A combined Brass-random walk approach to probabilistic household forecasting: Denmark, Finland, and the Netherlands, 2011–2041

Demographic Research Monographs

Alho

et al. 2019

Self Cite

Globally, the twenty-first century will witness rapid population ageing. Already in 2050, one out of five persons in the world, and one out of three in Europe, is expected to be 60 or over (UN 2015). Moreover, we have entered into a new stage of population ageing in terms of its causes, which have altered its consequences. In the first stage, lasting until the middle of the twentieth century in developed countries, population ageing was entirely due to the decline in fertility, with Sweden being commonly used as an example (Coale 1957; Bengtsson and Scott 2010; Lee and Zhou 2017). During this stage, the increase in life expectancy was primarily driven by declines in infant and child mortality. It worked in the opposite direction to

Section: From Scenarios To Stochastic Modellingmentioning

confidence: 99%

Introduction

Bengtsson

Demographic Research Monographs

Alho

et al. 2019

Self Cite

“…one-parent or two-parent family). A recent example is the random share model in Keilman (2017), which distinguishes the population by seven household positions. Four of these are family positions (child, living in consensual union, living in marital union, lone parent) and the remaining three are non-family positions (living alone, living in institution, other position).…”

Section: Membership Rate Modelsmentioning

confidence: 99%

“…Expected values of the shares were computed using a multi-state model of family and household dynamics. Keilman (2017) simplified the latter model and applied time series methods to predict the random shares into the future, thereby avoiding the complexities of a multi-state model.…”

Section: Deterministic and Probabilistic Modelsmentioning

confidence: 99%

Family Projection Methods: A Review

The Springer Series on Demographic Methods and Population Analysis

2018

Self Cite

I review methods for projecting the family structure of a population into the future. The focus is on Western countries. I define the concept of "family", and stress the close link between family and household. There is an important distinction between family position (the attribute of individuals) and family type (the attribute of families). I propose a typology of family projection models, and review a large number of approaches. Consistency problems (being a generalization of the two-sex problem of demography) arise when interrelated events experienced by members of the same family are modelled. Little is known about the accuracy of various types of family projection models. The chapter concludes with an evaluation and assessment of various types of family projection models, using criteria such as data requirements, output detail, ease of programming, and extrapolation of parameters. Agent Based Models and probabilistic projection models are promising avenues for future work.

Uncertainty in Population Forecasts for the Twenty-First Century

Annu. Rev. Resour. Econ.

2020

Self Cite

The aim of this article is to review a number of issues related to uncertain population forecasts, with a focus on world population. Why are these forecasts uncertain? Population forecasters traditionally follow two approaches when dealing with this uncertainty, namely scenarios (forecast variants) and probabilistic forecasts. Early probabilistic population forecast models were based upon a frequentist approach, whereas current ones are of the Bayesian type. I evaluate the scenario approach versus the probabilistic approach and conclude that the latter is preferred. Finally, forecasts of resources need not only population input, but also input on future numbers of households. While methods for computing probabilistic country-specific household forecasts have been known for some time, how to compute such forecasts for the whole world is yet an unexplored issue. Expected final online publication date for the Annual Review of Resource Economics, Volume 12 is October 5, 2020. Please see http://www.annualreviews.org/page/journal/pubdates for revised estimates.