New polling data on Indonesia's presidential election — and the lack of it from certain critical quarters — suggests that Jakarta Governor Joko Widodo's 38-point lead of three months ago has evaporated. He and his opponent, former military commander Prabowo Subianto, may be locked in a dead heat. 

If the race is as close as it appears, that would favour Prabowo. Local party officials who have kept a foot in both camps as the race tightened are now likely to board the Prabowo bandwagon as it picks up momentum. Moreover, businessmen are now likely to place new bets on Prabowo's already well funded campaign. As The Economist noted last week, those close to the Jokowi campaign say that a sense of panic has set in among his campaign staff.

Prabowo Subianto must now be considered the favourite to win the 9 July presidential election, a result that was unthinkable just a month ago.

Several  polls from the coalition negotiation period that ended on 22 May showed that Prabowo gained supporters as other contenders dropped out, but that Jokowi maintained a double-digit lead and shed few supporters. And press reports in late May generally overstated Prabowo's surge in support. But a recent series of polls strongly suggest that Prabowo has closed the gap completely. A poll by the respected Indonesian Survey Circle (LSI)* in early June put Jokowi's lead at just 6.3%. Since then, four polls have been released by less established organisations showing a slight Prabowo lead.

It is always difficult to know which polls to trust in Indonesia. As Karim Raslan points out, most pollsters are connected to or on retainer to specific candidates, and this can occasionally influence results (in a particularly egregious example, during the 2009 presidential election, former Vice President Jusuf Kalla's campaign team commissioned polls showing him in the lead; he received 12% of the vote). But three organisations known for their accuracy have generally served as a reference point: the Centre for Strategic and International Studies (CSIS), an international relations think tank; Saiful Mujani Research and Consulting (SMRC); and Indikator.

Yet none has released a face-to-face poll in the last month. This is highly unusual, but political affiliations could explain the silence.

Despite their reputation for accuracy, both CSIS and SMRC are run by Jokowi supporters. The executive director of CSIS, Rizal Sukma, is a leading adviser to Jokowi and largely responsible for his strong showing in Sunday night's foreign policy debate. Saiful Mujani, head of the epynomous consulting firm, has campaigned for the Governor. Indikator is a young outfit founded last year by former Indonesian Survey Institute pollster Burhanuddin Muhtadi, also a Jokowi supporter. He tweeted to my colleague Catriona and me on Saturday that the situation was 'critical' for Jokowi, and that he needed help.

Are these organisations holding back polls that show a more significant Prabowo surge than previously reported?

Perhaps they assume that if they were to release a poll showing a race too close to call or with a Prabowo lead, local officials would race to register their support for him. Golkar cadres at provincial and regency levels have been instructed to support the Prabowo ticket, in line with Golkar's formal affiliation, but often feel a stronger affinity for Jokowi's vice presidential candidate, former party chairman Jusuf Kalla. Many, seeing dispositive poll results, might decide to get on the winning train before it leaves the station. In a close race, Golkar's renowned party machinery could be key.

A tight race could also lead to higher levels of bloc vote-buying than we normally associate with presidential politics in Indonesia. There were certainly instances of vote-buying in 2004 and 2009, but because SBY won by 20% and then 34% in those years, it could not have affected the outcome. And because the result was scarcely in doubt, there was little incentive to engage in or fund electoral corruption to sway a few thousand votes.

But in a close race, tycoons may open their wallets in an attempt to move the result a critical hundredth of a percentage point in their favour. Though both sides have recruited businessmen with large war chests, Jokowi's reputation as an incorruptible reformer – and indeed perhaps his sincere convictions in that respect – strongly disincentivize his use of their funds to engage in bloc vote-buying. Prabowo, despite his angry declarations on the stump that Jakarta's political classes have been bought, does not have a similar reputation to worry about.

Why have Indonesians moved en masse to support Prabowo's candidacy? Former AusAID official Doug Ramage argues that Prabowo effectively defined Jokowi as a policy lightweight, and that Jokowi's reputation never recovered, even after his team released a detailed 41-page manifesto in late May. Anecdotal evidence also suggests Prabowo has captured the imagination of the middle class, who get much of their news from Indonesia's television news channels. The former general is clearly winning the air war; two big stations with a combined audience share of 40% are backing Prabowo, while Jokowi ally Surya Paloh's television station can claim only a 2% share. For the more tech savvy crowd, Prabowo's social media effort is years ahead of Jokowi's. His campaign team has released a series of slick videos designed to go viral among young voters, including this step dance and this Queen ripoff featuring Indonesian rockstar Ahmad Dhani. 

There are still two weeks left before the election, but Prabowo's momentum is now self-sustaining. All this suggests that a Prabowo victory is now likely, and is becoming more likely by the day. 

*This post originally credited the poll to the Indonesian Survey Institute (LSI), a similar polling organization which uses the same acronym.

* *Disclosure: I conducted my Fulbright research on Indonesian foreign policy at CSIS in 2010, but have not discussed polling with Pak Rizal.

Photo by REUTERS/Stringer.