I use a NET birth rate (births-deaths) for the 0-5 year group as without an earlier 5 year group I have no way of calculating a separate death rate for this group. It would require separate data on death rate.

The NET birth rate is therefore used. This still allows accurate calculation of projected population size but leads to slighly lower birth and death rate calculations.

To predict the number of births for any 5 year interval I multiply the number of women of reproductive age by the birthrate (total number of children per woman during the 25 years of reproductive life). This product is then divided by 5 to get the number of children born during the given 5 year period.

For instance starting with the 1950-55 data we take the total number of women in the 15-40 age groups (F1). They will have given birth to all children in the 0-5 year group of 1955 (B2).

this leads to :-
            net birth rate = B2/F1 (number of births per woman for 5 year period)

We multiply this by 5 to get the total births per woman over the period of her reproductive live (ie the 25 years from 15-40).

The process is then repeated for all groups to 2010 the get a series of birthrates.

The graph below shows the results from Algeria.