From b11cf586789dbd7956761d2e1b93d10763b376fa Mon Sep 17 00:00:00 2001 From: Lily Jordan Date: Wed, 13 Mar 2024 11:35:20 -0400 Subject: [PATCH 01/16] writing --- app/about/approach/page.tsx | 165 ++++++++++++++++++++++++++++++++++++ 1 file changed, 165 insertions(+) create mode 100644 app/about/approach/page.tsx diff --git a/app/about/approach/page.tsx b/app/about/approach/page.tsx new file mode 100644 index 00000000..512f5fb4 --- /dev/null +++ b/app/about/approach/page.tsx @@ -0,0 +1,165 @@ +import { ArrowLongRightIcon } from '@heroicons/react/20/solid' + +export default function ApproachPage() { + return ( +
+

+ Manifund's Approach to Thinking About Charitable Funding +

+

+ From 2023 retrospective: "Our initial thesis was that grant applications and screening largely can be done in public, and should be." +

+

+ Economic Experimentation +

+

+ Manifund's sister company, Manifold, runs experiments with prediction markets. (make own, contests, leagues, a dating app, etc. emergent behaviors) + Because Manifold runs on play money, and because users get to participate on the market-creation + side as well as the prediction side, +

+

+ Manifund is different, because lots of real money is moving through our programs. But we're + still small and fast-moving ([example of grant turnaround time]), +

+

Manifund aims to be similar for charitable funding

+

Playground for experimenting with funding mechanisms

+

+ Programs Manifund has run +

+

+

Regranting

+

ACX grants

+

Impact certificates

+

Assurance contract auctions

+

+

+ Impact certificates +

+

+ Take impact certificates, for example. [Explainer.] Then: what's neat about this? No need to learn new ways to evaluate charities, or make people more altruistic. It just introduces an extra market mechanism. +

+

+ Each individual step of this process is perfectly understandable if you understand the analogous thing in private markets. But altogether, it's a lot of new moving parts to conceptualize at once. Can we come up with a general framework for thinking about these kinds of market designs? +

+

+ The PFEEP framework +

+

+ the challenges of each +

+

+ for each: who does it? possibly, how does it happen? and what is it coupled with? +

+ +

The startup model

+

The bootstrapping model

+

The wealthy philanthropist model

+

The classical Effective Altruism model

+

The GiveWell model

+

+ stuff +

+

The impact market model

+

+ stuff +

+

+ Single-step mechanisms +

+

+ Some mechanisms are solely designed for use on one or two steps of the framework. For example: +

+

+ A large swath of public-goods and charitable funding mechanisms can be understood as innovations + on one or two steps of this framework. +

+

+ Challenge prizes – pay out +

+

+ S-process – evaluate. SFF already does this (check). Need to watch the video to learn more. +

+

+ Advance market commitments – pay out. Might not be the right fit for Manifund because what would we buy? +

+

+ Income share agreements – predict and front. No added cost to run. +

+

+ Schelling-point oracles – evaluate. Would entail a whole webapp, jurors, and ways to pay them. +

+

+ Patronage – predict and front. Cost the average "sponsor me to explore in my career" application is looking for, times maybe 7 or 8 as a minimum set of Manifund Fellows or w/e. +

+

+ Crowdfunding – pay out. No added cost to run. +

+

+ Regranting – predict. (Check current regranting budget) +

+

+ Coupling and decoupling +

+

+ Design-bid-build – decouple evaluation from execution +

+

+ Coupling is for aligning incentives. Decoupling is for allowing specialization. (Principal-agent problems arise + when you do too much decoupling.) +

+

+ Psychological +

+

+ Reputation +

+

+ Leverage/responsiveness/power to affect the world +

+

+ Caveat: different value systems +

+

+ Altruism toward non-economic entities: future people, animals, children, ... +

+

+ Is altruism a coordination problem? +

+
+
    +
  • $1,000 at 10k valuation
    • +
  • $3,000 at 8k valuation
    • +
  • $2,000 at 6k valuation
    • +
  • $3,000 at 5k valuation
    • +
+
+ +
+
+

First three bids go through at a valuation of $6k

+
+
+

+ Effective iteration +

+

Iteration speed

+

Iteration volume

+
+ ) +} \ No newline at end of file From bad70af789b9541778384eee19b5077a62cc1ff2 Mon Sep 17 00:00:00 2001 From: Lily Jordan Date: Thu, 28 Mar 2024 15:05:25 -0700 Subject: [PATCH 02/16] add Articles section --- app/sidebar-item.tsx | 5 ++++- app/sidebar.tsx | 7 +++++-- 2 files changed, 9 insertions(+), 3 deletions(-) diff --git a/app/sidebar-item.tsx b/app/sidebar-item.tsx index 3ddcfd40..a63b4970 100644 --- a/app/sidebar-item.tsx +++ b/app/sidebar-item.tsx @@ -12,7 +12,8 @@ import { UserGroupIcon, GlobeAltIcon, ChevronRightIcon, - ChevronDownIcon + ChevronDownIcon, + BookOpenIcon } from '@heroicons/react/24/outline' import { SiteLink } from '@/components/site-link' @@ -83,6 +84,8 @@ function findIcon(name: string, isCurrentPage: boolean) { return case 'People': return + case 'Articles': + return default: return } diff --git a/app/sidebar.tsx b/app/sidebar.tsx index 6fbc4b9a..06e88e09 100644 --- a/app/sidebar.tsx +++ b/app/sidebar.tsx @@ -29,7 +29,7 @@ export default async function Sidebar() {
) @@ -87,6 +87,6 @@ function findIcon(name: string, isCurrentPage: boolean) { case 'Articles': return default: - return + return } } From f9d50e9fb4632b58cb9570a7f3049f7decacbe78 Mon Sep 17 00:00:00 2001 From: Lily Jordan Date: Thu, 28 Mar 2024 18:36:47 -0700 Subject: [PATCH 06/16] copy everything over from the notion --- app/articles/approach/page.tsx | 513 +++++++++++++++++++++++++++------ 1 file changed, 427 insertions(+), 86 deletions(-) diff --git a/app/articles/approach/page.tsx b/app/articles/approach/page.tsx index 512f5fb4..11b6343a 100644 --- a/app/articles/approach/page.tsx +++ b/app/articles/approach/page.tsx @@ -4,162 +4,503 @@ export default function ApproachPage() { return (

- Manifund's Approach to Thinking About Charitable Funding + Manifund’s Approach to Thinking About Charitable Funding

From 2023 retrospective: "Our initial thesis was that grant applications and screening largely can be done in public, and should be."

-

+

Economic Experimentation -

+ +

+ Manifund’s sister company, Manifold, runs experiments with + prediction markets. Because Manifold runs on play money, and because + users get to participate on the market-creation side as well as the + prediction side, it’s arguably the best playground in the world for + experimenting rapidly with financial mechanisms. Since official + currency isn’t involved, both Manifold and its users (and the two in + combination) can run experiments rapid-fire without jumping through + many hoops. For instance, Manifold has tried various types of + automatic market-making algorithms, mechanisms for multiple-choice + questions, short selling, limit orders, contests, leagues, even a + dating app spinoff. And users have made bots, non-question questions + like “___ stock (never resolves)”, markets that serve as bounties to + incentivize particular events to happen, questions with meta + resolution criteria like “Will the total volume of YES trading on + this question be more than x mana?”. +

- Manifund's sister company, Manifold, runs experiments with prediction markets. (make own, contests, leagues, a dating app, etc. emergent behaviors) - Because Manifold runs on play money, and because users get to participate on the market-creation - side as well as the prediction side, + Manifund is different, because lots of real money is moving through our + programs. But we’re still small and fast-moving ([example of grant + turnaround time]), and our ethos is similar: we aim to be an incubator + for lots of small- and medium-scale experiments testing out funding + mechanisms for charity and public goods.

- Manifund is different, because lots of real money is moving through our programs. But we're - still small and fast-moving ([example of grant turnaround time]), + Effective altruism originally asked the question, “What properties make + a charitable project useful, efficient, and valuable?” Manifund is an + attempt to extend that question in two ways:

-

Manifund aims to be similar for charitable funding

-

Playground for experimenting with funding mechanisms

-

+
    +
  1. + What kinds of systems incentivize projects that have those properties? +
    2. +
  2. + How can we use improvements to the grantee experience to make funding + programs more effective? +
    3. +
+

Programs Manifund has run -

+

-

Regranting

-

ACX grants

-

Impact certificates

-

Assurance contract auctions

+

Regranting

+

ACX grants

+

Impact certificates

+

Assurance contract auctions

-

- Impact certificates -

+

+ How to think about funding +

- Take impact certificates, for example. [Explainer.] Then: what's neat about this? No need to learn new ways to evaluate charities, or make people more altruistic. It just introduces an extra market mechanism. + Is there anything general we can say about these kinds of experiments?

- Each individual step of this process is perfectly understandable if you understand the analogous thing in private markets. But altogether, it's a lot of new moving parts to conceptualize at once. Can we come up with a general framework for thinking about these kinds of market designs? + Take the system of impact certificates, for example. There are a couple + of differences from how charity is traditionally funded. The donations + come after the fact, for one thing, rather than in advance. And + there’s a middleman involved, who doesn’t even necessarily have + to have altruistic motives. +

+

+ Theoretically, this leads to neat results. There’s no need to learn + new ways to evaluate charities, or to make people more altruistic. The + addition of an intermediate market makes it more efficient to figure + out the most efficient allocation of funds by the donor’s own + metrics.

-

- The PFEEP framework -

- the challenges of each + But also, it seems a little contrived? Each individual step of this + process is perfectly understandable if you understand the analogous + element of private markets. But altogether, it’s a lot of new moving + parts to conceptualize at once, and it’s not obvious how to come up + with similar ideas or how to evaluate them.

- for each: who does it? possibly, how does it happen? and what is it coupled with? + So we’ve come up with a first pass at a general framework for + thinking about the design of novel funding mechanisms.

-
    +
    1. Predict +

      + What does it entail? Make a good guess about which + projects will be likely to achieve their aims, what impacts those + achievements will have, and which teams can achieve the best + results for the lowest cost. +

      +

      + What are the desirables? Predict accurately and at + low cost. +

    2. - Front + Front +

      + What does it entail? Make a good guess about which + projects will be likely to achieve their aims, what impacts those + achievements will have, and which teams can achieve the best + results for the lowest cost. +

    3. - Execute + Execute +

      + What does it entail? Carry out the project. +

      +

      + What are the desirables? Do something people + actually want. Do it well. Do it inexpensively. Don’t run off + with the money. +

    4. - Evaluate + Evaluate +

      + What does it entail? Determine how valuable the + project was. +

      +

      + What are the desirables? Align the measure with + the target – that is, [explain further]. Account for + externalities– for example, [example of private funding][example of + big donor funding for reputation]. +

    5. - Pay out + Pay out +

      + What does it entail? Allocate money as a function + of the evaluation of the project’s impact. In some cases, like + traditional grantmaking, the fronting is also the payout (although + to some extent this is a question of definitions; you could also + think of traditional grantmaking as just skipping the fronting + step). +

      +

      + What are the desirables? Incentivize the execution + of projects with the most impact. +

      6. -
-

The startup model

-

The bootstrapping model

-

The wealthy philanthropist model

-

The classical Effective Altruism model

-

The GiveWell model

+ +

+ Some examples +

+

The startup model

- stuff + Predict: investors

-

The impact market model

- stuff + Front: investors

-

- Single-step mechanisms -

- Some mechanisms are solely designed for use on one or two steps of the framework. For example: + Execute: founder

- A large swath of public-goods and charitable funding mechanisms can be understood as innovations - on one or two steps of this framework. + Evaluate: consumers

- Challenge prizes – pay out + Pay out: consumers

- S-process – evaluate. SFF already does this (check). Need to watch the video to learn more. + Benefit: founder (as profit), investors (as profit), consumers (as consumer surplus)

+

The bootstrapping model

- Advance market commitments – pay out. Might not be the right fit for Manifund because what would we buy? + Predict: founder

- Income share agreements – predict and front. No added cost to run. + Front: founder

- Schelling-point oracles – evaluate. Would entail a whole webapp, jurors, and ways to pay them. + Execute: founder

- Patronage – predict and front. Cost the average "sponsor me to explore in my career" application is looking for, times maybe 7 or 8 as a minimum set of Manifund Fellows or w/e. + Evaluate: consumers

- Crowdfunding – pay out. No added cost to run. + Pay out: consumers

- Regranting – predict. (Check current regranting budget) + Benefit: founder (as profit), consumers (as consumer surplus)

-

- Coupling and decoupling +

The government model

+

+ Predict: elected officials or bureaucrats +

+

+ Front: citizens (as taxpayers) +

+

+ Execute: contractors +

+

+ Evaluate: citizens (indirectly, as voters) +

+

+ Pay out: citizens (as taxpayers) +

+

+ Benefit: citizens, contractors (as profit), officials (as goodwill for + subsequent elections) +

+

The wealthy philanthropist model

+

+ Predict: donor +

+

+ Front: donor +

+

+ Execute: donor +

+

+ Evaluate: donor +

+

+ Pay out: donor +

+

+ Benefit: charity beneficiaries +

+

The classical Effective Altruism model

+

+ Predict: donor, with an Our World In Data tab open +

+

+ Front: donor +

+

+ Execute: donor +

+

+ Evaluate: donor +

+

+ Pay out: donor +

+

+ Benefit: charity beneficiaries +

+

The charity evaluator model

+

+ Predict: charity evaluator +

+

+ Front: donors +

+

+ Execute: founders +

+

+ Evaluate: charity evaluator +

+

+ Pay out: donors +

+

+ Benefit: charity beneficiaries +

+

The impact certificate model

+

+ Predict: investors +

+

+ Front: investors +

+

+ Execute: founder +

+

+ Evaluate: donor +

+

+ Pay out: donor +

+

+ Benefit: founder (as profit), investors (as profit), charity beneficiaries +

+

The quadratic funding model

+

+ Predict: contributors +

+

+ Front: contributors and matching-pool sponsors +

+

+ Execute: founder +

+

+ Evaluate: contributors +

+

+ Pay out: contributors and matching-pool sponsors +

+

+ Benefit: contributors +

+

+
+ There are a couple of things to notice here. +

+

+ Some distinctions don’t always occur

- Design-bid-build – decouple evaluation from execution + “Predict” and “evaluate” are often the same, as are “front” and “pay + out”. These are for cases where money is given out ahead of time, in a + way that’s – well, not no-strings-attached, the founders have to make a + reasonable attempt and not just grab the money and run off to hide out in + Antarctica; but it’s not contingent on the project actually succeeding. + Maybe the company goes bust, the idea doesn’t work, the nonprofit falls + apart, whatever. Or maybe the project does what it says it will, but it + turns out no one really likes or wants the result after all. But when + there’s not a separate evaluation step and payout step, then those + contingencies are beyond the purview of the funding mechanism to + influence. Trying to fund stuff that won’t fail is baked into the + prediction and fronting steps.

- Coupling is for aligning incentives. Decoupling is for allowing specialization. (Principal-agent problems arise - when you do too much decoupling.) + This is especially true for new endeavors. With well-established + charities and companies, the ability to secure future funding is an + indirect incentive to execute current projects well. But for new + organizations that have no track record at all (or sometimes, even + new projects started by existing organizations), you have to fall + back on looking at the track record of the people founding them, and + if the founders are relatively unknown, there’s not much of an + incentive to give them a chance as opposed to taking safe bets. Taking + big swings that you’ll usually miss on, but that will occasionally have + wildly outsized returns, is sometimes called “hits-based giving”; some + effective-altruist organizations explicitly try to use this strategy, + and the reason it’s even necessary for them to specify that – after + all, hits-based investing is how venture capital works be default – is + that the incentives for an individual employee of a typical + philanthropic foundation, like the incentives for an individual + bureaucrat, are to be risk-averse and lean toward safe bets rather + than innovation.

-

- Psychological +

+ Some mechanisms are “full-stack”; others exist at one or two layers

- Reputation + Some mechanisms, like the examples mentioned above, offer an + implementation of the entire framework.

- Leverage/responsiveness/power to affect the world + But a large swath of mechanisms can be understood as innovations on + one or two of the steps. For example:

-

- Caveat: different value systems +

+ + Coupling and decoupling are central tools of funding-mechanism design + +

+

+ No mechanism listed here has all its six steps carried out by different + actors. In other words, various subsets of the steps are coupled + together: for instance, quadratic funding is able to price demand for a + public good because potential beneficiaries put up some of their own + money, which ensures they’re credibly signaling their real preference + for the good. Or for a more obvious example, democracy is more + responsive to the needs of its populace than monarchy is because + elected officials have to predict which policies will be popular in + order to win re-election. That’s because coupling two steps aligns + their incentives: combining an information-gathering step with a + money-moving step ensures that decision-makers have skin in the game. +

+

+ But there’s a limit to the benefits of coupling. The ultimate fully + coupled funding system is “everyone builds everything for themselves + and pays for it themselves.” Then the incentives are very aligned! + But on the other hand, it’s tremendously inefficient – no + specialization, no division of labor. +

+

+ So decoupling turns the knob the other way. For example, in a market + for impact certificates, instead of donors doing all the work of + assessing applicants and allocating funds (by making grants in + advance based on who looks most promising), you split the process + into two layers: investors, who are purely speculators, and donors, + who only have to evaluate projects’ impact afterwards. +

+

+ Another example is the design-bid-build system of construction + contracting. Instead of the government choosing a single bid for a + contractor to design and then carry out a construction project, + with design-bid-build they accept one contractor’s bid to create a + design, then put out a separate call for a (potentially) different + contractor to implement it; see Asterisk Magazine’s + interview + with urbanist Alon Levy, who argues that this framework + is more cost-effective than the single-bid (”design-build”) system in + most places that run it. +

+

+ In general, decoupling allows for improved competition and + specialization, which is especially useful in philanthropic cases, + where it’s hard to operate a single nonprofit as efficiently as a + corporation of the same size. But too much decoupling can cause + principal-agent problems, as often happens in bureaucracy, and can be + outweighed by economies of scale. +

+

+ Design-bid-build – decouple evaluation from execution +

+ +

+ Psychological

- Altruism toward non-economic entities: future people, animals, children, ... -

-

- Is altruism a coordination problem? -

-
-
    -
  • $1,000 at 10k valuation
    • -
  • $3,000 at 8k valuation
    • -
  • $2,000 at 6k valuation
    • -
  • $3,000 at 5k valuation
    • -
-
- -
-
-

First three bids go through at a valuation of $6k

-
-
+ Aside from all the financial considerations, there are a couple of social + and psychological levers entailed in grantmaking. One lesson Manifund has + gleaned from Manifold, our sister prediction-market company, is that you + can operate a finance-like system without any actual money involved, + because people also value: +

+
    +
  1. + reputation – the glory of visibly performing well and accumulating + scarce prestige, and +
    2. +
  2. + agency – the feeling that their actions cause something to come into + being that wouldn’t have otherwise. This is the sensation of knowing + that when you push the button, something will actually happen. +
    3. +
+

+ The slower-moving and less accessible a philanthropic organization is, + the harder it is for many individuals to achieve either of those two. On + a crowdfunding platform like Kickstarter, users can get a sense of + agency, but not so much of accumulating reputation. Anything that + resembles a game, like investing or predicting, and that has real-world + effects, hits both these criteria: you can succeed relative to others, + and you can succeed in the sense that you did some actions on a website + and now a real physical project is getting accomplished. +

Effective iteration

-

Iteration speed

-

Iteration volume

+

+ Caveat: different value systems +

+

+ One limitation of this model is that it’s not obvious how well it applies + to entities that don’t naturally fit into standard economic models because + they can’t easily participate in markets: children, animals, people who + might exist in the distant future. For people whose value systems heavily + incorporate any of these groups’ preferences, *all* market-based funding + solutions might prove inadequate. In these cases, funding is closer to a + strict question of altruism than a coordination problem; donors have to + speculate on behalf of those who can’t (financially) speculate for + themselves. These causes lend themselves better to direct improvements + in prediction and evaluation than to financial mechanisms that couple + those steps with fronting and paying out; it’s not really clear how you’d + align your own incentives with the incentives of someone in the year + 2300, except by saying “I care abstractly about making future people + happy.” +

+

+ Even in these areas, though, there are often some smaller subproblems + related to coordination, and so market mechanisms can offer at least + some improvements. For example, grant proposals often become the + victims of “funder chicken,” where each potential donor, having a + limited budget, likes a proposal but hopes one of the others will fund + it first. And even if you can’t evaluate one year from now whether a + project intended to improve outcomes on a century-long timescale has + succeeded in that, you can at least check whether the project did what + it said it would, or whether it achieved the kind of result that you + already believe for other reasons would be valuable for the far future. + It’s hard to incentivize terminal goals this way, but for incentivizing + instrumental goals, market structures are pretty good. +

+

+ On some level, you can think of altruism itself as a public good: + like, ten thousand people are all sad about cows getting factory-farmed + and would rather the cows be freed to roam on beautiful verdant + pastures, and so cow liberation is a public good to all those people + insofar as it’s something that would make them all happy. This doesn’t + necessarily evaluate how much the cow would prefer freedom to its + current life; you could just as easily imagine that ten thousand other + people would enjoy seeing a herd of cows dressed up in identical little + bow ties, and so purchasing a pack of wide-necked bow ties would be a + public-goods-funding coordination problem in just the same way. The + “figuring out what would be best for cows” problem, and below it the + more fundamental “figuring out what my altruistic values are” problem, + aren’t easily solvable by markets. But in either case, once you make + up your mind as to what your goals are, market mechanisms become useful + tools to help you get there. +

) } \ No newline at end of file From a8d9d36855f605914f43a9a4d57d05125232046c Mon Sep 17 00:00:00 2001 From: Lily Jordan Date: Fri, 29 Mar 2024 13:06:37 -0700 Subject: [PATCH 07/16] sub-nested and default-open children --- app/sidebar-item.tsx | 7 ++++--- app/sidebar.tsx | 9 ++++++++- 2 files changed, 12 insertions(+), 4 deletions(-) diff --git a/app/sidebar-item.tsx b/app/sidebar-item.tsx index ce8cd671..71254cce 100644 --- a/app/sidebar-item.tsx +++ b/app/sidebar-item.tsx @@ -21,13 +21,14 @@ export type Item = { name: string href?: string children?: Item[] + childrenDefaultOpen?: boolean } export function SidebarItem(props: { item: Item }) { const { item } = props const isCurrentPage = item.href === usePathname() && item.href != null const icon = findIcon(item.name, isCurrentPage) - const [isOpen, setIsOpen] = useState(false) + const [isOpen, setIsOpen] = useState(item.childrenDefaultOpen) const sidebarItem = (
- {item.children.map((subItem) => ( - + {item.children.map((child) => ( + ))}
)} diff --git a/app/sidebar.tsx b/app/sidebar.tsx index c9c2fa6d..389e953f 100644 --- a/app/sidebar.tsx +++ b/app/sidebar.tsx @@ -68,7 +68,14 @@ export default async function Sidebar() { {user && ( Date: Fri, 29 Mar 2024 15:06:26 -0700 Subject: [PATCH 08/16] sidebar scroll --- app/articles/approach/1/page.tsx | 504 +++++++++++++++++++++++++++++++ app/articles/approach/2/page.tsx | 18 ++ app/articles/approach/3/page.tsx | 40 +++ app/articles/approach/4/page.tsx | 9 + app/articles/approach/5/page.tsx | 9 + app/articles/approach/6/page.tsx | 9 + app/sidebar.tsx | 90 +++--- 7 files changed, 640 insertions(+), 39 deletions(-) create mode 100644 app/articles/approach/1/page.tsx create mode 100644 app/articles/approach/2/page.tsx create mode 100644 app/articles/approach/3/page.tsx create mode 100644 app/articles/approach/4/page.tsx create mode 100644 app/articles/approach/5/page.tsx create mode 100644 app/articles/approach/6/page.tsx diff --git a/app/articles/approach/1/page.tsx b/app/articles/approach/1/page.tsx new file mode 100644 index 00000000..c941a057 --- /dev/null +++ b/app/articles/approach/1/page.tsx @@ -0,0 +1,504 @@ +export default function ApproachPage1() { + return ( +
+

+ Manifund’s Approach to Thinking About Charitable Funding +

+

+ From 2023 retrospective: "Our initial thesis was that grant applications and screening largely can be done in public, and should be." +

+

+ Economic Experimentation +

+

+ Manifund’s sister company, Manifold, runs experiments with + prediction markets. Because Manifold runs on play money, and because + users get to participate on the market-creation side as well as the + prediction side, it’s arguably the best playground in the world for + experimenting rapidly with financial mechanisms. Since official + currency isn’t involved, both Manifold and its users (and the two in + combination) can run experiments rapid-fire without jumping through + many hoops. For instance, Manifold has tried various types of + automatic market-making algorithms, mechanisms for multiple-choice + questions, short selling, limit orders, contests, leagues, even a + dating app spinoff. And users have made bots, non-question questions + like “___ stock (never resolves)”, markets that serve as bounties to + incentivize particular events to happen, questions with meta + resolution criteria like “Will the total volume of YES trading on + this question be more than x mana?”. +

+

+ Manifund is different, because lots of real money is moving through our + programs. But we’re still small and fast-moving ([example of grant + turnaround time]), and our ethos is similar: we aim to be an incubator + for lots of small- and medium-scale experiments testing out funding + mechanisms for charity and public goods. +

+

+ Effective altruism originally asked the question, “What properties make + a charitable project useful, efficient, and valuable?” Manifund is an + attempt to extend that question in two ways: +

+
    +
  1. + What kinds of systems incentivize projects that have those properties? +
    2. +
  2. + How can we use improvements to the grantee experience to make funding + programs more effective? +
    3. +
+

+ Programs Manifund has run +

+

+

Regranting

+

ACX grants

+

Impact certificates

+

Assurance contract auctions

+

+

+ how to think about funding +

+

+ is there anything general we can say about these kinds of experiments? +

+

+ take the system of impact certificates, for example. there are a couple + of differences from how charity is traditionally funded. the donations + come after the fact, for one thing, rather than in advance. and + there’s a middleman involved, who doesn’t even necessarily have + to have altruistic motives. +

+

+ theoretically, this leads to neat results. there’s no need to learn + new ways to evaluate charities, or to make people more altruistic. the + addition of an intermediate market makes it more efficient to figure + out the most efficient allocation of funds by the donor’s own + metrics. +

+

+ but also, it seems a little contrived? each individual step of this + process is perfectly understandable if you understand the analogous + element of private markets. but altogether, it’s a lot of new moving + parts to conceptualize at once, and it’s not obvious how to come up + with similar ideas or how to evaluate them. +

+

+ so we’ve come up with a first pass at a general framework for + thinking about the design of novel funding mechanisms. +

+
    +
  1. + predict +

    + what does it entail? make a good guess about which + projects will be likely to achieve their aims, what impacts those + achievements will have, and which teams can achieve the best + results for the lowest cost. +

    +

    + what are the desirables? predict accurately and at + low cost. +

    +
    2. +
  2. + front +

    + what does it entail? make a good guess about which + projects will be likely to achieve their aims, what impacts those + achievements will have, and which teams can achieve the best + results for the lowest cost. +

    +
    3. +
  3. + execute +

    + what does it entail? carry out the project. +

    +

    + what are the desirables? do something people + actually want. do it well. do it inexpensively. don’t run off + with the money. +

    +
    4. +
  4. + evaluate +

    + what does it entail? determine how valuable the + project was. +

    +

    + what are the desirables? align the measure with + the target – that is, [explain further]. account for + externalities– for example, [example of private funding][example of + big donor funding for reputation]. +

    +
    5. +
  5. + pay out +

    + what does it entail? allocate money as a function + of the evaluation of the project’s impact. in some cases, like + traditional grantmaking, the fronting is also the payout (although + to some extent this is a question of definitions; you could also + think of traditional grantmaking as just skipping the fronting + step). +

    +

    + what are the desirables? incentivize the execution + of projects with the most impact. +

    +
    6. +
+

+ some examples +

+

the startup model

+

+ predict: investors +

+

+ front: investors +

+

+ execute: founder +

+

+ evaluate: consumers +

+

+ pay out: consumers +

+

+ benefit: founder (as profit), investors (as profit), consumers (as consumer surplus) +

+

the bootstrapping model

+

+ predict: founder +

+

+ front: founder +

+

+ execute: founder +

+

+ evaluate: consumers +

+

+ pay out: consumers +

+

+ benefit: founder (as profit), consumers (as consumer surplus) +

+

the government model

+

+ predict: elected officials or bureaucrats +

+

+ front: citizens (as taxpayers) +

+

+ execute: contractors +

+

+ evaluate: citizens (indirectly, as voters) +

+

+ pay out: citizens (as taxpayers) +

+

+ benefit: citizens, contractors (as profit), officials (as goodwill for + subsequent elections) +

+

the wealthy philanthropist model

+

+ predict: donor +

+

+ front: donor +

+

+ execute: donor +

+

+ evaluate: donor +

+

+ pay out: donor +

+

+ benefit: charity beneficiaries +

+

the classical effective altruism model

+

+ predict: donor, with an our world in data tab open +

+

+ front: donor +

+

+ execute: donor +

+

+ evaluate: donor +

+

+ pay out: donor +

+

+ benefit: charity beneficiaries +

+

the charity evaluator model

+

+ predict: charity evaluator +

+

+ front: donors +

+

+ execute: founders +

+

+ evaluate: charity evaluator +

+

+ pay out: donors +

+

+ benefit: charity beneficiaries +

+

the impact certificate model

+

+ predict: investors +

+

+ front: investors +

+

+ execute: founder +

+

+ evaluate: donor +

+

+ pay out: donor +

+

+ benefit: founder (as profit), investors (as profit), charity beneficiaries +

+

the quadratic funding model

+

+ predict: contributors +

+

+ front: contributors and matching-pool sponsors +

+

+ execute: founder +

+

+ evaluate: contributors +

+

+ pay out: contributors and matching-pool sponsors +

+

+ benefit: contributors +

+

+
+ there are a couple of things to notice here. +

+

+ some distinctions don’t always occur +

+

+ “predict” and “evaluate” are often the same, as are “front” and “pay + out”. these are for cases where money is given out ahead of time, in a + way that’s – well, not no-strings-attached, the founders have to make a + reasonable attempt and not just grab the money and run off to hide out in + antarctica; but it’s not contingent on the project actually succeeding. + maybe the company goes bust, the idea doesn’t work, the nonprofit falls + apart, whatever. or maybe the project does what it says it will, but it + turns out no one really likes or wants the result after all. but when + there’s not a separate evaluation step and payout step, then those + contingencies are beyond the purview of the funding mechanism to + influence. trying to fund stuff that won’t fail is baked into the + prediction and fronting steps. +

+

+ this is especially true for new endeavors. with well-established + charities and companies, the ability to secure future funding is an + indirect incentive to execute current projects well. but for new + organizations that have no track record at all (or sometimes, even + new projects started by existing organizations), you have to fall + back on looking at the track record of the people founding them, and + if the founders are relatively unknown, there’s not much of an + incentive to give them a chance as opposed to taking safe bets. taking + big swings that you’ll usually miss on, but that will occasionally have + wildly outsized returns, is sometimes called “hits-based giving”; some + effective-altruist organizations explicitly try to use this strategy, + and the reason it’s even necessary for them to specify that – after + all, hits-based investing is how venture capital works be default – is + that the incentives for an individual employee of a typical + philanthropic foundation, like the incentives for an individual + bureaucrat, are to be risk-averse and lean toward safe bets rather + than innovation. +

+

+ some mechanisms are “full-stack”; others exist at one or two layers +

+

+ some mechanisms, like the examples mentioned above, offer an + implementation of the entire framework. +

+

+ but a large swath of mechanisms can be understood as innovations on + one or two of the steps. for example: +

+

+ + coupling and decoupling are central tools of funding-mechanism design + +

+

+ no mechanism listed here has all its six steps carried out by different + actors. in other words, various subsets of the steps are coupled + together: for instance, quadratic funding is able to price demand for a + public good because potential beneficiaries put up some of their own + money, which ensures they’re credibly signaling their real preference + for the good. or for a more obvious example, democracy is more + responsive to the needs of its populace than monarchy is because + elected officials have to predict which policies will be popular in + order to win re-election. that’s because coupling two steps aligns + their incentives: combining an information-gathering step with a + money-moving step ensures that decision-makers have skin in the game. +

+

+ but there’s a limit to the benefits of coupling. the ultimate fully + coupled funding system is “everyone builds everything for themselves + and pays for it themselves.” then the incentives are very aligned! + but on the other hand, it’s tremendously inefficient – no + specialization, no division of labor. +

+

+ so decoupling turns the knob the other way. for example, in a market + for impact certificates, instead of donors doing all the work of + assessing applicants and allocating funds (by making grants in + advance based on who looks most promising), you split the process + into two layers: investors, who are purely speculators, and donors, + who only have to evaluate projects’ impact afterwards. +

+

+ another example is the design-bid-build system of construction + contracting. instead of the government choosing a single bid for a + contractor to design and then carry out a construction project, + with design-bid-build they accept one contractor’s bid to create a + design, then put out a separate call for a (potentially) different + contractor to implement it; see asterisk magazine’s + interview + with urbanist alon levy, who argues that this framework + is more cost-effective than the single-bid (”design-build”) system in + most places that run it. +

+

+ in general, decoupling allows for improved competition and + specialization, which is especially useful in philanthropic cases, + where it’s hard to operate a single nonprofit as efficiently as a + corporation of the same size. but too much decoupling can cause + principal-agent problems, as often happens in bureaucracy, and can be + outweighed by economies of scale. +

+

+ design-bid-build – decouple evaluation from execution +

+ +

+ psychological +

+

+ aside from all the financial considerations, there are a couple of social + and psychological levers entailed in grantmaking. one lesson manifund has + gleaned from manifold, our sister prediction-market company, is that you + can operate a finance-like system without any actual money involved, + because people also value: +

+
    +
  1. + reputation – the glory of visibly performing well and accumulating + scarce prestige, and +
    2. +
  2. + agency – the feeling that their actions cause something to come into + being that wouldn’t have otherwise. this is the sensation of knowing + that when you push the button, something will actually happen. +
    3. +
+

+ the slower-moving and less accessible a philanthropic organization is, + the harder it is for many individuals to achieve either of those two. on + a crowdfunding platform like kickstarter, users can get a sense of + agency, but not so much of accumulating reputation. anything that + resembles a game, like investing or predicting, and that has real-world + effects, hits both these criteria: you can succeed relative to others, + and you can succeed in the sense that you did some actions on a website + and now a real physical project is getting accomplished. +

+

+ effective iteration +

+

+ caveat: different value systems +

+

+ one limitation of this model is that it’s not obvious how well it applies + to entities that don’t naturally fit into standard economic models because + they can’t easily participate in markets: children, animals, people who + might exist in the distant future. for people whose value systems heavily + incorporate any of these groups’ preferences, *all* market-based funding + solutions might prove inadequate. in these cases, funding is closer to a + strict question of altruism than a coordination problem; donors have to + speculate on behalf of those who can’t (financially) speculate for + themselves. these causes lend themselves better to direct improvements + in prediction and evaluation than to financial mechanisms that couple + those steps with fronting and paying out; it’s not really clear how you’d + align your own incentives with the incentives of someone in the year + 2300, except by saying “i care abstractly about making future people + happy.” +

+

+ even in these areas, though, there are often some smaller subproblems + related to coordination, and so market mechanisms can offer at least + some improvements. for example, grant proposals often become the + victims of “funder chicken,” where each potential donor, having a + limited budget, likes a proposal but hopes one of the others will fund + it first. and even if you can’t evaluate one year from now whether a + project intended to improve outcomes on a century-long timescale has + succeeded in that, you can at least check whether the project did what + it said it would, or whether it achieved the kind of result that you + already believe for other reasons would be valuable for the far future. + it’s hard to incentivize terminal goals this way, but for incentivizing + instrumental goals, market structures are pretty good. +

+

+ on some level, you can think of altruism itself as a public good: + like, ten thousand people are all sad about cows getting factory-farmed + and would rather the cows be freed to roam on beautiful verdant + pastures, and so cow liberation is a public good to all those people + insofar as it’s something that would make them all happy. this doesn’t + necessarily evaluate how much the cow would prefer freedom to its + current life; you could just as easily imagine that ten thousand other + people would enjoy seeing a herd of cows dressed up in identical little + bow ties, and so purchasing a pack of wide-necked bow ties would be a + public-goods-funding coordination problem in just the same way. the + “figuring out what would be best for cows” problem, and below it the + more fundamental “figuring out what my altruistic values are” problem, + aren’t easily solvable by markets. but in either case, once you make + up your mind as to what your goals are, market mechanisms become useful + tools to help you get there. +

+
+ ) +} \ No newline at end of file diff --git a/app/articles/approach/2/page.tsx b/app/articles/approach/2/page.tsx new file mode 100644 index 00000000..5e229b79 --- /dev/null +++ b/app/articles/approach/2/page.tsx @@ -0,0 +1,18 @@ +export default function ApproachPage2() { + return ( +
+

+ Manifund’s Approach to Thinking About Charitable Funding +

+

+ Programs Manifund has run +

+

+

Regranting

+

ACX grants

+

Impact certificates

+

Assurance contract auctions

+

+
+ ) +} \ No newline at end of file diff --git a/app/articles/approach/3/page.tsx b/app/articles/approach/3/page.tsx new file mode 100644 index 00000000..2d61632a --- /dev/null +++ b/app/articles/approach/3/page.tsx @@ -0,0 +1,40 @@ +export default function ApproachPage3() { + return ( +
+

+ Manifund’s Approach to Thinking About Charitable Funding +

+

+ How to think about funding +

+

+ Is there anything general we can say about these kinds of experiments? +

+

+ Take the system of impact certificates, for example. There are a couple + of differences from how charity is traditionally funded. The donations + come after the fact, for one thing, rather than in advance. And + there’s a middleman involved, who doesn’t even necessarily have + to have altruistic motives. +

+

+ Theoretically, this leads to neat results. There’s no need to learn + new ways to evaluate charities, or to make people more altruistic. The + addition of an intermediate market makes it more efficient to figure + out the most efficient allocation of funds by the donor’s own + metrics. +

+

+ But also, it seems a little contrived? Each individual step of this + process is perfectly understandable if you understand the analogous + element of private markets. But altogether, it’s a lot of new moving + parts to conceptualize at once, and it’s not obvious how to come up + with similar ideas or how to evaluate them. +

+

+ So we’ve come up with a first pass at a general framework for + thinking about the design of novel funding mechanisms. +

+
+ ) +} \ No newline at end of file diff --git a/app/articles/approach/4/page.tsx b/app/articles/approach/4/page.tsx new file mode 100644 index 00000000..faa7dbce --- /dev/null +++ b/app/articles/approach/4/page.tsx @@ -0,0 +1,9 @@ +export default function ApproachPage4() { + return ( +
+

+ Manifund’s Approach to Thinking About Charitable Funding +

+
+ ) +} \ No newline at end of file diff --git a/app/articles/approach/5/page.tsx b/app/articles/approach/5/page.tsx new file mode 100644 index 00000000..faa7dbce --- /dev/null +++ b/app/articles/approach/5/page.tsx @@ -0,0 +1,9 @@ +export default function ApproachPage4() { + return ( +
+

+ Manifund’s Approach to Thinking About Charitable Funding +

+
+ ) +} \ No newline at end of file diff --git a/app/articles/approach/6/page.tsx b/app/articles/approach/6/page.tsx new file mode 100644 index 00000000..d4797e37 --- /dev/null +++ b/app/articles/approach/6/page.tsx @@ -0,0 +1,9 @@ +export default function ApproachPage6() { + return ( +
+

+ Manifund’s Approach to Thinking About Charitable Funding +

+
+ ) +} \ No newline at end of file diff --git a/app/sidebar.tsx b/app/sidebar.tsx index 389e953f..969ac486 100644 --- a/app/sidebar.tsx +++ b/app/sidebar.tsx @@ -31,7 +31,7 @@ export default async function Sidebar() { aria-label="Sidebar" className="fixed top-0 hidden h-[100vh] w-56 justify-between divide-gray-300 self-start pl-2 pr-2 pt-10 lg:col-span-3 lg:flex lg:flex-col" > - +
} {user === undefined &&
} - - {user === null && ( - - )} - - - - - {user && ( - - {isRegranter ? 'Give a grant' : 'Create a project'} - - )} + + + {user === null && ( + + )} + + + + + {user && ( + + {isRegranter ? 'Give a grant' : 'Create a project'} + + )} + Date: Fri, 29 Mar 2024 15:51:40 -0700 Subject: [PATCH 09/16] all pages --- app/articles/approach/1/page.tsx | 453 +----------------------------- app/articles/approach/10/page.tsx | 59 ++++ app/articles/approach/4/page.tsx | 63 +++++ app/articles/approach/5/page.tsx | 53 +++- app/articles/approach/6/page.tsx | 158 ++++++++++- app/articles/approach/7/page.tsx | 62 ++++ app/articles/approach/8/page.tsx | 40 +++ app/articles/approach/9/page.tsx | 18 ++ app/sidebar.tsx | 2 - 9 files changed, 452 insertions(+), 456 deletions(-) create mode 100644 app/articles/approach/10/page.tsx create mode 100644 app/articles/approach/7/page.tsx create mode 100644 app/articles/approach/8/page.tsx create mode 100644 app/articles/approach/9/page.tsx diff --git a/app/articles/approach/1/page.tsx b/app/articles/approach/1/page.tsx index c941a057..1af18110 100644 --- a/app/articles/approach/1/page.tsx +++ b/app/articles/approach/1/page.tsx @@ -10,7 +10,7 @@ export default function ApproachPage1() {

Economic Experimentation

-

+

Manifund’s sister company, Manifold, runs experiments with prediction markets. Because Manifold runs on play money, and because users get to participate on the market-creation side as well as the @@ -48,457 +48,6 @@ export default function ApproachPage1() { programs more effective? -

- Programs Manifund has run -

-

-

Regranting

-

ACX grants

-

Impact certificates

-

Assurance contract auctions

-

-

- how to think about funding -

-

- is there anything general we can say about these kinds of experiments? -

-

- take the system of impact certificates, for example. there are a couple - of differences from how charity is traditionally funded. the donations - come after the fact, for one thing, rather than in advance. and - there’s a middleman involved, who doesn’t even necessarily have - to have altruistic motives. -

-

- theoretically, this leads to neat results. there’s no need to learn - new ways to evaluate charities, or to make people more altruistic. the - addition of an intermediate market makes it more efficient to figure - out the most efficient allocation of funds by the donor’s own - metrics. -

-

- but also, it seems a little contrived? each individual step of this - process is perfectly understandable if you understand the analogous - element of private markets. but altogether, it’s a lot of new moving - parts to conceptualize at once, and it’s not obvious how to come up - with similar ideas or how to evaluate them. -

-

- so we’ve come up with a first pass at a general framework for - thinking about the design of novel funding mechanisms. -

-
    -
  1. - predict -

    - what does it entail? make a good guess about which - projects will be likely to achieve their aims, what impacts those - achievements will have, and which teams can achieve the best - results for the lowest cost. -

    -

    - what are the desirables? predict accurately and at - low cost. -

    -
    2. -
  2. - front -

    - what does it entail? make a good guess about which - projects will be likely to achieve their aims, what impacts those - achievements will have, and which teams can achieve the best - results for the lowest cost. -

    -
    3. -
  3. - execute -

    - what does it entail? carry out the project. -

    -

    - what are the desirables? do something people - actually want. do it well. do it inexpensively. don’t run off - with the money. -

    -
    4. -
  4. - evaluate -

    - what does it entail? determine how valuable the - project was. -

    -

    - what are the desirables? align the measure with - the target – that is, [explain further]. account for - externalities– for example, [example of private funding][example of - big donor funding for reputation]. -

    -
    5. -
  5. - pay out -

    - what does it entail? allocate money as a function - of the evaluation of the project’s impact. in some cases, like - traditional grantmaking, the fronting is also the payout (although - to some extent this is a question of definitions; you could also - think of traditional grantmaking as just skipping the fronting - step). -

    -

    - what are the desirables? incentivize the execution - of projects with the most impact. -

    -
    6. -
-

- some examples -

-

the startup model

-

- predict: investors -

-

- front: investors -

-

- execute: founder -

-

- evaluate: consumers -

-

- pay out: consumers -

-

- benefit: founder (as profit), investors (as profit), consumers (as consumer surplus) -

-

the bootstrapping model

-

- predict: founder -

-

- front: founder -

-

- execute: founder -

-

- evaluate: consumers -

-

- pay out: consumers -

-

- benefit: founder (as profit), consumers (as consumer surplus) -

-

the government model

-

- predict: elected officials or bureaucrats -

-

- front: citizens (as taxpayers) -

-

- execute: contractors -

-

- evaluate: citizens (indirectly, as voters) -

-

- pay out: citizens (as taxpayers) -

-

- benefit: citizens, contractors (as profit), officials (as goodwill for - subsequent elections) -

-

the wealthy philanthropist model

-

- predict: donor -

-

- front: donor -

-

- execute: donor -

-

- evaluate: donor -

-

- pay out: donor -

-

- benefit: charity beneficiaries -

-

the classical effective altruism model

-

- predict: donor, with an our world in data tab open -

-

- front: donor -

-

- execute: donor -

-

- evaluate: donor -

-

- pay out: donor -

-

- benefit: charity beneficiaries -

-

the charity evaluator model

-

- predict: charity evaluator -

-

- front: donors -

-

- execute: founders -

-

- evaluate: charity evaluator -

-

- pay out: donors -

-

- benefit: charity beneficiaries -

-

the impact certificate model

-

- predict: investors -

-

- front: investors -

-

- execute: founder -

-

- evaluate: donor -

-

- pay out: donor -

-

- benefit: founder (as profit), investors (as profit), charity beneficiaries -

-

the quadratic funding model

-

- predict: contributors -

-

- front: contributors and matching-pool sponsors -

-

- execute: founder -

-

- evaluate: contributors -

-

- pay out: contributors and matching-pool sponsors -

-

- benefit: contributors -

-

-
- there are a couple of things to notice here. -

-

- some distinctions don’t always occur -

-

- “predict” and “evaluate” are often the same, as are “front” and “pay - out”. these are for cases where money is given out ahead of time, in a - way that’s – well, not no-strings-attached, the founders have to make a - reasonable attempt and not just grab the money and run off to hide out in - antarctica; but it’s not contingent on the project actually succeeding. - maybe the company goes bust, the idea doesn’t work, the nonprofit falls - apart, whatever. or maybe the project does what it says it will, but it - turns out no one really likes or wants the result after all. but when - there’s not a separate evaluation step and payout step, then those - contingencies are beyond the purview of the funding mechanism to - influence. trying to fund stuff that won’t fail is baked into the - prediction and fronting steps. -

-

- this is especially true for new endeavors. with well-established - charities and companies, the ability to secure future funding is an - indirect incentive to execute current projects well. but for new - organizations that have no track record at all (or sometimes, even - new projects started by existing organizations), you have to fall - back on looking at the track record of the people founding them, and - if the founders are relatively unknown, there’s not much of an - incentive to give them a chance as opposed to taking safe bets. taking - big swings that you’ll usually miss on, but that will occasionally have - wildly outsized returns, is sometimes called “hits-based giving”; some - effective-altruist organizations explicitly try to use this strategy, - and the reason it’s even necessary for them to specify that – after - all, hits-based investing is how venture capital works be default – is - that the incentives for an individual employee of a typical - philanthropic foundation, like the incentives for an individual - bureaucrat, are to be risk-averse and lean toward safe bets rather - than innovation. -

-

- some mechanisms are “full-stack”; others exist at one or two layers -

-

- some mechanisms, like the examples mentioned above, offer an - implementation of the entire framework. -

-

- but a large swath of mechanisms can be understood as innovations on - one or two of the steps. for example: -

-

- - coupling and decoupling are central tools of funding-mechanism design - -

-

- no mechanism listed here has all its six steps carried out by different - actors. in other words, various subsets of the steps are coupled - together: for instance, quadratic funding is able to price demand for a - public good because potential beneficiaries put up some of their own - money, which ensures they’re credibly signaling their real preference - for the good. or for a more obvious example, democracy is more - responsive to the needs of its populace than monarchy is because - elected officials have to predict which policies will be popular in - order to win re-election. that’s because coupling two steps aligns - their incentives: combining an information-gathering step with a - money-moving step ensures that decision-makers have skin in the game. -

-

- but there’s a limit to the benefits of coupling. the ultimate fully - coupled funding system is “everyone builds everything for themselves - and pays for it themselves.” then the incentives are very aligned! - but on the other hand, it’s tremendously inefficient – no - specialization, no division of labor. -

-

- so decoupling turns the knob the other way. for example, in a market - for impact certificates, instead of donors doing all the work of - assessing applicants and allocating funds (by making grants in - advance based on who looks most promising), you split the process - into two layers: investors, who are purely speculators, and donors, - who only have to evaluate projects’ impact afterwards. -

-

- another example is the design-bid-build system of construction - contracting. instead of the government choosing a single bid for a - contractor to design and then carry out a construction project, - with design-bid-build they accept one contractor’s bid to create a - design, then put out a separate call for a (potentially) different - contractor to implement it; see asterisk magazine’s - interview - with urbanist alon levy, who argues that this framework - is more cost-effective than the single-bid (”design-build”) system in - most places that run it. -

-

- in general, decoupling allows for improved competition and - specialization, which is especially useful in philanthropic cases, - where it’s hard to operate a single nonprofit as efficiently as a - corporation of the same size. but too much decoupling can cause - principal-agent problems, as often happens in bureaucracy, and can be - outweighed by economies of scale. -

-

- design-bid-build – decouple evaluation from execution -

- -

- psychological -

-

- aside from all the financial considerations, there are a couple of social - and psychological levers entailed in grantmaking. one lesson manifund has - gleaned from manifold, our sister prediction-market company, is that you - can operate a finance-like system without any actual money involved, - because people also value: -

-
    -
  1. - reputation – the glory of visibly performing well and accumulating - scarce prestige, and -
    2. -
  2. - agency – the feeling that their actions cause something to come into - being that wouldn’t have otherwise. this is the sensation of knowing - that when you push the button, something will actually happen. -
    3. -
-

- the slower-moving and less accessible a philanthropic organization is, - the harder it is for many individuals to achieve either of those two. on - a crowdfunding platform like kickstarter, users can get a sense of - agency, but not so much of accumulating reputation. anything that - resembles a game, like investing or predicting, and that has real-world - effects, hits both these criteria: you can succeed relative to others, - and you can succeed in the sense that you did some actions on a website - and now a real physical project is getting accomplished. -

-

- effective iteration -

-

- caveat: different value systems -

-

- one limitation of this model is that it’s not obvious how well it applies - to entities that don’t naturally fit into standard economic models because - they can’t easily participate in markets: children, animals, people who - might exist in the distant future. for people whose value systems heavily - incorporate any of these groups’ preferences, *all* market-based funding - solutions might prove inadequate. in these cases, funding is closer to a - strict question of altruism than a coordination problem; donors have to - speculate on behalf of those who can’t (financially) speculate for - themselves. these causes lend themselves better to direct improvements - in prediction and evaluation than to financial mechanisms that couple - those steps with fronting and paying out; it’s not really clear how you’d - align your own incentives with the incentives of someone in the year - 2300, except by saying “i care abstractly about making future people - happy.” -

-

- even in these areas, though, there are often some smaller subproblems - related to coordination, and so market mechanisms can offer at least - some improvements. for example, grant proposals often become the - victims of “funder chicken,” where each potential donor, having a - limited budget, likes a proposal but hopes one of the others will fund - it first. and even if you can’t evaluate one year from now whether a - project intended to improve outcomes on a century-long timescale has - succeeded in that, you can at least check whether the project did what - it said it would, or whether it achieved the kind of result that you - already believe for other reasons would be valuable for the far future. - it’s hard to incentivize terminal goals this way, but for incentivizing - instrumental goals, market structures are pretty good. -

-

- on some level, you can think of altruism itself as a public good: - like, ten thousand people are all sad about cows getting factory-farmed - and would rather the cows be freed to roam on beautiful verdant - pastures, and so cow liberation is a public good to all those people - insofar as it’s something that would make them all happy. this doesn’t - necessarily evaluate how much the cow would prefer freedom to its - current life; you could just as easily imagine that ten thousand other - people would enjoy seeing a herd of cows dressed up in identical little - bow ties, and so purchasing a pack of wide-necked bow ties would be a - public-goods-funding coordination problem in just the same way. the - “figuring out what would be best for cows” problem, and below it the - more fundamental “figuring out what my altruistic values are” problem, - aren’t easily solvable by markets. but in either case, once you make - up your mind as to what your goals are, market mechanisms become useful - tools to help you get there. -

) } \ No newline at end of file diff --git a/app/articles/approach/10/page.tsx b/app/articles/approach/10/page.tsx new file mode 100644 index 00000000..45789555 --- /dev/null +++ b/app/articles/approach/10/page.tsx @@ -0,0 +1,59 @@ +export default function ApproachPage4() { + return ( +
+

+ Manifund’s Approach to Thinking About Charitable Funding +

+

+ caveat: different value systems +

+

+ one limitation of this model is that it’s not obvious how well it applies + to entities that don’t naturally fit into standard economic models because + they can’t easily participate in markets: children, animals, people who + might exist in the distant future. for people whose value systems heavily + incorporate any of these groups’ preferences, all market-based funding + solutions might prove inadequate. in these cases, funding is closer to a + strict question of altruism than a coordination problem; donors have to + speculate on behalf of those who can’t (financially) speculate for + themselves. these causes lend themselves better to direct improvements + in prediction and evaluation than to financial mechanisms that couple + those steps with fronting and paying out; it’s not really clear how you’d + align your own incentives with the incentives of someone in the year + 2300, except by saying “i care abstractly about making future people + happy.” +

+

+ even in these areas, though, there are often some smaller subproblems + related to coordination, and so market mechanisms can offer at least + some improvements. for example, grant proposals often become the + victims of “funder chicken,” where each potential donor, having a + limited budget, likes a proposal but hopes one of the others will fund + it first. and even if you can’t evaluate one year from now whether a + project intended to improve outcomes on a century-long timescale has + succeeded in that, you can at least check whether the project did what + it said it would, or whether it achieved the kind of result that you + already believe for other reasons would be valuable for the far future. + it’s hard to incentivize terminal goals this way, but for incentivizing + instrumental goals, market structures are pretty good. +

+

+ on some level, you can think of altruism itself as a public good: + like, ten thousand people are all sad about cows getting factory-farmed + and would rather the cows be freed to roam on beautiful verdant + pastures, and so cow liberation is a public good to all those people + insofar as it’s something that would make them all happy. this doesn’t + necessarily evaluate how much the cow would prefer freedom to its + current life; you could just as easily imagine that ten thousand other + people would enjoy seeing a herd of cows dressed up in identical little + bow ties, and so purchasing a pack of wide-necked bow ties would be a + public-goods-funding coordination problem in just the same way. the + “figuring out what would be best for cows” problem, and below it the + more fundamental “figuring out what my altruistic values are” problem, + aren’t easily solvable by markets. but in either case, once you make + up your mind as to what your goals are, market mechanisms become useful + tools to help you get there. +

+
+ ) +} \ No newline at end of file diff --git a/app/articles/approach/4/page.tsx b/app/articles/approach/4/page.tsx index faa7dbce..0a24a9a6 100644 --- a/app/articles/approach/4/page.tsx +++ b/app/articles/approach/4/page.tsx @@ -4,6 +4,69 @@ export default function ApproachPage4() {

Manifund’s Approach to Thinking About Charitable Funding

+
    +
  1. + predict +

    + what does it entail? make a good guess about which + projects will be likely to achieve their aims, what impacts those + achievements will have, and which teams can achieve the best + results for the lowest cost. +

    +

    + what are the desirables? predict accurately and at + low cost. +

    +
    2. +
  2. + front +

    + what does it entail? make a good guess about which + projects will be likely to achieve their aims, what impacts those + achievements will have, and which teams can achieve the best + results for the lowest cost. +

    +
    3. +
  3. + execute +

    + what does it entail? carry out the project. +

    +

    + what are the desirables? do something people + actually want. do it well. do it inexpensively. don’t run off + with the money. +

    +
    4. +
  4. + evaluate +

    + what does it entail? determine how valuable the + project was. +

    +

    + what are the desirables? align the measure with + the target – that is, [explain further]. account for + externalities– for example, [example of private funding][example of + big donor funding for reputation]. +

    +
    5. +
  5. + pay out +

    + what does it entail? allocate money as a function + of the evaluation of the project’s impact. in some cases, like + traditional grantmaking, the fronting is also the payout (although + to some extent this is a question of definitions; you could also + think of traditional grantmaking as just skipping the fronting + step). +

    +

    + what are the desirables? incentivize the execution + of projects with the most impact. +

    +
    6. +
) } \ No newline at end of file diff --git a/app/articles/approach/5/page.tsx b/app/articles/approach/5/page.tsx index faa7dbce..d240c21d 100644 --- a/app/articles/approach/5/page.tsx +++ b/app/articles/approach/5/page.tsx @@ -1,9 +1,60 @@ -export default function ApproachPage4() { +export default function ApproachPage6() { return (

Manifund’s Approach to Thinking About Charitable Funding

+

+
+ There are a couple of things to notice here. +

+

+ some distinctions don’t always occur +

+

+ “predict” and “evaluate” are often the same, as are “front” and “pay + out”. these are for cases where money is given out ahead of time, in a + way that’s – well, not no-strings-attached, the founders have to make a + reasonable attempt and not just grab the money and run off to hide out in + antarctica; but it’s not contingent on the project actually succeeding. + maybe the company goes bust, the idea doesn’t work, the nonprofit falls + apart, whatever. or maybe the project does what it says it will, but it + turns out no one really likes or wants the result after all. but when + there’s not a separate evaluation step and payout step, then those + contingencies are beyond the purview of the funding mechanism to + influence. trying to fund stuff that won’t fail is baked into the + prediction and fronting steps. +

+

+ this is especially true for new endeavors. with well-established + charities and companies, the ability to secure future funding is an + indirect incentive to execute current projects well. but for new + organizations that have no track record at all (or sometimes, even + new projects started by existing organizations), you have to fall + back on looking at the track record of the people founding them, and + if the founders are relatively unknown, there’s not much of an + incentive to give them a chance as opposed to taking safe bets. taking + big swings that you’ll usually miss on, but that will occasionally have + wildly outsized returns, is sometimes called “hits-based giving”; some + effective-altruist organizations explicitly try to use this strategy, + and the reason it’s even necessary for them to specify that – after + all, hits-based investing is how venture capital works be default – is + that the incentives for an individual employee of a typical + philanthropic foundation, like the incentives for an individual + bureaucrat, are to be risk-averse and lean toward safe bets rather + than innovation. +

+

+ some mechanisms are “full-stack”; others exist at one or two layers +

+

+ some mechanisms, like the examples mentioned above, offer an + implementation of the entire framework. +

+

+ but a large swath of mechanisms can be understood as innovations on + one or two of the steps. for example: +

) } \ No newline at end of file diff --git a/app/articles/approach/6/page.tsx b/app/articles/approach/6/page.tsx index d4797e37..4e619dca 100644 --- a/app/articles/approach/6/page.tsx +++ b/app/articles/approach/6/page.tsx @@ -1,9 +1,165 @@ -export default function ApproachPage6() { +export default function ApproachPage4() { return (

Manifund’s Approach to Thinking About Charitable Funding

+

+ Some Examples +

+

The startup model

+

+ Predict: investors +

+

+ Front: investors +

+

+ Execute: founder +

+

+ Evaluate: consumers +

+

+ Pay out: consumers +

+

+ Benefit: founder (as profit), investors (as profit), consumers (as consumer surplus) +

+

The bootstrapping model

+

+ Predict: founder +

+

+ Front: founder +

+

+ Execute: founder +

+

+ Evaluate: consumers +

+

+ Pay out: consumers +

+

+ Benefit: founder (as profit), consumers (as consumer surplus) +

+

The government model

+

+ Predict: elected officials or bureaucrats +

+

+ Front: citizens (as taxpayers) +

+

+ Execute: contractors +

+

+ Evaluate: citizens (indirectly, as voters) +

+

+ Pay out: citizens (as taxpayers) +

+

+ Benefit: citizens, contractors (as profit), officials (as goodwill for + subsequent elections) +

+

The wealthy philanthropist model

+

+ Predict: donor +

+

+ Front: donor +

+

+ Execute: donor +

+

+ Evaluate: donor +

+

+ Pay out: donor +

+

+ Benefit: charity beneficiaries +

+

The classical effective altruism model

+

+ Predict: donor, with an our world in data tab open +

+

+ Front: donor +

+

+ Execute: donor +

+

+ Evaluate: donor +

+

+ Pay out: donor +

+

+ Benefit: charity beneficiaries +

+

The charity evaluator model

+

+ Predict: charity evaluator +

+

+ Front: donors +

+

+ Execute: founders +

+

+ Evaluate: charity evaluator +

+

+ Pay out: donors +

+

+ Benefit: charity beneficiaries +

+

The impact certificate model

+

+ Predict: investors +

+

+ Front: investors +

+

+ Execute: founder +

+

+ Evaluate: donor +

+

+ Pay out: donor +

+

+ Benefit: founder (as profit), investors (as profit), charity beneficiaries +

+

The quadratic funding model

+

+ Predict: contributors +

+

+ Front: contributors and matching-pool sponsors +

+

+ Execute: founder +

+

+ Evaluate: contributors +

+

+ Pay out: contributors and matching-pool sponsors +

+

+ Benefit: contributors +

) } \ No newline at end of file diff --git a/app/articles/approach/7/page.tsx b/app/articles/approach/7/page.tsx new file mode 100644 index 00000000..f5369c7d --- /dev/null +++ b/app/articles/approach/7/page.tsx @@ -0,0 +1,62 @@ +export default function ApproachPage4() { + return ( +
+

+ Manifund’s Approach to Thinking About Charitable Funding +

+

+ + coupling and decoupling are central tools of funding-mechanism design + +

+

+ no mechanism listed here has all its six steps carried out by different + actors. in other words, various subsets of the steps are coupled + together: for instance, quadratic funding is able to price demand for a + public good because potential beneficiaries put up some of their own + money, which ensures they’re credibly signaling their real preference + for the good. or for a more obvious example, democracy is more + responsive to the needs of its populace than monarchy is because + elected officials have to predict which policies will be popular in + order to win re-election. that’s because coupling two steps aligns + their incentives: combining an information-gathering step with a + money-moving step ensures that decision-makers have skin in the game. +

+

+ but there’s a limit to the benefits of coupling. the ultimate fully + coupled funding system is “everyone builds everything for themselves + and pays for it themselves.” then the incentives are very aligned! + but on the other hand, it’s tremendously inefficient – no + specialization, no division of labor. +

+

+ so decoupling turns the knob the other way. for example, in a market + for impact certificates, instead of donors doing all the work of + assessing applicants and allocating funds (by making grants in + advance based on who looks most promising), you split the process + into two layers: investors, who are purely speculators, and donors, + who only have to evaluate projects’ impact afterwards. +

+

+ another example is the design-bid-build system of construction + contracting. instead of the government choosing a single bid for a + contractor to design and then carry out a construction project, + with design-bid-build they accept one contractor’s bid to create a + design, then put out a separate call for a (potentially) different + contractor to implement it; see asterisk magazine’s + interview + with urbanist alon levy, who argues that this framework + is more cost-effective than the single-bid (”design-build”) system in + most places that run it. +

+

+ in general, decoupling allows for improved competition and + specialization, which is especially useful in philanthropic cases, + where it’s hard to operate a single nonprofit as efficiently as a + corporation of the same size. but too much decoupling can cause + principal-agent problems, as often happens in bureaucracy, and can be + outweighed by economies of scale. +

+
+ ) +} \ No newline at end of file diff --git a/app/articles/approach/8/page.tsx b/app/articles/approach/8/page.tsx new file mode 100644 index 00000000..55cc1f5c --- /dev/null +++ b/app/articles/approach/8/page.tsx @@ -0,0 +1,40 @@ +export default function ApproachPage4() { + return ( +
+

+ Manifund’s Approach to Thinking About Charitable Funding +

+

+ psychological +

+

+ aside from all the financial considerations, there are a couple of social + and psychological levers entailed in grantmaking. one lesson manifund has + gleaned from manifold, our sister prediction-market company, is that you + can operate a finance-like system without any actual money involved, + because people also value: +

+
    +
  1. + reputation – the glory of visibly performing well and accumulating + scarce prestige, and +
    2. +
  2. + agency – the feeling that their actions cause something to come into + being that wouldn’t have otherwise. this is the sensation of knowing + that when you push the button, something will actually happen. +
    3. +
+

+ the slower-moving and less accessible a philanthropic organization is, + the harder it is for many individuals to achieve either of those two. on + a crowdfunding platform like kickstarter, users can get a sense of + agency, but not so much of accumulating reputation. anything that + resembles a game, like investing or predicting, and that has real-world + effects, hits both these criteria: you can succeed relative to others, + and you can succeed in the sense that you did some actions on a website + and now a real physical project is getting accomplished. +

+
+ ) +} \ No newline at end of file diff --git a/app/articles/approach/9/page.tsx b/app/articles/approach/9/page.tsx new file mode 100644 index 00000000..d92e4d34 --- /dev/null +++ b/app/articles/approach/9/page.tsx @@ -0,0 +1,18 @@ +export default function ApproachPage4() { + return ( +
+

+ Manifund’s Approach to Thinking About Charitable Funding +

+

+ effective iteration +

+

+ The examples of the internet, cultural evolution, and the stock market – not to mention prediction markets! – suggest that the most accurate information is produced at scale by a decentralized mass of actors with an incentive for the right answers to bubble to the top. First-principles speculation only goes so far – it just isn’t possible for one person or team to process as much information as a network can. +

+

+ The way to use this principle to figure out which projects will work is to find ways to try lots of projects quickly: the higher the speed and greater the volume, the more effective the information-gathering. In 2023, Manifund’s first year of operations, the platform sent $2.06 million in funds to 88 projects. One of our explicit goals is to offer grantees turnaround times measured in weeks or even days instead of months. We hypothesize that timing is an underrated issue for projects that are smaller, newer, and/or run by individuals, as opposed to projects that are already ongoing or are part of existing organizations: it’s difficult to arrange your life so that you’ll be free in several months for an opportunity that might or might not materialize. In these cases, agility and speed fill a critical gap for applicants who would otherwise have been prohibitively inconvenienced by long turnaround times. +

+
+ ) +} \ No newline at end of file diff --git a/app/sidebar.tsx b/app/sidebar.tsx index 969ac486..5328c6ea 100644 --- a/app/sidebar.tsx +++ b/app/sidebar.tsx @@ -83,8 +83,6 @@ export default async function Sidebar() { {name: '8', href: '/articles/approach/8'}, {name: '9', href: '/articles/approach/9'}, {name: '10', href: '/articles/approach/10'}, - {name: '11', href: '/articles/approach/11'}, - {name: '12', href: '/articles/approach/12'}, ] }, ]}} /> From 3a8d3be55d6d4c576e3b59ae6099964392d343ea Mon Sep 17 00:00:00 2001 From: Lily Jordan Date: Fri, 29 Mar 2024 18:32:04 -0700 Subject: [PATCH 10/16] titles, formatting, uppercasing, etc --- app/articles/approach/1/page.tsx | 6 +- app/articles/approach/10/page.tsx | 26 ++-- app/articles/approach/2/page.tsx | 28 ++++ app/articles/approach/3/page.tsx | 40 ------ app/articles/approach/4/page.tsx | 130 +++++++++-------- app/articles/approach/5/page.tsx | 205 +++++++++++++++++++------- app/articles/approach/6/page.tsx | 232 ++++++++++-------------------- app/articles/approach/7/page.tsx | 26 ++-- app/articles/approach/8/page.tsx | 18 +-- app/articles/approach/9/page.tsx | 2 +- app/sidebar.tsx | 7 +- 11 files changed, 369 insertions(+), 351 deletions(-) delete mode 100644 app/articles/approach/3/page.tsx diff --git a/app/articles/approach/1/page.tsx b/app/articles/approach/1/page.tsx index 1af18110..a70c69d0 100644 --- a/app/articles/approach/1/page.tsx +++ b/app/articles/approach/1/page.tsx @@ -5,7 +5,7 @@ export default function ApproachPage1() { Manifund’s Approach to Thinking About Charitable Funding

- From 2023 retrospective: "Our initial thesis was that grant applications and screening largely can be done in public, and should be." + From 2023 retrospective: “Our initial thesis was that grant applications and screening largely can be done in public, and should be.”

Economic Experimentation @@ -29,8 +29,8 @@ export default function ApproachPage1() {

Manifund is different, because lots of real money is moving through our - programs. But we’re still small and fast-moving ([example of grant - turnaround time]), and our ethos is similar: we aim to be an incubator + programs. But we’re still small and fast-moving, + and our ethos is similar: we aim to be an incubator for lots of small- and medium-scale experiments testing out funding mechanisms for charity and public goods.

diff --git a/app/articles/approach/10/page.tsx b/app/articles/approach/10/page.tsx index 45789555..17defc20 100644 --- a/app/articles/approach/10/page.tsx +++ b/app/articles/approach/10/page.tsx @@ -5,31 +5,31 @@ export default function ApproachPage4() { Manifund’s Approach to Thinking About Charitable Funding

- caveat: different value systems + Caveat: different value systems

- one limitation of this model is that it’s not obvious how well it applies + One limitation of this model is that it’s not obvious how well it applies to entities that don’t naturally fit into standard economic models because they can’t easily participate in markets: children, animals, people who - might exist in the distant future. for people whose value systems heavily + might exist in the distant future. For people whose value systems heavily incorporate any of these groups’ preferences, all market-based funding - solutions might prove inadequate. in these cases, funding is closer to a + solutions might prove inadequate. In these cases, funding is closer to a strict question of altruism than a coordination problem; donors have to speculate on behalf of those who can’t (financially) speculate for - themselves. these causes lend themselves better to direct improvements + themselves. These causes lend themselves better to direct improvements in prediction and evaluation than to financial mechanisms that couple those steps with fronting and paying out; it’s not really clear how you’d align your own incentives with the incentives of someone in the year - 2300, except by saying “i care abstractly about making future people + 2300, except by saying “I care abstractly about making future people happy.”

- even in these areas, though, there are often some smaller subproblems + Even in these areas, though, there are often some smaller subproblems related to coordination, and so market mechanisms can offer at least - some improvements. for example, grant proposals often become the + some improvements. For example, grant proposals often become the victims of “funder chicken,” where each potential donor, having a limited budget, likes a proposal but hopes one of the others will fund - it first. and even if you can’t evaluate one year from now whether a + it first. And even if you can’t evaluate one year from now whether a project intended to improve outcomes on a century-long timescale has succeeded in that, you can at least check whether the project did what it said it would, or whether it achieved the kind of result that you @@ -38,19 +38,19 @@ export default function ApproachPage4() { instrumental goals, market structures are pretty good.

- on some level, you can think of altruism itself as a public good: + On some level, you can think of altruism itself as a public good: like, ten thousand people are all sad about cows getting factory-farmed and would rather the cows be freed to roam on beautiful verdant pastures, and so cow liberation is a public good to all those people - insofar as it’s something that would make them all happy. this doesn’t + insofar as it’s something that would make them all happy. This doesn’t necessarily evaluate how much the cow would prefer freedom to its current life; you could just as easily imagine that ten thousand other people would enjoy seeing a herd of cows dressed up in identical little bow ties, and so purchasing a pack of wide-necked bow ties would be a - public-goods-funding coordination problem in just the same way. the + public-goods-funding coordination problem in just the same way. The “figuring out what would be best for cows” problem, and below it the more fundamental “figuring out what my altruistic values are” problem, - aren’t easily solvable by markets. but in either case, once you make + aren’t easily solvable by markets. But in either case, once you make up your mind as to what your goals are, market mechanisms become useful tools to help you get there.

diff --git a/app/articles/approach/2/page.tsx b/app/articles/approach/2/page.tsx index 5e229b79..4da10724 100644 --- a/app/articles/approach/2/page.tsx +++ b/app/articles/approach/2/page.tsx @@ -13,6 +13,34 @@ export default function ApproachPage2() {

Impact certificates

Assurance contract auctions

+

+ Is there anything general we can say about these kinds of experiments? +

+

+ Take the system of impact certificates, for example. There are a couple + of differences from how charity is traditionally funded. The donations + come after the fact, for one thing, rather than in advance. And + there’s a middleman involved, who doesn’t even necessarily have + to have altruistic motives. +

+

+ Theoretically, this leads to neat results. There’s no need to learn + new ways to evaluate charities, or to make people more altruistic. The + addition of an intermediate market makes it more efficient to figure + out the most efficient allocation of funds by the donor’s own + metrics. +

+

+ But also, it seems a little contrived? Each individual step of this + process is perfectly understandable if you understand the analogous + element of private markets. But altogether, it’s a lot of new moving + parts to conceptualize at once, and it’s not obvious how to come up + with similar ideas or how to evaluate them. +

+

+ So we’ve come up with a first pass at a general framework for + thinking about the design of novel funding mechanisms. +

) } \ No newline at end of file diff --git a/app/articles/approach/3/page.tsx b/app/articles/approach/3/page.tsx deleted file mode 100644 index 2d61632a..00000000 --- a/app/articles/approach/3/page.tsx +++ /dev/null @@ -1,40 +0,0 @@ -export default function ApproachPage3() { - return ( -
-

- Manifund’s Approach to Thinking About Charitable Funding -

-

- How to think about funding -

-

- Is there anything general we can say about these kinds of experiments? -

-

- Take the system of impact certificates, for example. There are a couple - of differences from how charity is traditionally funded. The donations - come after the fact, for one thing, rather than in advance. And - there’s a middleman involved, who doesn’t even necessarily have - to have altruistic motives. -

-

- Theoretically, this leads to neat results. There’s no need to learn - new ways to evaluate charities, or to make people more altruistic. The - addition of an intermediate market makes it more efficient to figure - out the most efficient allocation of funds by the donor’s own - metrics. -

-

- But also, it seems a little contrived? Each individual step of this - process is perfectly understandable if you understand the analogous - element of private markets. But altogether, it’s a lot of new moving - parts to conceptualize at once, and it’s not obvious how to come up - with similar ideas or how to evaluate them. -

-

- So we’ve come up with a first pass at a general framework for - thinking about the design of novel funding mechanisms. -

-
- ) -} \ No newline at end of file diff --git a/app/articles/approach/4/page.tsx b/app/articles/approach/4/page.tsx index 0a24a9a6..590b6aac 100644 --- a/app/articles/approach/4/page.tsx +++ b/app/articles/approach/4/page.tsx @@ -4,69 +4,73 @@ export default function ApproachPage4() {

Manifund’s Approach to Thinking About Charitable Funding

-
    -
  1. - predict -

    - what does it entail? make a good guess about which - projects will be likely to achieve their aims, what impacts those - achievements will have, and which teams can achieve the best - results for the lowest cost. -

    -

    - what are the desirables? predict accurately and at - low cost. -

    -
    2. -
  2. - front -

    - what does it entail? make a good guess about which - projects will be likely to achieve their aims, what impacts those - achievements will have, and which teams can achieve the best - results for the lowest cost. -

    -
    3. -
  3. - execute -

    - what does it entail? carry out the project. -

    -

    - what are the desirables? do something people - actually want. do it well. do it inexpensively. don’t run off - with the money. -

    -
    4. -
  4. - evaluate -

    - what does it entail? determine how valuable the - project was. -

    -

    - what are the desirables? align the measure with - the target – that is, [explain further]. account for - externalities– for example, [example of private funding][example of - big donor funding for reputation]. -

    -
    5. -
  5. - pay out -

    - what does it entail? allocate money as a function - of the evaluation of the project’s impact. in some cases, like - traditional grantmaking, the fronting is also the payout (although - to some extent this is a question of definitions; you could also - think of traditional grantmaking as just skipping the fronting - step). -

    -

    - what are the desirables? incentivize the execution - of projects with the most impact. -

    -
    6. -
+

+ The Six Steps of Funding +

+

1. Predict

+

+ What does it entail? Make a good guess about which + projects will be likely to achieve their aims, what impacts those + achievements will have, and which teams can achieve the best + results for the lowest cost. +

+

+ What are the desirables? Predict accurately and at + low cost. +

+

2. Front

+

+ What does it entail? Supply the immediate + funding for the costs a team will incur in the process of + doing their project. +

+

+ What are the desirables? Predict accurately and at + low cost. +

+

3. Execute

+

+ What does it entail? Carry out the project. +

+

+ What are the desirables? Do something people + actually want. Do it well. Do it inexpensively. Don’t run off + with the money. +

+

4. Evaluate

+

+ What does it entail? Determine how valuable the + project was. +

+

+ What are the desirables? Align the measure with + the target – that is, [explain further]. Account for + externalities – for example, if the project was moderately beneficial + to the group that commissioned it, but also benefited lots of + other people as well, then we’d like that to be factored in. +

+

5. Pay out

+

+ What does it entail? Allocate money as a function + of the evaluation of the project’s impact. In some cases, like + traditional grantmaking, the fronting is also the payout (although + to some extent this is a question of definitions; you could also + think of traditional grantmaking as just skipping the fronting + step). +

+

+ What are the desirables? Incentivize the execution + of projects with the most impact. +

+

6. Benefit

+

+ What does it entail? Enjoy the benefits + of whatever it was that got funded. +

+

+ What are the desirables? Be helped by the project. + Do not be harmed by the project. +

) } \ No newline at end of file diff --git a/app/articles/approach/5/page.tsx b/app/articles/approach/5/page.tsx index d240c21d..28a86d8d 100644 --- a/app/articles/approach/5/page.tsx +++ b/app/articles/approach/5/page.tsx @@ -1,59 +1,164 @@ -export default function ApproachPage6() { +export default function ApproachPage4() { return (

Manifund’s Approach to Thinking About Charitable Funding

+

+ Some Examples +

+

The startup model

-
- There are a couple of things to notice here. -

-

- some distinctions don’t always occur -

-

- “predict” and “evaluate” are often the same, as are “front” and “pay - out”. these are for cases where money is given out ahead of time, in a - way that’s – well, not no-strings-attached, the founders have to make a - reasonable attempt and not just grab the money and run off to hide out in - antarctica; but it’s not contingent on the project actually succeeding. - maybe the company goes bust, the idea doesn’t work, the nonprofit falls - apart, whatever. or maybe the project does what it says it will, but it - turns out no one really likes or wants the result after all. but when - there’s not a separate evaluation step and payout step, then those - contingencies are beyond the purview of the funding mechanism to - influence. trying to fund stuff that won’t fail is baked into the - prediction and fronting steps. -

-

- this is especially true for new endeavors. with well-established - charities and companies, the ability to secure future funding is an - indirect incentive to execute current projects well. but for new - organizations that have no track record at all (or sometimes, even - new projects started by existing organizations), you have to fall - back on looking at the track record of the people founding them, and - if the founders are relatively unknown, there’s not much of an - incentive to give them a chance as opposed to taking safe bets. taking - big swings that you’ll usually miss on, but that will occasionally have - wildly outsized returns, is sometimes called “hits-based giving”; some - effective-altruist organizations explicitly try to use this strategy, - and the reason it’s even necessary for them to specify that – after - all, hits-based investing is how venture capital works be default – is - that the incentives for an individual employee of a typical - philanthropic foundation, like the incentives for an individual - bureaucrat, are to be risk-averse and lean toward safe bets rather - than innovation. -

-

- some mechanisms are “full-stack”; others exist at one or two layers -

-

- some mechanisms, like the examples mentioned above, offer an - implementation of the entire framework. -

-

- but a large swath of mechanisms can be understood as innovations on - one or two of the steps. for example: + Predict: investors +

+

+ Front: investors +

+

+ Execute: founder +

+

+ Evaluate: consumers +

+

+ Pay out: consumers +

+

+ Benefit: founder (as profit), investors (as profit), consumers (as consumer surplus) +

+

The bootstrapping model

+

+ Predict: founder +

+

+ Front: founder +

+

+ Execute: founder +

+

+ Evaluate: consumers +

+

+ Pay out: consumers +

+

+ Benefit: founder (as profit), consumers (as consumer surplus) +

+

The government model

+

+ Predict: elected officials or bureaucrats +

+

+ Front: citizens (as taxpayers) +

+

+ Execute: contractors +

+

+ Evaluate: citizens (indirectly, as voters) +

+

+ Pay out: citizens (as taxpayers) +

+

+ Benefit: citizens, contractors (as profit), officials (as goodwill for + subsequent elections) +

+

The wealthy philanthropist model

+

+ Predict: donor +

+

+ Front: donor +

+

+ Execute: donor +

+

+ Evaluate: donor +

+

+ Pay out: donor +

+

+ Benefit: charity beneficiaries +

+

The classical effective altruism model

+

+ Predict: donor, with an Our World In Data tab open +

+

+ Front: donor +

+

+ Execute: donor +

+

+ Evaluate: donor +

+

+ Pay out: donor +

+

+ Benefit: charity beneficiaries +

+

The charity evaluator model

+

+ Predict: charity evaluator +

+

+ Front: donors +

+

+ Execute: founders +

+

+ Evaluate: charity evaluator +

+

+ Pay out: donors +

+

+ Benefit: charity beneficiaries +

+

The impact certificate model

+

+ Predict: investors +

+

+ Front: investors +

+

+ Execute: founder +

+

+ Evaluate: donor +

+

+ Pay out: donor +

+

+ Benefit: founder (as profit), investors (as profit), charity beneficiaries +

+

The quadratic funding model

+

+ Predict: contributors +

+

+ Front: contributors and matching-pool sponsors +

+

+ Execute: founder +

+

+ Evaluate: contributors +

+

+ Pay out: contributors and matching-pool sponsors +

+

+ Benefit: contributors

) diff --git a/app/articles/approach/6/page.tsx b/app/articles/approach/6/page.tsx index 4e619dca..fd77205d 100644 --- a/app/articles/approach/6/page.tsx +++ b/app/articles/approach/6/page.tsx @@ -1,164 +1,86 @@ -export default function ApproachPage4() { +export default function ApproachPage6() { return (

Manifund’s Approach to Thinking About Charitable Funding

-

- Some Examples -

-

The startup model

- Predict: investors -

-

- Front: investors -

-

- Execute: founder -

-

- Evaluate: consumers -

-

- Pay out: consumers -

-

- Benefit: founder (as profit), investors (as profit), consumers (as consumer surplus) -

-

The bootstrapping model

-

- Predict: founder -

-

- Front: founder -

-

- Execute: founder -

-

- Evaluate: consumers -

-

- Pay out: consumers -

-

- Benefit: founder (as profit), consumers (as consumer surplus) -

-

The government model

-

- Predict: elected officials or bureaucrats -

-

- Front: citizens (as taxpayers) -

-

- Execute: contractors -

-

- Evaluate: citizens (indirectly, as voters) -

-

- Pay out: citizens (as taxpayers) -

-

- Benefit: citizens, contractors (as profit), officials (as goodwill for - subsequent elections) -

-

The wealthy philanthropist model

-

- Predict: donor -

-

- Front: donor -

-

- Execute: donor -

-

- Evaluate: donor -

-

- Pay out: donor -

-

- Benefit: charity beneficiaries -

-

The classical effective altruism model

-

- Predict: donor, with an our world in data tab open -

-

- Front: donor -

-

- Execute: donor -

-

- Evaluate: donor -

-

- Pay out: donor -

-

- Benefit: charity beneficiaries -

-

The charity evaluator model

-

- Predict: charity evaluator -

-

- Front: donors -

-

- Execute: founders -

-

- Evaluate: charity evaluator -

-

- Pay out: donors -

-

- Benefit: charity beneficiaries -

-

The impact certificate model

-

- Predict: investors -

-

- Front: investors -

-

- Execute: founder -

-

- Evaluate: donor -

-

- Pay out: donor -

-

- Benefit: founder (as profit), investors (as profit), charity beneficiaries -

-

The quadratic funding model

-

- Predict: contributors -

-

- Front: contributors and matching-pool sponsors -

-

- Execute: founder -

-

- Evaluate: contributors -

-

- Pay out: contributors and matching-pool sponsors -

-

- Benefit: contributors +
+ There are a couple of things to notice here. +

+

+ Some distinctions don’t always occur +

+

+ “Predict” and “evaluate” are often the same, as are “front” and “pay + out”. These are for cases where money is given out ahead of time, in a + way that’s – well, not no-strings-attached, the founders have to make a + reasonable attempt and not just grab the money and run off to hide out in + antarctica; but it’s not contingent on the project actually succeeding. + maybe the company goes bust, the idea doesn’t work, the nonprofit falls + apart, whatever. Or maybe the project does what it says it will, but it + turns out no one really likes or wants the result after all. But when + there’s not a separate evaluation step and payout step, then those + contingencies are beyond the purview of the funding mechanism to + influence. Trying to fund stuff that won’t fail is baked into the + prediction and fronting steps. +

+

+ This is especially true for new endeavors. With well-established + charities and companies, the ability to secure future funding is an + indirect incentive to execute current projects well. But for new + organizations that have no track record at all (or sometimes, even + new projects started by existing organizations), you have to fall + back on looking at the track record of the people founding them, and + if the founders are relatively unknown, there’s not much of an + incentive to give them a chance as opposed to taking safe bets. Taking + big swings that you’ll usually miss on, but that will occasionally have + wildly outsized returns, is sometimes called “hits-based giving”; some + effective-altruist organizations explicitly try to use this strategy, + and the reason it’s even necessary for them to specify that – after + all, hits-based investing is how venture capital works be default – is + that the incentives for an individual employee of a typical + philanthropic foundation, like the incentives for an individual + bureaucrat, are to be risk-averse and lean toward safe bets rather + than innovation. +

+

+ Some mechanisms are “full-stack”; others exist at one or two layers +

+

+ Some mechanisms, like the examples mentioned above, offer an + implementation of the entire framework. +

+

+ But a large swath of mechanisms can be understood as innovations on + one or two of the steps. For example: +

+

+ Challenge prizes: pay out. +

+

+ S-process: evaluate. +

+

+ Advance market commitments pay out. +

+

+ Income share agreements predict and front. +

+

+ Schelling-point oracles: evaluate. +

+

+ Patronage: predict and front. +

+

+ Crowdfunding: pay out. +

+

+ Regranting: predict. +

+

+ Prediction markets: predict.

) diff --git a/app/articles/approach/7/page.tsx b/app/articles/approach/7/page.tsx index f5369c7d..fe2a8803 100644 --- a/app/articles/approach/7/page.tsx +++ b/app/articles/approach/7/page.tsx @@ -6,31 +6,31 @@ export default function ApproachPage4() {

- coupling and decoupling are central tools of funding-mechanism design + Coupling and decoupling are central tools of funding-mechanism design

- no mechanism listed here has all its six steps carried out by different - actors. in other words, various subsets of the steps are coupled + No mechanism listed here has all its six steps carried out by different + actors. In other words, various subsets of the steps are coupled together: for instance, quadratic funding is able to price demand for a public good because potential beneficiaries put up some of their own money, which ensures they’re credibly signaling their real preference - for the good. or for a more obvious example, democracy is more + for the good. Or for a more obvious example, democracy is more responsive to the needs of its populace than monarchy is because elected officials have to predict which policies will be popular in - order to win re-election. that’s because coupling two steps aligns + order to win re-election. That’s because coupling two steps aligns their incentives: combining an information-gathering step with a money-moving step ensures that decision-makers have skin in the game.

- but there’s a limit to the benefits of coupling. the ultimate fully + But there’s a limit to the benefits of coupling. The ultimate fully coupled funding system is “everyone builds everything for themselves and pays for it themselves.” then the incentives are very aligned! but on the other hand, it’s tremendously inefficient – no specialization, no division of labor.

- so decoupling turns the knob the other way. for example, in a market + So decoupling turns the knob the other way. For example, in a market for impact certificates, instead of donors doing all the work of assessing applicants and allocating funds (by making grants in advance based on who looks most promising), you split the process @@ -38,22 +38,22 @@ export default function ApproachPage4() { who only have to evaluate projects’ impact afterwards.

- another example is the design-bid-build system of construction - contracting. instead of the government choosing a single bid for a + Another example is the design-bid-build system of construction + contracting. Instead of the government choosing a single bid for a contractor to design and then carry out a construction project, with design-bid-build they accept one contractor’s bid to create a design, then put out a separate call for a (potentially) different - contractor to implement it; see asterisk magazine’s + contractor to implement it; see Asterisk magazine’s interview - with urbanist alon levy, who argues that this framework + with urbanist Alon Levy, who argues that this framework is more cost-effective than the single-bid (”design-build”) system in most places that run it.

- in general, decoupling allows for improved competition and + In general, decoupling allows for improved competition and specialization, which is especially useful in philanthropic cases, where it’s hard to operate a single nonprofit as efficiently as a - corporation of the same size. but too much decoupling can cause + corporation of the same size. But too much decoupling can cause principal-agent problems, as often happens in bureaucracy, and can be outweighed by economies of scale.

diff --git a/app/articles/approach/8/page.tsx b/app/articles/approach/8/page.tsx index 55cc1f5c..3127ed2b 100644 --- a/app/articles/approach/8/page.tsx +++ b/app/articles/approach/8/page.tsx @@ -5,12 +5,12 @@ export default function ApproachPage4() { Manifund’s Approach to Thinking About Charitable Funding

- psychological + Psychological

- aside from all the financial considerations, there are a couple of social - and psychological levers entailed in grantmaking. one lesson manifund has - gleaned from manifold, our sister prediction-market company, is that you + Aside from all the financial considerations, there are a couple of social + and psychological levers entailed in grantmaking. One lesson Manifund has + gleaned from Manifold, our sister prediction-market company, is that you can operate a finance-like system without any actual money involved, because people also value:

@@ -21,15 +21,15 @@ export default function ApproachPage4() {
  • agency – the feeling that their actions cause something to come into - being that wouldn’t have otherwise. this is the sensation of knowing + being that wouldn’t have otherwise. This is the sensation of knowing that when you push the button, something will actually happen.

    • - the slower-moving and less accessible a philanthropic organization is, - the harder it is for many individuals to achieve either of those two. on - a crowdfunding platform like kickstarter, users can get a sense of - agency, but not so much of accumulating reputation. anything that + The slower-moving and less accessible a philanthropic organization is, + the harder it is for many individuals to achieve either of those two. On + a crowdfunding platform like Kickstarter, users can get a sense of + agency, but not so much of accumulating reputation. Anything that resembles a game, like investing or predicting, and that has real-world effects, hits both these criteria: you can succeed relative to others, and you can succeed in the sense that you did some actions on a website diff --git a/app/articles/approach/9/page.tsx b/app/articles/approach/9/page.tsx index d92e4d34..fa61daf1 100644 --- a/app/articles/approach/9/page.tsx +++ b/app/articles/approach/9/page.tsx @@ -5,7 +5,7 @@ export default function ApproachPage4() { Manifund’s Approach to Thinking About Charitable Funding

    - effective iteration + Effective Iteration

    The examples of the internet, cultural evolution, and the stock market – not to mention prediction markets! – suggest that the most accurate information is produced at scale by a decentralized mass of actors with an incentive for the right answers to bubble to the top. First-principles speculation only goes so far – it just isn’t possible for one person or team to process as much information as a network can. diff --git a/app/sidebar.tsx b/app/sidebar.tsx index 5328c6ea..2a6a6003 100644 --- a/app/sidebar.tsx +++ b/app/sidebar.tsx @@ -73,10 +73,9 @@ export default async function Sidebar() { name: 'Manifund\'s approach to thinking about charitable funding', childrenDefaultOpen: true, children: [ - {name: '1', href: '/articles/approach/1'}, - {name: '2', href: '/articles/approach/2'}, - {name: '3', href: '/articles/approach/3'}, - {name: '4', href: '/articles/approach/4'}, + {name: '1. Economic experimentation', href: '/articles/approach/1'}, + {name: '2. Programs we\'ve run', href: '/articles/approach/2'}, + {name: '3. The six steps of funding', href: '/articles/approach/4'}, {name: '5', href: '/articles/approach/5'}, {name: '6', href: '/articles/approach/6'}, {name: '7', href: '/articles/approach/7'}, From f61528565cb7a446318c7b5fc3ff26cfae855845 Mon Sep 17 00:00:00 2001 From: Lily Jordan Date: Fri, 29 Mar 2024 18:44:05 -0700 Subject: [PATCH 11/16] all sidebar titles --- app/articles/approach/10/page.tsx | 59 ------- app/articles/approach/3/page.tsx | 76 ++++++++ app/articles/approach/4/page.tsx | 171 +++++++++++++----- app/articles/approach/5/page.tsx | 285 ++++++++++++++---------------- app/articles/approach/6/page.tsx | 103 +++-------- app/articles/approach/7/page.tsx | 54 +----- app/articles/approach/8/page.tsx | 73 +++++--- app/articles/approach/9/page.tsx | 18 -- 8 files changed, 416 insertions(+), 423 deletions(-) delete mode 100644 app/articles/approach/10/page.tsx create mode 100644 app/articles/approach/3/page.tsx delete mode 100644 app/articles/approach/9/page.tsx diff --git a/app/articles/approach/10/page.tsx b/app/articles/approach/10/page.tsx deleted file mode 100644 index 17defc20..00000000 --- a/app/articles/approach/10/page.tsx +++ /dev/null @@ -1,59 +0,0 @@ -export default function ApproachPage4() { - return ( -

    -

    - Manifund’s Approach to Thinking About Charitable Funding -

    -

    - Caveat: different value systems -

    -

    - One limitation of this model is that it’s not obvious how well it applies - to entities that don’t naturally fit into standard economic models because - they can’t easily participate in markets: children, animals, people who - might exist in the distant future. For people whose value systems heavily - incorporate any of these groups’ preferences, all market-based funding - solutions might prove inadequate. In these cases, funding is closer to a - strict question of altruism than a coordination problem; donors have to - speculate on behalf of those who can’t (financially) speculate for - themselves. These causes lend themselves better to direct improvements - in prediction and evaluation than to financial mechanisms that couple - those steps with fronting and paying out; it’s not really clear how you’d - align your own incentives with the incentives of someone in the year - 2300, except by saying “I care abstractly about making future people - happy.” -

    -

    - Even in these areas, though, there are often some smaller subproblems - related to coordination, and so market mechanisms can offer at least - some improvements. For example, grant proposals often become the - victims of “funder chicken,” where each potential donor, having a - limited budget, likes a proposal but hopes one of the others will fund - it first. And even if you can’t evaluate one year from now whether a - project intended to improve outcomes on a century-long timescale has - succeeded in that, you can at least check whether the project did what - it said it would, or whether it achieved the kind of result that you - already believe for other reasons would be valuable for the far future. - it’s hard to incentivize terminal goals this way, but for incentivizing - instrumental goals, market structures are pretty good. -

    -

    - On some level, you can think of altruism itself as a public good: - like, ten thousand people are all sad about cows getting factory-farmed - and would rather the cows be freed to roam on beautiful verdant - pastures, and so cow liberation is a public good to all those people - insofar as it’s something that would make them all happy. This doesn’t - necessarily evaluate how much the cow would prefer freedom to its - current life; you could just as easily imagine that ten thousand other - people would enjoy seeing a herd of cows dressed up in identical little - bow ties, and so purchasing a pack of wide-necked bow ties would be a - public-goods-funding coordination problem in just the same way. The - “figuring out what would be best for cows” problem, and below it the - more fundamental “figuring out what my altruistic values are” problem, - aren’t easily solvable by markets. But in either case, once you make - up your mind as to what your goals are, market mechanisms become useful - tools to help you get there. -

    -
    - ) -} \ No newline at end of file diff --git a/app/articles/approach/3/page.tsx b/app/articles/approach/3/page.tsx new file mode 100644 index 00000000..590b6aac --- /dev/null +++ b/app/articles/approach/3/page.tsx @@ -0,0 +1,76 @@ +export default function ApproachPage4() { + return ( +
    +

    + Manifund’s Approach to Thinking About Charitable Funding +

    +

    + The Six Steps of Funding +

    +

    1. Predict

    +

    + What does it entail? Make a good guess about which + projects will be likely to achieve their aims, what impacts those + achievements will have, and which teams can achieve the best + results for the lowest cost. +

    +

    + What are the desirables? Predict accurately and at + low cost. +

    +

    2. Front

    +

    + What does it entail? Supply the immediate + funding for the costs a team will incur in the process of + doing their project. +

    +

    + What are the desirables? Predict accurately and at + low cost. +

    +

    3. Execute

    +

    + What does it entail? Carry out the project. +

    +

    + What are the desirables? Do something people + actually want. Do it well. Do it inexpensively. Don’t run off + with the money. +

    +

    4. Evaluate

    +

    + What does it entail? Determine how valuable the + project was. +

    +

    + What are the desirables? Align the measure with + the target – that is, [explain further]. Account for + externalities – for example, if the project was moderately beneficial + to the group that commissioned it, but also benefited lots of + other people as well, then we’d like that to be factored in. +

    +

    5. Pay out

    +

    + What does it entail? Allocate money as a function + of the evaluation of the project’s impact. In some cases, like + traditional grantmaking, the fronting is also the payout (although + to some extent this is a question of definitions; you could also + think of traditional grantmaking as just skipping the fronting + step). +

    +

    + What are the desirables? Incentivize the execution + of projects with the most impact. +

    +

    6. Benefit

    +

    + What does it entail? Enjoy the benefits + of whatever it was that got funded. +

    +

    + What are the desirables? Be helped by the project. + Do not be harmed by the project. +

    +
    + ) +} \ No newline at end of file diff --git a/app/articles/approach/4/page.tsx b/app/articles/approach/4/page.tsx index 590b6aac..28a86d8d 100644 --- a/app/articles/approach/4/page.tsx +++ b/app/articles/approach/4/page.tsx @@ -5,71 +5,160 @@ export default function ApproachPage4() { Manifund’s Approach to Thinking About Charitable Funding

    - The Six Steps of Funding + Some Examples

    -

    1. Predict

    +

    The startup model

    - What does it entail? Make a good guess about which - projects will be likely to achieve their aims, what impacts those - achievements will have, and which teams can achieve the best - results for the lowest cost. + Predict: investors

    - What are the desirables? Predict accurately and at - low cost. + Front: investors

    -

    2. Front

    - What does it entail? Supply the immediate - funding for the costs a team will incur in the process of - doing their project. + Execute: founder

    - What are the desirables? Predict accurately and at - low cost. + Evaluate: consumers

    -

    3. Execute

    - What does it entail? Carry out the project. + Pay out: consumers

    - What are the desirables? Do something people - actually want. Do it well. Do it inexpensively. Don’t run off - with the money. + Benefit: founder (as profit), investors (as profit), consumers (as consumer surplus)

    -

    4. Evaluate

    +

    The bootstrapping model

    - What does it entail? Determine how valuable the - project was. + Predict: founder

    - What are the desirables? Align the measure with - the target – that is, [explain further]. Account for - externalities – for example, if the project was moderately beneficial - to the group that commissioned it, but also benefited lots of - other people as well, then we’d like that to be factored in. + Front: founder

    -

    5. Pay out

    - What does it entail? Allocate money as a function - of the evaluation of the project’s impact. In some cases, like - traditional grantmaking, the fronting is also the payout (although - to some extent this is a question of definitions; you could also - think of traditional grantmaking as just skipping the fronting - step). + Execute: founder

    - What are the desirables? Incentivize the execution - of projects with the most impact. + Evaluate: consumers

    -

    6. Benefit

    - What does it entail? Enjoy the benefits - of whatever it was that got funded. + Pay out: consumers

    - What are the desirables? Be helped by the project. - Do not be harmed by the project. + Benefit: founder (as profit), consumers (as consumer surplus) +

    +

    The government model

    +

    + Predict: elected officials or bureaucrats +

    +

    + Front: citizens (as taxpayers) +

    +

    + Execute: contractors +

    +

    + Evaluate: citizens (indirectly, as voters) +

    +

    + Pay out: citizens (as taxpayers) +

    +

    + Benefit: citizens, contractors (as profit), officials (as goodwill for + subsequent elections) +

    +

    The wealthy philanthropist model

    +

    + Predict: donor +

    +

    + Front: donor +

    +

    + Execute: donor +

    +

    + Evaluate: donor +

    +

    + Pay out: donor +

    +

    + Benefit: charity beneficiaries +

    +

    The classical effective altruism model

    +

    + Predict: donor, with an Our World In Data tab open +

    +

    + Front: donor +

    +

    + Execute: donor +

    +

    + Evaluate: donor +

    +

    + Pay out: donor +

    +

    + Benefit: charity beneficiaries +

    +

    The charity evaluator model

    +

    + Predict: charity evaluator +

    +

    + Front: donors +

    +

    + Execute: founders +

    +

    + Evaluate: charity evaluator +

    +

    + Pay out: donors +

    +

    + Benefit: charity beneficiaries +

    +

    The impact certificate model

    +

    + Predict: investors +

    +

    + Front: investors +

    +

    + Execute: founder +

    +

    + Evaluate: donor +

    +

    + Pay out: donor +

    +

    + Benefit: founder (as profit), investors (as profit), charity beneficiaries +

    +

    The quadratic funding model

    +

    + Predict: contributors +

    +

    + Front: contributors and matching-pool sponsors +

    +

    + Execute: founder +

    +

    + Evaluate: contributors +

    +

    + Pay out: contributors and matching-pool sponsors +

    +

    + Benefit: contributors

    ) diff --git a/app/articles/approach/5/page.tsx b/app/articles/approach/5/page.tsx index 28a86d8d..904e97e5 100644 --- a/app/articles/approach/5/page.tsx +++ b/app/articles/approach/5/page.tsx @@ -1,164 +1,141 @@ -export default function ApproachPage4() { +export default function ApproachPage6() { return (

    Manifund’s Approach to Thinking About Charitable Funding

    -

    - Some Examples +

    + Observations About These Examples

    -

    The startup model

    - Predict: investors -

    -

    - Front: investors -

    -

    - Execute: founder -

    -

    - Evaluate: consumers -

    -

    - Pay out: consumers -

    -

    - Benefit: founder (as profit), investors (as profit), consumers (as consumer surplus) -

    -

    The bootstrapping model

    -

    - Predict: founder -

    -

    - Front: founder -

    -

    - Execute: founder -

    -

    - Evaluate: consumers -

    -

    - Pay out: consumers -

    -

    - Benefit: founder (as profit), consumers (as consumer surplus) -

    -

    The government model

    -

    - Predict: elected officials or bureaucrats -

    -

    - Front: citizens (as taxpayers) -

    -

    - Execute: contractors -

    -

    - Evaluate: citizens (indirectly, as voters) -

    -

    - Pay out: citizens (as taxpayers) -

    -

    - Benefit: citizens, contractors (as profit), officials (as goodwill for - subsequent elections) -

    -

    The wealthy philanthropist model

    -

    - Predict: donor -

    -

    - Front: donor -

    -

    - Execute: donor -

    -

    - Evaluate: donor -

    -

    - Pay out: donor -

    -

    - Benefit: charity beneficiaries -

    -

    The classical effective altruism model

    -

    - Predict: donor, with an Our World In Data tab open -

    -

    - Front: donor -

    -

    - Execute: donor -

    -

    - Evaluate: donor -

    -

    - Pay out: donor -

    -

    - Benefit: charity beneficiaries -

    -

    The charity evaluator model

    -

    - Predict: charity evaluator -

    -

    - Front: donors -

    -

    - Execute: founders -

    -

    - Evaluate: charity evaluator -

    -

    - Pay out: donors -

    -

    - Benefit: charity beneficiaries -

    -

    The impact certificate model

    -

    - Predict: investors -

    -

    - Front: investors -

    -

    - Execute: founder -

    -

    - Evaluate: donor -

    -

    - Pay out: donor -

    -

    - Benefit: founder (as profit), investors (as profit), charity beneficiaries -

    -

    The quadratic funding model

    -

    - Predict: contributors -

    -

    - Front: contributors and matching-pool sponsors -

    -

    - Execute: founder -

    -

    - Evaluate: contributors -

    -

    - Pay out: contributors and matching-pool sponsors -

    -

    - Benefit: contributors + There are a couple of things to notice here. +

    +

    + Some distinctions don’t always occur +

    +

    + “Predict” and “evaluate” are often the same, as are “front” and “pay + out”. These are for cases where money is given out ahead of time, in a + way that’s – well, not no-strings-attached, the founders have to make a + reasonable attempt and not just grab the money and run off to hide out in + antarctica; but it’s not contingent on the project actually succeeding. + maybe the company goes bust, the idea doesn’t work, the nonprofit falls + apart, whatever. Or maybe the project does what it says it will, but it + turns out no one really likes or wants the result after all. But when + there’s not a separate evaluation step and payout step, then those + contingencies are beyond the purview of the funding mechanism to + influence. Trying to fund stuff that won’t fail is baked into the + prediction and fronting steps. +

    +

    + This is especially true for new endeavors. With well-established + charities and companies, the ability to secure future funding is an + indirect incentive to execute current projects well. But for new + organizations that have no track record at all (or sometimes, even + new projects started by existing organizations), you have to fall + back on looking at the track record of the people founding them, and + if the founders are relatively unknown, there’s not much of an + incentive to give them a chance as opposed to taking safe bets. Taking + big swings that you’ll usually miss on, but that will occasionally have + wildly outsized returns, is sometimes called “hits-based giving”; some + effective-altruist organizations explicitly try to use this strategy, + and the reason it’s even necessary for them to specify that – after + all, hits-based investing is how venture capital works be default – is + that the incentives for an individual employee of a typical + philanthropic foundation, like the incentives for an individual + bureaucrat, are to be risk-averse and lean toward safe bets rather + than innovation. +

    +

    + Some mechanisms are “full-stack”; others exist at one or two layers +

    +

    + Some mechanisms, like the examples mentioned above, offer an + implementation of the entire framework. +

    +

    + But a large swath of mechanisms can be understood as innovations on + one or two of the steps. For example: +

    +

    + Challenge prizes: pay out. +

    +

    + S-process: evaluate. +

    +

    + Advance market commitments pay out. +

    +

    + Income share agreements predict and front. +

    +

    + Schelling-point oracles: evaluate. +

    +

    + Patronage: predict and front. +

    +

    + Crowdfunding: pay out. +

    +

    + Regranting: predict. +

    +

    + Prediction markets: predict. +

    +

    + + Coupling and decoupling are central tools of funding-mechanism design + +

    +

    + No mechanism listed here has all its six steps carried out by different + actors. In other words, various subsets of the steps are coupled + together: for instance, quadratic funding is able to price demand for a + public good because potential beneficiaries put up some of their own + money, which ensures they’re credibly signaling their real preference + for the good. Or for a more obvious example, democracy is more + responsive to the needs of its populace than monarchy is because + elected officials have to predict which policies will be popular in + order to win re-election. That’s because coupling two steps aligns + their incentives: combining an information-gathering step with a + money-moving step ensures that decision-makers have skin in the game. +

    +

    + But there’s a limit to the benefits of coupling. The ultimate fully + coupled funding system is “everyone builds everything for themselves + and pays for it themselves.” then the incentives are very aligned! + but on the other hand, it’s tremendously inefficient – no + specialization, no division of labor. +

    +

    + So decoupling turns the knob the other way. For example, in a market + for impact certificates, instead of donors doing all the work of + assessing applicants and allocating funds (by making grants in + advance based on who looks most promising), you split the process + into two layers: investors, who are purely speculators, and donors, + who only have to evaluate projects’ impact afterwards. +

    +

    + Another example is the design-bid-build system of construction + contracting. Instead of the government choosing a single bid for a + contractor to design and then carry out a construction project, + with design-bid-build they accept one contractor’s bid to create a + design, then put out a separate call for a (potentially) different + contractor to implement it; see Asterisk magazine’s + interview + with urbanist Alon Levy, who argues that this framework + is more cost-effective than the single-bid (”design-build”) system in + most places that run it. +

    +

    + In general, decoupling allows for improved competition and + specialization, which is especially useful in philanthropic cases, + where it’s hard to operate a single nonprofit as efficiently as a + corporation of the same size. But too much decoupling can cause + principal-agent problems, as often happens in bureaucracy, and can be + outweighed by economies of scale.

    ) diff --git a/app/articles/approach/6/page.tsx b/app/articles/approach/6/page.tsx index fd77205d..53781998 100644 --- a/app/articles/approach/6/page.tsx +++ b/app/articles/approach/6/page.tsx @@ -1,86 +1,39 @@ -export default function ApproachPage6() { +export default function ApproachPage4() { return (

    Manifund’s Approach to Thinking About Charitable Funding

    -

    -
    - There are a couple of things to notice here. -

    -

    - Some distinctions don’t always occur +

    + Non-Financial Elements

    - “Predict” and “evaluate” are often the same, as are “front” and “pay - out”. These are for cases where money is given out ahead of time, in a - way that’s – well, not no-strings-attached, the founders have to make a - reasonable attempt and not just grab the money and run off to hide out in - antarctica; but it’s not contingent on the project actually succeeding. - maybe the company goes bust, the idea doesn’t work, the nonprofit falls - apart, whatever. Or maybe the project does what it says it will, but it - turns out no one really likes or wants the result after all. But when - there’s not a separate evaluation step and payout step, then those - contingencies are beyond the purview of the funding mechanism to - influence. Trying to fund stuff that won’t fail is baked into the - prediction and fronting steps. -

    + Aside from all the financial considerations, there are a couple of social + and psychological levers entailed in grantmaking. One lesson Manifund has + gleaned from Manifold, our sister prediction-market company, is that you + can operate a finance-like system without any actual money involved, + because people also value: +

    +
      +
    1. + reputation – the glory of visibly performing well and accumulating + scarce prestige, and +
      2. +
    2. + agency – the feeling that their actions cause something to come into + being that wouldn’t have otherwise. This is the sensation of knowing + that when you push the button, something will actually happen. +
      3. +

    - This is especially true for new endeavors. With well-established - charities and companies, the ability to secure future funding is an - indirect incentive to execute current projects well. But for new - organizations that have no track record at all (or sometimes, even - new projects started by existing organizations), you have to fall - back on looking at the track record of the people founding them, and - if the founders are relatively unknown, there’s not much of an - incentive to give them a chance as opposed to taking safe bets. Taking - big swings that you’ll usually miss on, but that will occasionally have - wildly outsized returns, is sometimes called “hits-based giving”; some - effective-altruist organizations explicitly try to use this strategy, - and the reason it’s even necessary for them to specify that – after - all, hits-based investing is how venture capital works be default – is - that the incentives for an individual employee of a typical - philanthropic foundation, like the incentives for an individual - bureaucrat, are to be risk-averse and lean toward safe bets rather - than innovation. -

    -

    - Some mechanisms are “full-stack”; others exist at one or two layers -

    -

    - Some mechanisms, like the examples mentioned above, offer an - implementation of the entire framework. -

    -

    - But a large swath of mechanisms can be understood as innovations on - one or two of the steps. For example: -

    -

    - Challenge prizes: pay out. -

    -

    - S-process: evaluate. -

    -

    - Advance market commitments pay out. -

    -

    - Income share agreements predict and front. -

    -

    - Schelling-point oracles: evaluate. -

    -

    - Patronage: predict and front. -

    -

    - Crowdfunding: pay out. -

    -

    - Regranting: predict. -

    -

    - Prediction markets: predict. + The slower-moving and less accessible a philanthropic organization is, + the harder it is for many individuals to achieve either of those two. On + a crowdfunding platform like Kickstarter, users can get a sense of + agency, but not so much of accumulating reputation. Anything that + resembles a game, like investing or predicting, and that has real-world + effects, hits both these criteria: you can succeed relative to others, + and you can succeed in the sense that you did some actions on a website + and now a real physical project is getting accomplished.

    ) diff --git a/app/articles/approach/7/page.tsx b/app/articles/approach/7/page.tsx index fe2a8803..9bccb919 100644 --- a/app/articles/approach/7/page.tsx +++ b/app/articles/approach/7/page.tsx @@ -4,58 +4,14 @@ export default function ApproachPage4() {

    Manifund’s Approach to Thinking About Charitable Funding

    -

    - - Coupling and decoupling are central tools of funding-mechanism design - -

    +

    + Effective Iteration +

    - No mechanism listed here has all its six steps carried out by different - actors. In other words, various subsets of the steps are coupled - together: for instance, quadratic funding is able to price demand for a - public good because potential beneficiaries put up some of their own - money, which ensures they’re credibly signaling their real preference - for the good. Or for a more obvious example, democracy is more - responsive to the needs of its populace than monarchy is because - elected officials have to predict which policies will be popular in - order to win re-election. That’s because coupling two steps aligns - their incentives: combining an information-gathering step with a - money-moving step ensures that decision-makers have skin in the game. + The examples of the internet, cultural evolution, and the stock market – not to mention prediction markets! – suggest that the most accurate information is produced at scale by a decentralized mass of actors with an incentive for the right answers to bubble to the top. First-principles speculation only goes so far – it just isn’t possible for one person or team to process as much information as a network can.

    - But there’s a limit to the benefits of coupling. The ultimate fully - coupled funding system is “everyone builds everything for themselves - and pays for it themselves.” then the incentives are very aligned! - but on the other hand, it’s tremendously inefficient – no - specialization, no division of labor. -

    -

    - So decoupling turns the knob the other way. For example, in a market - for impact certificates, instead of donors doing all the work of - assessing applicants and allocating funds (by making grants in - advance based on who looks most promising), you split the process - into two layers: investors, who are purely speculators, and donors, - who only have to evaluate projects’ impact afterwards. -

    -

    - Another example is the design-bid-build system of construction - contracting. Instead of the government choosing a single bid for a - contractor to design and then carry out a construction project, - with design-bid-build they accept one contractor’s bid to create a - design, then put out a separate call for a (potentially) different - contractor to implement it; see Asterisk magazine’s - interview - with urbanist Alon Levy, who argues that this framework - is more cost-effective than the single-bid (”design-build”) system in - most places that run it. -

    -

    - In general, decoupling allows for improved competition and - specialization, which is especially useful in philanthropic cases, - where it’s hard to operate a single nonprofit as efficiently as a - corporation of the same size. But too much decoupling can cause - principal-agent problems, as often happens in bureaucracy, and can be - outweighed by economies of scale. + The way to use this principle to figure out which projects will work is to find ways to try lots of projects quickly: the higher the speed and greater the volume, the more effective the information-gathering. In 2023, Manifund’s first year of operations, the platform sent $2.06 million in funds to 88 projects. One of our explicit goals is to offer grantees turnaround times measured in weeks or even days instead of months. We hypothesize that timing is an underrated issue for projects that are smaller, newer, and/or run by individuals, as opposed to projects that are already ongoing or are part of existing organizations: it’s difficult to arrange your life so that you’ll be free in several months for an opportunity that might or might not materialize. In these cases, agility and speed fill a critical gap for applicants who would otherwise have been prohibitively inconvenienced by long turnaround times.

    ) diff --git a/app/articles/approach/8/page.tsx b/app/articles/approach/8/page.tsx index 3127ed2b..3548cf2e 100644 --- a/app/articles/approach/8/page.tsx +++ b/app/articles/approach/8/page.tsx @@ -4,36 +4,55 @@ export default function ApproachPage4() {

    Manifund’s Approach to Thinking About Charitable Funding

    -

    - Psychological -

    +

    + What About Different Value Systems? +

    - Aside from all the financial considerations, there are a couple of social - and psychological levers entailed in grantmaking. One lesson Manifund has - gleaned from Manifold, our sister prediction-market company, is that you - can operate a finance-like system without any actual money involved, - because people also value: + One limitation of this model is that it’s not obvious how well it applies + to entities that don’t naturally fit into standard economic models because + they can’t easily participate in markets: children, animals, people who + might exist in the distant future. For people whose value systems heavily + incorporate any of these groups’ preferences, all market-based funding + solutions might prove inadequate. In these cases, funding is closer to a + strict question of altruism than a coordination problem; donors have to + speculate on behalf of those who can’t (financially) speculate for + themselves. These causes lend themselves better to direct improvements + in prediction and evaluation than to financial mechanisms that couple + those steps with fronting and paying out; it’s not really clear how you’d + align your own incentives with the incentives of someone in the year + 2300, except by saying “I care abstractly about making future people + happy.”

    -
      -
    1. - reputation – the glory of visibly performing well and accumulating - scarce prestige, and -
      2. -
    2. - agency – the feeling that their actions cause something to come into - being that wouldn’t have otherwise. This is the sensation of knowing - that when you push the button, something will actually happen. -
      3. -

    - The slower-moving and less accessible a philanthropic organization is, - the harder it is for many individuals to achieve either of those two. On - a crowdfunding platform like Kickstarter, users can get a sense of - agency, but not so much of accumulating reputation. Anything that - resembles a game, like investing or predicting, and that has real-world - effects, hits both these criteria: you can succeed relative to others, - and you can succeed in the sense that you did some actions on a website - and now a real physical project is getting accomplished. + Even in these areas, though, there are often some smaller subproblems + related to coordination, and so market mechanisms can offer at least + some improvements. For example, grant proposals often become the + victims of “funder chicken,” where each potential donor, having a + limited budget, likes a proposal but hopes one of the others will fund + it first. And even if you can’t evaluate one year from now whether a + project intended to improve outcomes on a century-long timescale has + succeeded in that, you can at least check whether the project did what + it said it would, or whether it achieved the kind of result that you + already believe for other reasons would be valuable for the far future. + it’s hard to incentivize terminal goals this way, but for incentivizing + instrumental goals, market structures are pretty good. +

    +

    + On some level, you can think of altruism itself as a public good: + like, ten thousand people are all sad about cows getting factory-farmed + and would rather the cows be freed to roam on beautiful verdant + pastures, and so cow liberation is a public good to all those people + insofar as it’s something that would make them all happy. This doesn’t + necessarily evaluate how much the cow would prefer freedom to its + current life; you could just as easily imagine that ten thousand other + people would enjoy seeing a herd of cows dressed up in identical little + bow ties, and so purchasing a pack of wide-necked bow ties would be a + public-goods-funding coordination problem in just the same way. The + “figuring out what would be best for cows” problem, and below it the + more fundamental “figuring out what my altruistic values are” problem, + aren’t easily solvable by markets. But in either case, once you make + up your mind as to what your goals are, market mechanisms become useful + tools to help you get there.

    ) diff --git a/app/articles/approach/9/page.tsx b/app/articles/approach/9/page.tsx deleted file mode 100644 index fa61daf1..00000000 --- a/app/articles/approach/9/page.tsx +++ /dev/null @@ -1,18 +0,0 @@ -export default function ApproachPage4() { - return ( -
    -

    - Manifund’s Approach to Thinking About Charitable Funding -

    -

    - Effective Iteration -

    -

    - The examples of the internet, cultural evolution, and the stock market – not to mention prediction markets! – suggest that the most accurate information is produced at scale by a decentralized mass of actors with an incentive for the right answers to bubble to the top. First-principles speculation only goes so far – it just isn’t possible for one person or team to process as much information as a network can. -

    -

    - The way to use this principle to figure out which projects will work is to find ways to try lots of projects quickly: the higher the speed and greater the volume, the more effective the information-gathering. In 2023, Manifund’s first year of operations, the platform sent $2.06 million in funds to 88 projects. One of our explicit goals is to offer grantees turnaround times measured in weeks or even days instead of months. We hypothesize that timing is an underrated issue for projects that are smaller, newer, and/or run by individuals, as opposed to projects that are already ongoing or are part of existing organizations: it’s difficult to arrange your life so that you’ll be free in several months for an opportunity that might or might not materialize. In these cases, agility and speed fill a critical gap for applicants who would otherwise have been prohibitively inconvenienced by long turnaround times. -

    -
    - ) -} \ No newline at end of file From af2343f711de0bb2bf1268b948c86b85b6f34611 Mon Sep 17 00:00:00 2001 From: Lily Jordan Date: Sat, 30 Mar 2024 10:04:50 -0700 Subject: [PATCH 12/16] generic article page component --- components/article-page.tsx | 32 ++++++++++++++++++++++++++++++++ 1 file changed, 32 insertions(+) create mode 100644 components/article-page.tsx diff --git a/components/article-page.tsx b/components/article-page.tsx new file mode 100644 index 00000000..96e0d67e --- /dev/null +++ b/components/article-page.tsx @@ -0,0 +1,32 @@ +import Link from 'next/link' +import { ArrowLongRightIcon } from '@heroicons/react/20/solid' + +export default function ArticlePage(props: { + articleTitle: string, + pageTitle: string, + nextLink: string, + nextLinkText: string, + content: string +}) { + const { articleTitle, pageTitle, nextLink, nextLinkText, content } = props + return ( +
    +

    + {articleTitle} +

    +

    + {pageTitle} +

    + {content} + + {nextLinkText} + + +
    + ) + } +} \ No newline at end of file From 26aa7f1c899e7806119667e78961b419680f96b1 Mon Sep 17 00:00:00 2001 From: Lily Jordan Date: Sat, 30 Mar 2024 10:05:33 -0700 Subject: [PATCH 13/16] titles; how to get involved page --- app/articles/approach/1/page.tsx | 2 +- app/articles/approach/2/page.tsx | 2 +- app/articles/approach/3/page.tsx | 4 ++-- app/articles/approach/4/page.tsx | 2 +- app/articles/approach/5/page.tsx | 4 ++-- app/articles/approach/6/page.tsx | 4 ++-- app/articles/approach/7/page.tsx | 4 ++-- app/articles/approach/8/page.tsx | 14 ++++++++++++-- app/articles/approach/9/page.tsx | 15 +++++++++++++++ app/articles/approach/page.tsx | 2 +- app/sidebar.tsx | 18 +++++++++--------- 11 files changed, 48 insertions(+), 23 deletions(-) create mode 100644 app/articles/approach/9/page.tsx diff --git a/app/articles/approach/1/page.tsx b/app/articles/approach/1/page.tsx index a70c69d0..447ed538 100644 --- a/app/articles/approach/1/page.tsx +++ b/app/articles/approach/1/page.tsx @@ -2,7 +2,7 @@ export default function ApproachPage1() { return (

    - Manifund’s Approach to Thinking About Charitable Funding + A Framework for Funding Experiments

    From 2023 retrospective: “Our initial thesis was that grant applications and screening largely can be done in public, and should be.” diff --git a/app/articles/approach/2/page.tsx b/app/articles/approach/2/page.tsx index 4da10724..6e638ede 100644 --- a/app/articles/approach/2/page.tsx +++ b/app/articles/approach/2/page.tsx @@ -2,7 +2,7 @@ export default function ApproachPage2() { return (

    - Manifund’s Approach to Thinking About Charitable Funding + A Framework for Funding Experiments

    Programs Manifund has run diff --git a/app/articles/approach/3/page.tsx b/app/articles/approach/3/page.tsx index 590b6aac..7eb36ac7 100644 --- a/app/articles/approach/3/page.tsx +++ b/app/articles/approach/3/page.tsx @@ -1,8 +1,8 @@ -export default function ApproachPage4() { +export default function ApproachPage3() { return (

    - Manifund’s Approach to Thinking About Charitable Funding + A Framework for Funding Experiments

    The Six Steps of Funding diff --git a/app/articles/approach/4/page.tsx b/app/articles/approach/4/page.tsx index 28a86d8d..0acd9cf7 100644 --- a/app/articles/approach/4/page.tsx +++ b/app/articles/approach/4/page.tsx @@ -2,7 +2,7 @@ export default function ApproachPage4() { return (

    - Manifund’s Approach to Thinking About Charitable Funding + A Framework for Funding Experiments

    Some Examples diff --git a/app/articles/approach/5/page.tsx b/app/articles/approach/5/page.tsx index 904e97e5..a278df9a 100644 --- a/app/articles/approach/5/page.tsx +++ b/app/articles/approach/5/page.tsx @@ -1,8 +1,8 @@ -export default function ApproachPage6() { +export default function ApproachPage5() { return (

    - Manifund’s Approach to Thinking About Charitable Funding + A Framework for Funding Experiments

    Observations About These Examples diff --git a/app/articles/approach/6/page.tsx b/app/articles/approach/6/page.tsx index 53781998..031654e0 100644 --- a/app/articles/approach/6/page.tsx +++ b/app/articles/approach/6/page.tsx @@ -1,8 +1,8 @@ -export default function ApproachPage4() { +export default function ApproachPage6() { return (

    - Manifund’s Approach to Thinking About Charitable Funding + A Framework for Funding Experiments

    Non-Financial Elements diff --git a/app/articles/approach/7/page.tsx b/app/articles/approach/7/page.tsx index 9bccb919..37b48da8 100644 --- a/app/articles/approach/7/page.tsx +++ b/app/articles/approach/7/page.tsx @@ -1,8 +1,8 @@ -export default function ApproachPage4() { +export default function ApproachPage7() { return (

    - Manifund’s Approach to Thinking About Charitable Funding + A Framework for Funding Experiments

    Effective Iteration diff --git a/app/articles/approach/8/page.tsx b/app/articles/approach/8/page.tsx index 3548cf2e..4bc51869 100644 --- a/app/articles/approach/8/page.tsx +++ b/app/articles/approach/8/page.tsx @@ -1,8 +1,11 @@ -export default function ApproachPage4() { +import Link from 'next/link' +import { ArrowLongRightIcon } from '@heroicons/react/20/solid' + +export default function ApproachPage8() { return (

    - Manifund’s Approach to Thinking About Charitable Funding + A Framework for Funding Experiments

    What About Different Value Systems? @@ -54,6 +57,13 @@ export default function ApproachPage4() { up your mind as to what your goals are, market mechanisms become useful tools to help you get there.

    + + Next: How you can get involved + +

    ) } \ No newline at end of file diff --git a/app/articles/approach/9/page.tsx b/app/articles/approach/9/page.tsx new file mode 100644 index 00000000..7b4c1144 --- /dev/null +++ b/app/articles/approach/9/page.tsx @@ -0,0 +1,15 @@ +export default function ApproachPage9() { + return ( +
    +

    + A Framework for Funding Experiments +

    +

    + How You Can Get Involved +

    +

    + +

    +
    + ) +} \ No newline at end of file diff --git a/app/articles/approach/page.tsx b/app/articles/approach/page.tsx index 11b6343a..ac60ee48 100644 --- a/app/articles/approach/page.tsx +++ b/app/articles/approach/page.tsx @@ -4,7 +4,7 @@ export default function ApproachPage() { return (

    - Manifund’s Approach to Thinking About Charitable Funding + A Framework For Funding Experiments

    From 2023 retrospective: "Our initial thesis was that grant applications and screening largely can be done in public, and should be." diff --git a/app/sidebar.tsx b/app/sidebar.tsx index 2a6a6003..4f5b1cd5 100644 --- a/app/sidebar.tsx +++ b/app/sidebar.tsx @@ -70,18 +70,18 @@ export default async function Sidebar() { From b1906c86c416f699c398446de01c67e93054e4ca Mon Sep 17 00:00:00 2001 From: Lily Jordan Date: Sat, 30 Mar 2024 10:42:53 -0700 Subject: [PATCH 14/16] convert everything to use ArticlePage component --- app/articles/approach/1/page.tsx | 56 +++++++++++++++++--------------- app/articles/approach/2/page.tsx | 22 ++++++++----- app/articles/approach/3/page.tsx | 24 ++++++++------ app/articles/approach/4/page.tsx | 22 ++++++++----- app/articles/approach/5/page.tsx | 22 ++++++++----- app/articles/approach/6/page.tsx | 32 ++++++++++-------- app/articles/approach/7/page.tsx | 22 ++++++++----- app/articles/approach/8/page.tsx | 31 ++++++++---------- app/articles/approach/9/page.tsx | 25 +++++++------- components/article-page.tsx | 42 ++++++++++++------------ 10 files changed, 160 insertions(+), 138 deletions(-) diff --git a/app/articles/approach/1/page.tsx b/app/articles/approach/1/page.tsx index 447ed538..c214eb07 100644 --- a/app/articles/approach/1/page.tsx +++ b/app/articles/approach/1/page.tsx @@ -1,31 +1,26 @@ +import ArticlePage from '@/components/article-page' +import { Fragment } from 'react' + + export default function ApproachPage1() { - return ( -

    -

    - A Framework for Funding Experiments -

    -

    - From 2023 retrospective: “Our initial thesis was that grant applications and screening largely can be done in public, and should be.” -

    -

    - Economic Experimentation -

    + const content = ( +

    - Manifund’s sister company, Manifold, runs experiments with - prediction markets. Because Manifold runs on play money, and because - users get to participate on the market-creation side as well as the - prediction side, it’s arguably the best playground in the world for - experimenting rapidly with financial mechanisms. Since official - currency isn’t involved, both Manifold and its users (and the two in - combination) can run experiments rapid-fire without jumping through - many hoops. For instance, Manifold has tried various types of - automatic market-making algorithms, mechanisms for multiple-choice - questions, short selling, limit orders, contests, leagues, even a - dating app spinoff. And users have made bots, non-question questions - like “___ stock (never resolves)”, markets that serve as bounties to - incentivize particular events to happen, questions with meta - resolution criteria like “Will the total volume of YES trading on - this question be more than x mana?”. + Manifund’s sister company, Manifold, runs experiments with + prediction markets. Because Manifold runs on play money, and because + users get to participate on the market-creation side as well as the + prediction side, it’s arguably the best playground in the world for + experimenting rapidly with financial mechanisms. Since official + currency isn’t involved, both Manifold and its users (and the two in + combination) can run experiments rapid-fire without jumping through + many hoops. For instance, Manifold has tried various types of + automatic market-making algorithms, mechanisms for multiple-choice + questions, short selling, limit orders, contests, leagues, even a + dating app spinoff. And users have made bots, non-question questions + like “___ stock (never resolves)”, markets that serve as bounties to + incentivize particular events to happen, questions with meta + resolution criteria like “Will the total volume of YES trading on + this question be more than x mana?”.

    Manifund is different, because lots of real money is moving through our @@ -48,6 +43,13 @@ export default function ApproachPage1() { programs more effective? -

    + ) + return } \ No newline at end of file diff --git a/app/articles/approach/2/page.tsx b/app/articles/approach/2/page.tsx index 6e638ede..11b6a34c 100644 --- a/app/articles/approach/2/page.tsx +++ b/app/articles/approach/2/page.tsx @@ -1,12 +1,9 @@ +import ArticlePage from '@/components/article-page' +import { Fragment } from 'react' + export default function ApproachPage2() { - return ( -
    -

    - A Framework for Funding Experiments -

    -

    - Programs Manifund has run -

    + const content = ( +

    Regranting

    ACX grants

    @@ -41,6 +38,13 @@ export default function ApproachPage2() { So we’ve come up with a first pass at a general framework for thinking about the design of novel funding mechanisms.

    -
    + ) + return } \ No newline at end of file diff --git a/app/articles/approach/3/page.tsx b/app/articles/approach/3/page.tsx index 7eb36ac7..c1aba319 100644 --- a/app/articles/approach/3/page.tsx +++ b/app/articles/approach/3/page.tsx @@ -1,13 +1,10 @@ +import ArticlePage from '@/components/article-page' +import { Fragment } from 'react' + export default function ApproachPage3() { - return ( -
    -

    - A Framework for Funding Experiments -

    -

    - The Six Steps of Funding -

    -

    1. Predict

    + const content = ( + +

    1. Predict

    What does it entail? Make a good guess about which projects will be likely to achieve their aims, what impacts those @@ -71,6 +68,13 @@ export default function ApproachPage3() { What are the desirables? Be helped by the project. Do not be harmed by the project.

    -
    + ) + return } \ No newline at end of file diff --git a/app/articles/approach/4/page.tsx b/app/articles/approach/4/page.tsx index 0acd9cf7..34eab45b 100644 --- a/app/articles/approach/4/page.tsx +++ b/app/articles/approach/4/page.tsx @@ -1,12 +1,9 @@ +import ArticlePage from '@/components/article-page' +import { Fragment } from 'react' + export default function ApproachPage4() { - return ( -
    -

    - A Framework for Funding Experiments -

    -

    - Some Examples -

    + const content = ( +

    The startup model

    Predict: investors @@ -160,6 +157,13 @@ export default function ApproachPage4() {

    Benefit: contributors

    -
    + ) + return } \ No newline at end of file diff --git a/app/articles/approach/5/page.tsx b/app/articles/approach/5/page.tsx index a278df9a..b6c38216 100644 --- a/app/articles/approach/5/page.tsx +++ b/app/articles/approach/5/page.tsx @@ -1,12 +1,9 @@ +import ArticlePage from '@/components/article-page' +import { Fragment } from 'react' + export default function ApproachPage5() { - return ( -
    -

    - A Framework for Funding Experiments -

    -

    - Observations About These Examples -

    + const content = ( +

    There are a couple of things to notice here.

    @@ -137,6 +134,13 @@ export default function ApproachPage5() { principal-agent problems, as often happens in bureaucracy, and can be outweighed by economies of scale.

    -
    + ) + return } \ No newline at end of file diff --git a/app/articles/approach/6/page.tsx b/app/articles/approach/6/page.tsx index 031654e0..03f1c470 100644 --- a/app/articles/approach/6/page.tsx +++ b/app/articles/approach/6/page.tsx @@ -1,18 +1,15 @@ +import ArticlePage from '@/components/article-page' +import { Fragment } from 'react' + export default function ApproachPage6() { - return ( -
    -

    - A Framework for Funding Experiments -

    -

    - Non-Financial Elements -

    + const content = ( +

    - Aside from all the financial considerations, there are a couple of social - and psychological levers entailed in grantmaking. One lesson Manifund has - gleaned from Manifold, our sister prediction-market company, is that you - can operate a finance-like system without any actual money involved, - because people also value: + Aside from all the financial considerations, there are a couple of social + and psychological levers entailed in grantmaking. One lesson Manifund has + gleaned from Manifold, our sister prediction-market company, is that you + can operate a finance-like system without any actual money involved, + because people also value:

    1. @@ -35,6 +32,13 @@ export default function ApproachPage6() { and you can succeed in the sense that you did some actions on a website and now a real physical project is getting accomplished.

      -
    + ) + return } \ No newline at end of file diff --git a/app/articles/approach/7/page.tsx b/app/articles/approach/7/page.tsx index 37b48da8..eb111d12 100644 --- a/app/articles/approach/7/page.tsx +++ b/app/articles/approach/7/page.tsx @@ -1,18 +1,22 @@ +import ArticlePage from '@/components/article-page' +import { Fragment } from 'react' + export default function ApproachPage7() { - return ( -
    -

    - A Framework for Funding Experiments -

    -

    - Effective Iteration -

    + const content = ( +

    The examples of the internet, cultural evolution, and the stock market – not to mention prediction markets! – suggest that the most accurate information is produced at scale by a decentralized mass of actors with an incentive for the right answers to bubble to the top. First-principles speculation only goes so far – it just isn’t possible for one person or team to process as much information as a network can.

    The way to use this principle to figure out which projects will work is to find ways to try lots of projects quickly: the higher the speed and greater the volume, the more effective the information-gathering. In 2023, Manifund’s first year of operations, the platform sent $2.06 million in funds to 88 projects. One of our explicit goals is to offer grantees turnaround times measured in weeks or even days instead of months. We hypothesize that timing is an underrated issue for projects that are smaller, newer, and/or run by individuals, as opposed to projects that are already ongoing or are part of existing organizations: it’s difficult to arrange your life so that you’ll be free in several months for an opportunity that might or might not materialize. In these cases, agility and speed fill a critical gap for applicants who would otherwise have been prohibitively inconvenienced by long turnaround times.

    -
    + ) + return } \ No newline at end of file diff --git a/app/articles/approach/8/page.tsx b/app/articles/approach/8/page.tsx index 4bc51869..e71c344e 100644 --- a/app/articles/approach/8/page.tsx +++ b/app/articles/approach/8/page.tsx @@ -1,15 +1,10 @@ -import Link from 'next/link' -import { ArrowLongRightIcon } from '@heroicons/react/20/solid' +import ArticlePage from '@/components/article-page' +import { Fragment } from 'react' + export default function ApproachPage8() { - return ( -
    -

    - A Framework for Funding Experiments -

    -

    - What About Different Value Systems? -

    + const content = ( +

    One limitation of this model is that it’s not obvious how well it applies to entities that don’t naturally fit into standard economic models because @@ -57,13 +52,13 @@ export default function ApproachPage8() { up your mind as to what your goals are, market mechanisms become useful tools to help you get there.

    - - Next: How you can get involved - - -
    + ) + return } \ No newline at end of file diff --git a/app/articles/approach/9/page.tsx b/app/articles/approach/9/page.tsx index 7b4c1144..f0bc6ed9 100644 --- a/app/articles/approach/9/page.tsx +++ b/app/articles/approach/9/page.tsx @@ -1,15 +1,16 @@ +import ArticlePage from '@/components/article-page' +import { Fragment } from 'react' + export default function ApproachPage9() { - return ( -
    -

    - A Framework for Funding Experiments -

    -

    - How You Can Get Involved -

    -

    - -

    -
    + const content = ( + + ) + return } \ No newline at end of file diff --git a/components/article-page.tsx b/components/article-page.tsx index 96e0d67e..712d1b7a 100644 --- a/components/article-page.tsx +++ b/components/article-page.tsx @@ -1,32 +1,32 @@ import Link from 'next/link' import { ArrowLongRightIcon } from '@heroicons/react/20/solid' +import { ReactNode } from 'react' export default function ArticlePage(props: { articleTitle: string, pageTitle: string, nextLink: string, nextLinkText: string, - content: string + content: ReactNode }) { const { articleTitle, pageTitle, nextLink, nextLinkText, content } = props - return ( -
    -

    - {articleTitle} -

    -

    - {pageTitle} -

    - {content} - - {nextLinkText} - - -
    - ) - } + return ( +
    +

    + {articleTitle} +

    +

    + {pageTitle} +

    + {content} + + {nextLinkText} + + +
    + ) } \ No newline at end of file From 567dd01ea96bb7dd9d3227d2557997de4e15bafb Mon Sep 17 00:00:00 2001 From: Lily Jordan Date: Tue, 2 Apr 2024 04:03:42 -0400 Subject: [PATCH 15/16] previous button; added explanations to 4 and 5 --- app/articles/approach/2/page.tsx | 9 ++++---- app/articles/approach/3/page.tsx | 1 + app/articles/approach/4/page.tsx | 19 +++++++++++++++- app/articles/approach/5/page.tsx | 27 ++++++++++++++--------- app/articles/approach/6/page.tsx | 1 + app/articles/approach/7/page.tsx | 1 + app/articles/approach/8/page.tsx | 1 + app/articles/approach/9/page.tsx | 23 +++++++++++++++++-- components/article-page.tsx | 38 ++++++++++++++++++++++---------- 9 files changed, 90 insertions(+), 30 deletions(-) diff --git a/app/articles/approach/2/page.tsx b/app/articles/approach/2/page.tsx index 11b6a34c..d3e8580c 100644 --- a/app/articles/approach/2/page.tsx +++ b/app/articles/approach/2/page.tsx @@ -1,14 +1,14 @@ import ArticlePage from '@/components/article-page' import { Fragment } from 'react' +import Link from 'next/link' export default function ApproachPage2() { const content = (

    -

    Regranting

    -

    ACX grants

    -

    Impact certificates

    -

    Assurance contract auctions

    + Regranting + Impact markets + ACX Grants

    Is there anything general we can say about these kinds of experiments? @@ -43,6 +43,7 @@ export default function ApproachPage2() { return +

    + Let’s apply this framework to some existing fundraising models, as well as some more experimental ones. +

    The startup model

    + A startup founder raises money to run an early-stage company by selling equity in the company’s stock.

    Predict: investors

    @@ -24,6 +28,7 @@ export default function ApproachPage4() { Benefit: founder (as profit), investors (as profit), consumers (as consumer surplus)

    The bootstrapping model

    + A founder starts a company with their own money, rather than selling equity early.

    Predict: founder

    @@ -43,6 +48,7 @@ export default function ApproachPage4() { Benefit: founder (as profit), consumers (as consumer surplus)

    The government model

    + The government taxes citizens and uses the proceeds to pay for public-benefit projects.

    Predict: elected officials or bureaucrats

    @@ -63,6 +69,7 @@ export default function ApproachPage4() { subsequent elections)

    The wealthy philanthropist model

    + A wealthy person or family donates large sums of money to charity.

    Predict: donor

    @@ -82,6 +89,7 @@ export default function ApproachPage4() { Benefit: charity beneficiaries

    The classical effective altruism model

    + Altruistically-minded people try to donate to charities that will have the most marginal impact.

    Predict: donor, with an Our World In Data tab open

    @@ -101,6 +109,8 @@ export default function ApproachPage4() { Benefit: charity beneficiaries

    The charity evaluator model

    + An organization assesses the effectiveness of various charities, and then donors either donate to those charities + directly, or donate to the organization which then funds its chosen charities.

    Predict: charity evaluator

    @@ -119,7 +129,9 @@ export default function ApproachPage4() {

    Benefit: charity beneficiaries

    -

    The impact certificate model

    +

    The impact certificate model

    + Donors offer to retroactively award prizes for successful charitable projects. + The founder of a such a project funds their efforts by selling equity in whatever prize money the project later receives.

    Predict: investors

    @@ -139,7 +151,11 @@ export default function ApproachPage4() { Benefit: founder (as profit), investors (as profit), charity beneficiaries

    The quadratic funding model

    + Philanthropic sponsors fund a matching pool, whose funds are used to match donations to public-goods + projects in a way that incentivizes small donors to donate to the projects that most benefit other donors + too.

    + Predict: contributors

    @@ -162,6 +178,7 @@ export default function ApproachPage4() { return

    - Challenge prizes: pay out. + Challenge prizes: pay out. A funder agrees in advance to pay for a certain achievement.

    - S-process: evaluate. + S-process: evaluate. A funder enlists a group of recommenders to evaluate potential grants + according to a simulation process.

    - Advance market commitments pay out. + Advance market commitments: pay out. A large service provider, like a government, agrees in advance + to buy a product meeting certain criteria.

    - Income share agreements predict and front. + Income share agreements: predict and front. A funder gives cash upfront in exchange for equity + in the future salaries of people they predict will eventually have high incomes.

    - Schelling-point oracles: evaluate. + Patronage: predict and front. A patron philanthropically funds the career of a person creating + public goods like art.

    - Patronage: predict and front. + Crowdfunding: pay out. A group of people, often acting under a mutual assurance contract, collectively + fund a project.

    - Crowdfunding: pay out. + Regranting: predict. A donor distributes funds across several regrantors, each of whom makes grants to + promising projects in their area of expertise.

    - Regranting: predict. -

    -

    - Prediction markets: predict. + Prediction markets: predict. People bet on how future events, such as the results of a project, will + turn out.

    @@ -139,6 +143,7 @@ export default function ApproachPage5() { return +

    + If you’re interested in learning more about or getting involved with Manifund’s flavor of funding innovations, + we’d love to hear from you. +

    +
      +
    • + Apply for grant funding through our open call process, + or through one of our programs. +
      • +
    • + Donate to a project or invest in its impact certificates. +
      • +
    • + Keep up with Manifund on Substack and Twitter. +
      • +
    • + Join the Discord. +
      • +
    ) return } \ No newline at end of file diff --git a/components/article-page.tsx b/components/article-page.tsx index 712d1b7a..f9c5ed41 100644 --- a/components/article-page.tsx +++ b/components/article-page.tsx @@ -1,15 +1,17 @@ import Link from 'next/link' -import { ArrowLongRightIcon } from '@heroicons/react/20/solid' +import { ArrowLongLeftIcon, ArrowLongRightIcon } from '@heroicons/react/20/solid' import { ReactNode } from 'react' export default function ArticlePage(props: { articleTitle: string, pageTitle: string, - nextLink: string, - nextLinkText: string, + prevLink?: string, + nextLink?: string, + nextLinkText?: string, content: ReactNode }) { - const { articleTitle, pageTitle, nextLink, nextLinkText, content } = props + const { articleTitle, pageTitle, prevLink, nextLink, nextLinkText, + content } = props return (

    @@ -19,14 +21,26 @@ export default function ArticlePage(props: { {pageTitle}

    {content} - - {nextLinkText} - - +
    + {prevLink && + + + Previous + + } + {nextLink && nextLinkText && + + {nextLinkText} + + + } +
    ) } \ No newline at end of file From d8e1af87bece153d487cc63dd55dbf938445e402 Mon Sep 17 00:00:00 2001 From: Lily Jordan Date: Tue, 2 Apr 2024 05:10:10 -0400 Subject: [PATCH 16/16] explanations of programs --- app/articles/approach/1/page.tsx | 8 ++++---- app/articles/approach/2/page.tsx | 22 ++++++++++++++++++---- app/articles/approach/3/page.tsx | 6 ++++-- 3 files changed, 26 insertions(+), 10 deletions(-) diff --git a/app/articles/approach/1/page.tsx b/app/articles/approach/1/page.tsx index c214eb07..6dbd37c2 100644 --- a/app/articles/approach/1/page.tsx +++ b/app/articles/approach/1/page.tsx @@ -17,7 +17,7 @@ export default function ApproachPage1() { automatic market-making algorithms, mechanisms for multiple-choice questions, short selling, limit orders, contests, leagues, even a dating app spinoff. And users have made bots, non-question questions - like “___ stock (never resolves)”, markets that serve as bounties to + like “[thing] stock (never resolves)”, markets that serve as bounties to incentivize particular events to happen, questions with meta resolution criteria like “Will the total volume of YES trading on this question be more than x mana?”. @@ -25,9 +25,9 @@ export default function ApproachPage1() {

    Manifund is different, because lots of real money is moving through our programs. But we’re still small and fast-moving, - and our ethos is similar: we aim to be an incubator - for lots of small- and medium-scale experiments testing out funding - mechanisms for charity and public goods. + and our ethos is similar: we aim to iterate quickly on + small- and medium-scale experiments testing out the best ways + to fund charity and public goods.

    Effective altruism originally asked the question, “What properties make diff --git a/app/articles/approach/2/page.tsx b/app/articles/approach/2/page.tsx index d3e8580c..1db8a251 100644 --- a/app/articles/approach/2/page.tsx +++ b/app/articles/approach/2/page.tsx @@ -5,10 +5,24 @@ import Link from 'next/link' export default function ApproachPage2() { const content = ( + In 2023, Manifund ran a few different granting programs. +

      +
    • + Regranting: Donors delegate grantmaking budgets to regrantors, who + allocate the funds to the most promising projects they come across. +
      • +
    • + Impact markets: Donors offer retroactive prize funding to + successful and impactful projects, and project founders sell shares in these prizes to fund their projects. +
      • +
    • + Open call with assurance-contract donations: Manifund runs an open call for projects, + which operates like Kickstarter for charity: if a project receives donation offers totaling at least its minimum funding goal, + then the donations kick in and it receives funding. +
      • +

    - Regranting - Impact markets - ACX Grants + (You can read more about these programs in our 2023 retrospective.)

    Is there anything general we can say about these kinds of experiments? @@ -28,7 +42,7 @@ export default function ApproachPage2() { metrics.

    - But also, it seems a little contrived? Each individual step of this + But also, it’s easy to feel like the mechanism just came out of nowhere. Each individual step of this process is perfectly understandable if you understand the analogous element of private markets. But altogether, it’s a lot of new moving parts to conceptualize at once, and it’s not obvious how to come up diff --git a/app/articles/approach/3/page.tsx b/app/articles/approach/3/page.tsx index c0a5da99..86db20a6 100644 --- a/app/articles/approach/3/page.tsx +++ b/app/articles/approach/3/page.tsx @@ -40,8 +40,10 @@ export default function ApproachPage3() { project was.

    - What are the desirables? Align the measure with - the target – that is, [explain further]. Account for + What are the desirables? Figure out both how successful + the project was and how much impact that success had. Make the targets clear + so founders know what to aim for, but avoid using measures that are too easy + to game. Account for externalities – for example, if the project was moderately beneficial to the group that commissioned it, but also benefited lots of other people as well, then we’d like that to be factored in.