[17]   At p. 5 of its initial submission, CWHC sets out the following passage from Order No. 323-1999 (at p. 4) as a statement of the test under s. 19(1) of the Act:

Section 19(1) requires the head of a public body to be satisfied there is a reasonable expectation that disclosure of the requested information will threaten anyone else's mental or physical health or their safety or interfere with public safety. A reasonable expectation of a threat to health or safety requires something more than mere speculation. By importing into s. 19(1) the concept of 'reasonable expectation', the Legislature signalled its intention that speculation will not suffice to justify withholding of information. When faced with the reasonable expectation criterion - wherever it appears in the Act - the head of a public body must decide if a reasonable person who is unconnected with the matter would conclude that release of the information is more likely than not to result in the harm described in the relevant section of the Act. There must be a rational connection between the requested information and the harm contemplated by the Act, in this case as set out in s. 19(1).

[19]   In Order 00-02, I made the following observations, at p. 5, about s. 19(1):

Although s. 19(1) involves the same standard of proof as other sections of the Act, the importance of protecting third parties from threats to their health or safety means public bodies in the Ministry's position should act with care and deliberation in assessing the application of this section. A public body must provide sufficient evidence to support the conclusion that disclosure of the information can reasonably be expected to cause a threat to one of the interests identified in the section. There must be a rational connection between the disclosure and the threat. See Order No. 323-1999.

As I have said in previous orders, a public body is entitled to, and should, act with deliberation and care in assessing - based on the evidence available to it - whether a reasonable expectation of harm exists as contemplated by the section. In an inquiry, a public body must provide evidence the clarity and cogency of which is commensurate with a reasonable person's expectation that disclosure of the information could threaten the safety, or mental or physical health, of anyone else. In determining whether the objective test created by s. 19(1)(a) has been met, evidence of speculative harm will not suffice. The threshold of whether disclosure could reasonably be expected to result in the harm identified in s. 19(1)(a) calls for the establishment of a rational connection between the feared harm and disclosure of the specific information in dispute.

It is not necessary to establish certainty of harm or a specific degree of probability of harm. The probability of the harm occurring is relevant to assessing whether there is a reasonable expectation of harm, but mathematical likelihood is not decisive where other contextual factors are at work. Section 19(1)(a), specifically, is aimed at protecting the health and safety of others. This consideration focusses on the reasonableness of an expectation of any threat to mental or physical health, or to safety, and not on mathematically or otherwise articulated probabilities of harm. See Order 00-10.

... recognizes the need to provide evidence which is commensurate with a reasonable person's expectation that disclosure could threaten the safety, or mental or physical health, of third parties.

[23]   CWHC also relies, in its initial submission, on Order No. 7-1994, Order No. 18-1994 and Order No. 116-1996, all of which dealt with requests for information that would name or otherwise identify individuals involved in the provision of abortion services. These cases illustrate the importance of the specific evidence presented in each case in dealing with s. 19(1) (and with other exceptions under the Act). In approaching this case, I have been guided by the approach I adopted in Order 00-02 and Order 00-28. Order No. 323-1999 - which also dealt with a request for abortion statistics - is also of some relevance here.

[24]   In his initial submission, the applicant says the issue in this case is similar to that dealt with in Order No. 323-1999, i.e., "the release of statistics on the amount of abortions performed at" a public body. In his reply submission, he contends that this inquiry is "moot on the basis of Order No. 323-1999" and that the information should be disclosed accordingly. He argues that CWHC's evidence is similar to that presented in Order No. 323-1999, the result being that CWHC has not established that s. 19(1) applies. In his view, CWHC's case is based on "an imagined series of events" that would supposedly allow service providers to be identified through "non-identifying statistical information". The applicant argues that, because it is (according to him) "well known and in the public record" that CWHC "provides late-term abortions", the "issue of gestational age is also moot".

[25]   By contrast, CWHC says, at paragraph 15 of its initial submission, that Order No. 323-1999 does not "preclude the possibility that another institution could demonstrate harm from the release of statistical data for the purposes of the s. 19(1) test". It quotes the following passage from p. 6 of Order No. 323-1999:

The situation might be different if, unlike the case here, it is not publicly known that a particular hospital or clinic provides abortion services. If public confirmation of that fact alone could, in the circumstances, be reasonably be expected to threaten anyone else's health or safety, section 19(1) could well apply. This result may be even more likely if the hospital or clinic is in a small community and has minimal security arrangements available to it. The evidence in such cases would, of course, be determinative.

... (a) evidence of harm; (b) the unique circumstances of the public body; (c) the identity of the requester; and (d) nature of the information sought and evidence of the possible use of the information.

[28]   It is convenient to summarize the parties' evidence and argument at this point. The following summaries do not reflect findings on my part. The applicant's case can be summarized as follows:

• CWHC readily admits to performing abortions, including through information found on its Website and through other public information.

• CWHC has, in the past, announced that it would be "doing over 2000 abortions per year".

• CWHC has been actively involved in lobbying for the expansion of the availability of abortion services in British Columbia.

• The disputed information could not be used to identify those associated with abortion services or individual patients.

• Much of CWHC's evidence about harassment and threats, and notably respecting the applicant's activities, is based on hearsay and opinion and is seriously flawed.

• The applicant has used information about abortion statistics responsibly in the past, including by successfully asking for a coroner's inquest respecting live births after late abortions.

• Since I ordered VGH, in Order No. 323-1999, to disclose abortion statistics, there have been no reports of increased illegal activity and the events VGH predicted would occur have not occurred.

• CWHC evidence attempts to link anyone who is opposed to abortion with those who have wrongly used violence as a means to achieve their goals, amounting to guilt by association.

• There is a vigorous debate throughout North America about abortion and CWHC's attempt to withhold information is designed to shut down all debate and public accountability.

• British Columbia's political climate is often vigorous. This includes preventing those with pro-life views from speaking at universities and attacks by government members on those with pro-life views.

These are serious charges, I admit, but the affidavit evidence before you can offer no other conclusion. I am sorry to thrust your office into the middle of the political debate over abortion. But this issue, and the issue of late-term abortion, is what bioethics is all about. And bioethical issues can and must be debated in society. The medical and academic community for too long have laid claim to the right of debate on these issues. I disagree. In a free and open society all moral issues are open to debate. Private and, if the need be, public opposition to social policies are a healthy part to any democracy. How contemptuous to suggest that anyone who opposes abortion is somehow connected with fringe elements who would violate the first and foremost right that governs our cause - the right to life of everyone - a right, I may add, that even extends to those we philosophically disagree with.

Yet the tone and nature of the submission of the public body is one where all who oppose abortion are somehow suspect. The tone and nature of the submission of the public body is one where false statements and information are presented as factual. Where private political opinion and hearsay is presented as concrete truth.

The public body desires no public accountability on these issues. It seeks no public debate and wishes to hide its activities from the court of public opinion.

The evidence before you offers one conclusion. The private [sic] body has been unable to prove their case regarding release of the information requested.

[31]   The aspects of CWHC's evidence which I am at liberty to summarize, with CWHC's consent, are as follows:

• CWHC is a significant target for anti-abortion activists because of its unique and diverse responsibilities in maintaining and facilitating abortion services throughout British Columbia.

• CWHC's unique and diverse responsibilities place service providers, nursing staff and administrators at a far greater risk of harm.

• There is no other area of health care where patients and service providers are subjected to fear, intimidation and risk of harm as they are in the area of abortion services.

• Abortion service providers have been subjected to harassment, death threats and attempted murder. One physician was the victim of an attempt on his life in 1994. (I also take notice of the fact that a second attempt was made on that physician's life after the close of this inquiry.)

• Reactions to these attacks have ranged from disbelief that such a thing could occur, to fear, grief and anger for someone who was nearly killed.

• There is regular picketing activity close to the facility, on the CWHC site, where abortions are performed.

• Anti-abortion activists have entered CWHC property and left literature in public waiting areas and on vehicles in the parking lot. This literature causes extreme distress to families who are already suffering anguish.

• There is no access zone in place at CWHC under the Access to Abortion Services Act. (Such access zones restrict individuals' access to public and other property adjacent to designated abortion services facilities, to permit unimpeded and safe passage of patients and staff.)

• The applicant has published information on his Website concerning live births and has been successful in using data he obtained to have a coroner's inquest convened into his allegations concerning the alleged murder of infants at CWHC.

• The applicant arranged, earlier this year, for distribution of anti-abortion material, focusing on this inquiry, in the parking lot at CWHC. (The applicant strenuously denies any involvement in this act.)

• The requested information could be used with other information - as an example of the mosaic effect (which I discuss further below) - to determine how many doctors are performing abortions at CWHC and, through a process of elimination, to identify physicians whose identities are not known at present. This will place those service providers and CWHC staff at grave risk of harm.

Analysis of the Evidence

[32]   As a preliminary point, I agree with the applicant that aspects of CWHC's case tend to lump all anti-abortion activists together with those who resort to harassment, threats and violence to oppose abortion. Moreover, CWHC's case in some places comes close to suggesting that, because of the climate of violence and fear that surrounds abortion services, the disclosure of any information that is any way related to abortion satisfies the s. 19(1) test. For example, one of the individuals who swore an in camera affidavit deposed that the disclosure of "any information is an act of terrorism in evolution" and CWHC so argued. I readily acknowledge that abortion service providers have deep-seated concerns for their own health and safety (and that of their families), but I do not accept that disclosure of any abortion-related information by definition always satisfies the s. 19(1)(a) or s. 19(1)(b) test.

[33]   I have observed before that public bodies should, in light of the health and safety interests at stake, act with care and deliberation in deciding whether to apply s. 19(1) in a specific case. As I said in Order 00-28, one of the "contextual factors" relevant to s. 19(1) is that it deals with threats to the health or safety of individuals. This is to be contrasted with, for example, the financial and economic interests protected by s. 17 or 21 of the Act. In Ontario (Minister of Labour) v. Big Canoe [1999] O.J. No. 4560, the Ontario Court of Appeal considered that the purpose of s. 20 of Ontario's Freedom of Information and Protection of Privacy Act - a provision that is very similar to s. 19(1) - was relevant to application of that provision.

[34]   To echo what I said earlier, I am precluded from setting out as detailed an account as I would like of my thinking in this case. I can say that I have concluded, on the basis of the evidence provided by CWHC, that s. 19(1)(a) applies because: (a) there is a reasonable expectation that disclosure of the disputed information could be used (together with publicly available information) to identify physicians who provide abortion services at CWHC, but whose identities are not publicly known at this time and (b) once the identities of those service providers are known, there is a reasonable expectation of a threat to their safety or mental or physical health harassment, threats and violence. The reasons for this conclusion, to the extent I can articulate them, follow.

[35]   My assessment of CWHC's s. 19(1)(a) case on this issue - most notably as set out in its in camera evidence and argument - leads me to conclude that there is a reasonable expectation this information could be used, together with information available publicly or otherwise, to discern the identities of service providers. This conclusion depends entirely on the facts before me in this case, which stem from the unique role CWHC plays in abortion services and the manner in which those services are delivered. It also turns on the nature of how abortion services are provided at CWHC and by whom. This finding relates to the annual statistics sought by the applicant taken alone or in conjunction with the second trimester data covered by the request. CWHC's case may be more forceful if one considers both sets of data together, but I am persuaded in this case that the annual statistics alone are also excepted from disclosure under s. 19(1)(a).

[36]   The applicant disputes this. He says the requested statistics could not be used to identify abortion service providers. At p. 3 of his initial submission, he says "there is no attempt nor could anyone determine the personal identities of health care providers" from the statistics he seeks. At pp. 4 and 5 of his reply submission, he says the following:

I would submit that the information I have requested is of a non-identifying nature only. If we consider the fact that the public body proudly announced they would be doing over 2,000 abortions per year, and that this number has remained constant, there is no way to discover the identity of an individual provider anymore than I could knowing that the abortion rate for all of British Columbia is 16,000 per year. A fact on the public record.

[37]   The applicant makes the same point in relation to affidavit evidence provided by a police officer, on behalf of CWHC, on the issue of service provider identification.

[38]   In his further reply, the applicant describes CWHC's case on this point as a "novel theory" and a "ludicrous theory". He notes that a November 4, 1999 media release issued by the Ministry of Health "provides the exact number of practitioners in British Columbia that are directly involved in the provision of abortion services".

[39]   I do not suggest for a moment that the applicant would attempt to use the information to identify service providers or would threaten them with, or cause them, any harm. There is, however, evidence - including through the applicant's own words - that he exercises his freedom of expression by publicizing abortion-related information. It is reasonable to conclude, as was the case in Order No. 323-1999, that disclosure of information to the applicant amounts to disclosure to the world. Once the applicant disseminates information to the world at large, he ceases to have any control over its use. It is the use of information by others, together with other information, to identify CWHC service providers that is of concern here. In this sense, the identity of the applicant is, as was foreshadowed in Order No. 323-1999, relevant to the s. 19(1)(a) issue.

[40]   Again, there is evidence before me of a reasonable expectation that the disputed information could be used to identify abortion service providers. This is an example of what is often called the 'mosaic effect'. The term describes the result where seemingly innocuous information is linked with other (already available) information, thus yielding information that is not innocuous and, in the access to information context, is excepted from disclosure under the Act. The fact that, according to the applicant, the number of "practitioners" involved in abortion services in British Columbia has been published does not mean the information in dispute here - which relates to a single, arguably unique institution - could not be used to identify individuals through the mosaic effect.

[41]   The mosaic concept is usually encountered in intelligence and law enforcement contexts. In United States v. Marchetti, 466 F.2d 1309 (4^th Cir.), cert. denied, 409 U.S. 1063 (S.C.), for example, the U.S. Court of Appeals said (at p. 1318) that, due to the

... mosaic-like nature of intelligence gathering ... what may seem trivial to the uninformed may appear of great moment to one who has a broad view of the scene and may put the questioned item of information in context.

[42]   The mosaic effect is also encountered in access to information access cases. In Ternette v. Canada (Solicitor General), [1991] F.C.J. No. 1168, an individual had, under the Privacy Act of Canada, applied for access to his own personal information. At para. 45, the Federal Court of Appeal described the mosaic effect as what happens when

... one takes seemingly unrelated pieces of information, which may not be particularly sensitive individually, and compares them with each other to develop a more comprehensive picture.

[43]   More recently, in Ruby v. Canada (Solicitor General), [2000] 3 F.C. 589, the Federal Court of Appeal referred, at para. 85, to the expert evidence before the Court which described the mosaic effect as a process "whereby seemingly unrelated pieces of information could be compared with each other to develop a more comprehensive picture resulting in disclosure of" exempt information. Both Ternette and Ruby, while they are information access cases under the federal Privacy Act, relate to security and intelligence matters.

[44]   Closer to home, my predecessor found that the mosaic effect applied in Order No. 81-1996, where he found that "disclosure of one piece of the puzzle may disclose everything." The existence of the mosaic effect is also acknowledged in the Policy and Procedures Manual, published by the Information, Science and Technology Agency of the Ministry of Advanced Education, Training and Technology for use by public bodies. It is found at http://www.ista.gov.bc.ca/foi_pop/revised_manual/ToC.htm, in the discussion on the interpretation of s. 15 of the Act.

[45]   Let me emphasize that I am not, in finding that the mosaic effect applies in this case, expressing the view that disclosure of such statistics satisfies s. 19(1)(a) in all cases. As I said above, everything turns on the facts of each case, as established in the evidence before me as to the circumstances of the particular public body and the specific information in issue. As for the mosaic effect generally, a public body will be able to invoke it only where the evidence it has adduced establishes that it applies. Cases in which the mosaic effect applies will be the exception and not the norm.

[46]   Having found there is a reasonable expectation that the requested information could, in this case, be used to identify service providers, I also find, on the evidence, that, where the identities of physicians and other health care workers associated with CWHC are known, they have suffered harassment, stalking and violence (including attempted murder). It is reasonable to expect, therefore, that if the disputed information is used to identify service providers, service providers are threatened with these kinds of harm. There is, in this case, a reasonable expectation that disclosure of the information could be used to identify individuals and that threats to their safety or health will arise within the meaning of s. 19(1)(a). This satisfies the s. 19(1)(a) test. It is not necessary, in this light, for me to consider s. 19(1)(b) of the Act.

[47]   By way of conclusion, I should comment on a recent Ontario case on which the applicant relies, Order PO-1747 (issued January 26, 2000). In that case, an applicant requested access to statistical information respecting the number of obstetricians and gynecologists who billed the Ontario Health Insurance Plan for one or more therapeutic abortions, in each of several years, and the number of therapeutic abortions billed to that plan in each of those years. Senior Adjudicator David Goodis found that s. 20 of Ontario's Freedom of Information and Protection of Privacy Act had not been satisfied and ordered the Ontario Ministry of Health to disclose the requested information. Section 20 of that Act reads as follows:

A head may refuse to disclose a record where the disclosure could reasonably be expected to seriously threaten the health or safety of an individual.

[48]   Senior Adjudicator Goodis concluded the Ministry had failed to establish a reasonable expectation of a serious threat to anyone's health or safety. He noted that the requested information was "non-identifying province-wide statistical information" (p. 7). He also noted, at p. 8, that, in the United States, "generalized, statistical information similar in nature to the requested information regarding numbers of abortions is widely available, mainly on the basis of statutory requirements". He observed that, in some states such as Connecticut, "raw data" or "statistical information" respecting numbers of abortions performed must be disclosed where there is no "reasonable risk of identifying the subject of an abortion or the individual performing such abortion". See William C. O'Brien and Connecticut Right to Life Corporation and State of Connecticut (Commissioner, Department of Public Health) (Freedom of Information Commission, State of Connecticut, Docket FIC 1997-092, December 3, 1997). Senior Adjudicator Goodis concluded that the information in dispute in Order PO-1747 "could not be linked to any individual facility or person involved in the provision of abortion services", such that the necessary reasonable expectation of a serious threat to the life or safety of any person had not been established for the purposes of s. 20.

[49]   Order PO-1747 does not assist the applicant here. It is clear that the information in dispute there was generalized, province-wide statistical information that did not relate to or identify a specific facility or individual. By contrast, in Ontario Order P-1499 - a case relied on by CWHC - it was decided that information that would identify individual facilities where abortions were performed could reasonably be expected to lead to the harms described in s. 14(1)(e) of the Ontario Act (which is similar to s. 15(1)(f) of our Act). (This is not to say that Ontario Order P-1499 lays down a general rule on the disclosure of facility-specific statistics that applies in British Columbia.)

[50]   3.3 CWHC's Section 15 Case - In addition to relying on s. 19(1), CWHC argues that ss. 15(1)(f) and (l) authorize it to refuse to disclose the requested information. Those sections read as follows:

15(1)

The head of a public body may refuse to disclose information to an applicant if the disclosure could reasonably be expected to ... (f) endanger the life or physical safety of a law enforcement officer or any other person, ... (l) harm the security of any property or system, including a building, a vehicle, a computer system or a communications system.

[51]   In light of my findings with respect to s. 19(1)(a), it is not necessary for me to deal with s. 15(1) in this case.

4.0 CONCLUSION

[52]   For the reasons given above, under s. 58(2)(b) of the Act, I confirm the decision of CWHC to refuse access to the disputed information.

January 16, 2001

ORIGINAL SIGNED BY

---

David Loukidelis
Information and Privacy Commissioner
   for British Columbia

 January 16, 2000
Information and Privacy Commissioner of British Columbia