Will the 2019 elections see fake quick counts again?
itoday - Five years ago, Indonesia experienced the most polarised election since the introduction of direct presidential elections in 2004. After an often hostile campaign characterised by aggressive rhetoric, smear campaigns and personal denunciations, tensions culminated on election day when both candidates, Joko Widodo (Jokowi) and Prabowo Subianto, claimed victory.
The General Election Commission (Komisi Pemilihan Umum, KPU), as well as the Constitutional Court, later confirmed Jokowi as the winner, but for several days after the election, Indonesians had to endure uncertainty as both camps were adamant that they had won.
The 2014 precedent
At the heart of the controversy lay divergent quick count results presented by different polling institutes on live television shortly after voting had closed. While Indonesia’s most highly respected pollsters had all declared Jokowi the winner, four other institutes brazenly used manipulated results to claim that Prabowo had won.
The episode cast a shadow on Indonesia’s polling industry as it exposed not only the potential for political manipulation of quick counts and surveys but also the deep personal and political divisions between pollsters.
Five years on, these issues remain largely unresolved and the industry continues to be plagued by fragmentation and politicisation. Thus, to prevent another quick count debacle in the 2019 election, the KPU has tightened regulations that require survey institutes to register formally with the commission if they wish to publish a quick count on 17 April.
So far, a total of 33 polling organisations has registered with the KPU (link is external), including two of the four institutes at the heart of the 2014 scandal (Jaringan Suara Indonesia and Puskaptis).
The KPU is yet to confirm how many of those who registered will eventually be allowed to conduct their quick count, but if the commission applies its own new regulations strictly, then several pollsters could face disqualification.
At least one organization, for example, has openly admitted (link is external) that it does not fulfil the KPU requirement of membership in a professional association such as the Indonesian Public Opinion Survey Association (Persepi) or the Indonesian Association for Public Opinion Research (Aropi).
Others may be reluctant to reveal details about their methodology or their funding, another requirement stipulated in Article 28 of KPU Regulation 10 of 2018. However, even if the KPU decides to give the green light to all the registered pollsters – including those that do not comply with existing regulations – a repeat of Prabowo’s audacious political brinkmanship from 2014 seems unlikely this time, for three main reasons.
A changed political and media landscape
First, the political context in which this election takes place is vastly different from 2014 (link is external). Compared to five years ago, this year’s presidential contest is far less polarised and tense. In fact, it is almost subdued.
Religious tensions, for example, had been widely feared in the lead-up to the campaign, but were effectively contained early on by a coercive state apparatus and Jokowi’s acceptance of Ma’ruf Amin as his running mate. In response, Prabowo changed tack and tried to focus his campaign on the economy, but so far, most of his attacks on Jokowi’s economic record have fallen flat.
As a result, Jokowi has had little difficulty in maintaining his lead in the polls. Several surveys (link is external) published in early March confirmed that the incumbent still enjoyed an electability rate of around 50 per cent, while Prabowo’s support rate was stuck in the low thirties. The longer Jokowi can maintain this large gap, the less likely it is that Prabowo will resort to fabricated quick counts on election day.
Second, the media landscape has changed. In 2014, Prabowo enjoyed the support of two influential media tycoons, Hary Tanoesoedibjo and Aburizal Bakrie, who was then also the chairman of the Golkar Party.
Unsurprisingly, the controversial quick count results that saw Prabowo as the winner were broadcast on TV stations owned by Bakrie and Hary Tanoe, such as TV One and MNC TV. One of the pollsters involved in these quick counts later admitted (link is external) that Hary Tanoe’s MNC Group had paid his company for the quick count.
In 2019, by contrast, Prabowo’s influence over the mainstream media is much diminished. Hary Tanoe, who founded his own political party in 2015, switched allegiances to Jokowi for this election. Bakrie, meanwhile, is reportedly still close to Prabowo but is unlikely to allow his television stations to host yet another quick count controversy, especially now that Golkar supports Jokowi. In fact, TV One’s prominent talk show host Karni Ilyas recently apologised (link is external) for TV One’s role in the 2014 quick count saga.
Third, since the 2014 election, one of the industry’s professional associations, Persepi, has taken a number of active steps to restore public trust and enforce better ethical standards in public opinion research.
Immediately after the 2014 election, for example, it expelled two members (link is external) that had presented dodgy quick count numbers for violating the association’s ethics code. Subsequently, Persepi also intensified its communication with the media through joint workshops and seminars, with the aim of lifting reporting standards about public opinion surveys.
Further, individual members of Persepi have complemented the association’s initiatives by promoting a better public understanding of polling methodologies. For instance, when in the 2018 local leadership elections several results differed substantially from pre-election surveys, a number of prominent pollsters proactively explained these discrepancies (link is external) to the public via TV interviews and op-ed pieces in the print media.
Overcoming long-standing divisions in the polling industry
However, while these are steps in the right direction, Persepi’s reach in the polling industry is limited. In fact, only a third of the 33 pollsters registered with the KPU by mid-March are members of Persepi.
Of course, not all the others are necessarily unprofessional but there is a general perception among political observers that many of the pollsters who have opted to stay out of Persepi and join rival association Aropi instead are less interested in methodological rigour and ethical standards. Some pollsters have therefore acknowledged that to achieve improvements for the whole polling industry, the two main associations will first need to overcome their long-standing differences and find common ground on how to deal with key methodological and ethical questions.
This, of course, is homework for after the 2019 elections. For now, it is important to note that even with several dubious organisations registered with the KPU, it seems unlikely that the upcoming presidential election will be dogged by the kind of quick count controversy that occurred in 2014. However, it may yet be a different story for the simultaneous legislative elections, which have received far less media attention than the presidential election so far.
In the race for Senayan, several parties are fighting for survival, and the margins that will decide success or failure on election day might be very tight indeed. Pre-election surveys have indicated that a number of parties, both old and new, may struggle to clear the 4 per cent electoral threshold.
So the stakes are high, and not only for the parties, but also for the thousands of individual candidates who are competing with one another due to the open list electoral system.
It is therefore not inconceivable that at least some of these parties or candidates may be tempted to embrace desperate measures to secure a ticket to the legislature, especially if the number of votes that are missing falls within the standard margin of error for quick counts.
We should not be surprised if there is a pollster or two waiting to provide exactly these kinds of desperate measures.
source articel and image: indonesiaatmelbourne.unimelb.edu.a