Asymmetry in Civic Information: An Experiment on Tax Participation among Small and Informal Firms in Togo
How to cite
Blimpo, Moussa; Dower, Paul Castaneda. 2019. "Asymmetry in Civic Information: An Experiment on Tax Participation among Small and Informal Firms in Togo." AEA RCT Registry. AEARCTR-0004394. https://doi.org/10.1257/rct.4394-2.1
Download: BibTeX · RIS — for reference managers without a browser translator. Zotero and most others pick up this page's citation metadata directly.
* Primary investigator
Abstract
The often observed wedge between statutory policies and on-the-ground enforcement practices in low-income countries for the taxation of small, informal firms may stem from an asymmetry of information about the law. We test this hypothesis in a randomized controlled trial in Lome, Togo by training firms about the tax code and assessing the effect on their tax participation and economic activities. We find that tax participation decreased, driven bylow-revenue firms, and their economic activities expanded. Overall tax revenues increased, suggesting that aggressive taxation under asymmetric civicinformation is likely regressive and counterproductive for firms’ activities andgovernment revenues.
Intervention
The intervention consisted of the provision of information and training about
tax obligations and benefits of taxation to firm owners and society at large. The information included was based on the tax code as well as our interviews with senior tax officials concerning information that they would like these firms to know. Along with information about their tax obligations, the treatment also consisted of what taxes pay for and the cost of these services.
Experimental design
Many of the firms in the informal sector do not have physical addresses. Additionally, Togo does not have a recent firm census that could serve as a sampling frame. We resorted to a sampling method used by the Afrobarometer surveys, which uses a well-designed shoe leather and sampling on the ground. We used the census tracts as enumeration areas. For each enumeration areas, we randomly select a starting point on the map. Four enumerators then go to the nearest junction to the starting point
and choose five businesses in each of the four directions by counting businesses that they encounter on both sides of the streets and selecting every fifth. This first database is then used for the random assignment to the treatment or the control group.
Design details
Cited by
No citation data available for this registration yet.