The Deceitful Voter Register

The Deceitful Voter Register

According to the detailed report prepared on 25 August 2010 by the Electoral Reform International Services (ERIS) on the Preparation of the Voter Register during January-June 2010. ERIS was first contracted to review the state of the Voter Register prepared in 2008 and the readiness of the National Electoral Commission (NEC) to hold an election. Later, ERIS was asked to help NEC with the Presidential Elections. The report notes that on 26 June 2010, NEC held a successful Presidential election, which "generally met international standards" according to outside observers.

In this August report ERIS, however, concludes that the Voter Register on which the Presidential elections was based contained up to 30% fraudulent or duplicate records after all efforts to cleanup the system were exhausted.

According to ERIS, the registration process was marred from the very beginning with errors in "recording data and a high rate of false, illegal and duplicate registrations". ERIS also found system problems in the transfer of data to the main database and that some codes were duplicated and some data were lost. ERIS also found that the application AFIS (Automatic Finger Print Identification System) and FRS (Facial Recognition System) was unsuccessful in catching duplicates. The quality of pictures and fingerprint impressions may have been too poor as ERIS could not find inherent problems with the systems. In its later attempt to clean up the system, ERIS used lower thresholds, but the much touted FRS would not correctly identify known, control duplicates.

In a surprising finding, ERIS emphatically concludes that the existing voter registration system should not be re-used. Quoting from ERIS "The system was designed as a one-shot system to generate a register which is then thrown away after use". ERIS further concludes that biometric data collected through this system were not good enough to allow for the detection of duplicates. The system has no unique voter numbers, has known bugs and produces duplicate civil IDs and voter cards, and the "software was not fit for the purpose". The system did not have the functionality to correct errors, amend or transfer records and remove deceased voters.

ERIS's conclusion that the system was disposable is contrary to the often stated selling point- that Somaliland needed this expensive exercise to move its democracy forward and to sow the seeds for viable civil identity and voter registrations cards from this day forward; that this was a treasure. ERIS concludes "it would be easier to start again with an empty register and use the old [current] register merely as one source of supporting evidence". In other words a new registration may be inevitable and we have to either start from scratch or go back to our old system of no registrations.

Expert from Creative Associates (Independent American Organization) who had also scrutinized the register accepted same with ERIS that data was corrupted and cannot be re-used, their report showed that the software used to develop the list especially, the FRS software and the combining of the FRS & AFIS has serous defects and therefore cannot be considered of sufficient quality for the production of the list.

Hence it is clear from the report that NEC believes that elections in this current cycle (forthcoming elections for municipality) can still go ahead using the current Voter Register with its 30% fraudulent records and other warts.

NEC and International Somaliland Election Donors are going to ignore to find a way to accommodate the large number of voters coming of age each since 2008 (Four generations) and legitimate voters who missed the last registration, which is not fair election and will only create endless conflicts and destabilize the hard won democracy of Somaliland.

Other un-acceptable Findings of the Voter Register

Data of up to 14 centres is completely missing and is not included in the list, a proper audit of the data were not carried out, therefore, the completeness of the rest of the data is unclear.

The number of duplicates found in the same centres some of them totaling hundreds is also additional proof that the observed duplicates are not accidental.

More than 96,000 people who were struck off the register because appropriate technical enhancements of their photographs were not applied as agreed. This means that at least 80,000 genuine voters were being directly disenfranchised.

The disclosure of such sensitive personal data constitutes gross violation of the law, infringement of the right of the individual and fundamental breach of international standards of data privacy.

The data is overdue according to the constitution where the validity of the data to be only six months between the registration and the voting period.

The absence of the corresponding duplicate records of the data and the lack of summary reports shows the list was not intended to be transparent but to be taken on face value, as it is.

The comparison of electoral results of 2005 (parliamentary election) and the total number of registrations in districts where the elections were held in 2005. The average increase is around 106%, but the dramatic difference in the from -39%(Lughaya) to 373% (Caynabo) can be attributed to the levels of duplicate records in each district.

When 30% of the list are in fact fraudulent registrations, as approved. A register containing such high percentage of error cannot be regarded as being a credible voter list. It is clear that this system has failed quite substantially.

The total number of records with no fingerprint data was found to be 629,685 where only 96,780 were discarded.

Ultimately non compliance with all of the objective criteria & International standards would render this list as unreliable, incredible, and potentially a danger to Somaliland's democracy.

Thank you,

Abdi Halim Mohamed Musa

Acting Secretary-General of UDUB Party