Cause X: Public Stretch Prizes

Summary

All grants should be matched with a parallel public prize of ten times the original grant amount, but with a 10% chance of awarding. To activate the prize, grant applicants should spell out the quantifiable impacts of their grant so that forecasters can then tweak those numbers and fill out a prize template. The worst case scenario would be an extra inducement to the original grantee to go above and beyond the original goal, i.e., by creating a “stretch goal.” In the best-case scenario, the public commits extra resources in pursuit of the prize, similar to what happened with the X Prize or the Longitude Prize.

Intro to Prize Philanthropy

According to Robin Hanson, prize-like funding had its heyday in the 1700s. Most notably, John Harrison, a clockmaker by trade, worked in his spare time to solve the problem of calculating longitude while at sea. His invention of the maritime chronometer won him the £20,000 Longitude Prize and arguably extended British maritime supremacy by a hundred years. (Source: Longitude)

But if these prizes were so useful, why haven’t we seen more since? According to Hanson, as government patronage grew, funding became more non-local and non-autocratic (i.e., bureaucratic). Prize-like funding gave way to grant-like funding as patrons “were susceptible to distributive pressures from leaders of scientific societies, who preferred the ‘pork’ of increased discretion over the money that passed through their hands.”

So, if the barrier to prize philanthropy has always been the incentive structure of the status quo, why will this time be different? Fortunately, we have a new cultural innovation in the form of technocratic funding, most visibly in organizations like the Bill & Melinda Gates Foundation but also in effective altruistic organizations such as GiveWell and Open Philanthropy.

Technocratic funders have, at minimum, the goal to do philanthropy differently, and so they are the most likely candidates to change the way things are done. Furthermore, because this particular group of funders has a penchant for quantification, this proposal to create public stretch prizes at scale just might work.

I applied the ITN Framework below:

Importance

Impact Estimate

If we are interested in awarding a grant, then we should also be interested in matching funds for something that has at least as much impact as the original grant. Here is a back-of-the-envelope calculation of what that impact multiplier might be:

Example grant award: $250K
New public stretch prize: $2.5M
Stretch prize cost in expectation: $275K

Example grant USD per DALY: $1,110.00
Stretch prize performance: $242.00
Anti-malaria performance: $11.00

Improvement over standard grant-making: 4.6x

See spreadsheet for details.

Unquantifiable Impact

If the design of public stretch prizes was compact enough to be applied automatically to every grant, then norms around grant applications could change in general. Those who apply for grants would have to spell out their quantifiable impact to conform to the public stretch prize requirement. If applicants made this change enough times, the norm could spread beyond technocratic funding.

I wouldn’t be surprised if most grant applicants to OP do not contain impact estimates. We could politely ask them to provide such estimates, but without an incentive or rules-based reason, their behavior might not change.

Tractability

Real-World Examples

As an exercise, I went through OP’s recent grants database and created stretch prizes for the first ones I saw:

Farmed Animal Advocacy in Eastern Europe

Original grant:

Open Philanthropy recommended a grant of $100,000 over two years to Alliance for Animals to support farmed animal advocacy, including corporate outreach and undercover investigations, in Eastern Europe.

Example stretch prize:

Award $1,000,000 in 2026 if 90% of Eastern European farms and ranches meet standards for ethical animal welfare. If more than one organization claims credit for the improvement, award the top three claimants with 5x, 2x, and 1x portions of the total prize.

Claimants will be responsible for providing proof that a sample of farms meets the criteria, as well as for demonstrating their contribution. Claimants may collaborate on producing the sample.

Advocacy for Affordable Housing and Land Use Reform

Original grant:

Open Philanthropy recommended a contribution of $500,000 to Affordable Maryland, a new organization working to support the creation of new and more affordable housing in Maryland.

Example stretch prize:

Award $5,000,000 in 2030 if the number of affordable housing units with moved-in tenants in Maryland increases by 30% from 2022 levels.

The rest of the award details, including if there are multiple claimants, should follow the same pattern as above.

Malaria Vaccine Clinical Trial

Initially, I had skipped this grant, thinking maybe a stretch prize wouldn’t apply here, but with some creative adjustment, I came up with one.

Original grant:

Open Philanthropy recommended a grant of $630,123 over three years to the Institut de Recherche en Sciences de la Santé to participate in a phase III clinical trial of a malaria vaccine developed by the Jenner Institute at the University of Oxford.

Example stretch prize:

Award $6,301,230 for the completion by 2032 of a phase III clinical trial of a malaria vaccine that is 95% effective. This prize excludes the above trial by Institut de Recherche, but the institution may apply with a new trial. The prize will be awarded to the first successful claimant.

Possible Pitfalls

Prize philanthropy could be great for spurring innovation, but since innovation’s impact on charity is limited, the gains could be capped.

The simple “tweak-and-launch” model might be not-so-simple, and we could end up repeating all the challenges associated with the rest of prize philanthropy.

Neglectedness

Prize philanthropy has seen a resurgence in the past couple of years. According to Tate Williams (Inside Philanthropy): “It’s starting to feel like there’s a new high-dollar competition launched every week.” Here are a few notable examples of modern prize organizations:

  • 100&Change by the MacArthur Foundation is “a competition for a $100 million grant to fund a single proposal that promises real and measurable progress in solving a critical problem of our time.”

  • Lever for Change helps organizations launch competitions for $10 million-and-up prizes and grants.

  • Challenge.gov aggregates prizes and competitions across federal agencies.

  • InnoCentive is a crowd-sourcing platform for small-prize innovation (around $20,000). Their platform has awarded over $60 million in Innovation Awards.

There is a difference between a grant to create impact and a prize for rewarding impact. Within the latter category, there is an even smaller sub-category for inducement prize contests. Based on a rough understanding of the philanthropy world, prize philanthropy must be a tiny fraction of overall philanthropy. Also, since the resurgence in prize philanthropy is relatively recent, it’s unlikely every angle has been explored.

Technocratic funding organizations, like Open Philanthropy, are uniquely positioned to experiment with philanthropy at the highest levels.

Implementation Details

Here is one possible sequence:

  1. Conduct furthur paper research on prize philanthropy and produce normal distributions for the impact estimates above, using either Squiggle or some other modeling tool. Use the research to criticize public stretch prizes and consider stopping or pivoting the program.

  2. Create a mock prize, including an accompanying website, to get a feel for what would be involved. Ask one pending grant recipient to walk through the paces of quantifying their impact. Then simulate organizations pursuing these prizes and what a claim on a prize might look like. Finally, simulate judging and awarding these prizes.

  3. Bring in consultants from prize philanthropy organizations to analyze our trial and point out potential pitfalls.

  4. Run a parallel prize for one of our grants.

  5. If the launch and initial results of the first prize seem promising, designate one of OP’s programs or cause areas to launch parallel prizes for their new grants over a six-month period.

  6. A year after the first launch, conduct our own impact assessment. For example, check to see if any farms in Eastern Europe have since passed the threshold for ethical treatment. Furthermore, see whether those gains came from new contestants.

  7. Expand the scope of public stretch prizes and consider using a prediction market to generate prizes efficiently.