Raising Tax Revenues on a Tight Budget: A Field Experiment in the Pampas



Policymakers may have big plans like extending a highway or starting a new social assistance program. Then again, they may have more modest plans such as meeting payrolls or maintaining existing infrastructure and levels of service. Realizing any of those plans requires a dependable stream of revenue, usually including taxes on individuals’ income and property. In an ideal world all would pay their taxes in full and on time.

The world as we know it, though, falls far short of ideal. As shown by the experience of much of the developing world, and the current Greek crisis, a large share of revenues go uncollected due to low tax morale—a culture of widespread underpayment or even non-payment of taxes. This reduces governments’ ability to provide services and meet obligations, and it can even weaken countries’ negotiating position with creditors.

Since no government has the capacity to collect all taxes and punish all tax evaders, a second best option is to improve tax morale using available resources.  But how?

Relatively little research on this question has been undertaken so far in Latin America and the Caribbean, but Carlos Scartascini of the IDB Research Department and Lucio Castro of CIPPEC have found one possible answer. With the cooperation of authorities in the Municipality of Junín, a midsized and largely urbanized district in the north of Argentina’s Buenos Aires Province, these researchers conducted an experiment involving one 2011 bimonthly bill for a property-based tax that pays for services such as street lighting, trash collection and street cleaning. The control group, consisting of most of the municipality’s taxpayers, received a standard tax bill, while three treatment groups received tax bills with messages intended to increase compliance.

The first message, emphasizing deterrence, provided an example of the cost of noncompliance by showing the cost of a hypothetical unpaid debt of AR$ 1,000 after a year, including interest and penalties. The message also warned that the municipality would take administrative and legal steps in case of noncompliance. An image of a gavel was included to reinforce the message.

A second message, emphasizing reciprocity, featured information on how the Municipality used its revenues, particularly the number of streetlights and the number of water and sewerage connections installed in the previous six months. The message was accompanied by an image of a “men at work” sign.


A third message focusing on peer effects  discussed levels of compliance, asking whether the taxpayer was aware that only three out of 10 taxpayers did not pay their tax liabilities. It then added a sentence directly questioning the taxpayer: “What about you?” This message was illustrated with an image of seven larger figures, representing taxpayers who have paid over time, and three smaller ones, representing those who have never paid.

Only one of these messages made an overall difference in taxpayer behavior. Among the group that received the deterrence message, tax compliance increased by more than 4 percentage points. This change is particularly notable considering that some individuals may not have seen the message, possibly ignoring their bills altogether. The fairness and equity messages, on the other hand, made no significant difference on average.

Within those larger trends, though, not all taxpayers reacted to the messages in the same way. Most notably, taxpayers with multiple and/or larger properties are less likely to comply than their counterparts with smaller holdings. There is no definite explanation for this pattern, but it obviously poses a challenge to local officials in terms of both uncollected revenue and the example set for other taxpayers. This represents an especially important consideration in countries of the developing world, where trust in government and other institutions is chronically low.

A single short-term study can provide only a limited amount of information. The researchers could not, for instance, learn whether the short-term spike in compliance translated into long-term changes in behavior. Other unanswered questions include whether a measure of this type could successfully be scaled up to the provincial or national level, or what messages might prove most effective in other places and for other types of tax. Further research will be needed to obtain additional data that can be analyzed to design and implement policy measures.

Still, a long journey is made up of single steps, and this represents a necessary step in finding out how to collect the taxes in Latin America and the Caribbean that allow governments to provide important services and provide a setting in which countries can realize their development potential.


Print Friendly, PDF & Email

The Author

John Dunn Smith

John Dunn Smith is the Assistant Editor of the Research Department. His degrees include a BA in Latin American Studies from American University, an AM in the same area from the University of Chicago, and an MA in International Affairs from Carleton University in Ottawa. Writing as J.D. Smith, he has published collections of poetry, essays and humor as well as a children’s book.

Leave a Reply

Your email address will not be published. Required fields are marked *

Ideas matter © 2016