Building Blocks: an analysis

From time to time we want to study a project in-depth and analyze its positives and negatives, by going through its development, the partners, and our analysis framework. Such examination will usually be accompanied by an interview with one or more people significantly involved in the project; we have contacted WFP for Building Blocks, but without success. Some of our planned questions are included throughout, where relevant. 

For a quick rundown of Building Blocks, see our Dossier.

Building Blocks0 was bound to make waves from the beginning: not just another good-intentioned project from the tech world that could not withstand the intricacies of implementation in a real humanitarian context, but one born from within the humanitarian sphere, something still fairly rare at the time of writing. The results are quite positive and we would love to examine them more in detail!

Implementers

Who are the teams and organizations behind the project? What is their blockchain-related experience, or humanitarian action/development experience?

WFP needs no introduction: one of the largest humanitarian organizations in terms of budget (6.8 billion USD in 2017), working in over 80 countries, providing food assistance to populations in crisis.
As for the “blockchain credentials” of the project, they rest on the shoulders of Parity Technologies and Datarella: Parity Technologies is a well-respected company and the creator of one of the most widely used Ethereum clients (a client is simply the software that makes the Ethereum network function); Datarella was, back in 2016, a relatively unknown company that started out focusing on Quantified Self (the moniker for representing a person’s life and activities through personal collection of data), only to change focus to blockchain during 20151. Aside from WFP, it is also collaborating with UK’s Department for International Development (DFID). According to their website, they developed “a digital asset management and accounting system to distribute aid in kind as well as in cash directly to people in need. The system, called ‘WFP Building Blocks’ runs on a permissioned, private Ethereum network based on Parity.2

The project started from an idea by Houman Haddad (then Finance Officer at WFP, currently Middle East & North Africa Regional Cash Based Transfer Advisor), expanded thanks to funding from the WFP Innovation Accelerator, and was refined with Alexandra Alden during a Singularity University bootcamp3. A proof of concept was conducted in January 2017 in the Sindh province of Pakistan, though not much4 can be found about the lessons learned from it. Still, WFP was happy enough with the results to start a pilot project in Jordan. “As of October 2018, more than 100,000 people residing in camps redeem their WFP-provided assistance through the blockchain-based system. Thanks to the technology, WFP has a full, in-house record of every transaction that occurs at that retailer, ensuring greater security and privacy for the Syrian refugees. It also allows for improved reconciliation and significant reduction of transaction fees0.

Description

Main characteristics including phase, scale and scope of the project: the stage of implementation; the extent of the project in term of numbers (scale) and thematic areas (scope); geographical area of implementation.

Building Blocks is an accounting system put in place to monitor and record cash-based assistance to refugees, so that they can purchase goods at select  stores. This does not, at present, involve cryptocurrency.

WFP allocates a set amount of entitlements (approximately 30 USD) to accounts associated with the biometric identities of refugees. When refugees purchase goods, their iris is scanned, the system verifies that the account has sufficient funds, and the operation is authorized. All transactions are registered on the blockchain and accounts can be immediately consolidated. WFP regularly transfers resources (at present through normal banking) to the stores involved in the project5.

After an initial pilot in Pakistan (10,500 people, on the public Ethereum blockchain) and a more extensive pilot in Jordan (initially 100,000, then projected up to 500,000 people, on a private blockchain), Building Blocks seems to be well on the scale-up phase. Recently funded with 2 million EUR by Belgium (together with other innovations)6, the project is likely to be extended to other WFP settings, and beyond – a collaboration with UN Women is already expanding its scope to more refugee camps in Jordan7.

Relevance

What problem does the project aim to solve? Is it a “fundamental” issue (for example access to clean water or food security)? Is the project going to improve a wide range of markers, or does success mean minor improvements at best? How strategic is the project’s approach (is it the best way to try and solve this particular problem)?

Building Blocks aims to reduce and potentially eliminate transaction fees linked to account reconciliation for cash-based transfers (CBT), along with allowing for faster reconciliation of accounts. This can not be classified as a fundamental issue (like guaranteeing rights, or initiatives of similar import), but it has important ramification in its liberation of resources for other uses. 

With that in mind, given that the system basically eliminates the reconciling function of the intermediary (and the related fees), this is a sound approach, and we are not aware of better options to achieve the same goal. This project is the first of its kind in a humanitarian setting, and its relevance stems from opening the ground for similar and more sophisticated ones in the future.

Effectiveness

What results have been achieved so far (if any)?

According to WFP, transaction fees are reduced by 98% – which can only be considered an extremely positive outcome. No information is available on the speed of reconciliation in the previous voucher-based system, and it is therefore not possible to estimate the effectiveness of this aspect. A general evaluation would only be possible with access to data from both the project superseded by Building Blocks, and Building Blocks itself, which at the time of writing is unfortunately not available.

Impact

What kind of impact is the project going to have? What is (if any) the potential social impact of the solution proposed (for example on gender inequality, on reducing mortality, on enjoyment of fundamental rights)? Does it have an economic impact?

Transitioning the existing system to use blockchain would not, in theory, have any direct social impact – the end-user sees no change, and therefore any impact comes from cash programming itself, not the technology used to implement it. Indirectly, however, there is a potential impact in freeing more resources: as they don’t need to be used for various kinds of fees, they can be used for more CBT, benefiting more people.

As will be the case throughout, any more in-depth considerations (for instance to determine possible gender-related consequences) would need more information on how cash-based transfers are assigned and in general on how the project is designed.

Some of the questions we originally planned to ask WFP to better understand this project and to discuss its merits and challenges are included in the red boxes below.

Q> What proportion of an average household’s income does Building Blocks represent? Does this push households to buy only in select supermarkets and shops?

Q> How are supermarkets/shops selected to be part of the program (and thus receive IrisGuard machinery)? Are there any costs associated with IrisGuard, and do the supermarkets or shops sustain any of these?

Q> One can think of possible negative consequences, for example attempts at price gouging by stores that are part of the project, and a loss of sales for shops not selected to participate. What measures did WFP take to mitigate any potential negative impact? In other words, how have you included the Do No Harm approach in the project?

Who are the winners and losers if this solution is implemented?

Among the “winners”, the most important are certainly refugees: anything that makes CBT easier is a plus as it furthers its use; CBT is generally considered better and a more dignified solution than in-kind assistance.

Another winner is WFP, which pays less transaction fees; lastly, donors are also winners, as their money is arguably better spent as a consequence of less overhead.

Among the “losers”, we can count potential WFP implementing partners that would have managed the cash transfer programme, because they are unlikely to be able to set up a blockchain-based solution at this stage.

To the extent that only bigger/more formal stores are engaged in this project, this will push refugees to buy in those stores rather than in small shops, thus affecting the owners of these shops and reducing their opportunity for sales, and conversely enriching the owners of the major supermarkets – this aspect is however linked mostly to the CBT aspect of Building Blocks, rather than the use of blockchain as such.

Local banks and other Financial Services Providers can also be counted among those who do not benefit from the project, since their responsibilities (and gains) are reduced – WFP is not moving to get rid of their help, however: “while the system will not necessary replace Financial Service Providers (FSPs), it can reduce the cost of their services significantly by alleviating unnecessary costs through more efficient use of their unique value propositions in a win-win arrangement. Rapid and low-cost deployment after crises: In crisis situations where WFP does not have agreements with FSPs, blockchain-supported CBT is a rapid and secure way to ensure assistance reaches beneficiaries in their time of need” 8.

Q> What costs and responsibilities (direct and indirect) does WFP assume compared to before, and what are the implications?

Efficiency and value for money

If the project is concerned with the use of a resource, what improvement(s) does it bring in the use of resources?

According to information shared publicly by WFP5, Building Blocks resulted in a reduction of transaction fees by 98% – which was expected to amount to around 150,000 USD per month once the project was scaled up to the whole Syrian refugee population in Jordan 9 (it would be interesting to see the final outcome, now that the project is moving in that direction), and it is interesting that Datarella mentions a much larger figure in their homepage (see note below). While this sounds less impressive than the  widely cited “98% reduction in fees”, one can hardly fault WFP and the implementing partners for stressing the relative cost saving, as they have every reason to celebrate the outcome. Still, it is important to keep the savings in perspective to the total operational cost.

As no information can be found regarding the cost of setting up and maintaining the system, and its interoperability with traditional WFP systems, it is not possible to make further considerations on value for money and efficiency.

Q> Datarella mention on their website2 that “The estimated cost savings cut for the UN is 3.5m USD per year in Jordan alone.” – to date, how much has Building Blocks actually saved in absolute terms and as a percentage of total project costs?

Feasibility

Familiarity: is the proposed solution accessible? What about ease of use?

Haddad and team made a point of not changing anything in the process for end-users, so as to not disrupt the delivery of aid in case the Building Blocks failed. The experience is in fact very similar or even identical to the previous one, which makes it very familiar and easily adopted.

Q> Given that the user experience did not change after Building Blocks was implemented, did you explain the change to blockchain, data security and privacy implications to people participating in the project? If so, how did you explain these aspects? What questions did they ask, and what were their preoccupations? 

Sustainability and maintainability: how is the solution maintained? What are the ongoing costs, and who will shoulder them? How is the eventuality of a handoff dealt with (e.g. to actors not involved in the creation of the system but that will need to maintain it going forward). What are the long term implications for costs, operation and maintenance, ownership and so on?

At the moment WFP operates a private blockchain, and it recently started expanding control of the nodes to UN Women during a collaboration in two other refugee camps in Jordan. This reduces considerably the decentralized and trustless aspects of blockchain. Criticism has been leveled at WFP for this10 11, which we find rather misguided: the process of technology adoption has risks, and companies can’t simply turn a knob and switch to a completely new approach. Sane and non-disruptive uptake of a new technology, especially one requiring a different mindset such as blockchain, starts with a very controlled and limited test-run (a sentiment echoed by both Haddad12 and Robert Opp13, Director of Innovation and Change Management at WFP).

Having said that, the criticism is by no means invalid: if the infrastructural burden of Building Blocks does not end up being sustained by a number of different organizations, the project could rightly be criticized as having failed to deliver on the promise of blockchain. Still, as long as there is a quantifiable improvement for users, we doubt most would care. The inclusion of UN Women in a further iteration of the project might also expand control to other organizations and start operating in a more decentralized fashion.

Turning to costs, there is no information available on the operating ones, but we can assume that besides the cost of the technical expertise to set up and maintain the system, there are no other major costs specifically related to Building Blocks’s use of blockchain technology (other costs, like the funds for CBT, are not considered as they are the same with or without blockchain).

Regarding ownership, it would not make much sense to transfer this particular project to an external entity, as it is directly related to aid distribution from humanitarian organizations. However, more in general social protection schemes for vulnerable populations could be initially set up by humanitarian or development organization and later be handed over to national governments, in which case considerations of transfer and ownership are necessary from the start. 

What are, if any, the funding mechanisms of the solution? Is it self-sustaining, or do upkeep costs need to be covered by a well-funded entity?

The project seems to be fully funded by WFP, and is not self-sustaining. Participating organizations like Datarella and Parity might (in theory) provide support for free as part of their socially responsible work, but there is no indication whether this is the case or not in this particular instance.

Applicability to a humanitarian and/or development context: a context in which there are potentially scarce infrastructure, limited resources, low access to technology, or other contextual issues etc.

Fully applicable, as the project is born in a humanitarian context. It would be great to have more information about the reasons that prompted the change from the SMS/low tech option that was used in the proof of concept in Pakistan, as it’s an option that could be interesting to evaluate for low(er) tech settings than Syrian refugee camps – as the Syrian refugee crisis is one of the most visible humanitarian crises, funding (and technology) are more widely available than in other contexts.14

Scalability and potential for adoption: is the proposed solution likely to be adopted by other organizations in the humanitarian / development system?

No information can be found on potential scalability issues – and as the underlying consensus algorithm is Proof of Authority, which is inherently more perfomant than others when it comes to processed transactions per second, Building Blocks seems scalable and it has indeed been expanded.

If the project is extended to other organizations, areas, and countries there will be a need to train people on running and maintaining the system. Another possible issue is the lack of available biometric identification (a likely situation in other settings): there might be difficulties in linking this configuration with other ID systems. Other projects have used QR codes and e-vouchers with positive results.

The potential for adoption rests on two factors: first, if WFP allows other organizations to replicate  the system, and secondly, if other organizations join the existing deployment (as UN Women is doing), something WFP seems more interested in.

Q> Regarding WFP’s current collaboration with UN Women, and in view potential future expansion to other humanitarian or development organizations, what challenges do you see to a joint/common adoption of blockchain solutions?

Other key considerations

How is security set up and ensured in the system?

Aside from the obvious, that the system is under WFP’s control (and now partially UN Women’s), to our knowledge no details are available publicly about how this has been addressed.

Q> what aspects of information security are most relevant to the implementation of blockchain in humanitarian action and development? What aspects did you consider for Building Blocks, and what solutions did you find (or what reflections did you make)?

How is privacy guaranteed?

This is an important topic when using blockchain-based systems in a humanitarian context – the implications of having your identity recorded as a refugee in perpetuity are quite far-reaching, and other projects have been vehemently criticised15 for it.

We could find no information about the team’s approach to the issues of users privacy and informed consent.

Q> What data did WFP register on the Ethereum blockchain during the Pakistan proof of concept? We assume that each user had access to a wallet with their funds, was that the case? How was user identity determined in relation to their address?

Q> The ownership of users data, especially as it relates to having a permanent record of it, is a topic rife with ethical issues. What aspects of this did WFP consider, as they relate to the past, current, and also future of Building Blocks?

Q> There have been multiple instances of indirectly identifying an address thanks to related activity, whether on- or off-chain. To what extent are users aware that they purchase history could be publicly recorded, forever (even though in Building Blocks it might not be the case)? 

Q> You have mentioned in passing that you would like Building Blocks to be the first step towards a Self-Sovereign Identity solution. Could you share more about this vision, the steps and the timeline you’re thinking of?

Q> If it ever comes to SSI, users data will need to be transferred to them. Are you already planning how to deal with this portability issue?

To what extent have people been consulted and involved in the design of the project, its components, its solutions? And if they have been involved, who are they (women, men; refugees, local shop owners, others)?

No data is available on this topic, though it would be quite interesting to hear the users perspectives in relation to the questions on privacy and informed consent.

It has to be stressed that, reportedly, the long term vision for Building Blocks sees it as a jumping-off point for exploring digital identities for refugees – entering an innovative and far-reaching prospect, but also a minefield of ethical issues. While the current consensus algorithm allows more flexibility when dealing with bad or undesirable data and human error, the implications should be fully addressed before steps are taken to make that vision a reality.

Conclusions

Reflections on all of the above and further considerations on the merits, potential and challenges of the project.

Even though it might not be a fully-realized implementation of blockchain-based systems (and be used “as a regular database” for some time before further expansion), Building Blocks is one of the first and certainly the most advanced blockchain-based solution deployed in a humanitarian setting.  It is a promising approach, that uses one of the key aspects of blockchain – the ledger – effectively to cut costs. Its implementation has shied away from using a public blockchain after the proof of concept – which is probably the most problematic aspect, on the technological side. On the humanitarian side, the financial results are definitely positive, but many potential challenges of using blockchain as the basis for the system remain (at least publicly) unaddressed. More information is needed and an exploration of all relevant data would be useful to get a better understanding of advantages and shortcomings.

The greatest potential we see in this pilot is the integration with UNHCR and UN Women, meaning that if organizations are willing there can be a coordinated approach that could greatly expand potential uses and implications. This experiment also demonstrates that blockchain can be  cautiously integrated with existing systems, a fundamental requirement if the notion of using the technology across a wider variety of contexts and use cases is to be entertained. Additionally, more openness on challenges would be very welcome and useful.

Building Blocks might not be a particularly daring approach, or the revolution some were hoping for, but its cautious approach and the improvements it delivered on existing processes are there for all to see, and as a safe first test it will open the way to others. We look forward to all future developments, starting from the results of the collaboration with UN Women!

Sources

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.