Quick count, anxious voters' new reference

M. Taufiqurrahman, The Jakarta Post, Jakarta

In the midst of a fairly slow manual vote count used by the General Elections Commission (KPU), voters who are anxious to know the final result of an election have the so-called "quick count" to thank.

Developed for the Indonesian context by the Institute of Research, Education and Information of Social and Economic Affairs (LP3ES) in 1997, not only has the quick count displayed acceptable precision in predicting the final result of the polls, it also delivers the result in just a few hours after voting concludes.

In the first round of the presidential election in July, the quick count, jointly carried out by LP3ES and the Washington- based National Democratic Institute (NDI), extrapolated their samples from across the country and determined that Susilo Bambang Yudhoyono would emerge on top with 33.9 percent of the vote, while Megawati Soekarnoputri would likely get some 24.9 percent of the vote.

The KPU's official result placed Susilo on top with 33.2 percent and Megawati second with 26.6 percent of the vote.

The result of quick count was based on actual observation by observers at selected polling stations across the country.

In Monday's runoff election, LP3ES was able to monitor 2,000 polling stations. It said Susilo would win the presidency with 60.2 percent of the vote.

Sample determination plays a significant role in the quick count, as it will affect the validity of the data collected.

Because a polling station serves as the only unit of analysis in a quick count, the polling station surveyed will be determined only after the election organizers release the complete list of the stations.

As voters are mostly distributed in Java, quick counts will proportionately draw more of a sample from polling stations on the island, and less from sparsely populated provinces like Nanggroe Aceh Darussalam or East Nusa Tenggara.

The sample determination must also take into account urban and rural voters. The result will be flawed if it ignores any of these key factors.

"One quick count conducted by another association in the first round of the presidential election predicted that Amien Rais would end up second, while Wiranto would finish fourth. The results were inaccurate because its samples only comprised urban voters," LP3ES researcher Tatak Prapti Ujiyati told The Jakarta Post, without naming the association.