Thursday, September 26, 2013

Smarter than your average bear

My understanding of Obamacare was that if I got sick or injured and didn’t have health insurance, I would immediately be able to enroll via the exchanges. It turns out this isn’t true. The Wall Street Journal quotes the following explanation from a Salon piece:

One thing these conservatives are telling young people is that Obamacare will be different [from the current system]. Because it guarantees that people with preexisting medical conditions can buy insurance, they claim opting out carries no risk. Get sick or injured? Then you can reconsider your decision to opt out.

But this is a falsehood.

For the first year, Obamacare will have an unusually long open-enrollment period. It starts on Oct. 1, but eligible individuals can sign up through March, even if stricken by accident or illness. After that, though, anyone who decides, or is persuaded, to "skip" Obamacare will be as vulnerable as I would've been if I'd never applied for insurance and dropped that check in the mail. They'll be locked out of the system until the next open enrollment period begins on Oct. 14, 2014. That open enrollment period will last just 53 days. Break a leg or develop a serious illness in the interim and you get to choose between paying tens of thousands of dollars out of pocket or suffering for months and rolling the dice with a delayed diagnosis.

I hate Obamacare with a passion but the existence of a limited open-enrollment period each year means the law is considerably less stupid than I thought it was. It’s also a good reminder to me that when I hear about someone or something on the other side being unbelievably dumb, it’s probably best not to believe.

The WSJ argues that even this small window for enrollment doesn’t make buying health insurance a rational choice for “the ‘young invincible’ demographic.” Perhaps not but it does make forgoing health insurance more risky. And if not enough young, healthy people enroll, the government can up the pressure by changing the open-enrollment period to only once every two years (an idea I believe I first read about ages ago at JustOneMinute).

10 comments:

E Hines said...

Beutler has his anecdote; I have mine. Wife, breast cancer, no health insurance (by choice). We paid for the initial exam, the resulting biopsy, and the resulting bilateral mastectomy with our own resources. Which we had because we'd taken care of our money all our married lives (which decision made our bet even better). At the time, we were in the region of the Federal Poverty Guideline on family income.

But neither Beutler's nor my anecdote prove much of anything. Insurance companies make (or used to make, under Obamacare they're not insurance companies but privately funded Federal welfare agencies) their money on the likelihood of having to make a payout and being able to charge a commensurate premium. I like their odds. Buying insurance is a bad bet for the buyer if he's healthy.

Neither the existence of a signup period nor its duration have anything to do with those odds.

Eric Hines

Elise said...

I think about insurance - real insurance - a little differently. It's about spreading the risk so that if I'm the unlucky one, it doesn't cost me an arm and a leg. It should be a good bet for both sides - the insurer and the insured - since the premiums should reflect the likelihood the insurer will have to pay. The insurer's profit is the premium I pay to spread the risk.

Like your wife, I had breast cancer years ago. We could have paid for the standard treatment without health insurance but I ended up needing some cutting edge treatment that cost a fortune. This is why I like catastrophic health insurance: I figure out how much loss I can handle and buy insurance for the rest.

A limited-enrollment period affects the calculation because some part of that calculation should be: If I don't have health insurance and I get unlucky, what is my maximum exposure? The longer the time between enrollment periods, the more an illness can potentially cost me, in both dollars and risk of bad outcomes.

E Hines said...

We think about insurance in much the same way. We just differ in the risk we're willing to run. The enrollment periods only impact that; they have nothing to do with likelihood of payoff given a policy.

The longer time between enrollment periods certainly impacts the Expectation of the non-buyer's loss. But two times (e.g.) nearly zero still is almost zero.

It's a good bet for the buyer only if the buyer feels the need to lay off some of the risk. That's no knock on those who do; differences are what makes a market.

Glad you're a survivor. The alternative wouldn't be as much fun.

Eric Hines

Elise said...

First, I appreciate your being glad about my continued survival although I feel compelled to point out that we don't, actually, know that this is true:

The alternative wouldn't be as much fun.

:+)

In a side issue, I hate being referred to as a "survivor" for reasons that are beyond the scope of this comment and this post.

I'm no longer clear on the whole risk level multiplied by time elapsed thing. If 1% of the people my age and health status become seriously ill or are seriously injured in a year, that means 2% of them will be in the course of 2 years. But does that mean my odds increase to 2% over the course of 2 years? At a casual glance it seems like they should but there's a little voice in the back of my head telling me that at some point I learned differently.

Agree that whether health insurance is a good bet for the young and healthy depends on their tolerance for risk, both financial and physical. However, a free-wheeling health insurance market should be pricing policies for the young and healthy so low that it's always a good bet for them. Not that we *have* a free-wheeling health insurance market or have had for some time in most (perhaps all) States but that's how it should work.

E Hines said...

I'll take the bet on the truth of your so-far survival. The last AI that could hold a serious conversation (though she was a bit of a prude) was the Rogerian psychoanalyst Eliza, which (who?) ran under DOS.

If 1% of the people my age and health status become seriously ill or are seriously injured in a year, that means 2% of them will be in the course of 2 years.

A quibble, because the distinction is trivial for this sort of discussion: no, it doesn't. It means the likelihood has risen to 2% of them. And it doesn't quite get to that. Assuming all of the individuals in the population are independent of each other (and that means more than just unrelatedness; it requires that a set of factors outside the scope of the investigation don't impact groups of them similarly), the arithmetic is to add the probabilities (.01 + .01 in the present example), and then subtract off the (assumed) remaining overlap between the two years' groups: .01 * .01. Thus, the "true" likelihood is .0199.

[D]oes that mean my odds increase to 2% over the course of 2 years?

No, but as above, the distinction is trivial in this context. You're an individual. You might not even be part of the population the sample from which the (1%) estimate was drawn purports to represent. There's also the likelihood the sample itself does not represent the population, commonly represented with a line like p<.05. That .05 is the estimate of the likelihood that the sample does not represent the population.

These are subtle distinctions, though; it's reasonable in this kind of discussion to suggest that if a population has a 1% chance of getting sick in a year, your odds are 1%. And if it's 1% in one year, then it's 2% over a two-year period.

As to a free-wheeling market driving down the cost of an insurance policy to its lowest economically feasible level (that phrase is deliberate) producing a good bet even for the young and healthy, no, it doesn't (discounting risk tolerance).

The lowest true level for a premium, assuming no friction, would put the premium at a level that exactly equals the expected payout across the population of those buying insurance (eliding the question of whether these all truly look alike from a risk perspective--let's assume, heroically, that they do) over the time frame which the policy covers. If the probability of a payout equaled 1%, then the lowest premium for the risk transfer (assuming all of the risk is transferred; transferring less than all of it--deductibles, co-pays, etc--just complicate things) is 1% of the value of the payout divided by the number of folks who bought that policy (further divided by the number of payments required in the period and ignoring the cost of money over the period).

Now go to your favorite casino and throw down $1k on 00 at the roulette wheel. The odds of getting your 00 are 35:1 (we're assuming a fair wheel). You'd expect to collect 35 stacks on that success, yes? No. The casino takes a cut on every winning bet--that's how it makes money. If all it got was the losing bets, it would only break even, and it would have no money for expenses and profit.

So it is with insurance premiums, even in a truly free market. The insurer pluses up the premiums it must charge to cover its own house cut--its own expenses and profit--to reach its economically feasible level.

Insurance and casino gambling are always bad bets, for everyone, when we ignore risk tolerance. For the ones unwilling to run the risk, a free market, though, provides the least-bad bet.

Eric Hines

Elise said...

I understand that the insurance company takes a cut but I don't think that makes insurance a bad bet. We pay a fee for everything we value and if I value lack of risk, I'm willing to pay for it.

E Hines said...

It's a bad bet because you're not getting full value for your premium--that house cut. It's the extra you pay for the peace of mind you get from laying off some of your risk.

Like I said, it's the differences that make a market. I like the insurer's odds that they won't have to pay me--those are the odds I won't get sick.

Eric Hines

Elise said...

But my point is that doesn't make it a bad bet - actually it doesn't make it a bet at all. I'm paying for something I want - security, peace of mind. That has actual value for me, just like food or the experience of watching a movie or (for some people) the entertainment value they get from spending $20 playing the nickel slots.

To me, saying "this is not something Eric values" is far different from saying "this is a bad bet".

And, to go back for a moment:

I'll take the bet on the truth of your so-far survival. The last AI that could hold a serious conversation (though she was a bit of a prude) was the Rogerian psychoanalyst Eliza, which (who?) ran under DOS.

Wait, what? I can't quite figure out what AI has to do with my comment. And I think it's "which" - unless and until something like Eliza can tell me differently.

E Hines said...

From early in the thread: ...I feel compelled to point out that we don't, actually, know that this [my surviving] is true....

Thus, your not surviving means I'd be conversing with an AI on your blog. Or, I misunderstood your meaning, since my take centers on the square brackets.

That, which, who knows? My mother was a high school English teacher, and she never could get that straight for me.

Since you bring it back up, peripherally, shall I take you for a prosperer? If that term approaches your reason for disliking "survivor," then I think I get your rationale. If not (or even so), I'll drop the subject unless you want to post on the matter some later time. And I'll drop the "survivor" bit, too; although it's an old habit. You may need to kick me once or twice to remind me.

Back on topic: Of course "this is not something Eric values" makes it a bad bet. Doesn't the Universe center on me?

Eric Hines

Elise said...

Thus, your not surviving means I'd be conversing with an AI on your blog. Or, I misunderstood your meaning, since my take centers on the square brackets.

What we don't know the truth of is whether the alternative to my surviving wouldn't be as much fun as my surviving has been. I'm kind of hoping the alternative to survival is at least as much fun - and the older I get, the more I hope it.

shall I take you for a prosperer? If that term approaches your reason for disliking "survivor," then I think I get your rationale.

You currently surmise that I don't much like that one either. As for kicking you, I won't - outrage at common terms and well-meant statements is no longer part of my repertoire. I may restate my objection if appropriate but no violence - physical or verbal - will ensue.

Back on topic: Of course "this is not something Eric values" makes it a bad bet. Doesn't the Universe center on me?

Don't be silly. It's me the Universe centers on.