(dissenting). I most respectfully dissent. While I do not disagree with the points made in my fellow dissenter’s opinion, my own analysis of the legal and factual strengths to be found in the petitioner’s argument is somewhat different. Given that the regulation of abortion is the focus of enormous contemporary concern, I state that analysis.
I begin with the conviction that the city’s regulation compelling the exposure of all the women’s names and addresses to availability in a central data bank comes sufficiently within the constitutionally protected right to privacy elaborated in Roe v Wade (410 US 113) and Doe v Bolton (410 US 179) so that it must be justified not merely on a rational basis but by that higher standard of utter necessity which we denominate a compelling State interest. (See Griswold v Connecticut, 381 US 479; NAACP v Alabama, 377 US 288; Shapiro v Thompson, 394 US 618; Eisenstadt v Baird, 405 US 438, 460 [White, J., concurring].) As these cases point out, once a nexus is established between the Constitution and a particular right, any infringement of that right by the State must meet the higher test.
It is not now open to us to question whether abortion involves such a constitutional right. In Roe v Wade (supra) and Doe v Bolton (supra), the United States Supreme Court unequivocally declared that the right to an abortion is a medical matter within the range of the doctor-patient relationship and is to be accorded the constitutional dimensions of the right to privacy. Its recognition, in those cases, of the fact that the State has a strong interest in protecting women’s health and safety is to be read in the light of its extension of constitutional protection to abortion. In advancing a balance, the court did not sanction every curiosity a State might wish to express about the abortion process. Nothing in the two opinions suggests that every possible manifestation of the State’s interest has attained a status which could be utilized *250to put it beyond the scrutiny which must always be applied to State interference with a constitutional right.
Under the compelling interest test, not only must the State’s interest be one which is related to a vital and important need, but also the means utilized in order to achieve the desired goal must be necessary ones. Where less intrusive means exist they must be used, even if they are somewhat less efficient and even if their use, perchance, involves greater expense to the State. (NAACP v Alabama, supra, at p 307; Shelton v Tucker, 364 US 479, 488; Cantwell v Connecticut, 310 US 296, 304; NAACP v Button, 371 US 415, 438; and see Reed v Reed, 404 US 71.)
It is not disputed or at issue here that, without any centralized data bank, the names and addresses of women receiving abortions are already contained in the medical records kept by the hospitals themselves. Indeed such records, in accordance with standard hospital practice, cover all patients in such institutions, not just those who abort, and, under proper circumstances, such hospital records are available to city health officials. Since the presumed goal of the regulation is the safety of women who undergo abortions, however, and not the obtaining of their names, the requirement that names be centralized is itself but a means selected to achieve that goal. It follows that the city must meet the test of necessity by proving that no less intrusive means exist.
The city has failed to meet that test. In its attempt to show true necessity for mass centralization of. the data, it has offered the following reasons why it would like to be able to identify all of the women who have abortions:
1. Follow-up where the patient died and the doctor could not otherwise be found;
2. Follow-up where complications or coma occur;
3. Investigation of facilities to determine whether proper procedures are being used;
4. Research on the effects of multiple abortions and counseling on alternative means of birth control; and
5. Collection of vital statistics on abortion.
Careful analysis of both the logic inherent in these justifications for the city’s regulation and of the actual experience with it, as documented in the papers before us, indicates that wholesale collection of names and addresses is not warranted.
*251Though vital statistics on abortion are needed, obviously names are not essential for that purpose.
The desirability of follow-up of complications or coma may have superficial appeal. However, since these follow abortion very quickly, it is clear that the slow processing of hundreds of thousands of routine certificates at a central data bank will not result in meaningful help to a woman in such a condition. In most cases, treatment would be given by the abortion center and the data would be provided on the certificate itself.
The city nevertheless argues that there may be instances in which a woman might be admitted with complications to a hospital other that the one from which she received the abortion and might be unable or unwilling to tell authorities where the abortion itself was performed. But it has failed to demonstrate that the inclusion of names on the certificates is crucial to its ability to deal with such cases or that they occur with sufficient frequency so that other means of follow-up are not effective. Its very “Report from the City Department of Health”, though based on four years of experience with the regulation, in attempting to document a need for the names on certificates, proves instead that they are not necessary.
In fact, the entire report cites but five examples of such follow-up altogether. In the first, the woman herself identified the abortion hospital to the second hospital. In the second example, the woman again supplied the information needed; follow-up disclosed that the physician never filed her certificate. In the third example, data, not names, on certificates filed by a particular doctor was sufficient to reveal that he was performing abortions though he was only a general practitioner; on-site investigation turned up bad medical practices as to which he had filed false information on the certificates. The fourth example again involved an institution identified as guilty of repeated bad practices by information other than that on the certificates; on-site investigation revealed that one of the bad practices was failure to file these papers. In the fifth and final example, which involved an institution suspected of performing abortions after the twelfth week of pregnancy by methods not approved for use after that date, the problem was brought to the authorities’ attention by a patient herself; her certificate had never been filed. Interestingly, in a number of these examples, the women involved used false names at the abortion center in an attempt to maintain their privacy, and real ones at the hospital later, *252thus indicating that, if anything, the regulation tends to be self-defeating. What all of this information demonstrates vividly is that supervision of abortion practices can be and is carried out by investigative methods far more effective than the use of names on the certificates.
In this context, it is also important to point out that the city’s added contention that it needs the names of the women in order to follow up the possible existence of other diseases is also vulnerable to the test of necessity. Abortion is itself neither contagious nor dangerous to society in the way that venereal disease or tuberculosis is. Society possesses other means to follow up unrelated diseases which may occur in conjunction with pregnancy as, indeed, they may occur together with any other medical condition. A separate report on women with such diseases could be required, for example, if one is not already mandated. More significant, the invasion of women’s privacy in this constitutionally protected area in order to facilitate the achievement of health goals which are unrelated to abortion is so clearly prohibited by the Supreme Court decisions as to require no further comment here.
Finally, we address the city’s contention that the city has no other way to obtain data on the effects of multiple abortions or to counsel women who appear to be using abortion as their sole contraceptive to utilize other, less risky means. Surely, there are other and less objectionable ways to correlate data on such women; if, for example, the information other than name and address on the original certificates included specification as to the patients’ previous abortions, that would suffice. Indeed, as both parties have acknowledged, and as the city’s own examples have illustrated, the information supplied is likely to be more accurate and complete if its revelation is not inhibited by the name requirement.
Disturbing too is the implication that the regulation is intended to enable the city to seek out certain categories of abortion patients and advise them, paternalistically, of the best contraceptive procedures to follow. Even the city does not make so bold as to argue outright that it may so invade the constitutionally protected privacy of those who do not wish to be advised. Such advice would be more appropriately given by the woman’s own doctor, than by the city.1
*253In sum, not one of the reasons advanced as to why the city needs to know the names of all women who obtain abortions demonstrates the level of necessity required here.
I also note that the majority, in justifying the State’s interest in women’s health, emphasizes the problems associated with the second trimester of pregnancy. Indeed, as I read its opinion, the majority fully acknowledges the fact that the Supreme Court cases made abortion during the first trimester of pregnancy very much a private matter between a woman and her doctor. Those opinions indicate that the State’s interest reaches a compelling level only at the second trimester. I note further that the record indicates that most of the women who have abortions do so in the first trimester. Thus, the city’s argument here amounts to a contention that it may invade the declared privacy of women who have abortions in the first trimester simply in order to enhance its ability to regulate second trimester abortions. Such an argument obliterates completely the careful distinctions between the two trimesters drawn by the Supreme Court. I find this a classic example of regulation which sweeps too broadly.
In retreat from its unproved assertions, the city falls back on the argument that, since the names are recorded in hospital records anyhow, the harm which may arise from centralizing them is, in any event, de minimis. Building on that, it points out that there is no roster of women ready to come forward and testify that the existence of the data bank in fact chilled their exercise of the right to seek an abortion. It is well recognized, though, that such testimony is not essential; direct proof of a chilling effect is frequently elusive and its existence, especially in a constitutionally protected area, may often be inferred. (Gooding v Wilson, 405 US 518, 521; Coates v City of Cincinnati, 402 US 611, 616; Cantwell v Connecticut, supra; Griswold v Connecticut, supra; Friendship Med. Center *254v Chicago Bd. of Health, 505 F2d 1141.)2 Also, there is an enormous difference between the knowledge that one’s name exists, together with the names of all patients, not just abortion patients, merely in a hospital record somewhere in an immense metropolis, and the knowledge that the local political body has compiled a special file of all those who have sought this particular treatment.
Abortion is still an emotionally charged and heavily politicized topic today. Even were that not so, it is and always will be closely related to the most intimate decisions in the area of sex and family rearing an individual can make, and, because of the recency of the change in judicial and legislative attitudes, we know that many groups within our society still would stigmatize women involved in abortions if they could do so. Although this tendency may be expected to abate with time, one would have to blind oneself to reality to deny that the potential for stigmatization, intended or inadvertent, is multiplied by the unnecessary collection and centralization of such information.
This view was confirmed most recently in Roe v Ingraham (403 F Supp 931 [three-Judge court]). There, the court struck down a New York statute which required that the names and *255addresses of all patients obtaining drugs on the so-called "Schedule II”3 of sections 3331, 3332 and 3334 of New York’s Public Health Law be sent to Albany for inclusion in a centralized data bank. That statute was designed to identify both patients who go from doctor to doctor to obtain more drugs than their conditions justify and also doctors who over prescribe for individual patients.
In Ingraham, thé State’s goal was materially aided by possession of the patient’s name and was unquestionably a laudable one. Yet, on the authority of Roe v Wade (410 US 113, supra), and Doe v Bolton (410 US 179, supra), the centralized data collection system, the process the State chose to use there, was found to be unjustifiably violative of the right to privacy on two grounds. Noting first that in some 20 months of operation only two incidents of possible misuse of drugs had been uncovered (and even those cases were not clearly matters of illegality), the court held that the intrusive means was not justified by the results obtained. More significantly, the court emphasized the stigmatizing and, therefore, chilling effect such centralization may have upon the exercise of the right of privacy and explicitly found it irrelevant that patients’ names already existed on the registers of the pharmacies from which they bought their prescriptions. 4
In the case before us, there is not even such agreement that the possession of the names materially facilitates safer abortions. As noted above, the reasons advanced to justify possession of the information are not impressive; the addition of centralization, far from being de minimis, compounds the *256offense against the right involved. Because there appears to be no legitimate need for the name of every woman who seeks abortion, centralization of such information cannot be defended on the ground that any particular name or names could, with sufficient effort, be culled from hospital records.5 The modern technology which makes possible swift and efficient collection of data other than names and addresses in order to make abortions safer constitutes a welcome advantage; its capacity to provide grist for the mill of data banks, however, does not provide any justification for insensitivity to unnecessary intrusions on privacy.
The right to privacy, though slow and gradual in its articulation as constitutional doctrine, has come to play an increasingly powerful and pervasive antidotal role in keeping the potential abuses of electronic and other scientific devices in check. (Henkin, Privacy and Autonomy, 74 Col L Rev 1410; Konvitz, Privacy and the Law: A Philosophical Prelude, 31 Law and Contemp Prob 272.) It is no less important a doctrine because its reach cannot always be defined with precision. It surely applies here.
Thus, the cautious and careful step-by-step investigation of abortion as a newly-popular medical procedure, urged here by the city, ought also to be applied by the city in its use of technology in this area, for it too is new. And just as people may be divided over the unknown effects of legal access to abortion so also are they divided over and mistrustful of the potential of computers in government. Headlines in the recent past, and even in the present, provide ample basis for concern that centralized data banks may be misused and that neither legislation which attempts to control such use nor bland assurances that the public good is of foremost importance is any guarantee that the data will not be misused. Where the centralization of data made possible by technology serves a vital and otherwise unattainable goal it must be permitted, and, indeed, welcomed, but where, as here, it conflicts with and threatens a constitutional right and is not truly necessary, it ought not be used just "because it is there”.
*257Accordingly, the regulation challenged here should be declared constitutionally invalid and the order of the Appellate Division sustaining it reversed.
Chief Judge Breitel and Judges Jasen and Jones concur with Judge Gabrielli; Judges Wachtler and Fuchsberg dissent and vote to reverse in separate opinions in each of which the other concurs and in both of which Judge Cooke concurs.
Order affirmed, without costs.
. We note that many of the abortions performed in New York City are provided to women who are receiving some form of public assistance or free medical care. (See respondent’s brief, pp 32-33, wherein indigent women are specifically targeted for *253such advice.) The dangers inherent in even well-meant attempts to counsel such women are apparent from such recent cases as Self v Weinberger (372 F Supp 1196 [sterilization urged upon poor, black, allegedly retarded teenaged girls]) and Klein v Nassau County Med. Center (347 F Supp 496, vacated and remanded for consideration in light of Roe v Wade, 410 US 113 and Doe v Bolton, 412 US 925 [law which forbade use of public funds for elective abortions held violative of equal protection principles]; see, also, Babcock, Freedman, Norton & Ross, Sex Discrimination and the Law, pp 987-990; Doe vRose, 499 F2d 1112).
. The majority suggests that case law has overruled this principle; I do not so read it. Later cases, such as those cited by the majority, have merely suggested that defendants whose own conduct was not covered by the chilling portions of a statute do not always have standing to raise that question. The general rule set down in the cases cited- here, and in countless other cases, permits the courts to find a chilling effect where one plainly exists without the necessity for statistical proof of actual chill. In Laird v Tatum (408 US 1), also cited by the majority, the court found first that the petitioners had not themselves been subjected to the practices they sought to challenge (surveillance of private citizens on political grounds by the Army), thus reaching and deciding the merits, and then backed up to declare that the petitioners were without standing on that ground. I do not admire the decision, since, among other things, it violates that court’s own standing doctrines, and I see no reason to emulate it even if it were relevant; it is not, for there is no question of standing in the case before us. Moreover, the amicus brief filed by the National Organization of Women (NOW) indicates that, at least in the experience of NOW as an organization involved in the abortion issue, the number of women who seek information out of a concern that their names may be used to injure them is substantial. I also note that the present practice in courts, both State and Federal, of permitting suits in this area to be brought under protective pseudonyms provides at least some evidence that fears of stigmatization are sufficiently realistic to merit recognition by the courts. (See, for example, Doe v Doe, 314 NE2d 128 [Mass Sup J Ct]; Doe v Dunbar, 320 F Supp 1297; Doe v Scott, 321 F Supp 1385; Hall v Lefkowitz, 305 F Supp 1030; Poe v Menghini, 339 F Supp 986; Beecham v Leahy, 130 Vt 164; Doe v General Hosp. of Dist. of Columbia, 434 F2d 427; Coe v Gerstein, 376 F Supp 695; Doe v Hampton, 366 F Supp 189; Roe v Norton, 380 F Supp 726; Doe v Poelker, 497 F2d 1063; Roe v Fergusen, 515 F2d 279; Doe v Westby, 402 F Supp 140.)
. Schedule II includes such drugs as ritalin, codeine, percodan, morphine, and hycodan. There are drugs which have a great many vital and legitimate uses as well as possible addictive or other undesirable features.
. In Lamont v Postmaster Gen. (381 US 301), the court struck down a law which authorized the postmaster to withhold (p 302) "communist political propaganda” and to so notify potential recipients, who were then required to return a postcard if they wished to receive the mail in question. In rejecting the postmaster’s argument that this interference represented only an inconvenience since the addressee was ultimately able to receive the mail, Mr. Justice Brennan stated (p 309): "But inhibition as well as prohibition * * * is a power denied to government * * * In any event, we cannot sustain an intrusion on First Amendment rights on the ground that the intrusion is only a minor one. As the Court said in Boyd v. United States, 116 U. S. 616, 635: 'It may be that it is the obnoxious thing in its mildest and least repulsive form; but illegitimate and unconstitutional practices get their first footing in that way, namely by silent approaches and slight deviations from legal modes of procedure. This can only be obviated by adhering to the rule that constitutional provisions for the security of person and property should be liberally construed.’ ”
. This conclusion also suffices to dispose of the city’s belated effort, made first in the Appellate Division, to argue that, because its data is computerized, identification numbers supplied by the hospital and coded to its records of names and addresses could not be utilized successfully. Parenthetically, it should be noted that, though the argument was made, the city never adequately explained precisely why some such numerical system could not be implemented.