Sample count results:
Mr Tharman Shanmugaratnam: 70%
Mr Ng Kok Song: 16%
Mr Tan Kin Lian: 14%
The anticipation is palpable in Singapore as the nation awaits the outcome of its presidential election. While the final verdict may not be in until the stroke of midnight, Singaporeans will get an early indication of the results between 10.00 pm to 11.00 pm. This early peek into the election results is made possible through an official sample count, a practice introduced to enhance transparency and minimize speculation. In this article, we will delve into the significance of this sample count, its execution, and its accuracy.
The Introduction of a Sample Count
The Elections Department (ELD) has taken a proactive step by introducing the concept of a sample count in the presidential election. This innovative approach is designed to serve several crucial purposes:
1. Preventing Speculation and Misinformation
Speculation and misinformation can run rampant during the counting process, especially in a closely contested election. To counter this, the ELD website publishes a sample count result while the full counting is still underway. This preemptive measure helps quell any unofficial sources attempting to influence the narrative.
2. Ensuring Accuracy
The official election results may take several hours to confirm, particularly in tightly contested races. Sample counts serve as a reliable benchmark for election officials to cross-verify against the final tally, ensuring accuracy in the results.
The Mechanics of a Sample Count
Understanding how a sample count is executed sheds light on its reliability. Here’s how the process unfolds:
1. Random Selection of Ballots
At the counting center, a counting assistant randomly selects a bundle of 100 ballot papers from the votes cast at a polling station. This selection is performed in the presence of candidates and counting agents, ensuring transparency.
2. Weighted Tally
Votes for each candidate in the selected bundle are tallied. However, to account for variations in the number of votes at different polling stations, a weighting system is applied. This means that bundles from polling stations with higher voter turnout carry more significance in the sample count.
3. Excluding Spoilt Votes and Nursing Homes
The sample count considers only valid votes, excluding any spoilt votes. Additionally, votes cast at nursing homes, a new inclusion in the 2023 Presidential Election to facilitate elderly residents’ voting, are not included due to the relatively small number of voters in these institutions.
The Accuracy of Sample Counts
Ensuring the accuracy of sample counts is paramount. Singapore’s sample count process boasts a 95% confidence rate, with a margin of error of 4 percentage points. This means that the sample count estimate should not deviate from the actual voting result by more than 4% for 95% of the estimates made.
A retrospective analysis of the 2020 General Election revealed that, out of 31 constituencies, most sample count results had a difference within 2 percentage points of the official results. Only seven single-member constituencies had a larger difference, ranging from just over 2 percentage points to 5 percentage points. Among these, Punggol West, Pioneer, and Kebun Baru exceeded the 4 percentage point margin of error by a small margin.
In the 2015 General Election, all sample count figures closely aligned with the official results for each constituency, with the exception of MacPherson SMC, which had a difference of approximately 2.6 percentage points.
Conclusion
Singapore’s utilization of sample counts in its presidential election process is a commendable move towards transparency, accuracy, and the prevention of undue speculation. This practice not only provides an early indication of the election outcome but also serves as a valuable tool for election officials to validate the final results. As Singaporeans eagerly await the 2023 Presidential Election, they can take comfort in knowing that the electoral process is designed to be fair, accurate, and accountable.