There has been growing interest in reforming Medicare to make it operate in a similar way to the Federal Employees Health Benefits Program (FEHBP), the health plan enjoyed by Members of Congress and over nine million other federal workers and retirees and their families. One of the arguments made for doing so is that the FEHBP performs more effectively and efficiently.

A recent analysis by Mark Merlis, published by the Kaiser Family Foundation, disputes the claim that the FEHBP would be an improvement on Medicare.[1] Merlis's paper comes to the conclusion that although the FEHBP has many attractions, it falls short of Medicare in a number of ways. Among the most important:

• FEHBP's per capita spending, says Merlis, has risen faster than Medicare.
• FEHBP's administrative costs, he says, are much higher than those experienced by Medicare, suggesting that FEHBP is less efficient.

Merlis concludes that "the FEHBP experience highlights the difficulties in developing a competitive system that operates efficiently and equitably for a nationwide program."

The Merlis study, however, indicates the pitfalls associated with making comparisons without painting the full data picture. Unfortunately, his approach to the data presents a highly misleading comparison, and does not support his conclusions. 

Consider two measures that form the heart of his comparison:

Spending Growth in Medicare and FEHBP 

Comparing the experience of two programs depends a great deal on which time periods are compared, and how long a period is used for the comparison. Generally short periods, in which there may be a number of temporary influences on one of the programs, can give a misleading snapshot of the long-run picture. Moreover the selection of one particular period may give a very different impression from another period.

In his recent Kaiser paper, for example, Merlis confines his cost comparison to the years 1996-2002 - just seven years for programs that have been in existence for several decades. During the period he selects, FEHBP spending per enrollee increased at an average annual rate of 5.3 per cent; meanwhile spending per enrollee in Medicare rose at an average annual rate of 4.2 per cent. While Merlis does discuss some of the differences in the programs that would influence costs, such as the absence of drug coverage in Medicare, he nevertheless concludes that the record of Medicare is substantially better than the FEHBP.

One problem with this conclusion is that if one analyzes a different period one can reach the opposite conclusion. Merlis himself did this in a 1999 study for Kaiser also comparing the two programs.[2] In that study he looked at the years 1988-1997. Over that period, FEHBP's cost per enrollee rose 7.1 per cent, while Medicare's rose 8.1 per cent. Moreover, Merlis drew attention to the fact that from 1992-1997, FEHBP cost rose only an average of 3.4 per cent, compared with 8 per cent for Medicare, leading him to note in that study that "per capita spending did rise somewhat more rapidly under Medicare than under the FEHBP between 1988 and 1997, and much more rapidly between 1992 and 1997."[3]

Two Different Periods - Two Very Different Conclusions

1999 Merlis study: Conclusion - FEHBP Showed Better Cost Control

2003 Merlis study: Conclusion - Medicare Better Cost Control

Using the figures for the entire period covered in Merlis' two studies, 1988-2002 (a 15 year period), the difference between the FEHBP and Medicare turns out to be very small (and annual average of 6.9 per cent for FEHBP and 6.5 per cent for Medicare). 

But there are reasons why even that difference may be exaggerated, or FEHBP's cost growth might even turn out to be lower during that 15-year period. For example:

• Merlis notes that factoring in such things as the drugs available in private plans (drug costs have tended to rise rapidly in recent years) and the health and nursing home costs are services in Medicare, would narrow the cumulative difference between the two programs for comparable benefits.
• It is important not only to try to adjust for the range of benefits but also for trends in the generosity of benefits within the programs. American Enterprise Institute scholar Joe Antos, a former CBO health analyst, drew attention to this issue in a recent Heritage WebMemo.[4] During the 1980's and 1990's, noted Antos, the value of Medicare benefits did not keep pace with benefits in private plans such as those in the FEHBP. Antos and his co-author, Al Goyburu, studied the generosity of similar benefits in the two programs. Recently there have been growing concerns that Medicare's tighter price controls are leading to hospitals and doctors reducing their services, and many doctors have dropped out of the program. These changing features of the health benefits mean the value of similar benefits in private plans was rising more rapidly than in Medicare, making simple cost comparison misleading as a measurement of value for money. 

Administrative Costs

Misleading conclusions can also result from with overlooking important factors when comparing administrative costs. Critics of Medicare reform often argue that administrative costs of Medicare (defined as administrative expenses compared with spending on enrollees) averages roughly 2 per cent, and that this is well below the administrative costs of the FEHBP or private plans - suggesting that Medicare is more efficient. Merlis makes this argument, noting that administrative costs compared with benefits averages 7-10 per cent in preferred provider organizations (PPO's) and 15 per cent or more in health maintenance organizations (HMO's). On the face of it, therefore, Medicare management appears to be far more efficient, with private plans spending between 3 and 7 times as much on administrative costs for the same spending on beneficiaries. 

However, this simple statistic is quite misleading. For one thing, as Merlis notes in the study, a program serving predominantly retirees will necessarily spend more on benefits per enrollee than is typical in a plan serving primarily working-age people (such as FEHBP). Hence, there will be a larger denominator in the ratio of administrative costs to benefits for Medicare than for FEHBP just because of the age of the enrollees. Merlis does draw attention to this, and notes in his text that if one were to examine the cost of administrative cost per enrollee, rather than comparing it with the expenditures for each enrollee, it would be a better indicator of comparative administrative costs. He estimates that by doing this the difference between private plans and Medicare would be more like 2 to 1, rather than as much as 7 to 1.

But even this does not provide a real comparison of efficiency. There has been concern for many years that Medicare may actually suffer from dangerously insufficient spending on management and administration. Members of the National Academy of Social Insurances panel on Medicare Governance and Management, for example, cited an inadequate investment in Medicare administration as partly responsible for poor control of costs, including excessive waste and fraud by providers. An overhaul of management and increased funding for administration featured prominently in the panel's recommendations.[5] In addition, the General Accounting Office (GAO) issued a report in 2001criticizing the poor performance of Medicare's management in providing information to consumers and other services related to effective management.[6] The GAO's survey found Medicare's line managers to be more critical of their management structure and capacity than virtually any other agency. Thus, it appears that Medicare management is inadequately funded to carry out the tasks needed to provide quality service and careful control of expenditures. 

Paradoxically, inadequate financing of management falsely gives the impression of greater efficiency. This is because wasteful unnecessary expenditures actually serve to reduce the arithmetic ratio of administration to spending. In a similar vein, a private manufacturing company that spent nothing on inventory control, quality measures, or checking accounts receivable could boast it was clearly efficient because its ratio of overhead costs to sales was very low. But of course it would be highly inefficient. Medicare is probably also below the threshold of administrative spending needed to run an efficient program.

Thus, the comparison of administrative cost to spending between FEHBP and Medicare may well indicate that Medicare performs far less well than the FEHBP. 

Conclusion

The Merlis study thus does not make a convincing case that the FEHBP "highlights the difficulties in developing a competitive system."[7] Not only that, but the study fails to draw attention to the enormous weaknesses in the Medicare program. Among the problems:

1. There is a growing concern that price controls, payment formulas and congressional restrictions on Medicare's management spending are leading to reductions in the quality and availability of services in the program, particularly since the passage of the Balanced Budget Act of 1997. Worries about doctors leaving the program, and hospitals unable to maintain adequate services have led to calls in Congress for "givebacks." The more Congress responds to these concerns, the greater the additional future spending in Medicare and the greater the erosion of the apparent cost advantage of the program.
2. Medicare lacks the most basic benefits available in FEHBP and private-sector plans, such as a drug benefit and catastrophic protection. In the FEHBP, these and other benefits have been gradually added to the program over many years without congressional legislation because private plans competing in the FEHBP market must over time respond to consumer demands for modernized services or lose business. Medicare, on the other hand, sets benefits using the legislative process, and this is notoriously subject to the weaknesses of a political system in making rational changes in benefits. The lack of the basic benefits, in other words, is a direct and inherent characteristic of the design of the Medicare program and consequent politicization of Medicare benefits.
3. The chronic under-funding of management noted earlier is also a direct result of the design and political control of the program. Spending money on "overhead" is never popular with politicians when the alternative is to spend it directly on benefits - even if these cannot be adequately managed. The history of Medicare is one of congressional demands and micromanagement of the program, yet inadequate financing of the management obligations of the agency running the program.

Comparing Medicare and FEHBP from this wider prospective, therefore, highlights the difficulties of operating a tightly government-run system like Medicare. This is why there is growing interest in learning the true lessons of the FEHBP and using them to reform the operation of Medicare.

Stuart Butler is Vice President for Domestic and Economic Policy Studies at The Heritage Foundation.