LESSON FROM KENYA FOR ELECTRONIC VOTER BVR

Once voters’ details were captured by the BVR kit, an electronic applicant’s number was automatically generated by the kit that showed the BVR kit number, the year of registration, the date and time of registration.Tanzania is spending over 200 billion on next year election for 8000 BVR what can we learn from Kenya ?


1.The data from the BVR kits was later uploaded to the central server of the IEBC. This was used to generate the electronic poll book that was used on 4th March 2013.
2.IEBC procured Electronic Voter Identification Devices (EVID) to identify and confirm voter registration details on the elections day.2: Drawing a nationally representative sample of the registration centres. Our sample was 3,750 registration centres.
Step 3: Determination of the number of observers that ELOG was able to engage - based on the budgetary allocation. ELOG planned to engage 250 monitors.
Step 4: Determining how many days to engage one observer. Each observer was engaged for 15 days. This meant that each monitor was to report in one registration centre, remain there the whole day (7.30 am - 5.00pm) to observe and report. The reporting strategy included texting 3 SMS’s per day to the BVR database and completing one checklist an incident form for that specific centre.
Step5.Application of a short code from Safaricom and Airtel service providers. Our short code was 4156
Step 6: Recruitment of Constituency BVR observers. This was based on the sampled registration centres to reduce travel costs.
NB. As much as ELOG had planned to train and deploy 250 Constituency observers to observe BVR processes in 3750 registration centres, we were only able to train 214 observers who were deployed in 3,210 registration centres.
This is because it took too long for the IEBC to determine the timelines for the BVR voter registration exercise which depended on the procurement of BVR kits and training of the registration clerks. By the time the IEBC declared the timelines for the BVR process, ELOG was time constrained and could not train all the 250 BVR monitors as envisaged. Therefore we deployed only those whom we were able to train.
Step 7: Accreditation of BVR observers by the IEBC. The commission required details of observers for accreditation
Step 8: Development of BVR monitoring Tools. ELOG developed BVR checklist and incident form that had coded responses. Coding of responses enabled the observers to send their responses via SMS using a predetermined formula. We also developed a texting guide that was used to train BVR Monitors. Each day the observer was required to send three SMS’s to the BVR database.
Step 9: Development of the BVR database(this had observer details in regard to the registration centre codes that they were expected to observe, their contact details, the BVR Checklist and the incident form)
Step 10: Training and deployment of observers. All trainings were conducted at the national level. Deployment was done based on the sampled registration centres.
Observers stayed at each sampled registration centre the whole day. They were expected to text 3 SMS’s to the BVR database and complete one checklist per registration centre
Observers also used the incident form to text any incidences that occurred during registration process.
Completed checklists and incident forms were later sent to IED’s office after the observation period for final analysis and compilation of the final report.
On elections day, ELOG used the PVT Observers as well as the general observers to observe the E-day processes.
Procurement of the BVR kits was marred with controversies that eroded the confidence in the voters roll and the commission. A high profiled procurement process of $50million USD was cancelled occasioning public concern about the IEBC’s independence.
After terminating the BVR procurement process, Government intervened and 15,000 BVR kits were procured using loans negotiated between the government and the Standard Chartered Bank of London.
Delay in the procurement of BVR kits affected the elections timelines in general. This occasioned the amendment of the Elections Act 2011 which reduced timelines for voter registration and compilation of the principal register of voters from 90 days before elections to 45 days
According to the Elections Act, 2011, voter registration ought to have began in the month of September, 2012, while compilation, inspection and publication of the register would have commenced in December 2012 . However due to delay in the procurement of the BVR kits, BVR registration took only a month from 19th November 2012 to 18th December 2012.
Timelines for inspection of the voters register were also affected. Inspection took only 14 days instead of one month long period provided in the law.
The compressed timelines caused by delay in the procurement of BVR kits affected preparation for the 2013 General elections .
Malfunctioning of some of the BVR kits, low turn out of eligible voters especially in the 1st two weeks of registration, insecurity in some regions, insufficient power to drive BVR kits, lack of National Identification cards and misconception that BVR is harmful to the body.
Late commencement of Diaspora voter registration. By 30th November, 2012 Diaspora voter registration had not began. Two weeks to the closure of BVR.
Lack of information on BVR Kits movement. In total there were 24,613 registration centres. Therefore the BVR kits were shared between two or more registration centres. There was no operational schedule on the movement of the BVR kits a number of voters were disenfranchised.
This also affected the BVR observers who had now to observe based on the presence of BVR kits in the registration centres
Engaging all stakeholders to mobilize eligible voters to register was a good strategy for it increased voter registration from 33% in the 1st two weeks of registration to 79.9% by close of the registration period.
In total IEBC registered 14,352,545 voters in Kenya and 2,637 Kenyan in the East Africa Community(EAC) community.
Non-registration of those hospitalized and prisoners
Information that the Commission will not register voters outside the East Africa community came very late.
ELOG used two methodologies: Parallel Vote Tabulation and general observation. PVT was the method that was mainly used on E-day.
Both methods complimented each other to boost confidence in the fairness of the entire electoral process.
In order to support recruitment of PVT constituency supervisors and observers, ELOG recruited 9 Zonal coordinators.
ELOG also recruited and effectively trained 48 Master trainers who trained all the 580 constituency supervisors , 972 PVT observers and over 6000 general observers.
ELOG trained and deployed 580 Constituency Supervisors, and over 7,000 observers in all the 290 constituencies. Out of the 7,000 Observers, 976 were deployed as PVT observers in a nationally representative sample of the total 31,977 polling streams
ELOG received reports from 97.5% sampled polling stations on E-day
Highlights of E-day Observations on BVR
Failure of the Electronic Voter Identification Device(EVID) in 55.1% of all polling streams resulting to the use of the printed hard copies of the voters register by the IEBC.
Voters whose details were not on the register were not permitted to vote in 84.8% of the polling stations(according to the law)
Voters whose details were not on the register were permitted to vote in 15.2% of the total polling stations
Voters names were properly marked or crossed out once their details were confirmed in the voters register in 99.4%
In spite of comprehensive training, some participants could not pass the post-tests and therefore they had to be replaced. This meant organizing for additional trainings.
2.Late compilation of the official list of all polling streams by the IEBC, delayed the finalization of ELOG’s PVT sample. We had to recruit PVT observers on ward basis who were later deployed in specific sample polling streams. The extra PVT Observers were deployed as general observers
3. Suspicion from IEBC that ELOG would release the PVT results before the official announcement . ELOG had to sign an MOU with IEBC that provided that ELOG would only release PVT results after the commission had announced the official results.
4. Misunderstanding of PVT objectives by a number of stakeholders. Some thought it provides alternative results from that of IEBC. ELOG had to undertake extensive media awareness and understanding of what PVT is and what is not. We had to insist that only IEBC is mandated by law to announce results and PVT results are meant to verify the accuracy of the official results.
The controversies surrounding the procurement of the BVT kits affected our BVR monitoring processes. Due to compressed timelines we could not recruit and train the 250 BVR observers as planned.
Early planning and fundraising for the entire observation processes is crucial in ensuring resources are available on time. Probably a year to the E-day. By November 2011 ELOG observation proposal and budget were ready.
2.Planning for a run-off in highly contested elections is a good practice. ELOG was prepared in advance to observe the run-off even though it did not occur.
3.Consistently following up on the IEBC to share the final official list of the polling streams. ELOG never gave up.
4.Recruitment of competent and committed staff to work at the secretariat level is critical for the success of PVT and general observations
5. Training smaller groups of participants ensured quality and active participations. ELOG ensured each training group had not more than 25 participants.
6.Administering pre and post test to participants during training ensured only those who qualified were engaged as supervisors and observers. This minimized the level of errors during transmission of PVT data and completion of observer forms.
7. Having master trainers ensured that all trainings were standardized and the same delivery methods were used.
Technical advice and expertise provided by NDI staff played a key role in the success of PVT
9. Development of training materials in advance. Allowed for early review and ensured that trainers and participants had relevant materials on time.
10. Simulation of the PVT system before e-day is important. ELOG had a national wide simulation on 25th February 2013. This enabled us to address a number of issues that could have hindered the success of PVT
BVR is a better system for voter registration compared to manual processes.
However, the technology need to be introduced on time to ensure that it is tested on a large scale to ensure that it works.
Moreover training of staff to manage BVR is key. Proper training would reduce human error that could be associated with data entry process.