Date: 20170210
Docket: T-2587-14
Citation: 2017 FC 167
Ottawa, Ontario, February 10, 2017
PRESENT: The Honourable Mr. Justice Russell
BETWEEN:
|
KELLY A. O’GRADY
|
Applicant
|
and
|
ATTORNEY GENERAL OF CANADA
|
Respondent
|
JUDGMENT AND REASONS
I.
INTRODUCTION
[1]
This is an application under s 18.1 of the Federal Courts Act, RSC 1985, c F-7 [Federal Courts Act] for judicial review of a decision of the Chief of Statistics, dated January 7, 2011 [Letter of Agreement], whereby Statistics Canada entered into an agreement with McGill University’s Faculty of Medicine [McGill] to conduct a study examining perinatal outcomes in Canada.
II.
BACKGROUND
[2]
Statistics Canada is a statistical agency with a mandate to collect, analyze, abstract, and publish statistical information related to the social and general activities and conditions of Canadians. As part of this mandate, a census of Canada’s population is taken every five years, which is performed via a mandatory long-form or short-form questionnaire. In its analysis, Statistics Canada employs record linkages, which combine two or more micro-records to form a composite record. The use of record linkages is performed in a manner that minimizes privacy intrusions, requires approval from Statistics Canada’s Executive Management Board, and requires informing the public of such usage.
[3]
In 2011, Statistics Canada and McGill entered into a Letter of Agreement to conduct a study that would assess infant mortality and newborn health by examining perinatal outcomes in Canada according to risk factors related to socioeconomic position, ethno-cultural background, and environmental exposure [Study]. In connection with the Study, record linkages were used to link information from the national birth record database and the 1996 and 2006 censuses. In order to minimize the privacy intrusion, the record linkages were performed in accordance with s 6 of the Statistics Act, RSC 1985, c S-19 [Statistics Act] by Statistics Canada employees, or deemed employees, and the composite records were stripped of direct personal identifiers before they were made accessible to McGill. The composite records were also restricted to Statistics Canada’s premises. Additionally, the usage of the record linkages was publicly posted on the Statistics Canada website.
[4]
The criteria for inclusion in the Study included persons who were born between 1985 and 2010 to a Canadian mother, under which approximately 9.5 million individuals qualified. Since the Applicant gave birth in Ontario between 1986 and 1996, her census information could have qualified for usage in the Study. However, there was an issue with data quality concerns and Ontario birth records between 1994 and 1996, including the Applicant’s, were excluded from the Study.
[5]
On May 24, 2012, the Applicant filed a complaint with the Privacy Commissioner alleging that Statistics Canada had contravened the use provisions of the Privacy Act, RSC, 1985, c P-21 [Privacy Act] when her census information was linked with the birth records for use in the Study without her consent.
[6]
A decision sent by the Privacy Commissioner to the Applicant by letter dated November 21, 2014 determined that the Applicant’s personal information had not been improperly used by Statistics Canada.
[7]
The Privacy Commissioner agreed that the Applicant’s census information met the definition of personal information, as defined by s 3 of the Statistics Act. Additionally, the Privacy Commissioner found that usage of census information in the Study was beyond the scope of the purposes for which it was collected, which is prohibited under s 7 of the Statistics Act. However, there was no evidence to suggest that the Applicant’s information had actually been used in the Study as her information had been excluded. Furthermore, even if the Applicant’s information had been used, Statistics Canada had the authority to do so under the Statistics Act. Consequently, the Privacy Commissioner found that the Applicant’s complaint was not well-founded.
III.
DECISION UNDER REVIEW
[8]
Statistics Canada made a decision to enter into a Letter of Agreement with McGill to assess infant mortality and newborn health by examining perinatal outcomes in Canada according to risk factors related to socioeconomic position, ethno-cultural background, and environmental exposure. Under the agreement, McGill would pay $380,000 to Statistics Canada to perform record linkages and analyze census information. The record linkages would involve the use of birth records in the two years prior to 1996 and 2006 and would be conducted by Statistics Canada employees or deemed employees at Statistics Canada’s premises. The Applicant is seeking review of this agreement.
IV.
ISSUES
[9]
The Applicant submits that the following are at issue in this application:
What is the appropriate standard of review?
Did the decision-maker exercise his discretion in a manner that violated s 18.1(2) of the Statistics Act?
Does the Applicant have standing to seek judicial review?
What is the appropriate remedy?
[10]
The Respondent submits that, in addition to the Applicant’s submitted issues, the following are also at issue in this application:
Does the Court have jurisdiction?
If so, was the Privacy Commissioner’s non-binding finding reasonable?
V.
STANDARD OF REVIEW
[11]
The Supreme Court of Canada in Dunsmuir v New Brunswick, 2008 SCC 9 [Dunsmuir] held that a standard of review analysis need not be conducted in every instance. Instead, where the standard of review applicable to a particular question before the court is settled in a satisfactory manner by past jurisprudence, the reviewing court may adopt that standard of review. Only where this search proves fruitless, or where the relevant precedents appear to be inconsistent with new developments in the common law principles of judicial review, must the reviewing court undertake a consideration of the four factors comprising the standard of review analysis: Agraira v Canada (Public Safety and Emergency Preparedness), 2013 SCC 36 at para 48.
[12]
The Applicant submits that the standard of review is correctness. She says that the central issue is whether the decision-maker, a government official rendering an administrative decision, exercised discretion in a manner that violated the Statistics Act and failed to comply with the Privacy Act. She says these are questions of law and should be reviewed under the correctness standard.
[13]
The Respondent submits that if the Privacy Commissioner’s non-binding report of findings is reviewable, the standard should be reasonableness: Alberta (Information and Privacy Commissioner) v Alberta Teacher’s Association, 2011 SCC 61 [Alberta] at paras 48-55.
[14]
The Court is not here reviewing the Privacy Commissioner’s decision. As the Applicant clarified and confirmed at the hearing before me, she seeks review of the Chief of Statistics’ decision to enter into the agreement with McGill that resulted in the Study. This gives rise to questions of mixed fact and law and is reviewable on a standard of reasonableness.
[15]
When reviewing a decision on the standard of reasonableness, the analysis will be concerned with “the existence of justification, transparency and intelligibility within the decision-making process [and also with] whether the decision falls within a range of possible, acceptable outcomes which are defensible in respect of the facts and law.”
See Dunsmuir, above, at para 47, and Canada (Citizenship and Immigration) v Khosa, 2009 SCC 12 at para 59. Put another way, the Court should intervene only if the Decision was unreasonable in the sense that it falls outside the “range of possible, acceptable outcomes which are defensible in respect of the facts and law.”
VI.
STATUTORY PROVISIONS
[16]
The following provisions from the Federal Courts Act are relevant in this proceeding:
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
[17]
The following provisions from the Federal Court Rules, SOR/98-106 [Rules] are relevant in this proceeding:
|
|
|
|
[18]
The following provisions from the Statistics Act are relevant in this proceeding:
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
[19]
The following provisions from the Privacy Act are relevant in this proceeding:
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
VII.
ARGUMENTS
A.
Applicant
(1)
Violation of the Statistics Act
[20]
The Applicant submits that the Chief of Statistics exercised discretion in a manner that violated s 18.1(2) of the Statistics Act.
[21]
She says the Statistics Act and Privacy Act should be considered together as they contain overlapping responsibilities. Section 18.1 of the Statistics Act provides rules for how census information can be shared, with a clear distinction between census information collected pre- and post-2006. Under s 8(2)(a) of the Privacy Act, government institutions may only disclose personal information without consent for the purposes for which the information was obtained and compiled. According to the Privacy Commissioner’s Report of Findings, usage of the 2006 census for the Study “went beyond the scope of its original purposes.”
Thus, in the absence of consent, the Letter of Agreement is a violation of the Privacy Act.
[22]
The Applicant also submits that Statistics Canada’s method of obtaining consent from the public to use census information in the Study was not compliant with the consent provisions in the Privacy Act and Statistics Act. A single-page summary of the Study on the Statistics Canada website is an inadequate method to inform the public, considering that people first needed to be aware of the Study and it was the only means of notifying the public.
[23]
Furthermore, the information provided in the online notice was insufficient to allow the public to provide consent. First, the online notice failed to provide a complete and balanced discussion to enable an informed judgment on the usage of record linking, as it was not stated that indirect identifiers had a possible risk of revealing the participants’ identities. In the present case, postal code information remained on the final linked product, allowing individual participants and households to be identified. Second, the online notice did not reveal that McGill had been given exceptional permission to link the participants’ directly identifiable information with other personal and sensitive data, thereby providing McGill with unscreened access to personal and sensitive census information.
[24]
Statistics Canada justified the exemption of the Study from ethical commitments and legal requirements on the basis that it was not feasible to obtain consent from all participants and that the purpose of the Study furthered the public good. However, the Applicant submits that Statistics Canada erred when it rationalized the decision not to obtain consent based on the public good. It should have relied on s 8(2) (a) of the Privacy Act to decide whether usage of the 2006 census constituted an inconsistent and improper use. Statistics Canada also ignored the conflict of interest attached to the decision, as both Statistics Canada and McGill benefited from the Letter of Agreement. The use of record linking from the 2006 census to obtain the information used in the Study rather than actively recruiting participants provided substantial benefits to McGill. First, there was a substantial cost saving, as collecting new data would be “practically infeasible.”
Second, the usage of existing datasets reduced the potential bias and improved the Study’s reliability. Similarly, there was also a substantial cost savings associated with Statistics Canada not having to contact all the participants to obtain consent.
[25]
The legislative history of the Statistics Act demonstrates that Parliament had an intention to balance the right to access and the right of individuals to control their personal information, with informed consent viewed as a key principle of privacy protection that should be afforded to all. This is evident in the introduction of s 18.1 of the Statistics Act, which outlines the permitted usage of census information and differentiates between censuses performed between 1910 to 2005 and 2006 onwards. Consequently, Statistics Canada had an obligation to administer its program in a manner that would allow Canadians to change their responses to the consent question at any time during the 92-year period. In exercising its discretion to override the exemption clause on consent in s 18.1(2), Statistics Canada erred in law.
[26]
Under s 17(2)(b) of the Statistics Act, the Chief of Statistics is permitted to release certain types of identifiable personal information if written consent is obtained. However, the Applicant submits that this exemption does not supersede the protections provided in s 18.1(2) without express consent, which was not obtained in this case. The Applicant and other participants in the 2006 census should be treated as persons whose census information is not subject to the s 17(2)(b) exemption since there is nothing in the Statistics Act that allows Statistics Canada to dispense with the mandatory consent requirements of s 18.1(2), even if statutory approval processes are applicable to the Study. Furthermore, the Privacy Act does not allow Statistics Canada to contract out an individual’s consent to disclose information to the Chief of Statistics: see Friends of the Canadian Wheat Board v Canada (Minister of Agriculture), 2011 FC 1432.
[27]
Furthermore, the Applicant cites Bill C-626 as support for Parliament’s intention to address a gap in regulation of the care and control of the census. Although the bill was voted down, it defined the long-form census as one that “conforms, in length and scope and with such modifications as changing circumstances require—including new sources of data or data collection practices—, to the long form census...,”
which suggests that there is a gap in the current legislation. Additionally, since the current Statistics Act does not include such phrases, it suggests that the creation of a longitudinal census for administration purposes was not something that Parliament envisioned at the time the Statistics Act was last amended. Finally, the fact that Bill C-626 was voted down demonstrates that the government was sympathetic on principle to an outcome favourable to the Applicant.
(2)
Standing
[28]
The Applicant submits that she has standing to seek judicial review of the Letter of Agreement, and that the matter of whether or not her personal information from the 2006 census was included in the Study should not be used as a basis for the determination of standing.
[29]
The three-part test for public interest standing is met. This application for judicial review raises serious issues of statutory interpretation under the Statistics Act by focusing on the proper construction of s 18.1(2). There is also a historical record of public interest in all aspects of the subject-matter in these proceedings, as demonstrated by numerous Parliamentary debates and proposed bills.
[30]
The Applicant also has a genuine interest in the resolution of the matter, as the documentary evidence demonstrates that all of the eligible participants from the 2006 census were needed for the Study, which leads to the inference that the Applicant’s personal information was used at some point given that the entire 2006 census was required to create the longitudinal database and related files.
[31]
Finally, the Applicant says that there is no other reasonable and effective manner of bringing the legal issues in dispute to this Court. The remedies found in s 41 of the Privacy Act and s 18.1(3) of the Federal Courts Act are not appropriate to the present case, particularly because the Court has limited jurisdiction under s 41 to conduct a judicial review of the Privacy Commissioner’s findings and recommendations. Although there are other parties who could bring forth a challenge, this is a remote possibility. Neither the Privacy Commissioner nor the Attorney General has exercised their prerogative to challenge the usage of the 2006 census information, and while there are many persons directly affected by the Letter of Agreement, it is a remote possibility that these persons are aware of the Study’s existence given that the online notice was the only method of informing the public. Therefore, the Applicant has a choice of either requesting that her personal information be removed from the longitudinal database, although experience demonstrates that this approach is ineffective, or refusing to complete future censuses, which would constitute an offence under s 31 of the Statistics Act.
[32]
The Applicant also submits that the requested order, a declaration establishing the Applicant’s standing, would not prejudice the Respondent. Furthermore, the Applicant submits that Jane Badets’ affidavit evidence should not be accorded weight as she refuses to provide critical details about the Study.
(3)
Remedies
[33]
The Applicant says that the appropriate remedy is an order in the nature of a declaration that the Letter of Agreement is illegal, void and of no effect, by reason of having been issued without jurisdiction. The Applicant submits that this would send a strong message to the Minister of Industry and give effect to the plain meaning of consent under the governing legislation.
[34]
On the matter of costs, the Applicant says she has raised an important new principle in relation to the interpretation of the “use and disclosure”
code of the Privacy Act, which justifies an award of costs in her favour. Additionally, had the Respondent taken the opportunity to consider the jurisdictional issue and consequences of failing to observe the statutory duty of consent, this issue would not have had to be dealt with through a judicial review application. An award of costs would alleviate the unfairness of a self-represented litigant who has incurred large expenses in an effort to vindicate her rights.
B.
Respondent
[35]
As a preliminary matter, the Respondent submits that the Applicant contravenes Rule 302 of the Rules, which limits an application for judicial review to a single order. While the Notice of Application stated the application was a judicial review of the Privacy Commissioner’s Report of Findings, the Applicant also seeks judicial review of the Letter of Agreement.
(1)
Standing
[36]
The Applicant lacks standing to challenge the Letter of Agreement as it does not directly affect her. The Applicant was not a party to the Letter of Agreement and her personal information did not form part of the record linkages that were used in the Study.
[37]
On the matter of public interest standing, the Court must ensure a balance between access to the courts and preserving judicial resources. It is submitted that a direct challenge from individuals whose census information was used in the Study would best preserve judicial resources and ensure that the Court has the best evidence before it.
(2)
Jurisdiction
[38]
Under s 41 of the Privacy Act, Parliament has expressly limited the Court’s role in relation to complaints made to the Privacy Commissioner. Since this application does not fall under s 41, the Court does not have jurisdiction over this matter. The Privacy Commissioner’s non-binding Report of Findings issued under s 29 of the Privacy Act cannot be judicially reviewed by this Court.
[39]
The Applicant should not be permitted to use the Act to obtain judicial review as neither the Letter of Agreement, the Study, nor the Report of Findings satisfy the requirements of s 18.1(1) of the Federal Courts Act. Judicial review is available when an administrative body’s conduct affects the rights of a party or carries legal consequences, but this is not always triggered, particularly when a non-binding opinion or statement with no binding legal effect is issued.
(3)
Reasonableness
[40]
If the Letter of Agreement and the Study are reviewable, the Respondent submits that Statistics Canada complied with all statutory requirements and the decision was reasonable and correct.
[41]
If the Report of Findings is reviewable, the standard of review is reasonableness: see Alberta, above.
(4)
Report of Findings
[42]
While the Respondent denies that the Privacy Commissioner’s non-binding Report of Findings is subject to judicial review, the Court should find it to be reasonable upon review. The Applicant’s complaint was investigated by reviewing submissions from both the Applicant and Statistics Canada and found to be not well-founded for two reasons. First, there was no evidence that the Applicant’s personal information from the censuses had been used in the Study. Second, even if the Applicant’s personal information from the censuses had been used, such use was permitted by statutory authority.
(5)
Letter of Agreement and Violation of s 18.1(2) of the Statistics Act
[43]
The Study’s use of census information was a consistent use under s 7 of the Privacy Act. The test for consistent use, which was applied to s 8(2)(a) of the Privacy Act in Bernard v Canada (Attorney General), 2014 SCC 13 [Bernard], but is also relevant to s 7, requires only a sufficiently direct connection between the purpose and proposed use, such that an individual would reasonably expect that the information could be used in the manner proposed. The Census Population and National Household Survey Information Bank, in compliance with s 11(1)(a)(iv), stated that the information collected by the census may be used for research purposes as well as be combined with other survey databases for approved statistical purposes. Thus, the Letter of Agreement and the Study were a consistent use of census information, as census data was combined with the national database of pregnancy and birth outcomes to research perinatal health disparities. The Study and usage of census information was also consistent with Statistics Canada’s mandate to collect, analyze, and publish statistical information. While the Study could not have been contemplated at the time the 1996 and 2006 censuses were conducted, it is directly connected to both the censuses’ purpose and Statistics Canada’s mandate, thereby satisfying the test for consistent use.
[44]
The Applicant’s reliance on s 8(2)(a) of the Privacy Act is improper because that provision pertains to the disclosure of personal information; in this case, no personal information was disclosed in the Study. Consequently, Statistics Canada was not required to obtain participants’ consent to use their information in the 1996 or 2006 censuses in the Study. No information was disclosed contrary to s 17(1)(b) of the Statistics Act, since only Statistics Canada employees had access to the raw data and performed the record linkages. After the record linkages were completed, the composite file contained only the data needed for the Study and the linked data remained on Statistics Canada’s premises, accessible only by Statistics Canada employees and deemed employees.
[45]
Similarly, s 18.1 of the Statistics Act, which allows public disclosure of census information 92 years from the date of the census, is inapplicable to this matter. As there was no public disclosure of information apart from non-confidential aggregate data, consent was not required for the census information to be used in the Study. Despite this, Statistics Canada still informed the public about the Study and the usage of record linkages via an online notice on its website.
(6)
Remedies
[46]
The Respondent submits that this application should be dismissed and no remedy be available to the Applicant.
[47]
Alternatively, the remedy sought by the Applicant is improper and also unavailable under ss 18 or 18.1(3) of the Federal Courts Act because there is no basis for a declaration that the Letter of Agreement is illegal, void, and of no effect, by reason of having been issued without jurisdiction. The Study provides a public benefit by allowing researchers to better understand perinatal health disparities and does not have a direct effect on any individual since no personal information is disclosed. As there is no useful purpose to be achieved if such a declaration is made, a declaration should not be granted. Furthermore, the Privacy Commissioner issued a finding that the Letter of Agreement and Study were lawful in response to the Applicant’s complaint and the Applicant should not be able to use a collateral attack to obtain a remedy that she would otherwise not be entitled to.
VIII.
ANALYSIS
A.
Standing
[48]
The Applicant has no direct standing to bring this application. The evidence is clear that the Applicant’s records were not included in the Study. The Applicant did not give birth between 2004 and 2006. The Applicant did give birth between 1994 and 1996 in Ontario. The evidence establishes that no births which occurred in Ontario between 1994 and 1996 were included in the linkage conducted by Statistics Canada for use in the Study because of concerns about data quality. The Applicant has not refuted this evidence. This means that no records pertaining to the Applicant were included in the record linkage performed for the Study or in the Study itself.
[49]
The Applicant continued her assertion at the hearing before me that the evidence supported that her records had been used. However, when asked by the Court to identify any such evidence, she could not do so. On the other hand, the Respondent’s evidence on this point contained in the affidavit of Ms. Badets is clear and well-supported. The Applicant made no reference to this evidence and simply expects the Court to overlook clear, objective evidence and accept her unsupported assertions to the contrary. The Court cannot do this. The evidence is that the Applicant’s records were not used and she is not someone who is directly affected by the matter before me, or even someone who could be directly affected at some time in the future. Ms. Badets opined as follows:
20. The Applicant’s records did not form part of the Census-linked birth records for the Study. The Study linked births which occurred between 1994-1996 and 2004-2006 to Census data. Only the 1994-1996 period is relevant because, as the Applicant states in her affidavit, she did not give birth between 2004-2006. As the Applicant also states in her affidavit, during the 1994-1996 period she gave birth in Ontario. No births which occurred in Ontario between 1994-1996 were included in the linkage conducted by Statistics Canada for use in the Study due to concerns with data quality. Therefore, no records pertaining to the Applicant were included in the record linkage performed for the Study or in the Study itself.
21. The following demonstrates that no births which occurred in Ontario between 1994-1996 were included in the data linkage completed for the Study:
a) The Request for Record Linkage, a copy of which is attached as Exhibit “E”, shows that the decision not to include the records for that particular group of individuals as made at the planning stage of the Study before any record linkage occurred. This request was subsequently approved, as evidence by Exhibit “F”;
b) Kramer et al. state in their 2010 research application to the Canadian Institutes of Health Research (“CIHR”), a copy of which is attached as Exhibit “J”, under heading B.3.1 that:
We will exclude births from Ontario for data linkage to the 1996 census, owing to errors in birth weight and gestational age (key information for our outcomes) during the early and mid-1990s. However these errors have been corrected in recent data.
c) I have consulted with my colleague, Richard Trudeau, who was the Chief, Health Record Linkage Section, Health Statistics Division at the time the linkage occurred. In this role, Mr. Trudeau was responsible for the Statistics Canada team that conducted the birth-to-infant deaths linkage as well as the birth-to-birth linkage that were prerequisites to the births-to-census linkage. Mr. Trudeau has confirmed to me that the Applicant’s records did not form part of the data linkage performed for the Study. An excerpt of the computer code that created the files that were used in the births-to-census linkage is attached as Exhibit “L”. The highlighted portion of the computer code indicates the selection criteria <CBDB_BIRTHPLAC IN “910”, “911”, “912”, “913”, “924”, “946”, “947”, “948”, “959”, “960”, “961”)>, which means that only those records where the value of the CBDB_BIRTHPLAC variable was equal to one of the listed values were outputted. CBDB_BIRTHPLAC is the variable in the dataset that contains the coded place of birth, namely the province or territory where the birth occurred. The list does not include the code for Ontario, which is “935”. Therefore, this computer code demonstrates that when the linkage occurred, records of births occurring in Ontario were excluded.
22. Prior to the record linkages, Statistics Canada agreed to remove the Applicant’s data from the Study as a gesture of goodwill. However, once it became apparent that the Applicant’s information would not be included in the linkage due to data quality issues, there was no information to remove.
[emphasis in original]
[50]
The Applicant comes before the Court seeking judicial review under s 18.1(1) of the Federal Courts Act. However, she is not “anyone directly affected by the matter in respect of which relief is sought.”
The Applicant has an interest in privacy issues, and she may have an interest in privacy issues raised by the Study, but she is not “directly affected”
by the Study because her records were not used.
[51]
The Applicant has attempted to remedy this defect by laying claim to some kind of public interest standing. Public interest standing is not granted to everyone. As the Supreme Court of Canada made clear in Canada (Attorney General) v Downtown Eastside Sex Workers United Against Violence Society, 2012 SCC 45 [Downtown Eastside Sex Workers], there are various factors that must be taken into account:
[23] This Court has taken a purposive approach to the development of the law of standing in public law cases. In determining whether to grant standing, courts should exercise their discretion and balance the underlying rationale for restricting standing with the important role of the courts in assessing the legality of government action. At the root of the law of standing is the need to strike a balance “between ensuring access to the courts and preserving judicial resources”: Canadian Council of Churches, at p. 252.
...
[25] The most comprehensive discussion of the reasons underlying limitations on standing may be found in Finlay, at pp. 631-34. The following traditional concerns, which are seen as justifying limitations on standing, were identified: properly allocating scarce judicial resources and screening out the mere busybody; ensuring that courts have the benefit of contending points of view of those most directly affected by the determination of the issues; and preserving the proper role of courts and their constitutional relationship to the other branches of government. A brief word about each of these traditional concerns is in order.
[52]
The Applicant is not directly affected by this case and the evidence before me does not establish that others, whose records were part of the linkage for the Study and the Study itself, need have any legal concerns about the use of their records. The Applicant’s situation is not representative at all of persons whose information was used. The issues she raises and argues can only really be decided on a set of facts that includes an applicant or applicants who were directly affected, or who may be affected by the Study when it is eventually released. As yet, there has been no disclosure of personal information so that, at this point, it is not clear what will be, or could be disclosed that persons whose records were part of the linkage might wish to complain about. For all the Court knows, such persons may well approve of the Study and have no objection to the use of their records, on legal or on any other grounds. There is no indication that the Applicant’s position is anything more than her own personal position, born of her academic interests and her social activism. The Applicant has not demonstrated that her concerns are shared by anyone who was directly affected or that she has any more interest in the Study than any other member of the general public whose records were not used. It would be dangerous for the Court to attempt to pronounce upon complex issues without an appropriate record to do so. If the Applicant believes that those whose records were used should have concerns, then she can alert them to the issues and leave them to bring matters before the Court if they agree with her and believe that the situation requires judicial intervention. In terms of the jurisprudence, the Applicant is no more than a mere “busybody” who may once have believed that her records were used but has known for some time now that they were not.
[53]
The recent case of Lessard-Gauvin v Canada (Attorney General), 2016 FC 418 [Lessard] provides a good summary of the principles regarding public interest standing at paragraphs 14 to 15:
[14] In exercising the discretion to grant public interest standing, the Court must consider three factors: (i) whether the case before the Court raises a serious justiciable issue; (ii) whether the party seeking this standing has a real stake or a genuine interest in the case; and, (iii) whether the proposed case is a reasonable and effective way to bring the issue before the courts Canada (Attorney General) v. Downtown Eastside Sex Workers United Against Violence Society, 2012 SCC 45, at paragraph 20, [2012] 2 S.C.R. 524 [Downtown Eastside Sex Workers]).
[15] Even if a purposive approach is appropriate in assessing these factors, the Supreme Court of Canada, in Downtown Eastside Sex Workers, above, noted that the courts have long recognized that limitations on standing are necessary, and consequently, “not everyone who may want to litigate an issue, regardless of whether it affects them or not, should be entitled to do so” (Downtown Eastside Sex Workers, at paragraph 22). The considerations that favour such an approach are related to (i) properly allocating scarce judicial resources and screening out the mere busybody; (ii) ensuring that courts have the benefit of contending points of view of those most directly affected by the determination of the issues; and, (iii) preserving the proper role of courts and their constitutional relationship to the other branches of government (Downtown Eastside Sex Workers, at paragraph 25). Ultimately, the Court must seek to strike a balance between ensuring access to the courts and preserving judicial resources (Downtown Eastside Sex Workers, at paragraph 23).
[54]
The applicant in Lessard sought a declaration that Canadian human rights legislation violated the Canadian Charter of Rights and Freedoms as part of his legal studies at university. In finding that the applicant did not have public interest standing, the Court also commented:
Evidently, the applicant has an academic interest in being granted public interest standing, but that is not a valid justification.
[...]
Moreover, his interest in this part of his recourse is strictly theoretical and seems first and foremost the focus of an academic project. Furthermore, this part of the recourse is devoid of any basis in fact, since the legislation at issue has never been tested by the applicant.
[...]
On the other hand, there are other realistic ways to debate whether provincial and territorial human rights legislation complies with the Charter, in particular through remedies instituted by parties who, unlike the applicant, may have legal standing, in order to increase, in a more favourable context, the likelihood of a more effective and efficient use of judicial resources...the failure of the remedy instituted by the applicant, in the diffuse and disembodied context in which he presents himself, might impede future challenges by persons who have specific complaints based on facts. In other words, in a case like this one, our limited judicial resources should be allocated to support these persons.
[55]
Similarly, the Applicant in the present case has only an academic and social activist interest in being granted public interest standing since the Privacy Commissioner found that her information had not been misused and she has been unable to demonstrate otherwise on a factual basis. Additionally, there are other parties, namely those whose data were used in the Study, who may wish to bring a future challenge with a specific complaint based on facts.
[56]
In terms of the factors that have to be weighed in accordance with the Downtown Eastside Sex Workers case, this application does not raise a serious justiciable issue and the Applicant has no real stake in the outcome because her information was not used in the linkage for the Study. Her interest is no stronger than the interest of any other member of the public whose information has not been used. She asserts that others would be unlikely to discover what had taken place, or bother to take legal action if they did. However, the Applicant can make her concerns known to others whose records were used in the linkage so that they can decide whether to bring an application. The real danger here is that the Court is being used by a social activist who, no doubt, believes sincerely that she is doing good but who has no personal stake in this application and who may well create problems for other people who do have stake through her pre-emptive allegations. It is far safer to wait and see. The Study has not yet been released.
B.
No Disclosure of Personal Information
[57]
Alternatively, even if the Applicant had standing to bring this application, the record before me is clear that no personal information has been disclosed in this case, not only with respect to the Applicant whose records were not used, but also with respect to those persons whose records were used.
[58]
The evidence on this issue is found in the affidavit of Jane Badets who was, from October 12, 2010 until November 23, 2014 the Director General, Census Subject Matter, Social and Demographic Statistics Branch, which evidence has not been materially challenged by the Applicant. Ms. Badets was responsible for all Statistics Canada Research Data Centres [RDCs], including the RDC at McGill University where the data at issue in this application was accessed. Ms. Badets gives the following account of what occurred in this case:
9. The Study required record linkage of a sample of births records to Census data because previously existing statistical information failed to account for the interrelationship of successive births to the same mother and contained minimal, if any, information on parental socioeconomic, ethno-cultural and environmental health risks and protective factors known to affect perinatal health outcomes. By creating a record linkage for a sample of births to Census data, Statistics Canada was able to provide Kramer et al. with much of the missing information, for a reasonably large sample of births in two recent periods, while avoiding the need for additional data collection. Attached hereto as Exhibit “E” is a copy of the Request for Record Linkage submitted in support of the Study.
10. The Request for Record Linkage was reviewed and approved on February 22, 2012 by Statistics Canada’s Executive Management Board in accordance with the Directive on Record Linkage, a copy of which is found at Exhibit “C”. The Executive Management Board is the most senior governing body within Statistics Canada and is comprised of the Chief Statistician and the Assistant Chief Statisticians. Before approval was granted, the Executive Management Board would have reviewed the request to ensure that appropriate controls related to the protection of identifiable information were in place, that the study was for the public good and that there were no feasible alternative means of obtaining the same information. Attached hereto as Exhibit “F” is a copy of the approved Request for Record Linkage.
11. The Census-linked birth records for the Study consisted of approximately a 20% sample of births in the two years prior to the 1996’Census that were linked to the 1996 Census and in the two years prior to the 2006 Census that were linked to the 2006 Census, as well as records for subsequent sibling births up to two years after each Census and records for previous sibling births to the same mothers. The birth records were previously linked to corresponding infant deaths occurring up to 12 months following birth. The birth outcome information included in the linked files originated from data collected by Canadian provincial and territorial registries of vital statistics, which were validated by Statistics Canada.
12. The linked data files contained only those data items required to conduct the Study. Names were used only for linkage purposes and then removed from the linked files before access was given to Kramer et al. Attached hereto as Exhibit “G” is a copy of the Record Linkage Summary for the Study.
13. No record linkages were ever conducted by Kramer et al. with respect to Statistics Canada data used for the Study. Rather, all linkages were performed by Statistics Canada employees. Statistics Canada began the record linkage of these files in the fall of 2012. Once this process was complete, access to the linked micro-data analysis files was provided to Kramer et al. The linked data files were stored at Statistics Canada’s Head office and copies were available at McGill University & the University of Toronto RDCs. The linked data never left Statistics Canada premises.
14. RDCs are secure Statistics Canada environments located within university settings. RDCs are staffed by Statistics Canada employees at all times and operate under the provisions of the Statistics Act. Access to the linked data files was restricted solely to employees and deemed employees of Statistics Canada.
15. Certain members of Kramer et al. were deemed employees of Statistics Canada. Deemed employees are persons, contractors or federal public servants whose services are used by the Minister of Industry to carry out any function or to perform work pursuant to sections 5 and 10 of the Statistics Act. Deemed employees must take an oath under the Statistics Act and are subject to the same penalties under that Act as are all Statistics Canada employees. The Directive on the Use of Deemed Employees governs the use of Deemed employees by Statistics Canada. A copy of this Directive is attached as Exhibit “H”.
16. At all times, the linked data files remained on Statistics Canada premises. Access to the linked data was restricted only to Statistics Canada staff and deemed employees whose work activities required access.
17. The Study was not the first time that Statistics Canada provided access to Census information to deemed employees. Attached as Exhibit “I” is a spreadsheet evidencing other instances in which access to Census data was provided to deemed employees.
18. Based on my knowledge of this file and of Statistics Canada’s operations and my discussions with colleagues, I disagree with the Applicant’s statement at paragraph 18 of the her affidavit that Kramer et al. selectively recruited personal data from the approximately nine million records from the “Canada Live Birth, Infant Death and Stillbirth Database”. Rather, Kramer et al. conducted analysis solely on the database which had been linked by Statistics Canada employees.
[59]
The Applicant counters this evidence with an assertion that “indirect identifiers”
give rise to a “possible risk”
that the identities of individual participants will be revealed, but she has not demonstrated with convincing evidence how this could occur.
[60]
In her arguments before me, the Applicant relies heavily upon ss 17(1)(b), 17(2) and 18.1(1) of the Statistics Act. None of these provisions come into play on the facts of this case, either for the Applicant or anyone whose records were actually used.
[61]
This is because no information obtained under the Statistics Act was disclosed contrary to s 17(1)(b), so that s 17(2) does not come into play. As the evidence of Ms. Badets makes clear, only employees and deemed employees of Statistics Canada with a need to know had access to the raw data, and they were the ones who performed the record linkages. The linked data was then retained within Statistics Canada premises and, once again, could only be accessed by Statistics Canada employees or deemed employees, all of whom have been sworn under s 6 of the Statistics Act. All that Kramer et al have seen is non-confidential aggregate data. The Applicant has not shown how the aggregate data provided to Kramer et al can be linked to any individual. She has argued before me that Kramer et al will receive postal codes that will be used, for example, to examine the role that environmental factors could play in perinatal outcomes. She says that, once a postal code is known, it is then possible to identify individuals and families by accessing other connective sites and information. Once again, however, this remains a mere assertion by the Applicant. She has not led the Court to any evidence which would show how this could be done. In addition, no personal information, including postal codes, has yet been revealed to anyone outside of Statistics Canada and those who fall within the s 6 undertaking:
|
|
|
|
|
|
|
|
|
|
|
|
[62]
The Applicant also points to the Kramer et al research application that was made to Statistics Canada which she says shows that the eventual Study will be used to treat individuals. The Study would, of course, be of no use if it could not enable individuals to receive better care. This is why Canada spends money on such studies. However, the fact that the knowledge yielded by the Study will be used to treat individuals does not mean that personal information has been used in breach of the Privacy Act or the Statistics Act.
[63]
According to the research application, the researchers plan to achieve the objectives of the Study by creating an internally-linked, longitudinally-oriented file linking successive births to the same mother that will identify sets of multiple births. This file will permit the systematic examination of the extent to which a woman’s successive pregnancy outcomes tend to repeat and permit the assessment of the risk of adverse birth outcomes conditioned on her previous birth outcomes. The linkage provides additional information that can be examined, such as ambient air pollution, income and housing characteristics of the family, etc. The trends can then be used to inform social policy; specifically, the trends will be used to refine Canadian standards for exposures to various air pollutants and contribute to the formulation of evidence-based clinical guidelines and public health initiatives aimed at improving perinatal health.
[64]
In my view, the research application seems to indicate that the data sets will be used to identify trends which will form the basis of advice/guidelines delivered to the general public via bulletins issued by Health Canada or other government policy. For example, if the Study demonstrated that an area with heavy air pollution also contains a disproportionately high number of stillbirths, municipal governments may choose to prohibit residential zoning in highly polluted areas or doctors could advise their patients in highly polluted areas that they should reduce the amount of time they spend outside or relocate for the duration of their pregnancy, which would achieve the goal of improving public health. In this manner, personal information about an individual is not being disclosed; rather, information about a group of 9.5 million people (or 1.9 million since only 20% of eligible participants were used in the Study) is being disclosed.
[65]
The latter disclosure does not allow identification of individuals; i.e. a statement such as, “An analysis of mothers with low-income demonstrated a trend of below-average birth weights”
does not identify an individual in the sense that one could point to a mother with low-income and determine her child had a below-average birth weight.
[66]
The Applicant has also not shown how s 18.1 of the Statistics Act has any application to the present situation. That section deals with public disclosure of information obtained by a census 92 years after it has been taken. This section does not come into play here because the prohibition in s 17(1)(b) only covers “any information obtained under this Act in such a manner that it is possible from the disclosure to relate the particulars obtained from any individual return to any identifiable individual person, business or organization.”
On the evidence before me, it is not possible to relate the aggregate data disclosed to Kramer et al to any identifiable individual person, business or organisation. This means that Statistics Canada was not required to obtain consent for disclosure of the aggregate data.
[67]
The Applicant’s reliance upon s 8(2)(a) of the Privacy Act is also misplaced because no personal information has been disclosed in the Study. The only way that the Privacy Act becomes relevant is under s 7(a):
|
|
|
|
|
|
[68]
There is no doubt that census information is personal information, so the issue in this case is whether it was used “for a use consistent”
with the “purpose for which it was obtained or complied....”
[69]
The Supreme Court of Canada set out the “consistent use”
test in Bernard, above:
[31] A use need not be identical to the purpose for which information was obtained in order to fall under s. 8(2) (a) of the Privacy Act; it must only be consistent with that purpose. As the Federal Court of Appeal held, there need only be a sufficiently direct connection between the purpose and the proposed use, such that an employee would reasonably expect that the information could be used in the manner proposed.
(emphasis in original)
[70]
It is clear that Statistics Canada could not have contemplated the Study at the time of either the 1996 census or the 2006 census. Hence, the information collected by those censuses was not obtained specifically for the Study. However, the purpose of the Study is to compile and analyse statistics related to the health and welfare of Canadians, so that it complies with the purpose of the censuses and with Statistics Canada’s mandate.
[71]
On the 2006 census, the purpose reads as follows:
Community groups, businesses and governments develop programs and services such as education, health and other social and economic programs based on census data. The information is widely used in our schools to teach children about our country. Information may also be used by Statistics Canada for selecting samples or following up respondents for some of our surveys.”
[emphasis added]
Since the Study’s objective is to analyze and determine trends regarding perinatal health for usage in the development of social and health policies, the Study appears to comply with the census purpose of developing a health program or service.
[72]
Additionally, the Statistics Canada website describes the 2006 census as follows:
...The census also provides information about the characteristics of the population and its housing with small geographic areas and for small population groups to support planning, administration, policy development and evaluation activities of governments at all levels, as well as data users in the private sector.
Again, the Study’s objective appears to comply with the census description of supporting policy development of governments and data users in the private sector.
[73]
The purpose of the Study is to assess infant mortality and newborn health by examining perinatal outcomes in Canada according to risk factors related to socioeconomic position, ethno-cultural background and environmental exposure. This seems to me to be a consistent use of census information. The Study combines census data with record linkages (i.e. the national databases of pregnancy and birth outcomes) to yield information and allow research to be done on perinatal health disparities. It is part of Statistics Canada’s mandate to collect, analyse and publish statistical information, which includes statistics related to the health of Canadians, as set out in ss 3(a) and 22(c) of the Statistics Act.
[74]
In my view, then, the Study satisfies the Bernard, above, test for “consistent use.”
This means it is compliant with s 7 of the Privacy Act.
[75]
In my view, the Applicant has not shown an inconsistent use of the census information or a disclosure of personal information. That being the case, the Statistics Act does not require consent. No information has been disclosed under s 17(1)(b) of the Statistics Act by those persons sworn by s 6 of that Act.
[76]
The evidence before me shows that only employees and deemed employees of Statistics Canada with a need to know have had access to the data from record linkages and they have been sworn under s 6 of the Statistics Act. This linked data remains on Statistics Canada premises. The Applicant has not demonstrated that anything other than non-confidential aggregate data will be publicly disclosed when the Study is complete.
[77]
In order to overcome these problems in her application, the Applicant asserted that, in the end, this matter was not about disclosure, but was about what she called a “change to the democratic situation”
as embodied in s 18.1(2) of the Statistics Act and the debates surrounding Bill S-18 in Parliament. The following excerpts appear to be relevant to this issue:
Mr. Lloyd St. Amand (Brant, Lib.):
No doubt the House would agree that access to these records can provide a rich and valuable means by which both family and historical research is conducted...The other key aspect of this bill...is that it would enable Canadians a clear and unequivocal say in how they would like their personal information on census records to be used in the future. Canadians would be given the option of giving permission for the release of their personal information on the census questionnaire.
[...]
As it currently stands, at issue is the legal ambiguity as to the authority of the chief statistician of Canada to release records from the 1911 census population. That the records from this census have not yet been released has caused much outcry and consternation among the genealogical community...The bill then addresses the issue of historical access as it relates to censuses already conducted, it also sets the framework for the issue of access to future census records.
Beginning with the 2006 census and in any to follow, Statistics Canada would ask on the questionnaire for the consent of Canadians to release their census information, once again 92 years after each census. An individual’s census records would be released only when consent was given. If consent were not given, the census records of that individual would never be made publicly available.
[...]
Informed consent about the use of personal information is a key principle of privacy protection. Therefore, it follows that Canadians should have the right to decide for themselves if they want their personal census records to be made publicly available in the future. Bill S-18 would give Canadians the option for the first time to provide consent for the release of their own information 92 years after the census had been conducted. Plans call for such a question to be asked on the questionnaire of the next census of population, which will be conducted in May 2006.
The multitude of persons who serve as privacy protection advocates and supporters either in a formal or informal capacity would be the first to agree that informed consent about the use of personal information is a key principle of privacy protection. To many it is a right afforded to all. Bill S-18 will provide the legal basis under which this consent could be offered and upheld.
A third and equally important point of Bill S-18 is that it calls for parliamentary review in 2014. A committee of this House, the other place, or both houses of Parliament would review the administration and operation of the subsections related to the informed consent provision. By that time, the year 2014, there will have been two censuses in which Canadians will have been asked for their permission to release the personal information. How they have responded will provide an indication of how Canadians view this issue and how they wish their government to respect their wishes.
[...]
Now here we are with Bill S-18, which again attempts to resolve the 94 year old question on what to do with the release of census information from 1911 to 2001 and offers a solution to the privacy issues that will face future generations as they fill out their census forms.
[emphasis added]
[78]
After reading the debate in its entirety, it is clear that Parliament intended to address whether census records can be released in their entirety without consent after 92 years. The importance of consent is evident in such a situation because the record is personally identifiable. However, in the current situation, where parts of non-personally identifiable records are released, the importance of consent is less clear. The debates emphasize the importance of allowing Canadians to consent to release of their personal information in a specific context; namely, where the release of the information allows complete identification of an individual down to their name. That is not the case in the current situation and it does not follow from the debates that Parliament intended consent to be required in the release of partial, non-identifiable records.
[79]
What the Applicant appears to mean is that the decision of the Chief of Statistics to enter in the agreement with McGill and to authorize the Study and its uses without going back and obtaining the consent of individuals changed the fundamental balance between privacy and the public good that underlies the Statistics Act. In effect, she is saying that Statistics Canada changed the fundamental balance struck by the Act by using personal information without consent, and this undermines the democratic process because it was not what Parliament intended as evidenced by the debates on Bill S-18.
[80]
But the answer to this is that future studies in the health area cannot be identified at the time of a census, and that the public good does not require, and Parliament did not intend, that consent of the whole census population is required each time a new study is proposed. Instead, Parliament has built safeguards into the legislation to protect private information and for the use of census information for the public good. The Supreme Court of Canada considered the issue of consent in Bernard, and came up with the concept of “consistent use”
to overcome the obvious difficulties that will arise over time in deciding whether those who completed a census would provide consent for any particular use that arises in the future.
[81]
The Applicant cannot say in the present case that there has been, or will be, any unlawful disclosure of her personal information. Her personal information was not used. Nor has she established that the personal information of others has been, or will be disclosed, for the reasons I have given above. So she is thrown back upon the allegation that the Chief of Statistics failed to obtain consent in a context where consent should have been obtained. This assertion must fail on these facts. The Chief of Statistics did not change the law or upset the balance that Parliament and the Supreme Court of Canada have said is required in these situations. The Chief of Statistics acted in accordance with the mandate of Statistics Canada and the law which says that further consent is not required for a consistent use. On the facts, this was a consistent use in which no personal information has been, or will be disclosed.
[82]
In the end, I have simply reached the same conclusions as the Privacy Commissioner. The Privacy Commissioner told the Applicant clearly that his investigation revealed that her personal information was not used for the Study, and that, although the Study went beyond the research purposes referred to in the census, it did so only in a way that is authorized by the Statistics Act.
[83]
Before me, the Applicant has asked the Court to review these same issues and reach a different conclusion. She has continued to assert that her personal information was used when, on the evidence, it clearly was not used, and she has asserted some kind of public interest standing that she doesn’t have because she has failed to show the Court that her personal interests and views are shared by anyone whose records were linked for the purposes of the Study. In her written submissions, she does not engage with the applicable statutory provisions and, at the hearing before me, she relied upon what she called a “historical contextual approach”
which she feels is embodied in s 18.1(2) of the Statistics Act.
|
|
[84]
This section only deals with what is to happen to information 92 years after the census is taken. At that point, it says that ss 17 and 18 won’t apply, but only if the person consents.
[85]
In the present situation, no personal information was disclosed beyond the bounds of those sworn under s 6 of the Statistics Act, and the evidence does not establish that such information will ever be disclosed in any form that will fall into “information about an identifiable individual”
as defined by s 3 of the Privacy Act. Information provided in the censuses will certainly be used, but its use is a “consistent use”
as defined by the governing jurisprudence. The Applicant’s position appears to be that consent is always required unless it is specifically mentioned and consented to in the census. This is not the law and it would, practically speaking, be unworkable if it was. This is why we have the concept of “consistent use.”
[86]
The real problem with this application is that it is premature. The Study has not yet been released or used. The Applicant speculates that personal information will be used and disclosed, but has produced no convincing evidence to support that position. Whatever I have said in this application, which is based solely upon the record before me, should not prevent anyone whose personal information is inappropriately used or disclosed from bringing the matter before the Court in the future.
[87]
The application must be dismissed.
C.
Costs
[88]
Brief submissions on costs were made by both sides at the hearing, but I feel that, in fairness, further cost submissions should be made based upon my findings. Consequently, the parties should make fairly detailed cost submissions in writing and I will deal with costs in a supplemental order. The Respondent’s submissions, including a draft bill of costs, should be served and filed within 10 days of the date of this judgment and the Applicant should respond within 10 days of the receipt of the Respondent’s submissions on costs.
JUDGMENT
THIS COURT’S JUDGMENT is that
The application is dismissed.
The parties will make costs submissions to the Court in accordance with this judgment and reasons in writing and the Court will deal with them in a supplemental order.
“James Russell”
Judge
FEDERAL COURT
SOLICITORS OF RECORD
DOCKET:
|
T-2587-14
|
STYLE OF CAUSE:
|
KELLY A. O’GRADY v ATTORNEY GENERAL OF CANADA
|
PLACE OF HEARING:
|
Ottawa, Ontario
|
DATE OF HEARING:
|
November 15, 2016
|
JUDGMENT AND REASONS:
|
RUSSELL J.
|
DATED:
|
February 10, 2017
|
APPEARANCES:
Kelly A. O’Grady
|
ON HER OWN BEHALF
|
Abigail Martinez
|
For The Respondent
|
SOLICITORS OF RECORD:
William F. Pentney
Deputy Attorney General of Canada
Ottawa, Ontario
|
For The Respondent
|