Voter Behaviour and The Campaign Pattern of Candidates during Pandemics in Regional Head Election in Medan City, North Sumatra

The Abstract: Covid-19 has influenced the process of democracy in many countries, including Indonesia, especially in the local election. The political campaign involving mass mobilisation was restricted, so the candidates were encouraged to use online media to reduce the Covid-19 spread. Based on this problem, this study aims to analyse the campaign patterns between candidates in a regional head election in Medan during a pandemic and how the candidates increase voter turnout. This study used both the qualitative method by interviewing the campaign team of the candidates in charge of online and offline media and the quantitative method by distributing the questionnaires to measure the behaviour of the voters. This study showed that the perception of


Introduction
ne of the essential aspects of democracy is elections because elections can determine the quality of democracy. Elections are a part of a democratic process aiming to recruit politicians. This process starts with nominating candidates from political parties to the selection process in a general election (Heywood, 2013, p. 205). The nomination process in elections will also involve a campaign period given to candidates to demonstrate their abilities, leadership skills O Abstract: Covid-19 has influenced the process of democracy in many countries, including Indonesia, especially in the local election. The political campaign involving mass mobilisation was restricted, so the candidates were encouraged to use online media to reduce the Covid-19 spread. Based on this problem, this study aims to analyse the campaign patterns between candidates in a regional head election in Medan during a pandemic and how the candidates increase voter turnout. This study used both the qualitative method by interviewing the campaign team of the candidates in charge of online and offline media and the quantitative method by distributing the questionnaires to measure the behaviour of the voters. This study showed that the perception of popularity and electability influenced the pattern of candidate campaigns, and the low political awareness of the community influenced low participation in political campaigns and voter turnout, besides unhealthy campaigns and negative perception of local government. Therefore, although face-to-face was restricted, the candidates believed that direct meetings with the voters were only the way to influence the voter and increase political participation.
campaign strategy of the candidates in which the strategy can bring influence the escalation of voter participation and what campaign form the candidates use during the increasing case of Covid-19. The candidates' campaigns would differ between the campaign during the pandemic and the one in normal conditions. It is related to direct interaction restrictions between the candidates and the voters. But the development of technology information is seen to be capable of coping with the problems.
We would like to use hybrid media to analyse the model of the candidate campaign. The existing literature shows that online media positively affect candidates in the election. Candidates using online media such as Twitter received more preferential votes than candidates who did not use it (Kruikemeier, 2014;Lee and Shin, 2012;Spierings and Jacobs, 2014). Even though online media has become an effective tool, especially during pandemics, we found that in the case of pilkada in Medan, online media did not become the main focus. The main factor influencing candidates and their teams is the perception of online and face-to-face campaigns' impact on popularity and electability.
To analyse the pattern of candidates' campaign in Medan, this paper will be organised as follows: first, this chapter will discuss the online media and integration of modern and traditional media to describe the pattern of the modern campaign. The next is a research method that describes how the data was collected and analysed. Then, this paper will delineate the result of this research, and the last is the conclusion.

New Media and Campaign Strategy: A Review
Hybrid media combines old and new media as a communication tool; printed press and television have played an essential role in the traditional campaign model. Several studies showed that communication through conventional media could shape the attitude and behaviour of the voters (Graber, 2010;Holbrook, 1996;Patrick, 1989;West, 1994). After all, the evolution of media from traditional to modern, as seen from the emergence of the internet, has changed the pattern of communication between the candidates and the voters in which communication is no longer done by one-way communication but it can be done through two-way communication (Stromer-Galley, 2004, 2000Sundar et al., 2003).
In fact, the emergence of new media, particularly social media, does not reduce the importance of traditional media; as stated by Chadwick (2017, p. 131), "the campaign's significance lies not in its internet campaigning, but in how it so ruthlessly integrated online and offline communication, grassroots activism and elite control, older and newer media logics." Chadwick et al. (2018, p. 11) believed that integrated media could determine the candidates' success in the election.
The integrated media in the campaign can be seen from traditional media and the internet. For instance, though television only provides one-way interaction, the candidates still need television coverage to show their capabilities, power and personal characters (Chadwick et al., 2018). Meanwhile, the internet is not used to substitute the roles of television or newspapers but to support the traditional media. It can be used to promote the candidates, collect the voters' data, manage the volunteers, do fundraising, and build interaction with the voters (Chadwick, 2017).
The internet as an interactive communication device is a key variable in the online campaign (Sundar et al., 2003, p. 32). The interactive communication here is seen not only as a face-to-face conversation but also as an interaction through technology used to build a two-way communication (Stromer-Galley, 2004) to enable the public to hear the messages directly delivered by the candidates without any distance (Coleman, 2005, p. 9), reach the voter more widely (Jansen, 2004), control the information flow (Towner and Dulio, 2011, p. 167), define their political identity (Jensen, 2018), and encourage and motivate people to search the information about the candidates (Kaye and Johnson, 2002). Even some scholars see that the internet can also be used to predict the results of the election (Bermingham and Smeaton, 2012;Sang and Bos, 2012;Tsakalidis et al., 2015).
Another important aspect of the hybrid campaign is the target audience. The candidates or campaign team cannot focus themselves only on the internet. Although the internet can reach a wider mass, they also need to focus on certain audience groups such as ages and gender. Chadwick et al. (2018) exemplified the election in America, where Obama made some press advertisements on the web and YouTube to target different online demographic groups. This strategy could encourage young voters to support Obama, making him able to reach 66 per cent of the voters aged in the range of 18-29 years (Chadwick, 2017, p. 156).
Then, Cable TV is useful for influencing specific audiences through political advertisements in certain programs such as TV programs about lifestyle, food, or other preferred programs by certain groups. The data about the audiences can be bought through market research companies or cable television (Chadwick, 2017, p. 138). Furthermore, using Google's algorithms and ads becomean effective campaign that can be targeted in some areas. Google's algorithms can be used to get volunteers in certain areas. Meanwhile, Google advertisements can be synchronised into the main political programs, such as political debates (Chadwick, 2017, p. 139).
Furthermore, a hybrid media campaign can be an effective strategy in two aspects: first, it is used to capture input from the public. In this case, online media provides direct access to candidates to capture people's aspirations, where people can respond and give comments, suggestions, or criticisms in online media. Second, a hybrid campaign can mobilise people to engage in campaigns. The candidates can encourage the public to be involved in the campaigns through online activities and inspire the public to take action in traditional campaigns, such as donations, request support and facilitate offline campaigns, such as face-to-face campaigns (Chadwick et al., 2018 p. 11).
Even though hybrid media positively affect the candidate's success, we see that this strategy tends to be effective in the national election involving all segments of society. In the local elections, the hybrid media can be effective if society's awareness of politics is high, which can be seen from high political participation. Low awareness does not influence citizens to seek information about the candidate and their campaign online.
This research used interviews and survey data. The interview was conducted with five informants coming from the campaign team. Here, the informants were in charge of managing the campaign division either in printed media or online media. The interview was conducted to observe how the candidates influence and encourage the voters to participate in the election.
This research used a semi-structured interview by making a list of main questions relevant to the perception of the campaign team and the candidates towards the issues of the campaign during the pandemic, the use of social media and traditional media, political advertisement, target audience and the interaction pattern between the candidates or campaign team and the voters in social media. The results of this interview could describe the achievement of the candidates in the electionparticularly in this pandemic.
To analyse the community's behaviour, we surveyed 21 to 26 September 2020. The survey was conducted in Medan City with respondents who already had the right to vote. One thousand two hundred respondents were interviewed. This survey data was used to support the interview data and measure to what extent the candidates' strategies and issues could attract the voters. Also, this survey could describe the community's situation encouraging the candidates to do the face-to-face campaign.

The Pattern of Candidates Campaigns in Social Media
Even though the two campaign teams had almost the same understanding of the benefits and effects of social media, we found that the campaign teams did not integrate modern and old media in their campaigns, and they tended to focus on the traditional campaign in influencing the voters especially in encouraging the voter to participate in the election. The low level of community participation in Medan in two previous elections made it difficult for the campaign teams to invite the voters to use their rights to vote on Election Day.
One of the factors influencing the candidates' team is the difference in concept between popularity and electability. The campaign team of candidates see that online media is to increase the popularity of candidates. In contrast, online and social media cannot influence the electability of candidates and citizen participation in politics, but it can only be influenced by the face-to-face campaign. These two concepts contradict the findings of previous scholars who found that the use of social media can increase the electability of candidates and encourage people to participate in politics, engage in campaigns and initiate them to make campaigns (Ahmad et al., 2019;Chan, 2016;Chan et al., 2017;Chen, 2018;Gibson, 2015;Kruikemeier, 2014;Lee and Shin, 2012;Spierings and Jacobs, 2014).
Even though the two campaign teams had similar perceptions, the pattern of their campaign was different from each other in online media. The difference in pattern campaign will be seen in four patterns, i.e., the pattern of communication, social and online media selection, target audience and professional campaigner. First, this difference in pattern campaigns in online media is caused by the difference in popularity between Akhyar Nasution as the incumbent and Bobby as the new contender. Akhyar did not maximise the online and social media because he had popularity in the Medan, so the team campaign focused on increasing Akhyar electability and motivating their supporter to vote in the election time. Otherwise, Bobby and the campaign team used online and social media to increase the Bobby's popularity. The informant Y1 stated that based on their survey team, Bobby's popularity level was still low before the candidacy started. This condition made the campaign team frequently introduce the candidate through online media.
The different popularity also affected the communication pattern between Bobby and Akhyar in using online and social media. Effective communication in online media that could influence the voters is interactive communication. This pattern is considered capable of attracting the attention of voters, encouraging voters to participate in politics and empowering the voters in political campaigns (Aaker and Chang, 2010;Jensen, 2017;Opeibi, 2019). In developing communication with the voters, Bobby Nasution and the campaign team build two-way communication with the community. Here, they not only provide information about the candidate's activities or disseminate the candidate's programs but also create a special team to reply to the comments or answer the voters' questions on Facebook, Instagram and Twitter (Infotmant Y2, 2001).
Also, this team played an essential role in studying and analysing the information to be provided on social media and evaluating the contents of the information they delivered every week. It was because something they supplied on social media would be adapted to developments of the current issue. That is why the issue did not have to contain the candidate programs and the campaign activities, but it could be about humorous information that could attract voters' attention on social media. In addition, to provide information about candidates, the campaign team also used the issues to attack other candidates to strengthen the confidence of voters who were still unsure about their candidate. However, these issues were adapted to the latest campaign developments.
In contrast to Akhyar's team, where candidates even had a campaign team, including those in charge of managing the campaigns on social media and printed media, they did not have a special team to answer comments or provide feedback on social media. Informant Y4 stated that: We give volunteers the freedom to inform about candidate activities or disseminate candidate programs and give volunteers the freedom to answer and respond to comments on social media. (Informant Y4) The pattern adopted by Akhyar's campaign team aimed to create free space for volunteers to use social media to build communication in many directions. However, the problem in Akhyar's campaign model is that the message or content and language selection that the volunteer did could not be controlled well by the campaign team. It was unknown whether the language used by the volunteers in building communication was able to touch the voters.
Second, communication in social media and online media selection become essential aspects for candidates and teams, although many social media can be used as campaign tools, such as Twitter, Facebook and Instagram. However, both teams are more focused on using Facebook and Instagram because Facebook and Instagram are the two most used social media besides Whatsapp. Our surveys showed that the social media actively used by the people in Medan was Whatsapp (57.67%), followed by Facebook (48.75%) and Instagram (29.08%). Meanwhile, Twitter (6%) was not widely used by voters.
Of the two social media, Bobby had ten social media accounts managed by the campaign team and two social media directly controlled by the candidate. Meanwhile, Akhyar did not manage social media accounts submitted directly to the campaign team, but they empowered the volunteers to manage these accounts. Besides, the campaign team of Bobby also made news in electronic media to inform the candidates' campaign activities and disseminate the work programs to be carried out by candidates. Electronic newspapers were intended to reach market segments untouched by social media. Although social media has been widely used, according to the campaign team, not everyone likes to see information on social media, which is often considered hoax news.
To strengthen the spread of information, Bobby's campaign team also conducted research keywords on Google aimed to display news about Bobby Nasution in the top position on the Google search engine and to make acceleration enabling the information to be received directly by people who want to know more about the election in Medan. On the other hand, the strengthening of information on Google was not carried out by Akhyar, viewing that searching on Google had little impact on voters. Even though the impact was not so significant, Bobby's candidate team Informant Y3 continued to use it because : …there were certain groups of people who really wanted to know more about the competing candidates, and they were the educated people who wanted to know more about the candidates and their programs in more detail (Informant Y3).
In addition, to use social media and search engines, Bobby Nasution created two websites for voters who wanted to know more about candidates and the programs to be made when they win the election. A special team managed these websites to design and update information in the media. Meanwhile, the team of Akhyar did not create a particular website regarding candidate profiles.
Third, to strengthen the influence of online media campaigns, Bobby involved a professional consultant to design the issues that would be delivered on social media. Hence, the political campaign of Bobby involved two campaign teams: an internal campaign team established directly by the candidates and a team from a political consultant (Informant Y3 : 2021). In contrast, Akhyar did not use a political consultant as they highly relied on the competence of his team and the volunteers in doing the campaign. Two factors determined this; first, the cost of the political consultant was high, and the candidates and the supporters did not have sufficient funds to pay for it, and second, the candidates believed in the political consultant less and believed their friends and networks more. Those two factors were also influenced by the competitor that had strong finance and was supported by the power of the main parties in the coalition of the central government (informant Y4 and Y5, 2021). For this reason, it was considered that paying the political consultant was seen unable to cope with the problems they were facing (Informant Y4, 2021).
The consultant's involvement in Bobby's campaign team, as stated by his team, was in consideration that Bobby Nasution was a newcomer that was not widely known by the public. Therefore, the political consultant was deemed essential to analyse the candidate's needs to be known by all segments of society that the campaign team might not see as a whole. Hence, the involvement of the political consultant could also be a media for comparison in the campaign that was done by the internal team and professional team to cover any weaknesses of the internal campaign team.
Fourth, in online and offline media campaigns, the target audience becomes an essential aspect that the candidates or campaign teams must manage (Chadwick, 2017(Chadwick, , 2007Chadwick et al., 2018). The target audience in media is different from the offline campaign. The target audience in the media is aimed at special programs in both traditional and modern media such as sports programs on television or online media, cooking programs or soap operas (Chadwick, 2017). One of the modern campaign methods can be seen in the American presidential election, where the presidential candidate, Barack Obama, was able to come out on TV or cable TV programs. Here, Obama inserted advertisements designed according to the broadcasted programs. The result of this showed an increase in voter turnout in certain circles (Chadwick, 2017;Chadwick et al., 2018).
However, in Medan, two candidates did not have a target audience in the online media, and the specific segment of media did not become their attention. From the interview result, we conclude that the focus of the campaign teams on specific audiences, such as based on age, gender, or education level, tended to be seen based upon the mapping method in the media in which the campaign team assumed that the users of social media tended to be young people making campaigns on social media useful for attracting the young voters. Meanwhile, campaigns through printed media, such as reports in newspapers, were aimed at older voters, and campaigns through electronic media, such as online news, were intended to attract voters with a higher education level.
Interview results also showed that there were several problems faced by candidates not trying to use the advertisements on television or cable TV. First, KPU regulation restricted the advertisements made directly by candidates, where KPU managed advertisements and pamphlets. At the same time, the campaign team was only in charge of designing the advertisement or pamphlet. Second, the number of television and printed media in Medan was very limited, and third, the number of cable TV users is not high, and it is only owned by the middle to the upper class. From the above factors, the campaign teams did not focus on the use of television media, whether conventional television or pay-tv, in view of the relatively high cost and insignificant effect on the candidates.
Even though many tools in online media, including social media had been used by candidates and their teams in their campaign, the team campaign did not use online or social media to empower the audience in the campaign and encourage them to participate in politics and also did not integrate the traditional and modern media as an effective tool to influence the voter' participation. The candidates tended to use media separately, and some of the tools available on social media were not used. Say, the live feature on Facebook was not used by candidates, nor was candidate debate and interviews on television broadcast live on YouTube and Facebook. The lack of focus of candidates in using social media was caused by the understanding of the concept of popularity and electability. The campaign team emphasised that social media has a function to increase the popularity of candidates, but it cannot affect the electability level of the candidate (Informant Y5 and Informat Y3 :2021).
Popularity refers to the level of fame of the candidates in the community, including their political programs, characteristics, personality, or background. Meanwhile, electability refers to the level of electability of candidates in elections. Informant Y5 argued that : "There is no guarantee that the high popularity of the candidate in society will make him to have high electability" (Informant Y5). Similarly, the high electability of the candidate is not a guarantee for high popularity. Therefore, in the election, they still believe that dependence on social media does not affect the success of candidates.
The campaign team also distinguished the use of online and offline media, where online media was only a tool to introduce candidates to the public. In contrast, offline media such as television, newspapers and radio emphasise more on verifying information spread on social media so that the public can be more confident about the information on social media. This is because the level of public trust in conventional media is higher than the information available on social media. In addition, campaigns on social media are difficult to influence voters because social media can only show a few dimensions without fully showing the candidates. Therefore, the electability of candidates can only be obtained through their presence in the community, even though the pandemic has limited the space for them to meet with voters.
Therefore, we see that perceptions of popularity and electability are the main factors that cause candidates to use modern media with traditional campaign styles, i.e., social media is used only to broadcast candidate profiles and not to invite the public to be involved in political activities.

The Offline Campaign in Pandemic
Though campaigns on social media can influence voters, candidates' focus remained on face-to-face campaigns. Because the candidates viewed that such campaigns could influence the voters more than campaigns through online media. Campaigns on social media here became complementary to face-to-face campaigns, and they did not significantly affect the electability of candidates. Besides electability, one of the factors encouraging the candidates to focus on the face-to-face campaign, according to the informants Y3 and Y5, was the high low level of political participation in Medan in every mayoral election. Consequently, campaigns through online media were unable to change the voters' attitudes. It can be assumed that the more frequent the candidates meet the target voters, the greater the chance for the candidate to be known and accepted by the public. For this reason, there was a need for the two candidates to do the face-to-face campaigns with target voters still, though with different strategies.
The target audience in the offline campaign in both teams has a similar perception. First, they focus on areas with a large population to achieve the largest number of votes. Second, candidates focus on areas as their strong bases of support with an aim that voters remain loyal to support the candidate and not run for another candidate; third, they focus on areas where voters still tend to be volatile in providing support to the certain candidate based on the results of their internal surveys and finally, focus on areas with fewer voters. The target audience in both teams is a general strategy that candidates almost use both in the local and national election.
However, from our interview with informants, we also found different target audiences in both candidates' campaigns. The Akhar and Salam tended to focus on religious and ethnic groups, even though other groups also became their concern. Salman Alfarisi, as an Ulama, focused on filling out recitations in several places for Islamic groups. Meanwhile, Akhyar focused on visiting ethnic groups, especially non-Muslims. We see that the cleavage issues between Islam and secular nationalism in the Jakarta governor and presidential election influenced the pattern of their campaign. On one side, Salma as ulama wanted to strengthen the Islamic voters, while Akhyar visited ethnic and visited non-Muslims and wanted to eliminate cleavage issues between Islam and non-Islam or secularism. In contrast, Bobby and Aulia tended to focus on class and social organisations such as labour groups, traders or youth communities, relying on the popularity of Bobby as the president's son-in-law.
Besides, the change of duration in the face-to-face campaign during the pandemic has become an essential factor in candidates' campaigns. Face-to-face meetings with a short duration with a higher number of meeting points become focus candidates. Based on the interviews with the campaign teams, there were several patterns of face-to-face campaigns, such as the direct visit of the candidates to the voter without a need to gather the masses. Here, the candidate did some activities or meetings with a limited number of communities and then moved to other communities. For this, the candidates would try to visit more voters or do activities 10 to 12 times in one day on average. Our interview results showed that the direct visit of the candidates in public was due to not only the spread of the Coronavirus but both candidates persuade their encouraged their supporters and community to attend and vote at election time because the candidates and their team realised that the voter behaviour in Medan had different characteristics with other voters in other regions.

The Voter Behaviour in Medan
The candidates' campaign team used online and offline campaigns to increase candidates' popularity and electability and encourage the voters to participate in political campaigns and vote at election time. However, the impact of the political campaigns of candidates did not affect the community to participate in the political campaign.
The low voter turnout in Medan is in line with the low involvement of the community to be engaged in campaign activities. Our survey in Medan showed that most respondents did not want to join in a campaign team to help their candidates win in the election or to be involved in a political campaign such as attending a campaign. Of 1200 respondents interviewed, only 39.70 per cent were willing to attend a candidate's campaign, and 42.63 per cent were not willing to come. Similarly, when asked whether respondents were willing to be on a candidate's campaign team, most of them (73.41%) were not willing to be on a candidate's campaign team -only 26.59 per cent were willing.

Chart 1. The Willingness to Participate in the Campaign and Become the Campaign Team in Percentage
The low participation of the community in politics also impacted the public's unawareness of the programs offered by the candidates. Our surveys showed that 94.58 per cent of respondents did not know the Bobby's political program that he offered to society, and 95.92 per cent of respondents stated they did not know the programs presented by Akhyar. This result showed that three factors influence public ignorance of the candidate's political program. First, the political campaign of candidates in online and offline media focused on the candidates' popularity by showing the characteristics of candidates rather than the political program. Second, the candidate did not have a clear political manifesto that could be given to voters, and the community saw that the political program of the candidates could not have any effect on the community, and the candidate program did not meet community expectations. As a result, people have assumed that whoever the elected candidate is will not significantly impact their lives. Our survey showed that the majority of the community (52.51%) stated that the programs promoted by the candidates did not meet their expectations; only 6.28 per cent stated that the programs met their expectations.
Third, unhealthy political campaigns become the main problem that influences voter behaviour. The unhealthy campaign was the main problem that occurred in many political campaigns in Indonesia, including in Medan. An unhealthy political campaign can be seen from the prolonged practice of money politics in elections. During the pandemic, the community's increasing needs also made the community ask the candidates to offer a number of activities that were not only in the form of program delivery but also could meet their living needs. The public's demand for these valuable activities made the occurrence of highmoney politics during the campaign time. 42.58 per cent of respondents stated that they received money from candidates, while 57.42 per cent did not answer this. The amount of money given by the team to respondents varied; 21.08 per cent received more than IDR.200.000; 11.25 per cent earned approximately IDR.200.000; 8.83 per cent received IDR.100.000, and 1.42 per cent received IDR.50.000. From the survey we conducted, the community expected the candidates to provide basic food to people affected by . In comparison, the need for masks and hand sanitisers was not becoming a priority for the community (0.08%).
The reality among the people in Medan has shown that the pattern of the candidate's campaign during the campaign period through social media and face-toface principally is challenging to influence the public to vote for candidates during the pandemic. The increasing needs of the community during the pandemic have triggered the high money politics since people truly expected to get direct assistance from candidates, thus encouraging the candidates and campaign teams to play these politics as a form of exchanging money with votes.

Conclusion
Online and social media has become influential in modern campaigns, but in the local election, candidates and campaign team use these online media like traditional media such as TV and newspapers. The candidates only used online and social media to broadcast their views, and their backgrounds and online and traditional media were not integrated. The perception of popularity and electability is a factor that influences the pattern of the political campaign in online and social media. Hence, candidates and their team assume that online and social media is only used to increase popularity and not to increase electability. Even though existing literature has found that the candidate using social media as a tool to develop reciprocal conversation would have more preferential vote than candidates who did not use social media. The perception of popularity and electability caused candidates to focus on the offline campaign, even though regulations restricted the candidate from doing face-to-face campaigns. However, in the local election, we realised that on one side, the modern campaign could be effective only in some cases, especially at the national level, and the community have high technology literacy. However, on another side, it is not effective in the minor election such as in Medan because the candidates have the opportunity to meet the voter directly, and the use of modern technology is low, so the use of online and traditional media as campaigns tools needs high cost and disproportionate to the small level of the electoral district.
Besides, the low political awareness in Medan in several elections has encouraged candidates to use traditional rather than modern campaigns. The perception of the community who considered that the candidates' program did not meet their expectations and did not have any effect on the community became an obstacle to the candidates to convince the voters because in a face-to-face campaign with direct meetings with the candidates, the public still cannot be convinced by the candidates.
Therefore, the use of online media and integrated media between modern and traditional media can be done if the community has high political awareness and high technology literacy, and the elected candidate could change the perception of the community to their local government by implementing the political program that could meet the community expectations.