Introduction

Social networks are tools for identity construction as well as a means by which users establish connections with others, thus representing an important research context to investigate self-presentation (^{Boyd & Ellison, 2007}). This is due to its potential to increase the forms of self-presentation and get feedback from others (^{Belk, 2013}). In this context, brands or celebrities find on Instagram an opportunity to reach consumers (^{Lee, Lee, Moon & Sung, 2015}) by posting carefully planned pictures. Therefore, politicians can work in the same line. Instagram offers the possibility to easily connect with voters and allows self-expression, which explains the potential of this platform for online self-representation and impression management especially in the case of the political candidate.

Thus, Instagram appears as a key tool for personifying political ideologies through image-based emotional communication. This fact can be framed in the politainment era (^{Nieland, 2008}; ^{Schultz, 2001}, 2012), where the politic content in entertainment formats keeps the ideology to the minimum in favour of personification (^{McAllister, 2007}; ^{Campus, 2010}; ^{Bennett, 2012}; ^{Adam and Maier, 2010}). In this context, the political candidate can act as a celebrity politician (^{Street, 2004}), using celebrity techniques to approach the electorate. One of these techniques is precisely the use of social media, that allows the candidate to become, in terms of Wood, Corbett and Flinders, an everyday celebrity politician (ECP) when showing his or her more personal and human facet (2016). In this sense, McLaughlin and Macafee conclude that “as politicians increasingly use social media sites to tell their stories, share their thoughts and feelings”, the citizens’ commitment “should be driven more by a desire to see a particular plot outcome rather than deliberate consideration of what that candidate is likely to do in office” (2019, 596).

This study moves the electoral candidate’s self-representation on Instagram and the personification of politics to the Spanish context, trying to expand the knowledge about the candidate’s use of social media in political marketing studies (cfr. ^{Hong & Nadler, 2012}; ^{Karlsen, 2012}; ^{McGregor, Lawrence & Cardona, 2017}; ^{McGregor, 2018}; ^{Kreiss, Lawrence & McGregor, 2017}). In Spain, Internet users choose social networks as an information source more often when following the electoral campaign. According to the CIS (Sociological Research Center) report referring the 2019 general elections, Spaniards affirm to use the social media for this purpose, being, in general, their second choice after TV, and the first choice for 13,6% of Spanish citizens (^{CIS, 2019}).

Hence, following a quantitative methodology, this paper explores the degree to which Spanish political candidates apply dramaturgic strategies for impression management in the April 2019 national elections (28A). More in concrete, it follows Goffman’s (1959) principles to explain candidates’ self-representation performance on Instagram, that is, the conscious selection of photographs which represent the best version of themselves or, at least, an image corresponding to what that citizens may consider an ideal candidate. The main aim of this paper is to explore the self-representation of the main five Spanish political candidates on Instagram ー@sanchezcastejón (Pedro Sánchez, PSOE), @pablocasadoblanco (Pablo Casado, PP), @iglesiasiturrionpablo (Pablo Iglesias, Podemos), @albertriveradiaz (Albert Rivera, Ciudadanos) and @santi_abascal (Santiago Abascal, VOX)ー, in order to ascertain if the candidate is showing himself on this social media in a predominantly personal and private context or, on the other hand, in a more professional and political way.

Following the tendency of politainment, the politic content loses power against the humanization of the candidate who indeed adopts celebrity techniques to reach potential voters, such as the intense use of social media. Previous studies have been carried out to analyze the image of the political candidate on social networks and, specifically, on Instagram, some of them focusing also on self-presentation strategies. These show a growing relevance of the figure of the political candidate on Instagram, but our paper is trying to fill a gap in this self-representation of the politician in two ways: (1) by studying the candidate during a pre-election and an election period and (2) by carrying out a comparative study between candidates. Hence, the comparative nature of this paper will help us to understand the possible differences in the personification of Instagram regarding the period and the ideology.

For the purpose of studying the self-representation of these political leaders on Instagram in these two periods, five research questions (RQ) have been planned:

RQ1: Which context ーpersonal or professionalー is giving more interaction to each Spanish candidate on Instagram?

RQ2: Are there relevant differences in each Spanish candidate’s self-representation on Instagram in terms of the human or professional image that they show?

RQ3: Are there relevant differences between the pre-election period and the period in which the political campaign occurs regarding the self-representation of each Spanish candidate on Instagram?

Besides, following the literature that gives relevance to the personalization of politics in recent political campaigns (^{Lalancette & Raynauld, 2019}; ^{Filimonov, Russmann & Svensson, 2016}; ^{López-Rabadán & Doménech-Fabregat, 2018}; ^{Quevedo-Redondo & Portalés-Oliva, 2017}), and, concretely, the studies that attach this personalization to new political parties in particular (^{Turnbull-Dugarte, 2019}; Caro-Castaño & ^{Selva-Ruiz, 2017}), the following two research questions are also formulated:

RQ4: Are the five Spanish candidates showing themselves on Instagram in a private and personal context more frequently during the election period than during the pre-election period?

RQ5: Arte the candidates of the three newest Spanish parties (Podemos, Ciudadanos, VOX) showing themselves on Instagram in a private and personal context more frequently than the candidates of the traditional parties (PP, PSOE)?

Literature review: instagram, self-representation and candidates’ impression management

Social media represent dynamic objects that are tweaked in response to their users’ needs and objectives (^{Van Dijck, 2013}); in the case of Instagram, “the creation of an idealized virtual existence” (^{Caldeira, 2016}, 155). The success of Instagram is partly due to the symbolic power of photography, considered as “an element of social connection” (^{Serafinelli, 2017}, 109). By combining both visual and textual information, Instagram posts act as a language that enables social interaction (Van Dijck, 2008). The study carried out by ^{Lee et al. (2015}) reveals that social interaction and self-expression are among the main motivations for using Instagram. Users specially value keeping a good relationship with others and updated about family and friends, connecting with people with similar interests, presenting their personalities and lifestyles and expressing their actual self. These two facts explain the possibilities of this platform to perform online self- presentation.

In the current online context, “the self becomes the centre of a virtual universe” (^{Van Dijck, 2008}, 62). Modern photography focuses on construction more than on remembering, that is to say, on creating a certain identity more than on keeping memories. Individuals articulate and construct their social identities by taking, storing and exchanging pictures online (Van Dijck, 2008). Indeed, digital photography is better than texts for self-expression and impression management (^{Lee et al., 2015}), as it allows to materialize and share identitarian aspects in front of others.

Specifically, individuals participate in self-presentation performances directed to build a certain identity before others as part of their daily interaction routines (^{Goffman, 1956}). That considered, Instagram might be defined as an environment in which users can replicate this performance online through personal pictures. Recent studies point out that “engaging in social media is becoming a key self-presentation tactic and behaviour in online cultures” (^{Stokes & Price, 2017}, 160). One of the main advantages that Instagram offers to political candidates is the possibility to build a coherent identity ーone that meets the expectations of voters in any senseー by using several resources and applying different strategies for identity construction ーi.e. clothing, backgrounds, gestures or social relationshipsー all condensed in one single picture or the account taken as a whole. Offline interaction contexts can never permit such a controlled “impression management”, using Goffman’s terms. Every element appearing in the frame is previously prepared. The author behind the picture and thus behind conscious self-representation is always the candidate, the person who intends to influence the audience’s impression. Though he or she may have a communication team in charge of the Instagram account, the highest responsible person of the photo is still the politician.

Our study relies on ^{Goffman’s dramaturgical theory (1956}) to explain online self-representation. According to Goffman, individuals in social interaction express themselves in such a way as to give others a desired impression, a favourable self-image, and strive to manage the impressions that others have of them. The act of presentation, therefore, constitutes a strategy of interaction to influence the impression that is given. This struggle for individuals to project their selves occurs both off- and online (^{O’Donnell, 2018}). ^{Marshall (2010}) creates the concept presentational media to define a new scenario in which self-presentation occurs also via social media and introduces the idea of the private self for public presentation, considering that the private self is now revealed to the public word. A large body of research has examined online self-expression with special attention to social networks (^{Schau & Gilly, 2003}; ^{Krämer & Winter, 2008}; ^{Brivio & Ibarra, 2009}; ^{Mehdizadeh, 2010}; ^{Scolere, Pruchniewska & Duffy, 2018}) and to Instagram specifically (^{Gyorffy, 2013}; ^{Smith & Sanderson, 2015}; ^{Kavakci & Kraeplin, 2017}), all assuming that “there is intentionality to sharing the self with others online through photographs” (O’Donnell, 2018, 131).

In virtual environments, self-expression occurs in terms of representation rather than presentation. As users choose what to share and which self-presentations are appropriate (^{Rettberg, 2018}), it is more accurate to use the term “self-representation”, for it means to show only a certain aspect of ourselves by consciously selecting pictures and background elements to share on Instagram posts. Photography cannot be an objective representation of reality, as it “always depends of a specific subjective point of view that has a constructive function” (^{Caldeira, 2016}, 150). Users decide to represent themselves by different kinds of images with the aim to connect with others (^{Enli & Thumim, 2012}). In this conception, online self- representation implies that users are constantly reinventing themselves, “managing the details they decide to share or to hide” (Caldeira, 2016, 143). Our study follows Rettberg’s approach and explains candidates’ performance on Instagram in terms of online self-representation, as representation is closer to construction and subject to interpretation.

When sharing pictures on Instagram, individuals consciously apply dramaturgic strategies for impression management. Self-representation on Instagram consists of taking several photographs to select those that best represent the idealized more than the actual self (^{Caldeira, 2016}; ^{Deeb-Swihart, Polack, Gilbert & Essa, 2017}). Our research claims that Instagram stands as the perfect platform for online self-representation and impression management. Literature on social networks frequently relies on Goffman’s approach, being our postulate consistent with previous lines of inquiry (Shau & Gilly, 2003; ^{Marshall, 2010}; ^{Smith & Sanderson, 2015}; Deeb-Swihart et al., 2017). Scholars turn to Goffman to study how the user manages these different audiences (^{Rettberg, 2018}) or the manner in which sharing pictures on social media blends the online and offline selves together (Deeb-Swihart et al., 2017).

Self-representation on Instagram acquires a particular relevance in the case of the political candidate. The main difference regarding the standard citizen user is the use of marketing strategies to achieve notoriety and sympathy between the electorate and, ultimately, get the vote. The political candidate has recently being investigated on Instagram (^{Turnbull-Dugarte, 2019}; ^{Lalancette & Vincent Raynauld, 2019}; ^{Selva-Ruiz & Caro-Castaño, 2017}; ^{Russmann & Svensson, 2017}; ^{López-Rabadán & Doménech-Fabregat, 2018}; ^{Poulakidakos & Giannouli, 2019}; ^{Russmann, Svensson & Larsson, 2019}; ^{Rodina & Dligach, 2019}), as well as other social networks (^{Enli & Skogerbø, 2013}; ^{Dimitrova & Bystrom, 2013}; ^{Rossini et al, 2018}; ^{Pérez-Curiel & Limón, 2019}) to analyze the image he or she creates through photography. In this sense, it should be noted that this type of research will always be framed within the concept of photographic propaganda proposed by Ana I. Barragán Romero (2017): images emitted by a power entity whose main goal is to achieve or maintain a privileged power position are analyzed. Taking this candidate’s intent into account, an apparently spontaneous use of the image will always respond to a propaganda purpose.

The interest in the image of the politician on Instagram is such that ^{Russmann and Svensson (2016}) have created a model to analyze it. They establish four groups of variables: those related to (1) perception ーperspective, retransmission and mobilizationー, (2) image management ーpersonalization, privatization and celebritiesー, (3) integration ーhybridization, shared content and references to election campaignー and (4) interactivity ーa variable that was crossed with the rest in Russmann and Svensson (2017). A more compact version of this methodology was applied to the 2014 Swedish elections (^{Filimonov, Russmann & Svensson, 2016}), and later, those results were compared to the 2017 Norwegian elections (Russmann, Svensson & Larsson, 2019). Based on this method, ^{Lalancette and Raynauld (2019}) analyzed the image of Canadian Prime Minister Justin Trudeau using five groups of variables: (1) the composition of the published image or video ーpeople featured, filters...ー, (2) the content of the captions ーemojis, hashtags, mentions...ー, (3) the political issues referred, (4) the stages of performance ーmedia, professional, private...ー, and (5) the configuration of the setting ーlocation, symbols, colors... Also based on Russmann and Svensson’s method, Turnbull-Dugarte analyses the Spanish political parties on Instagram during 2015 and 2016 national elections, adding a variable named “Europeanisation” (2019). Methods based on the dramaturgical theory of Goffman are also noteworthy (^{Stanyer, 2008}; ^{Svensson, 2012}; ^{Selva-Ruiz & Caro-Castaño, 2017}), an idea that will be the cornerstone in this paper.

In general, the results of these studies show a growing relevance of the figure of the political candidate on Instagram, but that does not necessarily entail him or her humanization. In this sense, the distinction between the individualization and the privatization as the two sides of the candidate’s personalization strategies needs to be considered: individualization entails a “focus on individual politicians as central actors”, and privatization refers to a “person distinct from their public role” (^{van Aelst, Sheafer & Stanyer, 2012}). Thus, the humanization of the candidate will be attached to privatization strategies ーthat is, the openness of the private lifeー, more than the individualization ones. In their study of Justin Trudeau’s use of Instagram, Lalancette and Raynauld determined that the politician’s image belonged to a growing personalization campaign: “Trudeau’s visual rhetoric and storytelling emphasized the image (ethos) of a youthful, well-groomed, and positive person at ease in every personal or public situation” (2019, 29). In Sweden, ^{Filimonov, Russmann and Svensson (2016}) concluded that the candidates had a strong presence in Instagram images, though without showing themselves significantly in private contexts but in professional settings, apparently spontaneously. Similarly, in the case of Spain, López-Rabadán and Doménech-Fabregat state that the image that political leaders projected on Instagram during the conflict in Catalonia responded, among others, to an “intense dynamic of personalization” and an “appeal positive emotional” in the message (2018, 1026). However, this network was not used to humanize the leader by publishing photographs of the private sector (López-Rabadán & Doménech-Fabregat, 2018). On the other hand, focused precisely on the humanization of the Spanish politicians on Instagram, Caro-Castaño and ^{Selva-Ruiz (2017}) showed that 26.22% of the publications of the Spanish deputies belonged to unofficial contexts. They conclude that the new political parties make a more humanized use of Instagram than the traditional ones, but that a certain limitation belonging to the traditional media still remains (Caro-Castaño & Selva-Ruiz, 2017). In the same line, Turnbull-Dugarte’s study reveals that newest Spanish parties on Instagram show their candidate more often than the traditional ones (2019). Also, ideology as a significant factor is pointed out by Russmann, Svensson and Larsson (2019), who state that left-side parties are taking more advantage of the possibilities of Instagram than right-side ones. Nevertheless, Poulakidakos and Giannouli (2019) concluded that Kyriakos Mitsotakis, the leader of the conservative and traditional party New Democracy, was the one who posted more private and personal pictures on Instagram. Besides, the study of Rodina and Dligach established that the authoritarian Chechen leader Ramzan Kadyrov uses Instagram to create an outstanding humanized politician, combining “the tactics of a populist politician and an internet celebrity” (2019, 95).

The present study brings online self-presentation and the use of Instagram by politicians together. Considering that digital photography allows to materialize and share identitarian aspects in front of others and that engaging in social media is a key contemporary self-presentation tactic and behaviour, we assume that Instagram offers politicians the possibility to meet the expectations of voters by building a certain identity through the use of several resources such as clothing, backgrounds, or gestures. By doing so, this paper combines Sociology and Political Communication through SNS, showing that Instagram stands as a suitable platform for online self-representation and impression management. The individual behind the picture and thus conscious self-representation is always the candidate. Hence, in Goffman’s view, the political candidate may be defined as the person who intends to influence the audience’s impression, corresponding to a deliberate propaganda aligned with a defined political communication strategy.

Previous studies outline the growing importance of private pictures as a way to emphasize humanized politicians, making private life open to public. We develop this line of research by examining to which extend did Spanish candidates post pictures oriented to manage the audience impression towards a more professional or humanized politician during the April 2019 elections.

^{O’Donnell (2018}) claims that impression management strategies should be further explored in photo-sharing SNS research. Following this line of research, our study states that it is the ideal self that political candidates try to construct as an essential part of their social media strategy in the election campaign, as performance “is a critical component in any public figure’s identity. For the politician […], the actual performance in a public forum is often a factor in how policies are received by a public” (^{Marshall, 2010}, 39).

Method

This paper’s methodology is based on previous quantitative work about political candidates’ self-presentation on social media, more in concrete on those studying the humanization of the candidate on Instagram (Lalancette & Vincent Raynauld, 2017; ^{Selva-Ruiz & Caro-Castaño, 2017}; ^{Quevedo-Redondo & Portalés- Oliva, 2017}; ^{Muñoz & Towner, 2017}; ^{Filimonov, Russmann & Svensson, 2016}; Russmann & Svensson, 2016). A content analysis (Krippendorff, 2004) has been carried out toward their Instagram posts, taking as items the empirical operationalization of Goffman’s theory. We have departed from Goffman’s work The presentation of self in everyday life (1956) and confronted it with its operationalization in the mentioned studies by Selva-Ruiz and Caro-Castaño (2017) and Russmann and Svensson (2016). Specifically, we have chosen interaction and front as the two main concepts, includying in the latter other Goffman’s terms. As a result, the items included in the coding sheet are the following:

1. Interaction. Goffman refers to the reciprocal influence of individuals upon one another’s actions. Here, the audience is understood as those who contribute to the performance. For our study, likes and comments were measured. The act of giving a “like” is a quick form of validation, so an image can be judged by its quantity of likes. Commenting implies social validation, as it requires more effort from the audience than the “like” (^{Caldeira, 2016}). Likes and comments give the candidate information about the audience reaction, which may lead him to change his self-representation strategy. As Caldeira states, “moved by the desire to receive positive judgement from the other users, the photographer [...] tries to anticipate the responses that the viewers might have” (2016, 154). With this data, the engagement of each candidate in each period was considered.

2. Social frame. This is the first approach to Goffman’s front, considering a professional context, a personal context or an uncertain context. In our analysis, the front is the Instagram post itself, where we find the expressive resources employed by the individual during his performance (Goffman, 1956). The front is key for Goffman’s theory, as he gave a central role to an actor’s expression (^{McCall, 2013}).

3. Performance. Several kinds of performances were identified, according to the action the candidate is developing in the post: party or campaign affair, solidarity act, sport or cultural act, protest/public demonstration, private/familiar act, or hobbies. The three last performances will highlight the human condition of the leader more than the rest.

4. Role. Performed at the front, it is determined by the kind of performance itself. In Goffman’s terms, the social role is “the enactment of rights and duties attached to a given status” (1956, 9). To construct the items, we departed from Robert’s “symbolic roles” of the political candidate (^{Roberts, 1993}) and from the “ideal candidate frame” of ^{Goodnow (2013}), that have been bringing together in ^{Quevedo-Redondo and Portalés-Oliva’s study of the image of the candidate on Instagram (2017}). Here, we have considered that the candidate can act as a “leader” (most professional one), “citizen”, “father”, “friend”, “son”, or “partner”.

5. Setting. The group of physical elements or background items of the performance (^{Goffman, 1956}) were operationalized through predefined different stages such as “street”, “workplace”, “nature” or “candidate’s home”.

Goffman’s key concept backstage or “back region” is not included because it is defined as the place where “no member of the audience will intrude” (1956, 70). Given that Instagram users cannot access the communication strategy behind the posts, the backstage as has not been coded.

The five Instagram profiles and the political leaders that show are the following (Table 1):

This exploratory study focuses on two different contexts: one month of pre-election campaign (PEC) ーfrom 11th March 2019 to 11th April 2019ー, and the fifteen days of election campaign (EC) ーfrom 12th to 28th April, including in this period the 28A Election Dayー. The sample consists of 473 Instagram posts: 253 from the PEC and 220 belonging to the EC. The data was collected during the first week of May, a week after the national elections of April 28 2019. Thus, the number of followers, likes and comments of every candidate on Instagram belongs to May 2019. The empirical analysis was carried out by the authors of this paper, departing from a previous reliability test (Krippendorf’s α) that showed an average agreement of 0,84.

Results: the professional over the humanized candidate

Spanish political campaign of 28A was dominated by the tension generated by the quick growth of the far- right political party VOX. Created in 2013, VOX entered in Spanish political life at the end of 2018, when it gained 12 seats in the Andalusian Parliament. With a nationalist discourse, VOX had experimented a momentum due to the Catalan secessionist movement (^{Ferreira, 2019}). This region of Spain claimed for the independence at the end of 2017, which caused several demonstrations and riots for and against the Catalonia independence. Given this background, the claim of VOX campaign for 28A was “For Spain” (Por España). The party had a strong presence on social media, where it created accounts for a series of events called “Beer for Spain” (Cañas por España), which public were young people. The traditional right party PP took advantage of the growth of nationalist movements and it created the campaign “Safe value” (Valor seguro), using a conservative tone. With a party positioning, the liberal Ciudadanos claimed “Come on, Citizens!” (¡Vamos, Ciudadanos!). It also was an issues-centered campaign, and, as an originally Catalan party formed in 2007, it took the Catalonian conflict as a target for a discourse of unity. On the other side of the Spanish political spectrum, the traditional left party PSOE, that was the party that called for the elections, bet on a concept of campaign empty of ideology; the claims were “Make it happen” (Haz que pase) and “The Spain you want” (La España que quieres). Finally, the new party Unidas Podemos, located to the left of PSOE, presented its campaign called “You write the History” (La Historia la escribes tú), through with collected several social protests attached to the traditional social and economic system. Once known the concepts developed by these political campaigns, we can explore the self-representation of each candidate on Instagram.

According to Figure 1, the most popular candidate on Instagram is Abascal (VOX), with nearly 300k followers. This is linked with the strong social media strategy his party was following. In the same way, as Table 2 shows, Abascal gets the highest engagement rate1 in both PEC (82.87) and EC (101.3), followed in PEC by Casado (PP), Iglesias (Unidas Podemos), Rivera (Ciudadanos) and Sánchez (PSOE). All of the candidates see their engagement rate increased considerably in the EC period, when Rivera reaches the third position being the only candidate that is able to double his engagement.

1 The engagement rate was partially based on ^{Muñoz & Towner (2017}) and ^{Turnbull-Dugarte (2019}), and it was calculated using this formula: [(likes+comments) / followers] x 1.000. The result refers to the coding time (May 2019), rather than the actual date of the post.

Regarding the social frame, photos shared by candidates on their official accounts are found to depict them mostly in a professional context in opposition to a personal-human context. According to Figure 2, in the PEC all the candidates posted pictures in political or media contexts in more than 50% of the occasions. It is worth noting that, in the case of Sánchez, 89.74% of his posts in PEC show him in a professional context. Casado shows the lowest percentage with 53.85%. Considering the EC (Figure 3), Rivera is now the one who shows the lowest percentage (57.5%) and Sánchez the highest again (93.42%).

It is interesting to look at these data at the light of the interaction aspect on Instagram. When looking at users’ likes and comments (Table 3), it was found that in the PEC period interaction rises if the candidate appears in a personal context only in the case of Rivera and Abascal (Image 1 is his most liked photo), but not in the rest of candidates. In the EC period, on the other hand, all candidates see their figures augmented when posting personal photos. Of special relevance is the case of Abascal, whose average number of likes when posting photographs in personal contexts ascends to 48,287. Indeed, he is by far the most-liked candidate on Instagram.

The great majority of performances identified are campaign events. In the PEC period (Table 4), more than 50% of the actions and events shown on the pictures shared were public actions connected to the party or its campaign promises. Specifically, Rivera represents himself in campaign events 58.49% of the times; while Casado, Sánchez and Abascal show figures between 70% and 80%. Iglesias has the highest percentage with 90.63% of posts performing campaign events. It is important to draw attention to family events, where candidates represent themselves only in less than 10% of their posts. The highest figures in this case are Abascal’s (7.89%) and Casado’s (7.69%). It is interesting to observe that the former president Sánchez does not share pictures performing private activities. Regarding hobbies, only Rivera posts pictures in which a hobby is performed (5.66% of his posts).

Similar data are observed in the EC period (Table 5). Campaign events are the protagonists of most posts: all candidates except Rivera represent themselves performing campaign-related activities in more than 70%. Regarding the private/family events, Casado and Iglesias raise these kinds of performances up to 11.11% and 11.76% respectively; Sánchez ascends from 0% to 2.63% and Abascal, who showed the highest percentage of personal performances in the PEC period, falls to the lowest (2.33%). Remarkably, private/family events in Rivera rise up to 20%, which represents the highest percentage of private/family performances by a candidate in both PEC and EC periods. Indeed, he is also the candidate with more pictures representing himself performing hobbies (7.5%).

Our analysis found that the most preferred role by all candidates in both periods is “leader”, which is intrinsically linked to a professional candidate self-representation. In the PEC period (Table 6), all candidates represent themselves as a leader in more than 60% of their posts, Casado showing the lowest percentage (63.04%) and Iglesias the highest (81.25%). Candidates represent the role “citizen” unequally: Rivera has the highest representation of this role (30.19%) and Abascal the lowest (10.53%). Candidates also played family roles in their pictures, such as “father”, “son” or “partner” in certain occasions. Sánchez, who represented no private/familiar performances in the PEC period, does not play the role “father”, and only Casado shares pictures as a “partner”.

In the EC period (Table 7), Rivera, Casado and Iglesias decrease to 47.5%, 59.26% and 73.53% (respectively) their posts performing the role “leader”. On the other hand, Abascal increases his posts in this role up to 86.05%, while Sánchez shows the highest number of pictures as “leader”: 94.74%. The role “citizen” remains stable for Rivera, Casado and Abascal, while in the case of Iglesias it doubles regarding PEC. Other roles aligned with the human-candidate, though remaining low, are grown regarding PEC: Casado raised the “partner” role up to 7.41% and Rivera performed the “son” role in 7.5% of his posts, being the only one playing that role.

The setting is more irregular in terms of figures. Assembly halls or events rooms are generally preferred by candidates in both PEC and EC period, although the figures vary. In Table 8, Rivera and Casado show the greatest variety of settings in PEC. Rivera represents himself mostly at the street (32.07% of his posts), as well as Casado (24%). On the other hand, Abascal and Sánchez represent themselves mostly at assembly halls. The setting “home” deserves special attention, as it would identify the humanized-candidate self- representation more clearly. None of the candidates showed their homes in more than 4% of their posts, belonging the highest figures to Rivera (3.77%).

Following with the EC (Table 9), assembly halls are the preferred settings again, with the exception of Rivera: he only represents himself at an assembly hall in 12.5% of his posts, selecting the street as setting in 41.5% of his posts. Indeed, he chooses “home” in 10% of his posts while the other candidates simply eliminate this setting.

Discussion and conclusion

General elections of 28A 2019 gave the victory to PSOE, followed by PP, Ciudadanos, Unidas Podemos and VOX. Nevertheless, due to an absence of agreement, the general elections were repeated on November 10 2019, where PSOE won again but VOX reached the third position in the Congress and Ciudadanos experimented a great fall, loosing 37 seats in comparison to 28A results. In both election period, the combination of social and political tension on the street and the activity and impression management of the candidates on social media played a key role in the Spanish political life.

Our research emphasizes the idea that an apparently spontaneous use of the image is always corresponding to a propaganda purpose. The results of our study unveil that, in general terms, Spanish candidates in 28A 2019 national elections tend to show themselves more as professional politicians ーalthough some candidates of new parties applied humanization strategies. In other words, thet tend to perform self- representations closer to the professional candidate than to the humanized candidate. This is in line with the studies of ^{Filimonov, Russmann and Svensson (2016}), ^{López-Rabadán and Doménech-Fabregat (2018}) and Caro-Castaño and ^{Selva-Ruiz (2017}). Considering that all candidates ーat different levelsー prefer professional contexts, appear mostly at campaign events, represent themselves as the leader of the party and choose assembly halls as the main setting of the photographs they share on Instagram, it follows that candidates trust more on their public than private self to be depicted in both PEC and EC Instagram posts. Nevertheless, our analysis suggests that users tend to interact more when the post shows a personal/private more than professional context, as it was observed in Table 3. Regarding RQ1 (Which context ーpersonal or professionalー is giving more interaction to each Spanish candidate on Instagram?), differences between the PEC and EC periods and the candidates were observed. Before the campaign, only two of the candidates, Abascal and Rivera, got more interaction thanks to personal posts, while during the campaign all of the candidates obtained more likes and comments in personal pictures. It is also noteworthy that Abascal and Rivera have the highest number of followers.

Answering to RQ2 (Are there relevant differences in each Spanish candidate’s self-representation on Instagram in terms of the human or professional image that they show?), our findings reveal that, in general terms, the five candidates tend to maintain a professional self-representation on Instagram. The most performed role for all the candidates was the one of “leader”. Nevertheless, they performed self- representation in different ways, being some of them closer to the humanized than to the professional candidate. Rivera, leader of one of the new parties (Ciudadanos) and second in the rank of followers, is the candidate showing a higher balance between a human and professional image. This fact can be connected to the casual nature of his campaign “Come on, Citizens!”. Although he tends to represent himself in a formal way, he has the highest number of informal pictures: at cultural/sports events, at private/family events (where he has the higher percentage, 20%, in the EC period), performing hobbies or at the street as one of his favourite settings. Besides, he has the highest number of posts in the role of “citizen” in both PEC and EC. Conversely, Abascal, leader of the newest party (VOX) and first on the rank of followers, stresses his professional more than personal image. He prefers to represent himself as a leader, at campaign events (95.35% of performances in EC period) and at assembly halls. As he is the most unknown candidate in the 2019 national elections, his strategy might be oriented to depict an image of a professional and suitable candidate. Iglesias, leader of other of the newest parties (Unidas Podemos) and third in the rank of followers, also prefers a professional image, as he shows himself primarily in formal contexts and at campaign events (more than 80% of his posts for both cases and both periods). Although he represents himself in the role “leader” in more than 70% of posts (both periods), he intends to give a feeling of closeness as he is the candidate who represents himself at the street the most (alongside with Rivera), which corresponds to a strategy that combines the professional and the humanized candidate (Image 2). In fact, Unidas Podemos’ campaign is based on social movements appeared on the street. A balance between assembly halls and the street as settings is observed in his posts.

Looking at the traditional parties, Casado, leader of PP and presenting his candidacy for the presidency for the first time, is the last one in the rank of followers (half the followers that Abascal). He also shows a balance between the personal and professional candidate. Nevertheless, it is noteworthy that he tends to represent himself as a leader who feels close to citizens (Image 3), following the same strategy as Iglesias or Rivera. Lastly, Sánchez (PSOE), who is running for the reelection, is the penultimate in the rank of followers, close to Casado. A strong predominance of professional contexts is observed. Virtually all of his pictures (Image 4) represent him as a leader and at campaign events. He also tends to post campaign official materials, like posters or calls to vote, which gives a strong corporate style to his account as the leader of the party. Sánchez offers citizens no clues about his private life; indeed, our study found no pictures showing him in family-related roles (such as “father”, “son” or “partner”). Our analysis showed that he, as a president, is the candidate with a more serious and professional Instagram account, being these results aligned with previous studies (Poulakidakos & Giannouli, 2019). This fact can be related to the one of being the least followed leader.

In relation to RQ3 (Are there relevant differences between the pre-election period and the period in which the political campaign occurs regarding the self-representation of each Spanish candidate on Instagram?), Rivera and Iglesias are the candidates that incremented their humanized self in the EC, while the rest reduced it. Whereas Casado, Sánchez and Abascal incremented the number of professional pictures in the EC, Rivera and Iglesias posted more personal pictures, though with a very low increment regarding PEC (6.65% in Rivera’s case and 2.02% in Iglesias’). In the same sense, Rivera and Iglesias also reduced their pictures about campaign affairs in EC, but not in a significant way (-8.49% and -2.39% respectively), at the same time that the rest of the candidates incremented these kinds of posts. However, all candidates except Sánchez and Abascal reduced their role “leader” in EC. It is also noteworthy that, while the roles of “citizen”, “father”, “friend”, “son” and “partner” are distributed in the candidates during PEC, only Rivera and Casado represent these humanized roles significantly in EC. Lastly, all candidates except Rivera incremented their appearance in assembly halls in EC.

Thus, the answer to RQ4 (Are the five Spanish candidates showing themselves on Instagram in a private and personal context more frequently during the election period than during the pre-election period?) is partly negative: not all but only two candidates, Rivera and Iglesias, showed themselves on Instagram in a private and personal context more frequently during the election period than during the pre-election period. Among these two, Rivera is the candidate who follows this strategy more strictly (Image 5). This can be confronted with the fact that, as it was seen in Table 3, personal pictures get more interaction in EC, which may lead to conclude that the candidates don’t take this fact into account, since the majority of them tends to increment their professional self in EC. On the contrary, Rivera gets the double of engagement in EC.

Finally, the answer to RQ5 (Are the candidates of the three newest Spanish parties (Podemos, Ciudadanos, VOX) showing themselves on Instagram in a private and personal context more frequently than the candidates of the traditional parties (PP, PSOE)?) is also negative: the candidates of the three newest Spanish parties (Podemos, Ciudadanos and VOX) don’t show themselves on Instagram in a private and personal context more frequently than the candidates of the traditional parties (PP and PSOE). According to our study, the newness of the party in relation to the humanization of the candidate is not a relevant factor in the Spanish case, as it was in the studies of ^{Selva-Ruiz and Caro-Castaño (2017}) and ^{Turnbull-Dugarte (2019}). Taking PEC and EC as a whole, Rivera, leading a new party, is the candidate who shows himself in a more humanized way, but he is followed by Casado, the candidate of a traditional party. Hence, the ideology and not the newness might be a relevant factor, considering, in line with Poulakidakos and Giannouli (2019), the conservative ideology as the one that can explain this humanization on Instagram. However, Iglesias, leader of the most leftist and new party Unidas Podemos, is in the third position showing himself as a humanized politician.

A key practical implication of our study is the contention that ^{Goffman’s (1956}) dramaturgical theory on self- presentation is useful to explain the candidates’ use of Instagram for self-representation. Our study shows that Spanish candidates’ use of Instagram in both PEC and EC may be defined as an instrument for self- representation corresponding to their election campaign strategy, with the aim to construct a specific image of themselves. This paper applies Goffman’s dramaturgical theory to contemporary online communication through its combination with studies on online self-representation and impression management, as well as political communication studies about the image of the candidate on Instagram. On this basis, we define the political candidate as a person who intends to influence the audience’s perceptions about his image, thus building and transmitting an ideal self, which represents a major theoretical contribution. Like standard users, politicians post pictures on this network motivated by the needs of interaction and self-expression, which corresponds to the body of research on identity and social networks. The particularity in the case of politicians is that their motivations are not social-psychological but correspond to propaganda purposes, as they are oriented to build an ideal self-image that meets the expectations of voters, thus applying different strategies for impression management.

As a further theoretical implication, we define Instagram as an optimum social network for online self- representation and impression management, specifically in the case of the political candidate. Our results are aligned with previous studies regarding self-presentation as one of the key functions of modern photography (^{Van Dijck, 2008}) and Instagram as a context in which users portray an idealized version of themselves in front of their followers (^{Caldeira, 2016}). Regarding political communication, candidates get the opportunity to deliberately choose what to share on Instagram, which elements are appropriate to perform self-representation as closer to a humanized (e.g. children, pets, home) or a professional candidate (e.g. assembly halls). In general, the humanized candidate, in the same way as general users, would most post photos with family, friends or in informal contexts, as their objective is to connect with citizens (potential voters). On the contrary, the professional candidate would show himself in a suit or giving a press conference in order to show a sense of trustworthiness, although depicting a close image by, for example, finding time to greet kids. In general, the candidates’ fear of, as ^{Marshall (1997}) states, overplaying their ordinariness, is present in the 2019 Spanish general elections that took place on Instagram, as any candidate posted more private than professional images regardless the period of PEC or EC.

This study, as a continuation of other political marketing studies based on the candidate’s use of social media (^{Bennett, 2012}; ^{Hong & Nadler, 2012}; ^{Karlsen, 2012}; ^{Kreiss, Lawrence & McGregor, 2017}; ^{Lalancette & Raynauld, 2019}; ^{McLaughlin & Macafee. 2019}; McGregor, 2018; Poulakidakos & Giannouli, 2019; ^{Quevedo-Redondo & Portalés-Oliva, 2017}; ^{Wood, Corbett & Flinders, 2016}), contributes to develop the line of research on the image of the political candidate on Intagram as follows. First, the study shows that Spanish candidates trust more on their public than private self to be depicted, by selecting pictures in professional more than personal contexts and showing themselves as leaders more than fathers or friends. Nevertheless, and as a second implication, it is observed that some of them come closer to the humanized than to the professional candidate, with only two candidates incrementing their humanized self in the EC (while the rest reduced it) and most of them reducing their role “leader” in EC, although incrementing their appearance in assembly halls in EC. Third, our study confirms that the newness of the party in relation to the humanization of the candidate is not a relevant factor in the Spanish case; the ideology and not the newness might be a relevant factor. The conclusions of this study can be relevant for both political communication studies and political communication activity because it has shown that (1) the humanization of the candidate on social media increases interaction with users, (2) mostly during the election campaign: the only two candidates that incremented their humanized self during the election campaign got to double their engagement in this period.

As a final and key contribution, this paper shows that Instagram is an important part of contemporary election campaign strategies, essentially, as an instrument for self-representation used to build a specific image of the political candidate.

Our study is at the forefront of new and expanding paths of inquiry on online self-representation. Nevertheless, it has some limitations, being the scope the main one. Spain is not a representative sample of the political use of Instagram by all western democracies. This fact implies an impediment that prevent us to make an extrapolation of the results as an explanation for Instagram politics in other countries. Another limitation is the absent of a qualitative method that helps us to understand the motivation of the politician in his self-representation on Instagram. In this sense, interviews with someone in charge of the party’s communication would have been revealing. However, our study still reveals a prominent use of humanization on Instagram during electoral and pre-electoral periods and it shows relevant differences between candidates.

Further research should compare this data to future elections period in Spain in order to observe the evolution of the humanization of the candidate on Instagram. Making a comparison between Spain and other countries also can contribute to make a solid conclusion about the humanization status in Spanish politicians. Besides, it would be interesting to apply a mixed quantitative-qualitative methodology gathering focus groups in order to study public’s responses to candidates’ pictures on Instagram, along with the collected data about interaction.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Acknowledgement

The work presented in this article was performed in the context of the research group IDECO, Grupo de Investigación en Comunicación Política, Ideología y Propaganda (Universidad de Sevilla, SEJ-539).