Information and Privacy Commissioner, Ontario, Canada

Commissaire à l’information et à la protection de la vie privée, Ontario, Canada

ORDER PO-3233

Appeal PA11-507-2

Carleton University

July 25, 2013

Summary: A media requester submitted a request to Carleton University for access to specific student grade information from 1999 to 2011. Carleton denied access to the information, claiming that its disclosure would result in an unjustified invasion of students’ personal privacy under section 21(1) and prejudice the university’s economic interests or competitive position under section 18(1)(c). In his appeal of Carleton’s decision to this office, the requester challenged Carleton’s exemption claims, arguing that the grade data was anonymized. In this order, the adjudicator finds that the responsive grade information does not fit within the definition of personal information in section 2(1) of the Act and that it cannot, therefore, qualify for exemption under section 21(1). She also finds that section 18(1)(c) does not apply. As no exemptions apply to the grade data, the adjudicator orders that it be disclosed.

Statutes Considered: Freedom of Information and Protection of Privacy Act , R.S.O. 1990, c. F.31 , as amended, sections 2(1) (definition of personal information) and 18(1)(c).

Orders and Investigation Reports Considered: Orders P-644, P-1137, P-1389, PO-1805, PO-2204, PO-2404, PO-2713, PO-2726 and MO-1708.

Cases Considered: Ontario (Attorney General) v. Pascoe, 2001 CanLII 32755 (ON SCDC), aff’d 2002 CanLII 30891 (ON CA).

OVERVIEW:

[1] This order addresses the issues raised by a request submitted under the Freedom of Information and Protection of Privacy Act (the Act) to Carleton University (Carleton, or the university) for student grade information spanning a 12-year time period between 1999 and 2011.

[2] Upon receipt of the request, Carleton issued an interim access decision and a fee estimate to the requester. The requester appealed the fee estimate to this office, which opened Appeal PA11-507 to address the fee issues. During mediation of that appeal, the appellant removed grade information relating to graduate students from the scope of the appeal, and paid the requested deposit. In response, the university issued a final decision based on the narrowed request and Appeal PA11-507 was closed.

[3] In the decision letter denying access to the grade data, in its entirety, Carleton claimed that sections 18(1)(c) (economic and other interests) and 21(1) (personal privacy) apply. The appellant appealed the university’s decision. Appeal PA11-507-2 was opened and a mediator was appointed to explore resolution.

[4] During mediation, Carleton’s Computing Services Department advised that the data could be re-formatted in a machine-readable electronic format, consisting of five columns: year, course name, course code, final grade, and the number of students who obtained each grade. The appellant accepted the proposed format as responsive to his request, but maintained that the claimed exemptions do not apply or, alternatively, that there is a public interest in the disclosure of the records, as contemplated by section 23 of the Act .

[5] The mediator sought clarification of grade abbreviations appearing in the records, and obtained a written description of these notations from the university. As the abbreviation summary was not provided to the appellant, he also appealed Carleton’s apparent denial of access to it.

[6] No further mediation of the issues remaining in dispute was possible. The appeal was transferred to the adjudication stage of the appeals process, in which an adjudicator conducts an inquiry under the Act . The adjudicator formerly assigned to this appeal commenced her inquiry by seeking representations from Carleton, initially. The university provided representations on the issues set out in the Notice of Inquiry. Shortly after these representations were received, the appeal was re-assigned to me.

[7] Upon review of the university’s representations, I noted that the university had not responded to the former adjudicator’s request (in the Notice of Inquiry) to specifically address the possible application of exemptions to the grade abbreviation record. When staff from this office followed up with the university at my request, Carleton issued a revised decision letter respecting the grade abbreviation record, indicating that it is publicly available on the university website. In the circumstances, I am satisfied that the availability of this information on Carleton’s website effectively removes the issue of access to the abbreviation summary from the scope of this appeal.¹

[8] Next, I sought representations from the appellant by sending a modified Notice of Inquiry and relevant portions of the university’s representations. I received representations from the appellant. Several days later, the appellant provided brief supplementary representations addressing information that had recently come to his attention.

[9] As I concluded that the university ought to be provided with an opportunity to reply to the appellant’s representations, which included submissions on the possible application of the public interest override in section 23 of the Act , I sent a modified Reply Notice of Inquiry along with a complete copy of the appellant’s written representations.

[10] The university declined to submit reply representations, and I moved the appeal to the order stage.

[11] In this order, I find that disclosure of the grade data could not reasonably be expected to identify individual students. Therefore, since the grade data is not information about an identifiable individual, it is not personal information as that term is defined in section 2(1) of the Act and it is not eligible for exemption under section 21(1). I also find that section 18(1)(c) does not apply to the grade data. There being no other exemption claims with respect to the data, I order that it be disclosed to the appellant.

RECORDS:

[12] At issue is a CD-ROM containing information extracted from a database that consists of five columns: academic year, undergraduate course code, course number, letter grades and number of students who obtained each letter grade for the years 1999 to 2011.

ISSUES:

1. A. Do the records contain personal information according to the Act ?
2. B. Does the mandatory personal privacy exemption in section 21(1) apply?
3. C. Does the discretionary exemption for economic or other interests at section 18(1)(c) apply?
4. D. Did the university properly exercise its discretion under section 18(1)(c)?
5. E. Is there a public interest in disclosure that outweighs the purpose of the section 21 and/or 18 exemptions?

DISCUSSION:

A. Do the records contain personal information according to the Act?

[13] Carleton relies on the mandatory personal privacy exemption in section 21(1) to deny access to the student grade data requested by the appellant. Section 21(1), if it applies, would prohibit Carleton from releasing this information.² However, the appellant challenges Carleton’s denial of access, arguing that the data does not identify individual students and that its disclosure could not, therefore, result in an unjustified invasion of personal privacy under section 21(1).

[14] Since section 21 can only apply to personal information, as that term is defined in section 2(1) of the Act , I will start by determining whether the undergraduate student grade data at issue in this appeal qualifies as personal information.

[15] According to the definition in section 2(1), personal information means recorded information about an identifiable individual, including

1. (a) information relating to the race, national or ethnic origin, colour, religion, age, sex, sexual orientation or marital or family status of the individual,
2. (b) information relating to the education or the medical, psychiatric, psychological, criminal or employment history of the individual or information relating to financial transactions in which the individual has been involved,
3. (c) any identifying number, symbol or other particular assigned to the individual,
4. (d) the address, telephone number, fingerprints or blood type of the individual,
5. (e) the personal opinions or views of the individual except if they relate to another individual,
6. (f) correspondence sent to an institution by the individual that is implicitly or explicitly of a private or confidential nature, and replies to that correspondence that would reveal the contents of the original correspondence,
7. (g) the views or opinions of another individual about the individual, and
8. (h) the individual’s name where it appears with other personal information relating to the individual or where the disclosure of the name would reveal other personal information about the individual;

[16] The list of examples under section 2(1) is not exhaustive. Information that does not fall under paragraphs (a) to (h) may still qualify as personal information.³ Past orders have established that to qualify as personal information, it must be reasonable to expect that an individual may be identified if the information is disclosed.⁴

Representations

[17] Carleton submits that this request must be viewed within the broader context of the Act and that an interpretation showing due regard for [the Act ’s] purposes and objectives is required.⁵ Carleton acknowledges that disclosure is the rule under the Act , but asserts that the goals of transparency and accountability regarding the use of public funds must be balanced with the university’s valid objectives, including the ability to carry out its mandate of promoting excellence in research and education.

[18] With respect to whether the information at issue qualifies as personal information that must be withheld, Carleton highlights paragraph (b) of the definition of the term and submits that:

The record in dispute contains the personal information of Carleton University students from 1999 to 2011. Under section 2(1), information related to the education of an individual is considered to be personal information. The information is attributable to an identifiable individual, as the grades are assigned uniquely to an individual student based on their performance on various evaluation tools during the semester. It is indisputable that the grades received by a student at the University constitute information related to their education.

If the requested information was disclosed, individuals could be identified by using course calendars to pinpoint the necessary courses for graduation (which are public information) and dates of graduation (available in a number of public forums, such as LinkedIn, Facebook, Convocation booklets, announcements from Carleton regarding awards and scholarships or even a job applicant’s curriculum vitae). By combining these pieces of information, a viewer of the record could establish which mandatory courses a student had taken. Depending on the distribution of grades in the course, it may be readily apparent what grade the student received.

In particular, some upper year and specialized courses have very small enrollment. It would be possible for someone to identify a student in these courses, particularly high and low performing students. Such knowledge would affect the manner in which a student would be viewed by their classmates, [their] colleagues and other institutions. …

[19] Carleton refers to Orders PO-2713 and MO-1708, where grades and/or testing scores were found to qualify as personal information according to the definition in section 2(1) of the Act . Regarding Order PO-2713, Carleton notes that grade information for a single year of the University of Toronto’s law school class could be used to identify students at the upper and lower end of the grade spectrum. Relying on Order MO-1708, Carleton submits that in small classes, the IPC has recognized that students are often aware of their own relative standing in a small course and that given the low number of potential failing grades, release of those grades could effectively reveal the personal information of some students. Relying on these orders, Carleton maintains that the personal information of students at the high and low ends of the grade spectrum would be revealed with disclosure of the data at issue, particularly in competitive programs such a journalism, or in upper year seminars where students are potentially vying for a spot in graduate programs.

[20] In its submissions on section 21(1), Carleton states that it is not aware of any other university in the province that publishes the grades of their students. Carleton takes the position that publishing this data would set it apart from other universities that strongly guard the privacy of their students.

[21] The appellant responds to Carleton’s representations by stating that he did not request the names, addresses, student numbers or any other information that could identify any individual assigned any grade. In this context, the appellant maintains that the records are anonymized. The appellant adds that he responded to Carleton’s concern that disclosure of the grade records for graduate level classes could possibly lead to identification of some individuals, due to the smaller class sizes, by limiting the scope of his request to undergraduate level courses. The appellant submits, accordingly, that the remaining responsive grade data cannot reasonably be corresponded, tied, related or otherwise linked in any way to identifiable individuals.

[22] On the issue of class size, the appellant notes that Carleton failed to consider or apply the widely-accepted principle of small cell count, which establishes a minimum number threshold for identification of individuals. According to the appellant, had Carleton applied this standard, it would only have claimed the personal privacy exemption in relation to classes with five or fewer students, not indiscriminately with regard to all class sizes. The appellant submits that if I choose to impose the small cell count in this appeal, five students in a single class in a single year would be appropriate, given the standard codified in Orders P-644 and MO-1415.

[23] Further, citing Order P-1389, the appellant submits that the mere hypothetical possibility that an individual could be identified by comparing the records to other information is not sufficient to support exempting the information under the Act ; there must be adequate evidence. The appellant relies on Orders PO-1880 and P-1137, where this office did not uphold the institution’s claim for exemption of anonymized data related to the most sensitive possible information about medical practitioners who provide abortion services and hemophiliacs diagnosed with HIV because the evidence amounted to no more than unsupported speculation. The appellant contends that Carleton has failed to present any evidence to support either the supposition that students could be identified by comparing the data at issue with other sources of information or its position that other students could assess the academic performance of their classmates and somehow connect these to the anonymized grade data.

[24] Further, the appellant submits that:

Carleton contends that some individuals could be identified by combining grade records with course calendars, dates of graduation, mandatory course requirements, Facebook and other various sources to link some identified students to assigned grades. Carleton, however, does not offer any explanation of how this could be done nor does it offer even hypothetical examples to support this assertion. This submission is entirely speculative...

[25] The appellant also challenges Carleton’s submission that students’ keen awareness of their standing within a course would enable them to identify which students received what grades. He argues that none of the students would know with any certainty how another student had performed in the course relative to his or her classmates. The appellant explains that students could only speculate, or guess, and he submits that:

These guesses would be based on their own subjective assessments of how their classmates performed, drawn from their own subjective observations of their classmates’ participation in the classroom, based on the possibly incomplete set of observations made only on those days [both the] student and the observed classmate attended class. Moreover, students would not be apprised of their classmates’ grades on written assignments, tests, quizzes, exams, or course work outside the classroom, which typically account for the vast majority of final grades.

These highly speculative and subjective estimates by students of classmate performance could not be used to correspond disclosed grades to identifiable students with any reliability, certainty or accuracy. …

[26] The appellant also tries to distinguish the circumstances of Order PO-2713, noting that Adjudicator Jennifer James found that the LSAT⁶ score and grade information at issue qualified as personal information because:

… students could be identified in cases where their LSAT score was below the usual threshold for admission to law school because these would indicate that the student was admitted as a mature student, an Aboriginal student or another identifiable group not subject to the same required LSAT score for admission as regular students. Thus, [Adjudicator] James found the law school grade assigned to a mature or Aboriginal student, who could reasonably be identified by appearance alone, could be linked to an LSAT score…

[27] The appellant notes that the facts of this appeal differ from those in Order PO-2713 because the information sought – course number, year and grades – is not linked to any other information, and there is no way to link a single student’s grade in one course with a grade assigned in another course or another year.

[28] Respecting Order MO-1708, in which the grades of secondary school students were found to be exempt, the appellant suggests that the reasons do not account for the established small cell count standard of five. The appellant also submits that the secondary school context in Order MO-1708 is distinguishable from the much larger university setting, due to the prevalence of shared classes and social circles in the smaller community context of Order MO-1708 compared to the post-secondary context here, which is characterized by vastly bigger classes, less continuity in contact, and a correspondingly lower likelihood of accuracy in approximating classmate performance, all of which make identifiability more remote.

[29] In supplementary representations, the appellant provided samples of grade distribution information made publicly available by Simon Fraser University (SFU) and the University of British Columbia (UBC).⁷ The appellant submits that the proactive disclosure of student grade data by SFU and UBC suggests that these two universities reached the reasonable conclusion that publication of grades would not identify individual students…

[30] As noted previously, Carleton declined the opportunity to submit representations in reply.

Analysis and findings

[31] To begin, I accept Carleton’s submission that the Act  must be accorded a purposive interpretation.⁸ As I observed in Order PO-2726:

… [it] is well established that the act of statutory interpretation requires consideration of legislative purpose and that a purposive interpretation of the Act  necessitates the balancing of privacy protection principles with those related to access to government-held information [see, for example, Order PO-2693]. In this regard, then, any interpretation of the definition of personal information must be informed by the context in which it arises. Indeed, in past orders of this office where the notion of small cell counts has been relevant, findings have varied, and have been based on a contextual examination of the relationship between privacy and identifiability.⁹

[32] In Pascoe, the Ontario Court of Appeal endorsed the following comments by the Divisional Court below in its review of the purposes of the Act :

The presumption is then in favour of disclosure to promote the goals of government transparency and accountability and to permit public debate. Only if a competing individual interest outweighs those goals will the information be protected.¹⁰

[33] The court also identified that the burden of proof of establishing that an exemption from disclosure applies rests with the institution. In this appeal, therefore, as the first step in establishing that section 21(1) applies, Carleton was required to provide sufficient evidence to satisfy me that the responsive information fits within the definition of personal information in section 2(1) of the Act .

[34] As set out above, personal information means recorded information about an identifiable individual. In Order P-230, former Commissioner Tom Wright set out the basic requirements of identifiability as follows:

If there is a reasonable expectation that the individual can be identified from the information, then such information qualifies under subsection 2(1) as personal information.¹¹

[35] According to Carleton, the grade data is attributable to identifiable individuals because the grades are assigned uniquely to an individual student based on their performance on various evaluation tools during the semester. In my view, the fact that students attain unique grades based on how they perform in a course, or that information about these grades would be about that student’s education under paragraph (b) of the definition, is not really in dispute. Where the evidence provided by Carleton falters is in establishing a reasonable expectation that an aggregated data pool of such uniquely assigned grades can be linked with identifiable individuals.

[36] The question is whether it is reasonable to expect that an individual could be identified when the information at issue is combined with information from sources otherwise available.¹² In the circumstances of this appeal, I conclude that it is not. I agree with the appellant that Carleton’s argument that the data, when considered together with sources such as course calendars, Facebook or other unnamed sources, is speculative. I am not persuaded that the named (and unnamed other) sources could reasonably be used to identify individual students from the grade data at issue because the evidence itself does not point to this conclusion.

[37] In particular, I note that Carleton’s argument that these cross-referencing sources could be used to identify an individual student within the grade data pool is not sufficiently detailed to make the connection for me. Carleton relies on the keen awareness of students as to their own performance and that of others around them and argues that it is reasonable to expect that they could – indeed, would – seek to connect disclosed grade data with identifiable individuals. This argument is similar to one offered by the University of Toronto with some success in Order PO-2713 in relation to students enrolled in the Faculty of Law. I will review that order, below.

[38] The appellant dismisses Carleton’s position, citing Order P-1389, among others. In that decision, Adjudicator Donald Hale considered whether the total billing amounts for the ten highest billing general practitioners in Toronto qualified as personal information, if their names were removed. In that appeal, the Ministry of Health and Long-Term Care put forward a similar argument to that offered by Carleton about peer awareness and inquiry leading to identification. Adjudicator Hale set out the ministry’s arguments, as follows (at page 3):

The Ministry states:

For example, the GP [general practitioner] community in Metropolitan Toronto may have some belief or knowledge that a certain individual is, or is likely to be, one of the top GP billers, or that a select number of individuals comprises, or is likely to comprise, the top billers. Were this the case, and were the ministry to grant full disclosure of the total billings as requested by the appellant, there would arguably be some reasonable connection to be made between the listed billing amounts and one or more GPs.

It is notable that this connection need not link a specific GP to a specific billing amount in order to result in a disclosure of personal information. The ministry is persuaded that it would still constitute an unjustified invasion of personal privacy to disclose billing information that, in conjunction with outside knowledge, pointed to the conclusion that a particular GP must have billed one of the amounts within the range of the 10 amounts contained in the record, or, worse, within the range of, say, the top half or third of those ranked amounts.

Further on, the Ministry submits:

It is not unreasonable to suppose that information exists outside the ministry (within the GP community in Metropolitan Toronto, or elsewhere) that identifies certain GPs as probably amongst the top billers in Metropolitan Toronto. In this context, although it may not be possible to link the amounts on the record with any one GP, the fact that his/her billings are within these amounts on the record means the amounts should be considered personal information.

In my view, the Ministry’s arguments rely on the unproven possibility that there may exist a belief or knowledge of the type described. I have not been provided with any substantive evidence that information exists outside the Ministry which could be used to connect the dollar amounts to specific doctors. The scenario described by the Ministry is, in my view, too hypothetical and remote to persuade me that individual practitioners could actually be identified from the dollar amounts contained in the record. I find, therefore, that the information at issue is not about an identifiable individual and does not, therefore, meet the definition of personal information contained in section 2(1) of the Act  [emphasis in original].

[39] I agree with Adjudicator Hale’s disposition of the peer awareness argument in Order P-1389 for the reasons provided, and I adopt them here. In essence, Carleton’s argument respecting the nexus between the grade data and identifiable individuals is reliant on the reputed existence of an assiduous [student] inquirer, who, with his or her own keen mind and the assistance of one of the listed sources, could identify individual students from the grade data. I find the potential for identifiability contemplated by this scenario to be too remote and speculative.

[40] Further, in my view, Order PO-2713 does not assist Carleton in establishing that the particular grade data at issue in this appeal constitutes personal information. At issue in that appeal was concurrent grade information relating to the 2002, 2003 and 2004 first year law faculty classes and the final averages for the most recent first year class, correlated with LSAT scores. The first group represented approximately 500 students while the second group had approximately 180 students. As indicated previously, the University of Toronto argued that due to the highly competitive nature of law school admissions, performance and employment after graduation, the environment was such that individual students were reasonably capable of being identified by their peers with disclosure of the data.¹³ Adjudicator Jennifer James accepted these submissions, in part. However, Adjudicator James concluded nevertheless that neither the students in the first group of 500, nor those in the median intervals of the second group of 180, were identifiable. The adjudicator was only satisfied that those students whose grades fell into the top and bottom of the grade spectrum were identifiable. For those falling below the median intervals, she found both the competitiveness of law students and the visual identifiability of certain student populations to be persuasive factors.¹⁴ For those students whose grades were above the median intervals, the particular awareness and competitiveness of law students was singularly persuasive in Adjudicator James’ finding that individuals could be identified.

[41] In this appeal, however, I am not persuaded by Carleton’s evidence that the same keen peer awareness or competitiveness could be expected, such that these become relevant factors in assessing identifiability within the undergraduate data pool at issue here. Nor am I satisfied, on the facts of this appeal, that the thousands of students whose data is at issue in this appeal could be visually identifiable by age, race or self-asserted membership in a group, as submitted by the University of Toronto in Order PO-2713.

[42] A closer parallel might be found in Order MO-1708, where the requester sought the grades attained by students from two different cohorts at three schools for English, Science and Math courses. In that decision, Adjudicator Sherry Liang found that disclosure of the grade information could reasonably be expected to result in the identification of students who received a failing grade, even when the grade information related to more than five students. In particular, Adjudicator Liang considered the small class and course sizes¹⁵ and found that … students generally know their own relative standing in a class by the end of the semester, and that they will know the identity of the students most likely to have failed. She therefore concluded that in the particular format requested by the appellant, disclosure of the grades could reasonably be expected to identify individual students at these three secondary schools. I agree with the appellant that Order MO-1708 is distinguishable on the facts from this appeal, where the information at issue relates to grades received in undergraduate courses over a 12 year period.

[43] The appellant also seeks to distinguish Order MO-1708 on the basis that it did not address (as he thought it should) small cell count. In his representations, the appellant described the request that he submitted to the University of Ottawa for similar grade data. That request is not before me. Of note, however, is the evidence that the University of Ottawa expressed concern about the possibility of identifying students using the anonymized grade records in situations where the class had five or fewer students. During the inquiry into this appeal, the appellant proposed a minimum threshold for class size of more than five, suggesting that I might apply the small cell count to the grade data here in fashioning the provisions of my order.

[44] However, I have taken into account the flexibility in the concept of small cell count in reaching my decision not to impose a threshold in this appeal. As described by former Senior Adjudicator John Higgins in Order PO-2811:

the term small cell count refers to a situation where the pool of possible choices to identify a particular individual is so small that it becomes possible to guess who the individual might be, and the number that would qualify as a small cell count varies depending on the situation.¹⁶

[45] Given my conclusion that Carleton’s evidence is not sufficient by itself to conclude that individuals may reasonably be identified by disclosure, I turned to the content of the records to assist me in deciding whether to apply the concept of small cell count to the grade data. In reviewing the grade data, I paid particular attention to the data for the types of classes referred to by Carleton as being of particular concern: upper year seminars and competitive programs, such as Journalism.

[46] In my view, the context in which this data appears is a departure from the formative orders of this office that have considered and applied (or not) the small cell count concept. In many of the orders, adjudicators were called upon to review the Ministry of Health and Long-Term Care’s small cell count policy as the ministry sought to apply it to information, such as OHIP billing data.¹⁷ In Order PO-2204,¹⁸ former Assistant Commissioner Tom Mitchinson reviewed past orders dealing with similar requests and made the following observation:

It is significant to note that Adjudicator Pascoe [in Order PO-1880] rejected the Ministry’s position regarding the application of the small cell count policy for physicians providing abortion services on the basis that there was no evidence as to the likelihood of there being a small number of physicians in the Toronto area performing the types of services and/or the number of services identified in the record at issue in that appeal. Adjudicator Pascoe’s decision was upheld on judicial review by the Divisional Court and the Court of Appeal.

…

I have reached a different conclusion regarding affected party 1. This doctor practices in a highly specialized field, as reflected in both the record and his/her representations. As stated in the representations, he/she is the only doctor in Toronto providing this particular specialized service. Following the reasoning applied by Adjudicator Fineberg in Order P-644, I find that, given the small number of general practitioners who incorporate this area of specialization into their practice and the fact that the procedures outlined on their billing information would reveal this specialized practice, there is a reasonable expectation that the disclosure of the information related to affected party 1 would render him/her identifiable. Unlike the situation in Orders P-1137, P-1389 and P-1880, I am persuaded based on the evidence in this appeal that information in the public domain or in the general practitioner community could be linked to the information relating to affected party 1 in order to make a connection between a particular billing information and the specific doctor.

[47] Order PO-2204 stands as a warning against the rote application of the small cell count principle to all data sets. While the context of other orders dealing with small cell count may be distinct, all implicitly acknowledge that the analysis of whether information is about an identifiable individual is dependent on myriad factors, including the number of individuals and the nature of the information at issue. This is consistent with former Senior Adjudicator John Higgins’ comment in Order PO-2811 that the number that would qualify as a small cell count varies depending on the situation.¹⁹

[48] Accordingly, and with due consideration of the type of information, its particular format and the overall context of this request, I find that the grade data is not reasonably capable of disclosing information about identifiable individuals. Additionally, in the circumstances, I also conclude that it is not necessary to apply the small cell count principle to the grade data.

[49] As I have concluded that it would not be reasonable to expect that an individual may be identified as a consequence of disclosing the grade data, I find that the information does not qualify as personal information according to the definition in section 2(1) of the Act .

[50] Given my finding that the records do not contain personal information, it is not necessary for me to review Carleton’s claim of section 21(1), since the mandatory exemption for personal privacy can only apply to personal information.

C. Does the discretionary exemption for economic or other interests at section 18(1)(c) apply?

[51] Carleton also claims that the grade data is exempt under the discretionary exemption in section 18(1)(c), which states:

A head may refuse to disclose a record that contains,

information where the disclosure could reasonably be expected to prejudice the economic interests of an institution or the competitive position of an institution;

[52] For section 18(1)(c) to apply, Carleton must demonstrate that disclosure of the grade data could reasonably be expected to result in the specified harms. To meet this test, the institution must provide detailed and convincing evidence to establish a reasonable expectation of harm. Evidence amounting to speculation of possible harm is not sufficient.²⁰

[53] The purpose of section 18(1)(c) is to protect the ability of institutions to earn money in the marketplace. This exemption recognizes that institutions sometimes have economic interests and compete for business with other public or private sector entities, and it provides discretion to refuse disclosure of information on the basis of a reasonable expectation of prejudice to these economic interests or competitive positions.²¹

Representations

[54] Carleton submits that ordering the disclosure of the grade data would result in irreversible harm to the university’s competitive position and economic interests. Carleton explains the context of its harms argument, as follows:

Carleton University is one of twenty universities in Ontario competing for students from across the province, the country and the world. Application numbers have been rising, and Carleton is facing increased competition to attract high quality candidates. …

[55] Calling the competition for recruiting undergraduate students fierce, particularly for its specialized programs, Carleton states that students are becoming increasingly critical in their university selection. Carleton notes that there are a number of tools that students can use to assist them with their decision-making, including comparison surveys like those found in Maclean’s magazine with its annual university rankings. According to Carleton, these tools provide an adequate basis for a fulsome assessment of its programs and an informed choice about selecting a post-secondary institution. Carleton argues, however, that:

Disclosure of student grades would put this choice at risk. Instead of a fulsome consideration of all factors, choice for students may be reduced to a comparison of the number of As awarded in a particular program at a particular school.

If the records should show that Carleton is generally a tougher school, awarding fewer As, the near-term economic harm to the school is obvious. Carleton would see a drop in applications, which could be expected to result in enrollment of fewer students at all levels. This will have a trickle-down effect on the quality of research and teaching being done at the school. …

On the other hand, if the data shows that Carleton is generally a softer school, awarding more As in the near-term, this may result in more applications. However, Carleton would expect serious harm to result to its reputation in the long term. A reputation as an easy school would harm Carleton’s ability to attract investment in new programs or research by diminishing the credibility of the work being done at Carleton. …

In either case, the impact on the reputation of Carleton, its degrees, and its students would be grossly unfair. Grades represent one measure of the performance of students in an institution’s programs. They do not represent the full picture.

[56] Carleton also addresses the impact of the disclosure of the grade data from the perspective of consistency of student experience. Carleton submits that students expect that their grades will remain confidential,²² and the disclosure of such information would serve to undermine the confidence of Carleton students in the institution, creating a significant backlash by both former and current students. The balance of Carleton’s submissions directly addressing the application of section 18(1)(c) were held confidential.

[57] The appellant challenges Carleton’s claims under section 18 as speculative, unsupported by any evidence, and based on no research about how and why students choose to attend a particular university. In short, the appellant submits that Carleton has not met the burden of proof because the evidence provided is neither detailed nor convincing. Rather, the appellant submits that:

Prospective students [choose] a post-secondary institution … based on myriad factors that are more compelling and immediate than information about the long-term trends in average grades: the location of a college or university, programs of study offered, admission requirements, academic reputation, co-operative programs, tuition fees charges, the availability of on-campus and off-campus housing, the availability of financial assistance such as loans, bursaries and scholarships, facilities such as laboratories and libraries, recreational facilities and programs, the number and quality of instructors, class sizes, the rates of employment among graduates in their chosen fields of study, the diversity of on-campus entertainment and social events, the gender ratio of the student body, and countless other considerations are all weighed by prospective students. All of this information is either in the public domain or available freely upon inquiry. Adding information about general trends in grade assignment to this long list would not harm Carleton’s enrollment rates.

[58] The appellant submits that Carleton’s submissions on harm are premised on the speculative assumption that grades awarded in the past are predictive of grades that will be assigned in the future and that Carleton’s grades would diverge substantially from those at other Ontario universities. The appellant also maintains that the possibility that Carleton’s grades may be softer than other Ontario universities may result in embarrassment for the university, but that is not sufficient to support exemption of the grade data under section 18(1)(c).

[59] The appellant also submits that the publication of grades by SFU and UBC shows that neither of those universities shares Carleton’s view that their competitive or economic position would be compromised or damaged by disclosure of the grades. The appellant points out that these universities also compete aggressively for students, particularly in smaller or specialized programs, and he contends that their administrators must have concluded that grade publication would not harm their respective economic positions.

Analysis and findings

[60] The purpose of section 18 is to protect certain economic interests of institutions. Released in 1980, the report titled Public Government for Private People: The Report of the Commission on Freedom of Information and Individual Privacy²³ provides the following description of the rationale for including a valuable government information exemption in the Act :

In our view, the commercially valuable information of institutions such as this should be exempt from the general rule of public access to the same extent that similar information of non-governmental organizations is protected under the statute . . . Government sponsored research is sometimes undertaken with the intention of developing expertise or scientific innovations which can be exploited.

[61] Section 18(1)(c) does not require Carleton to establish that the information in the record belongs to it, that it falls within any particular category or type of information, or that it has intrinsic monetary value.²⁴ The exemption requires only that disclosure of the information could reasonably be expected to prejudice Carleton’s economic interests or its competitive position.²⁵

[62] Based on my review of Carleton’s representations, the appellant’s submissions, and the nature of the particular information at issue, I am not persuaded that its disclosure could reasonably be expected to prejudice Carleton’s competitive position or any corresponding economic interests, as contemplated by section 18(1)(c).

[63] I accept Carleton’s submission that grades represent [only] one measure of the performance of students in an institution’s programs. They do not represent the full picture. However, just as student performance is measured by more than grades, so too is the worth and attractiveness of a post-secondary institution to a prospective student. On this point, I agree fully with the appellant that a typical prospective student considers myriad factors and features in choosing which post-secondary institution to attend. The suggestion that disclosure of the grade data could reasonably be expected to harm Carleton in the manner described attributes an unwarranted one-dimensional quality to the decision-making process or, worse still, to the high quality candidates Carleton aims to attract, themselves. If anything, I would expect disclosure of the grade data to assist students to make a more informed decision, as it simply constitutes one dimension of many.

[64] Carleton’s submissions under section 18(1)(c) also rely on concerns about breaching the confidentiality of its past and current students and the resulting backlash if the university were ordered to disclose the grade data. Carleton alleges that disclosure of this information would undermine the confidence of Carleton students in the university, particularly in light of their reliance on the confidentiality provisions in its Undergraduate Teaching Regulations and Procedures and the Student and Applicant Record Policy, both of which articulate prohibitions on disclosure of students’ personal information, including academic performance information.

[65] However, I find this argument unpersuasive for several reasons. First, it is predicated, at least in part, on the assumption that the information sought is personal information. I agree that students have the right to expect that their personal information, including academic performance, will be kept confidential, but I have already rejected Carleton’s claim that the grade data constitutes the personal information of its students. Since the grade data is not personal information, it is not, therefore, about the academic performance of individual students, and its disclosure could not be expected to breach the regulations or policy. Further, the confidentiality and non-disclosure obligations outlined in the documents provided by Carleton arise in respect of voluntary disclosure and not disclosure pursuant to the operation of the access to information regime to which the university is clearly subject.²⁶

[66] Next, I have considered the evidence provided by the appellant respecting the publication of course grades by SFU and UBC in British Columbia. While the practice varies somewhat between the two institutions, I am satisfied that these two universities publish grade data that is essentially the same in format and content as that sought by the appellant from Carleton in this appeal. The practice of similar institutions voluntarily disclosing essentially similar information has been held to be a factor in determining the reasonableness of a claim to harm under section 18(1)(c).²⁷ In this appeal, I provided Carleton with an opportunity to comment on this aspect of the appellant’s representations. Presumably, Carleton could have tendered evidence of harm, including compromised student expectations and/or backlash by students of SFU or UBC. That Carleton was unable, or unwilling, to provide evidence in support of the reasonableness of its expectation of harm in this regard response to the public availability of SFU/UBC’s grade data further minimizes the strength of its position.

[67] Based on the evidence before me, it is unclear to me how the scenarios described could reasonably be expected to occur with disclosure of the grade data and lead to harm, particularly irreversible harm, as Carleton has alleged. As Carleton has failed to provide me with sufficiently detailed evidence to establish a link between the disclosure of the grade data and a reasonable expectation of either of the harms that section 18(1)(c) is intended to protect against, I find that it does not apply.

[68] Given my finding that the grade data information is not exempt, I will order Carleton to disclose it. Furthermore, because I have not upheld Carleton’s exemption claim under section 18(1)(c), it is not necessary for me to review the university’s exercise of discretion. Similarly, because I have found that neither section 21 or 18 applies, it is also unnecessary for me to determine whether the public interest override in section 23 of the Act  applies.

ORDER:

I order Carleton to disclose the grade data to the appellant by sending him an electronic copy of the information by August 30, 2013, but not before August 26, 2013.

Original signed by: Daphne Loukidelis Adjudicator

July 25, 2013

---

¹ Although Carleton’s October 29, 2012 revised decision letter did not specifically identify section 22(a) of the Act  (publicly available information), I interpreted this to be the university’s position with regard to the appellant’s access to the grade abbreviation record. In turn, the appellant’s representations suggest acceptance of this position, on the condition that the grade abbreviation summary available on the university’s website is satisfactory for interpreting the student grade data, if disclosed.

² That is, unless one of the exceptions in paragraphs (a) to (f) of section 21(1) applies.

³ Order 11.

⁴ Order PO-1880, upheld on judicial review in Ontario (Attorney General) v. Pascoe, [2002] O.J. No. 4300 (C.A.) (Pascoe).

⁵ Here, Carleton relies on well-established case law in this area: Rizzo & Rizzo Shoes Ltd. (Re), [1998] 1 S.C.R. 27; Dagg v. Canada (Minister of Finance), [1997] 2 S.C.R. 403.

⁶ Law School Admission Test.

⁷ SFU’s grade data is publicly available online, while UBC’s is provided on a course-by-course basis by request. UBC’s data features several more data fields than are at issue in this appeal; for example, the grades are set out in 10-point ranges. No small cell count is applied to the data. SFU applies a five or fewer small cell count to its data. More detail was provided in the appellant’s supplementary representations.

⁸ The Supreme Court of Canada has affirmed the purposive approach to statutory interpretation outlined in Rizzo, supra, on numerous occasions. For a recent case, see Re: Sound v. Motion Picture Theatre Associations of Canada, 2012 SCC 38.

⁹ The request in Order PO-2726 was for one-time data snapshot showing the length of sentence for individual inmates sentenced to terms of two years less a day, as well as the corresponding last known postal code for each inmate.

¹⁰ Pascoe, supra, at paragraph 12.

¹¹ [1991] O.I.P.C. No. 21.

¹² Pascoe, supra; as applied in many decisions, including more recent ones such as Orders MO-2407, PO-2551, PO-2726 and PO-2811. Order PO-2811 was upheld in Ontario (Ministry of Community Safety and Correctional Services) v. Ontario (Information and Privacy Commissioner), 2011 ONSC 3525 (Div. Ct.); appeal dismissed, 2012 ONCA 393 (C.A.); ministry’s appeal to SCC pending (SCC 34949).

¹³ As outlined in the order, the University of Toronto submitted that: "identifiability or knowledge of students’ membership in visually identifiable groups, persists long after graduation, not only in ongoing relationships with and the memory of classmates, but also in class photographs which they may have and which are available at the Faculty. In addition, there is an on-line and print directory of alumni which can be used to find classmates and as a basis for further inquiry, for example, through internet searches, including, where available, perusal of internet resources such as photos on law firm and other web pages (page 10).

¹⁴ For example, mature students were said to be visually identifiable, in that they look older and, therefore, different from their peers at the law school.

¹⁵ At page 5 of Order MO-1708, Adjudicator Liang accepted the school board’s evidence that each class could have as few as 8-10 students, while each course could have as few as 12 to 15 students.

¹⁶ Order PO-2811 at page 8.

¹⁷ At the time these orders were issued, the ministry’s policy provided that: when the processing of anonymized personal health information yields tabulations of less than five in which a possibility exists where an individual person could be identified, such information will only be released to an agency head or consultant/researcher and will not be included in the statistical report. There is some inconsistency in reference in the orders; the standard is variously referred to as five or fewer or less than five.

¹⁸ In Order PO-2204, the request sought: amounts that the top 10 billing general practitioners/family doctors in Toronto billed OHIP in the most recent fiscal year; fee codes of the top 10 items that each individual doctor billed most frequently and a brief description of those codes; and gross amount paid in the most recent fiscal year next to each of the 10 fee codes for every one of the 10 doctors. The ministry granted access to the total fees paid to each of the 10 doctors, but denied access to the fee codes, the itemized fee payments, and billing service descriptions.

¹⁹ See, for example, Orders P-644 and P-1137, where Adjudicator Anita Fineberg considered the small cell count in similar contexts and still reached different conclusions about the identifiability of individual physicians from the data.

²⁰ Ontario (Workers’ Compensation Board) v. Ontario (Assistant Information and Privacy Commissioner) (1998), 41 O.R. (3d) 464 (C.A.).

²¹ Orders P-1190 and MO-2233.

²² With its representations, Carleton provided copies of its Undergraduate Teaching Regulations and Procedures and Student and Applicant Record Policy, both of which address the confidentiality and non-disclosure of a student’s personal information, including academic performance information.

²³ Vol. 2 (Toronto: Queen’s Printer, 1980).

²⁴ This distinguishes the exemption from section 18(1)(a), which is arguably narrower in scope.

²⁵ Order PO-2014-I.

²⁶ See Order PO-1805, where former Senior Adjudicator David Goodis reviewed the possible disclosure of peer reviews of nuclear stations in the context of section 18 of the Act .

²⁷ See Order PO-2404, where Adjudicator John Swaigen reviewed the possible application of section 18(1)(c) to certain investment information regarding OMERS [the Ontario Municipal Employee Retirement System].