|Subject: The case of the missiing 166,000
The Case of the Missing Voters
Some election observers (myself included) - and East Timor's Prime Minister Jose Ramos-Horta have been distrubed by the large gap between the official number of registered voters in East Timor and the number of valid votes actually cast.
With 522,933 registered voters, there were only 357,766 valid votes cast, leaving a worrying gap of 165,167 voters having 'disappeared'. This is easily a big enough number to cast doubt over the whole voting proecss.
The failure of the Electoral Commission to acknowledge this discrepancy, much less provide an adequate explanation for it, has further increased quite legitimate concerns. However, it now appears that the 'missing' voters have been 'found', or at least their number accounted for.
East Timor' Electoral Commission, like many other institutions in East Timor, is still learning how to do things. Since its formation earlier this year, the Electoral Commission's learning curve has been steep, and it has faltered a few times along the way.
One of the mistakes it has made is that registered voters who have died since the last ballot two years ago have not been taken off the register. This could account for around six per cent of the registered voters, or perhaps more. Beyond this, it appears there has been a relatively high proportion of votes counted as 'invalid', either being black or not properly marked.
The differing criteria for what constitutes a valid mark on a ballot paper between polling stations, and the clear partisan allegiance of some polling staff, has meant that the Electoral Commission has now decided to re-assess all invalid votes in Dili. This will not significantly alter the outcome of the vote, but it is a confidence boosting measure and a good sign of transparency. Invalid votes, however, currently account for around 10 per cent of the ballots cast.
Beyond this, a significant number of voters did not attend the polls. No-one knows the exact figure, but it seems that at least seven per cent of registered voters did not cast their votes, and that number could well be higher. Beyond this, there were also a of registered East Timorese living overseas, who were not allowed to vote in this ballot.
Finally, when some voters sought new voter registration cards, their earlier registration numbers were not deleted from the lists, meaning that a significant number of voters were registerd twice. They thus could have also voted twice, except that the broadly applied method of dipping the voter's index finger in indelible ink precluded that.
So, in short, there was a big gap between the official number of registered voters and the actual registered valid vote. But we now seem to have a plausible explanation.
It is a pity, however, that the Electoral Commission was not able to provide this explanation. It would have saved considerable angst, and taken away one of the concerns over what was widely regarded as an imperfect but generally free and fair election.
Associate Professor Damien Kingsbury Coordinator, Victorian Local Governance Association Election Observation Team Director, Masters of International and Community Development School of International and Political Studies Deakin University Ph: +61(0)439638834
East Timor Journal
An online journal detailing the ups and downs of working for an international NGO in Dili, East Timor.
13 April 2007
"54-46 That's My Number"
A few new numbers rolled into my inbox this morning, and I thought that I would keep up the scratch-work until official figures lay this to rest.
The new number - 90,000 invalid votes (a little too round) - came via a FRETILIN spokesperson in a Thursday interview with the Jakarta Post (12.04.07). This number, I expect, is high.
Here is why: Adding together the total number from the CNE of valid votes counted (357,766) with FRETILIN's number for invalid votes (90,000) we get an estimate of total votes cast - 447,766. In this scenario, invalid votes make up 20% of all votes or 1 of every 5 votes cast. As I wrote before, this figure is generally expected to be below 5%.
Votes are invalidated for a number of reasons, including the lack of both a signature and a stamp on the ballot paper, no vote, and multiple votes. While a high percentage of invalid votes could reflect fraud, it also is attributable to a dearth of voter education programs and deficient poll worker training - both of which are ongoing issues in Timor Leste. The number of invalid votes here may be a little elevated, but 20% is freakish.
The other thing that the FRETILIN's figure for invalid votes would show is a 85% voter turnout rate (total votes cast [447,766] divided by the total number of registered voters [522,933] x 100%). While this rate basically matches earlier elections, the international trend is for voter turnout to decrease with each election. Additionally, getting around rural Timor Leste is not particularly easy, which could also depress voter turnout rates. Again, I think that 90,000 is too high.
Which leads me back to the 'mysterious' 30% gap between the number of valid votes counted and the number of registered voters (See 'By the Numbers'). If you're still confused, you are in illustrious company:
"East Timor presidential runoff candidate Jose Ramos-Horta said Thursday the United Nations should explain why 30 percent of voters did not cast a ballot in the first round and demanded a recount. 'It seems like at least 30 percent of voters did not vote,' he told reporters. 'Why? I don't know.'" (Agence France-Presse, 12.04.07)
Here is my answer: We need the number of invalid votes to tell us how much of that 30% is attributable to depressed voter turnout (JRH's concern) and how much can be attributed to invalid votes (FRETILIN's preoccupation), in addition to any uncounted* or lost ballots. The way Ramos-Horta presents it to the media assumes 0% invalid votes - an impossibility. Some of those voters did vote, they just, for instance, did not do it correctly or received invalid ballots.
It is nice to be able to say that those on both side of this fight are partially correct: there are invalid votes and some voters didn't show. We will just have to wait for official figures on either invalid votes or total voter turnout to get a sense of the real breakdown.
* It is important not to treat the 357,766 figure for valid votes released by the CNE as static. There is very little information on how many votes remained to be counted at the time that number was released (I've heard at minimum nine ballot boxes). As these valid votes are added, the 30% figure will shrink.