|
 |
 |
 |
 |
 |
 |
 |
 |
“Dynamic Scoring” in Health Care
Jeff Lemieux Earlier this year, the Congressional Budget Office (CBO) created “dynamic” estimates of tax cut proposals, which factored in the long-term economic effects of lower tax rates and higher federal deficits. Although it's not practical to perform additional economic analysis of all proposals, the extra analysis of tax cuts was helpful to Congress. Now, CBO should try to point out the long-term or other analytic considerations of Congressional health care proposals. For example, Congressional conferees would benefit greatly from additional analysis of several facets of the Medicare bills recently passed in the House and Senate. Earlier this year, Democrats howled when Republicans insisted on so-called dynamic estimates of tax cut proposals. Republican leaders in Congress argued that tax cuts would put the economy on a permanently higher trend. They claimed the extra growth would create additional federal tax inflows, partially offsetting the revenue lost from the tax cuts themselves. Democrats feared the Republican leadership was trying to force CBO and the Joint Committee on Taxation to produce unreasonably optimistic figures -- figures that would allow Congress to squeeze larger tax cuts into a given budget. Democrats’ fears proved unwarranted and Republicans’ hopes were dashed when CBO’s estimates factored in not only the positive effects of long-term tax reductions, but also the negative economic effects of correspondingly higher long-term federal deficits. In the end, dynamic analysis was a mixed bag for tax cut proponents: they would be forced to cut federal spending to get the economic result they wanted. For the analytic community, however, dynamic scoring was a success. Although there is always potential for abuse or gimmicks when the estimates get more complicated, Congress needs more than just a number or “score” on its complex proposals. Legislators also need to know how the proposals will work, and whether or not they would have important longer-term implications. This doesn’t mean all official cost estimates should include dynamic effects or additional technical analysis. That would be impractical for small policy changes. And the long-term impacts of some Congressional proposals would be too uncertain to estimate. However, when it’s impossible or inappropriate to add longer-term dynamic effects to official cost estimates, CBO should try to at least give Congress an idea of the possibilities, especially on complicated health care proposals. Avoiding the Short-Run Trap
A decade ago, when Congress was considering large-scale proposals to cover the uninsured, health economists developed a theory of “managed competition.” The theory said people should pick the health coverage they want from a menu of alternative health plans. All the plans would be subsidized (by employers or the government) at the same risk-adjusted level. That way, people could save money by choosing more efficient plans, but they would have to pay more for less efficient forms of health insurance. The result would be a gradual switch toward more efficient health plans, and reduced growth in health spending. Managed competition theory was very valuable in the policy debate. It sparked the consideration of socially run purchasing pools, which were part of the However, the economists designing managed competition systems fell into a timing trap. They correctly figured that the system would be most efficient in any given year if all plans had the same benefits. That way, people could make direct comparisons of prices across plans. But they didn’t consider what would happen in the long run. The menu system might indeed be most efficient in each year if benefits were uniform, but how would benefits change over time? Freezing a particular benefit design in place, or subjecting it to a bureaucratic decisionmaking process would be much less efficient over time than allowing health plans to experiment with alternative benefits to meet particular needs, and then subjecting those experiments to the market test: will people choose them, and like them? Medicare provides an analogy. Over time, private health plans have added drug benefits because that is what people wanted and needed. Drug benefits were both medically appropriate and efficient. Plans have also developed multi-tiered co-payment structures and preferred drug lists to try to hold down unnecessary drug spending. By contrast, Medicare’s benefits are determined by Congress; as a result, Medicare’s benefits haven’t adapted as quickly. Long-Term and Technical Questions About the Medicare Proposals
The most important long-term questions surrounding the Medicare bills are: (1) Would the federal employees-style premium system envisioned in the House bill lead to lower cost growth and more rapid benefit evolution over time? Would the government-run fee-for-service program flourish or flounder in a competitive system? (2) What formulas or conditions would entice comprehensive Preferred Provider Organizations (PPOs) into the Medicare program? Would those plans behave independently, or would they evolve into a new form of Medicare contractors, with benefits and payment mechanisms mostly directed by the government? There are also several analytic or technical issues that warrant further analysis: (3) The stand-alone drug benefit scheduled for implementation in 2006 is technically “voluntary," but there are penalties for delayed enrollment. If those penalties were relaxed, would the stand-alone benefit suffer “adverse selection,” with a disproportionate number of seniors with high drug costs enrolling in the program? Would private health plans continue to provide the benefit? (4) How would employers react to fact that their retiree drug benefits don’t “count” toward the catastrophic part of the stand-alone Medicare drug coverage? Would they drop retiree benefits suddenly, or slowly over time? (5) Will risk-adjustment methods be sufficient to compensate either private health plans or Medicare’s government-run fee-for-service program if they suffered from adverse selection; that is, if their enrollment consisted of less healthy, higher risk seniors? On question number 1, we believe managed competition theory is essentially correct (assuming that health plans are allowed to add or change benefits over the years). We believe that using formulas for Medicare premiums patterned after those used in the Federal Employees Health Benefits (FEHB) program would probably lead to greater price sensitivity among enrollees, and would spur both private health plans and the government-run fee-for-service program to find efficiencies and improve benefits over time. The fee-for-service program would almost certainly need additional administrative flexibility to compete in that manner. However, putting the program into direct competition with private plans is a political precondition for Congress to loosen the reins and switch from direct micromangement to an oversight role. However, managed competition theory is not certain. Medicare’s Office of the Actuary has implied that private plans would be considerably more efficient than the fee-for-service program. That would produce bargains for seniors where private plans offer coverage, but would raise the premium for fee-for-service enrollees. However, the actuaries do not seem to think there will be a longer-run improvement in Medicare efficiency or slower growth rates in Medicare spending. It would be interesting to see what CBO thinks. On question 2, we predict that technical rules limiting PPO plans to three per (large) region could possibly spur aggressive and cost-saving initial bids from large national health plans. But it would limit competition over time, and could lead to a lack of innovation and long-run cost savings. CBO would help Congress design the program if it weighed in on those sorts of short-run vs. long-run dynamics. Question 3 seems like a no-brainer. The rules charging penalties to late enrollees in the stand-alone Medicare benefit appear destined for political turmoil and repeal. Seniors just won’t know whether or not their employers will drop retiree coverage, and will feel coerced into joining a new program -- and paying an additional premium -- before they even know if they’ll need the coverage. However, relaxing the late enrollment penalties could lead to severe workability problems with the stand-alone drug benefit. Without those penalties, economic logic predicts adverse selection: that seniors with higher-than-average drug expenses would enroll in the stand-alone benefit, and seniors with low drug expenses would not. Adverse selection would drive up the premium and threaten the viability of the benefit over a period of time. If it agrees adverse selection is a potential problem, CBO should explain the issue in more detail. Question 4 also seems very straightforward. We believe that employers are much more likely to drop retiree drug coverage than to maintain it or “wrap around” the government coverage. Of course, employers are already reducing retiree coverage, and will continue to do so regardless. But it would be helpful to predict how rapidly employers will likely drop drug coverage for their retirees, and assess the degree to which the prohibition of retiree benefits from counting toward the Medicare drug benefit's catastrophic coverage would accelerate the process. Finally, future risk-adjustment methods will probably be adequate to protect health plans from adverse selection. Private plans still participating in Medicare (many have dropped out of the program) seem to have an increasingly sick and risky set of enrollees. Nevertheless, those plans tend to distrust Medicare’s risk-adjustment mechanisms, fearing that any risk-adjuster will somehow reduce their payments. (Whether it is true or not, health analysts and Medicare administrators have long argued that private plans had healthier-than-average enrollees, and that their payments should be reduced.) Meanwhile, advocates for the fee-for-service program consistently worry that the government-run plan would garner the sickest enrollees in a competitive system. In any event, a thorough discussion of the risk-adjustment issue would greatly assist policymakers. |
|
|
|
Centrists.Org is a non-partisan, non-profit, organization formed under section 501(c)(3) of the tax code, and dedicated to public education on vital public policy matters. Contributions to Centrists.Org are tax deductible. Centrists.Org |
