WHAT ARE THE IMPACTS OF THE EXPANSION OF HEALTH INSURANCE SUBSIDIES ON POPULATION MORTALITY?

Responsible researcher: Bruno Benevit

Authors: Sarah Miller, Norman Johnson and Laura R. Wherry

Original title: Medicaid and Mortality: New Evidence from Linked Survey and Administrative Data

Intervention Location: United States

Sample Size: 566,000 individuals

Sector: Health Economics

Variable of Main Interest: Mortality

Intervention Type: Health Insurance Eligibility

Methodology: Difference in Differences, Event Study

Summary

The relationship between health insurance and mortality is still a widely discussed topic in the field of health economics. In the United States, the Medicaid allows access to health insurance for the poorest populations, with heterogeneity in relation to the conduct of the program among the country's federal entities. In order to evaluate the effectiveness of Medicaid , this study aimed to evaluate how the increase in enrollment in the program after its expansion impacted the mortality of elderly individuals, in addition to identifying the groups of adults most likely to benefit from the program. The identified results revealed that individuals from states that experienced expansion of program eligibility increased health insurance coverage and reduced the annual death rate by approximately 10%.

1. Policy Problem

The relationship between health insurance and mortality is a constantly debated topic in the health economics literature. In this context, socioeconomic conditions play a critical role in access to health services, most of which are supported by health insurance. As a result, individuals with incomes close to the United States federal poverty level (FPL) have a mortality rate four times higher than individuals with incomes four times above the FPL.

In the United States, Medicaid is the main health program at the federal level, providing eligibility criteria for beneficiaries and aiming to expand access to health services for the poorest population in the country. With the aim of expanding health insurance coverage, the federal Affordable Care Act (ACA) was enacted in 2010, expanding health insurance coverage to individuals in several states. In this sense, measuring the impact of expanding health insurance coverage on the mortality of the affected population is fundamental to improving the effectiveness of health policies.

2. Policy Implementation Context

The expansion of eligibility for Medicaid through the ACA played a crucial role in reducing health disparities in the United States, marking the largest expansion of health insurance coverage since the creation of Medicare and Medicaid in 1965. Since the implementation of the ACA, More than 20 million people have gained coverage, with the majority being low-income adults receiving assistance through Medicaid .

According to the authors (MILLER; JOHNSON; WHERRY, 2021), related literature presents evidence that this expansion significantly improved access to effective medical care that beneficiaries would not otherwise have, including medications that reduce mortality, diagnostic and early treatment of treatable cancers, as well as visits to hospitals and emergency departments for conditions that require immediate care.

In 2014, the ACA's expansion of Medicaid included all adults in families with incomes below 138% of the FPL, expanding coverage that had previously been restricted to pregnant women, adults with disabilities, and very low-income parents. A 2012 U.S. Supreme Court ruling made expansion optional for states, resulting in just 29 states and the District of Columbia adopting the measure in 2014, with seven more states joining in the following years. This adoption by states represented a historic expansion in insurance coverage, with an estimated 13.6 million adults gaining Medicaid under the ACA.

3. Assessment Details

To perform the analysis, the study linked individuals from three different data sources. First, it used restricted data from the 2008 to 2013 waves of the American Community Survey (ACS). Respondents selected were those who, based on their pre-ACA characteristics, would likely benefit from the ACA's Medicaid Therefore, only individuals in families with income equal to or less than 138% of the FPL or who have less than a high school diploma were considered. This last criterion is used to identify individuals of low socioeconomic status who may not meet the income limit at the time of the ACS interview.

The main analysis was restricted to individuals aged 55 to 64 in 2014, an age group with relatively high mortality rates. Residents of four states and the District of Columbia that expanded Medicaid to low-income adults before 2014 were also excluded. In total, approximately 566,000 respondents met the main sample criteria. This data is then linked to the Census Numident file. The Census Numident file is derived from the SSA Numerical Identification file, which includes information about date and county of birth and date of death for individuals with a Social Security number. The annual death rate is about 1.4% for the sample on average, and about 1.3% among respondents in expansion states in the year before expansion.

4. Method

As an empirical strategy, the study adopted an Event Study model to evaluate the evolution of relative outcomes between states that experienced the expansion of Medicaid compared to other states. This model allows us to control fixed effects of state and national trends, identifying heterogeneous temporal effects of the ACA expansion associated with each period before and after the intervention. Observations were established at the individual-period level. The outcome variables considered identified the death and coverage of individuals through various measures in a given period (year). The variable of interest for the intervention identified the occurrence of expansion in the status of individuals in the respective period. All analyzes considered ACS sampling weights and robust errors were clustered at the state level.

Furthermore, the Differences in Differences (DD) method was adopted to estimate the effect of Medicaid expansions after implementation of the ACA. The analysis replaces the event study indicators with a single variable that denotes states with expansion during the post-implementation period. This indicator identifies the intervention from the year of expansion for each state. The fixed sample, consisting of adults aged 55 to 64 in 2014, ages over time, resulting in increasing mortality rates. The DD method considers the mortality trajectory of this group and constructs a "counterfactual" rate for expansion states, adjusting for both the aging of the sample and the reduction in mortality due to expanded Medicaid coverage.

5. Main Results

Medicaid expansion showed significant increases in eligibility and insurance coverage for individuals in affected states. Individuals in these states have experienced significant increases in Medicaid eligibility , program enrollment, and health insurance coverage compared to states that did not expand. The probability of an individual being enrolled in Medicaid increased by 12.8 percentage points (pp), while uninsurance decreased by 4.4 pp. These effects were substantially greater than those observed in previous studies, indicating that low-income individuals with less education were those most affected by the expansion, given the selection criteria considered in this study.

Regarding the intensive use of Medicaid , the identified estimates revealed an increase in the number of accumulated days and years of program coverage. Individuals in expansion states experienced an average of 43 additional days of Medicaid enrollment per year compared to those in non-expansion states. When accumulated, this resulted in a significant increase of 0.38 additional years of Medicaid , amounting to approximately 0.67 additional years at the end of the study period. These results highlighted how the ACA's expansion of Medicaid not only increased immediate eligibility and coverage, but also promoted greater exposure and ongoing benefits over time for covered individuals.

Regarding mortality, the results of the Medicaid revealed a significant reduction among individuals in affected states. Before expansion, death rates were similar between expansion and non-expansion states. However, from the first year of expansion, a significant drop in mortality was observed in expansion states compared to non-expansion states. The annual mortality probability decreased by 0.089 pp in the first year, more than 0.1 pp in subsequent years, and 0.208 pp in the third year. These reductions in mortality were statistically significant and indicated an increasing impact over time. On average, the expansion resulted in a 0.132 pp reduction in annual mortality, representing a decrease of approximately 9.4% in relation to the sample average. The analysis also revealed that mortality due to internal causes and controllable health factors showed a significant reduction, while deaths due to external causes did not change significantly.

6. Public Policy Lessons

This article assessed how the ACA's expansion Medicaid affected health insurance coverage and mortality. The estimates identified indicated that states that expanded Medicaid saw a significant increase in eligibility and program coverage and utilization at both the extensive and intensive margins. Furthermore, the results revealed a reduction in mortality. Specifically, a relevant decrease in the probability of annual mortality was observed, with the effects intensifying over the years. The analysis also revealed that the reduction in mortality was most pronounced for causes of death treatable with healthcare, while deaths from external causes did not change significantly.

The results of this article help identify the positive impacts of Medicaid on the health of beneficiaries, providing useful information for public policymakers seeking to expand health coverage for socioeconomically vulnerable populations and reduce mortality. This evidence highlights that policies to expand access to Medicaid can have great potential in improving the health of the population without access to such services.

References

MILLER, S.; JOHNSON, N.; WHERRY, LR Medicaid and Mortality: New Evidence From Linked Survey and Administrative Data. The Quarterly Journal of Economics , vol. 136, no. 3, p. 1783–1829, 30 June. 2021.