• Retirement Plans
• Executive Compensation
• Health & Wellness
• Instant Insight
• Insight Subscription

Email Print

Retirement income – fundamental issues

Oct 09, 2009

This article begins a series on retirement income. A number of issues are covered by the phrase "retirement income" – annuities vs. lump sums, defined benefit (DB) vs. defined contribution plans (DC), adequacy and "leakage" all come to mind. Overarching all these issues is the question: will American workers have enough money for their retirement years? And in that regard, a central issue is – what is the role of the individual, the employer and the government in making sure they do?

We want to review each of these issues, and although some of what we will cover will be policy-in-the-abstract, we will also review specific proposals and emerging initiatives with respect to each of them. Our goal is to describe the basic policies that inform critical design decisions – e.g., whether to amend or freeze a DB plan, what sort of matching strategy to pursue in a DC plan, whether to consider a cash balance plan – and how those policies may be changing.

In this article, we review what we consider to be the retirement income "basics" in the United States: (1) competing theories of adequacy (that is, what should the retirement income target be?); (2) key priorities from both the participant's and the sponsor's point of view; and (3) the role of the employer and the government (via Social Security and Medicare) in providing retirement income.

What is an adequate retirement income?

Sponsors and benefits design consultants have, when considering retirement income adequacy, "traditionally" begun with the concept of "replacement income." That is, the retirement income target is calculated as a percentage – typically 70% or more – of current income. This target might typically be described as: "In our view, when an employee retires, he or she will need 70% or 80% or 90% of current income." Thus, under this approach, retirement income (and retirement income adequacy) is strictly a function of current income. Replacement income theoreticians begin with a concept of replacement of 100% of current income and then discount (minimally) for, e.g., savings and work-related expenses that no longer apply in retirement.

The literature on replacement income is vast, and we are not going to review it here. We do want to point out, however, some (widely recognized) problems with the implicit premises of this model.

First, the income replacement model assumes that when saving for retirement the participant/employee has an almost inflexible preference for sustaining his or her current situation and assumes no willingness or ability to mitigate costs by, e.g., moving to a less expensive community. Thus, under this approach, adequacy – which might be taken as the minimum necessary for a dignified retirement – is "defined up" to mean the maintenance of the individual's current situation.

Second, the replacement income concept does not reflect the actual spending patterns of individuals in retirement. There is data, with respect to higher paid individuals at least, indicating that spending early in retirement is greater than later. The intuition is that retirees do "all the things they always wanted to do" in their 60s and early 70s, and then settle down to a more modest lifestyle. That pattern may also be interrupted (or in the worst case, terminated) by a major health event. Thus, retirement spending patterns are sometimes described as a "horseshoe" – elevated spending at the beginning of retirement and at the end of life, with more modest spending "in between."

Finally, discussion of retirement income targets and, particularly, replacement income is often idealized, as if saving for retirement involved no tradeoff against other needs or preferences.

With these problems in mind, there has emerged a significant alternative view of calculating retirement income adequacy. Its advocates generally find the notion of a simple income replacement ratio "retirement income/pre-retirement income" problematic. But if pushed, they would set the appropriate replacement target much lower than the traditional 70% - 90%, to perhaps as low as 40% of "pre-retirement" income.

From the individual's point of view

As an exercise, let's consider the issue of retirement income from the other end of the glass – that is, from the perspective of the individual. What are reasonable financial goals for and in retirement, from the individual's point of view? We offer the following as a more-or-less subjective, common-sense view of what is financially "necessary" for a "good" retirement.

Lifetime consumption. Providing for a baseline amount necessary, from the individual's point of view, in normal circumstances, for a "dignified retirement." While "dignity" here will be defined differently by different individuals, our assumption is that it is somewhat short of "maintaining the pre-retirement situation." Thus, this is a minimum amount, for normal times.

An emergency fund. Certain emergencies can trigger a need for more spending – a health crisis, divorce or the death of a spouse are often cited as examples. We would also include in this category "end of life financing," even though this is not necessarily an emergency expense.

Fun money. Providing for those pleasures deferred during the individual's working life time. Most often cited: travel.

Legacy. While different individuals in different circumstances will have different legacy preferences, for individuals this goal can often be very significant in retirement spending decisions. Often individuals may confront tradeoffs between a desire for a legacy and other goals.

Finally, we want to note that implicit in the retirement income policy project is a concept of individual autonomy that may not be shared by all individuals. In some cultures older parents simply live with their (income-producing) children.

The role of government policy

We will, in a subsequent article, be discussing specific retirement income policy initiatives, but as a general matter two elements of government policy affect almost every design decision made by plan sponsors – the tax incentives for retirement savings and Social Security and Medicare benefits.

Tax incentives

Tax Code section 401 provides a critical tax incentive for retirement savings. Whereas wages are both deductible to the employer and taxable to the employee when paid, employer contributions towards qualified retirement benefits are generally immediately deductible but not taxed until distributed. This is a valuable tax benefit that can only be "captured" by an employer providing a qualified retirement plan covering a broad cross section of employees. (A limited version of this tax incentive is available to individuals via IRAs.)

This tax incentive accomplishes two things: it mitigates the tax on savings and it requires employers to provide retirement benefits to lower paid employees who (because of their lower tax rate) find this tax incentive less compelling.

Social Security and Medicare

The Social Security system provides a benefit in the form of a life annuity beginning at "normal retirement age." That benefit replaces a higher percentage of the income of low-paid employees than of high-paid employees.

As of 2009, here's what Social Security will pay at normal retirement age:

90% of the first $744 of average monthly earnings, plus

32% of the amount between $744 and $4,483, plus

15% of everything over $4,483

That works out to a maximum benefit of $2,323 per month, payable at Social Security's "normal retirement age" (currently age 66, grading up to age 67 for those born in 1960 or later). For an individual who has made exactly the maximum taxable wage base for his or her entire career, that's 26% of current pay (the wage base is currently $106,800). For individuals making below the wage base, the percentage of pay replaced by Social Security goes up; for those making above the wage base, it goes down.

We should also note that a critical post-retirement expense will be health insurance. Currently, while employees under age 65 must provide for this themselves, after 65 Medicare will take care of much of this cost, and the value of that coverage should be added on to the income provided by Social Security.

The role of the employer

Against this background of individual employee retirement needs and preferences and government incentives and programs, sponsors must sort out what their role as a "benefit provider" should be.

While some companies may be able to work on a strict benefits = compensation model, for most companies benefits have a hybridized quality – they are part compensation and part something else. And often that "something else" has a 1 + 1 = 3 quality. For instance, it's possible that benefits may capture big-company buying power unavailable to individual employees. With respect to retirement savings, cheaper investment services are the most obvious example of this benefit. And, as noted, there are tax benefits that are only available from an employer.

Benefits are also a core element in building and driving corporate culture. The shift at some companies from DB to DC plans and the persistence of DB plans at other companies reflects this feature of retirement benefits.

Reconciling different objectives

Finally, in designing a retirement plan sponsors must consider that different employees will have different objectives. This is most obvious with respect to age differences, where older employees will typically have a much greater preference for retirement savings than younger employees. But also, for instance, as we discussed, employees of the same age from different cultures may have different preferences. Thus, outside of the idealized cafeteria-plan-in-the-sky, every benefits program involves a compromise between the different goals of different employees.

Keeping in mind that a commitment to a retirement program will defer compensation for at least some employees that have a perfectly legitimate preference for cash now, it's safe to say that most sponsors will want to disregard employees' "fun money" and "legacy" goals. Whatever the merits of these goals, it is hard for an employer to justify forcing all employees (many of whom will have more pressing needs) to pursue them. There are obvious exceptions to this rule – a low-cost (to the employer) 401(k) plan can further the objectives of some employees while letting others opt out.

The design equation

With all of these factors in mind, we can sketch out a sort of retirement income equation:

A sponsor's retirement plan should:

DB vs. DC, mandatory vs. voluntary

Briefly, DB plans (arguably) are more efficient and fit neatly into an income replacement equation. DC plans are more transparent – benefits = funding. Traditionally DB plans have provided greater benefits for older employees, whereas DC plans have provided a level (and, again, more transparent) benefit. Thus, DB plans lend themselves to a more "paternal" corporate culture, whereas DC plans tend to work better where benefits are generally equated with compensation.

That said, design innovations – cash balance and target benefit plans – have blurred these distinctions. And either sort of plan can be designed to target agreed upon goals.

Voluntary participation is a key feature of 401(k) plans but also may be a factor in some DB plans. We would observe that it allows participants to self-finance more personal goals (fun money or a legacy) without imposing those goals on others. As a way of meeting an agreed upon "minimum," e.g., 40% replacement income, in our view, voluntary plans do present problems.

*     *     *

Inevitably, this article reads a little like Plan Design 101. We thought it would be useful to begin, however, with a review of the basics so that when we take up more problematic issues, like the role of annuities, we can work from a shared understanding. In our following articles we will cover: the axiomatic case for annuities and possible reasons why, as a practical matter, employees reject the annuity option; and policy initiatives.