Reading the Manhattan Institute Part 2

Part Two of a quick analysis of the Manhattan Institute’s health care plan or ‘plan’:

The document can be found here.

I will be looking at Part Two “Reforming Employer-Sponsored Health Insurance” and Part Three “Medicaid Reform: Transforming Health outcomes for the poor” today.

The biggest thing to remember from Roy’s point of view is that Americans, or at least the class of Americans who have to give a damn about health insurance, have it too damn easy. We have too much coverage, we have too many choices, we don’t have to fear bankruptcy every time we make the decision about whether or not to go get that funny mole checked out, or that sharp, persistent stabbing pain on the left side of your abdomen examined. We don’t value health care enough because we don’t pay enough first dollar expenses. That is his diagnosis of the problem. And his solution set addresses that problem.

So let’s review.

Employer Sponsored Health Care

• Cadillac tax is moved up one year, and it is broadened to end the exemptions for most high risk non-law enforcement/non-fire professions. Other than that, there are no significant changes

This is an area of productive discussion. I would want to keep the high risk occupation exemptions (which are actually just add-ons to the base where the tax kicks in) but this is an area where liberals and conservative policy wonks could hash something out over a six pack as part of a wider set of negotiations.

• Cadillac Alternative

P. 35 An alternative to preserving the Cadillac tax, as described above, would be to apply a fiscally equivalent cap on the size of the employer tax exclusion.

Again, this could be a productive conversation. My inclination is to say no purely for path dependency reasons — employers, employee groups and plan administrators have been working in the same direction in anticipation of the Cadillac plan for a couple of years now. Economically, the two policy options are roughly the same, but the plumbing and prep work could be different. If that is the case, why junk the existing preparations.

• Employer Mandate

P. 35 THE ACA ALSO CONTAINS AN EMPLOYER MANDATE, Requiring firms with 50 or more full-time workers to offer federally defined “minimum essential coverage” or
pay a fine of $2,000 times the total number of full time- equivalent employees at the firm, less 30.

97% of large firms with 50 or more full timers already offer Bronze or better benefits. The employer mandate has been something that liberals have been willing to give up and the economic analysis is reasonable. Again, as part of a full set of negotiations between rational parties who are all interested in deal making, this could be part of a conversation.

• Benefit requirements

P. 36 Small group plans, in particular, are affected by the ACA’s requirements regarding essential health benefits
and medical loss ratios. The Universal Exchange Plan’s exchange-based reforms, as proposed in Part
One of this monograph—the ones that expand insurer flexibility around benefit design and financial structure—
will have the added effect of modestly lowering the cost of employer-sponsored coverage.

Small groups can resume offering lacy negligee coverage — it looks like it covers a lot but it really doesn’t.

It is amazing what the phrase “flexibility around benefit design” elides as that means employers don’t have to offer maternity coverage, they don’t have to offer mental health coverage at parity, they don’t have to offer preventative care at no cost share, they don’t have to cover certain classes of drugs. This reverts back to the 2009 status quo of good luck on figuring out which is actually covered. And this is where the core difference is — Roy is prioritizing lowering premium costs and his method of doing so throughout the monograph is minimizing what is covered and decreasing the probability that coverage actually kicks in. That works if and only if your goal is minimize specifically on one function. People are “covered” in Roy’s lexicon if they have a piece of shit policy for $22 a month that does not pay a dime until someone has spent their entire annual income on medical expenses. That is not coverage that is even vaguely useful in my mind.

• Cost accounting

• (1) by mitigating the phenomenon of cost-shifting, whereby health care providers charge commercial
  insurers higher rates to compensate for low reimbursements plans; and (2) by addressing the inefficiencies in Medicaid and Medicare that drive up overall health care
  costs.

  Cost shifting arguments chafe me. If we are to assume a reasonably free market that Roy wants to assume either explicitly or implicitly, we should assume that providers (sellers of medical services) will attempt to get the maximum rate that they can from any payer. And if that rate is not high enough, they won’t sell their service (as Roy argues in the Medicaid section, the MA rate is not high enough most of the time for all providers). So within this theoretical construct, sellers of services should already seek to maximize their rates from any payer irregardless of what any other payer is offering. Cost shifting is incompatible with the theoretical construct that Roy argues, and more importantly, there is evidence that it does not occur.

  Historically, private insurers have paid higher rates than Medicare for inpatient hospital care, and the gap between Medicare and private rates widened from 45 percent in 1995 to 57 percent in 2009, the study found. The average Medicare payment rate per discharge in 2009 was $11,031, while the average private rate was $17,286. There are two possible explanations for the gap—either hospitals charge private payers more to compensate for lower Medicare payments or hospitals can and do charge private payers more because of other factors, such as negotiating leverage.

  “The study found that when Medicare pays lower rates for inpatient hospital care, private insurers’ rates end up growing more slowly, too—it’s the opposite of what hospitals would have the public believe,” said HSC Senior Researcher Chapin White, Ph.D., the study’s author. “Hospital executives, understandably, want higher payment rates from private payers. To put a socially acceptable spin on higher rates, they blame Medicare for being a stingy payer—this study should put that notion to rest….”
  The study also found that a hypothetical 10-percent reduction in Medicare payment rates would lead to an estimated reduction of private payment rates of 3 percent or 8 percent depending on the statistical model used.

This makes sense if we use a Gini/HI ratio analysis of who has negotiating leverage as we did when I first started to write for Balloon Juice

Dominant insurers use their ability to steer hundreds of thousands patients to or from different providers as a stick to get rates that are based on Medicare with a small kicker. This works because a dominant insurer does not need to get every provider in a specialty or concentration into the network. They just need to get enough providers in network to satisfy state regulators. Providers who don’t take the low rates from the dominant insurer have a hard time making up at higher price points the revenue they lose in volume./

Medicare FFS is the biggest payer by far in the country.
Now onto the next point, Medicare does not have a structural pricing problem — they have a bit of a utilization problem. Medicaid has a pricing problem, they pay too low. Commercial insurers do a decent job of containing utilization but they have a structural pricing problem. As soon as I shit unicorns out of my ass, a fragmented commercial market will crack prices to Medicare or below in combination with utilization containment and we can all eat as much ice cream as we want without gaining weight. Using the simple Gini/HI model as a framework, I have a damn hard time seeing that happening. And knowing what goes into pricing decisions, the private sector does not have the data capacity to build a replicable and predictable rate schedule from scratch without Medicare as a baseline for several years. This is a massive disruption for no real gain even on Roy’s terms.

Medicaid

• Cherry picking to pull a fast one on us

P. 37 A LANDMARK STUDY PUBLISHED IN THE NEW ENGLAND Journal of Medicine compared health outcomes for Oregon
residents who had won a lottery to enroll in that state’s Medicaid program with demographically similar
residents who had lost the lottery and remained uninsured.
After following these individuals for two years, the authors found that Medicaid “generated no significant
improvement in measured physical outcomes” such as mortality, high blood pressure, high cholesterol, and diabetes.31

Let me just grab the Incidental Economist for a long minute:

Many people are contrasting the Oregon study—which didn’t find statistically significant effects of coverage on biomarkers associated with physical health—with the new Massachusetts study—which found a statistically significant mortality benefit of coverage. How could these two findings coexist in a rational world? There are various hypotheses, one of which is that the Oregon study was under powered (too small a sample size) to find physical health effects, as we’ve documented. (There are so many posts on this. Here’s just one.)

Just to illustrate the difference in power of the two studies, I did a thought experiment. What if we presume that the same mortality effect that was found in the Massachusetts study applied in the Oregon case? Would the Oregon study have been able to detect it with statistical significance?

The answer is no, and it’s not even remotely close. The Oregon (OR) study had a sample size about a factor of 100 below that of the Massachusetts (MA) study. That means the error bars would have been about 10 times larger. Here’s what that looks like:

Again the Incidental Economist on the limits of the Oregon study (it is too damn small for fine grained clinical measures as most people don’t have problems on most metrics)

The standard level of statistical significance is rejecting the null with 0.95 probability. Assuming the same baseline 5.1% elevated GH rate and a 20% reduction under Medicaid, what sample size would we need to achieve a 0.95 level of significance? Plugging and chugging, I get about 30,000 for the control group and a 7,500 treatment (Medicaid) group. (I’ve fixed the Medicaid take-up rate at 25%, as found in the study.) This is a factor of five bigger than the researchers had.

So Roy is engaged in at best truthiness or is engaged in outright bullshitting as he should know better as I know that he has been exposed to these arguments. And interestingly, he never mentions the Massachusetts mortality study which is actually big enough to make the claims that he wants to make, but they make the claims in the opposite direction.

He then goes on about relative risk rates of dying in the hospital. Does he happen to consider that there is a strong possibility that people who are on Legacy Medicaid may be different pre-hospitalization than people who have good to very good commercial insurance? Maybe their lives are more stressful, maybe the hospitalization is happening at a later point in the course of disease, maybe the easy/elective procedures are happening for a much higher rate for people with good insurance which inflates their denominator… Nahh.. can’t be, let’s use numbers to be authoritative.

• Low primary care provider reimbursement rates are a problem (p. 39)
  Yep, and PPACA has a plan in place to pay Medicaid PCPs, Medicare rates for 2013 and 2014 with discussion of extending this for another year or two.

• MIGRATING THE MEDICAID ACUTE-CARE POPULATION ONTO EXCHANGES

The thrust of this plan is to take the Indiana Healthy Indiana HSA style medicaid expansion waiver to national scale. The acturial value of coverage for the poor would decrease as Medicaid is typically a 97% to 100% acturial value program while he says cost-sharing subsidies currently available to the 100% to 150% FPL group would apply to people under 100% FPL. Those cost sharing subsidies makes a plan be worth around 94% of acturial value. He’ll top it off with HSA contributions so if you’re healthy, you’re okay, if you’re unhealthy (and we know the current Legacy Medicaid population is unhealthier than average) you’re screwed as you are also flat broke (as you already qualified for Medicaid via low income). He wants Medicaid members to move to the Benchmark Exchange plan (in his case a plan that covers 55% acturial value plus the top-up of cost sharing assistance subsidies) would have a benchmark deductible of $7,000 brought down to either a $500 to $1,000 deductible without any free preventative care mandates. So if you are poor, don’t get sick, or at least die quietly in the corner so that you don’t drag down the rest of your social network.

• Another important problem facing the Medicaid population is the problem of churn between different types of insurance coverage.

Churn is a problem. Moving everyone to annual or bi-annual contracts on the Exchange would solve the problem. However, it is more easily solved by mandating that income is verified once a year instead of every three to six months. That would mean accepting that someone who was broke and Medicaid qualified on the month of application but then has good luck five months and gets a good paying job without benefits would be kept on Medicaid instead of making too much money to qualify.

Higher costs on Exchanges for Medicaid members

migrating the Medicaid acute-care population onto the exchanges, over a ten-year period, would increase federal funding responsibilities by approximately $1.2 trillion, and reduce state spending by a corresponding amount, excluding the impact of higher per-member costs under the exchanges

Nice little magic asterisk right there (excluding the impact of higher per-member costs…). We know for a fact that an Arkansas style private option policy with the same benefit configuration costs significantly more than the same person on Medicaid because Medicaid pays significantly less. Back of the envelope calculations, Roy is eliding at least half a trillion dollars here. He addresses it later, but slick.

• Transfer Long Term care to state responsibility
• The states would give up the acute care Medicaid population to the Feds and take over the entire long term care/disability obligations of Medicaid with a proviso that they do no worse than they would have under joint Federal-State Medicaid until 2036.

 

If we did not have Mississippi, Texas, and Alabama in this country, this could make sense. Long term care for either the elderly or disabled is a very different beast than short term acute care for reasonable healthy people. It is a bit odd that we smash these two programs together into Medicaid, so separate programs with separate and distinct expertise could make sense. The political dynamic in three years or the first recession after this change goes into effect is that the Feds are dictating an un-funded mandate to the states, so the states are spending money on moochers instead of giving the productive another non-incentive changing tax cut. In Roy’s world, that is a feature and not a bug; in my moral calculus, that is a major bug and not a feature.

My very reasonable fear is that being old or disabled in Texas with Texas paying for everything would lead to far worse outcomes than being old or disabled in Massachusetts.

Moving Dual Eligibles to Exchanges

Under the Universal Exchange Plan, all of these “dual eligible” individuals would be migrated onto the exchanges, where they would receive an insurance benefit of the same actuarial value as that represented by their existing Medicare and Medicaid coverage.
This would amount to a benchmark exchange plan with the cost-sharing subsidies—in the form of health savings account subsidies—needed to achieve actuarial
equivalence. In this way, dual-eligible individuals could gain coverage from a single health plan managed by a single insurer, with a unified network of physicians

Dual eligible individuals are a problem from a high financing point of view as they are extremely expensive (on average). So unless the actuarial value is massively risk adjusted, a plan to give them an Exchange plan with cost sharing subsidy in the form of an HSA will gut their coverage as they can’t save. Furthermore, it is particularly wise to count on an individual who is in long term nursing care with a combination of dementia, diabetes and hypertension to be able to make cost effective decisions? Is it reasonable to count on their health care proxy to take on even more tasks to determine appropriate care? That is what this plan is asking.

An easier solution is to move dual eligible to special needs plans under Medicare Advantage and Medicaid MCOs which produces the unified network, unified back end administration and incentives for cost containment — but hey, that is not shiny and new and WOW.