Politics of Nostalgia and Populism: Evidence from Turkey

Abstract This article scrutinizes the relationship between collective nostalgia and populism. Different populist figures utilize nostalgia by referring to their country's ‘good old’ glorious days and exploiting resentment of the elites and establishment. Populists instrumentalize nostalgia in order to create their populist heartland, which is a retrospectively constructed utopia based on an abandoned but undead past. Using two original datasets from Turkey, this study first analyzes whether collective nostalgia characterizes populist attitudes of the electorate. The results illustrate that collective nostalgia has a significantly positive relationship with populist attitudes even after controlling for various independent variables, including religiosity, partisanship, satisfaction with life and Euroscepticism. Secondly, the study tests whether nostalgic messages affect populist attitudes using an online survey experiment. The results indicate that Ottoman nostalgia helps increase populist attitudes. Kemalist nostalgia, however, has a weak direct effect on populist attitudes that disappears after controlling for party preference.

AKP has come to power, Turkey has changed for the better. According to the AKP and its supporters, a 'New Turkey' has risen from its ashes. The AKP has structured its narrative with slogans like 'Resurrection once again, rise once again' (Yeniden diriliş, yeniden yükseliş), which means that the country was glorious and strong in the past and will be so again.
Following these examples, this study questions the relationship between nostalgia and populism using both a representative survey and experimental datasets. Many populists around the world exploit nostalgia by referring to their country's glorious past. Populists frequently emphasize that the elites hijacked the people's will long ago, and they promise to return power to the people. According to this logic, back in the good old days corrupt elites were not powerful enough to abuse the authentic people. However, many regulations and institutions now hinder people's will. Populists exploit nostalgia to create their populist heartland, which corresponds to a retrospectively constructed utopia based on an abandoned but undead past (Bauman 2017;Taggart 2004).
In the following section, I provide a theoretical background on populism, nostalgia and the link between the two. Next, I present the representative survey, items that were used to measure nostalgia and populism, along with the survey results. Then, I proceed to the survey experiment, a between-group design in which two groups received different types of nostalgic messaging in addition to a pure control group. The final section concludes the article.

Populism
This article follows the ideational approach to populism. According to Mudde (2007, 23), populism is 'a thin-centered ideology that considers society to be ultimately separated into two homogeneous and antagonistic groups, "the pure people" versus "the corrupt elite," and which argues that politics should be an expression of the volonté générale (general will) of the people'. More broadly, populism establishes an us vs. them identity by valorizing the people and its values while denigrating the elites. According to populists, the people's will is the cornerstone of politics; there is no higher authority.
Populism creates a duality in society based on a Manichean understanding of good vs. evil. According to Mudde (2017), populism differs from other ideologies because it is grounded in the Manichean duality on morality. According to populists, while the people are pure and authentic, elites are immoral and villain.
Many scholars agree that populism appears as a backlash to a sense of severe crisis (Rooduijn 2014;Taggart 2004) or discontent (Spruyt, Keppens and Van Droogenbroeck 2016). This sense of extreme crisis paves the way for raising concerns about the present conditions, which is a breeding ground for populism (Akkerman, Mudde and Zaslove 2014). Populists blame the establishment for the decline in the people's economic, political and living standards (Rooduijn, van der Brug and de Lange 2017). However, populism can also be generated by unsolved, long-lasting and cross-cutting political cleavages within a society. Somer and McCoy (2019, 8) define these cleavages as formative rifts that 'either emerged or could not be resolved during the formation of nation-states, or, sometimes during fundamental re-formulations of states such as during transitions from communism to capitalism, or authoritarian to democratic regimes'. Hence, while the sense of extreme crisis can be a recent change in a society, such as immigration for European countries, it can also be a secular-religious cleavage, as in Turkey (Aytaç and Elçi 2019;Somer 2019), or changes in gender and race relations, as in the United States (Duyvendak 2011;Mudde 2016). Populists aim to re-establish the present and future retrospectively by emphasizing the golden past of society to cope with contemporary challenges. Their goal is to return to a paradise lost in which life was better and more straightforward.

Nostalgia
The etymological roots of nostalgia come from Greeka combination of nostos 'return home' and algia 'a painful condition' and can be roughly translated as 'a painful yearning to return home' (Davis 1979, 1). Nostalgia first appeared as a problem of isolated monks who experienced extreme loneliness and melancholy due to their withdrawal from earthly life. The word nostalgia, however, was coined by Swiss medical student Johannes Hofer in 1688 when he observed the symptoms of soldiers fighting abroad (Boym 2001;Lowenthal 1985;Stauth and Turner 1988).
In line with Wildschut et al. (2014, 845), I define collective nostalgia as 'the nostalgic reverie that is contingent upon thinking of oneself in terms of a particular social identity or as a member of a particular group and concerns events or objects related to it'. According to Stauth and Turner (1988, 47), nostalgia is yearning for 'some golden age of heroic virtue, moral coherence and ethical certainty, a period in which there was no gap between virtue and action, between words and things, or between function and being'. They define four major components of nostalgia. First, 'there is the view of history as decline and loss, being a departure from some golden age of "homefulness"'. Second is the problem that there exists 'a sense of the loss of wholeness and moral certainty'. At this point, 'history is seen to be a collapse of values which had once provided the unity of social relations and personal experience'. The third is related to 'the loss of individual autonomy and the collapse of genuine social relationships'. Last but not least, nostalgia is 'the sense of a loss of simplicity, spontaneity, and authenticity' (Stauth and Turner 1988, 30-32; also see Bora and Onaran 2006 on this discussion).
'The nostalgic subject', for Tannock (1995, 454), 'turns to the past to find/construct sources of identity, agency, or community, that are felt to be lacking, blocked, subverted, or threatened in the present' because nostalgia is 'the search for continuity amid threats of discontinuity' (Davis 1979, 35). As a group-based emotion, nostalgia is related to social identity (Psaltis et al. 2017). According to social identity theory, identities are 'part of an individual's self-concept which derives from his knowledge of his membership of a group (or groups) together with the value and emotional significance attached to the membership' (Tajfel 1982, 2). The self-categorization of us vs. them is at the heart of social identity theory (Greene 1999). When absolute differences are salient between groups, people tend to perceive similarities of the in-group and exaggerate its differences from the out-group (Hornsey 2008). Nostalgia, or the way we remember our past, exacerbates in-group vs. out-group distinctions (Martinovic et al. 2018). Nostalgia also strengthens shared social identity and distinguishes the in-group from other groups (Brown and Humphreys 2002;Wildschut et al. 2014).
In the process of identity construction, history becomes a useful instrument for people to understand who they are, where they come from and where they should be going. The past emerges as a guideline during the constitution of group identity, its relationship with other groups, and as a defense mechanism for external threats (Davis 1979;Jetten and Hutchison 2011;Liu and Hilton 2005). In turn, nostalgia legitimizes the present action with references to the past (Gaston and Hilhorst 2018). According to Searle-White (2001, 49), group identity 'is not only its image of who it is now, but also who it has been in the past, and who its people hope it will be in the future'. References to past and nostalgic reveries help constitute people's identities and functions as an intermediary tool between memory and identity (Kalinina and Menke 2016). According to Smeekes (2015, 56), collective nostalgia triggers an in-group prototype 'as being part of a moral community based on shared past experiences' and out-group prejudice. The in-group and out-group stereotyping illustrates group boundaries between the authentic us and the alien them.

Populism and Nostalgia
Previous studies touched upon the relationship between nostalgia and populism in a relatively narrow sense. For example, Wiles (1969, 170) argues that populism dislikes 'the present and the immediate future' by seeking 'to mold the further future in accordance with its vision of the past'. For Betz and Johnson (2004, 324), right-wing populism is 'a backward-looking reactionary ideology, reflecting a deep sense of nostalgia for the good old days'. Furthermore, according to Sullivan (2017), the rise of populism is closely related to 'acute despair at the present moment and a memory of a previous golden age'. Populists aim to destroy the current status quo and 'return to the past in one emotionally cathartic revolt'. Populists constitute a new ideology of homea vision of the lost homelandwhich represents nostalgia for a reconstructed past and, in turn, provides a sense of security against the perceived loss of identity (Albertazzi and McDonnell 2008). Golden ages are crucial for populists because they provide sources of political legitimacy for the present, along with the rhetoric of authority and authenticity against troublesome changes, crises or decline (Elgenius and Rydgren 2019;Kenny 2017).
The golden age corresponds to the heartland in populist literature. Coined by Taggart (2004, 278), 'the heartland is a construction of the good life derived retrospectively from a romanticized conception of life as it has been lived'. It involves returning to the pre-lapsarian world, a Golden age before a catastrophic lapse or fall: in the post-lapsarian era, individuals feel lacking, deficient or oppressed (Tannock 1995). Populists aim to reach a retrotopia, which is based on 'visions located in the lost/stolen/abandoned but undead past, instead of being tied to the not-yet-unborn and so inexistent future' (Bauman 2017, 5).
The construction of the 'us vs. them' dichotomy is a function of cross-cutting conflicts in a given society (Barr 2009). During the construction of the antagonistic us and them, the heartland provides raw material where values are collected (Taggart 2004). There can be varieties of heartlands that no single theory can explain due to the unique characteristics of each society. Yet all such theories agree that society is divided along a single cleavage: the good, authentic people are separated from the corrupt, adversarial elites (Aslanidis 2020). Populists draw the borders of in-groups and out-groups by reconstructing the past and referring to the founding moment of the people (Cento Bull 2016). An escape to the heartland posits that 'a life has already been lived and so shown to be feasible. It assumes or asserts that there was a good life before the corruptions and distortions of the present' (Taggart 2004, 274). Building on these arguments, I expect that: HYPOTHESIS 1: The more nostalgic respondents are, the more populist attitudes they will express, because nostalgia creates a 'pure people vs. corrupt elites' duality.
HYPOTHESIS 2: Respondents in the nostalgia treatment groups should be more populist than those in the control group in the experimental setting.

Populism And Nostalgia In Turkey
In Turkey, what Şerif Mardin (1973) coined the center-periphery cleavage breeds both populism and nostalgia in Turkish politics (Aytaç and Elçi 2019;Elçi 2019). For Mardin (1973), the country's contemporary secularist vs. Islamist cleavage is a continuation of the socio-cultural and socio-economic duality between the ruling center and ruled periphery in the Ottoman Empire. While the former embraced secularism, Turkish nationalism, unitary state and state-supervised economy, the latter supports Islamism, conservatism, liberal economy and a decentralized state. According to the peripheral masses (which are composed of religiously, ethnically and regionally heterogeneous groups), the elites at the center (the foreign service, judiciary and military) have dominated Turkish politics and society as guardians of Kemalist values 1 and the secular Turkish Republic (Kalaycıoğlu 1994;Sunar 1990). Today, the center-periphery cleavage also overlaps with the left-right dimension in Turkish politics: the left appears as supportive of secularism and the right as the champion of Islamism (Çarkoğlu 2012;Çarkoğlu and Hinich 2006). The contemporary and significant representatives of the center vs. periphery cleavage are the Republican People's Party (Cumhuriyet Halk Partisi, CHP) and the AKP, respectively.
Ottoman nostalgia in Turkey constitutes a reaction to Kemalist modernization efforts and the cutting of ties to the Ottoman legacy in order to create a modern, secular nation-state (Karaveli 2010;Özyürek 2006). According to Bora (2011), the AKP inherited the Islamic nationalism of the National Outlook (Milli Görüş) tradition, 2 which imagines Turkey as the potential leader of the Islamic world. Nostalgia for the Ottoman past, at this point, appears as an imperial and irredentist fantasy. Consequently, Ottoman nostalgia can be summarized as discourse and actions 'on the past intended to legitimize contemporary neoliberal and cultural policies by drawing on anachronistic reinterpretations and the glorification of the Ottoman past in Turkey' (Iğsız 2015, 327).
For right-wing politics, mainly Islamist and ultranationalist factions, the Ottoman Empire represents the zenith of the Turkish people's history (Bora 2011;Çetinsaya 2006). As a response to the Kemalist establishment's secular nationalism, political Islam aggrandizes the Ottoman past as a tool for the reconstruction of 'the Great Turkey' today. The collapse of the Ottoman Empire corresponds to the absence of home. Recep Tayyip Erdoğan has referred to verses of his favorite poet, Necip Fazıl Kısakürek, 'the people are odd and a pariah in his own homeland', because they no longer live under the authentic Ottoman rule but the secular, modern Turkish Republic (Yılmaz 2017). Ottoman nostalgia reflects resentment towards the Republican elites, as it serves as a tool for revealing, transforming and mobilizing the desires, ambitions and anger of the Islamist masses (Tokdoğan 2019).
Kemalist nostalgia is more recent: it appeared during the 1990s, an era of significant transformations in world politics. The collapse of the Soviet Union and dismemberment of Yugoslavia created a power vacuum in the region, which led to the rise of ethnic conflicts. It was also a period of growing religious politics, which coincided with the rise of political Islam and religious extremism. These transformations were echoed in Turkey with the rise of the Welfare Party (Refah Partisi), which represents a continuation of the Milli Görüş ideology. This was concurrent with the establishment of pro-Kurdish political parties, which eschewed the ideology of the Kurdistan Workers' Party (PKK) and the bloody terror campaign it waged in the 1990s (Bora 2013;Özyürek 2006).
In addition to the rise of identity and religious politics, the increasing demands of the EU, economic privatization, and increasing interactions with the International Monetary Fund and the World Bank have also challenged Kemalist ideology. In turn, many Kemalists began to share 'the memory of a strong, independent, self-sufficient state and its secularist modernization politics which dominated the public sphere through the past century' (Özyürek 2006, 2). In other words, when foundational principles were at stake, Kemalists started yearning for 'the childhood 1 Kemalism is a quasi-ideology named after Turkey's founding leader, Mustafa Kemal Atatürk. According to Cizre-Sakallıoğlu and Çınar (2003, 310), Kemalism is 'a Westernizing/civilizing ideology whose incontrovertible maxims are secularism, understood as the separation of religion from political rule; a modern/Western identity and lifestyle; and the cultural homogeneity and territorial unity of the nation'. of the nation when everything was pure, and citizens were gathered around the authority of their father' (Özyürek 2006, 16).
In short, Ottoman nostalgia appears to be an effective tool for channeling the resentment of peripheral groups towards centralist elites and the establishment in Turkey. However, Kemalist nostalgia emerged in reaction to the rise of political Islam and Islamist populism. Hence, I expect that: HYPOTHESIS 3: The effect of the Ottoman nostalgia treatment on populist attitudes should be greater than the Kemalist nostalgia stimulus in the experimental setting.
In summary, following the ideational approach, this study claims that populists aim to facilitate the moralistic separation between the pure people (us) and the corrupt elite (them). While constituting the boundaries of us vs. them, populists create an in-group vs. out-group identity.
To this end, one of the primary tools that populists utilize is collective nostalgia. They aim to build the present and future retrospectively. Populists offer their constituencies a golden agea heartland where corruption, disruptions and enemies do not exist. The following two sections empirically analyze these arguments.

Study 1: Survey Analysis
This section tests the link between individuals' predisposition to nostalgia and populist attitudes (Hypothesis 1) using representative survey data (n = 1,954). I assume that after controlling for other independent variables, nostalgia should have a significantly positive impact on populist values.
The dependent variable is constructed from four survey items created by Hawkins, Riding and Mudde (2012). 3 These four items aim to 'capture key elements of populism, especially a Manichaean view of politics, a notion of a reified popular will, and a belief in a conspiring elite' (Hawkins, Riding and Mudde 2012, 7). I aggregated the four items and created an index ranging from 0 (least agreement) to 100 (most agreement): • Politics is ultimately a struggle between good and evil.
• The power of a few special interests prevents our country from making progress.
• The politicians in Parliament need to follow the will of the people.
• The people, and not the politicians, should make our most important policy decisions.
The primary independent variable is Nostalgia, which ranges from 0 (less agreement) to 100 (more agreement). It is constructed from four different survey items that seek to measure the importance of past values, past experiences, old tastes and the country's glorious past: • Listening to new pieces of music, I miss the old days and those beautiful folk songs.
• We must not forget the values that we had in the past.
• I miss my country's good, old glorious days.
• We can get rid of the problems we experience today as 'the people' only by looking to our past.
Without anchoring to any particular period or group, these items aim to capture collective nostalgia, which can be different periods for the respondents. For example, as discussed in the previous section, while CHP supporters could be nostalgic for the achievements of the Kemalist era, AKP supporters yearn for the glorious days of the Ottoman Empire.
I use three sets of independent variables to control for the relationship between nostalgia and populism. First, to connect the study to the broader populism literature and compare Turkey with previous studies on populism in different contexts, I used the support for EU membership, satisfaction with democracy, economic conditions, and life, and interpersonal trust questions. Secondly, to capture the effects of identity politics and the center-periphery cleavage in Turkey, I used variables such as Alevi, Kurdish and religiosity, where the cross-cutting cleavages in Turkish society occur. Lastly, I controlled for the effect of nostalgia with partisanship variables (see the Appendix for measurement of control variables). 4 Before moving to the analysis, it is necessary to discuss the control variables and hypothesized relationships. According to Almond andVerba (1963[1989]), the norms of interpersonal relationships, general trust and confidence in the social environment are fundamental dynamics for establishing civic culture and democracy. Thus I expect that respondents who expressed positive interpersonal trust have less populist attitudes. Euroscepticism can be defined as 'the idea of contingent or qualified opposition, as well as incorporating outright and unqualified opposition to the process of European integration' (Taggart 1998, 366). It is a symptom of both radical right and left parties in Europe and its periphery Rooduijn, Van Der Brug and De Lange 2017). I expect to find a negative relationship between populist attitudes and support for EU membership. Finally, according to scholars who support the idea of populism as a response to a political crisis, subjective dissatisfaction with economic, social and political conditions leads to an increase in populist attitudes (Akkerman, Zaslove and Spruyt 2017;Bowler et al. 2017;van Hauwaert and van Kessel 2018). Thus I expect that the more dissatisfied respondents are with their lives in general, subjective economic conditions and democracy, the more populist attitudes they will have.
Religiosity plays a pivotal role in explaining the party preferences of the Turkish constituency and is a key independent variable for explaining the center-periphery cleavage. While more religious people have typically been on the periphery of politics, they have always distanced themselves from the Western-oriented secular politics of the establishment since the late Ottoman period. Thus I expect more religious respondents to display more populist attitudes. While the Alevi minority in Turkey tends to side with the center due to political Islam's Sunni-order-oriented policies (Çarkoğlu and Elçi 2018;Çarkoğlu 2005), Kurdish people are historically located as an anti-establishment constituency (Celep 2018). Thus I expect Alevi respondents to have less populist attitudes because they are more supportive of secularism. However, I predict that Kurdish respondents will have more populist attitudes because they are more critical of the Kemalist establishment. Finally, the AKP constituency, located on the periphery of Turkish society, should have more populist attitudes. The Peoples' Democratic Party (Halkların Demokratik Partisi, HDP), as a radical democratic pro-Kurdish party, should also display more populist attitudes. I expect CHP and Nationalist Action Party (Milliyetçi Hareket Partisi, MHP) voters to have less populist attitudes since they represent the status quo. 4 The data used for analysis is a representative sample. Frekans Research (www.frekans.com.tr) conducted face-to-face interviews from 17 February to 2 April 2017. The sampling procedure starts with the classification of the Turkish Statistical Institute's (Türkiye İstatistik Kurumu, TUIK) twenty-six NUTS-2 regions. Data collection was carried out with a joint project. We distributed the target sample according to each region's share of the urban and rural population according to the current records of the Address-Based Population Registration System (Adrese Dayalı Nüfus Kayıt Sistemi). Next, we used TUIK's block data, with a block size of 400 residents. We aimed to reach twenty voters from each block without using any substitutions. We selected individuals in households, based on the reported target population of 18 years or older in each household, using a lottery method. Pollsters visited the same household up to three times until a successful interview was conducted. If for any reason that individual could not be reached, the household was dropped without applying a substitution.
Finally, I include demographic variablessex (female), age, education, income and urban residency. I also used age-squared to detect the nonlinear relationship of the age variable. Previous studies have not reported a significant relationship between populist attitudes and sex and age. However, some studies have found that populist attitudes are negatively correlated with urban living, higher education achievement and higher income levels in some contexts

Results
I created ten different models to control for the effect of collective nostalgia with party preferences and satisfaction with life, democracy and economic conditions. I controlled for the effects of satisfaction variables in separate models because once they enter the equation, they can affect party preferences due to a moderate correlation with party preference variables. It is not surprising that the AKP constituency is more satisfied with democracy, their lives and subjective economic conditions, since they voted for the winning party. Table 2 reports the standardized coefficients for comparing the magnitudes of the independent variables.
The results confirm the first hypothesisthat collective nostalgia positively correlates with populist attitudes in all models. A one-standard-deviation increase in nostalgia leads to a 0.26-standard-deviation rise in populist attitudes, which is twice the size of other significant control variables. Collective nostalgia remains significantly positive even after controlling for other independent variables, including religiosity and party preference. Unfortunately, many independent variables did not reach statistical significance, which limits my ability to draw conclusions about the determinants of populist attitudes in Turkish politics.
In line with expectations, religiosity also has a significantly positive impact on populist attitudes. However, standardized coefficients indicate that the magnitude of nostalgia is even higher than religiosity. EU supporters have lower levels of populist attitudes, which is also the case in the broader populism literature. Urban dwellers appear less populist than other respondents, which tracks with the urban vs. rural divide in the populism literature. However, contrary to expectations, respondents who are more satisfied with their lives have more populist attitudes. Finally, only AKP voters emerge as a populist constituency among the other party supporters. In line with expectations, voting for the AKP and HDP has a positive relationship with populist attitudes, while preference for the CHP and MHP are negatively correlated. However, none of these relationships with populist attitudes reach conventional levels of statistical significance except for the AKP.
In summary, although the results indicate a positive relationship between nostalgia and populism, they do not provide any information about which type of nostalgia stimulates respondents' populist attitudes. The best way to overcome this shortcoming is an experimental analysis in which participants receive particular messages to stimulate their populist attitudes. The next section reports the results of such an experimental test.

Study 2: Experimental Analysis
This section applies experimental analysis to capture the causal link between nostalgia and populism. I created a between-group experimental design in which two groups received different types of nostalgic messages: Ottoman and Kemalist. The control group did not receive any stimulus.

Data
The experimental data were collected via convenience sampling. I recruited respondents through Facebook and Instagram advertisements using the Qualtrics survey tool. I conducted the survey experiment between 23 and 24 June 2018. Participants did not receive any bonus or additional incentives.
I collected a total of 911 observations. After dropping respondents who failed attention checks, 803 observations remained. 5 The Qualtrics survey tool applied randomization with no external intervention. Likelihood ratio tests from a multinomial logistic regression suggest that the randomization was successful (χ 2 (10) = 12.92, p = 0.23) ( Table 3).

Experimental Design
After informed consent was obtained, the first part of the survey asked respondents warm-up questions, such as their level of satisfaction with democracy, their lives and economic conditions. Participants responded to party preference questions before receiving treatments to avoid posttreatment bias (Montgomery, Nyhan and Torres 2018). Later, respondents were randomly assigned to one of three groupsthe control group, the Ottoman nostalgia group and the Kemalist nostalgia group.
Next, I introduced experimental vignettes with the following message: 'Now you will be shown a text about the history of Turkey. Please read this text carefully and slowly. After reading, you will be asked some questions about the text.' Control group participants did not receive any notification and continued to answer the questions. The next button appeared after 3s of receiving treatment messages. Respondents spent an average of 25.8s on the stimulus (s.d. = 12.13). Overall, the average response time was 16.41 minutes (s.d. = 18.34).
The Ottoman and Kemalist nostalgia groups read texts that differed only in the underlined parts. The Ottoman nostalgia group received the following message: This heavenly homeland is the trust of our ancestors. Since the Ottoman Empire, this people have always stood firmly against several external outbreaks, and their collaborators within us, that have played games with our country. Remember the glorious Ottoman! Remember Fatih, [and] Abdülhamit Han! When I miss those times, I wish I had lived in these good old days. 6 Fatih refers to Mehmet the Conqueror (1432-1481), who was the seventh sultan of the Ottoman Empire. He conquered Istanbul and ended the Byzantine Empire. Abdülhamit Han (Sultan Abdülhamit II) (1842-1918) was the thirty-fourth sultan of the Ottoman Empire. He was dethroned following the 1908 Young Turk Revolution in 1909. Note: poststratification weights are applied. See the Appendix for the unstandardized coefficients and robust standard errors. + p < 0.1, * p < 0.05, ** p < 0.01, *** p < 0.001.
The Kemalist nostalgia group received the following message: This heavenly homeland is the trust of the founders of the Republic. Since the early years of the Republic, this people have always stood firmly against several external outbreaks, and their collaborators within us, that have played games with our country. Remember the glorious War of Independence! Remember İnönü, [and] Mustafa Kemal Atatürk! When I miss those times, I wish I had lived in these good old days. 7 All three groups then responded to nostalgia, identity and populism items, respectively. In the final part of the survey participants responded to demographic questions such as sex, age and education.

Results
I tested the effect of nostalgia treatments on populist attitudes by building a populism index with five items taken from Akkerman, Mudde and Zaslove (2014). 8 I summed all items and built a scale between 0 and 100, where 100 indicates higher populist attitudes (Table 4).
According to Figure 1, respondents who received the Ottoman treatment are more populist than the control group participants, which confirms the second hypothesis. Although Kemalist nostalgia also has a positive impact on populist attitudes, its difference from the control group is not statistically significant. 9 Lastly, the difference between Ottoman and Kemalist nostalgia is also nonsignificant, which rejects the third hypothesis.
Regression analysis highlights that the Ottoman nostalgia treatment has a positive impact on populist attitudes (Figure 2). However, the Kemalist nostalgia treatment has a barely significant positive effect on populism (p = 0.08). To control for heterogeneous treatment effects, I also interacted the experimental treatments with party preference. While the effect of Ottoman nostalgia stays robust, the effect of Kemalist nostalgia disappears when party preference enters the equation, as in the interaction models. Finally, the effects of nostalgia treatments do not differ according to party preference. None of the interaction variables is statistically significant (see the Appendix for the results).  , who was the second president of Turkey after Atatürk and a hero of the Independence War. However, Islamist groups in Turkey broadly criticize him due to his staunch secularist rule. 8 Populism items range from −2 to +2. The item 'The politicians in the Parliament need to follow the will of the people' was dropped due to low variation and extreme negative skewness, which makes it impossible to manipulate with nostalgia messages. The item 'What people call compromise in politics is really just selling out on one's principles' was also dropped because it decreased Cronbach's alpha from 0.61 to 0.57, which was already a poor score. 9 The ANOVA result between experimental groups is [F(2, 785) = 2.596, p = 0.08]. The mean difference between the control and Ottoman nostalgia groups is 3.66 (p = 0.08), and the mean difference between the control and Kemalist nostalgia groups is 2.96 (p = 0.19).

Conclusion And Discussion
I conducted two different analyses using two different original datasets, and the results confirmed that nostalgia has a positive relationship with populism. The nostalgic electorate displays more populist attitudes at the mass level, and particular nostalgic messages increase populist attitudes in an experimental setting.
Populism has different characteristics in Turkey than in Western Europe in at least three ways. First, in Turkey it does not stem from anti-immigration sentiment (like in France or the Netherlands) or economic downturns (like Spain). Instead, the primary driver of populism in Turkey is the Islamist vs. secularist divide (Aytaç and Elçi 2019). Secondly, the right-wing populist AKP has been in power since 2002, resembling the long-tenured left-wing populist governments in Latin American countries such as Venezuela, Argentina and Ecuador (Selçuk 2016). Finally, although Turkey is not a member of the EU, it gained negotiating candidate status in 2005 and these negotiations are still in progress. While populism in Turkey also shares the antagonism of other European populist countries towards the elites in Brussels, successive AKP governments have kept the membership goal as an objective despite losing enthusiasm for full membership (Öniş and Kutlay 2019). Thus we should keep these similarities and differences in mind when generalizing the results.
This article demonstrates that nostalgia based on long-term cross-cutting cleavages breeds populist attitudes. Nostalgia offers a return to a lost paradise free of disruptions and corruption. It promotes an 'us vs. them' dichotomy in which the 'us' aims to protect its moral coherence and ethical certainty. In turn, nostalgia affects populist attitudes where 'the authentic people vs. immoral elites' dichotomy occurs.
Contrary to the assumption that populism is a reaction to a sense of crisis (Taggart 2004) or discontent (Spruyt, Keppens and Van Droogenbroeck 2016), as is valid in the cases of West European populism, this study illustrated that respondents who are satisfied with their lives also display populist attitudes (Aytaç, Çarkoğlu and Elçi 2020). This result also demonstrates that contemporary populism in Turkey does not stem from discontent, but rather from the formative rift on the secularist vs. Islamist division (Somer and McCoy 2019). It may also suggest that respondents internalize populist attitudes; even though they are satisfied with their lives, they may still be populist because their supported populist party is in power. We also see a similar pattern in Latin America, where more economically satisfied respondents display less support for measures that introduce checks and balances and protect civil liberties (Singer 2018).
Populism built on the formative rift argument can also be valid in the case of Hungary, which has been under the populist rule of Viktor Orban and his Fidesz party for a decade. Orban has constructed exclusionary rhetoric based on the trauma of the Treaty of Trianon, which led to the loss of two-thirds of the country's territory and half of its population in 1920. In December 2004, Hungary held a referendum on granting dual citizenship to ethnic Hungarians residing outside the country, which was depicted as a way of healing the scars of Trianon. While the leftist-liberal  Note: based on R package ggplot2 (Wickham 2016). See the Appendix for the regression table with standard errors. + p < 0.1, * p < 0.05 referendum to generate a duality in politics and society based on 'good' and 'bad' Hungarians (Inotai 2019;Pytlas 2013). In other words, similar to the commemoration practices of the collapse of the Ottoman Empire, Hungary still keeps pre-Trianon memories alive in daily politics along with public monuments, museums and speeches. The experimental tests illustrate that Ottoman nostalgia has a significantly positive direct effect on populist attitudes. Mean plots and regression analyses show that respondents who received the Ottoman nostalgia treatment have significantly higher populist attitudes than control group respondents. Although Kemalist nostalgia also has a positive impact on populist attitudes, the result is barely significant. Nevertheless, once again, the results confirm the primary hypothesis: nostalgia has a significantly positive effect on populist attitudes.
Previous experimental studies scrutinizing the effects of collective nostalgia have focused on its mobilizing impact on in-group vs. out-group attitudes. For instance, Smeekes (2015) and Smeekes and Verkuyten (2015) illustrated that national nostalgia boosts in-group identification and protection as well as prejudice against non-Western immigrant groups in the Netherlands. Similarly, they demonstrated that autochthonic attitudes positively moderate the impact of national nostalgia on opposition to Muslim expressive rights. In addition to these examples, which take nativism as a dependent variable, this study also shows that collective nostalgia triggers populist attitudes in terms of generating a 'pure people vs. corrupt elites' intergroup cleavage. In other words, nostalgia generates an 'us' vs. 'them' duality in different societies in several different ways. However, when it taps into the moral distinction between the elites and the people, nostalgia increases populist attitudes.
While right-wing populists may use nostalgia more widely than left-wing populists, Bolivar nostalgia of Chavismo and the United Socialist Party of Venezuela is the most prominent counterexample to this argument (Roberts 2016). Another contemporary example is US Senator Bernie Sanders, who instrumentalizes nostalgia by emphasizing a period when the working class in the United States lived in better conditions than they do today (Gaston and Hilhorst 2018). In short, the content of nostalgia may change, but populists widely exploit it.
That being said, this study may shed light on contemporary populist politics in other countries. Multiple nostalgias in a given country may also lead to populism, which may in turn heighten polarization in politics and society. For example, Trump's nostalgia for an era of the so-called non-degenerated American values vs. Sanders' nostalgic reveries mentioned above can both catalyze polarization in the United States. In a similar vein, there are clashing nostalgias in the UK. Former Labour Party leader Jeremy Corbyn refers to the golden days of the welfare state system in the UK, while Prime Minister Boris Johnson's slogan of 'Take Back Control' is a typical example of nostalgia for the period before the UK joined the EU, which fueled the populist tendencies that paved the way for Brexit (Gaston and Hilhorst 2018;Kenny 2017).
The major shortcoming of this article is that it is a single case study. Other case studies can challenge the external validity of some of these assumptions. However, Ottoman and Kemalist nostalgias are both sui generis. In other words, it is not easy to test their effects in another context. Nevertheless, one can also argue that every country has its unique nostalgia for a specific period or periods. Very few common historical eras allow us to conduct a comparative experimental study, such as the era before the influx of immigrants to Europe, pre-Second World War or socialist rule. Still, although Ottoman and Kemalist nostalgias are unique to Turkey, this article suggests that future studies should focus on idiosyncratic nostalgias and their effects on populist attitudes in other contexts.
What is more, this study is conducted in a country where a populist party has been in power for almost two decades. Hence, I assume that 18 years is sufficient time to establish hegemony in all dimensions of the electorate's political, social and economic lives. Thus it is easy to manipulate populism for the AKP as the party in power. However, the results may be different in countries in which populist parties are in opposition and receive less media coverage.
From a theoretical perspective, if one can manipulate populist attitudes by exploiting nostalgia in a country where the populist electorate is satisfied with its current conditions, we can understand why populist politicians are extremely inclined to instrumentalize nostalgia for their own ends. Nostalgia is a false memory, a whitewashed past, a glorious era that everyone can rally around. Therefore, the cost of exploiting it is low. Methodologically speaking, if one can manipulate populist attitudes in an extremely populist environment such as Turkey, where populism has already become an established ideology, the people's will has become extremely sacred, and the people represent the authentic and pure wisdom of politics, then it would be easier to conduct a similar analysis in other settings where populism has not penetrated all areas of political and social life.