Noticing renewed interest in my March 2010 blog post, I should point out that I made some (relatively minor) updates to the calculations during the process of revising the paper and preparing it for publication. The paper is available here:
Are State Public Pensions Sustainable? Why the Federal Government Should Worry About State Pension Liabilities, National Tax Journal 63(3), 2010.
The main update was to use asset values as of September 2009, which reflect a recovery in asset values after the crash, and to implement some refinements to the model for benefit flows.
Table 1 of the paper shows the updated state-by-state table. The median runout year is 2026, and 20 states only have assets to pay for benefits through 2025. Illinois goes to 2018 and New Jersey to 2019.
Let me also take this opportunity to comment on a response I have heard from some states: that they are already making changes that make their systems more sustainable.
The first thing to note is that since the runout analysis here only looks at benefits that have been promised up until today, policy measures that hope to postpone these runout dates by cutting benefits would have to do so for current employees and/or retirees. So measures like those in Colorado in Minnesota, which are battling to reduce cost of living adjustments (COLAs) across the board, would indeed keep the funds alive for a little longer. My rough calculations show that reducing COLAs by 2 full percentage points for everyone (including retirees) in systems that run out in the early 2020s would make runout dates 3 years later. The effects of COLA reductions would be more powerful for states with a longer horizon, as those states have a longer period of time to pay reduced benefits before hitting a crisis.
Measures such as the introduction of a new tier of benefits for new hires (as was passed in Illinois in April) have no effect if the runout date is before the time that those new hires will retire. Even a “hard freeze” of defined benefit pensions and putting all future work on a 401(k)-type arrangement would still leave the legacy liabilities in the same situation as I document in my paper. The paper evaluates the adequacy of existing assets to pay for already-promised benefits.
Another set of changes that could postpone runout dates are contribution increases to pay for unfunded legacy liabilities. I would be skeptical of any state that argues they have solved their problems by pledging to increase payments in the future enough to compensate for the mistakes of the past. For example, Illinois and New Jersey have contribution requirements which at some point they promised they would meet. But Illinois is now paying them with borrowing and New Jersey is only paying a small fraction of the “required” amount (such as 6% of the $1.6 billion due to the New Jersey Teachers fund in 2009… better than 0%, I suppose). In states that have actually been meeting large contribution requirements, like New York, the question is now how they can continue to do in light of current budgetary problems.
What about raising contribution requirements for employees? Required employee contributions represent a more secure stream of revenue into the pension fund than proposed state contributions. The state can simply take the employee contributions out of employee pay rather than having to agree to raise taxes or cut spending on other programs. A number of states including Colorado, Iowa, Minnesota, Mississippi, Nebraska, New Mexico, Vermont, and Wyoming have increased employee pension contribution rates for current and future employees, while New Hampshire and Texas have increased them for new employees only.
Raising employee contributions forces younger workers, and especially new employees, to subsidize the older workers and the retired. For example, suppose a plan that is going to run out in 10 years increases contributions on existing employees in such a way that the runout date is now 20 years away. In that circumstance, employee money today is being used to pay for an additional 10 years of retirement payments. Many of the public sector employees who are now paying the higher contribution rate still face an unsustainable system, and a substantial group of them will be paying higher contributions now without seeing any personal benefit. In effect, the younger public employees are being taxed to pay for current retirees for 10 more years.
The solution to the state pension crisis must proceed in two steps. First, the growth of the unfunded liabilities must be stopped. The best way to do that is to pay future retirement benefits the way it is done in the private sector: though defined contribution plans and social security as opposed to the dangerous promises of underfunded defined benefit plans. Second, states must determine how much of the unfunded legacy liabilities will be paid for by the various parties, which include current taxpayers, future taxpayers, state retirees, current state employees, future state employees, and (it must be mentioned) bondholders and other creditors of the state.