Changing complex behaviors such as tax evasion may require several behavioral interventions, or nudges. We developed a tailored compound intervention approach to increase employers’ payroll tax compli-ance in Estonia’s construction industry. First, we used anthropological methods to gain insights into the deci-sion processes of employers and employees in the industry (Study 1, N=16). These insights were combined with behavioral decision-making principles to design an intervention e-mail with the aim to strengthen per-ceived risks and weaken descriptive norms of non-compliance as well as to strengthen collaborative and weaken adversarial construal of tax authority. The compound intervention was tested using a three-armed non-blinded randomized controlled trial (Study 2, N=4770) involving all employers whose declared wages were below 70% of the industry’s average. The intervention significantly increased declared payroll taxes by 5.1% to 6.1% on average across the two treated groups over a 3-month follow-up period (p< .0001) and was therefore effective. It remains unclear which constituent nudges of our compound intervention were crucial for this effect. These findings suggest that the compound intervention approach may be a feasible avenue for designing effective interventions in under-studied contexts.

Changing complex behaviours such as tax evasion often requires combining several behavioural insights. We developed a tailored compound intervention approach to increase employers’ payroll tax compliance in Estonia’s construction industry. First, we used anthropological methods to gain insights into the decision processes of employers and employees in the industry (Study 1, N=16). These insights were combined with behavioural decision-making principles to design an e-mail with the aim to strengthen perceived risks and weaken descriptive norms of non-compliance as well as to strengthen collaborative and weaken adversarial construal of tax authority. The intervention was tested using a three-armed (1:1:1) non-blinded randomised controlled trial (Study 2, N=4770) involving all employers whose declared wages were below 70% of the industry’s average. The intervention significantly increased declared payroll taxes by 5.1% to 6.1% over a 3-month follow-up period (γI=11.95,p= 0.0001; γE=15.38,p&lt; 0.0001). These findings demonstrate the feasibility of the compound intervention approach.

No abstract