We are living in a world where no digital initiative operates in isolation – and digital is increasingly influencing every part of our everyday lives, including the economic, political, and social. With such widespread impact, how do we ensure digital initiatives do no harm? And importantly, how do we anticipate and mitigate harms that may seem outside of our control?
At its best, the digital public infrastructure (DPI) approach is one that maximizes public value creation, using design principles that help drive the public and private sectors to deliver new services and benefits for people, while prioritizing the rights and aspiration of all people. At the same time, DPI is an approach to implementing specific digital systems – including digital ID, payments, and data exchange – each of which needs to be designed, deployed, and governed with people’s needs in mind in order to achieve this promise.
However, these systems are not always implemented as part of a people-centric DPI approach. In many countries, digital transformation is largely motivated by the desire to improve government efficiency. When efficiency is the priority, it is tempting to focus on metrics such as how much money is saved, how much revenue is raised, and how many services have been brought online. In this context, it is easy to overlook the experience of citizens being mandated to use these platforms.
To improve this situation, we need better quantitative metrics, and, crucially, we need to speak with actual users of these services to understand their experience. It is only with this knowledge that we can start to iterate and improve upon digital systems, and to turn DPI into something truly transformative – rather than a new term destined to repeat the mistakes of the past.
To share some of these stories and to discuss the insights coming from the everyday impacts of DPI, we invited 3 speakers to speak to our Digital Donors Exchange community:
- Nanjira Sambuli, Kenyan researcher, policy analyst, and strategist studying the unfolding gendered impacts of digitalization/ICT adoption.
- Teki Akuetteh, Founder & Executive Director of Africa Digital Rights Hub and former Executive Director of Ghana’s Data Protection Commission.
- Mansi Kedia, Research Fellow at the Indian Council for Research on International Economic Relations (ICRIER).
Stories of real people interreacting with payments, data exchange, and digital ID systems exposed common challenges.
Each speaker kicked off the discussion covering their research on people’s lived experience with DPI in Kenya, Ghana, and India respectively. Though their research spanned different political, cultural, and economic contexts, the group identified common themes among their findings:
People may not trust their governments or digital service providers.
Lack of trust is a pervasive issue: people often do not trust their government due to past mismanagement, lack of transparency, or corruption. In her research, Nanjira and her colleagues spoke with people whose experiences with Kenya’s digital ID efforts have been largely disappointing and even harmful. These people reported a lack of trust in Kenya’s new digital ID, a consequence which comes on the heels of multiple failed attempts to institute digital ID – none of which were properly tested, iterated, or designed with people. These previous failures, compounded with other poor experiences with government service delivery, create low public confidence that the purported digital ID benefits will materialize.
Unfortunately, the implementation of digital systems can also mean the replication of existing harms. In the example of Kenya, the existing processes for acquiring critical documentation like national ID or a birth certificate present challenges that still are not addressed by the new digital ID. This includes the need to bribe corrupt officials and a discriminatory ID vetting process that continues to marginalize vulnerable communities. From the discussion, it was clear that dwindling trust is a major risk and a prime determinant of how effective DPI will be in achieving its ideals.
People are not always informed enough to truly consent.
When digital systems are being rolled out to communities for the first time, it cannot be assumed that people fully understand what they agree to when they register. In her work as a DIAL fellow, Teki examined how female traders in Ghana navigate digital systems. Among the women she spoke to, several who had been issued national IDs did not understand that their biometric data was being collected in the process. There are, in fact, no direct translations of the words needed to describe biometric data in the local languages of Ghana.
These women were used to having their photographs or fingerprints taken as part of standard bureaucratic processes, and thus they did not understand that the digital ID was different and that data would be stored permanently. Similarly, these women were not aware of their data protection rights and did not understand where to go if they needed to correct their information.
This lack of understanding of how digital systems work was also cited by those interviewed in research into how rural communities in India are experiencing India’s digital payments system – Unified Payments Interface (UPI). Mansi and her team found that many users who had signed up for UPI did not understand what to do if, for example, they sent money to the wrong account. Without a robust understanding of how to use these tools, users are at risk of becoming victims of fraud or harm.
In these communities, digital tools are being rolled out using the same processes used in more urban, formalized environments, while many would-be users belong to the informal sector, with little or no formal education. Unfortunately, user research suggests that processes are not yet being adapted to meet the unique needs of people in these very different contexts.
Women face unique challenges, which are further exacerbated by challenges related to geography and income.
The group acknowledged that even with the growing awareness of gender divides and efforts to build more gender inclusive frameworks into policy positions, direct implementation of digital systems does not always consider the unique needs of women in on-boarding, accessing, and using digital services.
For example, a group of women in Ghana who primarily work in the informal economy, shared that the registration process for GhanaCard – mandatory for registering SIM cards and accessing mobile money services – was complex and time-consuming. As digital payments play a key role in how they earn their monthly income, such barriers to access mean they are excluded from engaging effectively in the economy.
In India, research showed that UPI usage is also gendered. Even when women had access to smartphones, and these women were literate, the use of digital payments and UPI was being driven by the male member of the household. Therefore, in this case, digital payments are expanding the existing gaps in social agency – where financial decisions, and those related to payments, are largely driven at the household level and not at an individual level.
Reflecting on these stories, the participants discussed how the design, delivery, and governance of DPI is at risk of overlooking the many intersecting layers of challenges that, for example, women in rural areas with low income and education levels are likely to face. This hinders the ability to deliver on its promise of positive impact at population-scale.
The financial burden of digital transformation is often passed onto individuals.
When it comes to taxpayer funded digital transformation in low-resource environments, the cost of digitizing can be steep.
When government incentives are driven mainly by efficiency, there is pressure to ensure that the cost of digital services is sustainable over time. This is a fair question, particularly in African countries, many of which are straddled with debt burdens that are higher than the foreign assistance they receive. At the same time, decisions about financial sustainability risk passing the cost of services onto individuals, who are often mandated to use these systems to access essential services. Costs may be passed on to the consumer via taxes, mandatory transaction fees, mobile data costs, or hidden price increases.
Some of the individuals interviewed in Kenya and in Ghana reported feeling the weight of the increased and mandatory costs, while also expressing frustration that it is unclear how the revenue from these fees is being used. Other interviewees, particularly those in India, where the cost of government services such as UPI has been kept low or even free up until now, reported their anxiety in knowing that this is not guaranteed to last. One participant in our discussion pointed out that the official policy for the citizen wallet DigiLocker reserves the right to charge for services in the future.
We discussed several ways to anticipate and mitigate harms – and these are only a sample of strategies to improve outcomes.
1. Agreement on the definition of DPI to make sure we are collectively taking the right approach
The consensus among the diverse group of implementers and funders during the session was that DPI cannot be a repackaging of old ideas about digital transformation. Digitalizing identity, payments, or other government services is not necessarily DPI and is just as subject to issues like corruption, inflated costs, and waste in government. The DPI approach is one that must include accountability through legislative engagement and oversight, strong judicial review, and even citizen pressure directly or through civil society. This enabling environment is just as important – if not more so – as any of the technologies themselves, reducing the risk of misuse and to achieve widespread social benefits.
Learn more about the components of good DPI in our recent article: What is Good Digital Public Infrastructure?
2. Outcomes-based policies and mandates to prioritize outcomes throughout implementation and continue to iterate when not meeting these standards
The group discussed the importance of repositioning DPI policies and statements to be more outcome-based. Using outcome-based policy statements and positions could enable DPI implementers to ask the important questions upfront: What are the outcomes that we are trying to achieve based on the specific issues within our environment? What are the outcomes we are expecting for specific demographics? These outcomes then become an obligation, whether by way of the policy, position, or the law. Using this outcomes-based frame could help governments, implementers, and civil society to better understand – and address – where projects are not delivering on desired outcomes or creating unexpected harms.
3. Communications campaigns to ensure people understand their choices and rights
The discussion highlighted that the communications efforts around a given project or service are essential to the success of the initiative. On one hand, communication is essential to user adoption and uptake. For example, digital payments like UPI had become ubiquitous in many parts of India. However, there was a clear urban-rural divide, where rural areas saw less uptake. Mansi and her team found that the service started to scale once time and resources were spent building awareness around the service and its benefits. On the other hand, thoughtful and user-centric communication is also central to securing meaningful and informed consent. In the case of the women in Ghana, they needed a breakdown of how their data was going to be used and their rights to control their own data – in their own language – in order to knowingly agree.
4. Effective redressal mechanisms baked into the technologies, systems, and processes
In the instances where digital systems may fail or may amplify human errors, online dispute resolution and grievance redressal are essential to protecting people and maintaining their trust. A strong ecosystem of redress options forms a buttress for DPI. In the case of UPI, research showed that rural users of the app were not necessarily comfortable with the redress mechanisms – largely handled via chatbots. In this instance, a parallel analog process would help better meet the needs of more users. These mechanisms are not nice-to-haves, but essential pieces of DPI necessary to build trust. Without trust, DPI systems simply won’t be used and the time, money, and political capital spent building DPI will be wasted.
We must work together to promote DPI as a truly impactful approach that maximizes public value.
The discussion ended with a clear message from each of the speakers: We must work together to get DPI right. Understanding people’s experiences with current digital systems – even when these experiences are negative – does not mean abandoning DPI entirely. Across all of the research, interviewees were nearly unified in their desire for digital services, even when expressing their frustrations. They understand that these services are critical to modern life – they just want tools to understand, trust, and benefit from these services.
Want more user stories? Check out the related articles below to read our impact series, where we are pursuing a deeper understanding of the lived experience of DPI users.
