Reforms of public financial management (PFM) systems – pursued by many countries and supported by development partners -- have attracted quite a bit of debate and analysis in recent years. Significant variation in progress achieved and lack of broad-based and sustained improvements in metrics of PFM performance, as reflected in CPIA ratings and PEFA scores, suggest to many observers that outcomes have not matched reform efforts and expectations.
This has led to a search for better solutions in two directions. First, grounding reform efforts in stronger problem analysis, and based on this, developing a better fit of reform approaches to specific country circumstances. Second, seeking a better understanding of non-technical aspects and, in particular, the role of political economy drivers in influencing which PFM reforms are pursued where and with what degree of success. ‘Doing things differently’ along these lines sounds promising – but reformers and development partners may well question whether we know enough to pursue such alternative approaches on a wider scale.
In a new working paper on the Drivers and Effects of PFM Reforms, we explore how much PFM performance varies across countries and over time depending on macro-social, economic, and political differences; and how much variation can overall be explained by considering such conditions, rather than being contingent on appropriate reform approaches. Such an exploration has only recently become possible due to the availability of cross-country and time series data about PFM systems through PEFA assessments. These are now available for 144 countries and a time period of about ten years. The paper builds on initial efforts in this direction carried out by de Renzio et al. (2011) for the Overseas Development Institute and by Matt Andrews of Harvard University.
A glass half full
Overall, the paper finds that about 50 per cent of the variation in the status of PFM systems can be explained by country conditions such as income level, current political stability, population size, and the degree of natural resource dependency. Income level matters most for PFM performance, but countries at comparable levels of development still have wide differences in the quality of their PFM systems. As figure 4 of the paper shows, for each of the three income groups considered – low, lower middle, and upper middle income – there is a wide dispersion in PFM performance, around an average PEFA score for each group that is about .5 higher for the highest income group compared to the lowest. Other important determinants are natural resource dependence and being a small island state (both negative) and having programmatic political parties (positive). The latter could reflect a more long-term approach and continuity in reform efforts in countries with programmatic parties, as well as a greater interest in ‘getting something done’ beyond just staying in power. Perhaps surprisingly, the paper finds that whether a country has a more democratic or more autocratic regime does not seem to matter, either for PFM performance overall, or for the specific scores addressing budget accountability.
Room for improvement
Rather surprisingly, improvement in PFM quality is highly correlated with where the country starts, but with a negative sign. Thus, countries that start off with worse PFM performance show stronger improvements over time than those with already better systems at the start. This is also consistent with the overall take-away that while country circumstances are important, they do not inevitably determine what can or cannot be achieved. How reforms are approached is likely to be important in making a difference between why some countries acquire good PFM systems ‘against the odds’ while in others, the opposite is true – an issue that we hope to explore through further qualitative work.
Overall, we hope the analysis can inform several policy and research questions around PFM. For us, the next step is to explore through a set of in-depth case studies how PFM reforms efforts have been initiated and sustained in particular countries, in order to better understand how political economy factors contribute to the fluctuation in reform efforts over time, and what might be done differently to sustain reform efforts.
Your suggestions about interesting reform experiences and ‘turns’ are very welcome!
Join the Conversation