Full Citation
Title: Deaths Without Denominators: Using a Matched Dataset to Study Mortality Patterns in the United States
Citation Type: Miscellaneous
Publication Year: 2018
ISBN:
ISSN:
DOI: 10.31219/OSF.IO/Q79YE
NSFID:
PMCID:
PMID:
Abstract: To understand national trends in mortality over time, it is important to study differences by demographic, socioeconomic and geographic characteristics. One issue with studying mortality inequalities, particularly by socioeconomic status, is that there are few micro-level data sources available that link an individual's SES with their eventual age and date of death. In this paper, a new dataset for studying mortality disparities and changes over time in the United States is presented. The dataset, termed 'CenSoc', uses two large-scale datasets: the full-count 1940 Census to obtain demographic, socioeconomic and geographic information; and that is linked to the Social Security Deaths Masterfile (SSDM) to obtain mortality information. This paper also develops mortality estimation methods to better use the 'deaths without denominators' information contained in CenSoc. Bayesian hierarchical methods are presented to estimate truncated death distributions over age and cohort, allowing for prior information in mortality trends to be incorporated and estimates of life expectancy and associated uncertainty to be produced.
Url: https://osf.io/preprints/socarxiv/q79ye/
User Submitted?: No
Authors: Alexander, Monica
Publisher: Center for Open Science
Data Collections: IPUMS USA - Ancestry Full Count Data
Topics: Fertility and Mortality, Population Data Science
Countries: