Causal effects of wealth on health: Evidence from lifetime follow-up of randomized land redistribution

点击翻译按钮获取中文摘要

PROJECT SUMMARY A significant body of research shows that having money is highly correlated with better health. But determining how much of the association is causal is challenging. The gold standard for determining causality is prospective random assignment of treatment, but practical and ethical considerations often make it hard to award money or equivalent resources in a truly random way. Thus, researchers have used longitudinal studies of parents’ income and children’s health or exogenous changes in social and fiscal policy to show that additional money lowers mortality. But studies with true randomization—lottery winners and people randomly given different levels of health insurance—show more modest influences of money on mortality. This study produces the first ever estimates of how randomly awarded wealth affects mortality over a lifetime of follow-up in two generations, for both men and women, and constructs a policy relevant metric of the incremental wealth required for an additional year of life expectancy. The project analyzes the long-term implications of a major land redistribution policy in New Zealand in which low interest loans for farms were awarded by lottery. Program terms ensured applicants were of similar economic status, and the policy operated like a true random experiment. Loan terms were similar to modern mortgages: low annual payments over a long term. Documents show randomization was adhered to for over 20 years. New Zealand’s universal, centralized death registration facilitates construction of a matched intergenerational sample of lottery winners and losers and their children. Ninety percent of applicants can be linked to death records, which is a significantly higher linkage rate than can be obtained in the US. These estimates will be a lower bound on how much money matters for health because New Zealand had low inequality, high life expectancy, and the subjects were born in an era when infectious disease was an important cause of death. This R21 project has four aims: 1) estimate the causal effects of winning land on lottery participants’ mortality, 2) estimate the causal effects of winning land on the mortality of children of lottery participants, 3) estimate gender differences in the effects of wealth on mortality, and 4) identify potential mechanisms that contributed to life expectancy differences. The study design combines the inferential strengths of true random assignment in a block design with additional contrasts between children born before and after their parents participated in the lottery to identify effects of wealth on mortality in two generations. Contrasts in effect sizes by sex, birth order, and generation are used to identify potential causal mechanisms. Causal mediation analysis of early adult health measures, mid-life occupational status, urban or rural residence, and lifetime wealth accumulation illuminate potential pathways between wealth and reduced mortality. The proposed research, with a credible causal identification strategy and lifetime follow-up, will contribute significantly to knowledge of how money and social status affect health, aging, and mortality.