As the Information and Privacy Commissioner, I conducted a written inquiry under section 56 of the Freedom of Information and Protection of Privacy Act (the Act) concerning a request for review received on November 25, 1993 under section 52. On May 2, 1993, the applicant requested access to agricultural land rental survey data in the custody of the Ministry of Agriculture, Fisheries and Food (the Ministry).

The applicant indicated in his request that he wanted the information in connection with his ongoing contract for appraisal services with the federal Department of Indian Affairs and Northern Development.

In denying the request for access to information submitted under section 4 of the Act, the Ministry cited sections 22(1), 22(3)(d), and 22(3)(f) of the Act. The Ministry initially denied the request for access on May 14, 1993 and subsequently reconfirmed this decision after the Act was proclaimed on October 4, 1993. The original request for review was made on June 5, 1993; following proclamation of the Act, the applicant submitted a second request for review to this Office on November 25, 1993.

The Portfolio Officer assigned to this case by the Office of the Information and Privacy Commissioner investigated the denial of access and concluded that the Ministry's position was appropriate and consistent with the Act. The applicant disagreed with the conclusion of the Portfolio Officer and requested a review by the Commissioner.

2. Documentation of the review process

Under sections 56(3) and (4) of the Act, each party was given an opportunity to make written representations to me. I also decided, under section 56(4), that both parties would have a further opportunity to reply to the other's representation during the week following the initial submissions. These replies were received. In reaching my decision, I have carefully considered both the representations and replies.

The Office of the Information and Privacy Commissioner provided both parties involved in the inquiry with a two-page statement of facts (plus appendices). The appendices consisted of a summary of the categories of information contained in the survey form and a sample blank survey form. The facts, the most essential of which are included in this Order, are not in dispute.

3. Issues under review at the inquiry

Under section 57(2) of the Act, the burden of proof is on the applicant to establish that disclosure of the information sought would not be an unreasonable invasion of a third party's personal privacy.

In the present case the Ministry chose not to give notice to approximately 750 third parties whose interests may be affected, since it did not intend to give access (see section 23[1.1] of the Act). Accordingly, this inquiry only involves the decision to deny access. If I were not to confirm this decision, I would refer the matter back to the Ministry in order to give the notices required under section 23 of the Act.

4. The data in dispute

The records in dispute consist of responses to an agricultural rental survey conducted by the Ministry in 1979. The eight-page survey form is titled "British Columbia Ministry of Agriculture - Lease Survey - 1979." Responses asked for and received by the Ministry consisted of the following personal data (among other items collected): 1) the landlord's identity, address, phone number, occupation, age, years of education, and market value of the buildings, machinery, and equipment owned by the landlord; 2) the tenant's name, address, phone number, work hours per week, spouse's work hours per week, age, years of education, value of farm and machinery owned, rent paid, acreage owned, leased land market value per acre, and relationship to the landlord. The survey instrument also asked for tenants' opinions about what changes to farming practices they would make if they owned the land. Further questions concerned the terms and conditions of specific leases and whether the tenant liked them.

It is my judgment that the data in dispute in this inquiry are overwhelmingly personal information, not least because they are associated on the survey forms with the names and addresses of specific landlords and tenants.

It is important to note that this survey was not a professional effort conducted by a government statistical office or trained statisticians. It was the work of the Abbotsford district office, Farm Economics Branch, of the Ministry of Agriculture. It was also not a scientific survey in that it was not directed to a formal sample of potential respondents. Thus some of the measures relating to the selection of a sample and documenting informed consent, that one would expect in a professional survey, were not followed.

According to a surviving form letter submitted to this inquiry by the Ministry, a particular regional farm economist sought assistance from various informants about names of farm managers who rented farm property anywhere in British Columbia. The purpose was a survey of existing lease agreements in order to understand their increased popularity due to the substantial financial benefits from renting.

The form letter stated that "Persons referred to us will be contacted and asked if they wish to participate. Information collected will be used to prepare a handout booklet on `Farm Leases in British Columbia.' All personal information is strictly confidential."

The sample survey form does not include a written statement about confidentiality, such as one would expect in a professional exercise, but I am satisfied from the last statement cited above that the intention of those collecting the personal information was to keep it confidential. This is a very significant consideration for me in reaching the conclusion expressed below in the Order, as is the fact that individual participation was voluntary and not mandatory.

Apparently, the survey information was used as background material for a 1982 booklet entitled Legal Aspects of Private Farm Leases (Ministry of Agriculture and Food, Victoria, B.C., 1982). However, the introduction to the booklet makes no reference to the survey or to the data obtained from the survey responses.

5. The applicant's case

In his submission to this inquiry, the applicant stated that he sought access only to "information regarding the particulars of land leased," not "the personal information of the landlord or tenant which is also included on the survey form." The problem with this point is that the applicant sought access to full copies of the completed survey forms of the Agricultural Rental Survey. My inquiry can only deal with the original request that was denied, not subsequent requests or various versions of requests that were subsequently contemplated. Furthermore, the applicant has requested review of the initial denial of access.

The applicant's second point is that he is interested in the location of a property in order to conduct an inspection. This would be an administrative use of information that could have a specific impact on the interests of individuals in contrast to a research or statistical use, from which no administrative consequences would follow for specific persons, as opposed to groups of persons represented in the aggregate as landlords or tenants in 1979.

The applicant admits, however, that as the survey responses did not include legal descriptions of the land leased, he would need the names and addresses of landlords or tenants in order to locate a specific property for inspection. Personal data are thus involved.

At a meeting on February 14, 1994, the applicant learned that a number of completed questionnaires do not contain personal addresses, leading him to ask, in his submission to this inquiry, "how confidential can information be if it cannot be identified?" The simplest answer to this query is that personal data from a survey form might be associated with specific persons even if names and addresses are removed from the forms. It would require more anonymization of such records, through the careful application of rules on disclosure avoidance, to ensure that the records in question would be truly anonymous. (See the discussion in George T. Duncan, Thomas B. Jabine, and Virginia A. de Wolf, eds., Private Lives and Public Policies: Confidentiality and Accessibility of Government Statistics [National Academy Press, Washington, D.C., 1993], chapter 6. I served, in a former capacity, as a member of the Panel on Confidentiality and Data Access that produced this report.)

The applicant raises the point that sections 16(1)(a)(i) and (iii) contemplate the avoidance of harm to relations among provincial, federal, and aboriginal governments: "It cannot be argued that the non-disclosure of material facts held by one of the parties, with denial of access to another, would most certainly dissolve the all important trust required in negotiations of native land settlements." I believe that in the context of the applicant's submission to this inquiry, he intended to use the word "can" rather than "cannot." In any event, the applicant has in fact misread this section. It provides reasons for the head of a public body to refuse to disclose information to an applicant if harm to the noted processes might occur. Thus it is a reason not to give out information rather than an argument for doing so. It is possible that alternative methods of obtaining access to the information sought might exist under the process for conducting negotiations with aboriginal governments.

The applicant further argues, in accordance with section 22(2)(d) of the Act, that disclosure of personal information would "assist in researching or validating the claims, disputes or grievances of aboriginal people." He added: "I have made my request as a professional appraiser on contract to Indian Affairs and Northern Development Canada with the knowledge that the information will assist in the settlement of land claims as contemplated as a reason for disclosure set out in Section 22(2)(d) in the Act."

The problem with this argument is that the applicant did not provide me with any evidence to support his assertion. I concluded in my first order that the application of any exception in this Act by a party holding the burden of proof required evidence that is detailed and convincing or detailed and persuasive in order for an exception to be successfully applied. This applies to the character or substance of the evidence. Once evidence is submitted, I will be in a position to weigh it against the appropriate standard of proof.

The Act provides, essentially, a civil code of fair information practices to be applied to identified public bodies. In a Part 5 inquiry under the Act, the standard of proof to be applied to the evidence is the civil standard, that is, on a balance of probabilities. This means that the party who bears the burden of proof must do so by tipping the scales in his or her favour.

The applicant has also stated that the information he seeks is about land, not persons. However, in the form in which it now exists, the information is about land and persons, making it personal data; and that is the information being sought.

The applicant also raises the fact that the information he seeks is more than fifteen years old. Since Cabinet confidences are disclosable after fifteen years, I have some sympathy for this argument as it applies to general, as opposed to personal, information. But the personal data currently included on the survey forms makes it impossible to accept this line of argument, unless the Ministry agreed to sever the personal data under section 4(2) of the Act. But this would also make it impossible for the applicant to undertake the property inspections that he wishes to undertake for appraisal purposes. Any severance practiced in a case like the present one would have to include an agreement that the recipient of the data would not try to identify specific parcels of property or specific landlords or tenants.

The applicant also raises the possibility of allowing him to review the information he seeks in order to determine if all or a portion of the data are usable. He suggests that I could rule on their release once these facts were known. There are two interrelated problems here. First, the applicant would have to sign a research agreement available under section 35 of the Act, which he states his willingness to do. However, section 35 specifically contemplates the use of such a device for research and statistical purposes only. The present applicant only wants the information in question for administrative purposes, which, in my judgment, does not make him an appropriate candidate for a research agreement. Secondly, I cannot contemplate ordering the release of personal data collected in a survey for administrative purposes, based on the facts of the current inquiry as I understand them.

It would be completely contrary to the fair information practices in Part 3 of the Act to order the release of the full records of a survey for use in an administrative function. To the best of my knowledge, this has never occurred in any Western society, especially one that benefits from having in place a data protection law like the present British Columbia Act.

6. The Ministry's case

The Ministry's submission indicates that there are 378 survey questionnaires in existence: 314 (83%) of the questionnaires list individuals as landlords; 331 (87%) of the questionnaires list individuals as tenants. Thus I concur with the Ministry's view that the records fall within the definition of personal information set forth in Schedule 1 of the Act. The Ministry relies on sections 22(1), 22(3)(d), and 22(3)(f) of the Act to conclude that disclosure of the personal information in question would be an unreasonable invasion of the personal privacy of third parties.

With respect to the applicant's case under section 22(2)(d) that disclosure of the records would assist in validating the claims, disputes, or grievances of aboriginal people, the Ministry makes the following points:

The Policy manual (C.4.13 page 17) defines validating the claims as a means of "confirming a statement or rights or contentions." To this date we [the Ministry] have learned/received nothing from the applicant that would substantiate his claim that the surveys may assist to "confirm a statement or rights or contentions". In fact, it is our opinion that the applicant is on a fishing expedition only and has no immediate need other than to help build a better understanding of leasehold values throughout the province during this time. If his purpose is not to validate or confirm anything at this stage[,] we feel that the Act requires us to give personal privacy a higher priority.

I agree with the Ministry that the applicant has not made his case on this point.

The Ministry raises the possibility that disclosure of this information might result in financial harm to third parties, contrary to section 22(2)(e). It also emphasizes the importance of the promise of confidentiality given to respondents to the 1979 survey. Regrettably, the Ministry has been unable to locate a copy of the form letter sent or given to respondents of the actual questionnaires. I note that failure to document such a communication in surveys undertaken while the Act is in force (after October 4, 1993) would be unacceptable.

I accept the Ministry's argument that failure to maintain promises of confidentiality given in the past would seriously impair the educational extension services that it is charged with providing and that require "the continued voluntary supply of information from our clients in order to meet our mandate and provide effective extension services."

7. The concept of functional separation

This case illustrates the important difference between research or statistical uses of data on one hand, and administrative uses of such data on the other. One of the fundamental premises of research and statistical work using government data, and of data protection in general, is the concept of functional separation between such uses. Using individual data for research and statistical purposes means using the information in such a manner that no direct action affecting a particular person is ever taken on the basis of the specific data, except in those rare instances where treatment is an additional, specified goal of a research activity. (Clinical trials and certain medical research activities sometimes identify treatment of selected persons as a necessary and desirable part of an approved research protocol.) Such a use can be contrasted with an administrative or regulatory use of information directly affecting an individual, as in the conferring of a health benefit. As appropriately defined in the Canadian Privacy Act, "administrative purpose" in relation to the use of personal information about an individual, means "the use of that information in a decision making process that directly affects that individual." (Privacy Act, Revised Statutes of Canada, 1985, c.P-21, s. 3.) Such an administrative use is also defined in section 28 of the British Columbia Act.

The centrality of the concept of functional separation is fully discussed in Duncan, Jabine, and de Wolf, eds., Private Lives and Public Policies: Confidentiality and Accessibility of Government Statistics, pp. 4-6, 34-5, 50, 53-4, 134, 135, and 223.

In my judgment, functional separation should be carefully observed by the heads of public bodies in British Columbia that are subject to the Act. Administrative data may be used for statistical purposes, but statistical information may not be used for administrative purposes that directly affect an individual.

8. Order

Under section 58(2)(b) of the Act, I confirm the decision of the Ministry of Agriculture, Fisheries and Food not to release to the applicant the information he requested.

February 23, 1994

David H. Flaherty
Commissioner