Preliminary reports on the results of the random manual audit indicate that the PCOS/VCMs have an overall accuracy rating of 99.884%. Based on the standard set by the Commission on Elections, the PCOS/VCMs should have 99.995% accuracy. This translates to one error mark out of 20,000 marks. The 99.884% accuracy rating translates to about 23 error marks out of every 20,000 marks.
The reported PCOS accuracy rate in 2010 was 99.6% or a discrepancy of 80 votes out of 20,000 vote marks. In 2013 it was 99.9747% or a discrepancy of 5 votes for every 20,000 vote marks.
The PCOS/VCM used in the recently held national elections is a new model and newly manufactured. It performed better, in terms of accuracy in reading vote marks, compared to the PCOS used in 2010. But the PCOS used in 2013 performed better than the PCOS/VCM used in 2016 despite of the discovery of digital lines in some ballot images in 2013.
About 43,613,248 registered voters actually voted in the recent elections. If we assume that all of them voted for a vice president, we can expect that there are 43,613,248 vote marks on the same number of ballots. Applying the 99.884% accuracy rate, this means that 50,591 vote marks were erroneously read by the PCOS/VCM. The standard 99.995% accuracy applied on the same population of vote marks, on the other hand, translates to 2,181 vote marks erroneously read. The erroneously read 50,591 marks is about 23.19 times the standard!
The 99.884% accuracy rate is still below the 99.995% accuracy rate required under the terms of reference published by the poll body.
* * *
An issue raised by certain camps is the matter of under votes.
An under vote is a valid vote, an expression of a voter’s preference not to select a candidate for a single position contest or selecting less than the allowed maximum number of candidates in a multiple position contest. It is basically an abstention, well, a case of partial abstention in multiple position contests.
Election experts opine that 1% under votes in relation to the number of ballots cast is acceptable. 2%-5% is a cause of concern. 6-10% an alarming cause of concern.
Election returns received from the transparency server show that there were 1,243,586 under votes in the presidential contest out of 43,612,673 ballots cast or an under vote rate of 2.8514% of ballots cast.
There were 2,907,404 under votes in the vice presidential contest out of 43,613,248 ballots or an under rate of 6.665% of ballots cast.
The under vote rates in the presidential and vice presidential contests are both beyond the acceptable limit of 1%.
The under votes counted have remained unexplained. The cause of these under votes can be determined only by conducting a complete audit of the automated election system. The audit should consider how the PCOS/VCM is configured since a miss reading of the vote marks may result in the recording of marks as under votes.
* * *
An issue that has escaped the attention of observers is the uncounted votes.
Election returns received from the transparency server shows:
For the presidential contest, 43,612,673 total ballots cast less 1,243,586 under votes and less 792,175 over votes yields a net valid vote of 41,576,912. Net valid votes less 41,278,124 sum of all candidates’ votes yields an uncounted votes number of 298,788.
For the vice presidential contest, 43,613,428 total ballots cast less 2,907,404 under votes and less 574,792 over votes yields a net valid vote 40,131,052. Net valid votes less 39,832,254 sum of all candidates’ votes yields an uncounted votes number of 298,798.
Just like the under votes, simply looking at the election reports will not yield the cause of uncounted votes. The cause of these uncounted votes can be determined only by conducting a completed audit of the automated election system.
* * *
Let’s face IT! The law authorizing the use of automated election systems requires that such systems operate accurately, properly, and securely. The numbers discussed above do not indicate that the automated election system, specifically the PCOS/VCM, performed accurately. It is about time that the Joint Congressional Oversight Committee on the Automated Election System consider other systems and look into Filipino developed systems rather than allow foreign entities to run and manage Philippine Elections.