Post–election Public Opinion Poll March 5 - 6, 2013 Background

AfriCOG commissioned this post-election survey as part of its work on the 2013 general elections, the first general elections under the new constitution.

Methodological Note

 Due to the prolonged nature of the March 4th elections, this survey differed from the classic exit poll  The exercise started on Monday 4th March, 2013 and ended on 6th March.  Very few successful interviews were conducted on 4th as many people who were randomly called had not yet voted.  This meant that not every voter was being provided with an equal chance of being interviewed, especially in more populous areas  The exercise was thus carried out on 5th March, achieving 1800 interviews.  The  On the 6th 1000 calls were conducted to validate the results gathered on the 5th.  When the validation was completed and the results seemed to be a reflection of each other, the data was married, giving the total sample of 2800. Methodology

 The poll was conducted between March 5 – 6, 2013 by Infotrak Research and Consulting.

 A sample of 2,893 respondents was interviewed to represent the Kenyan BVR registered voters of 14,337,399 translating into a minimum margin of error of -/+ 1.82 at 95% degree of confidence.

 Using preliminary BVR voter register as the sample frame, the sample was designed using Population Proportionate to Size (PPS) and mainly entailed; • Use of stratification, random and systematic sampling in drawing regions to be covered • Ensuring further distribution by area, age and gender • Using the Constituencies as the key administrative boundary • Ensured that every person in the sampled area had a known chance of being selected

 The interviews were conducted through CATI system • 25% of the interviews were back checked for quality control purposes

 Data processing & analysis was carried using IBM SPSS 20.0

 The questions asked of respondents are highlighted for each graphic presentation Margin of Error explained

 Margin of error decreases as the sample size increases, but only up to a certain point.

 A very small sample, such as 50 respondents, has about a 14 percent margin of error while a sample of 2,000 has a margin of error of 3 percent.

 By doubling the sample to 2,000, the margin of error only decreases from +/-3 percent to +/- 2 percent and +/-1.8 percent for a sample size of 4000.

 This illustrates that there are diminishing returns when trying to reduce the margin of error by increasing the sample size.

 What is imperative is to ensure that the sample is representative of the universe you wish to cover. This is why in a continent the size of USA, most sample sizes range between 1000 -3000 covering the entire population. And the results are more or less accurate

 A 95 percent level of confidence is the acceptable standard for social surveys. Margin of Error Illustration

16.0 Margin of Error 14.0 13.9 12.0

10.0 9.8 8.0 6.9 6.0 4.9

Variability Variability 3.7 4.0 3.3 2.7 2.2 2.1 2.0 2.0 1.5 1.4 2.0 1.2 1.0 3.0 2.5 1.1

0.0

50

200 400 600 800

1000 1200 1400 1600 1800 2000 2200 2400 3000 4500 6000 8000 10000 Sample size SAMPLING DISTRIBUTION

Region Sample distribution % allocation

Coast 234 8.1

North Eastern 88 3

Eastern 401 13.9

Central 442 15.3

Rift Valley 680 23.5

Western 294 10.2

Nyanza 394 13.6

Nairobi 360 12.4

Total 2893 100 Race to State House…. Presidential Candidate Voted For

100.0%

80.0%

60.0% 48.7% 45.7%

40.0%

20.0%

3.3% 0.6% 1.2% 0.2% 0.3% 0.0% as Musaila Peter Kenneth James Ole Mohammed President as President Mudavadi as as Presient as President Kiyapi as Abduba as President President president

Who did you vote for as president? N = 2893 (All respondents) Presidential Candidate Voted for By Region

Who did you vote for as President?

Region Uhuru Musaila James Ole Mohammed Base Raila Odinga Martha Karua Peter Kenneth Kenyatta as Mudavadi as Kiyapi as Abduba as as President as Presient as President President President President president

Coast 86.0% 11.5% 0.9% 0.4% 0.4% 0.0% 0.9% 234 North 53.4% 44.3% 0.0% 0.0% 0.0% 0.0% 2.3% 88 Eastern Eastern 55.1% 43.1% 0.5% 0.0% 1.2% 0.0% 0.0% 401

Central 5.4% 90.7% 0.5% 2.0% 1.4% 0.0% 0.0% 442

Rift Valley 24.7% 70.3% 2.8% 0.0% 1.2% 0.4% 0.6% 680

Western 70.4% 6.5% 22.4% 0.7% 0.0% 0.0% 0.0% 294

Nyanza 86.8% 12.2% 0.5% 0.3% 0.0% 0.0% 0.3% 394

Nairobi 55.2% 37.9% 0.6% 1.4% 4.2% 0.6% 0.3% 360

Aggregate 48.7% 45.6% 3.3% 0.6% 1.2% 0.2% 0.3% 2893

Who did you vote for as president? Who did you vote for as president? Raila Odinga vs Uhuru Kenyatta by Region

Raila Odinga as President Uhuru Kenyatta as President

100.0% 90.7% 86.0% 86.8%

80.0% 70.3% 70.4%

60.0% 53.4% 55.1% 55.2%

44.3% 43.1% 40.0% 37.9%

24.7% 20.0% 11.5% 12.2% 5.4% 6.5%

0.0% Coast North Eastern Eastern Central Rift Valley Western Nyanza Nairobi

Who did you vote for as president? N = 2893 Did you experience challenges as you voted?

Yes, 35%

No, 66%

Did you experience challenges as you voted? N = 2893 (All respondents) Challenges experienced during voting

Queues were too long 65%

BVR machines were too slow 24%

There were few polling booths 11%

The polling clerks were not competent 8%

Polling clerks were not helpful enough 5%

Polling clerks were too few 3%

There was commotion within the polling centre 3%

I was forced to vote for aspirants I had not intended to vote 1% for

0% 20% 40% 60% 80% 100%

If YES, what challenges did you experience? n = 997 Demographics Response Distribution by Gender

Female , 49.0%

Male , 51.0% Response Distribution by Age Group

50.0%

40.0%

30.0% 25.3% 21.5% 20.0% 16.2%

11.3% 10.1% 10.0% 6.0% 4.7% 3.6% 1.4% .0% 18-24 25-29 30-34 35-39 40-44 45-49 50-55 56-59 60+ Response Distribution by Highest Level of Education

100%

80%

60%

45%

40%

25% 21% 20%

7% 1% 1% 0% Primary Secondary College University Post graduate None