Property tax and waste management in Kanifing, the Gambia

In partnership with the Kanifing Municipal Council, we will leverage the ongoing adoption of a robust addressing system to develop new digital monitoring systems and improve property tax and waste collection.

Open Location Codes technology is a comprehensive digital addressing system covering the globe (https://maps.google.com/pluscodes/). This technology is open source, simple and free to use, and can be accessed offline. It allows households to receive mail, access emergency services, and verify their residency with public administrations and private entities like banks.
The Kanifing municipal council adopted Open Location Codes in 2021. These codes were linked to a georeferenced database and printed on the front-door of 35,000 compounds of the municipality (almost all of them except slums).

We will use Open Location Codes to develop a new property tax monitoring system in three steps. First, municipality agents will identify Open Location Codes of existing taxpayers. Second, tax data matched with the Open Location Codes data will be used to create non-compliance maps at the neighbourhood level (sub-ward level) showing unregistered, unvalued, and non-compliant properties. Finally, these non-compliance maps will be used to conduct systematic monitoring at the neighbourhood and property level.
This will include door-to-door tax-bill distribution as well as individually targeted enforcement for non-compliant taxpayers (tax agent visits, letters, and fines).

In addition, Open Location Codes will be used to improve the monitoring of the M’Balit waste collection program. M’Balit garbage trucks were recently equipped with GPS devices. These data are currently only used by the municipality to signal and manage truck breakdowns. We will design an automated system to match trucks monitoring data with Open Location Codes data and to identify which properties are served every week by the M’Balit program. By comparing these properties with the list of properties that should be served according to trucks targets, we will construct a reliable, trackable and high frequency measure of the quality of waste collection at the truck level. We will use this innovation to improve monitoring of the M’Balit teams and to implement a pay-for-performance system.

Finally, we will provide tickets for free M’Balit waste collection services to a sample of taxpayers, together with their tax notice (conditional on paying). This incentive aims to motivate tax compliance by linking payment to improved services. This could strengthen the social contract between the municipality and its residents. We hope to reinforce the reciprocal relationship between public revenues and the provision of essential services like waste collection.

2025-01-15

2026-02-28

Tax payment: amount of property tax paid, and the share of taxpayers who paid at least partially. From administrative tax data.
Frequency of waste collection: from GIS truck data.
Engagement with local institutions and leaders using survey data (see pre-analysis plan).
Waste burning and illegal dumping: high-frequency survey data
Air quality (PM2.5): high-frequency data from air quality monitors
Health status: prevalence of respiratory disease from survey data (see pre-analysis plan)

Mechanisms for tax compliance:
- Share of taxpayers receiving a demand note from both administrative and survey data.
- Perception of tax legitimacy: measured using survey data (see pre-analysis plan).
- Perception of state capacity: measured using survey data (see pre-analysis plan).
- Tax knowledge: measured using survey data (see pre-analysis plan).
- Perception of tax enforcement: measured using survey data (see pre-analysis plan).
- Perception of local tax compliance: measured using survey data (see pre-analysis plan).
- Tax morale: measured using survey data (see pre-analysis plan).
- Citizens’ perceptions that property tax revenues are well used: measured using survey data (see pre-analysis plan).
- Perceived corruption in the taxation process: measured using survey data (see pre-analysis plan).

Impact on other outcomes:
- Interactions with local administrations: measured using survey data (see pre-analysis plan).
- Attitudes toward local institutions: measured using survey data (see pre-analysis plan).
- Political participation: measured using survey data (see pre-analysis plan).
- Citizens’ engagement with leaders for topics other than taxation.
- Street cleanliness: high frequency survey data (photos).
- Perception of waste collection services: measured using survey data (see pre-analysis plan).
- Willingness to pay for waste collection services: measured using survey data (see pre-analysis plan).
- Household survey measures of perception of air quality, street cleanliness, and waste burning and illegal dumping (see pre-analysis plan).

We will also study the distributional effects of the tax by taxpayers' income, wealth, property value, and liquidity constraint (see the pre-analysis plan).

The experimental design includes the following components:

(1) What is the impact of tax compliance maps on tax payment?
Compliance maps will be created using QField and tax data matched with the Plus Code addressing system.
Design: 113 subwards randomly allocated between a Treatment group in which the tax administration will use tax compliance maps, and a control group.

(2) What is the impact of better trash collection?
224 blocks of properties with low access to trash trucks (according to Trucks GIS data), randomly allocated between a Treatment group receiving better trash collection and a control group where collection continued as before.

(3) Does a 30% tax reduction impact tax compliance?
1000 taxpayers randomly selected to receive a 30% discount on their tax bill.

(4) How do information and trash vouchers given with demand notes impact tax compliance?
1000 taxpayers randomly selected to receive a message on their demand note explaining that if they pay their tax, they will receive tickets for free M’Balit waste collection services worth D300. The tax bill mentions this incentive and the fact that property tax revenue funds waste collection services.

(5) Is it the impact of trash vouchers or just the information component?
1000 taxpayers randomly selected to receive a message with the same information that property tax revenue funds waste collection services, but without the free tickets as an incentive. The goal of this intervention is to separate the impact of the free tickets from the information that property tax funds waste management.

(6) What is the impact of summons?
1000 taxpayers randomly selected to receive targeted additional enforcement, among those who received a tax receipt but have not paid yet in May and August 2025.

Not available

Randomization done in office using a computer.

For intervention (1) the unit of randomization is the subward.
For intervention (2) it is the block (on average 20 properties in a block)
For interventions (3) to (6) is it the property level.

Yes

Intervention (1) 113 subwards
Intervention (2) 224 blocks

Intervention (1) 35,000 properties Intervention (2) 1,800 properties Intervention (3) 11,000 properties Intervention (4) 11,000 properties Intervention (5) 11,000 properties Intervention (6) 4,500 properties

Intervention (1)
T1 (Compliance maps): 56 subwards and 17,591 properties
C1 (no compliance maps): 57 subwards, 17,396 properties

Intervention (2) 2,000 properties
T2 (trash collection): 112 streets, 945 properties
C2 (business as usual): 112 streets, 842 properties

Intervention (3)
T3 (30% lower tax rate): 1,000 taxpayers
C3 (tax rate as usual): 10,000 taxpayers

Intervention (4)
T4 ((Vouchers + information): 1,000 taxpayers
C4 (no information): 10,000 taxpayers

Intervention (5)
T5 (only information): 1,000 taxpayers
C5 (no information): 10,000 taxpayers

Intervention (6)
T6 (summons): 1,000 taxpayers
C6 (enforcement as usual): 3,500 taxpayers