Much has been made of the findings from the symposium on media and Internet use during the 2015 general elections – organised by the Institute of Policy Studies on January 27, 2016 – including the relationship between social media sharing and knowledge about the elections, demographics of the swing voters, as well as the influence of open or closed social media groups. But what I thought was perhaps more interesting were questions raised by the five presentations, with opportunities for future research.
1. Who attends campaign rallies, and how do rallies influence decisions? Those who lower intensity of social media use were more likely to attend campaign rallies than those with average or higher intensity of use, according to Dr. Natalie Pang of Nanyang Technological University (NTU). This may appear counterintuitive, with expectations that political participation would be consistent online and offline. I asked Dr. Pang whether other control variables – such as age and education were considered – and from the sample of 2,000 respondents rally attendees were often older and more educated. Further profiling and comparison of these rally attendees could be useful.
Rally attendance was also curious for Dr. Zhang Weiyu of the National University of Singapore (NUS), who found that although swing voters – defined as those who voted for one party in 2011 and another in 2015 – participated in rallies more, these participants tended to participate in rallies held by the party from which they swung. For instance, 84 per cent of swing voters who voted for the Opposition in 2015 (and therefore, for the ruling party in 2011) had attended a rally organised by the People’s Action Party, while only 21 per cent had attended a Workers’ Party rally. More information about the reasons for this difference could provide insights into the psyches of voters.
2. What is the threshold for the use of personalised communication, and how might it improve the campaign strategies of political parties? High level of social media use did not necessarily enable voters to gain knowledge on more complex election issues. In other words, according to Dr. Debbie Goh of NTU, excessive use by some socio-economic and demographic groups – Indian, had secondary school education or less, earned a high income or low income, based on the data – had negative influence on political knowledge.
The question which follows is the threshold. That is, what is the ideal frequency of personalised communication, after which social media use would be counterproductive? Or does the quality of usage, beyond frequency, matter? And if so, to what extent?
3. How can researchers distinguish between open and closed social media groups? An interesting finding by Dr. Elmie Nekmat was the open social media platforms – in particular, Facebook, Twitter, and Instagram – had no influence on voting patterns, and that personal opinions, vis-à-vis societal opinions or both open and closed social media groups, was the greatest determinant of voting decisions. My query to Dr. Elmie was a more technical one: that the distinction between an open and closed social media group is not a straighforward, and that a secret Facebook group or a private Twitter and Instagram account could be regarded as a closed, not open, social media group. In this vein, finer distinctions in the survey choices could be helpful for future research.
4. Why are respondents hesitant to reveal voting decisions? Should more be done to assuage fears over voting secrecy, for instance? Of the 2,000 respondents, less than 50 per cent – just 943 – were willing to reveal their voting decision. The implication was that Dr. Zhang had a small sample to work with, since only 122 of these 943 respondents were swing voters. Survey design and limitations notwithstanding, it would be meaningful to understand why respondents were unwilling to share the party they voted for, and whether unfounded fears over voting secrecy remain a problem in Singapore.