FPSLREB Decisions

Heard at Ottawa, Ontario,
October 9 and 10, 2014, and June 15 to 19, 2015.

REASONS FOR DECISION

I. Policy grievances referred to adjudication

1        On November 1, 2014, the Public Service Labour Relations and Employment Board Act (S.C. 2013, c. 40, s. 365) was proclaimed into force (SI/2014-84), creating the Public Service Labour Relations and Employment Board to replace the former Public Service Labour Relations Board (PSLRB) as well as the former Public Service Staffing Tribunal. On the same day, the consequential and transitional amendments contained in ss. 366 to 466 of the Economic Action Plan 2013 Act, No. 2 (S.C. 2013, c. 40) also came into force (SI/2014-84). Pursuant to s. 393 of the Economic Action Plan 2013 Act, No. 2, a proceeding commenced under the Public Service Labour Relations Act (S.C. 2003, c. 22, s. 2; PSLRA) before November 1, 2014, is to be taken up and continue under and in conformity with the PSLRA as it is amended by ss. 365 to 470 of the Economic Action Plan 2013 Act, No. 2.

2        On June 19, 2017, An Act to amend the Public Service Labour Relations Act, the Public Service Labour Relations and Employment Board Act and other Acts and to provide for certain other measures (S.C. 2017, c. 9) received Royal Assent, changing the name of the Public Service Labour Relations and Employment Board and the titles of the Public Service Labour Relations and Employment Board Act and the PSLRA to, respectively, the Federal Public Sector Labour Relations and Employment Board (“the Board”), the Federal Public Sector Labour Relations and Employment Board Act, and the Federal Public Sector Labour Relations Act (“the Act”).

3        On July 29, 2013, the Association of Justice Counsel (“the AJC”) filed a policy grievance (file number 569-02-133) under s. 220(1) of the Act alleging violations of articles 5 and 13 and clause 19.19 of the collective agreement between the AJC and the Treasury Board (“the employer”) for the Law Group (All Lawyers) with the expiry date of May 9, 2014 (“the collective agreement”). The grievance specifically identified the Department of Justice (DOJ), the Public Prosecution Service of Canada (PPSC), and Veterans Affairs Canada as the departments and agency involved and stated as follows:

The AJC grieves all departments and agencies recent leave reconciliation initiative on the basis that management has exercised its management’s rights and discretion pursuant to the collective agreement in an unreasonable, arbitrary and bad faith manner, contrary to sections 5, 13 and 19.19 and such other sections as may apply.

While deductions are permitted for overpayment on account of salary, wages, pay or pay allowances under the FAA, there is no authority for management to unilaterally alter or recover leave credits or make unilateral pay deductions following reconciliations made pursuant to section 13. Nor is there a legal obligation to recover alleged overpayments on account of leaves of absence which were approved by management.

The members have detrimentally relied on prior reconciliation and leave reports.

[Sic throughout]

4        To remedy the violations alleged by the policy grievance, the AJC requested that a declaration be made that the departments’ and the agency’s exercise of discretion was unreasonable, that they be ordered to cease and desist from the leave reconciliation process, that they reimburse any and all AJC members or reverse any prior transactions involving the reconciliation or recovery efforts that are the subject of the grievance, that they apply clause 19.19 as appropriate and write off any alleged discrepancies, and that any other remedy be applied that may be deemed appropriate under the circumstances.

5        The employer denied the policy grievance on September 23, 2013, stating in part as follows:

...

... Management has authority to conduct this review and correct any inaccuracies pursuant to Article 5 of the LA collective agreement and section 155(1) of the Financial Administration Act (FAA). I believe that the Employer exercised their management rights in a reasonable, fair and good faith manner....

...

6        The AJC referred the policy grievance to adjudication on October 2, 2013.

7        On August 28, 2013, the AJC filed a second policy grievance, related to the leave reconciliation process (file number 569-02-137). It specifically identified the DOJ and the PPSC as the department and agency involved, respectively, and stated as follows:

The Department of Justice (DOJ) and/or Public Prosecution Services Canada (PPSC) have, in the context of their leave reconciliation initiative recalled, asked or insisted that LAs who are on approved leaves of absence, perform work while on such leave.

Such requests are considered to be unreasonable, arbitrary and in bad faith and a violation of the collective agreement provisions, including but not limited to sections 5.02, 11, 18, 19, 21.

...

8        As corrective action, the AJC requested that the department and agency cease such practices; that the employer provide direction to all relevant departments confirming that such requests are unreasonable and prohibited; that any employee on approved leave of absence who accommodated a request to work while on that leave be granted additional leave with pay equivalent to the time worked on the reconciliation process once the approved leave expires, pursuant to clause 19.19; that any disciplinary action taken be rescinded and removed in the event that it was taken in relation to an AJC member’s failure or inability to comply; and any other remedy considered appropriate under the circumstances.

9        The employer denied the second policy grievance on October 22, 2013. Among other things, it stated that employees on leave without pay were asked to participate in the leave reconciliation process but that employees who chose not to respond were not penalized and that those who spent time on the exercise were entitled to be compensated and could receive comparable time off at a later date.

10        The AJC referred the second policy grievance to adjudication on November 14, 2013, and asked that it either be consolidated with the first policy grievance or heard at the same time. The employer agreed with that request, and I ordered them joined.

11        Two days were allotted for hearing the grievances, which proved insufficient. Between the first hearing dates and the second scheduled dates, the AJC filed a third policy grievance, on a “without prejudice” basis, relating to the application of the leave reconciliation exercise to retired and former employees who worked for other departments or agencies. However, in correspondence with the Board, the AJC took the position that it had been unnecessary to file the third policy grievance because the existing policy grievances included former employees.

12        The employer did not agree with that and questioned whether the AJC had standing to file grievances on behalf of persons who were not members of the bargaining unit.

13        In response to that dispute, I ruled that the parties would have to specifically address it. If a breach of the collective agreement were established, a separate hearing could be scheduled to determine an appropriate remedy and to whom it should apply.

II. Summary of the evidence

14        The AJC called Sandra Guttman, AJC general counsel; Yaron Butovsky, lawyer, DOJ; Karen Pearce, lawyer, DOJ; Laurie Klee, lawyer, DOJ (retired); Aziz Saheb-Ettaba, lawyer, DOJ; and Mona Klinger, lawyer, DOJ. It filed 22 exhibits, including an exhibit book, “G-1”, containing 25 documents.

15        The employer called George Dolhai, Deputy Director, PPSC; Julie Betts, Chief Audit Executive, PPSC; Cheryl Driscoll, Chief Audit Executive, DOJ (retired); Joan Pratt, Director General Human Resources, DOJ (retired); Bev Richard, Head, Human Resources Statistics and Planning, DOJ; and Patrick Vezina, Director, Law Practice Management Division, DOJ. It filed 26 exhibits.

16        The parties requested that the personal record identifiers (PRIs) of witnesses or employees in any of the exhibits be redacted. In particular, they noted that Exhibits G-4, G-8, G-10, E-1, E-2, E-3, and G-21 contained employees’ PRIs that should be redacted. I agreed to the request.

17        The evidence presented at the hearing related to AJC members employed at the DOJ and the PPSC. As noted, the AJC is the certified bargaining agent for all Treasury Board employees in the Law (LA) group, as described in the bargaining certificate that the PSLRB issued on September 12, 2007, and as set out in clauses 1.03 and 2.01(a) of the collective agreement. The LA group includes lawyers, articling students, and Quebec notaries.

18        The grievances concern projects by the DOJ and the PPSC to reconcile data on paid leave found in two different and unrelated electronic systems. One system, iCase, is an electronic timekeeping system applicable only to lawyers, paralegals, and certain other designated employees (collectively known as “timekeepers”) in the DOJ and the PPSC, which manage and maintain their iCase systems separately.

19        The other system, PeopleSoft, is a government-wide human resources (HR) management system used by all government employees, managers, and HR personnel to request, approve, and process paid leave requests, among other purposes.

20        The requirement that DOJ and PPSC timekeepers track their work time in iCase is found in the DOJ’s National Timekeeping Protocol (“DOJ NTP”; Exhibit G-1, Tab 4, versions 8.4 and 8.5) or the PPSC’s National Timekeeping Protocol (“PPSC NTP”; Exhibit E-13, April 2013 version), as applicable. Mr. Dolhai testified that the PPSC used the DOJ NTP for part of the period in question but that it eventually developed its own protocol. The rationale for timekeeping is set out in article 2.2 of the DOJ NTP, which provides as follows:

Accurate and complete timekeeping information supports the effective planning, funding and management of departmental activities, such as managing and reporting on resources and work volumes, recovering for the cost of legal services provided, analyzing trends, and forecasting resource requirements.

21        A similar statement is found at article 2.2 of the PPSC NTP.

22        Both protocols make accurate and timely timekeeping mandatory for all counsel and paralegals. Timekeepers are required to record in iCase the time spent on all defined work activities during the course of a work day, including time spent on providing legal services, on management or corporate activities, on professional development, and on paid leave. Both protocols establish broad categories and subcategories of work activities and tasks to help timekeepers accurately classify their time and assign it to specific files.

23        During the period in question, timekeepers at the DOJ were expected to record paid leave, such as vacation leave, sick leave, and compensatory and management leave, in iCase under the general heading of “Administration Task”. Mr. Butovsky testified that the particular kind of leave taken was not specified in iCase and that neither his iCase leave summaries nor those of Ms. Klee identify the kind of leave taken (Exhibits E-1 and G-9). However, similar records for Ms. Pearce and Ms. Klinger show that they noted the kind of leave taken on some leave entries but not on all of them (Exhibits E-3 and G-18).

24        Although timekeepers are required to record paid leave in iCase, the official management of leave, paid or otherwise, falls within the jurisdiction of the HR branch, which has used PeopleSoft since 2003.

25        Ms. Richard testified that PeopleSoft is a position-driven system. Positions are created, classified, and staffed, and when a staffing process is completed, an employee is attached to a specific position. As a consequence, the paid leave and other benefits associated with the position’s group and level are also attached to it. PeopleSoft administrators create unique identifiers for each employee, and each employee receives a link to PeopleSoft with instructions on how to use the leave self-service system, along with hyperlinks to training modules.

26        Employees in the LA group are entitled to the leave benefits set out in the collective agreement. Both the collective agreement in force when the grievances were filed (Exhibit G-1, Tab 19) and the previous one, which expired on May 9, 2011 (Exhibit G-1, Tab 18), were entered in evidence. Vacation and sick leave are granted annually, and unused leave may be carried over in accordance with the terms of the collective agreement. Other forms of paid leave, such as for family related responsibilities, bereavement, or volunteering, for example, may be granted when employees meet the requirements of the relevant collective agreement provision. Those forms of leave are not carried over from year to year.

27        All forms of leave are assigned a unique code for the purposes of PeopleSoft. Referring to the PeopleSoft Quick Reference Guides for Managers and Supervisors and the one for employees (Exhibit E-23, Tab 1), Ms. Richard explained that an employee who wishes to apply for leave accesses his or her leave information page on PeopleSoft to complete the request electronically. The employee identifies the type of leave requested, the date, the number of hours of leave, and his or her supervisor’s name and ID. The system then emails the request to the supervisor, who can then, through the system, access the employee’s ID number, the leave code and description of the request, and the leave status. Additionally, the supervisor has access to the employee’s leave balances and a summary of all his or her leave transactions. With that information, the supervisor can approve or deny the leave or recommend a decision from a higher authority. At the end of the fiscal year, employees receive a statement of their leave usage for the year from PeopleSoft.

28        In cross-examination, Mr. Dolhai elaborated on the responsibilities of managers and supervisors in the PeopleSoft leave approval process. He stated that when a leave request arrives, front-line managers are responsible for ensuring that the employee has sufficient leave and that the request complies with the terms and conditions of the collective agreement. Those managers might be responsible for a handful of or 20 or 30 timekeepers. Offices are organized in teams, but ultimately, a manager is responsible on a day-to-day basis for knowing whether a person is at work and for ensuring that there are systems in place to make that knowledge possible. Nevertheless, he stated that given the nature of the work, managers might not know at the end of any given day that someone was absent. Regardless, he stated that managers are responsible for ensuring that employees do the work for which they are paid.

29        Mr. Vezina agreed that it is his responsibility as a manager to ensure that employees take the appropriate leave and to ensure that when they are not working, they enter their leave in PeopleSoft. He stated that managers generally make sure that leave applications are up to date for holiday purposes. His practice was to meet with the employees under his supervision at the end of every fiscal year to determine that all the leave they had taken had been properly entered into PeopleSoft.

30        Ms. Pratt testified that employees and managers have a joint responsibility to ensure that leave statements are accurate. She noted that there had always been an informal process in which employees attested to the accuracy of their leave balances at the end of the fiscal year and that following the reconciliation exercise, attestation became a formal requirement involving employees and their managers.

31        There was no dispute that PeopleSoft and iCase do not communicate with each other. Mr. Dolhai testified that the record of paid leave in iCase was simply an indication that leave had been taken. However, processes were in place to review and compare iCase records and PeopleSoft records.

32        In addition to the annual attestation process generated by PeopleSoft, timekeepers received an automated monthly timekeeping report. Ms. Klinger testified that the report was a printout of her iCase entries for the month, which she would review and approve and return to her supervisor. Mr. Vezina, who was Ms. Klinger’s supervisor, testified that he received and reviewed the timekeeping reports of the employees in his area. He explained that cost recovery was not a concern in his area. Therefore, iCase was a time management tool. He reviewed its reports to ensure timekeeping compliance. From time to time, he found it necessary to ask employees to bring their iCase records up to date.

33        Ms. Klinger testified that also, the practice was for employees to meet with their managers as part of the performance review process. She stated that that had been the practice for five to seven years before the reconciliation exercise and that she had done it with three different managers, including Mr. Vezina. Twice a year, she would meet with her manager and review, among other things, the timekeeping reports to determine how time was being allocated to files and to clear up any questions. She thought that because all time had to be accounted for in iCase, if there had been problems, they would have been addressed at those meetings.

34        Mr. Dolhai agreed that the monthly review of timekeeping records was an opportunity to identify any leave issues. If an employee had noted a day of leave in iCase for which there was no corresponding PeopleSoft record, the monthly review provided an opportunity to address it and to ensure that the leave balances were correct. He stated that managers had a responsibility to follow up on any identified issues. He also agreed that when managers conducted performance reviews, which took place at least twice a year, if not more often, iCase records were reviewed too. The annual performance review also provided an opportunity to ensure that leave balances were correct. It was part of the management function known to managers and employees.

35        Although as a general rule, employees used PeopleSoft to access and report paid leave, evidence was presented that when it could not be accessed, employees submitted paper leave forms to their managers for approval. Those forms were forwarded to HR to be processed and entered into PeopleSoft (Exhibit E-12). For example, Ms. Klee testified that PeopleSoft typically shut down for a time, often up to a week, at the end of the fiscal year. Furthermore, she stated that employees could not access it to make leave requests for a prior fiscal year. In those situations, employees submitted those paper leave forms for approval, which an HR employee would then enter in PeopleSoft.

36        When they were questioned in cross-examination about the use of paper leave forms, both Mr. Dolhai and Ms. Pratt agreed with Ms. Klee’s description of the practice followed when PeopleSoft was not available. Neither witness was able to identify who was responsible for ensuring that employees’ paper leave forms were properly entered into PeopleSoft. Nor were they certain if either employees or managers had any role or responsibility for making sure the requests were entered electronically in PeopleSoft. Ms. Pratt believed that a compensation advisor working on a particular account would have had the ability to enter a paper leave request into PeopleSoft, but she could not say whether managers would have been advised that that had occurred. She was also uncertain who would have had the authority to amend or correct a PeopleSoft record retroactively. Ms. Richard testified that information entered into PeopleSoft could be removed or amended only by certain designated people. For example, she noted that compensation advisors could add data but that for audit purposes, paper copies of the information would be required.

37        Although employees are responsible for ensuring that all paid leave is authorized by management, not all paid time off work had to be entered in PeopleSoft during the period in question. For example, management could grant up to a half-day of time off work with pay for personal, medical, or dental appointments, but because it was not considered leave with pay granted under the terms of the collective agreement, it was not recorded in PeopleSoft (Exhibit G-1, Tab 5).

38        Ms. Klinger testified that before the reconciliation exercise, an informal type of management leave was in place that required management approval but that was not entered in PeopleSoft. She stated that employees who had worked long periods of overtime could request up to five days of management leave annually, either orally or by email. The leave would also be approved orally or by email, at the manager’s discretion. She stated that the leave would be entered in iCase but not in PeopleSoft. It was an informal way of compensating people for unusual periods of overtime. She provided an example of a request for a day of leave in lieu of overtime for the purpose of a religious observance that her manager, Mr. Vezina, had approved (Exhibit G-22).

39        In cross-examination, Ms. Klinger acknowledged that the process for obtaining management leave became more formalized after the LA group became unionized and negotiated its first collective agreement because the leave then arose from the terms of the collective agreement. Before then, it had been discretionary.

40        Mr. Vezina also acknowledged the informal process for compensating employees for working large amounts of overtime. He stated that as a general rule, it would not have been recorded in PeopleSoft, and he did not believe that entering it in iCase was required. He believed that it amounted to a simple rearrangement of working hours. Hours worked on a weekend would be entered in iCase, so it was obvious that compensatory time off would not have to be entered in it. However, he also noted that paid management leave arising from the terms of the collective agreement would be recorded in PeopleSoft.

41        Mr. Dolhai testified that it was possible that informal leave arrangements were made, but he believed that they would have been monitored and that they did not relieve any responsibility from employees to make a leave request through their managers.

42        In cross-examination, he agreed that some managers had the discretion to grant up to five days of paid leave in compensation for overtime, but he stated that it depended upon the manager’s authority to exercise such discretion. However, he acknowledged that less-formal arrangements were possible in which a manager simply approved a day of leave in recognition of overtime. He stated that the PPSC directive on management leave (Exhibit E-14), which came into effect April 1, 2013, did not contemplate such an arrangement; nor was it the recognized practice before that directive came into effect. He stated that the directive was brought into effect to memorialize the official policy and to bring uniformity and consistency across the organization.

43        The official management leave policy outlined by Mr. Dolhai provided that employees could initiate a request for leave in compensation for overtime and that such a request could be verbal or made through email. No matter how it was initiated, he stated that it should have been entered in PeopleSoft as one of the five days of paid leave falling within managers’ discretion to recommend or to grant, depending upon their authority. He thought that all employees would be aware that such a day of leave had to be entered in PeopleSoft. In his view, this was different from a rearrangement of working hours.

44        No evidence was presented of any specialized training for timekeepers. Ms. Guttman testified that she was not aware of any specific iCase training for employees. Although she requested information on iCase training through an access to information request, she was told that there was no specific data or records on the training that had been given. However, in cross-examination, she acknowledged that she did not confirm with AJC members whether they had received timekeeping training. She acknowledged that the DOJ NTP was posted on the intranet.

45        Ms. Klee testified that she was not aware of any formal iCase training, although she received some informal training at a meeting. However, Mr. Dolhai testified that PPSC employees received timekeeping training at some point after the PPSC developed its timekeeping protocol. Furthermore, he stated that the PPSC NTP was posted on the intranet.

46        The AJC presented documentary evidence on consent that demonstrated employer concerns about the reliability of iCase reporting at both the PPSC and the DOJ. In 2011, the PPSC’s Internal Audit Division conducted an audit of vote-netted revenue and cost recovery. The final audit report was issued on June 9, 2011 (Exhibit G-1, Tab 21). It included observations on timekeeping at paragraph 3.3.1. It noted that while the majority of employees recorded their time in iCase over 66% of the time within the first week, no procedures or controls were in place to prevent inaccuracies or to ensure compliance. However, it was also noted that the PPSC had issued a Policy on Hours of Work, Overtime, Exceptional Leave, Travelling Time and Reimbursement of Meal Expenses for the Law Group (LA), effective February 20, 2010, which required lawyers to reconcile their hours worked at the end of each four-week period. The report indicated that PPSC management believed that timekeeping compliance would improve as a result of the policy.

47        Following an internal audit at the DOJ, a “Cost Recovery Framework” report was issued in July 2010 (Exhibit G-1, Tab 20). Its third section concerned timekeeping and noted the following at paragraph 3.1:

...

We found that some legal practitioners do not comply with NTP requirements regarding timely recording of their time into iCase. The practitioners interviewed noted that while they log their time on a daily basis, they often enter their time into iCase at the end of the month or later. LMPD [Law Practice Management Directorate] provided sample data for a random three-week period. These data showed that 33% of the total timekeeping hours were entered into iCase more than 10 days following the sample period. This can cause delays in processing billings, which, in turn, affects the cash position of the Department. The 2007 Internal Audit Report of iCase stated that time is often entered retroactively after a lengthy period and the lead counsel is not required to periodically provide an official record of assurance that the time recorded against files is correct. The report also noted the effect this has on data integrity: “There have been some poor timekeeping practices by certain sections or staff, such as bulk time entry against a few files and time recorded infrequently. These practices are likely to introduce inaccuracies.”

...

48        On February 3, 2011, a memorandum, with a copy of the DOJ NTP, was sent by Deputy Minister Myles Kirvan and Associate Deputy Minister Yves Côté to all LA employees at the DOJ to remind them that compliance with the timekeeping protocol was mandatory (Exhibit G-1, Tab 5).

49        On January 12, 2012, Mr. Kirvan and Mr. Côté sent another memo to DOJ employees in which they advised that the processes used for cost recovery would change. They also noted that timekeeping had to be strengthened in support of new cost recovery processes.

50        Ms. Guttman testified that AJC representatives and DOJ managers discussed inconsistencies with timekeeping and iCase entries. She stated that it was raised at a meeting that included Marco Mendicino, who was the AJC’s president at the time, Joel Oliver, who was at that time the acting assistant deputy minister in the DOJ’s Management Sector, and others, which was held on March 7, 2012. Unofficial notes of the meeting produced by the AJC but not approved by the employer referred to employer concerns about infrequent or inaccurate timekeeping (Exhibit G-1, Tab 5).

51        In late autumn 2012, the PPSC’s Internal Audit Division conducted an audit in the Atlantic region to review, among other things, the efficacy of the management control framework in the region. Mr. Dolhai testified that the division, which reported to Brian Saunders, Director of Public Prosecutions, conducted such audits regularly.

52        Among other issues, the audit uncovered a disconnect between leave recording in iCase and the leave records kept in PeopleSoft. In his report to Mr. Saunders on the findings of the audit (Exhibit E-10), Philip Morton, the PPSC’s chief audit executive, noted the following:

...

... As part of the audit, we conducted a comparison of leave entries in iCase with those in PeopleSoft. This analysis identified discrepancies and shortcomings in the internal controls over leave recording by timekeepers in PeopleSoft. On examination, it was discovered that this weakness in internal controls existed elsewhere in the PPSC. As a result, the Director of Public Prosecutions requested an extensive analysis of leave recording across the PPSC.

...

53        Mr. Dolhai testified that as a result of the audit findings, PPSC management decided that a reconciliation of leave recorded in iCase and PeopleSoft would be conducted in every region. Management letters were sent to the regional managers (Exhibit E-11) setting out the parameters of the reconciliation project. Mr. Dolhai testified that problems with leave monitoring had been identified as a factor that gave rise to the identified discrepancies. He stated that Mr. Morton had been tasked with determining whether faulty procedures contributed to the problem. As a result, the management letters identified the practices that managers should follow to ensure that leave monitoring policies and procedures were being followed.

54        He stated that the primary risk to the PPSC of failing to monitor leave properly was that employees could take paid leave or be compensated for leave to which they were not entitled and that, conversely, employees might not receive the leave to which they were entitled because they had made mistakes to their disadvantage. He testified that he knew of one employee who had entered leave twice in PeopleSoft and, as a result, had shortchanged himself two weeks’ leave.

55        Mr. Dolhai testified that management decided that timekeepers would be asked to explain the discrepancies identified between their iCase and PeopleSoft leave records going back six years. He stated that former employees were not asked to participate in the reconciliation exercise. Ms. Betts testified that the decision to reconcile leave data back to 2007 was made also because of the availability of that data. From 2010 onward, employees would be asked to explain any discrepancy of two hours or more. She testified that Mr. Morton and senior management were aware of inconsistencies in the practice of granting leave for personal and medical appointments and felt that a two-hour leeway would compensate for any such inconsistencies. For fiscal years 2007-2008, 2008-2009, and 2009-2010, employees would be asked to explain discrepancies only of greater than 75 hours in a year.

56        Mr. Dolhai explained that management had the most confidence in the records for the most recent period because iCase was established and collective bargaining was in place. He acknowledged in cross-examination that if employees had not kept records, it would be difficult to reconcile discrepancies from six or seven years earlier, and he agreed that employees were not required to keep extraneous records relating to their leave. He stated that that was the reason the PPSC established different thresholds for the reconciliation exercise. He believed that employees would remember the reasons for a discrepancy of 75 hours over the course of a year but acknowledged that it might be difficult to explain a discrepancy of 2.5 hours that occurred 6 years earlier.

57        Through the spring and summer of 2013, PPSC regional managers communicated with the timekeepers in their offices, either through team leaders or directly, about the reconciliation exercise (for example, see Exhibit G-1, Tabs 1, 2, and 3). Timekeepers with identified discrepancies between leave reporting in iCase and PeopleSoft were given packages of Excel spreadsheets that consisted of the identified discrepancies, their PeopleSoft records, iCase leave records, and PeopleSoft leave payouts. They were asked to review the documents and reconcile the discrepancies by either amending iCase or submitting the appropriate leave forms. Employees were also given an attestation form, which stipulated that the timekeeper had reconciled all the identified discrepancies in a manner consistent with the terms and conditions of employment in force at the appropriate time.

58        Mr. Dolhai stated that he understood that the reconciliation exercise was a strict comparison between the two electronic systems, based on the presumption that PeopleSoft was correct. He believed that employees would be able to determine if they had entered their leave in PeopleSoft simply by examining their leave balances in it for the year in question. He stated that he did not know of any employees who had to repay leave that they did not believe they owed. In cross-examination, he noted that the reconciliation process went both ways. Some had more leave recorded in iCase than in PeopleSoft, and some had less.

59        Mr. Dolhai also testified that PPSC senior management was emphatic that the reconciliation exercise was not disciplinary. It wanted to ensure that employees clearly understood that the employer was not suggesting that leave discrepancies had arisen because of dishonesty or fraud.

60        Ms. Betts, who succeeded Mr. Morton as the PPSC’s chief audit executive, testified that her office generated the documents, such as Exhibits E-1 and E-4, which identified the discrepancies between the PPSC timekeepers’ iCase records and the PeopleSoft leave records. Although the PPSC is a separate agency, the DOJ administers the PeopleSoft system applicable to its employees.

61        PPSC auditors asked the DOJ for the PeopleSoft data for all PPSC timekeepers for the period established by PPSC senior management. They received the requested information in an encrypted Excel spreadsheet. Ms. Betts stated that while the PPSC auditors could have accepted that the PeopleSoft data was inherently accurate, they did not. When they received it, they examined it to ensure that it met their parameters. In particular, they wanted to be certain that there were PeopleSoft transactions throughout the period that they requested. In essence, they conducted a “sore thumb” exercise, looking for gaps that would stand out obviously. Ms. Betts noted that by using that methodology, the auditors were able to identify three-month gaps in the data in some cases and therefore could ensure that they received all the data available. The PPSC auditors also requested the iCase records for all PPSC timekeepers, using the same parameters as for the PeopleSoft records.

62        However, Ms. Betts acknowledged that because of past audits, her office was aware that there were data integrity issues. She stated that when her office looked at a system, as a general rule, controls such as dropdown or mandatory fields had to be completed that would ensure that a field was completed correctly, but iCase has limited system controls. She acknowledged that as a result, there were problems with it.

63        Ms. Betts testified that in her opinion, the systemic problems in iCase arose not just from the lack of internal controls but also from the organization’s culture. She agreed that iCase recording was not always done daily as expected and that sometimes, at the end of the month, managers had to remind timekeepers to update their iCase records, which meant that data was being entered retroactively. She acknowledged that timekeepers did not always enter their data in iCase in the same manner. Some timekeepers provided much more detail than others.

64        Despite those inherent problems, Ms. Betts stated that in the year or two preceding the reconciliation exercise, iCase data had become more reliable because internal controls had been improved. Managers and timekeepers were required to review timekeeping records monthly and to attest to their accuracy.

65        To create the documents identifying the discrepancies, Ms. Betts explained that the auditors ran an audit program that allowed them to merge and select specific fields in PeopleSoft and iCase to identify any discrepancies between paid leave noted in iCase and recorded in PeopleSoft. In cross-examination, she acknowledged that the comparison between iCase and PeopleSoft was underpinned by the auditors’ assumptions that the information received from PeopleSoft was reliable and that there were no errors in the iCase records. She stated that from the audit perspective, the information was validated if no one said that it was incorrect. The information was validated via the reconciliation exercise. Timekeepers were asked to review the data and to confirm that the discrepancies were in fact discrepancies.

66        In cross-examination, Ms. Betts was shown copies of the spreadsheets (Exhibit G-18) sent to Ms. Klinger, a DOJ employee who had been employed by the PPSC between 2006 and 2009. The spreadsheet identifying leave discrepancies (Exhibit G-18, page 1) showed that between March 2007 and March 2008, a significant amount of paid leave reported in iCase had no corresponding PeopleSoft entries. The spreadsheet that set out all Ms. Klinger’s PeopleSoft records for the period of the reconciliation exercise (Exhibit G-18, page 2) showed no PeopleSoft entries for the period between March 19, 2007, and May 20, 2008, despite Ms. Klinger’s certainty that she had submitted leave requests for that period. Ms. Betts was told that Ms. Klinger was eventually able to obtain copies of the relevant PeopleSoft leave records. She was asked whether such a significant flaw in the data raised concerns for her.

67        Although not personally aware of the specifics of Ms. Klinger’s situation, Ms. Betts acknowledged that it appeared that there was a serious problem with the data and that had it been brought to her attention, she would have ensured that the data dump was corrected. She agreed that it might have been a difficult problem for employees to address who had not kept copies of their leave records. She acknowledged that validating the data was crucial because many potential issues could arise from the breadth of the auditors’ data request. She stated that ultimately, some of the data inconsistencies were explained.

68        In cross-examination, Ms. Betts acknowledged that as a result of the conclusions reached in the Atlantic regional audit report, even before the reconciliation exercise, the employer knew that there was a significant problem caused by the lack of controls built into PeopleSoft to ensure data accuracy. She agreed that before the reconciliation exercise, the controls were either so lax or so insufficiently implemented that there was an immediate and significant risk that the information in either iCase or PeopleSoft would be inaccurate. Therefore, as a result of the audit and reconciliation exercise, changes were made. Ms. Betts stated that while leave is still recorded in iCase, monthly meetings are held with timekeepers to ensure consistency. She stated that the integrity of the data in PeopleSoft has improved and that currently, it is at an acceptable level.

69        The PPSC’s leave reconciliation exercise took place through the late spring and summer of 2013. In May 2013, Ms. Driscoll, who was the DOJ’s chief audit executive from October 2011 to January 2014, presented a “Three Year Risk-based Audit Plan” (Exhibit E-19) to the Deputy Minister, in which she recommended that the DOJ’s Internal Audit branch scan the electronic systems. She testified that Mr. William Pentney, the deputy minister, was aware of the reconciliation exercise conducted at the PPSC and that he requested that the branch conduct a similar exercise as part of the environmental scans of systems recommended in the audit plan.

70        Ms. Driscoll testified that to prepare for the project, she contacted the PPSC to obtain a copy of the audit command language script that it had used. She explained that the auditors used the script, which was software (a computer program), to extract information from the systems. Because the DOJ and the PPSC shared systems, the department codes were the only difference in the script. She testified that the analyst who wrote and ran the script for the PPSC came to the DOJ to run the script.

71        The auditors ran the script and identified any anomalies in the data in PeopleSoft and in iCase. Ms. Driscoll explained that an anomaly was the absence of an expected result. In the context of the reconciliation exercise, an anomaly would arise when iCase showed that leave was taken on a particular day, but there was no corresponding PeopleSoft entry. She stated that initially, iCase data was the primary file searched to identify leave, and PeopleSoft was the secondary file. Later on, the audit team reversed the process and used PeopleSoft as the control. For many iCase entries, timekeepers had stated that they were on leave, but there were no corresponding PeopleSoft entries. In a smaller number of cases, timekeepers had recorded leave in PeopleSoft but had entered in iCase that they had worked that time.

72        In cross-examination, Ms. Driscoll acknowledged that when iCase was used as the primary point of comparison and it showed a leave entry for which there was no corresponding PeopleSoft leave entry, it left the impression that the employee in question owed leave. However, she stated that there could have been other explanations, such as a system error or an employee input error.

73        She also acknowledged that if a leave entry in PeopleSoft had no corresponding leave entry in iCase, the employee might be owed leave. However, she stated that when PeopleSoft was used as the primary file for comparison purposes, only about 10 discrepancies arose, and her office was able to verify them. Despite that, she confirmed that for the purposes of the Internal Audit branch, the reconciliation of PeopleSoft to iCase was, or should have been, part of the reconciliation exercise.

74        Asked if she was certain that her office caught all the discrepancies, she stated that it ran the data several times, always with the same results. She stated that she was not aware of all the issues that arose during the reconciliation exercise or of all the explanations employees provided, but she was confident that if data problems arose, she would have been notified. If managers had problems with the data, they could call her but were not required to, and she was not involved in the explanations or remedies that arose in individual situations.

75        The DOJ’s Internal Audit branch generated the documents that extracted the detailed PeopleSoft and iCase records of all DOJ timekeepers as well as the document that identified the discrepancies for each timekeeper (see, for example, Exhibits E-3 and E-4). The results were initially presented to the deputy minister, and then discussions occurred at the senior management level to determine how the reconciliation exercise would proceed.

76        When the data was presented to the deputy minister and senior management team, it was presented with and without the filters used by the PPSC. Ms. Driscoll stated that she had talked to the PPSC about the rationale of using a 75-hour filter for discrepancies arising in the first 3 years of the reconciliation exercise and a 2-hour filter for discrepancies arising in the more recent period, but she was not involved in any discussion on using such filters at the DOJ and could not say whether filters were applied. Senior management decided whether to apply them.

77        Ms. Driscoll testified that the Internal Audit branch was involved to a limited extent in the reconciliation exercise. She said that when anomalies were identified, they were run through a verification process, which entailed receiving evidence from managers and employees to determine what caused them. From the Internal Audit branch’s perspective, the expectation and intention were that the reconciliation process would test the systems’ accuracy and help ensure their integrity. It was an environmental scan of the systems rather than an audit, with the purpose of informing further auditing. She explained that to test the integrity of a system, the auditors would examine its design documents, flowcharts, and output, and they would then work backwards.

78        When asked if the Internal Audit branch tested iCase and PeopleSoft after obtaining access to the data and before the reconciliation exercise started, Ms. Driscoll stated that iCase had been audited a year-and-a-half earlier in a limited file testing based on sample data.

79        In cross-examination, Ms. Driscoll stated that she had been involved in the iCase audit at the reporting stage. She did not remember if the audit had concluded that there was a problem with the reliability of the data in iCase, but she agreed that the purpose of the reconciliation exercise was to test its integrity. PeopleSoft was also audited at about the same time. Because of those audits, she was confident that when the script was run on them both for the purposes of the reconciliation exercise, the data extracts obtained were correct.

80        In cross-examination, Ms. Driscoll was asked if she looked into the methods that timekeepers and managers used to deal with management leave and, in particular, the variations in the way it was administered. She stated that she was not aware of the issue specifically but acknowledged that she would have been concerned had discrepancies arisen from an inconsistent practice across the organization. However, she was not involved in any process to change the practices and systems controls arising from the reconciliation process.

81        The management of the DOJ reconciliation process fell under the jurisdiction of the Human Resources and Professional Development Directorate. Ms. Pratt, who was its director general during the period of the process, testified that the Office of Labour Relations and Compensation took the lead on handling any actions required during the exercise, with the assistance of the PeopleSoft team. She explained that once the Internal Audit branch identified discrepancies between employees’ iCase leave entries and their PeopleSoft entries, the Deputy Minister and Executive Committee determined that a reconciliation exercise would be undertaken at the DOJ and that timekeepers would be asked to account for all discrepancies identified for the fiscal years 2007-2008 to 2012-2013 (Exhibit E-20).

82        That directorate took the lead in writing talking points to help managers explain the exercise to employees (Exhibit E-21). Ms. Pratt stated that the Deputy Minister wanted a consistent message on the exercise across the organization. It was also considered important that employees understood that the exercise was not a witch-hunt and that they were not being accused of wrongdoing. The intent was to reduce any apprehension for the timekeepers asked to reconcile discrepancies. However, she testified that she believed that they arose because PeopleSoft’s records were not complete and accurate. She stated that she could not comment on whether there were other reasons for those inaccuracies, but to her, the obvious conclusion was that employees took leave that they failed to record in PeopleSoft. She stated that PeopleSoft was part of the department’s DNA and that she believed that professional, conscientious employees would use it.

83        The talking points developed by the HR directorate were sent to managers along with access to a shared electronic site containing the records and worksheets for each timekeeper in each manager’s section who had identified leave discrepancies. In cross-examination, Ms. Pratt clarified that the communications strategy developed by HR about the reconciliation exercise, including a fact sheet of questions and answers, was directed at the management team. Managers were expected to assemble and distribute the records and worksheets relating to leave discrepancies for the timekeepers in their sections. She stated that nothing prevented them from including the talking points or other information that they had received from HR in the packages that they prepared for the timekeepers.

84        DOJ timekeepers were expected to reconcile or explain all discrepancies arising from a comparison between their iCase leave entries and those in PeopleSoft going back to fiscal year 2007-2008. Management’s direction was that the exercise should be done as well as possible. Cases that fell within the “unexplained” category were to be taken to management for disposition. A range of things could be done in those circumstances within the terms and conditions of employment, although in cross-examination, Ms. Pratt stated that she was not aware of a single instance in which discretionary leave under clause 19.19 of the collective agreement was granted to cover an unexplained discrepancy; nor could she say whether managers were advised that leave reconciliation issues could be resolved by granting management leave.

85        In cross-examination, Ms. Pratt confirmed that if an employee could not explain a discrepancy, some kind of leave adjustment had to be made, such as drawing down existing leave entitlements to cover the unexplained absence or, if no leave entitlement was available, paying it off through future leave entitlements or salary deductions. She acknowledged that the exercise was predicated on the assumption that the iCase leave notation was correct, unless the timekeeper could prove otherwise. However, she also confirmed that it was discovered that in some cases, iCase was not correct.

86        Ms. Pratt testified that to reconcile leave discrepancies, timekeepers could go to past emails and calendars or to their past leave summaries in PeopleSoft. She did not recall being told that employees had difficulty accessing past data from PeopleSoft, but she was uncertain as to how many past fiscal years’ leave records were accessible.

87        In cross-examination, she could not confirm whether managers and timekeepers were advised that they could access paper leave records, and she was not certain how long such records would have been kept. She acknowledged that it was possible that employees were asked to clean out their email and calendars to reduce data storage. Although she could not recall being asked to get rid of old data from her calendar, she stated that she did not know how far back the records went. She never tried to access six- or seven-year-old data. She also stated that she personally did not keep such records.

88        Ms. Pratt testified that employees on long-term leave were also expected to participate in the reconciliation exercise. They were dealt with on a case-by-case basis and were invited to return to the office to participate in the exercise if they wished or to complete the exercise on their return from leave. She was not able to say how employees who chose to complete the exercise while on leave were compensated because she was not involved in that matter. Ms. Pratt was aware that the DOJ intended to ask former departmental employees to complete the reconciliation exercise, but that aspect of the exercise was completed after she retired.

89        Ms. Guttman testified that the AJC was advised about the reconciliation exercise at the PPSC in March 2013. She said that she received an email from an AJC member there about the possibility of the reconciliation exercise and that on March 27, 2013, the PPSC formally notified her about it at a labour-management consultation committee meeting. The DOJ notified the AJC about the reconciliation exercise through a conference call held with Mr. Pentney and others on July 12, 2013. She made informal notes of the call (Exhibit G-1, Tab 25).

90        In August 2013, Ms. Guttman became aware that the DOJ was asking employees on long-term leave to participate in the reconciliation exercise. She raised the issue with Colleen Laframboise, Director, Labour Relations and Compensation (Exhibit G-1, Tab 13), who advised that such situations would be handled on a case-by-case basis. Ms. Guttman testified that she believed that the DOJ would do whatever it needed to meet its objectives. She believed that managers would ask employees on long-term leave to complete the reconciliation exercise but that if they pushed back and refused, they would not be forced to. She acknowledged in cross-examination that she made certain assumptions about the employer’s position on employees on long-term leave but that she did not make them known to the employer because the AJC had already filed a policy grievance on the issue.

91        Mr. Butovsky, a lawyer employed in the National Security Group of the DOJ’s litigation branch, testified that he participated in the leave reconciliation exercise. Initially, he stated that he was asked to reconcile leave discrepancies in June 2012, when he had just returned to work part-time following a lengthy period of sick leave and was not yet fully functional. However, in cross-examination, he acknowledged that the reconciliation exercise took place in June 2013 and that he had returned to full-time hours in March 2013.

92        Mr. Butovsky received a package of spreadsheets that included a list of leave discrepancies (Exhibit G-2, page 3) in addition to a list of his PeopleSoft leave records from April 2006 to March 31, 2013, a list of his iCase leave records from April 2006 to March 31, 2013, and a list of his PeopleSoft leave payouts for the same period (Exhibit E-1). The list of discrepancies identified 3 days of leave (22.5 hours) in 2007, 1 day (7.5 hours) in 2008, 1 day (7.5 hours) in 2009, and 19 days (142.5 hours) in 2011 for which there were leave entries in iCase with no corresponding leave recorded in PeopleSoft.

93        Mr. Butovsky was able to explain that in 2007, on one of the days he had entered as leave in iCase, he participated in a Litigation branch conference, and that on the other two days identified as leave discrepancies in 2007, he was at work. Similarly, he stated that in 2008, he was at work and not on leave, despite his iCase entry. For the discrepancy identified in 2009, he was able to produce a leave form. However, the leave discrepancies for 2011 proved more difficult to explain.

94        Mr. Butovsky testified that he became ill in late 2010; as a result, he expected to undergo surgery in February 2011. On February 10, 2011, in anticipation of the surgery and expected recovery period, he entered in iCase that he would be on leave from February 11 until March 31, 2011. He did not formally request leave through PeopleSoft because, as he testified, he could not be certain of the terminal date of the leave, but his manager was fully informed about his situation. However, the surgery was postponed until he completed treatment for a different medical issue. Mr. Butovsky did not amend the iCase entries to reflect the change in his leave plans.

95        Due to the change in his medical situation, Mr. Butovsky was able to work some of the days that he had entered as leave days in iCase. He also took leave that was approved through PeopleSoft for some of the days he had expected to be on leave. In the end, he was able to explain all but 6 of the 19 days of identified discrepancies in 2011. Through discussions with his manager and the HR section, it was determined that in fact, he had gone on sick leave five days earlier than he had requested through PeopleSoft.

96        After a discussion with his manager, Mr. Butovsky submitted a leave form for five days of sick leave and a leave form for one day of volunteer leave to resolve the discrepancies. On the form that he submitted for sick leave, he requested that the leave be deducted retroactively from his 2011 leave bank. He testified that he made the request because if the leave had been deducted in the year that it was taken, his long-term disability benefits would have begun a week earlier. He acknowledged that he was partly responsible for the error but stated that he believed that the employer also bore some responsibility because his manager approved his sick leave forms even though she was fully aware of when he actually went on sick leave. He was told that the employer could not retroactively deduct the sick leave from his 2011 bank. Instead, it was taken from his 2013 sick leave bank, which left him in a negative position.

97        In cross-examination, he acknowledged that he lost no salary as a result of the reconciliation exercise and that he was not in a sick leave deficit as of the date of his testimony.

98        With respect to his volunteer leave request, Mr. Butovsky stated that he could not say for certain whether the day in question was in fact taken as a volunteer day or whether his manager just did him a favour to protect his sick leave. He explained that iCase does not specify what kind of leave is taken.

99        Ms. Pearce, a DOJ lawyer attached to the Parks Canada Agency, testified that she was informed about the reconciliation exercise in mid-July 2013. She stated that her manager convened a meeting of the lawyers in her section to explain the project and that a couple of weeks later, she received a package of documents outlining her leave issues from the acting manager of her section. She stated that she had to go through the package line by line to try to understand the issues and that she then had to find either paper leave forms or PeopleSoft records to explain or justify whether she had been on leave on the dates in question.

100        She testified that it was difficult to sort out what had happened. She had difficulty accessing some documents and information from her email server. She stated that it was not easy to obtain records from past years and that the data integrity was not good. The staff had been told to remove old data from their email accounts, to reduce the amount of stored email, and she did not keep old calendars. The passage of time between 2007 and 2013 made it difficult to reconstruct records, and no warning had been issued that employees should keep records to justify their leave usage.

101        The list of discrepancies (Exhibit E-4), which Ms. Pearce identified as being similar to the document that she had been given, identified 12 leave entries in iCase (85 hours) in 2007, 8 leave entries (35 hours) in 2008, 13 leave entries (46.38 hours) in 2009, 9 leave entries (54.25 hours) in 2010, 2 leave entries (5.5 hours) in 2011, and 1 leave entry (2.5 hours) in 2012, for which there were no corresponding PeopleSoft records. She stated that she was surprised that there were discrepancies because she always put in leave requests using either PeopleSoft or paper leave forms.

102        Ms. Pearce estimated that she spent a week trying to reconcile the discrepancies between iCase and PeopleSoft. She was able to resolve everything other than a two-week period in August 2007, a one-week period in August 2010, and one other day, which she did not identify. She stated that she simply could not explain the discrepancies. She knew what leave she had taken, but she did not know what had happened to those particular weeks. Because she could not explain what had happened, she stated that she was required to put in a leave form for that time, which depleted what remained of her 2013 bank of vacation leave and reduced her 2014 leave bank by four days.

103        In cross-examination, Ms. Pearce stated that some of the discrepancies were easily resolved but that it took a while because she had not kept records with the thought that there would be an audit in the future. She could not find any information relating to the vacation leave in question and could not say that she did not take it. She reiterated that she simply could not explain the discrepancies. She could not remember if she had planned holidays that she was subsequently unable to take because she had been required to work.

104        When questioned about the iCase leave entries for August 6 to 17, 2007, she stated that she would never have gone on leave without her manager’s approval and that it was her practice to submit leave applications through PeopleSoft, but she could not reconstruct what had happened so many years earlier. She acknowledged that sometimes managers wanted to know in advance when employees planned to take summer vacations and noted that on some occasions, she had to change leave plans because of work requirements. She stated that she never took long vacations and that she always had some leave left over at the end of fiscal years, although the amount would vary from year to year. Although she had never before reconciled iCase with PeopleSoft until the leave reconciliation exercise, she did keep careful track of her PeopleSoft balances, and she would have noticed had she had extra leave in her bank at the beginning of fiscal year 2008-2009.

105        Ms. Klee was a lawyer in the DOJ’s Legal Services at the Department of Indian Affairs and Northern Development between November 1991 and January 16, 2013, when she retired. In February 2015, Ms. Klee received a letter (Exhibit G-6) from Éric Trépanier, Director General and Deputy Chief Financial Officer at the DOJ, informing her about the leave reconciliation exercise. Spreadsheets containing a list of leave discrepancies (Exhibit G-7), Ms. Klee’s PeopleSoft leave records (Exhibit G-8), and her iCase leave records (Exhibit G-9) were attached to the letter, along with an attestation form. She testified that she understood that she was being asked to explain each of the entries on the discrepancy list (Exhibit G-7), which amounted to about 51 days.

106        Ms. Klee testified that in February 2015, she had almost no work records. She did not have her calendar and other related documents because she had thrown them out on her retirement, considering that they were no longer relevant. She managed to find a PeopleSoft printout (Exhibit G-11) that she had made before she retired, but that was the only record that she possessed. She contacted her former administrative assistant to see if there were any records of her leave. The assistant sent her two leave requests and two medical certificates (Exhibit G-10). Her supervisor had not signed either leave form.

107        In cross-examination, Ms. Klee explained that the leave applications that her former assistant sent her were paper forms, not electronic requests, and that they had been given to the administrative assistant for processing at the time the leave was taken. She stated that she had been told that the leave was approved, although she did not see the signed forms. She said that she would have followed up to ensure that the leave had been entered in PeopleSoft.

108        Those forms covered two periods of leave. Exhibit G-10, page C, is a leave application covering 112.5 hours of certified sick leave taken between February 29 and March 20, 2008. Each day covered by the leave application appears on the list of discrepancies (Exhibit G-7). In cross-examination, Ms. Klee stated that although that leave does not appear to have been recorded in PeopleSoft, she periodically reconciled her leave records and her calendars, and she would have noticed at the time if it had not been recorded in PeopleSoft.

109        The other leave form (Exhibit G-10, page A) that Ms. Klee’s former assistant sent her covered leave from February 28 to March 18, 2011, when Ms. Klee took 112.5 hours of certified sick leave, and from March 21 to March 31, 2011, when she took 67.5 hours of vacation leave (Exhibit G-10, page A). Although each day of leave taken between February 28 and March 31, 2011, appears on the list of discrepancies to be explained (Exhibit G-7), Ms. Klee stated that she believed that in fact the leave was entered in PeopleSoft but incorrectly. She noted that on the last page of the PeopleSoft records sent to her as part of the reconciliation exercise (Exhibit G-8), 112.5 hours of certified sick leave and 67.5 hours of vacation leave were entered as a single day of leave taken on April 1, 2011. There was also a leave entry for 22.5 hours of certified sick leave taken between April 1 and 5, 2011. She believed that her leave from February 28 to March 31, 2011, was wrongly attributed to a single day, April 1, 2011.

110        Ms. Klee was also able to resolve 9 out of 10 discrepancies alleged to have occurred in 2012 because she had the PeopleSoft leave printout for the 2012 fiscal year (Exhibit G-11) that she had made before she retired. It included her PeopleSoft leave summary for the 2012 fiscal year (Exhibit G-11, page B) and showed that 9 days of leave had been approved and recorded in PeopleSoft, although they were not included in the PeopleSoft leave records (Exhibit G-8) sent to her as part of the reconciliation package.

111        Ms. Klee testified that because her former administrative assistant was able to find two leave forms and medical certificates, and because she had a copy of her leave records for 2012, she was able to explain all the discrepancies on the list sent to her but for 7.5 hours on June 16, 2008, 15 hours on October 13 and 14, 2010, and 7.5 hours on December 31, 2012. Pressed in cross-examination, she would not acknowledge that the iCase leave entries for those days were necessarily correct. She stated that she did not remember using an iCase code entitled “Administrative/Management/Leave”, although she agreed that she had entered her leave in iCase. She contended that the leave entries that she could not explain might have been made in error. She stated that she was not prepared to accept that the iCase records sent to her were necessarily correct unless she could verify them.

112        On February 24, 2015, Ms. Klee received a second letter from Mr. Trépanier (Exhibit G-12), advising her of data anomalies in the reports sent to her as part of the reconciliation exercise. She was asked to wait for further instructions before completing the exercise.

113        Ms. Klee responded to Mr. Trépanier on March 2, 2015 (Exhibit G-13). She acknowledged the previous correspondence and asked for, among other things, copies of all her email exchanges with her managers and their assistants during the relevant period, copies of her electronic calendars for the relevant period, and copies of her iCase leave entries as she made them because she did not recognize the terminology used in the iCase report that she received. She also asked to be paid for the time to work through the reconciliation exercise, given that it would take her some time to recreate her leave information, and she was no longer at work. She testified that she received no response to her request from Mr. Trépanier or from anyone else. She was not provided with the documents that she requested. As of the date of her testimony, she had not been asked to finish the reconciliation exercise.

114        Mr. Saheb-Ettaba is a DOJ lawyer working in the Legal Services branch of the Department of Fisheries and Oceans. He received a package of documents (Exhibit G-14) relating to the leave reconciliation process from his manager, Daniel Roussy, sometime in the spring or summer of 2013. He also received a list of leave discrepancies (Exhibit G-17), which identified 3 totalling 15.25 hours in 2007, 3 totalling 8.75 hours in 2008, and 1 of 7.5 hours in 2010.

115        Mr. Saheb-Ettaba did not believe that he could refuse to carry out the reconciliation. He stated that he understood that he was to do it quickly and in the manner specified. He found the exercise’s instructions very narrow in that the solutions to any discrepancies were dictated and that no other solutions would be accepted. He stated that if an explanation for a discrepancy did not fit within the options provided, a leave form would be required.

116        In attempting to reconstruct his leave data for the period that the employer identified, Mr. Saheb-Ettaba discovered other errors and discrepancies between PeopleSoft and iCase that the employer had not identified. In particular, he found that leave times had been entered in PeopleSoft that he had not taken and that did not correspond with his iCase entries.

117        Mr. Saheb-Ettaba prepared Annex A (Exhibit G-14). In that document, he addressed the discrepancies the employer identified for leave reported in iCase but not recorded in PeopleSoft. He was able to explain that a day of sick leave on June 13, 2007, noted in iCase but not in PeopleSoft was in fact a day of compensatory leave approved by his manager. That explanation was accepted. However, he was unable to explain the remainder of the discrepancies identified by the employer, so he submitted the appropriate leave forms for those absences, which amounted to 24 hours. However, in Annex A, he also identified errors in PeopleSoft that were to his advantage. He noted that between 2008 and 2011, 29.25 hours of leave were recorded in PeopleSoft that he had reported in iCase as work time.

118        In cross-examination, Mr. Saheb-Ettaba acknowledged that the discrepancies in his favour that he identified amounted to a half hour here and there and were based on his belief that his iCase entries were accurate. He said that because he made the iCase entries, he knew that they were accurate. However, he did not necessarily accept the PeopleSoft entries as accurate because he had not made them. He stated that he used paper leave forms that someone in the compensation area had entered into PeopleSoft. He explained that he started using PeopleSoft only a few years ago. He also stated that he kept most of his calendars and records of work, although he did not keep copies of his leave forms. He did not know what happened to the forms once his manager signed them.

119        In the course of reviewing his leave and iCase records for the reconciliation exercise, Mr. Saheb-Ettaba noted that on a number of occasions between 2007 and 2013, he could have claimed compensatory leave for weekend travel time but did not. Before he submitted his attestation form and Annex A (Exhibit G-14), he emailed his manager (Exhibit G-16) and proposed to claim that time against any leave that he might owe as a result of the reconciliation exercise. He testified that Mr. Roussy refused his request because it did not fall within the parameters established by the DOJ for dealing with leave discrepancies. In cross-examination, he stated that he had not claimed all the travel time he believed that he could have claimed because it was a matter of principle to him not to make a retroactive claim, but he wanted to claim enough travel time to cover the debt that the employer said he owed.

120        Both Ms. Driscoll and Ms. Richard were asked about Mr. Saheb-Ettaba’s contention that his review of his iCase and PeopleSoft records revealed not only that he had failed to report some leave in PeopleSoft but also that he had reported more leave in PeopleSoft than his iCase records indicated he had taken. Ms. Driscoll stated that she did not know Mr. Saheb-Ettaba or his situation. She stated that she was not made aware of all the issues that arose as a result of the exercise. She believed that the reconciliation of PeopleSoft to iCase should have included corrections to both systems where mistakes could be demonstrated. Ms. Richard stated that she looked at Mr. Saheb-Ettaba’s records in preparation for her testimony. She brought to the hearing copies of his PeopleSoft records (Exhibit E-23, Tab 4) to show that he had entered the leave into PeopleSoft himself. She stated that in every instance that the record identifies the supervisor who approved the leave, Mr. Saheb-Ettaba would have entered the leave in PeopleSoft himself.

121        Ms. Klinger is a lawyer with the DOJ. She testified that when she started at the DOJ in 1991, she worked in the Federal Prosecution Service. When, in 2006, the Federal Prosecution Service became the PPSC, a separate agency, she became its employee. In 2009, she moved back to the DOJ to a position in the Legal Risk Management Division. In 2012, she moved to a position in National eDiscovery and Litigation Support Services, which was also part of the DOJ. She held that position at the time of her testimony. She worked part-time.

122        Ms. Klinger testified that her manager, Mr. Vezina, sent her a package of documents relating to the reconciliation exercise (Exhibit G-18) sometime around July 19, 2013. She was shocked by the size of the list of discrepancies because she believed that she had been conscientious about entering her leave forms throughout her career. The list of discrepancies (Exhibit G-18, page 1) identified 14 leave entries (63.5 hours) in 2007, 11 leave entries (61 hours) in 2008, 6 leave entries (28 hours) in 2010, 6 leave entries (27.5 hours) in 2011, 5 entries (30 hours) in 2012, and 2 leave entries (5 hours) in 2013 in iCase for which there no corresponding leave entries in PeopleSoft.

123        Ms. Klinger stated that she was not able to look at her old PeopleSoft records online. In cross-examination, she explained that she was able to access only the records for the current fiscal year. To view records from previous fiscal years, she would have had to have someone from PeopleSoft pull them for her. She stated that it was the same for iCase; she could not access old records. However, she stated that she kept some records of the emails approving her PeopleSoft requests going back to 2007. Although her record keeping was not perfect, she stated that her records were fairly good.

124        When she reviewed the PeopleSoft records of her leave that were attached to the reconciliation package that Mr. Vezina sent her (Exhibit G-18, page 2), she noticed that there were no leave entries in them between March 23, 2007, and May 20, 2008. She stated that a 15-month gap in leave jumped out at her because she knew that she took leave every year, but the PeopleSoft record made it look as though she had not taken any for that year. Reviewing the email records that she had kept, she found 15 to 20 approvals for leave taken between those dates.

125        She testified that she explained what she had discovered to her manager and that she asked that all her PeopleSoft records for the period in question be printed and given to her. She stated that she received those records (Exhibit G-20). They contained more information than did the records that she had kept. It was evident that the data was in the system but that it had not been extracted when the reconciliation documents were prepared.

126        Ms. Klinger stated that using the full PeopleSoft records that she had requested, she was able to demonstrate that many of the leave discrepancies in 2007 and 2008 identified by the employer in the reconciliation documents were in fact not discrepancies. She could prove that she was on approved leave for all but one of the days listed in the gap period, and on that day, she could demonstrate that the iCase record was incorrect and that she had not been on leave but on training (Exhibit G-19, page 3).

127        Ms. Klinger was also able to demonstrate that she had taken an approved day of management leave on September 29, 2011, which was also listed as a leave discrepancy (Exhibit G-22). She testified that she had requested the day off work in lieu of overtime and that Mr. Vezina had approved it. She said that she would not have entered such a day in PeopleSoft for that kind of leave. In an email to Mr. Vezina on August 13, 2013 (Exhibit G-19, page 3), she noted that she had been routinely granted time off work in lieu of overtime for the purpose of a religious observance, although she could not provide email evidence of approvals for any days other than September 29, 2011. She testified that she was certain that there was more than one such day of management leave but that she could not find any records to establish it.

128        Although Ms. Klinger was able to explain some of the discrepancies identified by the employer, she was not able to explain them all. She testified that she asked for her PeopleSoft records only for the period that was obviously missing. She believed that her iCase records were correct because she had made those entries, but she could not assert that PeopleSoft was correct.

129        Ultimately, she was required to submit leave forms for 10 days of leave or portions of that time (Exhibit E-7). In cross-examination, she explained that she reviewed her calendars and email to see why she was absent on the days for which she could find no PeopleSoft records. The leave application (Exhibit E-7) was based on that review if she had not noted in iCase why she was on leave.

130        Concerned about the missing data from her PeopleSoft records, and concerned that had she not discovered the problem, she would have been required to claim the leave in question again, Ms. Klinger sought an explanation from Mr. Vezina, who forwarded her concerns to HR (Exhibit G-19). She testified that she was told that leave records from the period she worked at the PPSC were not captured by the auditors when they created the reconciliation records. Ms. Klinger did not accept that explanation because the PeopleSoft records attached to the reconciliation documents (Exhibit G-18, page 2) showed leave entries before and after the gap that covered the time she worked at the PPSC.

131        Ms. Driscoll was questioned in cross-examination about the anomalies in Ms. Klinger’s PeopleSoft records. She explained that leave that Ms. Klinger took while employed at the PPSC probably would not have showed up in the records pulled by the auditors. However, she could not explain why some of Ms. Klinger’s leave requests while at the PPSC were captured and others were not. She testified that within a few days of starting the reconciliation exercise, the auditors became aware of problems in the data because people had moved into and out of the DOJ. They received phone calls from people who had left the department and had returned about problems in the data. Ms. Driscoll believed that managers were notified that they should handle employees who had transferred to the PPSC differently, but she was not directly involved. She testified that she informed senior management but that she did not know whether non-managerial timekeepers were also informed about the issue.

132        Mr. Vezina testified that he was the manager responsible for the reconciliation exercise in the division in which Ms. Klinger worked, although she reported directly to a different manager. On July 25, 2013, Ms. Klinger advised him about the gap in the PeopleSoft records that formed part of the reconciliation documents he had given to her (Exhibit E-25). His assistant, Louise Madone, provided Ms. Klinger with a complete set of her PeopleSoft records, which included her time employed with the PPSC (Exhibit G-20). Ms. Klinger completed the reconciliation exercise and emailed a copy of the completed documents to him for his review. When she brought hard copies of the documents to his office for him to sign, she also asked for an explanation of the missing PeopleSoft data (Exhibit G-19). He followed up on her request as quickly as he could after his return from leave (Exhibit E-24). He received a copy of the explanation provided to her by the HR branch (Exhibit G-19). As far as he was concerned, that was the end of the story. In his opinion, their interactions over the reconciliation were cordial.

133        In cross-examination, Mr. Vezina stated that he had management responsibility for between 6 and 10 people who completed the reconciliation exercise, and he believed that some of them had to repay leave. He noted that he also completed the reconciliation exercise and that he had had to repay some sick leave. He could not say whether he was wrongly charged. He stated that he could not say one way or the other whether he simply forgot to put in a leave form. In re-examination, he noted that the reconciliation exercise was not pleasant for anyone.

134        Asked in cross-examination whether he had the delegated authority to grant discretionary leave under clause 19.19 of the collective agreement, Mr. Vezina stated that he did not know. He could not remember ever seeing the collective agreement provision before it was brought to his attention in cross-examination, and he did not recall whether it was brought to his attention that he had the authority to grant such leave in the context of the reconciliation exercise. He did not recall what options were given to people who had to repay leave as a result of the exercise, although he noted that in one instance, an employee on secondment was sent to HR to deal with issues arising from the reconciliation. He noted that Ms. Klinger reported to a different director who would have been responsible for any discretionary leave granted as a result of the reconciliation exercise.

III. Summary of the arguments

A. For the AJC

135        The AJC stated that the issue to be resolved was whether the leave reconciliation and recovery process imposed by the employer violated the collective agreement, specifically clause 5.02, article 13, and clause 19.19. It clarified that it did not contest the employer’s right to review leave records with a view to determining whether additional controls were required to ensure the appropriate application of the collective agreement or departmental policies or to ensure that record keeping was appropriate and adequate. Rather, the AJC contended that the employer’s leave reconciliation process was unfair and unreasonable.

136        That process was unfair and unreasonable because it placed the burden of explaining any discrepancy between iCase and PeopleSoft on employees. It was based on the presumption that if the iCase records identified that an employee was on leave on a particular day and there was no corresponding PeopleSoft leave record, the iCase record was accurate, and therefore, the employee’s leave records had to be adjusted. However, the presumption that the iCase records were correct, absent proof to the contrary, is incompatible with the evidence. Furthermore, the corresponding presumption that PeopleSoft was correct when it identified a discrepancy is also incompatible with the evidence.

137        The evidence established that the reconciliation process began after an internal audit at the PPSC identified discrepancies between iCase and PeopleSoft entries. However, it is important to understand the purpose of the two systems. iCase is a file-management system, and PeopleSoft is an HR system that tracks and manages employees’ leave usage. Before PeopleSoft was introduced in 2003, employees submitted paper leave forms for approval. After 2003, they submitted their leave requests electronically to their managers through PeopleSoft. They did not request their managers’ approval for leave through iCase. The employer’s witnesses clearly testified that iCase is not a leave management tool, even though employees were required to note leaves of absence in it.

138        Although witnesses testified that employees were expected to track their time contemporaneously, they also acknowledged that from time to time, managers would have to remind employees to update iCase. Therefore, iCase entries could have been made at any time after the day in question. Even though Mr. Dolhai and others stated that iCase became more reliable, no one stated that in 2008, or in any subsequent year, it was reliable for the purpose of tracking leave.

139        Furthermore, not a single employer witness testified that it was legitimate to treat all references to leaves of absence in iCase as being presumptively correct. In the absence of that critical element, it was not reasonable to use iCase as the standard against which other information would be compared.

140        The execution of the reconciliation process was also unfair and unreasonable. Managers were provided with a package of information to explain the process, along with a list of questions and answers, but there was no evidence that the information was also provided to employees. Although the employer had identified problems with the data used in the process, there is also no evidence that employees were advised about it.

141        For example, Ms. Driscoll testified that early in the process, the employer knew that there was a problem with the data relating to employees who had transferred from the DOJ to the PPSC, but she did not know whether it had been brought to the attention of the managers responsible for implementing the process or to the employees’ attention. However, the evidence was clear that the problem was not brought to Ms. Klinger’s attention; she was left to figure it out on her own. She testified that if the gap in her PeopleSoft records had not been so long, she might not have noticed that there was a problem. Equally troubling is that she faced difficulties when she tried to obtain information and an explanation for the problem with her leave record.

142        Other problems arose with identifying discrepancies between iCase and PeopleSoft. For example, during periods when employees could not access PeopleSoft and had to use paper leave forms, their leave was entered into the system by employees responsible for PeopleSoft or HR employees. Exhibit G-8 demonstrates some of the problems that could arise as a result. On page 4 of that document, 112.5 hours of sick leave and 67.5 hours of vacation leave were entered as if taken on a single day, April 1, 2011, creating a false discrepancy because the PeopleSoft leave entries were not applied to the appropriate days. The problem was easily caught because it was obvious; no one takes 112 hours of leave on a single day. However, if it had been three or five hours, it might not have been as easy to explain the resulting discrepancy.

143        The methodology the employer used to test the data that identified discrepancies between leave entries in PeopleSoft and in iCase was not foolproof. Ms. Betts testified that the PPSC performed a “sore thumb” exercise via examining the data generally to determine if there were any significant gaps. She acknowledged that that exercise would not catch a gap of a month or two because it was looking for broad gaps that would stand out like a sore thumb.

144        Ms. Driscoll testified that the DOJ tested the data by running its script repeatedly. She believed that the data was accurate because the same information arose each time. However, that exercise did not increase the data’s reliability; it simply demonstrated its consistency. There is a difference between reliable data and consistent data. She did not testify that the data’s validity had to be tested internally because she believed that it would be validated through the reconciliation process.

145        There was also evidence that not all leave was recorded in PeopleSoft. There was an accepted practice, both before and after the coming into force of the collective agreement, in which a manager could tell an employee not to come to work on a scheduled work day in recognition of the fact that the employee had worked long hours over the course of a weekend or had travelled in addition to working on a work day. Mr. Vezina testified that such arrangements would not have been entered into PeopleSoft and probably not into iCase because they simply involved rearranging working hours. However, there was no standard practice. Such arrangements were informal and were not necessarily documented. In that context, employees had to determine whether to enter one as a day of compensatory leave or make no iCase entry on a normally scheduled workday.

146        A day of leave recorded in iCase without a corresponding PeopleSoft record of leave would have been identified as a discrepancy that had to be explained as a result of the reconciliation exercise. In the absence of any specific recollection of the event or supporting documents such as email or memos concerning such informal leave arrangements, employees who could not explain discrepancies were required to reimburse the employer for the leave taken.

147        There is a certain irony to a situation in which a manager, in exercising his or her discretion, proposed that an employee take a day off, in recognition of the long hours that the employee had worked, only to require the employee to reimburse the day years later because it could not be documented or explained. Such a situation could not be consistent with the reasonable administration of the collective agreement required by article 5.

148        The AJC noted that the leave reconciliation exercise began following an audit in the PPSC that was intended to identify inconsistencies in leave applications recorded in PeopleSoft. The purpose of the exercise was to make sure that internal controls were appropriate and that the data being used was as accurate as possible. It was not intended to be a moneymaking exercise. It was not intended that employees who were unable to prove that iCase was wrong should have to reimburse the employer. Mr. Dolhai testified that if an employee came to the end of the exercise and could not explain a discrepancy but maintained that he or she had properly applied for leave, then that person should not have been required to reimburse the employer. He believed that conclusion was implicit in the exercise.

149        However, the reconciliation exercise did not operate that way. Employees who discovered discrepancies between iCase and PeopleSoft that worked in their favour were not credited with that leave; nor were they entitled to set it off against any amount that the employer determined was owed as a result of the exercise.

150        Furthermore, employees were not advised that they could access other forms of discretionary leave, such as leave with pay under clause 19.19. For example, Mr. Vezina was not aware that discretionary leave was available or that it was an acceptable way to resolve discrepancies that arose from the reconciliation exercise. In fact, he testified that another manager would have had to approve the exercise of such discretion as it applied to Ms. Klinger, even though he was responsible for conducting the reconciliation exercise and for arranging any necessary leave reimbursements.

151        There was also evidence that documents such as emails, calendars, and memos that might have helped employees explain discrepancies were deleted over time. Employees were not told to maintain such records in the event that they might have to explain their leave records years later. Furthermore, there was evidence that the employees who administered PeopleSoft had the capacity to make manual entries and to overwrite and change data. While such actions would have required supporting documentation, none was provided to employees during the reconciliation exercise. Employer witnesses could not state whether such documentation still existed.

152        As a general premise, the process used could be considered fair only if all the available data was provided to employees, if the employer was transparent about potential problems with the data, and if memories were both intact and reliable. But that was not so with the reconciliation exercise.

153        It was not reasonable to require employees to reconcile leave usage with iCase entries going back to 2007-2008. The PPSC acknowledged the difficulty when it set a threshold of 75 hours on the understanding that memories from 6 or 7 years earlier were unlikely to be intact or reliable on the question of whether a person took a half-day of leave. But that threshold did not resolve the problem because if there was a single 65-hour discrepancy at any point during the period of the exercise, an employee could quickly reach the 75-hour threshold. Employees testified that it was difficult to reconstruct what had happened years earlier.

154        Citing a series of PSLRB decisions rendered by Adjudicator Pineau on grievances heard at the same time at expedited adjudication, the AJC contended that a considerable delay recovering overpayments put employees at a significant disadvantage (see Lafrance-Legault v. Canada Revenue Agency, 2009 PSLRB 162; Lepage v. Canada Revenue Agency, 2009 PSLRB 163; Marchand v. Canada Revenue Agency, 2009 PSLRB 164; Ouellette v. Canada Revenue Agency, 2009 PSLRB 165; Parazelli v. Canada Revenue Agency 2009 PSLRB 166; Alain v. Canada Revenue Agency,2009 PSLRB 167; Bolduc v. Canada Revenue Agency,2009 PSLRB 168; Butler v. Canada Revenue Agency,2009 PSLRB 169; Larose v. Canada Revenue Agency,2009 PSLRB 170; Therrien v. Canada Revenue Agency,2009 PSLRB 171; and Lachapelle v. Canada Revenue Agency,2009 PSLRB 173).

155        In the cited cases, the adjudicator considered the situation of employees who were asked to repay vacation leave overpayments following a reconciliation exercise that uncovered errors in the calculation of leave entitlements for employees who had breaks in service. In many cases, the employer had made the errors years earlier.

156        The adjudicator held that even though s. 155(3) of the Financial Administration Act (R.S.C. 1985, c. F-11; FAA) gave the employer the discretion to recover overpayments, in the cases before her, the exercise of that discretion had been unreasonable. She held that a distinction had to be made between salary overpayments and leave recovery and that the employer’s delay asserting its claims of leave overpayment meant that the grievors could not dispute them because they had not kept proof of their positions.

157        In Murchison v. Treasury Board (Department of Human Resources and Skills Development),2010 PSLRB 93, the adjudicator followed a similar approach in his analysis of s. 155(3) of the FAA, noting that a distinction had to be made between recovering salary overpayments and recovering leave credits.

158        The AJC noted that in all cited cases, the fact that the employer updated the employees’ leave entitlement statements annually was one of the significant operating factors, as was the fact that the employer had to approve leave usage.

159        The AJC also argued that the employer’s requirement that employees prove that there was a mistake in either the iCase or PeopleSoft entries violated most of the principles set out at paragraphs 34 and 35 of Lumber & Sawmill Workers’ Union, Local 2537 v. KVP Co. (1965), 16 L.A.C. 73.

160        Citing Lapointe v. Treasury Board (Department of Human Resources and Skills Development),2011 PSLRB 57, the AJC contended that when an employee relies on the employer’s actions or statements to his or her detriment, the employer cannot enforce its strict legal right.

161        The AJC noted that there was evidence that employees, such as Mr. Butovsky, had been specifically prejudiced by the employer’s actions. Employees relied on the accuracy of the leave information that they received annually from the employer. They planned their holidays, used sick leave, and accessed long-term disability benefits based on the information that they were given. There was also evidence that the reconciliation exercise was applied to employees on leaves of absence and on those who had left the department either for new employment or for retirement.

162        The AJC submitted that given the decision to bifurcate the hearing, all that was required at that point was a finding that the collective agreement had been violated. The question of the appropriate remedy, in addition to any jurisdictional questions, should be determined in the second phase of the hearing.

B. For the employer

163        The employer contended that it had the authority to carry out the reconciliation exercise in 2013. It is wrong to suggest that it was carried out too late. Given the employer’s objective and the nature of the exercise, it only makes sense that it covered a set period. Ms. Driscoll testified that electronic monitoring, such as that used in the reconciliation process, is useful to assure the employer that controls are in place in addition to providing a test of the reliability and integrity of the data in iCase and PeopleSoft, two key electronic systems used by the employer and its employees.

164        The leave reconciliation processes at the PPSC and the DOJ were conducted pursuant to the employer’s policies on internal audits and controls (Exhibits E-16 and E-17). They require departments to ensure that they regularly monitor, review, and audit their internal systems, to minimize risk.

165        Mr. Dolhai and Ms. Betts testified that the discovery of inconsistencies in leave reporting flowed from the PPSC’s Atlantic region audit results (Exhibit E-9). Ms. Driscoll testified that when she became the DOJ’s chief audit executive, she noted that the DOJ had not audited its electronic systems, as had been done elsewhere. She developed a three-year risk-based audit plan that included auditing electronic systems (Exhibit E-19).

166        As a result of the auditing at both the PPSC and the DOJ, the employer discovered leave anomalies that required explanation. To do that, it turned to the obvious constituencies within its workplaces who could provide the appropriate responses to the anomalies. Timekeepers, whose duties were of such a nature that they were required to track their time, formed one of the relevant constituencies. Managers, who were responsible for monitoring not only the tracked time but also leave entitlements, formed the other relevant constituency.

167        To obtain the necessary explanations, at the PPSC, the employer implemented the recommendations made by the Chief Audit Executive in his report following the Atlantic region audit. Management letters were sent to PPSC regional managers that identified the necessary corrective actions and that established the broad outline of the leave reconciliation process (Exhibit E-11). From those letters, emails were sent to employees that explained how the exercise was to be carried out. Examples are in Exhibit G-1, Tabs 1, 2, and 3.

168        A similar process was undertaken at the DOJ, as Ms. Pratt explained. The communication that launched the exercise there (Exhibit E-21) was sent to the Executive Committee and then forwarded to the managers. It provided instructions on how to deal with leave anomalies. The objective was to ensure that at the end of the process, there would be explanations for the anomalies, and that certain corrections and leave adjustments would flow from those explanations, as evidenced in the attestations that accompany the reconciliation documents (see, for example, Exhibits G-2 and G-14).

169        The employer noted that the AJC was informed early in the process about its intention to engage in the reconciliation exercise at both the PPSC and the DOJ. The AJC was given copies of the talking points and templates for the DOJ in July 2013. The PPSC also communicated with the AJC.

170        It is important to note that nowhere in the process did the employer communicate or insist on any presumption about the reliability of either iCase or PeopleSoft. To compare the two systems, the employer had to choose somewhere to start. It was reasonable to choose iCase as the primary system because the timekeepers had entered the information in it. Therefore, the employer trusted that iCase was accurate. Mr. Dolhai testified that employees diligently entered their time in iCase. Mr. Saheb-Ettaba also testified that he was confident in the accuracy of his iCase entries. In fact, none of the timekeepers who testified disputed the accuracy of their iCase entries.

171        The employer contended that the evidence demonstrated that the root data in PeopleSoft was also accurate. It is important not to confuse the root data in PeopleSoft with the data reports extracted by the auditors from the Internal Audit branch for the reconciliation exercise. In some cases, such as in the evidence about the missing data in Ms. Klinger’s reconciliation documents, there was a problem with how the data was extracted because there was an issue with the script. In fact, that example also demonstrates the reliability of the data in PeopleSoft because the employer was able to provide Ms. Klinger with the missing data fairly quickly.

172        To suggest that there was a glitch in the system or that PeopleSoft is unreliable would be to take the example too far. There was an issue with the script that extracted the data, which was resolved. Ms. Driscoll testified that she was aware of only 20 employees who had issues with the data similar to those of Ms. Klinger. It is not appropriate to extrapolate a widespread problem from a few cases.

173        Similarly, it would be incorrect to suggest that there were errors in Mr. Saheb-Ettaba’s PeopleSoft records. Ms. Richard testified that he made the leave entries in PeopleSoft himself. It was his mistake that his iCase entries did not reflect the leave requests that he had made and that had been approved by his manager. He could have amended his leave forms to reflect the fact that he took less time than he requested. The data in PeopleSoft was reliable.

174        The employer also contended that it is not consistent with the evidence to suggest that accesses by administrative or HR staff, like compensation advisors, to employees’ PeopleSoft records created problems. Ms. Richard testified that paper trails showed the actions that compensation advisors and others took on PeopleSoft files. Ms. Klee’s problems were resolvable. Based on what she entered in iCase, she could easily explain any inconsistency. It was an administrative issue, not a substantive problem.

175        The employer made no presumption about the accuracy of either system during the process because it accepted that both systems could be amended. It constructed the process to reflect its belief that the timekeepers, who are effectively the common denominator between iCase and PeopleSoft, would be the best people to assist it in arriving at an explanation for some of, if not all, the discrepancies. Unlike a manager, a timekeeper inputted information into iCase, such as leave information, and made leave requests in his or her PeopleSoft account. Managers may come and go, but with respect to the two electronic systems, timekeepers are the constant. The employer questioned whether under the circumstances it was unreasonable to ask a timekeeper and manager to meet to discuss discrepancies and arrive at an explanation.

176        The timekeeper and the manager both signed the attestation forms. Both parties stipulated that they had reviewed the reconciliation package and that they had reconciled the anomalies and addressed them in a manner that was consistent with the applicable terms and conditions of employment. In this context, it is important to emphasize that both parties are lawyers who are in the business of dealing with such documents. Nothing was hidden; nor was there any lack of transparency such that the result was unreasonable. For example, Mr. Saheb-Ettaba confirmed in his testimony that his records had discrepancies.

177        The timekeepers were given a number of options to address the leave anomalies in their records. For example, Exhibit G-6 set out a list of options. The employer did not tie the employees’ hands with respect to the possible explanations that could be provided. Included was one category simply titled “other”. Employees could have used that option to make a leave request under clause 19.19. It is also not correct to suggest that employees were required to make adjustments only when it benefitted the employer. Mr. Dolhai testified that he was aware of a case in which an employee was credited with leave. The process went both ways.

178        It was incorrect to assert that, as the AJC did, the employer should have advised employees about the possibility of using clause 19.19 to resolve discrepancies. That clause is not relevant to the facts. It is a residual provision that is exercised when no other leave provision in the collective agreement applies.

179        It is also important to note that there was no evidence that anyone made a leave request under clause 19.19. In fact, there was no evidence that anyone was unable to reconcile his or her leave discrepancies. If employees were adamant that they did not owe leave, they would not have signed the attestation forms. After all, they are lawyers. Furthermore, the AJC knew the options that were being given to employees. There was no evidence that it asked the employer to consider using clause 19.19 as an option to resolve discrepancies.

180        The employer stated that it is important to note that none of the witnesses testified that they could not explain the anomalies in their records because their computers crashed or they had lost their daybooks. Furthermore, as evidenced in documents relating to Mr. Saheb-Ettaba (Exhibit G-14), employees had access to historical PeopleSoft data to explain discrepancies in their leave records.

181        The employer also noted that it is important to be careful not to confuse different things related to compensatory time and management leave. There is a code for management leave in PeopleSoft, which is different from a rearrangement of working hours. As Mr. Vezina testified, a working hours rearrangement would not have been entered in iCase and therefore would not have triggered a discrepancy.

182        The employer argued that its right to recover leave overpayments has been upheld in such cases as Veilleux v. Treasury Board (Correctional Service of Canada),2009 PSLRB 152; Churcher v. Treasury Board (Department of Fisheries and Oceans), 2009 PSLRB 83; and Anderson v. Treasury Board (Transport Canada), PSSRB File No. 166-02-16440 (19871126), [1987] C.P.S.S.R.B. No. 339 (QL). The employer’s authority to recover leave arises under s. 155(1) of the FAA, which is a more general provision than s. 155(3). The Murchison decision can be distinguished on the basis that it considered leave recovery taken under the authority of s. 155(3), not under s. 155(1). It is important to note that on the facts in this matter, employees were not asked to repay leave that they did not take.

183        The employer also submitted that the KVP Co. decision is not applicable to the facts of these grievances. The reconciliation exercise was not a workplace rule in the sense of the jurisprudence. In any case, the process it undertook was reasonable, fair, and transparent.

184        The AJC has not established a violation of clause 5.02 or any other provision of the collective agreement. When the evidence is examined in context, it is clear that the employer acted reasonably and fairly and in a manner consistent with the purpose set out in clause 2.01. The purpose of the reconciliation exercise was to justify and correct leave anomalies and to ensure that adequate system controls were in place. Accordingly, the employer submitted that the grievances should be denied.

C. The AJC’s reply

185        The AJC noted that contrary to the employer’s assertion, it had specifically asked the employer to apply the provisions of clause 19.19 as a means of resolving leave discrepancies in the first policy grievance. Furthermore, the list of options for resolving leave discrepancies that employees received did not include applying for discretionary leave under clause 19.19, so it is not surprising that there is no evidence that anyone made a request under that provision.

186        Contrary to the employer’s suggestion in argument, employees were not told that they had the option of simply disagreeing with the list of discrepancies. They were never told that they could state on the attestation form that they did not agree. For example, Mr. Butovsky accepted that there were discrepancies, but he did not accept that he owed leave. It is a long way from accepting an apparent discrepancy to agreeing to have leave deducted.

187        Furthermore, although the employer referenced Mr. Dolhai’s evidence that he was aware of a circumstance in which a PPSC employee received the benefit of a discrepancy, there was no similar evidence of that occurring for DOJ employees. In fact, Mr. Saheb-Ettaba testified that he was told that having leave credited was not part of the process. His evidence was not contradicted or challenged.

188        The employer contended that it made no presumption that iCase entries were correct. However, the whole process of leave recovery was based on the inability of employees to explain the absence of PeopleSoft leave records, given their iCase entries. Clearly, there was an assumption that the iCase entries were correct.

189        The employer stated that leave recovery was made under the authority of s. 155(1) of the FAA, not under s. 155(3). However, no evidence was presented of the appropriate authority or delegation to conduct recovery. Furthermore, if leave recovery was undertaken under the authority of s. 155(1), the application of a limitation period becomes a strict defence. For example, in Ontario, no one can be subject to the collection of a debt that arose more than two years before the claim was made.

190        The cases cited by the employer can be distinguished on their facts. In Anderson, the grievor was found to have taken leave to which he was not entitled. In Veilleux, both the grievor and the employer contributed to the circumstances that led to an overpayment, but while the employer was found to have acted reasonably to resolve the situation, the grievor was found to have been responsible for allowing the situation to compound. He had been the author of his own misfortune.

IV. Reasons

191        The AJC filed the two policy grievances before me under s. 220(1) of the Act, which provides as follows:

220 (1) If the employer and a bargaining agent are bound by an arbitral award or have entered into a collective agreement, either of them may present a policy grievance to the other in respect of the interpretation or application of the collective agreement or arbitral award as it relates to either of them or to the bargaining unit generally.

192        These policy grievances arose as a result of the leave reconciliation exercises that the PPSC and the DOJ engaged in during the spring and summer of 2013, which affected all AJC members who were timekeepers. In the first policy grievance (file 569-02-133), the AJC complained that the reconciliation initiative was an unreasonable, arbitrary, and bad-faith exercise of management rights and discretion and that the leave recovery actions initiated as a result of it were neither authorized nor required. In the second policy grievance (file 569-02-137), the AJC complained that the requirement that employees on leaves of absence participate in the reconciliation exercise was unreasonable, arbitrary, and in bad faith.

193        At their cores, both grievances assert that the employer exercised its management rights unreasonably, unfairly, and in bad faith, in violation of article 5 of the collective agreement, which provides as follows:

Article 5

Management Rights

5.01 All the functions, rights, powers and authority which the Employer has not specifically abridged, delegated or modified by this Agreement are recognized by the Association as being retained by the Employer.

5.02 The Employer will act reasonably, fairly and in good faith in administering this Agreement.

194        Management rights in the federal public service are enshrined in ss. 7 and 11.1 of the FAA and 6 and 7 of the Act and give the employer broad authority to organize the public service, allocate resources, and assign duties. Clause 5.01 of the collective agreement is an acknowledgment that the employer retains all the rights, powers, and authority granted to it to the extent that they are not specifically modified by the collective agreement.

195        However, the employer’s unilateral exercise of its management rights is not unfettered. Although the employer argued in this case that the KVP Co. decision was not relevant, I do not agree. As recently noted by the Supreme Court of Canada in Association of Justice Counsel v. Canada (Attorney General),2017 SCC 55 at para. 20, which involved the same parties and the same collective agreement provision as these policy grievances, KVP Co. has long been interpreted as requiring employers to exercise their management rights reasonably and consistently with collective agreements.

196        Even had the arbitral jurisprudence following KVP Co. not established a general requirement that management rights be exercised reasonably, the collective agreement imposes one. Both the Supreme Court of Canada, in Association of Justice Counsel, and the Federal Court, in Association of Justice Counsel v. Attorney General of Canada,2013 FC 806, which also involved the same parties and collective agreement in a matter that was preliminary to the issue that the Supreme Court considered, found that clause 5.02 limited the employer’s exercise of its management rights by imposing a requirement to act reasonably, fairly, and in good faith in administering the collective agreement. The Federal Court specified at paragraph 51 that the requirement to act reasonably, fairly, and in good faith in administering the collective agreement would also include “... any of the employer’s policies or decisions taken or adopted by the employer under the purported authority of the managerial rights (5.01 of the collective agreement).”

197        It should be noted at the outset that the employer did not dispute that it had an obligation under the collective agreement to act reasonably, fairly, and in good faith when exercising its management rights. Furthermore, although the grievances alleged bad faith on its part, it must also be noted that the AJC did not assert bad faith by the employer in relation to either grievance at the hearing. Rather, the AJC argued that the employer’s exercise of its management rights had been unreasonable and unfair, which the employer denied.

198        While the collective agreement establishes the requirement that the employer’s exercise of its management rights must be reasonable and fair, it does not provide definitions or methods of determining what in fact are fairly vague concepts. What is reasonable or fair in any given circumstance is both a question of fact and context.

199        Since KVP Co., a large body of case law has developed, which has examined the issue in the context of health and safety and privacy rights. A consensus has developed in the jurisprudence that holds that a determination of the reasonableness of a unilateral exercise of management rights is best achieved by balancing the interests of employers and employees. The Supreme Court of Canada approved that consensus in such cases as Communications, Energy and Paperworkers Union of Canada, Local 30 v. Irving Pulp & Paper, Ltd., 2013 SCC 34, and, more recently, in Association of Justice Counsel.

200        When balancing the interests of the employer and employees to determine reasonableness, some of the factors that arbitrators and adjudicators have considered include the nature of the employer’s interest and its objective in introducing the unilateral policy or directive, whether the measures adopted were rationally connected to the employer’s objectives, whether there were less-intrusive means available to achieve the employer’s purposes, and the impact of the measures on employees. (See, for example, Irving Pulp & Paper Ltd.; and Peace County Health v. United Nurses of Alberta, 2007 CanLII 80624 (AB GAA).) However, assessing reasonableness and balancing interests with respect to an exercise of unilateral management rights must be placed within the context of the parties’ collective bargaining relationship and the collective agreement that they negotiated. In Canadian National Railway Co. v. C.A.W. - Canada (2000), 95 L.A.C. (4^th) 341 at 375, as quoted in Peace County Health,at 37, the arbitrator Michel Picher observed as follows:

An essential part of the balancing of interests is to determine whether an employer promulgated rule is reasonable. All of the parties before the Arbitrator acknowledge that, absent contrary language in a collective agreement, it is open to an employer to make policies and rules governing its employees, subject to certain generally recognized standards. Those standards, best articulated in [the KVP Co. decision], include the requirement, recently acknowledged by the courts, that a policy or rule must be related to the legitimate business interests of the employer, and that it must be reasonable (Metropolitan Toronto (Municipality) v. C.U.P.E., Local 43 (1990), 1990 CanLII 6974 (ON CA), 69 D.L.R. (4th) 268 (Ont. C.A.)). As further discussed below, an element to be considered in the assessment of reasonableness is whether a rule introduced by management is inconsistent with any substantive provision of a collective agreement, including the protection of employees against discipline or discharge without just cause.

201        In the two grievances before me, the employer’s unilateral assertion of its management right to require all bargaining unit members who were timekeepers to participate in the reconciliation exercise and to request, if not require, employees on long-term leave to also participate in the exercise had a direct impact on the administration and application of the collective agreement and in particular on its leave provisions.

202        The essential question in both grievances is whether on balance the employer’s assertion of its right to impose the reconciliation exercise on employees in the bargaining unit and on those on long-term leave, was reasonable, fair, and consistent with the terms of the collective agreement, bearing in mind the employer’s purpose and objectives, the means used to achieve the objectives, whether less-intrusive means were available, and the impact on employees.

A. The first policy grievance (file 569-02-133)

203        The reconciliation exercise imposed by both the PPSC and the DOJ was composed of two essential elements. First, all employees in the bargaining unit who were timekeepers were required to review and explain discrepancies in leave reporting uncovered by comparing the leave recording in iCase, the electronic time management system used at both the PPSC and the DOJ, and the leave reporting in PeopleSoft, the official leave management system used by HR branches across the federal public service.

204        At the DOJ, the exercise required all timekeepers, including former employees, to review all identified leave discrepancies between 2007 and 2013. At the PPSC, only current employees were required to participate. Furthermore, PPSC timekeepers were expected only to explain discrepancies arising between 2007 and 2010 of more than 75 hours, but they were expected to explain all discrepancies of more than 2 hours arising after 2010. Although not exhaustive, among the list of acceptable explanations provided to timekeepers included demonstrating that the iCase record was mistaken or that the PeopleSoft record was mistaken, that leave requests had been submitted, or that leave was not required in the circumstances.

205        The second element of the reconciliation exercise was that employees who acknowledged that they had failed to properly submit leave forms, which gave rise to a discrepancy, or employees who simply could not explain a leave discrepancy between the two electronic systems were expected to submit leave forms for the period of the absence noted by the discrepancy. Employees who did not have sufficient leave to cover the period could draw on future leave entitlements, or if that was not possible, they could repay the leave through salary deductions.

206        The employer’s objective in implementing the reconciliation was twofold. Concerned by the findings of a PPSC regional audit, which uncovered problems in the internal controls over leave reporting through a comparison of leave data recorded in iCase with that recorded in PeopleSoft, PPSC senior management worried that the problem was not localized, and it decided to reconcile the leave recording in iCase and PeopleSoft of all its timekeepers.

207        Mr. Dolhai testified that the employer was concerned that faulty leave monitoring procedures had contributed to the discrepancies identified between leave reporting in iCase and in PeopleSoft. The reconciliation exercise was designed to identify any weaknesses in internal processes.

208        However, the employer was also concerned that the leave discrepancies gave rise to another risk for the organization. Mr. Dolhai testified that failing to monitor leave properly gave rise to the risk that employees could take or be compensated for leave to which they were not entitled or equally that they were being charged for leave that they had not taken. Correcting the records was an essential part of the process.

209        The DOJ followed the PPSC’s lead in implementing the leave reconciliation process. Ms. Driscoll testified that she had recommended a scan of internal DOJ electronic systems as part of the regular audits the DOJ conducted for risk management. Mr. Pentney, the deputy minister, knew about the PPSC reconciliation exercise and asked that the scan of electronic systems include a reconciliation exercise like the one conducted at the PPSC.

210        In argument, the AJC acknowledged the employer’s legitimate interest in ensuring that internal controls over leave management were effective and appropriate. It did not question that the employer had the right to engage in a review of leave recording to determine whether additional controls were necessary and whether leave recording was appropriate and accurate. I agree. I do not believe that the employer’s objective in initiating the reconciliation process can be faulted.

211        However, for the following reasons, I share the AJC’s concerns that the design of the reconciliation process was seriously flawed, that its retroactive application placed an unfair burden on employees, and that the requirement that they reimburse the employer for leave anomalies going back to 2007 that could not be explained was unreasonable, unfair, and inconsistent with the terms of the collective agreement.

212        The design of the reconciliation exercise was virtually identical in the PPSC and the DOJ. In fact, Ms. Driscoll testified that the analyst who wrote the script used at the PPSC came to the DOJ to run it there. She stated that the only real difference in the software scripts were the departmental codes.

213        In both departments, the internal auditors created the documents that identified discrepancies between leave reported in iCase and in PeopleSoft. Ms. Betts testified that the auditors ran a script that allowed them to select specific fields from iCase and PeopleSoft to identify leave discrepancies. Ms. Betts acknowledged that the assumption was that the data in PeopleSoft and iCase was inherently reliable. It was tested at the PPSC to the extent that the auditors checked to make sure that there were transactions throughout the period that they had requested. Similarly, at the DOJ, Ms. Driscoll testified that the script was run several times using the same data to ensure that the results were consistent. She was confident that the data extracts obtained through comparing iCase and PeopleSoft were correct.

214        Although the auditors’ assumption underpinning the data extraction was that the data in both systems was inherently reliable, there was significant evidence to the contrary. An internal audit report at the DOJ in 2010 (Exhibit G-1, Tab 20) noted problems in iCase timekeeping that went back to at least 2007, when another audit of iCase also noted poor timekeeping practices.

215        Memos to timekeeping staff were sent in 2011 and 2012 that reminded them of the importance of accurate timekeeping, and Ms. Guttman testified that senior management at the DOJ and AJC representatives had discussed inconsistencies in timekeeping practices.

216        Similar concerns were noted at the PPSC. An internal audit report in 2011 touched on timekeeping and noted that no procedures or controls were in place to ensure accuracy or compliance (Exhibit G-1, Tab 21). Ms. Betts echoed this concern in her evidence. She stated that even before the reconciliation exercise, the employer knew of the significant problem with the data. She explained that iCase had limited controls. There were no mandatory fields to ensure that complete and consistent information was entered. She further acknowledged that iCase recording was not always done in a timely fashion, although she believed that it had improved somewhat in the year or two before the reconciliation process. She also confirmed that in PeopleSoft, the controls were either so lax or so insufficiently implemented that there was a significant risk that the information it contained was inaccurate.

217        Ms. Betts testified that from the auditors’ perspective, the information in the systems was validated by the reconciliation process. The expectation was that the timekeepers would be best placed to review the data and to confirm it. But that expectation was reasonable only if the timekeepers had the information at hand to confirm or explain discrepancies.

218        However, there was no requirement that timekeepers retain records of their leave usage. In fact, Ms. Pearce testified that in her office, the staff had been asked to delete old data from their email accounts to reduce the amount of stored email. She testified that she had difficulty accessing the information that she needed to reconstruct her leave records and that the integrity of her data was not good.

219        Similarly, Ms. Klee testified that she had few work records. She had not kept old calendars and emails following her retirement because she did not know that she would need them. Ms. Klinger also testified that she had difficulty accessing old records, although she had kept some records of emails approving leave requests going back to 2007.

220        Although Mr. Dolhai thought that employees would be able to explain identified discrepancies simply by checking their PeopleSoft leave records, in fact, the process was more complicated than that. The range of explanations accepted by the employer to resolve discrepancies included demonstrating that the iCase record was incorrect, that leave was not required for the time off, or that leave in fact had been approved for the time in question. Access to old calendars, emails, and leave and work records going back to 2007 was clearly essential if employees were to have the tools necessary to complete the task.

221        When questioned about the retention of leave records, Ms. Pratt, who was the director general of the DOJ’s Human Resources and Professional Development Directorate during the reconciliation process, testified that she did not know whether employees were advised that they could access paper leave records, and she was uncertain as to how long such records were retained. She also acknowledged that it was possible that employees were asked to clean out old emails and electronic calendar records to reduce data storage. She was unable to state how far back records were kept as she had never tried to access six- or seven-year-old data. She also noted that she did not retain her own records.

222        It should also be noted that timekeepers had no particular reason to keep leave and work records beyond the current fiscal year. The evidence established that processes were in place to regularly review both timekeeping and PeopleSoft records. Clauses 13.01 and 13.02 of the collective agreement, which applied between October 23, 2009, and April 1, 2013, included a provision for a monthly reconciliation of hours of work for the purpose of establishing overtime entitlements and included a requirement that timekeepers submit attendance and timekeeping records as required by the employer. Mr. Dolhai, Mr. Vezina, and Ms. Klinger each testified that monthly timekeeping records were reviewed. Furthermore, there was evidence that leave and timekeeping records were reviewed as part of the annual or biannual performance review process. Mr. Dolhai testified that the performance review process provided an opportunity to address issues with timekeeping and leave and to ensure that leave balances were correct. Ms. Klinger concurred with that assessment and stated that she believed that if there had been problems with her timekeeping and leave reporting, they would have been addressed at those meetings.

223        In addition to timekeeping records review, Ms. Pratt testified that at the end of every fiscal year, a process was in place for employees to attest to the accuracy of their leave balances. Mr. Vezina testified it was his practice as a manager to meet with the employees under his supervision at the end of every fiscal year to ensure that all leave had been properly entered into PeopleSoft.

224        Given those processes, it was not surprising that employees might have had difficulty reconstructing their records for the period between 2007 and 2013. As Ms. Pearce testified, there was no forewarning that employees should keep their records to justify their leave.

225        A number of other factors compounded the difficulty faced by employees trying to explain leave discrepancies identified through the reconciliation process. Although timekeepers were required to enter leave days taken in iCase, they were not required to specify the kind of leave taken. Neither Mr. Butovsky nor Ms. Klee entered that kind of information in iCase, as the summaries of the entries provided to them during the reconciliation exercise demonstrated (Exhibits E-1 and G-9). Ms. Pearce and Ms. Klinger sometimes, but inconsistently, noted the types of leave taken in iCase (Exhibits E-3 and G-18).

226        Furthermore, the evidence suggested that there was a general lack of clarity on the requirement to record certain kinds of leave. Management could grant time off work with pay for personal, medical, or dental appointments for up to half a day. Such leave might have been entered in iCase but was not recorded in PeopleSoft because it was not considered leave with pay granted under the terms of the collective agreement.

227        Similarly, time off with pay could be granted in lieu of overtime, but there was a lack of agreement as to when such leave would be recorded in either PeopleSoft or iCase. When it amounted to a simple rearrangement of working hours, Mr. Dolhai and Mr. Vezina believed that it would not have been recorded in either iCase or PeopleSoft, but they stated that management leave accorded under the terms of the collective agreement would have been entered. Nevertheless, both of them acknowledged that informal arrangements in recognition of overtime demands might have been in place.

228        Ms. Klinger provided an example of the confusion caused by the lack of clarity around recording time off with pay in lieu of overtime. As she noted in an email to Mr. Vezina (Exhibit G-19), she routinely requested such time off for a religious observance. She entered such leave in iCase but not in PeopleSoft. She provided an example of one such request made in 2011 (Exhibit G-22), which was entered in iCase but not PeopleSoft and that was subsequently identified as a leave discrepancy requiring an explanation.

229        It should be noted that clause 19.18 of the collective agreement provides for leave for a religious observance. Under the provision, employees have a number of options available to take time off for that purpose. They can use vacation leave, compensatory leave, or leave without pay, or at the employer’s discretion, they can take the time off with pay on the understanding that the hours of leave will be made up hour for hour within six months.

230        Ms. Klinger testified that although she was able to provide email evidence of one approval of leave for a religious observance, she was certain that she had taken other days for the same purpose, but she did not have the records to establish it.

231        In addition to evidence on the difficulty of explaining leave discrepancies going back six or seven years, evidence was presented that demonstrated that inaccuracies and problems appeared in some of the spreadsheets the employer provided to timekeepers for the exercise. The documents provided to Ms. Klee and Ms. Klinger contained quite significant errors that resulted in large numbers of discrepancies being identified. Because the errors were quite obvious, both employees were able to correct the record relatively easily.

232        The employer argued that widespread problems should not be extrapolated from those two examples, and I agree. However, Ms. Driscoll testified that DOJ auditors were aware of a number of employees in circumstances similar to those of Ms. Klinger. No evidence was presented to show any steps that the employer took to correct the problem or to warn the affected employees of the potential for false discrepancies arising because their PeopleSoft records from the PPSC had not been included in the data comparison. Ms. Klinger was not advised of the problem in advance.

233        Similarly, the PPSC management letters (for example, Exhibit E-10) that led to the reconciliation process noted that while iCase leave reporting was done daily, PeopleSoft leave entries could cover a number of days of leave entered on a single day, which gave rise to false discrepancies, as was the case with Ms. Klee. There was some evidence that PPSC employees were advised of that possibility (for example, Exhibit G-1, Tab 3), but there was no evidence that DOJ employees were similarly warned. Ms. Klee was not warned.

234        Management’s expectation was that employees would explain and correct the identified leave discrepancies. As noted, employees were given a number of options to explain and correct the discrepancies. The letter sent to Ms. Klee with the reconciliation package (Exhibit G-6) enumerated them: the iCase entry could be wrong; the leave might not have been entered in PeopleSoft, in which case a leave request was to be attached; a leave form was not required because it was a routine medical or dental appointment; a leave form was not required because it was a statutory holiday; a leave form had been submitted previously; or some other unspecified explanation.

235        Employees who acknowledged that they had failed to report leave in PeopleSoft were also given options for repayment. If they had leave credits available, such as annual leave, sick leave, family related leave, or personal or volunteer leave, they could draw on those credits to repay the leave they owed. If they did not have leave credits available, they could in some cases draw on future leave credits, or they could repay the amount owed through salary deductions.

236        Employees who could not explain a discrepancy were expected to repay leave on the presumption that the iCase entry was correct, unless they could prove otherwise. For example, Ms. Pearce was unable to explain a 2-week period in 2007, which amounted to 75 hours, a 1-week (37.5 hours) period in 2010, and 1 other day. She testified that because she could not explain what happened, she was required to submit a leave form for the time, which depleted her annual leave bank in 2013 and reduced her 2014 annual leave bank by four days.

237        She stated that she could find no records relating to the vacation leave in question and admitted that she could not say that she did not take it, but she was adamant that she never would have gone on leave without her manager’s approval. She was certain that she would have noticed had her leave bank had extra leave at the end of a fiscal year because it was her practice to keep track of her leave balances. She stated that it was possible that she had planned leave that she was subsequently unable to take because of work requirements, but she simply could not be certain. After so many years, it was impossible to reconstruct what happened.

238        Similarly, Mr. Vezina testified that as a result of his participation in the reconciliation exercise, he was required to repay some sick leave. He stated that he could not say whether he was wrongly charged or whether he had simply forgotten to submit a leave form through PeopleSoft.

239        Mr. Butovsky testified that he repaid a day of leave that was coded as volunteer leave even though he could not say for certain that it had been taken as a volunteer day or whether his manager had simply done him a favour to protect his sick leave.

240        These witnesses repaid leave that they were either not certain that they owed or not necessarily able to attribute to the appropriate leave code. Leave is provided under the collective agreement for specific purposes and is granted when employees meet the requirements. The requirement that employees retroactively make deductions from their leave banks or otherwise reimburse the employer for leave that they could not be certain that they took in my opinion is not consistent with the provisions of the collective agreement.

241        The requirement that employees reimburse the employer for leave discrepancies was predicated on the assumption that the leave notations made in iCase were correct. But given the evidence, that assumption was not reasonable or fair. Leave discrepancies were identified by comparing the data in the two electronic systems, which were designed for completely different purposes and that the employer knew were unreliable. The data was not independently tested or validated. Rather, the employer relied on the timekeepers to test it. But there was no requirement that the timekeepers retain records back to 2007, which were necessary to establish data reliability. The evidence presented to me established that there could be no confidence that timekeepers had kept the records necessary to complete the task.

242        Requiring timekeepers to reimburse the employer for leave discrepancies that could not be explained based on such faulty processes was unreasonable, unfair, and inconsistent with the leave provisions of the collective agreement. Because other possibilities could have explained the discrepancies, such as errors in reporting in iCase, problems with the data comparison, and mistakes in PeopleSoft, there could be no certainty that in fact the employees had taken the leave in question. The best that could be said was that a discrepancy could not be explained.

243        In my view, requiring employees to repay leave based on the mere possibility that they owed leave, rather than on the certainty that they did, was improper. Furthermore, even though the filters applied by the PPSC lessened the negative impact of the exercise on PPSC timekeepers, they did not eliminate it. As the example of Ms. Pearce demonstrated, it was possible to have a leave discrepancy of 75 hours in 2007 that could not be explained.

244        It should also be noted that I do not accept the employer’s argument that s. 155(1) of the FAA is applicable in circumstances in which employees repaid leave based on their inability to explain leave anomalies. I do not believe that the inability to explain a discrepancy in the circumstances of the leave reconciliation exercise can be said to give rise to “indebtedness”, for the reasons that I have already given.

245        Based on the evidence before me, I find that the retroactive application of the leave reconciliation process, which was a unilateral management directive, was unreasonable and unfair and that it contravened clause 5.02 of the collective agreement. I do not find that the process was unreasonable or unfair as it applied to the year in which it was implemented (2013) because the evidence established that employees would be expected to retain leave and work records to satisfy the monthly timekeeping processes and the annual and biannual performance reviews, which included reviewing leave usage and timekeeping. Therefore, they would have been in a position to address any inaccuracies in the records.

B. The second policy grievance (file 569-02-137)

246        The AJC asserted that employees on long-term leave were recalled from their leaves of absences and were required to participate in the reconciliation exercise and that such requests were unreasonable, arbitrary, and in bad faith, in contravention of the collective agreement, in particular clause 5.02 (Management Rights), article 11 (Leave With or Without Pay for Association Business or for Other Activities under the Public Service Labour Relations Act), article 18 (Sick Leave with Pay), article 19 (Other Leave With or Without Pay), and article 21 (Leave - General).

247        Ms. Guttman testified that she became aware that the DOJ had asked employees on long-term leave to participate in the reconciliation exercise in August 2013. She addressed the matter with Ms. Laframboise, the director of labour relations and compensation, who told her that such situations would be addressed individually and that employees who participated would be compensated (Exhibit G-1, Tab 13). Ms. Guttman believed that while the employer would ask employees on long-term leave to participate, it would not force them to.

248        Ms. Pratt testified that employees on long-term leave were invited to participate in the process while on leave but that they were also advised that they could complete the process on their return to work following their leave. She did not know how employees who chose to participate in the process while on leave were compensated.

249        No evidence was adduced from employees who had been recalled from leave to participate in the process.

250        The burden to establish a breach of the collective agreement in this matter fell on the AJC. I was provided with little evidence to suggest that it was unreasonable or unfair to simply ask employees on long-term leave if they wished to participate in the process while on leave or if they preferred to wait until their return. I was not directed to any provision in the collective agreement that would prohibit such a request. Furthermore, the employer confirmed that employees who chose not to participate were not penalized and that those who did participate were compensated, which answered some of the concerns addressed in the grievance.

251        Given all these circumstances, I cannot find that the AJC established a violation of the collective agreement.

V. Conclusion

252        I have found that the employer violated clause 5.02 of the collective agreement with respect to the retroactive application of the reconciliation exercise that is the subject of the policy grievance in file 569-02-133. Because issues with respect to the application of appropriate remedies were raised as a preliminary matter, I ruled that, should the grievance be allowed, the remedy would be addressed separately.

253        I urge the parties to take the opportunity to resolve the remedy question without resorting to a further hearing. However, should they be unable to, I believe that the outstanding matters concerning remedy can be addressed through written submissions that address the following issues:

1) whether the provisions of s. 232 of the Act in any way limit my jurisdiction with respect to remedy;

2) whether the grievance as filed applied to retired and former employees who worked for other departments or agencies; and

3) whether the AJC has standing to file grievances on behalf of persons who are no longer members of the bargaining unit.

254        For all of the above reasons, the Board makes the following order:

VI. Order

255        The policy grievance in file 569-02-133 is partially allowed. As it applied retroactively to the fiscal years 2007-2008, 2008-2009, 2009-2010, 2010-2011, 2011-2012, and 2012-2013, the reconciliation exercise violated clause 5.02 of the collective agreement.

256        At the parties’ request, a process for written submissions on the question of remedy, as set out in paragraph 253 of this decision, is to be established by the Board’s registry office.

257        The policy grievance in file 569-02-137 is dismissed. The file is to be closed.

May 3, 2018.