By Ana Ekelund

Somalia is home to an unknown amount of people. Population numbers that are presented in various textbooks and encyclopedias are in fact qualified guesses. This is in no way unique. In fact, most countries of the world have an unknown number of inhabitants. For some nations, including the rigorously catalogued Nordic countries or tiny nations such as former British colony Pitcairn Island (population: 50), these numbers are updated every time a person is born, dies, immigrates or emigrates. However, most states rely instead on estimations in between censuses. Census meaning to count each and every person residing within the territory of a country. It is an extensive and above all expensive – procedure. Thus, censuses are seldom carried out annually or even semi-annually.

What makes the Somalian population number more mysterious than others is the amount of time since the last census. The most recent figures were published in 1975. The current population size of Somalia is estimated based on numerous previous estimates during the past forty-four years, In other words: a guess based upon a guess. Which is not to say that the estimations are useless. They too are extensive and require complex analysis of birth and death rates, as well as migration patterns. But guesses, however qualified, still involve some bit of uncertainty. The less relevant the information the guess is based on, the larger the uncertainty.

There is, of course, no one more preoccupied with this uncertainty than the Somalian state itself. Accurate demographic information, including population size, location and age structure, is vital in order to tax and carry out state services. Up-to-date information is crucial in answering questions such as: How big is the tax income, and how big are the expenses? How many schools are needed, of what size, where and when?

A census is a complicated thing. Even in Great Britain, a nation with good prerequisites, such as a well-developed infrastructure, it is only done once every ten years. Somalia, in comparison, is home to many nomadic groups (constituting a rough quarter of the total population) and is currently in a seemingly endless civil war. Very few of its inhabitants are born or die in a hospital.

With the odds stacked against it: How can a census be carried out?

“No one really knows” is the answer given by a Swedish statistician working to help develop the Somalian statistics bureau. Due to the current situation he is stationed in a neighboring country, and on domestic trips he is escorted by a United Nations convoy. As of now, he says, a complete census is more or less off the table – but there are alternatives. With other surveys coming up, the plan is to use the opportunity to produce fresher estimates, namely household-surveys that estimate the population based on the demographics of several smaller samples. First, households are grouped by district. Second, a few are selected at random to be surveyed. The information gathered from these sample households is then generalized for the entire population using statistical methods of probability.

This method, while more simplified, is also associated with practical problems in the Somalian context, a clan-based society with armed conflict. Executive power is divided among clans based on the so-called 4.5 formula. This is a quota system where the four largest clans are represented by equally large parts of the parliament, and the rest of the clans are represented by the remaining 0.5 share of the seats. Because of this, clans often report having more members than they actually do, in an attempt to increase their claim to power. Furthermore, because regions Puntland and Somaliland are involved in a land dispute, they both include the inhabitants of the disputed regions when reporting their population size. This means that numbers reported are often questionable.

Despite all of this, many challenges of data gathering have also generated new and innovative methods. Due to the danger associated with working for the state in areas where the terrorist group al-Shabaab is active, the forms used in Somalian household surveys are much shorter than on average, in order to be carried out under two hours. The two hours represents the minimal amount of time in which information about the interview reaches al-Shabaab, a group is mobilized and then reaches the household in question. When tested, these forms actually provided data of higher quality than lengthier ones, simply by removing anything other than the very essential. Another innovative solution is the way in which nomads are interviewed. Because they lack a permanent residence, they are often hard to reach. Without an address to visit, the interviewers are instead posted for eight days close to water stations. Eight days being the maximal amount of days a camel can go without water, forcing the nomads to visit a water station.

As is evident, counting the population in countries is not an easy thing. The result has been very creative and innovative ways to acquire educated guesses. Somalia is a prime example of this phenomenon. But two important questions remain: How many people have been overlooked by the surveyors? And how many people do not in fact exist at all?

Illustration: Rickard Ström

Ana Ekelund is an intern at the International Cooperation Office at Statistics Sweden, where she is learning about statistical capacity development, as well as the mysterious ways of the copy machine on the fourth floor. Besides finishing up her Bachelor’s Degree in Peace and Conflict Studies, she devotes her time to in-door climbing, texting, and searching for the perfect satsuma (juicy and kind of sour).

Related Posts