Estimating Age-Specific Mortality Using Calibrated Splines

I adapt Schmertmann's Calibrated Spline method to the problem of estimating a complete mortality schedule from abridged data. I demonstrate its ability to fit the range of mortality profiles seen across the epidemiological transition. I find that it is as accurate as current popular methods across closed age groups and superior to them when extrapolating into the open age group. I show that the method is naturally suited to small populations where sample variance is significant.