Concurring. — I concur in the court’s decision affirming the Court of Appeal’s judgment ordering the superior court to sustain *358defendants’ demurrers to plaintiff’s first and second causes of action. I do so, however, on narrower grounds than the majority. The majority holds plaintiff’s allegations of proximate cause are both legally insufficient as to cause in fact and in contravention of public policies involving discretionary governmental decisions. (Maj. opn., ante, at pp. 355-357.) I agree with the first conclusion but not the second.
I
In a typical personal injury action, where the causal relationship between the breach of duty and the injury is relatively plain, “it suffices to plead causation succinctly and generally.” (Bockrath v. Aldrich Chemical Co. (1999) 21 Cal.4th 71, 78 [86 Cal.Rptr.2d 846, 980 P.2d 398].) “But when, by contrast, ‘ “the pleaded facts of negligence and injury do not naturally give rise to an inference of causation[,] the plaintiff must plead specific facts affording an inference the one caused the others.” ’ [Citation.] That is, the plaintiff must allege facts, albeit as succinctly as possible, explaining how the conduct caused or contributed to the injury.” (Ibid.)
In the present case, the causal connection between the State Department of State Hospitals’s failure to employ two mental health professionals to evaluate Gilton Pitre as a potential sexually violent predator (SVP), and Pitre’s subsequent rape and murder of plaintiff’s sister, is by no means plainly apparent. Both for this reason and because plaintiff’s cause of action under Government Code section 815.6 is a statutory one requiring every material fact to be pleaded with particularity (Bockrath v. Aldrich Chemical Co., supra, 21 Cal.4th at p. 78; Lopez v. Southern Cal. Rapid Transit Dist. (1985) 40 Cal.3d 780, 795 [221 Cal.Rptr. 840, 710 P.2d 907]), plaintiff bore the burden of pleading with particularity each fact necessary to show how the agency’s alleged breach caused her injury. The rule that a demurrer admits only properly pleaded material facts, not contentions, deductions or conclusions (Evans v. City of Berkeley (2006) 38 Cal.4th 1, 6 [40 Cal.Rptr.3d 205, 129 P.3d 394]), should be applied here with plaintiff’s burden of particularity in mind: to the extent the complaint included only conclusory or inferential claims of factual causation, rather than particularized allegations of fact, it was subject to demurrer.
Viewing the complaint in this light, I agree with the majority its allegations of cause in fact were overly speculative and conjectural. (Maj. opn., ante, at pp. 355-356.) Plaintiff alleges in a conclusory manner that had the agency employed two clinicians to perform a “full evaluation,” rather than a single professional to do a “paper screening],” the two evaluators, on reviewing Pitre’s criminal record and seeing the sadistic nature of his prior sex offense, would have determined he qualified as an SVP. But the complaint fails to *359include any specific facts as to what part of Pitre’s criminal record the single evaluator failed to review or what facts of the prior crime were therefore not before him. For the crucial first step in her hypothetical causative chain, therefore, plaintiff provides only a conclusion, and a rather speculative one at that.
To the extent plaintiff was unable to learn, before filing her complaint, what criminal records were and were not reviewed by the sole evaluator, but had reason to believe identifiable critical information was omitted from his screening that would have been before the required two evaluators pursuant to the agency’s “standardized assessment protocol” (Welf. & Inst. Code, § 6601, subd. (c)), she could have alleged such particular critical omissions on information and belief. (See 4 Witkin, Cal. Procedure (5th ed. 2008) Pleading, § 398, pp. 537-538.) If plaintiff had no information and belief with regard to particular critical omissions from the screening materials, as indeed may have been the case, she lacked an adequate basis for claiming the agency’s choice of a “screen[ing]” by one professional, instead of an “evaluation” by two, factually caused the agency’s failure to refer Pitre to the designated county attorney for an SVP petition. (See Welf. & Inst. Code, § 6601, subd. (d).)
The complaint’s deficiencies regarding the hypothetical causative chain’s first step, the lack of an evaluation by two mental health professionals, are compounded by its treatment of the final step, Pitre’s hypothetical commitment as an SVP. The allegation that Pitre “would have been civilly committed” is entirely conclusory, accompanied by no particularized facts to show that a unanimous jury would have determined, beyond a reasonable doubt, that Pitre was'an SVP. (See Welf. & Inst. Code, §§ 6603, subd. (f), 6604.)
The complaint may be read to claim that even the filing of a commitment petition would have prevented the injury because most such petitions allegedly go to trial and, had a petition been filed and Pitre held for trial past his scheduled release date, “he would not have had the opportunity to murder Alyssa only four days later.” But this type of coincidental causation — -an allegation that some breach created an opportunity for an injury to occur, without increasing the risk of that injury occurring — is insufficient. (See Rest.3d Torts, Liability for Physical and Emotional Harm, § 30 [“An actor is not liable for harm when the tortious aspect of the actor’s conduct was of a type that does not generally increase the risk of that harm.”]; Royal Indemnity Co. v. Factory Mutual Ins. Co. (Iowa 2010) 786 N.W.2d 839, 852 [“[E]ven where an act may be a factual cause, ‘we are convinced that an act which merely places persons in the position where they sustain injury from an unrelated event is not for that reason a legal cause of the injury.’ ”]; Berry v. Sugar Notch Borough (1899) 191 Pa. 345, 348-349 [43 A. 240] [rejecting as *360“sophistical” argument that streetcar motorman’s speeding was the cause of his injuries when tree fell on his streetcar as it passed: although “it was the particular speed at which he was running [the streetcar] which brought the car to the place of the accident at the moment when the tree blew down[,] . . . [t]he same thing might as readily have happened to a car running slowly, or it might have been that a high speed alone would have carried him beyond the tree to a place of safety.”].)1 The risk of Pitre reoffending was no greater because he was released without a trial than had he been released after a trial ending without commitment — the exact timing of his release did not increase the public safety risk he posed. Plaintiff thus cannot establish legal cause merely by showing that absent the agency’s breach of duty Pitre would not have been released in time to meet and attack the victim; she must show that absent the breach he would have been committed as an SVP. As just discussed, the complaint lacks particularized, nonconclusory factual allegations to that effect.
A plaintiff could plead factual causation in this context by alleging, on information and belief or otherwise, that through a breach of its mandatory evaluation duty the State Department of State Hospitals overlooked specific available facts regarding the potential SVP, facts so strongly indicative of mental illness and repetitive sexual predation that any decision maker would likely find the person to be an SVP. The allegations of the present complaint do not rise to that necessary level.
II
I concur in the court’s judgment on the grounds explained above. I do not, however, agree with the majority that a finding of liability in this case would contravene “policy considerations bearing on the question of proximate cause” (maj. opn., ante, at p. 356) because the hypothetical causative chain on which the agency’s liability depends — that two evaluators would have found Pitre to be an SVP, resulting in a referral to the designated county attorney, who would have filed a petition for commitment on which Pitre would have been held beyond his hypothetical release date and ultimately committed — involves discretionary decisions.
The majority cites no public policy stating the outcome of a discretionary decision may not form part of the chain of events a tort complaint hypothesizes as what would have happened but for the defendant’s breach of duty. Actual discretionary acts by public employees are protected under a public *361policy of preventing discretionary governmental decisions from interference or distortion through fear of liability, a policy codified in the immunity provisions of Government Code section 820.2. But we are concerned here only with discretionary decisions that allegedly would have been made, not with actual decisions. As discussed later in this opinion, the protective policy has no logical application to merely hypothetical decisions.
The majority observes vaguely that proving causation through the outcome of discretionary decisions that would have occurred but for the defendant’s breach is “in tension with the very idea of discretion.” (Maj. opn., ante, at p. 356.) To be sure, the existence of decisional discretion increases the plaintiff’s burden on causation, a burden I agree plaintiff here has not met even at the pleading stage. (See pt. I, ante) But the majority cites no principle of law or public policy barring the attempt, and draws no germane factual distinction between this case and those in which such causative proof has been approved, such as Landeros v. Flood (1976) 17 Cal.3d 399 [131 Cal.Rptr. 69, 551 P.2d 389], Henderson v. Newport-Mesa Unified School Dist. (2013) 214 Cal.App.4th 478 [154 Cal.Rptr.3d 222], and Alejo v. City of Alhambra (1999) 75 Cal.App.4th 1180 [89 Cal.Rptr.2d 768]. To the extent the majority means to distinguish these cases on the ground that there the but for chain involved only a single discretionary decision, while here a “series” of such decisions is hypothesized (maj. opn., ante, at p. 357, fn. 16), the distinction appears merely quantitative and factual, not a matter of public policy. If public policy allows a discretionary decision to form a link in the causative chain, a plaintiff with a sufficiently strong factual case — again, a case absent here — could presumably show that each link in the hypothetical chain was met.
There may be some discretionary governmental decisions so unrestricted by statutory or constitutional rules that one can say, as a matter of law, that their hypothetical outcomes could never be proven. Legislative budgeting decisions, certain sentencing choices and some conditional release determinations come to mind. But the majority makes no attempt to show the decisions involved in finding a person to be an SVP belong in this category. Indeed, the majority acknowledges these decisions are constrained by statutory factors and procedures. (Maj. opn., ante, at p. 356, fn. 14.) One might also note that the initial determination whether a person may be an SVP, and should be referred to the designated county attorney for a commitment petition, is made by mental health professionals using a protocol designed by and for such professionals. These are not the types of decisions whose outcomes can never be predicted.
While the majority refrains from expressly relying on the fact that the discretionary governmental decisions necessary for an SVP commitment *362would have enjoyed immunity under Government Code section 820.2 (discretionary acts of public employees) or 845.8, subdivision (a) (parole release determinations), the majority recites with apparent approval the reasoning of two Court of Appeal decisions that employed this rationale. (Maj. opn., ante, at pp. 353-355; see State of California v. Superior Court (1984) 150 Cal.App.3d 848, 859, fn. 8 [197 Cal.Rptr. 914]; Whitcombe v. County of Yolo (1977) 73 Cal.App.3d 698, 708 [141 Cal.Rptr. 189].) Neither court discussed the point in depth, however, and neither explained how a statutory immunity, or the protective policy that motivates it, can logically apply to a decision that never occurred, so as to preclude liability for breach of a mandatory duty.
How allowing causation to be traced through hypothetical decisions that did not occur would inhibit or distort real decisionmaking when it does occur is entirely unclear. To take an example from the present facts, as long as the designated county attorney knows that the decision not to seek a commitment against a potential SVP is immunized from liability, the attorney cannot be inhibited by the possibility that, in some other case that is never presented for prosecution because the State Department of State Hospitals does not make a referral, an injured plaintiff may allege the attorney would have sought a commitment had the decision been presented.
Neither the majority nor the lower court decisions on which it relies put forward a logical ground for concluding the proof of causation in circumstances like those presented here would violate an established public policy.
Ill
For the reasons given above, I concur in the result and in part of the majority’s reasoning, but not in the majority’s conclusion that policy considerations negate proximate cause as a matter of law.
Liu, J., and Kruger, J., concurred.The petition of real party in interest for a rehearing was denied July 22, 2015.
The Restatement Third categorizes this rule as one limiting “scope of liability,” otherwise known as proximate cause, but acknowledges it can also be considered an instance of the lack of factual cause. (Rest.3d Torts, Liability for Physical and Emotional Harm, § 30, com. a, p. 542; see id.., ch. 6, Special Note on Proximate Cause, pp. 492-493.)