This study explores the relationships between country characteristics and PFM quality, and between the quality of PFM systems and expected outcomes. It finds that country characteristics do not have a strong predetermining effect on the quality of PFM for most countries. And it finds evidence that stronger performance results in better budget credibility, but not in lower deficits. There is no clear evidence regarding operational efficiency.
The following characteristics of a country were investigated: economic factors (level, growth and resource dependency), population (including SIDS), levels and sources of revenue (tax revenue to GDP, the level of aid relative to GNI), and three macro-political characteristics – political stability, regime type, and the presence of programmatic parties.
Key findings:
- It was found that the quality of PFM systems is most significantly and robustly associated with a country’s income per capita and growth (positively) and having a high share of revenues that are obtained from natural resources (negatively). Statistically weaker associations exist with regards to the size of the overall population (positive) and being a small island developing state (negative).
- With regards to the political regime and tax revenue, the statistical significance as well as the potential impact is weak; while programmatic parties appear to possibly have a substantial impact, but again, the statistical significance is borderline (at the 10 percent level).
- There is a strong negative association between HIPC countries and the quality of PFM; in this regard it is assumed that the more likely direction of causation is from weak PFM systems to becoming highly indebted.
- There is no association with the volume of aid relative to GNI.
- Country characteristics do not appear to have a strong predetermining effect for most countries, and particular reform approaches including political economy considerations are likely to matter for reform success.
- However, for countries that combine being low income with being small island states or with being dependent on natural resources (or both) these characteristics appear to be relatively stronger constraints.
- In a second step, a first differences method is employed. In this model, except for SIDS, the study did not find any evidence that country characteristics pre-determine progress on the performance of PFM.
- Initial level of PFM quality is the only other statistically significant variable, with a negative sign. This indicates that countries with worse initial PFM systems have tended to achieve a greater degree of reforms.
- A strong negative effect of the change in population becomes significant, suggesting that countries with faster population growth have done worse with regards to improving the quality of their PFM systems.
- Resource dependence does not show a significant relationship in the first difference analysis using PEFA data, and a weak and barely significant relationship using CPIA data, this implies that being resource dependent does not prevent a country from progressing with PFM reforms relative to progress seen in other countries.
- Both the cross-sectional and first-difference findings suggest that to some extent, differences in levels and changes in PFM performance are associated with country characteristics; while only a few variables combine high statistical significance (at the 1 percent level or higher) and a tangible impact.
- Investigation into the relationship between PFM quality and key fiscal outcomes, including cost-effectiveness in the use of funds for public service delivery (technical or operational efficiency) found that better public financial management is associated with lower variances in overall budget execution rates as well as cross-sector variance (robust when controlling for GDP per capita), but not with lower deficits.
- There is no correlation between PFM quality and public service delivery outcomes in health or education relative to these sectors’ fiscal allocations—that is, available data do not show a significant relationship with the operational efficiency of spending.