Paying for medicinal weed

This is an amazingly dumb and naive question as my state will not be getting on this train anytime soon. Pennsylvania just legalized medical marijuana:

Pennsylvania governor Tom Wolf today signed a bill that legalizes medical marijuana in the state of Pennsylvania. The state is the 24th to enact a legal medical-marijuana program, and although it could take regulators up to two years to draft rules for retailers, a provision in the bill lets parents administer medical marijuana to their kids immediately. As Democratic senator Daylin Leach, who co-sponsored the bill, told the Associated Press, “Marijuana is medicine, and it’s coming to Pennsylvania.”

Like New Yorkers, residents of Pennsylvania will need a prescription to obtain medical marijuana, and they’re prohibited from smoking or growing it. They can, however, take it in pill, oil, vapor, ointment, or liquid form. The bill also puts a system in place for tracking marijuana plants and certifying physicians, growers, and dispensaries.

My question is does insurance come into play at all in the other states with medical marijuana laws? I am assuming prescriptions of some sort are needed, and insurers may pay for over the counter medication if there is a prescription.

Or is this just cash and carry?

Hacking the Exchanges

This is a non-ironic tweet from last night:

I spent some time last night playing the the actuarial value calculator and found a corner case that has significant negative public policy problems and a probable flaw in the CMS model.

This is a hack that companies can use to spam the Exchanges in states with very light regulation.  Here is the 2017 AV calculator to check my work.

A deductible can be thought of as a first dollar 100% co-insurance where the patient is responsible for all of the contracted rate expenses.  It is logically equal and it should be mathematically equal to a 100% co-insurance rate.  Logically a deductible of X which is also equal to the same out of pocket maximum  is the same thing as a 100% coinsurance rate with an out of pocket maximum of X.

The policy problem is the definition of substantial difference for plan differentiation allows for the addition of a plan design by a company if there is a significant difference in deductible.  going from a $3,750 deductible to a $0 deductible qualifies as a significant difference.     Companies can submit their first plan with a deductible only cost sharing design and then submit the same exact plan with a 100% coinsurance design to the same out of pocket maximum.  For plans that are not the first or second Silver there is no value from the point of view of the company.  However if this is done for the plan that is the first Silver, the cloned plan becomes the second Silver at the same price point.  It is an illusion of choice.

Health Sherpa search toolsI believe the 100% co-insurance design would significantly outsell the 100% deductible design because the current set of decision support tools prioritizes low deductible plans.  This is an Search exploit as well as a Subsidy exploit.

Ambetter in Chicago actively spammed the Exchanges with functional isomorphs of their Silvers so that cost-sensitive buyers could be fully subsidized on an Ambetter product only and then pay significant incremental premiums to go to any other insurer.  However there was at least some gap between the first Silver and the second Silver.  This plan design strategy eliminates that gap.  The AV calculator allows for this plan design strategy to occur.  There is a single “Are you sure” message box to ward against data entry errors on 100% co-insurance but no hard prohibitions.

plan design hack


This hack would not work in all states.

California with their active purchaser model has a very high substantially different standard.  A cloned plan design fails that standard miserably.  Washington state only allows a maximum 50% co-insurance rate.  My state’s regulators have broad discretion to laugh at us and reject one of the two plan designs.  However not all states have active or empowered regulators.  If the plan design meets model requirements, they will approve anything.  This is a special concern for the states that have outsourced all regulatory authority for the individual market to the federal government.  Those plans have to meet very minimal standards, and these plans will meet those standards.

Finally on a technical note, I think there is a problem in the AV calculator as the co-insurance design has a lower actuarial value than the full deductible design.  In my opinion, the model should produce the same actuarial value with these two inputs.

Plan design and cost sharing choices

This is a follow-up from this morning’s post on cost sharing.  I was curious what it would take to create a 70% actuarial value plan where there was only one form of cost sharing and no frills to it.

I used the 2016 CMS AV Calculator for Exchange as my tool.  You can play with it as a macro enabled Excel file.

Creating a deductible only plan was fairly simple.  The insurance would pay nothing out until the person spends $3,725 on cost-sharing eligible services.

The co-insurance routine was a bit harder to build.  There are a wide ranges of co-insurance rates that could be chosen.  For simplicity sake’s I chose a $0 deductible and a 40% co-insurance rate to start with.  That failed.  I could not design a Silver plan using only co-insurance at 40% before I hit the maximum out of pocket constraint of $6,850.

A 50% co-insurance rate creates a 72% Silver plan while a co-insurance rate of 53% creates a 70% Silver plan.  The out of pocket maximum for these plans are $6,850.

I will not even try to describe what I had to do to get a co-pay only plan.  Below is a screen shot

Co-Pay only Silver

Again, maximum out of pocket is $6,850 and most of that will be paid for by people who have inpatient admissions and high end specialty medications. The PCP co-pay is high enough that no one will ever go see their doctor for little things.

The problem is not the deductible.  The problem is low actuarial value of coverage combined with the high cost of services.

Exploding Deductibles

The Kaiser Family Foundation has put out a report


This looks like a crisis as deductibles have more than doubled. However percentages can be funny things, and thankfully the source data was in the report and I could play with it in a spreadsheet:. My big question was why did KHN break things out by type of cost sharing. So I combined cost sharing types and then got total benefit growth and total cost sharing growth using the same data but spliced a bit differently.

Cost Sharing

This is a different story.

There are two big things. First the cost of the total benefit which is cost-sharing plus what the insurer pays out has increased faster than the economy. Secondly, the actuarial value of used coverage has stayed fairly constant with a slight trend towards a higher proportion of all costs paid for by cost sharing.  The red line on the first graph as it increases at a higher slope than the blue line informs the conclusion on the right hand side graph.

From here the problem is not that deductibles are too high because if we hold cost sharing percentages constant but still seeing significant cost growth deductibles versus co-insurance versus co-payments is mostly an allocation and incentive issue.  The problem is that cost growth is too high and everything else flows from there. Larry Levitt made the point on the KFF chart that the interesting thing was the switch from quantity variant co-payments to total price variant co-insurance is a big deal as it does change incentives.

There are distributional concerns about cost-sharing choices.  I’m probably in the minority in that I think deductibles are preferable than co-insurance and co-pays for a given percentage of cost sharing that must be borne by the entire user pool:

Deductible plans favor the sickest people as the low utilizers pay for almost all of their care via deductible cash. That means the proportion of the pool’s individual responsibility amount is borne by healthy people.

Co-pay only plans favor people who use highly concentrated cost services. A co-pay does not differentiate between a specialist visit with a contract expense of $200 and a specialist visit with a contract expense of $600. It is the same fee. So people who use very costly services but only rarely are best off. People who use a lot of fairly low costs services on a regular basis pay more proportionally.

Co-insurance only plans favor low cost utilizers. They are not paying full price via their deductible, and unlike co-pays, the individual cost per unit matters.  Finally, No Use Nora is extremely valuable to the insurance company and the rest of the pool as she is fully cross subsidizing everyone else for this time period no matter how her benefits are built….

The other key insight of this exercise is that the non-covered actuarial value of a plan will get paid somehow by someone. The question is who pays and how much?  If the objection to a deductible level is the size of the level, the problem is not the deductible, the problem is the low actuarial value of the plan.

The problem is not the cost sharing choices per se.  The problem is medical care costs too damn much per unit of care.  Almost everything else derives from that problem.

Exchanges and population health management and the Health and Human Service (HHS) research group ASPE just released a summary of the shopping tendencies during the 2016 Open Enrollment Period for states.  There are lots of very interesting nuggets in it, but one table leaped out at me as it plays to my hobby horse of churn analysis:

Churn ASPE

How the hell do you do population health management on a population that is this transient***?

The short answer is you really don’t.

The long answer is that it is complicated. The stickers are more likely to be sick than the switchers so there is some value of population health management for the people who are actually sick and driving the expenses. But realistically, the Exchanges make up 4% of the US population. The individual market has historically been a high churn market where people stay in it until something better comes along. The Exchange market is similar to the pre-Exchange individual market in that aspect. The population health management focus of the ACA is far better targeted at the Medicare population as Medicare is where the probability of chronic conditions (ie being old) is very high over a very large population. That market has a good bit of churn but it also has mature risk adjustment systems in place with experienced players. Insurers will still see a good amount of churn but the process smooths it out plus the benefits of active population health management are fare more likely to be at least a break even proposition for the insurer in any given year.


*** I have a question into ASPE if they are including people who move from 2015 Mayhew Silver HMO to 2016 Mayhew Silver Narrow EPO in the mover category?  I think they are.  If that type of movement is being accounted for, the population is not quite as transient as the population health manager for Mayhew Silver Narrow EPO sits on the bathroom side of the 13th floor and the population health managers for Mayhew Silver HMO sit next to the kitchen on the 13th floor.