The effect of cash transfers on health in high-income countries: A scoping review
Robbie Brydon,Saud Bin Haseeb,Gum-Ryeong Park,Carolyn Ziegler,Stephen W Hwang,Evelyn L Forget,Navindra Persaud,Arjumand Siddiqi,James R Dunn
DOI: https://doi.org/10.1016/j.socscimed.2024.117397
2024-10-03
Abstract:High-income countries use cash transfer programs to mitigate poverty, in part to improve the health of low-income populations and potentially reduce their use of public health care. This review synthesizes evidence from studies that employed experimental or quasi-experimental designs to evaluate the effect of cash transfer interventions on health outcomes or health care utilization in high-income countries. We excluded interventions if they required prior contributions for eligibility, substituted cash transfers for in-kind services, or were contingent on specific health behaviours, and excluded studies published before 1970. We searched 14 academic databases on May 13, 2022 and April 18, 2023, identifying 20,978 unique records. After screening, 164 studies were included. These studies covered interventions in 14 countries, with the largest share from the United States. The most common health outcomes examined were fertility, birth weight, self-rated health, tobacco use, and depression. We classified studies into seven intervention categories and eight health outcome domains, and identified where systematic reviews may be possible. We found relatively few studies examining health care utilization as an outcome and identify this as a knowledge gap. We categorized effects as beneficial or harmful, except for fertility and health care utilization where effects were categorized as increase or decrease. With insufficient consistency of outcomes for meta-analysis, we employed a vote count and sign test to assess the presence of any effect. Across the six relevant health domains, 98 of 130 studies (.75; 95% CI: .67, .82) reported a beneficial median effect, significantly different from the null value of 50% (p = .000). Of 37 studies examining fertility, 23 showed increases (.62; 95% CI: .46, .76) in fertility, which did not clear our threshold for statistical significance using conservative assumptions (p = .094). However, a larger share of studies reported increased fertility for child/family benefits (.69, n = 26) than for employment-related cash transfers (.44, n = 9). Results for health care utilization were evenly distributed (5 increase, 4 inconsistent, 6 decrease), but these are difficult to interpret as outcomes include both preventive and acute care. Our study provides replicable methods to enable future meta-analyses.