EVENTS

Date:	Thursday, April 21, 2005
Time:	3:00 PM -- 5:00 PM
Location:	Wohlstetter Conference Center, Twelfth Floor, AEI 1150 Seventeenth Street, N.W., Washington, D.C. 20036

April 2005

Tort Reform: Learning from Asbestos Lawsuits

Led by President George W. Bush, the call for dramatic reform of the U.S. tort system has grown louder in recent months. At an April 21 AEI conference, Anup Malani, associate professor of law at the University of Virginia; and Charles H. Mullin, an economist at Bates White, presented their recent study on liability in asbestos-related lawsuits. The authors found that many companies bankrupted by lawsuits have passed their asbestos liabilities to their codefendants.

 

Anup Malani
University of Virginia

In our presentation we are going to cover one issue involved with tort reform that we believe raises some difficult and interesting problems: the doctrine of reallocation. The doctrine of reallocation is part of a larger concept, joint and several liability; the flip side of joint and several liability is simply several liability. Our presentation will go over what these concepts are and discuss some problems with reallocation--all in the context of the asbestos case.

In a case involving several liability, the plaintiff will sue two companies for damages. The plaintiff will collect the maximum amount of what he is entitled to from each company. If one company cannot pay off the amount, it goes into bankruptcy, and the plaintiff is unable to collect all the funds he or she is entitled to from that company.

In a case involving joint and several liability with reallocation and financially solvent defendants, the plaintiff will sue one firm and get the total amount awarded; in this case the defendant will pay his share plus the other defendant’s share. It is up to the first firm to go after any other firms that may be jointly liable in order to recover funds. This method reduces the litigation costs to the plaintiff by not requiring him to search out all the defendants. If the first defendant is financially insolvent, the plaintiff collects all possible funds, and the firm goes into bankruptcy. The defendant can go after the second defendant and recover all the funds he owes, plus the uncollected funds from the first defendant.

A situation of joint and several liability without reallocation and with solvent firms is exactly the same as the case with reallocation, with the only difference occurring with financially distressed defendants. The plaintiff goes after the first and recovers as much as he can; if the firm is unable to pay the amount, it goes into bankruptcy. The plaintiff then goes after the second firm and collects all that it is liable for; the defendant pays the required amount but does not have to pay the amount uncollected from the first defendant.

With reallocation the plaintiff is insured against defendant insolvency; any amount that goes uncollected from defendant one gets reallocated to defendant two. The threat of one defendant having to pay for the actions of another defendant will lead to closer monitoring between firms. This could prevent or deter firms from engaging in mischievous conduct. The cost of this policy increases the risk of companies going into bankruptcy. We estimate that each additional $1 recovered because of reallocation results in $0.23 to $0.66 in bankruptcy expenses.

We want to quantify the amount of insurance provided and compare that to the cost of the bankruptcies. We believe that the first bankruptcy wave and the reallocation associated with it caused the second wave. We are going to try to figure out the amount of money reallocated from companies that went bankrupt in the first wave to companies that went bankrupt in the second wave; this amount is the additional insurance provided. We also want to do is look at the companies that went bankrupt in the second wave and quantify the inefficiencies associated with the bankruptcies, reviewing how much social wealth was lost when those companies went bankrupt, for instance.

We conclude that $6.5 to $11 billion was reallocated from bankrupt companies to non-bankrupt companies. We get this estimate by taking a figure from Tillinghast-Towers Perrin revealing that between 1994 and 2002, $53.3 billion was paid out in asbestos claims. We then use a prior RAND study to figure how much money went to victims and how much went to administrative costs. Using this we find that two-thirds of the money was used for legal expenses. We use these two figures to get to the $6.5 to $11 billion amounts of plaintiff compensation. This is the benefit of reallocation.

To measure the cost of reallocation we look at the value of the companies that went bankrupt in wave two, and we apply a standard estimate of cost from the bankruptcy literature for every dollar that ends up in bankruptcy. We estimate that the total aggregate enterprise value of the companies that went bankrupt in wave two was $21.5 billion. We estimate the social loss is 12 to 40 percent of the pre-distressed company value. After all this we find that reallocation is very expensive insurance for plaintiffs against bankruptcy and against the insolvency of defendants; we estimate this cost as 23 to 66 cents per dollar.

There are several policy alternatives to reallocation. The first is to raise the priority of tort claimants in bankruptcy. A couple of options would be to give them the same priority as secured creditors or even to give them superpriority over all other creditors. Another alternative would be to eliminate corporate limited liability for tort claims; this would allow plaintiffs to go after company shareholders. A further alternative would be to require companies to purchase insurance. A final option would be to analyze whether bankruptcy courts have the authority to discharge tort liabilities.

We are going to focus on one alternative, the idea of superpriority with no reallocation. In this situation the onus is shifted from codefendants to secured creditors. This could have two further effects: it could lead to more monitoring of tortuous behavior and would insure plaintiffs against bankruptcy. This approach has three benefits: secured creditors are better prepared to bear risk than are codefendants, it eliminates bankruptcy contagion, and it could lower the cost of capital. This policy does have some costs however. It could potentially raise the cost of capital. In addition, secured creditors are often worse monitors. The conclusion to draw from all of this is that reallocation is very expensive and we should look for ways to reform it.

Charles Mullin
Bates White

The reason we are looking at asbestos as an example is that it is the biggest mass tort with reallocation. It involves over sixty companies going bankrupt, most which were the most liable. The number of companies going bankrupt has resulted in a large reallocation. What we are attempting to do in this study is quantify this reallocation and determine if it is dollar for dollar or some smaller amount.

The amount an average company has paid in claims has about tripled from 1990 to 2002. There were basically two waves of bankruptcies. The first occurred from 1989 to 1993 and the second occurred from 2000 to 2002. The second wave hit a bunch of big name companies, some of which paid out hundreds of millions of dollars. Between these two waves was when the Georgine v. Amchem Products case was going on, and lots of companies were anxiously waiting to see what decision was going to come out of it before the case was resolved in 1997.

Most of the asbestos producers went bankrupt in the second wave. It is difficult to find any firms named in current legislation that actually produced asbestos; most of them bought and incorporated it into their products.

There are two explanations for the growth in defendant-specific payments. The first is natural growth; this could be because of the discovery of new evidence, changes in liability rules, an involving litigation strategy, or growth in wages. The second explanation is the reallocation of payments. Our proposal is that defendant-specific growth is equal to reallocation growth times natural growth. We can estimate the defendant-specific growth from firm-specific settlement data; we estimate an upper and lower bound for the natural growth rate. Using the defined relationship we can estimate the reallocation growth.

The defendant-specific growth is estimated from 10-K filings and from 689 litigation filings that occurred between 1994 and 1998. We computed the annual growth rates from these two sources. To estimate the natural growth rates we assume that reallocation does not affect the sum of payments across defendants; as long as there are enough solvent firms to reallocate to, the plaintiff will collect the same amount of money. We can look at the total amount of money a plaintiff was awarded and look at compensation through time. By looking at the growth in total average payments from the 10-K sample and the growth in average payments per resolved claim within the 10-K sample we get the lower bound of the estimate. To get the upper bound we look at what we call the natural experiment, the time period around the Georgine case from 1995 to 1997; during this stretch, there were no significant bankruptcies. We estimate that reallocation caused at least 56 to 157 percent growth in specific claim value.

Francis McGovern
Duke University School of Law

There is no question that the direction of this paper is correct; there is no doubt that reallocation is very costly and expensive. The problem with any kind of empirical analysis in civil law is the availability of data. The precision of the paper is somewhat questionable--the case could be strengthened by softening some of the assumptions. Another point is the possible externalities that would arise for any policy change--I suspect that the cost of capital, for example, would increase dramatically.

The paper is absolutely correct on judgments, but its settlements are uncertain. For example, when looking at the growth in the value of cases you must keep in mind that lawyers have a tendency to overreach as far as determining the amount of claim to go after. A second aspect of the settlement process is the allocation among defendants; defendants are notorious for being unable to allocate among themselves. Plaintiffs do a lot better picking off defendants one at a time than defendants do coming together to agree on a settlement. If defendants come together they may be able to satisfy the plaintiff at a smaller amount because they are getting all the money at once.

One problem with the assumption of 12- to 20-percent reduction out of other bankruptcies is that in my experience there is a huge bounce in the value of companies after capital liability. Using the normal effect on the capital structure from other bankruptcies probably does not work in the mass tort system.

As a whole the paper is great, but I think the arguments could be strengthened by softening some of the numbers in the estimates and expanding the number of assumptions.

John E. Calfee
AEI

This paper is a crude first cut at a very big problem--I am a big fan of first cuts on big, important issues. A lot can be learned from the first paper on an important issue. This paper looks at one of the biggest problems with the liability system: its inefficiency. The system does not work very well as a deterrent.

One issue that does raise concerns for me is the idea of the monitoring system. I am not sure what we would regard as going on in this case. It would be hard imagine firms not wanting to do business with a firm in a particular industry because sometime down the road they may be involved in some litigation; it seems like everyone could be monitoring everyone else.

One of the biggest problems with asbestos is the sheer magnitude of the settlements being awarded. Nearly half of the money is going to non-impaired claimants. It seems like the system would work better if less money was changing hands and more money was being left for later payments. In addition there are some serious doubts about the medical science being used to validate some of the claims. It is inevitable that these kinds of claimants will get money when the courts process such huge amounts of claims. The larger problem is the amount that the system awards.

With the idea of superpriority it seems like the deep pockets will get even deeper. Once a plaintiff finds a firm with substantial assets, he will reach into those pockets as deeply as he can. An effect of all this is that even more money than is currently moving through the system could enter the system as a result of introducing superpriority. As has been stated earlier, superpriority could lead to sky-high interest rates for capital purchases.

This summary was prepared by AEI intern Michael Wilson.