This Web page has been archived on the Web.

1996 September Report of the Auditor General of Canada

Main Points

19.2 Each year under these programs about $8 billion in benefits flows tax-free to over 8 million recipients. Most are families with low or modest incomes.

19.3 Revenue Canada faced a significant challenge to launch the GST credit on time in 1990 and the child tax benefit in 1993, given the short lead times available to meet statutory deadlines. For child tax benefits, the challenge was shared with HRDC. Notwithstanding some important weaknesses in the initial systems, we were impressed with the ability of both departments to publicize the programs and meet the statutory deadlines. Since the programs were launched, millions of families and individuals have been receiving monthly or quarterly social benefit payments. To date, over $30 billion has been paid out under these programs.

19.4 Revenue Canada has been working to improve service and reduce costs by streamlining its administrative processes. Despite some successes, we have four concerns about the administration of these programs. First, we found that service needs to be substantially improved in some key aspects of delivering the programs, namely in providing access to the Department by phone and in processing child tax benefit applications. Second, the control environment needs to be significantly strengthened to provide reasonable assurance that only those who are eligible receive benefits and receive them in the correct amounts. (This was also the case with HRDC while it had responsibility for determining basic eligibility for child tax benefits, until August 1995.) Third, Revenue Canada has not developed a formal strategy for detecting and preventing fraud and abuse in the programs. Fourth, the Department does not monitor the programs to assess whether the total amount of benefits flowing from each program is reasonable in the aggregate, considering changes in population characteristics and family structure. Our analysis of program statistics against independent data on key variables suggests potentially serious problems that require immediate attention.

19.5 The Department of Finance is responsible for assessing the extent to which program policy is producing the desired outcomes. Despite their social significance and the dollars involved, neither program's performance has been evaluated against its stated objectives. Further, neither program is scheduled for evaluation in Finance's current three-year plan.

19.6 Although child tax benefits and GST credits are key components of Revenue Canada's income redistribution line of business, Parliament receives little information on the Department's performance and priorities in administering the two programs. The little information that is reported in the Department's Part III of the Estimates needs to be strengthened and improved.

Introduction

New programs represent an $8 billion annual share of Canada's social safety net

19.8 The two programs were introduced for different purposes. The GST credit was instituted to help families and individuals with low or modest incomes offset all or part of the GST that they pay. The child tax benefit was introduced to replace the former patchwork of child benefits delivered through the Family Allowance, Child Tax Credit and Dependant Tax Credit programs. Its objectives were to help families meet the cost of raising children and to be more responsive to the needs of families with low and moderate incomes by providing a unified child benefit that would be simpler, fairer and more generous to such families. Although the two programs have distinct objectives, both provide a cash supplement to families that increases with family size and diminishes when family income rises above a given threshold.

19.9 Until August 1995, Human Resources Development Canada (HRDC) and Revenue Canada jointly managed the child tax benefit but had distinct roles. Basic eligibility for benefits was determined by HRDC, and Revenue Canada was responsible for all other aspects of the program - primarily establishing how much individual recipients were entitled to receive, issuing benefit payments, maintaining participant benefit records (benefit rolls) and handling all telephone enquiries. On 28 August 1995, Revenue Canada assumed full responsibility for administering the child tax benefit, including determining basic eligibility for benefits. In the case of the GST credit, Revenue Canada has been responsible for all aspects of administration since the program's inception.

19.10 The Department of Finance is responsible for setting fiscal policy and evaluating the extent to which the two programs are producing the desired outcomes. Human Resources Development Canada continues to be responsible for social policy in general, which includes promoting and strengthening the income security of families with children. Changes in social policy could therefore have implications for future child tax benefits.

19.11 For a number of reasons, both of these programs warrant Parliament's interest. First, they represent a significant share of total federal direct social transfers to Canadians, particularly low- and modest-income families with children. Second, they represent a substantial shift in the way the federal government provides cash transfers to individuals. They rely on the tax system and its associated administrative machinery to assess program eligibility, determine entitlements and deliver periodic benefits to families. Third, unlike the majority of the Department's programs, these are not revenue programs but expenditure programs and hence present additional challenges and aspects of accountability for Revenue Canada.

Over $30 billion tax-free paid to date

19.13 To receive benefits under either program, participants must file a tax return and, for child tax benefits, so must the married or common-law spouse. Legislation requires individuals to apply if they want to receive benefits under these programs. The application for GST credits is found in the body of the personal tax return. To receive child tax benefits, the primary caregiver must apply for each child using a separate application form. Those children who were registered as eligible under the family allowances program at the end of 1992 were automatically included.

Marital status, family income and number of dependants determine benefits

19.15 An important feature of both programs is that the benefits, which are tax-free, are based on net income of the family, not of individuals. Family net income is the combined net income for tax purposes of the recipient and his or her married or common-law spouse. This measure of family income is important for both programs, as benefits are "targeted" to provide maximum assistance to families with net incomes of $25,921 or less. Benefits decrease when family net income exceeds this threshold. Both types of benefit may include an additional supplement. For the child tax benefit, the supplement is meant to reinforce incentives for low-income families to participate in the work force and to offset some of the expenses associated with holding a job; the maximum additional supplement for a family is $500 per year, expected to increase to $750 in 1997-98 and $1,000 thereafter. For GST credits, the supplement reaches a maximum of $105 per year, and is available only to single adults with low incomes. It is meant to recognize the extra costs that most single people incur, such as maintaining a household on their own.

19.16 Both programs follow a July-to-June benefit year. Each July, Revenue Canada recomputes the amount a family is to receive in child tax benefits during the new benefit year. The amount is based on the prior year's family income as assessed through the tax system and on the current family structure. For GST credits, benefits are computed according to the family income and family structure that existed on December 31 of the previous year.

19.17 The person primarily responsible for caring for and raising a child - usually the mother - receives the child tax benefit payment. However, the father or a grandparent or guardian could receive it instead. Benefits continue from the birth of a child until he or she turns 18, depending on family income levels and changes in family structure. For the GST credit, any resident of Canada who is 19 years of age or older (married or single) or who is a parent can apply. Recipients of GST credits must apply each year and indicate the number of dependants they have in order to continue receiving benefits.

19.18 With both child tax benefits and GST credits, the law provides for increases to basic benefits when inflation exceeds three percent. To date, this has not happened. As well, in the case of child tax benefits, provinces can modify the benefit structure within specified limits to better suit their needs. Both Quebec and Alberta have done so.

Focus of our audit

19.20 Our overall approach was to focus on key aspects of program administration for both GST credits and child tax benefits. In the latter case, the transition from joint administration to the transfer in August 1995 of responsibility for program delivery to a single department necessitated that we look at the program's administration from the global perspective of the federal government; we note that child tax benefits have been in place since January 1993. Further details on our audit objectives, approach and criteria are presented at the end of the chapter, in About the Audit .

Observations and Recommendations

Facilitation

The launches of both programs were well managed

19.23 The systems required to support both programs are complex. First, they must be able to deal with a huge volume of eligibility and payment transactions (about 70 million payments per year). Second, the processing itself is based on elaborate entitlement formulas in an environment where the status of recipients is subject to change, and where cross-matching an individual's taxpayer records with those of other family members is crucial. Coupled with the demands of those systems was the need for new systems and practices to establish eligibility criteria and, for child tax benefits, to identify a child's primary caregiver - a first in federal child assistance. Strategies were needed to publicize the programs and to inform Canadians about how to apply for and receive the new benefits. Finally, the systems also had to be able to respond to enquiries, complaints and requests from millions of Canadians.

19.24 We were impressed with Revenue Canada's and HRDC's separate and joint ability to respond to the challenge of launching the programs successfully in the short lead times available to meet statutory deadlines. Moreover, since the launches, these systems have formed the basis of delivery mechanisms that provide millions of Canadians with their monthly or quarterly benefits.

19.25 In the early months following the launch of each program, however, some important systemic problems did arise. For example, before child tax benefits were implemented in 1993, it was recognized that there would be problems in determining family income to compute benefits, because of the need to rely on individuals' tax returns of the previous year for family status and related information thought to be potentially incomplete and inaccurate. Revenue Canada states that all departments involved in the launch (the Department, HRDC and the Department of Finance) agreed that this was a risk that had to be taken in order to implement the program by the legislated deadline. Thousands of child tax benefit cheques were paid out improperly during the first several months of the program. In response to the problem, recipients were asked to notify Revenue Canada if family status had been determined incorrectly. Individuals who had received benefits to which they were not entitled (many for six months or more) were asked to pay them back. (See paragraph 19.92 on processing benefits with missing or inconsistent information and paragraphs 19.114 and 19.115 on monitoring overpayments.)

Actions are under way to assess the needs of clientele

19.27 In disseminating information about both child tax benefits and GST credits, Revenue Canada generally uses most of the same channels of communication it uses for its revenue-related lines of business. For example, the Department issues monthly or quarterly press releases as benefit payments are issued and provides separate toll-free 1-800 service lines for each program. As well, notices of determination are sent each new benefit year to all program participants, informing them of the benefits they will be receiving in the upcoming year. The Department also includes information on GST credits and child tax benefits in its various income tax returns, guides and pamphlets. Some of the publications target specific groups and include summary information about the programs, geared to answer some basic questions for the intended audience. Others, such as the tax returns, are more general in nature. Yet, as suggested in a 1993 departmental study on client enquiries, publications may not be the best vehicle for meeting the information needs of program clients. The study indicated that over 40 percent of GST credit and child tax benefit recipients surveyed who had contacted the Department about the programs were not aware of the existence of program publications that might have satisfied some information needs. Of those who were aware of the publications, over 40 percent were either not using them or had trouble understanding them.

19.28 To ensure a general level of awareness about the programs, the Department believes that providing information on the programs in tax return packages is an effective way to reach most Canadians. It has informed us that for the 1993 tax year, for example, almost 95 percent of all Canadian residents between the ages of 20 and 54, and 91 percent of older Canadians, filed returns and therefore are believed to have had at least some access to information about the programs. The Department has also been working to streamline and simplify the design of the tax packages to facilitate their use by a broad range of filers who have unique needs. For example, it has simplified return packages for lower-income seniors who file primarily for GST and provincial credits, for dependant spouses who file only for child tax benefits or other credits, and for Aboriginal people who have exempt income for benefit determinations. The Department has incorporated the annual application for GST credits on the personal tax return, making it simpler for many Canadians to continue qualifying for them.

19.29 Beyond the tax return packages and other publications, some regionally led initiatives are also under way to get more information on Revenue Canada services to a few local groups that serve people who, because of language and other barriers, may be less aware of the programs. One local office is also pursuing preliminary discussions with a provincial welfare agency to explore opportunities for sharing information on the programs and the potential for using common delivery networks.

19.30 Despite these activities, we are concerned that most of the efforts at communication to date, although useful, may not be sufficiently responsive to the needs of most eligible program participants. During our examination, program officers consistently indicated that the programs serve a clientele that differs markedly from those of Revenue Canada's other lines of business. For example, we were told repeatedly that clients of the two programs are heavy phone users who have many questions and often require a great deal of reassurance about their benefits, but who have major problems in contacting the Department by phone.

19.31 During our audit, we wanted to obtain specific information on the profiles of current participants in both programs, and on their specific information needs. We found that Revenue Canada did not have this kind of information available. We also found that it had not tested nation-wide the level of awareness about the programs, either in major centres or in remote communities, as a means of measuring the success of its overall approach to communication and program delivery. In the absence of this kind of information, we believe there is an added risk that the programs may be missing potential recipients, especially those who may be most in need of assistance.

19.32 Subsequent to our audit, Revenue Canada informed us that it had engaged outside consultants for a preliminary survey of GST credit and child tax benefit recipients who call the Department, to profile benefit recipient callers as well as to evaluate the effectiveness of the information they receive from other sources and to identify areas of communication that could be improved. A draft report was being considered by management.

19.33 Revenue Canada should review the current communication strategy for child tax benefits and GST credits, evaluate whether it is meeting the information needs of eligible and existing program participants and institute measures as required to ensure that communications are appropriate to support effective program delivery.

Department's response: Since the inception of the Child Tax Benefit and GST Credit programs, the Department has adopted effective communication strategies to address client needs. As the programs evolve, these strategies will continue to be evaluated and modified as appropriate.

Take-up rates are unknown

19.35 Program take-up provides decision makers with an independent measure of a program's success in reaching those who have been targeted for benefits. Further, as paragraph 19.86 elaborates, the take-up rate is also important for gauging the extent to which people are receiving benefits to which they are not entitled. Taken together with information on program enrolments, it provides an independent benchmark against which to measure the level of "leakage" or abuse in a program.

19.36 We found that neither Revenue Canada nor HRDC has measured the take-up rate for child tax benefits and/or GST credits. While it has not defined targets for take-up as indicators of program effectiveness, Revenue Canada has informed us that its implied target for take-up is 100 percent of all entitled participants except those who have chosen to opt out of the program. Its plans for the future include determining a method to measure the extent to which all truly eligible people are (or are not) receiving the benefits to which they are entitled.

19.37 The Department's inactivity to date in the area of determining take-up for either program, together with that of HRDC for child tax benefits, is worrisome from the standpoint of both facilitation and enforcement: these two benefit programs are among the largest at the federal level that make social transfer payments to families and individuals. They involve large expenditures, and target benefits primarily to those who are most in need. In an environment where individuals must complete applications and file tax returns before receiving benefits, and where controls over the processing of benefits are weak (see paragraph 19.78), in our opinion there could be a significant risk either that coverage of eligible participants is piecemeal or that families and individuals receive benefits for which they are not eligible and/or entitled.

19.38 The American experience with its Earned Income Credit Program, which has some similarities to our Child Tax Benefit Program, illustrates that take-up can be problematic even in a mature program that is widely publicized. Although the earned income credit has been in existence far longer than our child tax benefit, researchers in the U.S. have estimated that somewhere between only 75 and 86 percent of all entitled families actually received the credit for tax year 1990. This is an estimate of the true take-up rate, not inflated by ineligible participants.

19.39 There are some important differences in scope and practices between the earned income credit and the child tax benefit that have implications for program take-up. For example, the risk that benefits in Canada miss as many entitled families as those in the U.S. is reduced somewhat, because families who were receiving the universal family allowance were automatically included in the new program. However, children born since family allowance ended in December 1992 are not automatically included; consequently, fewer families may be participating than are eligible. Nonetheless, as paragraph 19.84 discusses, we expect that take-up for child tax benefits is higher than that experienced for earned income credits in the U.S.

19.40 We did some analysis focussing on the population of children born after the family allowance program ended in 1992. In paragraphs 19.81 to 19.88, we describe how different assumptions of the true take-up rate by families for children born after 1992 can lead to significantly different conclusions about levels of abuse or other non-compliance in the Child Tax Benefit Program.

19.41 We were unable to do a similar analysis for GST credits: there is insufficient detail in departmental records, and suitable third-party information is not available to make an analysis possible. We therefore cannot comment on the reasonableness of the level of program participation and/or quantify the extent of potential leakage in program delivery.

19.42 For both child tax benefits and GST credits, Revenue Canada should develop a methodology for determining actual program take-up rates and institute procedures for measuring and reporting periodically thereon.

Department's response: This is addressed in the Department's action plan. (See Special Insert )

Families may not be aware of their legal obligation to report changes

19.44 Recipients' reporting obligations are listed on the child tax benefit application form. However, primary caregivers complete the form only once in an 18-year period for a particular child. There are also children for whom no application was required - those who were included in the Child Tax Benefit Program from family allowances. Currently, they represent the majority of children on the child tax benefit roll. Since they did not have to complete an application form, there is a possibility that families with older children may be unaware of their reporting responsibilities. We also observed that even the monthly benefit payments do not clearly note recipients' reporting obligations. Although the notice that accompanies the monthly child tax benefit does contain a reminder on the back, it does not mention that there are penalties for failure to report changes.

19.45 In our review of practices of other jurisdictions administering broadly similar programs, we found that they place considerable emphasis on communicating recipients' responsibilities to report changes that could have an impact on entitlements. Formal processes have been established in these jurisdictions to collect this information. For example, the processes require recipients to confirm their eligibility annually in writing or to confirm at least biennially their understanding of reporting obligations that affect eligibility. By way of contrast, there is no such established mechanism for reporting changes in eligibility for child tax benefits. In our opinion, the potential financial consequences may be serious. The Department indicates that recipients may use the general 1-800 number for child tax benefits to communicate changes in family status and structure that may affect their eligibility and entitlements. Assuming that those who are aware of their obligations to report changes would want to do so by phone, we note in paragraph 19.57 that their ability to reach the Department could prove difficult. As a result, even these groups of individuals may not report some changes.

19.46 In our view, considerable opportunity exists both to state more clearly and explicitly recipients' obligation to report changes in their status that will affect their benefits and to determine that recipients are aware of their obligations and report changes as required.

19.47 Revenue Canada should take measures to ensure that recipients' obligations to report changes in status affecting child tax benefits are clearly understood and that recipients are provided with appropriate vehicles for reporting such changes.

Department's response: Recipients' reporting obligations will be included in the Department's ongoing review of communications requirements.

Standards of service have yet to be defined

19.49 In keeping with the federal government's Service Standards Initiative, Revenue Canada has made commitments to develop service standards for its lines of business. However, none have yet been established for the GST Credit and Child Tax Benefit programs. In the case of child tax benefits, the Department has indicated that it would have been inappropriate to establish standards for processing applications, given that it has implemented new systems to capture eligibility information in the short period following the assumption of this responsibility from HRDC. However, we note that Revenue Canada has had complete responsibility for telephone enquiries about child tax benefits since October 1993 and, at the conclusion of our audit, there were still no formal standards of service in place.

19.50 We decided to review the Department's performance in two key areas of service to clients. We looked at application processing for child tax benefits and at telephone enquiries about both programs. Both activities deal with high volumes and are highly visible to the public. (Standards for processing GST credit applications are less critical since this is done in conjunction with the T1 returns, and typically the first payment is not made until July.)

19.51 The results of our review were disappointing. We found that performance in these two key service areas could be significantly improved.

Backlogs in unprocessed child tax benefit applications have added to delays

19.53 The time it takes to process an application can be longer than the expected average of two to three months. For example, we noted that after the transfer of eligibility determination to Revenue Canada in August 1995, the volume of unprocessed applications reached a high of 70,000 by November 1995. This backlog caused some clients to wait up to four months before receiving their first benefit cheque. The Department states that factors contributing to the backlog included the transfer of an inventory of 19,000 unprocessed applications from HRDC, and a steeper-than-anticipated learning curve among new staff who had to learn and operate both the HRDC and Revenue Canada systems for processing applications. A similar situation of backlog developed after the Department introduced a new system in 1996 for processing child tax benefit applications. The Department has since managed to process the original backlog and to reduce the new backlog by deploying supplementary resources and taking other actions as appropriate. In our opinion, however, some of these backlogs could have been avoided by more effective up-front planning of requirements to ensure a smooth transition without affecting service to the public. Subsequent to our audit, the Department informed us that the turnaround time for processing most applications and issuing benefit cheques had been reduced to about two months.

19.54 Given that a benefit payment can represent for many a significant part of family income, we are concerned that the elapsed time before transactions are reflected in benefit cheques may still be unreasonably long and inflexible. We note that legislation allows for some flexibility in that it does not require benefit cheques to be released on a specific date in a month, as is the current administrative practice. However, benefits may not begin sooner than the month following the birth of a child. Revenue Canada does have some flexibility under the present system to issue benefit cheques manually on a priority basis when clients are in urgent need of benefits. Some 25,000 cheques were issued on this basis during the last benefit year. We are also concerned about the present lack of management information in this area of processing benefit applications, which hampers the Department's ability to ensure at least a minimum overall level of service in processing turnaround times and to pursue improvements, as appropriate, in a timely and informed way.

Access to phone lines is poor

19.56 The Department has not been able to cope with the tens of millions of phone calls it receives each year about child tax benefits and GST credits. As illustrated in Exhibit 19.4 , in the last fiscal year alone roughly 16 million calls were placed about child tax benefits. Although separate statistics were not available on the number of calls about GST credits (they are combined with statistics on calls about personal income tax refunds, for a total of 29 million calls in the last fiscal year), our estimates suggest that calls inquiring about the credits amounted to at least as many as those for child tax benefits in the last fiscal year. In total, we estimate that calls about child tax benefits and GST credits outnumber by more than one third the calls on the Department's heavily used general enquiry lines.

19.57 We found that telephone service accessibility for the two programs is significantly inferior to the accessibility on general enquiry lines (our concerns about the general enquiry lines are noted in Chapter 14 of this Report, paragraph 14.54). For this last fiscal year, some 90 percent of calls placed on the child tax benefit and GST credit lines got a busy signal, compared with 72 percent on the general enquiry lines. During peak times in three major cities, on or about the payment date, less than 5 percent of calls placed about child tax benefits and GST credits ever got through to the Department, compared with 12 percent on the general enquiry lines during peak periods in those same cities.

19.58 A 1994 departmental report on calling patterns in Toronto and Montreal for the busiest month during the year indicated that, on average, each caller tried nine times to reach the Department but only slightly more than half succeeded in getting into the queue for eventual contact with an agent. Recent departmental statistics show that, on average, callers in the queue can expect to wait almost two minutes before speaking with an agent, but about one quarter of them hang up before they do. It is unclear whether those who hang up do so because the wait is too long or because they receive the information they need from the Department's voice messaging system.

19.59 Revenue Canada recently instituted voice messaging on phone lines for child tax benefits and GST credits as part of its effort to improve telephone service. Voice messaging provides the same prerecorded message about benefits and/or credits to all callers while they wait to speak to an agent. In addition, an automated enquiries response system was piloted in Montreal and Toronto in the winter and spring of 1996 that provided all callers with full access to a message about the benefit payment. The system accepts up to 200 callers per minute and directs callers to dial another line if they need further information. This system is meant to free agents to handle the calls that require their attention. However, the results of the 1996 pilot tests in Montreal and Toronto suggest that most callers still want to speak directly to an agent. Clearly, telephone access remains a problem area. Subsequent to our audit, the Department informed us that it has expanded the use of the automated enquiry system to two other cities. It also informed us that it has completed a client services re-engineering study on departmental telephone enquiries in general, in an effort to improve service overall through expanded use of automation. The Department believes that, at least for simple calls, automation will also help improve accessibility of service on enquiry lines for child tax benefits and GST credits.

19.60 It is noteworthy that Revenue Canada's corporate plan details the Department's general service strategy, which includes consulting with clients on the development of policy and on decisions affecting the design and delivery of its programs and services. It recognizes that an open, consultative relationship with its clients builds a common understanding that strengthens client relations and benefits program delivery. We note that consultative committees are in place for most of the Department's lines of business. However, none have been devised yet specifically for child tax benefits and GST credits, although, through the Department's established Senior Advisory Committee, there have been some consultations with seniors on the delivery of the GST Credit Program. In view of its performance in some key service areas, as noted above, we believe the Department could benefit from the use of a consultative committee in delivering both programs, particularly because they are relatively new and serve a clientele that in many respects is different from those of the Department's other lines of business.

19.61 Revenue Canada should establish standards of service specific to the delivery of the Child Tax Benefit and GST Credit programs and a mechanism for measuring and reporting to decision makers its performance in achieving these standards.

Department's response: The regular and timely delivery of child tax benefits and GST credits is Revenue Canada's primary and uppermost standard of service. The backlog of unprocessed applications described was a temporary situation which followed the transfer of work from one department to the other. For both programs, the Department meets its most important standard by efficiently processing 70 million monthly and quarterly benefit payments each year on time.

The departmental action plan outlines our implementation schedule for other service standards.

Practices vary in judging disputes over primary caregiver status

19.63 We noted that there is little quality control in this area; there is no overall monitoring of the practices followed; and decisions made by a program officer in these sensitive cases are typically not subject to peer or supervisory review, even on a test basis. As a first step in determining primary caregiver status, the Department uses a standard questionnaire (originally prepared and used by HRDC) that each disputing party must complete. It asks them to provide detailed information on the nature and level of care they give to a child. Program officers use the responses as the basis for judging the case. During our field visits, we became aware of inconsistencies in the practices that program officers follow when the primary caregiver is not obvious from the information supplied on completed questionnaires. Some officers may interview both disputing parties, after which they will render a decision. Other officers may contact third parties to confirm or add to the information received. Still others will stop child tax benefit payments from being issued until the disputing parties come to an agreement on their own. The Department informs us that officers are able to refer any difficult determination to headquarters for consideration.

19.64 As noted, the financial consequences for the affected parties can be significant, and - particularly if benefits are stopped - can also directly affect the well-being of the children for whom the benefits are being paid. The Department's records do not track the number of adjudication cases it deals with.

19.65 Individuals who disagree with the Department's determination of either their eligibility or their entitlements may file a notice of objection and have their case reviewed by an independent party. Appeals officers informed us that some of the most contentious cases are those dealing with determination of primary caregiver status for child tax benefits. However, as we found with the initial determinations, Appeals officers also use different practices in handling these disputes. While the number of formal objections received overall is relatively small, we are still concerned about the potential effects of inconsistent practices on the fair and equitable treatment of all participants in the program.

19.66 Revenue Canada should establish uniform policies for the determination of primary caregiver status in cases involving disputes among applicants, together with monitoring practices as appropriate for the review of such decisions.

Department's response: In the determination of these cases, all facts provided by both applicants must be reviewed on an individual basis in order to render a decision. Revenue Canada requires flexibility in practices to ensure fair and equitable treatment of the array of different family situations. The variety of practices observed by the Office of the Auditor General are examples of appropriate solutions being applied to different situations.

The guidelines will be reviewed to ensure that all applicable practices are incorporated. Appeals Branch will review the operating guidelines for Appeals officers to ensure consistency in resolving disputes.

Initiatives to streamline operations and reduce program delivery costs

19.68 As noted in paragraph 19.59, voice messaging and the automated enquiry response systems are new initiatives. Recently, Revenue Canada has also experimented with cross-trafficking of phone calls among call sites as a means to better manage its accessibility by phone when one call site reaches full capacity. This initiative was deemed successful and the Department is looking at expanding it. It is also exploring ways to further enhance flexibility in responding to telephone enquiries about the programs.

19.69 The Department has been successful in pursuing the federal government's initiative to expand the use of direct deposits. Direct deposits generate savings in administrative costs. As Exhibits 19.5(a) and 19.5(b) show, over the years the Department's efforts to promote the use of direct deposit have greatly increased the number of child tax benefit and GST credit recipients who use this method. Currently, one family in three who receive either type of benefit receives its payments by direct deposit. This represents some 24 million benefit payments each year, and has been estimated by the government to generate approximately $7 million in annual savings in postage, banking and paper costs.

19.70 During this past year, the Department has been implementing the Individual Credit Determination system (ICD) and the Child Credit Management system. The ICD system was introduced to consolidate two separate systems that were processing similar information to determine entitlements to child tax benefits and GST credits. The Department estimates that the redesign of two systems into one will save $18 million in maintenance costs over five years. The new system is also meant to provide the Department with the ability to administer any new future credits, whether provincial or federal, through the tax system should that be required.

19.71 Although Individual Credit Determination may achieve efficiencies for the Department in the long run, its launch was hampered by systems problems that disrupted service to affected recipients of GST credits and child tax benefits and that incurred extra administrative costs. For example, when the system was first used in July 1995 to compute entitlements for the new benefit year, it missed more than 40,000 clients of the GST Credit and Child Tax Benefit programs. Payments were subsequently issued and mailed to affected clients by month's end. As well, about 90,000 married or common-law recipients of GST credits had their benefits calculated as though they were single, which reduced their maximum supplement from $199 to $105; adjustment cheques were issued thereafter. The Department indicates that most affected clients received their adjustment cheques within two weeks of the original mailing and that system problems identified in this paragraph have now been corrected.

19.72 To address concerns raised about the ICD system, the Department established a steering committee for benefit program systems. An initiative under the steering committee was to review the stability of the ICD system. At the completion of our audit, this review was still under way. Our observations on controls over systems development are found in paragraphs 19.124 to 19.128.

19.73 The second system introduced by the Department within the last year was the Child Credit Management system (CCM), which replaced HRDC's system for processing eligibility determinations. It is expected to yield administrative savings of $2.3 million per year over the cost of the separate systems previously maintained by HRDC and Revenue Canada, which contained overlapping, duplicate information. This system, like the ICD, is meant to accommodate potential provincial requirements. And like ICD, the system had start-up delays when it was implemented in February 1996. These led to some delays in processing child tax benefit applications, creating a new set of backlogs that caused some clients to wait several months before receiving their first benefit cheques.

Verification and Enforcement

Controls are necessary to ensure overall integrity and fairness of programs

19.75 Verification and enforcement activities are critical in administering large and substantial benefit programs because of the social and financial risks involved. These programs need appropriate systems to verify an individual's eligibility for benefits, to verify the amount of benefits involved and to detect and prevent fraud and abuse.

19.76 Our observations on the Department's performance in the areas of verification and enforcement lead us to conclude that Revenue Canada's existing control environment is not adequate to ensure equitable program delivery, and may be exposing the public purse to losses. These observations are equally valid for HRDC in its area of responsibility until August 1995 - determining basic eligibility for child tax benefits. We believe that verification and enforcement require much more attention than they have received to date. The current exposure to financial risk is large. The potential for errors, abuse and fraud to go undetected is high. During the audit, Revenue Canada indicated that it was in the process of finalizing a new divisional structure that would give a higher priority to its benefit programs, and would include a section dedicated to quality assurance as well as compliance activities. At the conclusion of our audit, work on this division was still under way.

Fundamental checks and balances are lacking

19.78 However, our examination also revealed five areas in which we believe fundamental checks and balances are lacking for GST credits and/or child tax benefits:

• enrolling children for benefits;
• accounting for deceased children on the benefit roll;
• processing benefits that are incomplete or contain inconsistent information;
• reconfirming eligibility; and
• maintaining source documents.

19.80 For child tax benefits, we attempted to analyze whether the number of newborn children for whom benefits were being paid was generally reasonable. We used the 1994-95 benefit year records for our analysis, since the 1995-96 benefit year was still in progress at the time of our audit. We also limited our analysis to children on the benefit roll recorded as born in 1993, because some children born in 1994 for whom benefits would be paid would not yet have been enrolled by the end of the 1994-95 benefit year. This is because some families may apply for benefits several months after the birth of child. There are also inherent time lags in the processing of applications received by the Department.

19.81 To assess the overall reasonableness of the number of children receiving child tax benefits who were recorded as born in calendar 1993, we compared departmental information with our own benchmark estimate of maximum expected enrolments. (This estimate was based on departmental data and relevant vital statistics and population information produced by Statistics Canada.) All children who are born are not necessarily entitled to a benefit. Therefore, in establishing our benchmark estimate we needed to determine an appropriate entitlement rate that would allow for a proportion of children who do not qualify for benefits either because they are not eligible, for instance if they reside outside of Canada, or because they are not entitled due to family income being too high. As illustrated in Exhibit 19.6(a) , the population of newborns for 1993 as per Statistics Canada records, adjusted for immigrants and neonatal deaths, amounted to about 389,000. Using what we determined to be the maximum reasonable entitlement rate of 88 percent (the actual entitlement rate could be lower), this translates into an entitled 1993 newborn population of 343,000. To arrive at a maximum benchmark estimate of the number of newborns one could reasonably expect to find on the benefit roll, we used the size of the entitled 1993 newborn population as a base and adjusted the figure to take into account plausible take-up rates for newborn enrolments.

19.82 We found that the number of children on the child tax benefit roll who were recorded as born in 1993 exceeded our range of benchmark estimates for plausible take-up rates. We are concerned that this may be an indication that some newborns on the benefit roll do not belong there. Such a problem could be caused either by incorrect reports of newborns or by the Department's failure to detect that families of some otherwise entitled newborns do not qualify for benefits because family income is too high.

19.83 The precise extent of the apparent problem we identified with newborn enrolment depends critically on the take-up rate of the program. (The lower the take-up rate among applicants who are genuinely entitled to receive benefits for their newborns, the higher the estimated number of newborns on the benefit roll who do not belong there.) Although the actual take-up rate has not been measured, we believe it is unlikely to be higher than 98 percent of entitled children and may be well below this level, for the following reasons:

• some families entitled to child tax benefits might not be aware of the program;
• other entitled families might incorrectly believe that their income is too high to qualify for benefits;
• still others may incorrectly perceive that obtaining the child tax benefit would result in the loss of other forms of social assistance; and
• the requirement to file a tax return in order to qualify for benefits could discourage some families from applying.

19.85 As Exhibit 19.6(b) shows, the number of children recorded on the Department's benefit roll as born in 1993 exceeds our maximum benchmark estimate of appropriate enrolment even if the actual take-up is as high as 98 percent. If the actual rate is 98 percent, the enrolment level exceeds our estimated maximum by 2 percent; if it is at the other end of the range, at 90 percent, the enrolment level exceeds our estimated maximum by nearly 10 percent.

19.86 The potential financial implications are quite serious. Our analysis suggests that between about 6,000 and 34,000 children enrolled as born in 1993 may have been improperly included on the 1994-95 benefit roll, for the range of our plausible take-up rates and assuming our estimated maximum level of enrolments. We are unable to give a more precise estimate, because the information we needed to do so was not available. For the estimated range of children enrolled as born in 1993 who may have been improperly included on the 1994-95 benefit roll, the improper payments for this one year alone would range from $5.9 million to $31.6 million. The problem takes on a multiplier effect in that once an unentitled newborn has been placed on the roll, benefits will continue to be paid over the years unless either the problem is detected and corrected or there is a change in family status or income that impacts on entitlement. Given that the control environment has not changed in the area of child enrolments since child tax benefits were first introduced, we are concerned that many of these apparently unentitled 1993 newborns likely have already been in pay status for three to four years. Moreover, there are also other unentitled newborns who could have been improperly added to the roll for 1994 and 1995. We recognize that in a major social benefit program such as child tax benefits, it would be unrealistic to expect a complete absence of non-compliance. Some degree of error, fraud, and abuse is to be anticipated even with a well-designed verification, monitoring and enforcement regime. However, we are concerned that given the current regime the exposure to risk may be excessive, and it needs to be reduced .

19.87 The actual extent of the problem with newborn enrolments will not be known unless the Department undertakes appropriate initiatives to measure program take-up and non-compliance. However, we think the potential consequences are sufficiently serious that the Department needs to investigate the causes and the extent of the apparent problem of some newborns being on the benefit roll who should not be there and, where appropriate, strengthen its procedures for verifying the eligibility of newborns and family income for purposes of entitlement.

19.88 Our concerns do not extend to the enrolment of older children for child tax benefits because most of these children have been subjected to more intensive verification efforts, either under the former Family Allowances Program from which many were automatically included or under the Child Tax Benefit Program. Furthermore, we found that overall statistics on total child enrolment were not implausible when compared with Statistics Canada estimates of the child population overall. But, as noted above, with respect to recent enrolment of newborns this is not the case.

19.89 We were unable to perform a similar analysis of child enrolments in the GST Credit Program: the Department maintains only limited information on program participation, and suitable third-party information against which to check overall enrolment levels could not be obtained. However, our findings on child tax benefits only heighten our concerns about the lack of verification of children for whom GST credits are claimed.

19.90 Accounting for deceased children on the benefit roll. The second area in which we found that appropriate checks and balances are lacking is accounting for deceased children who have been on the child tax benefit roll. (For GST credits, this is a lesser concern because of the legal requirement to apply annually, which also requires providing up-to-date information on numbers of qualified dependants.) Revenue Canada relies (as did HRDC until August 1995) almost exclusively on information volunteered by program participants concerning the deaths of enrolled children. While participants do have an obligation to report changes to the Department that could affect eligibility (and can be charged penalties for failing to do so), it is unclear to what extent they know about their duty to report the death of a child enrolled in the program.

19.91 We estimate that over one fourth of all deaths of enrolled children during calendar years 1993 and 1994 were not reported to the Department. Our analysis was based on a comparison of departmental data with relevant vital statistics information produced by Statistics Canada. Just as with newborn enrolments, the Department needs to immediately investigate the extent of this problem and consider strengthening its procedures for accounting for deaths of enrolled children.

19.92 Processing benefits with incomplete or inconsistent information. The third area where we think checks and balances need to be improved is the processing of GST credit and child tax benefit applications when information supporting entitlements is incomplete or inconsistent.

19.93 In both programs, the Department processes benefit payments even when crucial information affecting entitlements is uncertain. The most common type of missing or inconsistent information is spousal information (marital status, spousal social insurance number and spousal income). In the absence of complete information, the Department generally makes an assumption, which is then used as the basis for computing benefit payments. The Department does not routinely contact the applicant for clarification in such cases, even after the initial assessment. Consequently, the remaining payments over the benefit year will normally continue to be based on the Department's initial assumption. Such a policy creates a high potential for both overpayment and underpayment of benefits.

19.94 During our visits to tax centres, we observed that many child tax benefit applicants provide inconsistent details about their marital status. This raises questions about the quality of marital status information the Department receives overall. This is an important issue, because incorrect information about marital status not only can result in improper child tax benefit and GST credit payments but also can lead to inappropriate assessments of deductions and credits such as the federal child-care deduction, equivalent-to-spouse amount and spousal amount, as well as provincial sales tax credits.

19.95 The Ottawa Tax Centre recently reviewed inconsistencies in reported marital status between Ontario child tax benefit applications and previously submitted tax returns. The review focussed on new child tax benefit applications processed during the fall of 1995. The number of inconsistencies identified amounted to about 1,300 cases out of 35,000 applications processed in that period . At the conclusion of its review, the Tax Centre had determined that previous tax assessments and/or benefit payments needed to be reassessed for 283 families. For these families, the Centre concluded that either too little tax had been paid in the previous three years or overpayment of benefits had been made (or both) because the family had provided incorrect information on marital status. Tax, provincial credits, and GST credit and/or child tax benefit adjustments were computed; the computed three-year net total increase in taxes and reduction in previous benefits and credits amounted to more than $500,000, representing about $1,800 for each of the 283 families. Overpayments of GST credits and child tax benefits accounted for more than half of this total. If the applications reviewed are representative of those received nationally, similar reviews of other applications received during the 1995-96 benefit year alone could be expected to yield an additional five to six million dollars in tax reassessments and benefit adjustments.

19.96 The Ottawa Tax Centre initiative was undertaken as a pilot project. At the conclusion of our audit, the Department was considering plans to extend the project to a national level.

19.97 The Department continues to make assumptions when confronted with missing or inconsistent information on child tax benefit and GST credit applications, despite the clear risks involved. It is still not its standard practice to obtain clarification, even following the initial assessment.

19.98 Reconfirming eligibility. Because a child is enrolled for child tax benefits for up to 18 years on the basis of a single application, it is possible that the information on which eligibility is based may become outdated. (By way of contrast, GST credit recipients are legally required to apply annually, thereby "reconfirming" their eligibility status.) To deal with potential inaccuracies in its own departmental records for child tax benefits when it was responsible for determining eligibility, HRDC developed plans to conduct reconfirmation exercises for certain participants who met specified criteria. This exercise, which was to be carried out with the assistance of Revenue Canada, was deemed essential to reduce the risk of paying out benefits to recipients who no longer qualify for benefits. Up to 94,000 participants in a given year were to be contacted by phone to confirm their marital status and the eligibility of enrolled children. Although a reconfirmation exercise was planned for each benefit year beginning with 1993-94, the extent to which the exercise was carried out in the initial year is unknown (no records of this exercise exist). In the following year, due to resource shortfalls, only a small fraction of the planned reconfirmations were completed. For the 1995-96 benefit year, the reconfirmation exercise became Revenue Canada's responsibility. At the end of our audit, which coincided with the end of the 1995-96 benefit year, the Department had completed only a very small number of the planned reconfirmations.

19.99 We believe that periodic reconfirmation of eligibility information with child tax benefit recipients would be a prudent management practice, particularly if the recipients were required to provide written responses. To date, substantial effort has been directed at planning reconfirmation exercises, but only a small portion of those planned have been carried out. Clearly, Revenue Canada needs to devote more serious attention to ensuring that benefits for this multi-billion dollar program are paid on the basis of accurate and up-to-date information. The implementation of realistic and appropriate reconfirmation procedures would be a necessary first step.

19.100 Maintaining source documents. It is standard practice to maintain copies of source documents such as application forms and supporting documentation when an individual applies for social benefits. In the event that computer records on a client appear to be incorrect, are unintentionally modified or are destroyed, source documents provide an important paper trail to follow in resolving the problem. Moreover, these documents generally are essential for the successful prosecution of cases involving fraud and other program abuse.

19.101 To investigate whether child tax benefit payments were properly supported, we asked HRDC and Revenue Canada to supply us with supporting documentation on random samples of participants who received benefit payments during both departments' respective periods of responsibility for eligibility determination. Neither department was able to provide us with all requested documentation. In particular, HRDC was able to locate requested documents for only 148 participants of the 185 child tax benefit participants (80 percent) in the sample it was given. Revenue Canada provided all requested documents for only 26 of the 65 participants (40 percent) in the sample it was given. We note that the documents it did provide were those involving applications that were processed by the Department since taking over responsibility for child tax benefit eligibility. We were informed by Revenue Canada that the remaining documents could be located only by using a significant amount of additional resources.

19.102 Given the inability of the departments to provide all requested source documents, we are unable to confirm that all required documentation has been received from program participants. Moreover, we cannot verify that the information used to determine eligibility is consistent with the information supplied by applicants. Finally, we cannot comment on the authenticity of the eligibility information contained in the departments' data banks (HRDC's until August 1995 and Revenue Canada's thereafter). The inability of the departments to readily access all source documents for child tax benefits could impact on Revenue Canada's ability to undertake enforcement activities involving benefits paid (past, present and future) based on prior years' applications and related source documents. It is noteworthy that in the case of GST credits, we were able to trace supporting information contained in the applications on personal tax returns, as required, and observed no discrepancies.

Risks to GST credits may be higher

19.104 We note, for example, that the number of GST credit recipients grew by 8 percent between the 1992-93 and 1994-95 benefit years. Over this same period, the increase in total GST credits, from $2.5 billion to $2.8 billion (a 12 percent increase), has outpaced by 50 percent the growth in the number of recipients. The Department has done no analysis of the trends in benefit payments and participation levels. We think the causes and implications of these trends warrant careful investigation.

19.105 There are three respects in which GST credits may be even more vulnerable to error and abuse than child tax benefits as a result of the way that credits are currently processed. First, whereas a child tax benefit applicant is at least required to provide basic information about each child claimed for benefits (such as name, date of birth, relationship to applicant and period of residence with applicant), a GST credit applicant is required to state only the total number of children being claimed. On the basis of such limited information, it is not possible for the Department to check even whether the children for whom benefits are claimed meet the age and residency requirements of the program.

19.106 Second, the spouse of a GST credit recipient is not legally required to file a tax return in order for the applicant to receive benefits. If the spouse does not file, the Department must rely on the spousal income figure stated on the recipient's application to compute GST credit benefits. The amount reported by the applicant may not reflect true spousal income if, for example, the spouse does not keep accurate income records in the absence of a filing requirement or the applicant is not aware of the spouse's exact income. Moreover, in cases where the applicant does not provide any information about spousal income, the Department generally assumes that the spouse has no income rather than contacting the applicant for clarification.

19.107 Third, the family member who receives the GST credit may vary from one year to the next at the family's discretion, whereas the child tax benefit recipient in the family normally remains the same. This makes it more difficult for the Department to track the level of payments going to a family and to prevent duplicate payments when applications have been submitted by more than one member of the same family.

19.108 Revenue Canada should institute an adequate and appropriate control environment to have reasonable assurance that only eligible recipients receive child tax benefits and GST credits and receive them in the correct amounts. Attention should be given to the following areas: enrolling children for benefits, accounting for deceased children, processing benefits with incomplete or inconsistent information, reconfirming eligibility and maintaining source documents.

Department's response: The Department has already taken steps to build an appropriate control environment. In order to better account for deceased children on the benefit roll, Revenue Canada is negotiating with one province to obtain third-party information, and intends to expand the process, subject to agreement by each provincial jurisdiction. The review of inconsistencies in reported marital status, which was conducted as a pilot at the Ottawa Tax Centre, has been extended to a national project. The Department is also actively carrying out an eligibility reconfirmation project in each tax centre. No further action is required to improve the management of source documents, since this was resolved when Revenue Canada implemented the new processing system in February 1996.

Statistics used for monitoring are flawed

19.110 In reviewing quarterly statistics on GST credits, we noted serious flaws in the 1995-96 statistics generated for the Department's internal use to monitor the overall performance of the program. For example, the number of recipients during the benefit year was grossly overstated, suggesting approximately 2 million more than the actual number. Similarly, with child tax benefits we identified serious and persistent errors in the Department's monthly statistics on births and deaths of children for whom child tax benefits were being paid. A comparison of the birth statistics for enrolled children with benchmark estimates suggested that as many as one in three newborns on the 1993-94 child tax benefit roll did not belong there. Through our subsequent analysis of raw program data, we found that the departmental statistics were inaccurate and had been so since they were first produced. These statistics were generated not only for internal monitoring but also for use by other key stakeholders. The Department was not aware of the deficiencies in its statistical reports for child tax benefits or for GST credits.

19.111 The Department's poor performance in generating reliable statistical information essential for monitoring is a significant concern. The failure to maintain accurate data and program statistics not only impacts on stakeholders who rely on such information but also inhibits the Department's ability to properly monitor its programs and identify problem areas for timely corrective action.

19.112 Revenue Canada should take appropriate measures to ensure the reliable and accurate generation of key program statistics meant for either monitoring purposes or use by other key stakeholders.

Department's response: The Department is reviewing statistical information that is generated in order to improve its accuracy and completeness.

Program monitoring has been inadequate

19.114 We found that, until very recently, the Department had not used its statistics on GST credits and child tax benefits to develop client profiles or to evaluate program performance in any meaningful way. Recently, however, it has begun to analyze overpayments in the Child Tax Benefit Program. A draft departmental report reveals that outstanding overpayments since the inception of the program stood at $69 million as of 31 March 1996. The report indicates that there are obstacles to the collection of some of these overpayments. For example, child tax benefit participants who have received excess payments may receive no continuing child tax benefit payments or tax refunds against which the liabilities can be offset. Studies are being planned to explore the feasibility of improving debt collection and reducing the level of overpayments generated.

19.115 The Department has yet to undertake a similar analysis of GST credit overpayments. At our request, it produced an estimate of overpayments made during the 1994-95 benefit year. Based on a manual calculation of overpayment amounts from more than 75 quarterly and supplementary reports on cheque issues for the year, the Department estimated that overpayments amounted to $19.8 million. No information is available to track the recovery of GST credit overpayments, and no analysis has been undertaken to learn about the causes of overpayments or ways to reduce them.

19.116 While we recognize that these overpayment figures amount to less than one percent of total amounts paid, on either an annual or cumulative basis, in absolute terms the overpayments are still financially significant. Given that little analysis or scrutiny has been performed in this area, there could be cost-effective opportunities to reduce the balances, potentially by millions of dollars.

19.117 Overall, we believe the Department needs to devote more attention to monitoring its performance in these programs. As cited, one key aspect would be ensuring periodically that aggregate statistics on key program variables compare reasonably well with benchmark statistics on the same variables; another would be following up on known overpayments. We have already commented on the potential financial consequences if excess child enrolments exist on child tax benefit rolls. Similarly, as noted above, there are financial implications to insufficient monitoring of benefit overpayments. We do acknowledge the Department's draft report on child tax benefit overpayments as a necessary first step.

19.118 Revenue Canada should monitor overall performance of the programs periodically to ensure the reasonableness of total benefits being paid against benchmark statistics on key program variables, and develop improved systems for the control and collection of known overpayments.

Department's response: The Department has in fact been conducting a number of activities to monitor the performance of its benefit programs. A new divisional structure was initiated, which includes a section dedicated to quality assurance and compliance activities. From the Child Tax Benefit overpayment report, an action plan has been created and is being implemented. The Department has also initiated a thorough evaluation of its program monitoring requirements, including the integrity of internal statistics and the availability of external data sources. Finally, the collection of benefit overpayments will be expanded by automating links to the Department's well-established electronic collection system. These automated links will be in place in February 1997.

No formal enforcement strategy against fraud and abuse

19.120 The experience of the United States Internal Revenue Service (IRS) with its Earned Income Credit (EIC) program is also relevant. The EIC has been subject to wide-scale error and abuse. Prior to the 1986 tax reform in the United States, there was generally no requirement to provide documentation to substantiate dependency claims or to substantiate the claim for the Earned Income Credit. Starting with the 1987 tax year, new laws required social security numbers (similar to social insurance numbers in Canada) for all dependants aged five years or older claimed for personal exemptions on tax returns. In testimony before Congress, the IRS noted that the number of dependants claimed on tax returns fell by seven million between tax years 1986 and 1987 (many of whom likely would have been claimed for the Earned Income Credit). The drop in the number of dependants was estimated to equate to at least $2.8 billion in additional revenue for the Treasury. The Earned Income Credit has continued to be a problem area for the IRS. For tax year 1988, it was reported that of the $5.6 billion awarded in earned income credits that year, about $1.9 billion (34 percent) were awarded erroneously, due primarily to errors in taxpayers' filing status and the number of dependant children claimed. Since then, the law in the U.S. has been further amended to require social security numbers for all dependants claimed for either earned income credits or personal exemptions. While observed problems with non-compliance in the United States are not a sure indicator of similar problems in Canada, the evidence in the United States as well as numerous studies on compliance behaviour in general would suggest that some individuals in Canada may well attempt to cheat the system if given the opportunity to do so. Purely from a risk-management perspective, it would be prudent to ensure that sufficient controls exist to prevent substantial numbers of ineligible or unentitled participants from entering the child tax benefit roll. However, our audit found that this is not the case.

19.121 Exhibit 19.7 provides examples of practices of other jurisdictions in the area of verification and enforcement for programs broadly similar to the Child Tax Benefit and GST Credit programs.

19.122 Revenue Canada should establish and implement an appropriate enforcement regime for child tax benefits and GST credits to detect and prevent abuse and other non-compliance in the programs.

Department's response: This is addressed in the Department's action plan.

Quality of data and program logic supporting benefit payments is unknown

19.124 There are three key areas in which the Department has not adhered to generally followed standards in the area of systems development. First, it failed to maintain essential records of changes pertaining to the conversion of data on recipients from old systems to new systems. Documentation was also missing or incomplete in even such key areas as problem identification, resolution and follow-up activities. Incomplete documentation is a concern because it inhibits the Department's ability to effectively and efficiently correct program flaws and to modify systems later. It also limits the opportunity for independent reviews of new systems to ensure an acceptable level of quality. Second, the Department has compromised the testing of core applications in order to meet in-house deadlines. Third, the lines of responsibility for making changes to programs are not always clear, and formal approvals are not always present. This increases the risk that incorrect and unapproved changes will be made; indeed, we are aware of incorrect changes that have been made to these systems.

19.125 ICD development. We note that the ICD development schedule was shortened from 18 months to 9 months to meet internal deadlines. The time available for development and testing was significantly reduced, increasing the risk of inadequate quality control, errors and systems failure, particularly given the absence of contingency planning in the form of specific measures to mitigate the risks. Several problems did arise when ICD was implemented. We note that an internal review of the stability of the system was initiated, but not until several months after it was operational. At the completion of our audit, the review was still in progress.

19.126 CCM development. In the case of CCM, the system responsible for determining eligibility for child tax benefits, we noted that both system and user testers reported an inability to test the system fully before it went "live". The system was implemented according to schedule, but was immediately withdrawn for one week due to problems with its operations. For several weeks, users did not have full functionality, which built to a backlog of unprocessed child tax benefit applications. Full functionality was available two months later.

19.127 We find that Revenue Canada's approach to the above-noted aspects of generally followed standards for systems under development - such as ICD and CCM, which deliver billions of dollars in benefits to millions of individuals and families each year - has lacked the necessary level of rigor; consequently, the Department has accepted a higher degree of risk than typically would be accepted in implementing such systems.

19.128 We note that the Department established a Production Assurance Steering Committee and task force in the fall of 1995 to develop and oversee implementation of a new testing and quality assurance model for systems development projects. At the end of our audit, the Committee's work was still under way.

19.129 Revenue Canada should adhere to generally followed standards in the area of system quality assurance, problem resolution and management of change when developing and implementing key systems supporting the Child Tax Benefit and GST Credit programs and other similar systems.

Department's response: The discipline of testing in systems development is an exercise in risk management. When implementation dates must be met, Revenue Canada manages risk by deferring functionality that is not performing satisfactorily or immediately required. The final measure of testing success is the relative severity of the defects that are detected after production, and the proficiency with which they are handled and corrected. By this measure and by industry norms, these projects were successful.

Information to Parliament and Other Stakeholders

Neither program is a priority for evaluation

19.131 Although the Child Tax Benefit Program has been in place since 1993 and the GST Credit Program since 1990, and despite the fact that these programs are among the largest federal social benefit programs targeted toward low- and modest-income families, neither program's performance has been evaluated against program objectives. Nor are there any plans for the next several years to conduct such evaluations. Our concern is heightened further because a framework for evaluation has not been defined or clarified for either program.

19.132 The Department of Finance is responsible for policy-related analyses of the Child Tax Benefit and GST Credit programs. Finance has indicated that neither of these programs is scheduled for evaluation in its current three-year evaluation plan.

19.133 The stated policy objectives for the Child Tax Benefit Program include simplicity, fairness, responsiveness, effectiveness and efficiency. The stated policy objective of the GST Credit Program is to provide relief from the goods and services tax for families with low and modest incomes. There are a number of features in the structure of the two programs that are similar - such as the threshold for benefit phase-out, targeting low- and modest-income families, supplementing basic benefits and determining entitlements using family net income as a measure of well-being - that may warrant a combined evaluation.

19.134 In view of their social and financial significance, we are concerned that it may take many more years before Parliament learns of the extent to which the programs are meeting their intended objectives. Further, in the event that a decision is made to conduct such evaluations, the Department of Finance may not have the necessary information for the required analysis because a framework for evaluation has yet to be established for either program.

19.135 The Department of Finance should establish a framework for evaluating the Child Tax Benefit and GST Credit programs and ensure that the necessary information will be available when a decision is made to undertake an evaluation.

Finance's response: The Department of Finance will establish an evaluation framework for these credits and review the priority of evaluating them in relation to other evaluation topics.

Little accounting to Parliament for program performance and administration

19.137 For example, the information on the Child Tax Benefit and GST Credit programs provided in Part III of the Estimates is limited to general information and basic statistics on the volumes of payments and recipients. Also, most of the information provided about these programs served primarily to announce that the function of determining eligibility for child tax benefits was being transferred from HRDC to Revenue Canada. Part III does not indicate plans to establish standards of services for these programs; nor, in the absence of plans, does it report on the programs' performance in key areas. Nor does it elaborate on priorities for the administration of these programs.

19.138 Revenue Canada should report to Parliament periodically on overall performance in administering the Child Tax Benefit and GST Credit programs. Information should also include relevant details on program plans, priorities and initiatives.

Department's response: In the context of the Treasury Board's Improved Reporting to Parliament project, Revenue Canada will review the information provided to parliamentarians.

Conclusion

19.140 Our audit also revealed a number of matters that require immediate attention in the areas of verification and enforcement and the processing of child tax benefits and GST credits. We concluded that Revenue Canada's control environment requires significant strengthening, to ensure equitable program delivery and to safeguard the public purse from significant losses. The Department has not been monitoring program results on an overall basis to ensure that aggregate statistics on key program variables, such as births and deaths of enrolled children, compare reasonably well with benchmark statistics on the same variables. Our own analysis of program statistics against independent data on these variables suggests potentially serious problems that require urgent attention, at least with respect to child tax benefits. The implications for GST credits are unknown because we were unable to do a comparable analysis: the Department maintains only limited information on program participation and suitable third-party data were not available. However, we believe that the risks with GST credits are higher in many respects than with child tax benefits.

19.141 We also found that Revenue Canada accepted a high level of risk in not adhering to certain aspects of generally followed practices when it developed new systems to support the two programs. We also observed flaws in the tabulation of some key program statistics meant for use by not only the Department but also a number of key stakeholders. As a result, stakeholders may not be well served and the Department's ability to monitor program performance and identify problem areas is severely weakened.

19.142 We note the absence of plans for any policy-related evaluations of either program by the Department of Finance. Consequently, it may take several years before Parliament learns of the extent to which these key federal social transfer programs are meeting their intended objectives.

19.143 Revenue Canada informs us that it has a number of initiatives under way to improve its administration of the programs. We have noted some of these initiatives in this report.

About the Audit

Audit Scope

Objectives

• the level of GST credits and child tax benefits being provided through the tax system and information on the extent to which all those eligible for the benefits are receiving them;
• the adequacy of controls to protect the Crown from unnecessary losses and to report any significant deficiency in program delivery and/or controls;
• the efficiency of the Department's administration of the GST Credit and Child Tax Benefit programs;
• the systems in place for monitoring and evaluating both programs; and
• the quality of information reported to Parliament on the administration and performance of the programs.

Scope

We did our audit work at Revenue Canada's headquarters in Ottawa, five regional offices, five tax centres and seven tax services offices across Canada. In addition, we interviewed officials at HRDC and the Department of Finance to learn about their respective roles in administering and evaluating either child tax benefits or GST credits. We also met with U.S. officials from the Internal Revenue Service, the Office of Tax Analysis, the General Accounting Office and with officials from provincial social assistance and/or vital statistics agencies in Newfoundland, Quebec, Ontario, Alberta, and British Columbia to obtain a perspective on overall trends in social assistance administration and specific information on experiences, client profiles and current initiatives.

Criteria

• In administering new programs such as Child Tax Benefit and GST Credit through the tax system, the Department should prepare and adhere to an appropriate implementation strategy in accordance with its legislative mandate.
• The Child Tax Benefit and GST Credit programs should be supported by an effective and efficient facilitation process to ensure that all eligible individuals both understand their rights to these social benefits and are provided with means to apply for and receive them.
  
• Benefits under the Child Tax Benefit and GST Credit programs should be delivered promptly and be responsive to changes in recipients' status at least cost to the government, while providing for the fair and equitable treatment of recipients.
• The administration of the Child Tax Benefit and GST Credit programs should include an appropriate enforcement strategy.
• An adequate system of internal controls should be in place to ensure the authenticity of transactions; the reliability and completeness of key databases; and the processing of eligibility, entitlements, adjustments and elections in accordance with legislative provisions.
• The administration of the Child Tax Benefit and GST Credit programs should be carried out with efficient systems and procedures. The Department should search for and implement, to the extent feasible, cost-effective and innovative administrative practices.
• There should be a framework for monitoring and evaluating the Child Tax Benefit and GST Credit programs, and procedures for collecting and reporting appropriate performance information to decision makers.
• Parliament should be kept informed of the cost, results and departmental performance in administering the Child Tax Benefit and GST Credit programs.

Audit Team

For information, please contact Basia Ruta, the responsible auditor.

Revenue Canada provided the action plan below with its response to our recommendations.

THE DEPARTMENT'S ACTION PLAN

The Department will build on its tradition of client consultation and balanced service and enforcement by:

• the take-up rates for child tax benefits and GST credits;
• the level of awareness of entitlements and obligations under child tax benefits and GST credits;
• the degree to which related information needs are met; and
• compliance levels and potential areas of non-compliance.

• profiles of the information and service needs of child tax benefits and GST credit clients;
• compliance profiles to sustain effective, targeted verification and enforcement
  activities; and
• communications strategies for reaching all potential clients.

• client service and processing standards are clearly communicated and met;
• processing systems are added or modified in accordance with approved practices;
• risks and vulnerabilities are identified, quantified and addressed; and
• processing systems and work methods consistently provide accurate results and decisions.

3) Redesigning and improving telephone enquiries systems to

• facilitate the definition and achievement of appropriate service standards, at an affordable cost; and
• provide targeted services which meet identified clients' needs.

4) Enhancing its enforcement programs through compliance research and other measures, such as

• compliance measurement samples;
• directed statistical samples to quantify identified "problem areas"; and
• negotiating effective, economical exchanges of key information with the provinces.

5) Monitoring overall performance of major benefit programs by:

• developing and maintaining appropriate statistical measures which meet the needs of Revenue Canada, Finance, and other stakeholders;
• using data from Statistics Canada and other external sources to establish baseline indicators for the evaluation of key programs; and
• using client consultation and external data to quantify "take-up" rates, client errors or fraud, identify appropriate departmental responses and evaluate their effectiveness.