The Effectiveness of Different Tax Credits in Increasing the Demand for Legally Provided Household Services – A Discrete Choice Experiment

How to cite

Burgstaller, Lilith; Necker, Sarah; Doerr, Annabelle. 2022. "The Effectiveness of Different Tax Credits in Increasing the Demand for Legally Provided Household Services – A Discrete Choice Experiment." AEA RCT Registry. AEARCTR-0009800. https://doi.org/10.1257/rct.9800-1.0

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

Lilith Burgstaller* , Sarah Necker , Annabelle Doerr

* Primary investigator

Registry ID
AEARCTR-0009800
DOI
https://doi.org/10.1257/rct.9800-1.0
First registered
Last updated
Status
On going
Trial dates
2022-07-22 – 2022-07-30
Countries
Private
JEL codes
H26, C93, D10
Registry record
http://www.socialscienceregistry.org/trials/9800

Abstract

Tax compliance is difficult to enforce when services are provided to private households. In order to incentivize private consumers to act as third-party reporters, several countries have introduced monetary incentives such as tax credits for consumers that demand legally provided services. We study the causal effects of tax credit schemes for household services on consumers’ willingness to demand legally provided services. We implement a randomized control trial within a survey of consumers in Germany. Survey participants are randomly assigned to 1) a baseline scenario without a tax credit, 2) a tax credit of 20% that is granted “at source”, i.e., at the time of service consumption, as in Sweden, 3) a tax credit of 20% that can be claimed with the annual tax return, i.e., consumers receive the credit with some delay, as in Germany. To investigate the effect of the size of the tax credits we additionally implement two scenarios in which the tax credit is 30% via the tax return (scenario 5) and at source (scenario 6). Moreover, we are interested whether participants' willingness to pay differs when they do not have to calculate the tax refund themselves but the final price (decreased by the value of the subsidy) is displayed to them. Therefore, we add an additional scenario (scenario 4) where we show participants the final price. We aim to examine how consumers' preferences and their willingness to pay for legally provided services varies across these three conditions. We identify the willingness to pay with a discrete choice experiment. We hypothesize that the existence of a tax credit increases consumers' demand for legal services and their willingness to pay a higher price for legality. We expect that a tax credit granted at the time of consumption is more effective than the tax credit claimed with the annual tax return and that the tax credits of 30% increase the WTP to a greater extent than the tax credits of 20%. Moreover, we expect the the pre-calculation of the final price to increase consumers' WTP compared to a scenario where no such calculation takes place.

Experimental design

We study the causal effects of tax credit schemes for household services on consumers’ preferences for different attributes of the services and their willingness to pay for legality. Therefore, we implement a randomized control trial within a survey of consumers in Germany. Survey participants are recruited by a survey agency (Consumer field works) and interviewed with an online questionnaire.

The survey participants will be randomly assigned to six treatments. The first treatment is the baseline scenario. Respondents in this scenario will be informed that there is no tax credit available when they demand household services. The second treatment is a scenario in which the survey participants are informed about the existence of a tax credit that is granted “at source”, i.e., at the time of service consumption as in Sweden. If consumers decide to demand a legal service they can benefit from an immediate price reduction of 20% in form of a tax credit. The third treatment is a scenario in which the consumers are informed about the existence of a tax credit that can be claimed with the annual tax return, i.e., the consumers receive the credit with delay, as in Germany. If consumers choose the legal service they can deduct 20% of the price of that service from their tax bill. In the forth treatment, the final price (price - tax credit) in displayed to the participants. Apart from this feature, the treatment is the same as treatment 3. In treatment 5 we vary the size of the subsidy (30%). Apart from this feature the treatment is the same as treatment 2. In treatment 6 we, again, vary the size of the subsidy (30%). Apart from this feature the treatment is the same as treatment 3.

The outcomes of interest in this study are, first, the preferences of consumers for different attributes of the service, namely the receipt of an invoice, the price, whether the service provider was recommended by a friend or the availability of the provider. Second, we are interested in consumers' willingness to pay for a legal service. We identify this parameter by using a discrete choice experiment. We confront each respondent with seven choice sets. In each choice set, we contrast the offers from two service providers that may differ in the four mentioned attributes (legality, price, availability, recommendation). Overall we consider 70 different choice sets that are clustered in 10 blocks. Each respondent is randomly confronted with one block (7 choice sets). The answers to these choice sets allow us to identify how much weight is given to each attribute by consumers and we can also estimate their willingness to pay for legality, i.e., how much they are willing to pay more for a legal service.

Design details

Randomization method
The random assignment of respondents to the six treatments and the random assignment of choice sets to respondents in the discrete choice experiment is implemented with the survey software Qualtrics.
Randomization unit
Survey respondents are randomly assigned to six treatments. Furthermore, in the discrete choice experiment, we randomly assign one out of 10 blocks of choice sets to each respondent.
Clusters
We plan to recruit 1300 survey participants.
Observations
We plan to present 7 choice sets to 1300 survey participants. This leads to a planned number of observations of 9100 observations.
Arms
The 1300 survey participants will be assigned to six treatments. Thus, we plan to observe about 215 participants per treatment, which leads to 1505 observations per treatment arm.
Primary outcomes
We consider the weights for different attributes of the services, the willingness to pay for legal services as well as the probability to choose an offer with invoice as primary outcomes of the study. These parameters are identified from a discrete choice experiment in which respondents have to choose between two service offers that differ in four attributes: availability, recommendation, reciept of an invoice, and price. In the analysis, we will regress respondents choices on indicators of these attributes. From this regression, we can infer how consumers weight the different attributes in their decision-making. By dividing the coefficient of the invoice attribute by the coefficient of the price attribute, we can calculate the willingness to pay for a service with invoice, i.e., wich is legally reported to tax authorities.

Cited by

No citation data available for this registration yet.

Keywords