Latest predictions still show a parliament with no overall majority, though I'm sure this will shift. This is based on further reviews of candidate lists and regional/county polls, plus a few broader trends. Azimio is now slightly the larger coalition by 142-137, mainly due to my further downgrading ANC's performance in western and UDA's in north-eastern. 

The two coalitions have not succeeded in putting up candidates in much of their opponents' heartlands. Setting aside true independent candidates and non-aligned parties (who could still have secret deals), 39 of 290 constituency seats (13%) have already been conceded, as there is no party candidate from the other coalition standing. 

Azimio candidates (or independents) have already won 23 seats:

• 17 seats in Luo Nyanza
• 2 seats in Mandera
• 1 seat in Wajir
• 2 seats in Garissa
• 1 seat (Wundanyi) in Taita-Taveta.

Kenya Kwanza candidates (or independents) have already won 16 seats:

• UDA  - 16 seats in the Kalenjin Rift.

Most nominal independents are defectors or losers from party primaries from the local majority party, so 90% will return to or align with with their "home" party once elected. 
My best guess as to the share of the National Assembly (349 seats in total) between the parties is now as follows

UDA will certainly be the largest party, barring an extraordinary upset, and will dominate its smaller coalition partners. Azimio will be a six-way coalition but dominated by ODM. Jubilee's chances of the 30+ seats I've predicted depend on their performance in Mount Kenya (at best 10 seats) and the north and north east (circa 7). Elsewhere the party has few strong candidates and only patchy chances of success (for example in Taita-Taveta, Narok, Kisii). They have few safe seats, so will be the big "swing" variable in the house - a candidate vote just 5-10% smaller than I've predicted could leave them with few to no Mount Kenya seats and Kenya Kwanza with a working majority.

But for now, Kalonzo Musyoka's Wiper MPs - in Azimio but with serious reservations - could hold the balance of power.

A political history of southern central province (Kiambu, Thika and Murang'a) though the history of their elected representatives since 1957.

Just an incremental update, with a refreshed equal-vote county map, a few more Mizani opinion polls and articles, and deep dives into the IEBC's long-awaited all-seats candidate listing, gazetted 13 July but backdated to 1 July, which turned up some surprises.

Overall, no major change, still predicting:
1) a narrow 52-48 win for Ruto all other things being equal. Since my ethnically-driven model is predictive (educated guesswork) not statistically based, it has no margin of error (its either correct or not, we will find out). 
2) Azimio to take 24 governorships to Kenya Kwanza's 21, with two (nominal) independents 
3) In the Assembly, it's on a knife edge, with (among elected constituency MPs) 143 estimated for KK, 137 for Azimio and 11 independents/unknowns. I've not done women representatives in the assembly or the senators yet.

​So, all still to play for. Any one of a dozen factors could change those results. Though I note a little more confidence in the KK campaign Twitter feed recently and their language is moving gradually towards an assumption rather than an assertion of victory. Still, as Kenyans like to vote for winners, that itself is a campaign move - building a bandwagon in the media is a highly-paid dark art.

I've also updated the equal-vote (iso-suffragiem?) map with a couple of changes. I've enlarged the definition of "large majority" to 2-1 (66% to 33% or more) and partly as a result I've moved three counties from strong Raila to narrow Raila - Isiolo, Nyamira and Nairobi. In contrast I've reviewed the candidates and polls in Kakamega and concluded that Azimio will win the majority and consequently called it for Raila.

So, for me Nairobi, Kakamega and Kajiado are now key in deciding the result (as well as turnout of course).

​Kajiado I've reviewed again and still consider too close to call from here. I briefly considered whether Kirinyaga could be considered "narrow Ruto" but I think still not. The other Mount Kenya and Kalenjin counties look pretty safe to me right now. But that's the great danger of making public predictions - after the event they are either obvious if correct (even if no-one else said them at the time) or ridiculous and embarrassing if wrong. Of course, I wont be right. Its impossible. But I hope I'll be somewhere near...and if not, well, maybe I will cultivate my shamba next election. Meanwhile, please enjoy.


Disclaimer: its obvious but I'll say it again - no payments, no promises made by anyone. 

All Kenyan counties are not created equal. In the August 2022 Presidential election, numbers matter. I've therefore crafted what I think is Kenya's first equal-vote map of the 47 counties, based on the final registration figures from the IEBC.

This first graphic has been my main presentation format for the counties so far. This is my latest update,  with recent events, polls and predictions incorporated and now with Azimio (Odinga) in royal blue. I've also incorporated learnings from 20 plus county level opinion polls conducted by Mizani since March (https://mizaniafrica.com) which - while not perfect - give a better insight at a granular level to what's happening in under-served counties than any other source. It suggests the race is still neck and neck. Specifically, I've now rerated Kakamega as "in the balance", and Narok as small majority not large majority Ruto. I've kept all ex-Central Province as 'strong Ruto', despite gains made by Karua. But if I'm wrong (some experts have argued Kirinyaga could go Raila) its a decisive mistake and changes the overall result. Although the pro-Ruto PAA party has made gains in Kilifi for Ruto, I've kept the county in the 'Strong Raila' camp for now. I've also clarified that large (for me) means 60-40% or higher for the winning candidate.

But I now have a new presentation format to share for the first time. This new equal-vote map is still not perfect (its impossible to show the northern counties in any way relative to their real locations because of their huge size and tiny population) but it is striking nonetheless. The importance of a relatively few counties (Nairobi, Kiambu, Nakuru Kakamega, Machakos, Meru) in delivering the absolute numbers both parties need to win is clear. ​

One square is roughly 100,000 registered votes.
​
Nairobi's 2.4 million voters are probably the key to victory now, but the messages are confusing here. UDA candidates are doing well individually and may take nearly half Nairobi's parliamentary seats, while (controversial) UDA gubernatorial candidate Johnson Sakaja is leading too. However, opinion polls still show Raila as winning Nairobi by nearly 2-1 presidentially, putting it firmly in the 'strong Raila' bucket. In my opinion, Raila will win but it will be closer than that. Ruto's "its the economy", anti-privilege and class-based campaign has appealed directly to urban and peri-urban poor, and my suspicion is that these voters are less easily captured by telephone surveys and more likely to be fearful of pollsters if they are. We shall see!