«By JONATHAN D. KETCHAM, NICOLAI V. KUMINOFF AND CHRISTOPHER A. POWERS* We develop a structural model for estimating the welfare effects of poli- cies ...»
Section VII presents main results from the policy experiments and section VIII summarizes robustness checks. Concluding remarks are provided in Section IX.
I. Medicare Part D
US citizens typically become eligible for Medicare benefits when they turn 65. In 2006, Medicare Part D extended these benefits to include prescription drug insurance. A novel and controversial feature of Part D is that it created quasi-private marketplaces for delivering insurance.4 Part D created 34 state or multistate markets within which the average enrollee chose among 50 standalone prescription drug insurance plans (PDPs) sold by 20 private insurers.5 The default for new beneficiaries is to be uninsured.6 After an enPrior to the ACA, Part D was the largest expansion of public insurance programs since the start of Medicare.
5 Subject to CMS approval, insurers can sell multiple PDPs in each market and make annual changes to existing plans.
5 rollee chooses a plan she is automatically reassigned to the same plan the following year unless she switches to a different one during open enrollment. Enrollees pay monthly premiums as well as out of pocket (OOP) costs for the drugs they purchase and taxpayers subsidize the total costs of non-poor enrollees by an average of 75.5%.
PDPs differ in terms of premiums, OOP costs of specific drugs, and quality measures such as customer service, access to pharmacy networks, the ability to obtain drugs by mail order, and the prevalence and stringency of prior authorization requirements.7 The novelty of the market together with the complexity of the product led many analysts to speculate that consumers would struggle to navigate the market. Liebman and Zeckhauser (2008) summarize this view when they write that: “Health insurance is too complicated a product for most consumers to purchase intelligently and it is unlikely that most individuals will make sensible decisions when confronted with these choices.” Some analysts flagged Part D as a candidate for libertarian paternalism (McFadden 2006, Thaler and Sunstein 2008). Moreover, the government has expressed a desire to simplify health insurance markets and nudge enrollees toward cheaper plans. In 2014, CMS proposed limiting insurers to selling no more than two plans per region, which would reduce the average consumer’s choice set by about 20% (Federal Register 2014). The US Department of Health and Human Services also announced that it is considering redesigning federal health insurance exchanges to automatically reassign people to low-cost plans unless they opt out (Health and Human Services 2014). The welfare effects of these types of policies will depend on consumers’ preferences for PDP attributes, the cost of switching plans, and how the policies affect consumers’ decision processes and market outcomes.
Several prior studies have investigated the role of information and consumer behavior in Medicare Part D. Over the first five years of the program, the average enrollee could have reduced annual expenditures (premium + out of pocket) by 25% (or $341) by switching to their cheapest available plan (Ketcham, Lucarelli and Powers 2015). Yet, the implications for consumer welfare remain ambiguous. When enrollees are surveyed about 6 Enrollees who qualify for low-income subsidies are autoenrolled to certain plans, but we exclude them from our empirical analysis.
7 Many insurers require consumers to have prior authorization from a doctor in order to obtain certain drugs, but the stringency of these requirements differs from insurer to insurer.
6 their experiences in Part D most report being satisfied with the plans they chose (Heiss, McFadden and Winter 2010, Kling et al. 2012). Furthermore, Ketcham, Kuminoff and Powers (2015) demonstrate that most of the people who could have saved money by switching had chosen plans that were either superior in some measure of quality or provided greater protection from negative health shocks. These consumers could be making informed decisions to pay for quality and risk protection. On the other hand, when Kling et al. (2012) asked 406 Wisconsin enrollees how much they thought they could save by switching plans, most respondents underestimated the true figure. Kling et al. also found that sending enrollees a letter with personalized information about their potential savings increased the rate at which enrollees switched plans by 11.5 percentage points. Overall, the existing evidence suggests that some consumers are misinformed, but others may be choosing to pay more for plans with higher quality and/or greater risk protection.
II. Linking Administrative Records to Enrollee Surveys
For the first time in academic research, we have linked the Medicare Current Beneficiary Survey (MCBS) to the respondents’ administrative records at the US Centers for Medicare and Medicaid Services (CMS). The MCBS is a national rotating panel questionnaire that began in 1991 and is administered to approximately 16,000 people annually.8 It collects information about Medicare beneficiaries and their use of health care services. Each participant is interviewed up to three times per year for four consecutive years, regardless of whether they stay at the same address or move into and out of long term care facilities. Importantly for our purposes, participants are tested on their knowledge of the PDP market. The MCBS also asks participants if and how they searched for information about Medicare services and it provides rich demographic data.
8 A potential limitation of working with the MCBS sample is that it is not designed to be nationally representative without weighting, and selecting the appropriate weights is complicated by panel rotation and by our exclusive focus on respondents who participated in the standalone PDP market. Respondents who do not purchase a standalone PDP can instead obtain prescription drug insurance through an employer sponsored plan or a Medicare Advantage plan. Further, the MCBS does not sample individuals from 3 PDP regions: 1(Maine and New Hampshire), 20 (Mississippi), and 31 (Idaho and Utah). To assess whether using unweighted MCBS data might compromise the external validity of our results, we compared the unweighted demographics of the average enrollee in our linked sample with a random 20% sample of all Part D enrollees from CMS’s administrative files. Table A1 shows that the average enrollee in our linked sample is 1 to 2 years older. Otherwise, the two samples are virtually identical in terms of race, gender, rates of dementia and depression, number of PDP brands and plans available, expenditures on plan premiums and OOP costs, and the maximum amount of money that the average enrollee could have been saved by enrolling in a different plan. Given the strong similarity between the two samples, we expect that our findings from the linked MCBS-administrative sample can be generalized to the broader population of non-poor Part D enrollees.
7 Also of particular value for our study, the MCBS indicates whether a proxy responded to the survey, and whether the beneficiary makes health insurance decisions on her own, with help from someone else, or whether the proxy makes decisions for her.
For each MCBS respondent who purchased a standalone PDP between 2006 and 2010 we obtained administrative records on their prescription drug claims, the set of PDPs available to them, and their annual enrollment decisions. Then we calculated what each enrollee would have spent had they purchased the same bundle of drugs under each alternative PDP in their choice set. This was done by combining their actual claims with the cost calculator developed in Ketcham, Lucarelli and Powers (2015).9 Next we used administrative data from CMS’s Chronic Condition Data Warehouse to determine if and when each individual had depression or dementia, which are associated with diminished cognitive performance (Agarwal et al. 2009). Like prior studies of PDP choice we limit our analysis to enrollees who did not receive a low-income subsidy.10 Our linked sample includes 3,607 individuals who made 11,739 annual enrollment decisions between 2006 and 2010.11 Table A1 reports annual means of the key variables.
The typical enrollee is a retired high school graduate with living children. Approximately 22% are college graduates, 54% are married, and 54% have annual pre-tax household incomes over $25,000. Only 35% report that they ever personally use the internet to get information of any kind. However, among those who do use the internet most have used it to search for information on Medicare programs (25%). Another 18% report having called 1-800-Medicare for information.
The average beneficiary’s total expenditures on premiums and out of pocket costs increased from $1,203 in 2007 to $1,434 in 2010.12 This is a significant share of income 9 There is a correlation of.92-.98 each year between the out of pocket costs predicted for the actual plan and the realized cost observed in the administrative data. Differences between the calculator’s predictions and realized costs are due to changes in plan design or drug pricing that occur after open enrollment and are not observable to consumers at the time they make enrollment decisions.
10 We exclude those receiving low-income subsidies because they are autoenrolled into plans, they receive larger premium subsidies, and their copayments are much more uniform across plans. Hence, they are less relevant for our evaluation of prospective policies designed to alter choice architecture. Despite excluding them, our sample has similar income levels to the national average of people age 65 and above. In our sample 54% of households have annual income over $25,000 (weighted 2006-2010 dollars), compared with 63% (constant 2010 dollars) based on all householders 65 and older in the 2010 Census American Community Survey.
11 This excludes 3,890 observations on beneficiaries who reenrolled in plans they had originally chosen prior to joining the MCBS.
We drop these observations because we cannot observe the beneficiaries’ knowledge at the time they first selected their current plans.
12 The figure for 2006 is $1,020. It is smaller because during the inaugural year of the program open enrollment extended through May. More than half the enrollees in our sample were not enrolled for the full year. If we limit the sample to full-year enrollees to make it more comparable to later years, then the mean annual consumer expenditure is $1,373.
8 given that 45% of beneficiaries have household incomes below $25,000. The data also reveal that by the end of our study period significant fractions of enrollees had been diagnosed with dementia (12%) and depression (11%).
Given the relatively large amount of money at stake, the age range of the eligible population and the prevalence of cognitive illnesses it is unsurprising to find that 38% of enrollees did not make health insurance decisions on their own: 27% had help and 11% relied on a proxy to make the decision for them. Table 2 shows that beneficiaries who get help are likely to be older, sicker, lower income, less educated, and less internet savvy than beneficiaries who made decisions on their own. Those getting help are also more likely to have been diagnosed with depression or dementia. All of these differences are amplified when we compare beneficiaries who make their own health insurance decisions to those who rely on proxies to make decisions for them.
III. Identifying Enrollment Decisions Suspected to be Misinformed 9 Only 8% of the enrollment decisions in our data minimize ex post expenditures. In 2006 the average enrollee could have saved $460 by choosing their cheapest available plan.13 This is equivalent to reducing total expenditures by 45%. Potential savings declined to $349 in 2007 (or 29% of expenditures) and remained similar thereafter. Why are people leaving money on the table? We hypothesize that the answers differ from person to person. Some may be making informed decisions to pay more for plans that provide better risk protection and higher quality. Others may misunderstand how the market works or underestimate their potential savings. We must distinguish between these groups to evaluate the welfare effects of prospective choice architecture policies.
For the group we identify as informed, we maintain the standard assumption that their choices are informative to us and apply standard revealed preference logic to infer their preferences for cost reduction, risk protection, and quality. But revealed preference logic cannot be applied when consumers have latent beliefs about products that contradict the information we observe. With this in mind, we adapt two features of Bernheim and Rangel’s (2009) proposed approach to revealed preference analysis in the presence of latent heterogeneity in beliefs.14 First, we use theory and data to identify enrollment decisions that we suspect may fail to reveal preferences. We label these choices as suspect, using Bernheim and Rangel’s terminology. Second, for the consumers making suspect choices, we calibrate their preference relations using proxy measures derived from the behavior of observationally identical consumers who we observe making non-suspect choices. Thus, we implement Bernheim and Rangel’s proposal to respect consumer sovereignty and apply standard revealed preference methods unless theory and data suggest the standard approach may fail to reveal consumers’ preferences.
A. Defining Suspect Choices
Like prior Part D studies, we assume that consumer i’s utility from drug plan j in year t 13 This figure sums over premiums and out of pocket costs. See Table A1 for details. This average falls below the $520 figure reported by Ketcham, Lucarelli and Powers (2015) based on CMS’s 20% sample of 2006 full year enrollees because our average also includes people who only enrolled for part of the year. The primary reason for part-year enrollment in 2006 was the fact that the initial open enrollment period was extended through May (Heiss, McFadden, and Winter 2010).
14 Latent heterogeneity in beliefs is one case of what Bernheim and Rangel refer to as “ancillary conditions” on decision making.