Home » Elections » Coalition Trends » Historical Voting Patterns and How they Will Impact on Kenya’s Presidential Outcome 2013

Historical Voting Patterns and How they Will Impact on Kenya’s Presidential Outcome 2013


The political campaigns are in high gear after party primaries were completed on 21st January 2013. Every presidential candidate is promising his/her supporters that he/she will win in the first round regardless of the trends in recent opinion polls. It is obvious that the presidential race is a battle between Raila Odinga/Kalonzo Musyoka of the Coalition for Reform and Democracy (CORD) and Uhuru Kenyatta/William Ruto of the Jubilee Coalition. According to both Ipsos Synovate and Infotrack polls, Raila Odinga has consistently stood out as the front runner as Uhuru Kenyatta has continued to close the gap. Uhuru’s fortunes have risen partly because of selecting William Ruto as his running mate.

Despite previous prediction of an easy win for Raila Odinga, it is now evident that he can as well lose the coming election to the Jubilee Coalition candidate. My argument is premised on one key factor: voter turnout. I took the liberty to analyze the voting patterns by looking at voter turnout per region and ethnic-based alliances in the last five national electoral events which include: elections in 1997, 2002, 2007 and referenda in 2005 and 2010. I excluded the 1992 elections because there was no adequate data on voter turnout. I included the two referenda because the campaigns for and against the respective draft constitutions were largely used as future political platforms.

Voter Turnout Since 1997

The table below shows how different provinces faired in terms of turnout in the last 5 electoral events.




















Rift Valley




























North Eastern




























Central, Rift Valley, Nyanza and Eastern have registered above average voter turnouts. Of course, I am being generous to Eastern which is almost a point less than the national average. But I have to emphasize that Central and Rift Valley have consistently registered high performance throughout. We also see Nyanza performing much better than before in the 2007 elections and 2010 referendum.

The other provinces have generally been performing below average with very few outlier instances where they impress. For example, in 1997 and 2002, voter turnout in Western was same as national average. In the 2010 referendum, Nairobi hit the 70% turnout just below the national average while in 2002, North Eastern was at 57% just above 56% national average.

Voting Patterns and Lessons Since 1997

In Kenya, votes are largely cast on the basis of ethnicity. The various provinces represent the main ethnic blocs when it comes to electoral politics. Voters tend to vote for a candidate from respective ethnic blocs. Further, elections are won or lost depending on the strength of ethnic alliances that are forged during campaigns. The turnout is also high where leading ethnic blocs perceive the elections or referendum as having high stakes. For instance, Kibaki’s legacy and re-election must have motivated large turnouts in the 2005, 2007 and 2010 referenda and elections. The same can be said of Nyanza province in 2007 and 2010 where Raila Odinga had established credibility to win elections and also fomenting his constitutional reform credentials.

What is in store for the Jubilee Alliance? In 1997, Rift Valley voted for Moi as Central went to Kibaki. Similarly, in 2005, 2007 and 2010, Central Kenya overwhelmingly voted on one side while Rift Valley voted on the opposite side. It is only in 2002, when Kibaki (Central Kenya favourite candidate) and Uhuru (Rift Valley favourite candidate) shared votes across the two regions. Nonetheless, Kibaki received 68% and 43% in Central and Rift Valley respectively while Uhuru received 30% in Central and 53% in Rift Valley.

So 2013 must be an interesting year for both Central and Rift Valley. Thanks to the International Criminal Court cases against Uhuru Kenyatta and William Ruto. In 2013, Central and Rift Valley blocs, under the Jubilee Alliance will be voting on one side ever since the first multi-party elections in 1992. The two blocs have the highest number of voters registered so far – Central with 2.2 million and Rift Valley with 3.4 million. The two have also recorded the highest voter turnout compared to the rest.

What legacy does CORD have in terms of voting patterns? Except for Nyanza province, the voting patterns of Eastern and Western are checkered. Often, when we talk of Eastern, we are thinking of the Kamba bloc. However, it is important to note that a significant part of Eastern has a diverse population constituting Marsabit, Isiolo, Meru, Tharaka – Nithi and Embu Counties. While Marbit and Isiolo counties can be considered swing counties, Meru, Tharaka-Nithi and Embu counties have always voted with the Central bloc. Therefore, CORD can only count on Kalonzo’s Kamba votes and not the rest of Eastern.

So what happed to Eastern bloc votes when Kalozo took a position? In 1997, Kalonzo was in Kanu and Moi ended up receiving 36% of the Eastern bloc (This could have been Kamba votes and some of Marsabit and Isiolo). Charity Ngilu, running as a presidential candidate, received 34% (not bad considering that this was her first shot at the presidency). Kibaki (Central) received 29% of the votes which I think was mostly from Meru, Tharaka and Embu counties. In 2002, Kalonzo and Ngilu were on the same side of history which saw Kibaki carry Eastern with 72%. During the 2005 referendum, Kalonzo led the ‘No’ campaign while ‘Ngilu’ supported the ‘Yes’ campaign. Consequently, Eastern was divided in the middle with 49.5% voting ‘No’ and 51% voting ‘Yes.’ In this case, Ngilu can be said to be Kalonzo’s equalizer – by plucking off a few votes for the opposition. Come 2007, Kalonzo plunged himself into the presidential race while Ngilu supported Raila Odinga. He ended up getting 44 % of the Eastern vote compared to Kibaki’s 50%. Raila’s returns were insignificant at 5%; perhaps underscoring Ngilu’s outspent political capital in Eastern (we wait to see what happens in 2013). During the 2010 referendum, both Kalonzo and Ngilu supported the draft Constitution. 56% of Easterners voted “Yes” as 43% voted “No” Many analysts have concluded that it is because of Kalonzo’s middle ground position early in the campaigns that led to the below expectation performance. The strong “No” showing has also been attributed to the role of the church in the Eastern province.

You can conclude from these numbers that Kalonzo’s supremacy in Eastern is confined to the Kamba bloc which currently has approximately 1.1 million voters. Highest ratings for him came only when he joined Ngilu, Raila and Kibaki to vote against Moi’s choice, Uhuru Kenyatta, in 2002. He has also performed better when at the forefront of national politics as a presidential candidate. In 2013 elections, he is a deputy presidential candidate in CORD but he faces opposition in his backyard from Ngilu’s Jubilee Alliance. Like in the 2005 referendum, Ngilu might just pluck away some important votes from the 1.1 million Kamba votes. Eastern is always fourth when it comes to voter turnout. This does not auger well for CORD if it aims at winning in the first even the second the round.

Western Province has somewhat been referred to as swing bloc since 1997. Even when Wamalwa Kijana (a Westerner) was running for the presidency, Western could only give him 49% while supporting Moi with a good 44%. In 2002, the Western bloc rallied behind Kibaki giving him 76% of the votes. During this period, most of the Western elites were more or less on one side of history except Musalia Mudavadi. Musalia Mudavadi, who was Kenyatta’s running in 2002, could only marshal 22% of the votes for his ticket. In 2007, Raila Odinga selected Musalia Mudavadi as his running mate resulting in 66% of votes being cast for Raila Odinga. In 2005, 60% of the region opposed the draft Constitution while a good 40% supported it. What sway did the then Vice-President Moody Awori have on the referendum results? I will have to inquire further. And finally, in 2010 referendum, 84% of Westerners endorsed the new Constitution. Again, most political elites in the region were on the same side of history.

Western has been enticed by a vice-presidential position. In the two instances, the ticket lost (incidentally, it the two cases, it was Musalia Mudavadi being selected. Still debatable as to whether ODM lost in 2007). Western has delivered a majority of the votes to a single coalition when the political elites are united either by choice or circumstance. As it stands today, the entry of Musalia Mudavadi into the presidential race and having Wetangula as a principal in CORD will divide the Western bloc. The region has only 1.4 million voters and records an average of 59% voter turnout. Like Eastern, CORD has to work harder to improve on its numbers in Western for a win in the first even the second round.

Nyanza has historically stood by its man! In 1997, 57% voted for Raila Odinga. The 24% that voted for Moi was mostly the Kisii bloc. In 2002, as the Kisii bloc supported Nyachae with 30% of the Nyanza votes, 61% voted for Kibaki (whom Raila endorsed). In 2007, 82% stood out for Raila Odinga as a bloc. Similarly, Nyanza voted overwhelmingly with the winning teams in the 2005 and 2010 referenda.

Solid support for CORD in Nyanza is drawn from the Luo Nyanza. Since the departure of Nyachae from national politics, it appears the Kisii bloc has warmed up to Raila Odinga. However, it is still a swing bloc. With 1.9 million votes and consistent good turnout, CORD has an upper hand in Nyanza.

The Remaining Regions

Nairobi, Coast and North Eastern are regions to watch. The regions have a combined 3.4 million votes. In other words, their combined total is equivalent to Rift Valley (which is a Jubilee stronghold!). None of the three leading presidential candidates is an indigene of the three regions. In terms of the popularity of the alliances in the three regions, the Ipsos Synovate poll of 14th and 25th January 2013 shows Nairobi leaning towards CORD. But the support is not that convincing. Coast is also leaning towards CORD but slightly stronger than Nairobi. North Eastern is unpredictable for now. Historically, these provinces have registered an average turnout of 50% and below. It will become clearer as we approach the elections as where exactly these regions are leaning.

What is probable based on the above trends?

Jubilee Alliance has an upper hand in terms of number of registered voters and expected voter turnout within their strongholds. With a total of 5.6 million voters (Rift Valley and Central) within these blocs, a 70% turnout will mean 3.92 million voters of which 70% (2,744,000) are likely to vote for Jubilee.

CORD Alliance bastions Eastern, Western and Nyanza have combined 5.5 million voters. Average turnout based on history will be around 63% (3.47 million). Taking good turnout for CORD in Nyanza and a breakaway for the Jubilee and Amani coalitions in Eastern and Western, CORD is likely to net 63% (2.2 million).

None of the Coalition can win in the first round.

CORD has toughest challenge. It has to register a historical high voter turnout of not less than 80% in its strongholds, ensure over above 70% turnout in Eastern and Western and win convincingly with over 60% in Nairobi, Coast and North Eastern if it has to improve its chances of forming the next government.

The Western bloc will determine who becomes the next president after the second round. Both CORD and Jubilee have to tread carefully with Mudavadi.


5 thoughts on “Historical Voting Patterns and How they Will Impact on Kenya’s Presidential Outcome 2013

  1. Thia is a wonderful analysis. it is very important that the politicians from cord strongholds convinnce voters to turnou in large numbers for CORD to take upmleadership

  2. I was reviewing the article, and I am happy it provided an indicator of the likely winner. Thanks for your comments. I will provide some latest updates on why some of the assumptions did not hold.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s