هنجارهای اجتماعی و رشد اقتصادی در مدلی با فرار مالیاتی
Abstract
1- Introduction
2- The model
3- Productive public expenditures
4- Other features
5- Conclusions
References
Abstract
Empirical evidence suggests that low-income countries are characterized by high levels of labor and capital income tax evasion while the opposite is true for high-income countries. This paper proposes a model to study the relationship between economic growth and both types of income tax evasion. We show that the existence of a social norm towards tax compliance generates a complementarity between capital and labor income tax evasion which explains the decline of both the share of evaders in the population and the amount of tax evasion when countries accumulate capital. The model predicts that the level of tax morale is positively correlated with both types of income tax evasion and the level of income per capita, consistent with recent empirical evidence. Finally, a higher tax rate increases the share of evaders in the population and aggregate tax evasion.
Introduction
Tax evasion not only harms the possibility of governments to obtain public revenues and, thus, to finance public programs; tax evasion is also negatively related to economic performance. In a recent survey, Vasilopoulou and Thomakos (2017) provide evidence about a strong negative relationship between economic growth and tax evasion. Similarly, Schneider et al. (2011) and Crane and Nourzad (1986) find that the relative size of the shadow economy (and tax evasion) have decreased over time for a wide sample of countries whereas GDP per capita has increased. Besley and Persson (2014) and Gordon and Lee (2005) show that for similar tax rates high-income countries obtain much higher levels of tax revenue than low-income countries, suggesting that tax evasion is a severer problem in low-income countries. Moreover, empirical findings suggest that tax evasion is a widespread problem, especially in developing countries which show high levels of both labor and capital income tax evasion (e.g., Crivelli et al. (2016)).