Please could someone explain the difference between extrapolative methods for improvements in mortality assumptions, vs "alternative methods" i.e. projecting trends in e.g. causes of death, drivers of mortality (such as smoking), incidence and treatment of major diseases? It seems like the extrapolative methods also use lifestyle choices e.g. smoking and past data to project future improvements, so the two methods sound very similar to me. Is the difference that the extrapolation method uses the data blindly without consideration for how improvements will change whereas the 'alternative methods' contain more 'thought' e.g. consideration that people in the future can't reduce smoking by the same degree as was done over the past decades?
Hi Lucy Someone with more practical experience may be able to provide further information, but I reckon you are on the right lines here. Process based projections approach the problem from a bio-medical perspective, ie they aim to model trends and changes in the underlying causes of death and use these to estimate future mortality rates. Simple extrapolative methods just project mortality rates forward using historical trends in mortality. The SA4 Core Reading only touches on mortality projections - lots of variations in approach and alternative methodologies out there in the real world! Gresham
