Strengthening Property Tax Administration through Investments in Targeted Enforcement, Transparency, and Legitimacy

We have three main interventions. Described below.

1) Enforcement: We randomly assign owners of properties in the city of Kenema and default on their property tax to receive the “enforcement action.” The enforcement action consists of a set of four sequential steps that are part of the enforcement process. Each step is triggered by continued non-compliance after the previous step.
The four steps are:
i) sending an enforcement notice informing individuals that they might be subject to further enforcement action if they do not comply with the property tax and that they are subject to a 5% late payment penalty fee. It also informs individuals of how to pay. ii) Sending a formal court warrant to a specialized Court dealing with compliance with the property tax. This court will be led by representatives of the Paramount Chief and the Council of Chiefs in the Kenema District.
iii) On the day of the court, individuals will be reminded of the Court summon.
iv) The Court will settle the case and determine the course of action for Kenema City Council and the property owner in accordance with the Local Government Act.
For each of the steps described above, the information in the notices will be provided both in written form and in person by the individual delivering the written notification.

2) Religious sensitization campaign: We collaborate with the Interreligious Council of Sierra Leone (IRCSL) to recruit a team of religious experts representing the IRCSL and relevant religious communities. These religious experts will be assigned to visit the properties of the defaulters. During these visits, the religious experts will present a normative argument seeking to motivate tax compliance, following a previously prepared script. This script will emphasize three ideas:
A) The tax compliance is necessary to promote development,
B) That a good citizen must both comply with the tax and demand better public services, and
C) that tax enforcement is a necessary aspect of a well-functioning tax system.

Finally, the religious representatives will remind property owners how to pay property taxes.

3) Public information campaign: In collaboration with the Sierra Leone Association of Journalists (SLAJ), we are designing a campaign to inform the general public about enforcement activities implemented as part of the property tax administration of the city of Kenema. The information campaign is intended to increase awareness of enforcement activities and the new property tax administration system. We will manipulate the probability of exposure to this information via targeted distribution of local newspapers and information pamphlets.
The distribution of such types of papers and pamphlets is a common phenomenon in Kenema, and these are designed as independent media campaigns. As a result, this information will not be attributed to KCC, but rather to an independent third-party reporting on issues about the city.

2024-11-05

2025-07-31

The experiment has two main outcome variables.
The first is compliance with the property tax at the end of the study.
The second is measures of Tax Morale.

Compliance is defined as whether an individual has paid their tax liability by the end of the tax cycle defined in this study. Our primary test is therefore on the extensive margin of compliance.

Tax morale is measured through several waves of survey data collection.
Individuals are asked to rate their agreement with the following statements:

1) People have a duty to always pay their property taxes.
2) When someone does not pay their property taxes, they harm all other inhabitants of Kenema.
3) The City Council can only provide public services if everybody pays their property taxes.
4) In order to improve public services from the KCC, one must both pay their taxes and engage as a concerned citizen (examples: protest, vote, appeal to Council).
5) People have the right not to pay their property tax if the Kenema City Council failed to provide public amenities.
6) People have the right not to pay their property tax if they were not doing well financially.

As part of our secondary analysis, we have four sets of secondary outcomes: 1) perceptions of the local government’s capacity for enforcement, 2) perceptions of the state legitimacy and quality of governance, and 3) acceptance of enforcement as part of the tax system, and 4) substitute behaviours to tax compliance. Each of these measures is explained more in the following section.

We measure perceptions of the local government’s capacity for enforcement by asking the following questions:
1) Consider a typical citizen of Kenema, Mohamed. Imagine that Mohamed is a property owner. Now suppose that Mohamed does not pay their taxes during the current tax cycle. How likely or unlikely do you think it is that he will be penalized by the City Council?
2) Now I want you to consider a wealthy citizen of Kenema. Say, for example, Abdul. Abdul is also a property owner in Kenema. Suppose that Abdul does not pay their taxes during the current tax cycle. How likely or unlikely do you think it is that he will be penalized by the City Council?

We measure perceptions of the state legitimacy and quality of governance by asking the following questions:
1) The Kenema City Council (KCC) is supposed to develop the city and provide the people of Kenema with services to benefit everyone. How satisfied are you with the Kenema City Council? (Possible answers are a likert scale from “very unsatisfied” to “very satisfied”)
2) Please tell me if you think the Kenema City Council is “responsive” to your needs or not. By “responsive” I mean that the KCC takes action to address things that people want. (Possible answers are a likert scale from “very unresponsive” to “very responsive”)
3) The Kenema City Council is responsible for handling money that they need to use for public administration and development. Some people accuse the KCC of stealing money and acting corruptly. Other people say that the KCC is clean and doesn’t engage in this corruption business. Please tell me if you think the Kenema City Council is: (Possible answers are a likert scale from “very corrupt” to “very clean”)
4) We know that the Kenema City Council is responsible for handling money that is used to pay for public administration and development. How do you think the KCC handles the money? (the question is whether the KCC steals the money, but how it spends it) (Possible answers are a likert scale from “very wasteful” to “very efficient”)

We measure acceptance of enforcement action by directly asking respondents to rate agreement with the following sentence:
1) The City Council always has the right to penalize those that do not pay their property taxes.

Finally, the Kenema City Council provides access to an “Enquiries and Appeal Desk” where individuals are both able to ask questions about their tax liability, the general reform and present complaints about their specific cases. Individuals can potentially ask to change the administrative data regarding their properties, potentially changing their tax liabilities in future tax cycles. Thus, we argue that this may be an example of an alternative response to tax enforcement policy compared to paying the tax. We measure whether a property owner visits this Appeals Desk, using administrative data from the Kenema City Council that records both the purpose of the visit as well as the property ID for which the property owner is appealing.

Our experiment is conducted in two phases.

In the first phase, we will implement the two interventions, “enforcement” and “religious sensitization” following a 2x2 factorial experimental design. Under this design, each property (and their respective property owners) will be assigned to one of four treatment conditions: a pure control condition, an enforcement condition, a religious sensitization condition, and an interaction condition in which the individual is exposed to both the religious sensitization and to enforcement. Upon completion of the intervention, we will conduct another round of data collection to measure all key indicators.

The second phase of the intervention will commence upon completion of the new round of data collection and will consist of the public information campaign. We will randomly assign properties (and their respective property owners) to receive the paper/pamphlet detailing information about the previous enforcement actions that were implemented during the previous cycle as well as the different aspects of the reform. We then again measure both attitudes (through another wave of surveys) and compliance using administrative data.

Finally, at the end of the fiscal cycle, we measure these outcome variables again to estimate the long-term impact of the interventions.

Not available

For all properties that participate in the study, each observation will be randomly assigned to each intervention through a computer algorithm. For example, for the factorial experiment, for each property, we make a draw out of a uniformly distributed variable that can take one of four values. Each value is associated with one treatment condition. Similarly for the public information campaign, each property is assigned to receive the pamphlet based on a draw from a Bernoulli distribution.

We selected the experimental sample by first clustering all of the defaulter properties in the city based on geographical proximity to each other. Each cluster consisted of approximately 40 to 60 properties. Afterwards, we stratified by cluster.

Once we had the experimental sample, we simply randomized at the level of the individual property. Since, we only account for one property owner within each property, this is equivalent to randomizing at the level of the property owner.

In the public information intervention, pamphlets will also be randomized at the property level.

No

6357 properties;
Sample size of the public information intervention not yet defined.

6357 properties; Sample size of the public information intervention not yet defined.

1658 properties control;
1595 properties enforcement action;
1559 properties religious action;
1545 properties both religious action and enforcement action
Sample size of the public information intervention not yet defined.

Impact on compliance: - Expected standard deviation in control: 0.9 - Impact of the enforcement action relative to control group: 3.4 percentage points higher compliance in enforcement relative to control group. - Impact of the religious action relative to control group: 3.4 percentage points higher compliance in enforcement relative to control group. - Impact of joint religious and enforcement action relative to enforcement only: 3.9 percentage points higher compliance in the join group relative to the enforcement only group. Impact on Tax Morale - Expected standard deviation in control varies depending on the exact tax morale measure. However, it ranges from 0.09 to 0.24. - Impact of the enforcement action relative to control group: 4 percentage points higher in agreement in the enforcement group relative to the control group. - Impact of the religious action relative to control group: 4 percentage points higher in agreement in the religious group relative to the control group. - Impact of joint religious and enforcement action relative to enforcement only: Between 3.5 and 4.5 percentage points higher in agreement in the joint group relative to the enforcement group.