A Technological Convergence for Social Impact

Donors large and small are increasingly focused on funding organizations and projects that have a proven impact. However, such social impact funding is not without challenges. Notably, there is often a substantial lag between funding and reporting of key outcomes. Health and education projects may take years, engage thousands of personnel, and involve complex merges of data from many different systems and sources. Projects are often funded on the strength of pilot outcomes, which can be quite challenging to replicate at scale. Due to these challenges, some funders prefer easy-to-verify “hard infrastructure” projects. The Economist says it well: “…aid organizations often define success as creating tangible assets (building schools) as opposed to practical benefits (higher literacy).” 

Blockchain has ushered in a suite of enabling technologies that could redefine outcomes-based funding in the social sector. But aside from a few auctions of NFTs for nonprofits, there isn’t any evidence that blockchain is changing the face of philanthropy as we know it. One compelling application would be to use smart contracts and related technologies to anchor the dispersal of funds to key milestones as they are reached. Online learning could provide an early real-world demonstration of this concept.

How it works

A diagram depicting smart contracts for online learning.
1. The funder and trainer create a smart contract.
2. Key "milestone" metrics are routinely reported to the contract by an API.
3. The smart contract automatically pays out when milestones are met.

To make this concrete, consider an example involving literacy rates. English literacy is something that can be measured with high-quality, validated assessments. Let’s say that a foundation wants to fund Blingual, a (fictional) English learning app. They agree to pay Blingual for every learner who improves their English literacy by a certain amount. Blingual implements a literacy assessment to measure learner progress over time and builds an API that can report de-identified assessment data to the contract. All of this work needs to be done before the smart contract is created. Once the contract is in place, funds are automatically dispersed for each unit of improvement achieved.

What’s new here?

It’s pretty obvious how this differs from traditional outcomes-based funding, but here are some of the key differences.

  • Data transparency: In order for this to work, the training organization has to agree to routinely disclose de-identified data to the funder via an API. The data disclosed are the “milestone” metrics generated by the learning platform; these metrics are used to determine whether the contract conditions have been met.
  • Minimal latency: Data disclosure is real-time (updating, say, every minute or every hour), substantially shortening the feedback cycle and increasing the intelligence that funders have on the performance of their portfolio.
  • Fairness: The way these contracts work, once set in motion, they are immutable. Everything is established in advance, so both parties know exactly what they are agreeing to before the work has even begun. 

Let’s be clear: this overall concept is not new. It’s not even clear to me that blockchain per se is required to do this; blockchain just provides much of the technology to make this possible.

Why online learning?

  • Data: Online learning generates a lot of data; typically that data is more or less in one place (a learning management system). All that’s needed is some infrastructure to report relevant milestone metrics to the smart contract. It’s informative to compare this to other situations (medical records in the US, for example) in which data is held and managed by many different entities. 
  • Simplicity: In online learning, data collection is intrinsically tied to the learning process, so it’s relatively simple to collect and report milestone metrics. By contrast, many international development projects mostly exist in the “real world” rather than online. These projects require many personnel who need to enter a lot of data. This adds complexity, time, and cost. 

What could go wrong?

  • Technical issues: Smart contracts aren’t entirely self-contained; they are made by people and require data from the real world. If the milestones API reports inaccurate data to the contract, that’s a problem. This risk could be mitigated with technical measures and auditing procedures. For example, while the learning platform reports aggregate metrics via the milestones API to the contract, it might also provide raw data to an independent auditor who computes the same metrics and reports them to the contract. In the event of any discrepancy, the funds aren’t dispersed.
  • Misuse or fraud: With a strong financial incentive to report positive outcomes, there is a real risk of biased reporting or outright fraud. The previously-mentioned audit mechanism would help prevent this, but there would also need to be “real world” agreements to accompany the smart contract and provide procedures for mediating and resolving disputes.
  • Data privacy concerns: The outcomes of interest are anchored to students’ interactions with the learning platform, and those outcomes need to be reported to the smart contract, potentially exposing student data. Everything needs to be done in a way that protects students’ right to privacy.

What are the barriers?

It could take a fair amount of up-front funding to ensure that all of the pieces come together. There are technical problems to solve (building the milestones API, selecting and implementing learning assessments, setting up the smart contract) as well as theoretical challenges (building out a legal framework that enables the work). The “outcomes” measured, at least initially, would probably be fairly modest. Donors might not want to pay for learning when they are really interested in downstream outcomes, such as better jobs for low-wage workers. Funders need to be committed to the long game and recognize that this proof of concept is a stepping stone to a whole new funding paradigm.

Closing thoughts

This is just one example of automated outcomes-based funding in action. There are plenty of potential applications that go beyond online learning. If you’re one of the (many) online learning skeptics out there, consider how this concept might apply to other initiatives, like carbon capture and storage or hypertension control.

Thoughts or comments are welcome! contact@enactacademy.org