Taxpayer ledger · Public pensions
Idaho runs a single statewide retirement plan for nearly every public employee in the state, and it is 90.89% funded. A Colorado taxpayer is standing behind at least eleven separate plans, and they range from 56.63% funded to fully funded. This is what that looks like on one axis.
Sorted from worst funded to best. Colour marks the state. Idaho has exactly one bar because Idaho has exactly one plan: Boise, Twin Falls, Idaho Falls and the state itself all participate in PERSI.
The axis runs 0 to 100%. Bars are drawn to scale from the printed ratio, not estimated. Two Colorado entries describe the same fund on two different assumptions and are marked as such.
Colorado PERA, 2025 GASB Statements Nos. 67 and 68 for Division Trust Funds Valuation, Exhibit A. Idaho PERSI, Milliman GASB 67 and 68 Disclosure for Fiscal Year Ending June 30, 2025, p. 12. FPPA, Statewide Retirement Plan Actuarial Valuation January 1, 2026, Executive Summary. Denver and Aurora, each city’s Annual Comprehensive Financial Report pension notes.
A funded ratio is only as honest as the return it assumes. Assume a higher return and the same promises look cheaper today. Idaho reaches the better number while assuming less.
Idaho wins on the tougher yardstick. A lower discount rate makes the same pension promise look bigger and the plan look worse. Both rates are GASB 67 rates, net of investment expense, so they are directly comparable.
Put both systems on one rate and the gap widens. From each plan’s own published sensitivity table: at Idaho’s 6.55%, Colorado would be roughly 70% funded. At Colorado’s 7.25%, Idaho would be roughly 98%.
Two things that cut the other way, and belong in any honest version.
Idaho’s liability assumes only the 1.00% guaranteed retiree increase. Its actuary puts a 2.00% assumption at about $2.6 billion more in liability, which would pull PERSI’s ratio down materially. This is the same COLA-assumption question that decides whether FPPA is 103.5% or 80.1% funded, and Idaho is on the favourable side of it.
And the discount rate is not the only assumption. On mortality, Colorado is the more conservative of the two: PERA adjusts most of its base tables downward while PERSI loads its rates up. "Idaho assumes less across the board" would be false.
Every ratio on the ladder is the GASB accounting measure. Each system also publishes a funding valuation, which smooths asset values and is what actually drives contribution decisions. For Idaho the two barely differ. For Colorado they differ a great deal, and the funding basis is the worse one.
| Plan | GASB 67 (market) | Funding valuation | Asset basis of that ratio |
|---|---|---|---|
| Colorado PERA, five divisions combined | 75.31% | 69.1% | smoothed (actuarial) |
| Colorado PERA, same, restated on market value | 75.31% | ~74.4% | market |
| Idaho PERSI Base Plan | 90.89% | 90.6% | market (PERSI does not smooth) |
Do not put 69.1% next to 90.6%. PERA smooths its asset values and PERSI does not, so that pairing sets a smoothed ratio against a market one and overstates the gap by roughly five points. PERA’s shortfall on its own smoothed funding basis is $30.07 billion against an actuarial value of $67.28 billion, and that number is real. It just is not the number to hold up beside Idaho’s.
Matched on market value the honest pair is about 74.4% against 90.6%, which is effectively the GASB pair. The GASB comparison at the top of this page is the fair one.
Two scope corrections that belong on every quotation of these figures.
PERA’s 75.31% is an aggregation its own actuary labels for informational purposes only, because "the assets of each trust fund are for the sole purpose of its members and cannot be used by another fund." There is no single legal pool that is 75.31% funded. The divisions run from 66.1% to 90.3% on the funding basis and they are not fungible.
Idaho’s 90.89% is the Base Plan only. Add the Firefighters’ Retirement Fund, which holds a net pension asset of $291.9 million, and the Judges’ plan, and PERSI system-wide is 92.04% funded with a net liability of about $2.14 billion.
A funded ratio is a stock. The contribution rate is the flow, and it is the number that shows up in a budget every year. This is the same benefit type, priced by two states.
Employer contribution as a percentage of covered payroll. Axis runs 0 to 30%.
The naive comparison is 1.81x and it is wrong. These two rates do not buy the same product. PERA is a substitute for Social Security, so a Colorado agency pays no 6.2% employer OASDI tax. PERSI is a supplement to it, and an Idaho agency pays both.
Counted like for like, an Idaho state agency spends 18.16% of payroll on retirement income (11.96% PERSI plus 6.20% Social Security) against Colorado’s 21.65%. That is 1.19 times, not 1.81. The 6.2% never appears in either state’s pension line, which is precisely why side-by-side rate tables mislead.
A further trim: about 0.52 points of Colorado’s 21.65% is a Health Care Trust Fund allocation, which is retiree health care rather than pension, leaving 21.13% for the pension itself.
What is unique to Colorado is the debt component. The AED and SAED are amortisation payments layered on top of the base rate, still in force, worth 10.00 points of payroll in the State Division. Idaho has no equivalent line because Idaho has no equivalent debt, and it is paying down what it does have in 8.2 years against a 25-year statutory ceiling.
So the fair summary is not that Colorado pays nearly twice as much for a pension. It is that Colorado pays modestly more in total and roughly half of the gap is servicing a debt rather than buying a benefit.
Both states have a lever for when the numbers go wrong. They point in different directions.
Colorado’s direct distribution to PERA is $225 million a year by statute. It was suspended in full in 2020 by HB 20-1379, then repaid through a $380 million warrant under HB 22-1029 and a further $14.6 million under SB 23-056. A separate $500 million sits with PERA under SB 25-310.
PERA’s assumed return has been cut repeatedly, from 8.75% in 2002 to 7.25% today. Each cut makes the same promise cost more. The 2025 investment return was 14.1%, and the funded ratio still sits where it sits.
A raw dollar shortfall says nothing without the payroll behind it. Net pension liability measured against one year of covered payroll is the honest depth gauge.
| Plan | Funded | Net liability | As % of payroll |
|---|---|---|---|
| Colorado PERA, School Division | 71.71% | $15.60 bn | 222.18% |
| Colorado PERA, State Division | 72.61% | $8.27 bn | 203.12% |
| Colorado PERA, all divisions | 75.31% | $24.19 bn | 184.49% |
| Colorado PERA, Local Government | 96.32% | $0.25 bn | 25.56% |
| Idaho PERSI Base Plan | 90.89% | $2.42 bn | 50.69% |
Against a single year of payroll, PERA’s net pension liability is about 3.6 times as large as the PERSI Base Plan’s: 184.49% against 50.69%. (Not "3.64 times deeper", which would mean 4.64 times as large.) Both systems define covered payroll the same way, as the compensation contributions are based on.
But read the fourth row before drawing the obvious conclusion. PERA’s Local Government Division is 96.32% funded, with a hole worth about three months of payroll. Colorado’s $24 billion problem is a state and school problem. It is not, on PERA’s own numbers, a municipal one.
And neither figure is the whole state. Colorado’s excludes FPPA; Idaho’s excludes the Firefighters’ Retirement Fund and the Judges’ plan. Those two omissions push in opposite directions, so this is a comparison of two named plans, not of two states.
Colorado’s statewide fire and police plan is widely described as fully funded. That is true of exactly one version of the promise.
All three ratios are on the actuarial value of assets, which is 94.9% of market value at 1 January 2026. On market value the base-benefit ratio would be 109.0% and the COLA reserve $447 m, supporting a 0.77% increase rather than 0.30%.
On its actuarial basis the fund can support a long-term increase of 0.30%. Board policy allows a market value basis up to 0.50%, which produced a 2026 recommendation of 0.50%. The increase actually granted on 1 October 2025 was 0.27%, plus a one-off payment worth 2.63% of the annual benefit.
So the gap is real, and it is being closed by retirees absorbing sub-1% increases rather than by taxpayers funding a bailout. Whether that counts as a solved problem depends entirely on which side of the promise you are standing on.
FPPA, Statewide Retirement Plan Actuarial Valuation January 1, 2026, Executive Summary, p. 3; and Annual Update to the Pension Review Commission 2025, p. 9.
Colorado fire and police employers do not get to hold their contribution flat. The increase is written into statute by House Bill 20-1044 and steps up half a point every year.
| Effective | Member | Employer | Combined |
|---|---|---|---|
| Originally, at inception | 8.0% | 8.0% | 16.0% |
| 2025 | 12.0% | 10.5% | 22.5% |
| 2026 | 12.0% | 11.0% | 23.0% |
| 2030 and beyond, fixed | 12.0% | 13.0% | 25.0% |
The escalation is real, but only most of it is for an unchanged promise. The combined rate goes 16.0% to 24.0% by 2028, a 50% increase, and that part buys nothing new. The final point, taking it to 25.0% in 2029 and 2030, pays for a new benefit, the Rule of 80, which the actuary priced at about 1% of payroll. Calling the whole 56% the price of an unchanged benefit overstates it.
Nor was it one decision. Members voted in 2014 to raise their own share from 8.0% to 12.0%, phased over 2015 to 2022. HB20-1044 then raised only the employer share, 8.0% to 13.0%, to restore parity after the plan lowered its investment and longevity assumptions.
For contrast, an Idaho city pays PERSI 11.96% of pay for general employees and 13.98% for police and firefighters, rates set by the PERSI board rather than escalated by statute. Boise, Twin Falls and Idaho Falls all pay exactly the same rate, because they are all in the same plan.
FPPA, Annual Update to the Pension Review Commission 2025, Appendix B, pp. 20–21. PERSI employer rates effective 1 April 2025, confirmed in the Boise and Twin Falls budget books.
A 1977 study by the Colorado General Assembly found the state’s local police and fire pension plans carried more than $500 million in unfunded liabilities. Reform bills in 1978 and 1979 closed every one of them to anyone hired after 8 April 1978 and created the statewide plan in their place.
Those closed plans are the two worst bars on the ladder, at 56.63% and 58.44%. They are also finite, shrinking, and nearly fifty years past their last new member. The state funded the assisted old hire plans through 2013, when it made a final payment. There is no ongoing state old hire subsidy, and the statewide plan that replaced them has run on member and employer money alone since inception.
One correction worth making precisely, because it is easy to overstate. The state’s obligation to the old hire plans is finished and the Statewide Defined Benefit Plan has never taken state money. But Colorado does still pay FPPA $2,050,000 every year toward the Statewide Death and Disability Plan, after two catch-up payments of $6.65 million, against a shortfall recorded at $32.9 million as of 1 January 2026. That plan is 98.8% funded. So "no state assistance" is true of the pension and not true of the whole relationship.
Colorado has done this before, and it worked. The plan built in 1978 is the one sitting at the top of the ladder. The unfunded liability that remains is in PERA’s state and school divisions, not in the fire and police system that gets the attention.
The one number that says where FPPA actually stands: the blended statutory contribution is 22.81% of payroll, and sustaining a 2.50% annual increase would take 29.93%. A gap of 7.12 points of payroll, currently absorbed by retirees rather than by employers. Assume no future increase at all and the plan is 103.5% funded with a $172.4 million surplus.
Idaho PERSI members participate in Social Security. Most Colorado PERA members do not — PERA states plainly that "most PERA members do not contribute to Social Security."
So an Idaho public employee holds two independent retirement claims, one of them federal. A Colorado teacher or state worker frequently holds one, on a plan that is 69.1% funded on its own funding basis, with a retiree increase capped at 1.00% a year. The funded ratio is a measure of the taxpayer’s debt. This is a measure of the worker’s exposure, and it is the part that rarely makes the chart.