Individual- and State-Level Variations in Medicare Enrollment Among Older Adults Who Have Reached Their Full Retirement Age

Purpose:  Covering 54 million individuals, Medicare spends 14% of the federal budget and accounts for 20% of total healthcare spending in the U.S. As baby boomers started reaching retirement age, it has significant policy implications to understand the characteristics of the individuals who have reached their retirement age but have not enrolled in Medicare. One potential reason is that with increases in longevity and improved health, many older adults continue to work after age 65. Additionally, raising the full retirement age (FRA) eligible for full Social Security from 65 to 67 has facilitated continued employment into older ages. Many older adults who work beyond FRA may elect not to enroll in Medicare if they have coverage through their employment. The specific aims of this study were: 1) to assess the proportion of older adults who had reached FRA but not enrolled in Medicare; and 2) to identify individual- and state-level factors associated with Medicare enrollment. To control for cohort effects, this study examined the birth cohort of 1942 who all passed FRA for Social Security (65 years and 10 months), as well as the age of eligibility for Medicare (65+) as of 2009.

 

 Methods: This study utilized the 2009 American Community Survey (ACS, state-level) and the 2009 Public Use Microdata Sample of the ACS (individual-level). The two levels of data were linked by state ID. The sample included 28,678 individuals who were born in 1942 and residing across 50 states and the D.C. Replicate weights and the STATA SVY statements were employed to achieve accurate population estimates. Two-level multilevel modeling was conducted using the HLM software for multivariate analyses. The Level-1 model estimated the probability of being covered by Medicare at the time of interview (Yes=1; No=0).

Results: Among the birth cohort of 1942, about 28% were still working and 4.2% were not receiving Medicare in 2009. Among those without Medicare, 31% were uninsured (i.e., 1.3% of all sample). The unconditional model indicated the significant state-level variation in Medicare enrollment (variance component= 0.20967, df=50, p<.0001). The rates of Medicare enrollment varied from 100% to 87.2% at the state level. Among other characteristics, the percentage of foreign-born population in the state was the strongest state-level predictor (OR=1.03; p=.03) even after controlling for significant individual-level variables. Individual-level characteristics that were significantly associated with the lower probabilities of Medicare enrollment were: larger family size, not currently married, being Asian, being African American, being non-citizen, being covered by employment-based health insurance, having income from wages/salary and having no functional impairments. 

 

Implications:  This study indicates that older adults’ Medicare enrollment status among those who have passed FRA is associated with multi-level, multi-dimensional factors. Future studies should investigate the further sources of state-level variation, such as the variation in employment-based health insurance for older adults. Also, it is important to note that while delaying or not enrolling in Medicare beyond FRA may be voluntary for some older adults, some associated factors imply subpopulations of vulnerable older adults in terms of their financial, health, and legal status.