The trial court in this case ruled that Prisoners Union (Union) meetings are properly disallowed because “a prisoners’ Union is incompatible with current lawful policies of penological confinement. Its existence would violate institutional security.”
The court’s first ground, incompatibility with current policies, is inadequate justification under Penal Code section 2600 for depriving inmates of their rights. As for “institutional security,” though we may give weight to the court’s conclusion we are not bound by it. My own *456review of the evidence persuades me that respondent has not presented an adequate showing that reasonable institutional security requires the deprivation.
Respondent Director of Corrections states that, though he prohibits group activity by the Union, inmates are permitted to become members, to receive the Union’s newspaper, and individually to correspond with and be visited by outside members (unless they are prohibited from corresponding or visiting for valid reasons other than Union affiliation).1 However, since meetings in prison are allowed only to institution-approved activity groups, the Union now desires and seeks that formal status. Soledad prison officials have permitted group meetings with outside members for such groups as Alcoholics Anonymous, the Drug Abuse and Addiction Group, and the Duplicate Bridge Group.
Respondent’s argument that alternative means exist misses the point. The approved grievance procedure provides for processing of individual complaints but not for group discussion; also, the airing of grievances is but one of the Union’s aims. As for the approved Men’s Advisory Council, it is initiated by prison officials pursuant to regulation (Cal. Admin. Code, tit. 15, § 3230) and is completely subject to their control (Cal. Admin. Code, tit. 15, § 3232 (“the inmate is obligated to abide by all institution procedures and conditions for participation”)). Neither the grievance procedure nor the council can accurately be called an alternative means for the exercising of rights of assembly, association, and speech that petitioners seek through these Union meetings. Thus we turn to the question of the legality, under section 2600, of depriving petitioners of those rights.
Section 2600 and its interpretation
Originally the Penal Code contemplated “civil death” for prisoners. Even after a 1968 amendment, section 2600 provided that “sentence of imprisonment in a state prison for any term suspends all the civil rights of the person,” though the sentencing judge and the Adult Authority had discretion to restore certain rights. The 1968 amendment did provide, *457however, that four basic rights were retained: to inherit property, to correspond confidentially with lawyers and public officials, to own written material produced during imprisonment, and to purchase, receive, and read published writings (subject to specified exceptions). Also it was judicially proclaimed that “In this state we have long since abandoned the medieval concept of strict ‘civil death’ and have replaced it with statutory provisions seeking to insure that the civil rights of those convicted of crime be limited only in accordance with legitimate penal objectives.” (In re Harrell (1970) 2 Cal.3d 675, 702 [87 Cal.Rptr. 504, 470 P.2d 640].)
After 1968 several disputes concerning section 2600 reached this court. In all, California Department of Corrections (CDC) rules or actions were invalidated. (In re Harrell, supra, 2 Cal.3d 675, cert. den. 401 U.S. 914 [27 L.Ed.2d 814, 91 S.Ct. 890]; In re van Geldern (1971) 5 Cal.3d 832 [97 Cal.Rptr. 698, 489 P.2d 578]; In re Jordan (1972) 7 Cal.3d 930 [103 Cal.Rptr. 849, 500 P.2d 873]; In re Jordan (1974) 12 Cal.3d 575 [116 Cal.Rptr. 371, 526 P.2d 523]; Payne v. Superior Court (1976) 17 Cal.3d 908 [132 Cal.Rptr. 405, 553 P.2d 565].)
In 1975 the Legislature again amended section 2600, not only expanding the list of enumerated rights (now in § 2601) but also reversing the “civil death” presumption. It now commands that a prisoner may “be deprived of such rights, and only such rights, as is necessary in order to provide for the reasonable security of the institution in which he is confined and for the reasonable protection of the public.”
Since the 1975 revision the statute has guaranteed that inmates retain the same rights as nonprisoners and usually to the same extent. The only objectives that justify abrogating their rights more than those of other individuals are reasonable security of the institution and reasonable protection of the public. Further, it is not enough that the abrogation be desired or useful or consistent with institutional security or with public protection; it must be necessary.
The statute is clear, and courts should continue to require that corrections officials comply with it as did this court in In re Harrell, supra, 2 Cal.3d 675. There an earlier version of section 2600 specified reasons for which officials could impinge on a prisoner’s right to receive written material. Rehabilitation was not one of those reasons; CDC cited it as the justification for a rule excluding material; and the court invalidated the rule as violating section 2600. (Id., at pp. 703-704.)
*458Similarly, under present section 2600, no facts other than those regarding reasonable institutional security or public protection should be accepted as justifying deprivation of prisoners’ rights to hold Union meetings. Respondent’s references in his return to “rehabilitation” and “discretion to accommodate those groups which appear to promise the greatest benefit” are, therefore, irrelevant juridically to the issue of depriving prisoners of their rights. Further, the statute specifies reasonable security and protection. The Legislature thus recognized that goals of absolute security and protection might lead to intolerable deprivations.
The showing here
It seems clear that respondent’s main objection to the Union concerns its goal of collective bargaining. I discuss below his showing on that point. In addition he predicts certain security problems not necessarily related to the goals or content of Union meetings—potential for violence, possible takeover by a violent gang, possible use of force in soliciting members. The occurrence of any of those consequences could of course affect the security of the institution. Yet respondent’s showing to support those predictions and his conclusion that an absolute ban is necessary are indeed not adequate.
Respondent testified that “any kind of mass activity has a potential in it for destructive activity.” That general statement does not explain a ban in this specific case, particularly since other groups of inmates are allowed to meet. Respondent admitted that the prison has used various mechanisms to protect security when other groups meet (e.g., prior notice, search of outside participants, staff presence, suspension of meetings in case of security threat, disciplinary sanctions). He admitted also that he knew of no misbehavior by any Union members in or out of prison arising out of Union activities, and of no instance where Union officers had threatened or jeopardized prison security in any way. He gave no indication that he knew of any violence at prisoners’ union meetings in other states.
As for the possibility of takeover by a gang, respondent offered the unsupported conclusion that he was sure gangs would try to use the Union once it was recognizéd because gangs “are always looking for vehicles.” Again he failed to distinguish the Union from groups that do meet and indicated no knowledge of any gang involvement in any prisoners’ union. Petitioners’ witnesses, including the chief deputy director of CDC, expressed a belief that the Union’s open meetings and *459free, supervised elections would minimize the possibility of gang takeover.
Respondent testified that, in his experience, prison recruiting may take the form of threats and physical force. He also stated that it may take legitimate forms. He did not specify whether or why he expected violent forms in this case, and he testified that he had no knowledge of any threats or violence in past Union recruiting. He concluded his testimony by stating that he had not seen prison security threatened by the way the Union here has moved toward its goals.
On this record I find that Union meetings and discussions in themselves apparently pose no security threat that requires blanket disapproval at Soledad. Respondent contends that the case should turn on the broader issue of security risks in inmate collective bargaining. The trial court took a similar approach. However, petitioners in this proceeding are seeking only to hold meetings to discuss Union goals, one of which is collective bargaining. They do not now seek recognition as a bargaining unit2 or ask that respondent be required to bargain or meet and confer or negotiate with them, and it appears they have no immediate plans to do so. However, to the extent that ultimate goals of the group may be expected to affect their immediate actions the collective bargaining question is relevant.
I note first that respondent’s opposition to collective bargaining is based on his own understanding of that goal, which appears to be different from the Union’s. He testified to his conviction that collective bargaining would be based on inmate strikes and withdrawal of good behavior, which he believes would create a serious security problem. He admitted on cross-examination that he had no knowledge of the Union’s ever advocating such action.
Petitioners’ evidence, on the other hand, shows that Union goals are nonviolent; its methods have been to use available legal channels; and it actively has discouraged strikes. The idea for the Union as a group arose out of dissatisfaction with strikes, which seemed counterproductive in prison. Instances were cited of Union members’ dissuading other inmates from striking, and members testified that they will continue that policy in *460the future. The chief deputy director of CDC testified that his only experience with Union involvement in conflicts concerned a recent Susanville “situation” where the Union had an opportunity to act in an inflammatory manner but instead cooperatively helped to resolve the problems. The Union’s president pointed to its long-standing policy not to violate statutes or CDC rules, and respondent did not cite any instance to the contrary. Petitioners claim, without contradiction, that they consistently use legal channels, such as the instant proceeding, to oppose statutes, rules, and CDC action that they disapprove.
The Union witnesses discussed collective bargaining in terms of meeting and conferring or negotiating grievances through peaceful and rational persuasion. The testimony indicates that the Union hopes to develop a workable structure for negotiation with prison officials. The parties have pointed to authority on both sides of the question whether inmates should be involved in decision making on the conditions of life in prison. Whatever the merits of those policy arguments, I am far from persuaded that all forms of group representation in negotiation threaten security. Nor does it appear necessary to preclude meetings where such negotiation is discussed.
Petitioners predict that a prisoners’ union would improve communication between prisoners and staff and that prisoner frustrations thereby would be alleviated and the potential for violence reduced. Respondent wrote in his initial refusal letter that such beneficial results might occur and “I do not feel that the idea of an organization of inmates to deal with the institutional administration on common problems is entirely without merit.” Respondent’s other witness testified that he would recommend approval of any inmate organization providing the kind of access discussed in the trial court, except for collective bargaining; but his understanding of collective bargaining clearly did not include meet and confer or other negotiation forms considered by the Union. The CDC chief deputy director testified that in his opinion conventional notions of collective bargaining, inapplicable to a prison situation, were preventing acceptance of the Union; his testimony and that of all the other witnesses indicates that it is the prison officials, not the Union, who harbor the misunderstanding.
Finally respondent argues that no one can guarantee that the composition, leadership, and methods of the group will not change. I agree, taking note though of petitioners’ evidence of the Union’s stability over its 10-year history. At any rate, respondent has the authority to prohibit *461Union activity or otherwise protect institutional security whenever required in response to change.
In sum, though we must respect respondent’s professional judgment he has provided us with little or no support for his speculations, beliefs, and fears regarding the true security threat posed by the exercise of the important rights involved here. As the chief deputy director testified, changes in the prison system almost invariably have been anticipated with fear by CDC. We have witnessed the prison officials’ resistance in numerous cases in this court (and in the United States Supreme Court), and we have repeatedly ordered reforms in spite of those fears.3 Respondent’s fears and resistance in this case have not been supported by a showing that meets the requirements of section 2600. We should not permit the deprivation of prisoners’ rights on the basis of unsupported speculations and apprehensions. It is not a matter of our second-guessing prison officials’ judgment. Rather we should declare illegal their actions that violate the Legislature’s mandate.
Standard of review
The most relevant, recent prisoners’ rights case is Jones v. North Carolina Prisoners’ Labor Union, supra, 433 U.S. 119, where on facts similar to those before us the court held constitutional a ban on meetings and various other union activities. The court ruled that: a prisoner retains only those rights not inconsistent with legitimate penological objectives; since prison officials know best how to achieve those objectives, a court should uphold regulations justified by any rational beliefs of the officials; unless it is conclusively shown that those views are wrong the regulations should be upheld. Jones seems to be a return to a previously repudiated “hands-off” doctrine on prisons. (See Comment, The Future of Prisoners’ Unions: Jones v. North Carolina Prisoners’ Labor Union (1978) 13 Harv.Civ. Rights-Civ. Lib. L.Rev. 799, 802-804. Though the Jones majority opinion relies on earlier cases, none of them seems to set as low a standard of review of alleged infringement of prisoners’ rights as does Jones. Respondent suggests here that we interpret California law in accordance with those cases even if we shun the Jones analysis.
Two major decisions concern California prisons: Pell v. Procunier (1974) 417 U.S. 817 [41 L.Ed.2d 495, 94 S.Ct. 2800] (ban on face-to-face *462press interviews with inmates) and Procunier v. Martinez (1974) 416 U.S. 396 [40 L.Ed.2d 224, 94 S.Ct. 1800] (ban on paralegals and law students as interviewers of inmate clients). Both evaluate the constitutionality of restrictions by balancing individual rights against “the legitimate interests of penal administration.” (Martinez, supra, at p. 420 [40 L.Ed.2d at p. 244]; see also Pell, supra, at p. 822 [41 L.Ed.2d at p. 501].) They list as “legitimate interests” security, rehabilitation, deterrence, internal order, and discipline. The availability of any less restrictive alternative to accomplish the goals weighs against the regulation. As for deference to prison officials, Martinez places in balance “the proper regard that judges should give to the. expertise and discretionary authority of correctional officials” (Martinez, supra, at p. 420 [40 L.Ed.2d at p. 244]). Pell states that, if substantial evidence does not indicate an exaggerated response by the officials, courts should defer to their expert judgment in serving the policies and goals of penal administration. (Pell, supra, at p. 827 [41 L.Ed.2d at p. 504].)
The concurring opinions in Martinez pronounce as their approach that inmates generally retain the same rights as nonimprisoned citizens. (Procunier v. Martinez, supra, 416 U.S. at pp. 428 (Douglas, J.) and 422 (Marshall, J.) [40 L.Ed.2d at pp. 248, 245].) Section 2600 seems clearly to articulate that approach.
Some federal cases posit that prisoners come to court with fewer rights than free people, “[incarceration brings about the necessary withdrawal or limitation of many . '. . rights, a retraction justified by the considerations underlying our penal system.” (Price v. Johnston (1948) 334 U.S. 266, 285 [92 L.Ed. 1356, 1369, 68 S.Ct. 1049].) “[A] prison inmate retains those First Amendment rights that are not inconsistent with his status as a prisoner.” (Pell v. Procunier, supra, 417 U.S. at p. 822 [41 L.Ed.2d at p. 501].) Those words were quoted in Jones as support for the conclusion that courts should defer to the judgment of officials who are expert in the “complex realities” of administering a prison. The words also may support respondent’s assertion that “[p]ersons incarcerated in prison are, by definition, deprived of certain constitutional rights.” Such comments are, though, contrary to the words of section 2600.
In California the Legislature has created a powerful guarantee of prisoners’ rights. While reasonable restrictions may necessarily differ in nature and severity from those outside prisons, all rights are retained by prisoners except in the specific conditions provided by the statute. We *463should therefore require that prison officials make an affirmative showing that the statute permits their depriving prisoners of any right. We should not defer, as did the court in Jones, to the unsubstantiated “expert judgment,” “informed discretion,” and “reasonable belief” of the administrators. Section 2600’s declaration of prisoners’ rights does not permit deprivations on such bases.
This court’s opinions on the 1968 version of section 2600 provide some guidance as to the required showing. Harrell, supra, 2 Cal.3d 675 held that if a rule significantly impedes a prisoner’s right the burden of justification is on the prison officials. (Id., at p. 687.) “Remote and wholly speculative” predictions of security problems are not sufficient. (In re Jordan, supra, 12 Cal.3d 575, 579; In re Jordan, supra, 1 Cal.3d 930, 937.) Deprivations are not justified if there is no evidence that the claimed risk “is more than miniscule” (In re Jordan, supra, 7 Cal.3d at p. 938) or if there is no evidence of past security problems related to the risk. (Id.) The practices of other prison systems may be relevant. (Id., at p. 937.) Officials should show that there is no reasonably less-restrictive method for handling the risk. (Id., at p. 938.)
Respondent here has relied mainly on his and other prison officials’ opinions, conclusions, and speculations. He has not shown that the predictions of security risks in permitting meetings are supported in fact. He has not shown that his opposition to one Union objective requires that he prohibit discussion of that and all the other objectives. Nor, indeed, has he shown reasonable security grounds for prohibiting inmates from pursuing organized inmate-initiated prison reform and negotiation, which may well be a more efficient and effective means of securing needed reforms than are repeated court battles.
I would hold that on the basis of the record here respondent may not continue his absolute ban on Prisoners Union meetings. Bearing in mind that prisoners may be deprived of their rights only as section 2600 permits, he may of course impose reasonable restrictions on time, place, and manner. “Section 2600 cannot be construed as a straightjacket limiting the ability of prison authorities to deal with institutional realities.” (In re Harrell, supra, 2 Cal.3d 675, 698.) Monitoring, careful searching, and limits on the number of authorized groups and on numbers of outside participants are examples of conceivably reasonable regulations. If for any reason an actual security threat does develop,
*464respondent has discretion to take necessaiy action such as discipline, suspension, or even banning.
Respondent should be directed to cease enforcing his ban on Union meetings.
Bird, C. J., and Tobriner, J., concurred.
Respondent’s permissiveness concerning individual Union activity, though not a material issue here, is disputed by petitioners. In their traverse they allege and present evidence of denial of individual membership cards and correspondence. The recent case of In re Brandt, ante, page 136 [157 Cal.Rptr. 894, 599 P.2d 86] finds respondent defending his ban on inmate correspondence with a parolee on the ground that the correspondence involves Union business. In re Reynolds, ante, page 131 [157 Cal.Rptr. 892, 599 P.2d 83] concerns his ban on the wearing of Union buttons by inmates.
Some prisoners’ unions in other states have sought from state agencies recognition as bargaining units. Respondent presented one witness who testified- about recognition of a group by the Massachusetts Department of Human Resources. See also Rudovsky et al., The Rights of Prisoners: The Basic ACLU Guide to a Prisoner’s Rights (1977) pages 71-74.
See dissenting opinion of Marshall, J., in Jones v. North Carolina Prisoners’ Labor Union (1977) 433 U.S. 119 at page 142 [53 L.Ed.2d 629 at .p. 649]: “[Pjrison officials inevitably will err on the side of too little freedom.”