only smaller projects could get funded, and we would need to be more concerned about moral hazard (ie they lose all their money and then are sad). 

My thoughts: Institutional types probably won’t bite for a little while, so we'll need to be prepared for ordinary people.

11: Conclusion: What Kind Of Impact Market Should We Have?

At this point after talking to reviewers I am leaning towards a fractionalized-share-with-assurance-contract design, implemented through a mixed fiat-crypto market, aimed at relatively ordinary people, kickstarted by a few very large oracular funders who come up with their own decision procedures.

I’m most uncertain about how to handle issues of founder equity, moral credit, founders getting rich, and splitting surpluses. Here are some designs I’ve been thinking about:

A: The “Maximum Capitalism” Design

Founders start with 100% equity, representing all the credit for their project. They can sell however much equity they want to investors, for however much money the investors are willing to pay. They can spend some of that money funding their project, and if they have any left over, they can keep it as profits. If they sell off 100% of the equity, then in some sense that we will have to hammer out socially, they “get no credit” for their project, no matter how hard they worked on it or how good a job they did.

The main advantages are elegance, ease of use insofar as everyone knows how capitalism works, and potential for good charity founders to get rich.

The main disadvantages are that it requires some financial savvy (unsavvy people can potentially lose all credit for their project in a way that will make them very unhappy), and it diverts a lot of the surplus to founders and investors rather than funders/beneficiaries.

B: The “Project Funding Only” Design

Founders sell off an amount of impact shares equal to the cost of their project / constant, where the cost of their project includes paying themselves a fair salary. Investors speculate on these. The investors may get rich if they are unusually good at spotting promising projects, but the founders won’t.

The main advantage is ease of use for non-financially-savvy founders, plus people feeling more comfortable knowing that charity founders aren’t getting rich and buying yachts with their donations. In some cases, the oracular funders and their beneficiaries will keep more of the surplus.

The main disadvantage is that in cases where projects are very beneficial and everyone knows this, whoever snaps up the shares first will get most of the surplus. This is clearly unfair.

C: Hybrid Design: Some Kind Of Scaling Of Investor Equity

I dislike both A and B enough that I am trying to think of hybrid designs that avoid both extremes.

Consider our typical example: a malaria project that requires $1 million and will produce $5 million in benefits. If we know that the desired solution is to sell $1 million worth of equity, but we want people to be “bidding” for it instead of just snapping it up and “scalping” it, we could have them bid for smaller quantities rather than higher prices. For example, Bidder A could say “I’ll pay the $1 million for only 25% of the equity!”, and Bidder B could counter “No, I’ll pay it for only 21%!” Then Bidder B would win, pay $1 million, and when the final oracular funder bought it for $5 million, the investor would get $1.05 million - enough to make a fair profit, but most of the surplus would still go to the funder and beneficiaries.

The key would be that the original founder keeps 79% of the equity but - by agreement/collusion among oracular funders - can’t sell it. They can just keep it as moral credit for their hard work on the project. I don’t like the fact that it’s important they want to keep as much as possible but they have no real incentive to do so - maybe funders could give them 1% of what their equity entitles them to as a bonus, as they are buying other equity? I’m not sure; it’s kind of a hack.

The other problem would be the math. If there’s only one block certificate worth $1 million, it’s easy to imagine how the bidding works. If there are 1000 tokens for $1000 each, then bidding on an equity amount for each token sounds more complicated. This would be up to auction theorists and programmers to figure out a fair way to implement.

My thoughts: I really hate all these options, and I’m posting this partly to see if other people have better ideas. 

Some reviewers who are also potential final oracular funders feel like they might be able to solve the all-surplus-goes-to-making-people-rich problem with A by just not paying that much money; I haven't fully thought through the downstream effects of that.


A: Issues Around Unregistered Securities

The lawyer I consulted on this project says that impact certificates are unregistered securities. But he also says that in the US, it’s legal to sell unregistered securities to accredited investors - meaning (more or less) people who make more than $200,000 a year. Any impact certificate market could probably operate legally in the US as long as it screened buyers to make sure they were above this number. I’m not sure about other countries.

B: Issues Around Tax-Deductability

The same lawyer said impact certificates would never be tax-deductible. That makes investing in impact certificates less financially efficient than donating to a regular charity, from a would-be investor/donor’s point of view.

One possible retort is that potential investors and potential donors are different people, plus it’s more financially efficient because you might make money off of it, which you never do with normal charitable donations.

Maybe a better way to look at this is abstracting out the investor and comparing “final oracular funder donates directly to charity” vs. “final oracular funder gives money to investor, causing charity to get funded”. Here it becomes clear that it’s the oracular funder who’s losing out on the tax benefits. But do oracular funders (eg OpenPhil, Future Fund) pay taxes at all, or benefit from tax-deductability? I’m not clear on this.

C: Issues Around Oracular Funders’ Nonprofit Status.

Government bodies might think it’s weird for a tax-deductible charity, organized for the public good, to spend lots of its money in payouts to venture capitalists who are not themselves engaging in charity.

As far as I know there’s no legal precedent on this issue and we would just have to try it and see.

D: Issues Around Governance And Principal-Agent Problems

Suppose I invest $1 million in a project to cure malaria, on the assumption that it will work and an oracular funder will pay me back - and then the people involved are lazy and don’t even try to work on the malaria cure?

Realistically this is no worse than the current situation, where eg OpenPhil can fund someone to cure malaria and they can just not do it. If they deliberately embezzle the funds, it’s embezzlement and you can sue them. If they’re just really bad at their job, that seems like part of the risk you take in normal charity and normal investments.

In normal investment, stocks (sometimes) grant a right to control a certain amount of corporate governance. In theory, impact markets could work the same way. I’m against this because it would be extremely hard to enforce; we would need to get Congress to pass laws about it or something. Also, sometimes shares in a company don’t connect to governance, and that usually seems to go fine.

E: Issues Around Middlemen Holding Money

If there is an impact marketplace, probably it would take investors’ money (or maybe their credit card details, if it’s using assurance contracts), then shortly afterwards give it to the charitable projects.

Does the step where the marketplace acts as a middleman between the investor and the charity give it any legal responsibilities? For example, are these two transactions both separate taxable events? (I think no, but would like confirmation).

Appendix II: Ethical Issues

A: Accidentally Encouraging High-Risk Negative-In-Expectation Projects

Suppose I have a project where I spend $1 million to cure everyone in Senegal of malaria, creating $5 million in value. Also, there’s a 25% chance it has terrible side effects, giving everyone cancer and costing $100 million in lost value. So on average, the project produces $5 million + 0.25 * -$100 million = -$20 million in value: its expected costs are greater than benefits. A regular charity would avoid funding this, because they are altruistic and clearly don’t want to fund something that (in expectation) makes the world worse.

But a smart-but-amoral VC would fund it. In 75% of cases, it produces $5 million in value, and charities (who just a see a happy malaria-free country) pay them that money. In 25% of cases, it creates -$100 million in value, and nobody buys the impact certificate, for a total value of $0 (they’re not obligated to pay the negative value; how would you force them?) So in expectation, they will make 0.75*5 = $3.75 million on a $1 million investment. So they do invest in this net-negative project.

You could object that this is no worse than the situation with business investing now, but actually there are quite a lot of bad businesses that make the world worse, and since charities are explicitly trying to make the world better, they should hold themselves to a higher standard of not doing that.

Ofer and Owen Cotton-Barrett have discussed this here. Their conclusion is that final oracular funders should take this into account when deciding who to grant impact certificates to. This kind of softens the benefit of impact certificates - their whole point was supposed to be that funders shouldn’t have to be smart people who can figure out the results of various plans - but at least figuring out whether there were potential bad effects sounds like a somewhat easier problem than figuring out whether something will work.

This could also be handled at the market level, by refusing to list certificates/tokens in projects that were especially likely to be high negative-impact - although, again, this requires market officials to be smart, which is a strong requirement.

B: Incentivizing Reward-Hacking

Once an impact certificate marketplace is established, decisions of final oracular funders will determine whether savvy private investors get millions of dollars or not. That means that savvy private investors - who are very good at influencing things! who might have entire departments whose whole job it is to influence other large savvy organizations! - will want to gain power over organizations like Future Fund and OpenPhil and influence their grantmaking. A worst-case scenario for this might look like the current relationship between the US military and its defense contractors, where outright bribery is banned but the contractors use many other forms of soft influence to sway military decision-makers.

I have trouble imagining this really working with OpenPhil and Future Fund in their current forms, but part of the problem is that this provides a long-term incentive for private industry to try to twist them into other forms that are less resilient.

C: People Could Lose A Lot Of Money

The typical statistic is that nine out of ten startups fail. Of the 656 ACX Grants applications, I was only able to fund 38 (6%), though a few others got funded through other means. Future Fund has mentioned only funding 4% of its “open call” applicants. Unlike investing in regular stocks, where on average even if you can’t beat the market your stock will stay the same or even go up, the default outcome in buying impact certificates is that you lose all your money.

This is bad both because we don’t want people to lose all their money, and because this might create moral hazard on the part of final oracular funders to recoup some of people’s losses if they seem like an especially pitiful case.

I think we should promote a norm that people shouldn’t invest any money in impact certificates that they don’t want to lose. Otherwise, I’m not sure this is any worse than eg crypto, which already lets small investors lose all their money quickly.

D. Using Impact Certificates For Evil

For example, what if people use this technology to incentivize assassinating political figures they don’t like?

This concerns me least out of all the issues in this section. There is no advantage of retrospective assassination funding compared to prospective assassination funding. In fact, there’s a significant disadvantage: the prospective funding requires a conspiracy of two people (assassin and funder), whereas the retrospective funding requires a conspiracy of three (assassin, investor, and final oracular funder). Most investors would assume that most oracular funders wouldn’t support assassination, and on the very slim chance that an oracular funder publicly communicated the opposite, they could be immediately arrested for promoting assassination, the same as anyone else who publicly broadcasts “I WILL GIVE YOU $1 MILLION TO ASSASSINATE SOMEONE”.

Legal political movements that we disapprove of may eventually be able to use this technology, but this is a general side effect of any attempt to make charity more efficient, and most charity is not political.

Comments46
Sorted by Click to highlight new comments since: Today at 11:21 AM

Thanks for this incredibly comprehensive overview!

I’d like to use this chance to position our GoodX project “Impact Markets” in this context. (Thanks for linking to our MVP post!) I’m trying to write this in a way that is also understandable to people who haven’t thought about impact markets much, so sorry in advance if some of it is unnecessarily verbose.

Our main selling point is simply that altruists are more likely to have some sort of investor in their network (because there are countless ones ) than to have a funder in their network who we would trust (because there are like half a dozen or so – Open Phil, the Future Fund, the EA Funds, the SFF, et al.). Hence, to a funder, the altruist is going to succeed with the baseline probability of some suitably chosen reference class – say 10%. But to the investor, who has more knowledge of their altruistic friend, they are more likely to succeed – say 20%. There are also reasons to think that the investor can make the altruist succeed at a lower cost than the funder. The result is that the funder and the investor can find a reward size that saves the funder money and time (they don’t need to negotiate this in each case) and yet earns the investor a profit.

An added benefit is that funders can hire a wider range of staff. These staff still need to be excellent at priorities research or at applying priorities research, but they don’t need to be excellent at startup picking anymore.

What Is The Basic Format Of The Market?

We want to implement these in a sensible order, from simple to complex. We haven’t considered “block” impact certificates outside an early experiment, so they’re all considered to be fractional in our case.

If “block” certs are legally easier, that could be a reason to rethink that! Our vision for where impact markets might go is that they integrate with the classic financial market: Think tanks (incorporated as, say, public benefit corporations) issue impact certificates, generate a profit from their sale, and control their stock price with the profits, e.g., by paying dividends or doing buybacks. All the investment happens on the level of the classic stock of the think tanks. The impact certificates are just extremely simple products, consumables, that are so simple that it is dead obvious to the SEC that they are not securities. Our (Owen’s) metaphor is (consumable) bottles of wine.

This seems to us like it would make impact markets hard to use for independent researchers (the main audience we have in mind for issuers) because they would have to first start a corporation for their blogs. But maybe there are ways around that. (Maybe accredited investors can invest into the stock of unincorporated associations.)

Note that this consideration (the previous two paragraphs) is much more fundamental than the question whether certs should be fractional or not. It just relates to the question if non-fractional certs are legally easier. But if the investment happens on the level of classic stock anyway, then impact certs could also be issued retroactively, which makes them legally unproblematic according to what we’ve heard from the lawyers we talked to.

Note also that one of our cofounders, Dony, has thought a lot about dominant assurance contracts in the past. I remember him thinking about how they might combine with impact markets. That’s an interesting angle (dominant or not), but not one we’ve prioritized yet.

What Is Being Sold?

We want to keep it simple and so only focus on the core idea of making the “startup picking” part of funding delegatable. For that, the finer points of the metaphysical significance of impact purchases don’t matter. It just matters that there is a trusted retro funder that provides a prize pool and makes good decisions on what projects to reward. We want to leave the philosophy to philosophers or culture to figure out over time. (I can see that these are interesting questions, but I don’t think that that’s where the big counterfactual impact is.)

This sounds similar to what you write under “Credit for funding the project.” But I at least have moved away from the term “moral credit” for this because while some people interpret it simply as “I can get money for holding this contract,” others again interpret metaphysical significance into it (and in various conflicting ways too).

The funders that we envision serving with our solution are funders like the EA Funds, Open Phil, the Future Fund, et al., who are only interested in the counterfactual impact and not in being seen as virtuous by others or by God or some other variation.

We call what our certs certify a “right to retroactive funding,” so a right that is completely limited to only the entitlement to some share of the prize pool to the extent to which it is awarded to the project. (We want to eventually make it possible for people to attach other rights to their certificates, such as “bragging rights” if they so choose, but it’ll be up to them to define how that should work. This agnosticism about the rights that people attach to certs is just an implication of our using the Hypercertificate standard. The author of the Hypercertificates standard, David Dalrymple, and Protocol Labs are working on a solid, standards-based implementation of impact markets that we want to be compatible with.)

Another way to think about it is that if currently someone enters an article that they’ve written into a prize contest, then they can win a prize. If they’ve collaborated with five others on the article, they can agree to split the prize among themselves if they win it. They can (if they all so choose) make that agreement without ever clarifying the metaphysical significance of their agreement.

How Should The Market Handle Projects With Variable Funding Needs?

We’re going for the second solution from your list here because we put a strong emphasis on the precision and verifiability of impact certificates. This has various reasons:

  1. we want impact certificates to be like products that charities can sell and that investors can preorder,
  2. we want to make it easy for buyers to verify that the impact is not doubly sold under different framings, and
  3. we want to make it easy for buyers to verify that the set of issuers is complete according to strong cultural norms.

So “cure malaria in Senegal” is not a definition of an impact cert that would work on our market. It doesn’t have enough details about the concrete actions that’ll be taken, it doesn’t make clear that the issuers really owns the cert to the extent that they claim, etc.

Instead something like this could work: “We want to collaborate with Concern Universal on a distribution of 5 million long-lasting insecticide-treated bednets in the Matam region of Senegal between Jan. 1, 2023 and Apr. 1, 2023. We currently own 60% of the certificate; Concern Universal owns 40%. The census, the distribution, and the follow-up surveys are conducted by paid staff of Concern Universal who have forfeited any claim to the certificate in their work contracts. No one else, to our knowledge, can make a legitimate claim to the certificate.”

This clarifies ownership, timeframe, location, scope, etc. The organization would sell scores or hundreds of such certificates that can all follow a standard template.

Unrelatedly, I don’t think this impact certificate would work: LLIN distributions are an intervention that is highly likely to work. Funders may assign a 90% chance to their success; a highly involved investor might assign a 95% chance to the success. That gives the investor a very, very minimal edge over the funder. The funder would likely consider it unnecessary overhead to use impact markets for this, or, if they do end up using them, they’ll pay very little over the counterfactual prospective funding for the very slightly reduced risk. That’ll not be enough for the investor to beat some counterfactual investment like a standard ETF. That’s in contrast to hits-based giving where funders often assign a 10% chance of success, which leave a lot of (multiplicative) room for highly involved investors to be more optimistic.

Should Founders Be Getting Rich Off Impact Certificates?

We haven’t thought about this much and would love to be convinced one way or the other.

But if there are projects that everyone recognizes as good, then an impact marketplace will shift the surplus from final oracular funders (who are forced to pay prices more commensurate with their benefits) to someone else. If we let founders get rich, it shifts the surplus to founders. If we don’t let founders get rich, it shifts the surplus to fast investors who may not have added any value - that is, to the first people who snapped up the obviously-underpriced certificates. See Section 7 for more on this problem.

I don’t have a good solution to this latter issue other than either not using impact markets, or allowing founders to get rich. I sketch a hack-ish solution in Section 11C.

My take is basically the first, i.e. that impact markets won’t be used for projects where it doesn’t make sense for both sides (investors/founders and funders) to use them. These will just continue to use prospective funding. This just seems like a brute fact of the preferences of the market actors and not like something that is under the control of the marketplace at all.

Some more elaboration in case it’s not clear what I mean (but it probably is): Funders will pay as little as they can get away with paying. They may start by offering prizes that are just slightly more than what they would’ve paid prospectively for a single project (with, say, a 10% chance of success). Maybe no investor will be interested. Then they’ll raise their bid slightly again and again until investors start to be interested and the first few projects actually get funded from investor money. But this is still strictly less than what the funders would’ve otherwise paid in total for prospective funding. Sure, some of that money goes to investors who can buy yachts from it, but the counterfactual is not that that money would’ve gone to beneficiaries but that it (and more) would’ve gone to failed projects.

Or in made-up numbers: In the prospective-funding world, a funder pays out a total of $1m to 100 projects ($10k each), 10 of which succeed. In the retrospective world, the funder pays out $800k to 10 projects ($80k each), all of which have succeeded. These $80k are maybe split $40k to the founders and $40k to the investors, so that one could say that $40k or more have gone to waste in some sense. But that’s compared to a counterfactual that was never attainable in the first place. More realistically, the funder achieved the same impact while saving $200k that can now go to more projects to help beneficiaries.

Unrelatedly, established charities (as opposed to independent researchers or startup charities) often have track records that make impact markets redundant (see above) or have 18+ months of runway that retro funding would funge with, so that, one way or another, funders are probably not very interested in using impact markets for them.

If impact markets were “a success”, they might have to more than dectuple the amount they pay this person, with no change to their output.

Realistically, I think, a funder wouldn’t just dectuple the investment but would try to pay as little as possible that is just enough to incentivize the creation of this researcher. By construction, the six-figure salary (say, $200k) is not enough to create an additional engineer, but maybe $300k is enough, in which case the funder would bid $300k, not $10m. Of course it could just so happen that the only price that incentivizes the creation of another researcher is no less than $10m, but then the counterfactual is not another researcher for $200k but no researcher at all.

Again the funder is well-advised to keep their true valuation of the impact (if they can come up with such a thing) a secret and just bid as low as they can get away with. (Just like when you buy a second-hand laptop. You may privately decide that you’ll bid up to $1k for it and yet start by bidding $100 because it might just be enough to convince the seller to part with the laptop.)

Admittedly, the current AI safety researchers may get greedy and try to use whatever leverage they have to get more money, but they could do that already.

How Do We Kickstart The Existence Of An Impact Market?

The Committed Pot of Money is the main or most fundamental solution we have in mind. The lower-down solutions are more powerful and they may be the desirable end states but they also come with risks that we’re wary of.

If I put myself into the shoes of an early investor on the Committed Pot of Money impact market, I don’t model myself as judging my chances to break even as a random draw. Rather, by buying in early I can buy more at a lower price, so that I later have an outsized share in the expected returns compared to later investors. But I also have an opinion on what the retro funder will value, and that model has a much stronger effect on my subjective expected returns than the amount that has been invested in total in all projects.

(If I buy SOL tokens at $5 because I think Solana is a great technology with a high success chance, I don’t start to think that Solana is going to be much less successful only because there is heavy investment into Solana and many other technologies. I’ll Aumann-update a bit on all the investments into Avalanche, Cardano, etc., but not overwhelmingly much.)

Should The Market Use Cryptocurrency?

My thoughts: Crypto-literate reviewers say that crypto exchanges take fiat and handle transactions within themselves in a non-crypto way - but then the crypto exists in case someone wants to bring it to a different exchange. This seems like a best of both worlds scenario.

Agreed. Plus we can just start with simple non-crypto solutions and keep all other options open in case there are eventually strong arguments to transition to something more like FTX (centralized) or Serum (decentralized).

How Should The Market Price IPOs?

We have an auction process that A Donor developed for us  (see A Donor’s comment) that, we think, solves B neatly. We haven’t implemented it yet. (It is based on the view that it is desirable to reward people for aiding a fast price discovery.)

How Should The Oracular Funder Buy Successful Projects?

We’ve considered both – buying shares immediately if possible or creating the funding floor of limit orders. For now we’re going with the first solution simply for simplicity, but I definitely see the appeal of the second.

What Should The Final Oracular Funder’s Decision Process Be?

My thoughts: Guess this one is up to the final oracular funders.

Agreed. And by implication I think it’ll be an auction where the funders bid as little as they can get away with paying and only bid more if they don’t observe enough projects getting started or invested in.

Who Are We Expecting To Have As Investors?

My thoughts: Institutional types probably won’t bite for a little while, so we'll need to be prepared for ordinary people.

Agreed. Plus, institutional types are again fewer and have to optimize more for scale, which causes them to suffer from some of the same problems that funders currently suffer from. Investors should rather be the sort of people who are sufficiently many and sufficiently unconstrained by scale that every independent researcher or startup entrepreneur can so happen to know one or two of them quite well.

It’s apparently now possible to become an accredited investor by passing an exam, which lowers the bar. But impact certs as we conceive of them can be consumed (or “dedicated” or “burned”) by altruists, which plausibly turns them into consumables. Consumables, such as wine, can also be preordered and resold. I find it plausible that impact certs are so much more like wine and so much less like stock that they are not securities in the eye of the law or the SEC. Plus, all of this is easier in many countries other than the US and UK.

Conclusion: What Kind Of Impact Market Should We Have?

I think the above has elucidated the design that we’re interested in. It’s similar to the “Maximum Capitalism” design but without the metaphysical bits about “losing credit” and such. The founders have done their work, and if they (maybe unwisely) elected to sell 100% of their impact cert, they’ve still done their work and can get their recognition and kudos for it. (For comparison, if I start a successful startup and make an exit, I can still claim to have started a successful startup and get the reputational benefits.)

Issues Around Unregistered Securities

We have talked with lawyers about this and people with more experience with quasi impact markets. The devil is in the details. There are various ways to avoid the impression that impact certs are securities – as mentioned above, we conceive of them like bottles of wine that can be preordered, stored, resold, and consumed – but we don’t yet have full legal clarity on the issue.

Issues Around Tax-Deductability

In our view, investing in impact certs should not be tax-deductible. It’s done with a profit motive. But consuming the certificate is something that is altruistic, and it would make sense to see to it that that is tax deductible. We envision one potential system where you have the option to do all your cert trades through a nonprofit. The nonprofit just executes all your deals on your behalf. So long as you haven’t consumed your certs, the nonprofit does nothing else. But once you consume a cert, that nonprofit (not the issuer of the cert) writes the donation receipt for you.

Issues Around Oracular Funders’ Nonprofit Status

There may be larger prize contests (such as the XPrize) organized by nonprofits. One could research those to look for legal precedents.

Issues Around Governance And Principal-Agent Problems

Agreed.

Issues Around Middlemen Holding Money

We’re currently planning to let people transact directly with each other to avoid these questions for the time being to be lean and all.

Accidentally Encouraging High-Risk Negative-In-Expectation Projects

This is getting at some of the big crucial problems for us that we’ve been mulling over for a year or so. Here is a list of all the current risks and mitigation mechanisms that we see.

Attributed Impact (mentioned in the linked document ) also addresses moral trade, the lack of which I consider an even greater problem as well as a superset that contains the problem of the distribution mismatch.

Using Impact Certificates For Evil

The above document also addresses this.

Incentivizing Reward-Hacking

That is an interesting failure mode to look out for.

People Could Lose A Lot Of Money

I think we should promote a norm that people shouldn’t invest any money in impact certificates that they don’t want to lose. Otherwise, I’m not sure this is any worse than eg crypto, which already lets small investors lose all their money quickly.

Agreed. A marketplace could also set a default limit on investments and only raise it for investors who have somehow proven that they are financial savvy. Only allowing accredited investors to invest would have that effect, though it may be a bit too exclusive.

Nice!  If you end up with a predictive component, I—and perhaps my forecasting group—would be up for giving probability estimates of their chosen metric of success, and maybe estimates of their value. I imagine that these forecasts/estimates would at first be shitty, since they would be in a new domain, and it's hard to get intuitions, or baserates & distributions of success. But I'd be up for trying!

See also: I will bet on your success on Manifold Markets, which might be of interest to readers more generally.

DC
2y3
0
0

One suggestion we got at the EA Econ retreat was to have a prediction market somehow tracking the expected value paired to each certificate as a requirement so that the history of the expected value is known, reducing the risk of retro funding ex ante risky projects.

I imagine that this would at first take too much time and produce somewhat low-quality predictions.

I did a bunch of mechanism design on the details of #1 alongside the Impact Markets people!

Key desiderata:

  • Low overhead for participants in the market - Incentivising these people to spend lots of brainpower trying to work out when you buy and sell is massively suboptimal. You want a system that is buy and forget and just works.
  • Minimize hard feelings around selling your cert at a loss or seeing equivalent ones go for a very low price - Related to your #II - C
  • Impact authors automatically get some fraction of the rewards of a very successful project[1]
  • Ability to buy into certificates for flexible amounts, with minimal hassle and handling scaling well

We converged on the below model if there is a payments platform like FTX to automate microtransactions (crypto or not), and a different one which has slightly less nice properties but requires less payouts per pay-in. I can write that one up too if wanted.

Awesome Auto-Auction (better name welcome)

The core innovation is the queue of impact slices, with a "head" of buyouts moving along the queue as new money comes in.

Ownership of certificates in an auto-auction is divided into a series of slices, from earliest (the initial entry creating the auction), through all previous holders of slices of ownership, to the current owners’ slices.[2]

There is no action required to sell or decisions about price, the only available move is to buy and the only decision is how much you want to buy. When someone buys into a cert, the incoming money is split into three streams:

  • Creator - Directly to the impact creator
  • Capital repayment - Returning the capital to current holders of the cert and buying out their slice, with no opportunity for profit
  • Royalty - Giving profit to investors who have already been bought out, in proportion to the amount they raised price by (incentivising people to be early investors and raise the price)

Let’s look at an example of how this plays out with a $1000 purchase into a cert, with a 25% / 30% / 45% split.[3]

Some set % goes to the original creator (100% if they have not yet got their asking price, optionally held by an assurance contract until met), some goes to paying back the money current owners put in to buy their slices (with the oldest current owner getting priority), and some as a trickle income profit to already bought-out investors who funded the cert earlier in proportion to how much they raised the price (as a %) divided by the total amount that bought out investors raised the price (the %s added together).

I made a spreadsheet which you can copy to play around with split proportions and buy events and see how the money ends up distributed among the three groups at different investment levels.

This might sound complicated, but can be made very straightforward to operate and understand with a UI which lets buyers predict how much would be paid out by their purchase at any future level of total investment in the cert. This could work something like:
 

This mechanism seems to thoroughly succeed at all the desiderata, and would I think make impact markets much lower friction than other proposed market mechanisms (and as a bonus transitions smoothly into #5E!). 

Edit: Also might well help with some of the concerns raised in the reddit thread, as I think the action "sell" has weird emotional connotations in this domain which this mechanism might bypass.

I would be happy to have calls with anyone who might want to turn it into a reality.

  1. ^

    I come down on the "doing good should be a route to increasing wealth" side of #4, with the caveat that I'd like to see this happen not as a big sell off type event where a final funder directly buys impact from someone who has already got plenty back, but as an ongoing royalty stream which any funder can top up if they liked the project and is split fairly among investors and creators.

  2. ^

    If using crypto, each individual’s NFT represents the combination of all their blocks of ownership, and indicates both how large a share of the NFT they currently own and their share of past ownership visually in an unobtrusive manner (likely a thin bar, with two colors).

  3. ^

    The split can either be determined by the creator as they add the cert to the market, or we could just pick reasonable defaults and everyone sticks to that.

I think Yes on (4) - make the founders rich, because most of the value will come from incentivising people to have an impact. Hence (B) on (7) - you auction the shares. The main disadvantage is that the greater cost might dissuade investors, which I think is mitigated a bit by making the certs cheaper via (B) on (2).

Two points on (2): i) i'm not convinced that you (an initial buyer) get to ultimately decide how impact certs are valued. You at most get to write something about the asset in the contract, but ultimately it is the buyers and sellers who will decide how to value it. I feel if they are grantmakers seeking to incentivise good work at a low cost, they may be disinclined from buying the impact that would have happened anyway (even absent their existence as a funder). One way that preference could be satisfied is to give each share a number. Funders will value the first shares most, because they are fully "counterfactual", but if half of the value comes from a thing that the founder would have done anyway, then shares beyond halfway will be worth nothing. ii) so far we've talked about what happens when a funder buys a cert from a founder. There's also the question of what happens when the next guy buys the cert from the funder. Should you pay for the whole impact of the funder paying the founder, or only the impact of paying the funder (i.e. apply the solution recursively). The latter would arguably be a somewhat distinct solution (c).

I tried to think about (2) a bit here - https://forum.effectivealtruism.org/posts/eb28pDHzZz2RWh9Fh/will-impact-certificates-value-only-impact. It hasn't had much discussion, so probably I'm still missing a lot of considerations.

Thanks for the link, which I had previously missed and which does contain some important considerations.

I've been assuming that the people who set up the first impact market will have the opportunity to affect the "culture" around certificates, especially since many people will be learning what they are for the first time after the market starts to exist, but I agree that eventually it will depend on what buyers and sellers naturally converge to.

One way that preference could be satisfied is to give each share a number. Funders will value the first shares most, because they are fully "counterfactual", but if half of the value comes from a thing that the founder would have done anyway

I don't understand this - if you need $1 million to do the project, isn't the one millionth one-dollar share purchased doing just as much good as the first? I think I'm missing something about the scenario you're thinking of.

Agree on affecting culture.

Sorry, I could have been clearer. Suppose we want to choose (B) on (2). And I ask for retro funding for a painting I drew.

Let's consider three cases: a) it had $500 of impact, and I was 0% likely to make the painting anyway. b) it had $1k of impact, 50% likely anyway b) $2k of impact, 75% likely anyway.

The "obvious" solution is that in each case, I can sell some certs, say 100 for $5 ea.

Alternatively, we could say that in: case a) each cert 1-100 is worth $5 b) certs 1-50 are worth $10 ea, 51-100 $0 c) certs 1-25 are worth $20 ea, 26-100 worth $0.

The second solution allows the world to know how useful the painting/project was, though at the cost of some complexity.

The bottom-line I think is that the contract should be clear about which thing it purports to be doing.

I think the theoretically optimal way to distribute credit would be to compute Shapley values for each contributor (including the the founder) and distribute the funding accordingly. But this is of course usually not possible (because of limited information).

Whether you use Shapley values seems orthogonal to the question I'm asking - whether to price the shares non-uniformly, in order to estimate relative contributions.

This discussion would benefit from an empirical analysis of social impact bonds and previous efforts to quantify and financialize social impact. My not super deep understanding from talking to old school foundation COO types is that those examples are a very mixed bag and very difficult to measure. The market dynamics discussed here take well functioning measurement as a given but that is usually not the case. Consider all the debates about the impact of various education "reform" interventions in the United States for example. 

Here is the canonical first US social impact bond example: https://hbsp.harvard.edu/product/UV7311-PDF-ENG

By “measurement” do you mean the measurements of metrics that the payouts are conditional on (we don’t use those) or measurements of the extent to which the prizes encourage efforts that would not otherwise have happened (we’d be very interested in those)?

If we allowed founders to capture more of the surplus, then maybe the founder would end up with $50 billion, and the charity and its beneficiaries would only get $50 billion, which seems much worse for them. Given that our goal here is to help beneficiaries (and by implication charities, since any money charities save goes to the beneficiaries eventually), this seems pretty bad.

As far as I understand, your worry is that investors might pay so much for impact certificates that it lowers total impact (because it means less money to beneficiaries/programs). But shouldn't investors take such effects into account when they decide what to pay for an impact certificate - and when they estimate what other investors will be will be willing to pay for the impact certificate? As far as I can tell, this argument seems to assume that the investors would be irrational, and/or that impact certificate prices will for some other reason deviate from what's socially optimal. I guess the hope is that the socially optimal price will effectively be akin to a Schelling point, and that investors should try to guess what that price is. If so, they wouldn't overpay for impact certificates, but simply pay the socially optimal amount. But maybe there's some consideration here I'm missing - I'm not read up on these issues.

One thing that confused me was the assumption at various points that the oracle is going to pay out the entire surplus generated. That'll get the most projects done, but it will have bad results because you'll have spent the entire surplus on charity yachts.

The oracle should be paying out what it takes to get projects done. Not in the sense of labor theory of value, I mean that if you are having trouble attracting projects, payouts should go up, and if you have lots of competition for funding, payouts should go down.

This is actually a lot like a monopsony situation, where you can have unemployment (analogous to net-positive projects that don't get done) because the monopsonistic employer has a hard time paying those last few people what they want without having to raise wages for everyone else, eating into their surplus.

I agree it's a problem for the entire surplus to go to the seller. But that problem isn't impactful people getting rich. It's that if the certs are too expensive there'll be too few buyers to clear the market. So I agree that payouts should probably be tuneable. If you want the actual impact of a project to still be known, then you could have the "percent of impact purchased" be the tuneable parameter, if buyers aren't too sensitive to it.

Ofer and Owen Cotton-Barrett have discussed this here. Their conclusion is that final oracular funders should take this into account when deciding who to grant impact certificates to.

Speaking only for myself, I don't see that as my conclusion. Here are some excerpts from our post:

even if everyone believes that a certain project is net-negative, its certificates may be traded for a high price due to the chance that the project will end up being beneficial.

.

It seems especially important to prevent the risk from materializing in the domains of anthropogenic x-risks and meta-EA. Many projects in those domains can cause a lot of accidental harm because, for example, they can draw attention to info hazards, produce harmful outreach campaigns, produce dangerous experiments (e.g. in machine learning or virology), shorten AI timelines, intensify competition dynamics among AI labs, etcetera.

.

Currently, launching impact markets seems to us (non-robustly) net-negative. The following types of impact markets seems especially concerning:

  • Decentralized impact markets (in which there are no accountable decision makers that can control or shut down the market).
  • [...]

.

In any case, launching an impact market should not be done without (weak) consensus among the EA community, in order to avoid the unilateralist's curse.

.

To avoid tricky conflicts of interest, work to establish impact markets should only ever be funded in forward-looking ways. Retro funders should commit to not buying impact of work that led to impact markets (at least work before the time when the incentivization of net-negative projects has been robustly cleared up, if it ever is). EA should socially disapprove of anyone who did work on impact markets trying to sell impact of that work.

You wrote:

but at least figuring out whether there were potential bad effects sounds like a somewhat easier problem than figuring out whether something will work.

I think the oppositive may often be true. Quoting from our post:

Suppose that a risky project that is ex-ante net-negative ends up being beneficial. If retro funders attempt to evaluate it after it already ended up being beneficial, hindsight bias can easily cause them to overestimate its ex-ante EV. This phenomenon can make the certificates of net-negative projects more appealing to investors, already at an early stage of the project (before it is known whether the project will end up being beneficial or harmful).

You wrote:

Advantage: this is how we do everything else under capitalism, and it usually sort of works.

I'm not sure what "usually sort of works" means, from an EA perspective. Capitalism seems to be a contributing factor to x-risks from AI, for example.

This could also be handled at the market level, by refusing to list certificates/tokens in projects that were especially likely to be high negative-impact - although, again, this requires market officials to be smart, which is a strong requirement.

This solution also requires that the impact market will not be a decentralized market that no one can control. It also requires to sufficiently prevent the "market officials" from being influenced by conflicts of interest (e.g. in case they'll be able to sell the impact of their work to establish/control an impact market on that very market.)

Conclusion

Impact markets may incentivize people to carry out net-negative projects in anthropogenic x-risk domains, using EA funding. Conditional on impact markets being astronomically influential, the risk of them incentivizing net-negative projects can cause the entire EA movement to be net-negative. It's not a minor issue that we should just keep in mind while moving ahead with impact markets. As an analogy, suppose the NIH would prepare a long policy analysis document about how to fund more effective BSL-4 virology labs, with a section at the end titled: "Appendix II: safety issues".

I think these concerns are largely incorrect, and avoiding implementing this technology due to them would be a tragedy. Here's my reasoning:

  1. It's not that hard to see when a project was at risk of having large downsides, oraculars can avoid retrofunding these just as they would avoid normally funding them.[1]
  2. If you assume oraculars who are fine with incentivising doing large amounts of expected harm in order to get credit for something if it goes well (which I think is a flawed assumption), well.. they can do that with normal funding. Retrofunding does not add to it unless the early funders irrationally expect oraculars to buy up bad EV bets which paid off.

As for "weak consensus", we have Scott Alexander, Paul Christiano, and Eliezer Yudkowsky coming down on the side of "Yes, retrofunding is great". I'm not sure how that could be seen as anything other than strong consensus of key thought leaders, not to mention the many other people who've thought carefully about this and decided it is one of the most important interventions to improve the future.

  1. ^

    In reply to "hindsight bias"; retrofunders have strictly more information than normal funders, and can be explicitly made aware of this failure mode. I expect major oraculars to be capable of correcting for hindsight bias well enough to pass on any really risky projects. This is a mid-sized crux, necessary but not sufficient to reverse the sign of Impact Markets' value due to this failure mode.

Ofer
2y15
0
0

It's not that hard to see when a project was at risk of having large downsides

I strongly disagree. It's often extremely hard to judge whether a project related to anthropogenic x-risks was ex-ante net-negative. For example, was the creation of OpenAI/Anthropic/CSET net-positive or net-negative (ex-ante)? How about any particular gain-of-function research effort, or the creation of any particular BSL-4 virology lab?

Given a past project that is related to anthropogenic x-risks or meta-EA, it can be extremely hard to evaluate the ex-ante potential harm that the project could have had by, for example:

  1. Potentially drawing attention to info hazards (e.g. certain exciting approaches for developing AGI).
    • If a researcher believes they came up with an impressive insight, they will probably be biased towards publishing it, even if it may draw attention to potentially dangerous information. Their career capital, future compensation and status may be on the line.
    • Here's Alexander Berger (co-CEO of OpenPhil):

      I think if you have the opposite perspective and think we live in a really vulnerable world — maybe an offense-biased world where it’s much easier to do great harm than to protect against it — I think that increasing attention to anthropogenic risks could be really dangerous in that world. Because I think not very many people, as we discussed, go around thinking about the vast future.

      If one in every 1,000 people who go around thinking about the vast future decide, “Wow, I would really hate for there to be a vast future; I would like to end it,” and if it’s just 1,000 times easier to end it than to stop it from being ended, that could be a really, really dangerous recipe where again, everybody’s well intentioned, we’re raising attention to these risks that we should reduce, but the increasing salience of it could have been net negative.

  2. Potentially "patching" a problem and preventing a non-catastrophic, highly-visible outcome that would have caused an astronomically beneficial "immune response". Here's Nick Bostrom ("lightly edited for readability"):

    Small and medium scale catastrophe prevention? Also looks good. So global catastrophic risks falling short of existential risk. Again, very difficult to know the sign of that. Here we are bracketing leverage at all, even just knowing whether we would want more or less, if we could get it for free, it’s non-obvious. On the one hand, small-scale catastrophes might create an immune response that makes us better, puts in place better safeguards, and stuff like that, that could protect us from the big stuff. If we’re thinking about medium-scale catastrophes that could cause civilizational collapse, large by ordinary standards but only medium-scale in comparison to existential catastrophes, which are large in this context, again, it is not totally obvious what the sign of that is: there’s a lot more work to be done to try to figure that out. If recovery looks very likely, you might then have guesses as to whether the recovered civilization would be more likely to avoid existential catastrophe having gone through this experience or not.

  3. Potentially causing decision makers to have a false sense of security.
    • For example, perhaps it's not feasible to solve AI alignment in a competitive way without strong coordination, etcetera. But researchers are biased towards saying good things about their field, their colleagues and their (potential) employers.
  4. Potentially accelerating progress in AI capabilities in a certain way.
  5. Potentially intensifying the competition dynamics among AI labs / states.
  6. Potentially decreasing the EV of the EA community by exacerbating bad incentives and conflicts of interest, and by reducing coordination.
    • For example, by creating impact markets.
  7. Potentially causing accidental harm via outreach campaigns or regulation advocacy (e.g. by causing people to get a bad first impression of something important).
  8. Potentially causing a catastrophic leak from a virology lab.

You wrote:

unless the early funders irrationally expect oraculars to buy up bad EV bets which paid off.

Depending on the implementation of the impact market, it may be rational to expect that many retro funders will buy the impact of ex-ante net-negative projects that ended up being beneficial. Especially if the impact market is decentralized and cannot be controlled by anyone, and if it allows people to profit from recruiting new retro funders who are not very careful. For more important arguments about this point, see the section "Mitigating the risk is hard" in our post.

As for "weak consensus", we have Scott Alexander, Paul Christiano, and Eliezer Yudkowsky coming down on the side of "Yes, retrofunding is great". I'm not sure how that could be seen as anything other than strong consensus of key thought leaders,

The statement "retrofunding is great" is very vague. AFAIK, none of the people you mentioned gave a blank endorsement for all possible efforts to create an impact market (including a decentralized market that no one can control). There should be a consensus in EA about a specific potential intervention to create an impact market, before it is decided to carry out that intervention. Also, the EA community is large, so it's wrong to claim that there is a "strong consensus of key thought leaders" for doing something risky because a few brilliant, high-status people wrote positive things about it (and especially if they wrote those things before there was substantial discourse about the downside risks).

not to mention the many other people who've thought carefully about this and decided it is one of the most important interventions to improve the future.

Who are you referring to here?

DC
2y13
0
0

There should be a consensus in EA

This is prohibitively vague. How do you operationalize this exactly? Can you give examples of when EA has achieved a consensus that is analogical with what you desire in this situation?

If e.g. Future Fund and Open Phil were to use it, wouldn't that be a pretty strong signal, especially since they would want to derisk it pretty heavily before scaled up usage of it, with months of dialogue and planning? What are you looking for here that wouldn't already happen as a matter of course during the significant amount of downside mitigation work that would need to happen while building and scaling it up in concert with grantmakers, donors, and charities, who are pretty risk-averse on average and will generally incline towards wanting to be satisfied with at least interim solutions to at least some of the downside risks we, you, and others have identified (and probably ones not yet identified)?

I am pretty happy with Avengers Assemble-ing some kind of group discussion on impact markets as a consensing vehicle, perhaps a virtual event on the EA Forum using Polis, maybe a pop-up event during EAG SF, if it will please you or meet some objective criteria you specify. I find this sort of thing generally desirable regardless (cost-willing) but I additionally want to know what gets your thumbs up specifically given my impression is you want to stop people doing anything before achieving this consensus, whereas I view this downside work as a thing done concurrently alongside the long and arduous road of the empirical work of building that will provide spades of course-correcting feedback.

about a specific potential intervention to create an impact market, before it is decided to carry out that intervention

Decisionmaking by committee (my current impression of your ask) on specific product versions is not how things get built, especially early-on, especially with multiple parties involved, and is a recipe for not getting things done. The space of decisions is way too high-dimensional and things change based on feedback. Approaching a consensus on the important parts of the general theory of impact markets early on such that robust net-positivity is agreed-upon seems much more tractable and important in comparison, as well as generally keeping people working in the field coordinated and in communication as they iterate through different parameters of their visions.

I should have phrased differently: It's not that hard to pick out highly risky projects retroactively, relative to identifying them prospectively. I also think that the reference class which is most worrying is genuinely not that hard to identify as strongly negative EV.

Impact markets don't solve the problem of funders being able to fund harmful projects. But they don't make it differentially worse (it empowers funders generally, but I don't expect you would argue that grantmakers are net negative, so this still comes out net-positive).

I would welcome attempts to cause the culture of big grantmakers to more reliably make sure the recipients stay focused on the major challenges, but that is a separate project.

The classes of problem you list are all important questions of what should be funded, and it would be great to have better models of the EV of funding them, but none of them are impact-market specific. It's already true that the funder who is most enthusiastic about a project can fund it unilaterally, and that this will sometimes be EV-negative. People can already recruit new funders who are risk-tolerant.

We're making grantmakers generally more powerful, and that's not entirely free of negative effects[1], but it does seem very likely net-positive.[2]

I do think it makes sense to not rush into creating a decentralized unregulatable system on general principles of caution, as we certainly should watch the operation of a more controllable one for some time before moving towards that.

The community as a whole cannot come to consensus on each of the huge number of important decisions to be made, the bandwidth of common knowledge is far, far to low, and we are faced with too many choices for that to be a viable option. Having several of the most relevant people strongly on board is about as good a sign as you could expect currently. I'm open to more opinions coming in, and would be very interested in seeing you debate with people on the other side and try to double crux on this or get more people on board with your position, but turning this into a committee is going to stall the project.

  1. ^

    And your cataloging of them does seem useful, there are things we can adjust to minimize them.

  2. ^

    There are some other effects around cultural effects of making money flows more legible which seem possibly concerning, but I'm not super worried about negative EV projects being run.

It's not that hard to pick out highly risky projects retroactively, relative to identifying them prospectively.

Do you mean that, if a project ends up being harmful we have Bayesian evidence that it was ex-ante highly risky? If so, I agree. But that fact does not alleviate the distribution mismatch problem, which is caused by the prospect of a risky project ending up going well.

Impact markets don't solve the problem of funders being able to fund harmful projects. But they don't make it differentially worse (it empowers funders generally, but I don't expect you would argue that grantmakers are net negative, so this still comes out net-positive).

If the distribution mismatch problem is not mitigated (and it seems hard to mitigate), investors are incentivized to fund high-stakes projects while regarding potential harmful outcomes as if they were neutral. (Including in anthropogenic x-risks and meta-EA domains.) That is not the case with EA funders today.

There are some other effects around cultural effects of making money flows more legible which seem possibly concerning, but I'm not super worried about negative EV projects being run.

I think this is a highly over-optimistic take about cranking up the profit-seeking lever in EA and the ability to mitigate the effects of Goodhart's law. It seems that when humans have an opportunity to make a lot of money (without breaking laws or norms) at the expense of some altruistic values, they usually behave in a way that is aligned with their local incentives (while convincing themselves it's also the altruistic thing to do).

I do think it makes sense to not rush into creating a decentralized unregulatable system on general principles of caution, as we certainly should watch the operation of a more controllable one for some time before moving towards that.

If you run a fully controlled (Web2) impact market for 6-12 months, and the market funds great projects/posts and there's no sign of trouble, will you then launch a decentralized impact market that no one can control (in which people can sell the impact of recruiting additional retro funders, and the impact of establishing that very market)?

If the distribution mismatch problem is not mitigated (and it seems hard to mitigate), investors are incentivized to fund high-stakes projects while regarding potential harmful outcomes as if they were neutral.

I thought a bunch more about this, and I do think there is something here worth paying attention to.

I am not certain that there is a notable enough pool of the projects in the category you're worried about to offset the benefits of impact markets, but we would incentivise those that exist, and that has a cost.

If we're limited to accredited investors, as Scott proposed, we have some pretty strong mitigation options.  In particular, we can let oraculars pay to mark projects as having been strongly net negative, and have this detract from the ability of those who funded that project to earn on their entire portfolio. Since accounts will be hard to generate and only available to accredited investors, generating a new account for each item is not an available option.

I think I can make some modifications to the Awesome Auto Auction to include this fairly simply, and AAA does not allow selling as an action which removes the other risk of people dumping their money / provides a natural structure for limiting withdrawals (just cut off their automatic payments until the "debt" is repayed).

Would this be sufficient mitigation? And if not, what might you still fear about this?

If you run a fully controlled (Web2) impact market for 6-12 months, and the market funds great projects/posts and there's no sign of trouble, will you then launch a decentralized impact market that no one can control (in which people can sell the impact of recruiting additional retro funders, and the impact of establishing that very market)?

I don't see much benefit to a Web3 one assuming we can do microtransactions on a Web2, so I'd be fine with either not doing the Web3 or only doing it after several years of having a Web2 without any of those restrictions and nothing going badly wrong (retaining the option to restrict new markets for it at any time).

In particular, we can let oraculars pay to mark projects as having been strongly net negative, and have this detract from the ability of those who funded that project to earn on their entire portfolio.

I think this approach has the following problems:

  1. Investors will still be risking only the total amount of money they invest in the market (or place as a collateral), while their potential gain is unlimited.
  2. People tend to avoid doing things that directly financially harm other individuals. Therefore, I expect retro funders would usually not use their power to mark a project as "ex-ante net negative", even if it was a free action and the project was clearly ex-ante net negative (let alone if the retro funders need to spend money on doing it; and if it's very hard to judge whether the project was ex-ante net negative, which seems a much more common situation).
  1. Seems essentially fine. There's a reason society converged to loss-limited companies being the right thing to do, even though there is unlimited gain and limited downside, and that's that individuals tend to be far too risk averse. Exposing them to a risk to the rest of their portfolio should be more than sufficient to make this not a concern.
  2. Might be a fair point, but remember, this is in the case where some project was predictably net negative and then actually was badly net negative. My guess is at least some funders would be willing to step in and disincentivise that kind of activity, and the threat of it would keep people off the worst projects.

There's a reason society converged to loss-limited companies being the right thing to do, even though there is unlimited gain and limited downside, and that's that individuals tend to be far too risk averse.

I think the reason that states tend to allow loss-limited companies is that it causes them to have larger GDP (and thus all the good/adaptive things that are caused by having larger GDP). But loss-limited companies may be a bad thing from an EA perspective, considering that such companies may be financially incentivized to act in net-negative ways (e.g. exacerbating x-risks), especially in situations where lawmakers/regulators are lagging behind.

Yes, and greater GDP maps fairly well to greater effectiveness of altruism. I think you're focused on downside risks too strongly. They exist, and they are worth mitigating, but inaction due to fear of them will cause far more harm. Inaction due to heckler's veto is a not a free outcome.

Companies not being loss-limited would not cause them to stop producing x-risks when the literal death of all their humans is an insufficient motivation to discourage them. It would reduce a bunch of other categories of harm, but we've converged to accepting that risk to avoid crippling risk aversion in the economy.

Re 1. I think the key aspect of impact certificates which is 'un-mentioned' here is their continued tradability. At any stage, funders will seek to invest into projects with greater value in the future. This will cause development according to the market values. 

For example, someone can buy insect welfare farm prototyping certificates in 2022, because they think that in 2025, they will sell for 10x the value. If in early 2023, people are willing to pay twice that in 2022, in mid-2023 the price is 3x what it was in 2022. Continued investments into the org enable it to reduce the cost, make better use of, and gain more insights into the product. Thus, the org prospers.

So, founders (and any shareholders) should be getting rich off impact certificates, because richness denotes impact.

I suggest that full impact of the project is sold: if people compete in how much impact they have, then they can either advance projects traditionally (direct work or funding) and entitle what the consensus among the project organizers and funders is or buy certificates. If only funding impact is sold, then organizers cannot participate in funding ownership unless they buy shares, which creates an unnecessarily divided atmosphere.

If funding needs are variable (which should apply to the majority of cause areas or scalable projects), then 100% impact should be clearly defined and shares traded while it should be possible to purchase shares in bulk.

If beneficiaries are increasingly more costly to reach, then the price of additional shares increases. This motivates funders to buy early. If some intervention, such as malaria vaccine research, reduces the unit cost of the remainder of the problem, then the vaccine investors should be able to buy bulk of shares, for example until the cost of vaccinating very difficult to reach beneficiaries equates that of moderately difficult to reach beneficiaries currently given bednets (at the current impact share rate).

Some speculators may wait until cost-effectiveness increases, e. g. due to research, if they hypothesize that even the last person vaccinated enjoys malaria prevention at a lower cost than any other one receiving a bednet (current practice) would. This optimizes resource allocation between cost-effectiveness increasing and direct benefit actions.

To spare any Oracular Funders from having to fund projects while enabling purchases of any scope, no pot should exist. Anyone can sell and anyone can buy. For example, suppose that you express interest in 20 projects and get 500 notes of people starting working or thinking of them requesting more to keep going. You fund 5 which you like and think that can get funding elsewhere later which increases the value of your shares while someone else funds other 5 which later get back to you with more promising results asking you for more to increase their impact. This (also a form of ideal state) can be kickstarted by several funders expressing interest in the projects of which certificates they'd like to purchase. Sometimes, people post projects they'd like to see on the EA Forum. So, maybe a tag: 'Certificates Purchase Interest' can be a good start.

I think that there has to be an option of using fiat. For example, crypto taxation in the US is a new thing to many people and could discourage participants. Some people, on the other hand, enjoy crypto, or even try to see which of the few products which accept crypto they buy. So, there should be an option to pay with fiat or crypto, both with equally low (no) fees (e. g. no 3% credit card fee).

I am for direct purchases rather than action. Considering the speed at which impact funding progresses, I am not worried that rents would be captured by the fastest, who would immediately be able to resell higher. And even if they were (for example if they bought in a window of opportunity when price plummeted and solved the problem), good for them. Auctioning, which is unfamiliar to many, could prevent people from acting fast and taking the advantage of opportunities to solve great shares of problems.

Funding floors by large funders may influence the market sentiments. For example, if a funder states that they will buy a relatively unfamiliar solution for $5m, then other investors can start buying around $5m*percent. If they set the floor of a second project at $5m when currently funders invest $10m*percent, then market funding can more away from the second project.

By economic theory, this should reduce the price of the second project. In practice, this can happen. For example, if an NGO cuts relatively cost-ineffective aspects. However, for maximally cost-effective projects, the price cannot decrease so less impact in that area will happen.

A volatile funding floor can misallocate but also improve the allocation of investments. Using your example of the medicine factory, if a funder first announces that they value the disease eradication at $1m but after $400k is invested, they change the value to $500k, then if they invested the $400k, no disease is cured because factory is build but no medicines produced. If others invested the $400k, then 80% of the disease is cured by the additional $500k. Another investor can purchase the remaining $100k for a price which is not unreasonably high.

I am for purchases, pitches, or demand by anyone at anytime, while I think that giving ranges of both costs and price values with possible record of adherence to the ranges from both sides can increase efficiency. It may be that some prices do not converge at market value, but at 'niche' market value informed by the market participants costs and benefits knowledge (e. g. due to impact complementarity to different fields).

 From institutions, academia can be the most readily interested in impact funding, while it can retain the intellectual depth which could be decreased by commercial marketing. Considering the price and minimum wage differential among industrialized and industrializing nations, even a relatively average funder could hire a qualified FTE team for 10% of their salary. 

My stance is that founders should be able to sell any percentage at any price. If they seek to pay themselves a reasonable salary and produce results which have been agreed upon, they should be able to retain the 'action' impact while the funders the funding impact (this would be a traditional funding case). 

I am not sure about the legal issues but just like anyone can buy crypto in many jurisdictions, then people should be able to buy impact certificates? People invest into crypto but it is considered an asset thus does not follow investment law? I have not at all researched this.

Currently, anyone can purchase any good or service, even if its impact is negative. So, impact markets would not make this worse. Law should prevent criminal activity. It is likely that impact certificates would not fund (legal) projects which have clearly negative impact, such as a cage factory farm. The impact certificate would be, for example, to reduce such farming.

The main worry can be that a funder invests into risk-mitigation in a way that alerts malevolent actors of an opportunity, which increases the risk. Biosecurity and AI safety are the areas in EA that come to mind. This is worse for direct funding, where there is no feedback of other funders who would signal disinterest by divestment.

The impact market trying to influence funders should be optimal. In addition, the Future Fund and currently OpenPhil actively ask people for insights about projects. Both have structures which enable gaining value from the insights while preventing manipulation.

The emotions should be positive. If one is spending 40 hours working on a project application with 6% chance of acceptance, they know it. If someone invests $1,000 into a research project if they think that it will be bought for $100,000 later with 3% chance, they made a decision. I agree that people should have the time and money, otherwise it would not be fun. There are other fun impact opportunities paid more certainly.

Quick opinions on the choice branches:

1: What Is The Basic Format Of The Market?

Custom: Awesome Auto Auction, but closest to D: Fractionalized Impact Shares With Assurance Contract.

2: What Is Being Sold?

B: Credit For Funding The Project - Funding a project retroactively should not be considered morally different from funding it in any other way, creating a novel kind of moral accounting seems overcomplicating things and hard to get actual consensus / common knowledge on.

3: How Should The Market Handle Projects With Variable Funding Needs?

Custom: Awesome Auto Auction solves this elegantly, by letting the project absorb some fraction of the value of the cert if it continues to grow.

4: Should Founders Be Getting Rich Off Impact Certificates?

A: Yes as this provides good incentives, but again with Awesome Auto Auction presenting a somewhat cleaner way than the default, as the founders would not liquidate their shares and drop the price / eat a chunk of donor money, but instead receive trickle income if their project kept climbing in value.

5: How Do We Kickstart The Existence Of An Impact Market?

B: Committed Pot Of Money (aka The Original ACX Grants Plan) plus common knowledge that other funders can easily join later and will do so if they like what's happening

6: Should The Market Use Cryptocurrency?

The best of both world proposal you mention seems clearly the best.

7: How Should The Market Price IPOs?

Custom: Awesome Auto Auction. Combines most of the advantages of both, I think?

8: How Should The Oracular Funder Buy Successful Projects?

A. By Buying A Certain Number Of Shares, with Awesome Auto Auction I think removing the main concern of Monopsony? But either way, B does not fit the structure.

9: What Should The Final Oracular Funder’s Decision Process Be?

A: Fund Based On Regular Charitable Decision-Making Procedure seems like a reasonable default, but letting the oracular decide seems correct.

10: Who Are We Expecting To Have As Investors?

Custom: People who are basically just okay with donating the money, like me, but don't have infinite runway so would be happy to be topped up by major funders who like the bets we pick. I really want impact markets to exist for my own use case.

11: Conclusion: What Kind Of Impact Market Should We Have?

I really hate all these options, and I’m posting this partly to see if other people have better ideas. 

You bet I have an idea!  And I bet you can guess what it is! Awesome Auto Auction solves this neatly, the founder gets funding which scales with the value of the cert, but not in a form they can sell so there is no risk of them losing the credit for their project.

Happy to go into far more detail on any of these if requested.

I’m confused about how you trace whether a project was responsible for an outcome. Two examples:

  1. Many scientific discoveries are the result of many researchers working on a problem for a long time, making it difficult to trace back whose contribution had what fractional impact towards e.g. an AI alignment breakthrough. But if we’re only paying the person who did the last bit, we’re dramatically underfunding all projects that lay important groundwork, and overfunding work near the end.
  2. If malaria decreases in Senegal, but then two different charities claim this effect, one of which was doing mosquito gene drives and the other of which was distributing bednets. There is insufficient data to run an RCT good enough to provide high precision about the relative strength of these interventions. Do the VCs need to spend all their time trying to figure out how future RCT designs will or will not structurally favor their investment?

Thanks! I think those are important especially in situation where organizations more or less explicitly collaborate on something but then fail to include the other as owner of certificates that they issue. That could cause the other to respond by ending their collaboration.

I’ve written about similar problems in The Bulk of the Impact Iceberg and The Attribution Moloch.

Have you written about similar existing efforts to measure impact and pay just for that impact through SIBs / pay for performance frameworks and things of that nature? There are real world examples that would be good to investigate.

I haven’t but I’m aware of a forthcoming report by Rethink Priorities that covers prize contests. If you find more research on that – or on the similar dynamic of hopes for acquisitions incentivizing entrepreneurship – I’d be very interested!

Not a prize though this might be relevant: https://www.aqaix.com/case-studies/monterey-bay-stormwater-digital-master-plan

A flexible prize model is a much simpler design for creating a market for doing public good / charitable things. Seems that terminology being used is complicating matters - whether you call it Impact Certificates, Social Impact Bonds, Pay-For-Success contracts, Advance Market Commitments, Retroactive Public Goods Funding , etc. They are all flavours of conditional contracts . But you want to ensure that the parties cannot change the terms of the deal for receiving outcome payments unilaterally in the future (which IMO is a problem with RPGF). 

The part which is the most important is for a funder to put a price on some measurable outcome that they want and have clear and unambiguous criteria for determining whether the outcome is met (e.g. $10k per homeless person housed for at least 1 year).  Ideally, this outcome should provide more benefit to the funder than the price paid (e.g. $15k+ cost savings due to reduced policing, boost to economy etc), which makes it an arms-length negotiation and scalable business model. The superintelligence of the market does the rest (e.g. VC investing in companies competing for the payments, and automatically price in what they think are their future cashflows  etc). Where it can get more subtle is where you are paying a flexible outcome payment proportional to impact. 

The other concern is that if we assume funders will have a limited budget, so need to prevent blow outs. In that case, you can pay a fixed amount distributed between companies "registered" to the fund that achieve the criteria for outcome payments, proportional to the achievement of their impact measure (e.g. each person put in a home provides a number of "points" that determines the company's percentage entitlement for a pre-determined period of time such as 5 years). Market equilibrium is reached with a certain number of company registrations, and the proportion of rewards available proportionally increases as the "points" cease to be eligible for outcome payments after the pre-determined period of time . Social impact is then infinitely scalable with the size of the fund, and the market automatically allocates payments in an efficient way. 

So to illustrate this in an example:

A funder decides they will pay $2m p/a in a flexible prize fund, allocated between companies registered to the prize fund proportional to the number of homeless people housed for at least 1 year. Company A houses 30 homeless people for a year (30 points). Company B houses 90 (90 points). For that year, the £2m will be split 25% to Company A and 75% to Company B. They will keep those points for 5 years. Next year, Company A houses another 30 (30+30 = 60 points), and Company B houses another 90 (90 + 90 = 180 points). Same split. In the third year, Companies A and B they do the same, but a newcomer Company C puts in 100 people in a home, so it will be split 90/270/100 between A, B and C. After 5 years, Companies A and B lose 30 points and 90 points respectively as the 5 years is up, which provides more proportionally for everyone else.

If a funder wants to increase the number of homeless people housed, they just increase the annual payment, and a new "market equilibrium" is reached. 

This flexible prize model can be applied to any form of public good, as long as we can agree they are robustly measurable and comparable with each other (e.g. QALYs, % reduction in hospitalisations vs usual care, number of kids put in college, lives saved etc). 

Credit to Pogge and Hollis for the idea of a flexible prize fund:  https://en.wikipedia.org/wiki/Health_Impact_Fund

Prizes: Yes, totally! I’ve found that people understand what I’m getting at with impact markets much quicker if I frame it as a prize contest!

That point system sounds interesting. Maybe you can explain it again in our call as I’m not sure I follow the explanation here. But we’re currently betting on indirect normativity, so it won’t be immediately applicable for us.

What do we want from impacts markets?

Mostly, we want the prediction power of markets.

Do we want the market to take on some of the risk? Probably no.  Most current funds are ok with risks, as long as the odds are net positive. Scott seems to think that transferring risk to to investors is a bad externality of impacts market. 

This is bad both because we don’t want people to lose all their money, and because this might create moral hazard on the part of final oracular funders to recoup some of people’s losses if they seem like an especially pitiful case.

I have not thought at lot about this, but I think I agree with Scott here.


So what we want is a way of tracking and rewarding good predictions. But we already have a solution for that, i.e. prediction markets. Granted that there are still design details to be worked out regarding prediction markets, but 

  1. It seems like an easier problem
  2. There are already several experiments to learn from

 

In a prediction market, good predictors provide information and get rewarded for this. Funders can use the prediction market to guide what they fund. If they choose to fund something risky, some of the risk falls on the predictors (we can't remove this completely, and I don't think we want to either) but most of the risk falls on the funder, which I think is what we want.

This also solves some other problems. 

  1. Funders can just not fund things with obvious large downside risk.
  2. Investor can't capture almost all the value by just being fast at buying obvious opportunities.

Although I don't think 2 will be a big problem even in a impact market.  If it is an obvious good idea the people who wants to run the projects can just not go though the markets and instead reach out to a funder directly.

 

predictor = person who invest in prediction market
investor  = person who invest in impacts market
funder = person or institution who want's to fund altruism, with no expectation of financial return.

 

I'm I missing something that we want from a impacts market that we can't get from a prediction market?

(I've written about impact purchase before. However the thing I want out of such a system is completely removed from Scott's suggestion, and vise versa, so should probably not count as the same type of system.)

I think there were already prediction markets for future grants at the time when I did my first research into impact markets. Maybe they still exist.

The risk and prediction aspects that you mention factor somewhat into my value proposition for impact markets, but I’m not sure how important they are or generally how I feel about them.

My/our (impactmarkets.io) thinking is rather what I mentioned in the comment above, that we want to (1) radically reduce the time cost borne by funders, (2) expand the hiring pool for funders, and (3) enlist all the big networks of angel investors and impact investment funds in the search for the best funding opportunities. My thinking about the relative emphasis changes from time to time here as I become aware of new considerations. The risk angle could also make the top list, but of course only with fully informed investors who know exactly what they’re doing.

The biggest factor in my mind is currently the third one, followed by the first one. A quick Guesstimate model says that the third factor could 20x (5–85x) the accessible funding opportunities (access to thousands more social circles worldwide of people who are great at networking), and if hits-based funders currently aim for funding 10% successful projects, that translates to a reduction of about 10x in time cost.

I’m having a hard time guessing how the hiring pool might change. It will probably get vastly greater once grantmakers have to know only the priorities research and don’t need to be great judges of character and project-specific skill of others too, but funders might choose not to scale to absorb all of them.

I don’t know enough to exclude that this could be done with prediction markets, but to get angel investors and impact investment funds only prediction markets, they would have to involve real money, and that causes lots of legal hurdles. But even if we clear those, they would only put their money in them if they can expect the risky equivalent of a riskless 10–30% APY. I’m a bit hazy on how the initial liquidity gets onto new prediction markets that people create (e.g., on Polymarket or Augur) and how the initial value of Yes and No are set, so it’ll take me a lot more learning before I can repeat the profitability math for prediction markets. Also I don’t know yet how the markets would be resolved.

It seems all weird when I try to play through how this could work: You could establish the norm that researchers create markets on Augur that resolve positively if any of some set of retro funders endorse them. Then the researchers buy Yes. Then their impact investor friends buy Yes. Then the researchers sell out of some of their positions again to pay their rent etc. Then, if the project is successful, the retro funder comes in, buys just the right amount of No, and resolves the market to Yes. And by “just the right amount” I mean the sort of amount that just incentivizes investors to keep investing into projects like that without wasting money beyond that point. But that’s such a weird use of prediction markets… (On second thought, that wouldn’t work at all because there would not be enough liquidity on the Yes side for the researcher and the investor to buy it, right?)

Maybe someone with deeper knowledge of prediction markets can come up with a system that works well and has all the advantages though!

Update: Since there are maybe people here with more knowledge of prediction markets, it’s probably more productive to phrase this as a question: Would it be possible to create an Augur-like system where:

  1. altruistic funders can provide funds, a “prize pool” of sorts, with only a one-time effort,
  2. researchers can create markets at a low cost that try to predict whether any of a set of funders will endorse their research project and its results,
  3. investors can bet money on such endorsements,
  4. researchers can sell into those buys from investors to fund their project,
  5. funders can eventually resolve the markets with their endorsements and will thereby reward investor and researcher to just the right extent, and
  6. funders have no costs if markets don’t resolve or somehow resolve to no?

An alternative that we’ve been toying with are reverse charity fundraisers of sorts. You do your thing, and when you’re done, you publish it, and then there’s a reward button where anyone can reward you for it. “Your thing” can be doing research, funding research, copyediting research, etc.

I love the simplicity of it, but there are a few worries that we have when it comes to incentives for collaboration when participants have different levels of social influence. Still, it’s a very promising model in my mind.

I would very much want there to be a "money after the project" funding system for smaller projects in EA. 

https://forum.effectivealtruism.org/posts/7iptwuSyzDzxsEY5z/the-case-for-impact-purchase-or-part-1

Although after I wrote this post I updated towards this not being a good idea for anyone to get their income from this system long term. But I still think it would be a good alternative to other funding systems for new EAs. Retrospective funding have definite disadvantages, but for someone with out enough reputation, it may be better than the available alternatives. 

Indeed! I think this transition from impact markets to other sources of funding can happen quite naturally. A new, unknown researcher may enjoy the confidence in her abilities of some close friends but has little to show that would convince major funders that she can do high-quality research. But once she has used impact markets to fund her first few high-quality pieces of research, she will have a good track record to show, and can plausibly access other sources of funding. Then she can choose between them freely, is not dependent on impact markets alone anymore.

Someone who is exited about impact markets should do Goal Factoring on your preferred version of impact markets.

But do oracular funders (eg OpenPhil, Future Fund) pay taxes at all, or benefit from tax-deductability? I’m not clear on this.

In theory it would be great to get a lawyer/money manager from one of these orgs to comment on this, but I don't expect that to happen, so I'm going to give my guess as someone who runs a charity that has gotten money from both of these orgs.

I think most of Open Phil's money is stored at a DAF at SVCF. Dustin presumably got a big tax deduction when he donated to that DAF. Open Phil also sometimes distributes money in other ways, which may or may not benefit from tax-deductibility. But those should be thought of as "more expensive" since it costs money (taxes) to get money from Dustin's pocket into those funds. So I think impact certificates would be more expensive for Open Phil than if they'd funded the project directly.

I don't think the Future Fund has a DAF, but instead benefits from being based in The Bahamas, which doesn't have income or capital gains taxes. I expect that impact certificates would not really be more expensive for them because...FTX and SBF don't pay taxes at all?

I think most existing charitable funders look more like Open Phil here.

(Speaking for myself and not my employer.)

US tax law requires that US citizens pay income tax and capital gains tax, regardless of their physical/legal residency. Some limited deductions apply, but don't change the basic story.

I don't understand the incentives of the "oracular funder." What are some examples of organisations/individuals that would be oracular funders? Where would they have gotten their money? What would make them want to participate in this idea? What are their alternatives (in how they spend their money)? Would it be fair to say that if they decided not to take part in this, then the idea wouldn't work?

 

I appreciate your clarification!

Not the author, but this is my understanding assuming the idea holds:

OpenPhil/FTX/etc currently spend a lot of time & effort on evaluating grants. The idea here is that it's easier to evaluate a project after it's finished, e.g. "the area where bednets were given out had X% lower mortality, lives are valued at Y$, so this bednet intervention was worth Z$" - but current status quo is that you have to predict the value of X when evaluating the grant, and make a prediction on the likelihood that it works at all. (technically you'd predict a probability distribution over different values of X), which is much harder than measuring X after the intervention.

Investors (which are VC's and other people) take over the role of predicting the future, and oracular funders (openphil/FTX/etc) get an easier job and can make more accurate donations - making a larger impact than they would do otherwise. And good charity entrepenurs are rewarded more accurately, making more good for the world than if money was wasted on charities that at first glance seemed like a good idea but in fact has bad expected value.