That assumes both that the increase in reading scores due to increasing family income is linear, and that annual family income and birth spacing are completely independent of each other.