The impact of terrorism on tax enforcement effectiveness: a case study of the Basque Country and Navarre

ABSTRACT

This paper analyses the impact of terrorism on tax enforcement effectiveness by focusing on the case of the Basque Country and Navarre. The reduced-form model shows that terrorism negatively affects tax enforcement set by the regional administration and, consequently, the way it is perceived by residents in this area. These results are tested by using Spanish surveys and other data sources, finding evidence of the negative impact of terrorism on tax enforcement as it is perceived by residents in the Basque Country and Navarre. In particular, this effect is stronger for entrepreneurs and liberal professionals. Instead, no significant impact for individuals resident in the rest of Spain is found.

KEYWORDS:

• tax enforcement
• audit perception
• fiscal externalities
• terrorism

JEL:

• D74
• H23
• H26
• H83

ACKNOWLEDGEMENTS

The author is very grateful to José Maria Durán-Cabré, Alejandro Esteller-Moré, Amedeo Piolatto, Federico Revelli, Leonzio Rizzo, Albert Solé-Ollé, two anonymous referees and the journal editor for helpful comments and suggestions. The author also thanks the participants at the seminar of the Department of Economics of the Catholic University of Milan, 2016; the Royal Economic Society, 2015; the SAEe (University of Girona); the 27th SIEP conference (University of Ferrara); the 55th ERSA Congress (Lisbon); the IEB Workshop on Taxation (University of Barcelona); the 3rd annual TARC conference on Tax Administration (University of Exeter); the IEB seminar (University of Barcelona); and the seminar of the Department of Economics of the University of Umeå for useful comments. The author is also grateful to Mikel Buesa-Blanco for providing a copy of his book ETA S.A. The usual disclaimers apply. An earlier version of this paper was awarded the Epainos Award for Young Scientists by the 55th ERSA Congress Committee.

DISCLOSURE STATEMENT

No potential conflict of interest was reported by the author.

Correction Statement

This article has been republished with minor changes. These changes do not impact the academic content of the article.

Notes

1. This is the practice of several terrorist organizations, including Euskadi Ta Askatasuna (Basque Homeland and Freedom – ETA) in the Basque Country (Buesa & Baumert, 2013; Leonisio, Molina, & Muro, 2017); the Provisional Irish Republican Army (IRA) (Silke, 1998); and the National Liberation Front of Corsica (FCLN) (Sanchez, 2008).

2. The Basque framework is referred to in a wider sense to include Spanish so-called foral autonomous communities of the Basque Country and Navarre. The foral community of the Basque Country comprises three provinces (Álava, Biscay and Gipuzkoa), while the foral community of Navarre coincides with the homonym province. For a more detailed description of the Basque framework and ETA terrorist organization, see Appendix A in the supplemental data online.

3. The previous literature on tax externalities has demonstrated that sub-central tax authorities may employ tax enforcement as an additional instrument for strategic interaction (Cremer & Gahvari, 2000; Durán-Cabré et al., 2015, 2016).

4. An investigation conducted in 2004 by the Spanish anticorruption prosecution agency, reported by Buesa (2011) and by the national press (e.g., Korta, 2004; Bornstein, 2004), reported that the tax authority of the Basque province of Biscay formally exempted from being audited the tax returns of a group of entrepreneurs and liberal professionals who had treated payments to the terrorist organization as deductions in their tax forms. The consequent fiscal opacity might further distort the taxpayers’ incentives to resist extortion, particularly ‘if the payments to terrorists are mentally accounted for as an additional tax and, furthermore, if you are confident of obtaining a tax deduction from the tax authorities’ (Barbería, 2004).

5. Instead, cross-country and panel studies find mixed evidence for the macroeconomic effects of terrorism, showing that the average impact of terrorism on economic growth is either negative, but small in size (e.g., Blomberg et al., 2004; Tavares, 2004), or statistically insignificant (e.g., Gaibulloev, Sandler, & Sul, 2014).

6. Relaxing this assumption would not alter the results of the model, but it would add additional equilibria. Indeed, by being able to set two tax instruments, regions would compete in both instruments (Cremer & Gahvari, 2000; Durán-Cabré et al., 2015) and theoretically may employ both of them in trying to compensate taxpayers for the cost of terrorism, although the use of statutory tax rates for this aim seems to be unlikely.

7. Where $\left(\mathrm{\partial }{\text{β}}_i^e/\mathrm{\partial }{\text{β}}_i\right)>0$, $\left({\mathrm{\partial }}^2{\text{β}}_i^e/\mathrm{\partial }{\text{β}}_i^2\right)>0$ and $X$ is a variable exogenously collecting information about the individual and situational characteristics as well as the social context that might have an impact on the individual's perceived enforcement (e.g., Alm, 2000).

8. The psychic benefit from living in region 1 is then expressed as $(1-n)$. Thus, individuals indexed by$n\in \left(0,1/2\right)$ reside in region 1, while those identified by $n\in \left(1/2,1\right)$ reside in region 2.

9. The direct utility function is defined as $U=\left[1-t_i\times \left[\alpha +\left(1-\alpha \right)\times \tau \times {\text{β}}_i^e\left({\text{β}}_i,X\right)\right]-g\left(1-\alpha \right)\right]$, where $\left(\tau -1\right)>0$ is the exogenous tax penalty per unit of tax evaded such that $\tau \times {\text{β}}_i^e\left({\text{β}}_i,X\right)<1$ and the function $g\left(1-\alpha \right)$ represents the cost of tax evasion $\left(1-\alpha \right)$, such that $g^{\mathrm{\prime }}\left(1-\alpha \right)>0$, $g^{\mathrm{\prime }\mathrm{\prime }}\left(1-\alpha \right)>0,g\left(0\right)=0,g\left(1\right)\to +\mathrm{\infty }$.

10. In particular, it is possible to show that ${\text{β}}_1$ and ${\text{β}}_2$ are strategic complements. For a formal derivation, see Durán-Cabré et al. (2015), in particular equation (6).

11. All annually released surveys are based on personal interviews conducted with a representative sample of 2500 Spaniards over the age of 18. For the complete contents of the survey, see the CIS website (http://www.cis.es).

12. The original question in Spanish is ‘¿Cree Ud. que, en la actualidad. la Administración hace muchos, bastantes, pocos o muy pocos esfuerzos para luchar contra el fraude fiscal?’ (e.g., question 21 of survey no. 2994 released in 2013, as the numbering of the questions might change from year to year).

13. Apart from $Truc{e}_{jt}$, three alternative direct proxies of ETA's terrorist activity have been employed: the number of fatalities attributed to ETA in any Spanish province/year; the total pecuniary compensation for the damage caused by terrorism provided by the Spanish Ministry of the Interior; and an estimation of the total revenues obtained by ETA through the ‘revolutionary tax’ in the foral provinces.

14. An announced truce is considered as being implemented by ETA in a specific province/year if during that period ETA did not claimed fatalities in that province. The possibility to let this variable vary not only over time but also across provinces is of particular interest for the analysis since in the territories belonging to the foral regime the tax authorities are appointed to operate at this level of government. Information on truces and on fatalities is extracted from the data set of the Spanish Ministry of the Interior.

15. Since the variable $Truc{e}_{jt}$ is indirectly related to the level of terrorist activity, according to equation (9), its coefficient is expected to be positive and significant for foral communities.

16. The coefficients can always be interpreted as the marginal effects of the regressors on the latent variable, which is particularly useful in contexts such as that analysed here, where the latent variable can be given some easily interpretable meaning and it is not a mere modelling device (e.g., Wooldridge, 2002).

17. The issue of potential cyclicality of tax enforcement is theoretically analysed by Andreoni (1992) and empirically tested by Durán-Cabré et al. (2018). Their results confirm this prevailing counter-cyclical trend for Spain, but suggest that in the presence of a severe economic crisis, tax enforcement turns out to be pro-cyclical.

18. Probit models tend to outperform ordered response models, but since these models are based on a way the dependent variable was recoded into a binary one, they cannot be directly compared with the ordered probit models, and this is considered as a robustness check.

19. These results are available from the author upon request.

20. See note 13.

21. Compensations include personal as well as material damages. These data are extracted from the statistical report annually released by the Spanish Ministry of the Interior (for the 2013 report, see http://goo.gl/GEwg2R).

22. These data are aggregated at the national level and do not distinguish between the compensation paid out to the victims of ETA from that paid out to the victims of other terrorist organizations. Nevertheless, the author excluded data referring to the 2004 Al-Qaeda terrorist attack, and as 96.5% of the fatalities/injuries of terrorism in Spain are attributable to ETA, it seems these measures provide a reasonable approximation of the damage caused by ETA's activity.

23. This frequently used indicator has been criticized since it tends to underestimate the degree of terrorist activity (Frey, Luechinger, & Stutzer, 2007). As for $Truc{e}_{jt}$, this variable is defined by considering the survey year, and it is extracted from the data set of the Spanish Ministry of the Interior.

Additional information

Funding

This work was supported by the MCIU/AEI/FEDER. UE [grant number RTI2018-095983-B-I00]; as well as by the Generalitat de Catalunya [project number 2009SGR102] and the Economic and Social Research Council [grant number ES/S00713X/1].