Susan McGrath v. Superintendent of Financial Services, 2010 ONFST 5

2010 ONFST 5

Decision No. P0335-2008-2

Date: 2010/03/26

IN THE MATTER OF the Pension Benefits Act, R.S.O. 1990, c.P.8, as amended by the Financial Services Commission of Ontario Act, 1997, S.O. 1997, c.28 (the “PBA”);

AND IN THE MATTER OF the Registration of By-Law No. 7 of the OMERS Primary Pension Plan, Registration Number 0345983;

AND IN THE MATTER OF a Hearing in accordance with subsection 89(8) of the PBA;

-and-

SUPERINTENDENT OF FINANCIAL SERVICES, OMERS ADMINISTRATION CORPORATION and OMERS SPONSORS CORPORATION

-and-

POLICE PENSIONERS ASSOCIATION OF ONTARIO and IATSE, LOCAL 58

DATE OF HEARING:

January 18, February 8-9, 2010

BEFORE:

Elizabeth Shilton Member of the Tribunal and Chair of the Panel

David Short Member of the Tribunal and of the Panel

Ralph Scane Member of the Tribunal and of the Panel

Anthony McGrath, for the Applicant Susan McGrath

Mark Bailey for the Superintendent of Financial Services

Freya Kristjanson and Amanda Darrach for OMERS Sponsors Corporation

Jeff Galway and Rahat Godil for OMERS Administration Corporation

Paul Bailey for the Police Pensioners Association of Ontario

Tim Taylor for IATSE Local 58 (“IATSE”)

A. INTRODUCTION

This case considers the validity of an amendment to the OMERS Primary Pension Plan (the “Plan”) changing the method used to calculate inflation indexing. It addresses important issues of first impression before this Tribunal about the meaning of s.14(1) of the Pension Benefits Act (“PBA”), and the scope of a plan sponsor’s authority to amend pension plans.

In general, the Plan offers full inflation protection, with pensions fully indexed to increases in the Canadian Consumer Price Index (“CPI”). For several years, the Plan provided for inflation adjustments to be calculated according to a formula based on a comparison between the September CPI from one year to the next. We call this approach the Old Method, or the “OM”. In October of 2007, the OMERS Sponsors’ Corporation (the “SC”) amended the Plan to eliminate the OM and adopt the method of indexation used by the Canada Pension Plan (“CPP”), based on a year-over-year comparison of average monthly CPI increases for a 12 month period ending in October of each year. We call this CPP method the New Method, or the “NM”. The SC applied in November of 2007 to register the amendment pursuant to the provisions of the PBA. Registration was opposed by the Applicant, Susan McGrath, a retired member of the OMERS plan. The Superintendent of Financial Services (the “Superintendent”) proceeded to register the amendment. The Applicant requested a hearing before this Tribunal under s.89 of the PBA, challenging the Superintendent’s decision to register the amendment on the grounds that it reduced her accrued pension benefits within the meaning of s.14(1). She is opposed by the respondents, the SC, the Administration Corporation (the “AC”), and the Superintendent, who all take the position that the amendment does not reduce accrued pension benefits.

Two additional parties, the Police Pensioners Association of Ontario (“PPAO”) and IATSE applied for and were granted limited status as intervenors. They took no position on the outcome of the proceeding, and made no submissions.

At a pre-hearing conference, two issues were stated for hearing as follows:

Is the Amendment void within the meaning of section 14(1) of the PBA because it reduces the amount or the commuted value of a pension or pension benefit accrued under a pension plan?
Depending upon the answer to issue 1 above, what, if any, is the appropriate remedy and should it include revocation of the registration of the amendment?

Two additional issues were stated relating to whether the amendment was ‘adverse’ within the meaning of s.26 of the PBA, and whether, if it were ‘adverse’, proper notice had been provided. These two issues were subsequently abandoned by the Applicant and accordingly we do not address them in our decision.

We have decided that the amendment does not reduce the amount or the commuted value of the Applicant’s accrued pension within the meaning of s.14(1), and accordingly is not void.

Below we set out the reasons for our decision.

B. THE EVIDENCE

Much of the evidence in the case was filed by agreement through an Agreed Statement Facts (“ASF”) and an Agreed Book of Documents (“ABD”). This agreed-upon evidence was supplemented by sworn witness statements filed by various parties. The Applicant filed her own witness statement. The AC filed a witness statement from Jennifer Brown, Executive Vice-President and Chief Pension Officer of the AC. The SC filed a similar statement from Marianne Love, Co-Chair of the SC since its inception in 2006. All these statements were filed on consent and without cross-examination. Also on consent, the Applicant filed responses to certain interrogatories submitted to both the AC and the SC. At the hearing on January 18, 2010, this documentary evidence was supplemented by viva voce evidence from two expert witnesses, both actuaries: K. Paul Duxbury for the Applicant and Jill Wagman for SC. The Superintendent did not call evidence.

The evidence as a whole established the facts set out below. Much of this summary of the evidence is borrowed from the ASF.

1. General Background

The Ontario Municipal Employees Retirement System (“OMERS”) was first established under the Ontario Municipal Employees Retirement System Act, 1961-1962 and is now continued under the Ontario Municipal Employees Retirement System Act, 2006 (the “OMERS Act, 2006”). OMERS administers a number of different pension plans. The plan at issue in this case is the OMERS Primary Pension Plan (the “Plan”), a contributory defined benefit pension plan covering employees in the Ontario municipal sector. Up to 2006, the Plan was set out in Regulation 890; after the new statute came into effect, the plan text ceased to be a matter of regulation.

The Applicant, Susan McGrath, is a retired member of the Plan. She enrolled in the Plan on July 1, 1969 and remained a member continuously until she retired on September 1, 2000. She has been in receipt of a monthly pension since that date based on credited service of 34.25 years. The respondent SC is a statutory corporation established under the OMERS Act, 2006. The SC is made up of fourteen members of the corporation. Seven members are appointed by employers or employer associations who participate in OMERS, six are appointed by unions or employee associations whose members participate in OMERS and one is appointed by retiree organizations. Pursuant to s.16 of the OMERS Act, 2006, the SC determines the terms and conditions of the OMERS pension plans, subject to the restrictions set out in the OMERS Act, 2006. The respondent AC is a continuation of the Ontario Municipal Employees Retirement Board established under the original OMERS statute. Under the provisions of the OMERS Act, 2006, the AC acts as administrator of OMERS pension plans and trustee of the OMERS pension funds, provides for the actuarial valuation of OMERS pension plans, and advises and assists the SC. The AC is composed of fourteen members of the corporation who also function as the board of directors of the AC. Currently, there are seven employer or employer association representatives, six representatives of unions or employee associations whose members participate in OMERS and one representative of retiree organizations.

Prior to 1992, inflation increases applicable to pension benefits paid under the Plan were granted on a purely ad hoc basis. In 1992, the Board implemented guaranteed indexing. The indexation formula adopted by regulation produced 70% of the annual increase in the Consumer Price Index (“CPI”) (with a 6% cap and a minimum of 0%). The remaining 30% could be provided through ad hoc increases if there were sufficient funds available. From 1992 to 1998 the Board topped up the amount of the annual inflation increase each year to 100% of CPI. In 1999, the Board implemented 100% guaranteed indexing (with a 6% cap and a minimum of 0%) and removed the ad hoc payments. The regulation was duly amended from time to time to permit these changes.

From January 1, 1992 until December 31, 2007 the operative pension plan provisions addressing indexation, first set out in Regulation 890 and subsequently in s.31 of the Plan text, were as follows:

(1) In this section, the inflation increase of any adjustment year means the percentage increase of the Consumer Price Index for Canada for September of the previous year over the Consumer Price Index for Canada for September of the year which is 2 years prior to the adjustment year.

(2) The following is the inflation adjustment for the adjustment year indicated:

For each adjustment year not otherwise specified in this subsection, 70 per cent of the inflation increase for that year.
For 1999 and subsequent years, 100 per cent of the inflation increase for the applicable year.

(3) The inflation adjustment in excess of 6 per cent in any adjustment year shall be added to the inflation adjustment of the subsequent adjustment year.

(4) The inflation adjustment in any adjustment year shall never be less than zero.

(5) In this section and for the purposes of subsections 18(8), 18(9), 22(6) and 22(7), “adjustment year” means a year in which pensions are increased by the inflation adjustment.

(7) The year 1992 and subsequent years shall be considered adjustment years.

(8) The pension payable to a person during his or her lifetime shall be determined in accordance with this Plan or a predecessor thereof and,

(a) where a pension is payable to a person on the 1^st day of December of the year prior to an adjustment year in respect of a pension that was being paid on the 1^st day of December of the year which is 2 years prior to that adjustment year, the pension payable to the person on the 1^st day of January of that adjustment year shall be increased by the inflation adjustment; ....

Pursuant to this Plan provision, pensions in pay were increased every January 1 by a percentage identified by calculating the difference between the CPI for the September two years prior to the adjustment date, and the September in the year prior to the adjustment date: e.g., on January 1, 2005, pensions were increased by the percentage difference between the CPI for September 2003 and the CPI for September 2004.

While this method of indexing pensions to the CPI is perfectly acceptable from an actuarial perspective and meets requirements under the Income Tax Act, it is a volatile method. If the CPI in September of any given year spikes upwards or downwards, it can produce indexation increases which are either unusually high or unusually low as compared with an averaging approach. The Canada Pension Plan and other large public sector plans which provide 100% indexing to CPI typically use a less volatile ‘averaging’ approach. The evidence is that the differences between the OMERS OM and the averaging approach of other plans created some public relations problems for OMERS; in the years when the OMERS Plan produced lower increases, members complained because they perceived that they were getting less than full inflation protection.

In fact, based on the historical evidence it would appear that the two methods produced very similar results over time. All parties relied on the following comparative table, prepared by OMERS Staff and provided in November of 2008 to OMERS retiree organizations (ABD, Tab 45).

Indexation Start Year¹	CPP method Rate²	OMERS Method Rate³	Difference (CPP vs. OMERS)	Cumulative increase from start year to 2007⁴
1992	5.80%	5.47%	0.33%	0.9%
1993	1.80%	1.26%	0.54%	0.6%
1994	1.90%	1.90%	0.00%	0.0%
1995	0.50%	0.20%	0.30%	0.0%
1996	1.80%	2.30%	-0.50%	-0.3%
1997	1.50%	1.50%	0.00%	0.2%
1998	1.90%	1.62%	0.28%	0.2%
1999	0.90%	0.74%	0.16%	-0.1%
2000	1.60%	2.58%	-0.98%	-0.2%
2001	2.50%	2.70%	-0.20%	0.8%
2002	3.00%	2.60%	0.40%	1.0%
2003	1.60%	2.30%	-0.70%	0.6%
2004	3.20%	2.16%	1.04%	1.3%
2005	1.70%	1.79%	-0.09%	0.2%
2006	2.30%	3.36%	-1.06%	0.3%
2007	2.10%	0.70%	1.40%	1.4%

This table shows that in 6 of the 16 years involved, the OMERS method produced a larger increase than the CPP method would have produced. In 8 of the 16 years, the CPP formula would have produced a larger increase. In 2 years, the increase would have been the same. Overall, the CPP method would have produced slightly better results for pensioners over the 1992-2007 period.

2. OMERS Decision to Change the Method of Indexing

A possible change to the method of calculating indexation had been under consideration by OMERS for some time. In September 2005, in response to a desire expressed by the Board to explore the possibility of changing the plan indexing formula to mirror the model adopted by the Ontario Teachers’ Pension Plan, OMERS staff (“Staff”) prepared a report which it submitted to the OMERS Pension Committee (a committee of the Board) (ABD, Tab 4). In this report, Staff noted that other major public sector plans to which OMERS compared itself used different measurement periods and methodologies to calculate the plan’s annual inflation adjustment. The report argued that while the OM did not necessarily capture highs and lows occurring throughout the year, it tended to keep relative pace with the other plans.

While no change was made at this time, the issue did not go away. The Pension Committee considered the matter on November 23, 2006. In its report to the Committee, Staff recommended an amendment to the Primary Plan effective January 1, 2008 to change the indexing formula from a September-over-September comparison to one based on the average change in the CPI over a 12 month period from October to October (ABD, Tab 13). The principal basis for Staff’s recommendation to change the method was the volatility of the OM. The report noted the historical data showing that over time, regardless of the inflation indexation formula used, pension benefits were similarly indexed. Since a majority of members collecting an OMERS pension were or would be also collecting a CPP pension, Staff saw as a benefit of the change the fact that OMERS and CPP benefits would increase by the same amount and at the same time. Staff’s expectation was that this would significantly limit queries and complaints, and would also provide OMERS retirees with smoother, steadier inflation protection. The Pension Committee accepted Staff’s recommendation and recommended the change to the AC.

At an AC Board meeting held on November 24, 2006, the AC accepted the Pension Committee recommendation and resolved to recommend to the SC that OMERS adopt the CPP method. When the SC met on July 4, 2007 to consider this recommendation, Staff presented a report (ABD, Tab 24) recommending that the proposed change take effect on January 1, 2008. The reason provided for the recommendation was that “this change will eliminate the volatility that is inherent in the OMERS current formula”. The Staff report considered the pros and cons of implementing the change prospectively or retroactively to January 1, 2007. Among the considerations against retroactive application, the Staff Report noted that the actuarial gain to the fund related to indexation as at December 31, 2006 would be “significantly reduced”, from the current $371 million down to $110 million. It noted that “[a]pplying the increase retroactively will increase the going concern costs at a time when the Plan is dealing with funding issues”. The SC also discussed the proposed amendment at a meeting on September 5, 2007.

At a meeting of the SC on October 3, 2007, the SC passed By-Law No. 7 making a number of changes to the Plan text, including a change from the OM to the NM, the change at issue in this case. The SC’s decision resulted in an amendment to the wording of section 31 (1) of the Plan, repealing the old wording and substituting the following:

In this section, the inflation increase for any adjustment year means the monthly average for the Consumer Price Index (CPI) over the last 12 months of the 24 month period ending in October in the immediately preceding year compared to the monthly average for the CPI over the first 12 months of that period.

The Minutes of the October 3 meeting (ABD, Tab 30) record the reasons for the decision, as follows:

It had been discussed that this change will reduce the potential for volatility inherent in the current formula (i.e. September of one year compared to September of the previous year vs. averaging a 12-month period) and will bring the OMERS pension indexing in line with the Canada Pension indexing methodology.

Both the Applicant and several OMERS retirees’ organizations raised concerns about the impact of implementing the change from the OM to the NM on January 1, 2008. In response to these concerns, the retirees’ representative on the SC, Glen Mills, sought to have the SC reconsider the issue of the effective date of the amendment, bringing forward a motion to have the effective date postponed. On November 1, 2007, the SC met to discuss the issue of whether the effective date of the change should be altered. At that meeting, Staff presented a report which assessed a number of issues in relation to the NM and provided what was described as a “preliminary estimate” of the comparative impact on January 1, 2008 of the OM and the NM (ABD, Tab 33). The report concluded that:

Based on this preliminary comparison, the new methodology has rendered an increase that is smaller than that rendered by the old methodology. Unless there is a significant shift in October’s CPI figure, this result should remain relatively consistent through to next month when the actual calculation under the new methodology will be performed.

With respect to the funding implications of the change, the Staff report commented that:

There are no actuarial implications as the valuation assumes 100% indexing. However, with a long term inflation assumption of 2.5% there will be a smaller experience gain in the 2007 valuation if the old methodology is still used. (It is estimated that the old methodology will result in an experience gain of 10 million dollars, compared with a gain of 120 million dollars under the new methodology.)

The motion to change the effective date of the amendment failed to receive the 2/3 vote necessary to pass, however, and was accordingly defeated.

3. The Role of the Applicant in the Amendment Process

The Applicant had taken an active interest in the prospective change to the plan indexation since early in 2007. On April 30, 2007, she wrote to the Co-Chairs of the SC (ABD, Tab 20). In that letter, she was very critical of the volatility of the OM, and supported a change to the NM, effective January 1, 2008. She took the position, however, that the existing OMERS indexation methodology had operated to leave OMERS pensioners with less inflation protection than the CPP and the Ontario Teachers’ Pension Plan, and provided detailed calculations to back up her contention. She criticized calculations published in the March 2008 newsletter of the Municipal Retirees Organization Ontario showing that over the period from 1999 to 2007, the OMERS indexation formula had done better than the CPP’s; she argued the comparison was “misleading”, and pointed out that if the period compared was 2000 to 2007, OMERS had done worse. She was particularly concerned about the fact that OMERS pensioners had received a sharply lower increase in 2007, compared to the CPP. She argued that OMERS should pay a “catch up” increase to existing pensioners at the same time as it implemented the NM.

On September 24, 2007 the Applicant wrote again to OMERS (this time to Jennifer Brown of the AC; ABD, Tab 29) because, as she says in her witness statement, she “was beginning to get worried about the possibility that OMERS may make changes in the formula without a catch up.” In her letter, she argued once again for a catch up:

…if the Sponsors Corporation Board members approve implementation of the CPP method in 2008 without ensuring that there is a catch up to compensate for the low 2007 increase, they will lock in a permanent loss for pensioners. The answer, surely, is to calculate an additional increase for each pensioner to compensate them for the amounts that they have lost so far, before implementation of the CPP method.

She wrote again on October 29, 2007, this time to the Co-Chairs of the SC by email (ABD, Tab 32). By that time, she was aware that the amendment changing the methodology had been passed, to be effective January 1, 2008, with no catch-up provision for existing pensioners. She was also aware through her retiree organization that a motion to reconsider would be brought forward. By October 29, 2007, it appeared probable that the NM would yield a smaller increase for January 1, 2008 than the OM. In her email of that date, she argued that:

If the Sponsors Corporation Board changes OMERS methodology for January 2008, not only will the majority of OMERS pensioners lose out initially but, furthermore, the losses will be permanently locked in. With the exception of new OMERS retirees who will see their pensions grow at the same rate as the CPP, OMERS pensions will be outperformed by the plans that have always used the CPP method. (ABD, Tab 32)

She argued for a postponement or suspension of implementation “until it can be accomplished without losses to OMERS pensioners.”

4. The Registration of the Impugned Amendment

An application for registration of By-Law Number 7 containing the impugned amendment was filed by the AC with the Superintendent on November 28, 2007. A number of OMERS retirees’ organizations wrote to the Superintendent protesting the registration of the amendment. The Applicant likewise protested the registration. Her first letter to the Superintendent dated November 14, 2007 (ABD, Tab 41) pre-dated the registration application. In that letter, she made two main arguments against registration. First, she argued that OMERS had “purposely decided to change methodologies right after the OMERS methodology produced the lowest inflation increase in the history of the plan (0.7% in 2007 compared to the CPP increase of 2.1% in 2007).” Second, she argued that “[t]he SC, in full knowledge of the facts, switched methodologies precisely at the moment when this change would penalize pensioners permanently to the tune of 1.52%.” She argued that while maintaining the OM would have permitted OMERS pensioners to make up this shortfall in time, “[c]hanging methodologies at this point means we will never catch up with inflation”. As she had done in her correspondence to OMERS, she argued that the switch should not have been implemented without a “catch up” payment to OMERS pensioners “to bring OMERS in line with the CPP”.

On January 28, 2008, she wrote again to the Superintendent’s office, this time to the Pension Officer, reiterating her arguments against OMERS changing its method without a “catch up” for pensioners disadvantaged by the OM (ABD, Tab 68). Despite her opposition and that of the OMERS retiree organizations, on May 16, 2008 the Superintendent issued a Notice of Registration in relation to the impugned amendment. The Notice of Registration was re-issued on September 11, 2008 to correct the fact that the some of the amendments to the Primary Plan text contained in By-Law Number 7, other than the amendment to the indexation method, had an effective date of November 29, 2007 rather than January 1, 2008.

The retiree organizations chose not to pursue the matter. The Applicant, however, filed a Notice of Appeal to the Tribunal on October 8, 2008.

5. The Implementation of the New Methodology

As at January 1, 2008, the NM was used for the first time. The inflation adjustment to pensions in pay was 1.99%. Under the OM, the adjustment would have been 2.47%. Again in 2009, the NM yielded a lower increase than the OM would have yielded: 2.51%, as compared to 3.40%. In January of 2010, the NM yielded a higher increase then the OM would have yielded: 0.37% as compared to 0%.

6. The Expert Evidence

As noted above, both the Applicant and the SC called actuarial evidence in support of their positions. Paul Duxbury (called by the Applicant) and Jill Wagman (called by the SC) are both experienced expert witnesses with excellent credentials; their expertise was not challenged by any party and the Tribunal accepted them as expert witnesses. Mr. Duxbury’s evidence supported the Applicant’s claim that both the amount of the pension and the commuted value had been reduced, whereas Ms Wagman’s evidence supported the position of the responding parties that neither the amount nor the commuted value had been reduced.

Mr. Duxbury did not quarrel with OMERS’ basic decision to change its indexation formula from the OM to the NM. He testified that the two formulae were considered to be “actuarially equivalent” and that “over time” they would be expected to produce the same results. His critique was directed at the impact on the Applicant and other OMERS pensioners of the decision to implement the change to the NM on a specific date: January 1, 2008. He testified that under the OM, OMERS pensioners would have received indexation increases of 2.47% on January 1, 2008, 3.40% on January 1, 2009, 0.00% on January 1, 2010 and an estimated 1.74% on January 1, 2011.⁵ Instead, under the NM, they received 1.99%, 2.51%, 0.37% and an estimated 1.42% respectively, for a cumulative estimated net loss of 1.27%.⁶ He produced a table designed to demonstrate that by January 1, 2011, OMERS pensioners would have suffered a 1.27% cumulative loss resulting from the implementation of the NM, and that this projected loss would be “locked in” and carry forward indefinitely.

In his expert report, he stated that:

The old and the new methods perform differently depending on the year chosen to implement the new method. The worst year for implementation of the new method was 2008, the year chosen by OMERS for implementation....

Implementation in 2008 resulted in significantly smaller increases than the old method would have produced in the years 2008, 2009 and 2010. Implementation in 2008 produced a smaller pension than the old method would have produced that was 1.27% less by ratio in each of the years going forward from 2011.

Implementation in 2007 would have been relatively neutral for OMERS pensioners, as would implementation in 2004 and 2010. On the other hand, implementation in 2005, 2006 and 2009 would have produced smaller pensions than the old formula, though not as small as implementation in 2008 produced. (Exhibit 8, p.3)

He appears to have reached the conclusion that 2008 was the “worst” year by measuring the relative difference in increase produced by the OM v. the NM on January 1^st of each of the implementation years he considered.

Mr. Duxbury also testified on the significance of the fact that actuarial gains to the Plan varied depending on the date on which the change from the OM to the NM was implemented. He testified if the SC had agreed to implement retroactive to January 1, 2007, the net actuarial gain to the Plan for 2007 would have been reduced by $261 million. The decision to implement in 2008 produced an increase of $110 million in the net actuarial gain to the Plan. In his report, Mr. Duxbury concluded that:

Clearly the references to actuarial gains to be realized by changing the methods on January 1, 2008 indicate that the effect produced by the new formula in the future is not expected to offset the lower indexation increases granted upon implementation. This would appear to corroborate the finding that the reduction in indexation increases produced by implementing the new formula in 2008 are permanent. The new indexation experience gain of $261 million captured for the Plan by implementing the new formula in 2008 instead of 2007 reduces total pensioners indexation payments over their expected lifetimes by the same amount. (Exhibit 8, p.6)

In his view, then, the fact that actuarial gains resulted from implementation of the NM in January 1, 2008 demonstrated two things: first, that the Plan saw the savings from the implementation of the new method as permanent rather than temporary, and second, that these actuarial gains represented losses to the pensioners over their lifetimes.

Ms Wagman’s evidence did not focus on a specific time frame. The gist of her evidence can be found on page 9 of her expert report (Exhibit 4):

It can be demonstrated mathematically that, if the assumed long-term inflation rate is constant, the change to the inflation protection adjustment methodology under By-Law Number 7 has no effect on the projected future adjustment to the pension. In practice, inflation rarely behaves in such a uniform manner from month to month. Therefore, the variation between the current and previous calculations methods generates slightly different results. However, over the long term, neither [the OM or the NM] is designed nor expected to result in consistently higher increases, as illustrated in the November 26, 2007 report prepared by the OMERS Administration Corporation. In this report a historical comparison of the two formulas is provided, demonstrating that the current method (CPP) would have resulted in higher increases in 8 out of the 16 years from 1992 to 2007, with the remaining two years as neutral. [This is the table reproduced at p.6, above]. Since then, in 2008 and 2009, the previous (OMERS) method produced a higher result than the current method, but in 2010 the current method will produce an inflation adjustment of 0.37% whereas the previous method would have produced no inflation adjustment due to negative inflation over the period.

This gain/loss analysis will differ among the plan members, depending on their individual date of termination and/or retirement, and does not take into account future gains and losses that will emerge over their future lifetime. Therefore, it is inappropriate to compare or quantify the “promise of 100% inflation protection” at any fixed point in time. As stated above, however, over the long term the expectation is that the current method will produce an inflation adjustment that is not less than what would have resulted under the previous method.

In her report, she clearly acknowledged that at the level of individual plan members, losses and gains will be unevenly distributed. In the long term, however, the result will be equivalent.

Ms Wagman was also asked to comment on Mr. Duxbury’s evidence that the quantification of increased actuarial gains as a result of implementation in 2008 of the NM reflected an assumption on the part of the actuary that the gains would be permanent. She disagreed with that conclusion. As she explained it,

Actuarial gains or losses arise when emerging plan experience over short periods result in liabilities that are lower or higher than expected based on the long-term assumptions. In this case, an actuarial gain will emerge each time a valuation is performed and the inflationary increases granted since the previous valuation were less than assumed in the previous valuation. When setting an actuarial assumption, it is expected that, over the long term, actuarial gains will be offset by actuarial losses, thereby supporting the appropriateness of the underlying assumption over the long term.

With respect to the particular actuarial gains referred to in Mr. Duxbury’s evidence, she testified:

The actuarial gains referenced by Mr. Duxbury are measured at each valuation on a going-concern valuation basis, which assumes the plan is ongoing, and are related to the fact that the actual inflationary increase granted under the plan in 2008 was less than the assumed long-term annual inflationary increases.

In her view, actuarial gains in 2008 for the Plan do not represent permanent losses for pensioners:

[I]n future years, it is expected that the new formula will result in higher increases in some years and lower increases in others, resulting in lower actuarial gains (or higher losses) in others. By virtue of this fact, the larger gain experience in 2008 is not permanent, but rather a measure of the temporary deviation of plan experience from the long-term inflation assumption at a fixed point in time. (Supplementary Expert Witness Report, Ex. 6, p.8)

Ms Wagman was asked to comment on Mr. Duxbury’s evidence that OMERS pensioners would suffer a permanent 1.27% loss to the value of their pensions as a result of implementation of the NM on January 1, 2008. She was highly critical of the methods used by Mr. Duxbury in his expert report to demonstrate a 1.27% loss, and in particular of this conclusion that any early loss would continue indefinitely. She pointed out that Mr Duxbury did his projections in December of 2009 using actual inflation increases for the period during which CPI increases are known, and for which the new method produced losses. When he projected ahead into the future, however, he used the assumption that inflation would increase at a flat and uniform rate of 2% per annum, for all years, using both the OM and the NM. That assumption will inevitably lock in any early losses. In the real world, however, she testified that such a scenario would never unfold; indeed, Mr. Duxbury conceded as much in his cross-examination. Even if inflation in fact comes in at 2% per annum for all the years Mr. Duxbury considered (which is in itself improbable in light of earlier patterns), we can be sure that it will not come in at a uniform rate in every month of every year. There will inevitably be variations, and those variations will skew Mr. Duxbury’s uniform projections. In Ms Wagman’s view, the much more likely future scenario is the one consistent with the actuarial assumption on which both actuaries agree – that ‘over time’, the two formulae will produce the same total inflation increase. In her expert report she produced a projected scenario (described as “the Wagman Alternative”) which she characterized as “just as plausible as Mr. Duxbury’s”, in which monthly variations in the inflation rate during the year 2010 allow OMERS pensioners to quickly make up early losses imposed under the NM and come out ahead of where they would have been under the OM. While she conceded that the Wagman Alternative involved unusually sharp CPI fluctuations, she pointed out that similar patterns of sharp fluctuations had in fact occurred in recent economically turbulent years.

Mr. Duxbury, in response, did not disagree with the proposition that inflation was unlikely to behave in practice as he had projected it in his tables. He nevertheless defended the methodology behind his projection. He pointed out that his use of uniform projection was consistent with actuarial practice for projecting inflation, whereas the projections in the Wagman Alternative, which posited monthly variations, were not consistent with actuarial practice. In addition, he relied on basic probability theory to support his uniform projection, analogizing the likely pattern of monthly variation to the proverbial toss of a coin. With every new toss, there is an equal likelihood that heads or tails will come up. A run of ‘heads’ does not make it any more probable that the next toss will produce ‘tails’. In comparing the two methods, he testified:

So at any point in time, you don’t know which one is going to produce more, but we do know that one produced less than the other for two or three years, and that going forward from this date, we expect them to be the same, so therefore, there has been a loss to the pensioners. (Transcript of the hearing on January 18, p. 88).

In his view, if the NM imposes early losses on a pensioner, as had happened to the OMERS pensioners in this case, even if the outcomes over the next twenty years even out, the probability is that they will end up having to live indefinitely with those early losses.

We have identified above a number of differences of opinion in the evidence of the expert witnesses. On a number of very key points, however, the expert witnesses were in substantial agreement. They both agreed that:

The OM and the NM are “actuarially equivalent”. It is not possible to speculate on whether the OM or the NM will produce higher indexation rates in future years, but it is expected over time that the two methods will produce the same result.
The change from the OM to the NM did not affect either the amount or the commuted value of pension benefits for active members because both methods would be treated by actuaries as formulae providing 100% indexation to the CPI.
On October 3, 2007, the date of the SC’s decision to change the method effective January 1, 2008, an actuary calculating the commuted value of an OMERS pension in pay would come up with the same value regardless of which method was employed because the formulae are actuarially equivalent and the future impact of the change in method would not be known.
On January 1, 2008, the actual pension of an OMERS pensioner for 2008 would be lower under the NM than it would have been if the OM had still been in effect.
On January 1, 2008, an actuary determining the commuted value of an OMERS pension in pay on that date would find that the commuted value was lower under the NM than it would have been under the OM because the impact of the change would now be known for that year.

We turn, now, to the applicable provisions of the statutes.

C. STATUTORY PROVISIONS

As noted above, the Plan at issue here is established under the authority of the OMERS Act, 2006. Section 16(1) of that Act sets out the power of the SC to determine the terms of the Plan:

The Sponsors Corporation shall determine the terms and conditions of the OMERS pension plans, subject to the restrictions set out in this Act.

Section 18 spells out the SC’s power to make amendments to the Plan:

The Sponsors Corporation may amend the OMERS pension plans, including the contribution rates for employees, subject to the restrictions set out in this Act.

Notwithstanding its independent statutory basis, however, the Plan is subject to the terms of the PBA, and the SC’s powers of amendment under the OMERS Act must be exercised in compliance with the PBA. Section 14(1) of the PBA provides that:

An amendment to a pension plan is void if the amendment purports to reduce,

a) the amount or the commuted value of a pension benefit accrued under the pension plan with respect to employment before the effective date of the amendment;

b) the amount or the commuted value of a pension or a deferred pension accrued under the pension plan; or

c) the amount or the commuted value of an ancillary benefit for which a member or former member has met all eligibility requirements under the pension plan necessary to exercise the right to receive payment of the benefit.

“Pension” is defined in s.1 the PBA as “a ‘pension benefit’ that is in payment”. “Pension benefit” is defined as

…the aggregate monthly, annual or other periodic amounts payable to a member or former member during the lifetime of the member or former member, to which the member or former member will become entitled under the pension plan or to which any other person is entitled upon the death of a member or former member.

Section 14(1) protects both the amount and the commuted value of the “pension” from reduction. “Amount” is not a defined term. “Commuted value”, however, is defined in s.1 as follows:

“commuted value” means the value calculated in the prescribed manner and as of a fixed date of a pension, a deferred pension, a pension benefit or an ancillary benefit”.

The Applicant argues that the amendment in question reduces both the amount and the commuted value of her pension within the meaning of 14(1) of the PBA, and should not have been registered.

It is common ground among the parties that s.14(1)(a) is not applicable to employees who are retired and who are therefore not continuing to accrue pension benefits. There is no issue here with respect to ancillary benefits; therefore s.14(1)(c) is inapplicable. Because the Applicant is already retired, she is in receipt of a “pension”; accordingly, the sub-section that applies to her situation is s.14(1)(b).

D. POSITIONS OF THE PARTIES

The Applicant has summarized her position in her Written Submissions as follows:

The Applicant takes the position that, by implementing the change to the formula on January 1, 2008, the Amendment in question had the effect of reducing the amount and the commuted value of a pension benefit accrued by pensioners under the pension plan before the effective date of the amendment and the amount and the commuted value of a pension accrued under the pension plan. The “old” formula for calculating changes in the Consumer Price Index (CPI) formed part of the indexation provisions of the Plan. It was vested when pensioners retired. The “new” formula paid cumulatively less in 2008 than what the old formula would have paid, thereby reducing the benefit, the pension and the commuted value. The Superintendent of Financial Services should have declared the Amendment void under section 14(1) of the Pension Benefits Act (“PBA”). (Written Submissions of the Applicant, February 1, 2010, para. 1)

The Applicant emphasized in both her written and oral submissions that she does not challenge the right of the SC to change the indexation formula from the OM to the NM per se. In fact, she supports the change in principle. Her objection focuses on the implementation date for the change. She argues that the amendment is void because it was implemented on January 1, 2008; it is the timing, and not the nature of the amendment that has the effect of reducing accrued benefits.

In her oral submissions, the Applicant pursued three distinct lines of argument in support of her general position that implementation on January 1, 2008 resulted in a reduction of her pension benefits. Her first argument is that we must look at the impact of the amendment on its effective date, January 1, 2008. She points to the incontestable fact that on that date, OMERS pensioners got smaller increases than they would have got if the OM had still been in place. An actuary calculating the commuted value of the Applicant’s pension on that date would have produced a figure lower than would have been produced if the OM had still been in effect. Accordingly, she argues, both the “amount” and the “commuted value” of her pension have been reduced by the impugned amendment, and the amendment is therefore void.

Second, she argues that the shortfall she and other OMERS pensioners have experienced to date under the NM has been permanently “locked in” and will never be made up. This argument relies on the expert evidence of Mr Duxbury, who quantified an initial loss to the pensioners of 1.27% as a result of the implementation of the NM, and projected, using standard actuarial methods for projecting inflation, that this loss would be ‘carried forward’ indefinitely.

Third, the Applicant submits that even if the NM and the OM, looked at individually, may produce equivalent results, changing from one to the other results in what she called a ‘hybrid method’, a method which will not produce the same results as either of the NM or the OM standing alone. She points out that the OM tracked the CPI in a volatile manner: a low September result would likely follow (or be followed by) a relatively high September result. The NM, however, is a smoothing method based on averages and designed to eliminate exactly the kind of spike that would allow retirees to make up for the unusually low increase they had been awarded in 2007. She argues therefore that the adoption of the “hybrid method” introduced just after a year in which the OM produced a significant shortfall for pensioners will result in an overall reduction of benefits.

We note here that while the Applicant initially took the position that OMERS had deliberately implemented the change in method on January 1, 2008 in order to save money for the Plan at the expense of the pensioners, she withdrew that argument in the absence of any evidence to support it.

The SC, the OMERS body that made the decision to amend the plan, argues that the amendment does not reduce accrued benefits. The position of the SC is summarized in its written submissions as follows:

The plan continues to provide full inflation indexing to its members. The amendment simply implemented two changes to the method of calculating the annual inflation adjustment: (1) using a 12 month average rather than a September to September ratio, and (2) changing the effective adjustment month from September to October.

It is a methodological change, and does not reduce the amount or commuted value of a pension or pension benefit accrued under the plan. (Closing Submissions of the Respondent SC, February 1, 2010, para. 4, 9).

The SC argues that OMERS pensioners continue to enjoy 100% inflation protection, as they did before the change in methodology. It concedes that the amendment may have transitory impacts that result in lower pension benefits and lower commuted values at particular points in time. Over time, however, these effects will even out and OMERS pensioners will be in the same position they would have been in if there had been no change in methodology. The SC argues that for purposes of s.14(1)(b), the relevant date for measuring the impact of the amendment is the date the decision to amend the Plan was made: October 3, 2007. On that date, the evidence established that the change in indexation formula had no impact on the value of the Applicant’s pension.

The AC takes essentially the same position. To quote from the AC’s written submissions:

While the January 1, 2008 inflation adjustment was less than what the inflation adjustment would have been under the old OMERS methodology, the evidence is that over time, pension benefits under either formula will be similarly indexed. As a result, it cannot be said that the Amendment has reduced either the amount or the commuted value of a pension or pension benefit accrued under a pension plan. (Revised Written Submissions of the AC, February 1, 2010, para. 7)

The AC also argues that the relevant date to assess whether or not the amendment is void is the date the amendment was passed:

Otherwise, plan sponsors who wish to pass plan amendments one or two years in advance of the amendment’s effective date (e.g. where a long lead time may be necessary or desirable from a notice of administration perspective) will have to wait one or two more years before knowing whether the amendment is valid or void. (Revised Written Submissions of the AC, February 1, 2010, para. 64)

The Superintendent agrees. He rejects the fundamental premise of the Applicant’s attack based on changes in value on the effective date of the amendment. In his initial submissions he argues that:

… the analysis of whether or not the Amendment is void cannot depend upon when it is introduced. It is the overall formula that matters, not particular results in an isolated time frame. (Overview of the Legal Submissions of the Superintendent, January 11, para. 4-5).

He fleshes this position out more fully in his final written submissions:

The Superintendent submits that the preferred analysis is that provided by the SC’s expert which more fully takes into account the conclusion that both methods will provide the same level of inflation protection over time. Owing to the specific entitlements which are protected under section 14 of the PBA, it is the effect of the change of indexation methodology on the lifetime value of the benefits provided under the pension plan which must be assessed rather than evaluating the benefit at a particular point in time. The available historical evidence as well as the accepted approach to valuing indexation benefits adopted by the actuarial profession (as articulated by the SC’s expert) demonstrate that the lifetime benefit has not been affected. Accordingly, the Amendment is not void under section 14 of the PBA. (Final Submissions of the Superintendent, February 1, 2010, para. 6)

The Superintendent draws our attention to the definition of “pension benefit” which refers to “…the aggregate monthly, annual or other periodic amounts payable to a member or former member during the lifetime of the member or former member”. He highlights two aspects of this definition: the word “aggregate”, and the phrase “during the lifetime of the member or former member.” He argues “that the entitlement that is protected in section 14(1) is not the amount of specific payment but rather the aggregate value of the stream of payments made from the pension plan during the course of the member or former member’s lifetime” (Final Submissions of the Superintendent, para. 31). He points to the concurrence of the experts that the two formulae will produce the same result over time to support the argument that the Applicant will get the same “aggregate” benefits “over her lifetime” even if she did get reduced benefits on January 1, 2008, and again on January 1, 2009.

The Superintendent acknowledges that s.14(1)(b) protects both the amount and the commuted value of a “pension”. He submits, however, that under the PBA, the concept of commuted value is an “artificial notion” for pensions in pay. He argues that the PBA protects commuted values primarily for the purpose of protecting transfer options under s.42. Once a pension is in pay, the PBA does not contemplate any transfer of its commuted value out of the plan. Therefore, he argues, the concept of commuted value has no relevance under s.14(1) for pensions in pay. He submits that s.14(1)(b) refers to commuted values only to protect transfer values for deferred members on plan termination, and “not to protect some notion of the commuted value of a pension in pay” (Final Submissions of the Superintendent, para. 49-53). In the alternative, he submits that the relevant date for calculating any commuted value for pensions in pay is the date on which the SC made the decision to amend the indexation formula; any other date is simply too indeterminate to be meaningful, in view of the purpose of s.14(1)(b) and the rights it is intended to protect.

E. ANALYSIS

Section 14(1) is a very important protection for plan members, standing as a bulwark against plan amendments that deprive plan members or former members of pension benefits already vested or accrued. The language used by the legislative drafters to realize that straightforward purpose is not entirely transparent, however, and no cases have been brought to our attention which raise issues under s.14(1) similar to those which confront us in this case. Accordingly, we must deal with these issues as a matter of first impression.

For ease of reference, we set out again the provisions of the PBA most relevant to this case. Section 14(1)(b) provides that:

An amendment to a pension plan is void if the amendment purports to reduce,

b) the amount or the commuted value of a pension or a deferred pension accrued under the pension plan…

“Pension” is defined in the PBA as “a ‘pension benefit’ that is in payment”. “Pension benefit” is defined as,

Applying this statutory language to the problem before us, we must answer two questions: (1) is the OM part of the Applicant’s “accrued” pension within the meaning of s.14(1)(b)? and (2) If so, does the impugned amendment purport to reduce the amount or commuted value of the Applicant’s pension? We address these questions below.

1. Is the OM Part of the Applicant’s Accrued Pension?

The Applicant asserts that the OM for calculating changes in the CPI formed part of the indexation provisions of the Plan and that the OM “was vested when pensioners retired”. All parties appear to agree that the Applicant did have a vested right to a pension with 100% CPI indexation. They do not agree, however, that she had a vested right to indexation in accordance with the formula built into the plan when she retired. We must therefore address the question of whether the OM is part of the Applicant’s pension accrued within the meaning of s.14(1)(b).

While this issue touches upon the wording of the statute, it is primarily a question of interpretation of the pension plan. Under the OMERS Plan as it stood prior to the impugned amendment, members such as the Applicant were entitled to a base pension calculated in accordance with the provisions of the plan. In addition, they were entitled to have that base pension increased on an annual basis in accordance with s.31. Section 31(8)(a) of the Plan makes this clear:

The pension payable to a person during his or her lifetime shall be determined in accordance with this Plan or a predecessor thereof and,

Each pensioner in the position of the Applicant is therefore entitled to a pension, the amount of which is increased each year by the “inflation adjustment”. For years from 1999 on, the “inflation adjustment” is defined in ss.31(2)¶2 as 100% of the “inflation increase”. “Inflation increase”, in turn, is defined in s.31(1) as “the percentage increase of the Consumer Price Index for Canada for September of the previous year over the Consumer Price Index for Canada for September of the year which is 2 years prior to the adjustment year” [the OM]. It is clear, then, that prior to the impugned amendment, the Applicant was entitled not just to a pension indexed generically to 100% of CPI, but to a pension which included annual adjustments calculated in accordance with the OM.

Does the fact that the right at issue here relates to an annual adjustment to the base pension, rather than to the base pension itself, make any difference to the “accrued” nature of that right? As a matter of actuarial practice, it would appear not. Mr. Duxbury gave evidence on that point. He was asked to comment on a portion of Ms Wagman’s report in which she assumed that accrued benefits included indexing. He responded:

Well, a pensioner, the entire benefit of the pensioner has been accrued to the date they retire basically. Once they’re retired, you know, they’ve accrued all their pension and whatever happens after that is affecting their accrued pension. That’s – accrued is usually referred to in the sense of the ongoing employees who continue to accrue service after the valuation date. So a pensioner or deferred vested member, someone who’s terminated membership is considered to have accrued all their benefit up to the date of the, of the termination or retirement, or what have you.

Q: Is there anything in your experience where indexation increases to a pension would not be considered an accrued benefit?

A: Well, I can’t think of any off-hand. It’s, you know, the CIA Standard, the Canadian Institute of Actuaries’ standards require that, for example, if you calculate the commuted value of someone who terminates you, includes the index, if it’s a part of the, of the Plan, which it is in this case. There are provisions that allow it to be excluded under the Act for funding purposes, but that doesn’t mean that it’s not an accrued benefit, that’s just that’s a special dispensation from the, from the funding rules. (Transcript of the hearing January 18, 2010, pp.74-5)

Ms Wagman’s report proceeded on the same assumption: the fact that the Plan provided for indexing was an important factor in valuing the benefit.

This common actuarial understanding is supported by the case law. The Applicant relied on the decision of the Manitoba Court of Appeal in Dinney v. Great-West Life, 2005 MBCA 36, which involved a class action brought by former members of a Great-West Life Assurance Company pension plan. At the time the plaintiffs ceased to be members of the plan, the plan had provided for annual pension indexing on a formula which, according to the plan text, “shall be as the Company may from time to time determine and will be related to the investment performance of the Great-West Life Assurance Company Canadian Employees’ Pension Fund.” In the mid-1980s, the pension fund’s performance far exceeded both the annual rate of inflation and the rate of annual wage increases, and the plan trustees stopped basing inflation increases on investment performance. A few years later, the employer passed a series of plan amendments which ultimately resulted in pegging benefit indexation to the CPI. The affected former plan members brought action alleging inter alia that they had a right to benefits indexed on the basis of the investment-related formula. They argued that this right vested at the time of retirement and could not subsequently be divested.

Great-West Life argued that inflation indexing was merely a “contingent and discretionary entitlement”; it was not vested, and was therefore vulnerable to amendment under the employer’s general power to amend the plan. After an extensive review of scholarly authorities, and both American and Canadian case law, including the decision of the Supreme Court of Canada in Dayco (Canada) Ltd. v. CAW-Canada, [1993] 2 S.C.R. 230, the Manitoba Court of Appeal rejected that argument, concluding that the provisions of the plan gave the plan members a right “to their entitlements under the then existing plan, including annual pension increments”, which “accrued” or “vested”⁷ on their retirement. In determining that indexation was a vested right, the Court refused to distinguish between pension amounts already fixed on retirement, and pension amounts which would be fixed later in accordance with a formula:

I reject the argument of counsel for Great-West that benefits that are not quantified do not meet the definition of a “pension benefit” in the Act.⁸ The quantum of the annual increment need not be a fixed amount so long as it is calculable in accordance with the provisions of the plan itself. For example, a provision in a pension plan tying annual increments to the CPI would meet the requirement. [para. 71, emph. added]

Accordingly, the Court of Appeal upheld the decision of the trial judge that the company had violated the vested/accrued rights of the plaintiff class when it adopted a method of indexing which was unrelated to the investment performance of the pension fund.⁹

In our view, there is no doubt that Ms McGrath has a vested or “accrued” right to pension indexing based on the OM. However, s. 14(1) of the PBA does not “carve in stone” all accrued benefits. What it does is protect those benefits from reduction. Accordingly, the crucial question before us is whether the impugned amendment reduces the amount or commuted value of the Applicant’s pension, within the meaning of s.14(1)(b).

2. Does the Impugned Plan Amendment Reduce the Amount or Commuted Value of the Accrued Pension?

a. Purpose or effect?

Before we turn to the substantive question of whether the amendment has reduced the Applicant’s accrued pension benefit, it is necessary to dispose of a preliminary interpretive issue which arises on the face of s.14(1). The section nullifies any amendment that “purports to reduce” accrued benefits. As a matter of dictionary definition and grammatical placement in the section, the word “purports” is susceptible of more than one interpretation. One possible approach is to look simply at the amendment on its face: what does it “purport” to do? If we were to adopt that approach, this case could then be simply resolved. Both actuaries testified that on its face, the amendment did not change the level of CPI protection provided by the plan; an actuary, looking simply at the language of the OM and the NM, would conclude that they were functionally equivalent. If effect is irrelevant, we would need to explore the issues no further. To their credit, however, no party took that position. All parties agreed that s.14(1) protected plan members from amendments which had the effect of reducing benefit entitlements, and not just those that appeared to do so on their face. When specifically asked by the Tribunal whether the section would mean the same thing if it read “reduces” such entitlements rather than “purports to reduce” such entitlements, all parties answered that it would.

Why, then, does the phrase “purports to” appear in the section? Counsel for the Superintendent offered the suggestion that the wording was designed, in effect, to signal the fact that amendments which violate the section are void: since they cannot have actual effect, they can merely “purport” to have effect. This suggestion is quite plausible in the overall context of the s.14(1): see also Joseph v. Joseph, [1966] 3 All E.R. 186 in which Lord Denning M.R, faced with similar language in the Landlord and Tenant Act, 1954, held that in the context of that statute, “the word "purports"… does not mean "professes". It means "has the effect of".” In any event, it is our view that in the overall context and in light of its purpose, s.14(1) requires us to inquire into not only what a plan amendment says on its face, but also into the effect of that amendment, to determine whether or not it reduces vested pension benefits within the meaning of the section.

b. Does the Impugned Amendment have the Effect of Reducing the Pension Benefit?

Section 14(1)(b) is directed towards reductions in vested pension rights. Plan amendments which improve benefits, or have a neutral impact on benefits, do not offend s.14(1)(b). We have held that the Applicant has a vested right to inflation adjustments based on the OM. The impugned amendment, which substitutes the NM for the OM, certainly affects that vested right. The respondents argue, however, that because the NM is the actuarial equivalent of the OM, the amendment does not reduce the value of the Applicant’s accrued pension rights. It is to that question that we now turn.

There are situations in which the concept of a reduction in benefits is relatively easy to apply. For example, cuts to a fixed formula for calculating a base retirement pension or the elimination of early retirement benefits present situations in which there has been, at least prima facie, a clear reduction in either the amount or the commuted value of a pension. With respect to an amendment like the one at issue in this case, however, the matter is not so straightforward. The amendment deals with an ‘escalation factor’, a formula for increasing the pension on an annual basis after retirement. In addition, this escalation factor is based not on a fixed formula (e.g. 3% a year), but on a formula that cannot be quantified in advance. Under the terms of the amendment, the NM was applied for the first time on January 1, 2008. But it was also applied again on January 1, 2009, again on January 1, 2010, and will continue to be applied again every January 1 for the foreseeable future. When and how do we measure the effect of such a provision on the Applicant’s benefits to determine whether or not there has been a reduction in their value?

As noted above, the Applicant had made three arguments in support of her position. Her first argument is a “point in time” argument addressed to both the amount and the commuted value of her pension; she argues that the assessment of the impact of the impugned amendment should be made on its effective date, January 1, 2008. Second, she argues that the shortfall she and other OMERS pensioners have experienced to date under the NM has been permanently “locked in” and will never be made up. Third, the Applicant submits that even if the NM and the OM, looked at individually, may produce equivalent outcomes, changing from one to the other results in what she called a ‘hybrid method’, a method which will not produce the same results as either of the NM or the OM standing alone.

Before we deal with the Applicant’s first argument, let us address her second and third arguments, which can be disposed of solely on the evidence. The Applicant’s “lock in” argument, that shortfalls experienced in the first two years will continue in perpetuity, rests entirely on Mr. Duxbury’s expert evidence with respect to the “locked in” nature of the results in the first two years. The “lock in” argument is an empirical proposition. To succeed on this argument, the Applicant would have to persuade us that after two years of shortfalls under the NM, the NM will not produce results that will even out over time with the OM. While Mr. Duxbury’s evidence may be sound as an actuarial exercise in projection, in our view it does not stand up to scrutiny as a reliable prediction of what will actually happen in the real world for individual OMERS pensioners. Mr. Duxbury’s extrapolation of the current impact of the NM into the future depends entirely on an assumption about the behaviour of inflation in the future: an assumption that inflation will progress at a completely uniform rate not just from year to year, but also from month to month, over time. If that assumption is made, it is obvious that any initial loss will be ‘locked in’; simple mathematics can produce no other result. Both actuaries agree, however, that while the pace and behaviour of future inflation is unpredictable, it can be predicted with confidence that it will not behave in the uniform fashion postulated by Mr. Duxbury. As evidence of the comparative impact of the OM and the NM ‘in real life’, therefore, Mr. Duxbury’s projection has little probative value.

The Applicant attempted to bolster the “lock in” argument by pointing to the actuarial gains in the Plan resulting from the implementation of the NM. Mr. Duxbury testified that if the SC had agreed to make the NM retroactive to January 1, 2007, it would have cost the Plan some $261 million in reduced actuarial gains. He characterized the decision not to implement retroactively as a decision to “capture” that actuarial gain for the Plan. He likewise testified that implementation in 2008 produced an actuarial gain of $110 million, which he also characterized as a gain “captured” by the Plan. The gist of his evidence, and the Applicant’s argument on this point, is that these actuarial gains represented quantified and permanent gains to the Plan and losses to the pensioners arising out of the change to the NM. We are not persuaded by this argument. On this point, we prefer the evidence of Ms Wagman that actuarial gains (or losses) represent nothing more than temporary deviations from actuarial projections, and tell us nothing about the future impact of the amendment. Their presence or absence does not alter the fundamental proposition that over time we can expect the impact of the NM to be the same as the OM. Furthermore, we note that the largest of the “captured” actuarial gains referred to in Mr Duxbury’s evidence relates to 2007, prior to the implementation of the NM. In our view, losses or gains incurred prior to the implementation of the NM can have no relevance to the issue before us of whether the NM reduced the Applicant’s accrued pension. Accordingly, the Applicant has failed to prove that the short-term shortfalls experienced in the first two years as a result of the NM will be permanently “locked in”.

The Applicant’s “hybrid” argument, that the introduction of a smoothing method after using a volatile method will make permanent the losses experienced under the volatile method, has some superficial plausibility. It cannot be accepted, however, for two reasons. First, we are simply not in a position to evaluate this type of argument without the assistance of expert testimony. The Applicant’s expert did not testify with respect to inflation protection gaps created by “hybrid methods”. Instead, his evidence focused on the “lock in” argument, discussed above, the argument that early initial shortfalls imposed by the NM will continue into the future (based on an assumption as to the pattern of future inflation which both experts agreed is not realistic). As noted above, the evidence of both actuaries was that the OM and the NM were actuarially equivalent. Neither was asked to address the proposition now put forward by the Applicant, that a hybrid of these two methods produces a result that is not actuarially equivalent. Accordingly, the Applicant has not met the burden of proof with respect to this argument.

Second, the key factual underpinning of the “hybrid” argument is the shortfall produced by the OM in 2007. It is obvious from her 2007 correspondence with OMERS and the representations she made to the Superintendent that the Applicant has had difficulty accepting the very small increase the OM produced in the OMERS plan for 2007, compared with what the CPP method would have produced (.70%, compared to 2.1%). She argued vigorously to OMERS (and again to the Superintendent) that there should have been a “catch up” payment to pensioners to avoid locking in this “loss”. The “hybrid” argument depends for both its emotive and its logical force on the proposition that OMERS pensioners were unfairly treated under the OM in the years prior to the implementation of the NM. We cannot remedy any such perceived unfairness arising from the prior impact of the OM; our task requires us to focus only on the effect of the NM on the Applicant’s accrued pension.

We have rejected the “lock in” and the “hybrid” arguments based on the evidence before us. Had the Applicant succeeded in persuading us to her point of view, the practical consequences for plan administrators would have been decidedly unfortunate. Both the “lock in” and the “hybrid” arguments are based on outcomes that could not be determined at the time the plan amendment was adopted; indeed, in some circumstances they could not be determined for years to come. In his written submissions, the Superintendent argued that the Applicant’s analysis would give plan administrators no guideposts for determining in advance whether a plan amendment which was prima facie valid might subsequently become invalid. He highlighted the policy problems that would flow from that approach:

The approach adopted by the Applicant implies that the answer to the question of whether or not the Amendment is void will depend upon when the question is asked. If the question is asked at a point where the cumulative level of inflation protection under the new formula is lower than under the old method then the Amendment is void. In a following year, if the reverse is true, then the Amendment will not be void. Accordingly, the Amendment could be perfectly acceptable one year but contrary to the PBA the next only to be acceptable in the third year. (Final Submissions of the Superintendent, para. 40)

Who wins and who loses, he argues, would depend entirely on when the matter is litigated.

The Applicant’s response to this argument underscores the arbitrary nature of what she is asking us to do. Although she urged us in her initial written submissions to focus on the outcome on January 1, 2008, the date the NM was first used, in her final written submissions she sought to focus on the evidence as of the date the report of her expert witness was completed (December 2009). She argued:

While it is true that, as in all actuarial calculations, future CPI values may deviate from the present assumptions, a point in time must be chosen to calculate an actuarial value. For the purpose of the present case before the Tribunal, that time would appear to be the time the expert witness reports were written. Any party might suggest a later date hoping for a better outcome by random chance. The Applicant submits that this would potentially prejudice the results. (Written Submissions of the Applicant Incorporating the Evidence Led at the Hearing, para. 35; emph added)

To select the date at which expert witness reports were prepared for litigation would, in our view, be an equally random choice, unrelated to the purpose of s.14(1)(b) and to the benefits protected by that section.

Let us return, then, to the Applicant’s first argument, that we should evaluate the effect of the amendment on its effective date, January 1, 2008. On this argument, the amendment stands or falls depending on its impact on its effective date. In approaching this argument, we are mindful that s.14(1) protects two types of pension value: the “amount” of the accrued pension, and the commuted value of the accrued pension. As we have already pointed out, “pension” and its included term, “pension benefit” are defined terms in the PBA. While the “amount” of the pension is not separately defined, the PBA defines “pension benefit” in terms which refer to the “amount” of that benefit, as follows:

In other words, a “pension” or “pension benefit” is not the amount of a particular pension payment; it is an “aggregate” of the amounts of all the payments to which the Applicant is entitled. This is comprehensive language, directing us towards a comprehensive assessment of the nature and value of the Applicant’s pension as payable to her during her retirement. The question of whether a plan amendment reduces the “aggregate” amount of a pension is a very different question from whether it reduces a specific periodic payment. In responding to the question of whether the impugned amendment has reduced the “aggregate” value of the amount of the Applicant’s pension, we must make a practical and comprehensive assessment, on the basis of the evidence before us, as to the overall effect of the substitute method (the NM), compared to the effect of the original method (OM), on the amount of the Applicant’s pension.

In our view, the legislative intent is that the assessment of the longer-term impact of such an amendment on aggregate pension payments be made on the basis of information available at the time the amendment was adopted. A “wait and see” approach under which the assessment might be made as of a later date is simply not practicable, for three reasons. First, the result of that assessment could change over time, and it would be unreasonable and impractical to require that compliance with the requirements of s 14.1 of the Act be determined periodically for an indefinite period. Second, in the event that an amendment which initially appeared to be valid (and which initially might have increased pension payments) were found some time later to have had the effect of reducing aggregate pension payments and were therefore found to be void, the amendment could not practically be reversed, since pension payments including those to pensioners subsequently deceased would already have been made on the basis of the amendment. Finally, it is important that plan sponsors and the Superintendant be able to determine promptly and with finality whether or not an amendment is valid, taking the requirements of s. 14(1) into account – otherwise plan sponsors would be precluded from making amendments of the type whose impact could not be determined with certainty at the time of adoption, however desirable those amendments might be. Accordingly, our vantage point for assessing the longer-term impact of the amendment on the amount of the pension must be the date the decision is made, taking into account the information reasonably available to the plan sponsor at that time.

It would not be appropriate, however, to consider the longer-term impact of the amendment with respect to the commuted value of the Applicant’s pension. Although it is the commuted value of the overall aggregate “pension benefit” that must be assessed, the PBA mandates that the commuted value be calculated as at a fixed date. We must therefore determine the appropriate date on which to perform that calculation, in order to determine whether or not the commuted value has been reduced by the impugned amendment within the meaning of s.14(1)(b).

Focusing first on the issue of the amount of the Applicant’s pension, what does the evidence show? All parties agree that as of January 1, 2008 the Applicant’s pension was smaller by some .47% than it would have been if the OM had still been in effect, a minor but measurable reduction in the amount she would otherwise have received in 2008 if the OM had still been in place. Set against that “point in time” evidence, however, is the evidence of both expert witnesses that over time, the OM and the NM would be expected to have similar effects on the amount of the pension. The historical evidence about the relative impact of the two methods bears this out. Over any particular time frame, the two methods yield only minor differences. Over the entire period between 1992 and 2007 during which OMERS fully indexed its pensions to the CPI using the OM, the CPP method (i.e. the NM) would have yielded marginally better results. The snapshot taken by the Municipal Retirees Organization Ontario for the period from 1999-2007, a snapshot described as “misleading” by the Applicant in her April 2007 letter to OMERS, showed the OM ahead, whereas the Applicant’s own snapshot from 2000-2007 showed the NM ahead. The variable outcome of these different snapshots highlights the arbitrary impact of assessing the effect of either the OM or the NM at any specific date or over any randomly-selected period of time. It also highlights the fact that both the OM and the NM will have different effects on individual pensioners depending on their specific circumstances. Those who were OMERS pensioners from 1992 to 2007 would have been slightly better off if OMERS had used the NM instead of the OM over that period of time. For 2008 and 2009, OMERS pensioners would have preferred that the OM had stayed in place. But if we expand the lens a little, the picture is very likely to change. Already, in 2010, the NM has yielded a better result than the OM. It is conceivable that even as early as January 1, 2011 the Wagman Alternative will have kicked in and the Applicant will be ahead of the game under the NM. With the NM as with the OM, we simply do not know who will come out ahead in the longer run.

The Applicant is asking us to decide the case based solely on a single narrow snapshot that is unlikely, based on all the evidence, to be representative. In our view, s.14(1) does not dictate so arbitrary a result. We are persuaded that for amendments such as the one before us, the statute does not gauge whether or not the amount of a pension has been reduced based only on its immediate impact on the first periodic payment after it comes into effect (or indeed, only on its impact on periodic payments during the period between the date of implementation and the date of hearing). It instructs us to take a longer view. From that perspective, what does the evidence show us? As we have already noted, we do not accept the expert evidence of Mr. Duxbury that the short-term impact of the NM will be “locked in” in perpetuity. Indeed, it appears quite probable that the relatively minor deviations between the OM and the NM will be ironed out very soon. Likewise the evidence before us (both expert and non-expert) does not support the Applicant’s argument that shifting from one method to the other leads to a benefit reduction over the longer term. The persuasive evidence on the long-term (i.e. “aggregate”) impact of the amendment is the evidence of both actuaries, that the OM and the NM, are actuarially equivalent, and that “over time” they are expected to produce the same level of protection, 100% inflation protection as indexed to the CPI. On the basis of the overall evidence, then, the Applicant has failed to persuade us that the amendment has the effect of reducing the amount of her accrued pension within the meaning of s.14(1)(b).

However, we must still address the question of whether the impugned amendment reduces the commuted value of that pension within the meaning of s.14(1)(b). As noted above, this question raises somewhat different issues. The Superintendent has submitted that because the Applicant’s pension is in pay, we need not concern ourselves with the issue of commuted value. He argues, in essence, that the concept of commuted value has application under the PBA only with respect to active and deferred members whose seek to transfer their benefits out of a pension plan under s.42; since pensions in pay cannot be so transferred, they have no commuted value under the statute. While this argument has some force, we do not ultimately find it persuasive. The Applicant may have no right under the PBA to commute her pension; it is nevertheless clear that from an actuarial perspective, a pension in pay has a commuted value. The definition of “commuted value” in the PBA definitely contemplates the calculation of such values for pensions in pay. Neither expert witness had any difficulty with the concept of calculating a commuted valued for a pension in pay. We hold that s.14(1)(b) requires us to address the question of whether the commuted value of the Applicant’s pension has been reduced.

While the PBA’s definition of “pension benefit” supports an approach to calculating the amount of a pension within a flexible time frame, a commuted value cannot be calculated “at large”; it requires the identification of a specific date upon which the calculation should be made. The Applicant has submitted that the appropriate date upon which to make the calculation of commuted value is January 1, 2008, or alternatively on the date her expert prepared his report, both dates on which the expert witnesses agreed that the commuted value of the Applicant’s pension was lower as a result of the implementation of the NM than it would have been if the OM had still been in place. The responding parties have argued that the appropriate date is October 3, 2007, the date the amendment was adopted by the SC. On that date, the expert witnesses agreed that the commuted value of the Applicant’s pension was the same under both the OM and the NM. The choice of date upon which the effect of the amendment is measured is therefore critical to the outcome.

We have decided that the appropriate date upon which to measure the effect of the amendment on the commuted value of the Applicant’s pension is October 3, 2007, the date the amendment was passed. In making this determination, we have taken account of the nature of the amendment. As discussed above, the amendment will be applied not just on January 1, 2008, its ‘official’ effective date, but also on a series of dates thereafter. Furthermore, its impact on any January 1 cannot be quantified in advance. For the same reasons that it would be arbitrary to judge the impact of this amendment on the amount of the pension only on the first of the series of dates on which it will be applied, it would be arbitrary for purposes of s.14(1)(b) to measure the impact of the amendment on the commuted value of the pension on the first of that series of dates.

Commuted value, by its very nature, is a forward-looking measurement requiring an actuary, at a fixed point in time, to make predictions and judgments about a series of future events. The impact of the impugned amendment on commuted value can be meaningfully measured as of the date of decision. Neither expert suggested that any problems were posed for that exercise by the fact that the amendment was not yet governing the calculation of the CPI adjustment. On that date, it was possible to measure the “pure” effect of the amendment, untainted by the random impact of any particular application of the amendment on any particular January 1. On that date, the two formulae were actuarial equivalents. On that date, the plan administrator could make a reasoned assessment of the impact of the amendment on the commuted value of pensions in pay. We are persuaded, therefore, that the impugned amendment did not have the effect of reducing the commuted value of the Applicant’s accrued pension within the meaning of s.14(1)(b).

The Applicant’s interpretive approach to calculating both the amount and the commuted value of pension for purposes of s.14(1) would leave plan administrators in the untenable position of being unable to pass an amendment like this one – making a change that all parties agree is desirable – with any degree of confidence that the amendment is valid. The Applicant has made a number of suggestions about how, in her view, OMERS should have proceeded in implementing the impugned amendment. If her short-term approach to this amendment were to be adopted, however, none of her proposed alternatives for implementation would have truly insulated the Plan from potential attack under s.14(1)(b) of the PBA. The Applicant’s initial proposal was a special “catch up” increase for pensioners which would have put them in the same position they would have been in if the NM had been in place all along. As a practical matter, such a benefit enhancement might have avoided this litigation. But it would not have changed the fact that on January 1, 2008, the NM yielded less than the OM. Pensioners might simply have pocketed the enhancement and come forward to make the same argument the Applicant is now making: that the amendment was void based on its impact on its effective date. Similar problems arise with the Applicant’s proposal to implement the NM retroactive to January 1, 2007. Like the “catch up” payment, such a retroactive benefit enhancement might have avoided the litigation. But once again, if the Applicant’s approach were correct, it would not have guaranteed the validity of the amendment; pensioners might still have come forward in subsequent years to argue that the impact of the NM had the effect of reducing vested benefits. Her final proposal, that OMERS defer implementation “until it can be accomplished without losses to OMERS pensioners”, is clearly impracticable; on the Applicant’s theory of the case, how would OMERS ever be able to recognize that moment, except retrospectively after the passage of time?

We recognize that our decision leaves individual pensioners like the Applicant in a position in which they received indexation increases for 2008 and 2009 which were smaller than they would have received if OMERS had not decided to change the indexation formula. We also recognize that for some individual pensioners, events may unfold in a way in which the differences will not be made up to them over time. This is, however, the inevitable result of any change in the mechanics of an indexation formula. Furthermore, a DB pension plan cannot and does not produce identical results for all members. The value of DB benefits may well depend on individual lifespan. For example, members who die without a surviving spouse shortly after retirement will not get as good a ‘return’ on their contributions as members who live longer. The PBA recognizes that pension plans are collective instruments. It should not and does not force us to nullify an amendment of this type, designed to apply over time and passed in good faith for the benefit of all plan members, simply because it may have a modest negative impact in the short term.

The amendment at issue has left the Applicant and other OMERS pensioners where it found them: with a pension which is indexed 100% to increases in the CPI. We find that the Applicant has failed to meet the onus of proving that such an amendment reduces either the amount or the commuted value of her pension.

F. ORDER

For all these reasons, we dismiss the Application.

DATED at Toronto, Ontario, this 26^th day of March, 2010

“Elizabeth Shilton”

Elizabeth Shilton Member of the Tribunal and Chair of the Panel

“David Short”

David Short Member of the Tribunal and of the Panel

“Ralph Scane”

Ralph Scane Member of the Tribunal and of the Panel

Footnotes

The start year refers to the year the pension received full indexation.
Change in 12 month average October to October CPI figures.
Change in September over September CPI figures.
This column shows the cumulative difference in the inflation adjustment between the CPP method and the OMERS method from the start year to 2007. For example, cumulatively, the inflation adjustment over the period 1992-2007 would be 0.9% higher under the CPP method than the OMERS method.
His 2011 estimate was based on actual CPI data from November and December 2009, ‘topped up’ based on an estimated 2% overall annual increase in CPI thereafter in 2010.
This is not a straightforward mathematical calculation; it involves comparing the compounded cumulative increases under the two methods.
The Manitoba Court of Appeal saw the terms “vested” and “accrued” as interchangeable for purposes of this analysis: see paras. 30-32. With respect, we agree.
The Manitoba statute defines “pension benefit” in terms similar but not identical to the PBA: see para. 41 of the decision.
The Dinney decision discussed here deals solely with liability. The Dinney case returned to the Court of Appeal on the issue of damages: 2009 MBCA 29, [2009] M.J. No. 116. In that subsequent decision, the court held that while the pensioners had a right to a pension indexed in relation to the plan’s investment performance, they did not have a vested right to the precise formula which the company, in the exercise of its discretion, had previously applied. The plaintiffs’ application for leave to appeal to the Supreme Court of Canada was refused on December 17, 2009.