[[On the parameter estimation problems in the multiregional demographic growth model]].

"In this paper, we critically prove the formulas of Ledent-Rogers-Willekens.... Our procedure does not depend on multi-regional life table technique but on direct discretization of the continuous-time model. Here our main purpose is not to provide new results but to clear the conditions under which these methods for estimating the parameters can be applied. Finally, we give another method of calculation of parameters, which makes it possible to do a consistent disaggregated projection when we have already had an aggregated population projection." (SUMMARY IN ENG)