A historical glance at African patterns of migration exposes the drawbacks of humanitarian intervention.
Since the American army and the United Nations left Somalia in 1995, most Westerners and humanitarian organizations have also decamped, overwhelmed by kidnapings and the racketeering of warlords. During 10 years of chaos, Mogadishu has become a field of ruins, a lunar landscape in which combatants and displaced people wander aimlessly. Deserted by the international community, the country has no more embassies, no government, no recognized authority. A single French nongovernmental organization remained in the country until recently.
Permanently guarded by armed militia, this NGO's expatriate employees could barely leave their bunker to carry on their mission. The modus operandi of this group is typical of the way most NGOs operate in a context of war, and that has a significant effect on how a humanitarian crisis is perceived, followed, and evaluated by decision makers.
Since the fall of the Somalian dictatorship and the pillage of the city in 1991, this NGO regularly sent its fundraisers reports on the number of persons displaced by the conflict, who live in the ruins of the capital on sites reported as so many "camps." Year after year, a rapid tally seems to indicate an almost constant rise in the number of these camps and their occupants. These reports, however, are misleading.
In fact, unable to move around freely without armed escorts, humanitarian personnel have no clear picture of the entire city at any one time. Because of the insecurity and difficulty of living conditions, the rare foreigner stays in Mogadishu barely more than six months. Of course, the deployment of 30,000 American soldiers under the aegis of UNOSOM, the United Nations Operation in Somalia, did allow a more precise idea of the situation between 1992 and 1994. But our French NGO alone can't keep track of all the camps. So it continues to add, to the number of sites that close, the number of sites that open.
What's more, certain camps may be counted more than once. The Somali language, which borrows many elements from the Arabic, was transcribed into the Roman alphabet in 1972. The word Ali, for example, is spelled Cali; and Hashi, Xashi. So the same camp may be found more than once in an alphabetical listing, depending on whether the inventory's author chooses a Somali or English spelling. In addition, because of the high rate of turnover of on-site personnel, camp names change constantly. The name may change depending on the whim of the spokesperson of the day, or whether the name is given in Somali, Arabic, English, or Italian. Thus, displaced persons who are squatters in the industrial zone of Mogadishu have seen their sites inventoried under the English name for factory as well its Somali equivalent warshadaha.
In other words, no hard data can confirm that the number of displaced persons in Mogadishu has risen since the United Nations left in 1995. Quite the contrary, there is evidence the numbers stabilized.  In all war and emergency situations, the problem is the same, whether it's a question of evaluating needs, counting the dead, or estimating refugees. For the most part, to sensitize public opinion and obtain financing from their funders, humanitarian aid groups tend to exaggerate the extent of a crisis.
Local authorities, for their part, at first try to minimize the drama to mask their responsibility in the matter. Then they solicit international aid for their own gain through embezzlement. In either case, very complex political considerations blur the qualitative and quantitative evaluation of humanitarian crises.
In the Western world, the conflicts of the Third World have given rise to a very specific type of discourse that developed along with the rise of the "without borders" NGOs during the 1980s and later, at the end of the Cold War, with the establishment of partisan groups that claim a humanitarian right to interfere in the internal affairs of Third World countries. …