Chris Blattman’s fame as a development blogger is such that I think the rest of us development-types sometimes give short shrift to his published research. Thus it was with interest that I read his 2008 article on “Child combatants in northern Uganda: Reintegration myths and realities” (PDF) co-authored with Jeannie Annan. B&A identify a rather surprising natural experiment, arguing that LRA abduction of young men was so widespread as to be essentially random. This allows them to make relatively clean estimates of the impact of child soldiering on comparable groups of young Ugandan men, using a mixed-methods approach with approximately 1000 respondents.
What they find is more complex, and perhaps less dramatic, than many mainstream accounts of child soldiering suggest. Children were most likely to be abducted in early adolescence, as younger boys were inefficient fighters and older boys were more difficult to indoctrinate & posed greater escape risks. 80% of abductees eventually escaped, with most of the remaining 20% presumed dead.
Upon returning home, however, the former abductees were generally not received as the “damaged, uneducated pariahs” that the NYT had assumed them to be (quoted in B&A, p. 1). 94% of interviewed abductees said that their families had accepted them back without censure, and three quarters reported that they were generally treated well by their communities. Only one sixth of former child soldiers reported elevated levels of psychological distress, and on average they were no more likely to behave violently than non-abductees.
Unfortunately, the economic outlook for returned child soldiers was not quite as bright. B&A note that, whilst the average abductee only missed 9 months of education, abduction in early adolescence meant that that missing year of schooling was generally the 6th or 7th grade, when Ugandan students typically learn to read and write. Thus former abductees were “twice as likely to be illiterate” (p. 16) The educational gap also explained nearly two-thirds of the observed earnings shortfall of child soldiers, which found them “half as likely to be engaged in skill- or capita-intensive employment, and [to] have a third lower daily earnings” (p. 16) than non-abducted young men. Interestingly, shortfalls have also been found found in the earnings of American veterans of the Vietnam War. In both cases, “the source of this earnings gap appears to be time away from civilian education and work experience” (p. 22). The authors close with a set of useful recommendations for tailoring ex-combatant reintegration programs to these realities.