David R. Henderson  

The Economics of the Affordable Car Insurance Act

PRINT
My End-of-the-World Bet with E... Munger on the Basic Income Gua...

"The Affordable Car Insurance Act (ACIA), which President-elect Donald Trump and the Republican-controlled Congress have vowed to repeal, was crafted to overcome two basic problems in the provision of car insurance in the United States. First, the costs are incredibly skewed, with just 10 percent of drivers accounting for almost two thirds of the nation's spending on cars that have been in accidents."

This is from Dean Baker, "The Economics of the Affordable Car Insurance Act," January 17, 2017.

Actually, it's not from Dean's article. I've changed a few words.

But before you go and find his article, think through whether this 10 percent of drivers accounting for 2/3 of spending is a problem with car insurance. Once you've thought through that, go and read Dean's article.

HT2 Mark Thoma.


Comments and Sharing






COMMENTS (20 to date)
mike writes:

[Comment removed. Please consult our comment policies and check your email for explanation.--Econlib Ed.]

Nick writes:

How is this sentence in the linked article even printable: "The repeal of the ACA would also end the labor-market benefits of Obamacare. With workers no longer dependent on employers for insurance, there has been a jump of 2.4 million in the number of people choosing to work part-time"

Sure, maybe some individuals found it appealing to work part-time since their 15k/year benefit remained in tact no matter what they did; but I'd bet a lot more would have preferred to not have their hours cut by their employer seeking to avoid ACA coverage requirements.But either way, that's not a labor market benefit AND it acknowledges the incentive not to work created by implicit high marginal tax rates on the poor...

Dan writes:

You do realize that there is a mandate to buy car insurance. It's exactly like ACA.

What the article gets wrong is the asymmetric information part, which would lead to a complete failure. It's not hard to identify people who are already sick and deny coverage based on health history.

If 10% of the drivers were responsible for 2/3 of accidents and you could identify them based on past behavior, then the solution would be simple. Deny them coverage (or offer extremely high insurance prices)

The result would be risky drivers not being able to drive. The health care equivalent would be sick people dying. Dean considers the the latter outcome a market failure. It's not the way economists would use the term, but otherwise, I don't see anything wrong with the article.


Thaomas writes:

Just be be clear, is the basic objection to subsidizing health insurance, or just the specific way that ACA rearranged the subsidy so that more goes to those with low incomes?

Jay writes:

@Dan

Not exactly the same at all, it's a mandate to buy liability insurance to drive on public roads. There are two parts of that sentence that are very different than the ACA and for a different purpose: to take care of negative externalities of driving with other cars on the road.

Nick Bradley writes:

[Comment removed. Please consult our comment policies and check your email for explanation.--Econlib Ed.]

Re: " just 10 percent of drivers accounting for almost two thirds of the nation's spending on cars that have been in accidents "

Well, of course, that's the whole point of insurance. You collaborate to spread the risk, and hope that just some of you have accidents, so the average cost is small.

… think through whether this 10 percent of drivers accounting for 2/3 of spending is a problem with car insurance.
David Henderson does not mention a time frame, but I think this only makes sense in some given time frame. In a short time frame, one day, I guess maybe 0.01% of us account for all car accidents and thus all spending. Whereas in a long time frame, a century, 97% of us account for all accidents and spending.

Insurance is about spreading risk to entities (wealthy agents in a short time frame or populations over a long time frame) which have enough fat to absorb the variability, isn't it?

MikeP writes:

Just be be clear, is the basic objection to subsidizing health insurance, or just the specific way that ACA rearranged the subsidy so that more goes to those with low incomes?

Why must those be the only possibilities?

My first basic objection to the ACA is that it mangles the insurance market of the >75% of people who can get insurance in a free market for the sake of the <25% who cannot.

My second basic objection to the ACA is that its entire structure is built upon the lie that the cheapest way to provide health care to those who cannot afford insurance is to subsidize insurance for them.

That is not a complete list.

ZC writes:

@Richard O. Hammer

Hmmm...looks like you don't know much about healthcare spending. Health expense aren't random, totally unpredictable events that will revert to some mean over a long enough period of time. A very small percentage of the population incurs a disproportionately large percentage of public health care expenditures. In many cases, those individuals can be identified beforehand and yet, under the law, they can't be charged higher rates. so it's not truly an insurance market. That's the rub. That's a big part of the frustration from many intelligent people about the debates surrounding ACA and healthcare financing - various parties use terms erroneously to hide their true intentions, calling things like prepaid healthcare insurance and conflating 'insurance' with access to health care and conflating healthcare spending with health.

Mark Bahner writes:

Offered without comment (for now) ;-):

Lifetime distribution of healthcare costs

GregS writes:

I think what David’s getting at is that your insurability is based on your *expected* cost. If I’m reading the Dean Baker article right, he’s talking about disparity in *actual* cost. This is kind of a bait-and-switch. Baker is (I think) incorrectly implying that the most expensive 10% of patients have an *expected* cost of $54,000, which indeed would probably make them uninsurable. When people use actual after-the-fact spending figures to imply that a large segment of the population is uninsurable, they are exaggerating (perhaps even making up) a potential market failure.

It’s like grouping together a bunch of people who have already had a heart-attack (with all the associated medical expenses) and claiming that these people are uninsurable because they have such a high average expense. To do this correctly, you’d need to do some sort of predictive modeling/scoring to identify different tranches of “heart attack risk”; identifying after-the-fact those individuals who are 100% likely to have had a heart attack is cheating. It’s conceivable that *everybody* is insurable under an expected risk/cost measure, but a very large fraction of the population appears uninsurable under the actual cost measure.

acarraro writes:

Aren't medical expenses serial correlated? I would imagine car insurance premium are somewhat serially correlated, but to a much lower extent.

There are many chronic health conditions which last for many years. Would insurance companies be liable for all treatment costs emerging from all conditions that start while under insurance?

As far as I understand this is not generally the case: an insurance is only liable for treatment in the insurance period. If insurance companies can change premiums, wouldn't they raise premiums to the point that the insured is paying the bulk of the cost?

I can see this is not likely a problem for people that already are part of a pool through their employer (and the mandate that employers cannot fire people for sickness and that they must provide health insurance), but how is this going to work for people outside the pools...

GregS writes:

To acarraro’s point, I think insurance should pay out (or promise a future payout when bills come due) as soon as the patient gets a costly diagnosis. I believe this is what you’re describing. The insurable event *should be* that your health status changes, not that you got the surgery that a five-year-old diagnosis said you’d eventually need.

Your auto bodily injury coverage basically works this way. If you hit and injure someone, they can be incurring medical bills for years. We actuaries have lots of numerical methods for “developing” these claims to their ultimate value. Anyway, your insurance policy will cover those expenses regardless of the year the medical bills come due. The loss is *incurred* the moment you injure someone, and the insurer holds that liability.

I’ll concede that there’s a bit of an adverse selection problem; you might suspect you have diabetes, then obtain health insurance and promptly collect when you get your diagnosis. But this could be overcome with longer coverage periods, guaranteed renewals, etc. I don’t want to imply that it’s an easy problem, but I think economists tend to exaggerate the degree of market failure intrinsic to health insurance. I'm pleased to see David push back against this tendency.

Hazel Meade writes:

The result would be risky drivers not being able to drive. The health care equivalent would be sick people dying.

No, the health care equivalent would be sick people going bankrupt due to unpaid medical bills.
They aren't going to be denied treatment. Insurance protects you from bankruptcy, not illness.

Peter writes:

Since this is about comparing health to car insurance people might find this article interesting:
Josh Barro: Here's Why Health Insurance Is So Weird. Nov 2013
http://www.businessinsider.com/heres-why-health-insurance-is-so-weird-2013-11

Also interesting, in my opinion (given the type of health care insurance people like David Henderson favor, namely high-deductible catastrophic insurance which is what you can buy in the individual market):
Josh Barro: Here's The Real Reason People Hate Their Individual Market Health Insurance. Nov 2013
http://www.businessinsider.com/heres-the-real-reason-people-hate-their-individual-market-health-insurance-2013-11

Seth writes:

The skewed costs is the very reason for insurance. If costs were not skewed, it would not be insurance. It'd just be a cost.

We don't have oil change insurance because every car needs one periodically, so what would be the point of that?

It goes back to folks basic misunderstanding of what insurance is.

Peter writes:

[Comment removed. Please consult our comment policies and check your email for explanation.--Econlib Ed.]

AS writes:

[Comment removed. Please consult our comment policies and check your email for explanation.--Econlib Ed.]

Jon writes:

One issue with the ACA was its focus on "providing health insurance". The real problem was not about demand for healthcare; it was not about abstractly expensive health insurance. The core issue was healthcare costs. In this respect, the ACA contained only highly speculative cost reforms that didn't work.

- Supposedly insurance would decrease use of emergency care (it didn't)
- Supposedly digital heath records would reduce cost of care (they didn't)
- Supposedly more routine care would improve overall healthiness (it hasn't).

But even if these measures had made an impact, they ignore the structural problems:

- Drug approvals by other major developed country regulatory bodies are not recognized and duplicative clinical trials and applications must be completed in the US

- Importation of foreign drugs is illegal, enabling price discrimination on a massive scale

- Punitive licensing rules which bar foreign trained doctors from acquiring a license in the US unless they redo their residency, and a residency admissions rules which penalize foreign medical degree holders.

- A board certification regime which requires potential doctors to accumulate hundreds of thousands of dollars worth of debt, then undergo 4+ yrs of minimum wage work during which time they may be expelled or otherwise forced out with no marketable skills because medical training is primarily rote memorization of diagnostic and care procedures which cannot be performed unless board certification is obtained... for which incurring such risks doctors-in-training demand substantial potential future returns if they succeed.

- Exemption from anti-trust laws which allow hospitals to set minimum prices attending physicians will charge for their services.

- Medical payment systems built around elaborate catalogs of services which encourage doctors to employ 3:1 billing specialists on average to hack maximum revenue from each service rendered.

- Pervasive price discrimination in medical services. Pure High deducible plans are popular simply to have benefit of pre-negotiated prices and someone to bargin the price down on your behalf. But the ACA barred such plans for lack too few bundled freebies.

Comments for this entry have been closed
Return to top