An ally to the people: EU membership conditionality and Albanians’ attitudes toward EU membership

Do people's perceptions of the European Union (EU) membership conditionality's role in their country's democratization affect their support for their country's EU membership? If so, this would be an indicator that people view the EU as a democratization factor, an argument often overlooked by literature that links support for EU membership primarily with economic expectations (Tucker et al. 2002; Palmer and Whitten 1999; Gabel 1998). Thus far, the existing elite-centered approach in designing, applying and studying EU membership conditionality has failed to understand and account for the powerful attraction that the EU represents to the masses in EU membership aspiring countries from the former communist Eastern Europe. Often EU officials have shown little interest in directly engaging the public in the receiving end of the EU membership conditionality. Putting much faith on EU norm institutionalization in EU membership aspiring countries, EU enlargement policies have relegated to domestic governments the process of their societies’ socialization with those norms (Peshkopia 2014a). This approach might have denied EU membership conditionality tremendous potentials in driving not only policy change but also social change in countries that aspire to join the Union.

Policy makers and scholars alike have failed to see EU membership conditionality as a major ally to people's aspirations to improve the governance in countries that aspire to join the EU. Therefore, one should expect people who believe that EU membership conditions help their country's democratization to support their country's EU membership. While this seems a logically solid argument, previous research has demonstrated that concerns for national sovereignty might curb support for EU membership (Arikan 2012; Hooghe and Marks 2005; McLaren 2004). Moreover, arguably, concerns over economic development might top concerns over democratization and good governance (Anderson and Reichart 1995).

We test these claims with a probability simple random sample of public opinion data that we collected in Albania in summer 2015 through cellphone random digit dialing (RDD). We rely on our own data because we asked questions related to our key explanatory variables, but also because we consider the cellphone RDD data collecting technique to be an efficient tool for building simple random samples, especially in the Balkans urban chaos, where the intertwining of the single family and multifamily dwelling might make the household survey prone to significant selection bias.

This is a X-centered research. Support for EU membership conditionality and trust in country's politicians serve as key independent variables to predict support for their country's EU membership in EU membership-aspiring countries. A third key independent variable, which might capture oppositional feelings to EU membership conditionality is people's concern that EU membership conditions undermine country's sovereignty. We control for other variables, which we divide in three groups: socioeconomic variables, policy preference variables and benefit expectation variables. We found that, indeed, people's beliefs that EU membership conditionality helps country's democratization as well as their trust in county's politicians predict support for country's EU membership, whereas there is only weak evidence that viewing EU membership conditions as encroaching upon country's sovereignty negatively affects support for EU membership. Also, we found no evidence that prioritizing country's economic development predicts support for EU membership, although expectations of economic benefits could be such a predictor.

EU membership conditionality has been defined as a set of conditions sent from the EU to EU membership-aspiring countries for implementing reforms and pursuing policies in the direction prescribed by the Union (Peshkopia 2014b: 10). Introduced in the 1993 Copenhagen European Council, EU membership conditionality has been clearly designed as hard power, with conditions imposed by the EU on countries aspiring to gain membership in order to implement reforms and pursue policies prescribed by the EU (Grabbe 1999).

Two views define the purpose and goals of EU membership conditionality. The first, the demand side viewpoint, portrays conditionality as mainly concerned with minimizing the risk of new entrants becoming politically unstable and economically burdensome to existing EU members. These conditions serve to minimize the risks and costs of enlargement, which come with the difficulties to agree among a growing number of states (Grabbe 1999, 2002; Smith 2003). Other scholars tend to view EU conditions from the supply-side approach. For them, EU conditions and the programs attached to them provide material support for implementing reforms (Moravcsik and Vachudová 2005); shield moderate politics from populism and nationalism (Vachudová 2001); strengthen democratic forces in the face of authoritarian downturns (Schimmelfennig 2007); and serve as instrumental justification for domestic policies that Eastern European leaders need to implement in order to attain their own rational, power-driven goals (Brusis 2005). The conditions provide EU membership-aspiring countries from Eastern Europe with political and economic objectives, as well as guidelines for achieving these objectives. According to this view, the possibility of membership in the EU has created powerful incentives to shape transitional reforms (Pridham 1994; Smith 1997; Kubicek 2003).

It is obvious that those elite-centered approaches leave many factors affecting democratization in EU membership-aspiring Eastern European countries – most notably the masses – outside the policy and research focus (Piret 2001). Being designed and applied as a dialogue between political elites, EU membership conditionality has shown little concern for social change through the direct engagement of masses in this process (Peshkopia 2014a). Therefore, EU membership conditionality has been designed, implemented and sometimes studied as a shield of the presumed reform-oriented elites from the presumed reform-resisting masses. EU membership conditionality designers and implementers have applied this approach to such a great extent that they have persistently ignored masses’ attitudes and behaviors toward specific conditions imposed on their governments.

An alternative approach could assume that not the masses, but rather their political elites might represent the real threat to democracy. One can reasonably argue that the masses in EU membership-aspiring countries might have been affected by the EU promise, admire the EU democratic system, and are interested in promoting democracy in face of potential democratic downturns caused by power-oriented elites who might commit to democratic reforms only if and when the latter enhance their chances to attain and maintain power (Peshkopia 2014b). Just because the EU lacks formal engagement with the EU membership-aspiring countries’ publics doesn't mean that its ideas cannot permeate those societies, and inspire them to appropriate them.

Even though in several instances authors have pointed to the return-to-Europe psyche in Central and Eastern Europe (Gower 1999), as well as a perception of democratization and EU accession interconnectedness (Dimitrova 2004), most of the research conducted on EU membership conditionality has not considered it from a bottom up approach. The EU remains probably the most successful peace project in world history, bringing about an extension of its sphere of peace and stability each time new countries have joined its ranks (Tuomioja 2009). EU membership represents a powerful attraction for the masses in unstable and underdeveloped countries, and its perspective has been a powerful incentive for countries to shape reforms and appropriate EU norms and mores (Pridham 1994; Smith 1997; Kubicek 2003; Tuomioja 2009). Therefore, the real power of the EU is not the hard power it uses in negotiating with EU membership-aspiring countries’ domestic elites, but rather the very attraction it represents to the masses of those countries. That is, even though EU membership conditionality might have been designed and implemented as hard power, it could be better perceived as a tool of the EU soft power. Evidently, in the Western Balkans, the European perspective has already contributed to peaceful solutions to complicated ethnic conflicts, as well as domestic stability (Tuomioja 2009), one of the three major goals of EU membership conditionality.

As soft power, the European perspective permeates society, albeit perhaps not uniformly. Far from being a dialogue between Brussels and domestic political elites, EU membership conditionality could powerfully impact masses, and could serve as an ally in their quest to establish a functioning democracy. Oppositely, EU membership conditionality might be regarded also as an encroachment upon country's sovereignty, thus indicating lower support for country's EU membership. Therefore,

The already established argument maintains that the EU membership process remains an elite-driven process, and that support for EU membership remains embedded in people's evaluations of their domestic politics and government performance (Armingeon and Ceka 2014; Nelsen et al. 2011; Vetik et al. 2006; Franklin et al. 1995; Franklin and McLaren 1994). People use elites and institutions as informational shortcuts in order to inform their own policy perceptions and preferences, and help to locate policy responsibilities (Brader, et al. 2013; Druckman et al. 2013; Lupia and McCubbins 1998; Tilley and Hobolt 2011). Combining Albanians’ high support for EU membership with the fact that the EU membership process in Albania is an elite-driven process, and that there exists a national consensus around country's EU membership (Peshkopia 2014b), we expect a positive relationship between support for country's EU membership and trust in country's politicians.

We test these claims with a probability simple random sample of public opinion survey data that we collected in Albania in summer 2015. We operationalized the dependent variable, support for EU membership with responses ‘much’, ‘somewhat’, ‘little’ and ‘not at all’. One of the independent variables, people's beliefs on whether EU membership conditionality helps democratization (EU conditionality helps democratization) with responses ‘much’, ‘somewhat’, ‘little’, and ‘not at all’, tests Hypothesis 1. The main concern that EU membership conditions would raise involves the perceived threat that they might pose to country's sovereignty, an argument that parallels Arikan’s (2012) perceived symbolic threats to cultural identity and Turkishness. We operationalized concerns that EU membership conditionality questions country's sovereignty (a key independent variable in testing Hypothesis 2) as responses ‘yes’ and ‘no’. A third attitudinal variable, people's trust in their politicians, also measured as responses ‘much’, ‘somewhat’, ‘little’ and ‘not at all’ is a key independent variable in testing Hypothesis 3. Also, we test the effect of two interaction variables. One of them generated from the multiplication of the ‘EU conditionality helps democratization’ and ‘Trust politicians’ (Conditionality helps democratization*Trust politicians) tests the relationship between people's attitudes toward the EU membership and the combined effect of their attitudes toward EU membership conditionality and trust in their politicians. The other interaction variable, the multiplication of ‘EU conditionality helps democratization’ with ‘EU membership conditionality questions sovereignty’ (Conditionality helps democratization*Questions sovereignty) tests the relationship of people's attitudes toward the EU membership with the combined effect of their attitudes toward EU membership conditionality and their beliefs about EU membership conditions’ encroachment upon country's sovereignty.

We control for a number of socioeconomic variables derived from the extant literature on attitudes toward EU membership in both EU member countries and EU membership-aspiring countries. Variables such as age, gender and education derive from Schlenker (2012/3), Nelsen et al. (2011), Kentmen (2008), de Vreese et al. (2008), Elgün and Tillman (2007), Nelsen and Guth (2003), Carey (2002), and Gabel (1998); religion and attachment to religion derive from Nelsen et al. (2011) Çarkoğlu and Kentmen (2008), and Kentmen (2008); incomes and economic optimism, although differently operationalized and measured, derive from Nelsen et al. (2011), Kentmen (2008), and Elgün and Tillman (2007); and different occupation categories derive from Nelsen et al. (2011), and Elgün and Tillman (2007). Also, we added two sets of control variables, one of them measuring respondents’ policy preferences, and the other measuring respondents’ expectations from EU membership, which arguably affect people's attitudes toward their country's EU membership (Hooghe and Marks 2005; McLaren 2004).

We tested our hypotheses with the case of Albanians’ attitudes toward their country's EU membership. In spite of its maverick brand of communism, during the Cold War, the country shared similarities with other communist countries. With the collapse of communism in late 1990, Albania began its path to democratization, economic liberalization and EU membership. The legacy of an isolationist totalitarian communist regime has impacted all life sectors, and there was virtually no sector of cultural, economic, political and social life that did not need reforms (Peshkopia 2014b). The journey turned out to be much harder than anticipated, and the reform process was riffed with policy failures and intractable violent power struggle, putting the country in a perpetual state of political crisis, including cyclical electoral crises (Palickova 2019)—referring to the very latest political crisis related to the 2019 local elections, an Albanian political analyst observed that, differently from the West, where elections serve to resolve potential crises, in Albania elections always generate crises (Vangjeli 2019). However, distinctively, domestic and foreign observers alike have witnessed both the strong support among the Albanians for EU membership and the national consensus around the EU membership process (Peshkopia 2014b). There is virtually no political parties or policy advocacy groups in Albania that opposes country's EU membership (Peshkopia 2014b). The EU finally agreed to open negotiations for the Stabilization and Association Agreement (SAA) with Albania in January 2006 and concluded in June the same year. The SAA entered into force on 1 April 2009, and on 28 April 2009, Albania submitted its formal application for EU membership, until 2014, when Albania acquired the candidate country status, the EU has rejected Albania's application not once but twice (in 2010 and 2011), based mainly on the lack of progress in institutional reforms and a notorious lack of willingness of opposing parties to cooperate on those reforms, and offered only conditional candidate country in October 2013. Ever since, Albania has asked to open formal accession negotiations, but major EU actors’ lack of appetite for EU further enlargement and the most recent political crisis, with the opposition parties renouncing of their parliamentary seats in February 2019 and their boycott of the June 2019 local elections seem to have dimmed hopes that Albania would receive in October an EU invitation to open accession negotiations (Peel 2019). Keeping in mind that it has taken other countries one to two years to open accession negotiation after acquiring their EU candidate status (Jano 2018), the five years between it became an EU candidate country and the timing of this writing notoriously makes Albania a laggard outlier in the EU membership process. However, the country's unclear EU future seem to have not affected widespread pro-EU feelings on both sides of the aisle and among the Albanian public, and EU flags remain fixtures in opposition riots and rallies of the governing coalition supporters.

Therefore, Albania represents an interesting – and understudied – case to investigate the relationship between people's perceptions of EU membership conditionality and their attitudes toward their country's EU membership. In Albania, EU membership conditionality is not an abstract concept; the country's delay in gaining the EU candidate status, and most recently in opening up accession negotiations has made EU membership conditionality a very tangible policy not only to political elites, but also the masses.

We test our argument with a simple random sample of public opinion survey data that we collected in summer 2015 in Albania using a cellphone random digit dialing (RDD) technique through the iziSurvey digital platform. Our team of well-trained interviewers conducted the interviews on two major cellphone networks in the country, Vodafone and AMC, which combined for 78% of the country's cellphone users. A cellphone RDD technique is an efficient way to achieve a probability sample in a country with a severe lack of landline networks, especially in the countryside (Mohorko, de Leeuw and Hox 2013). In its 2015 Annual Report, AKEP (2015) stated that the penetration of landline telephony in Albania was only 8%, almost seven times lower than Southeastern Europe's regional average (40%) and almost twice as low as the world average (15.2%), and even lower than developing countries’ average (10%). Moreover, reportedly, the landline telephony penetration has been steadily decreasing in Albania since 2012 (AKEP 2016). Perhaps for that reason, the prevalence of cellphone SIM cards in the country in the period of our survey was 180% of the adult population, with many people using more than one cellphone. Specifically, during the second quarter of 2015, the period of this survey, there were 3,188,130 active SIM cards in a country with a population of 2,800,138, with the Vodafone sharing 47% of the market and AMC sharing its 31% (AKEP 2016; INSTAT 2011). This 78% of market share – with no evidence of socioeconomic differences between users of different cellphone networks – offers confidence that we reached a very good sampling frame.

Cellphone random digit dialing (RDD) sampling in the Balkans has several advantages compared to other sampling techniques. Cellphone RDD allows for unbiased randomization in cases with complicated residential patterns caused by the close proximity of single-family and multi-family dwellings (AAPOR Cell Phone Task Force 2010). Cellphone RDD helps sampling in cases where local norms, the family structure, or suspicion of state authority make difficult inviting strangers inside residences, and a household sampling method would create strong and systematic biases due to the difficulty of sorting out which family unit is the interviewer exactly sorting out, given that more than one of such reside under the same roof (Peshkopia and Voss 2016). Also, in most European countries cellphone providers use a single, national code number, which frees the sampling process from the need of survey stratification. Cellphone RDD sampling enables a better coverage of marginal groups, which are more difficult to reach through household sampling, in particular ‘the highly elusive young adult cohort’ (AAPOR Cell Phone Task Force 2010, 4). Also, luckily, the Balkans and the Albanian cellphone space has not yet been invaded by phone marketers in the same way as in the developed countries, keeping people receptive to calls from unknown numbers. Moreover, different from the United States, European cellphone providers do not charge their clients for incoming services, so Albanian cellphone users do not incur any costs for incoming calls, making them more inclined to participate in cellphone surveys.

We conducted our survey between the dates of 22 March and 14 July 2015. Our team of 30 well-trained interviewers (22 females and 8 males) contacted 4993 people, and 2839 of them opted to respond, hence a participation rate of 57.43%. Yet, for different reasons, only 2344 of them responded to the question that captures the dependent variable, namely ‘How much do you support country's membership in the EU?’, hence a partial completion rate of 46.95%. The survey generated missing values, some of them because respondents dropped the survey altogether, and others because they opted not to respond to that particular question. The maximum observation for any of the variables was 2366, including variables that measure age, gender and education. In order to avoid missing response biases, we decided to conduct the explanatory analysis through the multiple imputation procedure (outlined in the online appendix).

We measured age as an interval ratio variable (seven age groups); gender as a dummy variable with 1 = female and 0 = male; education as a ratio variable (years of education); Muslim, Catholic, Orthodox and not affiliated with any religion as dummy variables; practice religion as a scale variable measured in the range of 0 = not at all, 1 = little, 2 = somewhat and 3 = much; household economic trends during the last year as a scale variable measured as −1 = deteriorated, 0 = remained about the same, and 1 = improved; personal incomes as a categorical variable; and several employment categories as dummy variables. We also measured the ‘EU membership conditionality questions country's sovereignty’ variable as a dummy variable with values 1 = yes and 0 = no; and ‘Trust politicians’, and ‘EU membership conditionality helps democratization’ as ordinal variables measured as 0 = not at all, 1 = little, 2 = somewhat, and 3 = much.

Our survey generated some demographic discrepancies between our sample and the population, as shown in Table A.1 of the online appendix. Those discrepancies reflect different cellphone usage patterns by different age groups and genders. In order to account for those discrepancies, we built a post-stratification weight variable. However, as Table 1 shows, differences between unweighted and weighted data are negligible. Whereas there is a general consensus that weights should apply to descriptive statistics (Kish and Frankel 1970), same level of consensus does not exist on whether weights should routinely apply in regression analysis (Kott 2007; Winship and Radbill 1994). Therefore, whereas we used weights in our descriptive statistics (Table 1), we did not apply them in our explanatory models. We tested an ordinal dependent variable that measured people's support for Albania's membership in the EU with responses generated from the question ‘How much do you support country's membership in the European Union?’, with answers 0 = not at all, 1 = little, 2 = somewhat, and 3 = much. Therefore, ordered logit models would be most appropriate for this multivariate analysis of our ordinal dependent variable.

Table 2 is a contingency table between support for EU membership and the belief that EU membership conditionality helps democratization. Results reveal the high association between both variables, with the much-much cell representing the highest value (1017 respondents or 44.45%). Two tests, the Pearson chi-squared test, and the Kendall tau-b test confirm the strong relationship between these variables (we can apply the Kendall tau-b non-parametric test because both variables satisfy both the ordinal variable assumption and the monotonicity assumption, and the measurement scale of both variables is identical, hence a quadratic contingency matrix). Both tests offer the necessary confidence to reject the H₀ hypotheses of the independence of these variables.

Table 3 replicates the chi-squared test in the contingency Table 2 with adjusted residuals, with the goal of detecting response dyads that cause the lack of independence between variables (if the adjusted residual value is larger than 3.80). As the table shows, all response dyads in the matrix diagonal have adjusted residuals larger than 3.80, suggesting a substantial number of symmetrical responses for both questions.

Table 4 presents the ordered logit explanatory models with imputed data following the data augmentation procedure, and imputing data from the combination of 100 datasets. Model 1 offers predictions for people's attitudes toward their country's EU membership only by accounting the relationship between the socioeconomic variables and the dependent variable. The model employed two groups of variables, one of them measuring religion affiliation and the other one measuring employment category. None of those variables showed any significant effect on the dependent variable. Although not statistically significant, some directional signs suggest that women might be less supportive of the EU membership, as found by other scholars (Arikan 2012; Kentmen 2008; Elgün and Tillman 2007); and education might be positively correlated to stronger support for country's EU membership, as suggested by Kentmen (2008) and Elgün and Tillman (2007). The negative correlation of personal incomes with support for EU membership seems to validate Kentmen’s (2008) findings, whereas economic optimism coinciding with improved household economy seems to confirm findings of Nelsen et al. (2011). However, since none of those relationships carry any statistical significance, we can only report the trend of their correlation directional sign, but cannot offer them as empirical evidence of those variables’ effects. These results suggest that, except for age, support for EU membership in Albania spreads almost uniformly across genders, various levels of education, religion affiliation and practice as well as different employment categories. Age, the only variable in this model with statistically significant effect over the dependent variable seems to be a very strong and reliable predictor (p < .001). This finding suggests that, if there are people who hold skepticism toward their country's EU membership, they could be found more among the younger respondents, a feature noticed before by other scholars (Kentmen 2008).

Model 2 includes two more sets of independent variables, people's policy priorities and people's expectations from their country's EU membership. As it should be expected, prioritizing country's EU membership strongly predicts support for that membership (p < .001). It should be noted that the coefficient's negative sign reflects the order of values ranking for this variable (1–5), where 1 indicates that specific policy to be the top priority and 5 to be the lowest priority to the respondent. However, the model offers no evidence that any other policy priorities affect such support. Also, expecting no benefits from country's EU membership strongly predicts negative support for such a membership (p < .001), as it should be expected. However, the only other benefit expectation that predicts support for country's EU membership is the improvement of country's economy (p < .01). These findings are instructive about the effects of the combination of policy prioritization and benefit expectation on support for EU membership: people expect economic benefits from country's EU membership even when prioritizing economy does not affect their support for EU membership.

The theoretical discussion building up to Hypothesis 3 might have left room for inquiry on whether or not beliefs that EU membership helps country's democratization by trust in country's politicians affect support for country's EU membership. Model 3 tries to respond to such concerns by adding an interaction variable, ‘EU membership conditionality helps democratization*Trust politicians’. The role of this variable is to detect the effect of being simultaneously trusting in both EU membership conditionality as a tool of democratization and country's politician. Its effect does not acquire any statistical significance, failing to find the necessary evidence of any relationship between beliefs that EU membership helps country's democratization among those who trust more and less country's politicians, and support for country's EU membership. Indeed, Figure 3(a) shows only small differences in probabilities to support country's EU membership between those who believe that EU membership conditions help country's democratization and those who believe that those conditions do not help, grouped by those who believe that EU conditions question and those who believe that EU conditions do not question country's sovereignty Table 5.

Model 4 is a replication of Model 2 with one interaction variable, ‘EU membership conditionality helps democratization*EU membership conditionality questions country's sovereignty’. Similar to the interaction variable in Model 3, this variable's effect does not carry any statistical significance, and it also shows a negative relationship with the dependent variable. Meanwhile, again, similar to Model 3, only the effect of the constitutive variable, ‘EU membership conditionality helps democratization’ manages to carry statistical significance (p < .001), whereas the other constitutive variable's effect does not (p < .581). Figure 3(b) visualizes these small differences in probabilities to support EU membership by the belief that EU membership conditionality helps country's democratization grouped by the belief of whether or not EU membership conditions question country's sovereignty, while controlling the other variables for their modes.

The regression analysis brings about an important finding. The ‘EU membership conditionality helps democratization’ variable maintains its strong statistical significance throughout the models, thus emerging as a predictive powerhouse of people's support for EU membership. Its powerful impact remains intact even as a constitutive part of two interaction variables in two different models. This can be interpreted as a signal that people make a clear and direct positive connection between EU membership conditionality as a tool of democratization and their support for country's EU membership.

This research makes an empirical effort to offer an alternative to the dominating elite-centered approach to EU membership process in general, and EU membership conditionality in particular. From this perspective, our findings could contribute to a novel perspective. We argued that, although it was never designed and implemented as soft power, EU membership conditionality could powerfully impact masses, and can serve as an ally in their quest to expand and improve democracy. All in all, people in EU membership aspiring countries could rely on EU membership conditionality to prevent power-oriented politicians shirking their commitments to democratization, and force them onto the course of the much needed reforms to build institutions and policies compatible with those of the Union. These claims complement the argument that support for EU membership remains embedded in people's evaluations of their domestic politics and domestic governments’ performances.

The data analysis offered empirical evidence that people's view of EU membership conditions as being helpful to democratization strongly predicts support for their country's EU membership. The strong relationship between these variables attests not only to the strong attraction that the EU is able to exercise to societies who aspire to join it, but also to the belief that EU policies, including EU membership conditionality, enjoy wide support among the public in EU membership aspiring countries. Moreover, we found no evidence that people's prioritization of economy drives their support for country's EU membership, although they expect economic benefits from their country's membership in the Union. These findings would help to frame EU conditions as tools of the EU's soft power.

Such conclusions carry important policy implications. First, by knowing that strong support for EU membership conditionality positively predicts support for EU membership in EU membership-aspiring countries, the EU can tailor stricter conditions, thus forcing domestic political elites to improve their reform performances. Second, and more importantly, such a knowledge should make the EU more interested and engaged in public diplomacy, namely the relationship between its diplomats and politicians, and the foreign publics with whom they work. Distance between Brussels politicians, bureaucrats and diplomats from the public of current EU member states as well as the public in EU membership-aspiring countries has benefited neither the EU political system, nor societies of future member countries. As has been shown in other research, such a distance has not benefited social change in the latter societies either. In spite of reform implementation in line with EU conditions in many sectors, those societies lag behind in many world indicators, where the most blatant example is the human rights field. An active EU public diplomacy and EU membership conditions tailored to address societies at large, rather than simply ruling elites, would help to overcome the gap between legislation guaranteeing human rights and people themselves guaranteeing each other's rights on a daily basis. Our findings show that there might be room and reason for optimism in conducting such policies.

Our focus on Albania carries the double importance of shedding light on an understudied society, and taking into account one of the few remaining EU candidate former-communist countries still in the process of acquiring full membership (the other ones being North Macedonia, Montenegro and Serbia, with Bosnia and Herzegovina and Kosovo as potential candidate countries—a geopolitical region recently dubbed as the Western Balkans). However, the generalization problem stubbornly appears here as well. First there is the question of whether the Albania's bitter an unparalleled communist legacy would make its citizens more eager to embrace the EU, thus making their overwhelming support for EU membership to be an outlier. Second, differently from its EU membership-aspiring neighbors, Albania did not emerge out of the defunct Yugoslavia, and did not experience Yugoslav devastating ethnic conflicts during the 1990s and early 2000s. Literature suggests that both communist legacies and ethnic conflicts impact transition performances (Rupnik 2000), and through them, country's approach toward EU membership. Therefore, more empirical work is needed to see whether our models would apply to the other Western Balkans countries.

Under a different title, the author first presented this research at the 4th International Conference for Business, Technology and Innovation (ICBTI), Durrës Albania, November 6–7, 2015. I am grateful to Alfred Marleku, Armend Muja and Artan Tahiri of University for Business and Technology Kosovo (UBT) for suggestions, support and encouragement during the conference presentation and throughout the process of completing and polishing the paper. Douglas Page of Gettysburg College helped with the figures. Steve Voss of University of Kentucky helped with survey design and methodology, as well as partially financing the survey. Markela Muça of the University of Tirana's Department of Applied Mathematics helped with suggestions at the last stage of manuscript development. I feel indebted to all those who volunteered with conducting the public opinion survey, and who also often paid for their own expenses, namely Jonida Lika and Denisa Rira of University of Tirana, Albania; Rina Ajvazi, Saranda Berisha, Trëndelina Buja, Hermonda Kalludra, Blerina Miftari, Blerta Nuredini and Vesa Zatriqi of University of Warsaw, UBT Kosovo campus; Habibe Ademi, Arton Demolli, Arrita Deva, Përparim Hashani, Valmira Hoxha, Blerina Islami, Ardit Konjufca, Gresa Konjufca, Gazmend Obrazhda, Drenushë Osmani, Gentiana Sahiti, Brigita Rexha, Dynjah Xhakli, Vildane Xhemaili, and Blerina Zeqiri of UBT, Kosovo; Rrezart Dema, Erenik Dujaka Arbëresha Mehmetaj and Blerta Morina of Universum College, Kosovo; and Lekë Hoxha of Minnesota State University at Mankato. I thank my colleagues and fellow administrators with my institution, UBT, who took over more courses and more administrative work to open time for me to conduct this research. Eno Minka of the Vodafone company helped with resolving technical difficulties that emerged during the RDD sampling. Hannah Holmes of Wayne State University thoroughly copyedited the manuscript. Any lingering flaws remain author's own responsibility.

No potential conflict of interest was reported by the author.

Ridvan Peshkopia is Lecturer of Political Science at the University for Business and Technology, Kosovo. He received an MA in Diplomacy from the Patterson School of Diplomacy and International Commerce, and an MA and PhD in Political Science with the University of Kentucky; he also spent a year as a postdoctoral fellow with the George Washington University. He publishes in several disciplines and sub-disciplines including international relations, film studies, social theory, political behavior, migration studies and epistemology. He is currently pursuing an MSc in Applied Mathematics with the University of Tirana, Albania.