#2004-13 November 2004 The Medicare Prescription Drug Benefit: Potential Impact on Beneficiaries by Jack Rodgers and John Stell Health Policy Economics, PricewaterhouseCoopers The AARP Public Policy Institute, formed in 1985, is part of the Policy and Strategy Group at AARP. One of the missions of the Institute is to foster research and analysis on public policy issues of importance to older Americans. This publication represents part of that effort. The views expressed herein are for information, debate, and discussion, and do not necessarily represent formal policies of AARP. Copyright 2004 AARP. Reprinting with permission only. AARP, 601 E Street, NW, Washington, DC 20049 www.aarp.org/ppi Acknowledgments The authors would like to thank Rebecca Yip of PricewaterhouseCoopers for her significant contributions to the report. The authors would like to acknowledge the contribution of three independent reviewers: Marilyn Moon of the American Institutes for Research, Joseph Antos of the American Enterprise Institute, and Robert E. Moffit of The Heritage Foundation. -i- Foreword Regardless of one’s beliefs about the merits or shortcomings of the legislation, there are few who would disagree that the Medicare Prescription Drug, Improvement and Modernization Act of 2003 introduces one of the most significant changes in the history of the program: the availability of prescription drug coverage through Medicare effective in 2006. Since the passage of this major piece of legislation, there has been much debate and speculation about how the new benefit will affect Medicare beneficiaries but relatively little analysis to inform a better understanding of the issue. This paper was commissioned from Jack Rodgers and John Stell of Health Policy Economics at PricewaterhouseCoopers for the purpose of providing objective and relatively comprehensive information about the potential impact of the Medicare prescription drug benefit on beneficiaries. The paper aims to accomplish this by describing the key provisions of the legislation, presenting up-to-date information about characteristics of the Medicare population, analyzing the aggregate impact of the new drug benefit along a number of dimensions, assessing the impact on specific groups of Medicare beneficiaries, and highlighting illustrative cases of individual beneficiaries. For the most part, analyses are conducted so as to be consistent with assumptions of the Congressional Budget Office. As the report demonstrates, much of the potential impact of the Medicare prescription drug benefit is a function of an individual beneficiary’s income, the nature of his or her current drug coverage (if any), drug spending level, and even future behavior. Another key factor, however, is the future decisions of entities that currently sponsor drug coverage for Medicare beneficiaries, particularly employers. Because the outcomes of such future behaviors and decisions generally cannot be known in advance, the authors of this paper seek to illuminate the incentives that may influence behavior in one direction or another. In addition, given the increased uncertainty of projections over longer time frames, many of the estimates produced for this paper are for the single year of 2006. It is our sincere hope that, taken together, the information presented in this paper will provide a more informed understanding of the potential impacts of the drug benefit on Medicare beneficiaries. Susan O. Raetzman Associate Director AARP Public Policy Institute -ii- TABLE OF CONTENTS Executive Summary.............................................................................................................ES-1 I. Background on the Medicare Prescription Drug, Improvement, and Modernization Act of 2003............................................................................................................................1 A. Status of Prescription Drug Coverage and Spending before the MMA......................1 B. Description of the Medicare Prescription Drug, Improvement, and Modernization Act of 2003 (MMA)....................................................................................................4 II. Aggregate Impact of the MMA ..........................................................................................8 A. Federal Budgetary Costs of the MMA........................................................................8 B. Impact of the MMA on Total Prescription Drug Spending ......................................11 C. Allocation of Federal Subsidies by Use and by Income ...........................................13 D. Impact of the MMA on Distribution of Out-of-Pocket Spending.............................16 III. Impact of the MMA by Type of Pre-MMA Prescription Drug Insurance...................18 A. Self-Pay (No Prescription Drug Coverage Pre-MMA).............................................19 B. Individual Private Plans (Medicare Advantage and Medigap) .................................20 C. Employer-Sponsored Retiree Plans ..........................................................................20 D. Medicaid....................................................................................................................23 E. Other Federal Coverage ............................................................................................24 IV. Individual Variation in the Impact of the MMA ............................................................25 A. Illustrative Case Related to Population with No Coverage.......................................26 B. Illustrative Cases of Medigap Purchasers under the MMA ......................................27 C. Illustrative Cases of Retirees with Employer Coverage under the MMA ................29 V. Conclusion ..........................................................................................................................33 -iii- LIST OF TABLES AND FIGURES Table 1 Projected Source of Prescription Drug Coverage for Medicare Population in the Absence of MMA, 2006 (population in millions)………………………....... 2 Table 2 Projected Distribution of Medicare Beneficiaries in the Absence of MMA, 2006 by Total Drug Spending, Income, and Drug Insurance Status (share of total Medicare beneficiaries)………………………………………………………..3 Table 3 Components of CBO’s Cost Estimate for Prescription Drugs Under MMA, FY 2004-2013 (in billions of dollars)………………………………………….…...8 Table 4 Impact of the MMA on Prescription Drug Prices, Drug Utilization, and Overall Spending by Medicare Population, 2006 (in billions of dollars)……….....12 Table 5 Distribution of Medicare Subsidies by Income, FY 2004–2013………………......15 Table 6 Distribution of Out-of-Pocket Spending on Prescription Drugs, 2006 (pre-MMA and post-MMA)…………………………………………………..…...16 Table 7 Impact of MMA on Drug Use and Prices, by Source of Prescription Drug Coverage, 2006……………………………………………...……...……………...19 Table 8 Comparison of Out-of-Pocket Drug Spending for a Beneficiary Medigap Plan H (or I) vs. MMA Standard Benefit, 2006………………………………..….29 Figure 1 Participation by Medicare Beneficiaries in the MMA Drug Benefit, 2006 (participants in millions)………………………………….………………………..10 Figure 2 Distribution of Subsidies by Use, FY 2004–2013 (in billions of dollars)…….…...14 Figure 3 Change in Out-of-Pocket Drug Spending for a Retiree Who Loses Employer Drug Coverage, 2006…………………………………………………………........30 -iv- Executive Summary Background Drugs are becoming an increasingly important part of health care costs, and average per capita prescription drug spending by Medicare beneficiaries was estimated by the Congressional Budget Office (CBO) to be about $3,155 in 2006 (in the absence of the MMA). Although more than half of all beneficiaries will have total drug spending of less than $2,000 each, about 20 percent will have spending of more than $5,000. In the absence of Medicare drug coverage, PricewaterhouseCoopers (PwC) estimates that 28 percent of Medicare beneficiaries would have no drug coverage, and another 17 percent would lack any type of protection against catastrophic drug costs. As a result, significant numbers of Medicare beneficiaries are subject to substantial out-of-pocket expenses for prescription drugs. On December 8, 2003, the Medicare Prescription Drug, Improvement, and Modernization Act of 2003 (MMA), which added a prescription drug benefit as part of the Medicare program, was signed into law. Under the new law, Medicare will offer prescription drug coverage to beneficiaries who choose to enroll for the Part D benefit, beginning in 2006. Purpose This report examines the potential impact of the prescription drug benefit of the MMA on Medicare beneficiaries, with respect to its overall impact and its impact on specific categories of beneficiaries. Methodology The legislative provisions creating the Medicare drug benefit are described. Incentives and dynamics created by the new benefit are discussed in qualitative terms. Both publicly available data and privately developed computer models are used to quantify the impacts of the Medicare drug legislation. Aggregate impacts are typically quantified for the 10-year budget window of 2004-2013 and beneficiary impacts are typically quantified for the single year of 2006. For the most part, analyses are conducted so as to be consistent with the assumptions of the Congressional Budget Office. Findings In 2006, the new benefit will cover a portion of spending below $2,250 and almost all spending once the enrollee has spent $3,600 out of pocket (generally reached at $5,100 in total spending). Enrollees will have to pay out of pocket for all drug costs during this gap in coverage, popularly called the “doughnut hole.” Private providers will offer the new benefit ES-1 and will collect premiums from enrollees for 25 percent of benefit costs and rely on the Medicare program for the remaining 75 percent. Because of the relatively generous subsidy rate (and the penalties associated with late enrollment), both CBO and the administration expect that almost all Medicare enrollees will enroll in the new program. However, actual enrollment may lag behind projected levels. The overall impact of the MMA on Medicare beneficiaries represents a $407 billion shift in resources to Medicare beneficiaries, health plans, employers, and states over the period 2006– 2013, according to CBO. Much of this spending represents a transfer of private spending to the Medicare program, but the program also will provide new coverage to, and boost spending for, a large number of Medicare beneficiaries. • Many of the new subsidies paid under the Medicare drug benefit will replace spending formerly covered for individuals through their insurance plans or out of their own pockets. About $70 billion of the $407 billion total cost is estimated to represent new spending that would not have occurred in the absence of the MMA. The remaining $337 billion essentially replaces spending currently covered by Medicare enrollees (either directly or indirectly). • Approximately $145 billion of the $407 billion is estimated to go to Medicare beneficiaries who do not have Medicaid but meet the requirements for the low-income protections (about 8.5 million people in 2006). These individuals generally have little or no drug coverage. The new benefit will change the spending characteristics of the Medicare population. The new coverage will boost utilization, lower average drug prices, and lower out-of-pocket spending for the average Medicare enrollee. • PwC estimates that the Medicare population’s aggregate drug spending will decrease by 3 percent as a result of the new MMA benefit. Although total utilization of drugs is estimated to increase by almost 6 percent, PwC estimates that private plans’ additional discounts will lower the average drug price by more than 8 percent. These savings accrue to all spending, not just the spending covered by the new benefit. • The drug benefit subsidy under the MMA potentially reduces average out-of-pocket expenses from about $1,325 per capita in 2006 in the absence of the MMA to about $890 per capita under the MMA. • The new benefit will almost eliminate catastrophic drug costs for the Medicare population. Nearly all Medicare beneficiaries will have catastrophic coverage compared to only about half before enacting the MMA. The incidence of very high out-of-pocket expenses (i.e., greater than $5,000 in 2006) will be reduced from about 6 percent in the absence of the MMA to virtually zero percent under the MMA. ES-2 • Medicare beneficiaries who have incomes below 150 percent of the federal poverty level (FPL) and whose assets are below mandated threshold levels—accounting for about 30 percent of all beneficiaries—will be eligible for low-income subsidies that substantially reduce or eliminate cost sharing. • Beneficiaries with drug coverage through Medigap can reduce their out-of-pocket expenses by enrolling in Part D. This occurs primarily because the drug portion of the Medigap premium for which they were responsible is replaced by the government- subsidized premium. In addition, at most drug spending levels, the drug benefits under Medicare will be more generous than the Medigap drug benefit. The exception is at certain points in the “doughnut hole” where Plan J coverage results in lower cost- sharing than under the standard Medicare drug benefit. Although Medicare beneficiaries, on average, receive substantial benefits from the MMA, some beneficiaries may be worse off under the new benefit. PwC expects about half of the Medicare population to experience an improvement in drug coverage, but the impact on the other half of the Medicare population is not as clear. For instance, 2.7 million retirees are expected to shift from their former employers’ plans to prescription drug plans (PDPs) with the standard benefit. The loss can be substantial to beneficiaries with high prescription drug costs and who, in theory, could move from minimal cost sharing in their employer plans to as much as $3,600 or more, not counting the new premium under the MMA. Other Medicare beneficiaries, such as those with prescription drug spending below the premium and Medicaid beneficiaries from states with no cost sharing, may also experience modest increase in costs under the MMA. The program’s design also provides disincentives for plans to supplement the standard benefit package to offer more comprehensive coverage. The subsidies that plans receive from the federal government partially depend on the amount of catastrophic spending by enrollees. Because enrollees only reach the catastrophic benefit threshold by spending a specified amount out of their own pockets, plans will be reluctant to cover more than the standard benefit. Conclusion The new benefit represents a dramatic change to the Medicare program. Many Medicare beneficiaries will have to adjust to changes in their insurance coverage, some will experience higher out-of-pocket expenses, and others may have to adjust to more restrictive formularies. Overall, however, all Medicare beneficiaries will have access to drug coverage, every Medicare beneficiary will be able to enroll in a prescription drug plan with catastrophic insurance, and out-of-pocket expenses will be significantly lower. ES-3 I. Background on the Medicare Prescription Drug, Improvement, and Modernization Act of 2003 On December 8, 2003, the Medicare Prescription Drug, Improvement, and Modernization Act of 2003 (MMA), which added a prescription drug benefit as part of the Medicare program, was signed into law. Under the new law, Medicare will offer prescription drug coverage to beneficiaries who choose to enroll for the Part D benefit, beginning in 2006. This report examines the potential impact of the prescription drug benefit of the MMA on Medicare beneficiaries, with respect to its overall impact and its impact on specific categories of Medicare beneficiaries. The Medicare legislation offers subsidies to beneficiaries, to designated private sector prescription drug plans (PDPs), and to employers with qualified retiree plans that provide prescription drug coverage to Medicare beneficiaries. According to the Congressional Budget Office (CBO), the cost of the subsidies will amount to approximately $407 billion over the 2004–2013 period.1 The assumptions in this report are generally consistent with the assumptions underlying CBO’s budget cost estimate of the new benefit. The administration estimated that the new benefit will cost significantly more than the CBO estimate, or $502 billion over the 2004– 2013 period.2 Calibration to the administration’s totals or using specific CBO assumptions— where they are different from PricewaterhouseCoopers’ (PwC’s)—will not change the substantive conclusions of the report, but they might change the magnitude in certain cases. The authors have noted significant differences in assumptions throughout the report. A. Status of Prescription Drug Coverage and Spending before the MMA As Table 1 shows, projections are that most Medicare beneficiaries would have some form of prescription drug coverage in 2006 in the absence of the MMA.3 About 28 percent of 1 Because Medicare prescription drug coverage does not begin until 2006, this analysis focuses on the portion of the 2004 to 2013 budget window beginning in 2006. 2 Administration estimates placed the cost of the Medicare drug benefit at $502 billion over the 2004–2013 period, while CBO estimated a cost of $407 billion. The cost of the entire legislation, which includes additional payments to Medicare Advantage plans (formerly called Medicare+Choice plans) and cuts in payments to certain providers, was estimated at $534 billion by the administration and $395 billion by CBO. This paper does not discuss the differences in estimates and other nonprescription drug spending. 3 Estimates in Table 1 are based on Congressional Budget Office projections for 2006. The proportion of Medicare beneficiaries with prescription coverage is probably higher now than it would have been in 2006 in the absence of the MMA. This estimate is based on past patterns of drug coverage among beneficiaries and the assumption that, in the future, Medicare Advantage plans and employers may drop prescription drug coverage, and beneficiaries will find it increasingly difficult to afford individual coverage. The proportion of Medicare beneficiaries with prescription drug coverage also reflects the presence of coverage at any time during the year, which yields a higher number than when measuring the presence of continuous coverage throughout the year. -1- Medicare beneficiaries, however, would not have prescription drug coverage without the new Medicare benefit. In addition, even those beneficiaries with coverage for prescription drugs might not have adequate protection against high drug costs. Table 1 Projected Source of Prescription Drug Coverage for Medicare Population in the Absence of MMA, 2006 (population in millions) Primary Source of Drug Coverage Population Percentage Part B 39.9 94% Private 19.1 45% Medicare Advantage 3.0 7% Employer-Sponsored 11.8 28% Medigap 4.2 10% Government 8.9 21% Medicaid 6.3 15% Other 2.6 6% Self-Pay 11.9 28% Part A Only 2.7 6% Total 42.6 100% Source: PricewaterhouseCoopers, based on Medicare Current Beneficiary Survey data and CBO projections. Consistent with CBO, the Part A Only population is listed separately since it is not expected to participate in the new benefit. The new benefit offered under the MMA is intended to provide drug coverage to those currently without it and to address the lack of catastrophic drug coverage among a large share of the Medicare population. In 2006, in the absence of the MMA, 17 percent of Medicare beneficiaries are estimated to enroll in privately purchased plans, either Medicare Advantage (MA) or Medigap plans, generally with limited prescription drug coverage. The most comprehensive Medigap plan, “Plan J,” has coverage that is limited to $3,000 in benefits; therefore, Medigap does not pay for drug expenses above $6,250 in total.4 The MA plans also tend to have limited prescription drug coverage, usually with low limits on total coverage. Frequently, these plans have benefit caps of $1,000 or less, and many cover only generic prescription drugs.5 4 Plan J includes an annual $250 deductible and 50 percent coinsurance up to $3,000 in drug benefits (or $6,250 in total spending). 5 Medicare Advantage is the new name for Medicare+Choice plans—private plans that replace traditional Medicare and frequently offer additional benefits such as limited prescription drug coverage and lower cost sharing on medical services. -2- Thus, almost half of all Medicare beneficiaries—an estimated 45 percent in 2006—lack any protection against catastrophic drug costs in the absence of the MMA. The remaining Medicare beneficiaries have employer-sponsored plans provided by former employers, state Medicaid coverage, or federal coverage through the federal government, such as TriCare coverage for military retirees. The distribution of drug spending across Medicare beneficiaries varies based on many factors, but one important factor is whether a Medicare beneficiary has supplemental drug coverage. On average, drug spending by those with supplemental drug coverage exceeds that of those without coverage. Table 2 presents the estimated 2006 distribution of Medicare beneficiaries (Parts A and/or B) by income and supplemental drug insurance category assuming that the MMA did not pass. Approximately 37 percent of beneficiaries (15.8 million) will have incomes below 150 percent of the federal poverty level (FPL), and about a quarter of these would have had no drug coverage in the absence of MMA (9 percent overall, or about 3.8 million beneficiaries). The Medicaid program covers the drug spending of much of this population. Among beneficiaries with incomes over 150 percent of the FPL, about 30 percent (19 percent overall, or 8.1 million beneficiaries) would lack drug coverage in 2006 in the absence of the MMA. Table 2 Projected Distribution of Medicare Beneficiaries in the Absence of MMA, 2006 by Total Drug Spending, Income, and Drug Insurance Status (share of total Medicare beneficiaries) Income under 150% FPL Income above 150% FPL Total Drug Spending With Supp No Supp With Supp No Supp Total Drug Covg Drug Covg Drug Covg Drug Covg Under $250 5% 2% 7% 4% 18% $250–$1,999 8% 4% 13% 8% 34% $2,000–$4,999 8% 3% 13% 5% 29% $5,000 and over 7% 1% 10% 1% 19% Total 28% 9% 44% 19% 100% Source: PricewaterhouseCoopers based on MCBS data and CBO projections. Note: Numbers may not add to 100 percent due to rounding. The 2006 federal poverty threshold for a single person age 65 or older is estimated to be $9,300; 150 percent of FPL is $13,950. PwC estimates that, compared to those without supplemental coverage, a larger share of beneficiaries with supplemental drug coverage has total drug spending over $5,000. About a quarter of those with supplemental coverage (regardless of income) spend more than $5,000 (this is calculated as 7 percent/28 percent for the population with income under 150 percent of poverty and 10 percent/44 percent for the population with income above 150 percent of poverty). For the population without drug coverage, only about one in 20 of those whose -3- incomes are above 150 percent of poverty and only about one in 10 of those whose incomes are below 150 percent of poverty will spend more than $5,000. B. Description of the Medicare Prescription Drug, Improvement, and Modernization Act of 2003 (MMA) The MMA provides voluntary drug insurance through subsidized private sector plans. Medicare beneficiaries can choose either private, stand-alone prescription drug plans (also referred to as PDPs and available to those who remain in the traditional Medicare program or to Medicare Advantage enrollees whose plans do not offer drug coverage), or Medicare Advantage plans that offer both medical and prescription drug coverage (also referred to as MA-PDs). The MMA also provides subsidies to qualified retiree plans offering drug coverage that is at least as good as the standard benefit (in actuarial terms). The standard prescription drug benefit in 2006 will be as follows: • Deductible = $250 • Front-end Benefit = 25 percent coinsurance for drug costs up to an initial coverage limit of $2,250 in total drug spending • Doughnut Hole = 100 percent coinsurance between $2,250 total drug spending and $3,600 out-of-pocket spending, or $5,100 of total spending assuming no supplemental coverage • Catastrophic Benefit = greater of $2 generic/$5 brand copayments or 5 percent coinsurance after $3,600 in out-of-pocket spending ($5,100 in total spending assuming no supplemental coverage) The catastrophic benefit level is set at $3,600 in “true out-of-pocket” spending (also referred to as TROOP), meaning that supplementing coverage for Medicare cost sharing or the gap in coverage by other private drug insurance does not count toward the catastrophic spending level.6 Under the standard benefit with a $250 deductible and 25 percent coinsurance up to $2,250 in total spending, a Medicare beneficiary without supplemental drug coverage reaches $3,600 in out-of-pocket spending once the beneficiary’s total spending hits $5,100. To the extent that a supplemental plan pays any of the Medicare cost sharing below $5,100 in total drug spending, additional spending will be required before out-of-pocket spending reaches $3,600 and the catastrophic benefit threshold is met.7 6 Certain categories of payments count as true out-of-pocket spending for the purposes of calculating when the catastrophic benefit begins: low-income subsidies provided under the MMA, benefits from state pharmaceutical assistance programs, and payments on behalf of individuals by another person, such as a family member. 7 The standard benefit under the MMA does not have a true “stop-loss” such as is the case with many private sector health plans with catastrophic coverage. The Medicare beneficiary in a plan offering the standard benefit will continue to pay cost sharing even after crossing the catastrophic threshold, but the extra cost sharing will be relatively modest. For example, in 2006, out-of-pocket spending will rise from $3,600 at a total spending level of -4- In general, dollar values for the standard drug benefit deductible, initial coverage limit, and catastrophic threshold in years after 2006 will be indexed by growth in per capita drug spending. By 2013, CBO estimates that the deductible will reach $445, the initial benefit limit will reach $4,000, and the out-of-pocket threshold, $6,400. Because incomes are expected to grow more slowly than drug spending, out-of-pocket amounts that beneficiaries pay will increase as a share of income.8 Plans will be able to offer alternative packages with benefit structures that differ from the standard benefit. For example, an alternative plan can rely on copayments in place of 25 percent coinsurance for spending under $2,250. These plans must be actuarially equivalent to the standard package and meet several other specific requirements, including offering the same catastrophic benefit as the standard package. The requirements in addition to actuarial equivalence will limit the variety of alternatives, but these requirements generally will ensure that all enrollees have similar cost sharing and catastrophic protection. The federal government pays 74.5 percent of benefit costs, on average, to plans through two mechanisms: direct subsidies and reinsurance. The Medicare program will offer a reinsurance subsidy that pays 80 percent of all catastrophic costs (spending on prescription drugs after the Medicare beneficiary has met the $3,600 out-of-pocket threshold). A direct subsidy will cover roughly 60 percent of the remaining benefits not covered by the reinsurance subsidies. The direct subsidies will generally be the same for all plans, once adjusted for risk. This particular method of subsidizing plans has two major implications. First, insurance plans face less risk because the subsidies for high-cost benefit users are higher than those for low- cost users. To the extent that plans have higher-than-expected spending per enrollee, their subsidies will exceed the average 74.5 percent level. Second, the cost to plans of supplementing the standard benefit package, such as filling the doughnut hole, will be relatively expensive. Because of the true out-of-pocket provisions, a PDP or an MA-PD that enhances the benefit by supplementing the standard package not only increases total benefit costs but also reduces the amount of reinsurance subsidies to the plan because additional total spending is required before the catastrophic benefit threshold is reached.9 $5,100, to $3,845 at $10,000 total spending, and to $4,345 at $20,000 in prescription drug spending, assuming 5 percent coinsurance once out-of-pocket spending reaches $3,600. 8 Alternative indexes can be used to maintain a steady relationship between income and spending, but the overall cost of the program would increase if the basis for indexing the new Medicare drug benefit was changed to accommodate this protection. 9 For example, if an enhanced plan is designed to pay 75 percent of all charges up to a catastrophic limit, then Medicare beneficiaries who enroll in that plan will not reach $3,600 in out-of-pocket spending until total spending is $14,400 (and the plan will only receive reinsurance subsidies for spending above $14,400). In 2006, the plan will lose all federal subsidies between $5,100 and $14,400. Based on the projected distribution of drug spending, the actuarial value of filling the doughnut hole with 75 percent coverage will be about $540, and the plan will lose approximately $560 in reinsurance subsidies. Thus, the plan will have to charge enrollees at least an additional $1,100. This means an additional premium cost of $92 per month in return for $47 per month in -5- Employers and unions that provide drug coverage to retirees will be eligible for an alternative subsidy equal to 28 percent of enrollee spending between $250 and $5,000. To qualify for the alternative subsidy, the coverage must be at least as generous as the standard benefit. The legislation prohibits the sale of new Medigap drug plans, beginning in 2006. Current beneficiaries who purchased plans H, I, or J (Medigap plans with prescription drug insurance coverage) will be allowed to keep their current plans, but new enrollees will not be able to purchase such plans. In addition, beneficiaries with Medigap drug coverage must either drop their Medigap policy or drop the drug benefit portion of the Medigap policy to enroll in the Medicare drug benefit. Because enrollees pay the full cost of drug coverage under these plans, most are expected to switch to the subsidized MMA benefit. Low-income Medicare beneficiaries who meet certain income and asset tests are subject to lower cost sharing and premiums. For the purposes of low-income subsidies, Medicare beneficiaries fall into three categories: • Dual-eligible Medicare beneficiaries (those who have Medicaid coverage as well as Medicare) with incomes below 100 percent FPL—No premiums, no deductible, no doughnut hole, and copayments of $1 for generic drugs and $3 for brand-name drugs for spending below the catastrophic cap. No cost sharing above the catastrophic cap. No cost sharing for beneficiaries in nursing homes. • Income below 135 percent FPL and assets below $6,000 (single) or $9,000 (couple)— No premiums, no deductible, no doughnut hole, and copayments of $2 for generic drugs and $5 for brand-name drugs for spending below the catastrophic cap. No cost sharing above the catastrophic cap. • Income below 150 percent FPL and assets less than $10,000 (single) or $20,000 (couple)—Sliding-scale premium (starting at 135 percent FPL, up to full premium at 150 percent FPL), $50 deductible, coinsurance of 15 percent up to the catastrophic limit, and $2/$5 copayments above the catastrophic cap.10 On average, the standard benefit provided under the MMA will pay for about 53 percent of prescription drug costs, assuming no other supplemental drug coverage, and have a net value extra benefits, or almost $2 in premium cost for each additional $1 in prescription drug benefits. The financing mechanism will also influence the type of extra benefits offered. Consider a Medicare Advantage plan with a capitated payment that is $560 more per year than the cost of providing Part A, B, and D Medicare benefits. In this case, suppose the plan offers $560 in additional benefits. The plan can offer $560 in additional medical benefits and charge no extra premium. Alternatively, the plan can offer $560 in prescription drug benefits, as noted above, but then has to charge an additional premium of $540 to cover the lost subsidies. 10 As stated earlier, low-income subsidies are counted toward out-of-pocket spending to determine whether the catastrophic spending level has been reached. In 2006, at $5,100 in total drug spending, the beneficiaries receiving the 15 percent coinsurance benefit will have spent $807.50 out of pocket. -6- of about $1,250 per beneficiary, net of premiums paid, in 2006.11 The value of benefits to the lowest-income Medicare beneficiaries—that is, those with income below 135 percent FPL and assets below $6,000 (individual) or $9,000 (couples)—will average about $3,000 after accounting for lower cost sharing and waiver of premiums. 11 CBO estimated a gross cost per participant for the standard drug benefit, but no supplemental coverage, of $1,640 in 2006. Net of $420 in premiums, the $1,220 per capita cost is approximately the same as the $1,250 estimated by PwC. -7- II. Aggregate Impact of the MMA The Congressional Budget Office (CBO) estimates that the aggregate federal costs of prescription drug coverage under the MMA will be about $407 billion over the 2004–2013 period. This increase in federal spending, accounting for approximately 25 percent of total estimated prescription drug spending by Medicare beneficiaries over the same period, represents a significant expansion of the Medicare program. The new benefit also will affect prices enrollees pay, the number of drugs they use, and the insurance available to them from private sources. A. Federal Budgetary Costs of the MMA The overall 10-year federal budget impact of the drug benefit created by MMA is presented in Table 3. The new prescription drug plans will provide $507 billion in benefits, offset by premiums of $131 billion, reflecting a general subsidy of about 75 percent. These amounts do not include the $71 billion in subsidies to qualifying employer or union plans or the $192 billion in subsidies for low-income Medicare beneficiaries. Thus, the total amount of subsidies for prescription drugs will be $639 billion over the 2004–2013 period before accounting for Medicaid and other offsets. CBO estimates the net federal budgetary costs after accounting for existing government subsidies that are being replaced and the nonfederal dollars that are being captured by the new program to be $407 billion. Table 3 Components of CBO’s Cost Estimate for Prescription Drugs Under MMA, FY 2004-2013 (in billions of dollars) Components Spending/(Offsets) Payments to Medicare Drug Plans for Basic Benefits $507 Beneficiary Premium Payments ($131) Employer and Union Subsidies $71 Low-Income Subsidies $192 Federal Medicaid Spending ($142) Transfers from State Medicaid Programs ($89) Effects on Other Federal Programsa ($2) Total $407 Source: Congressional Budget Office, July 2004. a According to CBO, the legislation will reduce spending on the Federal Employees Health Benefits Program and other programs funded by the federal government. -8- Essentially all of the prescription drug spending will take place between 2006 and 2013, but approximately $2 billion will be spent in 2004 and 2005 for the discount card program established by the legislation. CBO estimates that roughly 87 percent of Medicare beneficiaries will enroll in the new private plans through a PDP, an MA-PD, or a qualified retiree plan; the exception is those who are in certain federal programs that already provide prescription drug benefits, such as military retirees. Figure 1 shows that roughly 28.8 million Medicare beneficiaries, or 68 percent of all Medicare beneficiaries, are expected to enroll in the new PDP and MA-PD drug plans, and 8.3 million, or 19 percent, are expected to receive qualified coverage through employer plans. The incentives to enroll are substantial, especially for those currently without drug coverage, those whose drug coverage is only minimally subsidized, and those with high drug spending. First, Medicare beneficiaries who enroll in these new plans will pay an annual premium estimated at about $420 in 2006 for the standard benefit, or, as discussed above, only about 25 percent of the full cost of prescription drug coverage. Second, a late enrollment penalty of at least 1 percent in additional premium per month of delayed enrollment is intended to discourage late enrollment in the plans. Part B of Medicare has similar characteristics— Medicare pays 75 percent of costs, and there is a similar late enrollment penalty—and roughly 95 percent of eligible individuals enroll in the program.12 Enrollment in Part B increased slightly in the early years of the program but quickly reached its current level (from approximately 90 percent of the total eligible population in 1966 to approximately 95 percent by 1970).13 About half of the nonparticipants in CBO’s estimates are Medicare beneficiaries who do not purchase Part B coverage. This population (or their spouses) generally has full-time jobs with employers that offer and contribute to health coverage. Medicare has an existing provision that allows these “working aged” individuals to delay enrollment, without penalty, until their active employment-sponsored coverage ceases. In a similar manner, Medicare beneficiaries who have prescription drug coverage through active employment that is comparable in value to the Medicare drug benefit (i.e., “creditable coverage”) and those who are covered by spouses as dependents are also exempt from the late-enrollment penalty. It is assumed that 12 Enrollment in the Medicare prescription drug plans will be lower than in Part B, according to CBO, because Medicare beneficiaries who have federal coverage under the Federal Employees Health Benefits Program (FEHBP) and Department of Defense are not expected to enroll in the drug plans but they do enroll in Medicare Part B. Some studies have shown that prescription drug spending is somewhat predictable, which may contribute to other healthy Medicare beneficiaries not enrolling despite the penalty. 13 CBO and the administration do not assume a phase-in period for participation in the general MMA drug benefit. Given the Medicare population’s broad participation in Part B, this assumption seems reasonable because enrollees will be familiar with the penalties associated with not enrolling and will view the drug benefit as an add-on to the Part B benefit, rather than an entirely new program. -9- those individuals will not enroll in PDPs or MA-PDPs unless and until their current coverage is no longer creditable or is terminated. 14 Figure 1 Participation by Medicare Beneficiaries in the MMA Drug Benefit, 2006 (participants in millions) Subsidized Employer Part B Plan Nonparticipants 8.3 Million 2.9 Million 19% 7% Nonparticipants 5.5 Million 13% Part A Only Part D Participants Nonparticipants 28.8 Million 2.7 Million 68% 6% Source: PricewaterhouseCoopers’ estimates based on CBO assumptions. A factor that might limit participation is uncertainty about how the program will operate. To the degree that Medicare enrollees are unfamiliar with or skeptical of the new benefit, enrollment might lag behind estimated levels. In such a case, enrollment will be lower than the levels assumed above. If enrollment is lower, the federal costs of the program will fall. For instance, if only half as many Medicare participants enroll in the new drug benefit, the total federal cost might be only $200 billion over the 2006–2013 period. 14 By way of comparison, the administration assumed that 94 percent of all Medicare beneficiaries will participate. Among those that the administration assumed will participate in the program are certain federal retirees, who already have generous coverage through the federal government, and Medicare beneficiaries who decline Part B. Since the federal government already bears the cost of prescription drug coverage for these beneficiaries, their participation does not significantly affect overall costs. -10- B. Impact of the MMA on Total Prescription Drug Spending The new Medicare drug benefit will have two primary effects on Medicare beneficiaries’ total drug spending. First, the additional insurance coverage provided under the MMA will increase the use of prescription drugs. Second, competition among prescription drug plans (PDPs) for enrollees is expected to result in relatively tightly managed plans that will deliver higher discounts on drugs than levels generally seen today in private plans, which will also increase use. The additional drug coverage provided under the new Medicare benefit is expected to lower the average dollar amount that beneficiaries who enroll in the program will pay out of their own pockets for drugs. From the perspective of the beneficiary, the cost of prescription drugs will fall for those beneficiaries whose insurance coverage improves because of Part D enrollment. The lower cost will cause, in turn, an increase in demand. PwC estimates that demand will increase drug use in 2006 by about 5.7 percent on average.15 The additional demand might result from several practical responses. Individuals who currently skip dosages or fail to fill prescriptions might be less likely to do so if the out-of- pocket costs of the prescriptions are lower. However, lower out-of-pocket costs might encourage enrollees to increase the use in cases where the drugs only provide marginal benefits. In addition, lower out-of-pocket costs could increase use of more costly drugs. Because the value and key features of the drug benefit are specified by law, plans will have few options to compete with one another other than through the premium they charge enrollees.16 Competition will force PDPs to establish relatively strict provisions that effectively manage the drug benefit. Because PDPs will bear much of the financial risk associated with providing the benefit, they will have a strong incentive to control costs. As a result, plans offering the Medicare drug benefit will be able to lower average costs relative to retail prices by around 25 percent, compared to savings of 18 percent under typical employer plans.17 The extra savings come from higher manufacturer rebates, narrower network 15 For calculations, the authors assumed, as did the CBO, that the “elasticity” of drug use to out-of-pocket spending is -0.3. This means that, for every 10 percent decline in the out-of-pocket cost of drugs, drug use will increase by 3 percent (Congressional Budget Office, “Issues in Designing a Prescription Drug Benefit for Medicare,” October 2002). 16 Plans can compete through formulary design and networks, but variation in these characteristics should be reflected in the plan premiums. The legislation allows alternative benefit structures, subject to actuarial equivalence requirements. As mentioned earlier, the actuarial equivalence requirements will have the effect of limiting plan variety so even alternative benefits will be similar to the standard benefit. For instance, because actuarial equivalence requirements mandate the same catastrophic coverage, alternative packages most likely will still have doughnut holes in coverage. The legislation encourages the use of copayments for the front-end benefit but requires the amount paid by the plan to equal the amount paid under the standard benefit at the initial coverage limit ($2,250 in 2006). 17 CBO assumes that current employer savings are around 15 percent and that savings under private plans will increase to 20 percent in 2006 and to 25 percent by 2013. See A Detailed Description of CBO’s Cost Estimate -11- arrangements, drug utilization review, generic substitution, and other methods plans and pharmacy benefit managers (PBMs) use to reduce costs. The additional savings applicable to the new drug benefit will lower costs in 2006 by an average of about 8.4 percent.18 CBO assumes that fallback plans, which contract with the government to offer the Medicare drug benefit in areas with no private plans, will achieve much smaller savings, only about 12.5 percent relative to retail prices. To the degree that fallback plans supply a significant share of the benefits, or the expected discounts in private plans are smaller than estimated, overall drug costs will be higher. Higher drug costs will increase the cost of the overall program above the $407 billion estimated by CBO and increase out-of-pocket amounts paid by enrollees. In such a case, the new Medicare benefit would still cover a similar percentage of total drug costs but the base would be larger.19 Combining the impact of the increased demand with the additional discounts from competition is projected to result in a small reduction, about 3.2 percent, in the overall dollar level of prescription drug spending (see Table 4). This is the result of a 5.7 percent increase in prescription drug use offset by an 8.4 percent reduction in prices for prescription drugs.20 Table 4 Impact of the MMA on Prescription Drug Prices, Drug Utilization, and Overall Spending by Medicare Population, 2006 (in billions of dollars) Total Drug Percent Spending Change Prescription Drug Spending (Pre-MMA) $125.9 Net Change in Prices ($10.5) (8.4%) Net Change in Utilization (post-MMA prices) $6.5 5.7% Prescription Drug Spending (Post-MMA) $121.9 -3.2% Source: PricewaterhouseCoopers’ estimates. Note: Numbers may not add due to rounding. for the Medicare Prescription Drug Benefit, July 2004. The administration estimates that average savings will slowly rise to 23 percent by 2010 as more beneficiaries move to more efficient plans. 18 The percentage change is calculated by comparing the discounted price (1 minus the average discount under the MMA) to the average discounted price before the MMA (1 minus the average discount before MMA). 19 Although the amount of spending in the doughnut hole would increase, increases in spending above the deductible and the catastrophic threshold would increase the overall cost of the Medicare benefit. 20 The overall impact of a 3.2 percent spending change results from multiplying the increase in utilization (1 plus 5.7 percent, or 1.057) and the lower price (1 minus 8.4 percent, or 0.916). Assuming the original spending level equals 100, the resulting spending is 96.8, or 3.2 percent less. In its July 2004 report, CBO estimated that drug use by enrollees in stand-alone prescription drug plans and MA drug plans will increase by approximately 9 percent. The 5.7 percent in the table above is consistent with CBO’s 9 percent because retirees with employer coverage and Medicaid recipients, who are expected to have a minimal utilization response, are not included in the CBO calculation (i.e., averaging the 9 percent for enrollees in stand-alone PDPs and MA drug plans with the negligible change for enrollees in employer plans and Medicaid yields an overall utilization change that is close to 5.7 percent. -12- For the single year 2006, lower prices under the MMA will lower spending by $10.5 billion. Greater drug utilization will increase spending by $6.5 billion (valued at post-MMA prices), yielding a net decrease in drug spending by the Medicare population of $4.0 billion. C. Allocation of Federal Subsidies by Use and by Income Most federal government spending under the new Medicare drug benefit is not for new utilization but displaces pre-MMA spending by both Medicare beneficiaries and third-party payers. Specifically, much of the funding replaces pre-MMA spending by Medicare enrollees on premiums for private prescription drug coverage and out-of-pocket costs. In addition, federal government spending will partially replace spending by employers and state Medicaid programs. Although the federal subsidies provided for the benefit generally are paid to insurance companies offering PDPs, their effects will alter the cash flow among all participants in the drug insurance market. Because this paper focuses on the impact on Medicare beneficiaries (as opposed to insurers or drug manufacturers), it looks at how the federal funds are used to provide benefits rather than which industry receives the funds. Specifically, the analysis allocates the $407 billion in two distinct ways: by category of use (reduced cost sharing and premiums, increased drug use, employer subsidies, and state government subsidies) and by income level of the Medicare beneficiaries. Currently, most Medicare beneficiaries already have some level of drug insurance coverage, albeit often limited, and even those who do not have drug coverage still buy some prescription drugs. As noted above, the use of prescription drugs will increase for those Medicare beneficiaries whose coverage improves as a result of the MMA; these are mostly Medicare beneficiaries who have no coverage and those in Medigap and MA plans who lack catastrophic protection. Figure 2 shows that just over half of the $407 billion in federal subsidies for prescription drugs under the MMA in the first 10 years are projected to accrue directly to beneficiaries: 35 percent ($142 billion) will be used to reduce cost sharing and 16 percent ($66 billion) will be used to reduce premiums paid by Medicare beneficiaries with individually purchased plans in the absence of the MMA.21 On average, the insurance coverage offered under the new benefit 21 The allocation of MMA subsidies to reduce premiums, shown in Figure 2, includes only the reductions in premiums for those with individual coverage from Medigap and MA plans. Premiums implicitly paid by Medicare beneficiaries in current MA plans are included in the premium estimate, i.e., it is assumed that MA plans will have offered other benefits of equal value to the prescription drugs so that Medicare beneficiaries implicitly pay for their drugs with reductions in other services or increased cost sharing. Premiums for Medicare beneficiaries who currently do not have any prescription drug coverage will increase from zero to about $420 on average in 2006. These new premiums paid by those who currently do not have coverage are reflected in Figure -13- will cover more drug spending than insurance would have otherwise, leading to the decrease in beneficiary cost sharing. Because individuals will pay for only approximately 25 percent of the cost of the new coverage, the amount of drug insurance premiums individuals pay will fall relative to premiums paid in the absence of the MMA. Figure 2 Distribution of Subsidies by Use, FY 2004–2013 (in billions of dollars) $17 4% Reduced Beneficiary Cost Sharing $111 $142 27% 35% Lower Premiums Paid by Individuals Increased Use of Drugs Subsidies to Employers Net Subsidies to State Medicaid Programs $72 18% $66 16% Source: PricewaterhouseCoopers’ estimates based on CBO. Note: Includes low-income subsidies. Lower Premiums Paid by Individuals only includes premium savings for those who previously purchased private drug insurance as individuals. Reduced Beneficiary Cost Sharing includes offsetting increases in Medicare Part D premiums paid by individuals who previously lacked drug coverage. Another 18 percent ($72 billion) of the subsidies will purchase additional prescription drugs beyond what would have been used in the absence of the MMA. Finally, 27 percent ($111 2 as an offset to reductions in cost sharing by the Medicare beneficiary. In addition, retirees with employer coverage are not included in the calculations of lower premiums. The majority of retirees with employer coverage are assumed to have no change in their current plans with respect to premiums or cost sharing. Those whose employers drop their current plan and enroll them in PDPs are assumed to pay the beneficiary premiums, about $420 annually in 2006. Such retirees may have lower premiums if the retiree contributions to the current employer plan are reduced to reflect removal of the large prescription drug component. Not only is this component difficult to estimate, the savings to retirees whose employers drop coverage and pay the Part D premium will be more than offset, on average, by changes in their out-of-pocket spending (as shown in the third section of this report). Since many of those changes are not included in the $407 billion federal costs, the employer component of premium changes is ignored in calculations for Figure 2. -14- billion) of the subsidies replace pre-MMA spending by employers and 4 percent ($17 billion) replace pre-MMA Medicaid spending by state governments. The results in Figure 2 assume employers keep all the subsidies provided under the new program. The total subsidy provided to employer (and union) plans represents the amount paid directly to employers under the alternative subsidy plus implicit subsidies paid to employers who enroll their retirees in PDPs and MA-PD plans. To the degree that employers take the subsidies and provide additional benefits, reduce cost sharing, or increase their own contributions to premiums for retirees who are Medicare beneficiaries, some of the $111 billion can shift to other categories of subsidy use. This analysis also allocated the $407 billion according to the income level of Medicare beneficiaries on whose behalf the funds were spent (see Table 5). Because the Medicaid program already covers virtually all drug spending by Medicaid beneficiaries, a very small share of the net subsidies under the MMA goes to that program. New spending on Medicaid beneficiaries under the MMA is offset by savings to states and the federal government in the current Medicaid program (most state savings are “clawed back” from state governments, i.e., paid to the federal government). Low-income, non-Medicaid enrollees receive a disproportionate share of the new subsidies. Roughly 35 percent of the subsidies go to 23 percent of Medicare beneficiaries whose incomes are below 150 percent of the federal poverty level (FPL) and who do not have Medicaid drug coverage.22 Table 5 Distribution of Medicare Subsidies by Income, FY 2004–2013 Income Relative to Fed. Participants Subsidies Subsidies as a Percent of Poverty Level (FPL) (millions) (billions) Beneficiaries Spending Medicaid (all incomes) 6.3 $25 17% 6% Less than 100% FPL 3.0 $54 8% 13% 100%–135% FPL 4.0 $70 11% 17% 135%–150% FPL 1.5 $21 4% 5% Greater than 150% FPL 22.2 $238 60% 58% All Participants 37.0 $407 100.0% 100.0% Source: PricewaterhouseCoopers’ calculations based on CBO estimates. Note: Numbers may not add to 100 percent due to rounding. 22 Before accounting for current Medicaid spending, 63 percent of subsidies (less premiums) go to low-income Medicare beneficiaries with incomes below 150 percent of FPL (40 percent to Medicaid beneficiaries and 23 percent to other low-income beneficiaries). The proportions of subsidies going to low-income beneficiaries will be higher as well if estimates are based on those of the administration rather than CBO. The administration assumes in its estimates that a higher proportion of Medicare beneficiaries receive the low-income subsidies. -15- The federal subsidies for low-income beneficiaries will average about $3,000 per Medicare beneficiary with an income below 135 percent FPL and meeting the $6,000/$9,000 asset test, compared to the average subsidy of $1,250 per Medicare beneficiary receiving the standard subsidy. D. Impact of the MMA on Distribution of Out-of-Pocket Spending As noted above, a large share of total federal subsidies under the MMA are used to reduce cost sharing, but the impact on individual beneficiaries will vary by current type of coverage and spending level. Table 6 presents estimates of the distribution of out-of-pocket spending on prescription drugs (not including Part D or private premiums) in 2006, before and after enactment of the MMA. (Of course, the total amount of money spent directly by beneficiaries includes both out-of-pocket amounts and enrollee premiums.23) The level of out-of-pocket spending is projected to fall dramatically under the MMA. For example, Medicare beneficiaries who spend $250 or less out of pocket on prescription drugs will increase from 36 percent of the total population in the absence of the MMA to 48 percent under the MMA. This increase in individuals with out-of-pocket drug spending at the lowest levels is due to the low-income subsidies, which will cover most or all of the $250 deductible under the standard plan. Table 6 Distribution of Out-of-Pocket Spending on Prescription Drugs, 2006 (pre-MMA and post-MMA) Out-of-Pocket Spending Percent of Beneficiaries on Drug Purchases Pre-MMA Post-MMA Less than $250 36% 48% $250–$749 21% 26% $750–$3,599 33% 18% $3,600–$4,999 4% 8% Greater than $5,000 6% * Average Out-of-Pocket Spending $1,325 $890 Source: PricewaterhouseCooopers. *Denotes less than 0.5 percent. Note: In this table, private and Medicare premiums are not defined as out-of-pocket spending for drug purchases. The impact of low-income subsidies is included. 23 Premium payments are not reflected in Table 6 because data on retiree contributions to prescription drug insurance are not usually separated from contributions for medical insurance. If Medicare premiums and contributions to prescription drug insurance are included in out-of-pocket expenses, the percentages will shift from lower to higher categories, under both the pre-MMA and post-MMA, but the patterns of change will be similar. -16- Correspondingly, at the other end of the spectrum, the proportion of Medicare beneficiaries with out-of-pocket spending greater than $5,000 in 2006 falls from 6 percent in the absence of the MMA to virtually zero under the MMA. The presence of the catastrophic benefit significantly reduces out-of-pocket drug spending by enrollees who would have had inadequate or no coverage for high levels of spending. As a result, virtually no beneficiaries will spend more than $5,000 out of pocket. One can also state the impact of the MMA in terms of the average dollar amount of out-of- pocket spending. Out-of-pocket spending falls not only because of the increase in insurance coverage but also because of the reduction in the price of prescription drugs; this decline is partially offset by spending associated with increased drug use from the improved coverage. Average out-of-pocket spending on drug purchases in 2006 will decrease from an estimated $1,325 before enactment of the MMA to about $890 under the MMA, a reduction of about one-third. -17- III. Impact of the MMA by Type of Pre-MMA Prescription Drug Insurance As described earlier, the new benefit will have two primary impacts: greater drug coverage and lower prices for drugs. Therefore, an important factor determining the impact of the MMA on beneficiaries is the type of prescription drug coverage they had before implementation of the new drug benefit. Other than health status, no other factor affects total prescription spending more than the type of prescription drug insurance. As discussed earlier, Medicare beneficiaries who do not currently have prescription drug coverage are the most likely to gain from the MMA, but even they will only see their out-of- pocket spending reduced if their current spending sufficiently exceeds the value of the new premium and deductible.24 At the other end of the spectrum, in the aggregate, Medicare beneficiaries who also have Medicaid coverage (so-called “dual eligibles”) and most retirees whose former employers provide prescription drug coverage are expected to see relatively little change in prescription drug benefits or out-of-pocket expenses for drug purchases. In the middle, Medicare beneficiaries who currently purchase prescription drug coverage in private Medigap and Medicare Advantage plans, as a group, are likely to have somewhat better coverage and pay somewhat lower prices than before. PricewaterhouseCoopers (PwC) estimates that the Medicare population’s share of total drug spending covered by insurance (private and public, not just the Medicare benefit) in 2006 will increase from 58 percent in the absence of the MMA to 71 percent under the MMA. As a result of the increase in insurance coverage, drug use will increase by 6 percent. Overall, PwC expects prices to fall by more than 8 percent as plans negotiate larger discounts in exchange for tougher benefit management. The level of both additional use and price discounts will vary based on current source of insurance coverage and average discounts under those plans. Table 7 shows the details for each category of current source of supplemental insurance. 24 Even if the amounts paid out of pocket by low-spending beneficiaries increase as a result of participation in the Part D benefit, the beneficiary will still receive an offsetting benefit from the insurance value of catastrophic protection. -18- Table 7 Impact of MMA on Drug Use and Prices, by Source of Prescription Drug Coverage, 2006 Percent of Total Drug Change in Primary Source of Beneficiaries Costs Paid by Insurance Change Change Total Drug Drug Coverage (millions) in Price in Use Pre-MMA Post-MMA b Spending c Self-Pay 11.9 0% 59% -25% 25% -6% Medicare Advantage 3.0 40% 60% -6% 12% 5% Medigap 4.2 42% 60% -15% 14% -3% Employer-Sponsored a 11.8 71% 67% -5% * -6% Medicaid 6.2 95% 95% 0% 0% 0% Other Federal 2.6 70% 70% 0% 0% 0% Total Beneficiaries 39.9 58% 71% -8% 6% -3% Source: PricewaterhouseCoopers’ calculations. Note: Numbers may not add to 100 percent due to rounding. *Denotes less than 1 percent. a Includes retirees with employer-sponsored insurance covered by qualified employer plans that receive the alternative subsidy. b Includes impact of low-income subsidies. c Overall percent change in spending equals [(1 + percent change in price) * (1 + percent change in use)] - 1. A. Self-Pay (No Prescription Drug Coverage Pre-MMA) In the absence of the MMA, and assuming no current use of private discount cards, the nearly 12 million Medicare beneficiaries who do not have drug coverage usually pay full price for drugs (their average discount is zero), and none of their prescription drug spending is paid by insurance.25 PwC estimates that, on average, under the MMA, the standard benefit will cover 53 percent of drug costs in 2006 for beneficiaries currently without coverage. Any additional coverage will increase the share of spending covered. For instance, the low-income subsidies at the individual beneficiary level cover between 85 percent and 95 percent of drug costs, depending on the type of low-income subsidy for which the individual is eligible. Incorporating these subsidies for the self-pay population overall results in about 59 percent of prescription drug spending being paid by insurance—in this case, Medicare—on average (i.e., the out-of-pocket spending share will fall from 100 percent of total drug purchase costs to 41 percent). The discounts and benefit management that are expected to average about 25 percent under the private prescription drug plans will also reduce spending. These Medicare beneficiaries who 25 Including a discount associated with drug cards has only a small impact on the estimates of coverage provided by the MMA to self-pay individuals. For instance, assuming the average discount for the self-pay population is 5 percent, the change in utilization will be 24 percent (rather than 25 percent) and the change in price will be -21 percent (rather than -25 percent). -19- are currently without drug coverage are expected to dramatically increase their use of prescription drugs: a combination of lower prices and insurance coverage will increase use by roughly 25 percent. After accounting for the reduction in price, net spending in dollars will fall by about 6 percent.26 B. Individual Private Plans (Medicare Advantage and Medigap) Roughly 7.2 million Medicare beneficiaries, or 18 percent, purchase private health insurance plans individually that include prescription drug coverage. In both MA and Medigap plans, enrollees find that health insurance covers only about 40 percent of drug costs. Under the MMA, insurance coverage is expected to cover about 60 percent of costs on average; the exact share of drug costs the Medicare plan covers varies slightly by prior insurance coverage based on the likelihood of receiving low-income subsidies and any other supplementation of the Medicare benefit. The discounts under Medicare Advantage plans vary based on the influence the plans have over prescribing physicians. Under staff-model HMOs, the plans can exert significant control over the drugs members use. As a result, the average discount under these plans is larger than under the typical employer plan. Other Medicare Advantage plans are less successful in controlling costs, so they achieve savings similar to those of typical employers. Overall, PwC estimates that Medicare Advantage plans currently achieve average total savings of around 20 percent, which will increase to 25 percent under the MMA benefit assuming that plans that are less strict today will be able to control drug costs more tightly under the MMA. Combining the lower average cost with the increase in use, average spending by the population in Medicare Advantage plans will increase by approximately 5 percent. Current Medigap plans with drug coverage have few if any controls over plan participants’ drug spending. PwC estimates that the typical Medigap plan can achieve total savings of approximately 12 percent now, but these savings will increase to 25 percent under the MMA benefit. Overall, total spending by current Medigap purchasers will decrease by 3 percent, after incorporating the 14 percent estimated increase in use.27 C. Employer-Sponsored Retiree Plans Retiree plans are the most common source of prescription drug coverage for the Medicare population. About 11.8 million or 30 percent of Medicare Part B beneficiaries currently receive drug coverage through these plans. Employer-sponsored plans have better coverage, 26 The resulting price level is calculated is as follows: (1-0.25=0.75) * (1+0.25=1.25) = 0.94, or a reduction of approximately 6 percent. 27 The estimated utilization increase in Table 7 is larger for the Medigap population (14 percent) than the Medicare Advantage population (12 percent) because the former is estimated to experience a larger decrease in the average price for drugs, which fuels additional induced demand. -20- on average, than MA and Medigap plans and most have catastrophic protection at some level of prescription drug spending. Under the MMA, employers have at least three choices: 1. Receive alternative subsidy—Employers may offer coverage that is actuarially equivalent to the standard Part D benefit (this may represent either a change in or a continuation of their existing retiree drug coverage) and receive a 28 subsidy on spending between $250 and $5,000 for each retiree who does not enroll in Part D. Based on a representative distribution of drug spending consistent with average drug spending by retirees, for a typical employer, the value of the 28 subsidy will be about $665 per retiree in 2006.28 2. Supplement PDP—Employers may encourage their retirees to enroll in the new private plans by providing only supplementary prescription drug insurance that pays for some or all cost sharing under the standard Medicare drug benefit. Providing such employer coverage will reduce the dollars covered under the standard Medicare benefit by increasing the amount of total spending required before the catastrophic benefit begins (due to the true out-of-pocket rules explained in previous sections). For example, a retiree receiving supplemental employer-provided protection with an out-of-pocket spending maximum of less than $3,600 will never reach the $3,600 out-of-pocket spending level required to receive Medicare’s catastrophic benefits, and for which the PDP will have received reinsurance subsidies. Assuming current employer plans cover about 75 percent of drug spending for retirees, an employer can maintain current levels of insurance by covering 75 percent of spending below the $250 deductible and in the doughnut hole. Such an employer will save about $710 per retiree before income taxes.29 3. Drop existing coverage—Employers may encourage their retirees to enroll in the new private plans by dropping all company prescription drug benefits. These employers may or may not choose to pay the Part D premium on behalf of their retirees.30 The value of the standard Medicare benefit for a typical retiree will be about $1,670 in 2006 after netting out the enrollee premium. (Because retirees with employer coverage generally spend more in total on prescription drugs, the gross value of the standard benefit exceeds the $1,260 overall average.) Even if the employers that drop their 28 Employers that provide coverage through company plans will be able to disregard this subsidy when deducting the cost of retiree prescription drug benefits for income tax purposes. The Congressional Budget Office estimated that the tax-related value of this subsidy will be almost $18 billion over the 2006–2013 period. In the absence of this provision, the after-tax value of the subsidy will be approximately 25 percent lower on average since any dollar received from the government will increase taxable income by a dollar. 29 The $710 subsidy assumes that employers require the same contribution by retirees for coverage before and after MMA (e.g., if an employer pays the entire premium for retiree coverage before MMA, it is assumed to do so afterwards). 30 CBO did not specify in its cost estimates whether employers who drop existing coverage will pay the premiums for enrollees. -21- current plan offer stingier drug coverage than other employers do in the absence of the MMA, they can still lower company costs significantly by dropping their coverage. For example, with a pre-MMA retiree drug insurance cost of $2,260 (representing coverage of about 60 percent of per capita drug spending), an employer can drop its plan, pay the MMA enrollee premium, and save $1,840 before income taxes ($2,260 savings minus the $420 premium). The employer can offer other nondrug benefits (a Medigap plan for medical expenses, for instance) to retirees to keep them whole (i.e., maintain the same overall level of benefits) and still save $1,250 per retiree. In this case, the employer provides additional benefits worth $590 so the average retiree benefit remains at $2,260 ($1,670 from the MMA benefit plus $590 additional benefit).31 Estimating the number of retirees whose employers will fall into each of these three categories is complicated by a number of unknowns. Because of union contracts or other binding agreements, many employers do not have the flexibility to change the company plan. Also, many employers are resistant to reducing benefits to individuals who have planned their retirement around pre-MMA plans. If this were not the case, employers could have already saved the cost of prescription drug benefits by dropping company prescription drug coverage (i.e., before subsidies are available under the MMA). Between legal requirements and reluctance to harm retirees, most employers are expected to continue existing prescription drug coverage, at least in the first few years under the MMA. Once the PDP market is sufficiently established, employers might be more willing to purchase PDPs rather than sponsoring their own plans. Employers also might have information about utilization characteristics of their retiree population that might influence their decision. In addition, the administrative costs associated with coordinating a benefit that wraps around PDP coverage could encourage employers to continue to offer the company benefit. 31 Although the subsidy amounts presented in this section differ from the net subsidies of the different options calculated by CBO in its July 2004 report, the basic conclusions are the same: employers receive significantly higher subsidies by dropping coverage. CBO estimated that the Medicare program will pay $692 (compared to $710 estimated by PwC) in 2006 per retiree whose employer wraps around the Medicare benefit; $766 (compared to $665 estimated by PwC) if the employer takes the alternative subsidy; and $1,201 if the employer drops its drug coverage (compared to $1,250 estimated by PwC if the employer drops its plan but offers other compensating benefits to retirees). PwC estimates differ from the CBO estimates because the base distributions of spending by retirees appear to be different. PwC figures presented are based on a representative employer, without adjusting for any systematic differences in retiree spending that may affect the decision the employer will make with regard to the options under the MMA. The CBO subsidies for each option appear to reflect different underlying average spending levels. Finally, for the wrap-around option, the net savings will depend on the amount of supplementation by the employer (more generous wrapping lowers the savings because it raises the amount of total spending required before the catastrophic benefit level is reached). Because deductible costs for retiree insurance would be lower if employers dropped coverage or wrapped around a PDP, the net savings to employers, after incorporating income tax effects, would be approximately 25 percent smaller than estimated above. -22- However, given that the ease of dropping drug coverage under a retiree plan increases with the new MMA benefit, employers will have an added reason to eliminate company-sponsored drug benefits. Additionally, employers that already planned to drop their plans in the future might do so more quickly, leaving retirees worse off than before. CBO estimates that roughly 2.7 million, or about 23 percent of retirees with employer coverage in 2006, will move from company plans to PDPs and MA-PDPs with the standard Medicare drug benefit. This change is in addition to any general deterioration in employer coverage that might have occurred in the absence of the MMA. The retirement of the baby boom generation will put significant stress on retiree health plans, which will cause general coverage to drop over the long term. Despite the expectation that some employers will stop offering drug coverage through their own plans, two factors may mitigate losses to retirees. First, among employers that currently offer coverage, those that choose to drop the company-sponsored prescription drug coverage are not likely to have the most generous employer plans. That is, any employer that currently has a plan that is not as good as the standard plan under the MMA will choose to drop the plan rather than spend more to become a qualified retiree plan eligible for subsidies. The retirees, in this case, will not be worse off under the PDP than they were under the employer plan. Second, employers may compensate workers who switch to the MMA benefit with increases in other benefits or supplementation of Part D, thus mitigating the impact of the switch on retirees. That is, many employers that drop plans are likely to try to use some of the employers’ windfall to make up the loss to the retirees. The movement from employer plan to PDP the new benefit causes is reflected in Table 7 by a drop in the amount of total drug spending covered by insurance from 71 percent to 67 percent. As a result, use will fall only slightly, but dollar spending is expected to decline by 6 percent because PwC assumes that discounts are higher under the PDPs than they are under current employer plans.32 D. Medicaid Medicare beneficiaries with Medicaid coverage, or dual eligibles, will no longer receive prescription drug coverage through state Medicaid programs but, instead, will enroll in private drug plans. These beneficiaries will have no deductible, and, unless states provide supplemental assistance with state-only dollars, they will pay nominal copayments equal to $1 for generic prescription drugs and $3 for brand-name prescription drugs. To estimate the prices and utilization shown in Table 7, PwC assumed that cost sharing by this population will 32 The estimate of the impact of the MMA on employers in this section is based on the assumption that only about 20 percent of employers drop their current plans and do not supplement. If PwC assumed that more employers dropped their current plans but most also provided some supplementation (but not enough to replace current benefits), the impact on utilization may be similar to what is estimated using the simpler assumption that 20 percent of retirees lose employer coverage with no supplementation. If more employers drop their plans, the change in price will be greater, even if employers provide supplementation. -23- remain roughly constant, or about 5 percent of total drug costs, under either state Medicaid or post-MMA private plans that enroll dual eligibles.33 As discussed in more detail below, this is a good approximation of the likely situation in the absence of, as well as under, the MMA. E. Other Federal Coverage Medicare beneficiaries with coverage from federal health programs such as TriCare have no incentive to enroll in the PDPs, so they are not expected to participate. This population generally has better coverage and larger discounts than it will under the MMA, and the federal government is unlikely—for both political and practical reasons—to drop the current plans and force them into the Medicare plan. If the Federal Employees Health Benefits Program (FEHBP), for example, enrolls its retirees in PDPs and provide supplementary coverage to make up the difference in benefits, the federal cost of subsidies will be similar to the existing subsidies for federal retirees and the overall value of benefits for retirees will not change substantially. For these reasons, PwC follows CBO by assuming that most of these beneficiaries will continue their existing prescription drug coverage. 33 Within this constant average, some Medicaid beneficiaries may have slightly increased cost sharing, and others may find that their particular prescription drug use has greater coverage under the MMA than it did under the existing state programs. Currently, some states do not have copayments under their Medicaid prescription drug programs. The yearly difference between no copayment and the $1/$3 copayments under the MMA will be about $100 for a typical Medicaid enrollee. However, some states that do not have copayments have other rules (e.g., prescription limitations, benefit limitations) that are more restrictive than those under the Medicare prescription drug benefit. In these cases, dual eligibles may incur lower cost sharing under the MMA plans than they did under their own state’s Medicaid program rules. -24- IV. Individual Variation in the Impact of the MMA The overall impact and the more disaggregated impact by supplemental health insurance category examined so far in this paper fail to reflect the individual variation that is key to understanding the implications of such a major change as the MMA has introduced. A complete census of Medicare beneficiaries and their circumstances is, unfortunately, not available. In fact, no public survey contains all the necessary information to assess the likely impact of the MMA by the four factors that most directly influence how beneficiaries will fare. These four factors are as follows: • Specific pre-MMA benefit—The adequacy of drug coverage that a beneficiary currently has (including how it is financed) affects the extent to which new drug coverage under the MMA is considered to be an improvement. Even within a category of supplemental insurance, prescription drug benefits vary. For example, although many employers offer relatively good retiree prescription drug benefits, some may have benefit limits and be no better than typical Medicare Advantage plans with respect to drug coverage. Likewise, some MA plans may have generous coverage that is more typical of employer coverage. • Income level—As discussed in detail above, low-income Medicare beneficiaries who meet the asset test are eligible for additional subsidies that allow for very limited cost sharing and, in most cases, no monthly premium. Individuals who do not qualify for these protections (e.g., those individuals in the middle of the income distribution or with assets exceeding the thresholds specified in the MMA) face the greatest risk of high out-of-pocket spending relative to their income. • Level of prescription drug use—Expected prescription drug spending is important because it affects the level of drug benefits provided. For example, Medicare beneficiaries with total drug spending that exceeds $2,250 in 2006 will encounter the doughnut hole during which they have no coverage from Medicare; less than half of beneficiaries are expected to reach this level of drug spending. Beneficiaries who do not qualify for the low-income protections and have sufficiently low spending will pay more in new deductibles, coinsurance, and premiums than their total drug bill in the absence of the MMA. On the other hand, individuals with moderate or very high drug spending will receive coverage for a larger share of their costs under the MMA standard benefit structure. • Post-MMA supplementation—The actual impact on beneficiaries will also depend on the response of state pharmaceutical assistance programs, employers, Medicaid, Medicare Advantage plans, and other potential sources of supplemental insurance. If these providers choose to supplement the standard MMA benefit or cut back on current coverage, the out-of-pocket costs to enrollees will change. -25- These factors will interact in complex ways. Although about 20 percent of Medicare beneficiaries are projected to have total spending at or above the $5,100 level in 2006 (at which point out-of-pocket spending is $3,600 if the beneficiary does not have supplemental drug coverage), considerably less than 20 percent of them will reach the out-of-pocket threshold in any given year. First, about 30 percent of Medicare beneficiaries will have low- income subsidies and, therefore, protection from the doughnut hole. Second, about 23 percent of Medicare beneficiaries will have employer coverage, which will protect most of them from catastrophic expenses. Finally, some Medicare beneficiaries will probably purchase enhanced coverage despite the higher premiums and receive insurance coverage in addition to the MMA benefit. In the absence of comprehensive data that allow precise examination of the impact by insurance, income, and spending levels, or the outcomes of decisions and actions regarding the future offering and generosity of supplemental drug coverage, this section looks at illustrative cases that may help to suggest the types of individuals who will fare better or worse than the average statistics for the group imply.34 A. Illustrative Case Related to Population with No Coverage A high-spender with little or no drug coverage, including an individual with no catastrophic drug coverage, will have a strong incentive to enroll in the new program.35 Individuals who spend more than the catastrophic benefit level will still pay the same premium as everyone else (estimated at $420 in 2006) but will receive at least $1,500 from the Medicare benefit (the amount paid out for low-end coverage). Use of any catastrophic benefits will only increase the plan’s payout. 34 This paper did not present examples of Medicaid enrollees or Medicare Advantage participants. As discussed in the previous section, the authors expect the impact on the Medicaid population to be small. Providing an illustrative example for the MA plans is difficult for several reasons. First, the drug coverage MA plans offer varies greatly, ranging from relatively sparse (say, $200 per year), to relatively generous but with limits (e.g., covers all spending on generic drugs). Illustrating the range of pre-MMA coverage with two or three case studies is difficult, if not impossible. Second, determining whether an individual with MA coverage is better or worse off under MMA should incorporate the other changes enacted to the MA program by the MMA (such as the increases in payment rates to MA plans for Part A and B coverage). Such an analysis is beyond the scope of this paper. 35 According to Table 1, almost half of Medicare beneficiaries lack drug coverage that protects against catastrophic costs, although only some of these beneficiaries have no coverage or are currently high spenders. Self-pay (28 percent), Medigap (7 percent), and Medicare Advantage (10 percent) are assumed to lack catastrophic coverage. Medicare enrollees with Part A only (6 percent), other public coverage (6 percent), and employer-sponsored coverage (28 percent) might also lack catastrophic coverage, which would boost the overall share currently without protection against high spending. -26- In 2006, at levels of total drug spending below $810, the premiums and cost sharing under the standard Medicare plan exceed prescription drug spending without the Medicare coverage.36 For Medicare beneficiaries who decide to purchase a PDP or MA-PD, and whose drug purchases remain below $810, the maximum loss will equal the value of the enrollee premium, estimated at $420 in 2006. Given this possible outcome, some analysts have questioned whether individuals who currently do not have drug coverage but are relatively healthy and, therefore, have low spending in any particular year will enroll in PDPs or MA-PDPs rather than continuing not to have any prescription drug coverage. The answer depends, in part, on whether a beneficiary focuses only on the current level of drug use or also considers his or her potential for greater drug use in the future. Because low spending in a particular year does not guarantee low spending in the future, insurance is a tool to control the risks associated with unexpected outcomes, in this case high drug spending. By definition, insurance has value for both those who know that their spending will exceed their premium contribution and those who know that their spending might exceed the premium. The late enrollment penalty creates the imperative for healthy beneficiaries to consider these factors when they are first eligible to enroll in Part D. As a result, even beneficiaries with low current spending are expected to enroll in the Medicare drug benefit in order to reduce their risk of large out-of-pocket expenses in the future. B. Illustrative Cases of Medigap Purchasers under the MMA Medicare beneficiaries purchase Medigap policies to provide catastrophic protection for high (nondrug) medical costs, to fill in some of the cost sharing, and, in some cases, to provide prescription drug coverage. Although there are 10 Medigap plans, A to J, only three (H, I, and J) cover prescription drugs. This discussion focuses on the drug benefit that is part of plans H and I.37 With respect to prescription drugs, plans H and I have a $250 deductible, 50 percent coinsurance for total spending between $250 and $2,750, and a benefit limit of $1,250, after which the plan ceases to pay for prescription drugs. Compared to the standard drug benefit under the MMA, plans H and I have the same deductible, less generous coverage at the front- end benefit (50 percent coinsurance required vs. 25 percent), more generous coverage for the first $500 of the doughnut hole (50 percent coinsurance compared to 100 percent), but no coverage at catastrophic spending levels. Table 8 shows the impact, at different drug spending levels, of prescription drug benefits under plans H and I compared to the MMA standard benefit package. The table portrays out- 36 Roughly 30 percent of all Medicare beneficiaries have total drug spending less than this amount; the portion without drug coverage who spend this much is, obviously, smaller. 37 Based on analysis by the General Accounting Office for 1999, 4 percent of Medigap enrollees chose Plan J, and 4 percent chose Plan H and I combined. See GAO, “Medigap Insurance: Plans Are Widely Available but Have Limited Benefits and May Have High Costs,” July 2001, GAO-01-941. -27- of-pocket spending on drug purchases and total out-of-pocket spending, including premiums for the drug component of plans H and I and for the standard benefit under the MMA. As shown in the first row, the extra premium for drug coverage under plans H and I is estimated at $950 in 2006. CBO estimated the premium for the standard benefit under the MMA at $420. The enrollee premiums for the drug component of plans H and I are higher than the premium under the MMA in part because Medigap plans are not subsidized (compared to the MMA plans, 74.5 percent of whose cost is federally financed). Another important factor in the premium difference is that Medicare beneficiaries who choose plans H or I tend to have higher medical and prescription drug costs than typical Medicare beneficiaries do. Because these enrollees generally have higher costs, premiums for these plans are correspondingly higher (the insurer must charge high enough premiums to cover the amount of insurance payments). An insurance product that attracts a relatively sicker population, and therefore must charge higher premiums, exhibits adverse selection. The extra costs of the drug benefit in plans H or I, if averaged over typical Medicare beneficiaries, is about $750; adverse selection into plans H and I will increase the cost to an estimated $950 in 2006.38 One factor to consider is that the Medigap premium is based on analysis of existing products, while the PDP and MA-PDP premium—which can vary across plans and markets—is only an estimate. The actual premium under the MMA benefit might be different from the estimated amounts. Cost sharing and the sum of cost sharing and premiums are estimated to be lower under the MMA drug benefit than under Medigap plans H or I at all drug spending levels (as Table 8 illustrates). 39 The difference in favor of the MMA benefit increases from a low of $530 at zero total drug spending (the difference between the $950 Medigap premium and the estimated $420 PDP premium under the MMA). The difference narrows between $2,250 and $2,750 in total drug spending and remains constant at $780 until catastrophic coverage begins under the MMA at $5,100. The difference in favor of the MMA plan continues to increase at higher spending levels. 38 This illustrative example is based on the assumption that adverse selection in Plan H (or I) increases premiums by about 10 percent more than the premiums that would be charged for typical Medicare beneficiaries with average spending. The entire adverse selection amount (equal to about $200 in 2006, or 10 percent of an estimated Plan H premium of roughly $2,000, which includes both medical and drug coverage costs) is paid by a beneficiary who enrolls in Plan H or I instead of enrolling in plans that have no drug coverage and, therefore, less adverse selection. 39 Although this case illustrates PwC’s best estimate of a typical situation, Medigap premiums vary widely by locality and by insurer. Since PDP premiums will also vary, in theory, some beneficiaries could have higher costs under the new Medicare coverage than under their previous plans. -28- Table 8 Comparison of Out-of-Pocket Drug Spending for a Beneficiary Medigap Plan H (or I) vs. MMA Standard Benefit, 2006 Plan H (or I) MMA Difference under MMA Prescription Drug Cost Total Enrollee Cost Total Enrollee Cost Total Enrollee Spending Level Sharing Cost Sharing Cost Sharing Cost $0 $0 $950 $0 $420 $0 ($530) $500 $375 $1,325 $313 $733 ($63) ($593) $1,000 $625 $1,575 $438 $858 ($188) ($718) $2,250 $1,250 $2,200 $750 $1,170 ($500) ($1,030) $2,750 $1,500 $2,450 $1,250 $1,670 ($250) ($780) $5,100 $3,850 $4,800 $3,600 $4,020 ($250) ($780) $10,000 $8,750 $9,700 $3,845 $4,265 ($4,905) ($5,435) Source: PricewaterhouseCoopers’ estimates. Note: Total Enrollee Cost is equal to cost-sharing plus premiums specific to drug coverage. PwC also calculated the out-of-pocket costs and total expenses of the other Medigap prescription drug plan (offered in Plan J) in 2006. The extra premium for drug coverage under Medigap Plan J is estimated at $1,500 in 2006. Plan J has the same deductible and coinsurance as Plan H and I but $3,000 in maximum benefits, a level that occurs at $6,250 in total prescription drug spending. Therefore, Plan J results in the same prescription drug cost sharing as plans H and I at spending levels below $2,750 but lower cost sharing at higher spending levels. Unlike plans H and I, there are certain points in the doughnut hole where Medigap Plan J results in lower cost-sharing than under Medicare. However, similar to the results for plans H and I, once the premium difference is incorporated, the total expense (i.e., the sum of premiums and cost-sharing) under the MMA benefit is lower than that for Plan J for all levels of spending. C. Illustrative Cases of Retirees with Employer Coverage under the MMA Certain large shifts in costs to retirees are possible, but not necessarily likely, under the MMA for the reasons discussed above (i.e., current long-term contracts, employers’ desire to protect retirees, alternative subsidies that encourage continuation of plans). Nevertheless, one potential outcome for a retiree with employer coverage will be the loss of that coverage and subsequent enrollment in a Medicare PDP with the standard benefit (the second option described in the earlier discussion).40 40 Employers with less generous retiree drug coverage are expected to be more likely to drop their retiree plans than are employers with generous coverage whose retirees will have the most to lose; however, the overall -29- Figure 3 illustrates the impact of an employer’s decision to eliminate drug benefits for retirees across a range of total prescription drug spending in 2006. The example is based on the assumption that the employer plan is one with cost sharing of about 24 percent at every level of spending such as will be the case with typical tiered-copayment plans. The retiree who is enrolled in a PDP after the employer plan is terminated will have higher cost sharing until total spending reaches $17,600 (not shown). The difference will especially increase in the doughnut hole and will be greatest for those whose spending reaches the catastrophic benefit level. At a total spending level of $5,100, a retiree with only the standard MMA benefit will pay $2,376 more out-of-pocket than he or she will under an employer plan that pays 76 percent of costs. Figure 3 Change in Out-of-Pocket Drug Spending for a Retiree Who Loses Employer Drug Coverage, 2006 Source: PricewaterhouseCoopers. Note: Analysis assumes that the employer pays for the PDP premium and the retiree plan covers a flat 76 percent of drug costs. Unless the employer has a lower catastrophic threshold than the Medicare drug benefit, the difference in out-of-pocket spending on drugs falls when Medicare’s catastrophic benefits begin. The difference in out-of-pocket amounts tapers off at higher levels of spending until, at average coverage level is used to illustrate the average-case scenario and to avoid presenting only a best-case scenario. -30- around $17,600 in total drug spending, out-of-pocket spending under the standard MMA benefit is less than under the employer plan. Based on a distribution of drug spending that represents all Medicare beneficiaries, not just those with retiree benefits, about 20 percent are projected to have total spending of $5,100 or more (or $3,600 in out-of-pocket spending for someone without supplemental coverage in 2006). Fewer beneficiaries will have out-of- pocket spending of this level if employers supplement the Medicare drug benefit. This example illustrates that the impact on retirees is likely to vary widely. Based on the distribution of drug spending consistent with a typical retiree population, about half of retirees with employer coverage have total drug spending below $2,250. Many retirees with low drug costs will see small differences between the standard MMA plan and employer coverage (although even those with low costs will notice the additional risk associated with the more basic MMA coverage). The retiree who has total drug spending of $500 will have out-of- pocket cost-sharing expenses of $120 under the employer plan (24 percent of $500) and $313 under the MMA benefit ($250 deductible plus 25 percent of spending above deductible), a difference of $193. At a total drug spending level of $2,250, out-of-pocket cost-sharing expenses under the Medicare benefit will be $210 higher than those under the pre-MMA employer plan. So, on average, retirees with up to $2,250 in total drug spending will pay at most about $200 more under MMA in the example above. This analysis assumes that the employer will pay the monthly Medicare drug premium for the retiree. As discussed earlier, employers most likely will save significant amounts by transferring retirees to the Medicare plan, even net of paying the Part D premium. They have also shown an interest in continuing to provide retirees with drug coverage (currently they can drop coverage to save money but they have not). Some retirees pay a portion of or the entire premium associated with health insurance. Because of the significant federal subsidy for the MMA drug benefit, it seems unlikely that employers will increase the premium contributions of enrollees as a result of the MMA (i.e., require retirees to pay more in Part D premiums than they paid under the employer plan pre- MMA).41 In light of the potential savings, employers could lower or eliminate the required contribution by retirees towards their drug coverage. Therefore, the changes illustrated in Table 8 above represent the upper boundary of the total change in spending paid directly by enrollees. 41 In the unlikely case that the MMA benefit is more generous than the employer plan, an employer might be justified in raising the enrollee premium for drug coverage. However, employer drug coverage tends to be more generous than the MMA coverage. Employers that decide to charge retirees the Part D premium in 2006 probably currently charge a premium for retiree drug coverage. Given that the federal subsidies significantly lower the enrollee cost of the MMA plan, even if the retiree has to pay the Part D premium, it seems unlikely that the federally subsidized premium will be higher than the premium for employer-provided drug coverage. -31- While the retirees in this example will be worse off under the standard MMA benefit than under the employer plan, the savings to the employer will be significant at each spending level. The employer that drops a typical company plan to enroll the retirees into a PDP with standard benefits will shift from paying about $2,260 for the benefits (not counting administrative expenses) to paying only the $420 annual premium for prescription drug coverage, a net gain of $1,840. Such employers can provide additional assistance to retirees and still realize large gains from dropping the company prescription drug benefit. Not shown in Figure 3 is an example where the employer drops the retiree drug benefit, pays the PDP premium, and provides $1,000 in cash or additional medical benefits to retirees. In this case, any retiree with an increase in out-of-pocket spending of less than $1,000 as a result of the switch from the retiree plan to a PDP will be better off on average. The windfall to the employer will be large enough to compensate retirees so the loss on average is zero.42 42 Even if the employer offers cash compensation to retirees for the less generous drug coverage to eliminate the losses on average, the effects on individual beneficiaries would vary by spending level. For instance, the $1,000 cash payment to beneficiaries will be sufficient to keep retirees whole on average, but as the previous discussion illustrates, some retirees will do worse and some will do better. -32- V. Conclusion Overall, and on average, Medicare beneficiaries are net winners from the Medicare Prescription Drug, Improvement, and Modernization Act of 2003 (MMA). Some Medicare beneficiaries, however, are expected to be worse off than they would be in the absence of the new Medicare drug benefit. Drawing conclusions about the number or proportion of Medicare beneficiaries who will fall into the “winners” and “losers” categories is difficult. Not only are the exact numbers difficult to estimate, the estimate also will depend on the definition of winner and loser. For example, Medicare beneficiaries who will not have prescription drug coverage in the absence of the MMA seem to be winners. However, even in that category, those who pay more in premiums than they receive in benefits in a given year might be labeled as losers for that year. Moreover, some of the winners and losers may only experience modest gains or losses. For example, dual-eligible Medicare beneficiaries are unlikely to see differences of more than $100 between the state Medicaid coverage and the private plans under the MMA. Medicare beneficiaries by type of prescription drug coverage can be grouped into three categories: winner, loser, or little or no change. Medicaid, as noted, mostly falls under little or no change. Beneficiaries with low incomes and few assets who are not also in Medicaid will be winners due to their eligibility for additional protections. Ignoring small differences in benefits, Medicare beneficiaries who are in Medigap plans and those who are Self-Pay mostly fall into the winner category. The situation for beneficiaries in Medicare Advantage plans is more complicated. In general, most MA enrollees have very thin coverage that they implicitly pay for in premiums or in lieu of other benefits. For that reason, most MA enrollees probably fall into the winner or little or no change category. The group that potentially faces the largest losses under the MMA are retirees in employer- sponsored plans. Table 7 indicates that, on average, they end up with less prescription drug coverage than they had at pre-MMA levels. Most retirees, especially in the early years of the MMA, are expected to have the same benefits under the MMA as they would have had if the MMA had never been enacted. Roughly 20 percent of employers are expected to drop their current plans in favor of the MMA coverage. This large group, accounting for about 6 percent of all Medicare beneficiaries, will be at risk for larger out-of-pocket expenses under the MMA than what they face now. Many such beneficiaries, however, might fare better than expected due to the mitigating factors discussed earlier. -33-