You can compare whether a family has < or > than 2.3 children though even though no single family is =
How would your method work on that?
My point being: statistics has developed arcane methods to describe a population. One should not blindly force the most common measure on every case.