Bureaunia Chronicles: Episode 22
Aithorix, Bureaunia’s Chief AI Advisor, and Auditax, Bureaunia’s Finance Minister, will host a side event at UNGA
With Friday almost upon us and as we prepare to launch our side event at this year’s UN General Assembly, Auditax is still unsure about the name I chose: Bureaunia’s Failure Stories. I understand the hesitation. While every other delegation in this building is most likely preparing its success stories, we are walking in with the other kind, and asking why we so rarely tell them.
I did not choose that name hastily. Ten years ago, at a public-sector innovation forum in Singapore, we looked back at Bureaunia’s own record on reform using technology and presented our failures.
From Big Vision to Small Failures
Our first attempt, in 2007, was built around large vision documents and long project plans. The documents quickly became shelfware, but we learnt something unexpected and valuable: the vision documents were not very useful, but the visioning process was very valuable. Because visioning built shared understanding, trust and mission-orientation in a small group of thinkers and actors across the government, civil society, private sector, and development partners.
A year later, we tried to go smaller: think big but plan small, and start execution right away, and keep refining based on citizens’ feedback. By 2013, that habit grew into something bigger: Service Innovation Teams in every ministry, every department and every district supported by a Service Innovation Fund, built on one governing principle: the innovators would enjoy administrative protection during execution and after failure provided they record their learning for the next cycle and next group of innovators. The concept we kept coming back to was blunt: the faster you fail, the sooner you are likely to succeed. Taken not from Silicon Valley, but from the villages and towns and streets of Bureaunia through the voices of its residents, particularly those who live on the margins.
It was an honest commitment from political leadership for a country-wide, whole-of-government experiment, with that protection built in.
Five years in, we looked back. We saw hundreds of failures. But in that list we found about 50 innovation initiatives that succeeded so spectacularly and scaled to population scale from a small pilot in a district or department that our hearts filled with joy and pride at having run that experiment. A health officer in a remote district developed a spreadsheet-based health record system that ultimately led to the Ministry of Health’s electronic medical record system that now boasts AI inclusion. An agriculture extension officer built a digital database of crop diseases which led to the Ministry of Agriculture’s e-Crop App which ultimately incorporated soil testing, fertilizer and pesticide recommendations, micro-weather prediction, now turbocharged by small language models using local dialects. A teacher’s group developed digital content to make classrooms more effective for joyful learning for primary and secondary school children. This led to an online community of a few hundred thousand teachers sharing digital content with each other, not waiting for some big-name education experts from the capital city and from the developed countries bestowing their wisdom.
Several land officers all of whom tried to produce the mythical digital land record system ultimately failed but not before documenting what they did and how they did it. This collective learning later formed the basis of the Land Ministry’s country-wide e-Land System which now serves millions saving the marginalized farmers and widows time and money they wasted on multiple visits to the faraway land offices and paying bribes to the intermediaries.
We learnt that allowing failures – actually, not just allowing but promoting failures – takes away fear and nurtures risk-taking in public officials. And 20% of the initiatives rise to fulfill citizens’ expectations. The rest of the 80% – which we call failures – are not failures if they record their lessons for posterity.
This year’s UNGA theme is “restoring trust, managing transformation.” I don’t think trust comes back by leaving things out. Before Auditax and I say another word about finance and infrastructure, we want to put the question to the room directly: how many delegations are prepared to name what failed, with their own institutions attached to it?
The Big Failure that We Structurally and Perpetually Promote
And if Bureaunia is asking others to name their failures, it has to begin with its own. One of ours was not a failed app or a collapsed procurement. It was more structural: we kept paying to build the same foundations again. Ministries built and re-built the same identity and payment systems in parallel, each defensible on its own, none of them talking to the others.
Here is why it keeps happening and surprisingly, it has nothing to do with technology. Development finance still funds a digital system the way it funds a bridge. A bridge is built once, in one place. And if you want another one somewhere else, you pay to build it all over again. But software does not work that way. Once a country builds something like a payment rail or an identity system, the marginal cost of a second ministry, or a second country, adopting it approaches zero. Yet the world keeps paying for it as if it isn’t.
No single actor in that chain is incentivized to reuse an existing system. Lending officers are assessed on volume disbursed, and a reused system is a smaller loan on a slower timetable. Finance ministers are celebrated for bringing in the largest financing and they sign for repayment due long after they have left office.
No party to that transaction represents the citizen who services the debt two decades later, a citizen who does not experience a duplicated registry as a financing flaw. She experiences it as being asked to prove she exists twice.
Auditax keeps returning to one line: no appraisal framework yet measures the difference between rebuilding and reusing, in avoided cost and in time to reach the first citizen served. Until that number exists, the case for doing this differently rests on argument alone. The case for the status quo rests on nothing at all. Bureaunia is no longer willing to accept nothing as an answer. We are asking who else in this hall is finally prepared to start counting.
We keep returning to one point: we know how much a new system costs, but we rarely count what we save by reusing what already works. We measure loans, builds and disbursements. We do not measure avoided debt, faster delivery, retained capability, or the cost of rebuilding the same thing twice.
That is why our UNGA session is about failure. For two reasons. One, are we allowing mushrooming of small failures to create the capacity for big successes? Two, are we perpetuating big financing failures knowing fully well the consequences to our citizens?
Anir Chowdhury
#BureauniaChronicles #UNGA81 #DPI #AIGovernance #DigitalPublicInfrastructure
Leave a Reply