Increasing the Probability of Comparing Between Traditional Medicare and Medicare Advantage

Abstract

Background:

Having knowledge of out-of-pocket cost and access to services constitutes an advanced level of health literacy within the Medicare population. The cost and services offered within Medicare change annually, yet less than 30% of beneficiaries compare their coverage options (between traditional or fee-for-service Medicare and Medicare Advantage)—an important application of health care knowledge. Failure to compare can expose the financing of the Medicare program and the beneficiary to financial risk if a beneficiary has not elected a coverage option that best suits their individual needs. The factors driving beneficiaries to compare are poorly understood.

Objective:

Our objective was to examine the association between different levels of information reviewed (exposure) and comparing coverage (outcome).

Methods:

This pooled cross-sectional study included 28,924 Medicare beneficiaries using data from the 2019−2021 Medicare Current Beneficiary Survey. Multivariable probit regressions were performed using Stata version 18.

Key Results:

Reviewing both cost and service information increased the probability of comparing coverage by 159% (44.4 percentage points). Exposure to service-only information increased the probability of comparing coverage by 97% (27.2 percentage points) and 68% (19.1 percentage points) for cost-only information, relative to those who reviewed neither cost nor service information. Forty-seven percent of beneficiaries did not review cost nor service information, followed by 39% who reviewed both, 8% cost-only, and 6% service-only. The probability of comparing coverage increased by 8.2% (2.3 percentage points) for beneficiaries who reviewed both cost and service information and those who had an education greater than high school.

Conclusion:

A targeted outreach and education campaign geared toward increasing the type of information reviewed could result in an increase in comparing Medicare coverage options.

Plain language summary

Plain Language Summary: We examined factors associated with beneficiaries comparing their option to elect either traditional Medicare or Medicare Advantage. In this study, review of both out-of-pocket cost and access to services increased the probability of comparing coverage options. These findings suggest policymakers and the Centers for Medicare & Medicaid Services should focus on crafting targeted campaigns to increase the level of health literacy for Medicare beneficiaries.

Medicare beneficiaries can maintain or change their preferred coverage choice during the annual open enrollment period, October 15 through December 7 each year (Centers for Medicare & Medicaid Services, 2024a), and should consider actively engaging in this comparison process because the cost and coverage of services change on an annual basis (Bruine de Bruin et al., 2024; Centers for Medicare & Medicaid Services, 2024a; Grabert, 2022; Jazowski et al., 2024; Miller, Parente, & Wilensky, 2022; Ochieng et al., 2024; Park et al., 2022b; Zhao et al., 2021). Beneficiaries can access assistance to manage coverage choice, including government websites such as the Medicare plan finder, a printed copy of the “Medicare and You” handbook, which is mailed annually to households, the 1-800-MEDICARE call center (staffed by trained customer service representatives), State Health Insurance Assistance Plans (SHIP) (offering unbiased assistance to beneficiaries and families, including analysis of tradeoffs between coverage options), and independent or affiliated private brokers (Medicare, n.d.-a., n.d.-b., n.d.-c., n.d.-d.; State Health Insurance Assistance Program, 2024; Tabor et al., 2025).

Despite these resources, less than 30% of Medicare beneficiaries compare coverage during open enrollment (Ochieng et al., 2024). Although policymakers have expressed concern regarding this lack of engagement, we are not aware of any studies that have examined the potential mechanisms behind what is predictive of comparing (Grabert, 2022; Ochieng et al., 2024). It is also important to note that although choice of a coverage option is a paradigm that may be unique to the US health care system, the concept of how choices are made is broadly applicable to international health care systems, even if choice occurs at a more detailed level, such as choices regarding individual clinicians.

If beneficiaries choose coverage that is misaligned with their needs, they risk being exposed to higher costs and/or limited access to services they may need. This could stem from misaligned Medigap coverage (too rich of a Medigap benefit, such as zero cost-sharing, relative to needs or lack of supplemental coverage), prescription drug coverage as part of a constructed traditional Medicare benefits package, or incompatibility of an integrated Medicare Advantage benefits package (such as accidentally selecting a plan where an existing specialist is out of network) (Bruine de Bruin et al., 2024; Grabert, 2022; Jazowski et al., 2024; Miller, Parente, & Wilensky, 2022; Ochieng et al., 2024; Park et al., 2022b; Zhao et al., 2021). One way to ensure beneficiaries are equipped to make coverage decisions is to increase individual health literacy.

The peer reviewed literature has previously defined a high level of health literacy as comparing coverage and/or reviewing cost and/or access to service information in the Medicare population (Park et al., 2022b). Park et al. (2022b) found that those Medicare beneficiaries that either review information or compare coverage are more likely to select a Medicare Advantage (MA) or private health insurance option to manage their annual coverage. Since the publication of Park et al. (2022b), the Medicare program has altered the survey questions related to coverage and now asks separate questions on review of information and comparing coverage (Park et al., 2022b). This new question scheme allows us to add to the health literacy literature by examining a sequence of actions, where we explore the impact of the reviewing information as an independent action that may impact the dependent action of comparing coverage.

Arguably, the most important coverage decision facing a beneficiary is whether to construct a holistic health benefits package through a multi-part benefit comprised of traditional fee-for-service (FFS) Medicare, Medicare supplemental coverage/Medigap plan, and a standalone prescription drug plan; or alternatively to elect a single product: an integrated MA plan which offers coordinated Medicare benefits comprising medical, pharmacy, and supplemental coverage. When considering these two options, scholars have regarded the MA option as simpler and more closely related to private insurance benefits offered by employers and exchange coverage offered under the Affordable Care Act (Grabert, 2022; Miller, Parente, & Wilensky, 2022). MA coverage is an integrated vehicle for a beneficiary to access their Medicare benefits, which includes Medigap and prescription drug coverage, and is available at a lower total annual out-of-pocket cost with respect to both premiums and cost sharing (Grabert, 2022). Additionally, nearly all MA plans may offer supplemental benefits which can include vision, hearing, dental, and in-home modifications, benefits unavailable in FFS (Freed, Ochieng, Cubanski et al., 2024; Grabert, 2022; Hammond, 2025).

In contrast, FFS Medicare offers an any-willing-provider network and has fewer restrictions on service use, while MA relies on utilization review tools such as prior authorization for medical services and the implementation of PPO (Preferred Provider Organization) or HMO (Health Maintenance Organization) network design to manage costs, features that are present in every ACA (Affordable Care Act) exchange plan and in a majority of private employer-sponsored insurance coverage offerings (Claxton et al., 2025; Grabert, 2022; Rae et al., 2024). FFS is regarded as a comparatively more complex decision since beneficiaries are required to coordinate multiple insurance benefit design features including a Part D plan and a Medigap plan for wraparound insurance (Miller, Parente, & Wilensky, 2022).

Beyond trade-offs, research has also focused on demographic characteristics that may predict preference for certain coverage options. Beneficiaries who are White (McWilliams et al., 2023), have multiple chronic conditions (Atherly et al., 2023; McWilliams et al., 2023), and have more than a high school education (Atherly et al., 2023), choose FFS more often than MA, while beneficiaries who are Black, Hispanic, have few chronic conditions, are urban dwelling, and married, choose MA (Atherly et al., 2023; McWilliams et al., 2023). Although these preferences are well documented, the mechanisms contributing to these associations are not well understood.

Given the importance of comparing coverage as a precursor to selecting FFS or MA, we were interested in exploring the potential mechanisms behind this action. Specifically, we aimed to test how review of cost and/or access to service information during the Medicare open enrollment period (October 15 through December 7) influences the decision to compare coverage options. We hypothesized that reviewing cost and/or access service information would have a positive influence on comparing coverage options.

Methods

Our analysis used pooled cross-sectional secondary data from the Limited Data Set version of the 2019−2021 US Medicare Current Beneficiary Survey (MCBS). The MCBS annually surveys Medicare beneficiaries for up to four years (Centers for Medicare & Medicaid Services, 2024b). The MCBS includes demographic information, geographic location, income, insurance coverage, and information related to health care services.

We limited our sample to those dwelling in the community who were surveyed from January 1, 2019, through December 31, 2021. Consistent with methods using the MCBS in other peer-reviewed manuscripts, beneficiaries whose responses, included “don't know” and “refused” to any variable were dropped from our sample (He et al., 2023; Ng & Park, 2021; Ng et al., 2022; Park et al., 2022a).

For our exposure (predictor variable x) the MCBS asks two questions about the type of information reviewed: “during the 2019 open enrollment period, did you review your Medicare insurance coverage to see if there were going to be changes in your monthly premium, deductibles, co-payments, or other out-of-pocket expenses?” and “during the last open enrollment period, did you review your Medicare insurance coverage to see if the kinds of treatment, drugs, and services covered will meet your health care needs?” For our outcome variable, the MCBS asks one question about comparing coverage: “during the last open enrollment period, did you compare your Medicare insurance plan with other Medicare plans that were available?”

To accomplish our aim, we estimated two probit regression models (Table A). In model one, we examined the impact of three review exposure behaviors: review of (1) both cost and service information, (2) cost-only information, and (3) service-only information on comparing coverage. In model two, we examined the impact of the interaction of each of the three exposure variables with demographic variables on comparing coverage. We included interactions with demographics that were found to be statistically significant under model one. We ran five separate regressions under model two.

Table A.

Probit Regression Models

Econometric equation #1 (applies to Figure 1):

$$Y_i={\beta }_0+{\beta }_{1}R{B}_i+{\beta }_{2}R{C}_i+{\beta }_{3}R{S}_i+\mathit{Dem}{o}_i{\beta }_4+{\gamma }_{sc}+{\alpha }_t+{\epsilon }_i$$

In our model, Yi is a dummy variable that equals 1 if individual i compared coverage. RBi denotes a high level of health literacy (review if both cost and service information), RCi represents review of cost only information, and RSi signifies review of service only information, both of which are defined as medium health literacy. Demoi includes indicator variables for individual observable characteristics (age categories, sex, race/Ethnicity, education level, marital status, residency in an urban environment, income categories, dual eligibility with Medicaid, at least one living child, and the presence of at least 1 of 9 common chronic conditions).1 State and county fixed effects denoted by γsc were also included. A dummy variable αt was included for each survey year. Finally, εi captured the error rate.

Econometric equation #2 (applies to Table 2):

$$\begin{array}{c}{Y}_i={\beta }_0+{\beta }_{1}R{B}_i+\mathit{Dem}{o}_i{\beta }_2+{\beta }_{3}R{B}_i*\mathit{Dem}{o}_i+{\beta }_{4}R{C}_i+{\beta }_{5}R{C}_i*\mathit{Dem}{o}_i+{\beta }_{6}R{S}_i+\\ {\beta }_{7}R{S}_i*\mathit{Dem}{o}_i+{\gamma }_{sc}+{\alpha }_t+{\epsilon }_i\end{array}$$

and

$${\text{H}}_{\text{Both}}:{\beta }_3=0;{\text{H}}_{\text{Cost}}:{\beta }_5=0;{\text{H}}_{\text{Service}}:{\beta }_7=0$$

¹These conditions included arterial hardening, hypertension, myocardial infarction, stroke, cancer, arthritis, osteoporosis, diabetes, or Alzheimer's disease/related dementias.

All regressions used a probit probability technique and included control variables for demographics (gender, race/ethnicity, urban/rural geography, education, and number of children, income, Medicaid status, age, marriage, and coverage status). All models were weighted using the continuous enrolled topical survey and replicate weights since the MCBS review and coverage questions used in our analysis came from the Medicare plan beneficiary knowledge topical segment of MCBS (Atherly et al., 2023). We included a control variable for state and county fixed effects to address variations in the availability of MA plan options. Finally, we controlled for time fixed effects by including a dummy variable for each year of survey data. This study followed the Strengthening the Reporting of Observational Studies in Epidemiology STROBE reporting guideline (EQUATOR Network, 2025).

Results

Consistent with prior studies, we found 28% (95% confidence interval [CI], 28−29) (Table B) of Medicare beneficiaries compared coverage during the annual open enrollment period (Ochieng et al., 2024). Thirty-nine percent (95% CI, 38−40) of our sample reviewed both cost and service information. Eight percent (95% CI, 8−9) and 6% (95% CI, 6−6) reviewed cost-only or service-only information, respectively. Forty-seven percent (95% CI, 46−47) of beneficiaries did not review cost nor service information.

Table B.

Sample Characteristics (Mean and 95% confidence interval)

	Overall		Cost & Service Review		Cost Only Review		Service Only Review		No Review
Characteristic
Compared Coverage	0.28	[0.28, 0.29]	0.59	[0.58, 0.60]	0.20	[0.18, 0.21]	0.29	[0.27, 0.32]	0.04	[0.04, 0.05]
Review			0.39	[0.38, 0.40]	0.08	[0.08, 0.09]	0.06	[0.05, 0.06]	0.47	[0.46, 0.47]
Male	0.45	[0.45, 0.46]	0.44	[0.43, 0.45]	0.44	[0.42, 0.46]	0.44	[0.41, 0.47]	0.47	[0.46, 0.48]
Race/Ethnicity
White	0.82	[0.82, 0.83]	0.85	[0.84, 0.86]	0.87	[0.85, 0.88]	0.78	[0.76, 0.80]	0.80	[0.79, 0.81]
Black	0.11	[0.10, 0.11]	0.10	[0.10, 0.11]	0.09	[0.08, 0.10]	0.15	[0.13, 0.17]	0.11	[0.10, 0.12]
Hispanic	0.03	[0.03, 0.03]	0.01	[0.01, 0.02]	0.01	[0.01, 0.02]	0.03	[0.02, 0.04]	0.05	[0.04, 0.05]
Asian	0.02	[0.02, 0.02]	0.01	[0.01, 0.02]	0.02	[0.01, 0.02]	0.02	[0.01, 0.03]	0.02	[0.02, 0.02]
Other	0.02	[0.02, 0.02]	0.02	[0.01, 0.02]	0.02	[0.01, 0.02]	0.02	[0.01, 0.02]	0.02	[0.02, 0.03]
Education
Less than High School	0.12	[0.12. 0.12]	0.08	[0.07, 0.08]	0.07	[0.06, 0.08]	0.14	[0.12, 0.16]	0.16	[0.16, 0.17]
High School	0.25	[0.25, 0.26]	0.23	[0.22, 0.23]	0.23	[0.21, 0.25]	0.28	[0.25, 0.31]	0.27	[0.26, 0.28]
More than High School	0.63	[0.63, 0.64]	0.70	[0.69, 0.71]	0.70	[0.68, 0.72]	0.58	[0.55, 0.61]	0.56	[0.55, 0.57]
Married	0.53	[0.52, 0.53]	0.58	[0.58, 0.59]	0.53	[0.51, 0.56]	0.49	[0.46, 0.52]	0.48	[0.47, 0.49]
Urban	0.82	[0.81, 0.82]	0.84	[0.83, 0.84]	0.81	[0.80, 0.83]	0.80	[0.78, 0.82]	0.80	[0.78, 0.80]
Dual	0.10	[0.10, 0.10]	0.06	[0.06, 0.07]	0.05	[0.04, 0.06]	0.14	[0.12, 0.16]	0.13	[0.13, 0.14]
Children	0.85	[0.85, 0.86]	0.86	[0.85, 0.87]	0.87	[0.85, 0.88]	0.85	[0.83, 0.87]	0.84	[0.84, 0.85]
Chronic Conditions	0.84	[0.84, 0.85]	0.85	[0.84, 0.85]	0.84	[0.83, 0.86]	0.87	[0.85, 0.89]	0.84	[0.83, 0.84]
Income
>$5K	0.01	[0.01, 0.01]	0.01	[0.00, 0.01]	0.01	[0.00, 0.01]	0.01	[0.01, 0.02]	0.01	[0.01, 0.02]
$5K-9,999K	0.05	[0.04, 0.05]	0.03	[0.02, 0.03]	0.03	[0.02, 0.03]	0.07	[0.05, 0.08]	0.06	[0.06, 0.07]
$10K-14,999K	0.09	[0.09, 0.10]	0.07	[0.07, 0.08]	0.07	[0.06, 0.08]	0.13	[0.11, 0.14]	0.11	[0.10, 0.12]
$15K-19,999K	0.08	[0.07, 0.08]	0.07	[0.06, 0.07]	0.06	[0.05, 0.07]	0.08	[0.07, 0.10]	0.09	[0.08, 0.09]
$20K-24,999K	0.07	[0.07, 0.07]	0.07	[0.06, 0.07]	0.06	[0.05, 0.07]	0.08	[0.06, 0.09]	0.08	[0.07, 0.08]
$25K-29,999K	0.07	[0.07, 0.07]	0.07	[0.06, 0.08]	0.07	[0.06, 0.09]	0.07	[0.06, 0.09]	0.07	[0.06, 0.07]
$30K-39,999K	0.10	[0.10, 0.10]	0.11	[0.10, 0.11]	0.10	[0.09, 0.12]	0.10	[0.08, 0.12]	0.09	[0.09, 0.10]

$40K-49,999K	0.09	[0.08, 0.09]	0.09	[0.08, 0.10]	0.09	[0.08, 0.11]	0.08	[0.07, 0.10]	0.08	[0.07, 0.09]
<$50K	0.44	[0.44, 0.45]	0.49	[0.48, 0.50]	0.50	[0.48, 0.53]	0.37	[0.35, 0.40]	0.40	[0.39, 0.41]
Age
<65	0.13	[0.12, 0.13]	0.10	[0.09, 0.11]	0.09	[0.08, 0.10]	0.15	[0.13, 0.17]	0.16	[0.15, 0.16]
65-74	0.50	[0.49, 0.50]	0.55	[0.54, 0.56]	0.52	[0.50, 0.54]	0.47	[0.44, 0.49]	0.45	[0.44, 0.46]
75-84	0.29	[0.29, 0.30]	0.29	[0.28, 0.30]	0.32	[0.30, 0.34]	0.30	[0.27, 0.32]	0.29	[0.29, 0.30]
85+	0.08	[0.08, 0.08]	0.06	[0.06, 0.06]	0.07	[0.06, 0.08]	0.09	[0.08, 0.10]	0.10	[0.09, 0.10]
Medicare Advantage	0.42	[0.42, 0.43]	0.51	[0.50, 0.52]	0.36	[0.34, 0.38]	0.46	[0.44, 0.49]	0.35	[0.34, 0.36]
Survey Year 2019	0.32	[0.32, 0.33]	0.31	[0.30, 0.32]	0.32	[0.30, 0.34]	0.32	[0.29, 0.34]	0.34	[0.33, 0.35]
Survey Year 2020	0.33	[0.33, 0.34]	0.33	[0.32, 0.34]	0.32	[0.30, 0.34]	0.34	[0.31, 0.36]	0.34	[0.33, 0.35]
Survey Year 2021	0.34	[0.33, 0.35]	0.36	[0.35, 0.37]	0.36	[0.34, 0.37]	0.35	[0.32, 0.38]	0.32	[0.31, 0.33]
Sample Size	28,924		10,920		2,372		1,738		13,894

Notes: Pooled cross-sectional 2019−2021 data from the Medicare Current Beneficiary Survey limited data set, using recommended survey wei

Reviewing both cost and service information increased the probability of comparing coverage by 159.2% (44.6 absolute percentage points; 95% CI, 44−45) (Figure 1 and Table C), relative to the mean of 28%. Review of cost-only and service-only information were also predictive of comparing coverage, where review of cost-only increased the probability by 68.6% (19.2 absolute percentage points; 95% CI, 17−21) (Table C) and 97.5% (absolute 27.3 percentage points; 95% CI, 25−29) for service-only review.

Figure 1.

Marginal effects of reviewing cost, service, or both cost and service information and observable characteristics on comparing coverage. Notes: Estimates are for a probit probability model and are weighted using the Medicare plan beneficiary knowledge topical segment of Medicare Current Beneficiary Survey (MCBS) using continuous enrolled topical survey and replicate weights. The outcome is comparing coverage and the independent variables reflect review of both cost and service information, cost-only, and service-only information compared to a reference population for no review. *** and ** represent statistical significance at the 1% and 5% levels, respectively. 95% confidence intervals are represented by the bars on each point estimate. Econometric equation presented in Table A. Coefficients for the full regression presented in Table C.

Table C.

Marginal Effects of Health Literacy and Observable Characteristics on Comparing Coverage

Characteristic	Coefficient	95% Confidence Interval	p-value
Health Literacy Level
High	0.446***	[0.44, 0.45]	(0.000)
Medium—cost only	0.192***	[0.17, 0.21]	(0.000)
Medium—service only	0.273***	[0.25, 0.29]	(0.000)
Male	0.005	[−0.01, 0.01]	(0.362)
Race/Ethnicity
Black	0.000	[−0.02, 0.02]	(0.950)
Asian	−0.039	[−0.09, 0.10]	(0.129)
Hispanic	0.001	[−0.03, 0.03]	(0.973)
Other	−0.004	−0.04, 0.04]	(0.854)
Highest Level of Education
Less than High School	0.010	[−0.01, 0.03]	(0.282)
More than High School	0.034***	[0.02, 0.05]	(0.000)
Married	−0.005	[−0.02, 0.01]	(0.438)
Urban Dwelling	0.031	[−0.01, 0.06]	(0.083)
Dual	−0.047***	[−0.07, −0.02]	(0.000)
Children	0.008	[−0.01, 0.02]	(0.306)
One or More Chronic Condition	0.020***	[0.01, 0.04]	(0.319)
Income
$5K-9,999K	−0.015	[−0.07, 0.04]	(0.570)
$10K-14,999K	−0.011	[−0.06, 0.04]	(0.653)
$15K-19,999K	0.011	[−0.04, 0.06]	(0.666)
$20K-24,999K	0.047	[−0.04, 0.10]	(0.073)
$25K-29,999K	0.013	[−0.01, 0.09]	(0.596)
$30K-39,999K	0.017	[−0.03, 0.07]	(0.494)
$40K-49,999K	−0.003	[−0.05, 0.05]	(0.915)
<$50K	−0.009	[−0.06, 0.04]	(0.721)
Age
65−74	0.031***	[0.01, 0.05]	(0.001)
75−84	−0.008	[−0.03, 0.01]	(0.385)
85+	−0.057***	[−0.08, −0.04]	(0.000)
Medicare Advantage Status	−0.002	[−0.01, 0.01]	(0.719)
Surveyed in 2020	−0.011	[−0.02, 0.00]	(0.068)
Surveyed in 2021	−0.001	[−0.01, 0.01]	(0.846)
Observations	28,924

Notes: Estimates are for a probit probability model and are weighted using the Medicare plan beneficiary knowledge topical segment of MCBS using continuous enrolled topical survey and replicate weights. ***and ** represent statistical significance at the 1 and 5% levels, respectively. The outcome variable is comparing coverage and the independent variable reflects level of health literacy with low health literacy (no review) as a reference. All model specifications include the following control indicator variables for male gender (female reference), race/ethnicity (White reference), education level (high school reference), marital status (not married reference), urban dwelling (rural reference) dual eligibility status (non-dual reference), presence of at least one living child (no children reference), presence of at least one chronic condition (no conditions reference), nine income categories ($5K reference) and age (under 65 reference) categories, Medicare Advantage coverage (Fee-for-Service reference), and surveyed in 2020 and 2021 (2019 reference).

We also found five of our demographic variables were predictive of comparing coverage (Figure 1). Having an education greater than high school (3.4 absolute percentage points; 95% CI 2−5), having more than one chronic condition (2.0 absolute percentage points; 95% CI 1−4), and being 65 to 74 (3.1 absolute percentage points; 95% CI 1−5) increased the probability of comparing coverage. Being dual eligible (having both Medicare and Medicaid) (−4.7 absolute percentage points; 95% CI −7 to −2) and being over 85 years old (−5.7 absolute percentage points; 95% CI −8 to −4) decreased the probability of comparing coverage.

Using the demographic predictors that were statistically significant in Figure 1, we created interaction variables with each of the health literacy responses for both, cost-only, and service-only. Only interactions with review of both cost and service information were statistically significant (Table 1). Among the interactions, three variables (both with having an education greater than high school, being age 65 to 74, or being age 85 and older) were statistically significant. Upon furthering testing (Wald) interactions, only being aged 65 to 74 was found to be statistically significantly different from zero. Therefore, we found the probability of comparing coverage increased 2.3 absolute percentage points (95% CI 2−6) among those beneficiaries aged 65 to 74 who reviewed both cost and service information.

Table 1.

Marginal Effects of the Interaction Between Reviewing Cost, Service, or Both Cost and Service Infomation and Observable Characteristics On Comparing Coverage

Characteristic	Health Literacy	Interaction Coefficient	95% CI	Hypothesis Test a
				Wald	p
Education greater than high school	Both cost and service	0.023*	[0, 0.05]	F = 3.45	.063
	Cost-only	−0.009	[−0.05, 0.03]	F = 0.21	.649
	Service-only	0.016	[−0.02, 0.06]	F = 0.61	.434
Dual eligibility	Both cost and service	−0.026	[−0.7, 0.01]	F = 1.58	.209
	Cost-only	0.051	[−0.03, 0.13]	F = 1.52	.217
	Service-only	0.013	[−0.05, 0,07]	F = 0.17	.677
One or more chronic conditions	Both cost and service	−0.222	[−0.06, 0.01]	F = 1.58	.201
	Cost-only	0.008	[−0.04, 0.06]	F = 0.10	.752
	Service-only	−0.036	[−0.11, 0.03]	F = 1.15	.283
Age 65−74 (y)	Both cost and service	0.041***	[0.02, 0.06]	F = 11.21	.000***
	Cost-only	−0.004	[−0.04, 0.03]	F = 0.05	.830
	Service-only	0.022	[−0.02, 0.06]	F = 1.09	.296
Age 85+ (y)	Both cost and service	−0.036**	[−0.07, −0.02]	F = 4.24	.039
	Cost-only	−0.033	[−0.09, 0.02]	F = 1.32	.250
	Service-only	0.002	[−0.06, 0.06]	F = 0.01	.940

Note. Estimates are for a probit probability model and are weighted using the Medicare plan beneficiary knowledge topical segment of Medicare Current Beneficiary Survey using continuous enrolled topical survey and replicate weights. The outcome of comparing coverage is assigned a 1 and not comparing was assigned a 0. ***, **, and * represent statistical significance at the 1%, 5%, and 10% levels, respectively. Observable characteristics were selected for interaction because they were found to be statistically significant in Figure 1. Econometric equation presented in Table A. CI = confidence interval.

^aHypothesis test confirms type of review coefficient does not equal zero. If the p value is statistically significant, null hypothesis is rejected.

Discussion

Our analysis confirmed that less than 30% of Medicare beneficiaries compare coverage. The health choice literature suggests the top reason for this lack of engagement may be inertia toward the status quo (Abaluck & Gruber, 2011). We found evidence that reviewing both cost and service information increased the probability of comparing coverage by nearly 160% and therefore exposure to this intervention may interrupt the inertia mechanism around comparing coverage.

An alternative potential mechanism contributing to lack of engagement around comparing coverage could be difficulty navigating the number of choices. For selection of an MA plan, beneficiaries faced an average of 43 options per county (Freed, Biniek, Damico et al., 2024). Evidence suggests simplifying choice can improve beneficiaries' enrollment decisions (McWilliams et al., 2011). Exposure to cost, service, or both cost and service information, may aide in simplifying the aggregate number of choices that need to be made when comparing coverage.

Beyond reviewing information, several subpopulations were also predictive of comparing coverage. First, we explored the mechanisms behind why dual eligible beneficiaries and those 85 years and older decreased the probability of comparing coverage. There are two possible mechanisms that may explain why those beneficiaries with dual eligibility for Medicare and Medicaid have a decreased likelihood of comparing coverage.

The first potential reason is that beneficiaries receiving Social Security disability income are auto-enrolled (Centers for Medicare & Medicaid Services, 2025) in FFS Medicare after the two-year waiting period that triggers Medicare eligibility. If a beneficiary is not given a choice regarding enrollment, it is far less likely that they would be motivated to compare coverage options. Though auto enrollment may be the most administratively simple way for the Centers for Medicare & Medicaid Services (CMS) to provide coverage to disabled dual eligible beneficiaries, FFS Medicare may not be the best fit. For example, a dual eligible special needs plan (D-SNP), a specific type of MA plan, may better fit the needs of disabled dually eligible beneficiaries. In 2022, 62% of duals in an MA plan were in a D-SNP, with many analysts highlighting fully-integrated D-SNP plans as frequently the best option for beneficiaries (Ma et al., 2023; Medicare Payment Advisory Commission, 2024). Emerging evidence supports this, with a 2024 survey of beneficiaries finding that a larger share of dual eligible individuals in MA plans felt that their coverage met their expectations compared to those in FFS (Jacobson et al., 2024).

The other potential reason dual eligible beneficiaries have a decreased probability of comparing coverage is that they may feel their needs are largely being met. For dual eligible beneficiaries in FFS, there are very few restrictions on where they can consume services, while those in a D-SNP MA plan, may also face relatively few restrictions on services because they are in a plan tailored to their needs. Dual eligible beneficiaries, regardless of the mechanism through which they access their health benefits, experience few or no out-of-pocket costs due to Medicaid wrap around coverage. Additional research is needed to better understand which of these two mechanisms is more likely.

Like dual eligible beneficiaries, those older than age 85 years also have a decreased probability of comparing coverage, but the reasons why may be different. One possible explanation is older beneficiaries' hesitation embracing technology. Several studies have linked advanced age (85+) and hesitancy to use of technology due to five main factors: physical and cognitive barriers, lack of exposure and health literacy, perceived lack of usefulness, privacy and security concerns, and peer social influences (Ahmed et al., 2021; Haase et al., 2021; Zhang, 2023).

Unsurprisingly, when Medicare beneficiaries were asked whether they ever used the official Medicare website (Medicare.gov), 22% of beneficiaries reported they did not have access to the internet (Ochieng et al., 2024). Further, 26% of beneficiaries reported they received information about Medicare from the hotline (1-800-Medicare) and 54% report getting information from the Medicare and You handbook (Ochieng et al., 2024).

Although we are not aware of an analysis that breaks down use of these resources by age group, it is possible those beneficiaries older than age 85 years are more reliant on the CMS hotline and handbook than other age groups. Our analysis in Table 1 for the 85-year-old sub-population reviewed both cost and service information without taking the additional step to compare coverage. More research is needed to understand how the 85-year-old subpopulation, and all beneficiaries, are using the variety of resources managed by CMS, as well as those resources sanctioned or permitted by CMS, such as assistance from SHIP counselors or brokers.

Three of our subpopulations increased the probability of comparing coverage, those age 65 to 74 years, having an education greater than high school, and having one or more chronic condition. One of the main drivers behind the 65- to 74-year-old subpopulation having a higher probability of comparing coverage may be related the same reason the over 85 subpopulation has a lower probability, level of comfort with technology. Medicare beneficiaries age 65 to 74 years are more likely to have the internet (15% more), own a computer (18% more), and own a smart phone (27% more) compared to those age 75 years and older (Koma et al., 2021). These differences in ownership of technology could be driving the different probabilities for these two age subpopulations. Other alternative explanations include greater familiarity with managed care from their more recent prior employment, a type of benefit that the oldest segment of Medicare beneficiaries may have less experience with.

The increased probability, as well as the predictability of the interaction of reviewing both cost and service information and having an education greater than high school may be attributable to the connection between education and reviewing information. Several studies have established a link between educational attainment and health insurance literacy (Braun et al., 2018; Edward et al., 2019; Ghaddar, 2021; Hero et al., 2019; Holst et al., 2023). The more educated individuals are, the easier it is to understand health insurance complexities, such as breadth of the service network and annual out-of-pocket costs.

Finally, one potential mechanism behind the increased probability of comparing coverage among those beneficiaries with one or more chronic conditions is recent research that found these beneficiaries have a decreased probability of sticking with their MA plan (Grabert et al., 2026). Several studies have linked enrollment in an MA plan with increased barriers to access for beneficiaries with complex medical conditions, compared to FFS (Achola et al., 2023; Ankuda et al., 2023; Grabert et al., 2026; Rahman et al., 2015). These barriers may include things like prior authorization of medically necessary services or limited network adequacy, both of which may prompt a beneficiary to review cost and/or service information and ultimately compare. However, more research is needed as not all chronic conditions have the same clinical and financial burden, with diabetes resulting in ophthalmology and podiatry visits, medication use and potential complications such as neuropathy, nephropathy, and retinopathy driving increased medical care, loss of functional status, and associated expenditures (Miller, Zima, & Wilensky, 2022). In contrast, gastroesophageal reflux disease is a chronic condition and far more benign clinically and financially (Snyder & Katzka, 2022).

Limitations

Our analysis is cross-sectional, associational, and does not imply causation. Like many self-reported surveys, responses may be subject to recall bias. We note 2 of our 3 years of data for this analysis included reviewing information and comparing coverage during the coronavirus disease 2019 (COVID-19) pandemic. One limitation related to the COVID-19 pandemic is the reliance on phone calls to complete the survey, where previous non−COVID-19 years relied on a mix of in-person interviews and phone calls. To address this potential bias, we included time fixed-effects dummy variables for 2020 and 2021, and we did not find either variable to be statistically significant compared to the reference year of 2019 (Table A and Table C). In addition, external influences (such as COVID-19 misinformation reported by the media) may have had an impact on whether beneficiaries compare coverage. However, it is difficult to surmise as to whether such an impact would result in a positive or negative effect on the coefficient. Since COVID-19 exposure can have a significant impact on overall health, it may induce beneficiaries to hyper-analyze their health care coverage decisions. But COVID-19 exposure is also correlated with mental health outcomes, such as increased anxiety, which may manifest in the desire to keep health-related factors constant, including coverage choice. Overall, it is difficult to determine what impact COVID-19, if any, had on whether a beneficiary compared coverage.

The survey data and results are only reflective of the Medicare population and therefore may not be reflective of other populations. Further, we note that previous literature has recognized review of cost and/or service information and comparing coverage is defined as a high level of health literacy for the Medicare population, only. Therefore, we urge caution when applying our results regarding health literacy outside of the Medicare population.

Conclusion

Comparing coverage is an important action all Medicare beneficiaries should consider on an annual basis. The greater the exposure to cost, service, or both cost and service information, the greater the probability of comparing coverage. Reviewing both cost and service information increased the probability of comparing coverage by nearly 160% and should be among the top mechanisms policymakers consider toward achieving the goal of increasing the percentage of beneficiaries who actively compare coverage. Our analysis indicated the probability of comparing coverage increased for those beneficiaries with an education greater than high school, age 65 to 74 years, and with more than one chronic condition and that the probability of comparing coverage decreased among those aged 85 and higher and those dually eligible for Medicare and Medicaid. Factors that may contribute to these correlations include different levels of acceptance of technology and plan benefit features associated with choice. A targeted outreach and education campaign exposing beneficiaries to cost and service information could increase the rate of comparing coverage. Future research is needed to determine which factors are predictive of comparing coverage.