By way of background, the LSAT is the Law School Admission Test administered by the Law School Admission Council (LSAC).  The LSAC prepare school specific correlation/validity studies for participating law schools in Canada and the United States.   The studies are commonly referred to as correlation studies and evaluate the effectiveness of the LSAT as well as other predictors, such as undergraduate grades, in predicting first-year law school performance.  The predictive statistical information contained in the study is based on the student’s actual LSAT score and undergraduate grade point average.  The correlation studies provide participating schools statistical data relating to the most recent academic year and the school’s last three academic years. 

The requester (now the appellant) appealed the University’s decision to this office.

RECORDS:

The information at issue consists of tables, graphs and an appendix contained in a 32-page document entitled “LSAT Correlation Studies Report on First-Year Performance”, dated Fall 2005.

The portions of that document which remains at issue are the tables, graphs and appendix found at pages 6-11, 14-16, 18, 20 and 28-32.  These pages contain information regarding the following five performance related variables:

• LSAT score (LSAT);
• Undergraduate grade point average (UGPA);
• Actual first year law school average (Actual FYA);
• Predicted first year law school average (Predicted FYA); and
• Minimum and maximum range of the predicted year law school average (Predicted minimum and maximum FYA range).

Pages 6, 7 and 8 contain frequency distribution and summary statistic tables for the LSAT, UGPA and FYA.  The frequency distribution tables record the number and percentage of students obtaining scores within a specified range.  These specified ranges are called intervals. 

Page 6 contains a table describing the number and percentage of the 2004 and combined first-year students who obtained LSAT scores within intervals representing about every four points.

Page 7 contains a table describing the number and percentage of the 2004 and combined first-year students whose UGPA scores fall within intervals representing about every 1.89 per cent.

Page 8 contains a table describing the number and percentage of the 2004 and combined first-year students whose FYA falls within intervals representing about every 1.39 per cent.

The summary tables found at pages 6, 7 and 8 identify the minimum, 25^th percentile, 50^th percentile, 75^th percentile and maximum scores relating to LSAT, UGPA and FYA scores along with information about the number of students, mean standard deviation and range for 2004 first-year and combined year students. 

The information relating to the 2004 and combined first-year students’ LSAT, UGPA and FYA scores contained in the frequency distribution tables are also depicted as bar graphs on pages 9, 10 and 11 of the record. 

Pages 14, 15 and 16 contain scatterplot graphs which show the correlation between the LSAT and FYA, UGPA and FYA, and LSAT to UGPA for the 2004 class.

Page 18 contains a scatterplot graph of the actual and predicted FYA for the 2004 class.

Page 20 contains a table which describes the correlation between the actual and predicted FYA for combined years.

Pages 28 – 32 is an appendix attached to the report which lists the LSAT, UGPA, actual FYA, predicted FYA and predicted FYA range for each student enrolled in the 2004 first-year class. The information is arranged from the highest to lowest FYA.  As noted above, the appellant does not seek access to the names of students contained on these pages.

DISCUSSION:

PERSONAL INFORMATION

The term “personal information” is defined, in part, to mean recorded information about an identifiable individual, relating to the education history of the individual [paragraph 2(1)(b) of the definition of “personal information” in section 2(1) of the Act].  Having regard to the definition of “personal information”, I am satisfied that information about an identifiable individual’s LSAT score and grades, including predictive grades based on their actual LSAT and undergraduate marks, would constitute that individual’s “personal information”. 

The issue raised in this appeal is whether the grade information at issue without students’ names would also qualify as “personal information” under the Act.  In Order P-230, former Commissioner Tom Wright stated:

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

To qualify as personal information, it must be reasonable to expect that an individual may be identified if the information is disclosed [Order PO-1880, upheld on judicial review in Ontario (Attorney General) v. Pascoe, [2002] O.J. No. 4300 (C.A.)].

Therefore, I must decide whether any of the 2004 and/or combined first-year students, who are not identified by name, but whose LSAT scores and grades are reported in the responsive record, could reasonably be identified given the information contained in the record and the surrounding circumstances.

Representations of the parties

As noted above, the University was asked to consider whether the information at issue constitutes “personal information” as defined under the Act, taking into account that the names of individual students are not at issue, and that the information at issue relates to combined years, as well as one specific year.

The University takes the position that there exists information in the public domain and in the legal community which could be used to link the information at issue to identify former students, particularly those who were admitted with grades and/or LSATs below the Faculty’s admission standards.  The University states that these identifiable students make up a sub-group within the general population that constitute Aboriginal, mature and special regular students. 

The University’s representations state:

Although this information does not contain names, it describes readily identifiable individuals – such as mature students – some of whom are admitted to the JD program with lower grades and/or LSAT scores than the rest of the class.

Some exempted information is tabular, in the form of graphs.  It shows how many students were admitted below the Faculty’s stated admission criteria, which are clearly set out on its website.  When combined with posted admissions statistics and data about special admissions categories, exempted information describes a significant proportion of individuals in identifiable small groups who are admitted below usual, posted admissions criteria …

Even if the names are removed … some of the remaining information would be identifiable to other members of the class who know, for example, that certain individuals are doing particularly well or that some individuals are trailing behind.

            …

The University of Toronto statistics in the record contain information about individuals who belong to a small special category.  Since admission requirements for those individuals differ from the main group of law students, releasing (even depersonalized) information about individuals at the ends of distribution would support the making of accurate inferences possible about their likely FYA, LSAT or UGPA.

The Faculty admission standard the University refers to in its representations states:

Very few candidates are admitted with LSAT scores at or below the 85^th percentile and cumulative undergraduate academic records below 78%, unless their backgrounds, other qualifications, or personal accomplishments would, in the opinion of the Admissions Committee, contribute specially and significantly to the class.

The above-noted admission standard is no longer posted on the Faculty’s website, as in recent years the University has attracted students with slightly higher LSAT scores and undergraduate grades.  It was, however, the Faculty’s admission standard at the time the record at issue was created and thus will be the one referred to in this Order.

The posted admissions statistics the University refers to in its representations are generated by the Faculty’s Admission Committee every Fall and set out the:

•         total number of applications received that year;

•         total number of students registered in the current first year;

•         total number of regular students registered in the current first year;

•         total number of mature students registered in the current first year; and

•         total number of Aboriginal students registered in the current first year.

The University also refers to the statistics contained in its Annual Report on Student Financial Support.  This report includes statistical information relating to admissions for the past ten years and information relating to the number of applications, offers, deferrals and registrations. 

The University takes the position that the identified sources of public information, when combined with the information at issue, could lead to the identification of students who were admitted below the Faculty’s admission standard (LSAT scores below the 85^th percentile and/or cumulative undergraduate grades below 78%).  The University also submits that it is reasonable to expect that disclosure of the information at issue would also lead to the identification of students who received top marks in their first year.

In support of its position, the University also provided information regarding the differences between regular, mature and Aboriginal applicants.  The University explained that the majority of its student population is made up of regular applicants who met the Faculty’s admission standard.  The University advises that there are two small sub-groups in the pool of regular applicants and explains in its representations that:

[o]ne sub-group can be characterized as students who belong to a visible minority group and/or come from dire socio-economic circumstances.  The second, smaller sub-group consists of students who had strong grades, attended highly regarded undergraduate institutions, and participated in challenging undergraduate programs, but whose LSAT scores were lower than those of other regular applicants (below the 85^th percentile).

With respect to mature students, the University states:

A small number of mature applicants are admitted each year, a substantial proportion of whom have LSAT scores and GPAs lower than posted Faculty standards.

The University, in the confidential portion of its representations, provided statistics in support of its position that a smaller proportion of special regular students are admitted below Faculty standards compared to the number of mature students admitted below Faculty standards.  The University did not make specific representations or provide confidential statistics regarding Aboriginal students. 

The University submits that disclosure of the information at issue along with posted information regarding the number of students in specific admission categories could lead to the identification of students belonging to small sub-groups.  The University’s representations state:

These individuals may be visually identifiable for example, because the mature students generally look older than the majority of the Faculty’s students and because the Faculty has a tradition of having many student clubs and activities, organized around ethnicity, culture and age.  In particular, the Faculty of Law has an active Aboriginal Law Students’ Association, a “Senator’s Club” (mature students) and a Black Law Students’ Association.  None of these groups promote the public disclosure of their grades and LSAT data individually or collectively.

The University advises that it “relies on a ‘small cell’ concept to withhold from disclosure information about groups small enough that individuals could reasonably be identifiable” and refers to Orders P-644 and MO-1415 which accepted the position that anonymized information relating to a group of five or fewer individuals could reasonably result in the identification of one of the individuals.  Most of the information at issue in this appeal, however, relates to groups of five or more individuals.  In any event, the University submits that the approach taken by Adjudicator Sherry Liang in Order MO-1708 is better suited to the information at issue in this appeal.  

In Order MO-1708, Adjudicator Liang found that disclosure of the grades unnamed high school students received in a specific course taught at a specific school could reasonably lead to the identification of some students.  Adjudicator Liang found that disclosure of the grade information at issue, even in cases where the grade information related to more than five students, could reasonably be expected to result in identifying students who received a failing grade.  In making her decision, she took into account the small class and course sizes 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”.

Like the institution in Order MO-1708, the University submits that the Faculty’s students have a special ability to ascertain the academic performance of their classmates.  In this regard, the University states that law students in general are:

…very interested in their academic performance and that of their peers and in other indicators, such as LSAT scores.  Such indicators and academic outcomes are related to later employment and career outcomes.  Law students are known to be competitive and focussed on relative performance and ranking within the group.  The members of this community can be expected to do what they are able to ascertain their relative position in the class and all that they can learn about the performance of their peers.

The appellant distinguishes the information at issue in Order MO-1708 from the information at issue in this appeal.  The appellant argues that the information at issue in MO-1708 represented one performance-related variable, the actual grade the unnamed student received in a specific course, as opposed to the five variables at issue in this appeal.  The appellant’s position is that disclosure of the information at issue could not reasonably lead to the identification of an identifiable individual as “…it would be impossible for an individual to identify a given individual, without names or student numbers, based on the myriad of different statistical factors included in the record.”  Further, the appellant argues that unlike the information at issue in Order MO-1708, none of the performance variables described in the record represent the actual grade a student received in a specific first year law course.

The appellant also questions the University’s claim that disclosure of the information could reasonably result in the identification of students, taking into consideration that the information at issue is dated and many of the students are no longer law students and have, in most cases, graduated.  The appellant also states that the University’s practice of accepting transfer students makes the possible identification of students unlikely.

In addition, the appellant takes the position that the University has failed to provide sufficient evidence to demonstrate that disclosure of the information at issue could reasonably result in the identification of students.  In support of his position, the appellant refers to Order P-1389 in which Adjudicator Donald Hale considered the Ministry of Health’s argument that linkages between the medical billing information and information in the public domain and/or general practitioner community could reasonably result in identifying the general practitioners who submitted the billings.  In that Order, Adjudicator Hale rejected the Ministry of Health’s argument and stated:

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]

The appellant argues that the evidence provided by the University relies on “assumptions and hypothetical scenairos, such as the ‘competitive’ nature of law school students” and “unfounded stereotypical assertions”. 

As noted above, the University was provided with the non-confidential portions of the appellant’s representations and was given an opportunity to provide reply representations.  The University, in its reply representations, responded to the appellant’s submission that the information was dated.  The University also provided details of how linkages can be made between the information at issue and external information.  In particular, the University states:

It is the University’s submission that the information contained on pages 28-32 of the record is personally identifiable, even with names removed, and that classmates could make very strong inferences with respect to several individuals at the top and bottom of the range for each of the statistics listed.

It is likewise the University’s submission that the information at pages 6-8, 14-16, 18 and 20 of the record can be used to infer or derive personal information of a significant proportion of the members of a small, identifiable subset of the class.

…

It is the University’s view that the release of the record would highlight too many members of small identifiable groups, rendering those individuals identifiable – or more accurately, disclosing information that can be linked to those individuals as a group.  As explained [in the original representations], potential identifiability of individuals in these groups is increased by lower admission standards than those generally applied to regular applicants without special circumstances – those standards are posted on the Faculty of Law website… To put it bluntly, it would be obvious that individuals admitted below the posted admissions standards are in all likelihood members of the three identified groups.  The only remaining question would then be which member.  This question would likely be answerable in the specific for individuals at the bottom of the distribution.  The number who can be specifically identified might vary with the quantity and quality of information on standing that has been garnered by classmates over time.

This identifiability persists after graduation, possibly for many years, depending on the relationships and ongoing contacts of graduates after their time at the Faculty.

…

After graduation, classmate relationships and understandings do not disappear but continue to develop as alumni find their way in the legal and related professions.  Classmates often stay in touch and/or come into frequent professional contact with each other after graduation, continuing to take an interest in each other’s careers and socializing, sometimes for the entire span of their careers or lives.

In fact, 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.  With these resources, it is not only possible, but easy, for interested parties – especially but not necessarily classmates – to revisit years later the question of membership in visually identifiable groups with special admission characteristics and to compare them with publicly available information.

In response to the appellant’s submission that the number of performance-related variables described in the record makes it impossible to identify individuals, the University takes the position that the variables are linked, as they relate to student performance and thus increase the likelihood of identifiability.

The University’s reply representations also responded to the appellant’s submission that it failed to provide the type of evidence described in Order P-1389 to establish that disclosure of the information at issue could reasonably result in the identification of former students.  In this regard, the University submits that the finding in Order P-1389 was “… specific to a vastly larger and more disparate community than Faculty of Law classes considered in the present appeal.” The University goes on to state:

It is reasonable to assume that family doctors [who were educated at different times at many universities] in Toronto would be far less likely to have detailed knowledge about each other and each other’s relative earnings than the much smaller group of law students in each year’s class at the faculty.  They are not the same sort of closely-knit community as a class at the Faculty.

For comparison, classes at the Faculty for years in question consist of some 180 students.

Each year or year of students at the Faculty is a small, tightly-knit community with strong networks of relationships and shared or understood information about each other.  Characteristics of each class relevant to their identifiability include that:

•               they are in daily or nearly daily contact,

•               they are competitive and strongly interested in rankings, grades and scores,

•               they have a good sense of grade distribution and of the relative ranking of individuals within the group, particularly the high and low ends,

•               the variables (UGPA, LSAT, actual and predicted FYA) for each individual compromise a list of linked factors (not a single factor such as income in Order P-1389) and,

•               [t]he groups with largest concentrations of low scores may be visually identifiable by age, race or self-asserted membership in a group.

Analysis and Decision

The University submits that disclosure of the information at issue could reasonably result in the identification of students, particularly taking into consideration the availability of public information relating to its Faculty admission standards, admission policies, admission statistics and alumni.  The University submits that the small, competitive, tightly-knit nature of its first-year classes increase the likelihood of interested parties identifying former students, particularly those that did especially well or poorly during their first year.  The appellant argues that the University has failed to provide sufficient evidence in support of its position that disclosure of the anonymized information at issue could reasonably result in identifying any specific individual.

As noted above, the grade information reported in the 2005 correlation study relates to students who completed their first year studies in the most recent year (2004) and the last three years (combined years). Accordingly, I will consider the grade information relating to the combined years and most recent year separately.

Grade information relating to combined years

The Notice of Inquiry sent to the University asked it to provide representations addressing the fact that some of the information at issue relates to combined years, as well as one specific year.  In particular, the University was invited to explain how disclosure of concurrent grade information relating to the 2002, 2003 and 2004 first year classes, representing a total of approximately 500 students, could reasonably result in the identification of individual students.  The University did not specifically address this issue in its representations. 

I have carefully reviewed the representations of the parties, along with the records themselves and am not satisfied that disclosure of grade information relating to combined years could reasonably lead to the identification of students who completed their first year during 2002, 2003 and 2004.  In making my decision, I took into account that the grade information relating to combined years represent a large group of students who, in my opinion, do not form a closely-knit or ‘small cell’ group.  In that regard, I note that only a third of the combined first-year students were in daily contact with each other as a result of taking their first-year law courses together.   In my view, disclosure of grade information relating to such a large group of students could not reasonably result in the identification of identifiable students. Accordingly, I find that this information does not constitute “personal information”, as described in the definition of that term in section 2(1) of the Act.  As a result of my finding, I will order the University to disclose the grade information relating to combined years 2002, 2003 and 2004 to the appellant as the exemption claimed for this information can only apply to personal information.

Grade information relating to the 2004 first-year class

I have carefully reviewed the tables, graphs and appendix and note that the majority of LSAT scores, UPGAs and FYAs fall within three or four intervals.  Most of the LSAT scores fall into three intervals and most of the UPGAs and FYAs fall into four.  In most cases, the grade information reported in these intervals fall just above or below the 25^th, 50^th and 75^th percentile grades identified in the summary statistic tables.  The range of grades in these intervals is very small.  The intervals show that the majority of the 2004 first-year class obtained similar LSAT scores and UPGAs and ended up obtaining similar first year averages.  The grade information in these intervals do not capture grades falling in the top or bottom of the grade spectrum.  For the remainder of this Order I will refer to these intervals as the median intervals.

In my view, disclosure of the grade information falling in the median intervals could not reasonably result in the identification of identifiable students and thus does not constitute “personal information” as defined in section 2(1) of the Act.  In making my decision, I took into account that the majority of students fall within these intervals and there is little disparity in the grade information reported in them.  In my view, disclosure of the large number of students reported in each of these intervals could not reasonably result in the identification of individual students.  Accordingly, I will order the University to disclose this grade information to the appellant as the exemption claimed for this information can only apply to personal information.

I will now go on to determine whether disclosure of the grade information relating to students falling outside the median intervals could reasonably result in the identification of identifiable students.   In doing so, I will consider the University’s position that disclosure of the information at issue could reasonably result in identifying former students at the top and bottom of the grade spectrum.

a)         Grade information relating to the 2004 first-year class below the median intervals

The grade information below the median intervals includes the frequency distributions and minimum score for the LSAT, UGPA, FYA, predicted FYA and predicted FYA range.

The admission statistics posted by the Faculty’s Admission Committee indicates that as of September 2004 a total of 182 students were registered.  However, the 2004 first-year grade information at issue relates to a smaller number of students.

Accordingly, it appears that some of the students registered for the 2004 first-year class did not complete their first year at the time the correlation study was issued.  The appellant already submits that the linkages the University argues can be made between the information at issue and public information are improbable, taking into consideration the fact that the University accepts transfer students, the information at issue is dated and there are too many performance-related variables to make linkages without the names of students.  In my view, the fact that some students registered for the 2004 first-year class did not complete their first year by the time the correlation study was issued would make it even more difficult to identify students at the top or bottom end of the grade spectrum.

The University submits that the potential for identifying individual students belonging to the three identified sub-groups – mature, Aboriginal and special regular students – is increased by lower admission standards.

With respect to mature students, the University submits that because mature students look older or self-identify by participating in student clubs or activities organized by age, their classmates could accurately discern who is a mature student.  The University also submits that students having ascertained their own academic standing, could identify which mature students trailed behind the class to make accurate inferences as to which mature students were admitted with lower LSAT scores and UGPAs.   The University also submits that Aboriginal and some special regular students can be identified from the general student population as a result of their appearance and/or participation in student clubs.   

The University provided statistics regarding mature students in the confidential portion of its representations to show that a “…small number of mature applicants are admitted each year, a substantial proportion of whom have LSAT scores and GPAs lower than posted Faculty standards”. The confidential statistics set out the number of mature students admitted in 2002, 2003 and 2004 and identify the number of mature students admitted with LSAT scores and UPGAs below the Faculty standard. 

The University did not provide specific representations or confidential statistics relating to Aboriginal students but did provide confidential statistics relating to special regular students.  These statistics set out the number of regular students for 2002, 2003 and 2004 and the number of special regular students admitted each year.  The University submits that these statistics establish that the number of special regular students who are admitted with lower LSAT scores and GPAs are much fewer than the proportion of mature students admitted with lower grades.  In fact, the number of special regular students admitted in 2004 is so low that it calls into question the University’s position that disclosure of the grade information below the median intervals could reasonably result in the identification of some special regular students.

However, having regard to confidential statistics provided by the University, I am satisfied that the number of mature students the University indicates were admitted with lower LSAT scores and UGPAs represents a number that could reasonably be expected to result in the identification of some mature students.  Accordingly, it is not necessary that I make a finding as to whether the grade information falling below the median intervals also constitutes the personal information of Aboriginal and/or special regular students.

In making my decision that the grade information falling below the median intervals qualifies as “personal information” of some mature students, I accept the University’s argument that accurate inferences as to which students look more mature and thus make up the mature student population of the 2004 first-year class can be made.  I also accept the University’s position that its law students are competitive and have the ability to ascertain which students in their class did poorly in their first year, even after graduation.  Given that a small number of mature students were admitted to the 2004 first-year class, I am satisfied that disclosure of the grade information contained in the intervals below the median intervals, including the minimum scores, could result in accurate inferences between the information at issue and mature students others perceive were trailing behind their class.  Accordingly, I find that disclosure of this information constitutes “personal information” of some mature students as described in the definition of that term in section 2(1) of the Act.

b)         Grade information relating to the 2004 first-year class above the median intervals

The grade information contained in the intervals above the median intervals include the frequency distributions and maximum score for the LSAT, UGPA, FYA, predicted FYA and predicted FYA range.

The University submits that disclosure of the information at issue could reasonably result in the identification of students who excelled in their first year.  The appellant rejects the University’s submission on the basis that the report does not contain information relating to a particular course, but rather the student’s overall first-year average.

The competitive law school culture described by the University is fuelled, in part, by the limited availability of prestigious summer and articling opportunities.  First-year averages are often used by law firms to identify top students during the summer recruitment process.  The recruitment process is competitive as coveted summer positions are well-paid and could lead to lucrative articling and associate positions.

Having regard to the importance of first-year averages to law school students, I accept the University’s position that its former students have the ability to ascertain which students in their class excelled in their first year.  I also accept that these students keep track of each other’s academic and professional careers, including knowing who was offered coveted summer, articling and associate opportunities.

I have carefully reviewed the grade information contained in the intervals above the median intervals and am satisfied that disclosure of the grade information contained in these intervals, including the maximum score, could reasonably result in the identification of the 2004 first-year students who obtained high LSAT scores, UGPAs and FYAs.  Though in some cases more than five students obtained grades reported above the median intervals,  I am satisfied that disclosure of this number of students, taking into account the competitive close-knit community of law students, is sufficiently small to lead to the identification of these students. 

Accordingly, I find that disclosure of the number and distribution of LSAT scores, UGPAs and FYAs above the median intervals could reasonably result in the identification of some students and thus constitutes “personal information” as described in the definition of that term in section 2(1) of the Act.

Summary

I find that the grade information relating to combined years does not relate to an identifiable individual and thus does not constitute “personal information” as described in the definition of that term in section 2(1) of the Act. As a result of my finding, I will order the University to disclose the grade information relating to combined years to the appellant.

With respect to the grade information relating to the 2004 first-year class, I find that the grade information falling within the median intervals does not relate to an identifiable individual and thus does not constitute “personal information” for the purposes of section 2(1) of the Act.  Accordingly, this information, if it can be reasonably severed, should be disclosed to the appellant.  In my view, the graphic nature of the scatterplot graphs and chronological order of the information contained in the appendix render these records difficult to reasonably sever.  This same information, however, is contained in the frequency distribution and summary statistics tables and related bar graphs.  Accordingly, I have decided to order the University to disclose only the grade information falling in the median intervals contained in the tables and bar graphs to the appellant.   I will provide the University with a highlighted copy of tables and bar graphs along with this Order.

With respect to the grade information falling below and above the median intervals, I find that disclosure could reasonably result in the identification of some mature students and students who did very well in their first-year and thus constitutes the “personal information” of these individuals as described in the definition in section 2(1) of the Act. 

As a result of my findings, I must go on to consider whether disclosure of the grade information I found qualifies as “personal information” for the purposes of section 2(1) would constitute an unjustified invasion of privacy under section 21(1) of the Act.

UNJUSTIFIED INVASION OF PERSONAL PRIVACY

Where a requester seeks personal information of another individual, section 21(1) prohibits an institution from releasing this information unless one of the exceptions in paragraphs (a) to (f) of section 21(1) applies. If the information fits within any of paragraphs (a) to (f) of section 21(1), it is not exempt from disclosure under section 21.  In the circumstances, it appears that the only exception that could apply is paragraph (f) (disclosure does not constitute an unjustified invasion of personal privacy).

Section 21(4) refers to certain types of information whose disclosure does not constitute an unjustified invasion of personal privacy.  If any of paragraphs (a) to (d) of section 21(4) apply, disclosure of the information is not an unjustified invasion of personal privacy and the information is not exempt under section 21(1).  The appellant has not claimed that any of the exceptions listed in paragraphs (a) to (d) apply in the circumstances of this appeal and I am satisfied that none of the exceptions apply.

Section 21(3) lists the types of information whose disclosure is presumed to constitute an unjustified invasion of personal privacy, if any of the paragraphs (a) to (h) of section 21(3) apply, disclosure of the information is presumed to be an unjustified invasion of personal privacy under section 21(3).  The University claims that the presumption at section 21(3)(d) applies to the information at issue.  This section reads:

A disclosure of personal information is presumed to constitute an unjustified invasion of personal privacy where the personal information, relates to employment or educational history

The appellant did not make representations regarding the possible application of section 21(1) as he takes the position that none of the information at issue contains personal information.

The University submits that the presumption at section 21(3)(d) applies “…since grades, LSAT scores and relative position in the class [amount to the] educational history of the individuals concerned.”   I agree and find that the presumption at section 21(3)(d) applies to the grade information I found constitutes “personal information” as described in the definition of that term in section 2(1) of the Act.  This information relates to identifiable individuals who did particularly well or mature students who were admitted with LSAT scores and UGPAs lower than Faculty standards.  In my view, disclosure of this information would reveal information about these identifiable individuals’ educational history, in particular, their LSAT score, UGPA and FYA.  I also find that disclosure of the predicted FYA and ranges is also presumed to constitute an unjustified invasion of privacy, as the predicted FYA and ranges are based on the student’s actual LSAT score and UGPA.

As I have found that the presumption at section 21(3)(d) applies to the remaining grade information, it is not necessary for me to consider whether the factors favouring disclosure or non-disclosure at section 21(2) of the Act also apply.  Accordingly, I find that the personal information qualifies for exemption under section 21(1) of the Act. However, I will go on to consider the appellant’s argument that the public interest override at section 23 of the Act applies in the circumstances of this appeal.

PUBLIC INTEREST OVERRIDE

Section 23 states:

An exemption from disclosure of a record under sections 13, 15, 17, 18, 20, 21 and 21.1 does not apply where a compelling public interest in the disclosure of the record clearly outweighs the purpose of the exemption.

In considering whether there is a “public interest” in disclosure of the record, the first question to ask is whether there is a relationship between the record and the Act’s central purpose of shedding light on the operations of government [Order P-984].  Previous orders have stated that in order to find a compelling public interest in disclosure, the information in the record must serve the purpose of informing the citizenry about the activities of their government, adding in some way to the information the public has to make effective use of the means of expressing public opinion or to make political choices [Order P-984].

The word “compelling” has been defined in previous orders as “rousing strong interest or attention” [Order P-984].  Any public interest in non-disclosure that may exist also must be considered [Ontario Hydro v. Mitchinson, [1996] O.J. No. 4636 (Div. Ct.)].

The existence of a compelling public interest is not sufficient to trigger disclosure under section 23.  This interest must also clearly outweigh the purpose of the established exemption claim in the specific circumstances.

The appellant’s representations state:

… there is a “public interest” element that lies in scrutinizing the methods of evaluating whether the LSAT is a reliable predictor of law school “success”.  Since the LSAT carries significant weight in the admission process of the Faculty of Law, it is reasonable to scrutinize the efficacy of such tools.

The appellant submits that the public interest is “compelling” as evidenced in growing scholarly debate about whether the LSAT is a reliable predictor of law school success and the need to collect and analyze data relating to Canadian law school admissions policies (Best and the Brightest?: Canadian Law School Admissions, (199) Osgoode Hall Law Journal, (Vol. 34, No. 4) by Dawna Tong and W. Wesley Pue).

The appellant also made confidential representations in support of his position that the public interest override in section 23 should apply to any information found exempt under the personal privacy provisions of the Act.

The University argues that disclosure of the withheld portions of the correlation report would not serve the purpose of informing the citizenry about its activities, taking into account that information relating to access issues including statistics of entering GPA and LSAT scores of regular applicants is already in the public domain.  The University’s position is that it has already discharged its reporting obligations to the public.  The public information the University refers to in support of its position is its Annual Report on Student Financial Support.  The statistical data in the report supports the University’s Office of the Vice-Provost’s position that “… the University continues to be accessible to students from minority and less-advantaged socioeconomic backgrounds”. 

The University also argues that there is a public interest in the non-disclosure of the information at issue in order to protect the privacy of students belonging to the identified sub-groups, such as “mature students and those with adverse circumstances and also by the interests of the groups themselves.”

Finally, the University submits that any interest that may exist in the disclosure of the record does not outweigh the purpose of the personal privacy provisions of the Act. 

Analysis and Decision

Having reviewed the records at issue and the representations of the parties, I find that the public interest override found in section 23 does not operate to override the personal privacy exemption which I found applies to some of the grade information.

Though I accept the appellant’s submission that there is a “public interest” in the disclosure of statistical information relating to the University’s admission policies, I am not satisfied that disclosure of the remaining grade information at issue would serve the purpose of informing the public about the University’s admission activities.  In my view, disclosure of the grade information I have ordered the University to disclose to the appellant already meets this purpose.  As a result, the appellant has not satisfied me that there is a “compelling public interest” in disclosure of the remaining grade information I found constitutes “personal information” as described in the definition of that term in section 2(1) of the Act.  In any event, I find that the public interest identified by the appellant is not sufficient to outweigh the purpose of the personal privacy provisions of the Act.

Accordingly, I find that the public interest override at section 23 does not apply in the circumstances of this appeal and uphold the Ministry’s decision to not disclose this information to the appellant.

ORDER:

1.                  I order the University to disclose those portions of the record that I found does not contain the “personal information” of an identifiable individual and can be reasonably severed, by October 3, 2008 but not before September 26, 2008.  For the sake of clarity, I have highlighted the portions of the record that should not be disclosed in the copy of the record enclosed with this Order.

2.                  I uphold the University’s decision to withhold the remaining portions of the record. 

3.                  In order to verify compliance with this Order, I reserve the right to require a copy of the information disclosed by the Ministry pursuant to order provision 1 to be provided to me.

Original signed by:                                                                              August 29, 2008        

Jennifer James