Development that Works
  • About

    This blog highlights effective ideas in the fight against poverty and exclusion, and analyzes the impact of development projects in Latin America and the Caribbean.
  • Who cares if you don’t pay your taxes?



    By Carlos Scartascini

    scartascini coverTax evasion is endemic in many countries. In particular, some developing countries like Argentina, the Dominican Republic, Ecuador and Guatemala do not collect even half of what they should if taxpayers complied with the written letter of the law, according to recent studies summarized in Corbacho, Fretes Cibils, and Lora (2013).

    This problem is sometimes compounded at the local level because the capacity to enforce the law is even lower. As local governments are given more and more expenditure responsibilities and central governments struggle to finance the budget, the need to increase revenues becomes more pressing for subnational authorities.

    Why Do People Pay Taxes?

    In order to understand why people pay or do not pay taxes we need to understand the underlying motives behind compliance.

    The literature has been very prolific on this issue. People tend to comply with the law when the expected utility or gains from complying surpass those from evading. In this computation, taxpayers compare the benefit from pocketing the money that should have been sent to the authorities versus the potential losses from being caught and having to pay interests and fines on the other.

    Of course, in the same way that people do not steal or cross a traffic light inappropriately even if the chances of being caught are zero, social norms and morals also play a role in people’s decisions to pay taxes.

    Otherwise, compliance would be much lower. Consequently, people tend to comply more if they believe it is the right thing to do. This belief tends to be reinforced if people think that being taxed is fair, and equitable. That is, if people believe that taxes serve the common good, they are being treated fairly by the government, and compliance is a prevalent behavior in society.

    Are there any solutions to this problem? 

    Plenty of specific solutions have been offered in the literature and many more have been tried by public authorities. Increasing and focusing enforcement, improving the collection and management of information, reducing the transaction costs of complying with the law, providing incentives for those who comply, and modifying tax bases and rates are some of the obvious examples.

    In some countries, the regular avenues for boosting tax compliance are largely dead-ends: the capacity to enforce the law by engaging in massive auditing or taking massive number of people to court is relatively limited, legislative changes are difficult to come by, and some of the more standard measures such as reducing transaction costs have already been taken. In this context, efforts to affect people’s beliefs and morals may be a worthwhile initiative.

    The evidence out there for developed countries is vast and shows that messages tend to affect people’s behavior and increase compliance. For example, in Minnesota (here and here) the mean increase in federal declared taxable income for the group receiving a letter designed to correct the erroneous perception of many taxpayers that cheating on taxes was common was $2,390. The results in Australia are similar (here and hereungated), and in Denmark and Austria positive effects have been found on self-reported income. The “Nudge Unit” at the UK Treasury has systematized and popularized the use of this type of instrument in the public sector.

    The evidence for developing countries —where trust in government is lower and tax evasion is higher— is scarce. Applying these techniques for the collection of a property tax in a subnational unit has not yet been documented. In a recent working paper, we (me and Lucio Castro from CIPPECC) explore the impact of the use of messages for affecting taxpayers’ compliance with a property tax in a developing country by conducting a large-scale field experiment.

    In the experiment, conducted in a municipality in Argentina, approximately 23,000 taxpayers of a tax levied upon individuals according to the size of the property and the services they receive from the local government, such as street lighting, trash collection, and street cleaning, were randomly divided into 4 groups. One of the groups received no treatment (the control group); the other 3 were treated by including messages in their tax bill. 

    scartascini F1

    The treatments were designed to affect people’s beliefs about deterrence (or beliefs about enforcement and fines), equity (or beliefs about other taxpayers’ behavior), and fairness (or beliefs about the use of resources by the government).

    The results indicate that introducing messages in the tax bill may be a useful instrument for affecting taxpayers’ behavior because those taxpayers who received the messages tended to behave differently than those who didn’t.

    The most effective message was the one on deterrence that listed the actual fines and potential administrative and judicial steps at the disposal of the municipality in the case of noncompliance. More precisely, tax compliance among the taxpayers that received this deterrence message increased by almost 5 percentage points with respect to the control group (which is equivalent to reducing tax evasion by more than 10 percent).

    We found no average treatment effects for the other two messages, which listed the investment works realized by the Municipality in the previous six months (fairness), and the average compliance in the Municipality (equity) (as shown in this Table).

    scartascini F2

    Beware of Universal Policies

    Interestingly, we found some heterogeneous effects across the population, which indicates that not everybody reacts to the messages in the same way. Own compliance behavior, the level of provision of public goods, and wealth seem to influence the effectiveness of the messages. For example, people with lower levels of wealth seem to react more to the messages than those with higher levels of wealth (who may not even react at all).

    This suggests that message targeting may be advisable; messages could be tailored by varying the fines according to the wealth levels of the taxpayer.

    Also, some of the messages had a negative effect on some people, which led to lower levels of compliance. For example, some taxpayers who used to comply stopped doing so after they learned about the actual rate of compliance in the population (people may have realized that fairness was actually lower than what they had previously thought)

    . This evidence provides further support for targeting the messages instead of sending common messages or running universal campaigns through newspapers or other media outlets.

    Parting Words

    Many nuances and details about the relevance of this research agenda, how to implement these types of policies, and several important results have been left out of this note for space reasons but they can be followed here. Hopefully, the policy implications have been duly highlighted.

    First, it is important to adequately manage the many opportunities available to policymakers to influence citizens’ beliefs because this cheap policy alternative could yield substantial benefits for the government. In the case of taxation, and the fight against tax evasion, this policy alternative minimizes administrative costs, a feature sometimes ignored –at least by the academic literature.

    Second, the messages should be appropriately designed to have a positive effect on taxpayers.

    Third, “universal” policies may backfire, as positive and negative behavioral responses cancel each other out. Therefore, policies (particularly nudges) should be tailored to taxpayers’ types. Finally, it is important to consider how to make any of the compliance (and revenue) gains permanent. On the one hand, any efforts to change beliefs through the use of messages should be accompanied by equivalent actions by government authorities such as actually stepping up enforcement

    . On the other, complementary mechanisms should be introduced to facilitate compliance among debtors who are willing to pay, and to reinforce the positive effect of complying for those who changed behaviors.

    In this short video we discuss some of the implications of our results.

    * This post is based on an article written for the IMF’s Public Finance Management Blog entitled “Reducing Tax Evasion in Developing Countries

    Carlos Scartascini is a Principal Economist at the Research Department of the Inter-American Development Bank where he specializes in Political Economy, Fiscal Policy, and Institutional Analysis. He is currently leading several projects on taxation, including field experiments for reducing tax evasion, and political economy analyses of tax policy and tax reform. He is the co-editor of several books, and has published in major academic journals. Links to his work can be accessed at:


    Comment on the post