We start this month with a new (and rather bullish) paper from the IMF on GovTech and public finance. We then pour cold water on this optimism by reflecting on more stories of IT failure in Canada, Ireland and the UK before concluding with a look at the flurry of high-level advocacy around digital public infrastructure (DPI) in September which was, as always, worryingly optimistic.

Without further ado, here are some of September’s top trends in digital and public finance.

1. Transforming public finance through GovTech

A new IMF paper called ‘Transforming Public Finance Through GovTech’ makes a case for ‘the potential to strengthen public finance operations and improve the delivery of basic services’ through digitalisation. Section III highlights lower digital adoption rates in lower-income countries, which are missing out on opportunities to raise more revenues and spend them more efficiently (though as always, there are real questions over correlation and causation in most of these findings).

While not wholly divergent from our own arguments for bringing public finance into the digital era, the paper’s public finance sections tend towards affirming a solution-driven approach rather than the approach based on meeting the needs of users that comes out in other parts. This may be due to ongoing (and understandable) difficulties in reconciling the GovTech philosophy of citizen-centric access to public services with strong beliefs about how technology should play a role in (good) public financial management and be part of a whole-of-government approach to digital transformation. Doing so is not easy, as many governments are still finding out.

2. IT cost overruns are out of control (Canada edition)

Why do government IT contracts have such huge cost overruns? This was the topic of a new article from the Investigative Journalism Foundation in Canada. Using data published by the Canadian government, researchers found that almost half of IT and telecom contracts overrun their originally agreed price by an average of 134% — far higher than their analysis for other sectors. Unsurprisingly, the article highlights the case of Phoenix: 'a system that was supposed to save the government $70 million a year by centralizing payroll [but] has ended up costing taxpayers more than $1 billion'.

Through interviews, the article points to multiple reasons for these cost overruns, including vendors lowballing on price to secure contracts, the government's failure to understand its IT needs, and a lack of project management know-how in the public sector. Perhaps the most interesting reason cited, however, is that ‘software is never really “done”'. This is of course true, but it does not fully explain the quantity and scale of the cost overruns reported here. We have argued elsewhere for a shift to agile funding and delivery for software development that recognises the fact that user needs change and that systems require ongoing iterative improvements. These methods are expected to save money, including by stopping projects that are not meeting user needs.

Speaking of which, our upcoming webinar is on the relationship between digital transformation and agile procurement. You can sign up here.

3. More financial management software failures

IT cost overruns are also common when it comes to financial management software. This month alone yielded two fresh examples: one in Ireland and the other in the UK courtesy of Birmingham City Council.

In Ireland, the Department of Finance suspended its own move to the new Financial Management Shared Services (FMSS) system owing to user dissatisfaction. It was hoped that the project would generate savings by replacing 31 legacy systems in 52 government bodies. Initially expected to cost €47 million over a period of five years, the budget has since ballooned to €115 million, the timeline has been extended by a further five years, and no cost savings are expected for the next decade. Meanwhile, Birmingham City Council’s Oracle implementation is in nine-digit cost overrun, contributing to the council's bankruptcy, and still does not work.

Both stories appear to be classic cases of the sunk cost fallacy whereby substantial up-front investments rule out a subsequent change in course, and lead to good money being thrown on top of bad. More fundamentally, they represent a failure to take into account user needs and incentives. Our recent paper cites both of these as major challenges to bringing public finance into the digital era.

As Rupert Goodwins argues in the Birmingham story, there is a different way of doing things: ‘Build an equivalent stack as a conceptual framework for local government needs and processes, and the things they all have in common will create a huge market for sustainable services despite no two organizations being the same.’ The same logic can be applied to government departments at the national level. As Goodwins notes, this is a ‘philosophical change of viewpoint that will be in direct conflict with billion-dollar global powerhouses’, but would be a change worth making.

4. DPI is front and centre at UNGA

As expected, in September the United Nations General Assembly (UNGA) highlighted the theme of DPI — perhaps most prominently through the launch of a new DPI Safeguards Initiative with the United Nations Development Programme (UNDP) during its High-Level Week. At the same time, the new 50-in-5 Initiative aims to help 50 countries scale up components of their DPI in five years. India’s G20 presidency concluded in September too, with data and DPI very much on the agenda and UNDP launching a supporting playbook around the DPI approach.

We must exercise caution in reading these DPI tea leaves and figuring out what initiatives work best so as not to incentivise countries to run into the wrong thing before they realise what they have done. Though defined ambiguously in their current guises — and potentially representing a risk for countries — we hope the DPI Safeguards and 50-in-5 initiatives will not only help codify the key principles for developing secure data exchange capabilities that finance ministries are crying out for, but also provide careful consideration of country-level architecture needs, as the needs for DPI can vary.

It remains to be seen if these international institutions sustain their enthusiasm for DPI. In particular, it will be interesting to see if the momentum gained under India’s G20 Presidency continues when Brazil takes the helm next year, and whether DPI remains on the agenda at future high-level fora.

News from our network

The 2023 ODI Public Finance Conference on navigating the poly-crisis took place last month. If you missed it, head to the event webpage to watch the recorded sessions. We particularly recommend Mike Bracken, Yamini Aiyar and Fantahun Belew Asfaw discussing the role of digital in targeted and responsive fiscal policy.

The day before the conference, the World Bank held a workshop at ODI to launch their ‘Reimagining Public Finance' initiative, marking 25 years since the publication of the Public Expenditure Management Handbook. The session gave us food for thought on the role of digital as a lever for bringing policy and delivery closer together, which we hope will be a key theme for the Bank and other international organisations in the months ahead.

And finally — our new LinkedIn Learning Group provides a space to share news, ideas and opinion on the intersection of digital and public finance. To connect and share content, visit our group webpage and request to join.