Do the catholic and protestant countries differ by their tax morale?

Abstract

Does the tax morale differ between various countries? The paper introduces a model of tax evasion. The equilibrium shadow production is determined by consumers’ and entrepreneurs’ tax morale, affected by the inherited culture or religion. The model suggests that in the conditions of the prisoners’ dilemma, shadow economies tend to be large once a moral code is violated. The implications of the model are tested in the OECD data on groups of countries with different religious denominations in two regimes, 1979–1992 and 1992–2003. We find evidence on a link between tax morale and shadow market activities, but none to support the view that tax morale differs between the catholic south and protestant north in Europe.

Appendices

Appendix A: Proof of Proposition 3

In the equilibrium with shadow economy, the number of tax paying entrepreneurs is ${\frac{m}{x_{1}}}={\frac{m}{\sqrt{2(\tau +\delta)}}}.$ Then, the total tax revenue is

$$ T=\tau {\frac{m}{x_{1}}}=\left(1-{\frac{\left(\sqrt{2(\tau +\delta)}-\sqrt{ 2\delta /\beta }\right) } {g\left(\alpha _{1}-\alpha_{2}\right) }}\right) {\frac{\tau } {\sqrt{2(\tau +\delta)}}}. $$

(10)

Developing the first derivative,

$$ {\frac{\partial T}{\partial \tau }}=-\left(\tau -{\frac{\left(\tau +2\delta \right) }{\sqrt{2\left(\tau +\delta \right) }}}\left(g\left(\alpha _{1}-\alpha _{2}\right) -\sqrt{2\left(\tau +\delta \right) }+\sqrt{2\delta /\beta }\right) \right) {\frac{1}{2\left(\tau +\delta \right) g\left(\alpha _{1}-\alpha _{2}\right) }}. $$

The revenue maximizing tax rate τ^* satisfies ${\frac{\partial T\% }{\partial \tau }}=0.$ Yet, no closed-form solution is available. However, it holds that T(0) = 0. Evaluating the first-order condition at the origin, $\left( {\frac{\partial T}{\partial \tau }}\right) _{\tau =0}={\frac{1} {g\left(\alpha _{1}-\alpha _{2}\right) \sqrt{2\delta }}} \left(g\left(\alpha _{1}-\alpha _{2}\right) -\sqrt{2\delta }+\sqrt{\frac{{2\% }{\beta }}\delta }\right) >0.$ Thereby, the Laffer-curve is increasing at the origin. Developing the second derivative, $\partial ^{2}T/\partial \tau ^{2},$

$$ -{\frac{2\left(\tau +2\delta \right) }{4g\left(\tau +\delta \right) ^{3}\left(\alpha _{1}-\alpha _{2}\right) }}-{\frac{\delta \sqrt{2}} {4g\left(\tau +\delta \right) ^{2}\sqrt{\tau +\delta }\left(\alpha_{1}-\alpha _{2}\right) }}\allowbreak \left(g\left(\alpha_{1}-\alpha _{2}\right) + \sqrt{2\delta /\beta }-\sqrt{2\left(\tau +\delta \right) }\right) <0, $$

reveals that the curve must have a global maximum. In addition, we know that at τ = τ ^max, where

$$ \tau ^{\max }={\frac{1}{2}}\left(g\left(\alpha _{1}-\alpha_{2}\right) + \sqrt{2\delta /\beta }\right) ^{2}-\delta, $$

the legal sector ceases to exist. Thereby, a unique revenue maximizing tax rate $0<\tau ^{\ast }<\tau ^{\max }$ must exist with $T\left(\tau ^{\ast }\right) >0.$

Appendix B: Data

1.1 Definitions

In Table 3, we list the variables used in this study. Column 1 lists the abbreviations for the variables and Column 2 the names of the variables as they appear in the (primary) data source. The main source of data is OECD Economic Outlook 76, except for the currency in circulation. For the members of the EU and the ERM I or II (Austria, Belgium, Denmark, Finland, France, Germany, Greece, Ireland, Italy, Netherlands, Portugal and Spain) we had to retrieve data from the IMF IFS and convert the currency by using the ERM I exchange rates.

Table 3 List of the variables and their sources

Table 4 presents the averages of each variable for both periods across countries. Reading from this table, it is generally the case that the growth rate of taxes has been greater in the first than in the second period. This is also true for the benefits variable, but not for the government consumption. Since the two activities of the government have grown faster in the first period, we name the periods as the regime of the growing public sector and of the mature public sector.

Table 4 Averages of each variables over the first and second period

1.2 Break point tests

There is casual evidence for the idea that there might be breaks in the public sector variables during the economic turbulence years of early 1990s. Therefore, we estimate the following model by OLS for all three country groups (t = calender time)

$$ q_{i,t}=a_{i,t}+b_{i,t}t $$

(11)

and test for the presence of a break, examining whether a _i,t = a _i,t-1 and b _i,t = b _i,t-1 for all t. The Quandt-Andrews unknown breakpoint test (Andrews 1993) seeks the presence of a breakpoint at any t and reports the maximum of LR-statistic and the year of the breakpoint. The results are presented in Table 5.

Table 5 Quandt-Andrews breakpoint tests

The results indicate that both the benefits in the North and in the Rest of the World exhibited a break in 1992 while the tax variables exhibited the break already in 1990. There is no evidence of a break in the government expenditures in these two groups. The results for the Catholic South suggest a break in 1983 both in the benefits and in the government expenditure variables while the break in the tax variable appears a year earlier. Our findings suggest that multiple breaks are present. Yet to have at least one period where no breaks occur, we choose to split our data in 1992.