Building tax capacity at scale: Evidence from technology investments in Ghana

Increased capacity to collect taxes leads to the funding of public goods which have high returns and contributes to the wider state-development process (Schumpeter, 1918; Besley and Persson, 2010). At the heart of any strong tax system lies the ability to accurately observe economic activities and use this information as inputs into an efficient billing, collection, and enforcement process. Constraints on these two capacities can have large adverse impacts:

• It can drive a wedge between the official, statutory policies, and the effective policies which are implemented.
• It increases the discretionary power of intermediaries (e.g. local tax collectors in the field).
• It increases the cost of collection and can decrease citizen trust and engagement with the wider tax system.

In the context of Ghana, where local property taxation is constrained along these two dimensions, strong adverse impacts in our baseline survey can be found. 76% of tax payments are made according to presumptive schedules, which are potentially regressive, rather than official, statutory tax schedules. The average cost of collection is 55% of revenue (compared to 0.2% in the US). Also, the majority of citizens report low trust in local tax officials and lack awareness of the local tax and public expenditure setting.

Appropriately designed information technologies (IT) have the potential to alleviate these capacity constraints. However, despite technology investments holding the promise to significantly improve tax and broader state capacities, there is no well-identified evidence on the returns to such investments.

This project will contribute towards filling this gap by presenting evidence from a large-scale experimental evaluation of the revenue collection and broader state-building impacts of improved technology in the local tax setting in Ghana. In collaboration with the Ministry of Finance and the Ministry of Local Government, the order in which 64 local governments are offered a domestically-developed IT system to collect property and business taxes is randomised. The randomisation is done at the level of entire local governments, covering 7.25 million people. The impacts of technology on revenue collection performance, cost of collection and leakage, citizen engagement, and public goods expenditure will all be assessed.