When the Uniform Law Conference of Canada decided, back in 1993, to address the legal effect of electronic communications, it started with the law of evidence. See 1993 Proceedings of the Uniform Law Conference of Canada, Appendix G, pages 34-35, 198 – 206. It was clear even then that more and more information intended to have legal consequences was generated, communicated and stored by electronic means. If the legal consequences were to be properly adjudicated, the information had to be capable of being put before the adjudicators.
The question was thought to be of interest both to barristers, who needed to manage, introduce and argue on the basis of this information, and to solicitors, whose clients were asking if their businesses could scan their records into a computer and destroy the paper. How the clients were going to be able to use those scanned records in disputes was a key question for the solicitors.
The first step was to commission a paper from Ken Chasse, who even then had a long record of publications on evidence and record management. His paper for the Conference in 1994 recommended that the focus of the law of evidence should shift from the individual document to the record-keeping system in which it was found. In particular, the notion of an ‘original’ document, as contemplated by the ‘best evidence’ rule, was not helpful in the electronic world, being either meaningless or irrelevant to the purpose of the rule.
The Conference set up a small working group that I chaired. The group also included a senior federal prosecutor who taught evidence law, and a lawyer who represented Canada on international delegations on the law of electronic transactions. The working group adopted Chasse’s recommendation to move to a system integrity focus, which became the core principle of the Uniform Electronic Evidence Act, ultimately approved by the Conference in 1998. The intervening period was spent working out the implications of doing so. (All of the Conference’s working papers and the Uniform Act itself are on the Conference web site.)
The Uniform Act was fairly successful, being adopted in six provinces and all the territories, as well as incorporated into the Canada Evidence Act, which makes it applicable to criminal prosecutions and federal administrative proceedings across the country.
Recently an article (published in Archivaria in 2010) was drawn to my attention that reviews the Uniform Act and finds it deeply wanting: “Electronic Records and the Law of Evidence in Canada: The Uniform Electronic Evidence Act Twelve Years Later,” by Luciana Duranti, Corinne Rogers and Anthony Sheppard of the University of British Columbia. The first two are archivists, the last a law professor. They are involved in a long-term project to review the law of evidence as it applies to digital evidence. Since the authors are still at work on the principles of the article, it may be useful to comment on them now.
Though the article is thorough and conscientious, I find it also to be wanting. It seems to me to misunderstand both the goal and the achievements of the Uniform Act and want something that the Conference did not set out to do and that may not be achievable in any event. It finds contradictions or conflicts where I still see none. At times in reading it, I found myself wondering if archivists are another profession in a parallel universe of discourse from lawyers, as I have written about engineers. The present note will touch on some of the highlights of my concerns.
The purpose of the Uniform Act
The Conference working group was aware of the risks presented by the growth of information, and thus evidence, in electronic form. It wanted to resolve some key issues before a creative counsel managed to blast holes in the current law that would cause serious problems in the courts.
The working group considered information in electronic form to be essentially documentary evidence– a point on which the authors of the article do not disagree. Such evidence is subject to three reviews before it is admitted for purposes of a legal proceeding: authentication, best evidence and hearsay. What, if anything, needed to be done to those testswhen the information was electronic? The project did not set out to revise the rules of record management or privacy or to tell people how long they should keep their records or in what form. The wish list of the authors of the article, and of the Digital Evidence project they are now engaged in, is much longer. The Conference aimed at resolving a much narrower set of issues.
As archivists, the authors find the Uniform Act’s definition of ‘electronic record’ deficient, both because of its reference to computer systems and because of its failure to distinguish between a record and a document. Despite their explanation of the difference of the latter two terms for archivists’ purposes, which I believe I understand, they do not say why the distinction may be important in law. No case has turned on it, so far as I know. Even in this note, I use the terms interchangeably, though that may not appeal to archivists.
The Conference varied its own terminology, in fact. When it adopted the Uniform Electronic Commerce Act in 1999, it used ‘electronic document’ as its key term. The main reason for the change is that ‘record’ presents challenges to translate into French. It can come out as ‘enregistrement’ or ‘inscription’, and neither really served the purpose. My guess is that translation issues caused the federal government to adopt the Uniform Electronic Evidence Act using ‘document’.
The reference to computer systems was intended to link back to language in the Criminal Code, to make the evidence rules fit with what those used to the Code would be looking for. There was no intention to limit the technology, and no such result, in my view. I am not aware of any cases where any implementation of the Uniform Act was refused because the technology did not fit its language.
The authors of the article say that the statute should provide rules on the weight given to electronic evidence. The Conference working group consciously decided that statutes never or almost never told judges how to make decisions, which is what assigningweight by statute would amount to. Once the evidence was in the door, the decision-maker would decide whether to trust it. Professor Hamish Stewart, in his 1996 paper for the Conference, described a trend in law to increasing the scope of admissibility ofevidence and an increasing reliance on the decision-makers to assign appropriate weight, but he did not support legislating this task.
Best evidence and authentication
Though the need for authentication and the best evidence rule are conceptually distinct, Canadian courts have tended not to make much of the distinction. Both tests are aimed at ensuring that documentary evidence is trustworthy.
The US Federal Rules of Evidence state expressly that authentication involves providing the basis on which the finder of fact could conclude that a document is what it purports to be. The leading Canadian textbooks agreed that this was the law here well, but there did not seem to be any Canadian cases of high authority for that rule, and it was not prescribed by statute. To make things clearer, the Uniform Act codified the rule, without changing it. (Codifying the rule does not mean, as the article seems to believe, that nothing can be said about how the rule operates.)
The best evidence rule requires that the document produced for the court be the original of the document, or some good explanation be given of why the original was not produced. The theory of this rule was clearly that it was harder to alter an original undetectably, so the content was more likely to be accurate.
The problem was that an electronic document or record either does not have a meaningful original or its copies may be completely identical, to the last bit. Thus having something that one might call the original did not add to the reliability of the information.
The working group considered recommending the abolition of the best evidence rule entirely, but was persuaded not to do so by the argument that since electronic records could be less reliable than paper ones, it was not appropriate to make it easier to get the electronic ones admitted in evidence.
In retrospect, I think that decision was wrong. The best evidence rule is meaningless for electronic records. There is no point deciding which one of completely identical and indistinguishable versions of a record is the ‘best’, for evidentiary purposes. The work – to focus on the record-keeping system – should have been done on the authentication test.
The article notes the official comment on the Uniform Act that says that the working group decided to apply its system test only once, to best evidence, rather than twice (i.e. repeating it as part of authentication). The authors say – correctly – (page 111) that the purpose of both authentication and best evidence rules are ‘interwoven to the point of being interdependent’ – but do not share my own current view that the best evidence label adds no value, only distraction.
The working group thought that for authentication, a witness would support the document, by giving the basis on which the judge could find it to be what it purported to be. But if opposing counsel – or the judge– asked ‘how do you know?’, that would bring in the principles that the rest of the Uniform Act describes.
The working group did not address its mind to the distinction between ‘real evidence’ produced by computers and documentary evidence produced from computers. The article mentions this distinction as set out by Underwood and Penner in Electronic Evidence in Canada, which has been taken up by some cases as well. A focus on the consequences of that distinction might have reinforced a preference to amend the authentication rule rather than to fuss with the obsolete best evidence rule, since authentication is needed for both real and documentary evidence, while the best evidence rule applies only to the latter.
On the other hand, it is not clear that courts that have focused on the distinction between real and other computer-generated evidence have had any difficulty in applying the version of the Uniform Act before them to the extent that it applied. Underwood and Penner warn against confusion; perhaps their warning is being duly heeded.
Hearsay is different
The third test that documentary evidence must pass to be admitted is whether it is banned as hearsay, or whether some aspect of the rules about hearsay will let it in anyway. The main ‘exception’ to the ban on hearsay in this area is the business records rule: records kept in the ordinary course of business are considered sufficiently trustworthy to be admitted even for the truth of their contents. The common-law version of this rule had a number of criteria to satisfy to achieve this status; statutory versions have tended to be much simpler.
The working group took the view, which I still believe to be the right one, that hearsay is a technology-neutral concept. A document is just as much somebody else’s knowledge, and just as much made in the ordinary course of business, whether it is on paper or in electronic form. Whether the information is on paper or in electronic form, it must be authenticated, as discussed above. Once it is authenticated as being what it purports to be, then the court can decide if it is a business record or otherwise exempt from the ban on hearsay according to the usual law.
The authors of the article want the law of evidence to set out how hearsay should work in the electronic world. This is one of the objectives of the Digital Evidence Project as well. I think this is misguided. All of the considerations of the record-management process that the authors appear to prefer for legislation on hearsay can properly be considered in authenticating the document. The hearsay analysis is a different order of question.
The Ontario Court of Justice in my view got it right inR v Mondor, 2014 OCJ 135. The court looked at s. 31.1-8 of the Canada Evidence Act on authentication of electronic records (i.e. the provisions that implement the Uniform Act federally), and held that the prosecution had complied completely. However, the evidence presented, though authentic and though satisfying the best evidence rule, was still hearsay, and without meeting the business records exemption described separately in s. 30 of the Act, the evidence was not admissible. In other words, hearsay was not an e-evidence issue but demanded a standard media-neutral hearsay analysis. Sections 31.1ff do not and are not intended to provide an exception to the hearsay rule. The decision reviews related cases and the Underwood and Penner text mentioned above concerning real evidence.
I might mention that this question – whether the nature of hearsay and the business records exception are changed by the use of electronic records – is one on which Ken Chasse and I have differed for the past twenty years. He has not persuaded me on the point, and I have not had more influence on his conclusions. He does agree with me, though – or came first to the conclusion – that the best evidence rule should go. See for example K. Chasse, “Electronic Records as Documentary Evidence”, (2007), 6 Cdn Jl Law and Technology 141, 156.
The authors of the article are keen supporters of the Canadian General Standards Board’s Standard on Electronic Records as Documentary Evidence, and say at one point that compliance with that Standard should be mandatory for admissibility. (My overview of the CGSB Standard is here.) The Conference’s working group was asked by a group of records managers to provide expressly for the use of the then-current CGSB Standard, on microfilming and scanning. The working group deliberately did not do so, on the ground that the Standard was more demanding than necessary for many cases.
It was common in the 1990s for businesses dealing with each other electronically to form ‘trading partner agreements’ that set out, among other things, how they would transmit, acknowledge and sign the electronic communications. These agreements tended to say that the parties would not dispute the admissibility of documents treated as agreed. There was some difference of views whether such agreements could bind courts, as evidence law is at least in part a matter of public policy.
It was pretty clearly the law before the Uniform Act that courts could look at industry standards in evaluating conduct. The reason that section 6 on standards was put into the Uniform Act was to try to ensure that such standards would be read broadly, toinclude if need be these private agreements on how electronic evidence would be handled. The purpose was to satisfy the parties’ reasonable expectations, not to please archivists.
A case of non-statute-based authentication at work, in the absence of standards, is Bell Expressvu Limited Partnership v. Rodgers (Tomico Industries), 2007 CanLII 50595 (ON SC). Though no reference was made to federal or provincial evidence laws, the plaintiff was held to have duly authenticated its evidence, despite the vulnerability of electronic documents to falsification. (paragraph 9 point 1) The court also noted that there were ‘no generally acknowledged practices for preserving, documenting or securing electronic evidence’, but on review, even the defendant’s expert described the plaintiff’s practice in this case as ‘picture perfect.’ (paragraph 9 point 9) In the light of such decisions, it seems imprudent to require adherence to a particular standard, however respectable.
To see how a case under the CGSB standard might look, one can review R. v. Oler, 2014 ABPC 130. The court held (wrongly, in my view) that the Alberta Evidence Act (which implements the Uniform Act) required compliance with the CGSB Standard, but the proponent of the electronic evidence persuasively qualified its witness and the witness showed the court that her employer had complied, and the evidence was admitted.
It can make sense to insist on the Standard for some purposes, as does the Canada Revenue Agency for the maintenance of scanned business records for excise tax purposes. (paragraph 15) However, for maintaining records that were originally electronic, the CRA sets out its own requirements (paragraph 17ff), and suggests referring to the (then pending) Standard as a help (paragraph 22). In my view, and that of the Conference working group, it would be too demanding to legislate the routine exclusion of relevant evidence because the Standard is not satisfied.
The Standard was, by the way, developed with a view to providing a way to satisfy the requirements of the Uniform Act, as well as to be a generally good way to keep electronic records. It is, after all, a standard on e-records as documentary evidence. So it would be wrong to treat the two documents as working at cross-purposes or being inconsistent with one another.
I have noted several of the main instances in which I think the authors of the Archivaria article have misconceived the Uniform Act and sometimes the law of evidence as well. Their approach to law reform in this field so heavily reliant on the common law seems to me unduly hopeful as well.
The article calls for ‘continuous and sustained updating’ ‘to keep law current with technological change, especially in the areas of evidence and procedure’ (page 100). Such a process may appeal to academics but would be very hard on record managers and on lawyers and judges trying to advise on or decide what records would be admissible from time to time and how. Attempts to legislate for particular technologies tend to fail, too, because legislators do not act as fast as technology evolves. Even standardsare slow work; a revision to the CGSB Standard has been under way for some years.
The authors lament that ‘the limitations of the [Uniform] Act have resulted in the courts’ continuing reliance on traditional, narrow common law rules rather than broader, new statutory rules.’ Besides rules on the supposed link between hearsay and e-records, they call for updated or wholly new rules on search and seizure of electronic records in both civil and criminal cases, the protection of privacy, retention and preservation duties, and spoliation of electronic records.(page 104)
What they do not actually demonstrate in the article is that people have trouble understanding the rules in the Uniform Act or that the rules lead to unjust results. Whether ‘broader, new statutory rules’ are desirable for the other areas they mention, which the Uniform Act was not aiming at, needs cautious review. Some, like privacy, are new under that name to the law of evidence but are beginning to make their appearance. Others, like spoliation, may be better developed for technology by common law principles.
Meanwhile, the Uniform Act may not work – for them– in theory, but I think a good case can be made for its working – for the rest of us – in practice.