Last week I posted a blog on Sweden's public financial management reforms. I asked what lessons we could learn from these reforms. Many people are looking for these lessons, with various attempts to mimic the fiscal rules Sweden employs, and the multi-year budgeting approach they use, and the performance based mechanisms they work with.
I argued that the lessons from Sweden are less about the mechanisms and systems and rules they have in place today, and more about the processes they went through to find and fit these mechanisms, systems and rules. These processes spanned decades and involved the basic dimensions that Lant Pritchett, Michael Woolcock and I see as important in any change process. We present these dimensions together in our paper on Problem Driven Iterative Adaptation where we call for development projects that are (i) problem driven, (ii) iterative--where future steps build on lessons learned from past steps, and (iii) involving many agents working together across space and time.
I see these dimensions in the Swedish story. They are evident 80 years ago in respect of the country's fiscal rules, where academic luminaries were engaged in offering ideas--and even nascent reforms--where what we would now call fiscal rules were discussed. The figure below shows how counter cyclical budgets were first attempted in the early 1930s, chewed on by intellectuals like Myrdal, and then introduced via reform in 1937. These rules were introduced because of major problems emerging because of the financial crises of the 1920s which raised serious questions about the role of state finances in the economy. They emerged gradually through the decade, with some experimentation, some learning, and then reform. They were the product of multiple agents in academia, various political parties, and the bureaucracy. PDIA in action even then. Most importantly, the lessons from this period carried forth into the 1990s and 2000s when coalitions of agents again picked up on their relevance given a new set of problems...
The 1930s experiments lost steam at the time because the country went through a difficult period during world war two. The problems of the 1930s had changed and the public financial management and budgetary issues of the day were different. After the war, one finds the government growing in size, and many bureaucrats and politicians started asking some even newer questions, reflecting new problems. The major problem seemed to be how they could allocate money effectively given the many new roles government was being asked to play in the economy. And so one finds experiments with program based reforms that also had a multi-year budgeting element to them. The country used pilots to introduce the reform and an alliance developed between the budget officials in the ministry of finance and officials in the national audit office. It is really interesting that these groups collaborated to do regular evaluations of the reform, with many lessons emerging from these evaluations in the early and late 1970s. By the late 1970s the reform was largely abandoned as a failure, partly because the problems warranting attention had again changed: the state was facing constraints and asking how to control spending, more than allocate it. So, one finds PPBS abolished in the same year that the government begins using a 'Cost slicer' to help control spending.
The reform may not have stuck, but once again I find it interesting to note the dimensions of PDIA in action. Problems driving reform. A reform introduced in steps with lots of learning and adjustment over a period of time. Multiple agencies involved. There was no five year project development phased, with loads of project preparation documents, a pre-designed intervention, limited wiggle room, and one central driver pushing reform from the top.
Most importantly, one sees this reform experiment leading to next steps almost immediately, and culminating in the more advanced multi-year and performance budgeting interventions we see today. The next steps were in the 1980s, accompanied by political transition and an abundance of new thinking about government. The reforms here were again all inspired by problems and all involved some experimentation, with quick feedback and slow formalization of interventions. We see experimentation with municipal reform, for instance, and multi-year budgeting, and civil service structures, agencification, etc. etc. The Swedes did not just jump into any of these reforms like many developing countries. They took their time, tried things out, gathered feedback, engaged multiple agents to learn, and then made interventions more formal.
The figure above shows some of the activity in the 1980s and early 1990s. You will see that it ends in 1992/3 when the country experienced a crisis that many present as the start of reform. Hopefully it is apparent that the reforms began way before this. No doubt much more happened after 1993 (and I will break some of this down in the next blog on this topic) but you need to see that the country has been involved in iterative muddling through with these reforms for a long time. New ideas build on past lessons and experiences, new groups or agents come out of the alliances developed in the past, new political support for reforms comes out of past efforts to show the viability of potential interventions, new capacities are the product of years of efforts to try new things in different ways.
This process of problem driven iterative adaptation has proved central to Sweden's reform experience and is, in my experience, the best practice we should learn from this experience.