The Biggest Health Reform Worry That No One Seems to be Watching; Labor’s Interests and Reform Prospects; and, Lessons from Massachusetts
December 2, 2009
The Biggest Health Reform Worry That No One Seems to be Watching; Labor’s Interests and Reform Prospects; and, Lessons from Massachusetts
- The White House now openly advocates certain features of the Senate health reform bill, including the Medicare Advisory Board, and the ‘Cadillac Plan’ tax on relatively high insurance premiums.
Medicare Advisory Board
- The Advisory Board is tasked with reducing per-beneficiary Medicare cost growth to certain levels – without reducing benefits or raising co-payments –beginning in 2015. We show that, according to legislative language in the Reid (Senate) bill, the Board apparently cannot reduce payments to either hospitals or physicians, who together represent 65% of Medicare payments.
- Thus other sellers to Medicare would be subject to inordinate cost pressures. Presumably, providers of commodity inputs (e.g. distribution services, generic drugs, disposables, etc.) face little pressure, as their margins should be at an efficient competitive equilibrium. This further focuses cost pressures onto the margins of innovators (e.g. pharmaceuticals, biotechnology, medical and surgical devices).
Labor, the Cadillac Plan Tax, and Prospects for Passage of Reform Legislation
- To our minds, the White House’s (and Senate’s) insistence on the Cadillac Plan tax reduces the odds that reform legislation passes. We show that the average union worker’s health benefits are twice as generous as a non-union worker’s, and that the average union worker would fall subject to the Cadillac Plan tax as early as 2014.
- Supporters of HR 3962, the House version of health reform, on average have an almost perfect (98.8%) AFL-CIO score. We struggle to see how pro-labor, pro-choice progressives (i.e. the House majority) vote for a watered down health bill that taxes union membership and may restrict choice.
The Massachusetts Experience
- We examine the Massachusetts health care system before and after its recent (’06 – ’07) reforms. Premiums were well above the national average before reforms, and since reforms have grown more quickly than the national average.
- We find evidence of queuing for primary care, but hospital admissions and lengths of stay are essentially unchanged. The primary driver of higher premium costs is real pricing by hospitals, as opposed to higher demand.
It’s not paranoia if they really are out to get you: Our take on the Medicare Advisory Board
The White House recently advocated several features of the Reid (i.e. Senate) bill. Included among these features is the bill’s Medicare Advisory Board. The Board is the same entity as the Medicare Commission we’ve mentioned previously in our review of the Baucus (Senate Finance) bill. The Board’s purpose is to limit Medicare spending growth to certain targets.
In short, we’re convinced the only feasible means of the Board approaching its savings goal – given the details of the legislative language – is for the Board to reduce the profit margins of manufacturers, particularly manufacturers of innovative products. Moreover, we’re convinced this interpretation departs from the consensus view; accordingly we deal with the Board and its probable effects here in some detail.
From the CBO score of the Reid bill (emphasis added):
“The legislation also would establish an Independent Medicare Advisory Board, which would be required, under certain circumstances, to recommend changes to the Medicare program to limit the rate of growth in that program’s spending. Those recommendations would go into effect automatically unless blocked by subsequent legislative action. For fiscal years 2015 through 2019, such recommendations would be required if the Medicare trustees projected that the program’s spending per beneficiary would grow more rapidly than a measure of inflation (the average of the growth rates of the consumer price index for medical services and the overall index for all urban consumers).”
Beyond 2019, the Board’s recommendations would be required to keep per-beneficiary Medicare spending growth within 1 percent of GDP growth. Exhibit 1 shows the history of real per-beneficiary spending growth, the current projections, the revised projection assuming the Board achieves its goal, and the associated savings in dollars. The last year of the projection (2019) is the first year during which the Board has to fully carry the burden of reducing Medicare cost growth to a longer-term sustainable rate; in this year the Board would have to reduce spending from baseline by 3 percent, or $30B – it’s a big lift.
The Board is specifically enjoined from reducing benefits, raising beneficiaries’ premiums or cost-sharing, or other such steps that directly affect beneficiaries. Less well understood is that, at least from 2015 to 2019, the Board is also precluded from making real reductions in provider (roughly speaking, hospital or physician) payments.
From the Reid bill itself; Sec. 3403, C(2)(a)(iii), pgs. 1004 – 1005:
(iii) In the case of proposals submitted prior to December 31, 2018, the proposal shall not include any recommendation that would reduce payment rates for items and services furnished, prior to December 31, 2019, by providers of services (as defined in section 1861(u))and suppliers (as defined in section 1861(d)) scheduled, pursuant to the amendments made by section 3401 of the Patient Protection and Affordable Care Act, to receive a reduction to the inflationary payment updates of such providers of services and suppliers in excess of a reduction due to productivity in a year in which such recommendations would take effect.
And from the Social Security Act, to which the Reid bill refers, for the purpose of de-coding the bill’s references to 1861 (u) and (d) (emphasis added):
Section 1861 (u): The term “provider of services” means a hospital, critical access hospital, skilled nursing facility, comprehensive outpatient rehabilitation facility, home health agency, hospice program, or, for purposes of section
, a fund.
Section 1861 (d): The term “supplier” means, unless the context otherwise requires, a physician or other practitioner, a facility, or other entity (other than a provider of services) that furnishes items or services under this title.
In plain English, this means that neither hospitals nor physicians will see payment reductions that exceed inflationary updates less productivity gains – which is the same as saying that (on average) hospitals and physicians will not see any real reduction in operating income, since the maximum payment reduction is equal to average productivity gains. For practical purposes, hospitals and physicians are insulated from the Board’s efforts to find savings during its first five years of operation.
Payments to hospitals and physicians represent roughly 45 percent and 20 percent, respectively, of Medicare outlays in any given year – i.e., the Board has to achieve its cost reductions in the first five years even though two-thirds of spending is effectively off-limits. Pharmaceuticals are the next largest Medicare payment category, at roughly 13 percent of outlays by 2015 (Exhibit 2). Notably, Medicare will by this time have become an even more predominant purchaser of total pharmaceuticals output (Exhibit 3).
If the Board cannot reduce payments to hospitals and physicians – and at least from 2015 to 2019 it clearly cannot – then the Board must either reduce the cost of things it purchases directly other than hospital and physician services (Exhibit 4); and/or, it must reduce the input costs that go into hospitals’ and physicians’ costs of producing care. Broadly speaking input costs would include labor, capital, and finished goods. Labor (physicians, nurses, non-professional staff, etc.) are an unlikely target of a Congressionally derived cost reduction; likewise, we see the odds of Congress reducing providers’ cost of capital – or further restricting access to capital (e.g. certificates of need) – as remote. Thus it stands to reason that the Board (whose actions Congress must approve) would focus either on non-hospital, non-physician direct purchases by Medicare (again, Exhibit 4); and/or, would focus on the cost of finished goods used by hospitals and physicians.
This leads to yet another branch in the tree – whether the Board would recommend reducing commodity input costs, or input costs that reflect an innovator’s premium. As regards the former, assuming commodity inputs (e.g. generic drugs, syringes, etc.) are in fact under efficient price / quality pressure, the Board could not expect to produce significant absolute dollar savings by forcing down these prices. In contrast – and setting aside the question of whether this is economically productive — the Board could reasonably expect to be able to reduce the prices of innovative inputs (branded pharmaceuticals, biotechnology products, higher-margin capital equipment, surgical implants, etc.) without having to worry greatly about those inputs no longer being made available. We conclude that innovators’ margins come under considerable pressure from 2015 – 2019 if a reform bill featuring the Medicare Advisory Board passes. Unless the reform effort fails, innovators very much need to either defeat the Medicare Advisory Board – which seems a remote prospect given White House advocacy and the critical need for anything that contributes to a better CBO score – or eliminate the insulation of hospitals and physicians from the Board’s cost recommendations during 2015 – 2019.
If reform legislation were to pass with the Advisory Board provisions intact, this would strengthen our belief that innovators, particularly branded pharmaceuticals, are likely to seek substantial real pricing gains for an extended period of time leading up to the actual implementation of reforms (2013/’14).
Labor’s Interests and the Prospects for Health Reform
Our belief that reform efforts are more likely to fail than succeed is anchored to a number of issues, prominent among them are abortion; immigration; shifting voter opinion, the conservative nature of many Democrats’ constituents and the proximity of mid-term elections; and, the conflicting pay-fors in the House (taxes on the wealthy) and Senate (taxes on relatively more expensive health premiums) bills. Here we focus on the issue of pay-fors.
The White House recently offered support for specific provisions of the Senate bill, including the Senate tax on relatively expensive health plans. This begins to cement the likelihood that the provision would be found in any bill that passes the Senate, and thus in any bill that might emerge from conference for a final vote in the House.
It’s worth considering the relative politics of labor in the House and the Senate, and the relative importance of the so-called Cadillac Plan tax to labor. Labor has rather dramatic influence in the House, and this influence correlates with yea votes on H.R. 3962, the House’s version of health reform legislation (Exhibit 5). Senate Democrats carry high AFL-CIO scores as well — with the notable exception of Majority Leader Reid – who has considerable power in shaping whatever legislation may emerge from the Senate (Exhibit 6).
More to the point, labor has conceded growth in cash wages in favor of more generous health benefits, and fears that the Cadillac Plan tax would erode these gains – and they’re right. Union workers’ health care benefits are more than twice as generous as non-union labor’s (Exhibit 7), so much so that the average union worker would be subject to the Cadillac Plan tax for (single-coverage) premiums by 2014 – the same year the health reform bill’s major coverage provisions would become effective.
It follows that we expect labor to care dearly about the presence or absence of a Cadillac Plan tax in any final legislation. Yet as we’ve said, given White House support for the tax, and Reid’s accommodative stance, we would expect the provision to be in any final conference report that might emerge.
This places pro-labor House members – and in particular pro-labor / pro-choice progressives – in the awkward position of taxing union members to pay for a health reform bill that lacks their preferred feature (a robust public option) and places restrictions on choice.
The Massachusetts Experience
In April of 2006, Massachusetts passed a series of health reforms, many of which are proposed as part of the current Federal reform effort. In particular, Massachusetts has an individual mandate that went into effect in July of 2007, provides subsidies for the purchase of health insurance to individuals and families with incomes below 300% of the Federal Poverty Level (pending Federal legislation provides subsidies to 400% of FPL), requires most employers to either provide insurance or pay penalties, and creates health insurance exchanges within which individuals and small groups can shop for coverage.
We examined several characteristics of the Massachusetts health insurance market before and after enactment of reforms. Massachusetts came into its most recent reform effort with relatively (as compared to the nation) low rates of un-insured, fractionally lower reliance on employers as a source of health insurance, and substantially greater reliance on public forms of insurance. Following enactment of reforms, Massachusetts has very few uninsured residents, fractionally greater reliance on employer-sponsored insurance than the nation as a whole; and, has further expanded its reliance on public forms of health insurance (Exhibit 8).
Despite very low relative rates of un-insurance, as is the case with the broader nation Massachusetts’ uninsured populations tend to be younger, non-white, and have limited education (Exhibit 9).
Massachusetts entered the ‘06/’07 reforms with average premiums for both single and family coverage that were 6% – 10% above the national average; since enactment of reforms, premium growth in Massachusetts has exceeded the national rate by a substantial margin, particularly for family coverage (Exhibit 10).
Critically, Massachusetts’ insurers were and still are predominantly not-for-profit; on a premium-weighted basis, not-for-profit insurers account for over 95% of the state’s health insurance coverage. Accordingly, we recognize the limits on our ability to compare Massachusetts insurers’ cost and profitability dynamics to the prospects of for-profit insurers under Federal reforms.
Notwithstanding these limits, we find underlying cost trends that we believe speak to health cost prospects at the Federal level, assuming enactment of comparable reforms. In short, Massachusetts’ greater rate of premium growth since reform ties entirely to greater rates of underlying health cost growth. Massachusetts insurers’ medical loss ratios (MLRs) were higher than the national average on the eve of reforms, and have since grown by nearly 2 percent (Exhibit 11). Administrative expense ratios (AERs) were below the national average on the eve of reforms, and have since fallen by roughly two-thirds of a percent (Exhibit 11). As a result, Massachusetts insurers’ premium – weighted profits fell by 150 bp following the roll-out of reforms, from 2.6% in 2005, to 1.1% in 2008 (Exhibit 11).
These underlying cost patterns are consistent with heightened pricing power at the provider level. We have no reason to suspect that inputs to care which are standardized on a national level – e.g. drugs, devices, supplies – have prices that increase more quickly in Massachusetts than in the rest of the nation. Rather, we would initially suspect that Massachusetts – specific providers (particularly hospitals and physicians) either find themselves over-run with marginal demand, able to raise prices more quickly because of heightened (insurance-related) consumer inelasticity, or both.
Waiting times for appointments have risen, particularly appointments with primary care providers (Exhibit 12); though at least for the moment we do not know whether physicians have succeeded in raising prices since reform. Conversely — and far more importantly as a percentage of total costs — hospitals have seen rather static admissions and lengths of stay, but fairly substantial pricing growth, even though average per- patient hospital charges in Massachusetts were already 26 percent above the national average in 2004. The average charges for an inpatient day grew at a compounded rate of more than 7 percent between ’06 and ’08 (Exhibit 13). Notably, Massachusetts’ total hospital capacity on a population adjusted basis is at roughly the national average, though we would note that Massachusetts’ ‘effective’ capacity is arguably below the national average, in that rates of insurance – and arguably thus per-capita bed demand – is on average much lower outside of Massachusetts. And, Massachusetts’ utilization of its hospital capacity is higher (Exhibit 14). This combination of average to below average capacity and above average utilization clearly favors hospital pricing power.
On top of this, we note that hospitals’ share of total Massachusetts’ admissions correlates with average charges per inpatient day (Exhibit 15) – i.e., as is true across the nation, larger hospitals charge more—and that the market power of the largest hospitals appears substantial, as just 20% of the hospitals generate more than half of the state’s hospital admissions (Exhibit 16). And, we’d note that hospitals’ margins have fared well during the post-reform period – in contrast to insurers’ margins (Exhibit 17). This clearly suggests hospitals’ higher charges are in fact a true reflection of pricing power, rather than a response to higher input costs.
More speculatively, we would note that the sudden burst of enrollment during the recent reforms marked Massachusetts insurers’ last best chance to make substantial gains in membership. Presumably insurers would prioritize enrollment over immediate measures of profitability; moreover, in prioritizing enrollment, insurers certainly might want to ensure that the best (or at least most preferred) hospitals were available in their networks. Accordingly, it stands to reason that on top of Massachusetts’ hospitals long-standing sources of pricing power, the enrollment period marked a time during which brand-name hospitals had even more market (and pricing power) than usual. The patterns of enrollment and comparative (hospital v. insurer) margins from ’06 to ’08 is consistent with this interpretation – note that flagship (i.e. teaching) hospital margins were expanding, and insurers’ margins contracting, as large numbers of enrollees entered the system during the roll-out of reforms (again, Exhibit 17).
Massachusetts’ near-total lack of for-profit insurers means the example offers little direct evidence regarding the large capitalization insurers’ prospects in a subsidized, exchange-oriented, mandatory-issue market. Nevertheless, we believe the example does speak to the broader structural realities of healthcare systems and prospects for their reform. We frequently assert that health costs grow faster than value for the simple reason that they can – nothing effectively links price and value. Massachusetts reforms expanded coverage, but have done little to increase cost / quality pressure on either insurers or providers (particularly hospitals). Temporarily setting aside the moral imperative of expanding coverage and focusing solely on the economics, absent tighter price / value or cost / quality links, we think it’s unreasonable to expect anything other than the incremental exercise of real pricing power when expanded coverage adds to already high levels of inelasticity. More to the point, absent structural reforms that link price and value, coverage expansion tends to raise the health system’s inflationary gearing.
Quick Updates: Flus and Blues
Recall our argument that much of the higher-than-expected claims costs in 2009 tie to employees’ tendency to over-consume care during periods of substantial job loss. In that job losses appear to be easing, we continue to believe that insurers’ medical loss ratios (MLRs) in 2010 will be much improved relative to 2009 – and in particular that the improvement exceeds consensus expectations.
An obvious wildcard to recent MLRs is the flu effect; here we simply update our view of the current season’s progress, and by extension the prospects for relatively greater or lesser flu effects on 2010 MLRs.
The apparent early peak in 2009 flu activity is now even more likely to be a ‘real’ peak with the benefit of two additional weeks’ data (Exhibit 18). We argue that 2009 saw a double season, an early end to the fall/winter season, and is likely to be more severe than the 2010 season. Clearly the trend is for the fall/winter season to end early, thus our belief that 1Q/10 flu effects will be modest. As regards relative severity of 2009 v. 2010, we note the late availability of vaccine in 2009 as compared to the peak of the epidemic (Exhibit 19). As the H1N1 season peaked, we had shipped roughly a tenth of the usual vaccine supply; even now, well past peak, cumulative shipments are just approaching half the usual season total of roughly 100 – 115M units. We expect greater vaccine coverage for H1N1 in 2010 v. 2009, thus our belief that 2010 severity is likely to be reduced relative to 2009.
We argued recently that the public option, as written in both the House and Senate bills, was almost indistinguishable from the typical non-profit Blue Cross / Blue Shield Plan – it has to pay back its start-up capital, negotiate prices with providers, and charge sufficient premiums to earn a reserve. We showed that for-profit commercial insurers could profitably co-exist with non-profit Blues, the point being that the public option, as embodied in current proposals, does not strike us as a dramatic threat to the commercial insurers.
Since then, we’ve been curious as to who enrolls in Blues plans – and, in particular, whether there is any difference between the typical enrollee in a non-profit health insurance plan as compared to a for-profit insurance plan. Where we were going with this was an attempt to further distinguish whether, and to what degree, current enrollees in the employer-sponsored plans of large capitalization insurers might or might not tend to shift toward a public option; or, conversely, whether the public option would simply pull enrollees from existing non-profits.
There is surprisingly little easily comparable information on enrollees in non-profit and for-profit plans. The one exception is somewhat dated (1997 – 1998 data), though we’d doubt there’s any reason non-profit v. for-profit enrollment patterns would have changed dramatically since then. The upshot is that there appears to be little if any difference between for-profit and non-profit beneficiaries (Exhibit 20). Accordingly, while we continue to believe that commercial insurers could co-exist with a non-profit public plan, we cannot argue that the public plan would pull its enrollees preferentially from existing non-profit underwriters.
- If you read the CBO score, specifically the sentences following our direct quote, you’ll see that CBO claims the post 2019 goal is the rate of growth in national health expenditures. This contrasts with the legislative language on pages 1013 – 1016 of the Reid bill. We assume the legislative language is correct.
- Note that despite carrying lower AERs as a percent of premiums, absolute administrative expenses per member grew, and are fractionally higher in Massachusetts than in the rest of the nation.
- “Controlling Health Care Spending in Massachusetts: An Analysis of Options” RAND Health, August 2009