Publication - Working Paper
Publication - Policy Brief
Making moves matter: Experimental evidence on incentivising bureaucrats through performance-based transfers
Many developing countries suffer from low levels of tax collection due to a mix of poor enforcement, tax evasion, and corruption. As a share of GDP, tax revenues are 45% lower in poor countries than in developed countries which has negative ramifications for public services. At the same time, governments face constraints in providing incentives to civil servants,...
Tournaments for postings: Using transfers and postings as an incentive for property tax inspectors in Pakistan
This randomised control trial examined the impact of incentivising tax collectors in Punjab through job transfers and postings.
Offering tax collectors a desirable non-monetary incentive – a merit-based opportunity to be transferred to a better location through a tournament – substantially raised tax revenues. For the treatment group in Year 1, the growth rate of tax revenues was 41% higher than in the control group.
- While tournaments as a performance incentive can be effective, applying them too often may be counterproductive. The researchers found that the effects of the postings disappeared in the second year if a tax collector was subjected to the tournament two years in a row.
This study was part of a long collaboration between the Government of Punjab and the research team on improving tax compliance in Pakistan. The first stage of the project involved the provision of performance pay to property tax collectors to incentivise increased tax collection while retaining/raising customer satisfaction. The results from this study were positive and indicate that performance pay can be effective in increasing tax revenues and in generating a positive return on investment.
This project sought to take forward lessons learned and test complementary interventions and focused on implementing and evaluating merit-based job transfers and postings schemes. The design of this intervention was a randomised control trial utilising a ‘dynamic incentives’ treatment that assigns tax collectors job postings for the following year through a tournament. Property tax collectors participating in the study were assigned to groups and their rank within their group was determined by their performance in raising tax collections. At the end of the year, the first ranked collector in a group is assigned to his most preferred posting, the second ranked to his preference among the remaining posts, and so on.
The transfers and postings intervention began with a small pilot in July 2013 that was later expanded. This intervention was designed in collaboration with the Punjab Government and in particular the Punjab Excise and Taxation Department, which have been enthusiastic collaborators throughout the study.
The results of this intervention will assist Punjab’s Excise and Taxation Department in developing an optimal human resource policy for its tax collectors. Overall, the lessons learned from this study will help improve tax compliance in Pakistan and in other developing countries.