Last week, Delta Air Lines announced it will charge employees who participate in its account-based health insurance program an extra $200 per month for insurance if they are not vaccinated against COVID-19. This is the first major organization I've seen announce that it will link vaccination status to insurance premiums, but the idea has been bubbling up elsewhere, including in a New York Times op-ed from Elizabeth Rosenthal and Glenn Kramon, floating but not-quite-endorsing the idea that health insurers could raise premiums for the unvaccinated.
There is a reason Democrats spent decades fighting to make it so insurers generally can't set the price for insurance based on a customer's risk of needing healthcare. "You shouldn't be charged more for a preexisting condition" has been a core health policy principle for a reason. We shouldn't roll it back out of zeal to induce vaccination.
Normal insurance is priced to risk, but medical insurance is not normal insurance
Property and casualty insurance premiums are generally set to reflect your likelihood of making insurance claims. If your house is on the ocean, you'll pay more for wind coverage. If you have a swimming pool, you'll pay more for liability. If you get a lot of speeding tickets, you'll pay more for auto insurance.
Pricing in these markets is not totally unregulated but this is generally how property and casualty insurance works: If you're riskier, you pay more.
That's considered fair because people have a lot of control over how much risk they take on. And if you didn't let insurers price this way, the markets would become dysfunctional without the addition of mandates and subsidies. (Indeed, because auto insurance is compulsory, states can and do regulate auto premiums more intensely than they do homeowner's premiums.)
Medical insurance is about the healthy subsidizing the sick – for a reason
There is lots of information available to insurers about an individual's likelihood of making claims for medical care. Some people are known to be expensive, for example because they have chronic illnesses. And while you can sell your oceanfront house if it's too expensive to insure, you can't change your body once you've had cancer.
So as a matter of policy and values, we consider it unacceptable for medical insurance to be priced based on known risk. As a matter of fairness, people like cancer survivors who are known to have high medical risk need to be offered insurance that costs less than their likely claims; as a result, people known to have low risk must pay premiums that are significantly higher than they would owe in a purely-risk based market.
That is, while property insurance is used to spread the costs of unknown events in the future – we know where hurricanes are more and less likely to hit, but we don't know exactly where a hurricane will hit each year – health insurance is also used to transfer costs from the known sick to the known healthy.
Of course, an insurance market where premiums aren't tied to likely claims is likely to be pretty screwy. People who know their risk is low are likely to find the pricing unattractive. And if they opt out of the pool, insurance gets more expensive for everyone, and fewer and fewer people get covered.
One way of dealing with this is that employer-based insurance is heavily incentivized through the tax code. Compensation in the form of health insurance is not subject to income tax, but employers must offer insurance to all similarly-situated employees, pooling risk across healthy and sick employees.
The Affordable Care Act sought to improve the availability of insurance to those who don't get it through work with a system of subsidies and regulations. Insurers were also told they had to offer comprehensive insurance to anyone who wanted it, and with limited exceptions, they were told they couldn't set prices based on medical risk.
That's one of the key achievements of the ACA: insurance prices not tied to your preexisting conditions, except age and smoking. Do we really want to add another category to that?
Vaccination status is not the only medical risk factor individuals can control
Smoking is both a significant risk to individual health and heavily socially disfavored, which is a reason public policy allows for charging smokers more. Not being vaccinated against COVID is also a significant risk to individual health and is socially disfavored. Should premiums also vary with vaccination status?
The problem with going down this road is, what else should be a matter for medical underwriting?
Should insurers be able to charge people more if they drink too much? If they are overweight? If they work in an occupation that puts them at a high risk for accidents? If they engage in high-risk sexual behavior?
I think we have had good, considered reasons for not going here with medical insurance. Questions about which medical risk factors people "should" change are inherently value-laden. Losing weight is harder for some people than others; doesn't that make it wrong to price based on obesity? But quitting smoking is also harder for some people than others, and we price based on that.
The other reason not to set insurance prices based on vaccination status is that it could induce some people to forego health insurance. The ACA was designed to try and get as many people on insurance as possible, so introducing a penalty that disincentivizes people from seeking coverage would be antithetical to the purpose of the law.
Financial penalties are fine, just don't tie them to insurance
In announcing its intention to tie health insurance premiums to vaccination status, Delta emphasized the high cost of COVID-related hospitalizations. Every time a Delta employee gets hospitalized for COVID, it costs the company tens of thousands of dollars.
And it's fair enough that that bothers Delta. But that complaint doesn't capture the whole universe of harms to Delta when its employees are unvaccinated.
An unvaccinated employee who gets sick might, in unusual cases, have an expensive hospital stay. More often, they will just miss work, which is also costly to Delta. They might give COVID to other Delta workers or to customers, including children 11 and under who aren't yet eligible for the vaccine.
All these risks apply to Delta employees even if they do not participate in the company's insurance plans and won't be subject to a premium increase. So why would Delta's non-vaccination punishment apply only to the employees who use its health insurance?
If Delta wants to incentivize non-vaccinated workers to get vaccinated, it could simply dock their wages – which would affect employees regardless of whether they get ther insurance through the company.
This analysis applies at the society-wide level, too. Most insurers are not even in a position to penalize plan participants who don't get vaccinated because the participants don't pay their own premiums at the margin; their employers do, or the government does.
There are ton of other financial touchpoints between individuals, their employers, and the government. If we're looking for somewhere to break out a financial stick, it doesn't have to be insurance premiums. So COVID doesn't provide a good reason to go back on the project of de-linking medical insurance premiums from medical risk.