Heterogeneous responses in Payments Under the Table: Experimental evidence from the Dominican Republic

In this Randomized Controlled Trial (RCT) we aim to study which interventions the government can implement to reduce informality and PUTs.

The experiment consists of a series of messages sent to firms, designed to highlight the potential risks associated with engaging in tax evasion schemes. It will be conducted in collaboration with the Dirección Nacional de Impuestos Internos (DGII), the tax authority of the Dominican Republic. If feasible, the study will be performed during the second semester of 2024 and will include 67,525 firms, representing the entire universe of formal enterprises in the Dominican Republic.

Firms will be randomly assigned to a series of control and treatment groups. We will stratify the firms to ensure consistent study results, with particular attention given to those incentivized to begin reporting payroll taxes after the trial. This includes firms reporting negative profits, those with a high assets-to-profits ratio, and those where the percentage of employees reporting salaries at or above the legal minimum is high.

In particular, the study will exploit the fact that labor lawsuits are one of employers' main risks when engaging in Payments Under-the-Table (PUT) with their employees or by hiring them off-the-books. According to the Dominican Republic's Tax Code, enterprises must withhold social security contributions from their workers' salaries and can be sued by employees for failing to meet their obligations if the PUT collusion breaks down. We will denote this as the Awareness Effect.

In addition, lawsuits are not the only consequence firms might face when engaging in various forms of tax evasion. The Dominican Republic's Judiciary can target audits at suspicious firms. Thus, labor lawsuits might trigger audits, serving as signals of tax evasion. We will refer to this as a Propagation Effect, which, when included in some of the treatments, is expected to increase payroll tax compliance.

Profit tax compliance is another significant issue for both the Tax Authorities and employees. Since 1972, the Dominican Republic has enacted a law, currently encompassed in Articles 223 to 227 of the Labor Code, which mandates most enterprises to provide a profit-sharing bonus to employees, amounting to 10\% of their profits. Consequently, firms have an incentive to misreport their profits to avoid this additional payment to employees. However, the Labor Code stipulates that if employees suspect that their employers are not accurately reporting their profits, they can notify the Tax Authorities, thereby triggering an audit. Therefore, we will investigate whether firms alter the profits they report and the income tax they pay when they receive messages informing them of the costs associated with misreporting benefits.

Thus, these treatments will provide variations in the information regarding the effects of employees reporting their employers and the consequences of initiating a lawsuit. Additionally, the study will emphasize how the risk of audits affects firms. Furthermore, given that firms have heterogeneous incentives to evade taxes, the study will also examine variations in compliance within the same treatment group.

2024-10-15

2025-02-28

We are mainly interested in Employees' wages and employment status, and in Firm' s payroll tax compliance.

We will also study Firm's revenues, costs, assets, profits and other taxes paid, and Employees' reported Income Tax.

We will separate our treatments into 7 groups. The Pre-Analysis Plan's appendix includes the treatment mailings to the firms in English. The original messages have been translated into Spanish and have been checked by tax authorities. Upon request, emails in Spanish are available. We will divide our experiment into two groups: firms that reported employees in 2023 and those that reported no workers during the same period.

Firms in the No Message group will not receive an email from the DGII and will remain untreated. In contrast, firms in the Control: Baseline group will receive a basic email that notifies them of their income tax submission date and provides a link to the DGII's website for additional details regarding their tax filings.

The main effects of the experiment will be studied in both the Treatment 1 and Treatment 2 groups. In the \textit{Treatment 1} group, firms will receive information about the potential lawsuits and audits they might face when engaging in PUT collusions, whereas in the Treatment 2 group, firms will be made aware of the same potential risks associated with hiring informal workers.

In particular, firms included in these treatments will be further divided into two subgroups. In the Awareness subgroup, we will examine the responses of employers who are informed solely about the risks of being sued by their employees due to Payments Under-the-Table (PUTs) or off-the-books hiring, depending on the treatment. In the Propagation subgroup, companies will additionally be informed that the tax authorities may use information about labor lawsuits to target audits, as both forms of informality violate the Tax Code.

Finally, while we anticipate that the direct effects of the previous treatments will be an increase in payroll tax compliance and a reduction in informality without necessarily affecting the reported profits or potentially reducing them, we also aim to include a treatment that does not have a direct expected effect on these variables.

Instead, Treatment 3 will address profit tax compliance and is expected to increase the reported profits. Since the Labor Code requires firms to provide an annual profit-sharing bonus to their employees, employers have an incentive to under-report profits. However, employees can report suspected profit misreporting, potentially leading to a full company audit. This treatment will emphasize this potential risk to firms, and we will compare the effects of this message with those of Treatment 1 and Treatment 2.

Not available

Randomization done in office by a computer.

Firms from the DGII tax base.

Yes

67,525 firms. Of those we will have 2 clusters: 17,123 firms have no reported employees, and 50,402 firms have at least one reported employee.

67,525 firms.

-Firms with Employees: 50,402 firms
*No Message: 8,902
*Baseline: 9,000
*Treatment 1: Awareness: 6,500
*Treatment 1: Awareness & Propagation: 6,500
*Treatment 2: Awareness: 6,500
*Treatment 2: Awareness & Propagation: 6,500
*Treatment 3: 7,500

-Firms with Employees: 17,123 firms
*No Message: 123
*Baseline: 7,000
*Treatment 2: Awareness: 5,000
*Treatment 2: Awareness & Propagation: 5,000

Specified in PAP.