FTCR notes that many insurers have falsely characterized loss data in statements to lawmakers, news media and the public, and cautions that lawmakers and regulators should not rely upon the insurance industry’s current loss projections, because those figures are not based on hard or otherwise reliable data. The study concludes that the insurance industry is in need of stringent regulatory and accounting reforms. Until such reforms are enacted, FTCR believes a moratorium is necessary on both rate increases and legislatively enacted limits on legal rights known as “tort reform.” Finally, whether insurance companies are intentionally inflating their reported losses is a question that can only be resolved by state regulators and state and federal law enforcement officials, who must invoke their authority to investigate the insurers’ accounting practices.2
False Accounting: How Medical Malpractice Insurance Companies Improperly Inflate Losses to Justify Sudden Surges in Rates and Tort Reform
Those who cannot remember the past are condemned to repeat it.
— George Santayana Introduction Doctors, hospitals and other health care providers purchase medical malpractice insurance to cover themselves if they negligently injure or kill a patient. Malpractice insurance companies suddenly and sharply increased rates in recent years, causing many health care providers to experience triple-digit rate increases, accompanied by the withdrawal of some insurance companies from the market entirely. This phenomenon, which began around 2001, is popularly known as an “insurance crisis.” Because the crisis affects all lines of insurance, including auto and homeowner, as well as business liability, it has had a destabilizing effect on the marketplace and on both national and local economies. However, physicians have been particularly vociferous in their complaints about higher malpractice premiums and have orchestrated local protests and job walk-outs.
The causes of the crisis – and hence the proper solution – are subject to intense debate in Congress and in state legislatures throughout the nation.
The insurance industry says soaring lawsuits and damage awards are responsible for the “insurance crisis.” Medical malpractice litigation, they say, has led malpractice insurers to suddenly experience high losses, necessitating quick and enormous premium increases. It is routine for insurance companies to claim that for every dollar of malpractice premium taken in, the insurers pay out far more in claims. One insurance group has asserted that in 2001, medical liability2 insurers nationally paid out $1.40 for every $1.00 they received in premiums. If true, that would be an economically unsustainable condition. Insurers use such2 See Sec. II. Incurred Losses vs. Actual Losses3 loss estimates as incontrovertible evidence for their argument that “tort reform” – caps on compensation to victims of medical malpractice, limits on plaintiff’s lawyers’ fees and various other restrictions on the right to sue – will solve the crisis by lowering losses, and thus premiums. In virtually every state in the nation, insurers, often joined by the medical lobby, are promoting compensation3 caps and other tort changes in order to control escalating insurance premiums. The insurance industry’s analysis of the crisis and its solution has been challenged by consumer organizations and commentators. Many note that the industry sustained enormous investment losses during the early part of the decade due to a severe drop in the stock market, low bond yields, poor investment decisions and record low interest rates. In 2001 alone, property-casualty insurers sustained an estimated $20 billion in stock market losses. Ten large insurance companies lost $274 million on investments in five corporate4 frauds: Enron, WorldCom, Adelphia, Global Crossing and Tyco. Additionally, in the wake of the September 11, 2001 terrorist attacks and the aforementioned market fall, reinsurance – on which many medical malpractice mutual insurers rely – has become extremely expensive.
Consumer advocates note that insurance premiums followed the same pattern during the last insurance crisis in the mid-1980s: massive premium increases occurred after interest rates fell from a highpoint of nearly 19% earlier in the5 decade and insurers’ investment income declined precipitously.
Finally, reports and other analyses have disputed insurers’ contention that6 lawsuit filings and damage awards have increased dramatically. Consumer34564 advocates and scholars find no need for proposed restrictions on access to civil courts or jury determinations.
Much is at stake in the malpractice insurance debate: billions of dollars in premium increases collected by insurers since 2001, doctors threatening to leave practice if premiums do not fall, and the right of malpractice victims to collect full compensation for their injuries.
The purpose of this study is to determine the accuracy of the insurers’ reported losses. These alleged losses are the predicate for the insurance premium hikes that have occurred in recent years. The loss data are also the basis for the insurance industry’s claim that legislation to limit malpractice suits and compensation is necessary to arrest the crisis and lower insurance premiums.7 I. Methodology This study compares the dollar amount medical malpractice insurers initially reported that they would pay out on policies in effect between 1986 and 1994, with insurers’ reports made ten years later of what they would actually pay out on policies in effect in each of those years as set out in Schedule P of the insurers’ Annual Statements filed with state insurance commissioners.
Insurance companies are required by law to file these Annual Statements in every state and to report certain data sets to state regulators. The contents of the Annual Statement are formally certified by an actuary.
A.M. Best’s, a data collection service headquartered in Oldwick, New Jersey, compiles data from these annual statements and publishes it in an annual volume entitled Aggregates and Averages. The data contained in this study are compiled from the Schedule P, Part 2F tables of all medical malpractice insurers published by Best’s in its 1996 through 2004 editions of Aggregates and Averages. Best’s describes the data it collects as “the world’s largest proprietary data-base of insurance industry information,” and it describes Aggregates and Averages as “the industry’s authoritative source for current and historical statistics on the property/casualty industry.” The Best’s data in this study represent aggregate data for the medical malpractice industry nationwide.75 II. Incurred Losses vs. Actual Losses The distinction between “incurred” and actual losses, commonly known as “paid losses,” is central to understanding an insurance company’s true financial condition and to evaluate the losses insurers report. It is a distinction insurers do not often make in public debate.
Insurers calculate their rates for a given year based on their “incurred losses” for that year. When insurers say they have “incurred losses” of a certain amount in a given year, however, they do not mean that they have actually paid out that amount in that year. Rather, insurers mean that they estimate they will ultimately pay out that amount on claims they predict they will receive that are covered by policies in effect in that year. In other words, “incurred losses” represent projected losses. Thus, if an insurer reports in 2003 that its “incurred losses” for 2002 were $100, the insurer has not paid out $100 for 2002 claims. Rather, the insurer estimates that it will ultimately pay out – over a period of several years – $100 for claims covered by policies in effect in 2002.
An insurer’s “incurred losses” are therefore, by definition, a guess. Statistical and mathematical methodologies have been developed which, using standard actuarial techniques, can be applied to make that guess an educated one. However, absent a regulatory formula that both mandates the use of such techniques and reviews insurers’ compliance, insurers have enormous discretion in determining incurred losses. (Presently, only California law under Proposition 103 contains such a mandate.)
Each year, the insurer receives more accurate information about the “incurred losses” it guessed it would ultimately pay for claims covered by policies in effect in a previous year. As time passes new claims are reported to the insurer, the insurer receives more details about existing claims, and the insurer ultimately pays a specific amount – or no amount – on each claim. As it receives this new information, the insurer adjusts the original guess it made. The more time that elapses, therefore, the less guesswork is involved and the more accurate an estimate for a previous year becomes.
In medical malpractice, the average claim is paid approximately 5 and 1/2 years after the claim arises; most claims are paid within 10 years. An insurer’s estimate of its true liability for claims it has incurred in a given year is therefore substantially accurate after 10 years. We will call these revisions “actual losses” throughout this report, as year-to-year revisions become nearly negligible by the end of a 10-year period, as the flattening of Figure 1 illustrates.6 This process can be observed by reviewing the “incurred loss” data reported for a set of years in the Annual Statements filed by each insurer.
Predicting the number of claims an insurance company must pay out, and the amount of those claims, and setting rates based on these guesses, is inherent in the nature of the insurance business. In exchange for a premium an insurer receives from an insured in the present, the insurer agrees to pay claims against that insured in the future. There is no way for the insurer to know at the time it receives the premium exactly how much it will pay for claims against the insured, nor even whether there will be any claims against that insured at all. Insurers therefore may not fairly be criticized for estimating their future losses8 and changing those estimates every year – that is the nature of the business.87 Insurers may fairly be criticized, however, when they mischaracterize these estimates of future losses as actual losses – which they do frequently. For example, the most commonly used measure of profitability in the insurance industry is the loss ratio: the ratio of an insurer’s incurred losses in a given year to its earned premiums in that year. While the earned premium number is the actual amount insurance companies collect from policyholders and does not meaningfully change over time, the incurred loss number is a guess and is certain to change. Yet insurers discuss the loss ratio as if each number were a hard number. For example, if an insurer reports a loss ratio for 2004 of 110, it typically characterizes itself as actually paying out $1.10 for each $1.00 in premiums it collects in 2004. The implication is that the company is losing money. In fact, it has not paid out $1.10 in 2004, but only guessed that when a final accounting of 2004 claims is completed years from now, it will have paid out $1.10.
Here is how a Florida coalition of insurance companies, hospitals and the medical lobby characterized the industry’s financial status: In 2001, medical liability insurers nationally paid out $1.40 for every $1.009 they received in premiums.
In fact, this dire portrayal is based on incurred losses, and is, by definition, only an estimate of what insurers will pay out in the future. Yet the statement expressly – and falsely – states that that amount was paid out.
Even analyses offered in “scholarly” publications misrepresent incurred losses, as does this claim published online by the health care policy journal Health Affairs: “Thus, by 2002 every premium dollar collected resulted in $1.29 in10 total expenses, awards, and settlements.”
Medical malpractice insurers had not paid all expenses, awards and settlements for 2002 by January 2004 when this article was published. The claim that every premium dollar resulted in $1.29 in costs is based on incurred loss estimates, not actual payments, and is false as such.
The description of projections as actual payments is a misrepresentation that has misled policymakers, the news media and the public.9108 A separate and far more serious issue is raised if insurers are intentionally inflating the incurred loss estimates beyond a reasonable projection of what will be necessary to pay claims in order to justify raising premiums.
The difference between an insurer’s initial estimate of its incurred losses for a given year’s policies and the amount of its actual losses on that year’s policies has important implications for the current medical malpractice insurance debate. This is because the rates an insurer charges for a given year are necessarily based on its incurred loss estimates for claims covered by that year’s policies, not on its ultimate paid losses on that year’s policies. Thus, if the amount an insurer ultimately pays out for claims covered by a given year’s policies is less than the amount the insurer initially estimated it would pay out for claims covered by those policies, the premiums paid by policyholders for that year would have been too high. Similarly, if the amount the insurer ultimately pays out is more than the amount the insurer initially guessed it would pay out, the premiums paid by policyholders for that year would have been too low.
In a weak economy, insurance companies stand to gain by reporting sudden and substantial increases in incurred losses. Big increases in incurred losses are used to justify sudden spikes in premiums, such as those in the current medical malpractice marketplace. The reported losses also yield tax breaks for insurers. And the increased estimates of incurred losses provide the foundation of the industry’s argument that only by enacting “tort reform” will premiums go11 down.
Whether the insurer charged a medical malpractice rate that was too low or too high, and the amount by which that rate was too low or too high, cannot be known with confidence until 10 years after the insured pays the premium. Physicians will not know for certain if their 2005 medical malpractice insurance rates are too low or too high until 2014.
Unfortunately, there is no opportunity to go back ten years and lower rates that, in hindsight, proved to be too high. Instead, insurers keep the extra premiums discovered when incurred losses are revised, and funnel the money into surplus or profits.
Nor will tort law restrictions put in place at the behest of the industry based on inaccurate loss estimates be retroactively repealed.119 III. Occurrence Coverage vs. Claims-Made Coverage There are two types of medical malpractice insurance coverage: “occurrence” coverage and “claims-made” coverage. An insurance policy that covers claims arising in the year the policy was written, even if the insured did not actually make the claim until several years later, provides occurrence coverage. Until the mid-1980s most medical malpractice policies were occurrence policies.
In contrast, so-called “claims-made” coverage – now the primary type of medical malpractice coverage sold – covers only claims made in a given year. With claims-made coverage, by the end of the year in which the policy is written the insurer knows of all the claims that will be reported, but must guess at the amount, if any, it will pay on those claims.
For purposes of loss projections with “occurrence” policies, the insurer must not only guess at the amount it will pay on a claim, but must also guess as to whether a claim will be reported at all. We would therefore expect the variation in future restatements of incurred loss estimates to be greater for occurrence coverage than for claims-made coverage.
This study reviews incurred losses for both forms of malpractice coverage. IV. Results – Incurred Losses For 1986-1994 Revised Over Ten Years A. Claims made coverage.
After 10 years, as claims information became more accurate, the initial incurred loss estimated for each year from 1986 through 1994 by the medical malpractice insurance industry has proved to be at least 25% overstated. Note that this period includes the last “insurance crisis” – 1986 through 1990. As Table 1 indicates:
- During the key crisis years – 1986 through 1990 – incurred losses were
initially estimated to reach $10.7 billion. Ten years later the reported losses for that period totaled just $7.1 billion, meaning that original loss estimates during the crisis were 51% higher than the actual losses reported ten years later.
- The initial incurred loss estimate for 1988 – the apogee of the crisis – has proved to be 58% overstated.
- In total, for the 9 years 1986 through 1994, malpractice insurers’ initial
incurred loss estimates were $23.4 billion. They reported incurred losses of10 $16.7 billion 10 years after the initial estimates, for a total overstatement of $6.7 billion, or 40%.
B. Occurrence coverage.
The extent to which the insurers’ initial incurred loss estimates have proved to be overstated is even more dramatic for occurrence coverage. As Table 2 indicates:
- During the key crisis years – 1986 through 1990 – incurred losses were
overstated by $3.7 billion, or 62% higher than the actual losses reported ten years later.
- Malpractice insurers overstated losses by 87% in 1989.
- Initial incurred loss estimates for the years 1986-94 totaled $16.1 billion,
while their reported incurred losses 10 years after the initial estimates were made were $10.8 billion, an overstatement of $5.3 billion, or 49%.11
| Year | Insurers’ initial reports of incurred losses for year | Insurers’ reported incurred losses 10 years later for year | Difference between initial estimate & 10th year estimate | % Inflated |
|---|---|---|---|---|
| 1986 | $ 1,430,307,000 | $ 1,053,046,000 | $ 377,261,000 | 35.8% |
| 1987 | $ 1,880,098,000 | $ 1,292,153,000 | $ 587,945,000 | 45.5% |
| 1988 | $ 2,192,936,000 | $ 1,390,953,000 | $ 801,983,000 | 57.7% |
| 1989 | $ 2,497,558,000 | $ 1,638,427,000 | $ 859,131,000 | 52.4% |
| 1990 | $ 2,705,808,000 | $ 1,743,368,000 | $ 962,440,000 | 55.2% |
| 1991 | $ 2,933,366,000 | $ 2,164,927,000 | $ 768,439,000 | 35.5% |
| 1992 | $ 3,117,994,000 | $ 2,267,284,000 | $ 850,710,000 | 37.5% |
| 1993 | $ 3,289,783,000 | $ 2,479,112,000 | $ 810,671,000 | 32.7% |
| 1994 | $ 3,364,855,000 | $ 2,697,617,000 | $ 667,238,000 | 24.7% |
Year Insurers’ initial reports of incurred losses for year Insurers’ reported incurred losses 10 years later for year Difference between initial estimate & 10th year estimate % Inflated 1986 $ 2,352,144,000 $ 1,595,897,000 $ 756,247,000 47.4%
Figure 2 illustrates the change in combined incurred losses (occurrence and claims-made policies) as reported by the nation’s medical malpractice providers for 1986 through 1994, over the course of ten years’ worth of loss revisions. Insurers projected a total of $39.5 billion in medical malpractice losses during this time, but revised that downward after 10 years to $27.5 billion, amounting to a 44% overestimate of losses by insurers between 1986 and 1994.00 The graph shows that the losses insurers initially reported are far higher than the actual losses reported ten years later. Even after revising the original 1988 projections upward in 1989 – perhaps to maintain the perception of an ongoing crisis(cid:1)– that year’s losses, along with every year’s losses, eventually fell precipitously as the incurred loss estimates were refined over time.12
| 1987 | $ 2,023,815,000 | $ 1,259,268,000 | $ 764,547,000 | 60.7% |
|---|---|---|---|---|
| 1988 | $ 1,699,015,000 | $ 1,046,396,000 | $ 652,619,000 | 62.4% |
| 1989 | $ 1,904,193,000 | $ 1,015,994,000 | $ 888,199,000 | 87.4% |
| 1990 | $ 1,764,533,000 | $ 1,101,090,000 | $ 663,443,000 | 60.3% |
| 1991 | $ 1,521,940,000 | $ 1,130,592,000 | $ 391,348,000 | 34.6% |
| 1992 | $ 1,608,852,000 | $ 1,161,299,000 | $ 447,553,000 | 38.5% |
| 1993 | $ 1,495,738,000 | $ 1,157,469,000 | $ 338,269,000 | 29.2% |
| 1994 | $ 1,765,645,000 | $ 1,341,849,000 | $ 423,796,000 | 31.6% |
Medical malpractice insurers overstated incurred losses for occurrence and claims-made policies combined between 1986 and 1994 by an average of 46% each year.
C. Loss estimates during and after the 1980s insurance crisis.
The data indicate that medical malpractice insurers overstated their anticipated losses for each of the years analyzed for this study. Additionally, it appears that the losses reported during the insurance crisis of the mid- to late-1980s were even more inflated than those of the early 1990s.
According to the data (claims-made and occurrence policies combined):
- In 1989, medical malpractice insurers announced losses for that year of $4.4 billion; by 1998, that number had been revised downward to $2.7 billion in losses – a 40% drop.
- For the crisis years, 1986 through 1990, insurers’ initial incurred loss
estimates were overstated by an average of 56%.
- During the following four years (1991-1994), initial incurred loss estimates were overstated by an average of 33%.
Figure 3 illustrates that loss inflation was higher during the 1980s insurance crisis than in 1991 through 1994, though loss inflation was still significant during those non-crisis years.0113
The Last “Insurance Crisis”
Insurers overstated losses during the crisis period by 56%, while pressing for rate increases and changes in tort laws, only to substantially reduce those loss projections a decade later, long after excessive premiums were collected. D. A case study: The Doctors’ Company.
Incurred loss estimates reported by the nation’s fourth largest medical malpractice insurer, the Napa, California-based Doctors’ Company, mirror the12 national trends. According to the company’s Annual Statements:
- During the key years of the last “insurance crisis,” 1986-1990, The Doctors’
Company’s initial estimate of incurred losses was an average of 49% higher each year than the revised report ten years later. This overstatement matches the industry’s average, for which the initial estimate (for claims-made policies) was also 49% higher than losses as reported 10 years later.
- For the nine-year period 1986-1994, The Doctors’ Company’s initial
incurred loss estimates were overstated by an average 39%; the industry overstatement (for claims-made policies) during the same period was 40%.
- The Doctors’ Company also had a higher loss inflation percentage during the crisis years of the 1980s – 49% – than between 1991 and 1994 – 27%, again following the industry-wide trend of greater overstatements during “insurance crisis” years.
VI.Reported Losses and the Present Crisis The current crisis is roughly four years old; there is no data to assess the accuracy of insurers’ “incurred loss” reports for recent years. Because we have fewer than ten years of restated incurred loss estimates for the years since the current crisis began, we cannot yet know what the ultimate payouts will be for claims incurred during this crisis with any reasonable degree of accuracy.
Nevertheless, we know medical malpractice insurers inflated losses by a minimum of 25%, and an average of 46%, every year between 1986 and 1994. We can assume that the insurers’ overstatement of incurred losses has continued in recent years. If in fact the trends identified in this analysis continue, it is1214 reasonable to estimate that malpractice insurers have overstated losses by approximately $15 billion since 1995.
We can also examine recent incurred loss reports to determine whether malpractice insurers have reported a sudden spike in incurred losses since the beginning of the most recent “insurance crisis” in 2001, following the pattern of the 1980s crisis.
As revealed by Table 3, there is a sudden increase in reported incurred losses between 2000 and 2001. After four years during which total malpractice incurred losses hovered between $5.07 and $5.26 billion, the estimate for 2001 jumped 22% to over $6 billion, an additional 9% in 2002 to $6.7 billion, and another 6% to $7.2 billion in 2003.
VII. Conclusion For each year, beginning with the previous insurance crisis, for which ten years of revised incurred loss information is available, the initial incurred loss estimates of medical malpractice insurers have proved to be substantially overstated, for both occurrence and claims-made coverage.
Loss inflation during the last insurance crisis – when insurers had multiple motives to show greater losses – was pronounced compared to the years that immediately followed.
Still, for those non-crisis years, insurers’ initial incurred loss estimates also were substantially overstated.15
| Year | Insurers’ initial estimates of incurred losses for year |
|---|---|
| 1997 | $ 5,259,038,000 |
| 1998 | $ 5,234,365,000 |
| 1999 | $ 5,097,796,000 |
| 2000 | $ 5,066,336,000 |
| 2001 | $ 6,196,047,000 |
| 2002 | $ 6,776,851,000 |
| 2003 | $ 7,198,259,000 |
As noted, insurance companies have a financial incentive to overstate losses during periods when their investments are performing poorly. By contrast, in periods of economic growth insurers will seek to maximize their investment income by lowering prices in order to attract capital and to expand market share. The need for high premium income, and the need to overstate losses, falls at such times.
In view of this data, there is no reason to expect that insurers’ incurred loss estimates for 2001-2004 – and thus their rates for 2002-2005 – are accurate. To the contrary, we now have certain evidence that the malpractice rates insurers charged during the last insurance crisis and the years following it were grossly excessive – by an average of between 40% (for claims-made coverage) and 49% (for occurrence coverage). We should expect to discover ten years from now that the incurred loss estimates medical malpractice insurers are reporting today, and the rates that they are charging, have been similarly inflated.
These results should raise a red flag for insurance regulators and lawmakers. The information presented here suggests that the industry’s accounting practices are in need of revision, including far greater scrutiny by insurance and financial regulators.
To protect against price gouging based on inflated estimates of incurred losses, regulators should follow the approach adopted in California with voter-approved Proposition 103. Proposition 103 requires incurred loss estimates and other projections by insurers to comply with a regulatory formula that disallows inflated losses. Moreover, Proposition 103 authorizes the insurance commissioner and/or citizens to challenge excessive rates and order rate rollbacks if necessary.
Lawmakers in many states who are contemplating the proposals made by insurance and health care lobbying organizations to limit the legal rights of injured patients risk casting their votes in favor of changes in tort law that are based on false loss and claims information disseminated by insurers and their allies.
Insurers may inflate premiums because they mistakenly believe that they will ultimately pay out the amounts they initially estimate. Or, insurers may knowingly inflate their initial incurred loss estimates, and thus the rates they charge, in order to compensate for a drop in investment income and, thereby, profits, and to pressure legislators to enact changes in tort law. The truth involves tens of billions of dollars and the public health and safety.16 The results of this study argue for far greater scrutiny of insurers’ practices using the tools available to state regulators and law enforcement officials – and a moratorium on both rate increases and tort restricting legislation until a definitive answer can be reached as to whether current losses are being properly calculated and correctly reported.17
