It does not make more sense no matter how I imagine it to refer to women. Once a woman has given birth to a child she is no longer childless. I cannot imagine any way of counting women to arrive at such dramatic changes in numbers of childless women in only three years.

It's a ratio of women who have had children to women that have not had children.

"...it doesn't extend to women who have already had children."

Then it isn't a ratio of women who have had children to women that have not, which is what it is described as.

Something is inexplicable about that number.