Information based tax enforcement measures on car tax evasion: evidence from field experiment

How to cite

Kosonen, Tuomas; Slemrod, Joel; Harju, Jarkko. 2017. "Information based tax enforcement measures on car tax evasion: evidence from field experiment." AEA RCT Registry. AEARCTR-0002591. https://doi.org/10.1257/rct.2591-1.0

Download: BibTeX · RIS — for reference managers without a browser translator. Zotero and most others pick up this page's citation metadata directly.

Tuomas Kosonen* , Joel Slemrod , Jarkko Harju

* Primary investigator

Registry ID
AEARCTR-0002591
DOI
https://doi.org/10.1257/rct.2591-1.0
First registered
Last updated
Status
Completed
Trial dates
2014-01-01 – 2017-11-20
Countries
Finland ()
JEL codes
H21, H23, H26, C93
Registry record
http://www.socialscienceregistry.org/trials/2591

Abstract

We study a tax evasion response to car taxes in Finland, where used car importers overstate the mileage to reduce the taxable value of the car. In the overall analysis we study many aspects of this phenomenon. We develop a tax evasion measure by comparing reported mileage information upon import with subsequent third-party information from vehicle inspections, and find that a decline in mileage — “missing miles” — is a common phenomenon. In the RCT we vary the salience of tax enforcement measures. For example, we inform in one treatment potential car importers that third party information is now more widely used in tax enforcement. The RCT was designed in co-operation with Finnish Customs. In the experiment different treatment letters were sent to potential car importers in the name of Customs. The baseline population in the treatment consisted of individuals who had imported a used car to Finland before. Four letter groups were randomized from this baseline population.

Intervention

The intervention is that Finnish Customs sent out information letters, and also may have conducted a tax audit to some car importers.

Experimental design

Four treatment letters were sent out by the Customs. First was a neutral letter explaining new services available to the customers of Customs. Second informed about public disclosure of information by the Customs, an internet service open to public where car tax decision of imported used cars are avaiable and searchable with VIN. Third a letter explaining that customs uses third party information about mileage of cars coming from mandatory car inspections. Fourth letter combined information in letters two and three.

Design details

Randomization method
Randomization in office by a computer.
Randomization unit
Randomization unit is an individual in baseline population.
Clusters
The experiment was not clustered, used car importers are treated as independent individuals
Observations
Baseline population of 35,000 out of which 20,000 received a treatment letter.
Arms
20,000 individuals were evenly allocated to receiving 4 different letters
Primary outcomes
The extent to which letters affect tax evasion from used cars. One measure of this is that reported mileage declines relative to comparison information. Another key outcome is the number of cars imported as used to Finland.

Cited by

No citation data available for this registration yet.

Keywords