I’ve said it before: Illinois cannot continue kicking the can on its pensions. I’ve said repeatedly (and fellow soapbox-stander Adam Schuster repeated the message in the Chicago Tribune last week) that the state’s spending on pensions is costly, not only in terms of dollar amounts spent – over 25% of the state budget – but in terms of spending that ought to be spent on needs such as services for those with disabilities, the unemployed, students, and so on, but isn’t, because it’s going to pensions instead.
And I’ve lamented that Gov. Pritzker just doesn’t get it, most recently in an interview in which he said that stretching out the funding target was not off the table, and seemed to suggest interest in the CTBA “solution” involving Pension Obligation Bonds and a reduced funding target of 70%.
All of which prodded me to look at the largest of the state pension plans, the Teachers’ Retirement System, in more detail. This plan is, like the other main plans, about 40% funded, and its unfunded liability is 60% of the total. It is also notable in having cut the benefits for the Tier 2 workers most sharply, so that, given the valuation’s assumptions and funding method, the benefit that they accrue is, in total, less than their required contributions. (Direct state employees mostly participate in Social Security so that their pension is a supplement, not a replacement, and the State Universities System has quirks of its own, including a pure Defined Contribution option for new hires.)
And I plowed through the data in order to assess, how reasonable are these funding targets? And what can go wrong?
Now, current legislation dictates that the state must make contributions as a level percentage of pensionable pay, to reach a target of 90% funded in 2045, but the contribution schedule is actually slightly lower than this target now, and jumps up later, so in order to do my own calculations, I replicated (approximately and in a more smoothed manner) the contribution projections. There are also some finer details which I simplified but the basic concepts below should still be clear.
Chart One: what happens if there’s a run of bad luck?
I calculated three hypotheticals, in each case keeping the dollar amount of contributions unchanged:
In the first, I projected asset growth in the case that the 7% assumption turns out to be 6% instead. In the second, I assumed a long-term 6.5% return but also adjusted the liabilities to reflect the changes that’d be needed for a new valuation interest rate. And in the third, I kept the assumptions unchanged at 7%, but dropped the assets immediately by 33% to reflect a market crash.
Not good, eh? A drop to a 6.5% asset return/valuation interest rate assumption is not just reasonable but likely, and this keeps the plan from reaching anything close to its 90% objective — instead, only 62.5%.
Chart Two: Bad luck with a 70% target
The colors shift because I’ve excluded the original target to illustrate the fact that setting a 70% target produces an even greater risk of poor funded status. (Note that the CTBA’s projections look different because of their POB proposal, and recall that I am, well, anti-POB, especially with respect to anyone who promotes pension obligation bonds as a form of “refinancing”.) In the wholly-realistic 6.5% discount rate scenario, it barely budgets from the present funded status, and maxes out at 46.6%.
Chart Three: The effect on the unfunded pension liability, with a 90% target and “bad luck.”
Note that while the funded status for the two low-return scenarios was similar, they diverge in terms of absolute pension underfunding. In the actuary’s current projection, the unfunded liability peaks in 2029 at $86 billion. With a discount rate drop, it peaks at $110 billion in 2037.
Chart Four: Unfunded pension liability with a 70% funding target.
Yes, you see that right, in the scenario in which the plan targets a 70% funded level, and doesn’t adjust its contributions afterwards, the unfunded liability reaches $139 billion for the teachers’ plan alone in 2045, and has not yet peaked and begun to decline.
Are you starting to see my point?
Even with a 90% funding target, the state is at risk of liabilities growing year after year — or (not modeled here) even more money being spent on pension funding rather than daycare subsidies for low-income families, for example. And with a 70% target, of course, those numbers are even worse.
But wait, there’s more!
The final two of the six charts are an attempt to approximate the impact on the pension liabilities of a “repair” of the Tier 2 benefits. As it is, there is effectively no employer contribution. I modeled the impact of a 6% employer contribution (less than a Social Security contribution) and a 9% contribution, such as in a notional account plan (a “cash balance plan” in the private sector), along with the removal of the pay cap that is lowered every year in inflation-adjusted terms. This is approximative, but is a very conservative scenario.
Chart Five: Tier 2 “repairs” with a 90% target
Chart Six: Tier 2 “repairs” with a 70% target
To be clear, these are based on the baseline valuation, with “original” representing the original contributions for the 90% target.
Are these six charts enough? It was tempting to produce “worst-case” scenarios but those would be too easy to brush aside, and even these hypotheticals, intentionally made realistic, should be enough to make it clear that it’s not remotely reasonable to shrug off pension debt as something for the next generation to deal with.
Added next-day comments:
I’m often asked, “how long until the plan runs out of money?” And, in fact, that’s a fair quesiton to ask, with respect for the even-worse-funded Chicago plans, where they’re still making their way up the funding “ramp” and are so poorly funded that if they chicken out and markets have poor returns, they will become insolvent. In an article in September 2019, I explained some of the math; to take one example, if the mayor replaces the scheduled “ramp” increases for the municipal plan with inflatoinary increases instead, the plan becomes insolvent in 2027, even without adding in any “bad luck” scenarios.
But there isn’t as dramatic an answer for the state plans — the 23% vs. 40% funded status ratios do make a difference. Nonetheless, I spent so much time typing up these numbers that I might as well make the most of them: the following table illustrates the insolvency year under the valuation assumptions, and a second scenario in which assets earn only 6% and there is a market crash.
If the state continues its current contribution level — which is, yes, 48% of pensionable payroll, and, yes, that’s a lot — the plan reaches 90% funding under current assumptions and avoids insolvency in the “bad luck” assumptions; however, its funded status is very shaky, staying 27% – 28% during the entire modeling period. And if the state were to drop its contribution down to significantly lower levels, then the plan would end up insolvent in either scenario.
Sadly, though, I doubt this table will persuade any of the politicians who need to be persuaded, many of whom would likely say, if off the record, “what’s wrong with insolvency anyway? We’re paying for the benefits our fathers and grandfathers promised; why not demand the same of our children and grandchildren?”
And one final comment: none of these scenarios take into account another likely issue in Illinois, that of declining population. A contribution of 48% of pensionable payroll directed almost entirely at paying off past debt, will grow to even higher levels, as a percentage of pensionable payroll, if the number of teachers declines over time in line with the number of residents. That’s not pretty, either.