Our Programmes



Sign up to our newsletter.

Follow ODI

Digital Public Finance - top trends in March and April 2024


Written by Cathal Long, Andrea Sissa Velandia

Hero image description: Abstract digital image consisting of yellow, orange, red and blue to represent degrees of heat Image credit:Dan Cristian Pădureț/ Unsplash

Welcome back to the Budgets and Bytes blog and our bi-monthly review of the top trends in public finance and digital.

This instalment covers recent research and discussions on: financial management information systems in OECD countries; whether digital government is as emergent as we think; recent developments on the digitalisation of tax administrations,  how to interpret ‘public’ in the term digital public infrastructure; and how and whether to make data open in the public interest. 

Here are the top trends in digital we found relevant for public finance in February and March: 

1. OECD countries face similar challenges with their Financial Management Information Systems (FMIS) as lower income countries

The term ‘financial management information system’ (FMIS) is widely used in international development. The World Bank and others have made substantial investments in FMIS in lower income countries, but the returns have often been below expectations. The term is used less widely in OECD countries, and it has been difficult to get an overview of how IT is used for public financial management in these higher income countries, whether it differs substantially from in lower income countries, and whether these systems work better compared to in lower  income countries. This new paper on FMIS in OECD countries provides a great service in starting to fill that knowledge gap.     

Based on a 2022 survey of senior officials the paper provides interesting insights on: 

  • Modernisation – a large majority of OECD countries are using systems that are more than 10 years old and more than half are considering modernising or replacing these systems. 
  • IT architecture – there is no clear consensus on whether it is best practice to take a centralised and fully integrated approach to FMIS. Most OECD countries are somewhere in the middle when it comes to these categorisations, institutional and functional fragmentation is common, and some processes are still manual.  
  • Technology choices – most OECD countries favour commercial off the shelf (COTS) solutions, and some are moving towards cloud-based solutions, but issues related to COTS providers are amongst the top challenges they face.  
  • New technologies – all OECD countries are actively implementing or considering business intelligence tools, and to a lesser extent cloud solutions and AI tools. In contrast relatively few are considering blockchain, so perhaps this has passed through the peak of the hype cycle.  

Our biggest takeaway from the paper was that OECD countries face similar challenges – around modernising business processes, lack of in-house skills, and data governance – as lower income countries do. This suggests that international guidance and coordination on software and data standards could benefit from greater cooperation between the so-called ‘global north’ and ‘global south’. Speaking of which, debt specialists at the World Bank have called for the creation of a ‘task force to coordinate the design of better debt-management systems’ to tackle ‘the worlds hidden-debt problem’. Perhaps this is a concrete area where some digital public goods for public financial management could emerge. 

2. How emergent is digital government? 

Not as emergent as some believe according to a new paper from Gerhard Hammerschmid and others. Analysing the aims, objectives and ‘imagined futures’ of the digital strategies of eight European countries for signs of a dominant paradigm, they conclude that digital government is ‘an “academic invention” and a normative framework describing how public administration should be organized rather than a model capturing the empirical reality’. That is a concern for those of us who view digital thinking and ways of working as a means to improve public finance and its management.    

Moreover, their finding rings true in the context of recent criticisms of the UK’s record of digital era governance. Fresh on the heels of the Post Office’s Horizon scandal, the Guardian recently revealed that Home Office software has merged the identities of 76,000 people with terrible repercussions for those affected. And Keegan McBride provides a broader perspective that ‘digital government is failing’ warning that ‘with little ability to share and move data between organizations and services, it is difficult to progress and build truly innovative and impactful digital services’. What does this say about emergence of digital government globally given that the UK was ranked third in the most recent OECD Digital Government Index?  

But for those that do imagine a better (digital) government in the future, there is some fresh guidance. Firstly, from the UK National Audit Office which released a guide for senior leaders and audit and risk committees on digital transformation in government, which follows on from their rather scathing report on barriers to efficiency from last year. Secondly, the The Radical How from Public Digital recommends 10 changes the next UK government should make ‘to build a mission-oriented government that delivers better outcomes, reduces risk, saves money, and rebuilds trust’. Lending further credence to the views of Hammerschmid et al., the authors note that these 10 changes in practice have been the exception rather than the norm to date. But they argue that where they have been practiced, they have produced exceptional results.

3. From e-filling to AI in tax administration  

One area where digital thinking and ways of working seem to be making more noticeable inroads is in tax administration, and the ICTD DIGITAX Research Programme , as well as others have been doing a great job of bringing these to our attention.  

This DIGITAX case study documents the digital transformation journey of the Togolese tax administration, including the benefits of introducing digital technologies (lower compliance costs, improved compliance, and improvements in collections, despite incomplete adoption), and the challenges of doing so (lack of IT skills and funding, staff resistance, and legacy IT systems). Meanwhile, Don Moynihan has been covering the Internal Revenue Service’s recent modernisation project, highlighting the political economy challenges of introducing free e-filing in the United States and the administrative burdens its tax payers and collectors have faced as a result. And, the Australian Tax Office (ATO), has provided something everyone has been asking for – a use case for AI in public financial management, with the ATO using AI to recover unpaid taxes.  

We have argued in the past, that the relatively more citizen facing nature of tax administration, compared to public expenditure management, makes it more fertile ground for digital transformation. Nevertheless, there is a lot that can be learned. Indeed, the case of Togo highlights that many of the challenges are the same as those we have previously identified

4. What puts the ‘public’ in digital public infrastructure? 

The increasing use of the term digital public infrastructure (DPI), as well as growing expectations of its impact on everything from service delivery and economic growth to responding to climate change, has been accompanied by calls for clarifying what is meant by the ‘public’ in DPI to avoid the term being captured. A new paper from the Institute for Innovation and Public Purpose at UCL aims to do just that.   

The paper undertakes a lot of useful conceptual clearing, by first setting out some definitions for traditional infrastructure (e.g. roads) and how those apply to digital infrastructure, and then providing a potted history of the evolution in thinking from e-government to DPI. They then provide an overview and examples of the two main ways that ‘public value’ is understood with respect to DPI – through its ‘attributes’ and ‘functions’. Ultimately the authors find this ‘public values’ framework for describing the ‘P’ in DPI limiting, because it is silent on the role of governance and the state. They instead propose a ‘common good framework’ for ‘public value maximization’, identifying five pillars that public institutions should aim to guarantee – purpose and directionality, co-creation and participation, collective learning and knowledge sharing, access for all and reward sharing, transparency and accountability.      

The first pillar, purpose and directionality, is particularly pertinent from a public finance perspective. As the authors argue throughout, DPI is not neutral, it has direction, and can lead to unintended and undesirable outcomes. In our last roundup we covered Yamini Aiyar’s concerns that DPI is making it easier for political parties in India to credibly promise benefits in exchange for votes. A new working paper from the National Bureau of Economic Research provides some evidence that this is coming at the cost of traditional infrastructure like roads and irrigation.

5. Making data open in the public interest 

An amendment to a bill currently making its way through parliament could make the UK’s postcode address file (PAF) more accessible. As James O’Malley writes, the PAF is a ‘critical public dataset for the digital age’ but it is currently privately owned by Royal Mail who ‘charge hefty fees if you want to access it’.  

The case of the PAF represents a larger problem of private firms holding data that ought to be open in the public interest. That requires better regulation and public interest data intermediaries that create the right incentives to make data open according to Simeon Duckworth. Governments are also responsible for making the data they control more open for others (both inside and outside of government) to use. But many governments ‘see their role with data to be “collect and protect,” not “collect, protect, and use”’ according to Charles Kenny. Open data is also becoming an increasingly outdated concept as was discussed in this Open Data Institute roundtable which emphasised a data spectrum for the accessibility of data.   

Finance ministries should care about data accessibility for at least three reasons:  

  1. Encouraging innovation and growth – according to James O’Malley, things like the PAF are ‘used every time you buy something online’ and ‘licensing fees amount to a tax on innovation’ which ‘are holding back economic growth and productivity.    
  2. Promoting better services – the efficiency and effectiveness of our public services are increasingly being determined by how well governments use data.   
  3. Saving money – as well as app developers and other gig economy startups, government agencies are amongst those paying for the privilege of using datasets like the PAF, meaning government (and taxpayers) are paying for the same thing over and over again. 

News from our network

In February Lauren Kahn and James Stewart from our partners at Public Digital attended the World Bank ‘Climate-smart Public Financial Management Middle East and North Africa (MENA) Conference’ with representatives of regional Ministries of Finance, Planning, Environment, Energy, and Transport. All the talks from the conference are now online. A recurring theme was (as always) the need for better measurement, and among the lessons from digital transformation that can be applied for climate transition we talked about the value of service-performance measures like the UK government’s much missed performance platform and Bangladesh’s Time Cost Visits approach provide a good starting point.