Somalia's first national census was taken in February 1975, and as of mid-1992 no further census had been conducted. In the absence of independent verification, the reliability of the 1975 count has been questioned because those conducting it may have overstated the size of their own clans and lineage groups to augment their allocations of political and economic resources. The census nonetheless included a complete enumeration in all urban and settled rural areas and a sample enumeration of the nomadic population. In the latter case, the sampling units were chiefly watering points. Preliminary results of that census were made public as part of the Three-Year Plan, 1979-81, issued by the Ministry of National Planning in existence at the time. (Because the Somali state had disintegrated and the government's physical infrastructure had been destroyed, no ministry of planning, or indeed any other government ministry, existed in mid-1992.) Somali officials suggested that the 1975 census undercounted the nomadic population substantially, in part because the count took place during one of the worst droughts in Somalia's recorded history, a time when many people were moving in search of food and water.

The total population according to the 1975 census was 3.3 million. The United Nations (UN) estimated Somalia's population in mid-1991 at nearly 7.7 million. Not included were numerous refugees who had fled from the Ogaden (Ogaadeen) in Ethiopia to Somalia beginning in the mid-1970s.

The Ministry of National Planning's preliminary census data distinguished three main categories of residents: nomads, settled farmers, and persons in nonagricultural occupations. Settled farmers lived in permanent settlements outside the national, regional, and district capitals, although some of these were in fact pastoralists, and others might have been craftsmen and small traders. Those living in urban centers were defined as nonagricultural regardless of their occupations. In 1975 nomads constituted nearly 59 percent of the population, settled persons nearly 22 percent, and nonagricultural persons more than 19 percent. Of the population categorized as nomads, about 30 percent were considered seminomadic because of their relatively permanent settlements and shorter range of seasonal migration.

Various segments of the population apparently increased at different rates. The nomadic population grew at less than 2 percent a year, and the seminomadic, fully settled rural and urban populations (in that order) at higher rates--well over 2.5 percent in the case of the urban population. These varied rates of growth coupled with increasing urbanization and the efforts, even if of limited success, to settle nomads as cultivators or fishermen were likely to diminish the proportion of nomads in the population.

The 1975 census did not indicate the composition of the population by age and sex. Estimates suggested, however, that more than 45 percent of the total was under fifteen years of age, only about 2 percent was over sixty-five years, and that there were more males than females among the nomadic population and proportionately fewer males in urban areas.

Population densities varied widely. The areas of greatest rural density were the settled zones adjacent to the Jubba and Shabeelle rivers, a few places between them, and several small areas in the northern highlands. The most lightly populated zones (fewer than six persons per square kilometer) were in northeastern and central Somalia, but there were some sparsely populated areas in the far southwest along the Kenyan border.

The nomadic and seminomadic segments of the population traditionally engage in cyclical migrations related to the seasons, particularly in northern and northeastern Somalia. During the dry season, the nomads of the Ogo highlands and plateau areas in the north and the Nugaal Valley in the northeast generally congregate in villages or large encampments at permanent wells or other reliable sources of water. When the rains come, the nomads scatter with their herds throughout the vast expanse of the Haud, where they live in dispersed small encampments during the wet season, or as long as animal forage and water last. When these resources are depleted, the area empties as the nomads return to their home areas. In most cases, adult men and women and their children remain with the sheep, goats, burden camels, and, occasionally, cattle. Grazing camels are herded at some distance by boys and young unmarried men.

A nomadic population also inhabits the southwest between the Jubba River and the Kenyan border. Little is known about the migratory patterns or dispersal of these peoples.

Somalia's best arable lands lie along the Jubba and Shabeelle rivers and in the interriverine area. Most of the sedentary rural population resides in the area in permanent agricultural villages and settlements. Nomads are also found in this area, but many pastoralists engage part-time in farming, and the range of seasonal migrations is more restricted. After the spring rains begin, herders move from the river edge into the interior. They return to the rivers in the dry season (hagaa), but move again to the interior in October and November if the second rainy season (day) permits. They then retreat to the rivers until the next spring rains. The sedentary population was augmented in the mid-1970s by the arrival of more than 100,000 nomads who came from the drought-stricken north and northeast to take up agricultural occupations in the southwest. However, the 1980s saw some Somalis return to nomadism; data on the extent of this reverse movement remain unavailable.

The locations of many towns appear to have been determined by trade factors. The present-day major ports, which range from Chisimayu and Mogadishu in the southwest to Berbera and Saylac in the far northwest, were founded from the eighth to the tenth centuries A.D. by Arab and Persian immigrants. They became centers of commerce with the interior, a function they continued to perform in the 1990s, although some towns, such as Saylac, had declined because of the diminution of the dhow trade and repeated Ethiopian raids. Unlike in other areas of coastal Africa, important fishing ports failed to develop despite the substantial piscine resources of the Indian Ocean and the Gulf of Aden. This failure appears to reflect the centuries-old Somali aversion to eating fish and the absence of any sizable inland market. Some of the towns south of Mogadishu have long been sites of non-Somali fishing communities, however. The fisheries' potential and the need to expand food production, coupled with the problem of finding occupations for nomads ruined by the 1974-75 drought, resulted in government incentives to nomad families to settle permanently in fishing cooperatives; about 15,000 nomads were reported established in such cooperatives in late 1975.

Present-day inland trading centers in otherwise sparsely populated areas began their existence as caravan crossing points or as regular stopping places along caravan routes. In some cases, the ready availability of water throughout the year led to the growth of substantial settlements providing market and service facilities to nomadic populations. One such settlement is Galcaio, an oasis in the Mudug Plain that has permanent wells.

The distribution of town and villages in the agricultural areas of the Jubba and Shabeelle rivers is related in part to the development of market centers by the sedentary population. But the origin of a considerable number of such settlements derives from the founding of agricultural religious communities (jamaat) by various Islamic brotherhoods during the nineteenth century. An example is the large town of Baardheere, on the Jubba River in the Gedo Region, which evolved from a jamaa founded in 1819. Hargeysa, the largest town in northern Somalia, also started as a religious community in the second half of the nineteenth century. However, growth into the country's second biggest city was stimulated mainly by its selection in 1942 as the administrative center for British Somaliland. In 1988 Hargeysa was virtually destroyed by troops loyal to Siad Barre in the course of putting down the Isaaq insurrection.

After the establishment of a number of new regions (for a total of sixteen as of early 1992, including Mogadishu) and districts (second order administrative areas--sixty-nine as of 1989 plus fifteen in the capital region), the government defined towns to include all regional and district headquarters regardless of size. (When the civil war broke out in 1991, the regional administrative system was nullified and replaced by one based on regional clan groups.) Also defined as towns were all other communities having populations of 2,000 or more. Some administrative headquarters were much smaller than that. Data on the number of communities specified as urban in the 1975 census were not available except for the region of Mogadishu. At that time, the capital had 380,000 residents, slightly more than 52 percent of all persons in the category of "nonagricultural" (taken to be largely urban). Only three other regions--Woqooyi Galbeed, Shabeellaha Hoose, and the Bay--had urban populations constituting 7 to 9 percent of the total urban population in 1975. The sole town of importance in Woqooyi Galbeed Region at that time was Hargeysa. Berbera was much smaller, but as a port on the Gulf of Aden it had the potential to grow considerably. The chief town in Shabeellaha Hoose Region was Merca, which was of some importance as a port. There were several other port towns, such as Baraawe, and some inland communities that served as sites for light manufacturing or food processing. In the Bay Region the major towns, Baidoa and Buurhakaba, were located in relatively densely settled agricultural areas. There were a few important towns in other regions: the port of Chisimayu in Jubbada Hoose and Dujuuma in the agricultural area of Jubbada Dhexe.