Researchers use regression discontinuity analysis to study how tax thresholds affect firms. Many tax systems set clear cut-offs based on revenue or size. Firms that cross these thresholds face higher tax rates or stricter compliance rules. As a result, some businesses adjust their reported figures or actual operations.
Tax thresholds create strong incentives. Managers often try to keep revenues just below the cut-off. This behavior produces a visible bunching of firms immediately under the threshold. At the same time, fewer firms appear just above it. The resulting firm size distribution shows a sharp discontinuity at the policy boundary.
Regression discontinuity design exploits this sharp cut-off. Analysts compare firms that fall slightly below the threshold with those that fall slightly above it. These two groups are similar in most characteristics. Therefore, differences in outcomes near the threshold can be attributed to the tax rule itself. The method provides a credible way to estimate causal effects.
Studies consistently find evidence of bunching. Reported revenues cluster tightly below the threshold. In addition, some firms appear to under-report sales or delay transactions to avoid crossing the line. Others may split operations into multiple smaller entities. These responses distort the observed size distribution of firms.
The analysis also reveals effects on reported revenues. Firms just below the threshold often show lower growth in subsequent periods. In contrast, firms that cross the threshold may expand more freely once they accept the higher tax burden. However, the initial jump in tax liability can reduce investment and employment in the short run.
Researchers examine several mechanisms. Some firms engage in pure reporting responses without changing real activity. Others alter actual production or sales timing. Moreover, the strength of the response depends on enforcement intensity and the size of the tax jump. Stronger monitoring usually reduces pure under-reporting but may still affect real decisions.
Policy implications follow directly from these findings. Sharp thresholds can create inefficiencies by discouraging growth near the cut-off. Smoothly graduated tax schedules may reduce such distortions. In addition, better third-party reporting and digital compliance tools can limit pure reporting manipulation. Policymakers therefore need to weigh revenue gains against the behavioral costs of rigid thresholds.
Overall, regression discontinuity analysis offers clear evidence on how tax thresholds shape firm behavior. It shows both the distributional distortions and the revenue effects that arise around these policy cut-offs. The approach continues to guide more effective tax design.