The Medicaid Expansion & Why Hospital Pricing Peaks in 2013

Print Friendly
Share on LinkedIn0Tweet about this on Twitter0Share on Facebook0

Richard Evans / Scott Hinds / Ryan Baum


203.901.1631 /.1632 / .1627 richard@ / hinds@ /


March 4, 2013

The Medicaid Expansion & Why Hospital Pricing Peaks in 2013

  • More Republican states are expanding Medicaid fully (to 138 pct of the Federal Poverty Level, or FPL) than we projected. Persons between 101 and 138FPL are now more likely to be in Medicaid than in exchange-based commercial plans
  • Medicaid pricing to hospitals currently is 37 pct lower than commercial pricing; accordingly full Medicaid expansion brings a less attractive payor mix than would have been the case had the 101 to 138 FPL beneficiaries gone to commercial coverage. Knock-on effects are likely to lower Medicaid hospital rates further:
    1. The larger Medicaid is as a pct of a state’s budget, the lower that state’s hospital payment rates tend to be. The 2014 Medicaid expansion, simply by (eventually) increasing the pct of state budgets going to Medicaid, should lower Medicaid hospital pricing by +/- 2.4 pct
    2. The Affordable Care Act (ACA) raises physicians’ Medicaid payment rates to Medicare payment levels for 2013 and 2014, thus any Medicaid savings from rate changes must come from hospital payment rates in this period
    3. The Administration recently has signaled its willingness to support states that have been challenged by hospitals for relatively large Medicaid rate cuts
  • On a weighted average basis across all payors, we believe hospitals’ pricing will be about 2.3 pct lower as a result of the full expansion of Medicaid to 138FPL, as compared to a partial expansion to 100FPL
  • Hospitals’ commercial pricing power arguably is as high as it has ever been as the kick-off of the exchanges approaches. However following an initial enrollment surge, we see commercial pricing weakening as exchange-based underwriters work to narrow provider networks, and as commercial beneficiaries (including employer-sponsored) shift to less generous plans (whose higher out-of-pocket liabilities are difficult for hospitals to collect, reducing net pricing)
  • We see further upside to Hospital share prices as volumes increase, whether on improving employment, the roll-out of the exchanges, or some combination of the two. However because we believe Hospital pricing will peak this year, we recommend rotating into Non-Rx Consumables (e.g. CFN, OMI) as 2014 approaches

It’s increasingly clear that many of the Republican states – particularly those with larger Medicaid populations – are likely to expand Medicaid eligibility fully to 138FPL in 2014, despite our expectation these states would limit the expansion to 100FPL (see Appendix 1)

There remain notable holdouts (mainly Texas), and even uncertainties as to whether Florida’s legislature will pass Governor Scott’s expansion proposal; nevertheless we now believe the Republican states will put substantially more 101FPL to 138FPL beneficiaries in Medicaid than we originally expected[1]

Under current policy states would retain the option to reduce eligibility back to 100FPL (or lower, but 100FPL is a natural stopping point) in 2016; however it’s unlikely policy will remain static. As evidence, states have been more or less trapped at pre-Lehman levels of Medicaid eligibility since the start of the recession, first by the American Reinvestment and Recovery Act (ARRA), then by the Affordable Care Act (ACA). We see every reason to believe federal policy would again seek to keep expansion states from lowering Medicaid eligibility from 138FPL in and beyond 2016

This forces the question of what these developments mean; our immediate focus is what this means for hospitals. Under our previous expectation, beneficiaries between 101 and 138FPL would buy coverage on the exchanges. Because exchange-based premium and co-pay subsidies are so generous for persons in this income range, we believed enrollment rates would be quite high, and out-of-pocket costs would have been manageable. The longer-term advantage to hospitals in this scenario would have been that beneficiaries in the 101-138FPL income range would be paid for at  commercial reimbursement levels, which are substantially higher than Medicaid (Exhibit 1)

Because the 101-138FPL beneficiaries now are more likely to be enrolled in Medicaid, weighted average payment levels to hospitals naturally will be less than if these beneficiaries had purchased commercial plans on the exchanges

Beyond this obvious payor mix-shift, we believe the Medicaid expansion will result in Medicaid hospital payment rates falling even further. States retain substantial latitude in setting Medicaid payment rates, and Medicaid payments to hospitals vary widely across states. Medicaid hospital payment rates appear to vary as a function of the percentage of states’ budgets allocated to Medicaid – states spending a higher percentage of their budgets on Medicaid tend to pay hospitals less (Exhibit 2; circle sizes correspond to the sizes of state Medicaid programs)


The expansion of Medicaid to 138FPL inevitably increases the percentages of state budgets spent on Medicaid. Using the eventual federal matching rate (90 pct) for expansion beneficiaries, we calculated how much Medicaid spending would increase as a percentage of states’ budgets, and assuming the relationship in Exhibit 2 holds, what effect this might have on average national Medicaid reimbursement levels for hospitals. The result is that we would expect the expansion to lower average Medicaid hospital payment rates by about 2.4 pct (Exhibit 3)

Pulling all of these observations together, we modified the existing National Health Expenditure (NHE) projections to reflect two distinct scenarios: 1) a Medicaid expansion by all states to 138FPL and a corresponding 2.4 pct fall in average Medicaid payment rates; and 2) an expansion by all states to just 100FPL, migration of (most) 101-138FPL beneficiaries to exchange-based commercial plans, and no change in Medicaid payment rates. The results are in Exhibit 4; we estimate hospitals’ aggregate payment to cost ratio would fall to a bit less than 1.10 in scenario 1 (full expansion to 138FPL), as compared to roughly 1.13 in scenario 2 (partial expansion to 100FPL). The precise value for the estimated difference is 2.3 pct — admittedly not a huge difference, but neither is it a trivial difference given the narrow margins and fixed costs associated with hospitals

The fall in Medicaid hospital payment rates actually may be larger, for a combination of three reasons:

  1. Per the ACA physicians’ Medicaid payment rates rise in 2013 and 2014 to the Medicare levels, which means spending on physician reimbursement will rise substantially. This places physician payment rates off-limits as a source of savings – and in fact makes them a driver of higher costs – thus hospital payment rates fall under marginally greater pressure
  1. The Administration is signaling[2] that states are largely free to set Medicaid payment levels as they choose, the main constraint on payment rates being that they are sufficiently high to assure beneficiaries have access to care. In the case of hospitals, this is very nearly no constraint at all – non-profit hospitals, which account for the large majority of the nation’s hospital beds, must accept Medicaid (and Medicare) patients and payment levels in order to retain their non-profit status
  1. The ‘provider-tax’ arbitrage appears to be closing. For decades states have taxed Medicaid providers to raise a portion of the funds states need for their share of Medicaid costs; in economic downturns this shell game is magnified. Since the beginning of the current downturn most states have become substantially more reliant on ‘Other State Funds’ (provider taxes) to cover their share of Medicaid payments (Exhibit 5). This has artificially inflated Medicaid payment rates to hospitals since a tacit element of the provider tax arbitrage is that states pay hospitals higher Medicaid rates as a (more than 100 pct) offset to the higher provider taxes[3]. As a result, in 2011 the Medicaid payment to cost ratio was higher than for Medicare (Exhibit 1, again), for the first time in twenty years. Provider taxes recently have been capped, and a lowering of the cap rate is a possible consequence of broader federal budget negotiations. Thus at best, the contribution of the provider taxes to higher Medicaid hospital payment rates is fully played out; more likely, revisions to provider tax rules and regulations may result in yet another source of pressure on payment rates

Stepping away from Medicaid payment rates for the moment, we continue to believe hospitals have considerable commercial pricing power into 2014. As the health insurance exchanges begin enrolling, many (presumably most) new enrollees will not have pre-existing primary care relationships, and will not recognize health plan brand names – however most will recognize hospital brand names, and are likely to choose among exchange-based health plans in large part based on which hospitals are in-network. This naturally translates into pricing power for hospitals up to and during the initial enrollment period; however once an initial enrollment surge has passed, we expect exchange-based plans to begin narrowing networks, substantially reducing hospitals’ commercial pricing leverage

Considering payment rates from all sources, the aggregate picture is one of hospital payment rates that peak in or around 2014. An added source of pressure beyond 2014 is likely to be a lowering of average actuarial values (AVs)[4] in employer-sponsored insurance[5]. Current employer-sponsored AV’s average 82; as some portion of beneficiaries are moved into exchanges, we expect average AV’s to fall to +/- 65. This has a potentially dramatic effect on hospitals’ net pricing. In the 82 AV scenario, the hospital collects 82 pct of charges from the beneficiary’s insurer at +/- 99 pct collection rate, and the remaining 18 pct of charges form the beneficiary at a +/- 35 pct collection rate, for a net of roughly 88.3 pct. In the 65 AV scenario, the hospital collects 65 pct of charges from the insurer at the +/- 99 pct collection rate, and the balance (35 pct) from the beneficiary at something presumably less than the current +/- 35 pct collection rate. At best the net collection falls to 77.3, ignoring any volume effects[6]

Despite these pressures, hospitals remain reasonably attractive given current low valuations (Exhibit 6), however we’d recommend an ever increasing emphasis on Non-Rx Consumables (especially CFN, OMI) at the expense of hospitals as 2014 approaches, as these names are likely to receive the full benefit of rising per-capita volumes (whether due to exchange roll-out, rising employment, or both) without being exposed to the same pricing pressures[7] faced by hospitals beyond 2014

  1. Please see: “Why Medicaid Eligibility Will (Still) Level Off at 100 FPL”, December 12, 2012, SSR Health
  2. New York Times: “States Can Cut Back on Medicaid Payments, Administration Says”, Robert Pear, 2/25/2013
  3. See the Congressional Research Service reported entitled “Medicaid Provider Taxes”, Alison Mitchell, 3/15/12, which can be found here:
  4. ‘Actuarial value’ refers to the average percentage of allowable costs covered by a plan. For example, an 82 AV plan is one in which the plan pays, on average, 82 pct of allowable costs
  5. For details, please see: “US Healthcare Demand Part 3: Reform Effects – ACA Looks Like a Headwind”, SSR llc, October 1, 2012
  6. Which are considerable, since the beneficiary’s out-of-pocket cost nearly doubled, from 18 pct of charges to 35 pct. Observed elasticities suggest this change in out-of-pocket costs would lead to a +/- 17 pct drop in unit demand from the affected beneficaries
  7. We recognize that pricing pressure on hospitals should translate into pricing pressure on hospital suppliers; however by choosing hospital suppliers whose prices are competitively / structurally stable, we believe we can have the volume benefit of rising admissions without facing substantial adverse changes to pricing
Print Friendly