Medicare Part D beneficiaries’ self-reported barriers to switching plans and making plan comparisons at all

Abstract

In the United States, individuals with disabilities and those aged ≥65 can supplement their Medicare with so-called stand-alone Medicare Part D prescription drug plans. Beneficiaries can switch their stand-alone prescription drug plans annually, but most do not. Indirect evidence has raised concerns that non-switchers do not even make plan comparisons (labeled “inattention”), but direct evidence is scarce. Therefore, we surveyed 439 beneficiaries of Medicare Part D plans from a nationally representative adult sample after the 2024 open-enrollment period. Overall, 53% self-reported making no comparisons. Of those who did not compare, 98% did not switch (vs 67% of those who did compare). Multinomial regressions revealed that beneficiaries who neither compared nor switched were more likely than switchers to report difficulties with comparing and switching, experiencing no plan-related discontinuation, changes, or dissatisfaction, not using advisors or the plan-finder website, and receiving potentially confusing mailings. Non-switchers who did compare were similar to switchers in reporting few difficulties and relying on advisors and the plan-finder website, but they were less likely than switchers to report plan-related changes, discontinuation, or dissatisfaction, while being more likely to report receiving mailings and having no college degree. We discuss insights for policy-making.

Medicare Part D, choice avoidance, inattention

Introduction

In the United States, government-subsidized prescription drug insurance is offered to individuals with disabilities and those aged ≥65 through a program called “Medicare Part D.” Beneficiaries can choose to have their prescription drugs covered through a stand-alone Medicare Part D prescription drug plan that supplements traditional Medicare, or as part of a Medicare Advantage plan that bundles prescription drug insurance with other types of health insurance.¹ If beneficiaries opt for a stand-alone prescription drug plan, they can choose from a multitude of plans, provided by private insurers who have contracts with the government and compete for beneficiaries. For their 2024 prescription drug coverage, the average beneficiary could choose from about 20 stand-alone Medicare Part D plans.¹ This market competition is meant to drive down prices and give beneficiaries the opportunity to choose the plan that is financially optimal for them. Henceforth, the financially optimal stand-alone prescription drug plan is defined as the plan that would require the lowest total beneficiary spending, given the beneficiary's prescription drug use as well as available plans’ premiums, deductibles, and a potential “donut hole” phase during which prescription drug coverage is temporarily limited.^2-7

In 2023, 22 million of the 65 million Medicare beneficiaries had a stand-alone prescription drug plan.¹ In the annual open-enrollment period that ran from October-December 2023, these beneficiaries were offered a choice to keep their stand-alone prescription drug plan, or to switch to another plan for the next year. This annual choice allows them to select the best plan in light of potential changes in available plans, as well as changes in their situation and medication needs.^8,9 Beneficiaries who do not make a choice are typically auto-enrolled into their existing plan, even if their insurance company is changing it or replacing it with another plan for the next year.^2-4

Yet, few beneficiaries switch their stand-alone Medicare Part D plans each year, which means they may be sticking with plans that become less financially optimal to them over time.² In 2009, only 5% of beneficiaries had the financially optimal plan given their prescription drug use, overspending on average $368 per year, with some overspending much more.³ If all beneficiaries chose the financially optimal plan, it would save the government an estimated $1.3 billion every 3 years.⁷ Failure to switch to a financially optimal prescription drug plan can have health consequences, because beneficiaries may try to manage their costs by limiting their use of prescribed medications.¹⁰

Traditionally, it was assumed that individuals would only stick with their default prescription drug plan if, after comparing the available plans, they decided that they wanted to keep the plan they had.^11,12 Some beneficiaries may indeed compare plans and decide not to switch, and then become more likely to switch their stand-alone prescription drug plan as their potential gains from switching increase over subsequent years.^13,14 However, indirect evidence from administrative data suggests that many Medicare Part D beneficiaries with stand-alone prescription drug plans do not switch because they are “inattentive” or make no plan comparisons at all.^2,7 A natural experiment with low-income beneficiaries suggested that most did not even switch when the potential gains of switching were large, perhaps in part because impending plan changes often go unnoticed.³ While these analyses of administrative data seem to suggest that many non-switchers make no plan comparisons at all, no direct evidence has been collected to support this claim.

Direct evidence about whether non-switchers may indeed avoid comparing stand-alone prescription drug plans is scarce. Interviews with 35 beneficiaries did reveal that they found plan comparisons difficult because they perceived plans as too similar, they did not have help from an advisor, and they felt overwhelmed by the many mailings from multitudes of plans.¹⁵ Surveys confirmed that beneficiaries often find it hard to identify the best plan, and that non-switchers are less likely to have help from advisors.^16,17 The psychology of decision making has posited that feeling overwhelmed with the number of options and the choice process can lead to making no comparisons at all.^18,19 Additionally, complex administrative burdens—also referred to as “frictions” or “sludge”—may undermine people's motivation to compare, or their motivation to switch if they do compare.^19,20

Here, we conducted surveys after the 2024 open-enrollment period in which we directly asked Medicare Part beneficiaries with stand-alone prescription drug plans whether or not they compared prescription drug plans. Specifically, we aimed to assess (1) the extent to which not making plan comparisons was associated with not switching, and (2) the features of non-switchers who did not compare, and non-switchers who did compare (vs switchers).

Method

Sample

Our participants were members of the nationally representative adult sample of the Understanding America Study (UAS), which is run by the University of Southern California's Center for Economic and Social Research.²¹ US adults age 18 and older were originally recruited into the UAS through random address-sampling, while oversampling individuals from underrepresented groups, and providing internet access and equipment as needed.²¹ UAS members are regularly invited to participate in surveys on a variety of topics. For our cross-sectional survey, we invited those 698 UAS members who had demographic information on file, and who had, on surveys conducted between 2019 and 2022, indicated they were enrolled in “Medicare Part D, also known as the Medicare Prescription Drug Plan.” Of those 698 invitees, 634 (91%) responded. Following previous research, we focused only on Medicare Part D beneficiaries who had a stand-alone prescription drug plan.^2-4,6,7 To be eligible to participate, invited UAS members had to answer “yes” (vs “no,” or “don’t know”) to a question asking whether they had a stand-alone Medicare Part D prescription drug plan before October 2023, and another question asking whether they would have one in 2024. A stand-alone Medicare prescription drug plan was defined as “a separate prescription drug coverage insurance plan that you purchase through a company such as Aetna, Cigna, Humana, Blue Cross, and others,” and “different from Medicare Advantage plans, in which private companies provide all of your health care, for example through an HMO or PPO plan.” Participants who had a Medicare Part D prescription drug plan as part of a Medicare Advantage health insurance plan were ineligible to participate, because their prescription drug plan choices are not solely based on their medication needs.⁴

Overall, 439 UAS members were eligible to participate. Table 1 shows that invited participants who responded (vs did not respond) to our survey invitation were largely similar in terms of demographics, except that those who responded were older and more likely to be college-educated. Table 1 also shows that responders who were eligible (vs ineligible) to participate were less likely to report living on low income, and more likely to report being non-Hispanic White, older, and college-educated.

Table 1.

Sample characteristics.

	Out of those invited (N = 698)			Out of those who responded (N = 634)
Demographics	Responders (N = 634)	Non-responders (N = 64)	Statistical test	Eligible (N = 439)	Ineligible (N = 242)	Statistical test
% race/ethnic minority	17%	20%	ꭓ(1) = 0.35, P = 0.56	13%	26%	ꭓ(1) = 15.78, P < 0.001
% age < 65	13%	22%	ꭓ(1) = 4.09, P = 0.04	9%	20%	ꭓ(1) = 15.39, P < 0.001
% no college degree	54%	69%	ꭓ(1) = 4.84, P = 0.03	51%	63%	ꭓ(1) = 8.00, P < 0.01
% female	56%	59%	ꭓ(1) = 0.35, P = 0.55	56%	55%	ꭓ(1) = 0.00, P = 0.95
% low income	18%	28%	ꭓ(1) = 3.90, P = 0.05	14%	27%	ꭓ(1) = 14.41, P < 0.001
% high income	12%	13%	ꭓ(1) = 0.04, P = 0.85	12%	10%	ꭓ(1) = 0.55, P = 0.46

Participants were recruited through the nationally representative Understanding America Study (UAS), a web-based survey panel run out of the University of Southern California's Center for Economic and Social Research. Eligible responders were included in the data analysis. That is, they had to answer “yes” (vs “no,” or “don’t know”) to 2 questions asking whether they had a stand-alone Medicare Part D prescription drug plan before October 2023, and would have one in 2024. Race/ethnic minority was defined as not identifying as non-Hispanic White. Beneficiaries at age < 65 were eligible for Medicare Part D due to disabilities. Low income was defined as <150% of the Federal Poverty Line,¹⁵ and high income as >500% of the Federal Poverty Line.²

Survey

Our survey was fielded from December 8, 2023 through January 13, 2024, following the open-enrollment period (October 15-December 7, 2023) during which beneficiaries could choose their plan for 2024. To ensure that survey questions covered relevant topics, they were informed by interviews we conducted with Medicare Part D beneficiaries about barriers to comparing and switching plans, following established methods.²² The survey and associated data are available from the Understanding America Study (https://uasdata.usc.edu/index.php, UAS #587).

Comparing and switching

Whether or not participants switched was assessed by asking: “Between October 15 and December 7 of 2023, did you switch to a different stand-alone prescription drug plan?” Response options were “yes,” “no,” or “don’t know.” If participants answered “yes” they were asked whether they “made the switch to a different plan” or “my insurance company switched out my plan.” Ten participants indicated the latter, suggesting that they stuck with a default provided by their insurance company. Only participants who made switches themselves were counted as having actively switched (=0), and the remainder was treated as not having actively switched (=1). Whether or not participants compared plans was assessed with the statement “I compared different plans for 2024” accompanied by response options “yes” (=0) or “no” (=1). Based on these questions, we identified 3 groups: non-switchers who did not compare (also referred to as “inattentive”²), non-switchers who did compare, and switchers.

Difficulties comparing and switching

Participants responded to the statements “I knew how to switch to a different plan for 2024” (“yes” = 0; “no” = 1), and “All plans seemed pretty much the same to me” (“yes” = 1; “no” = 0). They were also asked “how easy would it be for you to choose a new stand-alone Medicare Part D prescription drug plan today?” (“very easy” and “pretty easy” = 0; “not that easy” and “not easy at all” = 1).

Absence of plan-related issues

Participants responded to the statements “My 2023 plan will be discontinued in 2024” (“yes” = 0; “no” = 1), “My 2023 plan will be changed in 2024” (“yes” = 0; “no” = 1), “I was pretty satisfied with the plan I had throughout 2023” (“yes” = 1; “no” = 0), and “I hit the donut hole in 2023” (“yes” = 0; “no” = 1).

Absence of needs-related issues

Participants indicated whether their situation, their medication, or their state of residence had changed. Response options were “yes” (=0) and “no” (=1).

Information sources

Participants indicated whether (=1) or not (=0) during 2023 they had received information from: an advisor or insurance broker, Medicare plan-comparison website, friends or family, mailings, their pharmacy, an employer, the AARP, and the Medicare booklet.

Demographics

The Understanding America Study had information on file about participants’ race/ethnicity, age, education, gender, income, marital status, and state of residence. Based on reported income, marital status, and state of residence, we computed whether participants had low income (<150% federal poverty level [FPL]), which is a requirement for receiving low-income subsidies.²³ We also computed whether participants had high income (>500% FPL), following previous research on this topic.²

Results

Is not comparing related to not switching?

Among our 439 participating beneficiaries, 53% reported not comparing plans. Of those who did not compare plans, 98% did not switch, while only 67% of those who did compare ended up not switching, ꭓ²(1) = 75.09, P < 0.001 (Figure 1). This relationship held even when accounting for demographics in a logistic regression (Table S1). Overall, we were able to categorize 52% of participants as non-switchers who did not compare, 32% as non-switchers who did compare, and 17% as switchers. In previous research, non-switchers who did not compare were referred to as “inattentive,” while the other 2 groups have been referred to as “attentive.”² Indeed, non-switchers who did compare can be considered attentive because they presumably compared but decided not to switch.

Figure 1.

Percent of beneficiaries who did not switch, among those who did not (vs did) compare plans.

What are the features of non-switchers who did not compare and non-switchers who did compare (vs switchers)?

Descriptive statistics indicated that non-switchers who also did not compare (and have been referred to as “inattentive”²) were different from switchers in the following ways (Table 2). First, non-switchers who did not compare were more likely than switchers to report difficulties with comparing and switching, such as perceiving plans as the same (46% vs 12%), ꭓ²(1) = 26.86, P < 0.001 and not knowing how to switch (41% vs 12%), ꭓ²(1) = 20.88, P < 0.001. Second, non-switchers who did not compare were more likely than switchers to report that their plans were continued (93% vs 77%), ꭓ²(1) = 14.38, P < 0.001, unchanged (81% vs 32%), ꭓ²(1) = 59.94, P < 0.001 and satisfactory (88% vs 69%), ꭓ²= 14.70, P < 0.001—while changes in situation, medication, or state of residence were similarly uncommon for both. Third, non-switchers who did not compare were less likely than switchers to report relying on an advisor (21% vs 51%), ꭓ²(1) = 25.63, P < 0.001, the Medicare plan-finder website (12% vs 55%), ꭓ²(1) = 58.25, P < 0.001, or friends and family (16% vs 38%), ꭓ²(1) = 16.63, P < 0.001, and more likely to report receiving mailings (54% vs 35%), ꭓ²(1) = 7.92, P < 0.01. In terms of demographics, non-switchers who did not compare were more likely than switchers to identify with race/ethnic minorities (16% vs 7%), ꭓ²(1) = 20.88, P = 0.04, have no college degree (53% vs 38%), ꭓ²(1) = 5.19, P = 0.02, and report low income defined as <150% of the Federal Poverty Line,¹⁵ (20% vs 8%), ꭓ²(1) = 20.88, P = 0.02.¹⁵ Multinomial logistic regressions that accounted for all of these features found that all differences remained significant, except for those pertaining to perceiving all plans as the same, relying on friends and family for information, and demographic differences (Table 3).

Table 2.

Characteristics of non-switchers who did not compare, non-switchers who did compare, and switchers.

Potential correlate	Non-switchers who did not compare (N = 226)	Non-switchers who did compare (N = 139)	Switchers (N = 74)
Difficulties comparing and switching
% did not know how to switch	41%c,s	14%	12%
% perceived all plans as the same	46%c,s	24%s	12%
% perceived plan choices as hard	35%	25%	24%
Absence of plan-related issues
% previous plan was continued	93%s	91%s	77%
% previous plan was unchanged	81%c,s	67%s	32%
% satisfied with previous plan	88%s	86%s	69%
% avoided donut hole	92%	89%	87%
Absence of needs-related issues
% situation unchanged	98%	98%	97%
% medication unchanged	81%	75%	73%
% state of residence unchanged	99%	99%	99%
Information sources
% advisor	21%	45%n	51%n
% plan-finder website	12%	41%n	55%n
% friends or family	16%	26%n	38%n
% mailings	54%s	60%s	35%
% pharmacy	4%	5%	7%
% employer	4%	3%	3%
% AARP	16%	14%	14%
% Medicare booklet	24%	31%	19%
Demographic control variables
% race/ethnic minority	16%s	12%	7%
% age < 65	11%	8%	7%
% no college degree	53%s	54%	38%
% female	59%	52%	51%
% low income	20%c,s	9%	8%
% high income	13%	11%	12%

Race/ethnic minority was defined as no identifying as non-Hispanic White. Beneficiaries at age < 65 were eligible for Medicare Part D due to disabilities. Low income was defined as <150% of the Federal Poverty Line,¹⁵ and high income as >500% of the Federal Poverty Line.²

^cHigher than non-switchers who did compare (chi-square test, P < 0.05).

ⁿHigher than non-switchers who did not compare (chi-square test, P < 0.05).

^sHigher than switchers (chi-square test, P < 0.05).

Table 3.

Multinomial logistic regression predicting self-reports of being a non-switcher who did not compare and a non-switcher who did compare (vs switcher).

Potential correlate	Non-switchers who did not compare (vs switchers)		Non-switchers who did compare (vs switchers)
	OR	95% CI	OR	95% CI
Difficulties comparing and switching
Did not know how to switch	5.62**	(2.03, 15.52)	1.47	(0.52, 4.14)
Perceived all plans as the same	2.24	(0.83, 6.04)	1.34	(0.49, 3.62)
Perceived plan choices as hard	0.56	(0.23, 1.35)	0.70	(0.31, 1.59)
Absence of plan-related issues
Previous plan was continued	4.14*	(1.27, 13.52)	3.43*	(1.25, 9.42)
Previous plan was unchanged	8.98***	(3.97, 20.28)	4.04***	(1.95, 8.39)
Was satisfied with previous plan	5.83***	(2.19, 15.52)	3.17**	(1.34, 7.48)
Avoided donut hole	1.26	(0.40, 4.04)	1.27	(0.44, 3.69)
Absence of needs-related issues
Situation was unchanged	0.95	(0.09, 10.23)	1.28	(0.14, 11.43)
Medication was unchanged	1.02	(0.42, 2.47)	1.00	(0.44, 2.26)
State of residence was unchanged	0.02	(0.00, 1.77)	0.12	(0.00, 8.66)
Information sources
Advisor	0.08***	(0.04, 0.20)	0.47	(0.22, 1.02)
Plan-finder website	0.09***	(0.03, 0.21)	0.46	(0.21, 1.03)
Friends or family	0.55	(0.24, 1.29)	0.54	(0.25, 1.18)
Mailings	3.57**	(1.61, 7.95)	3.63**	(1.72, 7.67)
Pharmacy	0.92	(0.14, 6.17)	1.50	(0.27, 8.21)
Employer	0.62	(0.09, 4.24)	0.63	(0.09, 4.34)
AARP	0.62	(0.21, 1.86)	0.53	(0.18, 1.53)
Medicare booklet	1.55	(0.60, 3.99)	2.20	(0.91, 5.28)
Demographic control variables
Race/ethnic minority (vs not)	3.02	(0.70, 13.00)	2.91	(0.71, 11.94)
Age < 65 (vs not)	3.18	(0.75, 13.47)	2.41	(0.66, 8.78)
No college degree (vs degree)	2.08	(0.95, 4.55)	2.38*	(1.15, 4.91)
Female (vs not)	0.73	(0.34, 1.56)	0.61	(0.30, 1.25)
Low income (vs mid income)	1.93	(0.55, 6.76)	0.78	(0.22, 2.77)
High income (vs mid income)	0.93	(0.29, 2.95)	0.78	(0.22, 2.77)
Nagelkerke R2= 0.52, N = 439

Odds ratios reflect the ratio of the odds of an outcome occurring in 1 group relative to the odds of it occurring in a reference group, in line with our research question. An odds ratio that is significant and larger than 1.00 indicates a significant positive relationship. An odds ratio that is significant and smaller than 1.00 indicates a significant negative relationship. Marginal effects appear in Table S2. Race/ethnic minority was defined as not identifying as non-Hispanic White. Beneficiaries at age < 65 were eligible for Medicare Part D due to disabilities. Low income was defined as <150% of the Federal Poverty Line,¹⁵ and high income as >500% of the Federal Poverty Line.²

OR, odds ratio; CI, confidence interval.

^*P < 0.05.

^**P < 0.01.

^***P < 0.001.

We also compared non-switchers who did compare with switchers. While both groups may have actively paid attention to the plan choice,² any differences between the two may indicate reasons not to switch. Descriptive statistics indicated that non-switchers who did compare and switchers were similar in reporting that they knew how to switch, but non-switchers who did compare were more likely to perceive all plans as the same (24% vs 12%), ꭓ²(1) = 4.09, P = 0.04 and more likely to report plans continued (91% vs 77%), ꭓ²(1) = 8.44, P < 0.01, unchanged (67% vs 32%), ꭓ²(1) = 23.18, P < 0.001, and satisfactory (86% vs 69%), ꭓ²(1) = 9.25, P < 0.01 (Table 2). Both groups reported similar use of information sources except that non-switchers who did compare were more likely to report receiving mailings (60% vs 35%), ꭓ²(1) = 12.37, P < 0.001. Multinomial logistic regressions that accounted for all features found that these differences remained significant, except for differences pertaining to perceiving plans as the same (Table 3). In terms of demographics, the multinomial regression suggested that non-switchers who did compare were significantly more likely than switchers to report having no college degree (Table 3).

Discussion

Most Medicare beneficiaries do not switch their stand-alone prescription drug plans from year to year, and non-switchers are less likely to end up with financially optimal plans.^2-4,7 Indirect insights from administrative data analyses have suggested that not switching may partly be due to not even comparing stand-alone prescription drug plans—also referred to as “inattention.”^2,7 However, direct evidence is scarce. We therefore surveyed 439 Medicare Part D beneficiaries with stand-alone prescription drug plans from a national sample of US adults to directly ask them whether or not they compared, whether or not they switched, and why. Specifically, we assessed (1) the relationship between not comparing and not switching, and (2) features of non-switchers who did not compare and non-switchers who did compare (vs switchers).

Our first main finding was that, of the 53% who did not (vs did) compare, 98% (vs 67%) did not switch. These analyses confirm indirect insights from administrative data, which suggested that many beneficiaries do not switch their stand-alone prescription drug plans because they remain “inattentive” and do not even compare plans.² As a result of not comparing and switching, they may end up with less optimal stand-alone prescription drug plans over time.²

Our second main finding reveals 3 potential barriers to comparing plans at all. First, beneficiaries who did not compare or switch plans were more likely than switchers to report not even knowing how to go about making a switch. These findings align with the psychology of decision making, which has posited that choice avoidance, or not engaging with the choice at all, is a common response to feeling overwhelmed with the choice,¹⁸ and with administrative burdens also referred to as “frictions” or “sludge.”^19,20 Second, beneficiaries who did not compare or switch were less likely than switchers to report plan issues such as discontinuations, changes, and dissatisfaction. Indeed, people may simplify difficult choices by using an “if it ain’t broke don’t fix it” approach, resulting in “status quo bias” and sticking with the default.²⁴ Third, beneficiaries who did not compare or switch were less likely than switchers to report relying on advisors or the plan-finder website and more likely to report getting mailings. Indeed, interviews with small samples of beneficiaries have suggested that advisors can be helpful and plan mailings can be overwhelming.¹⁵

Our third main finding pertains to potential reasons for not switching, among participants who did compare plans. After comparing the available plans, these participants stayed with their current plan. These non-switchers who did compare were similar to switchers in perceiving few difficulties with switching, suggesting that it would have been possible for them to switch if they wanted to. However, we did find that participants who compared without switching were more likely than switchers to report experiencing no plan-related changes, discontinuation, or dissatisfaction. While it is possible that non-switchers who did compare had better default plans than switchers, it is also possible that non-switchers who did compare did not notice that alternative plans had become financially optimal.³ Additionally, while non-switchers who did compare were as likely as switchers to rely on the advisors or the plan-finder website they were more likely to report receiving mailings, and perhaps those mailings undermine switching due to being overwhelming.¹⁵ Non-switchers who did compare were also less likely than switchers to have no college degree, suggesting that, among individuals who make comparisons, overcoming barriers to switching may be harder for those without a college degree.

Like any study, ours has limitations. First, we lack the information needed to evaluate whether participants who did not compare plans were least likely to end up with financially optimal plans. However, analyses of administrative data have suggested that not comparing plans is related to not having financially optimal plans.^2-4,7 Second, likely because beneficiaries with stand-alone prescription drug plans were recruited from a nationally representative sample of US adults age 18 and older, our sample was relatively small. Third, despite having been recruited through a nationally representative sample of US adults, beneficiaries who ended up participating in our survey may not be entirely representative of all Medicare Part D beneficiaries with stand-alone prescription drug plans in the national population, due to for example being more likely to have college degree and higher income (Table 1). To the extent that participants with a college degree and higher income are more likely to report switching (Table 3), our findings may have slightly underestimated the likelihood of not switching. Indeed, 83% of our participants reported not switching their stand-alone prescription drug plan, as compared to 87%-90% of participants in previous studies of administrative data.² However, these administrative data were collected over a decade ago, before the implementation of various Medicare Part D reforms and an update of the Medicare Part D Plan Finder.^25-27 Moreover, the main focus of our analyses was not on estimating non-switching rates, but on predictors of not switching and not comparing, which may be less affected by the representativeness of the sample. Fourth, we had no direct information about whether participants received low-income subsidies, but participants who reported having low income were not significantly more or less likely to be non-switchers who did not compare, or non-switchers who did compare, or switchers (Table 3). Our main findings are also unaffected by removing low-income participants from the multinomial logistic regression (Tables S3 and S4). Fifth, we were unable to verify participants’ self-reports of (not) comparing, (not) switching, (not) having plan changes, or other experiences. Previous findings with low-income beneficiaries have suggested that plan changes may only sometimes catch beneficiaries’ attention and motivate switching.³ Fifth, presented analyses are correlational and do not warrant causal conclusions.

Yet, our findings suggest that there are barriers to making plan choices that may keep beneficiaries from switching to financially optimal plans, or even comparing plans at all. Various interventions have been suggested for helping to overcome barriers to making plan comparisons. It has been suggested to simplify plan choices in the plan-finder website, by making plan information easier to understand, and by reducing the number of plans,^28-31 as well as offering experts’ advice.^32-34 However, most beneficiaries who could benefit from switching still do not switch when they receive such interventions, perhaps because the interventions are not motivating them to make comparisons. Indeed, many do not even engage with the plan-finder website.³⁵

Since our findings suggest that many beneficiaries may not even bother to compare plans, they may welcome policies that reduce the need to make comparisons at all. It has been suggested to reduce the need to make comparisons by auto-enrolling beneficiaries into more optimal defaults with the option to opt out to another plan, and requiring plan choices only once every few years instead of every year.³⁶ While defaults have been successfully implemented to simplify some financial choices, suggestions to use auto-enrollment into defaults can sometimes be perceived as threats to consumer autonomy and consumer welfare.³⁷ Thus, the political feasibility of these strategies remains to be seen. Meanwhile, Medicare Part D beneficiaries continue to be faced with difficult stand-alone prescription drug plan choices, which may prevent many from making comparisons and lead them to stay with financially less optimal plans.

Acknowledgments

We thank Jill Darling and her team at the Understanding America Study for survey programming and data collection.

Supplementary material

Supplementary material is available at Health Affairs Scholar online.

Funding

This project was supported by a gift from the University of Southern California to the USC Schaeffer Center for Health Policy and Economics. Additional funding was provided by Deutsche Forschungsgemeinschaft (DFG, grant HE 6902/2-1).

Data availability

The survey and associated data are available from the Understanding America Study (https://uasdata.usc.edu/index.php, UAS #587).

Author notes

Conflicts of interest Please see ICMJE form(s) for author conflicts of interest. These have been provided as supplementary materials.