A View From The Press Box

While the notion of a benefits ratio at first blush has appeal, the reality is that it makes no sense to try to target any specific figure, at least in p/c insurance.

First, there actually are measures (such as loss ratios) that show the share of losses paid on behalf of policyholders.

The fact that these loss ratios can vary, sometimes signficantly, over very short periods of time suggests that a target ratio would be difficult to ever achieve as a practical matter. Catastrophes can push loss ratios up to several hundred percent in a given year. A year when mother nature decides to take a hiatus can push the same loss ratio below 50%. It is far too variable a number to try to target.

Moreover, every insurer has a different underwriting strategy, a different risk appetite and a different growth philosophy. Who would decide what the ratio should be on all these variables and more?

An insurer that specialized in insuring coastal homes, for example, could reasonably be expected to run a low ratio in most years to offset those years when the benefits paid well exceed the premiums earned.

Focusing on a loss or benefit ratio of some sort is also misguided because the resources insurers have available to pay claims includes not only premiums earned but also investment returns.

If investment results are poor (say from falling interest rates) and expected losses remain the same, the insurer will need to earn more in premium to maintain the same level of profitability (and financial strength rating).

The effect would be to push down the loss ratio even though the insurer is paying the same (or potentially larger) benefit.

Bottom line is that some target benefit ratio is a naive way to think about how quantity of benefits that should be paid.

In tightly regulated lines (and in other highly regulated businesses such as utilities) insurers target a specific rate of return (e.g., a return on equity) for the business.

The financial models used in such cases look at all sources of revenues, all costs and take in to account the variability in the line of coverage relative to the risk available in alternative investments.

I've see only a few plaintiff attorneys that have prepared realistic initial demands and that's okay--it's thier job.

But for a payment ratio measure to work there would need to be set standards for measuring the front-end demands.

I've seen similar broken arms bringing demands from $25,000 to $500,000. Will plaintiff attorneys agree to the former or latter?

I suspect they will want to leave things well enough alone.

Mr. Starkman's proposal, I think, exhibits a fundamental misunderstanding as to how claims are handled.

Many if not most homeowner claims are approached with the policyholder from the ground up, and are not packaged and submitted in the form of a demand. The exception would be when the policyholder engages a public adjuster

In most cases, the cost of repair or replacement will be determinative of the claim settlement, not a "demand" per se.

In the liability claim context, claimants who elect attorney representation will indeed have a demand made by their counsel. Such figures are starting points and invariably on the high side, and sometimes patently ludicrous, (just as they are with public adjusters).

There is certainly much data capturable in such cases, and it would show that settlement amounts are invariably less than the initial amount sought.

Depending on who views such data, some would say this difference is indicative of industry "gouging."

Believe it or not, most claim profesionals are only interested in paying what is fair and getting on down the road to the next claim.

Are insurance companies exploiting skilled plaintiff attorneys and public adjusters?

I think some imaginative data could help consumers.

For example, insurers could reveal, by size of claim paid, how long it took for the claim to be settled. Then comparative results could be made public showing how insurers do at settling claims in a timely fashion.

One way to get at something like the "Starkman" ratio would be if data was required to be collected on verdicts showing what was demanded, offered and the final verdict to see how reasonable the offers and demands were, effectively testing both plaintiff and defendant offers.

Surely there are ways to rank the claims practices of insurers if imagination is applied.

Starkman is absolutely right that the industry does not like disclosure of claims information and will fight any reasonable data availability to the death.

Here's an idea to get to the bottom of the Starkman ratio that I would like to see.

As someone stated, the question relates more to liability claims than material damage or homeowners. So let's focus on liability claims, most of which have legal representation to protect the claimant from those greedy insurers.

Since most insurers tend to closely estimate and monitor what they believe to be the true value of a liability claim (called case reserves) regardless of the initial demand of the claimant's attorney, why don't we require all liability lawyers to submit information regarding their initial damage request and the ultimate settlement.

They also could be required to document the reason they ultimately convinced their client to accept any amount lower than their initial fair demand.

Starkman might also demand information on the contingent fee of the attorney to get a clearer read on the net difference in their client's demand and ultimate payment since the insurer would not know this, and it would seem to be an important factor in determining the insurer's fairness.

Then we can get two measures for the price of one. We can get the Starkman ratio on insurer performance for all claims where there was legal representation AND a possibly useful Starkman Corollary ratio on law firm performance that consumers might find helpful.

Assuming the law firms would readily agree to provide this fairly simple information if you asked, Mr. Starkman, you can get what you want without having to fight those greedy insurers, as Mr. Hunter warns will fight to the death to hide?

The metrics of claims satisfactorily concluded would be tainted by the illusion that you received something you were not entitled too.

How do you factor in the costs of frivolous lawsuits and the untold millions spent defending them? How do you factor in the fraud rings, the arson cases and the like?

Keeping stats on legitimate claims improperly handled or paid would account for only an indeterminate percentage of those handled. A simple maintaining of stats proves nothing other than allowing manipulation to one's specific end.

Wouldn't the number of complaints to an insurance department, divided by the number of policies in the state, be a statistic that is available and suggests an answer to the question?

The stat should be calculated by line of business.

Mr Fleming has it right. Because many states--certainly the biggest insurance markets--for many reasons (including the political), are eager to hear about claims difficulties, there is plenty of access and established processes to correct a rare wrong.

Companies do not want to be political fodder and will do a fairly good job keeping their insureds happy.

On the other hand, the plaintiff constituencies may have a motive to try to force companies to increase unsupported payouts when they (the insurer's policyholders) are sued.

If this crowd can sell this rather dumb--for the reasons expressed in these posts--idea, then go for it. Anything to increase premiums.

If common sense and reason prevail, it was still a nice try. Sorry.

Sam, as an experienced catastrophe adjuster, your blog has hit the proverbial nerve of handling "cat" claims.

When many major insurance companies report a loss of earnings, they usually blame the loss on a larger than normal number of catastrophe claims.

Since Katrina, the number of "cat adjusters" has exploded. Unfortunately, only a few have the experience needed to properly handle a catastrophe claim.

Many larger insurance companies have had to resort to strict compliance guidelines to monitor the progression of the claim, requiring the use of adjuster buddies, quality control representatives, file reviewers, field support personnel, and the like to assure the claim is adjusted properly.

This support requires a huge monetary expenditure, and less money goes into the pocket of the policyholder. The more experienced adjuster can no longer make a living working for the larger companies because they receive fewer claim assignments and are required to meet absurb and redundant compliance guidelines, so many are gravitating to the smaller regional companies that hire the more experienced adjuster.

The insurance industry will experience a backlash of problems if we have a major hurricane this year, as most claims will be handled by unskilled rather than experienced adjusters.

All companies need to re-evaluate their catastrophe claims-handling procedures and focus again on retaining the quality adjuster, and forget the outrageous premise that "anyone" can do this work.

Dean Starkman and your blog responders should have listened to Henny Youngman when he was asked, "How's your wife?" His one-liner response was, of course, "Compared to what?"

The point -- and perspective -- everyone seems to be missing here is the perspective of the one party in the claims adjusting process that matters the most: the claimant.

The simple question to be posed to each claimant is:

Did your settlement meet, not meet, or exceed your expectations?

That's the only metric that counts -- that of the customer. (Even in 3rd party liability claims, the claimant is, by definition, a customer for these purposes.)

Obviously, responses can be stratified by type of claim and type of claimant -- but the goal of delighting customers by exceeding their expectations (without "giving away the store," of course) should be paramount.

If every insurance company CEO would carve out some professional development time for the Insurance Institute of America's one-semester course entitled, "Delivering Insurance Services" (AIS-25) -- and inculcate these quality disciplines into the organization the CEO leads, this kind of metric would evolve almost naturally!

(Other sources of this discipline including Six Sigma Quality, etc., are available through numerous sources -- but the IIA program is insurance-related and highly relevant.)

All too often, we lose our focus on our customer -- and fail to listen to the voice of our customer.

CEOs and branch managers should contact Pete Miller, CEO of the Institutes' in Malvern, Pa., and "get on the quality train" rather than throw rocks at it or miss it altogether.

This is THE major reason Japanese auto manufacturers are out-competing their U.S. counterparts. Just ask Don Petersen, former CEO of Ford. He tried. His successors didn't.

It is also a major reason other industries are performing more effectively -- and profitably -- than the insurance industry . . . not to mention the level of trust of our industry on the part of the general public.