A program of loans to troubled multiemployer plans has the potential to help preserve participant benefits and improve the financial condition of the plans.
The loans will represent a cost to the government if they charge plans an artificially low interest rate, and there will be additional costs if plans are ultimately unable to repay the loans.
There is a relationship between the ability of a loan program to protect benefits and the costs and risks borne by the government, with greater potential benefits generally corresponding to greater costs and risks.

1850 M Street NW
Suite 300
Washington, DC 20036
202-223-8196 |

Craig Hanna, Director of Public Policy
Ted Goldman, Senior Pension Fellow

© 2018 American Academy of Actuaries. All rights reserved.

Members of the Multiemployer Plans Committee include Jason Russell, MAAA, FSA, EA—chairperson; Christian Benjaminson, MAAA, FSA, EA, FCA—vice chairperson; Mariah Becker, MAAA, EA; James Dexter, MAAA, FSA, EA, FCA; James Donofrio, MAAA, FSA, EA; Aldwin Frias, MAAA, FSA, EA, FCA; Francis Gowen, MAAA, ASA, EA; Eli Greenblum, MAAA FSA, EA , FCA; Joseph Hicks, MAAA, EA, FCA, MSPA; David Pazamickas, MAAA, ASA, EA; and Peter Sturdivan, MAAA, FSA, EA.

 For a print-ready PDF of this page, click here.
The multiemployer pension system faces great challenges. Out of the roughly 1,200 multiemployer pension plans in the country, more than 100 plans, covering in excess of a million participants, are at risk of fully exhausting their assets within the next 20 years. Furthermore, the Pension Benefit Guaranty Corporation (PBGC) multiemployer insurance program is in jeopardy as a result, with the agency projecting that its resources will be depleted in less than 10 years, at which time the already low multiemployer guarantee would fall to a fraction of its current level.1 The American Academy of Actuaries’ Pension Practice Council has previously published issue briefs discussing these challenges in Overview of Multiemployer Pension System Issues and Honoring the PBGC Guarantee for Multiemployer Plans Requires Difficult Choices.

Policymakers and stakeholders recently have been exploring proposals that would authorize either direct government loans, or government-guaranteed loans issued by third parties, to troubled multiemployer plans in an effort to help them return to financial stability. These proposals are offered as mechanisms for preventing troubled multiemployer plans from becoming burdens on the PBGC and for protecting participants’ financial security in retirement by transferring cost and risk from the plans onto the federal government. This issue brief discusses the ways in which a loan program could benefit troubled plans and their participants, and also addresses the costs and risks associated with these proposals.

While a loan program would allow distressed plans to pay participants some or all of the benefits that might otherwise have been lost, it would not address the fundamental causes of the current funding crisis. There would also be no guarantee that a loan program would restore plans to long-term financial health, or that plans that currently expect to be able to pay all promised benefits will not become distressed in the future. The Academy’s Multiemployer Plans Committee looks forward to working with policymakers to help develop not only temporary relief measures (such as a loan program), but also longer-term reforms that will reduce the risk of future crises.


Under current law, federal credit programs are subject to complex federal budget scoring rules as well as other specific federal laws and regulations that govern their administration. Congress must of course evaluate the projected cost and feasibility of a loan program within the context of those rules, including how the federal agencies might implement a program after it is passed. An analysis of these issues is beyond the scope of this issue brief.

Overview of Loan Concept

The use of government-backed loans as a strategy to help troubled multiemployer pension plans would entail several key components.
First, the loan-application process should include criteria for determining whether a given plan is in sufficient financial distress to qualify for a loan, while also demonstrating that the loan is expected to return the plan to financial health. For this purpose, financial health might be defined as the ability to make all benefit payments and required loan repayments when due over the life of the loan, or it could be a higher standard that requires the satisfaction of certain funding metrics after the loan is fully repaid.
There would also need to be a formula or process for determining the amounts of the loans that plans would receive. Additionally, the terms of a loan would need to specify an interest rate (which may be below the rates available to creditworthy borrowers in the marketplace), as well as provisions governing the payment of the loan proceeds to the borrowing plans and repayment provisions. The provisions governing the payment of loan proceeds to the borrower could include either a single lump-sum payment or a series of loan payments over a period of time. The repayment provisions would include the length of time over which the loans would need to be repaid and the pattern of principal repayments, and may include a period during which the borrowing plan only needs to pay the interest on the loan amounts.
A loan program can be accompanied by a dedicated source of revenue to defray some or all of the costs of the program. Examples of potential sources of revenue include amounts paid by all multiemployer plans (similar to current PBGC premiums), amounts withheld from retiree benefits, or taxes levied against the industries that sponsor multiemployer plans. Plans receiving loans through a loan program could also have the authority to implement benefit reductions for their participants if that is necessary to satisfy the loan approval criteria. An additional possibility is that plans could receive some level of financial assistance from PBGC, which would presumably not be repaid2 in conjunction with a loan. Finally, an effective loan program will include a well-constructed administrative process under which loan applications can be assessed, administered, and appropriately monitored over the term of the loan.

Potential Benefits of Loan Programs

A low-interest loan program has potential value for troubled multiemployer pension plans and the various stakeholders in the plans, including participants, contributing employers, and the PBGC.3 Establishing such a program may improve the financial health of individual plans, the security of participant benefits, and the financial outlook for participating employers and the PBGC.
Plans would invest the loan proceeds in accordance with the statutory provisions of the loan program and the rules and regulations adopted by the government agencies administering it. These investments could be in asset classes that are expected to provide returns above the interest rate on the loans, potentially producing excess returns that plans could use to support promised benefits that they would otherwise be unable to pay or pay fully. Unless the loan proceeds are invested in very-low-risk asset classes, the expected returns would be subject to uncertainty. Due to the relationship between the volatility of investment returns and the expected level of returns, a loan program that has the potential to produce greater excess return revenue for plans will also entail greater financial risk to the lender or guarantor of the loan.
Certain design aspects of a loan program will affect the amount of additional income that plans may be able to generate from the loans. Most significantly, the additional income will depend on the amount of the loans. All else being equal, larger loan amounts will produce larger investment results. Charging a lower interest rate on the loans will also increase the ability of plans to generate net income from the loan programs. Lastly, a loan program with longer repayment terms will increase the potential investment gains by providing a longer time horizon. Each of these design features will also have an impact on the potential risks inherent in a loan program, with greater capacity for additional income generally corresponding to greater risks.

A loan program could require that plans invest the loan proceeds in a manner that de-risks the plan by annuitizing retiree benefits or implementing a retiree liability matching strategy. Investing the loan proceeds in this manner would remove much of the uncertainty regarding the future financial condition of the plan and its ability to pay benefits, but this predictability would come with a trade-off. If the loan proceeds are not available to generate investment returns above annuity contract or fixed-income rates, much larger loans may be needed to restore plans to financial health. Alternatively, the loan program might need to be accompanied by significant additional financial assistance measures, such as an expansion of the PBGC insurance program.
A loan can also improve the financial outlook of a plan by creating a longer time horizon over which the employers and active plan participants are able to finance the benefits. To pay benefits, a pension plan needs to accumulate sufficient income, either in the form of contributions or investment earnings, by the time those benefits are due to participants. The projected cash flows into a plan from employer contributions may be unable to support the benefits before they are due, but the same contribution stream may be sufficient to finance those liabilities over a longer time period. For a simple example, consider a payment due in 10 years. The expected contributions to the plan before this payment is due may be insufficient to support this payment. But if the plan receives a loan sufficient to cover the payment and then has 30 years to repay that loan, the expected contributions to the plan over the duration of the loan might be sufficient to support the repayments. A loan program that has a relatively long repayment period will be more effective at extending the time horizon over which the pension liabilities are financed.
In many cases, multiemployer plans that are deeply underfunded have difficulty maintaining the support of both the active participants in the plan and the contributing employers4 These stakeholders may have little desire to continue participating in or contributing to struggling plans, which can lead to further financial strain. A loan may prevent, or at least significantly postpone, the insolvency of the plan and the associated benefit cuts. Because a loan could ensure that participants will receive at least a portion of their benefits for a longer period of time, it could have a positive impact on maintaining participant support for the plan, which in turn can help make the plan more financially stable.
Similarly, employers considering withdrawing from a distressed plan before insolvency (so as to avoid any higher withdrawal liability assessments that might occur when the last employers leave) might choose to remain if the plan receives a low-interest loan, because the possibility of near-term insolvency would be reduced or eliminated. Maintaining the base of contributing employers could help stabilize the plan and further increase its prospects for recovery.
The PBGC, which currently expects to have requests for financial assistance from multiemployer plans far in excess of its available resources, could also benefit from a loan program as such a program would prevent, or at least significantly delay, need for financial assistance. When a multiemployer pension plan fails and the PBGC does not have sufficient resources to support the guaranteed benefit level, the taxpayers are likely to bear some or all of the cost, either directly through an effort to restore at least a portion of the pension benefits, or indirectly through social welfare programs. A loan program that allows plans to pay more benefits for longer would relieve some of that cost.

Potential Costs of Loan Programs5

There would be several potential sources of taxpayer costs in a program of government loans to troubled multiemployer plans. In many reform proposals, the interest rate to be charged to the borrowing plans would be substantially below the rates that apply to other financial transactions, potentially as low as 1 percent. When an institution lends money at an artificially low interest rate, it represents a cost to the lender, which is incurred even if the loan is repaid in full. If the government lends money to multiemployer plans using an interest rate that is below its cost of borrowing, there is a cost to taxpayers to subsidize the interest rate.
While the objective of a well-constructed loan program is for plans to repay the loans in full and maintain solvency thereafter, risks will remain. The ability of a plan to repay a loan and maintain solvency will depend on (a) future investment returns that can be subject to uncertainty; (b) the ability and willingness of employers and active plan participants to continue to contribute to the plan; and (c) potentially, additional financial assistance from the government in conjunction with a loan. If the government makes a loan to a distressed multiemployer plan, and that plan ultimately fails to repay some or all of that loan, then the amount that is not repaid is a cost to the government. While the default cost can be estimated when the loans are originated, it is likely to be subject to significantly greater uncertainty than the interest subsidy cost. In estimating default costs, it may be helpful to consider both deterministic projections that include a range of optimistic and pessimistic assumptions about future experience, and stochastic projections of future market conditions that assess the probability that plans will accumulate sufficient assets to repay the loans in full. We note that this type of analysis is currently part of the application process for benefit suspensions under the Multiemployer Pension Reform Act of 2014 (MPRA).
The design features of a loan program can have a large impact on the size of the default risk. A loan program will presumably only make loans to plans that are projected to be able to repay them. These projections will involve assumptions about future financial market conditions, the demographic experience of the plan, and the future health of the unionized industry supporting the plan. If these projections use conservative assumptions, fewer plans will qualify for loans, and those that do will be in a stronger financial position, resulting in a lower default risk than if more optimistic assumptions are used. However, the use of more conservative assumptions in the application process will produce larger loan amounts (because the amounts deemed necessary to avoid insolvency will be larger) or additional government financial assistance in conjunction with the loans, and may cause many plans to be ineligible for assistance from the loan program.

Key Projection Assumptions Relating to Multiemployer Loan Proposals

When evaluating whether a loan program is likely to return distressed multiemployer plans to long-term solvency, it is necessary to make several assumptions about future events. These include:
  • Investment returns on current plan assets
  • Investment returns on loan proceeds
  • Demographic experience of the plan population, including retirement patterns, mortality rates, disability experience, and employee turnover
  • Projected employer contributions into the plan; significant factors that affect these contributions are negotiated contribution rates, employment levels, and employer retention experience
  • Withdrawal liability collection experience
  • Administrative expenses paid from plan assets
The actual experience of a loan program will almost certainly deviate from the assumptions used to design the program and to assess applications, and the actual cost and level of the success of the program will be more or less than expected based on whether actual experience is more or less favorable than was expected.

As policymakers examine the benefits, costs, and risks of a proposed loan program, they will need analysis that incorporates carefully chosen assumptions, and will also benefit from stress testing and stochastic modeling that illustrates how the outcomes of the program would change if actual experience is different from what was assumed in a single deterministic projection.

A longer loan term may increase the default risk as projections far into the future involve greater uncertainty than shorter-term projections. If a loan program restricts the investment of loan proceeds to relatively low-risk investment strategies, it would significantly decrease the likelihood of defaults, but could require larger loan amounts to achieve the same expected benefits when compared to a program that allows more investment risk to be taken. It is important to note that the same design features that increase risk will also tend to increase the potential benefits of loans. As with most financial structures, there is an inevitable relationship between risk and potential reward.
Like all organizations, the federal government has limited resources. A loan program is likely to require large cash outlays to troubled multiemployer plans in the early years of the program. Unless the government entity responsible for making the loans has sufficient cash available, it must find a way to generate the additional resources to fund the principal amounts of the loans. The federal government could raise the necessary cash in several ways, including levying a new tax on the multiemployer system’s stakeholders, borrowing additional money through the Treasury Department, or through savings by canceling or delaying other programs and projects.
Alternatively, funding for the loans could come from the private sector, with the government’s role being primarily to back the loans in the event of default, and to provide oversight to the administration of the loans. While private-sector loans may eliminate or greatly reduce the upfront financing costs to the federal government, it does not substantially change the default risk if the government ultimately backs the loans.

Consequences of Failure to Repay Loans

The loan-based proposals currently under consideration generally require interest-only payments during some or all of the term of the loan, with the repayment of principal deferred to either the later years of the term, or to the final payment date. While anticipated default experience can be estimated when the loans are issued (using one or more deterministic projections, or using stochastic projections that measure the probability of failures), it is likely that the actual default experience of a loan program will be largely unknown for a long period of time. Monitoring the financial health of the receiving plans during the loan period may help assure they are on track to repay the loans or allow time to implement other strategies under the loan program to strengthen plan financials.

There will be two primary consequences of any loan defaults. As discussed in the previous section, the first consequence is that the default will represent a cost to the government. Secondly, if a plan is required to make scheduled loan repayments if it has sufficient assets to do so, a default will only occur when the plan assets are depleted. In such a case, only future contributions would be available to pay benefits, which would likely trigger a reduction in participant benefits to the PBGC guarantee level. It is important to note, however, that without a loan, this reduction in participant benefits to the PBGC guarantee level would likely occur years, or even decades, earlier.

A loan default is of course less desirable than full repayment, as it represents a combination of higher taxpayer cost and/or a reduced level of participant benefits. The situation is not as clear, however, when the default is compared to a scenario in which the loan was not made in the first place. Even a defaulted loan extends the solvency of the plan, which could significantly delay reductions to participants’ benefits. Viewed more holistically, loans will preserve benefit levels and delay the failure of plans, thereby generating savings to the PBGC and social safety net programs, which could offset a portion of the government’s costs associated with a default.


Loans may be an effective tool for assisting troubled multiemployer pension plans and the PBGC. An infusion of cash in the short term can give a plan more time to rebuild assets through employer contributions and investment returns, potentially leading to long-term solvency. However, loans also come with risk and may not necessarily deliver as planned.

In assessing a potential loan program, it may be helpful to consider how the potential benefits and risks are shared among the various stakeholders in the multiemployer system. The associated expenditure of governmental resources would compete with other immediate demands for those assets, even if the loans are ultimately repaid in full. Structuring the terms of the loan carefully and having a full understanding the level of risk associated with a program are critical to achieving the desired outcome of any loan program.



1. The PBGC’s projected insolvency reflects its estimate of the population of plans that are likely to run out of money in coming years and the financial resources PBGC will expend to provide benefits guaranteed to those plans under current law (before the additional reductions that occur when PBGC subsequently runs out of money).
2. Under current law, PBGC financial assistance to insolvent multiemployer plans is in the form of loans, though as a practical matter the loans are not expected to be repaid.
3. In a perfectly functioning market, a borrower and a lender would negotiate the terms of a loan. The agreed-upon interest rate would sufficiently compensate the lender for the opportunity cost of its capital and for the risk of default. The borrower would consider the rate affordable and worthwhile to access the needed capital. In the case of troubled plans, however, it is unlikely that such an agreement could be reached. Lenders would demand very high interest rates from troubled multiemployer plans facing significant challenges. The plans would not be in a position to assume the payments associated with such rates. The proposals considered therefore involve subsidized rates less than what would be available in the absence of such a program.
4. Because contributions to multiemployer plans are generated by collective bargaining, plans only receive contribution revenue (other than withdrawal liability payments) to the extent that employee representatives and employers negotiate contribution rates into the plans.
5. We note that the laws and regulations governing federal credit programs include specific requirements related to the estimated program costs for budget scoring purposes and to the approval criteria used when implementing a program. A discussion of these requirements is beyond the scope of this issue brief.