An Evaluation of Medicare’s Prescription Drug Policy


Chairman Coburn, Senator Carper, members of the Committee, thankyou for the opportunity to testify on the Medicare PrescriptionDrug Program. I feel very honored by it.1

I especially appreciate this opportunity because there is nomore important or scarier policy issue than how to design policiesto preserve the efficient operation of health care markets inattempting to pay for our growing health care needs. It is wellknown that designing policies to improve health-care marketefficiency is difficult. But it is not yet widely appreciated howhuge Medicare's future financial shortfall is. The MedicarePrescription Drug Improvement And Modernization Act of 2003 (MMA)substantially increases that shortfall and is likely to worsen theoperation of markets for prescription drugs and drug insurance. Assuch it deserves urgent reconsideration-a view that is shared bymany health care experts and policymakers including, I suspect, bymembers of this Committee.

I. Introduction

MMA offers prescription drug coverage to all retirees.The new law will benefit seniors on the whole but will exertseveral negative economic effects:

Five issues stand out:

  • Government intervention is usually justified when privatemarkets fail. With 75 percent of retirees already havingprescription drug coverage and 90 percent having access toprescription drugs prior to MMA, this market did not exhibit thesymptoms of "market failure." Indeed, passage of MMA is likely tocause market failure by displacing the private market'sprovision of drug insurance.
  • MMA will improve access to prescription drugs for poorerretirees - both those who are and those are not currently coveredunder Medicaid. Well-to-do retirees will also benefit in generalbut some may experience higher out-of-pocket costs if they losetheir private drug coverage and are forced to enroll into MedicarePart D. This law, therefore, appears designed to first displace theprivate market followed by sustained pressure on Congress toliberalize the MMA's benefit formula over time.
  • MMA will influence prescription drug prices in the privatemarket as the share of government-subsidized purchasers expands.Theoretical reasoning and empirical studies suggest that privatedrug prices would increase with additional government-subsidizedpatients entering the market. Most of the burden of this increasewill fall on workers covered by making employer-provided healthinsurance or private plans more expensive. That will reduce youngerworkers' likelihood of employment, cause lower wage growth,increase conversion from full- to part-time jobs, and reduce workeffort.
  • MMA makes a large addition to the already considerablefinancial shortfall in the rest of Medicare. Unresolved, thisshortfall will grow larger and impose higher fiscal burdens onfuture generations, further eroding their productivity and workincentives.
  • MMA will change workers' and younger generations' perceptionsabout the need to save for health-care expenses during retirement.Studies show that expansion in government entitlement obligationsleads to higher consumption and reduces national saving andinvestment-delivering a further negative impact on future workerproductivity and output.

MMA was hastily passed without a proper evaluation of its short-and long-term cost and it lacks appropriate measures to controlspending escalations. That means future Congresses may be inducedto regulate the actions of pharmacies, drug manufacturers,employers, and plan providers with regard to drug pricing andspending per person on prescription drugs. Such regulations wouldbe counterproductive because they would restrict prescription drugsupply, generate illegal prescription drug sales, and reduce thequality of prescription drug coverage for everyone - and not justfor retirees.

If MMA's repeal is deemed impractical, a financially andeconomically sensible course would be to scale it back to asustainable level by providing coverage only to those seniors whoare under financial pressure on account of their prescription drugexpenses. That effort needs to be combined with restoring the restof Medicare to financial sustainability.

II. Pre-MMA Prescription Drug Coverage ofRetirees

Prior to MMA's enactment, Medicare Parts A and B provided nolimits on out-of-pocket costs and did not insure retirees againstoutpatient prescription drug expenses.

The vast majority of retirees (75 percent) had prescription drugcoverage under private plans: Employer supplemental health coverage(33 percent), Medicaid and state drug programs (17 percent),Medicare+Choice Plans (15 percent), Medigap policies withprescription drug coverage (2 percent) or other sources(8%).2 New retirees wereguaranteed access to 10 alternative Medigap plans, three of whichcovered prescription drugs.

Some retirees, however, faced financial pressure on account oftheir prescription drug costs: Estimates as of 2000 suggest thataverage out of pocket costs for retirees in poor health took upabout 44 percent of their incomes.3 Low-income single women not covered underMedicaid spent about 52 percent of their incomes on healthexpenses, on average.

Enrollment into Medigap plans including prescription drugcoverage has been quite low. Such plans impose spending caps and sodo not cover catastrophic expenses. Their high premiums,deductibles and cost-sharing requirements make them expensive andtheir availability varies widely by geographic area. Premiuminflation among plans with prescription drug coverages has beenvery rapid. The plans also provided first-dollar coverage thatdiscouraged prudent use of services and prescription drugs.

These features made Medigap policies inferior to employersupplemental coverage, which generally had low co-insurancerequirements, no separate spending caps for prescription drugs, anddrug prices after negotiated discounts. Employer plans also do notprovide first-dollar coverage, thus promoting prudent use of healthservices including prescription drugs.

III. MMA, the Drug Market, and Retiree PrescriptionDrug Coverage

Drug treatments are becoming standard practice treating chronicconditions. Greater intensity of use of existing drugs and thedevelopment of new and more effective, but also more expensive,drugs have increased the entire population's dependence on drugstherapies. Higher drug development costs and higher demand for drugtreatments have caused drug prices to grow rapidly.

1. Is There "Market Failure" in the Prescription DrugMarketplace?

Data (cited earlier) show that a significant share of retireesalready had access to prescription drugs and drug insurance. About90 percent of seniors reported taking at least 1 prescription drug.Thus, MMA represents an increase in government intervention inprescription drug and drug insurance markets where there was noprior market failure.

Whether the provision of a good or service is financed by thegovernment or through private markets makes a large difference towhether the economy's scarce resources are allocated efficiently.Efficient allocation of resources implies their use in meeting themost important needs first-as signaled by peoples' willingness topay.

It is well known government intervention replaces resourceallocation through competitive forces by allocation through fiat.Because the government does not maximize profits, federal pricesetting and resource allocation decisions are not based on marketsignals efficient resource use. The usual result is a loss ineconomic efficiency. This is what will happen to the prescriptiondrug and drug insurance markets because of MMA.

That's not to say that market outcomes are fully acceptable. Ifthere is considerable inequality of wealth or of needs amongindividuals, market operation will provide goods and services tothe rich, whereas the poor will be unable to make their demandseffective. Because such outcomes may be socially unacceptable,government intervention could be justified - but only at the margin- to assist those in need of subsidies because of economicmisfortunes.

A study based on 2003 data indicates that only 25 percent ofretirees reported forgoing medications due to highcosts.4 The most vulnerablecategories of retirees on account of prescription drug expenses arethose without any drug insurance (50 percent spending $100 or moreon prescription drugs), those in low-income groups (34 percentspending more than $100 per month) and those with three or morechronic conditions (42 percent spending more than $100 permonth).

It is usually difficult to demarcate the appropriate extent ofgovernment intervention on account of wealth inequality. MMAclearly oversteps all reasonable limits, however, because itprovides a broad drug subsidy to all retirees regardlessof their economic status, previous access to prescription drugcoverage, and prescription drug needs.

MMAs generosity will significantly worsen the economy's abilityto allocate resources efficiently - directly by reducing the sizeof the private market, increasing drug prices, imposing larger thannecessary tax burdens on current and future productive citizens,and indirectly by reducing their ability and willingness to saveand invest for the future.

2. Who Will Benefit From MMA?

Dual eligible beneficiaries - those eligible for both Medicaidand Medicare coverage-will now receive drug coverage throughMedicare. The lowest income beneficiaries among them will receivepremium and cost-sharing subsidies as well-except for nominal drugco-payments. Low-income cost-sharing support would be phased outfor families with higher income and assets.

Dual beneficiaries will not lose the value of their coverage.Indeed, their drug coverage is likely to become more generous underMedicare Part D compared to Medicaid-especially as state budgetproblems increase the likelihood of stricter future costcontainment measures under Medicaid. Several states alreadyregulate the number of prescriptions filled per period, the numberof allowable refills, size of dosages, and drug dispensingfrequencies etc. These limitations will be disallowed whendual-eligible beneficiaries are shifted to Medicare Part D-makingtheir prescription drug coverage more valuable.

Many states facing budget pressures are likely to increase theircost-sharing requirements in the future making Medicaid benefitsless valuable. Hence, taxpayer costs of covering dual eligibles'drug insurance may be higher under Medicare Part D because Medicaidsavings "clawed back" by the federal government are likely to besmaller than the actual costs saved.

In addition, MMA will benefit seniors with poor health andconsiderable dependence on costly prescription drugs-includingthose who purchase Medigap plans offering prescription drugcoverage. As mentioned earlier, such plans' premiums, deductibles,and cost-sharing requirements can amount to thousands of dollars.In contrast, Medicare Part D's co-insurance rates are only 5%beyond expenditures exceeding $5,100. For example, under Medigapplan J, retirees must spend $6,250 out of pocket to attain themaximum benefit of $3,000 (implying total annual health carespending of $9,250). In contrast, Medicare Part D's cost-sharingformula would pick-up $5,059 of spending up to $9,250 leaving thebeneficiary better off by $2,058 per year.5

Medicare Part D will also benefit those retirees who choose topurchase Medigap plans without prescription drug coverage becausethey face restrictive choices among available plans. Suchpurchasers constitute the vast majority of Medigap clients.

3. Some Retirees May Pay More During the Long-term

Generally, employer provided retiree health coverage is broad,includes comprehensive drug coverage, requires low co-pay andco-insurance rates, and does not impose separate caps on drugexpenses. In contrast, Medicare Part D premium, deductible, andco-insurance costs will be substantial for those with drug expensesup to $5,100 per year. Hence, during the short-term many retireesmay choose to remain under employer-provided prescription druginsurance.

Over the long-term, however, MMA is likely to induce employersand other private providers to restrict or eliminate retiree drugcoverage. Those covered under such plans would then be forced tosign up for Medicare Part D and could face largerout-of-pocket costs-unless they qualify for additional low-incomesubsidies. This is likely to increase political pressure to shrinkor eliminate the "donut-hole" in the benefit formula. That, inturn, could prompt yet more seniors to drop their private coverageand enroll into Medicare Part D, increasing the program's alreadyhigh overall costs.

Thus, although retirees as a whole would gain considerably onnet from the implementation of MMA, some retirees may become worseoff over the long-term if employers cut costs by dropping retireedrug coverage. That means some of MMA's benefit won't stick withretirees but flow through to employers. Employers' overall gainscould be limited, however, as prescription drug usage expands anddrug prices increase. Those effects would increase the cost ofproviding health care insurance to workers.

IV. MMA's Impact On The Private DrugMarket

The government already subsidizes prescription drug use byMedicaid patients. The federal subsidy is provided through thestates' Medicaid programs. States possess set drug reimbursementrates within but must adhere to federally specified upper-paymentlimits. Drug reimbursement rates to providers, however, must be setto ensure drug provision consistent with the provision of othercomplementary medical services within each state. Rates must alsoensure that comparable service levels available to those eligiblefor Medicaid in all states.

Drug prices and federal and state drug spending under Medicaidhas escalated recently because of increased drug use andavailability of new, effective, but more expensive drugs forreplacing traditional medical treatments. Because prices ofestablished drugs are not allowed to rise by more that the ConsumerPrice Index, manufacturers have set high initial prices for drugsthat are technically "new" but work very much like older versionsalready on the market.

The entry of sizable additional government-subsidized patients(retirees) in the drug market means either that drug manufacturersmust ramp up drug production or substitute sales to Medicare inplace of sales to private purchasers including drug exports.

Some studies have estimated that post-MMA increases in drugdemand would be small. But they assume that those who alreadypurchase prescription drugs will not change their use ofprescription drugs. That assumption defies past experience.

Those who lack coverage today would increase their drug usage asthey obtain insurance against out-of-pocket costs. So also wouldthose with very high dependence on prescription drugs because MMAreduces their cost-sharing expenses. In addition, MMA is likely toreduce state restrictions on drug usage for dual-eligibles-whosedrug costs would now be met through Medicare Part D. And doctorswill hesitate less in prescribing drugs now that their retireepatients have acquired access to a new "third party" payer.

As mentioned in testimony by health care expert, Joe Antos,before you today, drug usage intensity is likely to increase as MMAexpands retiree budgets for prescription drugs. Consequently, thedemand for drugs is likely to increase considerably and will likelycause higher-than-projected program outlays.

If manufacturers can increase drug production withoutsignificant additional costs it may be feasible to accommodate theadditional demand without significant price increases. However, ina competitive marketplace where manufacturers must accept thehighest price offers first, pharmacies and, in turn, the federalgovernment may have to increase offer prices to manufacturers toobtain additional drug supplies for its new Medicare patients. Inthat case, prices charged in the private market must also increaseand the size of the private drug market must become smaller. Thus,theoretically, an increase in the drug market share of governmentpatients would increase drug prices and shrink the private drugmarket.

This theoretical expectation is supported by empirical evidenceon the relationship between the government's share in particulardrug markets and the private market prices of those drugs. A studycovering 200 drugs during 1997 and 2001 found that governmentparticipation in the drug market through Medicaid significantlyincreased drug prices faced by non-government payers.6 An increase in the government's marketshare by 10 percent was found to be associated with a 10 percentincrease in the drug's price. This finding remains true despite tothe addition of several controlling factors such as drugtherapeutic classes, the existence of generics, the number of closesubstitutes, and the time since the drug's first introduction.

Considering Medicaid's market share in the top 200 drugs, thestudy suggests that private-market drug prices would have beenlower by 13.3 percent on average in the absence of Medicaid.Greater intensity of drug use by retirees would, therefore, implyyet higher drug prices. Thus, now that all retirees will beguaranteed federal drug insurance, the higher prices willnegatively impact workers through employer-sponsored orprivately provided plans. As a consequence, employers may seek tocut back on wages, reduce workers' health-care coverage, increasehealth-insurance premiums, or convert full-time jobs to part-timepositions that do not include health benefits.

Another recent study documents that higher health insurancecosts are taking a heavy toll on workers.7 Each 10 percent hike in health insurance costsreduces the likelihood of being employed by 1.6 percent, and cutshours worked by 1 percent. Workers whose health insurance ismaintained are forced to accept smaller wage gains: A 10 percentincrease in premiums is offset by a 2.3 percent decrease inwages.

The prior study also demonstrates that the government's drugrebate program operated for Medicaid-that limits established drugs'price increases to no more than the Consumer Price Index-leads tolarger manufacturer incentives to introduce new drugs with slightperformance enhancements but with initial prices set a much higherlevels to compensate for the federal drug rebate program.

V. MMA's Financial Implications for Workers and FutureGenerations

CMS estimates that Medicare Part D's unfunded obligation (futureoutlays less enrollee premiums and cost-sharing) is zero. However,CMS assumes that Congress will continue to authorize generalrevenue transfers to Medicare Part D as and when needed to bridgethe gap between outlays and enrollee premiums. In presentdiscounted value, total future general-revenue infusions requiredare estimated at $18.2 trillion. That is, Medicare Part D promisesto provide net benefits to current and future generations ofretirees to the tune of $18.2 trillion in excess of the premiumsthey will pay for enrollment into Medicare Part D.

According to CMS, Medicare's Parts A and B combined areestimated to require total financial infusions of almost $50trillion in present value to meet benefit costs under current laws.MMA's enactment has, therefore, increased Medicare's fiscal burdenon current and future taxpayers to $68.1 trillion. The additionalcharge on federal general revenues from the new drug program issignificantly higher than Social Security's future financialshortfall-estimated by Social Security's Trustees to be $11.2trillion.

An $18.2 trillion figure is better understood as a share of thepresent value of GDP from which it must be financed. According toCMS's projections, that share equals 1.9 percent. That is, MMAcommits 1.9 percent of all future GDP to funding seniors' drugcoverage.

Because, the entire GDP is not (and will never be) subject totaxes, it is more instructive to compare MMA's general revenuecharge to the present value of the future income tax base fromwhich all federal general receipts are drawn. Unfortunately, thereis no official estimate of the present value of the income taxbase. However, if future taxable (personal and corporate) incomeaverages about 55 percent of GDP - its current ratio -- MedicarePart D's $18.2 trillion charge on general revenues would equal 3.5percent of the present value of the income tax base.8

Because Medicare Part D is not financed out of a dedicatedrevenue sources, it is impossible to know when the implied fiscalburden -- either higher taxes or federal spending cuts -- would beimposed. It is also impossible to know how this fiscal burden willbe distributed across different income groups and across living andfuture generations.

The calculation of MMA's fiscal burden above involves a criticalassumption: That GDP and the tax base will remain unchanged despitethe imposition of higher taxes or spending cuts. However, highertaxes will adversely impact work incentives and spending cuts maydegrade critical economic infrastructure, both of which wouldadversely affect productivity. Thus, financing the $18.2 trillioncharge on general revenues is likely to require an income tax-rateincrease exceeding 3.5 percentage points because the "feedback"effect of financing MMA benefits through higher taxes on nationaloutput would reduce future national output.

VI. The Impact of MMA on National Saving

The difference between what current generations earn by way ofincome each year and their annual consumption determines how manyresources are saved and invested. The more current generationsconsume, the less is available for investment. The $18.2 trillionestimate of the present value of Part D benefit encompasses theentire future without a time limit. That is, it includes benefitsthat will accrue to future generations.

Unborn generations, obviously, do not consume out of currentincome. The impact of Medicare Part D's net benefit oncurrent consumption depends on the share of it accruing tothose alive today. The Medicare program's Trustees' have estimatedthat federal general revenue infusions into Medicare Part D onaccount of living generations (both workers and retirees) willequal $6.7 trillion. That is, today's retirees and workers (thoseaged 15 and older) can, under MMA, expect to receive from thefederal government $6.7 trillion dollars on net by way ofprescription drug coverage.

As the drug law is implemented and as today's generations'expectations regarding their drug benefits become firmer, they willperceive an improvement in their total wealth position. Theirnatural response to higher perceived wealth would be to increasetheir consumption. As a consequence, national saving woulddecline.

Media Name: jg-figure1.GIF

Evidence from survey data confirms that retirees increase theirconsumption in response to receipt of additional entitlementbenefits.9 Figure 1 showsconsumption indices by age derived from the Consumer ExpenditureSurveys for four periods: 1960-61, 1972-73, 1984-87 and 1987-90. Ineach period, the consumption per capita of all age groups is shownrelative to the consumption of a contemporaneous 30-year-oldperson-whose consumption index is set equal to 1 in each of thefour periods.

The figure shows that consumption per capita of 70-year-olds in1960-61 fell short of 30-year-olds' consumption per capita in thesame period by 29 percent. However, by 1987-90, 70-year-oldsconsumed 18 percent more per capita than 30-year-olds in the sameperiod. More recent data also show the same pattern of increasingconsumption levels by retirees relative to the consumption of theiryounger contemporaries.

One of the most important elements driving the change inrelative consumption patterns by age appears to be the change inthe pattern of resource ownership by age. The expansion of federalbenefits by way of growing Social Security and Medicare outlayshave transferred resources from workers to retirees during the pastfour decades. That process is continuing today with liberalizedSocial Security benefits and the enactment of new entitlementbenefits -- such as Medicare Part D.

Media Name: jg-figure2.GIF

Those transfers have increased retirees' command over resourcesrelative to those available to younger generations. Figure 2 showstotal resource indices by age for the same four periods, wheretotal resources include current net worth per capita and presentvalues per capita of lifetime earnings, pensions, and governmenttransfers from all programs.10

Figure 2 shows that retirees' had more resources at theirdisposal compared to their younger counterparts' resources in1987-90 than did retirees in 1961-62. The passage of MMA willcontinue the trend of increasing retiree resources relative tothose of workers and younger generations. As a result, consumptionby retirees is likely to increase and national saving will continueto decline.

How large would be the impact of MMA's cross-generation resourceredistribution on saving? A Congressional Budget Office studyreviewed academic literature on this question and concluded thatfor every $1 increase in federal unfunded entitlement obligations,national saving declines by between 0 and 50 cents.11

That range indicates the considerable uncertainty surroundingsuch estimates. However, it suggests that the best estimate of theMMA's impact on national saving is negative. Taking the mid-pointof the range of estimates, national saving may be expected tocumulatively decline by $1.7 trillion by the time today's workersachieve retirement age. That is, by 2050 (when today's 15-year-oldswould approach retirement), the national capital stock would erodeby $1.7 trillion and future Americans' income and living standardswould decline correspondingly.

VII. Conclusion

MMA subsidizes retirees' prescription drug expenses but willprobably lead to considerable economic inefficiency. It willimprove prescription drug coverage for low-income seniors who werepreviously covered under Medicaid. It is also likely to benefitlow-income seniors without Medicaid coverage and those with highdrug expenses. It will also provide a substantial subsidy for thoseseniors previously covered against drug expenses under a Medigappolicy. However, out-of-pocket costs of those seniors previouslycovered under an employer-provided prescription drug plan arelikely to increase as employers increase their premiums to soak upthe subsidy or reduce, possibly drop, their coverage completelyleaving retirees to foot MMA's premiums and cost-sharingexpenses.

MMA will increase the share of government-subsidized patients inthe market for prescription drugs. That is likely to shrink theshare of privately purchased drugs via higher drug prices. Theadverse impact will mostly be on workers as the cost of employerprovided health insurance plans increases. That will trigger loweremployment, slower wage growth, reduced hours worked, andconversion of more full-time jobs to part-time jobs.

MMA's long-term costs represent a massive addition to thealready steep fiscal burdens implicit in current Medicare Part Aand Part B policies. This massive cost must eventually be met viatax increases or cuts in other federal spending such as defense,infrastructure, education, social welfare programs, R&D and soon. Meeting future health-care needs as projected under currentpolicies through tax increases alone appears infeasible as highertax burdens erode work incentives, lower employment, reducenational output and the tax base-requiring yet higher tax rates todraw the necessary revenues.

Past experience indicates that redistributing sizable amounts ofresources from workers and future generations toward retirees willerode national saving and investment, and increase our dependenceon foreign savings. Implementing MMA will induce a similarintergenerational redistribution of resources, causing higherconsumption by retirees and reducing national saving. This islikely to further reduce worker productivity and exacerbate theoutput-reducing effects of higher taxes.

Overall, MMA is a bad and shortsighted economic policy. Thisprogram needs to be re-evaluated and recalibrated from its currentfocus on covering all retirees regardless of their health-carecosts and ability to pay for prescription drugs. It should berefocused on those retirees who most need financial support againstprescription drug expenses.

1 I am Jagadeesh Gokhale,Senior Fellow at the Cato Institute. I was a visiting scholar atthe American Enterprise Institute during 2003, consultant for theU.S. Treasury Department during 2002, and a Senior Economic Advisorfor the Federal Reserve Bank of Cleveland since 1990. My researchhas focused on the effects of fiscal policy, especially entitlementprograms, on the economy. The views expressed herein are solely myown and not necessarily those of the Cato Institute.

2 See "Cost SharingPolicies Problematic For Beneficiaries and Program," Testimony byWilliam J. Scanlon before the Subcommittee on Health, Committee onWays and Means, United States House of Representatives, May 9,2001.

3 The remainder wasaccounted for by Medicare premiums, deductibles, co-payments andcost sharing. See "Medicare Cost Sharing Policies Problematic forBeneficiaries," Testimony by William J. Scanlon before theSubcommittee on Health, Committee on Ways and Means, U.S. House ofRepresentatives, May 9, 2001. [GAO-01-713T]. See also [GAO-01-941]

4 See "Prescription DrugCoverage And Seniors: Findings from a 2003 National Survey," byDana Gelb Safran and co-authors, Health Affairs, web-exclusive,April, 2005.

5Ibid, [GAO-01-713T].

6 See "The DistortionaryEffects of Government Procurement: Evidence from MedicaidPrescription Drug Purchasing," by Mark Duggan and Fiona ScottMorton, National Bureau of Economic Research, Working Paper No.10930.

7 See "Labor MarketEffects of Rising Health Insurance Premiums," by Katherine Baickerand Amitabh Chandra, National Bureau of Economic Research, WorkingPaper No. 11160, August, 2005.

8 The assumption that theshare of taxable income in GDP will remain constant at 55 percentis quite optimistic. Labor income - a large component of taxableincome - is expected to decline as a share of GDP as thebaby-boomers leave the work force and enter retirement during thenext two decades.

9 See "Understanding thePostwar Decline in U.S. Saving: A Cohort Analysis," by JagadeeshGokhale, Laurence J. Kotlikoff, and John Sabelhaus, BrookingsPapers on Economic Activity, Winter 1996, pp. 315-407.


11 See "Social Securityand Private Saving: A Review of the Literature" CongressionalBudget Office, July 1998.

Jagadeesh Gokhale

Subcommittee on Federal Financial Management, Government Information, and International Security
Committee on Homeland Security and Government Affairs
United States Senate