This Web page has been archived on the Web.

2004 March Report of the Auditor General of Canada Chapter 5—Canada Revenue Agency—Audits of Small and Medium Enterprises

2004 March Report of the Auditor General of Canada

Chapter 5—Canada Revenue Agency—Audits of Small and Medium Enterprises

Main Points

Introduction

Observations and Recommendations

Risk management

Information on compliance

Risk evaluation

Design and delivery of audit activities

Legislated penalties

Conclusion

About the Audit

Exhibits:

5.1—The small and medium enterprise audit program, 2002-03

5.2—Compliance activities and their cost, 2002-03

5.3—Estimated and expected audit coverage for the small and medium enterprise audit program

5.4—GST/HST refunds paid in the last five years

Main Points

5.1 The Canada Revenue Agency does not have the resources to be everywhere at once. It has to make trade-offs in deciding where to deploy resources to deal with competing threats to the tax base. To do this, it must ensure that all threats to the tax base are identified and considered in a consistent manner.

5.2 The Agency's aim is to improve compliance with tax legislation. However, it does not have the necessary information to enable it to measure compliance over time and, therefore, the effectiveness of its activities. For example, the Agency has not estimated the amount by which it overpays GST/HST refund claims. It is only now in the process of developing a compliance measurement framework.

5.3 The small and medium enterprise audit strategy could better reflect risk. Staffing in each tax services office for small and medium enterprise audits depends on the population of the different segments of taxpayers served by the office rather than the risk and complexity inherent in those taxpayers' affairs. To meet audit coverage targets, tax services offices can restrict up to 50 percent of their audits to one to three issues. These restricted audits take less time to complete but do not permit an overall evaluation of the taxpayer's compliance with tax laws. In addition, two studies conducted by the Agency found that 40 percent of all small and medium enterprises audited in a given year had little or no potential tax at risk according to the Agency's computerized risk evaluation system.

5.4 The Agency's own quality assurance reviews show that audits of small and medium enterprises need more rigour. We noted several areas where auditors must improve, such as using indirect methods of verification and considering the application of penalties more often.

Background and other observations

5.5 In December 2003, the government announced a major reorganization of some departments and agencies. The customs services operations of the Canada Customs and Revenue Agency were transferred to the newly created Canada Border Services Agency. The remaining organization operates as the Canada Revenue Agency.

5.6 The Agency defines small and medium enterprises as all self-employed individuals, small and medium-sized corporations, partnerships, and trusts. Small and medium enterprises include about 4.5 million taxpayers. Self-employed individuals accounted for almost 10 percent of the $90 billion in federal personal income tax revenue in 2003, while corporations with annual gross revenues of less than $15 million accounted for 36.5 percent of the $22 billion in federal corporate tax revenue that year. Over 60 percent of net goods and services tax (GST) and harmonized sales tax (HST) revenues, other than that collected by customs services on imported goods, comes from small and medium enterprises.

5.7 The small and medium enterprise sector is growing as the trend to self-employment increases. It is also the sector most at risk of participation in underground economy activities.

5.8 Computerized risk assessment systems, in conjunction with experienced auditors, are currently used to identify non-compliant taxpayers. Audits of taxpayers identified by the system as having high potential for non-compliance yield much higher reassessments than audits selected through other means.

5.9 The Agency has made changes to the computer-based risk criteria to select GST/HST refund claims for audit. In addition, it undertook several initiatives and pilot projects designed to strengthen GST/HST compliance. The validity of the risk criteria and the results of the new initiatives must now be measured.

5.10 The Agency has yet to complete a systematic review of the penalties provided in tax legislation to determine how effective these penalties are in improving compliance and achieving public confidence. We found that where the Agency reduced a certain penalty it applies to the late remittance of income taxes deducted from employee salaries by employers, it failed to get legislative authority to do so.

The Agency has responded. The Canada Revenue Agency agrees with all but one of our recommendations. In its responses, it provides background information and describes actions it will take to address the recommendations.

The Agency believes that one of our recommendations on better understanding compliance behaviour would divert resources from other activities and impose a burden on compliant taxpayers.

Introduction

5.11 In December 2003, the government announced a major reorganization of some departments and agencies. The customs services operations of the Canada Customs and Revenue Agency were transferred to the newly created Canada Border Services Agency. The remaining organization operates as the Canada Revenue Agency.

5.12 The majority of businesses in Canada fall into the small and medium enterprise category as defined by the Canada Revenue Agency. According to the Agency, this sector presents a higher-than-average risk of not complying with the requirements of the Income Tax Act and the Excise Tax Act when compared to other Canadian taxpayers, such as wage earners. As a result, we expect the Agency to have an appropriate compliance strategy for this sector, including an effective audit presence. The need for such a strategy is reinforced by an observation made in a study for the Department of Finance Canada's Technical Committee on Business Taxation about tax planning that occurs in small businesses:

A prevalent attitude among small-business owners is that, given the risks and uncertainty that come with self-employment, they are entitled to organize their affairs to pay less tax than someone who enjoys the comfort of a regular salary.

5.13 The Agency defines all self-employed individuals, small and medium-sized corporations, partnerships, and trusts as small and medium enterprises. Small corporations include those with annual reported revenues of less than $1 million. Medium corporations have annual reported revenues not exceeding $250 million. The Agency estimates there are about 4.5 million small and medium enterprises in Canada.

Did you know?

  • Federal personal income tax revenue in 2003: $90 billion
    Percentage of that amount accounted for by the self-employed: almost 10%
  • Federal corporate tax revenue in 2003: $22 billion
    Percentage of that amount accounted for by corporations with annual gross revenues of less than $15 million: 36.5%
  • Net GST/HST revenue (other than that collected by Customs on imported goods) in 2003: $14 billion
    Percentage of that amount attributable to small and medium enterprises: over 60%

 

5.14 The small and medium enterprise sector grows as the trend to self-employment increases. It is also the sector most at risk of participation in underground economy activities. In our April 1999 Report, Chapter 2, Revenue Canada—Underground Economy Initiative, we noted that the underground economy results in tax evasion and represents an estimated annual loss of federal and provincial tax revenues of $12 billion. Promoting compliance with tax legislation in such a large and diversified group of taxpayers is a challenge for the Agency.

The small and medium enterprise compliance strategy

5.15 The Agency seeks to promote compliance with tax legislation through communication, quality service, and responsible enforcement. As a result of its activities it expects that Canadians will pay the amount of tax properly owed under the law and that the tax base will be protected. To succeed, the Agency knows that it must use different methods to effectively target and address compliance issues.

5.16 In April 2001, the Agency launched an exercise to identify the elements that must be in place to ensure good performance in the future. This initiative resulted in a series of Future Directions reports aimed at different groups of taxpayers. For small and medium enterprises, the Agency identified increasing its outreach, expanding electronic services, improving the accuracy of its processing activities, and ensuring compliance.

5.17 In particular, it committed to the precise targeting of enforcement actions: focussing efforts on non-compliant taxpayers; quickly identifying complex cases of non-compliance; and then assigning those to specifically established units. The Agency also committed to contacting new enterprises to provide timely guidance and assistance and providing for more on-line filing of information returns and for secure Web access by taxpayers to their account information.

The small and medium enterprise audit program

5.18 The small and medium enterprise audit program is one component of the Agency's overall compliance strategy. Its mandate is to promote compliance with the Income Tax Act, Excise Act, and Excise Tax Act in small and medium enterprises. The Audit Directorate within the Compliance Programs Branch is responsible for this program. The small and medium enterprise audit program has the largest staff of all audit programs and is divided into units that focus on different-sized businesses and different taxes (Exhibit 5.1). The Assessment and Collections Branch is responsible for identifying and following up on enterprises that have not filed their tax returns or registered to collect the goods and services tax (GST) or the harmonized sales tax (HST).

5.19 Auditors and program officers carry out audits and other compliance activities in 44 tax services offices across the country. Headquarters personnel in Ottawa, with the support of six regional offices, provide program direction and administrative guidelines to the tax services offices. Within the Compliance Programs Branch, the Compliance Research and Business Management Directorate provides compliance research and risk assessment services to support the small and medium enterprise audit program.

5.20 Audits of income tax returns and GST/HST returns are combined for individuals reporting revenues of less than $800,000. For other individuals and all corporations, auditors who specialize in either type of tax perform separate audits, although the audits may happen concurrently.

5.21 The GST is administered in Quebec by the ministère du Revenu (Department of Revenue) under the guidance of a master agreement between the Government of Canada and the Government of Quebec dated 1 July 1992. More than a dozen subsidiary agreements outline terms and conditions covering different aspects of tax administration. New terms and conditions for carrying out audit activities took effect 8 August 2002. They call for the ministère to follow the Agency's policies, procedures, standards, and practices when it audits files. They also establish the process by which the Agency and the ministère are to co-ordinate their annual audit strategy.

Focus of the audit

5.22 The objectives of the audit were to assess how well the Agency

5.23 Although we covered all components of the small and medium enterprise audit program, we placed emphasis on those dealing with small enterprise income tax audits and GST/HST audits. Those components employ over 75 percent of the audit program staff and are responsible for more than 99 percent of the small and medium enterprises in Canada. Further information about the audit objectives, scope, approach, and criteria can be found at the end of the chapter in About the Audit.

Observations and Recommendations

Risk management

5.24 Compliance activities. To know where to allocate its compliance resources, including how much to assign to small and medium enterprise audits, the Canada Revenue Agency needs to have an overall view of the threats to the tax base and to know which compliance activity is most effective under what circumstances. It also needs a rational process to allocate resources to a variety of compliance activities.

5.25 Non-compliance with tax legislation can take the form of failing to file a tax return or register to collect the GST/HST, incorrectly reporting income or deductions, or failing to remit taxes owing. These are the risks to the tax base. They result from a variety of reasons: taxpayers may not know their obligations or understand parts of the Income Tax Act and Excise Tax Act; they may not interpret the law as intended; they may wilfully evade tax; or they may make mistakes on their returns. Reasons for non-compliance vary from one taxpayer to another and among taxpayers of different types, industry sectors, and revenue levels.

5.26 The Agency carries out a mix of education, facilitation, processing, audit, and enforcement activities to improve taxpayer compliance. It also suggests changes to legislation to the Department of Finance. Penalties also play a role in encouraging compliance. In its annual performance report, the Agency groups these activities in three broad categories—voluntary compliance, assisted compliance, and enforcement—and shows the proportion of resources allocated to each (Exhibit 5.2). The cost of compliance activities increases as they move toward the enforcement end of the spectrum. Audits, for example, are an expensive way of gaining taxpayer compliance.

Risk management needs to be improved

5.27 The Compliance Research and Business Management Directorate of the Compliance Programs Branch is the focal point in the Agency for understanding compliance and assessing the risks of non-compliance. However, many areas within the Agency research compliance behaviour, sometimes in partnership with the Directorate and sometimes in isolation.

5.28 These areas then develop compliance strategies based on their view of the risk in their particular area. But compliance activities should be designed with an integrated view of risks. Informed decisions could then be made as to how many resources should be devoted to, for example, examining disability tax credit claims, GST/HST refund claims, and requirements for employers to withhold income tax, as compared to developing and running awareness campaigns or providing seminars for new business owners, and so on.

5.29 The Agency does not have the resources to be everywhere at once. It has to make trade-offs in deciding where to deploy resources to deal with competing threats to the tax base. It must assess the various risks to the tax base and it must know the cost of its intervention and the costs its intervention imposes on taxpayers. The Agency was unable to provide us with an overall risk analysis to support its current resource allocation.

Comprehensive information about risks to tax base needed

5.30 The Compliance Research and Business Management Directorate is developing a compliance measurement framework. This is a major initiative. The Agency sees the framework as the foundation for the study of compliance. Plans are for the framework to lay out the key concepts, constructs, and operational definitions that will guide compliance research planning, monitoring, and reporting. The Directorate also plans to integrate compliance results from relevant work done by all areas within the Agency and from research done outside the Agency.

5.31 As part of the framework, the Directorate is developing a compliance measurement information system to collect and manage data on measures that the Agency believes show compliance. However, it is too early to determine whether the compliance measurement framework and related system will produce the comprehensive information on compliance needed for the strategic management of tax programs within the Agency, or whether it will help provide key measures of compliance to determine the Agency's overall effectiveness in improving compliance.

5.32 Recommendation. The Canada Revenue Agency should ensure that all threats to the tax base are identified and considered in a consistent manner to support its allocation of available resources. It should consider implementing an overall risk management function or responsibility centre for risk management to achieve this.

Agency's response: The Canada Revenue Agency fully recognizes the importance of identifying and considering threats to the tax base in a consistent manner when allocating available resources. Accordingly, we agree that it is necessary to periodically validate and refresh information on compliance risk for informed decision making.

The Agency has an overall tax compliance risk management process founded on the following key elements: identifying non-compliance and distinguishing compliant taxpayers from those who are non-compliant; understanding factors that drive or enable non-compliance; prioritizing compliance issues based on risk assessment; developing targeted approaches that focus on the underlying drivers of non-compliance and that maintain existing compliance; leveraging strategic partnerships; and measuring success in increasing compliance and refining our approaches through continuous learning.

The Office of the Auditor General notes that the Agency is developing a compliance measurement framework and recognizes that this is a major initiative. Once fully developed, this framework will allow us to monitor general trends in compliance and to study the effectiveness of an individual program or compliance initiative. The Agency is also undertaking a comprehensive review of its tax collection methodology. This review will examine the approach to risk management across the Agency from a program, cross-functional, and strategic perspective. The review will also examine other factors, in addition to risk, which guide the allocation of resources.

The outcome of these initiatives will assist the Agency in enhancing its overall risk management and compliance strategy development. It will also be considered in future deliberations on the merits of implementing a single responsibility centre for risk management.

Information on compliance

5.33 The Agency recognizes in its performance report that understanding compliance behaviour helps to

Audit coverage rates lack support

5.34 In 2001, the Treasury Board of Canada Secretariat completed a resource and management review of the Canada Customs and Revenue Agency. The Secretariat noted that the Agency had established what it considered to be optimum coverage rates for audit activities, informed by compliance research conducted by the Agency and experience, and concluded that additional audit resources would generate additional tax revenues. We agree that additional audit resources are likely to generate additional revenue. However, Agency officials could not provide us with a definition of an optimum rate of audit coverage and evidence to support that they have determined it.

5.35 As a result of the resource and management review, the Agency received funding to increase audit staff by up to 1,860 employees. These increases are for all audit programs, including small and medium enterprise audits, and are to be implemented over five years. The new funding was conditional on the Agency achieving increased audit coverage and generating additional tax revenues. The Agency determined audit coverage targets over those five years on the basis of a planned staff complement for several program components. Exhibit 5.3 shows the targeted audit coverage in 2005–06 and the estimated coverage for 2000–01 and 2002–03 for the small and medium enterprise audit program.

5.36 As a result of the additional resources, the Agency expected gross tax revenue to increase by $3.2 billion over five years. The Agency did not break down the increase by audit program. As a result, the small and medium enterprise audit program has specific audit coverage targets but no corresponding revenue targets.

The Core Audit Program has not produced desired results

5.37 The Audit Directorate of the Compliance Programs Branch also researches compliance. One of its major initiatives in collaboration with the Compliance Research and Business Management Directorate is the Core Audit Program. This program, begun in 1999, seeks to determine the degree of non-compliance in a given population of small and medium enterprise taxpayers as well as the nature and cause of that non-compliance. It does this by auditing a sample of taxpayers or registrants chosen at random. Results of randomly selected audits can provide compliance information not available through regular audits selected on some basis of risk.

5.38 Because of resource considerations, the Core Audit Program has focussed on segments of the small and medium enterprise population or on a specific tax or category of taxpayer each year. To compare the compliance behaviour of individual and corporate taxpayers and GST/HST registrants in different industry sectors and over time, the Agency needs a detailed, long-range plan to guide its Core Audit Program. Currently it has not identified, beyond next year, the specific segment of the population the program will cover.

5.39 The Agency did not complete its analysis of the 1999 Core Audit Program, which covered audits of tax and GST/HST returns from 1996 to 1999, until the summer of 2003. The analysis of subsequent years' results has not begun. The Agency told us that problems with capturing data, completing audits, and developing an acceptable definition of what constitutes non-compliance contributed to delays in analyzing and reporting results. The Agency suspended the Core Audit Program for the current fiscal year to allow officials to make corrections and take stock.

Overstatement of GST/HST refund claims has not been estimated

5.40 The Agency needs to know, among other things, how much revenue is likely at risk for each type of non-compliance so that it can determine an appropriate strategy to minimize the loss. Overpayments of GST/HST refunds are one example: for a variety of reasons, some registrants claim bigger refunds than those to which they are entitled. They might misunderstand the tax rules, make clerical and system errors, or willfully intend to deceive.

5.41 In 2002–03, the Agency received 2.6 million GST/HST refund claims on which it paid out $29.8 billion. It audited 1.6 percent of these claims before paying them and found $247 million in overstated claims. The Agency has not estimated the potential overstatement in the 98.4 percent of claims that have been subjected to a risk assessment process but that have not been audited. Exhibit 5.4 shows GST/HST refunds paid by the Agency in the last five years.

5.42 Of the 1.6 percent (or 42,000) of refund claims audited in 2002–03, 262 were chosen with some degree of randomness from the 1.5 million claims identified by the Agency as presenting little or no risk of error. The audits of the 262 refund claims found that 42 of these were overstated by a total of about $1 million. The Agency audits claims selected with some degree of randomness to gain insights into the behaviour of registrants and verify how well the audit selection system works. The Agency could modify the selection process for these types of audits to a proper statistical basis and develop a comprehensive sampling plan with a suitable sample size, so that good statistical projections of the amount of overstated claims could be made.

Information on completed audits lacking

5.43 The Agency also needs better information on audit results in order to understand non-compliance. Currently, the Audit Information Management System captures a variety of information on audits, such as the reason for selecting the audit, the section of an act that was not complied with, and whether the auditor completed a full or restricted audit. However, capturing additional information, such as the reason for the audit adjustment and the auditor's judgment of the nature and cause of the non-compliance, would increase understanding of non-compliance. In addition, comparing the estimate of someone's potential tax at risk against actual audit results—as soon as the audit is completed—would help to quickly validate the system that produces that estimate.

5.44 Recommendation. To better understand the compliance behaviour of small and medium enterprises and thus determine the right mix of compliance activities, the Canada Revenue Agency should do the following:

Agency's response: The Agency concurs with the recommendation to develop a multi-year plan and set deadlines for the completion of the Core Audit Program's work and the analysis of the results. Steps have also been taken to improve the integrity of the data collected.

The Agency recognizes the vital importance of identifying risks to the tax base and of developing effective strategies. For example, it has a GST/HST compliance strategy that provides service and support to those who wish to comply with the law, while taking responsible enforcement actions to identify and deal with cases of non-compliance. As recognized by the Office of the Auditor General, the Agency has implemented various risk assessment processes to evaluate the validity of refund claims before they are paid. These processes will be tested and reviewed for improvements on an ongoing basis.

A range of measurement methods can be used to quantify compliance. It is internationally recognized that all methods have specific benefits, limitations, and utility. The development of statistically valid estimates of the overstatement of refund claims would require the diversion of significant resources from other compliance activities and the imposition of burden on compliant taxpayers.

The Agency agrees with the need to better understand the compliance behaviour of small and medium enterprises and will continue to utilize a variety of methodologies to identify compliance risks and to assess the effectiveness of its compliance programs. These include monitoring tax data relative to indicators of socio-economic change, analyzing program data and results to identify compliance trends and risk factors, conducting random sample reviews to measure non-compliance in specific areas, evaluating the achievement of our programs to ensure that their objectives are being achieved, and conducting surveys to obtain the public's views on compliance. The Agency will also continue to carefully review studies undertaken by the academic community and work by other tax administrations to further inform our understanding of the compliance challenges facing tax administration.

The Agency agrees that additional information on audit results would be beneficial and we have identified possible improvements to the current information reporting systems. Two major projects have been initiated to improve the availability, consistency, and reliability of data, including capturing non-compliance information to refine our risk-profiling capability and better target compliance actions.

While additional information from audit results is useful for understanding reasons for non-compliance and methods for detection, it provides limited insight into taxpayer behaviour. Accordingly, the Agency also uses public opinion research to contribute to its understanding of the psychology of tax compliance and the factors affecting taxpayer behaviour.

Risk evaluation

5.45 Focussed audits directly reduce the burden on compliant businesses and try to ensure that the same rules apply to all. The key element of a focussed audit program is targeting areas of high risk. This involves identifying high-risk sectors and selecting high-risk taxpayers within these sectors. The Agency has several systems designed to assess the risk of non-compliance in the small and medium enterprise population.

The Computer Assisted Audit Selection System is useful

5.46 The Computer Assisted Audit Selection System evaluates the risk of non-compliance in individual, corporate, and GST/HST returns. For each income tax return filed and for each GST/HST registrant, the System calculates the amount of potential tax at risk. The System then classifies taxpayers as a low, moderate, or high risk for non-compliance and highlights potential areas of non-compliance. Studies completed by the Compliance Research and Business Management Directorate to validate and improve the System's results show a positive relationship between its estimate of tax at risk and actual tax reassessed when a taxpayer is audited. These studies also conclude that there is still a need for an experienced auditor's judgment in assessing the risk of non-compliance in any particular file.

5.47 Currently, the System estimates the tax at risk for each separate income tax and GST/HST return filed. That means that returns filed by members of groups that are not operating at arm's length are viewed separately. And, instead of estimating tax at risk for the whole group, when the potential tax at risk may well be greater than the sum of each return, the System makes individual estimates. In addition, the Agency is not applying the tax-at-risk calculation to large partnerships and trusts.

Validation required for criteria used to select GST/HST refund claims for review

5.48 Before paying a GST/HST refund claim, an automated processing system evaluates each return against a series of risk criteria. About 60 percent of refund claims pass all the selection criteria and are paid automatically. The other 40 percent, which account for about 70 percent of the dollar value of the refund claims received by the Agency, go to tax services offices for further appraisal. Available resources in the tax services offices allow 4 percent of the claims referred, or 1.6 percent of all claims submitted, to be audited.

5.49 Currently, the risk criteria used to refer refund claims to tax services offices result in many offices receiving a lot more claims than they can handle. A study by the Agency found that the number of referrals to tax services offices was unmanageable. Because auditors must review a large number of refund claims, they review each claim more quickly, thereby increasing the possibility of approving an unsupported or fraudulent return without doing an audit.

5.50 The Agency uses the amount being claimed as one of the risk criterion. It is the criterion responsible for most of the referrals of claims to tax services offices. Headquarters' staff has encouraged tax services offices to consistently set and maintain acceptable claim amount limits for each registrant. Despite this, the number of claims referred to tax services offices remains high.

5.51 The Agency has made changes to the risk criteria in the last two years in an attempt to make them more discriminating. The Agency intends to evaluate the current risk criteria. The effectiveness of the GST/HST System's selection criteria and of the screening activities in the taxation services offices can be measured by how often audits result in a change to the amount claimed. In 2002–03, auditors changed the amount claimed in 56 percent of the refund claims audited. In 2001–02, that amount was 55 percent.

5.52 In the last two years, special teams in each region and in selected taxation centres have also begun developing profiles of potentially fraudulent GST/HST registrants and following up on what they consider high-risk enterprises to try to identify fraudulent GST/HST refund claims before they are paid. The Agency has not yet measured the effectiveness of these activities.

5.53 The Agency has recently strengthened one of its procedures to deal with GST/HST refund claims. In the past, when a registrant could not be contacted or failed to respond to an auditor's request for information to support a claim, the auditor would simply deny the refund. Unscrupulous individuals could quickly realize that few refund claims were audited and, if selected for audit, they could abandon their claim with no consequences. New procedures require tax services offices to follow up on registrants who fail to contact the Agency when requested to do so. Agency officials have yet to measure the effectiveness of the new procedure.

5.54 In March 2003, the House of Commons Standing Committee on Public Accounts held hearings on GST/HST fraud and the reporting of losses of public money and property. The Committee's report contained five recommendations. Among them: the Agency should inform the Committee of recent initiatives and pilot projects designed to strengthen GST/HST compliance and enforcement activities. In response, the Agency described several activities but did not provide information on how it plans to measure the results of those activities.

5.55 Recommendation. The Canada Revenue Agency should improve its systems to identify non-compliant small and medium enterprise taxpayers. In particular, the Agency should

Agency's response: The Agency's Computer Assisted Audit Selection System currently links non-arm's length returns (for example, corporations and their controlled companies, and corporations and their shareholders). The Agency agrees to evaluate linking trusts and large partnerships in risk assessment processes, and evaluate assessing the risks of related groups.

The Agency recognizes the importance of undertaking regular studies to validate the effectiveness of computer-based and manual risk evaluation systems. Regular studies that measure the effectiveness of the Computer Assisted Audit Selection System and other initiatives are already performed, and will continue. The studies include analyzing anomalies in individual cases, modelling new criteria for identifying non-compliance, validating and refining existing criteria, and integrating new sources of data into the risk assessment systems.

Design and delivery of audit activities

Staffing of tax services offices could better reflect risk

5.56 With a few exceptions, the number of auditors assigned to a tax services office is proportional to the number of the different types of taxpayers served by that office. In that way, in all areas of the country, taxpayers within a given segment have an equal chance of being audited. However, information in the Computer Assisted Audit Selection System indicates that the average potential tax at risk is different in different regions. Some tax services offices must deal with taxpayers who generally present a higher risk of non-compliance, yet those offices have no more auditors than those where taxpayers appear more compliant.

5.57 As well, the Agency has little information on the complexity of taxpayers' returns. It estimates the time and the experience level of the auditor needed to complete an audit based on the gross revenues reported by the taxpayer. However, the complexity and the risks inherent in a taxpayer's return, rather than the revenue level, increase the time needed and the level of experience required to do the audit. There is a danger that, in areas of the country where there is more risk and complexity, auditors will audit a smaller percentage of risky and complex files.

5.58 Recommendation. The Canada Revenue Agency should study the possibility of assigning complexity ratings to small and medium enterprise files and of allocating audit resources to tax services offices to reflect the profiles of their files and their tax-at-risk estimates.

Agency's response: The Agency agrees that complexity ratings and tax at risk are important elements in allocating audit resources. Its risk assessment system permits the identification of tax compliance risks for all individuals, businesses, and GST/HST registrants who file income tax and GST/HST returns.

The majority of our audits are focussed on files with the highest risk of non-compliance. We also consider other important factors in allocating resources. These factors include maintaining a credible audit presence and high visibility throughout the country, meeting commitments to the provinces and territories to protect their portion of tax revenues, and providing service and support to those who wish to comply while taking enforcement actions to identify and deal with cases of non-compliance.

The Agency's current allocation process involves, among other things, segmenting the self-employed and business population along various lines, including type of tax (income and GST/HST), structure (individual and corporate), and size of entity. Within a given segment stratified by gross revenue range, and within the attributes related to that segment, the allocation is based on the population of that segment. For small and medium enterprises, the Agency's view is that segmentation by revenue ranges provides a fair, if not perfect, proxy for complexity at reasonable cost. Notwithstanding, the previously mentioned studies on a compliance measurement framework and tax collection methodologies will assist the Agency in enhancing its overall risk management and allocation of resources.

Meeting coverage rates drives audit activities

5.59 The strategy of the small and medium enterprise audit program is to audit high-risk taxpayers and to maintain selected levels of audit coverage for different groups of taxpayers. These are competing priorities because, as mentioned earlier, auditing the returns of high-risk taxpayers is generally more complex and takes longer to perform. Therefore, an office that performs a greater proportion of high-risk audits can perform fewer audits overall.

5.60 Currently, the Agency is putting emphasis on meeting the coverage targets agreed to with the Treasury Board Secretariat as a condition for additional funding. To meet these targets, tax services offices plan a mix of full compliance audits and restricted audits. Restricted audits focus on one to three issues and their design does not permit auditors to make an overall evaluation of the taxpayer's compliance with tax laws. Because full compliance audits take more time to complete than restricted audits, increasing the percentage of restricted audits increases audit coverage. National guidelines state that up to 50 percent of audits carried out in a tax services office can be restricted. To meet coverage rates, auditors may pass over taxpayers presenting greater potential tax at risk for taxpayers presenting a risk of non-compliance in a specific area—such as claiming higher than average expenses of one category or another. Despite the 50 percent limit placed on restricted audits, several taxation services offices exceeded this limit.

5.61 The Compliance Research and Business Management Directorate conducted two studies of audit results of individual tax returns for 2000 and corporate tax returns for 1997. The studies showed that the chance of a taxpayer being audited increased as the estimate of potential tax at risk increased. In fact, individual and corporate tax returns in the highest tax-at-risk ranges were 10 times more likely to have been audited than tax returns in the no-risk category.

5.62 Still, the studies noted that 40 percent of all taxpayers audited were identified by the Computer Assisted Audit Selection System as having little or no potential tax at risk. These audits of tax returns were initiated as a result of leads, taxpayer-requested adjustments, or special audit initiatives. On average, the studies show that when the results were subsequently compared to reassessments of higher-risk files, as identified by the System, they were three to four times lower. While the studies concluded that increased audits of taxpayers with a higher potential tax at risk would increase tax recoveries, the Agency believes that it needs to address these other workloads to achieve overall compliance with tax laws.

5.63 Discussions with audit section managers and team leaders in tax services offices across the country confirmed the importance that they place on meeting annual audit coverage targets set for their offices and regions. Team leaders evaluate auditors on their performance in relation to audit plans, audit tests, technical issues, quality of correspondence and reports, professionalism, and timeliness. In assessing an auditor's performance, substantial weight is also given to whether the auditor has completed the expected number of audits.

Audits need more rigour

5.64 Quality assurance reviews. The Agency has conducted national quality assurance reviews of the small and medium enterprise audit program to ensure that the audit work meets a national standard of quality. Experienced auditors examine audit files in tax services offices across the country and rate them against a number of criteria.

5.65 We looked at a cross-section of quality assurance reviews conducted in the last three years to see whether audits were meeting the Agency's quality standards. Specifically, we focussed on areas where, as a result of discussions with audit team leaders, we felt auditors might not be meeting those standards partly due to their desire to meet production targets. We looked at 16 quality assurance reports covering 891 small and medium enterprise audits performed by auditors in tax services offices across the country. Some of the findings merit further review.

5.66 Lack of audit plans. There were no written audit plans for 33 percent of the audits reviewed. With the focus on the ongoing evaluation of risk, a plan is essential to document the potential areas of risk that caused an auditor to select a particular file and the planned procedures to address those risks. As the audit unfolds and the initial evaluation of risk is confirmed, amplified, or refuted, there is a clear basis from which to modify audit procedures.

5.67 Compliance reviews. National guidelines require that auditors conducting audits of income tax or GST/HST complete a brief review of that taxpayer's compliance with the other tax. These are called compliance reviews. They are required because, except for the smallest taxpayers, auditors who specialize in only one of the taxes perform either income tax or GST/HST audits. Performing compliance reviews is one way the Agency is trying to more efficiently audit both taxes while minimizing disruption to the taxpayer. The quality assurance results show compliance reviews missing in 38 percent of the audits that should have had them.

5.68 Indirect methods of verifying income. Auditing small enterprises requires special techniques because they often lack the sophisticated internal controls on which auditors can rely to evaluate the accuracy of amounts reported. This is particularly true in industry sectors where cash transactions are common. To conduct a thorough audit, national guidelines instruct auditors to use indirect methods of income verification, including a technique called net worth audits. The quality assurance reviews found that auditors had not used net worth or other indirect methods of income verification in 55 percent of the audits that merited them.

5.69 Team leaders in many tax services offices told us that using indirect verification techniques takes time. They said that a net worth audit takes from 150 to 300 hours and that spending that much time on an audit would prevent them from meeting their coverage targets. In some tax services offices, team leaders cited lack of training as a reason for their auditors not using indirect methods of verification. Headquarters staff told us that a new comprehensive course on indirect methods of verification soon would be offered to auditors in the taxation services offices. Still, this will not address the issue that these audits take time.

5.70 Penalties not always considered. Finally, the quality assurance reports we reviewed found that auditors failed to consider the application of gross negligence penalties in 54 percent of the audits where such penalties may have been warranted.

5.71 Recommendation. The Canada Revenue Agency should

Agency's response: The Agency agrees to review the frequency and the circumstances in which it uses restricted audits to achieve compliance to determine whether the current policy of up to 50 percent restricted audits is too high or too low.

Restricted audits play a key role in increasing the Agency's visibility through increased coverage levels. This, in turn, enables it to maintain an audit presence throughout the country that is fundamental to its underlying compliance strategy.

Auditors continually conduct a risk assessment of files throughout the course of audits and tailor audit plans accordingly. Consequently, most audits are restricted to some degree. More comprehensive audits are undertaken in cases where serious non-compliance is encountered.

The Agency is committed to the conduct of high quality audits and has several measures in place to assess quality. Team leaders are responsible for ensuring that national standards of quality are achieved in every file. Technical advisors ensure that all issues are identified and assessed in a technically accurate manner.

The Agency's Quality Assurance program contributes to the attainment of high standards of quality outputs through quality assurance reviews. Quality Assurance plays a key role in identifying training needs and the application of best practices. Program-monitoring visits are regularly conducted by headquarters' staff to ensure that program policies and guidelines are being followed on a consistent basis. The Agency will continue to ensure that quality standards for carrying out audits are achieved.

The Agency has performance criteria for auditors that encompass quality standards as well as the completion of a reasonable number of audits. We agree to continue to review the current performance criteria and to emphasize the importance of using a full range of qualitative standards in reviewing an auditor's performance.

Legislated penalties

5.72 Tax legislation provides for penalties in cases of non-compliance. The Agency views a graduated penalty structure as a key component of a fair system that promotes compliance. In 2001, it planned to consult with Finance Canada to ensure that the penalty structure properly recognized prior good performance and imposed more severe penalties for systematic evasion and non-compliance.

5.73 The Agency told us that, as a result of preliminary discussions with Finance Canada, a decision was made not to review the income tax penalty structure at this time, with one exception. In July 2003, the Agency revised the penalty for all instances of late remittance of source deductions from 10 percent of the amount of the late remittance to a range of 3 percent to 10 percent, depending on how late the remittance is. While our audit did not set out to look at the Agency's compliance program that deals with source deductions, we found that the Agency lacks the legislative authority to implement the reduced penalty.

5.74 Penalties are one of many measures aimed at managing the risk of non-compliance. Reviewing the costs and benefits of different penalties is an integral part of maintaining an up-to-date risk management strategy. The following are two instances where the Agency could benefit from a review of the current penalty structure:

5.75 The Agency's risk assessment systems highlight situations where a GST/HST registrant is claiming a refund but is in default regarding the obligation to file a tax return. In 2003, in the southern Ontario region alone, over 20,000 GST refund claims were referred to tax services offices for review because registrants might not be current in filing their tax returns. At present, the Excise Tax Act does not allow the Agency to refuse a GST/HST refund to individuals and corporations for not having filed their income tax returns.

5.76 Recommendation. The Canada Revenue Agency should

Agency's response: The Agency held extensive discussions with small and medium enterprises on the penalty structure. The predominant issue identified was the penalty of 10 percent on all late remittances of source deductions. The concern expressed was that this penalty did not properly recognize prior good performance and did not provide an incentive for paying the late remittance at the earliest possible date. Our current pilot project on the graduated penalties will allow the Agency to respond to this issue.

The Agency agrees that GST/HST refunds should be denied to registrants who are late in filing their income tax returns. It is reviewing both its GST/HST and income tax systems to ensure that there is an effective and efficient manner to identify these situations before GST/HST refunds are paid. It has also commenced discussions with Finance Canada on a legislative response.

The graduated penalties pilot project currently running is designed, among other things, to identify the implications of reductions in these penalties, which will assist in drafting legislative amendments. The Agency is working with Finance Canada to identify an appropriate legislative response to this issue.

Conclusion

5.77 The Agency is unable to provide us with an overall risk analysis to support its current resource allocation. Neither does it have sufficient information with which to make decisions on the required mix of compliance activities. As a result, it is difficult to determine whether its current resourcing decisions are most appropriate for each of its compliance activities, including its small and medium enterprise audit program.

5.78 The Agency uses computerized risk assessment systems, together with experienced auditors, to identify non-compliant taxpayers. Improvements to the current systems would increase their effectiveness.

5.79 The Agency's audit strategy for small and medium enterprises seeks to target taxpayers suspected of significant non-compliance. It also seeks to serve as a deterrent against non-compliance through its presence. However, tax services offices are staffed based on the population of different segments of taxpayers rather than on the risk or complexity that population presents. In addition, the current focus on audit coverage to meet targets agreed to with the Treasury Board Secretariat as a condition for funding reduces efforts to address cases of significant non-compliance. The Agency needs to evaluate the effectiveness of its strategy.

5.80 The Agency's own quality assurance reviews show weaknesses in the conduct of audits. The Agency needs to understand the reasons why auditors are not meeting the quality standards set for audits of small and medium enterprise taxpayers and needs to take corrective measures.

5.81 The Agency must gather additional information on audit results and analyze that information to better understand taxpayer behaviour, reasons for non-compliance, and methods of detecting non-compliance.

About the Audit

Objectives

The objectives of the audit were to assess how well the Agency

Scope and approach

The audit covered the activities of the Compliance Research and Business Management Directorate concerned with small and medium enterprise compliance and the functional guidance provided by the Audit Directorate headquarters staff. It also included certain activities of the non-registrant and non-filer program within the Assessment and Collections Branch. The audit looked at the tax services office activities related to income tax and GST/HST audits of small and medium enterprises as well as activities aimed at getting individuals and corporations in the small and medium enterprise category to file their tax returns and register for the GST/HST.

We reviewed the Audit Directorate's plan for managing the risk of non-compliance in the small and medium enterprise population. We also reviewed the Compliance Research and Business Management Directorate's efforts at measuring the extent of non-compliance in the small and medium enterprise population and its systems designed to identify tax at risk for each taxpayer and GST/HST registrant. We examined documentation and conducted interviews at headquarters in Ottawa and in selected taxation services offices across the country.

We examined the Agency's quality assurance activities and reviewed quality assurance reports to gauge whether auditors complied with Agency standards in carrying out their audits.

The scope of our audit did not include other specialized Compliance Branch programs that cover all taxpayers, including small and medium enterprises. These programs include tax avoidance, international tax, scientific research, and other tax incentives. Nor did we look at the appeals program for reviewing contested audit decisions or at the tax collection activities undertaken for small and medium enterprises. Many of these programs have been subjects of past audits. Also excluded from the scope of this audit was the administration of GST in Quebec by the Ministère du Revenu du Québec.

Criteria

We would expect the Agency to

Audit team

Assistant Auditor General: Andrew Lennox
Principal: Barry Elkin
Directors: Denis Labelle (lead), John Pritchard, Brenda Siegel

Wilson Ford

For information, please contact Communications at (613) 995-3708 or
1-888-761-5953 (toll-free).


Definitions:

Goods and services tax (GST)—The tax on most goods and services consumed in Canada that became effective 1 January 1991. Basic groceries, residential rents, and most health and educational services are exempt. (Back)

Harmonized sales tax (HST)—The single tax that replaced the GST and provincial sales tax in Nova Scotia, New Brunswick, and Newfoundland and Labrador on 1 April 1997. (Back)

Potential tax at risk—The difference between how much income tax and GST/HST a person or corporation reports they should be paying and what the Agency estimates those amounts should be. The Agency uses a series of criteria and other information to arrive at its estimate. (Back)