Fauzi garnered a total of 438,393 votes in East Jakarta; 327,495 in West Jakarta; 170,006 in Central Jakarta; 214,449 in North Jakarta and 319,389 in South Jakarta.

The incumbent, however, won the most votes in Thousand Island regency, taking home 6,916 out of 12,555 votes cast there, while Jokowi fell to third place with 1,273 votes.

Preliminary vote recapitulation also showed that independent candidate Hendardji Soepandji received the least amount of public support, collecting only 25,042 votes in East Jakarta; 19,291 West Jakarta; 9,841 in Central Jakarta; 13,330 in North Jakarta and 18,361 in South Jakarta.

Jakarta’s second direct gubernatorial election was held last Wednesday with a total of more than 6.9 million people listed on the official voter’s roll.

Votes were counted immediately after polling stations closed. Ballots and vote count reports were handed to KPU Jakarta subdistrict committees, who later reviewed the documents and submitted them to district committees.

District committees follow the same process and report to municipal- and regency-level committees, who in turn report to the KPU Jakarta provincial office. KPU Jakarta is scheduled to officially announce the final election results on July 19.

As electoral regulations require a candidate to secure more than 50 percent of the total vote to win the election, Jokowi and Fauzi will face each other in a runoff, which had been scheduled for Sept. 20.

According to a number of quick counts by several pollsters, Jokowi finished first in the election and Fauzi second. Quick counts by the Indonesian Survey Institute (LSI), Kompas daily and Indo Barometer showed that Jokowi won the race with at least 42 percent of votes, followed by Fauzi with around 34 percent.

The quick count results turned heads as previous surveys predicted that Jokowi and Fauzi would be the candidates to square off in a second round of voting, but with the latter securing first place with a result far above the other five contenders.

Separately on Tuesday, Agus Herta Sumarto, a researcher from the Political Research Institute for Democracy (PRIDE) Indonesia, said the pollsters’ inaccurate prediction of the election results was an indication that the various surveys on the Jakarta election were biased and based on questionable methods.

Agus pointed out that a number of pre-election surveys had similar methodologies and samples but had significant disparities in their results.

He mentioned, as an example, an electability survey conducted by the Center for Policy and Strategy (Puskaptis) involving 1,250 respondents from April 2-7 that showed Fauzi to be leading with 47.22 percent, with Jokowi trailing far behind with 15.16 percent.

But another survey done by Cyrus Network, which involved 1,000 respondents and was conducted from April 8-16, showed Fauzi leading with 42.40 percent and Jokowi trailing closely with 31.8 percent. Both surveys used multistage random sampling and face-to-face interviews.

“Is it possible for public opinion to change dramatically within 24 hours?” Agus asked rhetorically during a discussion in South Jakarta, adding that the results of those surveys could be rigged by marking up the margin of error and steering the interview questions toward a specific response.

“What I’m questioning is how surveys using the same methodology, similar samples and conducted within same time period ended up showing very different results,” he said. (han)