Mobilizing for what? Polarized citizens and electoral turnout in transitioning Tunisia

ABSTRACT

In countries that have recently transitioned to democracy, what factors most drive citizens to mobilize and participate in early elections? Many comparative studies on democratization and elections stress the vital importance of early elections in new democracies – with voter turnout inexorably linked to a democracy’s long-term stability and legitimacy – however, much of this literature focuses on aggregate rather than individual-level behaviour, and very little targets the Middle East/North Africa region. This study closely examines individual voting behaviour in democratizing Tunisia’s critical second election in 2014. We argue that amidst great uncertainty, the polarizing issues of national and political identity created systematic disparities in participation – with the most ideologically polarized citizens/social groups more likely to vote. Using original data from a survey conducted in Tunisia right after its November 2014 elections, we find that Tunisians were sharply divided in their support for democracy, the previous regime, and Islamic governance. Specifically, Tunisians who were more ideologically polarized along its secular-Islamist divide and those more satisfied with the new democratic system were more likely to vote – overall suggesting somewhat uneven electoral participation in this critical election and, therefore, the potential for the kind of instability conducive to democratic breakdown.

KEYWORDS:

• Uneven turnout
• Middle East
• Tunisia
• regime preferences
• satisfaction with democracy
• polarization

Disclosure statement

No potential conflict of interest was reported by the author(s).

Supplementary material

Supplemental data for this article can be accessed online at https://doi.org/10.1080/13629395.2022.2083415.

Correction Statement

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

Notes

1. For more, see: https://www.washingtonpost.com/world/2021/07/27/tunisia-gulf-information-campaign, Accessed on August 9, 2021.

2. The first free and fair elections, named Constituent Assembly Elections, were conducted in 2011 to draft a new Constitution. The Constituent Assembly voted to adopt the Constitution on January 26, 2014. The 2014 election was the first after a new Constitution that set the boundaries of a democratic political system.

3. Postcommunist studies have extensively paid attention to turnout in early elections (see Kostadinova and Power (2007). Some exceptions include a detailed study of turnout in the Arab world by De Miguel et al., Rosenthal et al. (2018) and a few studies on Turkey and Israel (e.g., Rosenthal et al. 2018). Otherwise, the determinants of turnout at the individual level have not received much attention from MENA region studies.

4. The Egyptian case is also relevant here as the authoritarian Sisi regime initially ‘legitimized’ itself to ‘restore’ democracy through a coup and not-free/unfair elections.

5. We use Hegghammer (2013) and Volpi and Stein (2015) definition of Islamism. (Statist) Islamist actors are those who justified their (political) activism by primary reference to Islam. As seen in the responses (see Table 4), Ennahda is seen as part of the Islamist movements.

6. Also see Fornos et al. (2004), Kostadinova and Power (2007) and Kostelka (2014).

7. It should be noted that in the cases of Central/Eastern Europe, weak participation in the two decades following the transitions did not initially endanger democratic politics due to factors such as moderate economic development and robust early enthusiasm to be part of ‘Europe’ (Vachudova, 2005)—however, there is growing evidence that low early turnout in the region (especially Poland, Hungary, and the Czech Republic) has facilitated significant backsliding more recently given the rise of right-wing populist parties that have embraced anti-pluralistic, anti-rule of law, and repressive policies towards political opposition (Pirro, 2015).

8. We thank an anonymous reviewer who raised this important point.

9. A massive literature on citizen attitudes towards democracy in post-Communist countries shows that popularity for returning to some variant of authoritarianism increases when there is a strong sense of nostalgia and/or dissatisfaction with the new democratic regime’s institutional and/or economic performance (Evans & Whitefield, 1995; Haerpfer, 2008; Mishler & Rose, 1997).

10. Despite our limited expectations regarding affective partisanship in 2014 Tunisia, we actually find evidence that a positive attachment specifically to Nidaa Tounes resulted in being more likely to vote.

11. For more information on official electoral turnout rates in Tunisia: https://www.idea.int/data-tools/country-view/284/40.

12. Lustration – the purifying/purging of old-regime officials – can come in different forms depending on the context of the democratic transition, and the potential power of prominent actors from the former regime.

13. These survey results – when taken in conjunction with Albrecht, Bufano et al. (2021) finding that disillusionment with democratic procedures can breed support for expanding the military’s role in politics – help shed light on recent events in Tunisia and specifically the military’s embrace of Kais Saeid’s actions.

14. For robustness concerns, we used an oft-used question derived from the World Values Survey which asks whether democracy may have its problems but is better than any other form of government. 83 per cent of survey respondents agree with this statement while only 10 per cent opposed it.

15. Here, one’s represent about 21 per cent of the sample, while two’s represent about 25 per cent.

16. First, we created three nominal variables based on the distance between an individual’s answers for each group; high polarization (those who are coded above as 3), moderate polarization (those who are coded as 1 or 2), and low polarization (those with the same favourability for the two movements, coded as 0). Our second robustness measure examines whether an individual’s favourability towards one group mobilizes them to vote. We use the original survey questions to create two distinct, dichotomous variables-one for “favouring Islamists” and the other “favouring secularists.”.

17. See Kostadinova, Kostadinova (2003), Kostadinova and Power (2007), Pacek et al. (2009), and Turner (1993).

18. We ran an alternative model for robustness check, and the results did not change significantly. We separated non-partisans and other party supporters and included them as additional variables to draw the graphs in Figure 2, available in the Online Appendix.

19. For the city variable, we merged town and villages as the former had a much smaller percentage (8 per cent versus 30 per cent). Without merging, the city variable is insignificant. Nevertheless, the size of the coefficient is very small for both operationalizations.

20. See Albrecht, Bishara et al. (2022).