David Anderson

The New York Times reports on a new lawsuit brought against the University of Pittsburgh Medical Center (DISCLOSURE: I worked at UPMC Health Plan before I came to Duke).  The theory of the case is simple — UPMC has massive labor market power and uses that to drive down wages and make working conditions worse for employees.

After a series of acquisitions, it is Pennsylvania’s largest private employer with more than 40 hospitals, 800 doctors’ offices and clinics, and a health plan. With operating revenue of $26 billion last year, it employs about 95,000 people.

While antitrust cases frequently address how powerful organizations can operate as monopolies and unfairly raise prices, a company can also be accused of operating as a monopsony in which it exerts unfair leverage over suppliers, including employees.

From a basic economics point of view, concentrated employers can and do leverage their market buying power against labor, especially labor that has industry specific skills that don’t transfer well, to capture more of the surplus generated and drive down total compensation.  This means floor hospital workers like nurses are, under this theory, more likely to get squeezed than a generic business analyst who wrote SQL for six hours a day (which is what I did for years).

If this case goes to trial and the plaintiffs win, it creates a new and very powerful tool to challenge pre-existing health care consolidation.

Let’s keep an eye on this!

UPMC and labor market powerPost + Comments (32)

This is going to be geeky, even for me.

Yesterday, we talked about a new paper about the variance of the residuals after risk adjustment.  The TLDR of that paper is that risk is risky for insurers and therefore insurers will either charge higher premiums to compensate for residual risk or find ways to screen for risk on the residual.

This got me thinking last night.  I’m done with my comprehensive exams and can devote head space to fun projects again.  One of the projects that I’m revising this morning is reinsurance.  Reinsurance is a source of funds that pay for some segment of high cost claims. The ACA has an unusual reinsurance program in that it is not the only risk equalization measure.  The ACA also has substantial risk adjustment.  This produces weird incentives.

Let’s imagine a state has a reinsurance program where it pays 50% of the claims between $50,000 and $250,000.  The maximum payment is $100,000.  Let’s work through a simple example.  Imagine an idiosyncratic event such as a failed assassination attempt by a cat who was quite disappointed that you were five minutes late feeding them, and the recovery leads to a $200,000 allowed amount on the claim, the insurer pays the first $50,000, and then splits the next $150,000 in half.  The insurer will be on the hook for a total of $125,000.  This is straightforward.  Incentives aren’t particularly weird here.

Now let’s take another $200,000 claim in the ACA.  Let’s imagine that this is for maintenance therapy for someone with well controlled hemophilia.  The insurer again pays the first $50,000 on the claim, and then splits the next $150,000 so that the net spend for the insurer is $125,000.  This is just like the cat assassination scenario.  HOWEVER THERE IS SOMETHING ELSE GOING ON.  Risk adjustment comes into play for predictable expenses.  Risk adjustment is a transfer from insurers that cover populations that code as predictably low cost to insurers whose populations code as predictably expensive.  A hemophilia diagnosis is worth about 73 times the standard statewide average premium.  This means the insurer gets a transfer receivable between $400,000 to $600,000 depending on the state.

WHAT?

This is weird.

The insurer faces a $200,000 claim but gets at least $400,000 from risk adjustment!  This is okay-ish as risk adjustment is supposed to be about group fit.  But it creates a profitable residual for the insurer for this particular individual which is what the paper we talked about yesterday plays with.  Risk adjustment assumes that on average, an insurer that gets lucky with an individual with a profitable residual  will also get unlucky with a patient in the same risk adjustment bucket who costs more than the average payment.  The residuals are assumed to sum to zero over a large enough group.  But there is more.

The insurer also gets a $75,000 payment from the state reinsurance fund.  NOW THAT IS MESSED UP!

Reinsurance is supposed to eat risk.  For a profitable residual individual, the insurer faces absolutely no tail risk.  In this scenario, reinsurance is a pure transfer to the profit side of the insurer ledger without buying out risk or anything else.

This is problematic.

States that have reinsurance programs should modify their programs so that the reinsurance payment is only paying for the residual after risk adjustment.  In the scenarios I laid out  above, reinsurance would still kick in for the cat assassination attempt but it would not kick in for the individual with well managed hemophilia.  Instead, states should, for the same budget, devote more funds to conditions that poorly risk adjust or don’t risk adjust at all such as one off genetic diseases, and conditions with high residual variance like hemophilia.  Doing that reduces selection incentives within a risk adjustment category.

Reinsurance on the residualsPost + Comments (8)

Oskam et al have a new health insurance paper in the European Journal of Health Economics that looks at the risk of risk even after pretty good risk adjustment in the context of the Dutch health insurance market.  The Dutch market has some similarities to the US ACA market.   I read this paper last week and it has been deeply marinating in my head each time I’ve reread it.

We think that risk adjustment, compensatory payments to insurers that cover people with predictable high costs is needed to discourage or eliminate selection and screening incentives when we have a system of community rated and guaranteed issue insurance. Risk adjustment is a group level compensator. It pays an insurer the average incremental cost generated by a specific marker of predictable cost. However there is variance within group. Some people have cheaper than average experiences with a specific marker of predictable costs and other people have really expensive relative to the average experiences.

The easiest American example is hemophilia where the “cheap” years are $150,000 to $250,000 while the rare OMG year is equal to the pay of a competent veteran Major Leaguer. Yet risk adjustment will give the same transfer for hemophilia for both individuals. Insurers have strong incentives to try to figure out who could cost the equivalent of someone who can hit a curve reliably so that they can be avoided while the insurer would love to cover a bunch of hemophiliacs who are very unlikely to have million dollar bleeds because they risk adjust very well.

The reason is that the uncertainty about residual spending (i.e., spending net of risk equalization) differs across groups, e.g., the risk of substantial losses is larger for the chronically ill than for the healthy. In a risk-rated market, insurers are likely to charge a higher profit mark-up (to cover uncertainty in residual spending) and a higher safety mark-up (to cover the risk of large losses) to chronically ill than to healthy individuals. When such differentiation is not allowed, insurers face incentives to select in favor of the healthy….

In a third step, to quantify the heteroscedasticity of residual spending, we estimate the standard deviation (as a proxy for the uncertainty in residual spending) and the 99.5th percentile of residual spending (as a proxy for the risk of ruin) for each risk group. Variation in these measures across risk groups provides an indication of the heteroscedasticity in residual spending across risk types….

Although the mean residual spending equals €0 for both risk groups, the distributions are quite distinct. In addition to the visual contrast between the two distributions, some metrics are presented to summarize the divergence. The red [morbidity indicated — DA added ] distribution has a standard deviation of €11,378, far larger than that of the total population (€6978), while the blue distribution has a smaller standard deviation of €4646. The results for the group without morbidity are more concentrated around €0 than those for the group with morbidity, implying that insurers face more uncertainty regarding the ex-post financial result for individuals with a morbidity flag…

Insurers are in the business of aggregating individual level variance into far more stable . The more variance in a distribution, the more expensive it is to control even if the aggregate mean is the same as a lower variance distribution.

The authors also find that as the number of morbidity flags increase, the variance increases as well.  People who are very sick are both very expensive (which is fine as that is what risk adjustment and reinsurance are there for) and very variant — which produces selection incentives.  For conditions where there is a fairly large population like diabetes, an insurer that is big enough can eat enough risk to tamp down the individual level variance within a population but for conditions that are both expensive and rare, variance would still dominate a national single payer.  Insurers have strong incentives to either segment the market in some way to force the high cost high variance people to choose higher premium plans OR to avoid the segment entirely even if risk adjustment is perfect at the group level.

 

 

Risk is riskyPost + Comments (5)

JPal Poverty Lab has a recent literature review on the effect of nudges on health insurance take-up. Nudges are small changes to information, choice structures and other things that are attempts to lower administrative burdens without changing the underlying economic or legal structures. I study nudges. Several co-authors and I just submitted a grant on nudges. One of my papers was enacted into federal policy to shift the default plan for some people to get them out of strict domination. That was a nudge paper. It will lead to a few hundred thousand people seeing a different choice architecture.

However, nudges are limited.

A review of eleven randomized evaluations conducted between 2001 and 2019 that assessed the impact of various nudges aimed at increasing health insurance take-up among adults found these interventions to have a small but notable impact—both on encouraging individuals to take-up health insurance or to switch to plans with lower premiums and expected lower out-of-pocket costs to the individual.

These effects are bigger when the money is bigger:

To reduce potential choice error, consumers who were eligible for plans with higher subsidies but were enrolled in other plans were sent information regarding the higher-subsidy plans [9]. This information increased plan switching by 2 to 4 percentage points, for email-only and mail-plus-email interventions respectively (an 11 to 22 percent increase from a baseline of 18 percent). Letter recipients saved, on average, roughly $65 per year. Plan switchers saved, on average, $84 per month in premiums, or over $1000 annually, and $56 per month in reduced out-of-pocket expenses, or over $670 annually….

Choice architecture is worth doing, and it is worth doing right. Informational nudges are worth doing and worth doing right. Reminders are worth doing and worth doing right. However, they have modest effects in a complex choice space, so we should think about the choice space as a first best option and then think about nudges as second best options.

Nudges in Healthcare Insurance take-upPost + Comments (4)

Cancer statistics 2023 is out.

The incidence of cervical cancer 2012-2019 dropped a whopping 65% among women in their early 20s, the first cohort to receive HPV vaccine.

A vaccine that prevents cancer! It's amazing when you really think about it. 1/https://t.co/kmnmDwvoeS

— Noha Aboelata, MD (@NohaAboelataMD) January 22, 2023

The effort to get teens vaccinated against HPV started less than a generation ago. And it is paying off. Vaccinations provide direct protection and indirect protection in this case. Direct protection is a lower probability of a vaccinated individual getting HPV and then cervical cancer a decade or more later. Indirect protection means there are fewer opportunities for currently non-infected individuals to be infected from now vaccinated partners to then not spread it to future unvaccinated partners.

I would be surprised if we don’t see a decline in mouth and tongue cancers that can be attributed to HPV vaccination proliferation — could probably do a pretty cool analysis that leverages variation in county rate of vaccination a decade ago to get a plausible causal analysis.

Vaccines are amazingPost + Comments (63)