Which countries have the youngest populations?
rferl.org
rferl.org
Or you can look at it from a purely economic standpoint: if your military-aged men are mostly finding gainful employment and you want to raise an army you need to either conscript or pay competitive wages. Either way you're damaging your economy. If you have far more people that age then prospective jobs it's way cheaper to convince some of them to put on a uniform.
Even North Korea looks good in living standard and life expectancy compared to many African countries.
If I had to make an educated guess, I'd say it's probably done this way to make data more comparable between countries for whatever reason. I'm not exactly sure why that would be necessary, but generally when you see something weird like this, it's because something changes a lot from culture to culture, so they use wider buckets to capture everyone who is reasonably comparable. Maybe it's some combination of when different cultures start counting your age and a lack of good birth records in poorer areas, and they figure people can more or less estimate which 5 year block you're in. Another possibility is that the older data sets were collected using some form that had you blocked off in 5 year blocks for whatever reason, and they can't change it now without invalidating the comparability of the data sets.
EDIT: If anyone's interested, I found this manual [1] on estimating population age and sex where they go into detail about the assumptions in the five-year age group method. Doesn't really explain why this is desirable, though - I think it might have something to do with how census data are collected in different countries.
[0] http://esa.un.org/unpd/wpp/Excel-Data/population.htm
[1] http://www.un.org/esa/population/techcoop/PopProj/manual3/ch...
This is not necessarily true. Generally the number for "life expectancy" that is quoted is life expectancy at birth, which is roughly the average age of people at their death. In a lot of poorer countries, this number ends up being driven by infant mortality (to see why, consider a country where, if you survive infancy, you live to be 100 years old, but 50% of all infants die - the "life expectancy at birth" of such a country would be about 50).
Generally, I think it's understood that the "youth" of a population is a sort of measure of adult life expectancy. Though I suppose it can also be a measure of the birth rate (weighted by infant mortality) across countries. I think the two factors end up being fairly intimately tied to one another (likely by a third factor, wealth), so it's not absurd to conflate the two.