< Return To Hearing
Denise Martin, Ph.D.
Sr. Vice President
WRITTEN TESTIMONY OF
FOR THE HEARING
I am Dr. Denise Neumann Martin, Senior Vice President of NERA Economic Consulting ("NERA"). I received my undergraduate degree at Wellesley College and my Ph.D. in Economics at Harvard University. Since joining NERA in 1991, I have spent a substantial portion of my time estimating future liabilities of defendants involved in asbestos and other types of tort litigation, and I am co-author of a reference text, "Estimating Future Claims: Case Studies from Mass Torts and Products Liability".
The Congressional Budget Office ("CBO") has estimated that the cost of compensating asbestos personal injury claimants under the FAIR Act is between $120 and $150 billion. The CBO used standard methods for forecasting claims in the tort system, the same methods used to make claims projections when the FAIR Act was initially proposed in 2003. Similar approaches have been used by NERA in our consulting practice and have been used by Bates White to prepare forecasts for its clients.
In preparing its current FAIR Act estimate, however, Bates White rejected these standard methods. Instead, it adopted an approach that more than doubles the population of workers estimated to have been exposed to asbestos. Having expanded the population, Bates White also overestimated future lung cancer and other cancer cases by making no adjustment for the exposure requirements mandated under the FAIR Act and exaggerating the proportion meeting medical requirements. Finally, the Bates White study fails to apply a claiming rate, effectively assuming that 100 percent of those eligible will sue. The Bates White $300 billion estimate, which it describes as "conservative," adds more than 350,000 cancer cases (Levels VI and VII) that by standard methodologies, including that of the CBO estimate, would not add to the liability of the trust.
A. The Bates White analysis overestimates the eligible population and underestimates the effects of the exposure and medical requirements
A number of flaws inflate the Bates White estimate of the "eligible population" making it significantly higher than that of other published and accepted studies. First, the Bates White estimate of the exposed population includes many industries and occupations in which workers are likely to have had little or no exposure to asbestos. Second, Bates White assumes a higher "turnover rate"- i.e., people moving in and out of exposed industries and occupations. Use of this higher rate increases the number of exposed workers, but Bates White fails to adjust for the fact that, under this assumption, each individual will also have a shorter duration of exposure.
This oversight has important consequences because it ignores key provisions of the bill-particularly the 15-year and 12-year weighted exposure requirements for Level VI and VII, respectively--that cut down on the potentially eligible population. Finally, the Bates White analysis overestimates the prevalence of pleural changes because it relies on rates reported in studies in which the populations are more heavily exposed than the population that Bates White estimates. Making basic adjustments, described below, to the Bates White analysis reduces the rate to 5 percent, half the rate that Bates White asserts is "conservative." This overestimation, which leads Bates White to conclude that many more lung and other cancer cases would be eligible for compensation than the CBO estimates, occurs for several reasons:
B. Use of the Bates White Results to Value Liability Under the FAIR Act Relies on the Unrealistic Assumption that the Propensity to File a Claim is 100 Percent
The only way that the Bates White analysis is relevant to evaluating the FAIR Act is if it projects claims that will qualify for payment under the Act. In his responses to questions by the staff, however, Dr. Charles Bates of Bates White admitted that while he describes his estimate of $300 billion as "conservative", he has not calculated claims but entitlements; that is, Bates White has made no estimate of the portion of these entitlements that will result in claims. In its slide presentation, however, Bates White overlooks the distinction between eligible and filed claims, thereby implying without basis that the claiming rate would be 100 percent.
Given the long history of asbestos litigation in the U.S., it is clear that claiming rates will not reach 100 percent. In the current tort system, for example, the filing rate for mesothelioma claims--a terminal disease whose only known cause is asbestos--is less than 100 percent even according to the Bates White study. There are a number of reasons to believe, as CBO evidently did, that these observed tort system claiming rates are indicative of claiming rates under the FAIR Act.
The Bates White report estimates that an exceptionally large "exposed population" will meet the occupational and exposure criteria specified by the FAIR Act. Its estimate - more than 42 million people alive as of 1980 - far exceeds the 21 million figure estimated by Dr. William Nicholson in his 1982 seminal published study of asbestos-related illness. Bates White does not provide details on how its population was estimated. However, we have been able to determine that the Bates White estimate of the exposed population greatly exceeds other published, accepted estimates for two primary reasons:
A. The Bates White Estimate Includes Industries and Occupations with Little Or No Asbestos Exposure
Under the FAIR Act, eligibility will require 15 weighted years of "substantial exposure" to asbestos for Level VI claims and 12 weighted years of exposure for Level VII claims. The Bates White analysis does not impose these exposure requirements, however, instead including in its estimate of the "eligible" population industries and occupations with little or no exposure to asbestos.
The Bates White estimate of the exposed population alive as of 1980 is nearly twice the estimate contained in the Nicholson study. This increase occurs because Bates White adds many industries and occupations to the list identified by Dr. Nicholson. Some of these additions have been important sources of claims in recent years, and their inclusion is reasonable and has become standard practice. In his November 7 response to questions posed by Senator Specter, Dr. Bates properly points to the steel industry in this regard. But, Dr. Bates' addition of industries and occupations to the Nicholson list is not confined to industries, like steel, that have produced a substantial number of claims. Dr. Bates also includes in his asbestos-exposed group barbers, beauticians, bus drivers, taxi drivers, and parking attendants, among others. These groups are unlikely to qualify for compensation under the FAIR Act.
To qualify for compensation under the FAIR Act, a Level VI (other cancer) claimant would have to show 15 years of "working in areas immediate to where asbestos-containing products were being installed, repaired, or removed under circumstances that involved regular airborne emissions of asbestos fibers . . . ." (section 121(b)(16)(B)). A Level VII (lung cancer/pleural) claimant would have to show 12 years of such exposures. We find no explanation for the assumption that people in such occupations would qualify for compensation under the FAIR Act. And yet, these people are far more numerous in the Bates White expanded "exposed population" than steelworkers are. The large exposed population estimated by the Bates-White report - more than half of the male population 18 years or older in 1980  - suggests that all members of these industry/occupation groups were assumed to be exposed with sufficient duration to satisfy the FAIR Act requirements.
The estimated number of mesothelioma claims in the Bates White report provides further evidence that the expansion of the Nicholson exposed population should not, in fact, result in a major increase in the number of people who would meet FAIR Act requirements for other cancer and lung cancer. Although Bates White's exposed population is more than twice the size of Nicholson's, the forecasted number of mesothelioma claims, 49,000, is actually less than the Nicholson forecast of 66,518. This result would occur if the average level of exposure of the people that Bates White added to the exposed population was very low. However, with such low rates of exposure, it is highly unlikely that these additional individuals would be able to meet the exposure criteria of the FAIR Act for lung cancer and other cancer claims.
The second source of the increase in the size of the exposed population estimated in the Bates White report is the use of a higher turnover rate than was assumed by Dr. Nicholson. Annual turnover rates reflect the proportion of the workforce in a given occupation or industry that is replaced during the course of the year. Assuming a higher turnover rate when preparing a forecast of asbestos liability increases the number of people estimated to have been occupationally exposed to asbestos because more people rotate through industries with potential exposure. However, increased turnover also reduces the average length of time in asbestos-exposed groups: with higher turnover, workers spend less time in each industry. The Bates White analysis makes no adjustment for this reduced duration of exposure, however. This oversight has important consequences because the FAIR Act requires 12 years of exposure for lung cancer claims and 15 years for other cancer.
Dr. Bates implicitly acknowledges this consequence of changing the turnover rate, explaining that the higher turnover rates should not affect the incidence of asbestos-related disease. This conclusion follows from the assumption made by OSHA and others that lung and other cancers are related to exposure (dose) in a linear way.
In contrast, the number of background, non-asbestos related lung cancers and other cancers will increase proportionately with the population. The Bates White analysis assumes that claims for these diseases, which are not caused by exposure to asbestos, will be filed under the FAIR Act and this assumption accounts for a large proportion of the additional 350,000 lung cancer and other cancer cases estimated in the Bates White analysis.
C. The Bates White Forecast Does Not Apply the Exposure Requirements of the Fair Act
Our analysis indicates that most of the background cancer claims and some of the excess cancer claims counted in the Bates White report will not meet FAIR Act exposure requirements. Bates White has not provided a detailed description of their calculation of exposure durations. Using standard methods to measure duration of exposures in the industries and occupations identified in the Nicholson paper, we find that less than half the projected cases of lung cancer and other cancer would meet the FAIR Act exposure requirements. Using the asbestos exposed population estimate developed by Bates White, the proportion of future cases of lung cancer and other cancer that would meet these exposure requirements would be lower still. Indeed, by definition, very few lung cancer and other cancer cases arising in occupations and industries with little or no exposure to asbestos would meet the substantial exposure requirements.
While it is true that a larger population will exhibit a larger number of background cancers, in a population that has little or no exposure to asbestos, fewer of these cancers will meet the substantial exposure requirements or the pleural disease requirements (for Level VI and Level VII compensation) required under the FAIR Act. This length of exposure requirement is never mentioned in the Bates White report and does not appear to have been taken into account. Indeed, their analysis apparently counts every worker who had ever spent even a day in any of the supposedly exposed industrial or occupational groups, every taxi driver and beautician, as someone who could assert a valid claim under the FAIR Act.
D. The Bates White Forecast Does Not Apply Stated Medical Screens Properly,
The medical criteria of the FAIR Act require claimants to show evidence of pleural changes to satisfy the requirement for Level VI or Level VII compensation. The Bates White analysis overestimates the expected prevalence of pleural changes in future cases of lung and other cancer thereby overestimating the number of claims that will be filed and paid.
More specifically, the Bates White prevalence estimate for pleural changes of 10 percent is based on values from published studies. As detailed below, these studies use different definitions of pleural disease and examine groups of people that are different from the exposed population used in the Bates White forecast and so cannot be used directly to estimate the number of potential claimants arising from that population. Making basic adjustments to the Bates White interpretation of the studies lowers the pleural prevalence used to estimate the eligible lung cancers from 10 percent to 5 percent. These adjustments are described below.
1. Some of the study populations consisted of people seeking treatment and so likely overestimate overall prevalence
Some of the studies of pleural prevalence cited by Bates White yield an overestimate of the prevalence in the exposed population because participation was more likely among those with higher risks of health problems (a problem known as selection bias). For example, the Epstein study was conducted of subjects who were admitted to the hospital. The Frumkin and Wain studies were conducted of autopsy subjects.
It is not possible to quantify the extent of selection bias in these studies, however the prevalence reported in the Epstein study is three times higher than the prevalence reported in the other four studies, suggesting a stronger bias. We omit this study from our calculations both because of concern about selection bias and because it did not measure pleural changes but examined rates of interstitial fibrosis. We also adjust average prevalence for the McLoud study by excluding the study participants who were selected because they were clinic patients.
2. Average prevalence for some studies include unilateral pleural changes rather than only bilateral pleural changes as defined in the FAIR Act
Level VI and Level VII claimants must provide evidence of bilateral pleural changes, not just unilateral changes. Several studies (Rogan, Schwartz and McLoud) relied upon by Bates White did not distinguish between findings of unilateral and bilateral changes and, consequently, overstate the proportion of workers who would be able to meet this requirement. An average of 45 percent of the pleural changes reported by Michaels and Frumkin are unilateral. This proportion was used to adjust the study results reported by Bates White that did not distinguish unilateral cases.
3. The average duration of exposure of the study subjects is higher and so the pleural prevalence is higher than the exposed population in general
The prevalence of pleural changes increases with cumulative exposure to asbestos. Studies cited by Bates White of workers with medium or high intensity exposures generally include subjects with longer-term exposure to asbestos than the current surviving population of exposed workers. This is, in part, because these studies were typically sponsored by unions representing workers with the highest exposures and longest periods of exposure but also because they were mainly conducted in the 1960s and 1970s before occupational asbestos exposures began to decline.
4. Studies of pleural prevalence in the general population overstate the prevalence for workers with low intensity exposure
The average prevalence of pleural changes in studies of the general population cited by Bates White is used as an estimate of the prevalence for workers in occupations with historically low intensities of exposure to asbestos. The general population estimates are too high for this purpose because the general population includes workers with high and medium intensity exposures to asbestos that raise the overall prevalence of pleural changes by more than the workers with no exposure that lower it.
5. Use of overall pleural prevalence overstates the proportion who will make Level VII claims by including those with underlying asbestosis who will make Level VIII claims
The prevalence of pleural changes used to estimate future Level VII claims must exclude the proportion of workers with pleural changes and qualifying evidence of asbestosis. These workers will not file Level VII claims, instead filing Level VIII claims. The average prevalence reported in the studies cited by Bates White, however, does not exclude these cases.
The proportion of those with pleural changes and asbestosis can be inferred from six of the studies cited by Bates White (Abelda, Baker, Michaels, Robins, Schwartz and Sprince). The average proportion is 29 percent.
After overestimating the eligible population and the number of cancers that will qualify for payment, Bates White then unrealistically assumes a 100 percent claiming rate--that is, every single eligible claimant will file a claim. Such an assumption is unreasonable, however, supported neither by experience in the tort system nor by the experience of other trusts.
Dr. Bates admits in his response to the Senate questions that "[w]e do not estimate a filing rate. Rather we estimate the number of individuals who would qualify for compensation under the FAIR Act." However, his presentation slides state that the $300 billion figure is "an estimate of the value of claims" and directly compare this figure with the CBO estimate. While the presentation may leave the impression that the Bates White estimate is comparable to the CBO estimate, it simply is not.
Below, we review the evidence for the CBO's assumptions that claiming rates under the FAIR Act would be no higher than rates in the tort system. First, claiming rates historically have been well below 100 percent, even for mesothelioma, which is a signature asbestos-related disease. Second, the FAIR Act would not be expected to increase filing rates above these historically observed rates, because strong incentives exist in the tort system for plaintiffs' attorneys to file asbestos claims. Indeed, bankruptcy trusts, and many solvent defendants, already pay claims that would qualify for Level VI and Level VII payment under the FAIR Act. Third, certain of the medical and exposure criteria of the FAIR Act are stricter than the criteria used by trusts and defendants in settling claims currently, indicating the propensity might be lower under the FAIR Act, rather than higher.
A. Claiming Rates Have Never Been 100 Percent Whether in the Tort System or Against Asbestos Trusts
Asbestos is a mature tort litigation with a long history of filings, verdicts and bankruptcy trusts. Despite this history, data from both the tort system and existing asbestos trusts provide evidence that the propensity of claimants to file a claim is far lower than 100 percent. Bates himself has estimated that even for mesothelioma, for which asbestos exposure is virtually the only known cause, claiming rates have been only 60 to 70 percent.
In his responses to the Senate's questions, Dr. Bates provides the following estimates of historical filing rates for these diseases: "Recent lung cancer tort filings account for fewer than four percent of lung cancer incidence in the occupationally qualified population and approximately 12 to 22 percent of the lung cancer incidence in the occupationally qualified population with pleural conditions. Recent other cancer tort filings account for less than one percent of other cancer incidence in the occupationally qualified population and approximately three to seven percent of the other cancer incidence in the occupationally qualified population with pleural conditions. These filing rates are estimated using Manville Trust claims, which eventually include virtually every filed tort claim."
As these statistics make clear, while asbestos claims have been filed in the U.S. since the early 1970s and the litigation is described as 'mature', the historical filing rates have been well below 100 percent for most diseases, including mesothelioma.
B. The FAIR Act Will Not Create Incentives to Increase Filings Above Those That Exist Currently in the Tort System
The Bates White analysis presumes that a large number of potential lung cancer and other cancer claims exist but are not filed in the current system, but that these same claims would be pursued and paid under the FAIR Act. In other words, the Bates White analysis assumes that the propensity to bring claims of this type would be higher under the FAIR Act than it is under the current system. The only reason to believe that other cancer and lung cancer claiming rates would be higher under the FAIR Act is if the compensation of these claims in the tort system had not made it worthwhile for the plaintiffs' attorneys to pursue them in the past. This hypothesis is contradicted by two key facts: first, plaintiffs' attorneys have every incentive to file a claim in the current system; second, if anything, the FAIR Act requirements for payment of lung and other cancer claims are more stringent than those that exist currently. Below, we discuss these elements in turn.
1. Plaintiffs' Attorneys Have Every Incentive to Locate All Valuable Claims in the Tort System
Some lung cancer and other cancer claims of the type estimated by Bates White have been paid in the tort system, providing evidence that, to the extent such claims exist and have value, they have been filed in the tort system. To the extent claims have not been filed historically, evidence exists that the incentive does not exist for them to be filed.
Plaintiffs' attorneys discover, pursue and negotiate asbestos claims through either the tort system or a trust distribution process. Plaintiffs' attorneys' fees provided a powerful incentive for plaintiffs' attorneys to root out valuable claims. Currently, there are two major ways that a claim can be turned into cash flow for the plaintiffs' law firm:
From the perspective of the plaintiffs' attorney, the decision on how much to spend on a claim will depend on the ease with which liability and damages can be proven and the amount of damages that the claim will likely fetch in the tort system or from a trust. For example, mesothelioma claims are particularly attractive to plaintiffs' attorneys for at least two reasons: (1) asbestos is virtually the only known cause, and (2) mesothelioma is a terminal cancer. While the link between the development of lung and other cancers is somewhat less direct, these claims still have value and are paid in the existing system.
To maximize profits, a law firm should increase the number of claimants that it can represent up to the point where adding another claimant costs more than the expected contingency fee from that claimant. Following this logic, it would be irrational for plaintiffs' attorneys to leave potential lung and other cancer untapped while investing heavily in mass screening for nonmalignant claims, because the former would bring them a higher expected contingency fee. However, the Bates White analysis presumes that just this situation, in which plaintiffs' attorneys are leaving potential profits on the table, exists in the current system.
2. Claims of the type forecasted by Bates White have value in the tort system
Lung cancer and other cancer claims of the type estimated to surge under the Bates White analysis have historically been paid in the tort system, proving that they have been pursued by plaintiffs' attorneys. The Manville Trust database, for example, provides evidence that lung cancer claims with and without underlying asbestos-related diseases have been paid. In addition, the Trust has paid lung cancer claims who were smokers and those whose smoking conditions were undetermined. We matched these same claimants to the databases of solvent defendants and found that they were paid by those defendants as well.
Even though such claims may not be caused by exposure to asbestos, a defendant chooses to settle them when the costs of doing so are less than the expected costs of taking the case to trial and risking a large jury verdict. This risk is real: lung cancer and other cancer claims have commanded large jury verdicts in the tort system. Because these claims have value in the tort system, such claims would be filed if they existed. Yet the number of lung cancer and other cancer claims that have been filed in the past is only a fraction of the number predicted in Bates.
3. Many of the workers who develop lung cancer have characteristics that make them unlikely to claim
Bates White added a large number of lung and other cancers resulting from background risk which by definition, would have little evidence of asbestos causation. These individuals had the opportunity to make claims in the tort system. However, cancers among workers with little asbestos exposure and for diseases without established epidemiology evidence of causation tend to claim at lower rates. This means that even if the additional 350,000 lung and other cancers that Bates White added were eligible, their claiming rate would be low.
C. FAIR Act Requirements Are More Stringent Than Those Existing in Tort System for Types of Claims Bates White Assumes Will Surge
If anything, the FAIR Act requirements for payment of lung and other cancer claims are more stringent than those that exist currently. The FAIR Act requires proof of underlying pleural changes or asbestosis for both categories of lung cancer, for example, whereas the Manville 2002 Trust Distribution Plan has a lung cancer category that does not require underlying asbestos-related disease. So do the Celotex and National Gypsum Trusts.
Similarly, exposure requirements for lung cancer claims under the FAIR Act are more stringent than those in existing trust. The Manville, Halliburton and National Gypsum trusts require only five years of exposure. Celotex and Fuller Austin have no minimum exposure period.
For other cancers, the exposure requirements are similar. The Manville, Halliburton Celotex, Fuller Austin and National Gypsum trusts all require five or fewer years of exposure.
In the tort system, defendants, faced with the expense and risk of jury trials, typically cannot impose specific medical and exposure criteria to all. Weaker claims may have a lower value, but most cancer claims nevertheless are settled at some figure