Are Defined Contribution Pension Plans Fit for Purpose in Retirement?

• Publication year: 2013

SEATTLE UNIVERSITY LAW REVIEWVolume 37, No. 2, Winter 2014

Are Defined Contribution Pension Plans Fit For Purpose In Retirement?

Jeremy R. Cooper(fn*)

I. INTRODUCTION

The Australian superannuation system(fn1) has primarily had a "lump sum" approach to retirement, as distinct from a "replacement rate" or lifetime income stream approach.(fn2) After all, superannuation in Australia is primarily a defined contribution (DC) system under which the retirement benefit payable on retirement is a factor purely of the employee's own contributions (and those made by employers on the employee's behalf) as increased by the net investment returns attributable to the employee, less tax. Like the United States, Australia is now realizing the limitations of a DC retirement system insofar as it relates the provision of reliable retirement income for a population with increasing life expectancy.

In Australia, the core DC retirement plan product is called an account-based pension. This is a bit of a misnomer because it is only a "pension" because the legislators decided to give it that name. It is really just a mutual fund account that allows retirees to draw out, usually via a regular payment, their retirement savings to live on. Withdrawals are tax-free. Like in the United States, there is a minimum amount that must be drawn out each year to prevent perpetual exclusion of the balance from the taxation system. Unlike the United Kingdom, there is no limit to the amount that can be drawn out and hence no guarantee that the money will last.

Unlike DC plans, defined benefit (DB) plans provide a benefit based typically on time served and a predetermined proportion of either career average or final salary.

Towers Watson's 2013 Global Pension Assets Study shows that DC assets represent about 45% of total pension assets globally but are growing at an annual rate of about 8%, compared to a 6.6% growth rate for DB assets.(fn3) In Australia, DC assets represent over 80% of pension assets.(fn4)

So what is wrong with DC retirement plans? First and principally, the problem is that DC plans are not designed to produce a particular level of income in retirement, but just a terminal capital sum the size of which is subject to a range of uncertainties. Second, DC plans expose members to a much higher degree of risk than DB plans. Lastly, their individualized, rather than pooled, design can lead to a range of inefficiencies. The result is that DC plan contributors often fall short of their financial goals in retirement. According to the Organization for Economic Co-operation and Development (OECD), workers contributing a total of 10% to DC pension plans for 40 years only have a 53% chance of targeting 70% of their final salary.(fn5) At a 5% contribution rate, only 14% of retirees reach that target.

DC plans often lack a clearly stated goal that allows contributors to measure their progress toward a secure retirement. There is a tendency for DC plans, particularly in compulsory systems, to set a goal of building up a lump sum. For most retirees, their lump sum is not inherently suited to ensuring financial security in retirement. Retirees find it difficult to estimate how much they can spend each year to fund a retirement of an indeterminate length. The Australian system makes sense of this difficulty by using averages: average length of retirement, average investment returns, and average rate of consumption. The trouble is that almost nobody has an "average" retirement. It is a highly individualized experience.

Retirees face financial risks that are difficult to manage without specialized and sophisticated products. Central to the idea of a DC retirement plan is the idea that asset allocation can address all of the needs of the retiree.(fn6) In reality, there are risks facing retirees that are very difficult to ameliorate without using products that are specifically designed to deal with them. Such specialized products include a guarantee to protect against loss of capital or income, an indexed income stream to deal expressly with inflation, and a lifetime annuity to deal with longevity risk. All of these products involve some form of absorption of risk by intermediaries external to the DC plan itself or by the plan sponsor using its own balance sheet.

DC plans rest on a theory of consumer sovereignty. Success in a DC plan depends on the contributor making informed choices and demonstrating rational behavior. In the real world, information asymmetry, issues with financial literacy, and innate behavioral biases, often make it hard for the consumer to achieve any degree of "sovereignty." The result is a system heavily reliant on (potentially conflicted) agents.

DC plans also tend be more expensive to run than DB plans. The individualization that is central to the operation of DC plans means that they cannot easily pool risk. DC plans, therefore, typically accept less risk and hence receive lower average returns, particularly after investment and administration expenses.(fn7)

Lastly, in Australia there is no insurance scheme or government guarantee protecting DC plan members in the event of failure, except in the case of fraud where the government has discretion to compensate a plan to restore loss if it is in the public interest.(fn8)

A successful and sustainable DC system must address the shortcomings of DC plans. A properly integrated retirement plan should seek to protect contributors from three key financial risks: inflation, deviation from expected outcome, and longevity. This Article considers the historical basis for the shift from DB to DC plans, the structural and practical shortcomings of DC plans, alternate pension models, and adjustments to existing retirement plan models that may offer a degree of protection to plan contributors.

II. HISTORY OF DEVELOPMENT OF DC PLANS

How did DC plans become such a large part of the global pension system? DB plans were once widespread; employers contracted to provide a set amount (the "defined benefit") for the duration of the employee's retirement. However, governments and corporations around the world realized that retirement is expensive and making concrete promises about paying people in retirement is even more expensive. There are countless examples where the corporate pension plan became the core business-or brought down the business altogether-due to problems with funding liabilities.(fn9) Many governments had the same experience, with Greece being the latest example. Systematically, it became obvious that taking on market, inflation, and longevity risk for people in retirement was simply too risky other than for specialist asset managers.

In the face of failing DB plans, DC seemed like the perfect solution. The risks inherent to retirement planning can be made to look like they are not even present in a DC plan. Because the typical DC plan does not aim to provide the retiree with a particular level of income, there is no target from which there can be a shortfall.(fn10)

DC plans also allowed providers to transfer longevity risk(fn11) to plan participants. While DB plans pool longevity risk, DC plans leave it to individual participants to deal with. This forces each contributor to self-fund his retirement income. Contributors can be thought of as small insurance companies taking on their own longevity risk without the capital or the skills to do so.

III. COMPARISON TO DB PLANS

Amid the global trend of transitioning from DB to DC plans, there are many vocal proponents of one system over the other. Proponents of each system base their conclusions on different measures of success. The reality is that each system has its own advantages and disadvantages. Keith Ambachtsheer, a leading pension expert from Canada, made the point that both structures are flawed, and in his view, there is a desperate need for a fresh design for approaches within the pension sector.(fn12) While the economy may not be able to afford the promises made through a DB system, neither can the members of a DC plan take any real comfort from the building of assets without a specific goal for those assets.

In considering whether or not DC plans are suitable for retirement income purposes, it is important to remember that both DB and DC systems are flawed. We should not be asking whether or not DC plans can replicate DB plan outcomes; rather, we should ask whether DC plans can deliver what their members require. That is, can DC plans deliver a reliable stream of income in retirement to meet the needs of entire cohorts of retirees?

Part of the problem with DC plans can be traced to the transition from DB to DC plans. The conversion from DB to DC required a calculation of a lump-sum amount that was the present-day equivalent of the future benefits that had already been accrued by the member. The method used to convert DB plan members to a DC plan has become the (default) objective of the DC plan. While the DB was focused on the benefit (the ultimate income payments), the DC metric now focuses on the dollar balance that is accrued. In general, DC plans focus on this target balance at retirement. Even in Australia, where the contribution rate is currently mandated as 9.25% of a salary, increasing to 12% by fiscal year 2022, the focus of the system is the balance at retirement, not what level of income the retiree is going to need in retirement or how much reliable annual income that balance can produce. The problem...