What Election Is This? Candidate Versus Party Contests in Indonesia

In one of his works, Kölln (2015) evaluates the performance of individual representatives and parties, concluding that party performance generally surpasses individual representation in pluralist democracies. The design of electoral systems often plays a key role in such debates (see Benoit, 2007; Colomer, 2009; Cox, 1997; Duverger, 1963; D. M. Farrell, 2001; Gallagher & Mitchell, 2005; Sartori, 1986; Yan, 2022). In district systems, where competition centers on candidates, personal votes and individual representation are emphasized (e.g., Cain et al., 1987; Carey & Shugart, 1994; Lancaster, 1986), though this system comes with both advantages (Curtice & Shively, 2009; Kunicová & Rose-Ackerman, 2005; Rohrbach, 2022; Vargiu, 2022) and challenges (see Chang & Golden, 2006; Lancaster, 1986; Mainwaring, 1991). In contrast, proportional systems (see Cox, 1997; Mitchell, 2000; Shugart et al., 2005) place more emphasis on parties in elections (see Müller, 2000).

In practice, however, individual representation often becomes a priority. Indonesia, for instance, is a country working to strengthen individual representation, both in its bureaucracy (see Kriyantono et al., 2022) and in politics. The history of Indonesian elections is closely tied to the electoral systems they’ve adopted. Article 22E, paragraph one of the Indonesian constitution stipulates that elections must adhere to principles of fairness. “Fairness” here refers to Indonesia’s proportional electoral system, which has evolved over time, although it was once characterized by a closed-list system during the authoritarian era. A significant change occurred in the 2004 elections when Indonesia implemented a semi-open list system, allowing voters to select candidates on the ballot. By the 2009 elections, the fully open proportional system^[1] gained momentum, whereby candidates with the highest vote within their party earned seats.

The more open electoral system has been in place for four election cycles, from 2004 to 2024, yet several facts reveal a gap between theory and reality in Indonesia. First, PDI-P’s victory in the last three elections suggests the party remains in demand.^[2] Second, 94% of the previous legislators were candidates in 2024 (Muhamad, 2023), and 44% of those failed to win seats (Rizky, 2024). About 36% of the winners in the 2024 legislative elections were not incumbents (Muliawati, 2024), and many incumbent candidates and big name candidates lost their seats (Alinea, 2024), pointing to weak individual representation. This article presents the meaning of elections for voters as a foundation for understanding the gap. This article considers the meaning of elections in terms of the most important elections for voters. LaCombe and Juelich’s work (2019) showed how it relates to higher participation. In our 2019 survey of 800 respondents,^[3] 67.4% identified the National House elections as the most important. Based on these insights, this article will focus on the meaning of the most important legislative elections. What is the meaning of the election for voters? Do they view it as a contest between parties or between candidates?

There are several reasons why the questions are important to pose. First, voters are aware of the type of election they are facing, and this awareness shapes their understanding of the election’s meaning, which is then linked to the nature of the election they expect. Second, the questions relate to the ongoing conceptual debate about the role of parties in modern democracy. Third, the questions are pertinent to ongoing scholarly debates concerning the ideal form of political representation—specifically, whether representation should be vested in political parties or in individual legislators. Fourth, both parties and candidates are actors with the potential to compete against each other in the meaning of elections. By exploring the views of voters, this paper expands the electoral system discussion, which has so far been more focused on the concepts and perspectives of scholars.

Literature Review

Electoral System, Representation, and Meaning of Election

The design of electoral systems is one of the topics of representation studies (see Benoit, 2007; Colomer, 2009; Cox, 1997; Duverger, 1963; D. M. Farrell, 2001; Gallagher & Mitchell, 2005; Sartori, 1986; Yan, 2022). In majoritarian-district systems, where competition revolves around candidates, personal vote and individual representation are reinforced (e.g., Cain et al., 1987; Carey & Shugart, 1994; Lancaster, 1986), although it has some advantages (Curtice & Shively, 2009; Kunicová & Rose-Ackerman, 2005; Rohrbach, 2022; Vargiu, 2022) and challenges (see Chang & Golden, 2006; Lancaster, 1986; Mainwaring, 1991). This dynamic contrasts with proportional systems (see Cox, 1997; Mitchell, 2000; Müller, 2000; Shugart et al., 2005). Manin et al. (1999) present a more moderate term, focusing on a more open electoral system rather than distinguishing between the two systems mentioned above.

While it is theoretically argued that a more open electoral system can strengthen individual representation, scholars continue to debate whether parties or individuals dominate elections. Institutionalists argue that a political framework fosters party dominance in elections (Ezrow, 2010), with the structure of the electoral system often playing a significant role (Golder & Stramski, 2010; Klingemann et al., 2006). Several scholars point to party dominance in various forms: shaping the government system (Dalton, 2017), presenting issues and public policies (Budge et al., 2012; Dalton et al., 2011; Yu & Hsu, 2019), partisanship (e.g., P. G. Chen & Goren, 2016; Idid & Souket, 2014) and party spectrum (e.g., Albertazzi & van Kessel, 2021; D. Farrell & Scully, 2007; Miller & Klobucar, 2000; Rohrschneider & Whitefield, 2012).

While parties are often seen as the dominant actors, other studies highlight the significant role of candidates as key actors as well. The literature on the use of candidates to promote parties draws on the concept of party ambassadors, which ambassadors act as messengers (Marland & Wagner, 2020). Ambassadors, however, have individual characteristics, such as varying talents and interests, that reflect their authenticity and credibility (Guzman & Sierra, 2009; Speed et al., 2015), and the message does not resonate with grassroots or local areas (Pich et al., 2016). This creates both opportunities and risks for parties, as candidates can either enhance or damage the party’s image. The challenge for candidates lies in reconciling them (BouNassif et al., 2023; Loat & MacMillan, 2014; Speed et al., 2015). Several studies have explored whether voters are choosing parties or candidates (see Barberá, 2010; Walder & Strijbis, 2022). Explaining voters’ priorities is challenging when these two qualities cannot be simultaneously achieved.

Other studies also examine voters’ views on elections. Several scholars highlight a correlation between voters’ views and their participation (e.g., Franklin, 2001; Górecki, 2011; Orford et al., 2009). Voters’ perspectives on elections have also been associated with the emergence of social divisions (Bellucci & Heath, 2007). Additionally, Dragu and Fan (2016) analyzed voters’ views to explain the high level of party competition. These studies suggest that voters’ views on elections can shed light on the dynamics that shape electoral outcomes.

Research on Indonesia’s Fully Open Electoral System

Several studies have been conducted since the implementation of the open proportional election system in Indonesia. Some reveal dynamics that are increasingly focused on candidates. The lack of party involvement in voter mobilization (Tawakkal, 2009), the rise of candidate-based campaign teams (Aspinall, 2014), vote brokerage for candidates (Tawakkal, Kistanto, et al., 2017; Tawakkal, Suhardono, et al., 2017), and the growth of money politics in the name of candidates (Pradhanawati et al., 2019) all point to the open electoral system as a key driver.

In terms of party identification, during the authoritarian New Order era (1966–1997) and the initial elections following the democratic reforms (1999–2024), party identification remained strong (Aryanto, 2023; Rafni & Suryanef, 2004; Suryadi, 2005). However, in the post-reform period, marked by the introduction of executive elections and a more open proportional system, empirical studies have produced mixed findings.. Several studies have shown strong party identification. For example, research by Haryanto (2014) and Apriani (2019) demonstrates that party identification still plays a role in explaining candidate victories. However, other studies suggest that, in line with the open election design, party identification has weakened (e.g., Apriani & Irhamna, 2020; Liddle & Mujani, 2007; Sihidi et al., 2019).

Scholars have also studied parties in Indonesia, with their findings falling into two main categories. First, there are studies that argue party strength is a crucial factor in elections. For example, Hanafi (2018) highlights the importance of strengthening party programs to remain competitive in elections. Other research indicates that large parties continue to be a significant explanation for electoral victories (see Immanuel & Djuyandi, 2024; Nurdiansyah, 2018; Olifvia & Suryanef, 2023). Second, some scholars argue that the role of parties has weakened. Scholars like Gunanto et al. (2020) and Ansyari et al. (2019) suggest that party victories are often driven by non-party factors, such as patron-client relationships, candidate profiles, and non-party networks.

Theoretical Framework and Research Gap

Many scholars provide a theoretical foundation that a more open electoral system leads to individual representation and more individual-focused elections, with increased competition among candidates (Benoit, 2007; Manin et al., 1999; Sartori, 1986). This theory regarding the meaning of elections is grounded in scholars’ concepts and perspectives on electoral practices, but the question of how this system works in practice or how voters perceive it remains. It is important because understanding voters’ views can provide a more comprehensive explanation beyond the academic theories of concepts and practices. Competition between individuals seems to be confirmed in several studies in Indonesia (see Pradhanawati et al., 2019; Tawakkal et al., 2020; Tawakkal & Garner, 2017). Hypothetically, voters’ views on the election would also confirm this.

Based on the literature review above, the gap lies in the theory’s inability to explain the researchers’ findings concerning the strength of party identification and the consistent victories of certain parties in a district. How can we provide a scientific explanation for the differences in existing research regarding strong and weak party identification in an open election system in Indonesia? Additionally, existing studies on the meaning of elections have not connected the meaning of elections to the research gap. This article offers new insights by exploring how voters view elections, providing a foundation for further understanding the electoral system.

Methodology

To address the central question of this article, research was conducted through conversations with over 400 voters across three major islands in Indonesia—Java, Sulawesi, and Kalimantan—between 2019 and 2024. Each island represents distinct sociological characteristics. The participants in this study were eligible voters, categorized into two groups. The first group consisted of individuals who had participated in two elections conducted under different electoral systems—namely, those who voted in the 2004 election or earlier, and again in the 2009 election or later. The second group was comprised of voters who had only experienced one type of electoral system, specifically those who first voted in the 2009 election or thereafter. The first group was comprised of 115 male and 88 female participants. In terms of age distribution, 66 participants were 31–35 years old, 76 were 36–40, 44 were 41–50, and 17 were aged 51 or older. The second group included 103 male and 94 female participants, with 56 participants aged 21–25 and 141 aged 26–30. Participants in this study were recruited based on available opportunities rather than a predetermined sampling method. The number does not proportionally represent the population. Instead, this research employs a qualitative approach to public opinion, which is better suited for exploring nuanced meanings and perspectives. Rather than following a strict questionnaire format, this research employed open-ended conversations, allowing participants to guide the conversation. The conversations were held both individually and in groups. Data collection involved taking notes. I actively listened to informants and analyzed the direction of the conversations to determine whether it related to political parties or individual candidates. These insights were then documented by marking references to the party or candidate for each participant.

This study adopts a psychological framework that suggests that talk reflects thought. First, talking allows individuals to express their ideas and identity (Bellah et al., 1985). Second, talking is closely tied to thinking, since language can shape and reflect thought. Thus, speaking is generally equated with thinking, an essential activity. This view rests on the assumption of psychic unity (Bruner, 1996; Shweder, 1995). I adopted the framework by observing which topic is dominant, with dominant topics of discussion representing dominant thinking. To understand the meaning of elections, I introduce the term “top of talk” to describe the topic that dominates the conversation. Initial questions or statements set the stage for deeper exploration.

The “top of talk” offers a richer, more nuanced understanding compared to “top of mind” used in surveys, as it allows participants to discuss themes freely and highlights emerging conversation trends. The analysis of findings was conducted using two complementary models: the “kick-off” and “recall” approaches. The first focuses on initial themes spontaneously introduced by participants. The second examines participants’ responses and reflections in relation to specific themes introduced by the researcher. This article proposes the conversation as a source of public opinion, while scholars tend to use respondent surveys. I also developed in-depth interviews to understand the reasons behind the dominant topics they raised. This article will present the findings in five stages. First, it will outline the voters’ statements. Second, it will examine the reasons behind these statements. Third, it will discuss the results. Finally, it will provide a conclusion and address the limitations.

Results

Dominant Topics in Talk

After I started the conversation with “How is/was the election?” research participants led a discussion that varied. The conversations can be categorized into two groups: election participants and election winners. Both of these groups mentioned parties more frequently than candidates. I examined potential variations attributable to demographic differences in respondents; however, no sufficiently distinct patterns emerged to warrant categorization. Among election participant conversations, statements about parties were made more frequently than those about candidates. They exhibited a strong recognition of these parties, recognizing them consistently across different elections. For instance, one participant remarked, “There used to be numerous parties in the elections.” He referred to an election where 48 political parties participated. Participants who lived through the New Order era elections vividly recalled the presence of three parties during that time. Statements about candidates did occur, but they were infrequent. One example of these rare statements is, “I have a neighbor who was a candidate in the previous election.”

In the election winners conversations, statements about parties were also made more frequently than those about candidates. Some research participants vividly recall which parties emerged triumphant in each election. For example, one participant stated, “In the 1999 election, PDI-P emerged as the winner despite being a new party.” This statement underscores the participant’s clear understanding of that party’s dominance in the 1999 election. Another similar statement is, “Golkar remains a formidable force, continuing to be one of the major parties even after Soeharto’s resignation.” This remark highlights Golkar’s resilience in legislative elections despite public criticism of former President Soeharto. It is worth noting that Golkar’s reputation remains closely intertwined with Soeharto.^[4] Statements about candidates do arise, but they are infrequent. One such rare example is, “Mr. Bambang lost in the 2019 election, despite winning in the 2009 and 2014 elections.”

I attempted to steer the discussion toward the candidates. The conversations were dominated by two groups: candidate’s affiliation and candidate’s consistency. In the candidate’s affiliation conversations, the common statement that arises from participants is, for example, “Which party is he/she from?” Additionally, responses often include statements like, “He/she is from the Golkar Party,” or “He/she is from the PKB party.” In the candidate’s consistency conversations, participants delve deeper into evaluating the candidate’s allegiance to their party. Consistency refers to whether the candidate has remained loyal to a single party or not. For example, a participant stated. “He moved from the Gerindra party to the Demokrat party”. Statements about a candidate’s personal profile are very rare. One such rare example is, “He once campaigned here.”

Reasons Behind the Talks

Based on in-depth interviews, I identified three reasons why parties are a more dominant topic of conversation than candidates: news coverage, the performance of legislators, and candidate recognition. First, the media’s political discourse focused on political parties. Parties were discussed more than specific legislators or candidates, particularly during discussions concerning electoral dynamics. The majority of participants acknowledge the media’s coverage of the legislative elections. When I prompted discussions about the 2024 elections, most participants referred to party narratives. They often centered around the parties’ efforts to promote the 2024 elections. It is common in the media for statements regarding the 2024 elections to originate from political parties. Party coverage extends beyond elections to include offices targeted by elections, such as the National House of Representatives. Media outlets frequently mention political parties when reporting on policies or dynamics within the house. Despite referencing factions, party names are still associated with these terms. When I prompted participants to discuss a policy debate, participants cited media coverage and acknowledged party involvement. Voters tend to perceive dynamics within the house as primarily driven by parties. This information plays a significant role in shaping the view of the legislative elections among voters.

Apart from featuring parties in news coverage, the media often presents politicians with reference to their party affiliations. When politicians comment on election matters, the media typically links their party to their name. It attributes their statements to their party’s position. This association of party attributes with politicians extends beyond election-related issues to other subjects followed by legislators. For example, when a legislator appears as a guest on a television talk show, the host frequently mentions their party affiliation. While this media practice may seem straightforward, it can shape voters’ views. It prompts them to link parties with their thoughts while watching such programs. Several informants made statements such as, “Well, I know there is information about which party they are from.” This response came when I asked about a legislator’s affiliation, frequently seen on television. In fact, the participant solely identifies the figure based on their party affiliation. Other participants stated similar comments, citing media reports on legislator factions. Thus, this media coverage embeds party-related information in the minds of participants.

Second, the performance of legislators tends to be out of touch with voters, thereby weakening voters’ identification with their representatives. Several factors contribute to this disconnection. When I prompted participants to discuss legislators from their electoral districts, many participants mentioned not being familiar with these names. One participant stated, “I have just heard those names. Are they from this area? I have never seen them.” This response indicates that the participant has never had direct interaction with that member. A similar statement: “I have heard their name before, but they have never come down to this area.” In other words, voters lack interaction with the legislators or candidates in the elections. It makes it more difficult for voters to recall individual candidate information than recall their party affiliation.

The participants also perceive themselves as not experiencing the performance of legislators. When asked which institution’s performance they feel, the majority of participants mention other institutions. Upon further inquiry about the performance of the legislator, some participants stated that they felt no impact. This sentiment was evident when one participant mentioned the absence of legislative programs in their area. They understand that the performance of legislators relates to national performance. Participant statements reflect this understanding with phrases like, “They work in the capital,” or “Their responsibility is in Jakarta.”

Third, voters were not familiar with the candidates. During the national legislative elections, voters often lack adequate information about the candidates. Information about the candidates’ backgrounds or the performance of incumbent candidates is not well-known to voters. When I attempted to present the name of a candidate, participants did not provide recognition responses. Participants responded with questions like, “Who is that?” This signifies their lack of knowledge. The participants’ best response was to mention the party affiliated with the candidate. The candidate’s hometown also contributed to voters’ lack of familiarity. Typically, the candidates in the House elections were not from the area where the voters’ resided. In fact, candidates did not originate from the electoral district nor reside in it. They merely use the electoral district for candidacy. A participant articulated this situation. He stated, “If they are not from here, then I do not know them.” Another participant remarked, “It seems like there is no one from this area running for the elections.” Residence serves as a strong reference point for recognition to candidates.

I traced the social connections between candidates and participants who talked extensively about candidates. This pattern was observed among twenty-two participants. The topic of candidates dominated these conversation due to personal connections between the participants and the candidates. For example, one participant mentioned having a close friendship with the candidate’s spouse. Such personal connections were rarely found among participants who focused more on parties.

Discussion

Based on the dominance of topics in the conversation and in-depth interview data, participants placed the party as the “kickoff” of the conversation and the “recall” in the middle. This suggests that elections are more frequently interpreted as a party contest rather than a candidate contestation. Participant demographics did not account for any observable variation. Neither gender nor age, in either group, produced explanatory differences. This suggests that prior experience under two distinct electoral systems does not translate into differing interpretations of the meaning of elections—whether understood as a contest between parties or as a contest between candidates. Hypothetically, an open electoral system is expected to produce a candidate-centered election (Manin et al., 1999), a concept supported by some researches in Indonesia (e.g., Aryanto, 2023; Rafni & Suryanef, 2004; Suryadi, 2005; Tawakkal, 2009), and one that should be affirmed by voters as well. However, this study found the opposite. Voters’ views on the elections challenged the theory.

This challenge highlights the need for further theoretical development regarding open electoral systems. Despite nearly 20 years of implementation in Indonesia, the open electoral system’s design, which aims to promote candidate-centered elections, faces challenges influenced by the information voters receive. Mass media coverage, candidate performance, and candidate recognition have all contributed to voters maintaining the elections as arenas for inter-party contestation. This suggests that the electoral structure provided by the system alone is insufficient in shaping voters’ views.

This finding also addresses the gap in explaining why scholars have arrived at different conclusions about party identification in open electoral systems. In previous studies that confirm the theory, the meaning of elections for voters might be overlooked as a variable. On the other hand, other scholars have observed the strengthening of parties in elections (e.g., Apriani, 2019; Haryanto, 2014), noting repeated wins by certain parties in specific districts, and difficulties for incumbents in maintaining their seats, seem to contradict the theory of an open electoral system. Voters’ views on elections, in this case, cannot be disregarded. Voters’ views, which have been studied by scholars in relation to participation and social divisions, are further explored in this study, showing their influence on the dynamics of an open electoral system. Voters’ views not only affect participation and social divisions but also shape how parties are positioned within elections. The victories of challengers and the defeats of incumbents from the same party underscore how significant voters’ views are in shaping electoral outcomes.

Conclusion

This article argues that the meaning of elections for voters plays a crucial role in shaping the impact of the electoral system, which, in this case, can lead to distortion. Voters’ views help explain why an open electoral system, which theoretically leads to candidate dominance, presents anomalies. The gap in findings from previous party identification research is likely due to the insufficient consideration of voters’ view. The image-building tensions between parties and candidates (as party ambassadors) discussed in the literature (Marland & Wagner, 2020; Speed et al., 2015) are not confined to their dilemmas. I argue that the use of a particular electoral system as a means of addressing these dilemmas, in fact, generates new challenges at the voter level. While scholars have mentioned factors that cause the existence of parties (e.g., Aminuddin & Ramadlan, 2022; Andrews-Lee, 2019; P. G. Chen & Goren, 2016; R. Chen, 2021; Febres-Gagné, 2022; Ferrari, 2022; Grant et al., 2010; Lewis-Beck & Nadeau, 2014; Primo, 2013), this article offers the view of voters as another factor.

Indonesia’s goal of strengthening candidates encounters challenges at the voter level, particularly with candidate recognition. The findings suggest that incumbents should enhance their engagement with their constituents through active involvement with their constituents, while challengers should focus on building their social profiles. Additionally, the findings indicate that candidates should reside in their constituencies to increase visibility and recognition. In the context of party democracy, all of these efforts must be carried out within the party framework, despite the challenges involved.

While this article emphasizes the meaning of elections and helps explain the repetition of party identification, party victories, and expectations for party representatives in an open election system, it does not address the consequences of this meaning on the dynamics of post-election representation—such as how it affects voters’ assessments of benefits, post-election outcomes, or interactions with other factors like money in politics. Also, this article does not examine informants’ backgrounds, such as their party affiliations. However, it offers a potential avenue for future research to explore whether strong party ties help explain discrepancies between electoral systems. Further research is needed, but this study offers important insights. It highlights the need for democracy advocates, including governments, to develop policies that better align with voters’ expectations and understanding of the electoral process.