Living to age 100 in Canada in 2000

age. With the annual number of deaths by age in Canada, published by Statis-tics Canada, we first construct cohort life tables at ages 80 and over for people born before 1900, using the method of extinct generations. We fit some statis-tical models to these data sets, using maximum likelihood theory to estimate the parameters of the models and obtain estimates of the standard error for the mortality rates. Goodness-of-fit tests are performed to check model ade-quacy. We then project the mortality rates for people who will attain age 80 in the future. Finally, the distribution of the maximum age at death that will be attained by a Canadian is investigated. 1 1 DATA SOURCES Statistics Canada is a governmental organization responsible for making a detailed analysis of the Canadian population and publishing the data. From 1949 to 1997, we can find the annual number of deaths in Canada by sex and individual ages from 0 to 99 and grouped over age 100 (denoted 100+). This information is also available for each province and territory, but we will use only the total for Canada. Age is defined as the completed age in years. Vincent (1951) developed the method of extinct generations to construct life tables from death statistics. Assuming that no migration takes place after age 80, we can calculate the number of males and females aged x (x ≥ 80), living on January 1, Y, by summing the number of deaths in future years at successive ages, lYx = k=0 dY+kx+k, where lYx is the number of people aged x living on January 1, Y and d Y x is the number of people dying between ages x and x+ 1 in calendar year Y. An assumption on the fraction of people dying at age x in year Y born in year Y −x or Y −x − 1 is necessary: it will be assumed that 50 % are born in each year. The number lYx was therefore approximated by lYx