103 karmaJoined


This is awesome. Would love to hear some examples of how you think it can be used


Summary: The broad concept that Hollis' paper proposes ("outcome-based financing") has already been applied to several other areas such as reducing homelessness, improving specific health outcomes, etc. Recently, McKinsey, Meta, and a few others agreed to spend $925m to fund a similar mechanism to incentivize carbon capture technology innovation. Seems like there's lots of interest in expanding this type of financing model from big funders. Maybe something for the EA community to become more engaged with since there seems to be an appetite.   

More details: As I understand it, Hollis' paper's proposal fits into a broader concept known as "outcome-based financing".  The space is much more developed than I had thought when I wrote this previous comment. Two primary outcome-based financing models exist -- pay-for-success ("PFS") contracts (also known as social impact bonds) and advanced market commitments ("AMCs"). Hollis' paper (from 2004) describes an application of PFS contracts. Both, PFS contracts and AMCs, are already applied to several industries including health and clean energy. 


  1. PFS rewards innovators based on some per unit metric (e.g., QALYs per drug sold in Hollis' example).
  2. AMCs reward innovators in a pre-specified lump-sum fashion (e.g., the WHO, World Bank, a few countries, and the Bill and Melinda Gates Foundation funded a $1.5 billion AMC for entities that could create a vaccine for pneumococcal diseases).

Real-world Examples: 

  1. PFS
    1. Here's a link to Oxford's PFS database (~200 projects//$500m since the concept was formalized in 2010). PFS contracts are used most commonly for reducing prison rates, improving health outcomes (in developed and developing countries), reducing homelessness, and upskilling labor. Check out the database for more details.  
    2. Hollis' org is trying to set up a clean energy PFS fund -- seems promising, but I think doing this in cleantech is an extra tricky. 
    3. I've been engaged with a group that's trying to get funding to do this for a specific pharmaceutical application (see Crowd Funded Cures).
  2. AMCs are less common. However, last week McKinsey, Stripe, Meta, and a few others decided to finance a $925m carbon capture utilization and sequestration AMC.

Seems like there's a lot of momentum for outcome-based financing. Perhaps, the EA community should become more directly engaged in promoting this since it seems tractable. 

Yes, you're correct that each department produces its own CBAs. However, the Office of Management and Budget (OMB) does help standardize these CBAs to some extent. For example, the federal government sets up "interagency working groups" to standardize specific inputs (e.g. the social cost of carbon) across departments (more info on social cost of carbon standardization). And, when OIRA reviews CBAs, they make sure departments are following some general standard processes too. 

  1. 2020 Economists' Comparisons of Welfare Policies: A few economists (Hendren, Sprung-Keyser) aggregated the results of studies that calculate the impact and costs of various welfare policies (US only).  They develop a method to compare the impact of these policies. This paper has a lengthy description of this approach and contains the comparisons themselves (reference Table II. in the appendix for the full list and know that "MVPF" is the primary comparison metric). Hendren and, another economist, Finkelstein provide a less lengthy description here. There's also a great 7-minute explainer video by econimate here (I'd start with this!).
  2. Federal Government CBAs: Additionally, as @evelynciara mentioned earlier, the federal government uses cost-benefit analyses (CBAs) to evaluate regulations. Agencies (e.g. the Department of Energy) conduct these CBAs, which are then reviewed by OIRA (a group housed within OMB). You can find spreadsheets with aggregated results for different regulations here and can explore regulations.gov for specific CBAs.
  3. Other: State and local governments will publish CBAs too (and economists frequently perform CBAs on state/local programs obviously), but I'm not aware of any aggregation of these studies beyond the work by Hendren, Sprung-Keyser, Finkelstein, et al.


MVPFs and CBAs calculate the ratio of benefits-to-costs (or costs-to-benefits). On the other hand, GiveWell generally calculates "$/life saved" and/or "cost-effectiveness in multiples of cash transfers" for each program they analyze. At first glance, the"cost-effectiveness in multiples of cash transfers" metric appears to approximate a ratio of benefits-to-costs. However, this ratio should not be compared to MVPF calculations by Hendren, nor the US CBAs for multiple reasons. One primary reason they shouldn't be compared is that there are issues comparing cost-benefit ratios of programs that affect different income groups (a friend and I made a post about that issue here).

Happy to provide more info if you're interested (for example, I think there are some limitations to the approach used by Hendren/Sprung-Keyser, but overall I think it's awesome). 

These are great suggestions.

Perhaps, instead of the 3 vote suggestions for non-mod edits, there could simply be an option to make a suggested edit (i.e. delete or addition) that only gets implemented if the mod accepts it. Just like the suggested tool on Google docs/word.

Looks interesting! I think you might have some interest in MichaelA's shortform about impact certificates. I saw you mentioned some orgs that are in this space. You may also want to check out Dr. Aidan Hollis' paper, An Efficient Reward System for Pharmaceutical Innovation, and his organization which tries to pay for success, the Health Impact Fund.

The Health Impact Fund (cited above by MichaelA) is an implementation of a broader idea outlined by Dr. Aidan Hollis here: An Efficient Reward System for Pharmaceutical Innovation. Hollis' paper, as I understand it, proposes reforming the patent system such that innovations would be rewarded by government payouts (based on impact metrics, e.g. QALYs) rather than monopoly profit/rent. The Health Impact Fund, an NGO, is meant to work alongside patents (for now) and is intended to prove that the broader concept outlined in the paper can work. 

A friend and I are working on further broadening this proposal outlined by Dr. Hollis. Essentially, I believe this type of innovation incentive could be applied to other areas with easily measurable impact (e.g. energy, clean protein and agricultural innovations via a "carbon emissions saved" metric). 

We'd love to collaborate with anyone else interested (feel free to message me).