Census data on Kenyan Asians raises more questions than answers

When the census was taken last year, dissenting voices claimed that ethnicity should be left out of the questionnaire because it would lead to further polarisation in a country that was become increasingly tribalistic and because data on ethnicity would be manipulated by politicians to promote their own interests.

These arguments have been made in other countries as well, notably in India, where there is a debate about whether or not to undertake a caste census. Those opposing the census say that that it would promote casteism in the country, while those supporting it argue that a caste census would be an invaluable planning tool to promote equity.

At the time of the Kenya census, I argued in this column that enumerating ethnicity was not a statistical problem but a political one, and that the exercise of gathering data on the ethnic makeup of a country was not only normal, but desirable for planning purposes.

I was particularly excited about the current census because I was keen to know how many people of Asian origin resided in the country, as the only authoritative figure I had was one published over 40 years ago. Unfortunately, the figures on Asians released by the Kenya National Bureau of Statistics seem a little suspect.

According to an official text message that I received from the bureau, there are 35,009 people belonging to the “Asian tribe” in the country, who account for 0.09 per cent of the total population. This figure is questionable because, according to the bureau, there are also 53,393 Hindus in Kenya.

Now unless there has been a mass conversion of nearly 20,000 indigenous Africans to Hinduism in the last few years, it is impossible to reconcile these two figures. How can there be more Hindus than Asians in Kenya? Moreover, while it is very likely that the majority of Asians in Kenya follow the Hindu faith, not all Asians in Kenya are Hindus.

There are many who follow other religions, such as Islam, Christianity, Sikhism and Jainism. (Some are even atheists.) Zahid Rajan, founding member of the Kenyan Asian Forum, has alleged that since there was no specific code for Asians in the census, and that, in fact, Asians were coded as Somalis during the enumeration exercise, the data for both Asians and Somalis is flawed.

This might explain the rather large number of Kenyan Somalis that the KNBS says there are in the country. If the figure of 2.38 million is to be believed, then there are now more Kenyan Somalis in the country than there are Mijikendas or Kisiis. Government officials blame the high birth rate among Kenyan Somalis for the figure, but it is hard to believe that this group could have more than doubled in population since 1999, when it numbered roughly one million.

Even if you include the number of refugees from Somalia living in camps (who number around 350,000) in this figure, the rapid growth rate of this group is nothing short of extraordinary. These anomalies may appear minor, but they severely impact the credibility of the entire census data.

If the data on Asians appears to be false or questionable, then what other data might also be suspect? If the data has been manipulated, then the problem is even more serious; it is the reason, for instance, why Ravi Kanbur, the lead author of the World Bank’s World Development Report 2000/1, resigned just a few months before the report was to be published.

Kanbur claimed he was under “unreasonable pressure to tone down sections of the report” and resigned because he did not “submit to any substantive editing”. Apparently, Kanbur was not successful in convincing Bank officials that they must examine the detailed, disaggregated impacts (of poverty) on different population groups, rather than relying on general formulae.

A shoddy census is unforgivable in a day and age when data collection and analysis tools have become more sophisticated. If Kenyans do not believe the census results, and if planners, policymakers, researchers and the general public cannot use the data because it is seen to be inaccurate, then it means that a lot of taxpayers’ money was wasted on a futile exercise. I sincerely hope this is not the case.

rasna.warah@gmail.com