US Healthcare Demand Part 3: Reform Effects – ACA Looks Like a Headwind

ginna
Print Friendly
Share on LinkedIn0Tweet about this on Twitter0Share on Facebook0

Richard Evans / Scott Hinds / Ryan Baum

SEE LAST PAGE OF THIS REPORT FOR IMPORTANT DISCLOSURES

203.901.1631 /.1632 / .1627

https://twitter.com/images/resources/twitter-bird-blue-on-white.png richard@ / hinds@ / baum@sector-sovereign.com

@SecSovHealth

October 1, 2012

US Healthcare Demand Part 3: Reform Effects – ACA Looks Like a Headwind

  • We expect Affordable Care Act (ACA) health insurance and pricing effects to reduce real demand growth by roughly 30bp over the period from 2012 – 2021
  • Medicaid should expand to 100 FPL in most states; this translates to enrollment gains of 16% (8.4M persons), Medicaid spending gains of 10%, and 1.3% gains in national health spending. Because Medicaid prices are roughly 29% below commercial prices, Medicaid’s expanded share of third party payments reduces real price growth by just less than 10bp annually over the coming decade
  • Persons ineligible for Medicaid and with no offer of employer sponsored insurance (ESI) – whom we term the ‘middle market’ – see enrollment gains as high as 135.5% (19.5M persons), and spending gains of just less than 50% once the health insurance exchanges (HIEs) begin operating; the effect is to increase national spending by as much as 2.6%
  • Policies purchased on the HIEs will almost certainly be less generous than those offered under ESI. The prevailing ESI actuarial value (AV) is 0.82; we expect a 0.65 AV average on the HIEs. A likely outcome is that cheaper HIE plans catalyze shifting of employees out of ESI and onto the HIEs; and/or a downdraft in the average generosity of ESI plans to or toward the HIE average. If the ESI plans’ average actuarial values fell from 0.82 to 0.65, we would expect a nearly 17% drop in spending by ESI beneficiaries; such a drop would reduce national health spending by nearly 7%
  • Combining these three coverage effects — Medicaid expansion, HIE operations, and HIE effects on ESI – we see the potential for a 3% reduction to the baseline rate of healthcare demand growth. Our expectation is for a more moderate 1.5% reduction, which spread across the 2012 – 2021 period translates into an average annual headwind to real demand of roughly 20bp. We estimate the total of ACA effects on real pricing to be an average annual headwind of roughly 10bp across this same period; thus in total we expect the ACA to reduce real demand growth by roughly 30bp
  • We are bullish on Hospitals and (particularly commercial) HMOs ahead of the HIEs (which we expect in 2015 rather than 2014); Hospitals have more pre-reform pricing power than is generally believed, and we expect stable commercial MLRs where the market appears to anticipate rising MLRs. After the HIEs begin operating our view begins to reverse; we see less potential for longer-term volume growth in Hospitals (believing consensus expects a large tailwind from ACA), and steady declines in average contract values (plus rising per-member OPEX) for commercial HMOs
  • One of the largest and perhaps least appreciated effects of ACA is on pricing of innovative products (e.g. devices, drugs) to Medicare Parts A & B; beginning in 2015 we expect rebates on sales to Medicare of +/- 25%.  DNDN, CBST, AMGN, MDCO, THOR are most susceptible; each relies on Medicare for 40% or more of sales

Summary and Conclusions

The Affordable Care Act (ACA) expands Medicaid, improves offers of insurance to persons who are both ineligible for Medicaid and lacking an employer offer, and changes a multitude of incentives in the employer-sponsored insurance (ESI) market. Additionally, the Act makes other non-coverage related changes that affect real healthcare demand; prominent among these is the likely effect of the Independent Payment Advisory Board (IPAB) on innovators’ pricing

The demand impacts of ACA-driven coverage changes depend heavily on the extent of the Medicaid expansion, the actual acceptance rates of coverage offered on the health insurance exchanges (HIEs), whether employers shift employees from ESI to the HIEs, and whether the lower average values of HIE-based policies lead to lower average ESI policy values. The estimates of reform effects provided in Exhibit 1 assume that most states eventually expand Medicaid eligibility to 100 FPL, that HIE uptake rates are marginally lower than ESI acceptance rates, that a third of employees are shifted from ESI to HIEs, and that the average value of ESI policies falls considerably as the result of cheaper plans being sold on the HIEs. On net, we expect ACA-driven coverage changes to actually lower average per-capita intensity over the coming decade by roughly 20bp / year. We believe risks are to the downside of this estimate, and provide a full range of sensitivities in our summary

Separately, we expect the ACA will modestly temper real price growth. Medicaid grows as a percentage of total third party payments; because Medicaid rates are substantially lower than commercial rates this obviously offsets a portion of system-wide real price growth. And, the IPAB’s likely approach to its mandate may involve rebates on innovative products. Together, the Medicaid and IPAB effects reduce our expectation of real pricing growth in the coming decade by roughly 10bp

Taken together, we expect the coverage and pricing provisions of the ACA to reduce per-capita real demand growth by roughly 30bp over the period 2012 – 2021

Medicaid Effects

We expect (<65) Medicaid enrollment to expand by 16% as a result of the ACA, and Medicaid spending to grow by 10%. Net spending growth in the expanded Medicaid population is slightly reduced by these new enrollees’ modest prior spending; the expansion raises total national health spending by 1.3%, and reduces the number of uninsured persons by 8.4M, which in 2012 would be equivalent to reducing the national rate of uninsurance from 16% to 13%

In sizing the expanded Medicaid program we explicitly assume that most states eventually will make all persons under 100 FPL[1] Medicaid eligible, and with the exception of pregnant women and children aged 0-5 will eliminate eligibility for persons above 100 FPL. The basic rationale is Keynesian; at or below 100 FPL states’ best economic option is to attract federal Medicaid matching dollars by raising eligibility to 100 FPL; above 100 FPL states are better off attracting the much larger (than Medicaid) federal subsidies for health insurance purchased on the exchanges by making these persons ineligible for Medicaid[2] (and thus eligible for exchange subsidies)

Exhibit 2 provides relative health spending for age- and health status-matched cohorts according to health coverage status (Medicaid v. uninsured); Exhibit 3 provides details on the changes in enrollment and related impacts on total national health spending that we expect as a consequence of the ACA’s Medicaid expansion

Importantly, prices paid by Medicaid for healthcare products and services generally are lower than those paid by commercially insured (or uninsured) persons, thus the expansion of Medicaid will have a modest deflationary effect on overall national pricing growth. Hospital, physician services, and drug pricing are roughly 27%, 42%, and 19% lower in Medicaid than in the commercial market, respectively. On a weighted basis, Medicaid prices are roughly 29% lower than commercial insurers’ prices. Across the 10-year period from 2012 to 2021, the effect of the Medicaid expansion should be to reduce real health price growth by 8bp per year

‘Middle Market’ Exchange Effects

In crude terms, the effect of the health insurance exchanges (HIEs) is to provide an affordable offer of health coverage to persons and households that do not have an offer of health insurance from an employer (or other large group), and who also have incomes above the cut-off for Medicaid eligibility –we refer to this population as the ‘middle market’. After accounting for the Medicaid expansion as outlined above, we estimate 37M persons in the middle market as of 2012

The current rate of uninsurance in the middle market is 61%. By contrast, the rate of uninsurance for persons with an ESI offer is approximately 8%. If we start with the simplistic assumption that the effect of HIEs is to bring an ‘ESI-quality’ offer to persons who do not currently have an ESI offer, then we might further assume the rate of uninsurance in the middle market falls to (or nearly to) the rate of uninsurance among persons who do have an ESI offer. Under these assumptions the net effect of bringing HIEs to the middle market would be to add 19.5M persons to health coverage

The assumption that HIE offers are ‘equal’ to ESI offers is imperfect, and we believe the risks lay to the downside. Employees’ premium costs as a percent of income increase as incomes fall under ESI; under HIEs the opposite is true: premiums as a percent of income fall as incomes fall. In this regard the HIE offer is substantially better than an ESI offer, particularly for persons with lower incomes who make up a disproportionate share of the uninsured. Conversely, under ESI premiums are deducted from paychecks, whereas HIE enrollees have to pay their share of premiums monthly with after-tax earnings from their cash account(s); in this regard the HIE offer is ‘worse’ than the ESI offer. As a related distinction, under ESI participants are enrolled by default; under HIE participants have to actively ‘opt-in’; this too would tend to reduce HIE participation relative to ESI participation. Finally, unlike the ESI market, on the HIEs persons and households face no restrictions with respect to pre-existing conditions; on net this would tend to reduce uptake on the HIEs relative to ESI

At present, 39% of the 37M persons in the middle market purchase non-group coverage; under the simplistic assumption that HIE and ESI offer rates are equivalent we would expect as many as 92% of those in the middle market to have health insurance under the HIEs, an increase of 19.5M persons. The average actuarial value[3] (AV) of policies held by those currently insured in the middle market is roughly 0.65, and we expect marginal enrollees gained under the HIEs to purchase policies of similar values. For each currently uninsured person (or household) purchasing coverage on the exchange, we expect an increase in health spending of about 150%. This translates to the potential for 48% growth in spending within the middle market, and a potential 2.5% increase in national health spending. Still assuming similar rates of HIE and ESI acceptance, the availability of health coverage on the HIEs potentially adds 19.5M persons to coverage, which in 2012 would have the effect of reducing the uninsurance rate from 16% to 10% (Exhibit 4). Enrollment risks are to the downside; these are addressed in sensitivity analyses provided in the summary

Employer-Sponsored Insurance (ESI) Effects[4]

We take the view that employers fashion compensation (wage + benefit) packages as bids to secure the best available talent at the lowest possible price. This frames the question of whether employers shift employees out of employer-sponsored insurance (ESI) and onto the health insurance exchanges (HIEs) very simply – employers will take whichever path allows them to produce a target compensation package at the least possible cost

Because federal subsidies to subsidy-eligible (> 400 FPL) households generally are larger than the sum of penalties and tax effects faced by employers that shift employees to the HIEs, the net incentive is to shift employees out of ESI

Employers A and B each want to offer a compensation package with a given actuarial value (AV) of health coverage (0.82), and a given amount ($53,370) of after-tax, after-health premium wage. Employer A uses ESI; Employer B uses the HIEs. ‘A’ pays a pre-tax wage of $71,250, ‘B’ pays a pre-tax wage of $75,362; however after ‘B’ employees purchase the same plan on the HIEs that they would have been given under ESI, ‘B’ employees have the same amount of after-tax (and after premium) wage left over (Exhibit 5). From the perspective of the employee, B’s offer is at least as good as A’s[5]

Even though A and B have produced compensation packages of equal value, B has done so for less cost – using the wage and health coverage values in our example, B spends $2,666 less per employee than A (Exhibit 6)

This basic rule (HIE costs employers less than ESI) holds true across the income range of subsidy-eligible households (100 – 400 FPL) for households purchasing family coverage (52% of ESI premiums), and holds true across nearly all of the income range (up to +/- 350 FPL) for households purchasing single + one coverage (21% of ESI premiums). The rule fails above +/- 200 FPL for persons buying single coverage (26% of ESI premiums)

We conclude that because employers can realize considerable savings on the majority of their subsidy-eligible employees’ compensation packages by grossing up wages and shifting employees to HIEs, that most employers eventually will do so, if the HIEs are viable[6], and if no new restrictions to such a shift emerge[7]

If employees are shifted from ESI to HIE, two demand effects are realized: 1) persons covered under ESI may not enroll on the HIE; and 2) those who do enroll on the HIEs on average buy plans cheaper than the ones they had under ESI

The loss of enrollment in an ESI to HIE shift is best explained by the differences in opt-out and opt-in behaviors. Declining ESI requires potential beneficiaries to opt-out, i.e. the default ‘status’ for employees in firms that offer ESI is to be insured. When the insurance choice is framed as an opt-out, the level of participation is higher than when it’s framed as an opt-in – as it is in the case of the enrollment choice on the HIE. And, ESI participation levels benefit from the tendency of persons to place less value on amounts deducted from payroll than on amounts withdrawn from their cash account(s). In the very best case we assume no enrollment losses in any shift from ESI to HIE; however the risk to this assumption plainly is to the downside, and we provide the corresponding sensitivity analyses in the summary

The second demand effect of reforms on ESI is to reduce the average value of coverage purchased. Under ESI, employees typically have very few coverage options, and the average actuarial value (AV) offered is roughly 0.82. On the HIEs, beneficiaries can chose plans across an AV range of 0.60 to 0.90 from any of several underwriters

Because: 1) beneficiaries have a reasonably good ability to predict forward health risks; 2) beneficiaries on lower AV plans (or uninsured) can always upgrade to a higher AV plan at the next annual enrollment period; and 3) the marginal costs of premiums for more generous plans (i.e. higher AV) than the plan covered by subsidies are paid entirely by the beneficiary using after-tax dollars, we expect more healthy beneficiaries to enroll in the 0.60 AV plans, and less healthy beneficiaries to enroll in the 0.90 AV plans (Exhibit 7). Thus on the HIEs, we expect a weighted average AV of roughly 0.65, as compared to the prevailing 0.82 under ESI

Even if employers don’t shift employees to HIEs, we expect 0.65 AVs to prevail on the HIEs, and we expect this to put downward pressure on the generosity of coverage offered under ESI – in fact we expect ESI generosity to fall to or at least very much ‘toward’ 0.65. There are multiple reasons to expect this outcome, chief among these is adverse selection. An employer offering ESI in a market where a similar employer offers higher wage (and where employees buy health coverage on the HIEs) is more likely to attract older, less healthy workers. Conversely the higher wage / HIE employer is likely to attract younger / healthier workers. Also, recall the Cadillac tax provisions of the ACA, which tax premiums above a certain level; over time these should serve to lower average contract values

Thus whether or not employers shift employees out of ESI and onto the HIEs, the average AV of employees’ health coverage is likely to fall to (or toward) the HIE-based average; this has a major effect on demand, and we can frame this in terms of price elasticity. An uninsured person has a plan with an AV of zero; a person on the HIE has an (average) AV of 0.65, and a person under ESI has an average AV of 0.82. The cost of $100 worth of healthcare to the uninsured person is $100; to the HIE beneficiary the cost is $35 ($100 – 0.65×$100); and to the ESI beneficiary the cost of $100 worth of healthcare is $18 ($100 – 0.82×$100)

Estimates of the price elasticity of demand for healthcare cluster around –0.20, and reasonably well accepted estimates can range as high as -0.60. For our immediate purposes we’ll use an assumption of -0.20, both because this is supported by the weight of research evidence; and because this is the most consistent with our own observations[8]

We then simply treat the shift from an 0.82 AV to a 0.65 AV as a 94% price increase[9]; at an elasticity of –0.20 this implies an 18.8% decline in real (i.e. unit) demand

To estimate reform effects across the ESI population, we assume that the average AV of ESI policies falls from 0.82 to 0.65 over a five year period beginning in 2015. We do not make any explicit assumption regarding a shift of enrollees from ESI to HIE. Our reasoning is two-fold: despite the financial benefit of shifting subsidy-eligible employees out of ESI, uncertainties related to federal funding for subsidies[10] may limit employers’ willingness to make the shift; and, the downdraft in average AV is by far the biggest effect of reform on total ESI demand, and this downdraft is likely to occur whether or not employees are shifted out of ESI. Sensitivities around the enrollment-loss and AV ‘downdraft’ assumptions are provided in the summary

Exhibit 8 provides our ‘point estimates’ of the impact on enrollment and spending among the ESI population, and on national spending and enrollment in total

Summarizing ACA Coverage Effects on Real Demand

Exhibit 9 rolls-up the Medicaid, ‘middle market’ and ESI effects of the ACA to a ‘potential’ point estimate of change in national health spending, and national rates of uninsurance. This ‘point estimate’ scenario assumes no shifting of employees from ESI to HIE (or conversely that all of those shifted enroll), and a drop in average AV under ESI from 0.82 to 0.65 over a five year period beginning 2015

The uptake rate of HIE, the change in average AV of ESI, and the rate at which employees are shifted from ESI to HIEs are by far the most impactful assumptions. The assumed uptake rate of HIE ‘flows’ in two directions. The first is the percent of middle market uninsured persons with no ESI offer who choose to enroll once the HIEs begin operating. Our ‘point estimate’ assumes a best case, i.e. that this rate of uptake is that same that we observe in the ESI market – about 92%. This implies that the HIE and ESI offers are equally attractive, and this assumption spills over into the other ‘side’ of the uptake rate – namely the percentage of persons shifted out of ESI who purchase coverage on the HIE. If we assume ESI and HIE are equally attractive, naturally we have to assume everyone shifted out of ESI enrolls on the HIEs. We believe the risks here are to the downside, i.e. that the uptake rate of HIE insurance offers is lower than the ESI uptake rate; and, correspondingly that not all persons shifted out of ESI choose to purchase coverage on the HIEs. The relevance of the uptake rate assumption obviously is compounded by our assumption regarding the ESI to HIE shift: the more employees shifted at a low uptake rate, the greater the number of persons becoming uninsured, and vice versa

The third major assumption is the decline in average AVs among current ESI participants; the current average is 0.82, and our point estimate assumes this falls to 0.65 over the five year period beginning 2015

Assuming zero shift of employees from ESI to HIEs, Exhibit 10 shows the range of outcomes (change in total national real health demand growth) of varying our uptake rate and average AV assumptions across reasonable ranges: uptake rate varies from 80% to 92%[11]; average AV under ESI varies from 0.65 to 0.82. Each cell contains two values; the upper value is the total one-time change in real national health demand as the result of ACA-related reforms, the lower value is the change in real national health demand expressed as an effect on the 5 year CAGR (2015 – 2019) of (total national) real demand growth. Our ‘point estimate’ is in the upper-left hand corner; i.e. if uptake rates of HIE are equal to ESI uptake rates, and average AV’s on ESI fall to 0.65, then real national health demand falls by 3.0%, or if we calculate the change as a 5-year CAGR, by 0.6%. Conversely, if we assume the acceptance rate of HIE offers is only 85%[12], and that ESI AV’s only fall to 0.74, then the one time effect of reforms[13] on real health demand is to increase demand by 10bp

Exhibits 11 and 12 are identical to Exhibit 10, however under Exhibit 11 we assume that one-third of ESI beneficiaries are shifted to HIEs; in Exhibit 12 we assume two-thirds of subsidy eligible employees are shifted to HIEs

The mid-point of the three major assumptions – rate of HIE uptake, change in ESI AV, and rate of shift of persons from ESI to HIE – produces the values in the center cell of Exhibit 11: that ACA’s coverage related provisions reduce real healthcare demand by 1.5% when calculated as a ‘one-time’ effect, or by 30bp when calculated as a 5 year CAGR

Rather than lose ourselves in basis points or even whole percentage point levels of difference, we think the strategically relevant conclusion is that the ACA is more likely to have a flat to negative effect on real demand than it is to have any significantly positive effect on real demand

The Independent Payment Advisory Board (IPAB)

The IPAB is required to propose measures that reduce growth in per-beneficiary Medicare spending to within a range of either CPI (2015 – 2017) or GDP (2018 – 2019). By law IPAB cannot reduce hospital (and many other facility) payment rates; for practical reasons IPAB cannot meaningfully reduce physician payment rates. Also by law, IPAB cannot reduce eligibility or raise premiums[14]

To our way of thinking this leaves IPAB with no option other than to reduce the input costs to care by some means other than lowering hospital or physician pricing. In turn, this narrows IPAB’s options to reducing the costs of materials[15] used in delivering care, and these materials can be further sorted into two broad categories: commodity, and innovative. There’s little use attempting to reduce the cost of commodity materials used in Medicare beneficiaries’ care; by definition commodities are competitively priced. By process of elimination we conclude IPAB has few (if any) options other than to reduce the pricing of innovative inputs to care. We envision this happening much the way Medicaid rebates are processed. States collect data on the drug products consumed by Medicaid beneficiaries, and send corresponding bills for rebates to the manufacturers of these products on a quarterly basis. We anticipate rebates on innovative inputs (e.g. devices, drugs) to care provided under Medicare Parts A and B[16] that are processed in much the same manner – e.g. regional payers compile data on the devices and drug products used by hospitals and physicians under Parts A and B, and send manufacturers periodic invoices for rebates on those products

To estimate the potential rebate burden, we estimate both the likely savings targets for IPAB, and the sales of devices and drugs under Parts A and B. Per-beneficiary Medicare spending growth is likely to exceed the CPI caps in 2015-2017, but at least for the moment is unlikely to exceed the GDP caps in 2018-2019). The extent to which spending exceeds the CPI caps in 2015 – 2017 also exceeds the percent cap on reductions IPAB is required to produce. The result is that we expect roughly $4.2B, $9.3B, and $12.3B in IPAB cost reductions in 2015, 2016, and 2017 respectively (Exhibit 13)

We estimate innovator sales to Parts A and B of $34.3B, $37.1B, and $40.3B in these same years; this implies an effective rebate on innovators’ Medicare sales of roughly 23% across this period (Exhibit 14). Even though incremental savings may not be required in years beyond 2017, innovator rebates applied in 2015 – 2017 almost certainly will not be reduced in later years

IPAB pricing risks fall disproportionately on those companies whose global sales are heavily reliant on sales of innovative products to Medicare Parts A and B. The cardio and ortho device innovators are very highly exposed, particularly THOR (Exhibit 15). Select biotechs (e.g. DNDN, CBST, AMGN, MDCO, VPHM) also are highly exposed (Exhibit 16). With the exception of Roche (Exhibit 17) large cap pharmaceutical manufacturers have limited exposure; similarly with the exception of ALXN specialty pharmaceutical companies have limited exposure (Exhibit 18)

Investment Conclusion

We continue to favor Hospitals and commercial HMOs ahead of the reform roll-out, which we anticipate in 2015. Innovators with large exposures to Medicare Parts A & B should be avoided (see Exhibits 15-18)

In Part 1 of this series we estimated baseline real demand growth – defined as the rate of growth one would expect in the absence of secular, cyclical, or reform effects – of 4.8%. In Part 2 we estimated a sustained secular headwind of 90bp; just less than half of this headwind is from declining uptake rates and average generosities in employer sponsored insurance (ESI), the balance is attributable to declining real growth in R&D spending. In Part 3, we’ve estimate a roughly 30bp headwind from coverage and pricing effects of the Affordable Care Act (ACA). Taken together, these estimates suggest sustainable real healthcare demand growth of roughly 3.6% (4.8% baseline – 90bp secular – 30bp ACA). For comparison; since 1960 real demand growth has averaged 5.2%; from 1990 to pre-Lehman growth was 4.0%. The forward ten-year period should benefit from per-capita intensity gains as a consequence of an improving jobs market (more on this in Part 4). Expressed as a one-time effect these cyclical intensity gains should raise real demand by +/- 2%; across the 2012 – 2021 period the average annual effect is to add +/- 20bp to real demand. All-in, our working expectation for 2012-2021 real demand growth is roughly 3.8%. This modestly reduced level of real demand is generally consistent with healthcare valuations in total; US-listed healthcare names trade at a slight discount to the SP500, versus their long-term history of trading at a slight premium

Volume bellwethers (e.g. diagnostic laboratories and manufacturers / suppliers of interchangeable products) trade at or near a market multiple; this implies the market has faith in underlying ‘unit’ demand for care. Hospitals and commercial HMOs trade at steep discounts to the market and their Healthcare peers; this implies some combination of price erosion and/or disintermediation. We expect Hospitals’ valuations are pressured by near-term concerns over net pricing power. We note that Hospitals would have to lose all pricing power in the near-term to justify current valuations; a more realistic scenario is that rising commercial prices at least mitigate – and may offset entirely — the effect of 2013 Medicare rate cuts. Commercial HMO valuations appear to suffer from concerns over rising MLRs; we expect stable MLRs in an environment of reasonably steady employment and predictable changes to medical pricing. After reforms roll-out (expected date 2015) Hospital volume gains are likely to eventually dis-appoint, and commercial HMOs are at great risk of declines in both market share and average contract values. The conventional wisdom appears to view post-reform prospects favorably for both Hospitals and commercial HMOs, thus our willingness to favor these sub-sectors in advance of 2015

  1. ‘FPL’ = federal poverty level; ‘100 FPL’ implies an income equal to 100% of the then prevailing FPL
  2. “Medicaid Eligibility Capped at 100 FPL: The Logical Outcome of the SCOTUS ACA Ruling” SSR, 7/27/12
  3. The percentage of allowable costs paid by the insurance plan
  4. For more detail, please see: “Why Employers Are Likely to Drop Health Insurance – A Simplified View” SSR; 7/11/11
  5. In reality, B’s (HIE-based) offer is more attractive. Under ‘B’, the employee has a much greater range of options with respect to how much health insurance to buy and from whom – including the option not to buy insurance. Because of this, ‘B’ is likely to attract younger and healthier employees than ‘A’
  6. Discounting the likelihood of repeal in the wake of a Romney victory, the primary risk to the HIEs is the ACA’s cap on federal subsidies to persons purchasing coverage on the HIEs. The cap is set at 0.504% of GDP. When the cap was set, the expectation was that Medicaid would expand to 138 FPL. Because an expansion to 100 FPL is more likely, persons and households between 100 and 138 FPL now are likely to seek coverage on the HIEs. Because subsidies are larger for persons and households with lower incomes, the claim on subsidies by persons between 100 and 138 FPL, combined with even the most modest assumption of enrollment by others on the HIEs, exceeds the 0.504% cap immediately. Even if Obama wins in November, the House is likely to remain under Republican control; since raising the cap almost certainly requires a budget initiative from a Republican-controlled House, there is every possibility that the employers will keep employees under ESI until and unless it’s clear that sufficient federal subsidies will be available
  7. The employer penalties are unlikely to go higher in the near- to mid-term. These penalties arguably were the highest (politically) possible at the time they were passed, even under a Democratic president with historic levels of approval and bi-cameral majorities. A more likely but still uncertain rule change would be for the IRS to disallow the tax deductibility of white collar health benefits for employers shifting employees to HIEs
  8. On an age and health status matched basis, the insured consume 2.7x more healthcare than the uninsured. Thus if an uninsured (zero AV) person consumes $100 of care, her cost is $100. If she has insurance (0.82 AV), she consumes $270, but pays $48.6. Her (insured) price per unit of care is $18; the uninsured price per unit is $100. Uninsured, her consumption (delta Q) falls 63% (from $270 to $100) in response to a 456% change in price (from $18 to $100, delta P); crudely this implies a price elasticity of demand of -0.14. Because the uninsured actually pay higher prices than the commercially insured, the delta Q is in fact larger, which easily puts our crude estimate of price elasticity in the range of -0.20
  9. Price is 1-AV, so under an 0.82 AV price is 0.18, and under an 0.65 AV price is 0.35 – a difference of 94%
  10. See footnote 6
  11. The uptake rate is expressed as the rate at which uninsured persons with no ESI offer choose to purchase coverage on the HIEs. In the case of the 92% rate, we are assuming HIE offers are accepted at the same rate as ESI offers – since the known ESI acceptance rate is in fact 92%. Looking at this from the perspective of persons shifted out of ESI, the 92% rate implies 100% of persons shifted out of ESI enroll on the HIE
  12. Which also implies that 8% of persons shifted from ESI to HIE do not enroll (1-(0.85/0.92))
  13. Or more precisely the one-time effect of reform-related coverage provisions
  14. Congress is not compelled to rubber-stamp IPAB’s recommendation; however if Congress does not pass the IPAB recommendation (which benefits from reconciliation rules in the Senate), Congress must pass its own legislation that achieves the applicable savings target
  15. Labor is very nearly out of the question – it’s captured under hospital and physician pricing schedules
  16. Part D (prescription drug benefit) is at least temporarily protected by the exclusionary clause which precludes the Secretary from setting drug prices – and by the practical reality that effective Part D discounts on retail brands already are quite large
Print Friendly