In celebration of Business Talk’s 300th column in the Business Daily, let us take an evidence-based look at public sector job allocations in Kenya. Famous author Arthur Goldsmith broke down perceptions about the African public sector employment that exploded following independence as a way to offer patronage to those who supported various political leaders.
However, empirical evidence is often not available to dissect the reality of public sector job allocation favouritism in sub-Saharan African countries.
Researchers Emanuele Colonnelli, Edoardo Teso, and Mounu Prem out of Harvard University and Stanford University found that supporting the political party in power increases the probability of securing a public sector job by 47 percent in Brazil. University of California-Berkeley social scientists James Robinson and Thierry Verdier call such public sector job allocation as “clientelism”.
In so doing, a job is selective and can be removed and thus ties the job holder to continue in the political process to ensure the success of the specific politician.
But what about here in Kenya? Are our public sector jobs distributed through clientelism? Do some communities benefit from receiving public sector jobs more than others?
London School of Economics economic historian Rebecca Simson dissected public sector job allocations in Kenya, Uganda, and Tanzania for two highly respected recent publications in the journals Economic History Review as well as African Affairs.
The stunning research challenges our perceptions of the public sector. The Government of Kenya used to retain huge numbers of public workers. In 1990, public employees comprised 7.7 percent of all workers in the country.
But government has steadily decreased public labour staff since 1990, hitting 3.7 percent of total labour force in 2007 and inching up to 3.8 per cent in 2017 into 2018. Which Kenyan communities lost out the most with the nearly 52 per cent decline in government workers?
The London School of Economics dissected Kenyan census and other government publication figures to find the proportion of Kenyans born in different constituencies in 1972 compared to 2009. The proportion of Government of Kenya workers born in Nairobi totalled 23 percent of the government workforce in 1972. By the 2009 census, Nairobi-born government of Kenya workers only comprised 17 percent.
The former Coast Province also experienced sharp declines in representation in the government workforce with a decline from 14 percent down to nine per cent in the same period. Those born in the former Central Province reduced their government workforce proportion from 15 to 14 .
All other former provinces saw increases in public labour force representation between 1972 and 2009 with Nyanza, North Eastern, Western, Eastern, and Rift Valley increasing substantially. Rift Valley-born government workers comprised 20 percent in 1972 and then 25 percent in 2009, by far the highest represented area.
Many things also changed in Kenya between 1972 and 2009. The proportion of secondary school completion skyrocketed from 15 to 79 percent. Between 1965 and 2009, the percentage of women in government employment increased from eight to 37 percent. But government employment represented more of a lure to the independence generation more than generation X.
In 2009, Kenyans at age 55 who also held a university degree had a 60 percent probability of working for the government. But also in 2009, those Kenyans aged only 30 and held a university degree were only 30 percent likely to work for the government.
Next, looking at the 2009 Kenyan Census, those born in which counties represented the largest proportion of government workers?
The London School of Economics researchers disaggregated the census data to those born in each county. They then looked at the ethnic composition in each county. They then cross-checked the ethnicity of government workers compared to the total population of that ethnicity to come up with surprising results. Since Nairobi County and Mombasa County hold high levels of ethnic diversity, those born in those counties were designated as “mixed” ethnicity counties.
Not surprisingly, the ethnic group based on place of birth with the largest percentage of workers in government compared to the overall population in their counties by the 2009 census were indeed those born in “mixed” ethnic counties with just over eight per cent of all government workers.
In terms of those born in more ethnically homogenous counties, the Kalenjin community comprised the highest proportion of their 25 to 55-year-old citizens in government roles with over seven per cent. The next largest ethnic community in the public sector workforce was the Embu community followed by the Luo, Kamba, Kikuyu, and then Kisii communities with between six and seven per cent of their adult populations between ages 25 and 55 working for the government.
The Luhya, Meru and Mijikenda communities each totalled between five and six percent in government. The communities with the lowest proportion of citizens aged between 25 and 55 in government were the Somali at 2.75 per cent and the Turkana at only 1.8 percent.
However, if considering the proportion of adults in a community aged between 25 and 55 years old who also completed secondary school, then the numbers show a dramatically different picture. The Turkana community goes from the lowest proportion in government to the highest with 33 percent between ages 25 and 55 with secondary school completion working in the government while the Somali community under the same criteria comprise 31 per cent of their citizens.
The disparity of secondary school education seems to tell a larger story than possible clientelism on the macro scale. The Kalenjin make up the third highest proportion at 22.5 percent followed by the Mijikenda at 22 percent.
The communities with the lowest proportion of their 25 to 55 years olds with secondary school completion who also work for the government in public sector employment are Kikuyus with only 14 percent and Kisii with 15 percent.
In summary, public sector employment proved more evenly distributed than expected. While cronyism based on ethnicity surely can exist at different levels, the overall picture of government employment was more equally distributed than in more cronyism societies, such as arguably Brazil at political party levels. It will be interesting for researchers to uncover trends in the new upcoming Kenyan census.
[email protected] or on Twitter: @ScottProfessor