What a world-class park system would cost Portland
Built bottom up from Portland's own records, the Atlas's targeted path to a world-class park system costs $1.71B up front and $116M a year more once it is built, about $177 per resident a year. It fixes what Portland already owns, puts a park within a 10-minute walk of 99% of residents, and adds what people notice in winter: a restroom that is open, a dry place to meet, and someone on site.
Portland does not need to outspend the country's best park systems; it already matches them. It needs to spend in a different shape: permanent money, steady reinvestment in what it builds, and published standards.
Atlas model in 2026 dollars, from PP&R's FY 2026–27 adopted budget, its August 2026 backlog, 2020 Census blocks and PortlandMaps. Ranges run from low to high unit costs. Not a City plan.
The targeted path at a glance
- One-time capital
- $1.71B$1.29B–$2.20B
- Per year during a 15-year build-outcapital spread evenly, plus the new running costs
- $230M$177M–$324M
- Per year once builtoperating and reinvestment
- $116M$91M–$177M
- Per resident per year once built
- $177$139–$272
What it leaves out
Maxing every measure in the national park ranking would cost $4.80B up front and $231M a year, mostly for sports fields, land and the last 1% of walk access. The path skips those and keeps them priced, so any can be added back.
Why not chase the rankingsPortland already spends like a leader. Its gaps are where parks are, how long they last and how the money is raised.
On the one comparison that counts every agency the same way, Trust for Public Land's City Park Facts, Portland spent $351 per resident in FY 2025, level with Minneapolis and more than twice the median large city. It ranks 9th on ParkScore, where Washington, DC is first.
- The repair backlog is 30% of what the system is worth. PP&R reports $759M of unfunded repairs (August 2026) against a $2.50B replacement value. It reinvests about $6.6M a year, 0.26% of that value; the City's 2024 asset report calls for at least 2–3%.
- A five-year levy carries the operations. The 2025 levy raises $86.3M in FY 2026–27 and ends after FY 2030–31. Measure 5 compression took 27% of the City's local option levy yield in FY 2025–26 (TSCC).
- East Portland is furthest behind. In District 1, 4.5% of residents live within a half mile of an indoor room or community center, against 28.4% citywide.
- Recreation programs are the clearest spending gap. Portland spends $78 per resident on programs, against $117–128 in DC, Minneapolis, Arlington and Seattle (TPL, FY 2025).
$1.71B up front and $116M a year buy eight things residents would notice
Each line is a quantity from the model's siting curves or Portland's own records, times a sourced unit cost. The foundation, clearing the backlog and reinvesting every year, is 45% of the capital and 54% of the yearly cost.
1Fix the foundation
Clear the $759M backlog; reinvest 2.5% of replacement value a year, from 0.26% today; publish independent condition and cleanliness scores; winterize 94 restroom buildings and open all 203 every day of the year.
Today: Backlog $759M; reinvesting 0.26% a year
After: Backlog cleared; 2.5% a year; scores published
One-time$764M$555M–$811MPer year once built$62.6M/yr$49.7M–$88.1M2A park within a 10-minute walk
39 new two-acre neighborhood parks, sited where the most residents lack one.
Today: 90.8% within a half mile
After: 99% within a half mile
About 53,700 more residents reached, about $6,000 of capital each
One-time$323M$227M–$554MPer year once built$8.4M/yr$4.9M–$16.3M3Restrooms that are open and clean
52 new year-round restrooms (41 at existing parks), and every restroom cleaned twice a day.
Today: 78.7% within a half mile
After: 95% within a half mile
About 107,200 more residents reached, about $290 of capital each
One-time$31.2M$14.6M–$52MPer year once built$6.2M/yr$4.9M–$8.2M4Covered places to gather all year
76 covered, heated gathering places (74 at existing parks); District 1's 23 built as enclosed rooms.
Today: 59.9% within a half mile
After: 90% within a half mile
About 198,500 more residents reached, about $750 of capital each
One-time$149M$144M–$266MPer year once built$6.4M/yr$5.1M–$13.9M5Community centers within reach
4 full-service community centers, the need PP&R named in its 2022 Level of Service guidance.
Today: 4 full-service centers
After: 8, one within 3 miles of everyone
One-time$200M$140M–$240MPer year once built$19.6M/yr$15.6M–$24.3M6People on site
A park keeper at each of the 56 parks with 8,000 or more residents within a half mile.
Today: No site-assigned keepers; 30 rangers citywide
After: 56 keepers at busy parks
One-timenonePer year once built$8.1M/yr$7.3M–$9M7Grounds care to a published standard
Grounds care per developed acre at the Royal Parks and Helsinki level. Modeled spending already meets it, so the cost appears only in the high case.
Today: About $22,000 per developed acre (modeled)
After: A published standard and inspection score
One-timenonePer year once built$0/yr$0–$12.6M8Trees on park land
79,028 trees planted and established on park land, its share of the City's 45% canopy goal.
Today: Canopy 32.4% citywide
After: Park land's share of 45%
One-time$245M$213M–$277MPer year once built$4.3M/yr$3.6M–$5.1M
Total, parks budget
Outside the parks budget: the City's 45% canopy goal also needs about 560,700 trees on streets, schools and private land, about $1.74B plus $30.8M a year of care. The 39 new parks also need safe sidewalks and crossings, which PBOT has not priced. Neither is in the totals above.
Every cost line, with its quantity
| Cost line | Quantity | One-time | Per year |
|---|---|---|---|
| Fix the foundation | |||
| Clear the repair backlog | 1 | $759M | — |
| Reinvest in today's assets: yearly gap to the target rate | $55.9M | — | $55.9M |
| Lifecycle for already-approved new projects | $4.7M | — | $4.7M |
| NYC-style park inspection program | 2 | — | $382K |
| Winterize seasonal restrooms | 94 | $5.4M | — |
| Open existing restrooms year-round (cleaning and supplies) | 30,653 | — | $1.7M |
| A park within a 10-minute walk | |||
| Land for new parks | 78 | $50.2M | — |
| Build new parks | 78 | $273M | $8.4M |
| Restrooms that are open and clean | |||
| New restrooms at existing parks | 41 | $24.6M | $1.4M |
| New restrooms at new park sites | 11 | $6.6M | $385K |
| Clean every restroom twice a day (the extra cleaning) | 255 | — | $4.3M |
| Covered places to gather all year | |||
| Covered all-season hubs at existing parks | 53 | $68.9M | $3M |
| Covered all-season hubs at new park sites | 0 | — | — |
| Enclosed, heated rooms in District 1 | 23 | $80.5M | $3.4M |
| Community centers within reach | |||
| Full-service community centers | 4 | $200M | $19.6M |
| People on site | |||
| Park keepers at busy parks | 56 | — | $8.1M |
| Grounds care to a published standard | |||
| Raise grounds care to the world-class standard | 1,580 | — | high case $12.6M |
| Trees on park land | |||
| Trees on park land | 79,028 | $245M | $4.3M |
Quantities are counts, acres or restroom-days; the reinvestment line's quantity is its yearly dollar gap.
Lifecycle reinvestment is charged at 2%, 2.5% and 3.5% of non-land capital a year (low, typical, high). Covered hubs, District 1's enclosed rooms and restroom winterization use Atlas assumptions where no Portland figure exists. Grounds care already costs about $22,000 per developed acre in the Atlas's modeled allocation, above the Royal Parks and Helsinki, so that line costs nothing except in the high case.
Within a 10-minute walk, 99% would have a park, 95% a restroom and 90% a covered place to meet
The siting curves place each new park, restroom and covered hub where it reaches the most residents who lack one. District 1 gains the most.
| Within a half mile of… | City | District 1 | District 2 | District 3 | District 4 |
|---|---|---|---|---|---|
| A public park | 90.8% → 99.0% | 88.3% → 99.5% | 93.1% → 98.8% | 93.9% → 99.7% | 88.0% → 98.2% |
| A park restroom | 78.7% → 95.2% | 69.7% → 94.8% | 87.7% → 96.1% | 83.9% → 98.5% | 73.5% → 91.2% |
| A covered or indoor gathering place | 59.9% → 90.3% | 38.6% → 86.7% | 71.8% → 93.1% | 74.0% → 97.6% | 54.5% → 83.6% |
| An indoor room or community center | 28.4% | 4.5% | 37.0% | 41.5% | 29.6% |
- About 53,700 more residents near a park. The 39 new parks go where the gaps are largest: 14 in District 4, 11 in District 1, 8 in District 2 and 6 in District 3.
- About 198,500 more near a covered place. 74 of the 76 hubs sit in parks Portland already owns, so no land is bought for them. District 1 gains about 76,600 residents, the most of any district.
- Indoor rooms where there are none. District 1's 23 enclosed rooms would put at least 76,600 more residents within a half mile of an indoor room (siting-curve estimate).
- Every restroom open every day. All 203 existing restroom buildings stay open through winter, as Seattle is doing for all 129 of its park restrooms by 2028 (Seattle Parks).
Maxing every ranking measure would cost $4.80B, mostly for sports fields and land
ParkScore gives full points at about twice the national median, and Washington, DC ranks first at 85.1 while scoring 55 on acreage (TPL 2026). The model prices four ambitions side by side; the targeted path is the one that buys what residents feel.
| Tier | One-time capital | Per year once built | Per resident per year | Years |
|---|---|---|---|---|
| Fix the foundationClear the repair backlog, reinvest 2.5% of replacement value a year, inspect parks and publish the scores, and open every restroom all year. | $764M$555M–$811M | $62.6M$49.7M–$88.1M | $96 | 10 |
| Top of ParkScoreThe foundation plus every 2026 ParkScore maximum-points threshold at once: 100% walk access, parkland on 18.7% of city area, and TPL's per-resident counts of restrooms, fields, hoops, centers and playgrounds. | $4.80B$3.21B–$7.91B | $231M$155M–$382M | $354 | 15 |
| World-classBoth, plus the strictest credible global standards: restrooms and covered places near nearly everyone, 2 more indoor pools, 142 more trail miles, Portland's 45% canopy goal and park keepers. | $5.69B$3.84B–$9.41B | $269M$182M–$465M | $413 | 20 |
| Targeted pathThe foundation plus the gaps residents feel most, at a reasonable cost per resident reached. No ParkScore counts of fields, land, hoops or neighborhood centers. | $1.71B$1.29B–$2.20B | $116M$91M–$177M | $177 | 15 |
The world-class tier also carries $1.74B of citywide tree planting outside the parks budget.
Left out of the path, at world-class prices
| Left out | One-time | Per year | Why |
|---|---|---|---|
| The last 1% of walk access: 70 more parks for about 6,200 residents | $580M | $15.1M | About $93,000 of capital per resident reached. |
| 224 athletic fields, to reach ParkScore's count per resident | $1.12B | $37M | A count relative to other cities; natural turf at $2.6M a field would halve the capital. |
| Parkland to 18.7% of city area (2,269 more acres) | $1.33B | $39.6M | Portland already has 19.6 acres per 1,000 residents; its gap is where parks are, not how much land. The most assumption-sensitive line. |
| 32 neighborhood recreation centers | $432M | $46.9M | Center size is an Atlas assumption; the four full-service centers come first. |
| 2 more indoor aquatic centers | $180M | $7.5M | The North Portland Aquatic Center ($91.5M) opens in 2029; revisit for District 1, which has one pool. |
| 142 more miles of regional trail | $355M | $8.9M | Portland's own 2020 Vision goal and a natural fit for Metro money; trail upkeep is not yet priced. |
| 297 basketball hoops and 28 playgrounds | $30.5M | $973K | Cheap ParkScore gains that could be folded in. |
A bond can fix what exists. The yearly cost needs a permanent source.
The path's $116M a year once built equals a park district rate of $1.34 per $1,000 of assessed value, close to the Tualatin Hills Park & Recreation District's permanent $1.3073, which would raise about $113M a year in Portland (THPRD).
| Cost | Best fit | What it yields | What it takes |
|---|---|---|---|
| Backlog, $759M once | General obligation bond, outside Measure 5 compression | About $58.3M a year over 20 years at 4.5%. A $600M bond polled at 57% informed support in 2023 (survey). | Council referral and a vote. Council has committed to explore a major-maintenance bond (Resolution 37710). |
| Yearly cost, $116M | Permanent park district, or a larger levy | District: $1.34 per $1,000. Levy: $1.88 per $1,000, about $416 a year for the median home, on top of renewing today's $86.3M; added levy cents compress first. | A district needs City consent, county hearings, a vote and its own elected board (ORS 266). A Council-governed district like Seattle's needs a state law change; HB 3515 died in 2023. |
| New parks, restrooms, hubs and centers, $704M once | System development charges, Metro, a second bond | Development charges bring $6.7M–$12.2M a year (2024–25 report); Metro's local share is $10.3M in FY 2026–27. | Charges pay only for added capacity; an updated methodology; a later vote. |
| Trees on park land, $245M once | Portland Clean Energy Fund | $26.9M a year in PP&R's FY 2026–27 budget (PCEF). | Council amends the Climate Investment Plan. |
- Keep new money from replacing old money. Seattle's park district comes with a General Fund floor indexed to inflation, $126.7 million in 2026 (Seattle Park District).
- Guarantee a yearly repair amount, allocated by formula. Minneapolis sets a minimum for capital and rehabilitation each year, $13.1 million in 2026, and ranks projects on published equity and condition criteria (NPP20).
- No new asset without its running cost. The City Auditor found eight approved projects carrying $4.7 million a year of unfunded lifecycle cost (October 2025).
- The window is short. Council must adopt a long-range financial plan by June 30, 2028, and the levy ends after FY 2030–31.
The leaders share rules more than budgets
The Atlas reviewed 21 park systems in the U.S., Europe and Asia-Pacific. Five practices recur in the ones that work.
- Permanent money. Seattle voters created a park district in 2014; it levied $131.2 million in 2026 (Seattle). Minneapolis guarantees its neighborhood-park rehabilitation money by ordinance (NPP20).
- Condition measured and published. New York inspects random parks and rated 87% acceptable overall and 93% for cleanliness in FY 2026, against targets of 85% and 90% (NYC). Helsinki prints its maintenance cost per square meter in every budget (Helsinki); Singapore reports walk access beside spending (Singapore).
- Capital steered to underserved parks. New York's Community Parks Initiative rebuilds parks that received less than $250,000 of capital in 20 years (APA).
- Parks usable all year, with people in them. Minneapolis budgets about 154 park keepers for about 160 neighborhood parks (MPRB). Vienna funds youth workers who run free afternoon sessions in parks (Wien). Paris keeps about 140 parks open around the clock (Sortiraparis).
- The running cost priced up front. Paris created 30 posts in 2025 to maintain new green space (Paris budget). Melbourne turned its 40% canopy goal into at least 3,000 trees a year (Melbourne).
Six PP&R records would turn this model into a plan
The model leans on assumptions wherever Portland does not publish a figure. Each of these records would replace one.
- Work-order actuals by park. The $22,000-per-acre grounds figure is an allocation, not an accounting fact.
- The winter restroom list. Which restrooms close in winter, and which can run in freezing weather.
- Replacement values by asset. The reinvestment line uses the bureau-wide value.
- Current costs and funded status for each backlog item. See Next in line.
- Current land values. The model escalates 2013 values.
- PBOT sidewalk and crossing costs. Safe routes to new parks are not yet priced.
Source: Atlas world-class parks model (scripts/park_cost_model/world_class.py (Portland world-class parks model)), exported 2026-10-25. Unit costs from Portland and Northwest sources where they exist; targets from TPL ParkScore 2026, NYC Parks, Seattle, Minneapolis, the Royal Parks, Helsinki and Portland's Urban Forest Plan.