Media use,information reliability and political efficacy in Tunisia,2011

Abstract

Information access and open communication – through in person and mediated information and communication technology – are critical to an informed citizenry in democratic societies. The 2011 Arab Spring uprising that originated in Tunisia and resulted in the overthrow of long-time Tunisian authoritarian president Ben Ali, established a new transitional government with more democratic institutions and more open press and political expression. In this paper, we explore changes over time (2011–2019) in the use by young, educated Tunisians of different political information sources, the perceived reliability of these sources, their information sharing behavior, and sense of being politically well-informed (i.e., political information efficacy). We report here results from the third of three surveys we administered of an online questionnaire to three different but comparable opportunity samples of young, educated Tunisians. The first two surveys conducted in 2012 and 2015 have been previously reported. We compare results from the most recent survey regarding 2019 elections with findings from the two prior surveys. Our findings confirm increasing perceived reliability of government information sources during the 2014 and 2019 elections, and decreasing reliability of social media. Results also confirm that higher perceptions of information reliability along with information sharing, lead to greater political information efficacy which is an important predictor of further democratic political participation.

Keywords

Political participation social media Tunisia political efficacy information sharing

1. Introduction

The Arab Spring began in Tunisia with protests and violent demonstrations originating in the rural town of Sidi Bouzid on December 18, 2010, and spread throughout Tunisia to its capital, Tunis, largely due to the participation of civil society (Klass & Dirsus, 2018) and young, well-educated adults (Wulf et al., 2013; Honwana, 2013), labor unions, and satellite-distributed news and videos broadcast by the Qatar-based Al-Jazeera Television Network (Dahdal, 2012; Ghannam, 2012; Salem & Mourtada, 2014). The Arab Spring spread to Egypt, Libya, Bahrain, Yemen and Syria, as well.

Tunisian government-controlled media down-played the development of the 2010–11 uprising. Under the authoritarianism of the 23-year regime of President Zine El Abidine Ben-Ali, the mass media were strictly censored, and political talk and discussion repressed (Browne, 1999, Siebert, 1963). After 29 days of violent street protests, on January 14, 2011, President Ben Ali resigned and went into exile in Saudi Arabia, ending his 23-year regime. He died there in exile, aged 83, in September 2019. Following Ben-Ali’s downfall in 2011, Tunisians elected a National Constituent Assembly, after which Islamist and secularist political parties formed a transitional coalition government (Arieff, 2020). The Assembly drew up and ratified a new Constitution in 2014, followed by Presidential and Parliamentary elections the same year. The transitional government, led by an acting president (briefly former Prime Minister Mohamed Ghannouchi, followed by the Speaker of the Parliament) allowed greater press freedoms and an increase in the number and diversity of news outlets for broadcast media, such as television, radio and newspapers. Access to online news sources also increased, as some previously blocked or filtered websites opened up and new ones were created (Haugbolle & Cavatorta, 2012, Marzouki et al., 2015).

Tunisia has many typical weaknesses of post-revolutionary states, including lingering corruption, opaque funding of political parties, disappointment with democratic initiatives, and the need for regulatory reform (Grewal, 2019; Elswah and Howard, 2020; Sghaier et al., in press). But an increasing reliability (i.e., trustworthiness, factual accuracy) of political information sources (offline and online) are also among the most important conditions for political participation in a democratic society (Verba & Nie, 1972). Access to information that is reliable and opportunities for political discussion about political developments are critical, especially during times of upheaval or crisis, such as revolution and post-revolutionary transitions in government (Starbird & Palen, 2012; Al-Ani, et al., 2012; Ghareeb, 2000; Garon, 2003; Kavanaugh et al., 2016). Political talk and discussion whether formal or informal and online or offline are fundamental forms of democratic participation (Kim et al., 1999, Shane, 2004). Discussion and interaction take place not only face-to-face and over the telephone, but also online, and increasingly via social media, such as Facebook, Twitter, blogs, Instagram, YouTube and other user content-generating and user-sharing applications (Chadwick, 2006; Zuniga, et al., 2012). According to many studies over two and half decades (Shane, 2004; Chadwick, 2006; Howard, 2011; Zúñiga et al., 2012; Ghannam, 2012; Chun & Noveck, 2020; Scholl, 2020; Schuler, 2020) the use of Internet-based communication technologies opens up channels that can strengthen democracy, expand civil society and contribute to the democratization of the state.

Political talk depends on government tolerance, if not active support, of a relatively free press (for reliable information), information sharing, and freedom of expression among the populace (for discussion and deliberation) (Zúñiga et al., 2012). Citizens’ risk-free conversation in political discussion networks and related contexts is the defining criterion of a deliberative, democratic political system (Meikeljohn, 1948; Kim et al., 1999).

Political conversation is riskier in heterogeneous discussion networks because disagreement is more likely among diverse members who may respond with disapproval, criticism, or confrontation (Zúñiga et al., 2018). Under authoritarian regimes, such as pre-2011 Tunisia, where censorship and repression were routine, political expression was risky and constrained, since criticism of the government was typically met with reprisal, arrest, or worse (Miladi, 2021). The 2011 uprising in Tunisia encouraged citizens to voice their opinions and values (Skandrani et al., 2020). It also influenced citizens’ relationship to state institutions and policies exhibited through their emancipatory claims, e.g., for inclusion, moderation and participation through dialogue and compromise (Kéfi, 2015; Thomas, 2018; Gabsi, 2019). More recently, the global freedom index of Freedom House, rated Tunisia “very high” in terms of freedom of expression, based on responses to questions such as the following: Are individuals free to express their personal views on political or other sensitive topics without fear of surveillance or retribution (Miladi, 2021). In a national study of Tunisians (Thomas, 2018), 70% replied ‘freedom of the press is guaranteed to a great or medium extent”; 72% replied “freedom of expression is guaranteed to a great or medium extent”; and the majority of Tunisians (80%) – the highest in the MENA region – replied “freedom to criticize government is guaranteed to a great or medium extent”.

News media use, political conversation, opinion formation, and political participation are four interrelated components of deliberative democracy (Kim et al., 1999; Moy and Gastil, 2006). Some scholars argue the first three components are all forms of political participation (Dahl, 1989; Verba and Nie, 1972; Almond and Verba, 1963; Rothenbuhler, 1991). Use of news media is associated with more frequent political conversation; use of news media and political conversation have positive effects on the quality of opinions (e.g., argument quality); news media use and political conversation affect willingness to argue with opponents; and news media use and political conversation are associated with participation, notably with “campaigning” rather than “complaining” (Kim et al., 1999).

One known social impact of sharing what users perceive as more reliable information (regardless of source) is a sense of being politically informed and knowledgeable, that is, having higher political information efficacy (Kaid et al., 2007; Michaelson, 2000; Martin et al., 2018). Studies have shown that citizens with a higher political information efficacy are more likely to participate politically in general (Kaid et al., 2007; Zúñiga et al., 2017).

Our research question is: Are there significant changes over time in young, educated Tunisians’ use and perception of the reliability of different political information sources, their information sharing behavior, and their political information efficacy, from the period of the 2011 revolution through two post-revolution election cycles in 2014 and 2019? We ask whether the results of our multi-year survey studies regarding the 2011 revolution, and the 2014 and 2019 elections, suggest a possible trend toward transition and consolidation of civic life and democracy in Tunisia, at least among a subset of young, educated Tunisian adults?

We have reported results from the 2011 and 2014 survey studies earlier (Kavanaugh et al., 2016; Kavanaugh et al., 2017) and focus here on the results from our third survey study about the 2019 elections, in order to compare them with previous findings. We test for these differences over time using survey data and statistical analyses.

Given prior research, although empirical studies are limited, we expect significant differences in the use and evaluation of government-controlled versus more independent media between 2011 and 2014. We expect fewer significant differences in these measures between the 2014 and 2019 elections, since the conditions are more stable and the press more open and free after 2011. Nonetheless, across the initially more independent media (i.e., the Internet, including social media), we seek to determine empirically whether there are significant differences among these information sources in terms of use and perceptions of reliability over time in post-revolutionary Tunisia. For example, following greater press freedoms, is there evidence that government sources of information in post-revolution Tunisia are perceived to be increasingly more reliable?

We further test for changes over time in information sharing behavior, since prior studies suggest that in countries with authoritarian regimes and tight controls over the press, citizens are likely to turn to each other for information, knowing that official sources are less reliable, due to censorship (Gibson, 2001; Al-Ani et al., 2012; Kavanaugh et al., 2013). If the press becomes more open following a revolution, as it has in Tunisia, is there evidence that citizens perceive official sources as more reliable? Do they feel increasingly better informed politically (i.e., report higher political information efficacy)? Given that political information efficacy is an important predictor of political participation (Kaid et al., 2007; Michaelson, 2000; Zúñiga et al., 2017), we continue to test for changes in this construct in the third of our three survey studies related to the 2011 uprising, the 2014 elections and the 2019 elections, as an indicator of a possible strengthening and consolidation of civic life and democracy in Tunisia (Jebril et al., 2013; Karolak, 2017; Grewal, 2019; Richter, 2017).

2. Political participation theory and related work

Political participation theory explains the various motivations among different types of people and organizations to engage (or not) in civic and political life (Dahl, 1989; Almond and Verba, 1963; Verba and Nie, 1972). Keeping up with news, sharing information, and discussing issues with others are all forms of political participation (Rothenbuhler, 1991), including online variations of these behaviors (Shane, 2004; Chadwick, 2006). An informed citizenry is essential to opinion formation and consensus building – two major building blocks of democratic participation (Meikeljohn, 1948; Verba and Nie, 1972). We test the theoretical construct of ‘keeping up with news’ with our survey measures on the frequency of use of different political information sources; we test the theoretical construct ‘sharing information’ with our survey measures on the frequency of exchanging online political news with family and friends (described further below); we test the construct ‘discussing issues with others’ as part of the same set of survey measures, since the questions are specifically related to social media use to share information. Social media are highly interactive, and constitute a conversation (boyd et al., 2010), among discussion network members, and sometimes a wider audience.

Political discussion network theory explains political conversation and social interaction about politics and issues of interest (Gibson, 2001; Moy & Gastil, 2006; Zúñiga et al., 2012). Most people never or rarely discuss politics with anyone (Zúñiga, et al., 2018). Some people only discuss politics with strong social ties, that is, trusted members of their social network – close friends and family (Granovetter, 1973). A minority speaks with both strong and weak social ties, including public contexts. The wide use of social media has made it easier for the people who only speak about politics with strong social ties to engage more frequently in political talk online, especially via social network platforms, such as Facebook or Twitter, and to share information with their weaker social ties, such as acquaintances and friends of friends (Gibson, 2001; Zúñiga et al., 2018).

The exchange of information and opinions in social media, with its built-in combination of stronger and weaker social ties, can facilitate political discourse and help strengthen civil society in transitional democracies (Gibson, 2001). The combination of strong and weak ties among most members of a society facilitates transitions from more authoritarian to more democratic forms of government (Gibson, 2001). Studies in the US have shown that political expression leads people to participate further politically in other ways, both online and in person, such as, campaigning for an issue or candidate (Pingree, 2007; Rojas and Puig-i-Abril, 2009). As noted, we test the theoretical constructs of information sharing with our survey measures of respondents’ information sharing with members of their social networks, face-to-face, by telephone, and via Facebook, Twitter and photo/video-sharing sites.

Having access to reliable political information sources, and sharing information also leads to a greater sense of being politically informed, i.e., political information efficacy or PIE (Kaid et al., 2007). PIE is related to (but distinct from) internal political efficacy, which is a person’s belief that s/he is personally competent to participate in politics in a broad sense (not just because one has adequate information). External political efficacy is also important in predicting participation; it is a person’s belief that the government authorities are genuinely interested in and responsive to his or her needs and interests, whether individually or as an identity group, such as a minority, or as a citizen of an authoritarian regime (Zúñiga et al, 2017; Martin et al, 2018). Studies have shown that people with higher levels of PIE are more likely to participate politically in general (Kaid et al., 2007).

Citizens’ perceived political efficacy is among the factors that build a participatory political culture, which is a key element in established democratic societies (Almond & Verba, 1963) and in the democratic transition of post-autocratic societies (Garon, 2003; Breuer & Groshek, 2014). We tested the theoretical construct of PIE (as well as internal and external PE) with our measures of PIE and PE discussed in detail under the Methods section.

A number of studies have examined the use of the Internet and/or social media during or after the revolution in Tunisia and its relation to a democratic transition and consolidation (Jebril et al., 2013; Breuer & Groshek, 2014; Zayani, 2015; Richter, 2017; Karolak, 2017; Grewal, 2019; Elswah & Howard, 2020).

These studies wisely emphasize the importance of contextualizing investigations of media use in the social, political and economic conditions and history which are shaping its use and its effects, especially among transitional democracies (Jebril et al., 2013; Bellin, 2012). There is some agreement that liberalizing authoritarian political systems and controls of political communication can lead to alternative platforms for information access and exchange of viewpoints (Jebril et al., 2013), although disillusionment with democracy has been growing in Tunisia (Grewal, 2019; Elswah & Howard, 2020). Moreover, in the hands of authoritarian governments, the Internet can be a tool for social control (Kavanaugh, 1998; Karolak, 2017).

A study of over 300 Facebook users during the uprising and revolution showed that it served several functions: political, informational, and as a media platform, possibly in contrast to constrained official sources (Marzouki et al., 2012). A Web survey of over 600 Tunisian Internet users finds that the political use of the Internet during the 2011 uprising/revolution influenced their levels of internal political efficacy, and may have positively affected their attitudes toward participation in the 2014 elections (Breuer & Groshek 2014).

An interview study of 12 Tunisian social media activists in 2015 finds that the activists considered the Internet to be an important tool in democratizing politics in Tunisia (Karolak, 2017). Nonetheless, these activists lamented, the online public sphere is chaotic and polarized which counteracts the positive effects of social media, e.g., social mobilization, information diffusion, opinion formation. Interviewees also stressed the importance of the legitimacy of the government in strengthening citizens’ sense of external political efficacy (Karolak, 2017).

Facebook (FB) was already embedded in users’ everyday lives and routine communications with their social networks at the time of the uprising and revolution in 2010–11 (Zayani, 2015). They continued to use FB to exchange news and information about the transitional government, street demonstrations, and the elections, as freedom of the press and political expression increased. Citing Howard (2011, p. 11), Zayani (2015) notes “Only a few brokers are needed to use social tools to keep others up to date”.

These studies inform our work and guide the interpretation of our findings. These relevant studies of Tunisia, however, are largely conceptual and contextual, or they are qualitative with a different focus (e.g., activists, civil society organizations). We build on their work and contribute an empirical investigation over time, using survey data, to test the constructs with an opportunity sample of young, educated Tunisians that are predictive of democratic political participation under Tunisia’s emerging new government, namely: reading news and information, seeking information sources with greater reliability, and sharing information with the strong and weak ties in social networks, which is known to lead to feeling politically well-informed.

These constructs are well-known predictors of further political participation in established democracies. In transition democracies, which Tunisia may be, an important predictor of further political participation is the legitimacy of the government itself (that is, a government that is responsive to the needs and interests of its citizens, corresponding to their sense of external political efficacy). We test these additional political efficacy constructs in our trend study, as they shed light on young, educated Tunisians’ perceptions of the possible transition to and consolidation of democracy.

Before turning to our Methods section, we summarize briefly the larger context of the media systems and national penetration of the Internet and social media in Tunisia at the time of the 2019 elections, and in comparison with our earlier survey studies.

3. Political and economic context

Tunisia is a small, lower middle-income country in North Africa with a population of 11.7 million in 2019 (National Institute of Statistics, 2020). It is a former French colony whose languages are predominantly French and Arabic. During the period of the uprising and revolution (2010–11) adult literacy was fairly high at 79%, rising to 82% by 2018 (National Institute of Statistics, 2020; The World Bank, 2016). GDP per capita in 2011 US dollars was $4,305, and fell slightly to $3,872 by 2015 and dipped further to $3,318 in 2019.

Total unemployment just before the uprising had been steady through the late 2000s at about 13% and rose to 18% a year later, with the loss of foreign investment and tourism after the revolution. In 2019, unemployment was 15% (Tunis Afrique Presse, 2019; Grewal, 2019). Youth unemployment (ages 15–24) as a percentage of total labor force has been high, rising from 29% in 2011 to 31% in 2015, and 34% by 2019 (Tunis Afrique Presse, 2019). Unemployment has been especially high for educated youth (Egel & Garbouj, 2013; IBRD, 2016) and has been identified as one of the motivations for youth participation in the 2011 revolution (Honwana, 2013).

Not only in Tunisia, but across the MENA region during the Arab Spring, young, educated people took to the streets and to the Internet to revolt against authoritarian structures, economic hardships, and social exclusion, and to advocate for respecting the dignity of citizens and their rights in political decision-making (Howard & Hussain, 2013, Miladi, 2021).

Based on these findings, we have focused our attention on the communication behavior and effects of an opportunity sample of young, educated adults in Tunisia, for all of our survey participants (described under Methods section).

3.1 Parliamentary and presidential elections (2014 and 2019)

Local and international monitors describe the 2014 elections for members of parliament (October 26) and president (November 23 with a run off on December 21) as “free and fair”. The secular parties won in both parliament and the presidency, and the Islamist parties conceded the election peacefully (AP, 2014; McCarthy, 2019).

In the parliamentary elections, the Nida Tunis (or Nidaa Tounes, Tunis Calls) party, running on an explicitly more secular, anti-Islamist platform, won a plurality, 85 of the 217 seats. This plurality in parliament gave it the right to name a prime minister and to lead a coalition government.

The moderate Islamist Ennahda party (or al-Nahda), which had previously dominated the parliament, came in second with 69 seats (AP, 2014; McCarthy, 2019). The Free Patriotic Union was a distant third with 16 seats, followed by the left wing coalition of parties known as the Popular Front (extremists assassinated two of its members in 2013). The Presidential elections went to a second round in a runoff between incumbent Moncef Marzouki, the Nidaa Tounes party’s founding leader, and candidate Beji Caid Essebsi, who won with 55.68% of the final vote (British Broadcasting Corporation, 2014).

In the 2019 elections, Tunisia again held two rounds for president (September and October) and one round for members of parliament (October 6). Over 7 million voters (out of an estimated 8 million eligible voters) registered for the 2019 elections (about 1.5 million more than 2014 elections), according to the Independent High Authority for Elections (ISIE). However, actual turnout for the 2019 presidential election was lower, with 54% versus over 60% for the 2014 presidential elections (Election Guide, 2021).

Originally scheduled for November 2019, the presidential elections were re-scheduled for September, due to the death in July 2019 of the incumbent President Beji Caïd Essebsi. Twenty-four men and two women were candidates for the presidential election. Among these, two candidates were from the main parties: Youssed Chahed, Prime Minister from Tahya Tounes, and Abdelfatah Mourou, the interim speaker of parliament from the Islamist party Ennahdha. However, the independent candidate, Kais Saied, a conservative (and retired professor of constitutional law), and Nabil Karoui, owner of the popular Nessma TV, secured the highest percentage of votes, leading to a second round runoff election for president a few days after the parliamentary elections. The independent, conservative candidate, Kais Saied, won with 72% votes versus 27% for media mogul, Nabil Karoui. Based on exit polls, an estimated 90% of young people supported Kais Saied (British Broadcasting Corporation, 2019).

In the 2019 parliamentary elections, about 220 registered political parties and many independent candidates ran for the 217 seats of the Assembly of the Representatives of the People (ARP) (Yerkes, 2019). The elections led to a different picture of the Parliament. The ruling secularist party Nidaa Tounes founded by Beji Caid Essebsi to oppose the Islamist party Ennahda and which won most seats in the 2014 parliamentary elections (86), not only lost two-thirds of its members to other parties, but also won only 3 seats. The Islamist party Ennahda, despite a decrease of its popularity and the number of seats won, claimed the lead with 52 seats followed by a new secular party Qalb Tounes, or “Heart of Tunisia,” officially created in June 2019, and led by media mogul Nabil Karoui. The fourth secular party, Free Destourian Movement (PDL)1

¹
PDL French acronym of Parti Destourien Libre.

won 17 seats. The PDL is headed by Abir Moussi, who was a senior figure in former President Ben Ali’s party (Arieff, 2020).

3.2 Internet and social media use

The change to a new transitional government in 2011 gave birth to greater press freedoms and news outlets for broadcast media, such as, television, radio and newspapers, as well as new online news sources, and opened up access to some blocked or filtered websites, such as, www.nawaat.org, YouTube.com and dailymotion.com (Haugbolle & Cavatorta, 2012; Marzouki et al., 2012). Internet penetration among adults (people aged 18 and above) in Tunisia in 2019 was 67%, up from 43.8% in 2013 and an estimated 36.8% in 2010–11 (Lotan, et al. 2011, The World Bank, 2016; Dataportal, 2019; Pew, 2019; Radcliffe & Abduhmaid, 2020). As with prior years, Internet use in 2019 among young adults (18–29 years old), was much higher (94%) than adults 30 and older.

Among Internet users, Facebook (FB) penetration in Tunisia dropped to 55% in 2019 (Pew, 2020) from a high of 85% in 2015 (and a mere 17.6% in early 2011 at the time of the revolution) (Mourtada, 2011; Salem & Mourtada, 2014). In the Middle East and North Africa (MENA) the percentage of FB users that are young people (whether measured as aged 15–29, or 18–35) has been higher than the percentage of total FB users over the age of 29 or 35 (Mourtada, 2011; Salem & Mourtada, 2014; Thomas, 2018). That is, as in many other parts of the world, in the MENA region, young adults have been the predominant users of FB. In 2019, 88% of young adults (people 18–29 years old) used FB, compared with 58% aged 30–49 years old, and 24% aged 50 and above.

There has also been a long-time youth bulge in Tunisia, as in many other MENA countries (Dhillon, 2009), making young adults a predominant segment of the population in these countries. This demographic predominance gives young adults, who are affected especially by policies on social issues, such as employment, marriage, and civic life, a loud voice via the Internet and social media. The primary language on FB is Arabic, followed by French, with English a distant third.

Twitter penetration in 2019, at 7% of Tunisian adults (aged 18 and above), is low, but rising slowly from less than 1% between early 2011 and late 2014 (0.34%) (Salem and Mourtada, 2014). We summarize these user population percentages for the Internet and social media in 2019, nationally, for the age cohort of our sample, and for our sample in Table 1a.

4. Methodology

We have sought to use the same methods as closely as possible for each of the three survey studies in order to compare results and identify any emerging patterns. We conducted the original study related to the 2011 revolution in early 2013 using survey items derived from prior studies (on media use, information sharing, political discussion, information reliability and political information efficacy) and our interest in understanding the use and impact of various information sources in the political context of Tunisia (Kavanaugh, et al., 2016). In February 2015, following the September-November 2014 elections, we conducted a second survey using the same survey items, but referencing the ‘2014 elections’ (Kavanaugh et al., 2017). In this paper we report findings from our third survey conducted in 2020 after the elections in the Fall of 2019. We compare the 2020 results to the results from the previous two survey studies (2011 and 2014). In each of the three studies we asked the same survey items, however, we referred to a different event, as appropriate, for each survey: in the 2013 survey we referred to ‘the uprising’; in the 2015 survey we referred to ‘the 2014 elections’; in the 2020 study, we referred to ‘the 2019 elections’. (See Appendix 1 for the 2019 survey items). In the following analyses we focus on comparing the results from the third survey to those from the 2011 and 2014 surveys.

4.1 Survey samples for studies

The sample for the first survey regarding the 2011 revolution consisted of 150 participants (a 21% response rate); the sample for the second survey regarding the 2014 elections consisted of 156 participants (a 47% response rate). The sample for the third survey regarding the 2019 elections is described in detail below.

The participants in all three survey studies were university students of co-author, Hamida Skandrani and for the 2019 survey, co-author Malek Sghaier, as well. Their students were undergraduate and graduate students in the Higher Institute of Management (marketing and business administration) at the University of Tunis in Tunis (a public university in the capital of Tunisia) for the 2011 survey, and the University of Manouba (just outside Tunis, also a public university) for the 2014 and 2019 surveys, where Prof. Skandrani had moved. Thus, the participants are similar in age, studies, and locale, but are different cohorts of students completing each of the three surveys. Our study could be classified as sequential research or a trend study, which is a form of longitudinal study (Malhotra et al., 2012).

For the 2019 elections survey, as for the prior surveys, Professor Skandrani (and for the 2019 survey Professor Sghaier) invited students in their courses and others in their department to participate in the survey. As with the previous surveys, the 2019 survey was in French, because most courses in these universities are taught in French. We administered the survey for the third study in two rounds: April and November 2020. In April, our data collection was disrupted by the widespread emergence of the COVID-19 pandemic during that time. To increase our response rate, Professors Skandrani and Sghaier invited additional students in November 2020 to complete the same survey. The survey was online, hosted on a server maintained by co-author Sheetz.

Professors Skandrani and Sghaier invited 270 students to participate in the online survey in April-May 2020 (Spring semester) through in-class requests, FB groups, class Google groups, email, or private messages on FB. Eighty-four (84) students submitted completed surveys, a response rate of 31%. Due to Covid-19 increasing at this time, we considered that we may have had fewer respondents than we would have had otherwise. So, we sent a second invitation in November 2020 (Fall semester) to 130 additional participants, resulting in 31 completed surveys, a response rate of 24%. Thus, a total of one hundred and fifteen (115) respondents completed the 2019 election survey in two data collections (April-May 2020 and November 2020). The November 2020 supplemental survey is admittedly a long time after the events, which means respondents’ recollections may not have been as accurate as the May 2020 participants. We tried to mitigate this limitation by inviting the November participants to remember their communication behavior and perceptions during the 2019 election period as best as they could. We conducted t-tests of the responses on all variables across both collections for the third survey to test for any differences.

We found significant differences for several variables including: the amount of reading and writing of blogs, opinions of newspapers, use of social networking sites, and use of photo/video sharing sites. Since we analyzed many additional variables, we combined the groups for analyses and comparisons to the 2011 revolution and 2014 election surveys despite these differences. The number of completed surveys and the response rate for the 2019 elections (115 total completed surveys, a combined 29% response rate) is lower in participants, but higher in response rate than the 150 respondents in the first survey related to the 2011 revolution (a 21% response rate). The highest participation is 156 respondents in the 2014 election survey (a 47% response rate).

4.2 Survey instrument for all three studies

All three surveys (2011, 2014, 2019) ask the same questions, stated the same way, and only the “event” reference changes: in the 2011 survey the event is referred to as “during the uprising”; during the 2014 elections, it is “during the 2014 elections”; for the 2019 elections, it is “during the 2019 elections”. We focus in this paper on the 2019 survey, and compare results with the previous two survey findings reported earlier (Kavanaugh, et al., 2017).

There are multiple sections to the online survey which takes about 20 minutes to complete. The 2019 survey items are included in Appendix 1. Apart from face-to-face (F2F) and phone communication with family and friends, the types of communication we ask about fall into two broad areas: more government-controlled media and more independent media.

Media Use: Information Sources and Frequency of Use: Among government-controlled media (which are primarily traditional, mass media) we asked about TV and radio networks and print newspapers. For each of these information sources, we specified broadcast networks and newspapers by name. We asked which of the listed sources they watched/listened to/read most often to obtain information about political developments during the 2019 elections. The specific TV networks we asked about are shown in Table 2. The specific newspapers we asked about are: Assabah, Essahafa, La Presse, Le Temps, Echourouk, Agence Tunis Afrique Presse, and an open category ‘Other’. In the 2019 survey we added two information sources to the questionnaire, specifically: 1) Attessia TV: a private Tunisian TV channel founded on May 18, 2015 by journalist Moez Ben Gharbia offering variety shows, talk shows, series, and sports shows; and 2) IFM: a private Tunisian FM radio station founded by Hamed Soyah on November 4, 2011; it is the country’s first radio station on the theme of laughter and music. Attessia TV is new since our second study (for the 2014 elections), so it was not included at that time.

For the more independent information sources, we ask about the frequency of using the Internet and social media, including social network sites or SNS (e.g., Facebook), Twitter, photo and video sharing sites, blogs, and Internet communication services, such as Skype and Instant Messaging (IM). Our scale for frequency of use of information source is: almost never (1), less than once a month (2), almost once a month (3), more than once a month (4), almost once a week (5), more than once a week (6), almost once a day (7), multiple times per day (8).

Information Source Reliability: We ask respondents to rate their perceived reliability of the government-controlled versus more independent sources they used to obtain information during the 2019 elections. See Appendix 1 for items. As with the ‘Media Use’ and ‘Frequency of Use’ questions, we listed each type of source individually (e.g., Attesa TV, Twitter, SNS, photo/video sharing sites). We asked them to rate the reliability of each of these sources on a 4-point scale from: 0 (not at all reliable) to 4 (reliable most of the time). Capturing these perceptions of the reliability of each information source provides the basis for comparing perceived reliability of information sources across surveys.

Overall Information Reliability: We also collected two measures of overall information reliability that do not directly mention any specific information source, on a four-point Likert scale of agreement; these are: 1) Overall, over the past year, the information I have been able to obtain has been accurate, and 2) The information I have been able to obtain during the past year was not reliable. These measures allow us to determine which of the underlying information sources (specified in the other questions) are associated with overall information reliability.

Information Sharing with Social Networks: To measure the extent of information sharing by respondents with their social networks, we asked (with a 4-point agreement scale) whether the respondent’s family and friends regularly shared online information about the 2019 elections with the respondent. We also asked whether the respondent regularly shared online information about the 2019 elections with friends and family, with the same agreement scale. See Appendix 1 for precise wording of items.

Political Information Efficacy: There are four standard questions for the construct of political information efficacy (Kaid et al., 2007; Zúñiga et al. 2017), which we adapted slightly for this political context, as follows: 1) I consider myself well qualified to participate in politics of the current Tunisian government; 2) I think that I am better informed about Tunisian politics and government than most people; 3) I feel that I have a pretty good understanding of the important political issues facing Tunisia; and 4) If a friend asked me about the 2019 Tunisian elections and related political events, I feel I would have enough information to help my friend figure out whom to vote for.

We also asked three standard questions measuring internal and external political efficacy with a 4-point agreement scale (Michaelson, 2000; Martin et al., 2018; Zúñiga et al., 2017); these are: 1) Since the first government of Mohamed Ghannouchi, I feel that sometimes politics and government seem so complicated that persons like me can’t truly understand what’s going on; 2) I don’t think officials in the current Tunisian government care much what people like me think; and 3) There are plenty of ways for people like me to have a say in what the current Tunisian government does.

Demographics: We asked respondents about basic demographics, including age, gender, highest level of education completed, occupation (e.g., student) and marital status.

4.3 Data analyses

We analyzed the survey results with standard statistical techniques, including correlations, factor analysis and regressions, using SPSS software. For the factor analyses, we used Principal Component Analysis extraction method, and Varimax with Kaiser Normalization rotation method.

5. Results

In each of the following subsections we present the results of the 2019 study then compare them to the results of the two prior survey studies.

5.1 Demographics

Of the 115 respondents in the 2019 election study, 80% are female and 58% are single. Participants have substantial education: 26% with bachelor degrees, 9% having master degrees, and 65% hold diplomas from technical, commercial or industrial schools. The average age of respondents was 28 years, ranging from 19 to 59 years old. Almost all (97.8%) of the respondents report that they have smartphones. Most participants (87%) report using French when interacting online, while 40% report using Arabic and 24% Franco-Arabic. Only sixteen percent of respondents report using English when interacting online.

For the 2014 election and the 2011 revolution studies, the demographics are very similar. Including a high majority that is female, 76.4% and 61% respectively, well educated, and slightly younger at 24 versus 28 years in the 2019 study.

5.2 Internet use for all purposes

Table 1a shows a comparison of our sample population with other young, adult Tunisians (age 18–29) and with all adult Tunisians for 2013–2019. For penetration measures (use percentages) of Social Network Sites (SNS), typically Facebook, and microblogging sites (e.g., Twitter) these are percentages of Internet users who use these platforms.

Our sample is above the national average, but only slightly above the national average for their age cohort. Table 1b shows the percentages of participants from the three studies that report using different information sources for ALL purposes, not just to get information about the elections or political developments during the uprising. Our respondents are above the national average on measures of social media use, but not much higher than averages for their age cohort of young adults (that is, not controlling for education). There is a high correspondence in the use of the information sources across all three surveys (shown in Table 1b), including the Internet, SNS, TV, video sharing sites, and face-to-face interactions, which are all used at very high levels. The only differences across the three surveys are the lower level of radio use during the revolution and a lower use of newspapers by the participants in the 2019 election study.

Table 1a
Our sample compared with all adults and young adults 2013–2019

Use internet
	2013	2015	2019
Total population	60	44	72
Age 18–24	85	84	98
18–24 w/University	100	100	100
Age 18–34	80	74	95
18–34 w/University	96	98	98
Our respondents	99	98	96
Use SNS/FB
Total population	59	40	48
Age 18–24	85	80	68
18–24 w/University	100	85	78
Age 18–34	80	69	68
18–34 w/University	96	88	78
Our respondents	99	96	97
Use twitter
Total population	16	12	3
Age 18–24	25	26	4
18–24 w/University	27	31	13
Age 18–34	20	22	3
18–34 w/University	26	28	4
Our respondents	25	29	23
Use YouTube
Total population	na	25	31
Age 18–24	na	52	41
18–24 w/University	na	85	54
Age 18–34	na	44	43
18–34 w/University	na	76	55
Our respondents	79	89	93

Source: Dennis et al., 2019.

Table 1b

Use of different information sources during revolution and elections

Information source	% Used source: Revolution	% Used source: 2014 election	% Used source: 2019 election
Internet	99.2	97.9	96.2
SNS/FB	98.7	95.9	92.3
TV (broadcast)	90 (est.)	92 (est.)	95.7
Face-to-Face	84	92.9	86.1
Video sharing	79.1	89.6	84.8
Radio (broadcast)	29.6	82.5	80.9
Newspaper (print)	26.4	30.3	19.8
Twitter	25	24.1	22

Table 1c

Differences in frequency of use of information sources: Revolution and elections

Information source	Mean for 2011 revolution (a)	Mean for 2014 elections (b)	Mean for 2019 elections (c)	Mean diff. and sig. (a-b)		Mean diff. and sig. (a-c)		Mean diff. and sig. (b-c)
Video sharing	7.05 (1.81)	5.14 (2.55)	6.36 (2.15)	1.	91 ${}^{***}$	0.	69 ${}^{***}$	$-$ 1.	23 ${}^{***}$
SNS/FB	7.79 (0.84)	6.71 (2.09)	7.46 (1.55)	1.	08 ${}^{***}$	0.	33 ${}^{**}$	$-$ 0.	76 ${}^{***}$
Internet news	7.71 (0.95)	6.41 (2.20)	7.63 (1.27)	1.	25 ${}^{***}$	0.	09	$-$ 1.	17 ${}^{***}$
Twitter	1.53 (2.72)	1.92 (1.93)	2.04 (2.19)	$-$ 0.	38	$-$ 0.	51	$-$ 0.	13
TV (count of channels)	2.68 (1.37)	2.38 (1.41)	1.73 (0.99)	0.	30 ${}^{*}$	0.	95 ${}^{***}$	0.	64 ${}^{***}$
Radio (count of stations)	1.69 (1.03)	1.37 (0.86)	1.45 (1.11)	$-$ 0.	40 ${}^{***}$	$-$ 0.	34 ${}^{***}$	0.	06
Newspapers (count of)	0.52 (0.83)	0.41 (0.75)	0.27 (0.62)	0.	11	0.	25 ${}^{***}$	0.	14

$p<$ 0.1, ${}^{**}p<$ 0.05, ${}^{***}p<$ 0.01.

5.3 Media use for information about 2011 revolution vs. 2014 and 2019 elections

The majority of 2019 election study respondents (86%) report talking face-to-face with family and friends to obtain political information and about a quarter (27.8%) report talking by phone with family and friends. These results are similar for the 2014 election study respondents, where (92.9%) report talking face-to-face (F2F) with family and friends, and (25.8%) talking by phone with family and friends. For the 2011 revolution study face-to-face interaction is also high at 86%.

As in the previous surveys, we used t-tests to compare the frequency of use of the information sources across the 2019 results (see Table 1c). We find that during the revolution there was significantly more use of SNS, video sharing, TV (Al-Jazeera overwhelmingly), and Internet news sites, but less radio, than during the 2014 elections, see column (a-b). Similarly, comparing the revolution to the 2019 survey we have significantly more use of all sources except Internet news sites and radio stations, column (a-c). The results of the two election studies also are significantly different in measures of use, column (b-c), with the 2019 having higher frequency of use of SNS, video sharing, and Internet news sites, but less TV use.

5.3.1 Broadcast media use: Revolution vs. elections

The popularity of TV stations, that is, the percentage of respondents that report watching a given station, for each of the three studies, is shown in Table 2. The stations that respondents reported they most commonly watched to obtain information about the 2019 elections (Table 2, last column) are the new private channel El-Hiwar Ettounsi (68.5%), followed by the private channel, Nessma (57%), and the government channels El Watania 1 and 2 (56.5%).

Table 2
TV stations watched most often during revolution vs elections

TV stations	% Who watched during revolution	% Who watched during 2014 elections	% Who watched during 2019 elections
Al Jazeera	75.6	15.5	9.6
TV Nationale	40.2	56.5	60.9
Al Arabiya	37.0	7.5	6.1
Nessma	33.1	57	23.5
Other (Fr 24)	29.9	24	17.4
Hannibal	25.4	21	5.2
BBC Arabic	7.7	5.5	2.6
El-Hiwar Ettounsi	NA	68.5	65.2
Attessia	NA	NA	46.1

5.3.2 Internet use including social media for information during revolution vs elections

Almost all respondents reported they used the Internet, averaging 97.7% across all three studies; use of SNS/FB averaged 97.3%, followed by video sharing sites averaging 87%. Twitter use averaged only 25.6% of respondents, although this is much higher than the national estimate of 3% in 2019.

Given this low percentage indicating that they used Twitter, it is not surprising that it was not used frequently, as shown in Table 1c. The average across all three studies of 1.83 means that among those who do use Twitter, most use it infrequently (almost never or less than once per month). On the other end of the frequency of use scale are SNS at 7.32 and Internet news at 7.25, that is, they were used daily or multiple times per day for the respondents in all three studies. Video sharing sites average 6.18 across all three studies, indicating these sites were used more than once a week.

5.4 Perceived reliability of information sources

We investigated perceptions of the reliability of different information sources with the expectation that reliable information would be harder to obtain and assess during the revolution (Gibson, 2001; Al-Ani et al., 2012; Wang & Mark, 2013) than during each of the two election periods. Table 3 is organized by decreasing reliability values for the 2011 revolution study, and shows that, as expected, government sources had the lowest reliability of any information sources in 2011. Indeed the value of 0.88 is the lowest reliability value across all three studies. This is significantly lower than the perceived reliability of government sources in the two election studies, increasing to 1.41 in 2014 and reaching 1.97 in 2019, which is the second most reliable source in the 2019 study. Government sources in 2019 are lower only than face-to-face communications; they are higher than Internet news, SNS (FB), and video sharing sites. This suggests that over the years, information from government sources is perceived as much more reliable than it was during the uprising and revolution (2010–11).

The perceived reliability of face-to-face communication during both the 2014 and 2019 elections is significantly higher than the perceived reliability of face-to-face communication during the 2011 revolution. The reliability scores of information from SNS/FB and video sharing sites during the 2014 and 2019 elections have fallen significantly from reliability scores of these social media sources during the 2011 revolution. The perceived reliability of ‘Internet News’ sources shows no significant differences across all three studies. Only government sources are significantly different from 2014 to 2019 studies (Table 3, last column). The perceived reliability of broadcast media sources (typically government controlled before the revolution) rose after the revolution, including TV, radio, newspapers, and a category generically named ‘government sources’. Finally, the average of all sources for each survey is highest for the 2019 study at 1.60, next highest for the 2014 study at 1.51, and lowest for the revolution study at 1.40. This is driven largely by the increases in the reliability of broadcast media sources including TV, radio, newspapers, and (generic) government sources in the 2014 and 2019 studies.

Table 3
Mean reliability of information sources during revolution vs elections

Information source	Mean for 2011 revolution (a)	Mean for 2014 elections (b)	Mean for 2019 elections (c)	Mean diff. and sig. (a-b)		Mean diff. and sig. (a-c)		Mean diff. and sig. (b-c)
Video sharing	1.76 (0.84)	1.42 (1.00)	1.36 (0.95)	0.	34 ${}^{***}$	0.	40 ${}^{***}$	0.07 ns
Face-to-face	1.73 (0.79)	2.06 (0.87)	2.01 (0.82)	$-$ 0.	33 ${}^{***}$	$-$ 0.	28 ${}^{**}$	0.46 ns
SNS/FB	1.67 (0.89)	1.47 (1.07)	1.25 (1.02)	0.	20 ${}^{*}$	0.	42 ${}^{***}$	0.23 ns
Internet news	1.55 (0.85)	1.43 (0.88)	1.56 (0.77)	0.	12 ns	$-$ 0.	01 ns	$-$ 0.12 ns
TV	1.47 (0.89)	1.68 (0.89)	1.74 (0.93)	$-$ 0.	21 ${}^{**}$	$-$ 0.	28 ${}^{**}$	$-$ 0.07 ns
Radio	1.37 (1.01)	1.61 (0.88)	1.81 (0.85)	$-$ 0.	24 ${}^{**}$	$-$ 0.	44 ${}^{***}$	$-$ 0.20 ns
Newspapers	0.99 (0.96)	1.01 (0.88)	1.11 (0.85)	$-$ 0.	02 ns	$-$ 0.	12 ns	$-$ 0.09 ns
Government	0.88 (1.00)	1.41 (1.07)	1.97 (0.90)	$-$ 0.	53 ${}^{***}$	$-$ 1.	09 ${}^{***}$	$-$ 0.57 ${}^{***}$

${}^{*}p<$ 0.1, ${}^{**}p<$ 0.05, ${}^{***}p<$ 0.01, ns not significant.

We conducted paired-samples $t$ -tests comparing all of the information sources within the revolution study (Table 4 Panel A), the 2014 election study (Table 4 Panel B), and the 2019 election study (Table 4 Panel C). Comparisons in Table 4 read from row to column, so that a positive mean difference indicates that the row variable has a higher reliability rating than the column variable. Conversely, a negative mean difference indicates that the row variable has a lesser value than the column variable.

In the 2011 revolution study, the perceived reliability of SNS/FB, photo-video sharing sites, and Internet news sites is not significantly different from the perceived reliability of F2F communication with family and friends (Table 4, Panel A column 2). However, in the election surveys (Table 4, Panels B and C), the perceived reliability of these social media and Internet news sources is significantly lower than F2F communications. The transformation in the reliability of government sources is also revealed by comparing Table 4, Panel A column 3, to Table 4, Panel B column 3, and to Table 4, Panel C column 3, which shows that the perceived reliability of government sources is significantly lower than all other sources at the time of the revolution in 2011. It rises in the 2014 study, and during the 2019 election period, the perceived reliability of government sources is significantly higher than most other sources. While the perceived reliability of social media and video sharing sites is significantly higher than most other sources during the revolution (Table 4, Panel A rows 6 and 7), the perceived reliability of social media and photo-video sharing sites falls in 2014, and by the 2019 elections, social media and photo-video sharing sources are perceived as less reliable than most other sources (Table 4, Panel C rows 6 and 7).

Table 4

T-test comparisons of mean reliability of information sources

Panel A: T-tests of perceived reliability of information sources used during 2011 revolution
	F2F	Government	Newspaper	Radio	TV	Internet news	SNS
Government	$-$ 0.84 ${}^{***}$
Newspaper	$-$ 0.71 ${}^{***}$	0.11
Radio	$-$ 0.41 ${}^{***}$	0.43 ${}^{***}$	0.33 ${}^{***}$
TV	$-$ 0.26 ${}^{**}$	0.58 ${}^{***}$	0.48 ${}^{***}$	0.15 ${}^{*}$
Internet news	$-$ 0.13	0.68 ${}^{***}$	0.58 ${}^{***}$	0.25 ${}^{**}$	0.09
SNS/FB	0.06	0.88 ${}^{***}$	0.77 ${}^{***}$	0.45 ${}^{***}$	0.32 ${}^{***}$	0.18 ${}^{**}$
Video sharing	0.08	0.92 ${}^{***}$	0.79 ${}^{***}$	0.49 ${}^{***}$	0.34 ${}^{***}$	0.21 ${}^{**}$	0.04
Panel B: T-tests of perceived reliability of information sources used during 2014 elections
Government	$-$ 0.64 ${}^{***}$
Newspaper	$-$ 1.03 ${}^{***}$	$-$ 0.38 ${}^{***}$
Radio	$-$ 0.42 ${}^{***}$	0.22 ${}^{**}$	0.61 ${}^{***}$
TV	$-$ 0.36 ${}^{***}$	0.28 ${}^{***}$	0.64 ${}^{***}$	0.06
Internet news	$-$ 0.59 ${}^{***}$	0.04	0.42 ${}^{***}$	0.19 ${}^{**}$	0.23 ${}^{***}$
SNS/FB	$-$ 0.56 ${}^{***}$	0.08	0.48 ${}^{***}$	0.15	0.19 ${}^{**}$	0.04
Video sharing	$-$ 0.62 ${}^{***}$	0.04	0.39 ${}^{***}$	0.21 ${}^{**}$	0.26 ${}^{***}$	0.01	0.06
Panel C: T-tests of perceived reliability of information sources used during 2019 elections
	F2F	Government	Newspaper	Radio	TV	Internet news	SNS
Government	$-$ 0.04
Newspaper	$-$ 0.46 ${}^{***}$	$-$ 0.88 ${}^{***}$
Radio	$-$ 0.19	$-$ 0.18	$-$ 0.64 ${}^{***}$
TV	$-$ 0.28 ${}^{**}$	$-$ 0.26 ${}^{*}$	$-$ 0.92 ${}^{***}$	$-$ 0.07
Internet news	$-$ 0.92 ${}^{***}$	$-$ 0.41 ${}^{***}$	$-$ 0.48 ${}^{***}$	0.22 ${}^{**}$	0.17
SNS/FB	$-$ 0.78 ${}^{***}$	$-$ 0.71 ${}^{***}$	$-$ 0.18	0.56 ${}^{***}$	$-$ 0.28 ${}^{**}$	$-$ 0.46 ${}^{***}$
Video sharing	$-$ 0.66 ${}^{***}$	$-$ 0.61 ${}^{***}$	$-$ 0.26 ${}^{*}$	0.42 ${}^{***}$	0.38 ${}^{***}$	0.18 ${}^{*}$	0.09

Cells contain mean difference and significance. ${}^{*}p<$ 0.1, ${}^{**}p<$ 0.05, ${}^{***}p<$ 0.01.

To determine if different information sources grouped together based on their reliability scores, we conducted principal components exploratory factor analyses using the reliability variables in all three studies. We used Varimax with Kaiser Normalization rotation. The factor analysis for each study produced a valid solution.

For the 2019 election study, the Kaiser-Meyer-Olkin measure of sampling adequacy is 0.57, which is very close to the recommended value of 0.6; Bartlett’s test of sphericity is significant 181.99, $p<$ 0.01). The face-to-face and government sources variables have low communalities of 0.19 and 0.15, respectively. All of the communalities for the other factors are greater than 0.60, confirming items share some common variance.

The results of the factor analyses for the three studies are presented in Table 5. The 2011 revolution study resulted in a two-factor solution (Table 5, columns 2 and 3). The first factor groups together the perceived reliability of information from (unspecified) government sources, newspapers, radio, TV, and Internet news websites; it accounts for 35% of the variance (i.e., largely government-controlled broadcast media, with the exception of Internet news sites). The second factor includes F2F communication, Facebook/SNS, unspecified Internet sites, and video sharing sites, and accounts for 22% of the variance (i.e., largely sources based in social networks, including social media, with the exception of general Internet sites).

Table 5

Factor analyses of source reliability: Revolution and elections

	2011		2014			2019
	1	2	1	2	3	1	2
Internet news	0.544	0.389	0.656	0.417	0.240	0.449	0.630
Radio	0.791	0.089	0.777	0.154	0.125	0.370	0.740
Newspapers	0.820	$-$ 0.120	0.413	$-$ 0.010	0.739	$-$ 0.326	0.771
TV	0.655	0.320	0.320	0.133	0.794	0.718	0.353
Government	0.704	$-$ 0.018	0.744	$-$ 0.092	0.062	0.191	0.330
Video sharing	$-$ 0.028	0.847	0.195	0.895	0.045	0.785	0.166
Face-to-face	0.071	0.690	$-$ 0.285	0.256	0.596	0.421	0.101
SNS/FB	0.089	0.782	$-$ 0.031	0.891	0.208	0.885	0.019

The 2014 election study analysis produces a three-factor solution (Table 5, columns 4, 5, and 6) for perceived reliability of information sources. The first factor includes Internet news sites, radio, and government sources. The second factor includes SNS/FB and photo-video sharing sites with 19% of the variance. The third factor consists of TV, Newspapers, and F2F communication, accounting for 13% of the variance. This three-factor structure in the 2014 election study analysis is different from the two factors produced for the revolution study which may reflect greater press freedoms after 2011.

The 2019 election survey analysis produced a two-factor solution (Table 5, columns 7 and 8). The first factor includes F2F, TV, SNS/FB, and photo-video sharing sites, which is similar to the social factor of the revolution study, with the exception of TV. The second factor includes Internet news, newspapers, government sources, and radio, which is a close match with the official sources factor from the revolution study.

5.5 Information sharing behavior

Most respondents reported sharing information with their social networks in all three studies: during the 2011 revolution (77.9%), during the 2014 elections (81%), and during the 2019 elections (80.7%). The information sharing variables are presented in Appendix 1. As with the 2011 and 2014 survey analyses, with the 2019 data we computed a new variable combining the two items for information sharing with family and friends. Since most respondents agree or strongly agree that they regularly shared information about the 2019 elections, we applied a median split to create roughly evenly sized groups to test for differences among respondents reporting “more information sharing” and “less information sharing” with members of their social networks. We discuss additional analyses of information sharing in the next section due to its close relationship with Political Efficacy (PE) both internal and external, and Political Information Efficacy (PIE).

5.6 Political efficacy (PE) and political information efficacy (PIE)

The 2019 PIE items are presented in Appendix 1. We adopted a standard scale for the four PIE items adapted to refer to the 2011 revolution, and the 2014 and 2019 elections (Cronbach alpha $=$ 0.75, 0.67, and 0.69, respectively). To test different information sources as predictors of PIE, we used regression analysis of the frequency of use of different information sources to predict the PIE scale responses. The frequency of Internet use and the frequency of SNS/FB use are significant predictors of PIE for the 2011 revolution and the 2014 election studies, and those two variables plus the use of photo-video sharing sites are significant predictors of PIE in the 2019 elections study (adj. R Square $=$ 0.25, $p<$ 0.05, adj. R Square $=$ 0.19, $p<$ 0.01, and adj. R Square $=$ 0.18, $p<$ 0.01, respectively).

We test for differences in our internal and external political efficacy (PE) and PIE variables with information sharing in one’s social network, using t-tests of the individual PE and PIE variables across the two information-sharing groups (higher, lower) within each survey year. These results are presented in Table 6.

Table 6
T-test differences in PIE, PE by information sharing: Revolution and elections

	2011 Revolution			2014 Elections			2019 Elections
	Share more (a)	Share less (b)	Mean diff. (a-b)	Share more (c)	Share less (d)	Mean diff. (c-d)	Share more (e)	Share less (f)	Mean Diff. (e-f)
PIE better informed	3.44 (0.84)	3.23 (0.85)	0.21	2.68 (0.82)	2.79 (0.90)	0.11	3.25 (0.96)	2.68 (0.95)	0.57 ${}^{*}$
PIE understand issues	2.58 (0.98)	2.61 (0.86)	$-$ 0.03	2.49 (0.93)	2.50 (0.83)	0.02	2.58 (1.16)	2.48 (0.95)	0.11
PIE could advise a friend	2.84 (0.85)	2.54 (0.84)	0.31 ${}^{**}$	2.92 (0.64)	2.34 (0.79)	$-$ 058 ${}^{***}$	2.50 (0.80)	2.64 (0.71)	$-$ 0.14
PIE Feel Q ualified	2.31 (0.86)	2.38 (0.75)	$-$ 0.07	2.59 (0.83)	2.30 (0.65)	$-$ 0.29 ${}^{**}$	2.17 (0.94)	2.30 (0.72)	$-$ 0.13
PE doesn’tcare re people like me	2.79 (0.83)	2.73 (0.76)	0.70	3.08 (0.60)	2.57 (0.63)	$-$ 052 ${}^{***}$	2.75 (0.97)	2.55 (0.73)	0.21
PE so complicated for people like me	2.13 (0.89)	2.04 (0.82)	0.09	2.84 (0.87)	2.18 (0.81)	$-$ 066 ${}^{***}$	2.75 (0.96)	2.42 (0.89)	0.33
PE plenty of ways for people like me	3.18 (0.85)	2.82 (0.81)	0.36 ${}^{**}$	2.67 (0.76)	2.64 (0.80)	$-$ 0.03	3.00 (0.85)	2.82 (0.75)	0.18

${}^{*}p<$ 0.1, ${}^{**}p<$ 0.05, ${}^{***}p<$ 0.01.

In the 2011 revolution study, there are significant differences between the level of information sharing and one of the four measures of PIE: “If a friend asked me about political developments during the 2011 uprising, I feel I had enough information to help my friend figure out what’s going on.” The higher information-sharing group is significantly higher, but only at $p<$ 0.1 (Table 6). This same measure of PIE in the 2014 elections study, is significantly higher for respondents in the higher information-sharing group, as well as the PE measure: “I consider myself well qualified to participate in politics in the current Tunisian government.” For the 2019 elections study, there are significant differences between the information sharing groups on one PIE question “I think I am better informed about Tunisian politics and the current government than most people.”

We suspect that the relatively high levels for the PE variables in both information sharing groups could be masking differences in PIE among the groups. To check this, we conducted t-tests to compare the means for information sharing, PE and PIE across the three surveys (Table 7). The mean differences for information sharing variables in the revolution period are significantly lower than in both the 2014 and 2019 elections (bottom of Table 7). This indicates that more respondents reported sharing information during the revolution, than during the two periods of the elections.

Table 7

Mean differences: PIE, PE and information sharing: Revolution and elections

	2011 revolution (a)	2014 election (b)	2019 election (c)	Mean diff. (a-b)		Mean diff. (a-c)		Mean diff. (b-c)
PIE better informed than most	2.35 (0.85)	2.37 (0.88)	2.27 (0.75)	$-$ 0.	02	0.	08	0.	10
PIE understand issues	2.76 (0.79)	2.69 (0.66)	2.56, (0.78)	0.	07	0.	10	0.	13
PIE could advise a friend	2.66 (0.85)	2.48 (0.79)	2.60 (0.74)	0.	18 ${}^{*}$	0.	59	$-$ 0.	12
PIE feel qualified	2.08 (0.85)	2.33 (0.86)	2.34 (0.95)	$-$ 0.	25 ${}^{**}$	$-$ 0.	37 ${}^{***}$	$-$ 0.	12
PE Govt doesn’t care about …me	3.32 (0.84)	2.76 (0.88)	2.76 (0.96)	0.	55 ${}^{***}$	$-$ 0.	55 ${}^{***}$	0.	00
PE Too complicated for people like me	2.59 (0.92)	2.50 (0.92)	2.50 (0.97)	0.	09	0.	09	0.	00
PE Plenty of ways for people like me	2.98 (0.84)	2.65 (0.79)	2.81 (0.78)	0.	33 ${}^{***}$	0.	16	$-$ 0.	17
Info sharing: You with family	3.15 (0.86)	2.76 (0.86)	2.58 (0.75)	0.	39 ${}^{***}$	0.	56 ${}^{***}$	0.	19 ${}^{*}$
Info sharing: Family with you	3.14 (0.80)	2.78 (0.89)	2.75 (0.81)	0.	35 ${}^{***}$	0.	39 ${}^{***}$	0.	04

${}^{*}p<$ 0.1, ${}^{**}p<$ 0.05, ${}^{***}p<$ 0.01.

Results for PIE and PE variables for each of the three surveys are also shown in Table 7. The internal political efficacy (internal PE) variable “I consider myself well qualified to participate in politics $\ldots$ ” is significantly higher in both of the 2014 and 2019 elections studies than in the revolution study. The external PE variable “I don’t think officials in the current Tunisian government care much what people like me think” was significantly higher during the revolution than both of the election periods. There are no significant differences between the election studies in any of the PIE or PE variables, which suggests a sustained pattern of media use and information reliability after the revolution.

To further test for differences in information sharing behavior, PE and PIE, we combined all the responses from respondents to these questions from all three surveys (see Table 8). For almost all PE and PIE questions, participants who report more information sharing are higher on measures of PIE and PE than those who report less information sharing.

Table 8

Differences in PIE, PE and information sharing: All respondents combined

	Information sharing
	More sharing (a)	Less sharing (b)	Mean diff. (a-b)
PIE better informed than most	3.16 (0.91)	2.89 (0.92)	0.27 ${}^{***}$
PIE understand issues	2.55 (0.98)	2.53 (0.87)	0.02
PIE could advise a friend	2.83 (0.76)	2.48 (0.79)	0.35 ${}^{***}$
PIE feel qualified	2.39 (0.86)	2.32 (0.70)	0.06
PE Govt doesn’t care about …me	2.88 (0.79)	2.61 (0.70)	0.27 ${}^{***}$
PE Too complicated for people like me	2.43 (0.95)	2.20 (0.84)	0.24 ${}^{**}$
PE Plenty of ways for people like me	2.99 (0.85)	2.74 (0.79)	0.25 ${}^{***}$

${}^{*}p<$ 0.1, ${}^{**}p<$ 0.05, ${}^{***}p<$ 0.01.

Significant differences exist for two of four the PIE questions: ‘I feel I have enough information to advise a friend’ and ‘I think I am better informed about Tunisian politics than most people’; and for all three PE questions: ‘sometimes politics and government seem so complicated that persons like me can’t truly understand what’s going on’, ‘I don’t think officials in the current government care much what people like me think’, and ‘there are plenty of ways for people like me to have a say in what the current Tunisian government does’. This indicates that respondents with higher levels of PE and PIE also exchange information more often with more members of their social networks.

6. Limitations of the study

One limitation of our study is that our participants in the 2019 elections study completed surveys in April ( $N=$ 84) and November 2020 ( $N=$ 31). While April 2020 is not an unreasonable amount of time after the September-October 2019 elections, the November group completed surveys a full year after the 2019 elections. This means some of our respondents’ recollections may not have been as accurate as they might otherwise have been. As noted above, we tried to mitigate this limitation by inviting participants to remember their communication behavior and perceptions during the 2019 election period as best as they could. We conducted $t$ -tests of the responses on all variables across both collections for the third study to test for any differences.

Another limitation is that our samples of survey participants are not representative of Tunisian society as a whole. While they are also not representative of young Tunisian adults as a whole, we have shown how their use of the Internet and social media is only slightly above the national averages for their age cohort (18–29). Since our sample is from a pool of university undergraduate and graduate students (in a marketing and business administration department), they are above average in education, which accounts for some of the differences with their age cohort. Our study can only provide some insight into the behavior and political impacts of some young, educated Tunisians. This is an important population segment of Tunisian society, however, as young, educated adults have been active and instrumental throughout the uprising and the transition to a new form of government.

7. Discussion

Political conditions affect citizens’ and organizations’ willingness and opportunity to participate in the civic life of a nation (Dahl, 1989; Almond & Verba, 1963; Verba & Nie, 1972). In the case of post-authoritarian and transitional governments, the legitimacy of a new government (as perceived by citizens as a form of external political efficacy) is among the known predictors of political participation in established democracies (Jebril et al., 2013; Breuer & Groshek, 2014; Zayani, 2015; Karolak, 2017). Other known predictors include keeping up with news and political discussion (i.e., sharing information).

During the first three years after the 2011 revolution that overthrew an authoritarian regime, Tunisia formed a transitional government, adopted a new Constitution, and increased freedom of the press and free speech. In Fall 2014, the country elected members of parliament and held its first regular presidential elections – the first free and fair presidential elections since independence from French colonialism in 1956.

Access to reliable news, and opportunities for discussion through information sharing face-to-face, by telephone and on social media, became critical factors not only during the chaos of the Tunisian uprising and revolution, but also during the transition toward potentially more democratic processes, practices, and institutions (Gibson, 2001; Al-Ani et al. 2012; Howard, 2011; Wang & Mark, 2013). These critical factors are known to strengthen political efficacy (internal and external), as well as political information efficacy among citizens (Martin et al., 2018; Karolak, 2017). Studies of established democracies find that increases in political information efficacy lead to increases in political participation overall (Kaid et al., 2007; Zúñiga et al., 2018).

Our study findings contribute to understanding the role of information and communication technology in the political participation of young, educated adults in post-authoritarian Tunisia. More of our respondents in both post-revolution election surveys reported they believe that officials in the government care about what they think than did respondents in the 2011 revolution survey. It is notable that the uprising began in part due to citizens’ frustration with an unresponsive government (Zayani, 2015; Breuer & Groshek, 2014; Richter, 2017).

We attribute the differences in information reliability values to greater press freedom after the revolution that allowed previously restricted sources to provide more independent, less censored, content. That is, these sources gained more legitimacy for respondents. As a result, government sources, national newspapers and Internet news sites were perceived as more reliable than during the revolution. We suggest that due to the lack of reliable information during the uprising and revolution period of 2010–11, respondents relied more on their social networks for news and updates about political developments. As the information sources, including government sources, became more reliable after the revolution when censorship was relaxed and the press was more open, respondents had greater access to reliable information from official sources of 2014 and 2019 election news.

It is possible, although we do not have any evidence in our data, that the drop in perceived reliability of social media from its high in our 2011 revolution study to 2014 and 2019 elections studies, may be due to users’ concerns about misinformation. US studies in recent years have pointed to the threats to democracy of social media spreading misinformation (Marwick & Lewis, 2017; Luther, et al., 2019). In the 2019 elections, Elswah and Howard (2020, p. 2) found that “Tunisian civil society organisations raised concerns about the spread of disinformation and polarising content on Facebook. Civil society organisations found that there were orchestrated campaigns on Facebook to discredit candidates and spread hate speech before and during the 2019 presidential elections.” Ritcher (2017) noted that during 2012 and 2013, a bulk of quite alarmist reports voiced serious concerns about the transition facing an authoritarian backlash stemming from the vision of liberal democratic media that did not seem to materialize when the Troika transitional government under the leadership of the Islamist Ennahda party had taken over power in late 2011 until 2014.

However, we attribute the drop across our surveys in perceived reliability of social media between the 2011 uprising and the 2014 elections, to the rise in new press freedoms for broadcast and print media (TV, radio, newspapers) rather than misinformation in social media. Moreover, much of the misinformation in social media started to appear around the 2016 US elections. In any case, our respondents reported an increase in the reliability of social media by 2019 which is indicative of media of all types coming into greater overall balance with other sources, including government sources, in terms of their reliability.

Tunisia has many typical weaknesses of post-authoritarian states (e.g., lingering corruption, non-transparent funding of political parties, need for regulatory reform), but it has persevered in some of the major institution building processes that facilitate a transition to a more participatory democracy. Among these are the opening of the press and other sources of information (online and offline) and allowing greater freedom of expression. In the process, over time, the young, educated adults in our study show greater trust in government information sources, and judge those sources to be increasingly more accurate and trustworthy over time.

Overall, our survey participants reporting higher levels of information reliability and information sharing also report higher levels of political efficacy and political information efficacy. This is consistent with other studies finding that information reliability and information sharing predict political information efficacy, leading to further political participation (Kaid et al, 2007; Michaelson, 2000; Zúñiga et al., 2017). Given the increased legitimacy of the Tunisian government, despite shortcomings and setbacks, our findings suggest that communication behavior and political participation among young, educated adults, contribute to a slow transition toward greater democratic practices.

8. Conclusion

In three survey studies covering key political developments and events in Tunisia (2011 revolution, 2014 elections, and 2019 elections) we examined changes over time in respondents’ use of different sources of political information and the perceived reliability of those sources, information sharing behavior, and respondents’ sense of being politically informed (political information efficacy). Each of the three survey studies invited a comparable, but different, opportunity sample of university students to participate, as the country transitioned from authoritarianism to a new form of government.

Our study is unique in that it contributes an empirical analysis of key predictors of political participation, using the same survey over an extended period (2011–2019) with different cohorts of similar university students of the authors. We contribute to survey findings on important constructs measuring communication behavior and effects among young, educated adults in Tunisia.

We ask in this study whether young, educated Tunisians report changes over time in their use of different political information sources, their perceived reliability of these sources, their sharing of information, and their feeling of being politically well-informed. Answers to these questions suggest a tendency toward a strengthening and consolidation of democracy.

The perceived reliability of government controlled information sources available during the 2010–11 uprising/revolution was lower than online sources, especially social network sites (SNS), such as Facebook (FB), and face-to-face and telephone communication. This is consistent with other studies of information reliability in authoritarian states (Gibson, 2001; Al-Ani, 2012; Wang & Mark, 2013). Between the 2011 revolution and the 2014 elections, the use of different information sources changed significantly, with more use and perceived reliability of official sources than of social media sources, as press freedoms increased.

The change in the percentage of respondents reporting which TV stations they watched during the elections to obtain political information is in stark contrast to the pattern during the revolution, when the large majority of respondents reported they watched Al Jazeera, followed distantly by the National TV channels, and Al Arabiya. During the 2014 and 2019 elections, the percentage of respondents watching Al-Jazeera plummeted as national and government TV sources became freer and more open after the revolution. Al-Jazeera was also providing less coverage of elections in Tunisia than other sources. Tunisia’s ‘TV Nationale’ and a new, private channel (El-Hiwar Ettounsi) maintained high percentages of viewers during the 2014 and 2019 elections.

The perceived reliability (i.e., accuracy, credibility) of government information sources rose significantly between the 2011 revolution and the 2014 elections, and stayed higher in the 2019 elections. The reliability of SNS/FB and video sharing sites, by contrast, fell significantly from the 2011 revolution survey to the 2014 election survey, but was not significantly different between the 2014 and 2019 election surveys. That is, the pattern of information reliability among different sources stabilized between 2014 and 2019 studies, confirming a new pattern.

We have confirmatory analyses for the use of information sources and their perceived reliability between the 2014 and 2019 elections surveys, that is, the 2019 results are generally similar to the 2014 results. Given that the favorable conditions for political participation have endured, despite some violent disruptions over time (e.g., assassinations, terrorist attacks), the official sources have continued to be popular and perceived to be generally reliable for political news and election information.

Interestingly, the two election surveys are significantly different on use measures of SNS, photo-video sharing sites, TV, and Internet news sites. The 2014 survey shows lower frequency of use for SNS, video sharing sites, and Internet news sites, and less TV use than for the 2019 elections. This suggests a reinstatement of the value – a ‘bouncing back’ – of social media and Internet news sites as sources of political information, following their decreased use in the 2014 election survey. That is, the government sources and the non-government sources are more in balance in terms of providing reliable political information among our respondents.

Given our findings that the 2019 measures of information use, sharing, reliability and political information efficacy are largely confirmatory of those in 2014, we note that the pattern has been fairly stable while becoming more balanced across government, private and social network sources of political information. These findings suggest, like prior studies in the US showing political expression leading people to participate politically further in other forms, both online and in person, there may be some strengthening of these democratic practices in Tunisia.

In future work, we would like to conduct additional survey studies in order to track media use and effects related to political participation. We would also like to incorporate some of the measures from related studies on media use, information reliability, and political efficacy with possible effects on democratic transition and consolidation. In future work, we would also like to expand our sample to a more representative population of young, educated Tunisians.

Footnotes

Acknowledgments

We would like to thank our colleagues Edward A. Fox, Donald Shoemaker, John Tedesco, and Sunshin Lee for their assistance and guidance with various aspects of the work described in this paper. We are grateful for support from the National Science Foundation for the ‘Global Event and Trends Archiving Research’ project (III-479619) under the auspices of which this longitudinal study was developed.

Appendix: Item name and text

The items listed in this appendix are for the 2019 election study. For the 2014 election study the only differences in the items are the year of the election. For the 2011 revolution study the items used the phrase “the uprising” instead of the phrase “the 2019 elections”

Name	Description
	Use of information sources
TVUse	Which TV channels did you most often watch for information relating to the legislative and presidential elections (September, October and November 2019)? –Please check all that apply to your experience– Al Arabiya Al Jazeera National Television (El watania 1 and / or El watania2) BBC Arabic Hannibal Nessma El-Hiwar Ettounsi Attessia TVOther
RadioUse	Which radio stations did you most often listen to for information relating to the legislative and presidential elections (September, October and November 2019)? –Please check all that apply to your experience– National Radio Radio-Tunisienne ChaÃ®ne Internationale (RTCI) FM mosaic Shems FM Express FM Zitouna FM Radio Regional Tatouine/Monastir/Sfax/El Kef IFM Other
NPUse	During the legislative and presidential elections (September, October and November 2019), which newspapers did you read most often (not online)? –Please check all that apply to your experience– No Assabah Essahafa The Press Time Echorouk Tunis Africa Press Agency Other newspapers not listed above

Name	Description
	Use of information source frequency
intFreq ${}^{\text{a}}$	How often did you use the Internet during legislative and presidential elections (September, October and November 2019) to learn about candidates, their ideas, positions and political events?
twFreq ${}^{\text{a}}$	How often did you use Twitter, during legislative and presidential elections (September, October and November 2019) to learn about candidates, their ideas, positions and political events?
SNSFreq ${}^{\text{a}}$	How often did you use social networking sites, e.g., Facebook, during legislative and presidential elections (September, October and November 2019) to learn about candidates, their ideas, positions and political events?
vdFreq ${}^{\text{a}}$	How often did you use video sharing sites, e.g., YouTube, during legislative and presidential elections (September, October and November 2019) to learn about candidates, their ideas, positions and political events?
blgwrfr ${}^{\text{a}}$	How often were you writing on blogs during legislative and presidential elections (September, October and November 2019) to learn about candidates, their ideas, positions and political events?
blgrdfr ${}^{\text{a}}$	How often were you reading blogs during legislative and presidential elections (September, October and November 2019) to learn about candidates, their ideas, positions and political events?
	Reliability of information sources
SNSRel ${}^{\text{b}}$	How reliable was the political information you have received from social networking sites, e.g., Facebook, you used during the 2019 elections?
VshrRel ${}^{\text{b}}$	How reliable was the political information you have received from video sharing sites, e.g., YouTube, you used during the 2019 elections?
IntRel ${}^{\text{b}}$	How reliable was the political information you have received from Internet news sites, e.g., sites of Television or Newspaper sites, you used during the 2019 elections?
GovRel ${}^{\text{b}}$	How reliable was the political information you have received from government sources you used during the 2019 elections?
TvRel ${}^{\text{b}}$	How reliable was the political information you have received from television stations you watched during the 2019 elections?
RaRel ${}^{\text{b}}$	How reliable was the political information you have received from radio channels you listened to during the 2019 elections?
NPRel ${}^{\text{b}}$	How reliable was the political information you have received from the newspapers you read during the 2019 elections?
F2FRel ${}^{\text{b}}$	How reliable was the political information you have received from face-to-face conversations you had during the 2019 elections?
	Internet information sharing
SGSA1 ${}^{\text{c}}$	I regularly shared the information I obtained from the Internet (including YouTube Facebook, Twitter, etc.) with family and friends during the legislative and presidential elections (October and November 2019)?
SGSA2 ${}^{\text{c}}$	Members of my family and friends have regularly shared with me information obtained from the Internet (including Facebook, YouTube or Twitter) during the legislative and presidential elections (October and November 2019).
	Information availability
Need ${}^{\text{c}}$	In general, I have been able to get the information I needed in the past year.
Accurate ${}^{\text{c}}$	Overall, over the past year, the information I have been able to obtain has been accurate
	Political efficacy (internal and external) and political information efficacy
PIE1 ${}^{\text{c}}$	I consider myself well qualified to participate in politics of the current Tunisian government.
PIE2 ${}^{\text{c}}$	I think I am better informed about Tunisian politics and the current government than most people.
PIE3 ${}^{\text{c}}$	I feel that I have a pretty good understanding of the important political issues facing Tunisia.
PIE4 ${}^{\text{c}}$	If a friend had asked me about the 2019 Tunisian elections and related political events, I feel I would have enough information to help my friend figure out whom to vote for.
PE1 ${}^{\text{c}}$	Since the first government of Mohamed Ghannouchi, I feel that sometimes politics and government seem so complicated that persons like me can’t truly understand what’s going on.
PE2 ${}^{\text{c}}$	I don’t think officials in the current Tunisian government care much what people like me think.
PE3 ${}^{\text{c}}$	There are plenty of ways for people like me to have a say in what the current Tunisian government does.

Name	Description
	Demographic and other variables
Gender	Gender What is your gender? (Male/female)
Marital	What is your marital status? (Single/Married/Divorced/Widower)
Age	How old are you? (Enter a number of years)
Proj	What is your profession?
Education	What is the highest level of secondary or university education you have completed? (High School Diploma (technician, commercial, industrial)/ Undergraduate degree/Graduate degree (Master or Doctorate))
IntYrs	How many years have you been using the Internet? (enter a number of years, if you have used the Internet for less than a year enter (1)):
Lang	What language (s) did you use most often in your search for information on the Internet (including social media or Twitter) about the legislative and presidential elections (September, October and November 2019)? –Please check all that apply to your experience– French Arab Franco-Arab English Other
CPOwn SmrtCP	Do you have your own cell phone? (Yes/No)If you answered “yes” to question above, are you using a Smartphone (for example: iPhone, Blackberry, Android, etc.) (Yes/No/ I do not know)

${}^{\text{a}}$ Measured on an eight-point scale of: almost never, < once per month, about once per month, several times per month, about once per week, several times per week, about once per day, several times per day. ${}^{\text{b}}$ Measured on a four point scale of: 0 $=$ not at all reliable, 1 $=$ unreliable (the information was not generally reliable), 2 $=$ moderate (Information was reliable from time to time), 3 $=$ high (the information was reliable most of the time). ${}^{\text{c}}$ Measured on a four-point scale of: strongly disagree, disagree, agree, strongly agree.

References

Al-Ani

Mark

Chung

, & Jones

(2012). The Egyptian Blogosphere: A counter-narrative of the revolution. Proceedings of the 2012 ACM Conference on Computer Supported Cooperative Work (CSCW ’12), 17-26. ACM Digital Library. doi: 10.1145/2145204.2145213.

Almond

G.A.

, & Verba

(1963). The Civic Culture: Political attitudes and democracy in five nations. Princeton University Press.

Arieff

(2020). Tunisia: In Brief. Congressional Research Service. https://www.everycrsreport.com/files/20200316_RS21666_5a6e0d290f8920461bf85c50138f53d81c407327.pdf.

Associated Press. (2014, October 30). Tunisia Election Results: Nida Tunis wins most seats sidelining Islamists. The Guardian. https://www.theguardian.com/world/2014/oct/30/tunisia-election-results-nida-tunis-wins-most-seats-sidelining-islamists.

Bellin

(2012). Reconsidering the Robustness of Authoritarianism in the Middle East: Lessons from the Arab Spring, Comparative Politics, 44(2), 127-49.

Bennett

(1998). The Media and Democratic Development: The social basis of political communication. In O’Neil

P.H.

(Ed.), Communicating Democracy: The Media and Political Transitions. Lynne Rienner Publishers.

Boyd

Golder

, & Lotan

(2010). Tweet, Tweet, Retweet: Conversational Aspects of Retweeting on Twitter. Proceedings of the 43rd Hawaii International Conference on System Sciences (HICSS-43). IEEE Computer Society.

Breuer

, & Groshek

(2014). Online Media and Offline Empowerment in Post-Rebellion Tunisia: An Analysis of Internet Use During Democratic Transition. Journal of Information Technology & Politics, 11(1), 25-44.

British Broadcasting Corporation. (2014, December 22). Tunisian Election: Essebsi wins presidential run-off. Retrieved March 8, 2015 from https://www.bbc.com/news/world-middle-east-30579662.

10.

British Broadcasting Corporation. (2019, October 14). Tunisia election: Kais Saied to become president. Retrieved January 11, 2021 from https://www.bbc.com/news/world-africa-50032460.

11.

Browne

(1999). International Broadcasting in Arabic: Tunisia. Iowa State University Press, 261-278.

12.

Chadwick

(2006). Internet politics: States, citizens, and new communication technologies. Oxford University Press.

13.

Chun

S.A.

, & Noveck

B.S.

(2020). Editorial: Digital Democracy: 25 Years Look Back, Look Ahead. Digital Government: Research and Practice, 1(1) Article 1: 1-3. doi: 10.1145/3374904.

14.

Dahdal

(2012). Social Media and the Arab Spring: The historical context and the role of Aljazeera satellite station. Donau-Universitat Krems, Vienna.

15.

Dahl

R.A.

(1989). Democracy and its Critics. Yale University Press.

16.

Datareportal (2019). Digital 2019: Tunisia. https://datareportal.com/reports/digital-2019-tunisia.

17.

Dennis

E.E.

Martin

J.D.

, & Hassan

(2019). Media use in the Middle East, 2019: A seven-nation survey. Northwestern University in Qatar. https://ww.mideastmedia.org/survey/2019.

18.

Dhillon

, & Youssef

(2009). Generation in Waiting: The unfulfilled promise of young people in the Middle East. Brookings Institution, Washington, DC.

19.

Egel

, & Garbouj

(2013). Socioeconomic drivers of the spread of the Tunisian revolution. RAND Corporation and the Graduate Institute.

20.

Elswah

, & Howard

P.N.

(2020). Challenges of Monitoring Social Media in the Arab World: The case of the 2019 Tunisian elections. COMPROP Data Memo 2020.1/23 March, 2020. https://demtech.oii.ox.ac.uk/wp-content/uploads/sites/93/2020/03/Tunisia-memo-English.pdf.

21.

Gabsi

(2019), Tunisia’s youth: Awakened identity and challenges post-Arab Spring. British Journal of Middle Eastern Studies, 46(1), 68-87.

22.

Garon

(2003). Dangerous Alliances: Civil society, the media and democratic transition in North Africa. Zed Books.

23.

Ghannam

(2012). Digital media in the Arab world one year after the revolutions. Center for International Media Assistance (CIMA), National Endowment for Democracy, Washington, DC.

24.

Ghareeb

(2000). New Media and the Information Revolution in the Arab World. The Middle East Journal, 54(3), 395-418.

25.

Gibson

J.L.

(2001). Social Networks, Civil Society, and the Prospects for Consolidating Russia’s Democratic Transition. American Journal of Political Science, 45(1), 51-68.

26.

Granovetter

(1973). The Strength of Weak Ties. American Journal of Sociology, (78)6, 1360-1380.

27.

Grewal

(2019). Tunisian democracy at a crossroads. Policy Brief. Washington, DC: The Brookings Institution.

28.

Haugbølle

& Cavatorta

(2012). Vive la grande famille des médias Tunisiens: Media reform, authoritarian resilience and societal responses in Tunisia. Journal of North African Studies, 17(1), 97-112.

29.

Honwana

(2013). Youth and Revolution in Tunisia. Zed Books.

30.

Howard

(2011). The Digital Origins of Dictatorship and Democracy: Information Technology and Political Islam. Oxford University Press.

31.

Howard

P.N.

& Hussain

M.M.

(2013). Democracy’s Fourth Wave? Digital media and the Arab Spring. Oxford University Press.

32.

International Foundation for Electoral Systems. Election Guide: Democracy assistance and election news: Country Profile Tunisia. https://www.electionguide.org/countries/id/217.

33.

Jebril

Stetka

, & Loveless

(2013). Media and Democratisation: What is known about the role of mass media in transitions to democracy? University of Oxford. https://www.politics.ox.ac.uk/news/media-and-democratisation-what-is-known-about-the-role-of-mass-media-in-transitions-to-democracy.html.

34.

Kaid

L.L.

McKinney

M.S.

, & Tedesco

J. C.

(2007). Political information efficacy theory and younger voters. American Behavioral Scientist, 50, 1093-1111.

35.

Karolak

(2017). The Use of Social Media from the Revolution to Democratic Consolidation: The Arab Spring and the case of Tunisia. Journal of Arab & Muslim Research, 10(2), 199-216.

36.

Kavanaugh

(1998). The Social Control of Technology in North Africa: Information in the global economy. Praeger/Greenwood Press.

37.

Kavanaugh

Sheetz

Sandoval-Almazan

Tedesco

, & Fox

(2016). Media Use during Conflicts: Information seeking and political efficacy during the 2012 Mexican elections. Special Issue on Smart Government. Government Information Quarterly, 33(3): 595-60.

38.

Kavanaugh

Sheetz

Hassan

Yang

Elmongui

Fox

Magdy

, & Shoemaker

(2013). Between a Rock and a Cell Phone: Communication and Information Technology Use during the 2011 Uprisings in Tunisia and Egypt. International Journal of Information Systems for Crisis Response and Management, 5(1), 1-21.

39.

Kavanaugh

Sheetz

Skandrani

Tedesco

Sun

, & Fox

E.A.

(2016). The use and impact of social media during the 2011 Tunisian Revolution. Proceedings of the 17th Annual International Conference on Digital Government Research, 20–30. ACM Digital Library. doi: 10.1145/2912160.2912175.

40.

Kavanaugh

Sheetz

Skandrani

, & Fox

E. A.

(2017). Media use by young Tunisians during the 2011 revolution and 2014 elections. Information Polity, 22(2-3), 137-158.

41.

Kéfi

(2015). Tunisia: Civil society, the driving force behind the democratic transition, IEMed (Ed.), Mediterranean Yearbook 2015, IEMed, Barcelona, 237-40. http://www.iemed.org/observatori/arees-danalisi/arxius-adjunts/anuari/med. 2015/IEMed%20Yearbook%202015_Panorama_TunisiaDemocraticalTransition_RidhaKefi.pdf.

42.

Kim

Wyatt

& Katz

(1999). News, Talk, Opinion, Participation: The part played by conversation in political participation. Political Communication, 16, 361-385.

43.

Klaas

& Dirsus

(2018). Non-Governmental Organizations, International Organizations, and Civil Society in Tunisia. NATO OPEN Publications, 2(1), 1-20. https://reliefweb.int/sites/reliefweb.int/files/resources/b3eb9d_3bbd6c2ab6564a3e980b221f565f5175.pdf.

44.

Lotan

Graeff

Ananny

Gaffney

Pearce

& Boyd

(2011). The revolutions were tweeted: Information flows during the 2011 Tunisian and Egyptian revolutions. International Journal of Communication, 5, 1375-1405.

45.

Luther

Kavanaugh

Thebault-Spieker

, & Antin

(2020). Guest Editorial for Special Issue on Negotiating Truth and Trust in Socio-technical Systems, in Transactions on Social Computing, 3(1). ACM Press.

46.

Malhotra

N.K.

Birks

D.F.

, & Wills

(2012). Marketing Research – An Applied Approach, fourth edition. Pearson Education Limited.

47.

Martin

J.D.

Martins

R.J.

, & Naqvi

(2018). Media use predictors of online political efficacy among Internet users in five Arab countries. Information, Communication & Society, 21(1), 129-146.

48.

Marzouki

Skandrani-Marzouki

Bejaoui

Hammoudi

, & Bellaj

(2012). The contribution of Facebook to the 2011 Tunisian revolution: A cyber-psychological insight. Cyberpsychology, Behavior and Social Networking, 15(5), 237-244.

49.

McCarthy

(2019). The politics of consensus: al-Nahda and the stability of the Tunisian transition. Middle Eastern Studies, 55(2), 261-275.

50.

Meikeljohn

(1948). Free Speech and its Relation to Self-Government. Harper & Brothers.

51.

Michaelson

(2000). Political efficacy and electoral participation of Chicago Latinos. Social Science Quarterly, 81(1), 136-150.

52.

Miladi

(2021). Tunisia: The Transformative Media Landscape after the Revolution. In Richter

and Kozman

(Eds.) Arab Media Systems, Open Book Publishers, 267-284. doi: 10.11647/OBP.0238.16.

53.

Moy

, & Gastil

(2006). Predicting deliberative conversation: The impact of discussion networks, media use and political cognitions. Political Communication, 23, 443-460.

54.

National Institute of Statistics. Tunisia. Institut National des Statistiques (National Institute of Statistics). http://www.ins.tn/en/themes/population.

55.

Pew Research Center. (2019). Mobile Connectivity in Emerging Economies. https://www.pewresearch.org/internet/2019/03/07/mobile-connectivity-in-emerging-economies/.

56.

Pingree

(2007). How messages affect their senders: A more general model of message effects and implications for deliberation. Communication Theory, 17(4), 439-461.

57.

Radcliffe

, & Abuhmaid

(2020). Social Media in the Middle East: 2019 in Review. SSRN Electronic Journal. doi: 10.2139/ssrn.3517916.

58.

Richter

(2017). Media policy in times of transition. Publizistik, 62, 325-337. doi: 10.1007/s11616-0170340-x.

59.

Rojas

, & Puig-i-Abril

(2009). Mobilizers mobilized: Information, expression, mobilization and participation in the digital age. Journal of Computer-Mediated Communication, 14(4), 902-927.

60.

Rothenbuhler

(1991). The process of community involvement, Communication Monographs, 58(1), 63-78.

61.

Salem

, & Mourtada

(2014). Citizen engagement and public services in the Arab world: The potential of social media. Mohammed Bin Rashid School of Government, Dubai, UAE.

62.

Scholl

H.J.

(2020). Digital Government: Looking Back and Ahead on a Fascinating Domain of Research and Practice. Digital Government: Research and Practice, 1(1) Article 7: 1-12. doi: 10.1145/3352682.

63.

Schuler

(2020). Can the Internet Support Democracy? Digital Government: Research and Practice, 1(1), Article 3: 1-14. doi: 10.1145/3352462.

64.

Sghaier

Skandrani

, & Robson

(2021). Repairing Political Trust in Tunisia. Qualitative Market Research: An International Journal, in press.

65.

Shane

P. M.

(2004). Democracy Online: The prospects for political renewal through the Internet, Routledge, New York, NY.

66.

Siebert

F. S.

Peterson

& Schramm

(1963). Four Theories of the Press, University of Illinois Press, Urbana, IL.

67.

Skandrani

Kooli

, & Doudech

H.N.

(2020). Inhibitors of Non-for Profit Organisations’ activities and survival in a crisis context, Qualitative Market Research: An International Journal, forthcoming.

68.

Starbird

& Palen

(2012). (How) will the revolution be retweeted? Information diffusion and the 2011 Egyptian uprising. In Proceedings of the 2012 Conference on Computer Supported Cooperative Work (CSCW ’12) (Seattle, WA, February 11-13). ACM, New York, NY, 7-16.

69.

The World Bank. World Development Indicators (various years 2010–2019). International Bank for Reconstruction and Development (IBRD), Washington, DC.

70.

Thomas

(2018). Civic Engagement in the Middle East and North Africa, Arab Barometer. Retrieved January 17, 2020, from https://www.arabbarometer.org/wp-content/uploads/PublicOpinion_CivicEngagement_MiddleEast_NorthAfrica_2018.pdf.

71.

Tunis Afrique Presse. (2019, May 15). Unemployment rate down 0.1% to 15.3% in Q1 2019 (INS).” https://www.tap.info.tn/en/Portal-Society/11451211-unemployment-rate.

72.

Verba

, & Nie

N.H.

(1972). Participation in America: political democracy and social equality. Harper & Row, New York, NY.

73.

Wang

, & Mark

(2013). Trust in online news: comparing social media and official media use by Chinese citizens. In Proceedings of the 2013 conference on Computer Supported Cooperative Work (CSCW ’13). ACM Press, 599-610.

74.

Wulf

Mitase

Atam

Randall

, & Rohde

(2013). ‘On the Ground’ in Sidi Bouzid: Investigating social media use during the Tunisian revolution. In Proceedings of the 2013 conference on Computer Supported Cooperative Work (CSCW ’13). ACM Press, 1409-1418.

75.

Yerkes

(2019). Tunisia’s Elections, Explained: The world’s youngest democracy votes. Carnegie International Endowment for Peace. https://carnegieendowment.org/publications/interactive/tunisian-elections-2019.

76.

Zayani

(2015). Network Publics and Digital Contention: The Politics of Everyday Life in Tunisia. Oxford, UK: Oxford University Press.

77.

Zúñiga

H.G.D.

Jung

, & Valenzuela

(2012). Social media use for news and individuals’ social capital, civic engagement and political participation. Journal of Computer Mediated Communication, 17 (2012), 319-336.

78.

Zúñiga

H.G.D.

Trevor

, & Ardévol-Abreu

(2017). Internal, External, and Government Political Efficacy: Effects on News Use, Discussion, and Political Participation, Journal of Broadcasting & Electronic Media. 61(3), 574-596.

79.

Zúñiga

H.G.D.

Liu

J.H.

Barnidge

Huber

(2018). Social Media as a Sphere for “Risky” Political Expression: A twenty-country multilevel comparative analysis. The International Journal of Press/Politics, 23(2), 161-182.

Media use,information reliability and political efficacy in Tunisia,2011–2019

Abstract

Keywords

1. Introduction

2. Political participation theory and related work

3. Political and economic context

3.1 Parliamentary and presidential elections (2014 and 2019)

1 PDL French acronym of Parti Destourien Libre.

4. Methodology

4.1 Survey samples for studies

4.2 Survey instrument for all three studies

4.3 Data analyses

5. Results

5.1 Demographics

5.2 Internet use for all purposes

Table 1a Our sample compared with all adults and young adults 2013–2019

5.3.1 Broadcast media use: Revolution vs. elections

Table 2 TV stations watched most often during revolution vs elections

5.4 Perceived reliability of information sources

Table 3 Mean reliability of information sources during revolution vs elections

5.6 Political efficacy (PE) and political information efficacy (PIE)

Table 6 T-test differences in PIE, PE by information sharing: Revolution and elections

7. Discussion

8. Conclusion

Footnotes

Acknowledgments

Appendix: Item name and text

References

¹
PDL French acronym of Parti Destourien Libre.

Table 1a
Our sample compared with all adults and young adults 2013–2019

Table 2
TV stations watched most often during revolution vs elections

Table 3
Mean reliability of information sources during revolution vs elections

Table 6
T-test differences in PIE, PE by information sharing: Revolution and elections