Where the competition lies

Paul Michael Cohen is running some interesting data on Exchange competition and premium pricing. He has tested two theories of competition. Does more insurers on a given exchange lead to lower average premiums? Does more insurers on an Exchange lead to lower minimum premiums (ie the Second Silver benchmark?)

My first analytical approach was to plot the mean monthly premium in each state (by coverage level) against the number of unique insurers in each market (see R code on Github). A basic linear regression of premiums by number of unique insurers and by coverage level revealed no significant relationship….

Average premium rates are insensitive to the number of insurers on the Exchange. I am not too surprised.

However, it is possible that the important unit of analysis is not the mean premium in each market, but instead the minimum premium in each market….
There is a significant negative relationship between increased competition and health insurance premiums for all coverage levels in 2014 and 2015. The regression model (collapsing 2014 and 2015 rates) suggests that each incremental market entrant corresponds to a 3% reduction in monthly premiums. Moreover, this trend continued as more entrants entered the market in 2015. However, linear regressions by year suggest that the size of the competition relationship is decreasing: the effect of an incremental insurer on premiums was -3.7% in 2014 and -2.8% in 2015…

This result makes sense to me as the relevant problem that insurers, or at least insurers in my market, have been trying to solve is who offers second lowest priced Silver.

The subsidy structure of the Exchanges determines the nature of the competition. Subsidies are determined by the price of the second least expensive Silver in a given market. A family is expected to contribute a certain percentage of their income as determined by the Federal Poverty Line. A family at 101% of FPL is expected to pay 2% of their income, and a family at 399% FPL is expected to pay 9.5% of their income for health insurance. The gap between the expected contribution and the price of the second least expensive Silver is the subsidy level. Most people on Exchange are premium price sensitive. These two facts lead to significant segmentation in Silver;

there is a strong incentive for insurers to offer at least a Silver plan that is either the cheapest two Silvers or very close to the subsidy cut-off.

This segment in a competitive market should see a cluster of plans that are at the subsidy line plus or minus a couple percentage points. These plans are the first segment. They tend to be very restrictive in all modifiable aspects. HMO’s with gatekeeper and strict authorization processes are likely to be here while open access PPO networks are unlikely to be in this segment. The networks will tend to be very narrow as the pricing model is Medicare plus a small kicker….and insurance companies are avoiding the high cost providers if they can. These are the super narrow networks where the goal is to get a Silver plan that is either top 2 or really close to top 2 in pricing. They are aimed at people who are getting subsidies are extremely aware of every additional dollar they have to spend on monthly premiums. If we had a public option plus 5% scheme in place, it would fall into this segment….These segments were haphazardly defined in 2014 as companies were mostly shooting blind on both what the risk pools looked like and what their competitors’ strategies are. 2014 is a successful beta testing year. I think the Silver segmentation will be much clearer in 2015 and very obvious in 2016 as more data and experience comes into play.

Most of the customers are in the Subsidy Silver segment. That is where more companies are shooting for. They want to get either a Top 2 Silver plan or something within a couple of bucks of the Top 2 Silver plans where marginal differentiators (networks, customer service, reputation for not fucking people over etc) can push people to the slightly more expensive product.

There are two mechanical points to note.

From a mechanical point of view, once a basic narrow or super-narrow network is set-up, approved and priced, expanding offerings is fairly simple. The insurance company knows that if its super narrow network meets both basic regulatory approval and marketing approval, then throwing another 20% of the entire provider population at the network will also meet regulatory approval and marketing approval. It knows that removing a gatekeeper requirement and converting an HMO plan to either an EPO or PPO configuration will also meet approval. It knows the benefit design can be built into the claims system without too much hassle.

Secondly, a company will offer the same network, the same plan type (HMO, EPO, PPO etc) at most if not all benefit levels. The difference at that point is benefit configuration. A Bronze plan on Mayhew Super Narrow HMO could get a $5,500 deductible while the Platinum on Mayhew Super Narrow HMO gets a $300 deductible. The basic architecture of a Second Silver competing plan will propogate to all metal bands.

Once a Second Silver competing plan is built on the narrowest and tightest specification possible, the marginal cost of loosening constraints from a system architecture and system plumbing perspective is fairly low. Loosening constraints leads to higher prices for three reasons. The first is the sickest people will often choose the broadest coverage. The second is PPOs and indemity plans have minimal constraints on seeking expensive care while thirdly, all the expensive providers are in the network for routine care.

The cost of developping broad plans with minimal constraints are fairly low. These plans are reasonably well defined and generic so the pricing converges among most competitors becaue there is no differentiation between insurers. A broad PPO is a broad PPO. As Austin carroll notes, this is expected:

I think that premium costs are likely being lowered by narrowing networks and making deals, and that’s more likely to occur with more insurers.

But the incentive to make deals only applies to the segment of the market where the subsidies drives cost consciousness as everywhere else the marginal cost of offering a broad product is fairly low, and it provides some market segmentation for the insurers.