Uhuru contributed roughly half of Jubilee votes

[RV Pundit]

• Tweet

1) Central - biggest base - had roughly 3m registered votes (with half from Kiambu leading at 1.6M) - although kiambu has more than 20% being non-kikuyus -  we are talking about 2.7m.
2) Meru-Tharaka-embu - is 1.3M registered votes with Meru at 0.7m, Embu 0.3m and Tharaka (0.25).
3) Nakuru-Laikipia-Kikuyu diaspora in rv -is about 0.9m Nakuru roughly 0.55m of the 1m registered votes, about 150k from Laikipia, about 50K from Uasin Gishu and maybe 150k elsewhere with Kajiado at 120k.
4) Nairobi - GEMA votes are roughly 30% of the 2.25m - 0.7m registered votes.
5) Coast - Lamu about 50k and maybe another 50k elsewhere in Mombasa and such areas.

So generously we are talking about  2.7+1.3+0.9+0.7+0.1=5.7M registered votes.

If you assume 85% turn out (higher than national average of 80%) and assume everyone voted Jubilee (we know more than 10% meru didn't) - that gives you about 4.8M votes - against Jubilee 8.3m - which is 57% of Jubilee vote - roughly 55% if you take out KenyaPlatos who will vote opposition no matter what.

In short Uhuru's GEMA contribute about 4.5m out 8.3M Jubilee votes - slightly more than half - about 55%.

Outside GEMA Uhuru is a pretty hard sell - maybe in Gusii - he can sell on is own - elsewhere Ruto did the donkey work.

[Nowayhaha]

https://www.africaresearchinstitute.org/newsite/blog/kenyas-2017-elections-preliminary-results-analysis-county/

Kenya’s 2017 presidential elections: a preliminary results analysis, by county

Opinion polling prior to Kenya’s August 2017 presidential elections suggested a dead heat between the incumbent, Uhuru Kenyatta of the Jubilee Party (JP), and his main opponent, Raila Odinga of the National Super Alliance (NASA). Election results released by the Independent Electoral and Boundaries Commission (IEBC) show a comfortable victory for Kenyatta, who took 54.2% to Odinga’s 44.7%. Both presidential candidates increased their share of the vote over that of the last election in 2013, when Odinga took 43.7% and Kenyatta won a narrower majority of 50.5%.

Alleging the elections were rigged, Odinga is contesting the results at the Supreme Court.  The Court is required to hear the case and deliver a verdict by 1 September.

NASA has yet to publicly produce compelling evidence of fraud.  A partial analysis of the constituency tabulation results forms (Form 34B), shows a scattering of mathematical errors and inconsistencies.  These do not appear to be of the scale required to overturn Kenyatta’s margin of victory.  Further analysis of these forms, and their counterparts issued at the polling station (Form 34A), may indicate localised electoral fraud, mismanagement, or human error.

If the IEBC figures are broadly correct, what statistical trends does the data reveal?  What may explain the scale of Kenyatta’s victory?  A preliminary analysis of the election data available at county level suggests four trends:

Kenyatta increased his share of the vote in a larger number of the most populous counties, whether or not these areas were JP strongholds.
Across the country, Kenyatta increased his share of the vote in more counties than Odinga.
Kenyatta generally achieved higher turnout in his areas of core support; turnout figures for Odinga were weaker in key areas of his support, notably the coast.
Kenyatta was more consistent than Odinga in picking up support in areas where a third party candidate had performed well in 2013. There were no strong third party candidates for president in 2017.
 

Kenyatta increased his share of the vote in more of the most populous counties

Kenya’s 47 counties vary considerably in population.  In 2017, 11 counties accounted for more than seven million of the valid votes cast, nearly half the total cast nationwide.  More than 400,000 valid votes were cast in each of these counties, which span Kenya’s political spectrum.  They include big cities and more rural areas and are spread through western, central and eastern areas of the country.  They are nearly evenly divided between areas of support for Kenyatta/JP and areas of support for Odinga/NASA.

Compared to the 2013 election, Kenyatta increased his share of the vote in all but one of these counties, as shown in Table 1.  Even a modest increase in vote share is significant given the number of votes on offer in these areas.  Meanwhile, Odinga’s share of the vote increased in only six of these counties, and decreased in five counties.

Table 1: Counties with more than 400,000 votes cast, and swing in vote share by candidate, 2017 election

 
County

Valid votes cast   Swing to Kenyatta (%)   Swing to Odinga (%)
Nairobi City    1,635,416   1.34   1.58
Kiambu    985,417   1.92   -0.93
Nakuru    754,537   3.84   -2.48
Kakamega    555,581   8.84   22.56
Meru    543,580   -1.48   2.67
Murang’a    508,808   1.43   -0.65
Machakos    471,112   7.92   -6.11
Kisumu    439,423   0.43   0.76
Bungoma    421,060   17.76   14.46
Kisii    405,872   15.43   -13.43
Homa Bay    402,836   0.25   0.08
 

Kenyatta increased his share of the vote in more counties than Odinga

Kenyatta increased his share of the vote in 41 of 47 counties, and lost ground in six.  Odinga increased his share of the vote in only 16 counties, and did worse in the remaining 31.

In nine counties, Kenyatta achieved a 9.5% or better improvement in his share of the vote, as shown in Table 2.  Most of these counties – six of nine – were won by Odinga.  The results suggest Kenyatta was able to make in-roads in areas where he performed poorly in 2013.

Table 2: Kenyatta’s share of the vote, 2013 versus 2017, and swing in share of the vote

 
County

2013
(%)

2017
(%)

Swing
(%)

Marsabit   47.35   83.63   36.28
Nyamira   29.84   52.07   22.24
Bungoma   12.44   30.20   17.76
Kisii   27.78   43.22   15.43
Turkana   29.97   44.94   14.97
Taita Taveta   13.33   27.69   14.36
Wajir   39.05   51.20   12.16
Tana River   34.99   46.10   11.12
Kwale   14.15   23.71   9.56
In comparison, as seen in Table 3, Odinga achieved a similar or better swing in his share of the vote in only four counties: Vihiga, Kakamega, Bungoma and West Pokot.  In contrast to the shift in support for Kenyatta, in three of these four counties Odinga was already strong, having performed well in 2013.  Only in West Pokot was Odinga able to increase his share of the vote at Kenyatta’s expense.

Odinga’s improved performance in these counties was largely at the expense of third party candidates.  Crucially, Kenyatta was also able to increase his share of the vote in three of these four counties, again largely at the expense of third party candidates.

Table 3: Odinga’s share of the vote in 2013 and 2017, swing in share for Odinga and Kenyatta, and swing in share of the vote for third party candidates

County   Odinga share of the vote, 2013
(%)

Odinga share of the vote, 2017
(%)

Swing for Odinga
(%)

Swing for Kenyatta (%)   Swing against third parties
(%)

Vihiga   47.03   89.64   42.62   7.61   -50.23
Kakamega   64.80   87.36   22.56   8.84   -31.40
Bungoma   53.64   68.10   14.46   17.76   -32.23
West Pokot   23.10   34.44   11.34   -9.04   -2.30
 

Higher turnout in areas of Kenyatta’s core support

18 counties reported voter turnout of 80% or more. Kenyatta triumphed in 14 of these 18 counties.

Of the five counties with turnout of 70% or lower, Odinga won four of them, including, critically, the highly populated coastal counties of Kilifi and Mombasa.  Odinga/NASA voters appear not to have been as effectively mobilised as supporters of Kenyatta/JP.  In coastal counties such as Kwale and Mombasa, which voted heavily for Odinga, turnout fell from the levels recorded in 2013.

Table 4: Comparison in turnout in selected coastal counties, 2013 versus 2017

 
County

Turnout, 2013 (%)   Turnout, 2017 (%)
Kwale   73.59   67.08
Mombasa   66.62   60.49
Kilifi   64.91   66.23
National   85.91   79.37
 

Kenyatta took more of the 2013 third party vote than Odinga

In 2013, third party candidates took 5% of the vote or more in 12 counties, as shown in Table 5.  In 2017, only one county, Isiolo, saw a third party candidate take more than 5% of the vote.  Discounting Isiolo, Kenyatta increased his share of the vote in all 11 other counties, by an average 8.2%.

Odinga’s performance was more mixed.  His share of the vote decreased in four of 12 counties.  In Vihiga, Bungoma and Kakamega, most, but not all, of the support for 2013 third party candidate Musalia Mudavadi, now aligned with NASA, did transfer to Odinga.

Table 5: Third party share of the vote, 2013 and 2017, and swing to Kenyatta and Odinga, 2017

County   2013
(%)

2017
(%)

Swing to Kenyatta
(%)

Swing to Odinga (%)
Vihiga   51.44   1.21   7.61   42.62
Bungoma   33.93   1.70   17.76   14.46
Kakamega   32.54   1.13   8.84   22.56
Trans Nzoia   14.74   1.15   6.42   7.17
Isiolo   14.46   15.57   -6.43   5.32
Wajir   11.09   4.59   12.16   -5.65
Busia   9.75   0.73   8.71   0.30
Nandi   8.92   0.62   4.51   3.79
Lamu   6.65   1.75   8.32   -3.42
Garissa   5.46   3.42   2.70   -0.68
Mombasa   5.40   1.05   5.04   -0.68
 

Preliminary conclusions

Analysis of these election results suggest Kenyatta and the JP ran a better campaign, even in areas historically opposed to the party and its candidate. JP better mobilised voters in areas of both strength and relative weakness, and focused on areas with the greatest number of registered voters.  By increasing his support in almost all areas of the country, Kenyatta appears to have demonstrated a broader national appeal to Kenyans.  These trends suggest that voters did not only consider ethnic or tribal identity in casting their votes, as the swing towards Kenyatta cannot solely be explained by shifts in demography.

In contrast, Odinga and NASA failed to mobilise more voters in their areas of strongest support. Despite potentially favourable economic and political circumstances, the opposition was apparently unable to significantly increase its appeal to Kenyans in key electoral districts across the country. Odinga failed to win over enough voters who had supported third party candidates in 2013. Collectively, even if Odinga increased his share of the vote relative to 2013, it appears he underperformed in too many areas of the country to win in 2017.

N.B. All data was sourced from the IEBC’s public results database.  All findings are preliminary and subject to confirmation of the final figures and cross-referencing against original Forms 34A and 34B.

Aly Verjee is a visiting expert at the U.S. Institute of Peace and a fellow of the Rift Valley Institute.

[Njuri Ncheke]

1) Central - biggest base - had roughly 3m registered votes (with half from Kiambu leading at 1.6M) - although kiambu has more than 20% being non-kikuyus -  we are talking about 2.7m.
2) Meru-Tharaka-embu - is 1.3M registered votes with Meru at 0.7m, Embu 0.3m and Tharaka (0.25).
3) Nakuru-Laikipia-Kikuyu diaspora in rv -is about 0.9m Nakuru roughly 0.55m of the 1m registered votes, about 150k from Laikipia, about 50K from Uasin Gishu and maybe 150k elsewhere with Kajiado at 120k.
4) Nairobi - GEMA votes are roughly 30% of the 2.25m - 0.7m registered votes.
5) Coast - Lamu about 50k and maybe another 50k elsewhere in Mombasa and such areas.

So generously we are talking about  2.7+1.3+0.9+0.7+0.1=5.7M registered votes.

If you assume 85% turn out (higher than national average of 80%) and assume everyone voted Jubilee (we know more than 10% meru didn't) - that gives you about 4.8M votes - against Jubilee 8.3m - which is 57% of Jubilee vote - roughly 55% if you take out KenyaPlatos who will vote opposition no matter what.

In short Uhuru's GEMA contribute about 4.5m out 8.3M Jubilee votes - slightly more than half - about 55%.

Outside GEMA Uhuru is a pretty hard sell - maybe in Gusii - he can sell on is own - elsewhere Ruto did the donkey work.

You are good with estimates even though some figures are way off generally respectable data bearing in Mind KNBS hasn't provided any.

[RV Pundit]

Which figures are off.

You are good with estimates even though some figures are way off generally respectable data bearing in Mind KNBS hasn't provided any.

[RV Pundit]

Most of the increment came from Ruto incessant campaigns - and now after Wanjohi was secured in his 2nd term - he wanted Ruto to go slow. It's the ultimate betrayal. Ruto did a lot of donkey work in western, coast, and all over kenya.

https://www.africaresearchinstitute.org/newsite/blog/kenyas-2017-elections-preliminary-results-analysis-county/

Kenya’s 2017 presidential elections: a preliminary results analysis, by county

Opinion polling prior to Kenya’s August 2017 presidential elections suggested a dead heat between the incumbent, Uhuru Kenyatta of the Jubilee Party (JP), and his main opponent, Raila Odinga of the National Super Alliance (NASA). Election results released by the Independent Electoral and Boundaries Commission (IEBC) show a comfortable victory for Kenyatta, who took 54.2% to Odinga’s 44.7%. Both presidential candidates increased their share of the vote over that of the last election in 2013, when Odinga took 43.7% and Kenyatta won a narrower majority of 50.5%.

Alleging the elections were rigged, Odinga is contesting the results at the Supreme Court.  The Court is required to hear the case and deliver a verdict by 1 September.

NASA has yet to publicly produce compelling evidence of fraud.  A partial analysis of the constituency tabulation results forms (Form 34B), shows a scattering of mathematical errors and inconsistencies.  These do not appear to be of the scale required to overturn Kenyatta’s margin of victory.  Further analysis of these forms, and their counterparts issued at the polling station (Form 34A), may indicate localised electoral fraud, mismanagement, or human error.

If the IEBC figures are broadly correct, what statistical trends does the data reveal?  What may explain the scale of Kenyatta’s victory?  A preliminary analysis of the election data available at county level suggests four trends:

Kenyatta increased his share of the vote in a larger number of the most populous counties, whether or not these areas were JP strongholds.
Across the country, Kenyatta increased his share of the vote in more counties than Odinga.
Kenyatta generally achieved higher turnout in his areas of core support; turnout figures for Odinga were weaker in key areas of his support, notably the coast.
Kenyatta was more consistent than Odinga in picking up support in areas where a third party candidate had performed well in 2013. There were no strong third party candidates for president in 2017.
 

Kenyatta increased his share of the vote in more of the most populous counties

Kenya’s 47 counties vary considerably in population.  In 2017, 11 counties accounted for more than seven million of the valid votes cast, nearly half the total cast nationwide.  More than 400,000 valid votes were cast in each of these counties, which span Kenya’s political spectrum.  They include big cities and more rural areas and are spread through western, central and eastern areas of the country.  They are nearly evenly divided between areas of support for Kenyatta/JP and areas of support for Odinga/NASA.

Compared to the 2013 election, Kenyatta increased his share of the vote in all but one of these counties, as shown in Table 1.  Even a modest increase in vote share is significant given the number of votes on offer in these areas.  Meanwhile, Odinga’s share of the vote increased in only six of these counties, and decreased in five counties.

Table 1: Counties with more than 400,000 votes cast, and swing in vote share by candidate, 2017 election

 
County

Valid votes cast   Swing to Kenyatta (%)   Swing to Odinga (%)
Nairobi City    1,635,416   1.34   1.58
Kiambu    985,417   1.92   -0.93
Nakuru    754,537   3.84   -2.48
Kakamega    555,581   8.84   22.56
Meru    543,580   -1.48   2.67
Murang’a    508,808   1.43   -0.65
Machakos    471,112   7.92   -6.11
Kisumu    439,423   0.43   0.76
Bungoma    421,060   17.76   14.46
Kisii    405,872   15.43   -13.43
Homa Bay    402,836   0.25   0.08
 

Kenyatta increased his share of the vote in more counties than Odinga

Kenyatta increased his share of the vote in 41 of 47 counties, and lost ground in six.  Odinga increased his share of the vote in only 16 counties, and did worse in the remaining 31.

In nine counties, Kenyatta achieved a 9.5% or better improvement in his share of the vote, as shown in Table 2.  Most of these counties – six of nine – were won by Odinga.  The results suggest Kenyatta was able to make in-roads in areas where he performed poorly in 2013.

Table 2: Kenyatta’s share of the vote, 2013 versus 2017, and swing in share of the vote

 
County

2013
(%)

2017
(%)

Swing
(%)

Marsabit   47.35   83.63   36.28
Nyamira   29.84   52.07   22.24
Bungoma   12.44   30.20   17.76
Kisii   27.78   43.22   15.43
Turkana   29.97   44.94   14.97
Taita Taveta   13.33   27.69   14.36
Wajir   39.05   51.20   12.16
Tana River   34.99   46.10   11.12
Kwale   14.15   23.71   9.56
In comparison, as seen in Table 3, Odinga achieved a similar or better swing in his share of the vote in only four counties: Vihiga, Kakamega, Bungoma and West Pokot.  In contrast to the shift in support for Kenyatta, in three of these four counties Odinga was already strong, having performed well in 2013.  Only in West Pokot was Odinga able to increase his share of the vote at Kenyatta’s expense.

Odinga’s improved performance in these counties was largely at the expense of third party candidates.  Crucially, Kenyatta was also able to increase his share of the vote in three of these four counties, again largely at the expense of third party candidates.

Table 3: Odinga’s share of the vote in 2013 and 2017, swing in share for Odinga and Kenyatta, and swing in share of the vote for third party candidates

County   Odinga share of the vote, 2013
(%)

Odinga share of the vote, 2017
(%)

Swing for Odinga
(%)

Swing for Kenyatta (%)   Swing against third parties
(%)

Vihiga   47.03   89.64   42.62   7.61   -50.23
Kakamega   64.80   87.36   22.56   8.84   -31.40
Bungoma   53.64   68.10   14.46   17.76   -32.23
West Pokot   23.10   34.44   11.34   -9.04   -2.30
 

Higher turnout in areas of Kenyatta’s core support

18 counties reported voter turnout of 80% or more. Kenyatta triumphed in 14 of these 18 counties.

Of the five counties with turnout of 70% or lower, Odinga won four of them, including, critically, the highly populated coastal counties of Kilifi and Mombasa.  Odinga/NASA voters appear not to have been as effectively mobilised as supporters of Kenyatta/JP.  In coastal counties such as Kwale and Mombasa, which voted heavily for Odinga, turnout fell from the levels recorded in 2013.

Table 4: Comparison in turnout in selected coastal counties, 2013 versus 2017

 
County

Turnout, 2013 (%)   Turnout, 2017 (%)
Kwale   73.59   67.08
Mombasa   66.62   60.49
Kilifi   64.91   66.23
National   85.91   79.37
 

Kenyatta took more of the 2013 third party vote than Odinga

In 2013, third party candidates took 5% of the vote or more in 12 counties, as shown in Table 5.  In 2017, only one county, Isiolo, saw a third party candidate take more than 5% of the vote.  Discounting Isiolo, Kenyatta increased his share of the vote in all 11 other counties, by an average 8.2%.

Odinga’s performance was more mixed.  His share of the vote decreased in four of 12 counties.  In Vihiga, Bungoma and Kakamega, most, but not all, of the support for 2013 third party candidate Musalia Mudavadi, now aligned with NASA, did transfer to Odinga.

Table 5: Third party share of the vote, 2013 and 2017, and swing to Kenyatta and Odinga, 2017

County   2013
(%)

2017
(%)

Swing to Kenyatta
(%)

Swing to Odinga (%)
Vihiga   51.44   1.21   7.61   42.62
Bungoma   33.93   1.70   17.76   14.46
Kakamega   32.54   1.13   8.84   22.56
Trans Nzoia   14.74   1.15   6.42   7.17
Isiolo   14.46   15.57   -6.43   5.32
Wajir   11.09   4.59   12.16   -5.65
Busia   9.75   0.73   8.71   0.30
Nandi   8.92   0.62   4.51   3.79
Lamu   6.65   1.75   8.32   -3.42
Garissa   5.46   3.42   2.70   -0.68
Mombasa   5.40   1.05   5.04   -0.68
 

Preliminary conclusions

Analysis of these election results suggest Kenyatta and the JP ran a better campaign, even in areas historically opposed to the party and its candidate. JP better mobilised voters in areas of both strength and relative weakness, and focused on areas with the greatest number of registered voters.  By increasing his support in almost all areas of the country, Kenyatta appears to have demonstrated a broader national appeal to Kenyans.  These trends suggest that voters did not only consider ethnic or tribal identity in casting their votes, as the swing towards Kenyatta cannot solely be explained by shifts in demography.

In contrast, Odinga and NASA failed to mobilise more voters in their areas of strongest support. Despite potentially favourable economic and political circumstances, the opposition was apparently unable to significantly increase its appeal to Kenyans in key electoral districts across the country. Odinga failed to win over enough voters who had supported third party candidates in 2013. Collectively, even if Odinga increased his share of the vote relative to 2013, it appears he underperformed in too many areas of the country to win in 2017.

N.B. All data was sourced from the IEBC’s public results database.  All findings are preliminary and subject to confirmation of the final figures and cross-referencing against original Forms 34A and 34B.

Aly Verjee is a visiting expert at the U.S. Institute of Peace and a fellow of the Rift Valley Institute.

[Nowayhaha]

Most of the increment came from Ruto incessant campaigns - and now after Wanjohi was secured in his 2nd term - he wanted Ruto to go slow. It's the ultimate betrayal. Ruto did a lot of donkey work in western, coast, and all over kenya.

https://www.africaresearchinstitute.org/newsite/blog/kenyas-2017-elections-preliminary-results-analysis-county/

Kenya’s 2017 presidential elections: a preliminary results analysis, by county

Opinion polling prior to Kenya’s August 2017 presidential elections suggested a dead heat between the incumbent, Uhuru Kenyatta of the Jubilee Party (JP), and his main opponent, Raila Odinga of the National Super Alliance (NASA). Election results released by the Independent Electoral and Boundaries Commission (IEBC) show a comfortable victory for Kenyatta, who took 54.2% to Odinga’s 44.7%. Both presidential candidates increased their share of the vote over that of the last election in 2013, when Odinga took 43.7% and Kenyatta won a narrower majority of 50.5%.

Alleging the elections were rigged, Odinga is contesting the results at the Supreme Court.  The Court is required to hear the case and deliver a verdict by 1 September.

NASA has yet to publicly produce compelling evidence of fraud.  A partial analysis of the constituency tabulation results forms (Form 34B), shows a scattering of mathematical errors and inconsistencies.  These do not appear to be of the scale required to overturn Kenyatta’s margin of victory.  Further analysis of these forms, and their counterparts issued at the polling station (Form 34A), may indicate localised electoral fraud, mismanagement, or human error.

If the IEBC figures are broadly correct, what statistical trends does the data reveal?  What may explain the scale of Kenyatta’s victory?  A preliminary analysis of the election data available at county level suggests four trends:

Kenyatta increased his share of the vote in a larger number of the most populous counties, whether or not these areas were JP strongholds.
Across the country, Kenyatta increased his share of the vote in more counties than Odinga.
Kenyatta generally achieved higher turnout in his areas of core support; turnout figures for Odinga were weaker in key areas of his support, notably the coast.
Kenyatta was more consistent than Odinga in picking up support in areas where a third party candidate had performed well in 2013. There were no strong third party candidates for president in 2017.
 

Kenyatta increased his share of the vote in more of the most populous counties

Kenya’s 47 counties vary considerably in population.  In 2017, 11 counties accounted for more than seven million of the valid votes cast, nearly half the total cast nationwide.  More than 400,000 valid votes were cast in each of these counties, which span Kenya’s political spectrum.  They include big cities and more rural areas and are spread through western, central and eastern areas of the country.  They are nearly evenly divided between areas of support for Kenyatta/JP and areas of support for Odinga/NASA.

Compared to the 2013 election, Kenyatta increased his share of the vote in all but one of these counties, as shown in Table 1.  Even a modest increase in vote share is significant given the number of votes on offer in these areas.  Meanwhile, Odinga’s share of the vote increased in only six of these counties, and decreased in five counties.

Table 1: Counties with more than 400,000 votes cast, and swing in vote share by candidate, 2017 election

 
County

Valid votes cast   Swing to Kenyatta (%)   Swing to Odinga (%)
Nairobi City    1,635,416   1.34   1.58
Kiambu    985,417   1.92   -0.93
Nakuru    754,537   3.84   -2.48
Kakamega    555,581   8.84   22.56
Meru    543,580   -1.48   2.67
Murang’a    508,808   1.43   -0.65
Machakos    471,112   7.92   -6.11
Kisumu    439,423   0.43   0.76
Bungoma    421,060   17.76   14.46
Kisii    405,872   15.43   -13.43
Homa Bay    402,836   0.25   0.08
 

Kenyatta increased his share of the vote in more counties than Odinga

Kenyatta increased his share of the vote in 41 of 47 counties, and lost ground in six.  Odinga increased his share of the vote in only 16 counties, and did worse in the remaining 31.

In nine counties, Kenyatta achieved a 9.5% or better improvement in his share of the vote, as shown in Table 2.  Most of these counties – six of nine – were won by Odinga.  The results suggest Kenyatta was able to make in-roads in areas where he performed poorly in 2013.

Table 2: Kenyatta’s share of the vote, 2013 versus 2017, and swing in share of the vote

 
County

2013
(%)

2017
(%)

Swing
(%)

Marsabit   47.35   83.63   36.28
Nyamira   29.84   52.07   22.24
Bungoma   12.44   30.20   17.76
Kisii   27.78   43.22   15.43
Turkana   29.97   44.94   14.97
Taita Taveta   13.33   27.69   14.36
Wajir   39.05   51.20   12.16
Tana River   34.99   46.10   11.12
Kwale   14.15   23.71   9.56
In comparison, as seen in Table 3, Odinga achieved a similar or better swing in his share of the vote in only four counties: Vihiga, Kakamega, Bungoma and West Pokot.  In contrast to the shift in support for Kenyatta, in three of these four counties Odinga was already strong, having performed well in 2013.  Only in West Pokot was Odinga able to increase his share of the vote at Kenyatta’s expense.

Odinga’s improved performance in these counties was largely at the expense of third party candidates.  Crucially, Kenyatta was also able to increase his share of the vote in three of these four counties, again largely at the expense of third party candidates.

Table 3: Odinga’s share of the vote in 2013 and 2017, swing in share for Odinga and Kenyatta, and swing in share of the vote for third party candidates

County   Odinga share of the vote, 2013
(%)

Odinga share of the vote, 2017
(%)

Swing for Odinga
(%)

Swing for Kenyatta (%)   Swing against third parties
(%)

Vihiga   47.03   89.64   42.62   7.61   -50.23
Kakamega   64.80   87.36   22.56   8.84   -31.40
Bungoma   53.64   68.10   14.46   17.76   -32.23
West Pokot   23.10   34.44   11.34   -9.04   -2.30
 

Higher turnout in areas of Kenyatta’s core support

18 counties reported voter turnout of 80% or more. Kenyatta triumphed in 14 of these 18 counties.

Of the five counties with turnout of 70% or lower, Odinga won four of them, including, critically, the highly populated coastal counties of Kilifi and Mombasa.  Odinga/NASA voters appear not to have been as effectively mobilised as supporters of Kenyatta/JP.  In coastal counties such as Kwale and Mombasa, which voted heavily for Odinga, turnout fell from the levels recorded in 2013.

Table 4: Comparison in turnout in selected coastal counties, 2013 versus 2017

 
County

Turnout, 2013 (%)   Turnout, 2017 (%)
Kwale   73.59   67.08
Mombasa   66.62   60.49
Kilifi   64.91   66.23
National   85.91   79.37
 

Kenyatta took more of the 2013 third party vote than Odinga

In 2013, third party candidates took 5% of the vote or more in 12 counties, as shown in Table 5.  In 2017, only one county, Isiolo, saw a third party candidate take more than 5% of the vote.  Discounting Isiolo, Kenyatta increased his share of the vote in all 11 other counties, by an average 8.2%.

Odinga’s performance was more mixed.  His share of the vote decreased in four of 12 counties.  In Vihiga, Bungoma and Kakamega, most, but not all, of the support for 2013 third party candidate Musalia Mudavadi, now aligned with NASA, did transfer to Odinga.

Table 5: Third party share of the vote, 2013 and 2017, and swing to Kenyatta and Odinga, 2017

County   2013
(%)

2017
(%)

Swing to Kenyatta
(%)

Swing to Odinga (%)
Vihiga   51.44   1.21   7.61   42.62
Bungoma   33.93   1.70   17.76   14.46
Kakamega   32.54   1.13   8.84   22.56
Trans Nzoia   14.74   1.15   6.42   7.17
Isiolo   14.46   15.57   -6.43   5.32
Wajir   11.09   4.59   12.16   -5.65
Busia   9.75   0.73   8.71   0.30
Nandi   8.92   0.62   4.51   3.79
Lamu   6.65   1.75   8.32   -3.42
Garissa   5.46   3.42   2.70   -0.68
Mombasa   5.40   1.05   5.04   -0.68
 

Preliminary conclusions

Analysis of these election results suggest Kenyatta and the JP ran a better campaign, even in areas historically opposed to the party and its candidate. JP better mobilised voters in areas of both strength and relative weakness, and focused on areas with the greatest number of registered voters.  By increasing his support in almost all areas of the country, Kenyatta appears to have demonstrated a broader national appeal to Kenyans.  These trends suggest that voters did not only consider ethnic or tribal identity in casting their votes, as the swing towards Kenyatta cannot solely be explained by shifts in demography.

In contrast, Odinga and NASA failed to mobilise more voters in their areas of strongest support. Despite potentially favourable economic and political circumstances, the opposition was apparently unable to significantly increase its appeal to Kenyans in key electoral districts across the country. Odinga failed to win over enough voters who had supported third party candidates in 2013. Collectively, even if Odinga increased his share of the vote relative to 2013, it appears he underperformed in too many areas of the country to win in 2017.

N.B. All data was sourced from the IEBC’s public results database.  All findings are preliminary and subject to confirmation of the final figures and cross-referencing against original Forms 34A and 34B.

Aly Verjee is a visiting expert at the U.S. Institute of Peace and a fellow of the Rift Valley Institute.

And thats why I insist  looking at the results not from a tribal  perspective but from a County level , One will get the true picture . The improvement in North Eastern , Ukambani, Coast , Western and Kissii was an elaborate  strategy from as  early campaigns as from  2014 to , defections and goodies . Ruto was instrumental in this  .
From Ukambani  Counties  Uhuru Bagged around  180 K  ,in western More than 200K  IN Coast around 290 K .

[Nefertiti]

Fair enough. Add that to Raila 70% of NASA votes. Where does that leave Mobutu? If you do such simplistic math it just expose Mobutu soft underbelly. Oh wait, tribes are dead - now it hustler vs dynasty.

[RV Pundit]

Don't get you. You mean Raila already won GEMA votes? Tribes matters - but only where there is a strong tribal king. We can say Ruto and Raila have their tribes. Will a retiring Uhuru manage to herd GEMA while he is on his way to Ichaweri the verdict is out there - but why would GEMA listen to him - especially if he is egging them on a war with RV - that could see their people suffer.

Don't be suprised that nobody will listen to Uhuru - not even his MP - Moses Kuria.

Ruto has is gema votes - that are committed to Jubile original vision. There are new dynamics every election - we don't know if Kalonzo will manage to keep Kambas again - we don't know if maDVD will be strong - it look likely Weta is going for governorship - we know Matiangi if he quit statecraft could be a new factor in Gusii.

Those are dynamics - they change every election. For GEMA - it the same - 92 till 2002 - Mt kenya east were mostly with Moi - except for Imenti/Embus. GEMA itself were disunited except in 1997 when Kibaki got most of them. Obviously 2007, 2013 and 2017 - they were united.

This 2022 we really don't know how GEMA will pan out - we can predict for now two camp - Pro-Uhuru and Pro-Ruto. There is NO pro-Raila camp. Raila is hoping to inherit Pro-Uhuru as a gift and so is Matiangi, Kalonzo, PK and even Mutua - all these they have been promised by Murathe - the village drunkard.

What is likely to happen - without Uhuru - nobody can really unite GEMA - not even the kikuyus alone - their is no tribal heir acceptable to everyone - PK is dead on arrival - and maybe fly only in Muranga at best -  so we migt see 2 or even 3 camps - each doing their own thing. Kiunjuri group will stick with Ruto and get DPORK. Someone as desperate as Waiguru could go for Raila - and others could arrange harambee house project with Matiangi.

Will Uhuru have the time and money to engage in useless campaign that doesn't benefit him - we will see. If he stay out - Ruto makes mince meat of Pro-Uhuru group by inciting a GEMA poor revolt against their elite. It's already happening...

Fair enough. Add that to Raila 70% of NASA votes. Where does that leave Mobutu? If you do such simplistic math it just expose Mobutu soft underbelly. Oh wait, tribes are dead - now it hustler vs dynasty.

[Njuri Ncheke]

Which figures are off.

You are good with estimates even though some figures are way off generally respectable data bearing in Mind KNBS hasn't provided any.

Isiolo eti Meru&Kikuyu 10% laughable firstly The following are major tribes in isiolo.
1.Borana 27%
2.Turkana  20%
3.Meru 20%
4.Somali 17%
5.Samburu 13%
Others 3%

[RV Pundit]

Isiolo town (south) or county. Isiolo North has lot more Boranas and Somalis. What you got there look like Isiolo south.

Isiolo eti Meru&Kikuyu 10% laughable firstly The following are major tribes in isiolo.
1.Borana 27%
2.Turkana  20%
3.Meru 20%
4.Somali 17%
5.Samburu 13%
Others 3%

[RV Pundit]

NJuri I did research then - but it pretty obvious Borana, Somalis, Samburu,Turkans and Merus - are in that order.
The report ranks the Borana as the most dominant ethnic community in Isiolo county, followed by the Somali, then Samburu.
https://www.the-star.co.ke/counties/eastern/2019-11-13-meru-community-in-isiolo-dismisses-census-report/

[Njuri Ncheke]

Isiolo town (south) or county. Isiolo North has lot more Boranas and Somalis. What you got there look like Isiolo south.

Isiolo eti Meru&Kikuyu 10% laughable firstly The following are major tribes in isiolo.
1.Borana 27%
2.Turkana  20%
3.Meru 20%
4.Somali 17%
5.Samburu 13%
Others 3%

Pundit you confusing, Isiolo North constituency is the one inhabited by those 5 major kabilas, thats where Isiolo town is, Isiolo south constituency is the remote part of Isiolo occupied 50% Borana, 20% samburu (west) 20% somalis (east) 5% Meru 5% Turkanas

[RV Pundit]

I didn't know that - that sounds weird - I guess North has small tip where it connect to meru?. Okay so what the figures for Isiolo as county - and some source please.

Pundit you confusing, Isiolo North constituency is the one inhabited by those 5 major kabilas, thats where Isiolo town is, Isiolo south constituency is the remote part of Isiolo occupied 50% Borana, 20% samburu (west) 20% somalis (east) 5% Meru 5% Turkanas

[Njuri Ncheke]

NJuri I did research then - but it pretty obvious Borana, Somalis, Samburu,Turkans and Merus - are in that order.
The report ranks the Borana as the most dominant ethnic community in Isiolo county, followed by the Somali, then Samburu.
https://www.the-star.co.ke/counties/eastern/2019-11-13-meru-community-in-isiolo-dismisses-census-report/

This so called census report which was doctored and No one knows where it came from was disputed by Merus simply because other 4 kabilas always gang up against Merus. What they do to Merus in isiolo is like what happens to kikuyus in other parts of Kenya It was rumors being spread around as you can see only star reported what Merus had raised. Take it from me, Merus are either 2 or 3. Other Kabilas hate Merus because they dominate business with a choke hold and at same time controlling Land thick and thin, Merus are also capable of defending their interest ruthlessly and engaging in the guerilla warfare like those pastoralist  communities hence the warfare situation in Isiolo ends up a draw no community wins. Thats why I told you Merus would drive you out of RV in just 2 days, we are used to simultaneously attacks from those communities frequently along the boarder front and we drive them back. Mark you they are not armed with bows and arrows like you people but with state of the art automatic rifles and machine guns.

[RV Pundit]

Interesting but if you don't keep cattle - pastoralist will never care about you. That only way to deal with them. Just avoid cattle.Maybe you'll have conflict in drought season as they bring their cows to graze by force in Meru

This so called census report which was doctored and No one knows where it came from was disputed by Merus simply because other 4 kabilas always gang up against Merus. What they do to Merus in isiolo is like what happens to kikuyus in other parts of Kenya It was rumors being spread around as you can see only star reported what Merus had raised. Take it from me, Merus are either 2 or 3. Other Kabilas hate Merus because they dominate business with a choke hold and at same time controlling Land thick and thin, Merus are also capable of defending their interest ruthlessly and engaging in the guerilla warfare like those pastoralist  communities hence the warfare situation in Isiolo ends up a draw no community wins. Thats why I told you Merus would drive you out of RV in just 2 days, we are used to simultaneously attacks from those communities frequently along the boarder front and we drive them back. Mark you they are not armed with bows and arrows like you people but with state of the art automatic rifles and machine guns.

[Njuri Ncheke]

Interesting but if you don't keep cattle - pastoralist will never care about you. That only way to deal with them. Just avoid cattle.Maybe you'll have conflict in drought season as they bring their cows to graze by force in Meru

This so called census report which was doctored and No one knows where it came from was disputed by Merus simply because other 4 kabilas always gang up against Merus. What they do to Merus in isiolo is like what happens to kikuyus in other parts of Kenya It was rumors being spread around as you can see only star reported what Merus had raised. Take it from me, Merus are either 2 or 3. Other Kabilas hate Merus because they dominate business with a choke hold and at same time controlling Land thick and thin, Merus are also capable of defending their interest ruthlessly and engaging in the guerilla warfare like those pastoralist  communities hence the warfare situation in Isiolo ends up a draw no community wins. Thats why I told you Merus would drive you out of RV in just 2 days, we are used to simultaneously attacks from those communities frequently along the boarder front and we drive them back. Mark you they are not armed with bows and arrows like you people but with state of the art automatic rifles and machine guns.

Merus keep cattle alot and its a frequent source of envy from those kabilas simply because merus have other sources of income apart from cattle who the others 100% depend on them, so when fighting or cattle rustling begins while other kabilas are hellbent on making sure cattle get safely to their lands merus are happy to shoot the cows, ngamias etc dead, this makes them deranged and hate the Merus with a passion Merus have no dear attachments to these stuff except land. You can google conflicts in Isiolo and you will see they are always based on cattle and land,and since Meru is more developed of these areas most of the news reporting will cover there.

[RV Pundit]

Remind me of Kikuy 1992 molo wars - most pastoralist value cattle so much - more than even human life that is not their tribe. I think Kalenjin value a fellow kalenjin - then cattle next - and then other tribes. So when you kill a cow inhumanely people feel it. This pastoralist thing. The worse a kalenjin
rustler will do if they cannot drive a cow is cut their tail or ears.

For pastoralist - cattle is simply everything - the source of daily bread. Growing up - every of our cattle had a nick-name - and were adored.

Merus keep cattle alot and its a frequent source of envy from those kabilas simply because merus have other sources of income apart from cattle who they 100% depend on them, so when fighting or cattle rustling begins while other kabilas are hellbent on making sure cattle get safely to their lands merus are happy to shoot the cows, ngamias etc dead, this makes them deranged and hate the Merus with a passion Merus have no dear attachments to these stuff except land. You can google conflicts in Isiolo and you will see they are always based on cattle and land,and since Meru is more developed of these areas most of the news reporting will cover there.

[Njuri Ncheke]

Remind me of Kikuy 1992 molo wars - most pastoralist value cattle so much - more than even human life that is not their tribe. I think Kalenjin value a fellow kalenjin - then cattle next - and then other tribes. So when you kill a cow inhumanely people feel it. This pastoralist thing. The worse a kalenjin
rustler will do if they cannot drive a cow is cut their tail or ears.

For pastoralist - cattle is simply everything - the source of daily bread. Growing up - every of our cattle had a nick-name - and were adored.

Merus keep cattle alot and its a frequent source of envy from those kabilas simply because merus have other sources of income apart from cattle who they 100% depend on them, so when fighting or cattle rustling begins while other kabilas are hellbent on making sure cattle get safely to their lands merus are happy to shoot the cows, ngamias etc dead, this makes them deranged and hate the Merus with a passion Merus have no dear attachments to these stuff except land. You can google conflicts in Isiolo and you will see they are always based on cattle and land,and since Meru is more developed of these areas most of the news reporting will cover there.

Yes that's precisely the point merus play against these pastoralist, they know they can't win the cattle rustling /bandit activities so they engage in psychological warfare, at times they will maim the animals as a sign of warning and defiance and let them limp back to their owners who are hiding in the hills, but the price is high as these kabilas will shoot dead any meru they come across. There are no sorries

[Nefertiti]

Oh Njuri. Are these Ameru related to the Ashkenazi?

[Nefertiti]

Yes but your punditry is dodgy. The dynamics that change have metrics - influencers - and I have shown you how BBI is a credible carrot. Many influencers have a straw to raft to soft landing- from governor or VP or tribal god - to D/PM. That how Kalonzo, Mdvd, Weta will resist Mobutu vehemently despite matharao from Raila or Sifuna. Hustler is such a novel dynamic it amazing how much faith you have invested in the unicorn. Sorry I don't see the genius behind hustler concept.

If Luhya will ignore Mdvd and back Mobutu - that so reaching - it a big gamble at best. NARC or FORD or ODM was tribal blocks goldwrapped by "message" - propaganda, mana, cattle insurance, etc - not hustler abracadabra. I hold the same view in GEMA - Meru/Embu vote Kikuyu cause it closest shot to home. Realism. Kikuyu vote their own cause they are biggest tribe - at least in theory - and are very organized elite. Once Matiba bit the dust they all went Kibaki. Then Uhuru. Then Mobutu??

Mobutu is not with Muthama, Khalwale, Omar by masterstroke - it all he got. He crave Kalonzo or Mdvd or Joho but nope - no takers. Cause he got nothing to offer them - that the key reason. Already it a tussle for Mobutu - with 1M hostages, MOU, mlolongo - when they formally annoint a new prince your boy won't stand a chance. You are right post-Uhuru there is no strong GEMA - which is bad for Mobutu. Noone can trust mere MOU word of honor. Mobutu could pull it off because he had iron grip on Kalenjin.

Raila has a better offer than Mobutu. Class vs tribe - what is more entrenched?

Don't get you. You mean Raila already won GEMA votes? Tribes matters - but only where there is a strong tribal king. We can say Ruto and Raila have their tribes. Will a retiring Uhuru manage to herd GEMA while he is on his way to Ichaweri the verdict is out there - but why would GEMA listen to him - especially if he is egging them on a war with RV - that could see their people suffer.

Don't be suprised that nobody will listen to Uhuru - not even his MP - Moses Kuria.

Ruto has is gema votes - that are committed to Jubile original vision. There are new dynamics every election - we don't know if Kalonzo will manage to keep Kambas again - we don't know if maDVD will be strong - it look likely Weta is going for governorship - we know Matiangi if he quit statecraft could be a new factor in Gusii.

Those are dynamics - they change every election. For GEMA - it the same - 92 till 2002 - Mt kenya east were mostly with Moi - except for Imenti/Embus. GEMA itself were disunited except in 1997 when Kibaki got most of them. Obviously 2007, 2013 and 2017 - they were united.

This 2022 we really don't know how GEMA will pan out - we can predict for now two camp - Pro-Uhuru and Pro-Ruto. There is NO pro-Raila camp. Raila is hoping to inherit Pro-Uhuru as a gift and so is Matiangi, Kalonzo, PK and even Mutua - all these they have been promised by Murathe - the village drunkard.

What is likely to happen - without Uhuru - nobody can really unite GEMA - not even the kikuyus alone - their is no tribal heir acceptable to everyone - PK is dead on arrival - and maybe fly only in Muranga at best -  so we migt see 2 or even 3 camps - each doing their own thing. Kiunjuri group will stick with Ruto and get DPORK. Someone as desperate as Waiguru could go for Raila - and others could arrange harambee house project with Matiangi.

Will Uhuru have the time and money to engage in useless campaign that doesn't benefit him - we will see. If he stay out - Ruto makes mince meat of Pro-Uhuru group by inciting a GEMA poor revolt against their elite. It's already happening...

Fair enough. Add that to Raila 70% of NASA votes. Where does that leave Mobutu? If you do such simplistic math it just expose Mobutu soft underbelly. Oh wait, tribes are dead - now it hustler vs dynasty.