Adrian E Raftery1, Nevena Lalic2, Patrick Gerland3. 1. Departments of Statistics and Sociology, University of Washington, Seattle, Washington, USA. 2. Institutional Research, University of Washington, Seattle, Washington, USA. 3. United Nations Population Division, Population Estimates and Projection Section, New York, New York, USA.
Abstract
BACKGROUND: The United Nations (UN) produces population projections for all countries every two years. These are used by international organizations, governments, the private sector and researchers for policy planning, for monitoring development goals, as inputs to economic and environmental models, and for social and health research. The UN is considering producing fully probabilistic population projections, for which joint probabilistic projections of future female and male life expectancy at birth are needed. OBJECTIVE: We propose a methodology for obtaining joint probabilistic projections of female and male life expectancy at birth. METHODS: We first project female life expectancy using a one-sex method for probabilistic projection of life expectancy. We then project the gap between female and male life expectancy. We propose an autoregressive model for the gap in a future time period for a particular country, which is a function of female life expectancy and a t-distributed random perturbation. This method takes into account mortality data limitations, is comparable across countries, and accounts for shocks. We estimate all parameters based on life expectancy estimates for 1950-2010. The methods are implemented in the bayesLife and bayesPop R packages. RESULTS: We evaluated our model using out-of-sample projections for the period 1995-2010, and found that our method performed better than several possible alternatives. CONCLUSIONS: We find that the average gap between female and male life expectancy has been increasing for female life expectancy below 75, and decreasing for female life expectancy above 75. Our projections of the gap are lower than the UN's 2008 projections for most countries and so lead to higher projections of male life expectancy.
BACKGROUND: The United Nations (UN) produces population projections for all countries every two years. These are used by international organizations, governments, the private sector and researchers for policy planning, for monitoring development goals, as inputs to economic and environmental models, and for social and health research. The UN is considering producing fully probabilistic population projections, for which joint probabilistic projections of future female and male life expectancy at birth are needed. OBJECTIVE: We propose a methodology for obtaining joint probabilistic projections of female and male life expectancy at birth. METHODS: We first project female life expectancy using a one-sex method for probabilistic projection of life expectancy. We then project the gap between female and male life expectancy. We propose an autoregressive model for the gap in a future time period for a particular country, which is a function of female life expectancy and a t-distributed random perturbation. This method takes into account mortality data limitations, is comparable across countries, and accounts for shocks. We estimate all parameters based on life expectancy estimates for 1950-2010. The methods are implemented in the bayesLife and bayesPop R packages. RESULTS: We evaluated our model using out-of-sample projections for the period 1995-2010, and found that our method performed better than several possible alternatives. CONCLUSIONS: We find that the average gap between female and male life expectancy has been increasing for female life expectancy below 75, and decreasing for female life expectancy above 75. Our projections of the gap are lower than the UN's 2008 projections for most countries and so lead to higher projections of male life expectancy.
Authors: Leontine Alkema; Adrian E Raftery; Patrick Gerland; Samuel J Clark; François Pelletier; Thomas Buettner; Gerhard K Heilig Journal: Demography Date: 2011-08
Authors: Adrian E Raftery; Nan Li; Hana Ševčíková; Patrick Gerland; Gerhard K Heilig Journal: Proc Natl Acad Sci U S A Date: 2012-08-20 Impact factor: 11.205
Authors: Patrick Gerland; Adrian E Raftery; Hana Sevčíková; Nan Li; Danan Gu; Thomas Spoorenberg; Leontine Alkema; Bailey K Fosdick; Jennifer Chunn; Nevena Lalic; Guiomar Bay; Thomas Buettner; Gerhard K Heilig; John Wilmoth Journal: Science Date: 2014-09-18 Impact factor: 47.728