The centre for tax analysis in developing countries

Overseas Development Institue Institue for Fiscal Studies

Attacking tax evasion in production networks: theory and evidence from Paraguay

Your word against mine: using discrepancies in taxpayers’ reports to target tax enforcement

The creation of a tax system that raises substantial revenues efficiently and equitably is one of the central challenges in economic development. In large part, this relies on the creation of the capacity to enforce taxes effectively---to reduce tax evasion. Doing this requires a detailed understanding of the drivers of tax evasion and the optimal allocation of extremely scarce tax enforcement resources. Our project combines the development of new theory and a Randomized Controlled Trial (RCT) in Paraguay to provide new insights into the nature of tax evasion by firms, the strength of enforcement spill overs through production networks, and how governments should optimally target enforcement activities.

We focus on reporting discrepancies in the Value Added Tax (VAT). The VAT taxes firms based on the difference between their sales and their purchases of materials. To reduce their tax liabilities, firms can either under-report their sales or over-report their purchases. However, when these transactions are between two businesses, this type of misreporting creates discrepancies between what the buyer and the seller report. Of course, when a discrepancy exists, the tax authority does not know whether it has come about because the seller is under-reporting, or because the buyer is over-reporting. And with scarce resources to audit taxpayers, it is not typically feasible to audit both sides of transactions.

Our project builds new theory to shed light on how differences in taxpayers’ incentives to misreport can give rise to discrepancies between reports and how these can be leveraged to target enforcement activities. We then use a series of RCTs to inform the model and, guided by the model, develop new targeting tools and test them. These tools can help tax administrations take full advantage of their limited enforcement resources and put them to maximal use reducing tax evasion.

Get your story straight: can electronic invoicing and a few, well-placed audits improve compliance?

Paraguay is an upper-middle income country in South America. Informality and tax evasion are central barriers to the development of the economy: 65% of workers are informally employed, and tax revenues are a paltry 10% of GDP (World Bank, 2017). Moreover, the tax administration has extremely scarce resources with which to enforce the collection of taxes.

However, VAT taxpayers in Paraguay make detailed reports on the identity and value of their trading relationships with other VAT-registered businesses. When both the buyer and the seller make these reports (small firms do not have to report), cross-referencing the reports allows the tax authority to detect discrepancies and act on them.

Working with the tax authority (Subsecretaría de Estado de Tributación, SET) we are using a series of RCTs to evaluate the use of targeted desk audits and electronic invoicing to eliminate discrepancies and increase tax compliance. Both audits and electronic invoicing have been shown to have the potential to improve tax compliance (see, for e.g. Kleven et al. 2011, Bérgolo et al. 2021 on audits, and Bérgolo et al. 2018 and Eissa & Zeitlin 2015 on electronic invoicing) but they are costly for the tax authority to implement. As a result, proper targeting of these costly resources is key.

To learn how best to target these interventions, we designed a two-step experiment. The first step of the experiment is designed to estimate the effects of the interventions on the targeted taxpayers, but also crucially, on their trading partners whose reports are used in cross-referencing. Then, in the second step, these direct and indirect effects are combined with our theoretical framework to learn how these interventions should be targeted at particular taxpayers to maximize their total impact on tax compliance.

Published on: 31st August 2021

Bergolo, M L, R Ceni, G Cruces, M Giaccobasso, and R Perez-Truglia (2017), "Tax Audits as Scarecrows: Evidence from a Large-Scale Field Experiment", National Bureau of Economic Research Working Paper Series 23631: 1-54.

Bergolo, M L, R Ceni, and M Sauval (2018), "Factura electrónica y cumplimiento tributario: Evidencia a partir de un enfoque cuasi-experimental", Banco Interamericano de Desarrollo, 1-25.

Kleven, H J, M B Knudsen, C Thustrup Kriener, S Pedersen, and E Saez (2011), "Unwilling or Unable to Cheat? Evidence From a Tax Audit Experiment in Denmark", Econometrica: Journal of the Economic Society 79(3): 651-692.

World Bank (2018), "Paraguay: Systematic Country Diagnostic", World Bank Group, 1-130.

Zeitlin, A, and N Eissa (2015), "Evaluation of Electronic Billing Machines: Effectiveness of Tax Administration", International Growth Centre: Project Memo, 1.


Skip to main content