This blog is based on remarks to the National Coordinating Committee for Multi-employer Plans, Miami, Florida, September 21, 2019.(Edited for Length.) I have the original transcript , supplied by Dr Con Keating and published by C.D. Howe Institute’s Pension Policy Council. Randy Bauslaugh is a Partner at McCarthy Tétrault LLP,
My Experience with a Contingent Pension Plan
and the Lessons Learned
For many years, I have had the privilege of being legal counsel to the board of trustees of a multi-employer pension plan (MEPP) established in 1972. The employers are Canadian members of an association of schools in the US and Canada. The Canadian plan is modelled on the defined-benefit (DB) MEPP established in 1948 by the school association for US schools and their employees.
Both plans are administered by joint boards of trustees, one in Canada, and one in the US. The boards include representatives of participating employers and employees.
The obvious difference between the US plan and the Canadian plan is that the defined benefits promised under the Canadian plan are contingent on the funded status of the plan. If plan liabilities get too far ahead of the assets and the agreed fixed rate of contributions, the trustees of the Canadian plan have a fiduciary obligation to consider reducing accrued benefits to bring the assets and liabilities back into balance.
In Canada, this type of plan has many names, including “negotiated cost,” “shared risk” or “target benefit.” In other countries it might be called a “defined ambition,” “collective defined-contribution,” or in the US, a “composite” plan. Whatever the label, these plans all share the same characteristics. They all are a hybrid of DB and defined-contribution (DC) design. In my view, they bring together the best of both DB and DC plans, namely, the benefit predictability and cost efficiency of DB plans and the cost certainty of DC plans.
The actual operation of both school plans is carried out in the US. The plans are substantially similar, but increases in life expectancy, an aging workforce, the fall-out from the 2008-2013 recession and sustained low interest rates have resulted in two very different consequences for the Canadian and US plans.
In the period following 2013, both plans struggled to meet minimum legislated funding
Unfortunately, funding pressures forced the US plan to convert to a DC arrangement in 2019. But trustees of the Canadian plan had another option. In 2015, they voted to reduce accrued early retirement rights by increasing the age at which an unreduced early retirement benefit would be available.
Changing the age from age 61 to age 63 was enough to relieve the immediate funding pressure. By 2017, they were able to partially restore those rights by moving the age
of unreduced early retirement to age 62.
Because legislation in Canada accommodates a shared-risk model, the Canadian plan was able to bend, not break when faced with economic and demographic adversity. TheUS plan had no such option, so it broke.
Today, the Canadian plan continues to improve its funding position. Based on a 20-year actuarial projection study completed this year, it is highly probable that the Canadian plan will continue to deliver predictable and cost-efficient lifetime retirement income on its current cost basis for the foreseeable future.
But getting back to our tale, how did stakeholders (in the Canadian plan) react to the reductions?
Surprisingly, only a very small number of employers and plan participants expressed any concerns at all; fewer than 10 percent. But that minority included one
of the largest employers in the plan. Pressured by local financial advisors in their communities to move to DC retirement savings arrangements, some members
threatened to leave the Canadian plan and adopt DC arrangements.
The trustees responded by organizing a cross-country road trip with a couple of trustees, the plan’s Executive Director, the plan actuary and me to deal directly with the employers and participants who had expressed concern. We explained the nature of the
plan. We pointed to other plans that had to break in recent years because they couldn’t bend – plans like those of Nortel, steel companies and airlines. The messaging made much of the fact that the winds of economic and demographic change wouldn’t break
this plan because of the strength to be found in its pooling and sharing of risk. That pooling and risk sharing made the plan cost-efficient and sustainable.
The messaging highlighted the cost efficiencies of DB-style plans over DC arrangements. We pointed to expert studies in both the US and Canada thatshowed that contributions to a DC plan would have to be twice as much to provide the same retirement
income as the Canadian plan already provided. We shamelessly borrowed the phrase used in a National Institute on Retirement Security paper to argue that the Canadian plan is able to deliver a much “better bang for the buck” than the average DC arrangement.
The math the actuaries did for these meetings also showed very clearly that if the employer had the proposed DC arrangement in place over the relevant period with the same rate of contribution, the accrued benefits would be much smaller than those under the plan, even with its cuts.
We occasionally got help from some of the meeting participants. At least a couple of the participants observed that the factors affecting the plan’s conditional DB liabilities would also affect their own savings arrangements and expectations. Another pointed to a scholarly article indicating that less than 4 percent of Canadians can meet or beat the average rate of return of the average DB plan. And another commented that the change was “just” an unreduced early retirement benefit, something most plan participants never took advantage of anyway.
The good news was that participating employers and participants accepted the change. No one left the plan. But the news gets even better.
The really surprising part of this story is the number of Canadian plan stakeholders – employers and participants – who actually wrote in to compliment the trustees on their decision to “protect the plan.” In fact, based on surveys conducted by thetrustees after the benefit reduction, the vast majority of plan participants and participating employers were more satisfied with the plan than they had been priorto the cuts!
What? How could that be?
I think a lot of the positive reaction was simply the by-product of effective communication by the trustees and plan staff around the conditional nature
of the benefit promise over a period of more than two decades.
I started providing advice to the trustees of the Canadian plan sometime in the late 1990s. Initially, they hadn’t fully realized that accrued benefits could be reduced while the plan remained ongoing. Evidently, their prior lawyer had never said anythingabout that, or maybe they didn’t want to hear it. They also didn’t realize they didn’t have to make payments to Ontario’s Pension Benefit Guaranty Fund (PBGF.)
That’s because the benefits can be reduced, so the PBGF would never apply. Eventually they came around. They got a PBGF refund and started to modify the tone of their communications.
In my early days as their counsel, the trustees were understandably apprehensive about referencing the conditional nature of the DB promise.
“If we tell people that benefits can be cut, it will ruin trust in the plan by
employers and employees. Maybe we have that right, but we can’t be so explicit or employers will leave. Thanks for your legal advice, Randy. We agree. But we
can’t be naïve. We can’t just stick this in their faces.”
They were also concerned that the parent organization in the US would have grave concerns if the Canadians were suddenly communicating anything that stepped back from the rock-solid promise to provide accrued benefits, such as existed
in the US. They did get some pushback, but they persevered with a transitional communication plan over a period of years that subtly and gently
acknowledged the contingent nature of the benefit.
I think this escalating approach helped to alleviate suspicion that a sudden change of communication might have prompted, and helped tremendously when
the cuts had to be made in 2015.
And then, late in 2018, Canadian plan employers and participants heard through the association grapevine (not from the trustees) the shocking news that the US would convert to a DC arrangement in 2019. Suddenly, the target-benefit nature of the plan was
not a defect to be apologized for; it was a strength to be celebrated.
I wanted to relay this tale not just because it is a good news story. Not just because it actually happened. And not because it refutes the fears often expressed by those concerned about the conditional nature of a contingent DB promise; it also delivers
some useful lessons.
Lessons in Legislation
There is legislation in several western countries that accommodates shared-risk plan designs. Some of that legislation has been in existence for decades, as it has in the Canadian province of Ontario, in the Netherlands and in Iceland. A little known statistic
is that in Ontario contingent plans are the dominant form of DB pension provision measured by number of participants.
But Iceland must surely be the poster child for contingent plans. There are only 25 workplace pension plans in all of Iceland but they cover 95 percent of
the workforce. All are industry-wide MEPPs. All are target-benefit plans. All are jointly trusteed. Iceland’s 2008-11 economic meltdown resulted in 50 percent cuts to accrued pensions, including pensions already being paid. Almost all of the cuts have since been
restored; proof that the contingent design works when supported by a legislative framework that accommodates it.
A key to appropriate legislation does not seem to be prescriptive rules, but a focus on fiduciary duty. Fiduciary duty is after all the highest duty known to the common law, and the one that comes closest to resembling a moral code of conduct as well as a legal
duty. It is also a really good sort of legal polyfilla that can effectively fill in legislative gaps, so the legislation doesn’t have to be so prescriptive, which in turn
enables fiduciary discretion to respond appropriately to a broad variety of circumstances and challenges.
Legislation that leaves considerable latitude to fiduciary governance seems to be a key to successful outcomes.
But fiduciaries like safe harbours, and governments like control. I understand the desire for prescriptive rules. If there are to be prescriptions, my advice would be to go light on tick boxes and heavy on fiduciary responsibility. I would also advise that legislators
rein in any urge to enact overly prescriptive rules relating to governance structures and disclosure obligations; that they allow wide funding corridors before triggering mandatory benefit reductions and improvements; and that they never dictate the type or
ordering of benefit adjustments.
Let me explain the last point.
For almost 50 years, the rules in the Province of Ontario simply required trustees to ensure that a contingent DB plan was provisionally funded ona going concern actuarial basis (with an ability to amortize emerging deficits over 15 years). But in the 1980s “solvency funding” rules were introduced. Those rules required plans to ensure that the plan would be fully funded if the plan were wound up immediately, and emerging solvency funding deficits had to be amortized over 5 years. Thankfully, contingent plans
were, until the mid-2000s, exempted from solvency funding, except that they had to do “solvency testing” and then take the recommendation of an actuary on how best to address any solvency deficit as fiduciaries.
The theory was that in a MEPP environment they were less likely to fail, so making sure assets equaled liabilities in the event of an immediate wind up was useful as a gauge, but not a strict funding requirement.
In my experience, that was a good approach, except that in many cases it often resulted in failures on the part of representative fiduciaries to make hard and timely decisions about cutting accrued benefits. It also resulted in what in hindsight were clearly improvident decisions about benefit improvements. The practical reality is that well-intentioned fiduciaries in a representative governance model are subject to the
influence of their appointing constituencies. Union trustees are under enormous pressure to avoid any benefit reductions. Management trustees are often under considerable pressure to promote benefit improvements as a means of enabling employers to avoid current wage increases.
In my experience, appropriate legislative thresholds that expand representation to other beneficial interests, and that promote or require independent and expert trustees as part of the mix can help. But funding-related rules that trigger benefit cuts or prohibit improvements may be a better solution to enable representative trustees to act more independently of their union or management constituents.
When the Canadian province of New Brunswick introduced its shared-risk plan legislation in 2013 it provided some prescribed thresholds. First, it included a requirement that an actuary certify, based on stochastic and deterministic modelling, that the plan assets and rate of contribution would support a 95 percent probability of paying the base benefit and a 75 percent probability of paying the ancillaries.
Anything below those probabilities would trigger mandatory accrued benefit reductions (unless contributions could be increased).
The New Brunswick legislation also imposed an ordering of benefit cuts and benefit restorations. New Brunswick’s probabilistic approach received some criticism; mainly, that it was too prescriptive or too sophisticated. The objection to sophistication seemed to be from smaller consultancies that were not equipped to do the necessary scenario testing. I can also tell you from my experience that many such plans elsewhere in the country were already doing such probabilistic modelling as a sort of fiduciary reality
check. The trustees of the Canadian schools plan ha been doing it for many years.
Probabilistic testing is at the very least a good fiduciary idea. New Brunswick’s probabilistic approach is not a bad idea, but its decision to regulate the order of
benefit cuts and benefit restorations is. That ordering limits fiduciary discretion in a manner that can result in more harm than good because it doesn’t take into
account the specific facts and circumstances.
Not all plans and not all industries are the same. The tick box approach to cutting and restoring benefits ignores circumstances, ignores plan design (or assumes theyare all the same) and alleviates fiduciaries from exercising discretion in a manner that is in the best
interest of the specific plan stakeholders.
In contrast, the trustees of the Canadian plan at the centre of this tale considered several different forms of adjustment. In the end they took away unreduced early retirement
at age 61 knowing with a high degree of certainty it would not likely affect many participants and would improve the funded status of the plan.
It should be noted that the pure fiduciary approach in Ontario has over the past decade been replaced by evolving legislative developments that impose mandatory thresholds for benefit reductions and mandatory restrictions on benefit improvements.
There is some good and bad to this. Like the proposals for contingent plans in the US, Ontario is now requiring that benefit improvements not supported by contribution increases cannot be made unless going concern funding is at 100 percent, plus a cushion. The cushion is not a flat rate of 20 percent as seems to have been proposed in the
US. Instead, it is a stipulated percentage (depending on whether the plan is ongoing or closed to newmembers); plus an additional percentage based on the plan’s asset mix(fixed income assets to non-fixed income assets); and another percentage reflecting the plan’s going concern valuation interest rate over the plan’s benchmark rate.
Benefit reductions do not have to be made if the cushion is not there; but it seems
to me that fiduciary duty suggests that the trustees should at least consider benefit reductions whenever a going concern valuation discloses a funding ratio
of assets to liabilities that falls below the legislated minimum for permitting improvements.
Under proposed legislation, benefit cuts are not absolutely required unless the funding ratio falls below 85 percent. The Canadian school plan is presently in the 90s and improving.There have been some objections to imposition on such thresholds as intruding on labour-management relations – which may be proof that they were needed.
Fiduciary concerns should not be dependent on, or guided by, collateral bargaining interests. They should be done in the best interests of the plan stakeholders– employers, employees and others with rights to benefits. The imposition of these kinds of thresholds
certainly helps representative trustees in certain industries manage the practical pressures imposed by their management and labour appointers. But any legislation ought to provide wide corridors, so the result is more likely to be independent fiduciary
decision-making that can be responsive to the particular needs of the particular plan and its current membership.
In my view, the fact that there is a lot of “wiggle room” between 85 percent and the funded ratio plus the cushion is a good thing. It means plan benefits
are not constantly in a state of adjustment. It seems to me that the US proposals for contingent plan legislation could result in annual adjustments – putting a plan in a continuous unsteady state and its participating employers and beneficiaries in a state
of high anxiety.
Stability is something the trustees of the Canadian schools plan strive to achieve because they feel it promotes confidence in the plan to its stakeholders. It is also consistent with the long-term nature of the arrangement. But this wiggle room also reflects the reality that economic and demographic factors are in a constant state of flux and that
actuarial science cannot predict the future wit certainty. It is not an exact science. It is not a crystal ball. It is a sophisticated estimate.
Governance is the key to a successful contingent plan. I have been attending part of the Canadian schools plan trustee meetings every quarter for about 20 years. Even today, I cannot say for sure who is a management and who is a labour trustee. They work
together seamlessly and collaboratively.
Frankly, that is my personal litmus test for excellence in governance.
Going beyond minimum standards and embracing a willingness to consider innovations in governance is also an attribute of excellence in governance. More than a decade ago, the trustees of the Canadian school plan reshaped their eight-member board of trustees from a purely representative board of four employer and four employee trustees.
One of the legislatively required employee seats is often allocated to a retiree or deferred vested participant. One of the employer seats is filled by an independent trustee with expertise in finance.
The Canadian schools plan trustees have also modified their trustee selection process over the years to try to weed out the labour-management and sponsor organization politics and pressures.
Even though the legislation dictates a representative governance approach – one more appropriate to resolution of labour disputes, than governance of financial institutions – the trustees have embraced governance concepts more common in the boards of
directors of financial institutions.
What do I mean?
First, trustees are not appointed because of key positions they hold within a union, or
any other employee or employer association. Some years ago, the trustees developed a skills matrix for new trustees that identifies critical competencies, like relevant professional experience in finance, governance, law, public relations and human resources.
They also look for personal effectiveness skills and specialized pensions or benefits knowledge as well as representation. The trustees use the skills matrix
as a guide to trustee recruitment, selection and replenishment. The trustees also promote and support ongoing trustee education.
In Canada, legislators are also reviewing the basic rule that contingent plans be administered by a joint board of trustees, at least one-half of whom represent
employees. They are considering rules requiring participation of other plan beneficiaries such as retirees and persons with deferred vested entitlements.
They are also considering whether to require independent or expert trustees to be appointed. Other jurisdictions, like the Netherlands, already have such requirements, and they continue to tinker to achieve continuous improvement in governance.
Design is a critical aspect of composite plan success.
The design of a DB plan will almost naturally have certain “levers” that can be pulled without changing the core benefit promise, such as early retirement subsidies, indexing, bridge benefits, even the “normal form” of benefit. Consciously designing a plan with
“levers” will enable fiduciaries to adjust “ancillaries” rather than the core or basic defined-benefit promise.
It is important to take into account “levers” that respond to the particular characteristics of the industry the plan covers.
Some “levers” like indexing not only provide trustees with a collateral benefit they can take away before touching the core benefit, they are also “levers” that can be inserted in a conditional or ad hoc way to improve core benefits from time to time. The trustees
of the Canadian schools plan use updates to career average earnings to improve benefits periodically; they also employ ad hoc indexing of benefits. These types of “levers” are really useful because in any year the career average earnings base is not updated or indexing is not granted, it doesn’t feel like a take away. It just wasn’t granted.
I think governments should resist the urge to legislate in this area. Dictating the types of
adjustments that can be made, and when, is counterproductive. Trustees need discretionary space to operate effectively. Dictating the order of benefit reductions and restorations poses other problems, as can be seen by what happened under certain aspects of New Brunswick’s shared risk plan legislation.
In my respectful view, it is better to leave the reduction decisions to fiduciary responsibility so fiduciaries can appropriately respond to particular facts and
circumstances, even if most fiduciaries might prefer the safe harbour that a prescription or tick box approach might provide.
Legislation should also leave some fiduciary scope for making ad hoc adjustments
to other ancillaries.
Carefully Consider Critical Plan Impacts
Trustees should identify mission-critical impacts, and devote meeting time to deal with them as they come up. These impacts might relate to experience that differs from assumed rates of return, interest rates, or mortality; or to state-of-the-industry developments or trends relating to the nature, health, innovation and outlook for the industry that the participants work in; or to specific material incidents, like employer
withdrawal. In connection with some of these, they may even want to develop written policies to guide and memorialize decision-making.
Just take the example of employer withdrawal. Every employer withdrawal or proposed withdrawal is brought to the attention of the board of trustees of the Canadian schools plan for consideration. If an employer withdraws while the Canadian schools plan is in deficit, the employer is asked to remit or make provision to pay for the shortfall. If it will not, or if it demonstrates it cannot, the trustees make a fiduciary decision based on all relevant facts and circumstances to either cut benefits associated with that employer’s
participation or not.
The trustees identified this as a critical issue and devoted meeting time to develop a list of factors to ensure thoughtful consideration of employer withdrawals. They did this without a live case in front of them so they wouldn’t need to make one-minute to-midnight, gun-to-the-head decisions when a withdrawal did happen.
Some of those factors include the size of the employer and its materiality to the plan
as a whole; whether terminated participants intend to transfer commuted values out of the plan or not; how many new employers have joined the plan and other legal, economic and demographic factors. The trustees also use experience with specific employer withdrawals as a learning opportunity that sometimes results in new factors being added to the list.
In the case of any employer withdrawal during a period when the plan is in a deficit funding position, the trustees of the Canadian plan have a policy bias to provide the accrued benefit, but they also have full fiduciary discretion – from doing nothing, to
partially or fully reducing benefits associated with the
There have been three employer withdrawals during periods of deficit in the recent past in the Canadian schools plan in circumstances in which each of the withdrawing employers provided evidence they were unable to make additional contributions. After careful deliberation, the trustees recommended no adjustments, generally on the basis that the plan could sustain the termination without harm to other stakeholders – i.e., plan participants and participating employers. No such option is available to trustees of the US schools plan.
Effective Plan Communications
Plan communications are another key that cannot
Almost everything from Summary Plan Descriptions (SPDs), newsletters, internet portals to benefit statements.
The trustees of the Canadian schools plan make use of respected communication experts, lawyers and actuarial consultants to help them make legal and practical disclosures that promote the plan’s flexibility and responsiveness to circumstance. The communications
focus on the goal – predictable lifetime income and the positive benefits of the conditional nature of the promise.
One thing I have learned from experience is that communicating the true nature of a contingent plan with a positive attitude, rather than an apologetic one, does not solve all problems or fears, but it does seem to have a measurable and positive impact on plan engagement, understanding and appreciation.
Based on survey results, in the case of the Canadian schools plan, positive, repetitive and simple messages that avoid pension jargon and emphasize the value proposition seem to have worked very well amongst a broad range of participants, including those who are
years away from retirement.
The Tale of Two Plans
The biggest surprise for me in this tale of two plans was the positive reaction that arose out of what was effectively a benefit reduction. I would never have expected increased engagement and satisfaction withthe plan to be a result. It has really been remarkable
how “tuned-in” and pleased plan participants seem to be with the Canadian plan.
Prior to the reductions, it seemed like quiet indifference prevailed.
Following the reductions, there were measurable increases in plan engagement, not to mention unsolicited compliments. I attribute that mainly to the governance factors set out above – and the thoughtful way in which the trustees communicate on an ongoing basis. Comparisons with the sad fate of the US plan cannot be ignored, but at the same
time, much of the increases in plan engagement and satisfaction had surfaced prior to news of the US plan converting to a DC model.
I should also mention that the 2015 cut is the only cut made to the Canadian plan that I am aware of. The trustees routinely implement improvements, every three
years or so. But these generally relate to upgrading the career average base for the core benefit or indexing pensions already being paid, as described above.
I know that for each trustee of the Canadian plan, the decision to reduce accrued benefits was an extremely personal and difficult decision. They exercised extreme caution; but they had the force of character to take decisive and timely action. I feel certain that for every one of them that period in 2015
“was the best of times, it was the worst of times, it was
the age of wisdom … the spring of hope … the winter
Ultimately, it turned out to be faith in the logic of the sustainability of the contingent plan
design that prevailed. A sustainable way to deliver lifetime retirement income security that is predictable, cost efficient and cost certain.