*474ON MOTION FOR REHEARING
GONZALEZ, Justice,dissenting.
The court has glossed over the fact that this is a mandamus proceeding. Since the trial judge was following over 100 years of precedent, it is preposterous to conclude that he clearly abused his discretion. I would grant the motion for rehearing, and deny the writ. In the alternative, we should adopt some guidelines and/or make rule changes in order to avoid some of the practical problems that will arise as the bench and the bar struggle to implement this decision.
We have considered mandamus to be proper in some cases to compel a trial court to allow discovery. Jampole v. Touchy, 673 S.W.2d 569, 572-573 (Tex.1984); Allen v. Humphreys, 559 S.W.2d 798 (Tex.1977); Barker v. Dunham, 551 S.W.2d 41 (Tex.1977). In addition, mandamus has been issued to correct improper allowances of discovery by a trial court. See, e.g., General Motors Corp. v. Lawrence, 651 S.W.2d 732 (Tex.1983); West v. Solito, 563 S.W.2d 240 (Tex.1978); Crane v. Tunks, 160 Tex. 182, 328 S.W.2d 434 (1959). However, mandamus is an extraordinary writ that should be used only when there has been a violation of a clear right possessed by the relator. Neville v. Brewster, 163 Tex. 155, 352 S.W.2d 449, 452 (1961); See State Bar of Texas v. Heard, 603 S.W.2d 829, 833 (Tex.1980).
Under Young v. Kuhn, 71 Tex. 645, 9 S.W. 860 (1886) and its progeny, the trial court in this case did not abuse its discretion in disallowing the discovery. Our most recent cases establish that a relator who attacks a trial court ruling as an abuse of discretion “labors under a heavy burden. ... The relator must establish, under the circumstances of the case, that the facts and law permit the trial court to make but one decision.” Johnson v. Fourth Court of Appeals, 700 S.W.2d 916, 917 (Tex.1985). In our case, the court had but one choice — to follow our pronouncements that a defendant’s net worth was neither discoverable nor admissible to prove punitive damages. It would have been an abuse of discretion for the trial court to grant the discovery request under the law as it existed at that time. Therefore, the mandamus should be denied.1
The majority opinion does not decide the question of admissibility of net worth evidence. Without guidelines and/or corresponding rule changes, we have needlessly planted the seeds of confusion that will result in years of litigation as practitioners and the bench strive to comply with this opinion.
In order for the benefits sought to be achieved by the court’s opinion to fully inure to the citizens of Texas, the procedure employed to offer evidence of net worth must allow the defendant’s conduct to be judged as much as possible in a prejudice-free atmosphere. It is clear that the ability of a defendant to pay has no relevance to the issues of liability or compensatory damages. Accordingly, there is no legitimate need for a jury to be made aware of a defendant’s net worth when determining these issues.
To preserve the right of all litigants to a fair trial, we should adopt procedural guidelines for cases where punitive damages may be awarded and pre-trial discovery and presentation of net worth evidence is permissible. The utilization of a bifurcated trial procedure would prevent net worth evidence from prejudicially impacting liability and compensatory damage findings when punitive damages are claimed. The idea of a bifurcated trial procedure to separately determine issues of liability and damages is not new. Federal *475district courts are empowered with discretionary authority to order a bifurcated trial for this purpose. Fed.R.Civ.P. 42(b); see 9 Wright & Miller, Federal Practice and Procedure §§ 2388-2390 (1971); see also Annotation, Propriety of Ordering Separate Trials as to Liability and Damages, Under Rule 42(b) of Federal Rules of Civil Procedure in Actions Involving Personal Injury, Death or Property Damage, 78 A.L.R.Fed. 890 (1986). This procedure has recently been upheld for use in Texas diversity actions. Rosales v. Honda Motor Co. Ltd., 726 F.2d 259, 260 (5th Cir.1984).
The Supreme Court of Wyoming in Campen v. Stone, 635 P.2d 1121, 1131 (Wyo.1981) recently instituted a bifurcated trial procedure to remedy a situation similar to that which has been created by the majority opinion. The “Wyoming Plan,” which requires that plaintiffs make a prima facie showing that a viable issue exists for punitive damages before pre-trial discovery is permitted, provides a good model for Texas. A bifurcated trial procedure would work as follows:
1. The plaintiff would claim in his petition a right to punitive damages and then seek pre-trial discovery of defendant’s net worth.
2. The defendant would move for a protective order requiring the plaintiff to make a prima facie showing to the trial court that a viable issue exists for punitive damages. Upon such a showing, the pretrial discovery would be allowed. If plaintiffs claim for punitive damages is groundless and brought in bad faith, Tex.R.Civ.P. 13 authorizes the trial court to impose sanctions.
3. At trial, if sufficient evidence is produced establishing a prima facie case for punitive damages, the jury charge would make provision for compensatory damages and additionally ask the jury whether punitive damages should or should not be awarded. However, no provision would be made for the jury to determine the amount of punitive damages to be awarded at that point.
4.If the jury finds that punitive damages should be awarded, it would then hear evidence of the defendant’s net worth and return a separate verdict setting the amount of punitive damages.
See Campen v. Stone, 635 P.2d at 1132; see also Annotation, Necessity of Determination or Showing of Liability for Punitive Damages Before Discovery or Reception of Evidence of Defendant’s Wealth, 32 A.L.R. 4th 432 (1984). Under this plan, the admission of net worth evidence would constitute reversible error only during the first stage of a bifurcated trial.
I am also concerned that the court’s opinion will create uncertainty regarding a number of other issues. The most apparent area of uncertainty is the failure to define the term “net worth”. No fewer than eighteen times does the court’s opinion refer to “net worth”. However, despite the repetitious use of this term, the court has failed to inform the bench and bar what “net worth” is or how it should be calculated. Is a single balance sheet sufficient to identify “net worth” or is additional financial information necessary? Since “net wealth” was what the petitioner actually requested to be discovered in this case, is this synonymous with “net worth?” How do we measure net worth? Do we prove “net worth” by profit and loss statements, income tax returns, cash liquidity, a Fortune 500 listing, Standard & Poor’s rating, and the like? “I know it when I see it” is not much of a standard. Without objective criteria, a case by case determination will undoubtedly yield a wide disparity of results. Perhaps we should refer all of these questions to our Rules of Procedure and/or Evidence committees for recommendations. In the absence of guidance, confusion will prevail as practitioners and judges attempt to ascertain the components of “net worth”.
Aside from definitional problems, the respondent raises many questions in his motion for rehearing. For example:
Does a defendant’s net worth include the cash surrender value or the limits of liability of an insurance policy? If *476 the insurer is defending under a reservation of rights, would the insurance still be includable in the calculation of the assets? Likewise, would it make a difference if the defendant's insurance policy did or did not provide coverage for exemplary damages?
Assuming that a plaintiff attempts to offer net worth evidence that includes insurance coverage, the defendant should be able to keep this out pursuant to Tex.R.Civ. Evid. 411 which provides that liability insurance is inadmissible to prove negligent or otherwise wrongful conduct. Alternatively, such evidence could be kept out on the theory that the insurer’s duty to indemnify depends on a liability adjudication against the insured without respect to the insurer’s potential liability. The trial court judge would also have the discretion under Tex.R.Civ.Evid. 403 to exclude the evidence as misleading or unfairly prejudicial.
In an action against a corporate division or subsidiary should the net worth of the of the parent be considered? Would a different rule apply to a nonprofit defendant?
Once again, these questions involve considerations that properly should be balanced by the trial court judge pursuant to Tex.R.Civ.Evid. 403 when deciding the issue of admissibility.
Will a plaintiff be entitled to only an interrogatory answer stating what defendant’s net worth is, or will a plaintiff be entitled to all of the underlying financial data necessary to make his own calculations?
During discovery a plaintiff should generally be entitled to copy, at his own expense, all of the relevant financial documents. However, this will be problematic since the components of “net worth” are unknown. Consequently, the trial court will need to determine exactly what constitutes “net worth” and then decide which documents are relevant to calculate “net worth”. As discussed previously, this situation is unsatisfactory and needs to be remedied by a clear definition of the term “net worth.”
At what point in time is a defendant’s net worth relevant? Should the jury receive evidence of net worth as of the time the conduct occurred or at the time of trial which may be several years later?
Generally, assuming liability for punitive damages, evidence of defendant’s net worth at the time of the conduct, as well as subsequent gains and losses, is at least relevant and may be considered by a jury. However, since this issue also involves considerations of admissibility it would need be resolved by the trial court on a case by case basis. Tex.R.Civ.Evid. 611 provides that the trial court “shall exercise reasonable control over the mode and order of ... presenting evidence as to (1) make the ... presentation effective for the ascertainment of the truth....”
What safeguards exist to ensure that a relatively poor defendant in a multi-defendant case will not be unjustly punished by a jury on the basis of information of the other defendant’s ability to pay a large judgment?
Recent tort reform legislation provides the answer to any possible problem in this area.
In any action in which there are two or more defendants, an award of exemplary damages must be specific as to a defendant, and each defendant is liable only for the amount of the award made against that defendant.
Tex.Civ.Prac. & Rem.Code § 41.005 (Vernon Supp.1988). This recently enacted statute codified what undoubtedly was the common law. It provided that no defendant should be subject to primary or contributory liability for exemplary damages based upon conduct attributable to another tortfeasor. Similarly, the financial resources of any one defendant should not be relevant to punitive damages awarded against another defendant. See also Tex. R.Civ.Evid. 105(a) (when evidence is admissible as to one party but not admissible as to another, the court, upon request, shall restrict the evidence to its proper scope and instruct the jury accordingly).
*477In summary, I would grant the motion for rehearing and deny the writ. In the alternative, I would adopt the above guidelines.
. It doesn’t make any sense to say that the purpose of punitive damages is to deter others and to punish wrongdoers and then keep evidence of wealth from the jury. So, generally, I agree with the court that a jury should be able to consider the financial condition of the defendant in order to determine exemplary damages. Thus, the question is not if this evidence is relevant but when it is relevant. However, the more basic question here is whether the writ of mandamus is a proper way or vehicle to make this substantive change in the law. I don't think so. If we had intended to overrule Young v. Kuhn, and its progeny when Tex.R.Civ. P. 166b was changed, we certainly would have announced our intention.