Uniform Acts Drafting Conventions
1994 Charlottetown PE
Civil Section Documents - Electronic Evidence: Computer Produced Records in Court Proceedings
Records generated by or stored in a computer are increasingly used in the public and private sectors. Both common law and statute provide rules for the use of documentary evidence in court. This paper examines whether the existing rules for documentary evidence apply readily to computer-produced records, or whether those rules should be changed or supplemented.
After a brief look at some very fundamental principles of evidence law and of computer records, the paper looks at the common law and the existing statutes that deal with business records. Less attention is paid to public documents, though the principles of the discussion may be relevant to them as well. It concludes that the current law is unsatisfactory in principle.
A number of statutes elsewhere do deal expressly with computer records in evidence, as does the Civil Code of Quebec. These are set out in some detail. Appendices contain a closer analysis of the provisions of Canadian and other legislation.
The Uniform Law Conference dealt with computer records when it created the Uniform Evidence Act. The paper reviews the debates at that time, including a detailed appendix on the subject. At that time the Conference elected not to impose special rules on computer-produced evidence.
Several options face the Conference now. It could do nothing, on the ground that the current law works in practice, even if uncertain in principle. It could adopt a short statement on the evidentiary status of electronic records. It could insert a list of criteria for the admissibility or weight of these records. It could revamp all the documentary evidence sections of the Uniform Evidence Act to conform with the proposed rules for computer records. It may wish to consider whether new rules should apply differently to bank, business and public records.
Electronic imaging is a related subject, and a new technology since the Uniform Evidence Act was adopted in 1981. It straddles the rules for microfilm and those for business records. Arguably its status should be clarified by statute.
It appears desirable that the law should enforce private agreements setting standards for evidence in litigation between the parties. This will give some comfort to enterprises doing business electronically under the terms of a trading partner agreement.
Any new rules should promote the admission of reliable evidence and be clear, fair and workable.
Computer-Produced Records in Court Proceedings
- some principles of documentary evidence
- electronic evidence
- some legal guideposts
- the common law: a brief reminder
- statutory rules
- some statutory shortcomings
- the case for legislation
- a qualification
Options for harmonized rules:
- Do nothing
- Adopt a limited, facilitating rule for electronic evidence
- Add a special list of factors touching electronic evidence
- Rewrite all the documentary evidence rules on principles consistent with the rules for electronic records
- banking documents
- public documents
- microfilm rules
- civil and criminal evidence
APPENDIX B - Statutes on Electronic Records
APPENDIX C - Law Reform in Canada
APPENDIX D - Imaging and the Law
 As we rely more and more heavily on computers in all aspects of our life and work, the products of computers are more and more widely relevant to litigation to sort out our legal relations. This paper addresses the use of computer records as evidence in such court proceedings, both civil and criminal.
 Electronic records are documents, and the rules of documentary evidence apply to them, though not always easily. The paper starts with a very basic overview of the law on documentary evidence. It then discusses computer technology and how the technology challenges the application of the usual rules for documents.
 The law has not been completely silent on electronic records. In fact, these records are regularly admitted in both criminal and civil proceedings across Canada. The paper looks at what has happened so far, in the courts and in statutes here and abroad. The adequacy and fairness of the present law are evaluated in some detail.
 We then look at options for law reform in this area and the case for harmonization of a reformed legal regime across the country. What the Uniform Law Conference has already thought on this subject is obviously relevant to this part of the discussion.
 The paper concludes with recommendations for uniform action in the area.
some principles of documentary evidence
 The basic rule of evidence is that it must be the (sworn) oral account of facts of which the witness has personal knowledge. The witness is available in person in court to have his or her account tested by cross-examination.
 Information of which the witness does not have personal knowledge is "hearsay", that which the witness heard someone say (or which he/she learned in some other way without experiencing it first hand.) (The witness's opinions make up a distinct class of evidence, which is sometimes relevant to electronic records too.)
 Information in documents or other records is hearsay, since the person presenting the information as evidence in court does not have personal knowledge of that information. If the witness had the personal knowledge, it would not be necessary to use the documents to prove the facts in them.
 The traditional rule is that hearsay evidence may not be admitted. This general rule has been eroded substantially in recent years by a series of exceptions. An argument can even be made that there is no longer a ban on hearsay evidence. There is simply a rule that hearsay evidence must be demonstrated to be reliable and its admission necessary to the proper adjudication of the case. For the purposes of this paper, we do not need to decide whether the present admissibility of hearsay evidence is a matter of exception or a matter of rule.
 Two subsidiary rules of evidence apply to documentary evidence. The first is the "best evidence" rule: to prove something in court, you must use the best evidence that can be produced. "Best" means closest to direct sworn oral evidence. This produces a hierarchy of documents judged on such criteria as when they were made, by whom they were made, their status as "original" documents or copies, and the like.
 The second subsidiary rule is an "exception" to the rule barring hearsay evidence. Courts have long agreed that documents should be admitted to prove the information they contain. The common law developed criteria for admissibility, such as that the documents were produced at the same time as the events they recorded; that they were produced in the ordinary course of the business of the party creating them; and that they were used and relied on by the creator in his/her/its business.
 These rules have been replaced or overlaid with statutory rules in the federal and provincial Evidence Acts. Many Canadian statutes classify documentary evidence depending on its origin and its form. Different rules apply to each. Government and other public documents are treated in one way, business records in another, business records that are also banking records in a third.
 The courts have interpreted these sections in inconsistent ways, sometimes appearing to apply "bank" standards to other documents, or "other" standards to bank documents. They have also used the common law tests of contemporaneity and the like in applying the statutory tests. This may be in part because they have not distinguished in every case when they were deciding admissibility and when they were judging the weight of the evidence.
 These criteria are aspects of the one of the tests for the admission of hearsay evidence: reliability. The other test is necessity. One of the main reasons admitting a document is necessary is because the oral evidence is not available. The person who has direct knowledge of what is reported in the document is not available, or the information is such a routine bit of data among much else that no one could reasonably be expected to recall learning or receiving the particular information to be proved.
 The "necessity" test has given rise to some confusion because the term is also used to justify using a copy of a document instead of an original. However, this justification is an aspect of the best evidence rule, not the hearsay rule. Applied to documents, the best evidence rule means that an original document is the preferred evidence. Sometimes this too has been altered by statute. Some statutes provide that a photograph of a cheque is admissible without proving how it came to be produced or used.
 Additional provisions have been made for other photographic and microfilmed documents. Some of them show their origins by requiring that the original (paper) documents must be retained for a period of years as well as the microfilm. (Presumably this allows parties to test the reliability of the microfilm by looking at the originals, even though the microfilm is admissible as is.)
 As a result, the law is somewhat confusing in theory. Not all the contentious issues have been mentioned here. However, documentary evidence is regularly used without serious problems of principle.
 Electronic evidence is a version of documentary evidence. As a result, both of the above tests for the use of documentary evidence apply to it: best evidence rule and hearsay rule. How this is done and how it should change, if at all, is the subject of this paper.
 Electronic evidence is information that is recorded electronically. It may be created electronically or simply stored electronically. It may be on paper at one or more stages of its "life" and electronic at others, such as a fax (though faxes are generally treated as copies of paper records rather than as computer records). It may exist in more than one place at a time - in two computers, for example. An incomplete sample of electronic records would include those in or created by single computers, computer-to-computer communications, with or without intermediaries and with or without transformation of the messages at both ends, magnetic strips on plastic cards, microcomputers on plastic cards (smart cards), electronic mail, bulletin boards and international communications networks.
 The different ways in which computers are used to create, store and retrieve business records involve either communications between computers and humans or computer to computer communications, the latter being merely a variety of the former with the intervention of a second computer or multiple computers. Therefore it is not necessary to describe all of the different ways in which computers are used to create, store and retrieve business records. From the point of view of the law of evidence the different applications of computer technology will not affect the type of evidentiary provisions necessary to accommodate them, if those provisions concern operations common to all computer applications.
 Such operations are, for example, the sources of data and information used in databases, the entry of such data and information, business reliance upon such databases, and software reliability. General or specific references to such operations placed in the business record provisions for example, would be applicable to all computer-produced business records. On the other hand, the relation between computer technology and microfilm could change evidentiary provisions because traditional microfilming has its own provisions in the Evidence Acts in Canada.
 Businesses and their lawyers express considerable interest in EDI (electronic data interchange). This can be defined as computer to computer transmission of data in structured forms, i.e. paperless trading. It does not require special treatment apart from Evidence Act provisions that apply to other computer-produced records. EDI's special legal issues concern contract law, not evidence law, e.g. trading partner agreements containing terms as to establishing the communications network, allocating costs and risks, determining security procedures, and procedures for verifying content, timing and authenticity of messages. They might also contain evidentiary provisions for settling disputes but they do not require intervention by an evidence statute. The validity of a private code of evidence might be addressed in a statute. More on this later in the paper.
 From an evidentiary point of view EDI is simply an example of a computer used to receive externally created records and data and therefore it does not provide anything unique. Any specific factors recommended for proof could apply equally to EDI systems. Those factors, particularly sources of data and information and entry and verification procedures would apply equally to proof of the handling of data and information between computers as well as between people and computers.
 A different example is electronic imaging. The word "imaging" is commonly used in the information and image management industry itself to mean electronic imaging, which is the capture of exact images or pictures of documents onto optical or magnetic disk by means of an image scanner. It is expected to replace microfilming before the next century. The electronic records so scanned become part of a computer memory. Technically they may be able to be altered once in the memory in undetectable ways, depending on how they are recorded. As a result, those interested in using imaging technology have prescribed standards for handling the information to increase the security of the information. Appendix D of this paper discusses the legal impact of imaging in greater detail. For the moment, we need only to note that it straddles the common law and statutory rules relating to microfilm records and business records. Since these two are not consistent, some new statutory rule on imaging itself is desirable.
some legal guideposts
 With this much by way of background, we can now look more closely at the legal issues posed by electronic evidence. Here are some of the key points on which the discussion will turn.
* Admissibility and weight: Should the electronic record be allowed into the courtroom discussion at all? If so, what factors are relevant in determining its effect? Some statutory rules provide that some features of the production of a record may not affect admissibility. Some may affect both admissibility and weight, at which point the question becomes one of clarity of legal rule: how can one manage one's records in a way to ensure their best use in litigation, or how can one challenge the use of records produced by the other side?
* Statute and common law: The common law rules for documents were detailed and narrow. Statutes have generally been more flexible and broader. However, they have also been vague enough that courts have reverted to the common law, or created a new common law of statutory reading, to interpret them. Electronic records challenge the vagueness of the statutes even more thoroughly. Should new statutory provisions pick up some of the common law standards and apply them expressly to electronic records, or define the new rules in more detail by some other means?
* Types of document: Is the classification of rules by type of document (business, bank, government) adequate for electronic records, or does the electronic nature of the record unite the statutory classes so similar rules should apply to all?
* Criminal and civil proceedings: The current law, both common and statutory, does not distinguish between evidentiary rules in different types of proceeding, though of course the Canada Evidence Act applies largely to criminal actions and the provincial statutes to civil. Is there any case for deliberate variation?
* Role of consent: Is there anything in the nature of evidence that would prevent private parties from setting out by contract what criteria will be used for the admission and the weighing of evidence in litigation between themselves? Is the only concern one of equality of bargaining power in creating such a contract?
 Rules of law on this topic should seek to achieve three ends: accurate evidence of reliable records; fairness between proponent and opponent of the evidence; and workability in practice.
 Three approaches are possible in reforming and harmonizing the law in this area.
1. i) to add special subsections for computer-produced records to the existing business record provisions and for imaged records to the microfilm provisions:
* the existing provisions seem to be adequate for traditional precomputer business records and therefore do not have to be disturbed for those records. However most current records are now computer-produced so special rules may be useful for them.
2. to re-write those provisions into an single integrated set of provisions having common definitional, procedural and other support provisions (as the Uniform Law Conference, among others, proposed in the early 1980s; see discussion below under the heading, "Law Reform in Canada"):
* this would produce the same legislation for all types of business (and other) records, rather than having separate provisions for computer-produced records and imaged records. If imaging becomes widely used as expected, the two sets of provisions will often have to be used together. Therefore they should be integrated for efficiency so as to reflect that reality.
3. to do nothing and let the existing business record provisions deal with computer-produced records as best they may:
* computer-produced records are being admitted under the existing provisions without the creation of court decisions or new statutes that could inhibit their admissibility or weight.
 To determine the merit of each of these three approaches to legislative amendment, the current law for each of the different types of business record provisions has to be briefly analyzed.
the common law: a brief reminder
 The common law made an exception to the rule against hearsay for business records because of "circumstantial guarantees of reliability", in Wigmore's phrase. Those guarantees were:
- records made contemporaneously with the events recorded;
- by persons with direct personal knowledge;
- during the course of business;
- by a person under a duty to make the record;
- without any motive to misrepresent;
- who is since deceased.
 The common law rule was restated in Canada in Ayres v. Venner 7 It held that hospital records made contemporaneously by someone having personal knowledge of the matters then being recorded and under a duty to make the entry or record should be received in evidence as prima facie proof of the facts stated therein. It has been stated to apply to business records in general.Pacific Railway Company v. City of Calgary,  4 W.W.R. 241 (Alta.CA.). It clearly can still be used, even though many of its rules have been supplemented or often even made unnecessary by statute.
 In addition, the best evidence rule generally required the documents to be originals.
 The evidence statutes of all the jurisdictions in Canada contain some rules about documentary evidence, most of it not specifically designed to apply to computer records. (The new Civil Code of Quebec is an exception. It is discussed below in the section on legislation dealing with electronic records. See page 22.) A detailed examination of the statutory regimes appears as Appendix A.
 As a very general rule, we can say that most of the Canadian statutes focus on the circumstances in which the document was made and the fact that it was used in the usual and ordinary course of business. Some statutes expressly require that it was made reasonably contemporaneously with the facts it describes. Several statutes provide that the circumstances of making affect the weight of the document but not its admissibility.
 It should be noted that the best evidence rule will continue to apply to the documents admitted by statute, so that the original document will have to be produced if it is available. Statutory exceptions are confined to provisions on banking documents and some microfilmed records.
 It is fair to say that little uniformity exists, and thus little certainty for those who carry on business across provincial or territorial lines - and little enough of the latter for those within a single jurisdiction either. The necessary evidence to be marshalled for court is uncertain, as are the standards to be applied. As a result, the common law principles have been invoked as a source of interpretation, though selectively and not in every case.
 Even with these rules, the treatment of electronic evidence is not clear. In particular, the present law lacks a list of specific requirements for admissibility. The following statement from R. v. McMullen (Ont. C.A., 1979), is the closest the law has come to providing a specific test for the admissibility of computer-produced records:
I accept that the demonstration of reliability of computer evidence is a more complex process than proving the reliability of written records. I further accept that as a matter of principle a Court should carefully scrutinize the foundation put before it to support a finding of reliability, as a condition of admissibility (see McCormick's Handbook on the Law of Evidence, 2nd ed. (1972), p.734), and that the admission procedures in s. 30 [of the Canada Evidence Act] are more finetuned than that in s. 29. However, this does not mean that s. 29(2) is not adequate to the task. The four conditions precedent provided for therein, the last one being that the copy of the entry offered in evidence is a true copy of what is in the record, have to be proven to the satisfaction of the trial Judge. The nature and quality of the evidence put before the Court has to reflect the facts of the complete record-keeping process - in the case of computer records, the procedures and processes relating to the input of entries, storage of information and its retrieval and presentation: see Transport Indemnity Co. v. Seib (1965), 132 N.W.(2d) 871; King v. State ex Rel Murdock Acceptance Corp. (1969) 222 So.(2d) 393, and "Note, Evidentiary Problems and Computer Records", 5 Rutgers J. Computer Law (1976) p.355, et seq. If such evidence be beyond the ken of the manager, accountant or the officer responsible for the records (R. v. McGrayne, Ontario Court of Appeal, March 14, 1979 [since reported 46 C.C.C.(2d) 63] then a failure to comply with s.29(2) must result and the printout evidence would be inadmissible.
 This statement was refined in R. v. Bell and Bruce (Ont. C.A., 1982) and the concept of "the original" was refined to be more compatible with computer-produced records:
McMullen is authority for the proposition that information stored in a computer is capable of being a "record kept in a financial institution", and that the computer print-out is capable of being a copy of that record, notwithstanding its change in form. It is not authority for the proposition that the stored information is the only record, or that a computer print-out is only a copy of that record.
Because of the rapidly changing nature of the technology, it would be impossible to lay down general rules to govern every case. It is always a question of fact whether any recorded information (in whatever form) is a "record kept in any financial institution", but I think the following general propositions have so far emerged:
1. A record may be in any, even an illegible form.
2. The form in which information is recorded may change from time to time, and the new form is equally a "record" of that kind of information.
3. A record may be a compilation or collation of other records.
4. It must have been produced for the bank's purposes as a reference source, or as part of its internal audit system and, at the relevant time must be kept for that purpose.
Before computers were used by banks, a teller's journal was the original record. The entries in that journal were posted to a ledger, and that became a second record. I have no doubt that the ledgers of all accounts in a branch were collated so as to produce a ledger for the branch, and that became a record. So it makes no difference that the original information changes form, or becomes absorbed in some larger record. The authenticity of the record as evidence is sufficiently guaranteed by compliance with s-s. (2) of s. 29.
 We see from these quotations that the courts want to look at foundation evidence, i.e. testimony about the records that establishes the facts on which admissibility may be based. The key question is the nature of the foundation evidence that is needed. It is significant that both these cases dealt with bank records, which have traditionally been received more readily into evidence than documents from other kinds of businesses. To the limited extent that they do set a standard, it is arguable that the standard is too liberal to apply universally.
some statutory shortcomings
 In short, the law as to the admissibility and weight of business records appears to be based upon three concepts, two of which are without fixed definition and the third of which needs to be revised for computer-produced records. The undefined concepts are "the usual and ordinary course of business" and "the circumstances of the making of the record." They are in most of the evidence legislation in Canada. The third is the concept of "original" document.
 The absence of fixed definitions of usual course of business and circumstances of making of the record gives the courts complete flexibility in applying them, but leaves litigants and the business community uncertain as to what is required to prove business records as admissible and credible evidence. The courts do not have adequate principles by which to judge the "usual and ordinary course of business", and the "circumstances of the making" of computerized record-keeping.
 A surprising number of questions about documentary evidence remain unanswered, even by the combination of statute and common law. Consider these examples:
Whether the present statutory language requires that admissible records need only be made by a person under a "business duty' to make such records, or whether the supplier of the information recorded, as well as the maker of the record must have been acting pursuant to such "business duties".
Whether it is sufficient if the making of the record was part of the ordinary routine of the business, or whether not only the making of the record but also the events being recorded must be part of the business routine.
Whether contemporaneity between the making of a record and the events recorded as part of the "usual and ordinary course of business" must always be considered.
Whether records are inadmissible because of the interest or bias of the maker of the records, or whether such a requirement is not to be read into the statute.
 These "hearsay" questions could be resolved by statute to allow the business records exemption to be compatible with computer-produced business records. Whether one imports and adapts some of the common law to help make the current statutes compatible, or returns to first principles, is less important than the result.
 The third unsatisfactory concept in the present law is that of "the original", i.e proof of a record requires proof of the original record, an acceptable original being one made at or near the time of the events it records. This is the best evidence rule, rather than a hearsay rule as were the previous two. The best evidence rule states that the absence or alteration of the original must be adequately explained or proof will fail. The concept of "the original" provides some compensation for the vagueness of the "usual and ordinary course of business" concept.
 However, computerized record-keeping does not produce such originals unless the computer is being used merely as a storage device for records of original entry. A computer printout is usually created at the end of the information creation and handling process and not at its beginning as is the traditional "original". Designating the printout as the "original" as was done in Bell and Bruce is artificial. Arguably it should not displace the need for proof of the procedures and sources of information that went into creating the record.
 On the contrary, the absence of a traditional "original" in computerized record-keeping should be compensated for by a more detailed analysis of the usual and ordinary course of business and the circumstances of the making of business records. In other words, the loss of the traditional "original" evidence, being a record made at the time of or near the event recorded, in favour of a printout made at the other end of the information handling process, makes more necessary some proof of the system that went into making that printout.
 In short, the law should move from "original" to "system", that is, from a dependence upon proof of the integrity of the original business document to a dependence upon proof of the integrity of the record-keeping system. This means that the best evidence rule loses most or all of its application in this field, since the same factors that will be relevant to the hearsay exception will affect the use of the evidence as equivalent to an "original".
the case for legislation
 We have seen that the current statutes provide several different tests of admissibility and weight of computer-generated records, and the key terms are often either not defined or defined in ways that do not work well for electronic documents. For a number of reasons, it is arguable that these divergences and ambiguities will not work themselves out through the case law.
 The judicial process of case-by-case decision-making is a very poor process by which to develop binding, uniform principles for records management and data-processing. A statute that requires evidence on key points of computerized record-keeping would prevent a haphazard development of case law principle in the courts. The statute could elicit evidence on the main features of a computerized record-keeping system; the courts can then analyze it in detail. An individual case is a good vehicle for establishing a principle to govern a particular set of facts. But it is a poor method for establishing a major, dominant principle that is to govern many cases, unless that individual court case has been preceded by many court decisions that it can draw experience from in drafting that broad, dominant statement of principle. There does not exist a large volume of court decisions that have analyzed computerized record-keeping.
 It appears that it will take decades for Canada to develop a sufficient body of case law from its courts. For example, the present business record provisions have been in the evidence statutes for more than 25 years but they have produced very little analysis of what the courts should look for by way of the "usual and ordinary course of business", and the required "circumstances of the making of a record". The development of computer technology has greatly accelerated, and with it the pressure from the business community for legislation that makes clear how the new technology will produce admissible business records.
 A number of other jurisdictions have legislated on electronic evidence. We may get some guidance from those precedents for appropriate action for Canada. Amended evidence statutes could then "occupy the field" by making specific rules that would clearly pre-empt common law glosses on existing statutes. This is arguably more manageable than the present collection of unresolved issues.
 The issues and criteria by which they may be judged were listed previously, at paragraphs 25 and 26. The essential questions may be these:
* Should the admissibility of computer printouts into evidence requires as a condition precedent a detailed examination of the computerized record-keeping system that produced them, and if so what features should be proved?
* When is a computer printout to be treated as an original record and when is it to be treated as a copy, or does this question still make sense for electronic records?
 Evidence that does not meet the statutory test for documentary evidence - business or other records - may nevertheless be admissible for some other reason. First, the common law rules still apply, and they may allow some material in that would not fall under the statute. Second, other exceptions to the hearsay rule may apply. The evidence may be part of the res gestae,or be adduced for some other reason than to prove the truth of its contents. Third, some kinds of computer-generated records may be expert opinion rather than evidence of facts, and admitted under the rules for opinion evidence.
Legislation dealing with electronic records
 Specialized legislation for computer-produced records from other countries can help to determine whether similar provisions should be enacted here. Consider the following legislative succession. First came legislation from England whose requirements for admissibility were, as stated here in point-form:
The Civil Evidence Act 1968, s. 5(2) (U.K.)
(a) regular use of the computer for activities regularly carried on;
(b) computer regularly supplied with information of the kind in the statement;
(c) computer operating properly, or defective operation did not affect production or accuracy;
(d) the information is derived from that supplied to the computer in the ordinary course of those activities.
 Next, in 1972 the South Australia Evidence Act added a specialized computer printout provision whose admissibility requirements were, in point form:
(a) computer correctly programmed and regularly used to produce output of the kind tendered in evidence;
(b) computer output produced from data prepared from information, "that would normally be acceptable in a court of law as evidence of the statements ... in the output";
(c) no reasonable cause to suspect departure from the system or error in the preparation of the data;
(d) from input to output, the computer was not subject to a malfunction affecting accuracy;
(e) no alterations to mechanisms or processes of the computer that might affect accuracy;
(f) records have been kept by a responsible person of alterations to the mechanisms and processes of the computer;
(g) accuracy or validity of output not adversely affected by improper procedure or inadequate safeguards.
 Both of the above lists establish an important safeguard or guarantee of accuracy in requiring that the data or information upon which the tendered printout is based be such as is regularly fed to the computer as part of the regular activities of the organization. This requirement is comparable to the business document requirement of records "made in the usual and ordinary course of business." The desired result is records or printouts prepared according to established procedures and pursuant to established business duties. The Civil Evidence Act 1968 is more direct and therefore clearer in establishing this "business as usual" requirement.
 The importance of such a requirement would be given further emphasis if another of its underlying reasons were expressly stated - reliance upon such records in business decision-making. The South Australian criteria are superior because they direct attention to the computer program used. But both direct too much attention to the mechanical fitness of the computer. Such concern is more appropriate to the conditions of admissibility of the evidence produced by breathalyzer machines and radar devices. Since these provisions were enacted, it has become clear that intentional falsification and negligence by human operators is almost the totality of the threat to computer printout accuracy, and that computer mechanical fitness is a minuscule source of inaccuracy. The admissibility criteria laid down by the courts or the legislatures should be designed accordingly.
 For similar reasons a contemporary law journal draft provision placed emphasis in the required proof upon: input procedures; reliance upon the database in business decision-making; and upon the computer program. These criteria focus more closely upon the human parts of a computerized record-keeping system, where the errors and falsehoods are likely to occur, instead of upon the mechanical and electronic parts of the system where they rarely occur. They give much more guidance to a court, to a lawyer preparing a case, and to a records manager attempting to determine what the law requires of records keepers, than do such vague phrases as, "the usual and ordinary course of business", and, "the circumstances of the making of the record."
 Next came the South African Computer Evidence Act 1983. It provides for an authenticating affidavit from someone knowledgeable about the computer system and the record to be adduced, certifying to the proper operation of the computer.
 The effect of such a legislative approach to the foundation evidence of computer printout admissibility is to place upon the adducing party an onus to demonstrate the reliability of the record-keeping system, instead of placing an onus upon the opposing party to disprove reliability once the adducing party has adduced some evidence of records made in the "usual and ordinary course of business." Such systems can be too complicated to justify casting upon an opposing party a burden to disprove reliability. The sources of evidence for proving reliability are within the custody of the adducing party, therefore it is with that party that the burden of making out proof should lie.
 The witness need not necessarily have personal knowledge of the basic data as entered into the system, nor personal knowledge of the actual physical operation of the data processing equipment, so long as he or she is generally familiar with and accountable for the methods employed by the company in processing the business records. Indeed, the courts seem to be reluctant to require the proponent of computer evidence to call more than one foundational witness.
 The purpose in setting out such lists of points is this: specifying detailed criteria greatly increases the probability that the witness used to introduce computer records will be a person with detailed knowledge of the records system as a whole, i.e. detailed criteria make necessary the use of witnesses having supervisory responsibility over the record-keeping system that produced the records that are to become the evidence.
 By thus making more necessary the use of witnesses who are held accountable for the record-keeping system, the business record provisions of the evidence statutes that apply to computers can more closely approximate rules requiring expert evidence. Experts, like computer printouts, represent information systems. They are allowed to give opinion evidence and base that evidence upon hearsay because of the integrity of information systems they use. It is not sufficient that the expert witness testify that he obtained the evidence he gives in the "usual and ordinary course" of his business. He puts forth his professional responsibility and opinion in certifying the accuracy of the evidence he gives.R. v. George,(1993), 14 Alta.L.R.(3d) 106 (Prov. Ct.). (The same would be true of paper documents; those created with the litigation in mind would not be admissible under the documentary evidence rules.)
 The dominant test in Canada as to the admissibility of computer printouts is still the McMullen rule: "the nature and quality of the evidence put before the Court has to reflect the facts of the complete record-keeping process." In order to argue which "facts" and how "complete" the foundation evidence for admissibility must be, guidance can be gained from legislation such as the South African and the U.K. Acts because they use that kind of detail in their legislative language.
 A more detailed description was contained in the Police and Criminal Evidence Act, 1984 (U.K.) Section 69 of that Act provides for certificate evidence on the operation of the computer, the contents being spelled out by rules where necessary. It is an offence to make a false statement in such a certificate. The section also contains a discussion of the weight attached to the evidence.
 There have been a number of decisions in relation to s. 69 but none provides commentary that would be helpful here, such as commentary on the need for legislative reform or as to the type of generalized or specialized provision desired.R. v. Minors,  1 W.L.R. 441,  2 All E.R. 208 (C.A.); R.v. Neville,  Crim. L.R. 288 (C.A.); R. v. Spiby (1990), 91 Cr.App.R. 86 (CA.); R. v. Burke, Crim.L.R. 401 (H.C.); R. v. Mather,  Crim.L.R. 285 (C.A.C.D.); R. v. Robson, Mitchell and Richards,  Crim.L.R. 362 (H.C.).
The same can be said of the case law in relation to the other computer provisions referred to above. Case law has not been very helpful in the development of the principles of admissibility and weight applicable to computer printouts in particular and business records in general.
 The above pieces of legislation appear to accomplish the following effects:
1. They shift the onus of demonstrating the reliability of the record-keeping system which produced the computer printouts sought to be adduced onto the proponent of their admissibility, thus preventing an onus of adducing evidence of unreliability being placed upon the opponent of admissibility merely because the proponent has presented some superficial evidence that the printouts were created in the usual and ordinary course of business, as is the case under our present evidence statutes.
2. They establish criteria that are compatible with computerized record-keeping and replace those of the pre-existing statutory and common law business record exceptions to the hearsay rule that were not.
3. They establish the criteria of admissibility for computer printouts in an authoritative document so as to "occupy the field", thus preventing the caselaw from developing conflicting admissibility criteria or developing criteria only slowly and in a fragmented fashion.
4. They force a meaningful accountability into the foundation evidence, i.e. they have the effect of requiring that the person who supervises the record-keeping or data-processing system that produced the printout that is taken to court, be the witness who supplies the testimony or swears the affidavit that is to be used to gain admissibility and credibility for that printout.
 In the United States the Federal Rules of Evidence came into force on July 1, 1975. Their provisions are spelled out in Appendix B. Most states have adopted the Federal Rules or the very similar Uniform Rules of Evidence of 1973. They are not specialized computer provisions although they do make specific reference to computer-produced records by means of phrases such as "data compilation" and "electronic recording".
 The case law has produced very little analysis that is helpful here. It produces a repetition of the requirements of Rule 803(6), reproduced in the Appendix. An overview of the main themes is found in the same Appendix. That Rule refers to the trustworthiness of records in relation to "source of information" and the "method or circumstances of preparation", but without specifying what particular methods or circumstances a court should be considering, and without assigning an onus of proof. In brief, it shares some problems with its Canadian counterparts.
 Regardless of the requirements, commentators have found U.S. courts to be liberal in granting admissibility:
 It is important to note that the determination of whether, in all circumstances, the records have sufficient reliability to warrant their receipt into evidence is left to the sound discretion of the trial judge. Manycourts take a generous view of Rule 803(6) and construe it to favour admission of a document into evidence rather than exclusion if the document has any probative value at all. The trier of fact can usually identify self-serving or untrustworthy records and discount the weight of the evidence accordingly. 
 Under the Rules, a trial court rarely excludes an offer of computerized business records or reports not specially prepared for trial. Moreover, courts of appeal almost always uphold a lower court's finding of proper foundation for computerized business records, even when the lower court's finding is questionable. Appellate court opinions offer two rationales for their rubber stamps of approval. First, they grant trial judges broad discretion in admitting evidence. Thus, the party opposing admission must carry the formidable burden of persuading an appellate court that the trial court judge abused her discretion. Second, federal judges define objections to admissibility as arguments about probative value. In practical terms, this means that the opponent is left with a Sisyphean task - arguing that the judge erred in granting any probative value to the business records and thus that she should not have permitted the jury to consider them at all. Given the context of trial court discretion, the probability of carrying that burden approaches zero.
 The shift of the disagreement between the circuits to the domain of probative value can be described as a pragmatic movement to resolve the dilemma raised by the doctrinal confrontation of necessity and reliability - as a reconciliation between the two pre-Rules lines of cases that reflect that dilemma. In particular, the need for computerized business records is served by their easy admissibility, while the issue of system reliability can be joined before the jury under the rubric of probative value. Certainly the trial process calls for pragmatic solutions.
 The main criticism of the U.S. caselaw and Rule 803(6) offered by law journal commentary is that admissibility is made too easy for the proponent of the records adduced and objecting is made too difficult for the opponent. In compensation it has been suggested that proof of the reliability of the data processing techniques and equipment used be required under Rule 901(b)(9). Federal rule 901 states in part:
Rule 901. Requirement of Authentication or Identification
(a) General provision. The requirement of authentication or identification as a condition precedent to admissibility is satisfied by evidence sufficient to support a finding that the matter in question is what its proponent claims.
(b) Illustrations. By way of illustration only, and not by way of limitation, the following are examples of authentication or identification conforming with the requirements of this rule: ...
(9) Process or system. Evidence describing a process or system used to produce a result and showing that the process or system produces an accurate result.
 The Civil Code of Quebec deals with computerized records of juridical acts, i.e. documents intended to have a legal effect on their makers, such as contacts (Articles 2837 - 2839). The full text is in Appendix B. It allows these records in evidence if they are intelligible and their reliability guaranteed. It contains a mix of traditional common law rules and rules specific to computers. For example, it refers to the circumstances under which the data were entered, which echoes other Canadian statutes.
 It then says that the reliability of the entry of data is presumed to be sufficiently guaranteed "where it is carried out systematically and without gaps and the computerized data are protected against alterations." In other words, evidence on the operation of the computer system is needed. The Code expressly states that a document producing these data may be contested on any grounds, so the opposing party may try any line of attack on the reliability of the electronic record.
 On the other hand, it also states that the reliability of documents "drawn up in the ordinary course of business of an enterprise ... is, in particular, presumed to be sufficiently guaranteed". Electronic business records that are not "juridical acts" may well be admitted under this provision, which is part of the rules on hearsay (Article 2870). The weight of any evidence is always left for the trier of fact (Article 2845).
 A somewhat different approach to law reform is being taken by the United Nations Commission on International Trade Law (UNCITRAL). Its draft model law on electronic records (mainly focussing on EDI) simply provides that the electronic form of a record shall not affect its admissibility in litigation. It suggests a couple of factors that might go to weight, namely the reliability of the manner in which the record was created, stored or communicated and also the manner in which it was authenticated, if this is relevant. The current draft text is in Appendix B.
 In addition, the UNICTRAL draft model law deals with the concept of "original". It supplants the strict concept with a "functional equivalent", i.e. with criteria that a record must fulfil to perform the same functions as does a paper original. The main function is to ensure the integrity of the record.
options for harmonized rules
 The main discussion in this section will relate to business records, with some overflow to banking records. Public records will occupy a smaller section. Rules for microfilm and imaging are also discussed separately.
 Four options seem to encompass the main points of discussion: do nothing; make a small gesture to acceptance of electronic formats; create a list of special rules for electronic records; integrate all the provisions of the statutes into a thematically consistent treatment of all documentary evidence. We should consider whether any new rules adopted should affect admissibility or only weight of the record once it is admitted.
1. Do nothing
* The current rule, basically one of showing actual business reliance on the record, is right. If the creator of the record trusts it, why should courts not trust it, at least to the point of admitting it for consideration?
* Neither judges nor litigators are asking for change in the present law, under which electronic records are normally admitted.
* Businesses, especially financial institutions, have come to expect the open standard of reliance and would be unhappy to have more rules to satisfy to get their records in.
* Listing specific factors to consider may risk impeding changes to technology and thus to business efficiency.
* The current rule has several areas of uncertainty that the case-by-case development of the law will not resolve for years.
* Many solicitors have difficulty advising their clients about the proper conduct of their businesses because of uncertainty about the use of electronic records in litigation.
* A very liberal use of electronic evidence may be unfair to someone wishing to attack the reliability of the evidence; the onus of proof of reliability should be on the proponent of evidence not on its opponent.
* Current rules are based on inconsistent statutes and thus create a diversity of law across Canada. Case law is not capable of resolving this in the short run.
* National and industry standards for the proper use of computer technology are widely accepted and do not inspire fear of unduly limiting the technology or systems to which they apply.
2. Adopt a limited, facilitating rule for electronic evidence
 Such a rule would reflect the UNCITRAL model law mentioned earlier, the full text of whose relevant provisions is in Appendix B. In addition, it should allow parties to contracts to choose their own rules for admissibility and weight in litigation between themselves, subject to contractual defences.
* A limited permissive rule is all that is needed to end a lot of uncertainty about the use of electronic records in court.
* A limited rule is less likely to limit technology than more precise requirements.
* A limited rule will not require detailed proof of the operation of computer systems even when electronic evidence is not seriously disputed.
* The limited rule appears to be the latest international legal standard, that of the United Nations.
* The basic fact of admissibility of computer records in general is not in doubt. As a result, an UNICTRAL-type rule adds little or nothing of use to our current law (except the ability of the parties to consent to an evidence rule).
* The limited rule does not resolve many areas of uncertainty discussed earlier in this paper.
3. Add a special list of factors touching electronic evidence
 These factors would concern practices that should be considered in judging the organization's usual and ordinary course of business and of the circumstances of the making of the records in question. A description of the sources of information for the database would be required. Transactions should be recorded contemporaneously before memory fades. The information going into the database should be routine business information. The reliability of the resulting database could be shown by evidence of business reliance upon it in making business decisions. And there should be some evidence of the reliability of the software used and of the security procedures protecting the organization's record-keeping system. By keeping the list of factors short and basic it will be applicable to small and simple systems as well as large and complex ones. The court and parties could choose to use other factors in addition to those expressly listed if an open ended and permissively inclusive wording were used.
 How should such listed factors be left to the courts to be applied? One would not want them applied in the same way to a small office having a single personal computer as to a large mainframe computer installation, or in the same way to proof of a single business letter as to proof of a large complex database. Therefore the applicability of each factor has to be left to the court in regard to each record adduced. The application of each factor would be flexible but would have to be considered in each case. (The list is an adaptation of one in the national standard for electronic imaging, reproduced in Appendix D.)
Records produced by computers may be accepted as original records. In determining the admissibility and weight of records produced by computers regard shall be had to the circumstances of the making of the records including the following factors:
- Sources of Data and Information -- The sources of data and information recorded in the databases upon which the record is based.
- Contemporaneous Recording -- Whether the data and information in those databases was recorded in some fashion contemporaneously with, or within a reasonable time after, the events to which such data and information relates (but contemporaneous recording within those databases themselves is not required).
- Routine Business Data and Information -- Whether the data and information upon which the record is based is of a type that is regularly supplied to the computer during the regular activities of the organization from which the record comes.
- Data Entry -- Whether the entries into the databases were made in the regular course of business.
- Business Reliance -- Whether there has been reliance upon those databases in making business decisions within a reasonably short time before or after producing the records sought to be admitted into evidence.
- Software Reliability -- Whether the computer programs used to produce the output, accurately process the data and information in the databases involved.
- Security -- The security features used to guarantee the integrity of the total information or record-keeping system upon which the output is based.
 The evidence that these points will produce will vary with each information management or record keeping system. However, a single supervising officer of any well-run information or record-keeping facility should be a sufficient witness. An additional witness might be required for software that is unique to the system if that supervisor cannot testify to its history of reliability. If not, the programmer who wrote it should be available to certify its reliability until it does have a history of reliability.
* The list would allow the court to require the proponent of the records to prove their reliability rather than transferring the onus of proof to the opponent just because the proponent has provided some superficial evidence as to the records having been made in the usual and ordinary course of business;
* It would permit an analysis of system integrity instead of paper originals;
* It would provide a mechanism for achieving some uniformity in the court decisions as to the factors affecting admissibility and weight;
* It would provide a framework for conducting voir dires as to admissibility;
* It would provide guidance to the parties to proceedings as to the necessary evidence to be marshalled to prove the admissibility and weight of computer-produced records.
* It would provide guidance to business on the proper development and maintenance of electronic record keeping systems and on electronic transactions.
* It would not limit the development of technology or the kind of record keeping used, because the rules apply to information handling, not to the computer systems that handle the information. They do not require specific practices or levels of performance.
* Proponents of electronic records might have to submit extensive foundation evidence even when the records were not seriously contested.
* In any event litigation would become more time-consuming and costly to introduce evidence now routinely introduced without these foundations [but is the present easy-to-introduce regime fair to those who want to dispute the evidence?]
* The only real rule needed is that the proponent actually relies on electronic records of the class to be admitted - without demonstrating an actual incident of reliance on the particular records at issue in the litigation.
* An acceptable supplementary rule may be that the records were made for routine business purposes, though not necessarily that they record only routine business events - rather than being prepared with a view to the current litigation, for example.
4. Rewrite all the documentary evidence rules on principles consistent withthe rules for electronic records
 The statutory rules for all types of documents would be revised on common principles, and rewritten rules on microfilm and imaging would be integrated into them. This possibility was extensively debated in the Uniform Law Conference and law reform bodies in the late 1970s and early 1980s. The essence of these debates and the results at the time are set out in Appendix C.
 Differences among traditional business records, computer-produced records, microfilm records, electronically imaged records, and banking records can be dealt with in definitions and special subsections where needed rather than having completely separate provisions for different types of records as at present. Separate provisions have worked so far because creating business or banking records and later microfilming them have been separate distinct functions.
 Now, computer technology has integrated the creation of business records and banking records with the use of microfilm and other storage media. If as expected, electronic imaging becomes widely used for both current and archived records, then two sets of separate provisions will frequently have to be used--the imaging/microfilm provisions for the imaging operations, and the business record provisions for the computer operations. Instead, imaged records could be part of the definition of an "original" record in an integrated set of provisions.
* Unified rules would deal comprehensively with all business records and not just computer-produced records;
* Such rules would separate and make clear the three traditional classes of evidentiary issues concerning business
records - best evidence rule, authentication, and hearsay rule issues - and provide detailed provisions concerning them;
* They provide uniform supporting procedural provisions such as notice, necessary witnesses, cross-examination, and production;
* In particular, records involving microfilm and electronic imaging will often involve computer functions that will invoke the computer provisions of the business record provisions. More often than not, both the business record and the imaging/microfilm provisions would be relevant. Therefore a single set of integrated provisions would be helpful.
* The other business record provisions should be uniform across Canada, for similar reasons as the electronic record provisions. Provisions unified in principle would best serve that purpose.
* They are the product of an extensive consultation and law reform process, although not recent.
* Unifying the whole field will take too much time and effort, as can be seen from the debates when we tried it last time.
* The rules on paper documents are working well, despite gaps in principle, and they do not need revision.
* If what is needed for electronic records is no rule or a simple permissive rule, then there is no need to revisit all the other provisions to solve this narrow problem or set of problems.
- Next >>