Paul
Your reasoning that projections are 'purely mathematical ...', whereas forecasts 'imply extra knowledge ...' is spot on - but then your interpretation of this seems to be competely contrary.
Forecasts are a subset of projections - not the other way around. Therefore the phrase 'projections are not forecasts' is certainly true most of the time, although they can sometimes be the same thing - ie forecasts are a subset of projections.
One may make many demographic projections - high, median, low variants - projection of AIDS deaths under various theoretical conditions etc. - but only one or two of these may be realistic and then, with added judgement thrown in, they become forecasts.
Projections cannot be wrong, forecasts usually are wrong.
Forecasts involve putting your professional neck on the line, projections just involve going through the motions. Forecasters are as often wrong as anybody else but they have to have good and complicated reasons for being so.
Andy Sloggett
Lecturer in Medical Demography
and MSc Course Organiser
Centre for Population Studies
London School of Hygiene & Tropical Medicine
50 Bedford Square, London, WC1B 3DP
Tel: 020 7299 4628
Fax: 020 7299 4637
|