Aharon Friedman has worked on tax and pension policy issues for more than a decade. He served as senior adviser to the assistant Treasury secretary for tax policy and as senior tax counsel to the House Ways and Means Committee.
Senator Sanders’s Budget Committee made a little-noticed change to the House budget bill that may lead to the termination of union multiemployer pension plans receiving a massive taxpayer bailout. The change removes a “Hotel California” provision in the House bill calculating the withdrawal liability exiting employers must pay without counting the bailout. This will dramatically lower or eliminate the cost for withdrawing employers, possibly leading them to terminate the plans. Even under the questionable assertion that taxpayers should be forced to bailout pension promises they did not make, Treasury should send checks to participants while shutting down the plans. Instead, the bill uses byzantine mechanics to preserve the plans that may backfire unpredictably, leaving plans to resemble Weekend at Bernie’s, astronomically higher costs, and years of litigation.
The 1980 MPPAA requires an employer exiting a plan to pay “withdrawal liability” based on the employer’s share of underfunding, including that attributable to bankrupt employers. However, MPPAA limits this joint liability by capping withdrawal liability at 20 annual payments equal to the employer’s past annual contributions.
Political unwillingness to hold employers responsible for any of the $673 billion in existing multiemployer underfunding is a key obstacle to reform. On the one hand, plans measure new promises at forty cents on the dollar, a practice facilitated by convincing Congress to permit “less stringent” funding rules for multiemployer plans because all employers would have joint liability. Employers should have known the level of contributions they were making for the benefits they were promising was too good to be true. On the other hand, most employers have little or no say in running the plans and forcing employers to foot the entire bill at this point is neither fair nor viable. Contributions based on the MPPAA compromise should be the starting point to balance the burden on employers against benefit cuts and/or taxpayer bailouts.
Instead, the House bill fully covers participants by sticking the whole bill on taxpayers, with no required contributions from employers. But there is one hitch for employers – if they exit within 15 years of the bailout they would be responsible for plan underfunding as if the bailout hadn’t occurred. This makes exiting much more expensive than remaining given that the plans allow employers to contribute less then new pension costs. [A similar provision in 2014 MPRA prevents benefit cuts from effectively reducing withdrawal liability.]
Under the Senate changes, plans receiving bailouts will be dramatically better funded for purposes of calculating withdrawal liability, allowing employers to wash their hands of the union pension mess for good at low or no cost and terminate the plans. The cost to employers is further reduced because the bailout measures pension promises covered by taxpayers as costing much more than as measured by the plans. If the end result is plan termination, laundering taxpayer funds through skeleton plans instead of paying retirees directly serves no constructive purpose. The impact will differ across plans, illustrating other flaws in the existing rules and the bill.
Many multiemployer rules are based on projections by plans of future unknowable events, instead of current conditions. Similarly, the bill bases the amount of taxpayer assistance on a plan’s projections about unknowable future events such as future contributions and pension accruals, instead of just looking at existing underfunding. Like under current law, plans have every incentive to manipulate the projections.
The bill half-heartedly tries to prevent game-playing by requiring plans to use prior 2020 assumptions in making these projections unless “unreasonable.” But in evaluating applications for taxpayer assistance, the government must accept plan changes to these projections unless “unreasonable.” And the usefulness of past assumptions for future projections is questionable given the vastly changed circumstances created by the bailout, especially given the Senate change reducing or even eliminating the exit cost for employers.
Despite the bailout having an $86 billion CBO score, the amount of taxpayer assistance, which may run into the trillions if exploited by plans, is not actually capped. And the score is based on a stringent interpretation of ambiguous statutory text that plans and the Biden Administration may read as providing tens of billions more. It seems doubtful the Biden Administration would attempt to challenge questionable assumptions by plans seeking to increase their bailout. If employers terminate plans after the bailout, the assistance would be based on phantom accruals, possibly leaving plans with more taxpayer funds than liabilities. The complexity, uncertainty and immense amounts at stake will likely result in years of litigation.
But termination is far from certain. Some employers may choose to keep riding the tiger, attracted by the ability to promise benefits on the cheap and the expectation of future bailouts should anything go wrong. In addition, unions generally have a veto over employers withdrawing, and union bosses may attempt to convince their members to strike over pension withdrawals. Courts are still split over liability valuation for withdrawal liability purposes under MPPAA, which may give some employers pause. And the Biden Administration may attempt to administratively overwrite the statute to effectively block withdrawals.
Chairman Sanders completely letting big companies off the hook for billions of dollars of promises they made to employees puts an exclamation point on the bailout’s immense flaws and failure to provide the reforms needed to protect workers, retirees and taxpayers. But if the House version providing bailouts with no reforms is the baseline, facilitating termination of plans instead of rehabilitating them to make new promises payable through 2051 funded by taxpayers and afterwards by no one might actually be an improvement.
Views of our Guest Bloggers are theirs alone, and not of the Pension Research Council, the Wharton School, or the University of Pennsylvania.
