Policy Issue:
Revenue collection and public sector efficiency is a central question for developing countries. The low level of tax revenues raised in these countries can result in the under-provision of public goods, heightens vulnerability to economic crises, and may constrain growth. Poorer countries collect on average only two-thirds or less of the amount of tax revenue as a fraction of GDP that richer countries do1, with an estimated $285 billion per year loss due to tax evasion. Past work on this topic attributes the weak performance of tax collection to the poor incentives for proper tax collection and administration. It is thought that incentives, accountability and monitoring can raise public sector efficiency, but there is little rigorous evidence on anti-corruption measures that systematically address the incentives faced by government bureaucrats.
Context of the Evaluation:
Punjab is Pakistan’s most populous province with a population of over 80 million. International comparison reveals that the present level of property tax collection in Punjab is roughly a fifth of the level of comparable countries. Evidence suggests that the main way tax evasion takes place is through several distortions such as granting exemptions to widows, the disabled, owners of small plots, retired federal and provincial government employees, and religious charitable institutions. Because officials may employ significant discretion in applying valuation to individual properties and determining exemptions, the system leaves considerable opportunities for leakages, collusion, and low collection.
Tax collectors in Pakistan are part of the provincial career bureaucracy with wages determined by salary band and length of service. Tax officials have few avenues for vertical mobility through promotions, and wage levels are not tied to performance in any way. As a result of these factors, tax officials suffer from low motivation, and evidence suggests that tax evasion and rent seeking are prevalent.
Details of the Intervention:
Geographical areas serviced by a set of tax collectors (called tax circles) were randomly assigned one of the following four wage and incentive programs (or no treatment):
Wage schemes:
Pure Wage Increase: the base salary of tax inspectors, clerks, and constables will be increased by a fixed amount. The usual monitoring and control systems from the Excise and Taxation Department (which also apply to the control group) remained in place in this and all other treatments.
Pure Wage Increase Plus Audit: In addition to the increase in wages, tax officials were told that a random sub-sample of the properties under their jurisdiction would be visited by an independent government unit outside the tax department and audited. Officials were told that they would be rewarded or penalized based on the accuracy of their audit relative to audits in other circles, as well as on taxpayer satisfaction with the quality of interactions with tax officials. The reward / penalty will be in the form of job assignments: the best performing circles would be given the highest priority in postings in the subsequent year, and the worst circles will be given lowest priority in reassignment in the subsequent year and denied the wage increase.
Output-Based Incentives: In this scheme, tax officers were rewarded on the basis of revenue collected. Specifically, a percentage of revenue collected in the circle above a historical benchmark will be given to the tax officers as performance honorarium.
Output-Based Incentives Plus Audit: In this scheme, the output incentives were combines with the audits, with the penalty that if the audit performance is good/poor the circle officials are given highest/lowest priority in reassignment to tax circles and forfeit future incentive payments.
Results and Policy Lessons:
Results forthcoming.
