Public Spending Efficiency determinants New evidence using Simar and Wilson two-stage efficiency analysis
Description
Data from 147 countries for the years 2010 and 2022 gathered from various sources, allow the calculation of public spending performance (PSP) on the basis of opportunity and Musgravian indicators. Public spending Efficiency (PSE) scores are then derived using Data Envelopment Analysis using the level of public spending (as % of GDP) as input and PSP as output. Simar and Wilson (2007) approach is finally employed to analyze the determinants of public spending (in)efficiency using a host of environmental variables. The results indicate that inefficiency averages 31% in 2010 and 28% in 2022, meaning that countries have reduced inefficiency by 3% over the period, mainly by middle-income countries, and could reduce it further. Economic complexity, financial development and trade openness are shown to be powerful efficiency enhancers, in contrast to natural resources rents. The benefits of private investment are inconclusive, while debt and FDI inflows are insignificant, indicating the complexity of the relationship between these factors and PSE, and highlighting the need for targeted development projects, better regulatory frameworks, and accountability mechanisms to reap their benefits.
Files
Steps to reproduce
As a first phase, we construct the Public Expenditure Performance Index (PSP) using the methodology proposed by Afonso et al. (2003). This entails using a set of Opportunity and Musgravian sub-indicators. To facilitate meaningful comparisons between countries, the 14 sub-indices are normalized to constrain all data between 0 and 1. These sub-indicators are then merged into the seven indicators, and finally PSP is calculated as the simple average of the seven sub indicators, by giving a weighting of 1/7 for each. PSE scores are estimated by following the DEA approach using PSPs as outcome and public spending (as % of GDP) as input, adopting the output orientated approach, to measure the proportional increase in outputs while holding input constant and assuring variable-returns to scale (VRS) to account for the fact that countries might not operate at the optimal scale. Finally, the command Simarwilson in Stata is employed, under Algorithm 1 and Algorithm 2, to account for the impact of environmental variables.