Biometric recompilation and SMS voters verification in Cameroon

Biometric recompilation and SMS voters verification in Cameroon

A sound voter registration process is crucial to a successful election. Yet voter registration is also often the most expensive part of conducting elections, and therefore there is a need to look at how cost effective voter registration can be.  There is no ‘one way’ or ‘best way’ to conduct elections and, for that matter, voter registration. What works in one country does not necessarily work in another. Each country has its own political and socioeconomic contexts, its own resource limitations and its own needs to take into consideration when designing a voter registration system. Much attention has been given lately to the use of technology, particularly the use of biometrics, in voter registration. Some argue that biometrics will make the voter registry ‘accurate’ and ‘fool proof ’. But there never can be an absolutely accurate voter register, even in the most developed countries, due to the difficulties of constantly capturing deaths, movements of populations, etc. In addition, biometrics in itself will not increase confidence in a voter registration process if, for example, the entity that conducts voter registration does not have the confidence of the populace as a competent and impartial body. (Craig Jenness)

Those of us involved in electoral assistance must simply be aware of the new pitfalls that technology can introduce. For example, vendors and experts related to them may hide ‘future costs’ when certain technologies are being marketed to election commissions. We should also be aware of who will have rights to the information gathered and the software used. Electoral processes– for both good and bad – become ‘big business’ to a number of entities ( corporate, NGOs and governments alike).

A biometric registration system uses one or a combination of facial and finger prints of individuals as a unique identifier in a voter’s database.  It may be good to know that no two persons have the same finger prints or the same facial formation if digitally analysed. This uniqueness is therefore what is deployed in doing clean electoral registers free of duplicate data.  The biometric information is however not the only unique identifier in such a system.  National identification numbers as well as names sequence and birth information (date of birth, place of birth and parents names) are additional security matching conditions.

Bribery and corruption of elections commissions by big companies and influence by governments related to the allocation of markets for electoral processes are well known facts in many developing countries. For an election to be transparent and acceptable, the election commission itself must be independent, transparent and accountable.

Finally the decision to recompile the electoral list has been announced.  But how realistic is this in terms of time and money?  What is biometric registration in the first place? A lot of persons I talked to simply think that a voters list with pictures and fingerprints means biometric. There is more to this than such a simple definition. But my worry here is “how likely will ELECAM do a biometric registration?" For this to happen within reasonable time before the elections of 2012, the nation will need to deploy a tested technology with over 4000+ fixed and mobile  registration centres or terminals, put in place the required funds to pay for such a service.  Within the 2011 national budget, ELECAM had 11 billions to run the presidential election excluding recompilation of the voters list. In the 2012 budget, ELECAM has 11.5 billion for the twin elections and now including voters’ registration.  Analysts will definitely reach at the conclusion that if the 11 billion Francs CFA allocated for the  2011 was completely used, then the current elections will not be a possibility given the fact that over 10 billion out of the 11.5 billion allocated to ELECAM in the 2012 budget will be allocated for voters registration alone. So the different stakeholders need to know if the budget announced by ELECAM from the government is the 11.5 billion allocated in the finance bill of 2012 or is an additional sum to take care of the recompilation.

I understand so well that by making public the figures may influence pricing of the technology to be deployed but I also think that the public deserve the right to know those figures once the bit has been finalised.  Transparency and accountability should start with the body’s operations.

The same exercise in Nigeria cost 81.9billions Francs and about 18.7 billions in the Philippines. After comparing the cost of the different registration processes in Nigeria, Kenya, Malawi, Lesotho, Ghana, Sierra Leone, Burkina Faso, Benin, DRC, Mali,  I am more or less certain that if the 2012 elections are based on the current 11.5 billion in the finance bill, it is  likely that we will be heading to a real disappointment.  Also, any move by ELECAM to do another manual recompilation will also be another waste of time and will generate the same problems we are trying to solve.  There are many methods of recompilation ranging from the current manual system through the Optical Mark Readers (OMR)  to the most effective active biometric registration system.

Let me take this time to give very brief description of the different options:

Manual registration (used by MINATD and continued by ELECAM): This is the simplest and the most ineffective system of registration in terms of time and accuracy. With this method, people are deployed to specific registration centres and collect voters’ data on forms or registers. These registers are then sent to the regional centres where they are keyed into a computer with or without a database. This regional data are then sent to the central point where they are merged to form a national electoral list.  In the case of ELECAM for 2011, some form of database was used to input the data. Evidently, this database was so poorly developed that it allowed duplicates with basically zero variance in the data fields.  The fact that the data is collected manually and then typed later; already give a major chance for data entry errors thereby making this the most ineffective and inefficient system of voters registration. In terms of speed of processing, this is the most ineffective. In terms of cost, it’s the cheapest.  In terms of ability to promote fraud, this method has the biggest chances.

DRS ( OMR) used in  Mali,  Kenya, Burkina Faso, Sierra Leone:

This system comprises of specially designed forms that are filled and later scanned into a database by an Optical mark reader (OMR).  The main registration centre receives the completed forms from all over the country and then scanned then into a database.  After all data is captured, it is then matched and duplicates ejected from the system.  While this is comparatively cheaper than the biometric system, it has its own problems. This system will not be able to print voters’ cards on the spot, it takes time and it is also vulnerable to fraud if anyone will so desire to complete additional forms between the registration centers and the final central processes Office.  However, this problem can be solved if all political actors at all the registration centers are given the statistics(eg total number of forms) at the time registration is closing at each of the centers.  These statistics can then be compared to the final voters list when it is finally released.  Also, as a general control, the 2012 election MUST completely make it illegal to add voters’ names in the electoral list manually during the voting process. In Mali, it took 3 months to register 13 million voters using this system.

Inactive Biometric system:

Inactive or passive Biometric registration system uses a non interconnected system that operates a full biometric registration program.  The matching operation to eliminate duplicates is done on each terminal before consolidating into a national database. This system will also be able to print out voters cards on the spot. Because the systems are not interconnected, anyone wishing to register twice can register successfully at different registration centres and receive two different cards.  However, the second matching done when all the different data are brought together( consolidated matching)   will make it possible to identify the duplicates coming from different registration centres.

Eg.  Mr. John Mbah registers in point “A”, get a voters card and rush to point “B” and register again and get the second card. Because point “A” is not communicating with point “B”, this is possible to occur.  However, when the data from point “A” and “B” are eventually brought together and sorted ( marched), John Mbah will be picked by the system as a duplicate and one of the names eliminated. So the passive biometric system is still an effective method

                                                                    

Active biometric registration system:

This system functions exactly like the inactive biometric system but for the fact that all the 4000+ registration centres are connected to a central server in real time via GPRS where mobile phone networks exist and by satellite in zones or villages not covered by mobile phones.  The big advantage here is that the data is sent to the main database in real time and the national matching is done.  This method is the most effective but has its own problems if the GSM network is poor.  Double registration using this system is relatively unlikely as compared to all the other systems.  It is also more expensive in terms of the required technology.

Whichever system is adopted, basic human ethics and moral standing remain a principal factor in ensuring an effective outcome.

Mobile Voter’s verification system

Early this Year, ELECAM also approved a proposal that will allow Cameroonians anywhere in the country to check their names in the electoral list without having to pay Taxi fares, stand in long lines and spend hours looking through thousands of names.  With the mobile verification system, voters will simple send their ID card numbers by SMS to a special short code 8000. The system will verify the ID card in the ELECAM database and send back a response to the voter by SMS stating if the ID card is found in the database or not.  The SMS will also tell the voters all details about his registration such as:

Names:

Polling station:

Date of Birth:

Registration Number:

If the ID is not found, the system will send back an SMS encouraging the voter to verify that the number he sent was correct OR to go to the nearest ELECAM registration centre and get him or herself registered.  This is a great innovation all aimed at contributing to a more transparent election in Cameroon.  This system works wherever you maybe in Cameroon.  This allows potential voters to know their polling stations long before the polling day and hence will reduce the known situation of persons running from one polling station to the other looking for their names.  This system has been tested and is now ready to be deployed once the recompilation begins

However, both systems will be available during this process.

By

Tanyi Christian E
CEO, Martin Luther King Jr. Memorial Foundation

Mobile 237 77 94 74 49
Post Box 1348, Limbe, SWR, Cameroon,  Africa
Skype: Christian24035;   E-mail: christian@lukmefcameroon.org 
Website: www.lukmefcameroon.org
Mission: To Promote Peace, Nonviolence, Social Justice and Sustainable Development in Cameroon