A reply to Mr Palloni's comments

Abstract The protracted and inconclusive debate on the cause of the post-war mortality decline in Ceylon reflects our ignorance of this complex historical event and although I am reticent to prolong this already lengthy discussion, I feel that it is necessary to reply to certain points raised by Mr Palloni. The object of my paper 'The Decline of Mortality in Ceylon and the Demographic Effects of Malaria Control'(9) was to re-examine some of the past work on this subject in order to attempt a synthesis of previous theories, was not, however, intended to provide a definitive account of all the causal mechanisms underlying the decline of mortality as it is my view that the data are insufficient for such an undertaking. In the reappraisal I was mainly concerned with the validity of Newman's regression model and, as far as the data would permit, an assessment of Meegama's thesis that there were significant disturbing variables which confounded the simple regression of mortality decline and malaria prevalence. I will try first to respond to Mr Palloni's specific substantive points and then go on to consider the broader question of regression models.