How can countries improve tax collection? Lessons from Liberia

“Out of all the people in Monrovia, how will the government know that I have not paid?” This was the question from a participant in a focus group discussion I held with property owners in Monrovia about their experiences with real estate tax in Liberia. This property owner, and many others, were well aware that the government was barely collecting real estate taxes across the city. 

Liberia, like many low- and lower-middle-income countries, faces serious challenges in collecting taxes. At the time this project began, a staggering 95 percent of residential properties remained outside of the tax net . It was difficult to identify the tax base because there was no comprehensive database on property locations, ownership, and values. Even for properties on the tax register, there was difficulty in collecting the taxes owed from year to year due to weak administrative and enforcement practices. 

Taxes lie at the crucial intersection between the provision of public goods, income redistribution, social safety nets and government accountability. Building capacity to collect taxes efficiently and equitably is an important policy goal  for governments. While this project focuses on real estate tax in Liberia, similar challenges with identifying the tax base (detection capacity) and in collecting tax liabilities (enforcement capacity) are common across other tax categories and across countries. 

I had the opportunity to work with the Liberia Revenue Authority (LRA) to study the impact of detection and enforcement capacity in tax compliance in Monrovia, the capital city. First, the LRA developed a new property database for a pilot community. Youth were trained to use open-source software on tablets to capture the location, ownership, photograph, and GPS location of properties. Starting in 2014, the project covered about 1,000 properties and expanded to over 40,000 properties in later years. With the collected information, tax authorities were able to fill a massive gap in the tax authority’s ability to identify properties that were outside the tax registry. Figure 1 shows an aerial photograph of a section of the city.

Figure 1: Satellite Image from Monrovia, Liberia

Next, using this newly established database, we conducted a randomized experiment to examine the impact of different approaches to increase compliance among property owners whose properties were not on the tax register. One group received a generic notice that provided information on property tax requirements (Control). A second group received a similar notice, but it was personalized with the photograph of the property and the name of the owner (Detection). A third group received a similar notice as the Control but it included information on penalties for non-compliance (Penalty), and a fourth group received a notice that included the features of both the Detection and Penalty notices (Detection and Penalty).

Tax payments barely budged in the Detection and Penalty treatment groups. However, the combined Detection and Penalty treatment group paid significantly more in taxes, reaching almost 10 percent payment rate compared to the 2.2 percent payment rate of the control group (Figure 2). The increase in property tax payment persisted even four years after the experiment, and the additional tax revenues in the first year alone more than covered the cost of establishing the property database and sending notices.

Figure 2: Impact of Detection and Penalty Notices in First Experiment

This result reveals that people’s decision to pay is affected by the combined knowledge that the tax authority is aware of their non-compliance and people’s awareness of the associated penalties . However, a puzzle remains. Whereas standard tax models would predict that the Detection and Penalty notice would lead to extremely high rates of compliance, despite the increase noted above, the vast majority of recipients did not pay. Thus, we used a second experiment to test the hypothesis that many people did pay because they believed that even though there were stated legal penalties, these penalties would not be applied to them.

For the second experiment, the sample was delinquent taxpayers who were on the tax register but had not paid for at least one year. The control group received a notice similar to the Detection and Penalty notice in the first experiment.  The treatment group received a notice that, in addition, alerted them that they had been selected for intensive follow-up if they did not comply (Enforcement notice). This additional message to increase the perception of property owners that they will face the consequences for non-compliance increased the tax payment rate from 20 percent to 27 percent (Figure 3). 

Figure 3: Impact of Enforcement Notice in Second Experiment

Taken together, these two experiments highlight the importance of both detection and enforcement capacity for collecting taxes. In much of the existing economics literature on tax compliance, enforcement is assumed and the only problem the tax authority needs to solve is identifying the tax base. However, information alone does not collect taxes . In many countries, tax authorities are unable to collect large amounts of tax liabilities even though they are aware of these liabilities. In order to ensure tax compliance, detection and enforcement capacity must work in tandem .

Improving detection and enforcement capacity requires different types of investments. Investing in detection capacity typically involves deploying information technology solutions to collect information on the tax base, such as creating taxpayer databases and linking third party information from employers, trade partners, government agencies, and financial services providers.  In contrast, investing in enforcement capacity typically involves working with actors outside the tax authority, such as the judicial system, to expand and implement the range of sanctions that the tax authority can take against delinquent taxpayers. In addition, enforcing sanctions typically requires high level political support. Although investing in detection capacity is often politically easier, it may have limited impact on improving compliance if not paired with enforcement capacity to maximize the newly available tax information. By identifying the constraints in the system and responding with the appropriate investments governments are better poised to improve tax compliance.