Retiring a Defined Benefit Pension Plan

Data from the U.S. Department of Labor's Bureau of Labor Statistics (BLS) reveals that participation levels in DB plans and IAPs have moved in markedly different directions during the past 15 years. In 1991, 39% of all full-time workers participated in employer-provided DB plans, and the same number of workers participated in employer-provided IAPs. By 2003, only 24% of full-time workers participated in employer-provided DB plans, representing a nearly 40% decline. In stark contrast, participation by full-time employees in employer-provided IAPs had increased to 48% by 2003 ' that is, twice the participation level seen in DB plans at the same time.

Overview of the DB Dilemma

A principal reason for the shift away from DB plans toward IAPs was summarized by IBM officials when that company announced that it would freeze the accrual of additional benefits under its domestic DB plans at the end of 2007 and redirect its retirement contributions to the company's IAPs. IBM's DB plans constituted a very unpredictable element of the company's cost structure; an element that IBM's competitors did not have. The Employee Benefit Research Institute (EBRI, www.ebri.org), based in Washing-ton, DC, has stated that, in 2004, only 12% of private sector employees participated in both a DB plan and an IAP, down from 16% in 1984. The EBRI has identified several factors resulting in this decline, including: 1) cost; 2) unpredictability of cost; and 3) difficulty in communicating to current and prospective employees a working understanding of DB plans.

Retirement Costs

A study published in February 2006 by the U.S. Government Accountability Office (GAO) reported that private employers' average real cost of employee benefits increased by 18% between 1991 and 2005 (compared with a real cost increase of 10% in wages). The study also found that the real cost of wages remained relatively flat between 2002 and 2005, while the cost of benefits continued to rise. The rise in benefit expense between 1991 and 2005, according to the GAO report, was due in part to a 47% increase in retirement costs. A closer examination of these numbers reveals that more than half of this increase in retirement costs came in the final 3 years of the review period (2003-2005).

The BLS has offered an explanation of this dramatic increase in retirement costs. It notes that DB plans, like other investors in the stock market, enjoyed favorable asset performance in the 1990s. The return on assets during this period was sufficient to fund the liabilities of many DB plans without the need for employer contributions. The adverse change in overall market performance that began in 2000, coupled with the use of IRS-mandated methods for computing DB plan benefit liabilities, moved many DB plans from over-funded to under-funded positions. Consequently, employers who had for 1 or more years enjoyed a funding 'holiday' suddenly found themselves obliged to make contributions, frequently in substantial amounts. Employers who failed to do so could face additional funding charges and enforcement action by the IRS and PBGC.

The GAO study noted that a 'panel of experts from a variety of backgrounds agreed that rising benefit costs are forcing private employers and their employees to make increasingly difficult tradeoffs between wages and benefits.' It should be noted that the rise in benefit costs is prompting reconsideration of DB plans by financially sound companies as well as employers in more precarious fiscal situations. A 2006 analysis by Mercer of a selection of S&P 500 companies that had announced their intention to freeze their DB plans found that these companies had a slightly better credit rating than the median S&P company.

The Unpredictability of Cost

A major risk concern with DB plans arises from the recent (and anticipated future) volatility in annual contributions required from employers. Recent volatility flows from the actuarial methods and IRS rules used to compute the present value of DB plan assets and liabilities. Actuarial smoothing techniques that average asset values over several years worked to employers' advantage in the immediate aftermath of the 2000 economic downturn, but smoothing has also slowed the recognition of more recent improvements in investment returns. On the liability side, the rate of investment return assumed for valuing benefits under IRS rules has lagged behind the actual return rate for many plans, effectively overestimating the present liability for funding purposes. As a result of these methods and rules, an employer's DB plan can move in a single year from a funded to an under-funded position (or from an under-funded to a more under-funded position), notwithstanding an increase in asset value during that period.

From a budgeting standpoint, the unpredictability of return on assets from investment, and uncertainty regarding the applicable interest rates to be established by the IRS for calculating benefit liabilities, render it virtually impossible at the commencement of a fiscal cycle to project what a DB plan's funded status will be at the end of that cycle, and consequently, how much the employer may be required to contribute to the plan for that year. To further sour employers' taste for DB plans, legislative and accounting changes presently under consideration by Congress and the Financial Accounting Standards Board raise the specter of increased funding obligations.

The Employee Benefit with Little or No Employer Benefit

Another significant consideration leading employers away from DB plans is the limited utility of these plans as a means of attracting and retaining new employees in today's marketplace. Several factors produce this limitation. First, it is difficult to explain to a current or prospective employee precisely how much he or she will receive from a DB plan. The calculations are complex, and benefit estimates often must contain a dizzying number of projections, qualified statements and disclaimers. In contrast, IAPs ' which bear many resemblances to a standard savings account ' are relatively easy to explain and understand. The complexity of DB plans is aggravated by their relative scarcity in the marketplace. As already noted, only one-quarter of today's full-time workers participate in a DB plan. As such, it is likely that many prospective employees will be unfamiliar with DB plans, and therefore less able to determine their value as a component of compensation. Consequently, all other things being equal, an employer with a DB plan rather than an IAP could find itself losing a prospective (or even current) employee to a rival with an employee-funded 401(k) plan, even though the DB plan provides a benefit at no cost to the employee.

A second factor that limits the utility of DB plans as a means of attracting and retaining new employees flows from the fact that these plans provide the greatest benefit to older employees and employees with the longest period of service. As a result, the program is often regarded as a 'silent' or forgotten element of employee compensation for younger and short-service employees. By contrast, IAPs typically provide employees with an opportunity to accrue benefits without regard to age or service seniority.

Freezing vs. Terminating

An employer that decides to shift its retirement program design from a DB to an IAP model must first decide whether to freeze or terminate the DB plan. The decision is generally dictated by the funded status of the DB plan. A 'freeze' refers to an amendment to the DB plan that ceases the accrual of additional benefits under the plan (not the reduction of existing benefits), and/or prevents the admission of additional employees into the DB plan. A termination refers to the process of dissolving the plan (and its associated trust) and paying all accrued benefits. While under-funded plans cannot be terminated without PBGC approval, they can be frozen.

Freezing an under-funded DB plan does not eliminate the employer's obligation to make contributions to the plan. In addition, it does not necessarily freeze the amount of the plan's benefit liabilities. This is so because DB plan liabilities remain subject to change so long as they remain unpaid. The dynamic nature of DB plan liabilities is due primarily to the actuarial adjustments required as plan participants age (eg, to reflect the diminishing period preceding normal retirement age during which assets can appreciate in value), and to changes in the applicable interest rates used for determining the present value of the benefit liabilities. That said, employers typically freeze DB plans with the expectation that a plan's investment gains will exceed any increase in benefit liabilities, so that the plan can eventually become fully funded and then be terminated.

The Need to Replace a Frozen or Terminated Plan

In general, private sector employers are not required to provide retirement benefits to their employees. However, employers that elect to provide these benefits in the form of tax-qualified retirement plans must ensure that these plans satisfy the coverage, participation, non-discrimination and other requirements imposed by the federal tax code on such plans. If an employer uses a DB plan as part of a retirement program composed of more than one tax-qualified plan, then a freeze or termination of the DB plan may upset the tax-qualified status of the remaining plans ' unless a replacement plan (or additional benefit under an existing plan) is provided.

For example, assume that an em-ployer provides a DB plan for its non-highly compensated employees, and a tax-qualified IAP with an employer contribution for its highly compensated employees. If the employer froze or terminated the DB plan, it would not be able to make any additional contributions to the IAP without causing the IAP to lose its tax-qualified status. This result could be avoided if the employer created a replacement plan, such as an IAP, for the non-highly compensated employees (or allowed these em-ployees to participate in the existing IAP). In that event, the employer could continue to make contributions to the IAP for highly compensated employees so long as contributions at comparable rates and on comparable terms were made to the replacement plan (or replacement feature within the existing plan) for non-highly compensated employees.

It is important to note, however, that if the frozen DB plan is under-funded, then the employer would need to make contributions to both the DB plan and the replacement plan (or feature) ' with DB contributions computed to meet the minimum funding requirements for under-funded plans, and IAP contributions computed to meet the tax code's non-discrimination requirements. Thus, freezing an under-funded DB plan may actually result in an increase in retirement costs (at least during the short term) if the employer wants to provide additional contributions or accruals under another tax-qualified retirement plan while it simultaneously attends to the DB plan's under-funded status.

Law Firm Concerns

Law firms experience the DB dilemma in several atypical ways. First, the volatility of funding costs can translate directly into reductions in the amount of income distributable to partners from year-to-year. Second, partner turnover raises issues of fairness in allocating the cost of funding liabilities that may predate a partner's admission or become due after a partner's withdrawal. Why, a partner elected in 2006 might ask, should I be required to fund a liability accrued during the 1990s, when a recently retired partner who enjoyed the benefit of that accrual without a corresponding contribution obligation, does not?

Third, if a firm uses a DB plan for non-lawyers or non-partners as part of a multiple plan retirement program that also includes one or more tax-qualified plans exclusively for partners, then it may not be possible to freeze or terminate the DB plan un-less a replacement plan (or replacement benefit feature within an existing plan) is established. As noted above, the use of a replacement plan (or benefit feature) may increase the cost of retirement benefits if the frozen DB plan is under-funded.

Conclusion

As law firms and other employers examine their options for getting rid of DB plans (or, more precisely, the economic risks they represent), they must be careful to consider the costs associated with this course. In particular, if a replacement plan or feature must be established and funded while a frozen under-funded DB plan is prepared for termination, then an employer's retirement expenditures will increase ' at least temporarily. Nevertheless, many firms will doubtless choose this path in the hope and expectation that short-term economic inconvenience will be rewarded over the long term by greater self-determination regarding retirement costs.

Robert D. Webb, Esq., manages the Tax Department of Nutter McClennen & Fish, LLP, in Boston. As a member of the firm's Employment, Labor and Benefits practice group, he also maintains an active practice in the areas of employee benefits, ERISA and executive compensation law. He can be reached at [email protected], or 617-439-2552.

According to a 2005 report of the Pension Benefit Guaranty Corporation (PBGC, the agency that administers the federal insurance program for DB plans), more than 2700 DB plans ' nearly 10% of all PBGC-insured plans ' were frozen as of 2003, and more than 165,000 DB plans were terminated between 1975 and 2004.

This same period has seen a rise in popularity of defined contribution (DC) and other individual account plans (IAPs), such as 401(k) plans. This article explores the thinking that has led many employers to freeze and/or terminate their DB plans as a means of managing the risk/reward attributes of their tax-qualified deferred compensation programs, as well as some special considerations for law firm financial managers.

The Trend

Data from the U.S. Department of Labor's Bureau of Labor Statistics (BLS) reveals that participation levels in DB plans and IAPs have moved in markedly different directions during the past 15 years. In 1991, 39% of all full-time workers participated in employer-provided DB plans, and the same number of workers participated in employer-provided IAPs. By 2003, only 24% of full-time workers participated in employer-provided DB plans, representing a nearly 40% decline. In stark contrast, participation by full-time employees in employer-provided IAPs had increased to 48% by 2003 ' that is, twice the participation level seen in DB plans at the same time.

Overview of the DB Dilemma

A principal reason for the shift away from DB plans toward IAPs was summarized by IBM officials when that company announced that it would freeze the accrual of additional benefits under its domestic DB plans at the end of 2007 and redirect its retirement contributions to the company's IAPs. IBM's DB plans constituted a very unpredictable element of the company's cost structure; an element that IBM's competitors did not have. The Employee Benefit Research Institute (EBRI, www.ebri.org), based in Washing-ton, DC, has stated that, in 2004, only 12% of private sector employees participated in both a DB plan and an IAP, down from 16% in 1984. The EBRI has identified several factors resulting in this decline, including: 1) cost; 2) unpredictability of cost; and 3) difficulty in communicating to current and prospective employees a working understanding of DB plans.

Retirement Costs

A study published in February 2006 by the U.S. Government Accountability Office (GAO) reported that private employers' average real cost of employee benefits increased by 18% between 1991 and 2005 (compared with a real cost increase of 10% in wages). The study also found that the real cost of wages remained relatively flat between 2002 and 2005, while the cost of benefits continued to rise. The rise in benefit expense between 1991 and 2005, according to the GAO report, was due in part to a 47% increase in retirement costs. A closer examination of these numbers reveals that more than half of this increase in retirement costs came in the final 3 years of the review period (2003-2005).

The BLS has offered an explanation of this dramatic increase in retirement costs. It notes that DB plans, like other investors in the stock market, enjoyed favorable asset performance in the 1990s. The return on assets during this period was sufficient to fund the liabilities of many DB plans without the need for employer contributions. The adverse change in overall market performance that began in 2000, coupled with the use of IRS-mandated methods for computing DB plan benefit liabilities, moved many DB plans from over-funded to under-funded positions. Consequently, employers who had for 1 or more years enjoyed a funding 'holiday' suddenly found themselves obliged to make contributions, frequently in substantial amounts. Employers who failed to do so could face additional funding charges and enforcement action by the IRS and PBGC.

The GAO study noted that a 'panel of experts from a variety of backgrounds agreed that rising benefit costs are forcing private employers and their employees to make increasingly difficult tradeoffs between wages and benefits.' It should be noted that the rise in benefit costs is prompting reconsideration of DB plans by financially sound companies as well as employers in more precarious fiscal situations. A 2006 analysis by Mercer of a selection of S&P 500 companies that had announced their intention to freeze their DB plans found that these companies had a slightly better credit rating than the median S&P company.

The Unpredictability of Cost

A major risk concern with DB plans arises from the recent (and anticipated future) volatility in annual contributions required from employers. Recent volatility flows from the actuarial methods and IRS rules used to compute the present value of DB plan assets and liabilities. Actuarial smoothing techniques that average asset values over several years worked to employers' advantage in the immediate aftermath of the 2000 economic downturn, but smoothing has also slowed the recognition of more recent improvements in investment returns. On the liability side, the rate of investment return assumed for valuing benefits under IRS rules has lagged behind the actual return rate for many plans, effectively overestimating the present liability for funding purposes. As a result of these methods and rules, an employer's DB plan can move in a single year from a funded to an under-funded position (or from an under-funded to a more under-funded position), notwithstanding an increase in asset value during that period.

From a budgeting standpoint, the unpredictability of return on assets from investment, and uncertainty regarding the applicable interest rates to be established by the IRS for calculating benefit liabilities, render it virtually impossible at the commencement of a fiscal cycle to project what a DB plan's funded status will be at the end of that cycle, and consequently, how much the employer may be required to contribute to the plan for that year. To further sour employers' taste for DB plans, legislative and accounting changes presently under consideration by Congress and the Financial Accounting Standards Board raise the specter of increased funding obligations.

The Employee Benefit with Little or No Employer Benefit

Another significant consideration leading employers away from DB plans is the limited utility of these plans as a means of attracting and retaining new employees in today's marketplace. Several factors produce this limitation. First, it is difficult to explain to a current or prospective employee precisely how much he or she will receive from a DB plan. The calculations are complex, and benefit estimates often must contain a dizzying number of projections, qualified statements and disclaimers. In contrast, IAPs ' which bear many resemblances to a standard savings account ' are relatively easy to explain and understand. The complexity of DB plans is aggravated by their relative scarcity in the marketplace. As already noted, only one-quarter of today's full-time workers participate in a DB plan. As such, it is likely that many prospective employees will be unfamiliar with DB plans, and therefore less able to determine their value as a component of compensation. Consequently, all other things being equal, an employer with a DB plan rather than an IAP could find itself losing a prospective (or even current) employee to a rival with an employee-funded 401(k) plan, even though the DB plan provides a benefit at no cost to the employee.

A second factor that limits the utility of DB plans as a means of attracting and retaining new employees flows from the fact that these plans provide the greatest benefit to older employees and employees with the longest period of service. As a result, the program is often regarded as a 'silent' or forgotten element of employee compensation for younger and short-service employees. By contrast, IAPs typically provide employees with an opportunity to accrue benefits without regard to age or service seniority.

Freezing vs. Terminating

An employer that decides to shift its retirement program design from a DB to an IAP model must first decide whether to freeze or terminate the DB plan. The decision is generally dictated by the funded status of the DB plan. A 'freeze' refers to an amendment to the DB plan that ceases the accrual of additional benefits under the plan (not the reduction of existing benefits), and/or prevents the admission of additional employees into the DB plan. A termination refers to the process of dissolving the plan (and its associated trust) and paying all accrued benefits. While under-funded plans cannot be terminated without PBGC approval, they can be frozen.

Freezing an under-funded DB plan does not eliminate the employer's obligation to make contributions to the plan. In addition, it does not necessarily freeze the amount of the plan's benefit liabilities. This is so because DB plan liabilities remain subject to change so long as they remain unpaid. The dynamic nature of DB plan liabilities is due primarily to the actuarial adjustments required as plan participants age (eg, to reflect the diminishing period preceding normal retirement age during which assets can appreciate in value), and to changes in the applicable interest rates used for determining the present value of the benefit liabilities. That said, employers typically freeze DB plans with the expectation that a plan's investment gains will exceed any increase in benefit liabilities, so that the plan can eventually become fully funded and then be terminated.

The Need to Replace a Frozen or Terminated Plan

In general, private sector employers are not required to provide retirement benefits to their employees. However, employers that elect to provide these benefits in the form of tax-qualified retirement plans must ensure that these plans satisfy the coverage, participation, non-discrimination and other requirements imposed by the federal tax code on such plans. If an employer uses a DB plan as part of a retirement program composed of more than one tax-qualified plan, then a freeze or termination of the DB plan may upset the tax-qualified status of the remaining plans ' unless a replacement plan (or additional benefit under an existing plan) is provided.

For example, assume that an em-ployer provides a DB plan for its non-highly compensated employees, and a tax-qualified IAP with an employer contribution for its highly compensated employees. If the employer froze or terminated the DB plan, it would not be able to make any additional contributions to the IAP without causing the IAP to lose its tax-qualified status. This result could be avoided if the employer created a replacement plan, such as an IAP, for the non-highly compensated employees (or allowed these em-ployees to participate in the existing IAP). In that event, the employer could continue to make contributions to the IAP for highly compensated employees so long as contributions at comparable rates and on comparable terms were made to the replacement plan (or replacement feature within the existing plan) for non-highly compensated employees.

It is important to note, however, that if the frozen DB plan is under-funded, then the employer would need to make contributions to both the DB plan and the replacement plan (or feature) ' with DB contributions computed to meet the minimum funding requirements for under-funded plans, and IAP contributions computed to meet the tax code's non-discrimination requirements. Thus, freezing an under-funded DB plan may actually result in an increase in retirement costs (at least during the short term) if the employer wants to provide additional contributions or accruals under another tax-qualified retirement plan while it simultaneously attends to the DB plan's under-funded status.

Law Firm Concerns

Law firms experience the DB dilemma in several atypical ways. First, the volatility of funding costs can translate directly into reductions in the amount of income distributable to partners from year-to-year. Second, partner turnover raises issues of fairness in allocating the cost of funding liabilities that may predate a partner's admission or become due after a partner's withdrawal. Why, a partner elected in 2006 might ask, should I be required to fund a liability accrued during the 1990s, when a recently retired partner who enjoyed the benefit of that accrual without a corresponding contribution obligation, does not?

Third, if a firm uses a DB plan for non-lawyers or non-partners as part of a multiple plan retirement program that also includes one or more tax-qualified plans exclusively for partners, then it may not be possible to freeze or terminate the DB plan un-less a replacement plan (or replacement benefit feature within an existing plan) is established. As noted above, the use of a replacement plan (or benefit feature) may increase the cost of retirement benefits if the frozen DB plan is under-funded.

Conclusion

As law firms and other employers examine their options for getting rid of DB plans (or, more precisely, the economic risks they represent), they must be careful to consider the costs associated with this course. In particular, if a replacement plan or feature must be established and funded while a frozen under-funded DB plan is prepared for termination, then an employer's retirement expenditures will increase ' at least temporarily. Nevertheless, many firms will doubtless choose this path in the hope and expectation that short-term economic inconvenience will be rewarded over the long term by greater self-determination regarding retirement costs.

Robert D. Webb, Esq., manages the Tax Department of Nutter McClennen & Fish, LLP, in Boston. As a member of the firm's Employment, Labor and Benefits practice group, he also maintains an active practice in the areas of employee benefits, ERISA and executive compensation law. He can be reached at [email protected], or 617-439-2552.