December 13, 2022


Regulatory Affairs Division
Office of the General Counsel
Pension Benefit Guaranty Corporation
445 12th Street, SW
Washington, DC 20024–2101

Re: Actuarial Assumptions for Determining an Employer’s Withdrawal Liability, RIN 1212–AB54

To Whom It May Concern:

This is the response of the U.S. Chamber of Commerce, the Association of Food and Dairy Retailers, Wholesalers, and Manufacturers and the Associated General Contractors of America (undersigned) to the Pension Benefit Guaranty Corporation’s (PBGC) Actuarial Assumptions for Determining an Employer’s Withdrawal Liability proposed regulation (Proposed Regulation). The Employee Retirement Income Security Act of 1974, as amended (ERISA), specifically ties an employer’s withdrawal liability to the plan’s actual experience and reasonable expectations, only to be disregarded if a mass withdrawal occurs. The Proposed Regulation attempts to circumvent this requirement by allowing the use of a discount rate untethered to a plan’s history or reasonable expectations. The PBGC’s rationale for its proposed approach conflates an individual employer’s withdrawal with the mass withdrawal of, all or substantially all, contributing employers, which are two different events that Congress chose to treat differently. It also raises constitutional concerns.  We therefore request the PBGC withdraw the Proposed Regulation.

 Background on Withdrawal Liability

“Most of the participating employers in multiemployer plans are small businesses. Multiemployer plans offer small businesses a way to provide a pension plan with a minimum of administrative burden. They offer the employees of small businesses portability, [and] a way to have an assured pension without being tied to a particular employer.”[1] Because of this unique makeup, Congress treated single-employer plans differently than multiemployer plans for purpose of ERISA Title IV and the PBGC insurance program. For example, “[i]n the event a single employer [plan] is unable to meet its funding obligations, the PBGC must step in and take over the plan liabilities and administer the plan. For multiemployer plans, however, the pool of contributing employers assumes that role and the agency acts as insurer of last resort, becoming involved in the funding of a plan only when it becomes insolvent.”[2]

However, due to a confluence of events, the makeup of multiemployer plans was not enough to make many of them sustainable. In 1980, Congress found that the “characteristics of multiemployer plans make them susceptible to industry declines or employer withdrawals which may result in a substantially increased funding burden for the remaining employers which can adversely affect labor-management relations…” and “economic problems in some industries supporting multiemployer plans make plan continuation highly uncertain….”[3]

To address this, Congress amended ERISA with the Multiemployer Pension Plans Amendments Act of 1980 (MPPAA), and withdrawal liability was a major part of this reform. However, nothing in the legislative history or Congressional findings indicates that withdrawal liability was meant to be punitive on the employer or a windfall for the plan. Instead, withdrawal liability was meant for the employer to “fund a reasonable share of the plan's unfunded vested obligations.”[4]

In MPPAA, Congress meticulously laid out what withdrawal liability is, the process by which it is calculated, and how an employer can contest the calculation. “Specifically, a withdrawing employer must pay the plan its proportional share of the plan's ‘unfunded vested benefits,’ which is ‘the difference between the present value of the plan's vested benefits and the present value of its assets.’”[5] To make this determination, an actuary must make numerous assumptions, such as mortality, vesting, and the interest, or discount, rate to determine the present value of future benefits. “The discount rate is the weightiest assumption in the overall withdrawal liability calculation.”[6]

Although Congress provided that the PBGC “may prescribe by regulation actuarial assumptions which may be used by a plan actuary in determining the unfunded vested benefits of a plan for purposes of determining an employer’s withdrawal liability…[emphasis added]”[7], which PBGC failed to do for over 40 years,[8] Congress did not authorize the PBGC to provide blanket approval for any assumptions that an actuary chooses without regard to an individual multiemployer plan’s history or reasonable expectations.  Rather, Congress still required that any assumptions, whether determined by a plan’s actuary’s best estimate of the plan’s experience and reasonable expectations or prescribed  (not “clarified” or “approved”) by the PBGC, must still be reasonable taking into account the experience of the plan and its reasonable expectations.[9] As discussed below, nothing in Congress’s grant of authority to the PBGC to prescribe assumptions in one section of ERISA eliminated the requirement that assumptions be reasonable and based on the individual experience and expectations of the multiemployer plan set forth in a separate section of ERISA. Congress certainly did not authorize the PBGC to declare sections of ERISA inapplicable thereby usurping Congress’s role.

Instead, in promulgating MPPAA, Congress recognized that in determining the withdrawal liability assumptions, plan data is imperative, and the plan actuary must rely on the most recently completed minimum funding actuarial valuation and reasonable estimates for the interim years of the unfunded vested benefits. If such data is unavailable, the actuary may rely “on the data available or on data secured by a sampling which can reasonably be expected to be representative of the status of the entire plan.”[10] However, any assumptions or methods must be related to the specific plan.

Although a plan actuary is given considerable leeway in determining withdrawal liability, and although the PBGC may prescribe by regulations actuarial assumptions that may be used, neither is unfettered. ERISA explicitly provides a withdrawal liability dispute resolution system, under which any dispute must first go through arbitration. A plan’s determination of its unfunded vested benefits (and ultimately withdrawal liability) is presumed correct unless a party contesting the determination shows by a preponderance of the evidence that “the actuarial assumptions and methods used in the determination were, in the aggregate, unreasonable (taking into account the experience of the plan and reasonable expectations)” (emphasis added).[11]

Although the statutory scheme dictates that an employer must pay withdrawal liability, there is nothing in ERISA that dictates how a plan must use or invest these funds. Congress certainly could have dictated this given that Congress did so for purposes of terminated single-employer plans which requires the plan administrator to either purchase annuities or otherwise fully provide all benefits under the plan.[12]  

Congress Distinguished Mass Withdrawal from Withdrawal Liability 

Congress recognized that there is a distinction between one employer leaving a plan and all (or substantially all) employers leaving a plan, which is why there are separate rules for each event. The PBGC rates are used to determine the present value of future liabilities for plans that have a “substantially all” mass withdrawal (generally 85% or more of a multiemployer plan’s contributors or contribution base units withdraw) or that are terminated by a mass withdrawal, which occurs when all employers completely withdraw from a multiemployer plan.[13] Even in a “substantially all” mass withdrawal, trustees often involuntarily terminate the participation of the few remaining participating employers. A mass withdrawal will typically result in the formal termination of the plan. In a termination setting, employer contributions and accruals permanently cease. And almost always, a mass withdrawal occurs when the multiemployer plan has little to no remaining assets, so investment returns will not be available or sufficient to pay for future benefits. By using mass withdrawal assumptions, the multiemployer plan sets withdrawal liability at large enough amounts so that withdrawn employers will owe annual withdrawal liability payments indefinitely to cover benefits (usually with PBGC financial assistance) until no participants are owed any benefits.

Courts have distinguished between using the mass withdrawal interest rate, where the plan is essentially going out of business and is settling its liabilities, and a different interest rate where only one employer withdraws and the plan is ongoing. In Sofco, the fund argued that blending the actuary's growth-rate assumption with the interest rate on risk-free annuities is fair because this transfers investment risk from the pension fund to the withdrawing employer. However, the Sixth Circuit found that “. . . [p]erhaps this is a laudable policy proposal. But, even if it is, we see no indication in the statute that Congress has adopted it. ERISA does incorporate a risk-shifting regime for mass withdrawals, which the statute does treat as a ‘settlement’ [under 29 U.S.C. § 1341(a)(c)(2)]. And certain parts of the statute do direct actuaries to make calculations ‘without taking into account the experience of the plan.’ See id. § 1084(c)(6)(E). But there is nothing in the statutory text to indicate that ERISA adopted a similar regime for individual withdrawals from multiemployer plans.”[14]

“Thus, risk-free rates might be appropriate if a plan were invested in risk-free assets, or perhaps if it planned to invest the withdrawal liability payments in risk-free assets. But if the plan is currently and projects to be invested in riskier assets, the discount rate used to calculate withdrawal liability must reflect that fact.”[15] It would be a rare plan that only invested in risk-free assets.  “While a minority of advisors began to suggest movement to ‘immunized’ and ‘risk-free' portfolios, most advisors and plan fiduciaries rejected that advice as failing to adequately recognize the equity premium which was historically realized by long-term ‘patient’ investors.”[16]

The PBGC Proposed Regulation

Section 4213.11(a) provides that withdrawal liability may be determined using the actuarial requirements in this section. Section 4213.11(b) of the Proposed Regulation provides that the interest rate assumption for withdrawal liability purposes must either be the interest rate under Section 4044.52 (mass withdrawal rates) on the date withdrawal liability is determined, the interest rate under ERISA section 304(b)(6) (funding rate) for the plan year within which withdrawal liability is determined, or a rate between these two rates. Section 4213.11(b) does notprescribe any required interest rate (even though prescribing a rate is the limit of the PBGC’s authority under the statute), but, in effect, authorizes actuaries’ historical practice of selecting a rate anywhere from the mass withdrawal rates to the funding rates. Under section 4213.11(c), all other assumptions must each be reasonable (taking into account the experience of the plan and reasonable expectations) and, in combination, offer the actuary’s best estimate of anticipated experience under the plan.


1.      The PBGC’s Proposed Regulation Exceeds Its Authority

Section 1393 amended ERISA 42 years ago, but the PBGC did not issue any regulation in this time because it knew that there was no single interest rate that could take into account the experience or reasonable expectations of every multiemployer plan. However, the PBGC now purports to have the authority to issue a Proposed Regulation that rewrites Title IV’s withdrawal liability provisions by removing the requirement that the withdrawal liability determination be reasonable taking into account the experience and reasonable expectations of the plan, contrary to Congressional intent. The PBGC’s authority to prescribe assumptions does not equate to the authority to authorize a wide range of interest assumptions unrelated to the plan in question or void the application of statutory provisions requiring assumptions to be based on a plan’s experience and expectations. The PBGC admits in the preamble that under the Proposed Regulation, the plan trustees’ investment risk appetite, asset allocation choices or the actuary’s best estimate of the plan’s future investment returns are not relevant to the withdrawal liability assessment, notwithstanding express statutory language to the contrary.  As such, and in direct conflict with statutory requirements, the Proposed Regulation has  divorced any plan experience or reasonable expectations from the interest rate used in determining the amount of unfunded vested benefits in the withdrawal liability calculation.[17]

The requirements of the withdrawal liability dispute resolution system make clear that the PBGC’s authority under 29 U.S.C. § 1393(a)(2) does not permit the PBGC to divorce the interest rate under 29 U.S.C. § 1393(b) from the plan’s experience and reasonable expectations, and the DC Circuit Court of Appeals in Energy West made this clear.  Specifically,

The dispute resolution provision permits vacating an arbitration award if the actuarial assumptions were unreasonable in the aggregate "taking into account the experience of the plan."  § 1401(a)(3)(B)(i). The Aggregate Reasonableness Requirement, both for dispute resolution and for withdrawal liability in Section 1393(a)(1), does not just require assumptions that are reasonable in the abstract; it requires assumptions that are reasonable relative to the plan, taking the plan's experience into account. If the actuary is not basing the assumptions on the plan's characteristics, the assumptions will not be reasonable ‘taking into account the experience of the plan.’ In other words, not only must the actuary's assumptions be reasonable, but they must also be aimed at the right calculation, namely the predicted future of the plan.[18]

The requirements under § 1401(a)(3)(B)(i) do not exclude assumptions prescribed by the PBGC under § 1391(a)(2) from the requirement to be reasonable taking into account plan’s experience and reasonable expectations. 

Finally, ERISA provides that the PBGC may prescribe by regulations actuarial assumptions to be used in calculating withdrawal liability. Under this language, the PBGC’s limited authority is to “lay down” specific assumptions and methods to be used.[19] The Proposed Regulation does no such thing, and, instead, according to the PBGC itself, is meant to clarify that a plan actuary’s use of Section 4044 rates, funding rates, or anything in between, represents a valid approach to selecting an interest rate assumption in all circumstances, and the Proposed Regulation will eliminate the cost-shifting effects of impediments to actuaries’ use of such rates.[20]  “Clarifying” a range of assumptions that can cause vested liabilities in a multiemployer plan to increase five-fold or more, depending on the plan’s interest rate selection, is not prescribing assumptions. Instead, the Proposed Regulation attempts to provide blanket approval for multiemployer plan actuaries to select virtually any rate with no correlation to the plan’s experience or expectations, which is in direct conflict with the relevant statutory provisions.

2.      The PBGC Justification Does Not Hold Merit.

Throughout the preamble to the Proposed Regulation, the PBGC justifies the use of the mass withdrawal rates by stating that “it is reasonable to base the amount needed to settle the employer’s share of the liability on the market price of settling pension liabilities by purchasing annuities from a private insurer.”[21] The PBGC states that “[c]onsideration of the anticipated experience of the plan in selecting withdrawal liability interest assumptions is not necessarily appropriate in light of a withdrawing employer’s lack of continued shared investment experience.”[22]

However, as noted above, nothing in ERISA requires a plan to annuitize and actually settle the liabilities associated with the withdrawing employer’s participation or to use a withdrawing employer’s payments to annuitize and settle any plan liabilities. Furthermore, there is no indication whatsoever that plans actually do so. In fact, according to an actuarial firm that specializes in multiemployer plans, multiemployer pension plans “are usually invested in a well-diversified mix of stocks, bonds, real estate, and alternative investments structured to meet the goals of the Trustees. This typically involves maximizing returns over the long term while minimizing return volatility.”[23] There was nothing in the Horizon report to indicate that withdrawal liability payments are invested any differently. 

To the extent a withdrawing employer is not continuing to share in a plan’s investment risk,  the withdrawing employer neither shares in losses nor gains. Withdrawal liability should be based on each employer’s allocable share of unfunded vested benefits at the time of withdrawal, not on the prospect of future unfunded liabilities that may not be realized, which is what the Proposed Regulation attempts to do. In the ordinary course, an actuary’s best estimate of future investment experience should bear an equal risk of future losses or gains. Allowing the use of mass withdrawal rates to calculate an employer’s withdrawal liability creates a significant opportunity for bias. Trustees, in managing their risk appetite, and actuaries, in determining their best estimate assumptions of plan experience for funding purposes, could be incentivized to take on greater risk because withdrawing employers would be saddled with liability that absorbs the additional risk. There is no indication in the Preamble that the PBGC has examined the opportunity for actuarial or other bias and the policy consequence of it in evaluating its Proposed Regulation. See also, infra, Section 3, regarding the constitutional implications of the opportunity for bias in the Proposed Regulation.

The PBGC also inaccurately asserts that the Proposed Regulation is needed because there is a split between the courts as to the appropriate interest rate to be used. However, no such split exists. To date, three different Circuit Courts have ruled on cases involving the appropriate withdrawal liability interest rate, and all three Circuit Courts have reached virtually the same conclusion, namely, multiemployer plan actuaries must use a rate that considers the plan’s experience and reasonable expectations.[24] No Circuit Court has reached the opposite conclusion. The one case where the district court reached the opposite conclusion was settled with the appeal dismissed. 

3.     The Proposed Regulation Raises Constitutional Concerns

The Proposed Regulation raises serious concerns under the Due Process Clause of the Constitution, and could threaten the constitutionality of withdrawal liability overall. The U.S. Supreme Court previously upheld the withdrawal liability statutory scheme only because it concluded that plan actuaries did not have unbridled discretion to select discount rate assumptions. By giving actuaries precisely that boundless authority, the Proposed Regulation undermines the premise of the Court’s decision and revives the due process objections to the MPPAA.

In Concrete Pipe & Products of California, Inc. v. Construction Laborers Pension Trust for Southern California, 508 U.S. 602 (1993), the U.S. Supreme Court considered a series of constitutional challenges to the withdrawal liability provisions of MPPAA. One challenge focused on the statutory presumption of correctness afforded to the actuary’s calculation of unfunded vested benefits under 29 U.S.C. Section 1401(a)(3)(B). The employer contended that the presumption violated the Due Process Clause by effectively allowing a biased adjudicator to make the critical decisions bearing on an employer’s withdrawal liability. The argument was that the plan’s actuary is selected by the plan trustees, and the plan trustees are fiduciary-bound to maximize assets of the plan, not to be fair to withdrawing employers.[25] 

The Supreme Court rejected that challenge because it reasoned that actuaries do not have meaningful discretion to select biased assumptions, and are not “vulnerable to suggestions of bias or its appearance.”[26]  “The technical nature of an actuary’s assumptions and methods, and the necessity for applying the same assumptions and methods in more than one context, as a practical matter limit the opportunity an actuary might otherwise have to act unfairly toward the withdrawing employer.”[27] The Court specifically focused on the “statutory requirement (of ‘actuarial assumptions and methods—which, in the aggregate, are reasonable . . . ‘)” and observed that it “is not unique to the withdrawal liability context, for the statute employs identical language in 29 U. S. C. § 1082(c)(3) to describe the actuarial assumptions and methods to be used in determining whether a plan has satisfied the minimum funding requirements contained in the statute.”[28] Citing that “use of the same language to describe the actuarial assumptions and methods to be used in these different contexts,” the Supreme Court identified a “check” on “the actuary's discretion.”[29] Indeed, because “arguably the most important assumption” in the withdrawal liability calculus is “the critical interest rate assumption” (that is, the discount rate), and because that interest assumption “must be used for other purposes as well” (that is, for minimum funding purposes), the Supreme Court concluded that there was no “significant opportunity for bias to operate.”[30] Therefore, the presumption that the plan actuary’s calculation is correct did not violate the Due Process Clause.

The Supreme Court upheld the statute because it understood that there existed a sufficient “check” on actuarial “discretion” in the selection of assumptions, and most particularly the selection of “the critical interest rate assumption.” That check derived largely from the requirement that the actuary employ the same (or at least similar) assumptions across a variety of contexts so that the actuary could not select a particular discount rate solely for withdrawal liability purposes and solely to maximize the amount that a withdrawing employer must pay. 

The Proposed Regulation would eliminate that check and render that discretion virtually boundless by allowing the actuary to select any assumption from an extremely wide range of interest rates without any link to actual plan experience or expectations. Under the Proposed Regulation, actuaries would be free to select any discount rate between the PBGC mass withdrawal rates and the plan’s minimum funding rate, a spread that for some plans in recent years has reached nearly 500 basis points.[31]  By taking away the constraints the Supreme Court understood to exist, the Proposed Regulation would  allow a “significant opportunity for bias to operate,” and revive the due process concerns that the Court laid to rest.[32]  These constitutional problems are sufficient reason to withdraw the Proposed Regulation.[33]

4.     The PBGC’s Economic Impact Analysis Is Flawed

Small Employer

Section 603(a) of the Regulatory Flexibility Act (RFA) provides that:

Whenever an agency is required by section 553 of this title, or any other law, to publish general notice of proposed rulemaking for any proposed rule, or publishes a notice of proposed rulemaking for an interpretative rule involving the internal revenue laws of the United States, the agency shall prepare and make available for public comment an initial regulatory flexibility analysis. Such analysis shall describe the impact of the proposed rule on small entities. 

The term small entity has the same meaning as the term “small business” or small organization.[34] The term small business is defined to include “small business concern under section 3 of the Small Business Act,” which varies according to size and or industry. 

The RFA requires an agency to describe why it is considering the regulatory action, the small entities the rule will apply to, and, where feasible, an estimate of the number of small entities and the projected reporting, recordkeeping, and other compliance costs. The RFA also requires a description of any significant alternatives that would accomplish the statutory objectives but minimize the cost to small entities.

The PBGC provided limited analysis of the Proposed Regulation’s impact on small employers, but instead based its analysis on small plans with less than 100 participants. The PBGC found that out of the approximately 1,360 multiemployer plans, there are only 38 with less than 100 participants. “Based on its definition of small entity, PBGC certifies under section 605(b) of the Regulatory Flexibility Act (5 U.S.C. 601 et seq.) that the amendments in this proposed rule will not, if promulgated, have a significant economic impact on a substantial number of small entities.”

The PBGC’s assessment ignores the fact that small entities under the RFA include small business, and as noted earlier in this comment letter and by the PBGC, the vast majority of employers in most plans are, in fact, small businesses. Given that under the Small Business Administration’s regulation, the smallest employer has 250 employees, it is more likely than not that the vast majority of employers that contribute to multiemployer plans would meet this definition.[35] However, the PBGC provides little to no analysis of the impact on small employers, and, as discussed later, ignored alternatives that could accomplish the statutory objectives and have less impact on small employers.

The Manhattan-Lincoln withdrawal liability case is instructive of the large impact the Proposed Regulation will have on small employers. Although unclear from the decisions, based on current public records, Manhattan-Lincoln Ford would be considered a small employer with only 29 employees and $55 million in revenue.[36] Using the plan’s funding rate, the plan was overfunded, and the employer owed no withdrawal liability. However, using the Segal blend method the plan owed $2.55 million.[37] If the actuary in this case had used the mass withdrawal rates, which are even lower than the Segal Blend, the withdrawal liability would have been even higher.[38] It is unfathomable to think that a proposed rule that would increase a small business’ withdrawal liability by millions of dollars should not only include the dollar impact, but also other economic impacts, such as the ability to get a loan, to hire or retain new employees, or to sell or merge the business.

The PBGC states that for small plans that adopt the Section 4044 rate, the Proposed Regulation would have a positive economic impact by increasing the amount of withdrawal liability and deterring employers from withdrawing. The PBGC also states that because small employers have smaller contribution levels, they will see smaller dollar increase in withdrawal liability. However, this ignores that “small employers are often less able to absorb contribution rate increases than large employers in the same plan. Also, in industries where small employers bid for projects, increased costs related to their participation in multiemployer plans can put them at a competitive disadvantage against nonunion competitors.”[39] Finally, the Proposed Regulation  provides no incentives for employers that can afford to leave to stay. Thus, the Proposed Regulation makes it is more likely that larger employers who can leave, will, which would make the plans more financially unsound for smaller employers that cannot afford to leave.

Increasing the assessed amount of withdrawal liability against large employers that will now elect to leave (possibly before their multiemployer plans have the opportunity to move to the Section 4044 rates) does not improve the plan’s funding status. Rather, such an employer’s withdrawal liability payment will approximate its current contributions, will not be subject to future contribution rate increases, and will only last for 20 years. Adopting this Proposed Regulation will expedite, not forestall, the decline of multiemployer plans, leaving only small employers that cannot afford to withdraw from these plans as the contributors. When these small employers do leave the plan, the increased withdrawal liability assessment by virtue of the Proposed Regulation will mean nothing because they will likely only be able to withdraw upon filing for bankruptcy protection.

The PBGC acknowledges that inaction would have been less impactful on small employers, but it states inaction would have contributed to plan underfunding. However, as discussed below, the PBGC did not provide an analysis of any alternatives or provide any analysis to support the conclusory statement that it would lead to plan underfunding.


In its analysis, the PBGC assumes that because the Proposed Regulation provides increased certainty,[40] it would reduce arbitration and litigation costs for plans and employers with an annual savings of $500,000 to $1,000,000. This estimate does not consider the costs of the inevitable years of litigation between employers and multiemployer pension plans (and possibly the PBGC) that will follow if the Proposed Regulation becomes final. The PBGC’s analysis also does not compare the cost of an arbitration settlement (estimated at $82,500 to $220,000) to the additional cost of withdrawal liability. As shown by the case law cited in this letter, the use of the mass withdrawal rate versus the funding rate can triple the amount of withdrawal liability. As is the case in Energy West, the $75 million withdrawal liability increase is far more than the cost of arbitration.

This increase will impact nearly every employer. The PBGC states that in the 20 years following the effective date, there will be a nominal increase in cumulative withdrawal liability payments ranging between $804 million and $2.98 billion. However, the PBGC does not explain 1) how this would be nominal for the thousands of small employers in these plans; or 2) why all plans will not move to the mass withdrawal rates given that PBGC forewarns that those who do not will be able to pay benefits.

Other Economic Harm

With no analysis supporting its conclusions, the PBGC states that larger withdrawal liability increases plan solvency and ensures benefits are received, “which would enhance [participants’] income security and benefit the communities, including small businesses within those communities….” The PBGC’s analysis does not consider lost future contributions from employers that withdraw because of the Proposed Regulation. Further, as noted above, these withdrawal liability increases are not insignificant, even for large employers. For example, in Energy West, the use of the funding rate resulted in withdrawal liability of $40 million, but the use of the Section 4044 rates increased the withdrawal liability to $115 million, which, even for a large company, is significant. Not only will the increased withdrawal liability exposure impact a company’s balance sheet, but its active employees, its shareholders, its ability to obtain loans, and its ability to sell any part of the entity. As the PBGC noted in its report to Congress, this is especially difficult for smaller employers that “are often surprised by the presence of withdrawal liability when they are ready to close or sell their business or to retire. The cost of potential withdrawal liability – particularly if the plan’s unfunded liabilities are significant (due usually to events beyond the small employer’s control) – may discourage a prospective buyer from the purchase of a small business.”[41] However, none of this was discussed in the PBGC analysis.


In the preamble to the Proposed Regulation, the PBGC states that “[n]one of the alternatives were as cost-effective as the proposed rule.”[42] However, the PBGC does not state to whom or what entity this is the most cost-effective and why. Furthermore, the only alternative the PBGC discusses is inaction, which it says “could contribute to plan underfunding, benefit losses, cost shifting to the remaining employers, and higher claims on PBGC’s insurance system.”[43]  This analysis fails to explain how inaction would lead to any of these results. Moreover, multiemployer pension plans that currently project insolvency, which are the plans that will become claims against the PBGC insurance system, are eligible to receive Special Financial Assistance from the PBGC. As a condition of receiving the PBGC Special Financial Assistance, these multiemployer plans have to use the Section 4044 rates. This means that this Proposed Regulation would only impact plans that are not going to receive Special Financial Assistance. As explained in our recommendations, there is plenty of middle ground between allowing an interest rate that is completely unrelated to the plan and, in most cases, will provide a windfall to the plan, and doing nothing.

Our Recommendations

We recommend the PBGC withdraw the current Proposed Regulation. If the PBGC believes it is beneficial to reconsider this issue, then we recommend it proceed through an Advanced Notice of Proposed Rulemaking to solicit public comments on alternative proposed rules. In formulating a new Proposed Regulation, if any, the PBGC should be mindful of its statutory authority and the underlying premise of withdrawal liability requiring the actuarial assumptions to take into account the plan’s experience and reasonable expectations. For example, the PBGC could analyze and allow actuaries to use a discount rate for withdrawal liability that is within so many basis points of the funding rate, perhaps to account for a withdrawing employer not paying future administration costs. The PBGC also could determine a rate that would be considered neutral to both the plan and the employers, and then allow that for the withdrawal liability rate. A study from one actuarial firm that works with multiemployer plans indicates that a 7.0 percent withdrawal liability interest rate is risk neutral over a 20-year period because there is a 52 percent probability that future investment returns will meet or exceed that 7.0 percent interest rate assumption.[44] Therefore, the employer causes no harm to the fund by withdrawing. Accordingly, that interest rate might be a floor for a withdrawal liability interest rate where a plan actuary also adopts a higher rate for minimum funding purposes.

Another alternative is to tie the interest rate to the 20-year average of investment returns for some plans and a 10-year average for others, depending on the characteristics of the plan.[45] Another alternative is to allow for the interest rate assumptions in the Proposed Regulation, but to require that under Section 4213.11(c), all assumptions, including the interest rate, must each be reasonable (taking into account the experience of the plan and reasonable expectations) and, in combination, offer the actuary’s best estimate of anticipated experience under the plan. Under such an approach, it may be appropriate for a critical and declining plan (whether it applies for SFA or not) that should be focusing on asset protection to use the Section 4044 rate, while the funding rate may be the appropriate rate for a plan not projecting insolvency. However, allowing any plan to use the Section 4044 rates notwithstanding its experience or reasonable expectation does not comply with ERISA.

If the PBGC insists on allowing an actuary to use an interest rate that is completely unrelated to the plan or the plan’s reasonable expectation, the PBGC must explain how that aligns with ERISA Section 4221, which requires an arbitrator to uphold the withdrawal liability calculations unless it is shown that “the actuarial assumptions and methods used in the determination were, in the aggregate, unreasonable (taking into account the experience of the plan and reasonable expectations).”[46] The PBGC should explain how it would be possible for a plan with, for example, a funding rate of 7.5 percent based on the plan’s historic investments and reasonable expectations, that elects to use the mass withdrawal rate of 2.8 percent, which could increase unfunded vested benefits five-fold, to be able to adjust other assumptions, such as mortality and vesting, so that the end result is reasonable in the aggregate, taking into account the experience of the plan and reasonable expectations.[47]

Finally, any cost analysis must include the impact on small employers. To ignore this, ignores the fundamental design of these plans that are made up of hundreds of small contributing employers that are the main source of income for these plans. Without employers, these plans could not exist.


If the goals of the Proposed Regulation were to encourage employers to stay in these plans and attract more employers, neither were met. The Proposed Regulation will encourage those employers that can get out now to do so, while leaving more financially vulnerable employers in these plans, with no hope of any new employers joining knowing that they could be on the hook for millions of dollars for only participating for a few years.

We encourage the PBGC to withdraw the Proposed Regulation and work with the entire community to find a solution that balances the financial needs of plans, employers, active participants and retirees.


U.S. Chamber of Commerce

The Association of Food and Dairy Retailers, Wholesalers, and Manufacturers

Associated General Contractors of America

[1] Multiemployer Pension Plans Report to Congress Required by the Pension Protection Act of 2006, prepared by the Secretaries of the Department of Labor and the Department of the Treasury and the Director of the PBGC available at

[2] Examining Building a Secure Future for Multiemployer Pension Plans, Focusing on Long-Standing Challenges that Remain for Multiemployer Pension Plans, Hearing before the Committee on Health, Education, Labor and Pensions,  S. Hrg. 111-1140 (Statement of Randy G. DeFrehn, Executive Director of the National Coordinating Committee for Multiemployer Plans) available at

[3] 126 Cong. Rec. 12180 (May 22, 1980).

[4] Id.

[5] United Mine Workers of Am. 1974 Pension Plan v. Energy W. Mining Co., 39 F.4th 730, 735 (D.C. Cir. 2022) (quoting Connors v. B & H Trucking Co., 871 F2d. 132, 133 (D.C. Cir. 1989)).

[6] Energy W. Mining Co., 39 F.4th at 735 (D.C. Cir. 2022), (citingCombs v. Classic Coal Corp., 931 F2d 96, 101 (D.C. Cir. 1991)) (explaining that an "erroneously low" discount rate, without appropriate offsetting assumptions, might "destroy the validity of the entire calculation" of unfunded vested benefits).

[7] 29 U.S.C. § 1393.

[8] In 1980, the PBGC issued a list of regulations it proposed to issue after MPPAA and requested comments from the public. The actuarial assumptions under 4213(a) were included in that list because “the regulation is needed to provide standards that may be used for the computation of withdrawal liability.” 45 Fed. Reg. 72,213 72,215 (Oct. 31, 1980). Although the actuarial standards were on the initial list, no proposed regulations were issued.

[9] 29 U.S.C. § 1401(b)(3)(i).

[10] 29 U.S.C. § 1393(b).

[11] 29 U.S.C. § 1401(a)(3)(B).

[12] 29 U.S.C. Section 1341(b)(3). With respect to terminated multiemployer plans, Congress dictated that benefits generally may only be paid in the form of an annuity, unless “the plan assets are distributed in full satisfaction of all nonforfeitable benefits under the plan.” 29 U.S.C. § 1341a(c)(2). The PBGC has determined that purchasing a commercial annuity will satisfy this requirement. 29 C.F.R. § 4041A.42(a).

[13] 29 U.S.C. § 1341a(a)(2)(plan deemed terminated when all employers completely withdraw). 

[14] Sofco Erectors, Inc. v. Trs. of Ohio Operating Eng'rs Pension Fund, 15 F.4th 407, 422 (6th Cir. 2021).

[15] Energy W. Mining Co., 39 F.4th at 738.

[16] Examining Building a Secure Future for Multiemployer Pension Plans, Focusing on Long-Standing Challenges that Remain for Multiemployer Pension Plans, Hearing before the Committee on Health, Education, Labor and Pensions,  S. Hrg. 111-1140 (Statement of Randy G. DeFrehn, Executive Director of the National Coordinating Committee for Multiemployer Plans) available at

[17] See GCIU-Employer Retirement Fund, 2022 WL 15579987, *7 (“Because …[the PBGC rate] overlooks the plan's expected returns, it does not satisfy the “best estimate” standard”).

[18] Energy W. Mining Co., 39 F.4th  at 741.

[19] The definition of prescribe is to “to lay down as a guide, direction, or rule of action.”

[20] 87 Fed. Reg. 62316, 62319 (Oct. 14, 2022).

[21] 87 Fed. Reg. 62316, 62318 (Oct. 14, 2022).

[22] 87 Fed. Reg. 62316, 62318 (Oct. 14, 2022). The PBGC’s assertion that a “withdrawing employer . . . is settling its liabilities once and for all and bears no risk of future losses,” 87 Fed. Reg. 62316, 62317 (Oct. 14, 2022) (emphasis added), is not accurate. When an employer engages in a partial withdrawal, it continues to bear the future risk of complete withdrawal liability, and a completely withdrawing employer often bears the continued risk of mass withdrawal liability. The amounts of such future liability events will be affected by the plan’s future investment experience.

[23] See Survey of Capital Market Assumptions 2022 Edition, p. 7, Horizon Actuarial Service available at

[24] Energy W. Mining Co., 39 F.4th at 743 (interest rate must be similar to funding rate); Sofco Erectors, Inc., 15 F.4th at 423 (invalidating use of Segal Blend and requiring use of funding interest rate); GCIU-Employer Retirement Fund v. MNG Enterprises, Inc. 2022 WL 15579987, **8, 9, n2 (9th Cir. Oct. 28, 2022) (expressing no view on the validity of the Proposed Regulation and finding that multiemployer plan had to use interest rate that considered experience and expectations of the plan).

[25] See Concrete Pipe, 508 U.S. at 617-18; see also Nat’l Ret. Fund v. Metz Culinary Mgmt., Inc., 946 F.3d 146, 151 (2d Cir. 2020) (“Actuaries unwilling to yield to trustees’ preferred interest rate assumptions can be replaced by others less reticent.”)

[26] Concrete Pipe, 508 U.S. at 632.

[27] Id.

[28] Id.

[29]Id. at 632-33.   

[30] Id. at 633.

[31] SeeUnited Mine Workers of Am. 1974 Pension Plan v. Energy W. Mining Co., 39 F.4th 730, 742 (D.C. Cir. 2022).

[32] Concrete Pipe, 508 U.S. at 633. 

[33] The spread in recent years between the PBGC mass withdrawal rates and funding rates also means that the Proposed Regulation will lead to arbitrary and capricious results. For example, two employers, with similar contribution histories to the same plan, that withdraw one year apart may incur liability amounts that are many multiples different, simply because in the first year the actuary uses the funding rate as the discount rate and in the next it reverses course and uses the mass withdrawal rate.

[34] Regulatory Flexibility Act § 601(6).

[35] 13 CFR § 121.201.

[36] See

[37] Manhattan Ford Lincoln, Inc. v. UAW Local 259 Pension Fund, 331 F. Supp. 3d 365 (D.N.J. 2018)

[38] See MNG Enterprises, Inc. dba, Digital First Media,  2022 WL 15579987, *5 (“Generally, using the PBGC rate results in a higher amount of withdrawal liability because it assumes a lower rate of growth.”)

[39]Multiemployer Pension Plans Report to Congress Required by the Pension Protection Act of 2006, prepared by the Secretary of the Department of Labor, the Secretary of the Department of the Treasury and the Director of the PBGC available at

[40] The Proposed Regulation does not provide for increased certainty, and employers will continue to challenge withdrawal liability calculations. Under the Proposed Regulation a plan may use the PBGC rates under the Proposed Regulation. If the actuary uses these assumption, the interest rates must either be the mass withdrawal rate, the funding rate or somewhere in between. However, if this rate does not take into account the experience of the plan, it will not be upheld by an arbitrator under 29 U.S.C. § 1401(a)(3)(B)(i).

[41] Multiemployer Pension Plans Report to Congress Required by the Pension Protection Act of 2006, prepared by the Secretary of the Department of Labor, the Secretary of the Department of the Treasury and the Director of the PBGC available at

[42] 87 Fed. Reg. 62316, 62321 (Oct. 14, 2022).

[43] 87 Fed. Reg. 62316, 62321 (Oct. 14, 2022).

[44] See Survey of Capital Market Assumptions 2022 Edition, Horizon Actuarial Service, available at

[45]A major actuarial firm that works with multiemployer plans has noted that for “for less mature ongoing pension plans without solvency issues, we believe a horizon of 20 years or more is appropriate for evaluating the reasonableness of the long-term investment return assumption. A shorter horizon, such as 10 years, may be more appropriate for evaluating the return assumption for a plan that is more mature or has solvency issues.” See Survey of Capital Market Assumptions 2022 Edition, Horizon Actuarial Service, available at

[46] 29 U.S.C. § 1401(a)(3)(B).

[47] The Preamble provides that the Proposed Regulation would apply to withdrawals that occur on or after the effective date of the final rule. 87 Fed. Reg. 62316, 62317. The Preamble, however, goes on to state that “the proposed rule does not preclude the use of an interest rate assumption described in proposed § 4213.11(b) to determine unfunded vested benefits before the effective date of the final rule.” Id.  The liability for any withdrawal that occurs before the effective date of the final rule must be determined under the best estimate of anticipated plan experience standard set forth in 29 U.S.C. Section 1393(a)(1). Throughout the Preamble, however, the PBGC acknowledges that the mass withdrawal rates per proposed Section 4213.11(b) are not derived from the best estimate of anticipated plan experience. See, e.g., 87 Fed. Reg. 62316, 62317 (noting that mass withdrawal rates approximate the market price of buying annuities, and that “[f]rom this perspective, the plan trustees’ investment risk appetite, asset allocation choices, or the actuary’s best estimate of the plan’s future investment returns following the withdrawal are not relevant to the withdrawal liability assessment”); 62318 (“consideration of the anticipated experience of the plan in selecting withdrawal liability interest assumptions is not necessarily appropriate”).  Given the PBGC’s recognition that mass withdrawal rates are not derived from and do not satisfy an anticipated plan experience standard, the PBGC should not state that any proposed Section 4213.11(b) assumptions are not precluded to determine unfunded vested benefits before the final rule. Such a statement irreconcilably conflicts with the PBGC’s reasoning for its rule. If the PBGC does not withdraw the current Proposed Regulation it should make clear in the preamble to any final rule that interest rates derived from mass withdrawal rates are precluded from pre-final rule withdrawal calculations under 29 U.S.C. Section 1393(a)(1), or at least withdraw and make clear it is not making any statement about calculation of withdrawal liability before the effective date of the final rule.