Kenya Survey Finds Police Corruption

A recent survey in Kenya reported in Transparency International’s 2010 Global Corruption Barometer Report yesterday indicated that ninety two percent of Kenyans believe that the police force is the most corrupt institution in Kenya.  The index which has similar ratings from around the globe, put the police at 4.6 on a scale of one to five – with five being most corrupt.  It also reported that 45 percent of Kenyans had given a bribe to a public official in order to access services.

I found the results of the survey interesting, and performing surveys myself, I looked at their survey methodology.  The survey, which was performed by Synovate, had a sample size of  1,000 Kenyans, which is substantially enough to obtain a significant result assuming that the population was randomly selected and representative of the population.  The sample, however, was performed by computer-assisted telephone interviewing (CATI).  While an excellent tool in developed countries where most people have telephones, its use in Kenya puts into question the sample’s ability to represent the population since many Kenyans do not personally have a phone.

So instead of saying the sample is representative of the general population, Transparency International should say that it is representative of the population in Kenya with telephones.

9 comments to Kenya Survey Finds Police Corruption

  • Elise

    The article makes good points. The CIA World Factbook says that Kenya has 664 thousand land lines and 19 million cell phones – the cell phone number is likely skewed since people have two lines, while others with limited minutes wouldn’t be willing to use minutes on a phone interview. What about those with cell phones, but no electricity to charge them? So Transparency Intl should also make the clarifying point that Kenya survey is more representative of the urban population which has infrastructure than the rural population.

  • Elise,

    Great point, Transparency International should indee clarify that their Kenya survey is more representative of the urban populations. Thanks for the input.

  • Barbara

    Great article about the Kenya survey. It made me think of differences in rural vs urban surveys. CATI is definately a good tool, but not useful in many rural settings

  • Thank you. Transparency International’s corruption survey in Kenya is help and would be more so if not for the unfortunate bias.

  • Craig W.

    It is interesting that this Kenya survey used CATI considering that roughly two thirds of Kenyans live in rural settings. Perhaps those designing the survey thought rural cell phone penetration was larger than it actually is?

  • Craig,

    Thank you for your comment. I think part of the confusion is that the number of mobile phone subscribers in Kenya isn’t too far from the overall population. However, this number is skewed in that some people in cities have multiple phones (e.g. business and personal).

  • Mike

    I understand that the survey is biased in favor of urbanites, but do you think the survey is also biased in terms of socio-economic conditions? I imagine that urbanites below a certain poverty level would not be financially able to own home or cell phones.

  • Mike,

    Thank you for your comment. Having a phone is definately tied to a person’s economic level. Land lines are out of reach for many in Kenya, and are not reliable. Mobile phones are alot less expensive, but to someone earning two dollars a day, it is still a very significant expense.

  • [...] two percent of Kenyans believe that the police force is the most corrupt institution in Kenya. We looked into this survey methodology to report some weaknesses in the study that likely bias the results towards those living in urban [...]