STUMP » Articles » Happy Halloween! Some Scary Public Pension Projections » 31 October 2017, 12:09

Do you want to see something really scary?

I have updated my public pension cash flow projection spreadsheet. You can get a copy of the latest version here.

I ran the 170 plans with updated data up to (possibly) fiscal year 2016, using data from the Public Plans Database.

The three scenarios:

• A baseline, where all the assumptions are 0 (contributions, expenses, and benefits all remain constant; no investment return) – all plans are given the same assumptions
• A run where the assumptions are set to the 10-year average growth rate and return for each individual plan
• A run where the assumptions are set to the 5-year average growth rate

Those last two differ based on the plan, obviously.

After weeding out plans that lacked data for the 5- and 10-year averages, and removing all the plans that survived past year 2040 for the baseline, I was left with 52 plans.

Here is the ranked results (ranked on the baseline result).

RANK OF HORROR

Not many surprises at the top of the list.

The Portland Fire and Police Disability Retirement Fund is an odd case, but looking at the data, it looks like it’s a pay-as-it-goes fund. I tried looking for this plan in the interactive data browser and the state data… and it doesn’t show up. It’s just in the downloadable data.

PUBLIC PENSIONS IN DANGER: WORST ARE NO SURPRISE

But let’s look at the others:
Kentucky ERS and Chicago MEABF I’ve projected out before:

• Watching the Money Run Out: A Simulation with a Chicago Pension
• Geeking Out: Testing Chicago MEABF to Destruction
• Geeking Out: Testing Kentucky ERS to Death

Both Kentucky ERS and Chicago MEABF have been grossly underfunded for decades, greatly due to deliberate underfunding.

SUSPICIOUS ONES

Milwaukee City ERS is one I’ve not covered before. Let’s take a look!

Ah, I see the problem is the huge non-benefit growth rate. Still, running out of cash in 2023 as a baseline is very bad. According to this page of the Public Plans Database, this pension is 97% funded… but it doesn’t match up with what I’m seeing re: cash flows.

Looking again, I see the non-benefit expense numbers are HYUUUUGE compared to benefit amounts, and perhaps there was a mistake in the database.

Let’s look at something suspicious in a different way: New Jersey Teachers. It fails in 2024 on the baseline and in 2025 when matching 10-year trends, but seems to do just fine matching the 5-year trend.

The stats tell the story:

Well yeah, if you increase contributions by 20% per year, you’re going to be able to sustain the pensions. The question is whether those contributions are sustainable. (spoiler: they’re not.)

Anyway, have a little fun and try it out yourself. These are very simplistic projections, but based on cash flows and initial market value of assets for the plans.

Happy Halloween!