Abstract
We study the effects of low-intensity tax enforcement on VAT compliance among newly registered firms in Norway using a large-scale randomized field experiment. Around 20,000 firms were randomly assigned to desk-based audits, informational letters, phone-based guidance, or a control group. Desk-based audits uncover non-compliance in over nine percent of cases, corresponding to an estimated VAT gap of about 6.3 percent. However, results show no evidence of persistent increases in reported net VAT following any of the three interventions, and the estimates are sufficiently precise to rule out even small compliance effects. The results suggest that low-intensity VAT enforcement, in isolation, is insufficient to generate behavioral change in the firms’ average VAT reports.