I pulled the most recent adopted budget for ten cities across Colorado and Idaho, expecting to sort them into the ones with a pension problem and the ones without. Denver, Aurora, Parker, Greenwood Village, Centennial and Breckenridge in Colorado. Boise, Twin Falls and Idaho Falls in Idaho.
The sort worked. It just did not predict anything.
The finding that ruined the thesis
Centennial, Colorado contracts its policing from the Arapahoe County Sheriff's Office. It does not employ a single sworn officer. Its own employees, all 92 of them, are in a 401(a) money purchase plan. Its audited financial statements contain no net pension liability and no OPEB liability at all.
By every rule of thumb about municipal finance, this city should be fine.
Its sheriff's contract went from $34,179,580 in 2023 to $43,589,360 in 2026. That is 27.5% in three years, it consumed 70.2% of all general fund growth, and it now equals 58.9% of general fund operating expenditures. The city prints the same figure in its own budget book as "Public Safety (59%)".
So the most pension free city in the study is the most squeezed one in it.
That result reframes everything else. The pension is the tail. The payroll is the dog. Wages, staffing levels and the market price of a sworn officer drive a municipal budget whether the city employs that officer or rents one from the county. Centennial moved the liability off its balance sheet and kept every dollar of the cost.
One precision, since "Centennial has zero debt" is the city's own framing and its audit says something narrower. The consolidated statements show $8,184,871 of outstanding debt at the end of 2025, nearly all general obligation bonds of three general improvement districts the audit folds into the primary government. City only direct debt is $450,982. The zero pension and zero OPEB findings hold exactly as stated.
What the public safety share actually looks like
| City | Police and fire, share of general fund |
|---|---|
| Centennial, CO | 58.9% (sheriff's contract) |
| Aurora, CO | 58.8% (operating basis) |
| Idaho Falls, ID | 55.1% |
| Twin Falls, ID | 51.7% |
| Boise, ID | 50.9% |
| Parker, CO | 32.4% (police only) |
| Greenwood Village, CO | 31.7% (police only) |
| Denver, CO | 26.4% |
| Breckenridge, CO | 17.3% (police only) |
Do not read that as a league table. The denominators are not the same kind of thing, and the differences are structural rather than managerial.
Denver is a consolidated city and county. Its general fund also carries the jail, the courts, the district attorney, human services and public health. In Idaho those sit with the county on a separate levy. Denver's denominator is simply larger in kind.
Parker, Greenwood Village, Centennial and Breckenridge have no municipal fire department at all. Fire is a separate taxing district that bills residents directly and appears nowhere in the city budget. Their numbers cover police only and understate what the household pays.
The honest reading is narrow and still useful: a plain city's general fund is mostly police and fire, somewhere around half, and the cities that look cheaper are usually the ones that moved a service out of the budget rather than out of existence.
Aurora, and a number I had to take back
Aurora publishes four consecutive comparable years, which makes it the best case study in either state. It is also where I got it wrong first.
Measured against total general fund uses, police and fire look like they took 83.9% of the growth between 2023 and 2026 and climbed from 45.7% to 49.4% of the fund. I published that. It is arithmetically correct and materially misleading, because total uses include transfers out to other funds, and those transfers fell $19.0 million as a capital transfer tied to use tax receipts shrank. A falling denominator was doing much of the work.
On operating spending, which is what the city actually spends running itself, the picture is this. Police and fire were about 58.3% of general fund operating spending in 2023 and are about 58.8% now. They took 61.8% of the operating growth. Police and fire together rose 19.6%, which is solid on any basis, against operating growth of 18.53%.
The corrected story is worse, not better. A share that barely moved because it was already at 58% is a harder problem than a share climbing through the forties. Aurora disclosed a 2026 shortfall of about $21.2 million and closed part of it with $10,205,516 of cuts.
Denver has the worst funded pension in the study
Not PERA. Denver's own plan.
The Denver Employees Retirement Plan is 59.97% funded, with the city's proportionate share of the net pension liability at $1.62 billion at the 31 December 2024 measurement date. Denver paid in $161,613,000 during 2025. That is a lower funded ratio than any PERA division, including the School Division at 71.71%.
Meanwhile Denver's general fund appropriation fell from $1,773,899,000 in 2024 to $1,663,080,000 in 2026, down 6.2% over two years, while police and fire rose 1.7% and their share went from 25.0% to 26.4%.
Denver's police and firefighters are covered through the statewide Fire and Police Pension Association, including two closed Denver old hire funds. Its civilian plan is the one carrying the weight.
Aurora runs five retirement plans, and two are in trouble
This one surprised me enough that I had it checked twice. Aurora's sworn police are not in the FPPA statewide defined benefit plan. New hires go into a money purchase plan, which is defined contribution, with the city putting in 12.0%.
What Aurora actually operates:
| Plan | Funded ratio |
|---|---|
| FPPA Old Hire Fire, closed | 56.63% |
| FPPA Old Hire Police, closed | 58.44% |
| General Employees' Retirement Plan | 88.22% |
| FPPA Statewide Retirement Plan, fire sworn | 100.00% |
Across all its defined benefit plans Aurora carries a $191,212,991 net pension liability, plus $20,262,694 of retiree health liability. The two worst are closed plans, shrinking, with no new members since 1978. They are a finite problem being paid off, which is a genuinely different thing from an open plan drifting.
The cities that structurally cannot have this problem
Three Colorado municipalities have essentially removed defined benefit exposure.
Parker puts every full time employee who is not a sworn officer into a 401(a) defined contribution plan, contributing $3,789,061 in 2025 against $961,201 to FPPA for its police. Just under 80% of Parker's retirement spending carries no unfunded liability by construction, and the town's proportionate share of the FPPA net pension liability is $0.
Greenwood Village has no defined benefit plan at all. Both city plans are defined contribution, so no funded ratio exists to report. Employer contributions were $1,955,714 to the 401(a) and $509,155 in 401(k) match. Its only defined benefit exposure is the FPPA Statewide Death and Disability Plan, $257,099 a year, which is 98.8% funded.
Breckenridge runs a single employer defined contribution plan, $1,444,274 in 2026.
These are real structural wins and they are permanent. A defined contribution plan cannot accrue an unfunded liability. But look at Centennial again before concluding that this solves the budget problem, because Centennial did the same thing and is the most squeezed city here.
Breckenridge does not fit any of it
The resort town breaks the pattern completely, and it is the most interesting budget in the set.
Police are $5,801,816, just 17.3% of the general fund, the lowest share in the study. The Workforce Housing Fund is $22,863,450, with a $30,640,000 capital plan behind it. Breckenridge budgets 3.9 times more for housing its workforce than for policing it.
The revenue side explains why. A permanent population of roughly 5,078 funds a town built for a far larger peak, out of a 2.5% sales tax raising $33,505,000, a 3.4% lodging tax raising $5,508,000, a 1% real estate transfer tax raising $6,000,000 and a 4.5% tax on lift tickets raising $4,535,000. Property tax contributes $5,367,034. Only 30.9% of its 7,364 housing units are occupied year round.
Its general fund actually fell, from $34,820,801 to $33,463,168. The pressure is not pensions and not public safety. It is the cost of housing the people who staff the town, and it is the single largest discretionary line in the budget.
Meanwhile the fire district serving Breckenridge went from 9.531 to 14.019 mills in one year, a 47% increase, pushing the total levy on a Breckenridge property from 53.412 to 58.061 mills. Nothing the town council did explains that.
Two completely different ways to fund a city
This is the cleanest divide in the whole study, and it is not about competence.
Boise raises 62.1% of its general fund from property tax. Parker raises 4.4%.
Colorado suburbs are sales tax entities that collect a little property tax on the side. Parker gets 69.7% of general fund revenue from sales tax, Centennial 60%, Greenwood Village 59.2% of all revenue. Idaho cities are property tax entities: Twin Falls funds 53.2% of governmental funds that way.
That single fact drives most of what follows. It decides who actually pays for the city, whether a recession hits the budget through consumer spending or through assessed values, and crucially whether a resident can see what they are buying.
Which is why the Colorado household cannot see the bill
Every parcel inside Parker town limits pays a common floor of 85.123 mills. The Town of Parker's share of that is 2.602 mills, about 3.1%. On a $500,000 home the town collects roughly $81.
South Metro Fire Rescue takes 12.250 mills, 4.7 times the town's own levy, entirely outside the town budget, with its own board and a 2026 appropriation of roughly $303 million. On 4 November 2025 its voters approved Ballot Issue 7A, adding 3.000 mills and about $50.7 million a year.
There is not even one Parker tax bill. The town contains 118 certified tax districts running from 85.123 to 210.035 mills, because metropolitan districts sit on top in many neighbourhoods.
Greenwood Village makes the point from the other direction. Its municipal levy has been 2.932 mills, identical every year from 2016 through 2025, roughly $185 per $1 million of home value. A resident could reasonably conclude their city has held costs flat for a decade. It has. It is also about 6% of their revenue and a rounding error on their tax bill.
A Colorado household can watch its town hold the line and still see its property tax climb, because the expensive part of local government is a district most people cannot name.
Idaho does the opposite. A city's property tax budget can rise 3% a year plus new construction, and that is the whole conversation. Boise took 2.7% for FY2027 and 3.0% the year before. Twin Falls took $984,689 as the 3% increase, $473,951 on growth, and recovered $342,816 of previously forgone authority. Idaho Falls has $5,527,161 of forgone balance banked, which is a deferred tax increase sitting on a shelf.
The Idaho bill is legible. The average Boise homeowner pays $1,614.75 to the city, up $9.39 even though the levy rate fell. The Twin Falls median home pays $1,247.79.
The hidden tax in Idaho Falls
One Idaho quirk deserves naming. Idaho Falls owns its electric utility. Idaho Falls Power transferred $4,347,810 out to the general fund in FY2025, up from $3,788,064 in FY2024 and $3,723,372 in FY2023.
That transfer is a tax on ratepayers that does not appear on anyone's tax bill and is not subject to the 3% cap. It is legal, it is common, and it is invisible.
So what actually separates them
Not pension design. Centennial settles that.
Whether the expensive services are inside or outside the budget. The cities with low public safety shares mostly moved fire to a district. The cost did not fall. It moved to a line item with less scrutiny and, in South Metro's and Red White and Blue's cases, rose sharply.
Whether the city can see its own trend. Aurora publishes four comparable years and can therefore prove police and fire took 61.8% of its growth. Several cities here cannot produce that series from their own documents.
Whether the revenue base matches the obligation. Idaho cities fund fixed public safety costs from property tax, which is stable. Colorado suburbs fund them from sales tax, which is not. That is fine until it isn't.
And whether anything is genuinely restrained. Twin Falls carries no general obligation debt, holds a Moody's Aa2 rating, and put a 0% health insurance premium increase in its FY2027 budget. Boise's health insurance fund rose 30.6% between adopted budgets over the same period. Colorado's state employee health line rose 20.3% in one year. That is the fastest moving cost in this entire study and almost nobody watches it.
If I had to name the best run budget here on the documents alone, it would be Twin Falls: no debt, a restrained health line, a legible property tax, and public safety at a share it can explain. If I had to name the one facing the hardest arithmetic, it would be Centennial, which did everything the textbook says and is watching a contract it does not control take seven of every ten new dollars.
I wrote up the state level comparison behind these cities separately, including why Colorado's pension gap is larger than it looks and why its employer contribution rate is smaller than it looks.
City by city
Each of these cities got its own write up, because the summary above flattens differences that matter.
Denver: DERP at 59.97% funded and a $1.62 billion liability, against a general fund that shrank 6.2% in two years.
Aurora: police and fire at 58.8% of operating spending, and why the widely repeated 83.9% growth figure is a shrinking denominator rather than a finding.
Parker: a $0 pension liability, and a town that collects 3.1% of its own residents' property tax bill.
Greenwood Village: a municipal mill levy unchanged every single year from 2016 through 2025, and what that does and does not tell you.
Centennial: no pension liability at all, and the most squeezed budget in the study.
Breckenridge: 3.9 times more budgeted for workforce housing than for policing.
Boise: 62.1% property tax funded, and why the levy rate fell while the bill went up.
Twin Falls: no general obligation debt, an Aa2 rating, and a 0% health premium increase.
Idaho Falls: a $4.35 million utility transfer that appears on nobody's tax bill.
Two structural pieces sit underneath all of them:
The fire districts nobody votes for: four of the six Colorado municipalities here have no fire department at all.
Idaho's property tax buy down: House Bill 292 moved school costs onto the state general fund, which is a transfer rather than a saving.
The charts
One line item ate the growth: the public safety shares above with their footnotes attached, the Aurora series on both bases, and the Centennial contract.
The city is a small part of the bill: the property tax stacks drawn to scale, the two revenue models, and what each Idaho city took against the 3% cap.
One plan, or eleven: every pension plan in both states on one axis, including Denver's DERP and Aurora's closed old hire funds.
Fact-check notes and sources
Colorado cities: adopted budget books for Denver (2026), Aurora (2023 through 2026), Parker (2026), Greenwood Village (2026), Centennial (2023 through 2026) and Breckenridge (2026), plus each city's Annual Comprehensive Financial Report for pension and OPEB notes. Aurora's plan detail is Note 12 of its 2024 report. Centennial's zero pension and zero OPEB position was verified against the audited statements, not the budget narrative.
Idaho cities: City of Boise FY2025 and FY2026 adopted and FY2027 proposed budgets; City of Twin Falls FY2026 adopted and FY2027 recommended budgets; City of Idaho Falls FY2026 adopted and FY2027 proposed budgets, plus its FY2025 audited report for the utility transfer. Idaho city fiscal years run 1 October to 30 September, so FY2027 figures are proposed budgets rather than adopted ones and are labelled as such throughout.
Boise's public safety share: 50.9% is the city's own published figure, police 28.0% and fire 22.9%, against its gross departmental base. A 54.9% figure can be derived but divides by a net general fund total reduced by a negative $27.1 million intergovernmental cost allocation, which inflates every department's share.
Aurora's growth capture: the 61.8% is $44,632,653 of $72,239,590 of general fund operating growth, from the General Fund Summary in the 2026 adopted budget. The 83.9% figure that circulates uses total general fund uses including transfers out, which fell $19.0 million over the same period.
Boise health insurance: the 30.6% rise is adopted budget to adopted budget, FY2025 to FY2027. Measured against FY2025 actual spending of $31.28 million the two year increase is 9.5%. Both are true; the basis matters.
Mill levies: Douglas County 2025 Tax Districts and Mill Levies; Arapahoe County Abstract of Assessment and Levies; Summit County levy certifications. Parker's common stack is 85.123 mills; a 90.894 figure also circulates and includes Parker Water and Sanitation, which only 47 of the 118 districts levy.
Fire districts: South Metro Fire Rescue 2026 Budget Report and Ballot Issue 7A, approved 4 November 2025. Red, White and Blue Fire Protection District 2026 budget.
Pension plans: Colorado PERA 2025 GASB Statements Nos. 67 and 68 for Division Trust Funds Valuation; Fire and Police Pension Association Statewide Retirement Plan Actuarial Valuation January 1, 2026 and Annual Update to the Pension Review Commission 2025; PERSI employer rates effective 1 April 2025, which are identical for every Idaho city because they are all in the same plan.
Every figure was read from the primary document and checked against that document's own printed totals. The claims were then run through an adversarial review whose brief was to refute them. All survived, and several are stated more narrowly here than in my first draft, including the Aurora figure I corrected above.
Related reading
Two states, two budgets: the Colorado and Idaho state level comparison behind these cities, including why the pension gap is wider than it looks.
The 2026 entitlement infographics: the federal programmes these budgets increasingly compete with.
The non-wage paycheck: what public and private employers actually spend per hour on benefits, and why the public figure is mostly debt service.
Public money surprises: more cases where the published number and the real number are not the same.
This post is informational, not legal, financial or municipal advisory advice. Budget figures are appropriations as adopted or proposed and change through the year. Mentions of named municipalities, districts and retirement systems are nominative fair use. No affiliation is implied.
