Project Summary
The correlation between health and income is one of the most persistent observations in the social sciences,
supported by research across countries, demographic groups, and time periods. The relationship between
income and health is large, emerges early in life, and is present across many indicators considered proxies for
health. Despite the striking nature of the association, there is relatively little evidence in the U.S. context on
whether income has a causal impact on health outcomes, and less information on what mechanisms may be
responsible for such a causal relationship if one exists. Income and health may be correlated because higher
income has a direct impact on health (e.g., by allowing individuals to purchase medical care), or health may
directly affect income (e.g., because poor health makes it difficult to work). It may also be the case that a third
factor—such as education—influences both income and health, driving the observed correlation between the
variables. Without the ability to randomize variation in income, discerning the true nature of this relationship is
difficult. We propose to provide new evidence from a large-scale, randomized intervention in the U.S. on the
relationship between income and mental and physical health through a randomized evaluation of a sustained
unconditional cash transfer program being implemented by two non-profit organizations. This program will
randomly assign participants to a treatment group that receives $1,000 per month or to a control group that
receives $50. Both groups will receive the transfer monthly for three years. We will examine treatment/control
differences to estimate the effect of the additional income on behaviors and environmental factors (e.g.,
nutrition, food security, exercise, substance use, sleep quality, and neighborhood characteristics) that could
affect longer-term health outcomes (Aim 1) and on medical care access and utilization (Aim 2). In order to
understand the effect of the increase in income on health, we will assess treatment/control differences in
subjective measures of physical and mental health as well as blood pressure, height, weight, and blood spots that
can be analyzed to provide cholesterol, A1c (a measure of glycated hemoglobin that indicates diabetes risk), and
C-reactive protein (Aim 3). We will measure the outcomes associated with these aims using a combination of in
person and online survey questions, nutrition diaries (ASA 24), time diaries, and biomarkers collected in person
by trained enumerators. We will also conduct long term follow up after the end of the payments. This study will
provide groundbreaking, policy-relevant new information on the impact of income on health, healthy behaviors,
and well-being.