Walking with Feet Tied to the Past: Childhood Exposure to the 1918 Influenza Pandemic and Later-Life Male Mortality
中文摘要
既有许多研究通过跨队列比较并聚焦于宫内暴露，记录了1918年流感大流行对幸存者晚年结果的持续影响。其中一部分研究考察了其对老年健康和死亡率的影响，但所得结果不一。本文将分析扩展至评估生命早期和童年时期暴露于该大流行对老年寿命的影响，并考察一组潜在作用机制，从而为相关文献作出贡献。我们将社会保障署死亡记录与1940年人口普查全量数据相链接，并采用事件研究法和双重差分回归，根据1918年流感相关死亡率所衡量的暴露程度，比较出生于高暴露城市和低暴露城市、暴露年龄不同的各队列的寿命。我们的处理组平均处理效应（ATT）计算表明，暴露年龄为3—8岁和9—11岁的队列，其老年寿命分别缩短6.4个月和11.2个月；需要特别指出的是，这些队列在更广泛的文献中经常被用作“对照组”。受数据限制，我们的主要分析聚焦于男性，但补充结果表明，对女性的影响更小，估计精度也更低。对人口普查数据的分析表明，健康、教育和社会经济状况得分的下降可能是潜在机制。
Abstract
Many previous studies have used cross-cohort comparisons and a focus on in utero exposure to document the lingering effects of the 1918 influenza pandemic on survivors’ outcomes later in life. A subset has examined impacts on old-age health and mortality, finding mixed results. This paper contributes to the literature by extending the analysis to evaluate early-life and childhood exposure to the pandemic on old-age longevity along with a set of candidate mechanisms. We employ Social Security Administration death records linked to the full-count 1940 census and employ event-study and difference-in-difference regressions to compare longevity of cohorts with different age-at-exposure who were born in cities with high versus low exposure based on 1918 influenza-related mortality. Our treatment-on-treated calculations suggest that cohorts with age-at-exposure of 3-8 and 9-11 experience 6.4 and 11.2 months reductions in old age longevity; importantly, we note that these cohorts are often used as “controls” in the broader literature. While our main analysis focuses on males due to data constraints, supplementary results suggest smaller and less precise effects for females. An analysis of census data suggests reductions in health, education and socioeconomic scores as potential mechanisms.