Lessons from the New CBO Medicare EstimatesBy Jeff Lemieux
PRELIMINARY 7/28/2003
Last week, the Congressional Budget Office (CBO) released full estimates of the House and Senate Medicare bills. The new CBO report does not fully explain the basis of the new estimates, but it does contain a clear and comprehensive description of the bills, and an equally helpful set of tables illustrating how the numbers add up. The descriptions alone are worth reading, just to get a sense of the enormous scope of the bills and of the Congressional micromanagement involved. The numbers and analysis of the stand-alone drug benefit, scheduled to be implemented in 2006, are equally chilling.
It's a daunting task estimating the costs and impacts of bills the size of the Senate and House Medicare proposals. For example, the Senate bill's Table of Contents alone runs to 8 small-print pages.
And it's certainly not fun for the public servants at CBO to hear complaints from Capitol Hill groupies (like Centrists.Org, for example) that the agency stumbled badly in recent months on Medicare. First CBO failed to provide an economic analysis of the various types of drug benefits proposed in Congress last spring. Then it failed to estimate to costs of alternative proposals, effectively squelching any possible amendments to the House and Senate leadership bills. Finally, the agency fumbled yet again by making an unexplained correction to its hasty estimate of the Senate bill in June, and by not providing much more than an a one-page sketch of the House bill (and that barely prior to the final vote).
But all is not lost. The new CBO estimates provide many key insights. Here are some preliminary lessons (and some new questions) that reflect CBO's new work:
1. Illusory Savings Not So Bad. The "illusory" savings pattern within the fee-for-service section of the Medicare bills is not as bad as we feared. The new CBO estimates contain line-by-line details of the costs of myriad fee-for-service policy changes contained in the House and Senate bills. By contrast, the preliminary estimates from June aggregated those costs and savings to one line. The earlier one-line figures showed new costs of about $6 billion in the early years of the estimate, and savings of $13-24 billion in the later years. That pattern is a red flag. When Congress proposes to spend money now but pushes the savings provisions off into the future, it usually means that the saving provisions aren't really intended to occur. That may still be case for several items in the House and Senate bills, especially regarding certain physician payments. But the new detailed CBO estimates show that the magnitude of the illusory savings is probably more like $5-10 billion, not the full $13-24 billion we had previously reported.
2. Permanent Acceleration of Spending. The new estimates are higher -- meaning the Medicare bills cost more than CBO estimated in June -- but overall pattern is the same: a permanent increase in spending growth. Based on the preliminary CBO estimate of the Senate plan in June, Centrists.Org extrapolated the impact of a continuation of accelerated growth in Medicare spending to 2030, showing that it would amount to about 1 percent of GDP. The new CBO estimates of either the Senate or House plans contain virtually the same permanent increase in Medicare spending growth. The long-term acceleration isn't much -- about one-quarter of one percent a year -- but after 20 years, it implies a very large commitment of federal resources.
3. Late-Enrollment Penalties Are Key. The CBO estimates assume most eligible seniors will voluntarily join the new stand-alone federal drug benefit, in spite of its substantial premium and limited benefits. The estimates explain that this is largely due to late-enrollment penalties. In other words, the new benefit is not really so voluntary after all. Beneficiaries who decline coverage when it is first offered would face considerable premium penalties. Those sorts of penalties wouldn't cause much controversy if the benefits were generous. Seniors would enroll anyway. But because the benefits are not generous in these bills, and the premiums are high, the coercion factor of late-enrollment penalties becomes very important. Not only would seniors hesitating to pay for the new benefit be socked with penalties, but employers would have a strong incentive to force them to enroll.
4. "True Out-Of-Pocket" Rule Hard to Administer. The House and Senate bills don't allow retiree or supplemental coverage to count toward the federal benefit's catastrophic coverage. Only "true" out-of-pocket expenses (and those paid by families and public programs, it turns out) are eligible. This is a terribly wrongheaded provision, but it is included in both the House and Senate bills because it saves money and thereby makes the drug benefit seem more cost effective. However, CBO says that the process of disallowing retiree or supplemental coverage from counting toward the federal benefit's catastrophic coverage will require coordination between multiple federal agencies. That's a red flag of a different sort: possible administrative snafus or invasions of privacy.
5. Employers Will Drop Retiree Coverage. Fears that employers would drop retiree coverage were fully justified. Under the Senate bill, CBO estimates that 37 percent of retirees with employer-based drug coverage would be dropped from those plans. The figure for the House bill is 32 percent. Moreover, employers would essentially force remaining retirees to pay the $35+ a month premium for the new federal coverage (without giving them anything in return); otherwise employers would miss out on new federal subsidies.
6. Employers Will Not Supplement Federal Coverage. CBO's report spells out the problem: "...both [the House and Senate] acts would target greater federal assistance toward those beneficiaries who lack supplemental private drug coverage -- most noticeably, through the requirement that third-party reimbursements not count toward the catastrophic threshold ... As a result, it would provide a clear financial disincentive for employers to supplement the Part D [federal drug] benefit." (page 21)
7. Political Disputes With Retirees Are Likely. CBO's report doesn't talk about future political pressures, but the combination of late-enrollment penalties and incentives for employers to coerce retirees into purchasing the federal plan will cause narrow, but possibly virulent political disputes. Seniors will probably demand a repeal of the late-enrollment penalties once the newspapers explain the cost and limits of the proposed coverage. Seniors with no coverage now will be happy to take the new plan -- it is better than nothing. But seniors with generous retiree coverage stand to lose a lot.
8. Repeal Late-Enrollment Penalties? Get Adverse Selection. If the late-enrollment penalties are repealed, the federal benefit will probably have much reduced enrollment and higher premiums, due to adverse selection. CBO does not discuss this possibility directly, but it is implied. CBO says that late-enrollment penalties are a very important part of their estimate that most seniors would sign up for the federal benefit, even if their drug costs would not otherwise seem to make it worth their while. That essentially means that if the late-enrollment penalties were repealed, fewer seniors with low drug expenses would enroll, those enrolling would have higher drug costs, and the premium would be higher than CBO's estimate of roughly $35 a month. Over time, adverse selection problems -- literally enrollment of sicker-than-average seniors -- could threaten the viability of the stand-alone drug benefit by creating a cycle of rising premiums and reduced benefits. This is not just a remote possibility. Medigap plans with drug benefits have already reached the point where premiums equal the cost of maximum payable benefits. To be fair, Medigap coverage isn't subsidized at all -- because it is subsidized, the federal plan would not suffer such extreme adverse selection. But selection problems even in the presence of subsidies could effectively kill the federal benefit from a political point of view.
9. Will Retirees and Workers Recoup the Employer Subsidies? CBO's economic logic on how employers receiving federal subsidies would pass those savings along to retirees and workers seems a little strained. Economists generally agree that employers facing a competitive labor market make decisions based on the overall compensation workers would receive, not on the split between benefits and wages or salaries. Therefore, employers are assumed to pass any savings in benefit costs back to workers in higher wages. Likewise, employers would respond to an increase in benefit costs by holding down employees' raises. However, the situation with retirees is somewhat more murky. Employers do not face a competitive market for retirees. Retirement expenses for those already retired are essentially a sunk cost, not a labor market decision. When the government begins to subsidize retiree health coverage, is it logical to assume pensions for those retirees would automatically rise? It certainly seems odd to assume that savings on retiree health coverage would pass through to higher wages or benefits for unretired workers. For estimating purposes, it may not matter much if federal subsidies were passed back to retirees and workers (and taxed as income), or if instead the federal subsidies went into increased profits (and were taxed at the corporate profits rate). Either way, the federal subsidies would be partially taxed back into the Treasury. But this section of CBO's new estimates does seem to merit a bit more explanation.
10. No Systematic Savings from Premium Support. CBO has taken a very a programmatic or formulaic view of the House bill's "premium float" system, which would be phased-in beginning in 2010. The CBO discussion of that section of the bill -- probably the best hope for long-run savings in the program from competition (as opposed to benefit cuts) -- talks about assumptions that certain rebates will save money in some areas and cost money in others. But it does not address the key question, which is: Will direct competition between private health plans and the government-run fee-for-service program cause a change in health plan, consumer, or government behavior that leads to greater cost-consciousness and continuing efficiencies over time? Centrists.Org and some other think tanks believe it will. However, CBO was silent on that point.
Links:
Centrists.Org Medicare's Fee-For-Service "Savings" Are Really Costs (originally published July 14, 2003, revised July 28, 2003)
Congressional Budget Office Estimate of S.1 (without the Cantwell Amendment) and H.1 (released July 22, 2003)
New CBO Testimony on the Long-Term Cost of Federal Spending Obligations (July 24, 2003). This statement contains the same CBO long-term budget baseline as last June (which does not include a Medicare drug benefit). It does not discuss tax cuts, interest expenses, or deficits.
CMS Office of the Actuary (Rick Foster) Letter to Rep. Charles Rangel (June 26, 2003).
Senate-passed Medicare bill's Table of Contents alone is a graphic illustration of continued Congressional micromangement of Medicare, not reform.
Roll Call On Medicare and Budget Numbers, CBO is Failing by Jeff Lemieux (July 7, 2003)
Centrists.Org Medicare Drug Benefit Costs 1 Percent of GDP by 2030 (If CBO is Right) (June 25, 2003)
CBO's Blurry Medicare Estimates of the Senate drug bill, with links to CBO's estimates (June 20, 2003)
Congressional Budget Office (one paragraph) estimate of the House drug bill (June 25, 2003)
