Mythbusters: Digital Public Infrastructure & Digital Public Goods

From distributing life-saving shelter, food, and medical attention after a disaster; to delivering government subsidies for seeds and fertilizer to smallholder farmers; to enrolling students for national examinations— these are the kinds of “public goods” not unlike roads that require a non-invisible hand to ensure citizens have equal and fair access. At a foundational “building block” level, this includes digital systems that give citizens:

• The ability to verify their identity
• The ability to share data accurately and securely
• The ability to make and receive payments

Look no further than the US Department of Education’s disastrous handover of its loan repayment servicing to private-sector debt collectors like Mohela to understand the impact on student borrowers, let alone the Department of Education’s ability to collect payments. Or consider that the average American spends 13 hours and $240 on third-party tax software each year just to file taxes; and despite it being ostensibly free, standardized, and compulsory, 33% of Americans “completely” or “mostly” disagree that they trust the IRS to help them understand their tax obligations. In sum, the consequences of a free market approach to government digital transformation are felt most severely by citizens, which can foster and in some cases exacerbate mistrust in government technology systems.

You’ve probably been using some very high-quality open-source technologies yourself and just didn’t know it. Some of the most widely-used examples of open-source software, both inside and outside of government, include the Linux operating system, Apache web server, and PostgreSQL database. Open-source tools specifically for DPI, also known as digital public goods, include DHIS2 for health; Moodle for education; openCRVS for civil, birth, and death registrations; the Open Contracting Platform for government procurement; and OpenStreetMap for GPS and mapping data. 

The demand for these tools speaks to their quality: DHIS2 alone has been adopted in over 74 national health systems, and will be adapting its model for the education and agriculture sectors. And even if one of the existing open-source DPI tools doesn’t meet your exact quality standards, the beauty of open source is that you can customize or build on top of the foundational infrastructure to build any solution you want, much like we did for local health districts in Tanzania and the Ministry of Agriculture in Ethiopia.

Source: AfrICT (Africa ICT)

Piecemeal approaches to digital systems can make governments more vulnerable to potential risks. While open-source technology does have vulnerabilities (like any software), using open-source for DPI is arguably more secure than the alternative because it allows for cross-system interoperability, data validation, and cross-referencing. Public code review forums like GitLab (also open-source) and GitHub give users access to the latest security upgrades, bug fixing, and risk mitigation measures that would otherwise require extensive long-term support contracts or subscriptions. India regularly invites the best hackers in the country to intentionally break into the open APIs that power its digital public infrastructure, India Stack, and these hackathons have enhanced the security and protection of the technology. 

While OSS can reduce licensing costs, there are still costs associated with implementation, customization, maintenance, and training. DPI projects also require substantial investments in infrastructure, development, and ongoing support. Taipei, for example, recently invested US$6.27 million to roll out a digital national ID program but had to pause in 2021 due to security concerns. Other experts in India estimate that the average cost of a digital identity program is about US$1 per person; meanwhile, costs for digital payment infrastructure can range from US$100 million in Thailand to only about US$2 million in Brazil. 

However, the long-term economic impacts for society greatly outweigh the costs of deploying DPI. A McKinsey study recently estimated that Digital ID alone can boost economies by 3–13% of GDP, with an average of 6% growth in emerging economies. Governments that opt for open-source DPI can benefit from additional long-term cost savings by not having to renew licenses or software agreements, and adjusting or maintaining the technology in-house rather than contracting a third party. 

A recent Deloitte study confirmed that DPI can strengthen the government’s digital systems and facilitate equitable access to digital services. Take for example the successes of Ghana Post GPS, a digital addressing system that provides every location in Ghana with a unique digital address. Aside from ensuring your maps app will navigate you to the correct destination in Ghana (“Eye” Street in DC, I’m looking at you), this tool has improved citizens’ ability to access services, including emergency response, mail, and e-commerce; enabled more inclusive access to services for residents in remote and informal settlements; and facilitated business operations and logistics by providing precise location data.

DPI, when implemented with open data principles, can increase government transparency and accountability, enabling citizens to access and analyze public information.  If implemented responsibly, DPI makes anonymized and real-time data available to citizens and civil society watchdogs, and the transparency provided by DPI can help improve civic engagement, increase trust in government, and credibility in a country’s decision-making. Further, by allowing anyone to contribute, open-source technologies foster innovation and collaboration across different sectors and geographies. Digital Public Goods communities have global reach and active online communities, which further support the reputation, neutrality, comparability, and trustworthiness of these systems.

In addition, open-source DPI solutions allow governments to avoid being locked into a single vendor’s ecosystem. This gives governments more flexibility and control over their IT systems, preventing dependence on high-cost, long-term contracts with the private sector. By using and developing their own digital infrastructure, governments can reduce dependency on foreign technology providers and protect their data sovereignty.