Norway is conducting a large-scale, real-world experiment on earned income tax allowances from 2026–2030. The study will involve 100,000 randomly selected taxpayers aged 20–35, who will receive a new deduction. The aim of the randomised, controlled trial (RCT) is to examine how new tax incentives influence labour supply and employment. The Finnish Tax Administration (FTA) regularly uses RCTs to evaluate the efficiency of its activities (Hyytinen et al., 2022). It is well known that RCTs offer an attractive study design characterised by simplicity and clarity. However, if a tax experiment analogous to that in Norway were implemented in Finland, certain questions would need to be considered in advance: The FTA’s statutory duties include tax collection, tax control, and safeguarding taxpayers’ rights. It must also ensure correct and consistent taxation, develop its service capabilities and perform certain other duties assigned to it by the Ministry of Finance. Academic research is not one of the FTA’s responsibilities, and a study similar to the one in Norway could not be implemented. In addition, due to strict constitutional requirements (The Constitution of Finland (731/1999) 81 §), any tax experiment must be based on a specific piece of legislation that stipulates clear rules on tax liability, calculation bases, and the legal remedies available to the persons or entities liable to taxation. They cannot rely on broad administrative discretion or open-ended clauses citing “special reasons”.
To conduct a tax experiment equivalent to that in Norway, the political decision-makers and the Finnish Ministry of Finance would need to support the initiative. Implementing the experiment would require new legislation authorising and detailing the trial. In practice, the statute would have to: 1) state compelling public interest and a clear evidence‑generation purpose, ensuring that fundamental rights are not infringed; 2) be enacted by Act of Parliament and set precise rules on the target group, selection/area criteria, the scope of the trial, and the transparency of the randomisation method (including code); 3) observe proportionality and necessity, allowing only those deviations from ordinary law that are needed, and avoiding undue advantages or excessive gaps between treatment and control; 4) apply fair, non‑discriminatory selection; person‑based limits (e.g., by age) are acceptable only when well‑justified and non‑arbitrary; 5) set out the geographic/target scope and duration in the legislation, with a strict, time‑limited mandate; and 6) pre‑plan evaluation, governance and reporting to ensure credible, policy‑relevant results (Finlex, n.d.-a, n.d.-b; Kangas & Pulkka, 2016; Raitano et al., 2021).
Sufficient funding would be needed as the experiment would entail development needs for the FTA’s information and taxation systems. An RCT would also require sufficient staffing of the FTA’s customer services, as demand for telephone services and physical visits to its service points would probably increase. The FTA would also need to draft detailed guidance regarding the RCT and train its staff. Long-term technical change management would also be required, given that taxpayers would retain the right to appeal their previous years’ tax assessments for several years after the end of the experiment.
However, randomised controlled trials raise a more fundamental issue: the unequal treatment of taxpayers inherent in the RCT design. The Finnish constitution explicitly supports equal treatment for all, a position reinforced by other legislation, such as the Tax Administration Act (Finlex, 2010), which stipulates that the FTA must promote consistent and equitable taxation. As RCTs inevitably result in differential treatment between the experimental and control groups, researchers must address the tension between the need for the high-quality knowledge RCTs provide and the need to protect basic rights. In Finland, the Constitutional Law Committee is the parliamentary body that pronounces on the constitutionality of bills. At the time of the widely publicised basic income experiment in Finland (Helsinki University Library, n.d.), the Committee considered randomised assignment and differential treatment permissible because the enabling Act set out, in sufficient detail, the experiment’s objective, target group, selection and randomisation criteria (including the transparency of the method), scope, and duration, along with legal safeguards. Consistent with its broader practice (e.g., in the trial of relief from social security contributions for temporary employers), the Committee accepted randomisation and unequal treatment only when these core parameters were defined by statute with adequate precision and openness. In the basic income case, compulsory participation was also deemed acceptable to secure reliable results, provided the statutory justifications were sufficiently robust (Finlex, 1999).
Before conducting a similar experiment in Finland, it is important to assess whether the study would yield sufficient new knowledge and insights to justify the effort. It is essential to determine whether Finnish and Norwegian societies are sufficiently comparable in the underlying causes of youth unemployment, particularly whether it stems primarily from a lack of willingness to work rather than from structural barriers to employment. If the situations are sufficiently similar, the findings from the Norwegian experiment could be applied directly in Finland without the need for a new trial.
It might be possible to mitigate the inequality inherent in RCTs by adjusting the experimental design. Randomisation could be applied multiple times during the study, increasing the likelihood that more participants benefit from lower tax rates, though possibly for a shorter period. Finally, at the conclusion of the experiment, the control group could also be offered the trial group’s tax rate for a fixed period, thereby reducing the disparities caused by randomisation.
Should Finland proceed with a similar experiment, it could plausibly be designed to generate more comprehensive knowledge than the Norwegian trial. For instance, multiple experimental groups could be included to examine taxpayer behaviour under different combinations of tax rates and income levels. The study could also provide insights into how people comply with their tax responsibilities when rates vary.
The experiment Norway is conducting is expected to provide valuable insights into how targeted tax system adjustments may influence employment opportunities for young adults, and its findings will be closely studied in Finland.