Economics of Auto Classification

[cgott42]

Can someone explain what they mean on the top of pg 66 - that if rate classification was removed and both high and low risks paid he same rate per unit of coverage - that the extra coverage purchased by the high risks (as they are now getting a bargain) would be inefficient b/c "the value to buyers of the extra coverage purchased will be less than its marginal expected cost. Otherwise, they would have purchased the extra coverage prior to the restriction and the rate reduction"

I don't understand as before the rate reduction (higher rate) it wasn't efficient to purchase more (marginal expected loss pmt < rate), but now with the lower rate it is efficient because the rate is less than the marginal exp loss pmt

[Pokeytax]

It now benefits the high risks to buy more coverage, because like you say, it's worth more to them than they pay for it. You could say that it's efficient for their budgets. But it isn't broadly efficient, or societally efficient, because what they pay doesn't reflect the full expected cost of providing coverage and sufficiently deter high risks from driving.

Before, it was worth V to them, the actuarial expected cost was C, and the price was set at C, so

V < P = C,

and they don't buy coverage because V < P.

Now, the price is set below V and C, so

P < V < C

and they buy coverage because V > P.

[cgott42]

I see from the perspective of the high risk (H) why he would purchase more coverage, however why is this not efficient for society?
The reason given in the paper assumes that before the change was 100% efficient, thus any change in coverage by definition must be inefficient (of else they would have done so before).

I could understand the additional coverage being inefficient (as it causes more high risk driving, bringing up the price for everyone - as they mention).
I just don't understand why it's inefficient b/c "if it was efficient they would have done so before" - at the old prices the lower amt of coverage was efficient, and now at the new price, the new amt of coverage is equally as efficient

[cgott42]

Likewise, they explain that it's inefficient b/c the value to the high risk driver purchasing the extra coverage at the new (lower than what he was previously paying) rate is less than the expected marginal cost.

Isn't the value higher than the cost? i.e that's why he's buying it b/c really he shoudl be paying $2 per unit of coverage (if based on actuarial rates), however due to the restriction on actuarial rate classification they're only pay $1.5 for the same unit of coverage.
So the benefit ($2 worth of coverage) is now greater than the marginal cost ($1.50)????