dissenting.
I respectfully dissent. Although I believe the majority has done a thorough job analyzing opinions from the various courts of appeals which attempt to reconcile Hughes and Murphy, it fails to recognize that the two cases cannot be reconciled, for if the requirements of Murphy had been applied in Hughes, the very tolling doctrine it announced would have been *471inapplicable. Moreover, I believe the majority opinion has focussed so intently on the cases applying Murphy, it has neglected to recognize the public policy which should be a touchstone in attempting to resolve the conflict we are faced with here.
Hughes, not Murphy, is directly on point and should control
This is theHughes exception to the two-year statute of limitations in legal malpractice claims, as announced by the Supreme Court of Texas:
[W]e hold that the statute of limitations was tolled until all of the Hugheses’ appeals in the [the underlying] termination action were exhausted.1
The court stated two reasons for recognizing the exception: requiring a malpractice suit to be filed earlier would result in the parties having to take inherent inconsistent positions,2 and “limitations are tolled for the second cause of action because the viability of the second cause of action depends on the outcome of the first.”3 The rule was clearly defined, based on two independent policy considerations, and never hinted that the client-plaintiffs must retain their original attorney throughout litigation for tolling to apply. Indeed, the Hughes’ attorney withdrew as their representative more than two years prior to filing of their malpractice suit.4 Murphy v. Campbell5, by restricting this tolling doctrine to the period in which a defendant attorney continues to represent client-plaintiffs, wholly ignored both the facts in Hughes and the second policy consideration upon which the doctrine was grounded.
Edwards v. Kaye
In the recent case of Edwards v. Kaye,6 the Houston Fourteenth court examined the same question we face today:
[W]hen a lower court is faced squarely with clearly applicable precedent, and a subsequent opinion by the same higher court attempts to modify by dicta its own precent, which road are we or the trial court to take?7
Edwards, an attorney malpractice case with facts similar to both Hughes and the case before us, engages in a careful analysis which neither the Murphy opinion nor the cases following it have applied. It outlines the rules of stare decisis, noting that it “only applies if the facts on which the prior case is founded are substantially the same as those in the subsequent suit.”8 It further notes that dicta, even judicial dicta, although persuasive and a source of guidance, is not binding if erroneous under the facts before the court.9 Thus, because Murphy involved account malpractice and its language limiting the Hughes rule (which applies only to attorneys) was not necessary to the outcome of the case, Edwards comes to the following conclusion:
Because the Murphy court’s modification of the Hughes test, if applied to the facts of Hughes, would result in an opposite result in Hughes, we find the dictum *472to be inapplicable and not binding on our facts.10
I would hold, as the Edwards Court did, that the statute of limitations against Turpin-Smith was tolled until exhaustion of the underlying suit. As the majority has concluded that date was July 11, 1994, I would conclude that the malpractice suit against Turpin-Smith was not barred by limitations.
Ripeness and judicial economy
It bears repeating that Murphy (and the cases applying it) recognize only one of two policies upon which Hughes was based, and ignore the second, equally important one: a case must be ripe before it is litigated. Hughes acknowledges that a legal malpractice claim is premised on a bad outcome in the underlying representation; to require a client to file suit before the outcome is certain encourages unnecessary litigation. Under Murphy, a plaintiff must file suit before the claim is ripe, and indeed, in some cases where no suit need ever be filed at all. Certainly some potential malpractice actions may be resolved favorably to the client in the course of appellate proceedings or retrial; certainly others may find the malpractice, although present, was harmless. In these cases, litigation against the former attorneys will ultimately prove needless, but Murphy nonetheless requires it.
Moreover, the solution Murphy suggests for this dilemma, abatement of the malpractice case pending resolution of the underlying suit,11 underscores my point that a malpractice case filed before underlying claims are resolved is not ripe. The law generally requires ripeness before a suit may be brought.12 Why apply a different rule here? I conclude that policy considerations, as well as long-standing legal principles on ripeness of a controversy, favor the original Hughes tolling doctrine, without its narrow Murphy interpretation.
Reliance on plain language of Hughes
Finally, I observe that the Eilands filed suit in 1995, when the plain language of Hughes taught that the statute of limitations in an attorney malpractice suit was tolled until all appeals were exhausted.13 Murphy was not decided until 1997. Although I recognize that the general rule is that decisions of the Supreme Court apply retrospectively,14 this is not always true. Where there is extensive public reliance on the old law, and a change is unforeseen, fairness and policy preclude retroactivity.15 I would hold that the Eilands justifiably relied on what appeared to be a settled exception to the statute of limitations, decided in the highest civil court of the state on facts clearly applicable to their case.
CONCLUSION
For these reasons, I dissent.
. Hughes v. Mahaney & Higgins, 821 S.W.2d 154, 156 (Tex.1991).
. Id. at 157.
. Id.
. Id. at 156. The court does not specify the exact date on which the attorney withdrew, an indication it was not dispositive of any issue on appeal. Nevertheless, from the record it is clear he ceased to represent the Hugheses more than two years before the malpractice suit was filed. He withdrew "[b]efore the motion to dismiss was heard, and denied” by the trial court. The Waco court of appeals reversed the trial court’s decision on March 7, 1985. The malpractice suit was filed on May 21, 1987.
. 964 S.W.2d 265, 272 (Tex.1997).
. 9 S.W.3d 310, 314 (Tex.App. — Houston [14th Dist.] 1999, pet. filed).
. Id. at 313-14.
. Id. at 313.
. Id.
. Id. at 314-15.
. Murphy, 964 S.W.2d at 272.
. Id. at 275 (Spector, J. dissenting); see Casarez v. NME Hospitals, Inc., 883 S.W.2d 360, 366 (Tex.App. — El Paso 1994, writ dism. agr.) (discovery rule does not require lawsuit be brought on unconfirmed fear of HIV exposure before positive test results are possible; to require otherwise would encourage needless litigation); City of El Paso v. Madero Dev., 803 S.W.2d 396 (Tex.App. — El Paso 1991, writ denied) (court did not have jurisdiction to review adverse condemnation judgment where landowner had not sought zoning variance; case was not ripe until administrative decision was final).
. Hughes, 821 S.W.2d at 156.
. Sanchez v. Schindler, 651 S.W.2d 249, 254 (Tex.1983).
. Id.