Developing countries often adopt public financial management reforms as a signal to garner external legitimacy. Not enough attention is paid to the local context, meaning that the impact is limited

The public financial management community has emerged aggressively in the past few decades, offering the world a set of practices that did not exist 20 years ago but are now commonplace.

We often know them simply by their acronyms; MTEF (Medium-Term Expenditure Frameworks), IFMIS (International Financial Management Information Systems), and IPSAS (International Public Sector Accounting Standards) are prominent examples. These products are the good or better or best practices that our community advises governments to adopt.

And governments do adopt these practices, especially in developing countries. Official reports by professional organisations show that up to 100 countries have adopted IPSAS, for instance, and many of these adopters are in the developing world.  A blog post on this site recently provided an example, noting that Nigeria plans to introduce IPSAS in one year.

Such a high level of adoption sounds like a good story. But we should not be congratulating ourselves too fast. More often than not, what you see is not what you get with these new PFM practices; they are adopted as signals to garner external legitimacy but often cannot be implemented and seldom provide real solutions to the PFM problems of different countries.

My perspective is based on evidence I discuss at length in my new book, The Limits of Institutional Reform in Development.  The book grew from research into PFM performance and other reforms in developing countries. I found that:

• Many governments were pursuing the same best practices, often with the support and advice of international organisations. Medium-term frameworks are now in place across the developing world, as are integrated financial management systems, modern internal audit mechanisms and competitive procurement regimes.
• Projects introducing the new practices were commonly considered successful, with public financial management seen as a bright spot in the otherwise dim world of governance reform.
• Governments from Afghanistan to Liberia and Uganda showed improvement on overall measures of PFM processes – like Public Expenditure and Financial Accountability (PEFA) scores.
• But these governments often lacked better outcome scores on these and other indicators – budgets were not more reliable than they had been in the past, and money was still not flowing effectively to foster the kind of service delivery countries needed.
• And these countries were still open to major problems with their public financial management. Even with ‘better’ systems designed to improve fiscal management, for instance, Uganda continues to experience regular instances of fiscal mismanagement and corruption.
• When one scratches the surface, the ‘new processes’ are not properly implemented in many countries. Medium-term frameworks make budget documents look better, for example, but these budgets are not executed as they are drawn up. Laws have improved, making it appear as if governments account and report in certain ways, but the laws are not implemented. Central agencies like treasuries, budget departments and procurement bureaux have strengthened their capacities and obviously support reforms; but decentralised agencies that have to live with and implement reforms still have limited capacity and do not buy into the new processes.

Evidence from some developed countries is quite similar to this. In the past few decades, for instance, the US Federal Government has arguably pursued more PFM reform than in the 70 preceding years.  It has adopted various versions of performance budgeting, created and strengthened bureaux that perform multi-year forecasts intended to influence spending decisions and regularly changed accounting rules to try to improve conformity with international standards. But the budget has grown seemingly without constraint in this period, and few observers would say it is more strategic, performance based, long-term, or manageable.

The bottom line is that many governments – developed and developing – adopt the new good, better and best practices in form but do not see improved functionality as a result. The reforms are limited and should thus not be considered successful.

This is not a good story. It should be a story that worries those who really believe in the power of MTEF, IFMIS, IPSAS and other PFM reform products. All of these (and other) best practice reform variants are subject to the same kinds of limits and may very well be adopted to much acclaim without providing solutions to PFM problems in reforming countries.

These limits manifest because PFM and other institutional reforms are often adopted as signals – to make governments look better – and not as realistic reforms designed to actually make governments better.  This tendency to signal is common to all governments but is stronger where:

• Governments are under pressure from outside parties or dependent on external support. Think of Greece, perhaps, and the pressure it faces to clean up its public finances; or consider developing countries negotiating with donors looking for evidence that the country can manage its money; or think of the US Federal Government trying to send a message to taxpayers that it can spend resources effectively.
• The problems countries face are really complex, involving multiple parties, deep politics, and with no clear answer.  Think about the fiscal problems of post-2008 Europe, or the political struggles that underpin the fiscal profligacy in the US, or the major challenges developing countries like Afghanistan, Liberia and Uganda have of improving spending in the face of unreliable revenue streams, messy and unstable economic and political contexts.
• Outside agents have identified good, better or best practices that are deemed legitimate. Agreeing to adopt these practices will make any government look legitimate, even if the practice is not a good fit to the country’s context or challenge (and adopting the practice makes no sense). In sad contrast, deciding that one will not adopt these practices makes a government look like it is not legitimate, even if the practice is not a good fit to the country’s context or challenges (and rejecting the practice makes sense).

Where governments are dependent, problems are complex, and ‘best practices’ are already being marketed, one should expect that countries will adopt reforms as signals. And one should expect that the results will be limited. This is because of three major weaknesses of ‘reforms as signals’:

• First, reforms as signals are often introduced with scant attention to the complexities of context. Unfortunately, contextual constraints – in the politics, capacities, ways of thinking and seeing – always end up influencing what can and cannot be done and often undermine reforms because they are poorly considered when reforms are introduced. An example is the way Argentina’s fiscal rules failed to bring control to the country’s finances because political leaders ignored the rules. But any observer of Argentine politics could have predicted that this would happen and that a ‘best practice’ fiscal rule probably would not work in Argentina. Because the fiscal rules were adopted as signals, however, this observation was not made.
• Second, reforms as signals often over-specify solutions but oversimplify what it takes to put these solutions in place. Most African countries have been encouraged to adopt international private sector accounting standards, and the mechanisms for reform typically involve new laws, professional bodies and oversight agencies. The reforms do not address the ‘softer’ or unseen content required to make an accounting system work according to international standards. Reform designs constantly take for granted the cultural and normative dimensions informing such standards, assuming, for example, that there is a cultural appreciation of the need for and value of disclosure in countries slated for reform.
• Third, reforms as signals are frequently introduced through narrow sets of agents – champion-like ministers or treasurers or budget directors – who often lack the reach and influence to access support from the ultimate implementers in line ministries, provinces, and districts. This manifests in accounting and budgeting processes and systems that look good on paper but are not given life by those who actually make transactions and input data. This latter set of distributed agents are crucial to the success of any real institutional reform but are not important if the goal is just to introduce ‘reforms as signals’.

My biggest concern with our public financial management community is that we are all about reforms as signals. Many countries consider reforms when they are under pressure to do so from outsiders. If we are honest, we don’t really know what to do in many of the over 200 different countries where reform is being considered, but we have a rigid and highly marketed set of one-best-way best practices for these countries to adopt. We push these solutions on countries regardless of context, and with very little attention to the softer content requirements of these reforms, and with very little connection to end users.

But there is a better way. In the book I discuss this as PDIA (problem driven iterative adaptation). PDIA is just an amalgam of good ideas that others have suggested about how institutions really change and, more fundamentally, is the way in which most functional public financial systems we currently see actually emerged. It has a number of fundamental dimensions:

• First, PDIA emphasises starting reform by asking what the problem is. It is not solutions-based and does not begin with a good, better or best practice in mind. There are various reasons why starting with problems matters but, primarily, it is because agents embrace change more regularly when faced with problems that they must address. And if one forces a conversation about problems at the start of reform, the chances of mindless ‘reform as signals’ is partly mitigated. I believe functional systems emerge when agents address problems in their context. For instance, Sweden came up with various reforms in the 1990s as it constructed a story about the failings of its public finances during the 1992 crisis. It drew attention to problems, used these to argue that change was needed, and focused its reform search on finding solutions to the problems.
• Second, PDIA emphasises an iterative process of reform and change. Simply put, this means going step by step in the process. Each step involves experimenting with various ideas, building capacity for new ideas, gradually establishing political support, capturing lessons about what works and why, and moving to the next step. This approach does not guarantee ending with one ‘best practice’ reform, but it does hold the assurance that whatever emerges as the reform hybrid is likely to be something that addresses the problem it was designed to tackle, has political support, and can be implemented.
• Third, PDIA emphasises working in groups of agents. These groups are not your typical bureaucratic agencies but look more like networks of leaders. These leaders all play different functional roles required to foster change. Some help define the problem, others provide authority to address such, some bring new ideas to the table, others offer the implementation perspective, others motivate, convene and connect.

PDIA is the ideal antidote to ‘reforms as signals’ and offers a way of doing reform and pursuing change that is less limited than the current public financial management reforms we currently pursue. It requires us spending less time marketing global solutions, like MTEF and IPSAS, and spending more time asking country specific questions about problems and potential entry points for real change: What is the problem in Greece? Where can reform start, with what small step? When would we be able to think about evaluating the results of this step, gather lessons learned about what worked and why, and pursue a next step?

This kind of approach may ultimately lead countries to adopt the kind of solutions offered in mechanisms like MTEF, IFMIS and IPSA, but in a functional sense and not just in form. This should, surely, be our goal as a community.

Matt Andrews is associate professor of public policy at Harvard University’s John F Kennedy School of Government. The Limits of Institutional Reform in Development is published by Cambridge University Press