How to rectify the Kenyan Elections going forward

In close to call elections, Kenyan Elections will always have problems as long as a manual system still exists. The reason 2002 elections were seen as free and fair is because the gap between Mwai Kibaki and Uhuru Kenyatta was so huge that no amount of rigging could bridge it. However, in elections where the leading candidate is leading by less than 1 million votes, the Kenyan Elections with its manual components will always present opportunities for the incumbency to do all within its power to enforce its wish on the will of the majority.

Technology was introduced in the Kenyan Elections after 2007 when Kriegler commission suggested for a results transmission system that would allow for all Kenyans to monitor the results as they streamed in from all polling stations, a beautiful suggestion that worked well during the 2010 referendum and all bi-elections prior to and post 2013 general elections. The transmission of results would however terribly fail during the 2013 general elections, with IEBC admitting that the transmission system could have been hacked, forcing the commission to revert to the manual tallying.

In 2017 general elections, although the transmission system hasn’t failed, there have been claims that the system has been compromised. When looking closer at the data for presidential results, this claim seem to be true as the difference between the two leading candidates appear to be dictated by a certain algorithm, something a Facebook user calling himself Charles Karanja claimed to be impossible if result streamed in from random polling stations across the country. The belief that the transmission technology was compromised led Raila Odinga to reject the results, and as I write this, the verification of the presidential results has gone back to the manual tallying system through collation of 40,883 form 34As.

Given that the same political players have previously had issues with manual transmission and tallying of results, do not expect the losing candidate to accept the results arrived at after all form 34As have been collated. Already, rumours are being spread online to the effect that there are people stationed in certain hotels whose main work is to doctor form 34As so that the figures in the forms can match the figures already transmitted.

In our Kenyan Elections therefore, we are doomed if we use technology, and doomed if we don’t. So, what’s the way forward? This is the same question asked by Bobby Mwirigi on his Twitter timeline.

Going forward, what can we do to sort this out? We can’t wait another 5 years them expect IEBC to be on top of it’s game at the last minute.

— igiriwM@ (@Mwirigi) August 9, 2017

Following the tweet, three broader suggestions emerged namely: thorough audit of the entire electoral process to find where we have always gone wrong in each election with an aim of rectifying them, disbanding the entire electoral commission and start a fresh, and lastly adoption of 100% electoral technology. It is 100% adoption of electoral technology, which was suggested by me, that I think can totally and completely rectify the Kenyan Elections going forward.

The technology in place today allows for voter registration, voter identification, and being used as first but backup transmission of results. The actual voting process itself is still manual and the official results are still manually transmitted. These two manual components should be done away with, although the hard copy versions of results per polling station can still be produced at the polling stations for the purposes of court cases.

In a 100% technological electoral process, this is the set up I would recommend. In any given polling station, a device like the KIEMS will not only be used to identify the voter, but also be available to present a virtual ballot paper to the voter. The voter would therefore walk into a polling station, press a finger print scanner to be identified, and immediately after be presented with a screen showing all presidential candidates. The voter will then touch the presidential candidate he or she intends to vote for, the vote is validated, then the next screen shows him or her the candidates for the next elective positions. The process continues until the voter has voted for all candidates he/she wants.

In the meantime and in the same polling station, there is a screen that displays the tally of the results as the voting progresses. When a voter walks into a station, he or she will check on the screen and see all the votes that the presidential candidates have so far garnered in that particular polling station. The agents of the candidates will also know, at that particular moment, the votes each candidate has garnered.

Let us assume that the voter in question is the first voter in that polling station, then on display will be 0 votes for all candidates. After the voter has voted, he or she will look up on the screen and check that his/her preferred candidate now has 1 vote, all other candidates still have 0 votes, and total votes cast is 1. Immediately the voter finishes the voting, the tallying technology in the polling station simultaneously sends its data to the constituency, county, and national tallying centers.

At the end of the voting in a particular polling station, the agents together with presiding officer will counter check their local tally against that polling station’s tally as recorded in the constituency, county and national tallying centres. Once every agent and IEBC official is satisfied that all data is correct in all tallying centres, they will print out their local tally, sign the form, stamp it, and put it in a safe envelop ready to be delivered to the national IEBC office. Of course each agent will also have their copies of the signed hard copy results.

The above system as described immediately does away with the possibility of hacking, as each election official inside the polling station can monitor the progress of election as it happens. The voter too will also know that by the time he/she got into the polling station, the screen displayed that candidate X had 35 votes and when he/she left the station, that candidate had 36 votes.

The immediate setback one would see from the above system is the lack of secrecy, and possibility of tilting the ballot in favor of the leading candidate in that polling station. To work around secrecy, the polling station should allow only one voter insider the station at a time, and the election officials inside the polling station, including the observers, security, and agents, must be people from far away regions who have no idea who the voters in that polling station are.

To work around the possibility of tilting the ballot in favor of the leading candidate in that polling station, voter education should thoroughly be carried out. For example if you are a voter in Murang’a and your preferred candidate is Raila Odinga, the fact that in your polling station Uhuru Kenyatta has already scored 509 votes against Raila’s 11 votes shouldn’t discourage you, as you will be educated that possibly Raila is already leading in other polling stations in other regions of the country, so there is no way your vote in your polling station shall end up being a useless vote.

The adoption of 100% technology in Kenyan Elections is not a pipe dream, as the KIEMS has already demonstrated that it can be trusted with both voter registration and identification. Going a step further to incorporate a virtual ballot paper in KIEMS should be a walk in the pack for the software engineers that programmed them.

The next hurdle is transmission of results of every polling station, including those in far flung areas. In the suggested system, there will be no reason to transmit scanned documents, so 2G network should suffice. Over 90% of the country is already covered by 2G. In addition, the Mobile Network Operators have been paying 0.5% of their revenues to Universal Service Fund, a fund set up to help erect networks in places where the MNOs would find no economic reasons to set up such networks.

If there will be no network at all in certain polling stations by 2022 (which should be very few and in very remote places), then the micro computers together with the KIEMS that will be handling the tallying in that polling station, which will be encrypted and write only, will be taken to constituency tallying center from where they will transmit their results. This will however happen after the polling station has printed the results of that polling station and the agents of the candidates have signed the printed form.

A huge advantage adoption 100% technology in Kenyan Elections will offer is lowering the cost of the elections. In the 2017 elections for instance, the country spent more than Kshs 2 billion in printing of the ballot papers. Given that the country has already acquired the KIEMS kits, the 2022 elections should be very affordable if the suggestions herein are adopted. A tiny 21 inches screen for displaying the results per polling station shouldn’t be that expensive.

If there are loopholes in the system suggested above, feel free to point them out in the comments section below.