This Web page has been archived on the Web.

1993 Report of the Auditor General of Canada

Assistant Auditor General: Shahid Minto
Responsible Auditor: Jim Ralston

Main Points

21.1 In an effort to enhance productivity and improve service to taxpayers, the Department of National Revenue instituted a program for electronic filing of personal income tax returns, which was launched nationwide this year. This program is known as EFILE. With EFILE, the Department has demonstrated a continued commitment to pursue and develop new, innovative tax administration technologies.

21.2 Departmental statistics reveal good initial acceptance of the program. About 2.1 million taxpayers filed their 1992 returns using EFILE, surpassing by a considerable margin National Revenue's own predictions.

21.3 The Department tested the data capture and transmission systems on which EFILE is based through a phased-in implementation. This proved successful.

21.4 EFILE is also paperless. Taxpayers using EFILE are not required to submit any supporting documentation unless the Department asks for it. The Department made this decision based on limited knowledge and assessment of the impact that such a move would have on voluntary compliance. The potential for revenue loss if taxpayers' voluntary compliance deteriorates could be significant. The Department needs to expand and improve its efforts to monitor changes in voluntary compliance.

21.5 New verification programs related to the introduction of EFILE are evolving. Most of the work of those programs takes place after refund cheques or initial assessments are issued. To mitigate the risks of revenue loss associated with paperless electronic returns, National Revenue should conduct its present document matching program earlier, at the time that tax returns are filed.

Introduction

Electronic filing is central to National Revenue's vision of efficient tax administration

21.7 The electronic filing of personal income tax returns is one such initiative. Launched nationwide this year, electronic filing has caused profound changes to tax administration practices. The implications are far-reaching, since electronic filing is the forerunner of changes that will also affect the processing of paper income tax returns. Departmental officials have indicated that returns processing is undergoing the most significant restructuring of the last 30 years.

21.8 Electronic filing (EFILE) provides taxpayers with an alternative to paper tax returns. With EFILE, personal income tax information is sent electronically to National Revenue, using communication lines. EFILE is paperless - taxpayers who use it are not required to submit any supporting documentation unless the Department asks for it. However, taxpayers are expected to retain the information for a period of six years.

21.9 EFILE offers substantial opportunities for National Revenue to streamline its operations. The Department believes that EFILE will reduce the cost of entering information from tax returns into its computer files. EFILE means less paper to handle, fewer keying errors and fewer arithmetical errors on tax returns as originally submitted. It also means reducing the Department's storage facility requirements. Other benefits made possible with EFILE include providing a new vehicle for the Department to communicate with the tax preparation and software industries. ( see photograph )

Taxpayer acceptance of electronic returns has exceeded National Revenue's expectations

• have the return completed by the tax preparer, who will also electronically file the return;
• complete the return on paper and then have a tax preparer convert it to an EFILE format; or
• use EFILE microcomputer software to complete the return and then provide the tax preparer with an electronic copy of it.

21.11 For taxpayers, electronic filing means that they receive their assessments and tax refunds in less time. Electronic returns bypass various manual processing steps - such as sorting, batching and data transcription - that are necessary for paper returns (see Exhibit 21.2 ). Those processing steps, which the Department estimates can take as long as three weeks if done manually, are automated for electronically filed returns.

21.12 Assessments for electronic returns are processed within about 14 days, compared to the usual three to eight weeks for paper returns. Where taxpayers have asked to have refund cheques deposited directly to their bank accounts, the wait under EFILE can be as brief as eight days. The system also provides authorized transmitters with National Revenue's acknowledgment of receipt of transmission within hours.

21.13 National Revenue statistics show the program's appeal (see Exhibit 21.3 ). In its first year of nationwide operation, about 2.1 million taxpayers used EFILE (about 11 percent of personal tax return filers), surpassing by a considerable margin the Department's own predictions. In the United States, it has taken the Internal Revenue Service three years to reach this percentage. The Department estimates that individuals who used EFILE were assessed $6.4 billion in federal tax revenues for 1992, representing approximately 12 percent of the total for the year. Currently, about 95 percent of personal tax returns are eligible to be filed electronically. Generally, all but the more complex returns (such as those involving deceased taxpayers, or non-residents) are eligible.

21.14 Statistics show that individuals using EFILE are mostly those expecting larger tax refunds. For 1992 returns, 78 percent of taxpayers who used EFILE received refunds averaging $1272, compared to a $987 average for paper and electronic returns combined. Over the last three years the average refund for EFILE returns has consistently been higher - by at least 20 percent - than the average refund overall.

21.15 National Revenue anticipates that EFILE volume will double next year, reaching four million returns, or roughly 20 percent of all personal income tax returns. In the future, the Department intends to make EFILE more widely accessible to individuals at the lowest cost possible while maintaining high security. As a long-term objective, the Department wants to expand the scope of electronic filing to include corporate and trust tax returns.

Audit Objectives and Scope

21.17 We conducted our audit work in National Revenue's head office, six taxation centres, and eight district offices.

Observations and Recommendations

New Returns-processing Regime

Using a phased-in approach, National Revenue put EFILE in place nationwide a year ahead of schedule

21.19 The Department proceeded in a systematic way to develop the data capture and transmission systems on which EFILE is based. It tested the technology and public acceptance of EFILE through phased-in operations in western Canada, the Northwest Territories, the Yukon and Quebec. Gradual implementation proved successful on other fronts as well. It enabled the Department to check the computer and telecommunications systems supporting EFILE for capacity and response time at different volume levels, and provided a good test for built-in systems controls, including the production of acknowledgments.

21.20 During the phase-in period, the Department surveyed Canadians on their receptiveness to EFILE as a filing alternative and kept pace with legal considerations affecting electronic paperless returns. It also consulted with representatives of the tax preparer, transmitter and software industries.

Absence of timely cost-benefit analysis weakens accountability for EFILE

21.22 The absence of an analysis of expected costs and benefits at the time the decision was taken weakens the accountability for the EFILE project. The fact that an analysis was done later, albeit for a purpose other than supporting the decision and establishing a basis for accountability, mitigates our concern somewhat.

21.23 Also required for adequate accountability is an appropriate mechanism to monitor costs and to verify that planned benefits are achieved. The Department indicates that, although it monitors the costs of various activities and components of its operations, it does not think it feasible, practical or meaningful to track costs specific to initiatives such as EFILE. We appreciate the Department's concern that an elaborate accounting system may not be worth the effort needed to maintain it for some projects. However, we think it possible, given today's technology and software, to find a suitable, economical accounting system in most circumstances.

21.24 With respect to benefits, the Department most recently estimated, in May 1993, that it will save 62.5 person-years or $1.8 million in front-end mailroom, keying, storage and error correction activities for every one million EFILE returns processed. The Department has begun to redeploy staff savings to other areas. ( see photograph )

21.25 The Department should perform a cost-benefit analysis on a timely basis for projects such as EFILE and should put in place appropriate systems for monitoring costs.

Department's response: The Department agrees that a cost-benefit analysis is normally a basic ingredient when recommending an initiative. As noted, EFILE offers substantial opportunities to streamline operations. The Department already has appropriate systems in place to accurately cost various activities and components of overall operations; but it considers tracking costs specific to initiatives such as EFILE to be neither practical nor meaningful, until they become support tools for the new lines of business.

Impact on Voluntary Compliance

Introduction of EFILE may have an impact on voluntary compliance

21.27 In a paperless environment, information slips and receipts are not available immediately to National Revenue to check the legitimacy and accuracy of income and deduction amounts reported by taxpayers. With paper returns, information slips and receipts are available at the time of initial assessment and are subject to some checking. If National Revenue finds an error, or suspects an intentional misstatement, it can correct it or follow it up immediately to keep the error or misstatement from entering the assessment. The absence of this information would seem to increase chances that assessments based on incorrect information may be issued. Incorrect assessments can, in turn, lead to incorrect social benefit entitlements where benefits received by taxpayers are tied to tax return information.

21.28 Also, it seems likely that taxpayers who are required to file information slips and receipts will be more careful to ensure that the amounts they record on tax returns agree with the supporting documentation. In a paperless system, taxpayers might be less diligent about keeping and using supporting documents and thus more likely to make errors or misstatements. Granted, the information should still be available and taxpayers will still have to present it if National Revenue asks them to.

21.29 Finally, there may be taxpayers who see the elimination of the requirement to file information slips and receipts as an opportunity to cheat on their tax returns. Some may be individuals who would deliberately try to evade taxes in any case; others may simply find that an innocent error one year goes undetected, and then try the next year to slip a not-so-innocent error past National Revenue. Our concern is further heightened by recent literature on the Canadian underground economy, which seems to suggest that an increasing number of Canadians are turning outside of the formal economy to evade paying taxes.

21.30 We are concerned that National Revenue's paperless EFILE initiative may ultimately lead to a reduced level of voluntary compliance by taxpayers, an increased enforcement problem, and a significant loss of tax revenue for the government. We recognize that the Department generally has three years to reassess personal tax returns and that it intends to devote increasing resources to post assessment verification programs. Lost tax revenue could be recovered. But the Department does not have the resources to ensure that every erroneous tax return is reassessed. Moreover, even for those that are reassessed, collection is not guaranteed. However, we note that the Department can offset refunds in any one year against taxes owing from prior years.

21.31 In 1991 the Department reported on a qualitative threat-risk assessment of paperless EFILE, after which it removed most of the paper filing requirements for EFILE returns for that year. The Department considered it crucial to review the 1991 returns for impacts prior to nationwide implementation. However, in the end it decided to eliminate the requirement to file supporting documentation, without first having seen the results of the test (see paragraphs 21.35 and 21.36).

21.32 The United States and Australia both have systems for electronic filing of personal tax returns. In the United States, the Internal Revenue Service still requires taxpayers who file their returns electronically to send in third-party information slips on employment earnings and income tax withholdings. In Australia, the same type of information slip as well as third-party information slips supporting pension contributions are required either to be sent to the Australian Taxation Office or to be presented to a tax preparer, who will retain them on the Taxation Office's behalf.

21.33 It is significant that neither Australia nor the United States has removed completely the requirement to provide supporting documentation for income tax withholding amounts at the time of filing. Income tax refunds represent mainly the repayment of excess amounts of tax withheld from employees at source. An error that overstates tax withholdings can cause an overstated refund to be issued. In not requiring evidence of income tax withholdings before issuing refunds, Canada is clearly accepting a greater level of risk of incorrect refunds than has been accepted by the United States and Australia.

21.34 The Department has put in place new verification programs to compensate for the change in the control environment related to the introduction of EFILE. However, these were developed only recently and are still evolving, and some aspects were untested prior to nationwide implementation. As well, the resulting verification activities largely take place after the refund cheques are issued. Given these circumstances, National Revenue's ability to contain the risks associated with paperless EFILE returns has not been proven, particularly in this first year of nationwide implementation and in view of the unexpectedly high volume.

The Department needs to expand its efforts to monitor changes in voluntary compliance caused by introducing paperless returns

21.36 In our opinion, however, the preliminary results are far from conclusive. First, they are based on a comparison of only five items (the EFILE program looked at 13 items but comparative data for paper returns were obtained for only five). Second, the five items did not include information on income and income tax withholdings, which often are claimed by people filing electronic returns and which U.S. experience shows are frequently implicated in frauds involving electronic returns. Third, there may be differences in the characteristics of the EFILE and paper return filing populations that render the respective samples incomparable. Fourth, the data from only one year are incapable of revealing trends. Changes in voluntary compliance may vary substantially from one year to the next and may be cumulative over time.

21.37 It might be argued that the exclusion of information on income and income tax withholdings from the Department's monitoring program is not significant, since these items are covered in its matching program (which compares information on T4 slips filed by employers with information on personal tax returns). But at present, the matching program is not set up to compare the incidence of discrepancies between EFILE and paper returns. We understand that, subsequent to our audit, the Department requested that this information be captured from its systems.

21.38 We also observed that the Department is changing the composition of items targeted for monitoring of taxpayers' voluntary compliance. As shown in Exhibit 21.5 , for 1992 the Department has selected 16 items to monitor on the EFILE tax return. But only about half of these are in common with the 13 items selected for monitoring from the EFILE returns for the program in the previous year. Moreover, to compare the compliance rates between EFILE and paper returns for 1992, the Department will have data for ten items. However, only four of these items are in common with the five reported on in the previous EFILE-to-paper comparison. This will limit the Department's ability to evaluate changes in voluntary compliance over time as the EFILE program expands and evolves.

21.39 In monitoring taxpayers' voluntary compliance, the Department should increase the scope of items covered, ensure that the most significant items are included, and ensure the year-to-year comparability of sample characteristics.

Department's response: The Department agrees that it is essential to monitor the level of voluntary compliance and make periodic comparisons. To that end, the Department will continue to use current compliance monitoring methods and pursue any further improvements in compliance monitoring techniques, as they become available.

Legal Matters

Legal advice was sought on a regular and timely basis

21.41 The Department was required to seek amendments to the Income Tax Act to accommodate changes brought about by the EFILE initiative. The required amendments were tabled in the House of Commons in June 1992 and received royal assent in June 1993. The amendments provide a legal basis for EFILE returns and define the weight that will be given to EFILE data in a court of law.

21.42 We noted that the Department proceeded with full implementation of the initiative nationwide in advance of the necessary legislative changes. Although legislation was passed in June of 1993 with retroactive application to 1992, the Department had completed its processing of the 2.1 million EFILE returns for the 1992 tax year, the first year of nationwide implementation, while legislation was still pending.

EFILE requires a new way of linking a taxpayer with a return

21.44 Taxpayers wishing to use EFILE are required to sign the form and to retain a copy; the Department may ask to examine it at a later time. EFILE tax preparers are also required to keep copies of their clients' T183 forms for a period of six years. Failure to do so may result in a tax preparer's rejection from the EFILE program.

21.45 The Department checks that tax preparers are maintaining their copies of the T183 form. For the 1992 returns, the Department contacted 75 percent of its authorized EFILE tax preparers and asked them to provide the forms for 22,500 EFILE tax returns. It found that 95 percent of the T183 forms were completed in accordance with the Department's requirements. The sampling method was not designed to ensure a specific degree of confidence that preparers were complying with the requirement as designed. For 1993, the Department plans to use a statistically based monitoring process to select returns for T183 monitoring. This should provide it with greater confidence that EFILE tax preparers are in compliance.

21.46 Although we are unable to determine the risk of loss associated with having this potentially important document kept outside the custody of National Revenue, we are concerned that the evidentiary link between tax returns and taxpayers may be broken. Our concern is heightened by the fact that the Department is silent on security and confidentiality standards for the safekeeping of T183 forms at EFILE tax preparer premises. As well, the Department has no procedures to follow with respect to the forms in the event of sale or closure of an EFILE tax preparer's business.

21.47 The Department has indicated that the lodging of taxpayer authorization forms processed off-site is an interim position. It is pursuing new cost-effective technologies to receive electronic transmissions of taxpayer authorizations and signatures at the time of filing.

21.48 The Department should specify standards for security and confidentiality of the T183 forms lodged at tax preparer premises, as well as procedures to follow in the event of sale or closure of a tax preparer's business.

Department's response: Ongoing monitoring of form T183 has not revealed any problems associated with the security and confidentiality of these forms. We are also investigating voice recognition and electronic signature technology that should render this form obsolete.

Control Environment

Screening of tax preparers and transmitters is a first line of defence

21.50 We noted that the application form asks potential tax preparers and transmitters for only limited information. For instance, unlike the Internal Revenue Service (IRS), the Department does not require disclosure of associates, and does not require applicants to state if they have been assessed penalties in the past, if they have been convicted of a monetary crime, if they have unpaid tax liabilities, or if they have ever been convicted of a criminal offense under the Income Tax Act . Disclosure of such information on National Revenue forms would provide useful data for the screening process. We also observed that, unlike the IRS, the Department has no authority to invoke penalties against applicants who submit false or erroneous information.

21.51 EFILE tax preparers and transmitters must apply for authorized status each year, but they might undergo suitability screening only once every three years. In the intervening period, the Department may accept them on the basis of their performance in the program in the prior year(s). For the 1992 tax year, the Department rejected 15 applicants and accepted a total of 7100 into the EFILE program. Of these, 4520 were preparers, 75 were transmitters and 675 were both transmitters and preparers. The remainder were inactive. EFILE has expanded considerably over the past few years and the Department foresees significant growth in the number of EFILE applicants. It is critical that the Department have sufficient resources in its district offices to ensure proper screening for suitability, and to monitor EFILE tax preparers and transmitters in the years between their suitability screenings. Exhibit 21.6 provides statistics on tax preparer volume for the 1992 EFILE returns.

21.52 The Department should improve the design of its application form to facilitate the screening of applicants for EFILE tax preparers and transmitters. The Department should assess the extent to which established criteria for suitability screening are being followed up. Criteria that are not being followed up should be reviewed by the Department and removed from the established list if found to be invalid. The Department should also consider seeking legislative authority to invoke penalties against applicants submitting false or erroneous information.

Department's response: The Department is in the process of redesigning the application form to facilitate screening of potential applicants. Suitability screening criteria, and procedures associated with their application, are in the process of being revised. National Revenue presently has the authority to deny an applicant entry into, or revoke participation in, the EFILE program if false or erroneous information is submitted, thus seriously impacting the economic livelihood of the applicant. A tax preparer penalty is under consideration to provide a less draconian alternative.

Our audit focussed on three sets of controls over EFILE returns

21.54 In the first category are computer-based controls over the transmission of data from preparers or transmitters to National Revenue. The objective here is primarily to ensure that the information received by the Department has been transmitted by authorized third parties, is complete and is in a usable format. We refer to controls of this type as front-end edit checks.

21.55 The second category consists of controls meant to highlight returns that may not comply with the tax law. These controls can be partly computerized and partly manual. Returns are highlighted before an assessment or before a refund is issued but may not be followed up until afterward. The controls of this type include confidence validities, error clues and the fraudulent refund detection program.

21.56 The third category of controls differs from the first two in that the controls operate only after assessments have been completed and refunds issued, but their purpose is the same - to detect possible noncompliance. The processing review and matching programs fall into this category.

The Department has appropriate, automated front-end edit checks built into the EFILE system

21.58 During our audit, we tested the automated edit checks built into the EFILE system, which are designed to ensure that the data as initially submitted are arithmetically correct, the tax preparer's and transmitter's numbers are valid and the EFILE transmissions that the Department accepts for subsequent processing, including the generation of corresponding acknowledgments, are complete. The results of our tests confirmed that appropriate controls exist and that these controls are effective in meeting their design objectives. Further, the Department asked the Royal Canadian Mounted Police to conduct a security review of EFILE front-end systems prior to nationwide implementation. The Department has made changes to the EFILE systems as a result of this review and has contacted the RCMP to assess the adequacy of these security improvements.

Some limited checks are done on electronic returns as part of initial assessing

21.60 Departmental statistics reveal that the number of electronic returns subjected to follow-up activity for confidence validities was small - about 6 of every 1000 returns filed this year. Selection criteria for confidence validities and the volume of transactions the Department reviewed were based in part on availability of staff and on the need to achieve a balance between the objectives of enforcement and maintaining taxpayer acceptance of the program.

21.61 At the time of initial assessment, electronic returns are also subjected to the same automated systems checks for reasonableness that are in place for paper returns. These checks are known as error clues. When error clues are combined with confidence validities, the volume of EFILE returns requiring follow-up action increases to about three percent of electronic returns filed for 1992.

21.62 If following up on unusual amounts and potentially excessive claims raises a suspicion of fraud in an electronic return, it will be examined in the fraudulent refund detection program. Paperless electronic tax returns may necessitate the development of new procedures to identify fraud, since examiners' procedures for detecting fraud on paper returns may not be appropriate for electronic tax returns. In the U.S., the IRS is developing better ways to identify fraudulent electronic returns. We note that National Revenue's detection program has not been modified for electronic returns. The Department has indicated that enhancements to the program are in the planning stage.

21.63 Although confidence validities, error clues and the fraudulent refund detection program are applied at the front end, the combined rate at which EFILE returns are examined is less than for paper returns during initial assessment.

Department's response: The Department considers that the controls, i.e., confidence validities, on-line error clues and the risk assessment profiling, will contain the risks associated with EFILE. In addition the Department continues to work actively on enhancements to these controls as well as investigating other compliance mechanisms.

Processing review can be used to improve the Department's ability to identify erroneous or fraudulent EFILE returns

21.65 In 1991 the Department reported a study, based on 1989 taxpayer return information, to assess the benefits of manual up-front verification and follow-up of automated systems reasonableness checks prior to initial assessments or the issuance of refund cheques. The study indicated that those procedures had resulted in the collection of approximately $119 million in additional tax revenue by the Department. Of that amount, $61 million was the result of manual up-front verification of documents supporting income and deduction amounts.

21.66 From the Department's perspective, returns processing would be more cost-effective if effort were focussed on examining returns that have the most potential for additional revenue. In its view, the key would be to ensure that any enhanced back-end review process would generate at least as much revenue as would have been generated by the up-front verification steps that were eliminated.

21.67 To implement this strategy, the Department created a new verification program called processing review, which addresses paper and EFILE returns separately. The program is run several months after returns are first processed and refund cheques issued. For EFILE returns, the Department verifies information by asking taxpayers to submit supporting documents. Its overall goal is to maintain the integrity of the tax system by ensuring that tax returns are properly assessed.

21.68 One objective of processing review is to provide information that can be used to update the criteria used by National Revenue to identify potential areas of abuse for further investigation. Part of the processing review program involves verification of a random sample of returns, and another part involves verification of a non-random "compliance" sample of returns that are considered most likely to contain errors. These samples will cover 22 items on the 1992 return.

21.69 The compliance or non-random sample currently is selected using criteria that are plausible but arbitrary, and that have not been empirically tested prior to implementation. The random sample is intended to identify and test such criteria. If proven effective, they can be incorporated into the routine for selecting subsequent compliance samples.

21.70 We wanted to find out whether the new back-end review process (processing review) had generated as much revenue as would have been generated by the up-front verification steps that were eliminated. We found that the Department had not collected the information necessary to answer this specific question. As an alternative, we looked at information that the Department had gathered for other purposes to see whether it could shed light on the question.

21.71 We found that, in 1993, the Department estimated the maximum additional revenue that might have been found in a sample of certain types of unverified items on 1991 paper returns if the returns had still been subjected to manual up-front verification for these types of items. The amount was estimated at $39 million but, due to the fact that taxpayers were not contacted to provide explanations or other supporting documents, the Department feels that this figure overstates the maximum revenue foregone (possibly by as much as 50 percent). By way of contrast, we note that the actual reassessments resulting from the new back-end review of the paper returns for the same types of items amounted to only $1.2 million.

21.72 The significant variation suggests that the focus on enhanced back-end verification may not be as effective in generating additional revenue as was anticipated. We assume that the experience with respect to the same types of items on electronically filed returns would be similar to the above results relating to paper returns.

21.73 This also illustrates the need for the Department to establish appropriate, reliable measures for evaluating the effectiveness of the back-end verification process. The proxy measures that we made use of are adequate for our purpose but we would expect the Department to have better information for its own monitoring requirements.

21.74 The verification of both the random and compliance samples is meant to fulfil yet another objective of the processing review program - to heighten taxpayers' awareness of the Department's enforcement activities. However, the Department is conscious that pursuing this objective for EFILE returns would impact on another departmental objective: to make using EFILE attractive to taxpayers. In establishing the extent of contact with taxpayers for processing review of the 1991 EFILE returns, the Department balanced those objectives and decided on a combined random and compliance sample of 43,000 returns. For the 1992 returns, the sample will increase to 104,000 although, due to growth in the EFILE population and the availability of resources, the percentage coverage will decline.

21.75 The Department should establish appropriate, reliable measures for evaluating the effectiveness of the back-end verification process.

Department's response: The Department's processing restructuring plans are a multi-year process. The Report indicates that the audit focussed on the control environment in place for the 1992 EFILE returns. Consequently, this "snapshot" review does not take into consideration the significant enhancements to the processing system that have been implemented for 1993 returns and the further enhancements scheduled for implementation in subsequent years.

Earlier matching would reduce risks related to paperless filing

21.77 Matching is a long-established enforcement program performed on virtually all returns. It involves the verification of income and deductions reported by taxpayers against information slips with valid social insurance numbers received from third parties, for example, T4 slips from employers and T5 slips from financial institutions. Matching generated over $194 million in reassessments in 1992-93.

21.78 Currently, matching is done in the fall, after returns have been assessed and refunds issued. The main reason is that, even though the Department receives the information from third parties in the spring, a significant proportion is received on paper and has to be converted to electronic format so that the computer can do the matching.

21.79 In our 1990 Report, we recommended that the Department take steps to increase the proportion of T4 and T5 slips filed in magnetic format by employers and financial institutions. Since then, the proportion of T5 slips filed in magnetic format has increased from 38 percent to about 90 percent, but for T4 slips the proportion has increased only from 20 percent to 27 percent. The Department is committed to reviewing the need for mandatory filing of T4 slips on magnetic media at the end of 1993.

21.80 In our opinion, elimination of the requirement for supporting documentation for EFILE returns, the increased speed in processing these returns and the Department's focus on back-end compensating verification procedures significantly increase the overall risk that fraudulent or erroneous tax returns will go undetected at initial assessing. Because of this, we believe it is more important than ever to perform matching as early as possible - at the time returns are first assessed and before refunds are paid. Having third-party information to match with taxpayer returns at the time the returns are filed would enhance significantly the Department's control environment over initial assessing and payment of tax refunds. To this end, we think the Department should be exploring ways to implement early third-party matching as soon as possible. As a first step, we again urge the Department to take measures to substantially increase the percentage of T4 slips filed on magnetic format. For those T4 slips that are not filed on magnetic format, we note that the Department is considering the use of electronic scanners for data conversion. It has recently started using this technology for T5 slips. In addition, the Department is exploring the feasibility of electronic transmission of third-party information slips over communication lines.

21.81 To mitigate potential risks associated with paperless returns, the Department should perform matching at the time of initial assessing.

Department's response: The Department confirms that this is consistent with its plans for the Matching Program. Indeed, the Department's T1 Simplification Initiative for 1993 returns will result in the pre-printing of selected third-party information for a selected number of taxpayers on customized returns. Additionally, significant progress has been achieved in obtaining information slips in magnetic or scannable format. The Department is committed to using third-party data in our on-line processing of individual returns both to simplify the filing process and to enable automatic verification during assessment.

Information for Parliament

Information to Parliament can be improved

21.83 Other information about EFILE has appeared in Part III of the Estimates each year, starting with 1990-91. Disclosure has included information on human resources costs and other expenditures, and discussion of expected benefits such as faster refunds and assessments, more accurate returns and streamlined operations. Part III does not indicate standards of service for EFILE, nor does it report on performance.

Summary and Conclusion

21.85 Regarding the paperless aspect of EFILE, we found that the Department acted on very limited information about the risks involved, and with only a qualitative assessment of the effect this may have on voluntary compliance by taxpayers. Although National Revenue has identified and put in place important compensating verification and monitoring programs, we are concerned, given the newness of these programs in the year of nationwide implementation, that their likely effectiveness in achieving desired results is unknown. At this stage, it is unclear what the overall impact on tax revenues will be.

21.86 The Department has focussed primarily on achieving gains in operational efficiency. In contrast, we found that it has not placed the same emphasis on assessing the effect of the new program on taxpayers' voluntary compliance or on the performance of the new verification programs. The Department could mitigate many of the risks associated with paperless filing by having third-party information matched with taxpayer returns at the time of initial assessing.

Department's response: The Department has a strong commitment to pursue and develop new, innovative tax administration technologies. Electronic filing (EFILE) is a clear example of an innovative use of technology that is improving client service and enhancing productivity.

This success story provides a simple, cost-effective way for Canadians to have their income tax returns processed. At the same time, the Department continues to invest in and enhance complementary compliance mechanisms. This not only increases compliance, but also strengthens the Canadian self-assessment income tax system.