S L P P N , A
RESEARCH
RETIREMENT
NEW BRUNSWICK’S NEW
SHARED RISK PENSION PLAN
By Alicia H. Munnell and Steven A. Sass*
I
* Alicia H. Munnell is director of the Center for Retirement
Research at Boston College (CRR) and the Peter F. Drucker
Professor of Management Sciences. Steven A. Sass is an as-
sociate director of the CRR. They would like to thank Keith
Ambachtsheer and W. Paul McCrossan for extremely useful
comments on earlier drafts and Jean-Pierre Aubry for insight on
public plans.
Employer defined benefit pension plans have long
advance. The Netherlands certainly oers one model
been an important component of the U.S. retirement
of risk sharing; this brief discusses an adaptation of
system. Although these plans are disappearing in the
the Dutch approach closer to home – namely New
private sector – replaced by (k)s – they remain the
Brunswick’s Shared Risk Pension Plan introduced in
prevalent retirement plan arrangement in the public
May .
sector. But these public sector defined benefit plans
The discussion proceeds as follows. The first sec-
are currently under financial pressure, as two finan-
tion reviews the problem of risk in employer defined
cial crises since the turn of the century have caused li-
benefit plans. The second section describes New
abilities to soar and assets to plummet. The response
Brunswick’s response – the Shared Risk design and
so far among state and local plan sponsors has been
the regulatory framework for supervising such plans.
to suspend or eliminate cost-of-living adjustments,
The third section discusses the response of union rep-
cut back sharply on benefits for new employees, and
resentatives of workers covered by the new program.
raise employee contributions. Some states have also
The fourth section considers what lessons U.S. plans
introduced a defined contribution component. While
can draw from the New Brunswick approach. The
the cutbacks have sharply reduced future costs, they
final section concludes that the Shared Risk approach
have been ad hoc and unexpected. The question is
is an important evolutionary step, and potentially an
whether a more orderly and predictable way can be
attractive alternative to the traditional defined benefit
devised to share risks, and perhaps head o trouble in
plan design.
LEARN MORE
Search for other publications on this topic at:
crr.bc.edu
Center for Retirement Research
T R D B P
Defined benefit plans promise workers a fixed pen-
sion payment that lasts as long as they live, providing
retirees a valuable source of security when their work-
ing days are done. The cost of these plans, however,
has risen sharply. One reason is that retirees now live
longer: U.S. workers retiring in can expect to
live about four years longer than workers retiring in
.
Most plans also have generous early retire-
ment provisions, with less-than-actuarial reductions
for the increased length of time that early retirees
collect a pension.
Defined benefit plans have also become increas-
ingly risky as equity holdings have risen and the
maturation process has increased the number of
retirees and older long-service workers relative to the
plan’s funding base. Risky means outcomes can be
good as well as bad. In the s, when the stock
market boomed, the value of pension assets generally
rose well above the value of plan obligations. As a
result, many employers, in both the private and public
sector enjoyed “contribution holidays” and increased
benefits. In the s, when financial markets
tanked, significant underfunding suddenly became
the norm. As a result, employers had to sharply
increase contributions(see Figure ).
F . S L P C
P O S R, -
2.9%
3.0%
3.4%
4.2%
3.8%
3.7%
3.9%
4.2%
4.5%
4.6%
0%
1%
2%
3%
4%
5%
2001 2002 2003 2004 2005 2006 2007 2008 2009 2010
Source: Authors’ calculations from U.S. Census Bureau
(-a) and (-b).
In both Canada and the United States, govern-
ment regulations require private employers to
eliminate underfunding within a specified number of
years. As defined benefit risks increased over time,
governments sharply reduced that timeframe – in
Canada from years in the mid-s to five by the
end of the century. This demand for large sums of
cash, when cash is hard to come by, has further in-
creased the risks to sponsors of defined benefit plans.
Given the dicult financial conditions since the
turn of the century, defined benefit sponsors have
been hard pressed to quickly fill funding shortfalls.
The Canadian provincial governments, which set the
minimum funding rules for private defined benefit
plans, responded by relaxing or even exempting
employers from making “required” deficit-reduction
payments. As these make-shift measures dragged
on, the New Brunswick provincial government in
created a “Task Force on Protecting Pensions,”
and added the province’s own public sector plans
to the agenda one year later. The members of the
Task Force then asked for and received a mandate to
“fix” the problem, not just issue a report. Their “fix,”
designed to make both private and public employer
plans “secure, sustainable and aordable for both
current and future generations,” was the Shared Risk
Pension Plan, announced in May .
S R
New Brunswick’s Shared Risk program has three key
elements: ) a new design that splits plan benefits
into highly secure “base” benefits and moderately
secure “ancillary” benefits; ) protocols that require
pre-determined actions to change future benefits,
contributions, and asset allocations in response to
changes in the plan’s financial condition; and ) a
new risk management regulatory framework to keep
these plans on track. The “base” and “ancillary” ben-
efit design is based on the widely admired approach
developed in The Netherlands. The new regulatory
framework is largely based on Canada’s “stress-
test” methods for supervising banks and insurance
companies.
The key innovation is to combine these
elements into a coherent pension program.
N B’ S R P D
The Shared Risk plan guarantees base benefits, but
only grants ancillary benefits if allowed by the plan’s
financial condition. The funding program is then de-
signed to ensure that both base and ancillary benefits
will be paid with a high degree of likelihood. But the
plan sponsor also specifies protocols for responding
to changes in the plan’s financial condition: how to
Issue in Brief
increase contributions, change asset allocations, and
reduce benefits in response to funding deficits; and
how to reduce contributions, change asset allocations
and restore benefits, including restoring previous
benefit reductions, and grant ancillary benefits when
the plan’s financial condition improves.
When the new program was announced, four Ne
Brunswick defined benefit plans – three public sec-
tor plans and one private sector plan – declared that
they were adopting the Shared Risk design. These
plans introduced a new benefit formula that made a
significant portion of future benefits ancillary and de-
pendent on the plan’s financial condition. The plans,
which had initially based benefits on the employee’s
final salary, now adopted formulas that base benefits
on the employee’s much lower career average salary,
but provide much the same “final salary” pension by
indexing employee earnings to wage growth or infla-
tion. These indexing increments are ancillary and
granted only when the plan’s finances exceed speci-
fied benchmarks. For an example, see the Box for th
sequence of steps to be taken in the case of under- or
over-funding for one New Brunswick plan.
,
w
e
N B’ S R R
F
Existing regulations in both Canada and the United
States assess safety and soundness based on a plan’s
current funded ratio. A fully funded plan, with pen-
sion fund assets equal to the present value of plan
obligations, is only required to contribute amounts
needed to cover the additional benefits currently
earned by active workers. Underfunded plans must
eliminate shortfalls within a specified number of
years. This approach ignores how risky a plan’s assets
and obligations might be or how they might change
over time. It makes no dierence whether the plan’s
assets are invested in penny stocks or government
bonds, or whether its obligations could spike, say if
large numbers of workers suddenly claim sweetened
early retirement benefits.
New Brunswick’s regulatory program for Shared
Risk uses a “stress test” to assess the ability of Shared
Risk plans to pay promised benefits. To test the
plan’s ability to pay benefits, each year the plan actu-
ary must run at least , -year simulations using
H R S
This example gives the sequence of actions to be taken by the New Brunswick Hospitals’ plan (Canadian
Union of Public Employees Local ) in response to changes in its financial condition.
If the funded ratio falls below percent for If the funded ratio rises above percent, a
two years in a row or the plan fails to meet the portion of the surplus can be used as follows, if
risk management goals of a Shared Risk plan: the plan can still meet the risk management goals:
. Increase contributions up to percent of . Reverse previous deficit-recovery measures in
earnings, split evenly between workers and the following order:
the employer. a. Reverse any increase in contributions.
. Change the rule for calculating early retire- b. Reverse any reduction in base benefits.
ment benefits, for those not currently eligible c. Reverse any reduction in early retire-
for such benefits, to a full actuarial reduction. ment benefits.
. Reduce “base benefit” accrual rates for future . Index pensions and base benefit accruals up
service up to percent. to the full Consumer Price Index (CPI).
. Reduce base benefits for all members, includ- . Increase individual benefits, as needed, so
ing benefits based on past and future service, that all retirees receive a benefit based on final
in equal proportion until the plan meets the five-year average salary, indexed to the CPI.
risk management goals. . Provide lump-sum payments to oset past
shortfalls relative to a benefit based on final
five-year average salary, indexed to the CPI.
Source: Mann ().
Center for Retirement Research
reasonable estimates of relevant financial parameters.
The simulations must show that over this -year
horizon: ) base benefits will be paid in full at least
. percent of the time; and ) at least percent
of ancillary benefits will be paid, on average, over all
scenarios. Plans that fail this forward-looking test
must modify their investment, funding, or benefit
rules until they pass. This annual review is expected
to identify changes in the plan’s financial condition
much earlier than the traditional funded ratio ap-
proach and produce smoother responses to changing
conditions.
The Shared Risk regulatory program also includes
funded ratio yardsticks. However, the measure of
plan obligations used as the funded ratio denomina-
tor excludes ancillary benefits, such as indexing. The
Shared Risk program requires the actuary to project
the plan’s annual funded ratio over the next years.
For new plans, the projected ratios must never be less
than percent of plan obligations. In subsequent
years, the projected ratio must equal or exceed
percent of plan obligations at the end of the -year
planning horizon and must never fall below that level
two years in a row. These requirements, and the
mandatory stress test, eectively raise current fund-
ing targets for Shared Risk plans to about percent
of base benefit obligations.
U S S
R P
It is easy to see why employers would welcome New
Brunswick’s Shared Risk program – the new ap-
proach would allow them to better manage their
finances. Workers, however, were switching from a
defined benefit to a “target benefit” program. Nev-
ertheless, the new plans were adopted with union
support.
The unions embraced the shift to Shared Risk
plans because:
• As private sector workers were losing defined
benefit plan coverage, public employees feared
that taxpayers would target their plans.
• Their defined benefit plans were stressing the
finances of plan sponsors, raising the prospect o
significant benefit cuts.
• Young workers increasingly viewed rising pen-
sion contributions as funding the pensions of
older workers and retirees, not their own
benefits.
f
• The Shared Risk plan oered much greater
security than the likely alternative – a (k)-type
retirement savings plan.
The Government of New Brunswick has been
negotiating with public sector unions to move all of
its plans to the Shared Risk design, and many of these
unions have agreed to the change. Two municipali-
ties, Saint John and Fredericton, have also adopted
Shared Risk plans with union support. In addition,
one private sector plan has applied for registration
as a Shared Risk plan. And a half dozen others are
exploring conversion to the new design.
L U S
U.S. state and local governments have responded
to the large deficits in their pension programs in
dramatic and unpredictable ways. Some states have
cut or eliminated cost-of-living adjustments for cur-
rent as well as future retirees. Some have increased
mandatory employee contributions, sometimes on all
employees and sometimes only on new employees.
Many have sharply reduced future benefits – primar-
ily for new employees – by raising age and tenure
requirements, lengthening the average salary period,
and/or reducing pension benefit factors (see Figure
). In many cases, the cuts for new employees will
produce lower benefits than provided before the
financial crisis.
And just like the permanent benefit
F . P C M S
S-A P, T C
15
14
24
18
9
3
0
10
20
30
New employees
All employees
COLA
Contribution rate
Age/tenure
requirements
Average salary period
Benefit factor
No changes
Source: Munnell et al. ().
Issue in Brief
expansions that occurred during the good times of
the s, nearly all states made the cuts permanent
adjustments to their pension program.
New Brunswick’s Shared Risk program oers a
dierent approach. First, it makes responses to pen-
sion shortfalls far more predictable. The plan design
clearly spells out how the sponsor would respond,
and by sharing the burden among the employer (i.e.
the taxpayer), employees, and pensioners, it moder-
ates the burden borne by each. Moreover, ancillary
benefits not granted in bad years can be expected to
be fully restored in good years. In fact, pensioners
will receive checks in good years that will compensate
for COLAs missed in bad years. Second, the required
risk management tests function as an early warning
system, helping plans minimize the size of any need-
ed adjustments by heading o trouble in advance.
For U.S. state and local plans to adopt a risk shar-
ing approach, sponsors must develop specific rules
for adjusting benefits, contributions, and investment
allocations; these rules need to be fair to workers in
dierent cohorts and income groups; and they need
to be communicated eectively to plan participants.
Establishing such rules is a thorny and dicult
task, but is likely to produce a much more sensible
outcome than lurching toward generous benefit
expansions when times are good and dramatic benefit
reductions when times are bad.
C
New Brunswick’s Shared Risk program is a promis-
ing innovation. It makes changes in benefits and
contributions more orderly, moderate, predictable,
and reversible.
The New Brunswick program does not solve all of
the challenges facing defined benefit plans. A large
enough shock would no doubt also overwhelm the
adjustment rules and risk management procedures of
Shared Risk plans. Nevertheless, Shared Risk plans
could be expected to weather such shocks much better
than traditional defined benefit programs or individu-
als with a (k).
The New Brunswick Shared Risk approach pro-
vides an attractive model for governments, employ-
ees, and pensioners that find current procedures for
handling risk in defined benefit plans unworkable,
and for workers who might otherwise end up with a
(k).
Center for Retirement Research
E
U.S. Social Security Administration (). Public sector plans adopting the Shared Risk
design covered workers represented by “Certain
The three members of the Task Force were Susan Bargaining Employees” of New Brunswick Hospitals
Rowland (Chair), a pension attorney with extensive and the Canadian Union of Public Employees, New
experience in restructuring troubled plans; W. Paul Brunswick Hospitals. A private sector plan cover-
McCrossan, former head of the Canadian Institute ing union workers, the New Brunswick Pipe Trades
of Actuaries; and Pierre-Marcel Desjardins, a Ph.D. Pension Plan, also adopted the Shared Risk design.
economist active at the Canadian Institute for Re- The Task Force consulted with unions representing
search on Public Policy and Public Administration. the hospital workers when developing the Shared
The addition of public plans to the Task Force agenda Risk program, and these unions endorsed the shift
in was in part a response to downgrades of New to the Shared Risk design. The pension plan cover-
Brunswick government debt by bond rating agencies ing members of the New Brunswick legislature also
due to the Province’s mounting pension liabilities. adopted the Shared Risk design: the politicians sup-
See Government of New Brunswick (, ). porting the new design felt it important to accept the
same risks borne by other Shared Risk plan partici-
Government of New Brunswick (). pants, a decision supported by an all-party consensus.
The Task Force recommended the conversion of all
For a discussion of the Dutch models and related government plans to the Shared Risk design, and
“Defined Ambition” plan designs, see van Rieland the government also announced its desire to do just
and Ponds (); Ambachtsheer (); Kocken that. The Task Force report also called for changes
(); and Kortleve (). in public sector plans to reduce their costs, including
an increase in the retirement age and reductions in
If all these steps have been taken and the funding subsidized early retirement benefits, albeit introduced
ratio is greater than percent of plan obligations, over a -year period. These changes were accepted
the plan would establish a reserve to cover years’ in the public sector plans that adopted the Shared
contingent indexing; then reduce contributions by Risk program. See Government of New Brunswick
up to percent of earnings; then improve various ().
benefits, including early retirement benefits. The
plan document also suggests more permanent adjust- As the government of New Brunswick noted,
ments could be in order. “Many in the public complained openly about the
generosity of these schemes and the fact that they
Dutch plans strengthened their risk management have to pay extra taxes to cover pension deficits for
methods in the early s, adopting a Value-at-Risk benefits that they cannot aord for themselves.” See
yardstick with a one-year horizon and a confidence Government of New Brunswick ().
level of . percent (Kortleve, Mulder, and Pelsser
). Dutch plans, however, failed to accommodate The New Brunswick government declared its de-
the sharp financial shock that accompanied the Great fined benefit plans to be “no longer sustainable.” See
Recession and, like plans in New Brunswick, were Government of New Brunswick ().
forced to relax their “deficit recovery” rules (Kortleve
and Ponds ). The Canadian Oce of the Su- Government of New Brunswick ().
perintendent of Financial Institutions, which had no
regulatory authority over employer plans, in had Munnell et al. ().
recommended the use of stress tests to manage risks
in defined benefit pension plans. New Brunswick’s Of those plans that have increased employee
regulatory program for Shared Risk plans was the contributions since , only a few have stated that
first to require such tests, at least in North America; the increases are likely to be temporary. Of those
see Government of New Brunswick (, ); that have cut pension COLAs during this period, only
Oce of the Superintendent of Financial Institutions a handful explicitly link future COLAs to the plan’s
Canada (); and McCrossan (). financial condition. One state, Wisconsin, has always
linked its COLA to the rate of return on pension as-
sets.
Issue in Brief
R
Ambachtsheer, Keith. . The Pension Revolution: A
Oce of the Superintendent of Financial Institutions
Solution to the Pension Crisis. Hoboken, NJ: Wiley.
Canada. . Guideline: Stress Testing Guidelines
for Plans with Defined Benefit Provisions. Available
Government of New Brunswick. . Pension Reform
at: http://www.osfi-bsif.gc.ca/app/DocReposi-
Background Document. Available at: http://www.
tory//eng/pension/guidance/defined benefitst_e.
gnb.ca/content/gnb/en/corporate/promo/pen-
pdf.
sion.html.
U.S. Census Bureau. -a. State and Local
Government of New Brunswick. . Rebuilding New
Public-Employee Retirement Systems. Washington,
Brunswick: The Case for Pension Reform. Available
DC.
at: http://ppforum.ca/sites/default/files/NB_Pen-
sion%Reform%Case_EN.pdf.
U.S. Census Bureau. -b. State and Local
Government Finances. Washington, DC.
Kocken, Theo. . “Why the Design of Maturing
Defined benefit Plans Needs Rethinking.” Rotman
U.S. Social Security Administration. . Annual
International Journal of Pension Management ():
Report of the Board of Trustees of the Federal Old-
-.
Age and Survivors Insurance and Federal Disability
Insurance Trust Funds. Washington, DC.
Kortleve, Niels. . “The ‘Defined Ambition’ Pen-
sion Plan: A Dutch Interpretation.” Rotman Inter-
van Rieland, Bart and Eduard Ponds. . “Sharing
national Journal of Pension Management. (): -.
Risk: The Netherlands’ New Approach to Pen-
sions.” Issue Brief -. Chestnut Hill, MA: Center
Kortleve, Niels and Eduard Ponds. . “Dutch Pen-
for Retirement Research.
sion Funds in Underfunding: Solving Generation-
al Dilemmas.” Working Paper -. Chestnut
Hill, MA: Center for Retirement Research.
Kortleve, Niels, Wilfried Mulder, and Antoon Pelsser.
. “European Supervision of Pension Funds:
Purpose, Scope and Design.” Netspar Design
Papers . Tilburg, NL: Netspar.
Mann, Troy. . Personal Communication with
Mann, Director of Human Resources, Corporate
Services Section, Government of New Brunswick.
McCrossan, W. Paul. . “The Whys Behind The
What: New Brunswick Pension Reform Shared
Risk Plan.” Powerpoint delivered at Ryerson
University, November . Available at: http://www.
ryerson.ca/content/dam/clmr/publications/con-
ferences/conf_pensions_paul_mccrossan.pdf.
Munnell, Alicia H., Jean-Pierre Aubry, Anek Belbase,
and Josh Hurwitz. . “State and Local Pension
Costs: Pre-Crisis, Post-Crisis, and Post-Reform.”
State and Local Plans Issue in Brief . Chestnut
Hill, MA: Center for Retirement Research at Bos-
ton College.
A C
The Center for Retirement Research at Boston Col-
lege was established in through a grant from the
Social Security Administration. The Center’s mission
is to produce first-class research and educational tools
and forge a strong link between the academic com-
munity and decision-makers in the public and private
sectors around an issue of critical importance to the
nation’s future. To achieve this mission, the Center
sponsors a wide variety of research projects, transmit
new findings to a broad audience, trains new schol-
ars, and broadens access to valuable data sources.
Since its inception, the Center has established a repu-
tation as an authoritative source of information on all
major aspects of the retirement income debate.
s
A I
The Brookings Institution
Massachusetts Institute of Technology
Syracuse University
Urban Institute
C I
Center for Retirement Research
Boston College
Hovey House
Commonwealth Avenue
Chestnut Hill, MA -
Phone: () -
Fax: () -
E-mail: [email protected]
Website: http://crr.bc.edu
Center for Retirement Research
The Center for Retirement Research thanks AARP, Advisory Research, Inc. (an aliate of Piper Jaray
& Co.), Charles Schwab & Co. Inc., Citigroup, ClearPoint Credit Counseling Solutions, Fidelity & Guaranty
Life, Goldman Sachs, Mercer, National Council on Aging, National Reverse Mortgage Lenders Association,
Prudential Financial, State Street, TIAA-CREF Institute, T. Rowe Price, and USAA for support of
this project.
© , by Trustees of Boston College, Center for
Retirement Research. All rights reserved. Short sections of
text, not to exceed two paragraphs, may be quoted without
explicit permission provided that the authors are identified
and full credit, including copyright notice, is given to
Trustees of Boston College, Center for Retirement Research.
The research reported herein was supported by the Center’s
P
artnership Program. The findings and conclusions ex-
pressed are solely those of the authors and do not represent
the views or policy of the partners or the Center for Retire-
ment Research at Boston College.
