Two interesting stories have surfaced in the news over the past few days.

First, there’s the entrepreneur with school-age kids who couldn’t stand the dysfunctional IT solution the city of Stockholm spent millions developing. Frustrated, he took matters into his own hands and built an open alternative. That didn’t go down so well, in fact city officials called the cops on him.

Then the next day a story about another rogue citizen. This guy was frustrated because of the long waiting lists to get a new passport. But more specifically he was pissed off because he couldn’t even book an appointment with the passport authorities, since their system wasn’t built to handle the amount of incoming requests.

So instead of spending hours trying to click at the exact right time when the interface was responsive, he built a simple script that automated the process for him. Then published it on GitHub, and before long it had received improvements from five other contributors.

These stories are revealing. While Sweden has become a unicorn factory and programming is now one of the most common professions, successful public sector IT projects are exceptionally rare. In fact, there’s a long list of initiatives that have failed spectacularly, often after costing hundreds of millions.

What exactly went wrong here, and how can the system be fixed?

It’s clear that public sector IT projects suffer from a lack of agile development processes. ‘Big bang’ implementations are the default, instead of delivering incremental value through short sprints. And with big shipments come great risk of course, especially in projects where the end deliverable is unclear as you set out, which is pretty much always the case in software.

But simply prescribing ‘more agility’ is bound to fail. It’s like telling a drowning person who never learned to swim to ‘try harder.’

Instead, I believe the one factor that could actually bring leverage, would be a mandate that forces all public sector players to adhere religiously to open source, open data and open API’s

Only then could knowledgable and committed citizens who want to contribute be seen as a resource and only then can the non private sector digital infrastructure stand a chance to start developing an anti-fragile resilience in the face of ever nastier cyber threats.

What would it take to execute such a paradigm shift? First and foremost, political leadership. Just as we have a minister for infrastructure, we need a national CTO (or a similar role). Because IT is fundamentally different from building roads, bridges and railways.

It would also make sense to create an agile task force that could be deployed as needed to address the fact that each public sector silo is currently held hostage by private sector IT consultants. This isn’t just a waste of taxpayer money; it also contributes to a constant brain drain, preventing institutional learning.