THE National Election Committee yesterday presented the results of an independent audit of the nation’s voter lists, reporting that only three per cent of eligible voters’ names checked could not be accounted for.

But the NEC only arrived at that figure by cross-checking their results in a private central database after their initial field test revealed 9.7 per cent of voter names missing, a number close to earlier independent surveys that slammed the accuracy of the list.

NEC president Im Suosdey presented the audit yesterday – conducted at a cost of $30,000 – as a refutation of an earlier survey from the National Democratic Institute, among others, that highlighted a 9.4 per cent rate of missing names.

According to Suosdey, another 5.9 per cent of voter names were found in further searches conducted at the NEC Computer Center, the methodology of which was not shared in the audit report.

“Compared to the 2011 commune elections, we see that the findings of [election monitor] Comfrel were good, but the NEC still doesn’t believe them,” said Suosdey, referring to a 2011 survey by the election watchdog that found that 1.5 million registered voters were missing from the lists. “NEC found that only three per cent of names were missing.”

However, Hang Puthea, executive director of the Neutral and Impartial Committee for Free and Fair Elections in Cambodia, said the NEC had conducted the audit as a clear response to the one released by his organisation and the NDI.

“If it is a real independent company.… I think that the result is maybe not quite different from the results which civil society has found.”

NDI senior director Laura Thornton was also quick to note that the results of her study – and those carried out by Comfrel – weren’t so different from those of BMRS Asia, the Bangkok-based firm contracted for the audit.

“The field audit of this group found exactly what we found,” she said. “It’s statistically the same.”

The NEC didn’t disclose the methods it used to find the initially missing names, saying simply that auditors at NEC headquarters “had more time to conduct the search”, and that the search centre had “advanced capabilities”.

The election body also didn’t say whether these capabilities would be available to polling stations on Election Day. NEC officials could not be reached for comment, despite repeated attempts.

Phoung Vuthy, director of BMRS Asia’s Cambodia office, said that the double-checking process was usually necessitated by patchy rural internet penetration.

“The internet was slow in many provincial villages, and therefore, some checks were too slow to make, so we had to record the data, names, etc and then do a further final check when back in [Phnom Penh],” Vuthy said via email.

“After the fieldwork … we took the names that we could not find during the fieldwork to double check at the NEC computer centre in Phnom Penh, along with our own staff to witness.”

The “missing” names uncovered during the field tests actually turned out to be the result of misspellings, he added.

Thornton, however, said that if discrepancies existed in spellings and registration precincts, those discrepancies would still cause hang-ups come polling time.

“We stand by our audit, so we do believe that people are going to show up on Election Day and not be able to vote,” she said.

Nonetheless, she added, “I’m really, really happy that they did this audit.”

The NEC’s report on the audit didn’t outline the steps the body would take to amend the mistakes that it did find, but intimated that any such changes would come after the election.

It said the organisation would “look into solutions based on past experiences, suggestions of concerned stakeholders, and the post-election conference”.

The audit’s objectives also didn’t explicitly include repairing the voter list. Rather, the audit was conducted to “demonstrate the quality of the 2012 Voters’ Lists”.

However, NDI’s Thornton maintained that “the objectives of an audit should be to do an objective assessment of your list, and if you need to make changes, to make changes”.

According to Thornton, the NEC’s audit, though a positive step, was not complete.

An “authentic voter registry audit”, she said, needed to ensure that a sample of people’s names appeared on the list and verify that names appearing on the list belong to actual people.

In its survey, NDI found that 10.4 per cent of voters on the NEC’s list could not be located or accounted for.

Cambodia’s opposition has long pointed to discrepancies in the NEC’s lists, and frequently demands changes. However, senior leaders of the opposition could not be reached for comment.