(concurring). I agree with the court that G. L. c. 231, § 60B (1988 ed.), directs a medical malpractice tribunal to make a collective determination of the sufficiency of a plaintiff’s offer of proof, and that the judicial member of the panel may not overrule the decision of the other two. I also agree that this statutory scheme is not an improper interference with the judicial power in violation of art. 30 of the Massachusetts Declaration of Rights. I disagree, however, with the court’s reasoning which, in my opinion, unnecessarily confuses the legal standards under which these tribunals are to operate.
As the court points out, ante at 413-414, the plain language of the statute (“said tribunal shall determine if the evidence presented ... is sufficient”), read in conjunction with G. L. c. 4, § 6 (1988 ed.) (directing that when statutes confer authority on three or more persons the authority is to *420be exercised by majority decision), clearly indicates that the medical malpractice tribunal’s decision is to be collective. Instead of relying on this straightforward interpretation of the statute, however, the court attempts to buttress its holding by characterizing the tribunal’s decision as a factual rather than legal determination. Apparently the court is trying to fit its interpretation of the statute into the legal framework set up in Paro v. Longwood Hosp., 373 Mass. 645 (1977), in response to the art. 30 challenge posed therein. I disagree with this effort.
First, attempts to draw a coherent distinction between factual and legal determinations (to mention nothing of mixed questions of fact and law) historically have been illusory. See 5 K.C. Davis, Administrative Law § 29.9, 365-369 (2d ed. 1984); Levin, Identifying Questions of Law in Administrative Law, 74 Geo. L.J. 1, 9-12 (1985). The distinction can be both manipulable and conclusory. See State ex rel. Wisconsin Inspection Bureau v. Whitman, 196 Wis. 472, 498 (1928) (noting that in an attempt to modify strict separation of powers doctrines, courts have used “the guise of power to find facts” to uphold laws delegating judicial power). The case at bar illustrates the difficulties in separating questions of fact from questions of law. The court holds, ante at 415, that the decision whether the offer of proof “ ‘is sufficient to raise a legitimate question of liability appropriate for judicial inquiry’ or whether the plaintiff’s case is merely an unfortunate medical result” is a question of fact. G. L. c. 231, § 60B. To me, the decision is more analogous to a conclusion of law — the tribunal is not asked to determine what the defendants did, but whether what the plaintiff offers to prove they did meets some standard of sufficiency. Obviously reasonable people can and do differ on whether a given question is one of law or fact. Thus, not only is its value as a guidepost questionable, but the distinction lends itself too readily to result-oriented analysis.
Second, the court’s insistence on the law-fact distinction needlessly confuses the task before the tribunal. Our cases have strived to clarify the standard to be applied by the tri*421bunal in determining sufficiency. See Blood v. Lea, 403 Mass. 430, 433 (1988) (“the tribunal must use the same standard that a judge would use ‘in ruling on a defendant’s motion for directed verdict’ ”), quoting Little v. Rosenthal, 376 Mass. 573, 578 (1978); Kapp v. Ballantine, 380 Mass. 186, 193 (1980) (setting forth three-part test to determine the sufficiency of the offer of proof). This effort was not gratuitous; it was intended to set forth sufficient guidance for malpractice tribunals so that their decisions would not be arbitrary, and so that the standards applied by different tribunals to different plaintiffs would approach uniformity as nearly as possible. I fear that the court’s opinion undermines that effort. Its characterization of the tribunal’s task as a fact-finding mission suggests a much higher discretionary role for the panel than I believe the Legislature intended.1 The court’s attempt to disavow the directed verdict standard leaves one in the dark as to what standard a tribunal should apply. In short, the opinion weakens the very safeguards that have established the fairness and efficiency of the procedure.
Third, relying as Paro did on a distinction between questions of law and questions of fact to determine whether judicial power has been improperly delegated is misguided. Agencies adjudicate disputes all the time, determining “legal rights, duties or privileges.” See G. L. c. 30A, § 1 (1) (1988 ed.). The classification of agency action into legal or factual categories does not aid in a determination whether the action is an improper exercise of judicial power. Thus, the mere fact that the tribunal may or may not be making a factual determination does not determine the constitutionality of its actions.
*422I would uphold the tribunal procedure simply because the panel does not usurp judicial powers. The panel “ ‘has no authority to determine violations of substantive rights.’ ... [It] cannot make a binding determination that a crime, a tort, or any other actionable event has occurred.” Human Rights Comm’n of Worcester v. Assad, 370 Mass. 482, 487-488 (1976), quoting Bloom v. Worcester, 363 Mass. 136, 158 (1973). See Commonwealth v. Jackson, 369 Mass. 904, 921-922 (1976) (no separation of powers violation where statute does not interfere with “inherent judicial power”). The major impact that a panel decision may have on a plaintiff is to make it more difficult and risky for her to pursue her claim. See G. L. c. 231, § 60B (if panel finds for defendant, plaintiff may pursue the claim in court only upon filing of $6,000 bond payable to defendant for costs if defendant prevails). In order to keep this burden within reason, the judge has discretion to reduce the amount of the bond required in cases of indigency. See id:, Gugino v. Harvard Community Health Plan, 380 Mass. 464, 469 (1980) (abuse of discretion not to reduce sufficiently the amount of bond for indigent plaintiff). Moreover, the panel’s decision is subject to judicial review, either before trial or on appeal. See McMahon v. Glixman, 379 Mass. 60, 63-64 (1979).2 These safeguards help to ensure that whatever adjudicatory discretion the panel does have is sufficiently insulated from arbitrariness. See Arlington v. Board of Conciliation & Arbitration, 370 Mass. 769, 776 (1976) (statutory guidelines, standards, and procedures “protect against arbitrary action and provide also a proper occasion for delegation” to an arbitration panel of private individuals). Clearly the panel does not exercise the kind of power to make binding adjudications of substantive rights that would implicate art. 30.
A review of the cases from other jurisdictions considering the constitutionality of medical malpractice screening tribu*423nals confirms my view that the law-fact distinction is not relevant to the art. 30 inquiry. Only one State has invalidated the procedure on separation-of-powers grounds, and the perceived constitutional flaw was not that the panel determined questions of law. See Wright v. Central Du Page Hosp. Ass’n, 63 Ill. 2d 313 (1976).3 The overwhelming majority of courts that have considered the question have held that the procedure is not an unconstitutional interference with judicial power. Their holdings are not based on the panels’ role in finding facts as opposed to determining law. Rather, these courts usually rely on the fact that the panel determination is not a final adjudication of the parties’ rights. See Keyes v. Humana Hosp. Alaska, Inc., 750 P.2d 343, 355-357 (Alaska 1988); Eastin v. Broomfield, 116 Ariz. 576, 582 (1977); Lacy v. Green, 428 A.2d 1171, 1178 (Del. Super. Ct. 1981); Kranda v. Houser-Norborg Medical Corp., 419 N.E.2d 1024, 1036 (Ind. Ct. App. 1981); Attorney Gen. v. Johnson, 282 Md. 274, 283-290 (1978); Prendergast v. Nelson, 199 Neb. 97, 110 (1977); Suchit v. Baxt, 176 N.J. Super. 407, 423-426 (1980); Comiskey v. Arlen, 55 A.D. 2d 310-311 (N.Y. 1976), affd, 43 N.Y.2d 696 (1977); Simon v. St. Elizabeth Medical Center, 3 Ohio Op. 3d 164, 170 (1976); State ex rel. Strykowski v. Wilkie, 81 Wis. 2d 491, 520-522 (1978).
*424Because I believe that the factual-legal distinction relied on by the court is unworkable, unnecessary, and unconstruc-tive, I concur only in the court’s conclusion that the panel decision is collective and constitutionally permissible. I agree with the court that the judge correctly determined that the offer of proof was sufficient as a matter of law to permit the plaintiff to go to trial without posting a bond.
The flaw in the court’s analysis is illustrated most clearly by the court’s review of the offer of proof. The court correctly sets forth the facts in the light most favorable to the plaintiff and then applying the appropriate standard (directed verdict) to the facts, the court concludes that the offer of proof was sufficient. The court’s analysis of sufficiency of the offer of proof is consistent with my reasoning, not with the factual-legal distinction relied on by the court in its opinion.
Indeed, the court’s characterization of the issue decided by the tribunal as one of fact might tend to diminish our standard of review on appeal, because we usually review determinations of fact much more deferentially than conclusions of law.
The Wright opinion disapproved of the nonjudicial panel members’ being “vested with authority, equal to that of the judge, to determine and apply the ‘substantive law.’ ” Wright, supra at 322. Nevertheless, a later case in the same court reviewing the legislative revision of the law in response to Wright makes clear that it was not merely the nonjudicial members’ role in making legal as opposed to factual determinations that was offensive. The revised law clearly assigned to the judge the sole responsibility for determining all questions of law. Yet the court again struck down the procedure merely because the judge had to share some authority — even if only fact-finding authority •— with nonjudicial personnel. See Bernier v. Burris, 113 Ill. 2d 219 (1986). The approach of the Illinois court has been strongly criticized by courts and commentators alike as rigid, conclusory, and practically unworkable. See, e.g., Attorney Gen. v. Johnson, 282 Md. 274, 290 (1978); 1 K.C. Davis, Administrative Law § 3:10, at 184-185 (2d ed. 1978); Redish, Legislative Response to the Medical Malpractice Insurance Crisis: Constitutional Implications, 55 Tex. L. Rev. 759, 794-795 (1977).