This section presents and discusses the results of the estimates.
Source: authors.
The stationarity of the variables indicates the existence of the unit root. The stationarity test helps to avoid spurious regressions. Table [2] gives the results of the variable stationarity tests. In this table, we can see that the variables tax revenue, terrorism and inflation are stationary at the level tested by Maddala and Wu (2002) and Pesaran (2007). On the other hand, the variables trade openness, balance of power and industry are stationary at level with the Maddala and Wu (2002) test. In addition, the population variable is stationary at level with the
Source: authors.
4.2. Results and discussion of the relationship between the effect of terrorism on tax revenues
The objective of this study is to determine the effects of terrorism on tax revenues. Eq. (1) is estimated using fixed effects. The Hausman test gave the fixed-effect estimate as an indicator (see Appendix 3). However, the presence of heteroscedasticity (see Appendix 4) was observed using the Breusch-Pagan (1979) LM test. Serial autocorrelation was also found (see Appendix 5) using the Wooldridge (2002) test. In order to correct for the presence of heteroscedasticity and autocorrelation, the fixed-effect estimator with Driscoll and Kray (1998) is used. Using this approach, the model produces unbiased parameters and consistent standard errors for all forms of heteroscedasticity, autocorrelation, cross-sectional and temporal dependencies (Abaidoo & Agyapong, 2022).
Table [3] reports the results of the estimation of the effects of terrorism on tax revenues. In model [1], the terrorism variable is the only variable. In model [2], the variables trade openness and natural resources are introduced. In model [3] the inflation variable has been introduced. In model [4] we add the variables industry and population. Finally, in model [5] the variable measuring the quality of democratic institutions, that is the balance of power, is introduced.
Table [3]: Terrorism and tax revenues: Fixed-effect model
Note: The numbers in parentheses are the standard errors of the regression coefficients. ***P<0.01, **P<0.05, *P<0.1.
Source: authors.
The results show that terrorism reduces tax revenues significantly at the 1% threshold. Thus, a 1% increase in terrorism in Sub-Saharan Africa leads, all other things being equal, to a 0.0001% reduction in tax revenues. The negative effects of terrorism are directly in line with the theories developed on terrorism and violence. The presence of violence, and terrorism in particular, hampers economic growth and, by extension, tax revenues. The significance of the results can be explained by the political, economic and even geographical environment that is conducive to the development of terrorism. The countries of sub-Saharan Africa have taken certain steps to curb terrorism and violence of all kinds. However, these actions remain limited, hence the negative effects. The results are in line with those found by Mukhtar & Jehan (2021). The latter showed that terrorism has a negative effect on tax revenues in Pakistan. However, with a better quality of institutions, this phenomenon is reduced.
The inflation variable has a positive but insignificant effect on tax revenues. A 1% increase in inflation leads to a 0.001% increase in tax revenue. This insignificance can be explained by the reforms in terms of inflationary policy in these countries. Indeed, despite the generalised rise in prices, the state authorities have taken action to curb the various effects of inflation on the economy. Thus, the results on the effect of inflation on total tax revenues run counter to those found by Muttaqin and Halim (2018) and Prowd and Kollie (2021).
The trade openness variable has a significant positive effect on tax revenues. In this situation, a 1% increase in trade openness leads, all other things being equal, to a 0.111% increase in tax revenues. This result can be explained by the economic dynamic that the countries of Sub-Saharan Africa have undertaken. Policies to open up to the outside world have boosted countries' exports. This action has led to an increase in tax revenues through higher productivity in these countries. This result is consistent with those found by Terefe and Teera (2018). The latter showed that trade openness has a positive effect on fiscal performance. Thus, to arrive at these results, they use a multivariate panel analysis of cointegrating data.
The variable measuring natural resources improves tax revenues in Sub-Saharan Africa. However, this improvement is not significant. In this situation, a 1% increase in natural resources leads, all things being equal, to a 0.100% increase in tax revenue. The insignificance of this variable can be explained by the natural resource curse theory. According to this theory, countries rich in natural resources have a tendency to divert revenue from natural resources to their own ends. To do this, not all the revenue from natural resources is paid into the state accounts. Some of it is embezzled by rogue agents. The results corroborate those of Brun and Diakité (2016).
The balance of power control variable reduces tax revenues. However, this reduction is insignificant. In fact, a reduction in the number of opposition MPs in the National Assembly leads, all other things being equal, to a reduction in tax revenue of 0.764%. This can be explained by the fact that the absence of a number of political parties in the National Assembly will result in the de facto domination of the presidential party. As a result, there will be no oversight of the State's actions, which will be able to do whatever it wants with public revenue because it is relying on their large number in the assembly to block the motion for oversight. The results are in line with those of Rashid et al (2021).
Industrialisation in Africa, and Sub-Saharan Africa in particular, remains in an embryonic stage. For this reason, the industrial sector plays very little part in the economic dynamics of these countries. The results of this research show that the industry variable has a negative impact on tax revenues. However, this effect is not significant. Consequently, a 1% reduction in the industrial sector leads, all other things being equal, to a 0.0095% reduction in tax revenues. This result is in line with those found by Teera and Terefe (2018).
The final control variable in the study is the population variable. This has a positive but insignificant effect on tax revenues. In fact, a 1% increase in the population leads, all other things being equal, to a 0.567% increase in tax revenue. The variable's insignificance can be explained by the lack of knowledge about the validity of tax revenues. The authorities in sub-Saharan African countries communicate very little about the need to pay taxes. What's more, some citizens who are aware of the validity of tax revenues are discouraged by the misappropriation of public funds. This situation no longer motivates them to fulfil their tax obligations. Despite this, there is a willingness to participate in citizens' tax revenues. The results run counter to those found by Boukbech et al (2018).
Source: authors.