David Fud raised a good point in comments yesterday on how we decide to price premiums:
I have to say that I really believe that insurance premium differentiation should broadly apply to risky behaviors. Drive fast with a meter in the car? Higher automobile premiums. Don’t take vaccines? Pay a higher premium. Have a gun? Have to insure it/higher premium/no coverage for gun related health problems.
I realize that these are taboo costs in American society. However, if someone started a pool of folks who were up front about their cheaper behaviors (speeding, guns, vaccines, whatever), you would think it could bifurcate the insurance market between the two risk pools and force the increases indirectly on the more expensive pool.
What he proposes is to use risk-rating and experience rating to price health insurance. This is wicked common in the entire US insurance system.
As a teenager, I was a truly shitty driver, even relative to my peers who, as a class, were shitty drivers. My auto insurance premiums were horrendous. Now that I have not caused a claim to be filed in the past twenty years and I’ve aged into a far less likely to generate claims group, my risk and experience rated car insurance rates are far lower now than in 1998.
Car insurance is a risk and experience rated product. And we, as a society, think that this is just. The economists will argue that this is an efficient allocation as it removes moral hazard where people would engage in destructive or bad behavior if they don’t bear the full costs of bad behavior.
In health insurance, we aren’t often making that same decision that experience and risk rating is just. We, as a society, think that there is a significant component of luck that is randomly distributed that is highly correlated with costs. And if we think that luck matters a lot and that the provision or non-provision of affordable and useful insurance produces lots of suffering that we don’t want to see, then pure risk rating goes out the door. We’ve mostly made the decision as a society that pure risk rating produces bad things in the healthcare setting. So we don’t allow pure risk rating in large group health insurance via HIPAA, we don’t allow risk rating for Medicare or Medicaid except in the sense that very ill individuals may more easily qualify for coverage than healthy, low cost folks, and with the ACA we don’t allow for pure risk rating in fully insured individual and small group markets. We allow pure risk rating in products that fill gaps and live in the cracks between payment systems.
This is a political and social decision.
Going to a broader sense of risk rating in health insurance is a political and moral decision.
It could work.
It could have consequences that we either do or do not want.