Principal investigator: Eduarda Miller de Figueiredo
Authors: Janet Currie and Jonathan Gruber
Article: Health Insurance Eligibility, Utilization of Medical Care, and Child Health
Location of the Intervention: United States
Sample Size: 225 thousand children
Sector: Health Insurance
Primary Variable of Interest: Measurement of use per individual
Type of Intervention: Health insurance
Methodology: Instrumental Variable
Summary
The high infant mortality rates in the United States prompted the authors to investigate whether this was related to the quality/quantity of medical care these children received. In 1984, eligibility for the Medicaid program, targeting low-income children, was expanded. Based on this expansion, the authors estimated the effects and magnitude of this expansion on medical care utilization and mortality. Using a linear probability model and a model with instrumental variables, the results suggest an increase in the number of consultations and hospitalizations, as well as a reduction in the infant mortality rate.
- Policy Problem
High rates of infant mortality and morbidity suggest that American children do not receive the same quantity or quality of medical care as children in other developed countries. For example, when compared with Canadian children, children aged 1-4 years in the United States have a 14% higher mortality rate.
According to Bloom (1990), one possible explanation for this is that up to 30% of poor children have no health insurance of any kind, in a country where the uninsured receive less healthcare than the insured. Therefore, the debate on reforming the American healthcare system increasingly emphasized health insurance for children, however, there was no convincing evidence that increasing eligibility for public insurance would actually improve children's health (Boston Globe, 1994).
Taking this into account, the authors sought in this article to identify the effects of insurance coverage based on expansions of eligibility. Medicaid for low-income children. In which the Medicaid It is a federal-state program that provides health insurance for the poor.
- Implementation and Evaluation Context
Historically, the eligibility of Medicaid It was linked to receiving cash welfare payments in the program. Aid to Families with Dependent Children (AFDC). Therefore, eligibility was effectively limited to women and children of very low income in single-parent families. In 1984 there was an obligation to extend coverage to Medicaid For other groups of children, and in 1992, states were required to include in the coverage children under 6 years old in families with income up to 133% of the poverty line and those aged 6-19 years old with family income up to 100% of the poverty line. Furthermore, there was the option to include infants with income up to 185% of the poverty line; thus, there was variation in eligibility due to different rates among states.
Naturally, the first question to be asked after the changes in the policy of Medicaid It is whether they had a significant effect on the fraction of the population eligible for the program.
- Policy/Program Details
Starting with the Deficit Reduction Act of 1984 (DEFRA '84)[1], the link between AFDC coverage and eligibility for the Medicaid The requirement was reduced because DEFRA '84 eliminated the family structure requirements for Medicaid eligibility for young children, requiring states to provide coverage for children born after September 1983 who live with families whose income qualifies for AFDC.
- Method
The authors used data from research by Current Population Survey (CPS) from 1984 to 1992, which enabled the collection of information on demographic characteristics, income and work data. And also data from National Health Interview Survey (NHS), which also includes information on the use of medical care, covering a total sample of 225.000 children in the study period.
To begin analyzing the effect of eligibility for the Medicaid In the utilization study, the authors used a linear probability model, measuring utilization per individual. i as the dependent variable. In addition, the authors added control variables that include gender, race, ethnicity, education, and housing, an indicator of an individual's eligibility for the program. Medicaid and binary variables for state and year.
However, the authors highlight the existence of an endogeneity bias. A sick child can lead to lower parental income – if one parent is forced to leave work to care for the child, for example – leading to a spurious positive correlation between eligibility and use of the [system/program/etc.]. MedicaidTherefore, there may be a substantial error in the eligibility measure. To correct for this potential bias, the authors used a “simulated instrument” that varies only with the state’s legislative environment and not with its economic or demographic characteristics.
- Main results
First, the authors observed that there was a "dramatic" increase in eligibility for the program. Medicaid, in which one-third of all children in the United States were eligible at the end of the period. To separate the effects of the economic cycle from the effects of legislative change, the fraction of the 1984 population that would have been eligible under the laws of each year was also estimated. The results suggest that most of the increase in eligibility was a result of legislative changes.
Increases in eligibility do not automatically mean increases in insurance coverage. When examining the acceptance of coverage... Medicaid For newly qualified children, the authors demonstrated that coverage increased, but not as abruptly as eligibility. Making a child eligible for the MedicaidThis increased the likelihood of her being covered by insurance by approximately 30%.
Based on the results obtained from the linear probability model, the authors found that the Medicaid This significantly reduces the probability of missing a doctor's visit in the year prior to the survey by 2,5 percentage points, but has no statistically significant effect on the probability of having visited a doctor in the last two weeks. When compared to the baseline probability of missing a visit for eligible respondents... MedicaidMaking a child eligible reduces the likelihood of them not receiving a visit by 12,8%.
However, when estimating using the instrumental variable, the results suggest that becoming eligible for the Medicaid This is associated with a 9,6 percentage point drop in the probability of missing a visit in the previous year. In other words, the results found by the authors, using this methodology, suggest that making children eligible for the program reduces the probability of them missing a visit by half. A very large and significant increase of 4 percentage points in the probability of hospitalization in the previous year was also observed. This coefficient implies that becoming eligible for the program... Medicaid It almost doubles the likelihood of being hospitalized.
Another finding was that the increase in the eligible fraction for the Medicaid It has a significant negative effect on the infant mortality rate. For every 10 percentage point increase in the fraction of children eligible for the... Medicaid, reduces mortality by 0,128 percentage points, which is 3,4% of the baseline mortality rate of the sample.
- Lessons in Public Policy
The results suggest that expanding eligibility for the U.S. federal-state health insurance program increases consultation utilization relatively efficiently. However, it also demonstrated significant increases in the incidence of hospitalization, which, according to the study authors, may reflect inefficiencies in how care is delivered to patients in the health insurance program.
References
Bloom, Barbara, “Health Insurance and Medical Care,” Advance Data from Vital and Health Statistics of the National Center for Health Statistics, No. 188 (Washington, DC: Public Health Service, 1990).
[1] Deficit Reduction Act of 1984.
