LEDGER GOOD DEEDS SHINE: USING BLOCKCHAIN TO MEASURE, RECORD AND PREDICT SOCIAL IMPACT

1 RECORDING EXISTING SOCIAL IMPACT DATA

The most immediate application would obviously be to use blockchain technology to record existing data on social impact, outcomes etc. This would basically take the approach pioneered by organisations like 360 Giving and push it further. Why, you might ask, would you want to use blockchain to do this? Well, partly because it is cool (obviously…), but also (and more soberly) because it brings relevant benefits in terms of being distributed (and hence can enable the reconciliation of information entered by multiple parties), providing a shared infrastructure (and thus far greater interoperability), being immutable (as a result of timestamping and the increasing security of blocks over time), being secure (as there is no single, central point of attack), and being public (which may be good in itself, but also makes third party innovation that much easier).

2 HUMAN RECORDING OF FUTURE SOCIAL IMPACT DATA

Once we have established a blockchain-based infrastructure for recording existing social data will that also be sufficient to record all such data in the future? I suspect not. The existing methods of measurement and capture reflect existing systems and modes of governance; and as such are centralised and rely on expert organisations (in the form of charities and NGOs) to do the job. However, blockchain is an inherently decentralising technology, so we must consider whether its application in this context would also be likely to result in social impact measurement becoming decentralised and democratised.

What would this mean in practice? Well, I have written before at some length (some might even say excessive length) about the possibility that charitable organisations themselves could become decentralised. The adoption of new blockchain-based governance models like Distributed Autonomous Organisations (DAOs) could enable networks of individuals to coordinate and act together to address social and environmental problems without the need for a centralised structure. In this scenario, responsibility for measuring and recording impact would also have to be distributed in some way among the nodes that make up the network.

“But hold on”, I hear some of you saying, “are we supposed to just let everyone make up their own social impact data and enter it on the blockchain? Isn’t that going to be a total car crash?” Well, probably yes, if we assume that individuals and systems operate in the same way as they do now and we just open it up into a decentralised free-for-all. That’s why we might have to think a bit smarter.

One possibility (which brings things back a bit closer to where they are now) is to create levels of status within our distributed network; so that some nodes are accorded a higher degree of trustworthiness. These “oracles” would then play a key role when it comes to verifying that information recorded on the blockchain accurately reflects facts in the real world. Given that charitable organisations (and the people who work in them) are likely to have most of the experience when it comes to measuring social impact (at least for now), they would seem to be in a prime position to act as oracles within social purpose DAOs.

New incentive mechanisms

In the longer term, however, would it be possible to get away from the need to accord individual nodes within the network privileged status on the basis of their prior reputation or experience? The answer might well be yes, if we can create new incentive mechanisms that reward those who consistently record data accurately and punish those who do not. These mechanisms might be financial, but they wouldn’t have to be. One thing that should be obvious if you cast even a passing glimpse at what is currently going on with the boom in ICOs and token sales is that you can use blockchain technology to create tokens to represent pretty much anything. (NB: this is putting aside concerns about where the current ICO craze is going which I covered in ICOs, crypto tokens, Swiss foundations and philanthropy and Crypto tokens, ICOs and blockchain philanthropy).

Interestingly, the use of crypto-tokens to reward truthful reporting of facts is something that has been mooted in the context of the current battle against fake news; and this reads across fairly neatly to the idea of incentivising accurate social impact reporting.

How would tokens work?

The question in the minds of perceptive readers (which I’m sure you all are), is probably: how exactly would these tokens work then? As I mentioned above, they need not be straightforwardly financial, and I would personally prefer that they weren’t. I don’t have entirely concrete reasons for this, but it seems as though the potential for creating perverse incentives would be significant, and also the idea that any incentives for socially positive behaviour necessarily have to collapse into financial ones is just quite depressing. Unfortunately, I don’t have the answer to the question of how you could use crypto-token systems to incentivise the accurate reporting of social impact. I am aware of some interesting projects starting to look at this, but I think there is still a lot more to be done in this area (not least because I think it is one of the most fascinating questions about the application of blockchain to philanthropy!)

One potentially fruitful avenue might be Mechanism Design Theory. This is an area of economics closely aligned to Game Theory; the development of which led to Eric Maskin, Roger Myerson and Leonid Hurwicz being awarded the 2007 Nobel Prize for Economics (or, as my former colleague Adam doggedly points out, not actually the Nobel Prize, but just a prize given at the same time…) Mechanism Design Theory is in some ways the mirror image of game theory. Whereas game theory considers is concerned with finding optimal strategies for success within the framework of systems where the rules are already set, mechanism design theory is concerned with how you structure systems in the first place to maximise the attainment of certain outcomes or the allocation of scarce resources. A number of people have pointed out that this seems unbelievably germane to the use of crypto-tokens as incentives, and that maybe economists who understand mechanism design should be playing a bigger part in shaping our understanding of how this can work most effectively. I would love to see someone do this for the specific case of incentives for altruistic behaviour or social impact. So if there are any economists out there already doing it, let me know. And if not, consider the gauntlet thrown down!

3 AUTOMATED RECORDING OF SOCIAL IMPACT DATA

Just briefly, before we come back and develop the idea above, what if there was a way of bypassing the need for human involvement altogether? That would immediately remove the need for either oracles or incentive systems. So is it possible? Well, the answer in the medium term is likely to be “yes”, if the Internet of Things (IoT) continues to expand in the way most experts predict. (For anyone new to these parts, the IoT is the system of ‘smart objects’ that are connected and contain sensors so that they can monitor and adapt to their own environment and communicate information to each other).

Blockchain is often spoken of in connection with IoT (and usually AI as well), because it seems likely to provide the missing piece in terms of the infrastructure that will allow transactions between smart objects to take place at scale. If these objects are connected and represented on the same blockchain that underpins a social purpose DAO, then they could record things like health data or information on environmental conditions directly, without any human members of the DAO needing to be involved. Clearly this would not work for all instances. It is also arguable that the sorts of data recordable in this way would be always be outputs, and thus at best proxies for actual data on outcomes. However, it is easy to see that the convergence between IoT and blockchain could have a significant impact in terms of automating the collection of large volumes of relevant data.

4 PREDICTING SOCIAL IMPACT

Coming back now to human involvement, can we take the idea suggested above – of creating mechanisms that incentivise accurate reporting of impact data – one step further? What if, instead of rewarding people for capturing the past effectively, we could reward them for predicting the future effectively too? This is something that already exists in many fields in the form of prediction markets. And whilst these are not reliant on blockchain, there has been a lot of interest in them in this context because some of the features of blockchains (e.g. their ability to record any kind of asset, their trustlessness, automation via smart contract) means that they could make a far wider variety of prediction markets far more practical and efficient in the future. We are already seeing this develop in the shape of blockchain-based platforms for decentralisd prediction markets like Augur and Gnosis.

So could we create prediction markets for impact? The idea is that people would be able to choose an outcome (or set of outcomes) and then “bet” on how likely they think a given approach or intervention is to deliver that outcome. There would then be a mechanism (possibly financial, or possibly based on some non-financial crypto-token) to reward those whose predictions turned out to be accurate. So, over time, those who proved to be better at identifying the most effective ways of delivering outcomes would “win bigger”. This would mean both that they would have more to bet on future predictions, and that others would also certainly start to follow their predictions (much as people do with successful gamblers or financial investors).

Futarchy

The economist Robin Hansen has explored the application of prediction markets in similar contexts in his work on the idea of a “Futarchy”. His focus has been on using such markets to replace existing models of democratic governance, but his exploration of how they could be relevant to public policy choices is pretty close to what we are talking about. As Hansen notes:

"Policy disputes arise at all scales of governance: in clubs, non-profits, firms, nations, and alliances of nations. Both the means and ends of policy are disputed. While many, perhaps most, policy disputes arise from conflicting ends, important disputes also arise from differing beliefs on how to achieve shared ends."

He suggests that by using prediction markets we could improve the process of selecting the most effective public policies, once we had decided what our goals were and elected representatives who shared those goals. (He catchily sums this up as “vote on values, bet on beliefs”). Hansen makes a number of theoretical arguments in favour of a prediction market model, but is also careful to point out that we needn’t rely on theoretical justification because:“the main reason to believe in speculative market accuracy…is the robust and consistent empirical track record.”  i.e. You don’t need to have any ideological commitment to the idea of prediction markets as a way of addressing information failure, because the available evidence is just that they do work.

He is also clear that his idea is not the same as the “wisdom of crowds”, and that there is still quite a lot of inherent value in expertise:

“Speculators tend to rely more on crowds when crowds know more, and on experts when experts know more. Yes, the best track bettors have no higher IQ, but speculative markets if anything still over-emphasize experts, both public and private….Instead, the main reasons for superior speculative market accuracy seem to be incentive and selection effects: stronger accuracy incentives tend to reduce cognitive biases, and those who think they know more tend to trade more, and specialists are paid to eliminate any biases they can find.”

This should be reassuring (in one sense) to existing charitable organisations when it comes to considering the possibility of social impact prediction markets, as they would surely have an advantage as a result of their knowledge of what works in terms of delivering social outcomes (wouldn’t they…?) I suppose the really interesting thing is that this model would provide a concrete test of that supposition, by gauging exactly how much confidence each participant actually had in their own “theory of change”.

This gives me amazing visions of a huge (metaphorical) poker table, around which a range of philanthropic funders are sitting and assessing how strong their own hand is to gauge whether they think they can win the big social impact pot in the middle. There are some high-rolling foundations dotted around the table as well as a mix of innovative smaller charities, philanthropists, individual social entrepreneurs, corporates and social purpose start-ups. So whose idea about how to solve the problem at hand is going to prove to be most effective, and ho much are they willing to stake on their belief in it? Ooh, look: the Effective Altruist on the end of the table has just gone all in on her hand- but is she bluffing…?

There may be those who recoil in horror at the picture I’m painting, either because of an inherent distaste for the idea of using incentives in this context, or because they are uncomfortable with the analogies with gambling (although it has to be said that gambling has a pretty strong heritage as fertile ground for economic thinking!) However, I think this is another one of those seemingly far-out ideas that we actually need to pay some attention to. Not only would the system be far more reassuringly above-board in practice than my poker analogy might suggest, but also it seems certain that if prediction markets do start to be adopted in other fields as a means of democratising and rationalising decision-making over how to deliver desired goals (which is definitely possible), then it will be increasingly likely that they will be used in the context of public goods or charitable outcomes as well. I can certainly see how many philanthropists and funders would be very attracted to idea of mechanisms that effectively identified the highest-impact approaches, even if that was highly disruptive to the status quo.

So I guess we should all ante up and decide how much we want to stake on our own theory of change…