Kicking pension problems into the future is popular with politicians, enabling them to make promises and let voters worry later about borrowing costs.
But large, unfunded state pension liabilities are a costly problem—and the cost is already reflected in current bond prices, research by Chicago Booth PhD candidate Chuck Boyer suggests. The public pension funding crisis is not merely about future insolvency,” he writes. “Future obligations are having an effect on debt spreads right now.”
To many Americans, it may seem unimaginable that states would fail to fully pay pensions promised to teachers, firefighters, and other public-service workers. It has been almost 90 years since the last state default: during the Great Depression, Arkansas owed over $160 million to debt payments, which was nearly half of the state’s annual revenue (and equivalent to roughly $3 billion in 2019 dollars). The debt was restructured and “debtholders were eventually made whole,” Boyer writes in recounting this history.
However, pension obligations are mounting in many states, and officials are struggling to cut costs and raise taxes to pay what is owed. And he argues that the effects can be seen in the $3.8 trillion capital market for US municipal bonds, which includes bonds issued by 50,000 state and local governments.
When a company defaults, there is a clear legal framework for who gets paid back first. This isn’t the case for states, however, as there is no such legal structure, nor much precedent. The markets’ expectations, then, are built into bond prices. Bondholders, wary of how a default could play out, demand a premium.
Using annual fiscal reports released by state governments, Boyer looked at the ratio of unfunded pension liabilities to GDP from 2002 to 2016 and estimates that every 1-standard-deviation increase is associated with a 27–32 basis-point increase in bond spreads over the Treasury rate, up to a fifth of the average total spread. Unfunded pensions cost US states more than $2 billion in lost bond-issuance proceeds in 2016, he calculates, adding that he considers that a conservative estimate.