탈세의 동태이론모형과 정책 시사점

The analytic literature on tax evasion mostly focuses on the static framework and abstracts noneconomic factors including psychological aspects. Supplementing such shortcomings, this paper analyzes tax evasion decision of heterogeneous taxpayers regarding tax morale and practice based on comparative statics, steady state equilibria, and phase lines from two-period models. Several conclusions of the standard static models with a representative taxpayer are modified as follows. ➀ Tightening enforcement in one period alone may encourage cheating in the other period due to the intertemporal substitution effect and the income effect. ② As the time preference (discount factor) inversely affects tax compliance, tighter liquidity constraints induce tax evasion. ③ In a steady state, the proportion of tax evaders is composed of taxpayers at two extremes alone: those who never cheat and those who always cheat. It declines with an increase in nonmonetary costs of tax evasion, penalties, and the intensity of monitoring. ④ The minimum penalty to achieve full compliance in a steady state diminishes with the audit probability at a decreasing rate and vice versa. Hence, a mixed bundle of the two measures is desirable as long as both are costly. ⑤ The possibility of multiple stable equilibria in the spread of tax evasion arises due to externalities generated through psychic disutility of tax evasion. As a result, a modest policy change could not reverse the situation once positioned at either equilibrium and a temporary overshooting or big-bang policy is required to get out of the social trap, which contradicts to the conclusion drawn without externalities.