FUND · 2018 · Metro Measure 26-199
Metro Affordable Housing Bond
Unit counts have been adjusted downward relative to original targets; per-unit costs have grown well above the projection range; "preserved" vs "built" classification has shifted to maintain headline numbers.
Voter intent. In 2018, voters across three counties approved a $652.8M bond. The promise: build or preserve 3,900 affordable homes for working families, seniors, veterans, and people with disabilities.
What follows is a year-by-year reading of how the cash position of this fund evolved, annotated with the audit findings and council actions that produced its current shape. Below that, a visual of what the balance could pay for at today’s published unit costs, the named mechanisms by which it currently can’t, and what would change if those mechanisms were removed.
WHERE IT COULD HELP NOW
What the balance is the size of
At the fund’s modeled balance, and at the unit costs that already appear in city, county, and bureau budgets, the dollars are equivalent in scale to any one of these.
$436.6M sitting today. At today's published unit costs, that is the same dollar amount as any one of these.
1,060 × Newly-built affordable units at corrected unit cost
$410K per unit · Metro bond program 2024 unit cost
1,100 × Preservation of existing affordable units
$165K per unit · Metro preservation program per-unit
3,790 × Site acquisition for the next round of units
$115K per unit · Metro land acquisition program average
Unit counts are rounded against published references. Real procurement and ramp time would shape the exact numbers. The point is the order of magnitude.
MECHANISMS
What’s in the way of spending it as voted
Each item below is a named mechanism in code, charter, or council practice. The defense is the line routinely offered for it. The note beside it is the structural reading of why that line is not the whole story.
Mechanism 1
Cost growth was absorbed by lowering the unit count
When per-unit construction costs outran the bond assumption, jurisdictions kept the dollar total fixed and reduced the home count. Voters approved a unit count.
- Where the discretion sits
- Metro Council and partner-jurisdiction bureaus
The defense
“Construction inflation is outside Metro's control.”
Structural reading
Yet Metro chose to absorb cost growth by lowering targets rather than supplementing the bond. Voters approved 3,900 homes, not $652.8M.
Mechanism 2
"Preserved" units count toward the 3,900-home headline
The mix between newly built and preserved units has shifted toward preservation while the headline number stays the same. The two are not equivalent.
- Where the discretion sits
- Metro reporting practice; jurisdiction project selection
The defense
“Preservation prevents displacement and is counted by HUD as a housing strategy.”
Structural reading
The ballot title said build or preserve. The bureau choices have shifted the mix away from build without returning to voters.
If those mechanisms were removed
A supplemental bond and a binding split between built and preserved would land 3,900 homes within the original schedule plus two years.
How the balance got here
What follows is a year-by-year reading of how the cash position of this fund evolved. Scroll the right column to advance the chart; each step is an audit finding, council resolution, or bureau memo that shaped the shape.
Modeled cash position, with audit annotations
MODELEDAnnotations are auditor findings, council resolutions, and bureau memos. The active step is shown with a vertical rule.
Data table for Modeled cash position, with audit annotations
| Year | Balance | Obligated | Inflow | Spent |
|---|---|---|---|---|
| 2019 | 79390412 | 49222055 | 96817575 | 17427164 |
| 2020 | 167577932 | 103898318 | 107545757 | 19358236 |
| 2021 | 233045825 | 144488411 | 121236838 | 55768945 |
| 2022 | 303841029 | 188381438 | 131102230 | 60307026 |
| 2023 | 346580589 | 214879965 | 147377794 | 104638234 |
| 2024 | 394654706 | 244685918 | 165772816 | 117698699 |
| 2025 | 436573096 | 270675320 | 190538139 | 148619748 |
2020
Audit: per-unit cost growth
Per-unit construction cost climbs above the bond program assumption, narrowing the achievable unit count.
2022
Discovery: classification shift
Reporting begins counting "preserved" units alongside newly built units to maintain the 3,900-home headline.
2024
Council: timeline extension
Delivery timeline extended past original sunset; jurisdictions request supplemental funds.
Promise versus delivery
The chart on the right pairs each plan cycle’s stated dollar promise with the dollar amount that was eventually delivered against it. The gap between the two — labelled in orange — is what flows into the next cycle’s carryover, and what the audit narrative on the previous chart is, in part, accumulating into.
Promised vs. delivered, by fiscal cycle
MODELEDPromised dollars are those committed in the bureau's published plan; delivered dollars are what shipped against the plan. Modeled.
Data table for Promised vs. delivered, by fiscal cycle
| Cycle | Promised | Delivered | Gap |
|---|---|---|---|
| FY 2020 | 110000000 | 28000000 | 82000000 |
| FY 2021 | 130000000 | 67000000 | 63000000 |
| FY 2022 | 145000000 | 102000000 | 43000000 |
| FY 2023 | 160000000 | 121000000 | 39000000 |
| FY 2024 | 175000000 | 137000000 | 38000000 |
Reserve growth
When delivery lags collection, the residual accumulates as an unobligated reserve. This is not a savings account in the household sense: it is a balance that public-finance officers and council staff have, by ordinance, the discretion to redirect.
Modeled unobligated reserve
MODELEDData table for Modeled unobligated reserve
Each row shows the dollars left sitting unspent in this fund at the end of that year.
| Year | Unspent reserve (USD) |
|---|---|
| 2019 | 30168357 |
| 2020 | 63679614 |
| 2021 | 88557414 |
| 2022 | 115459591 |
| 2023 | 131700624 |
| 2024 | 149968788 |
| 2025 | 165897776 |
Drift from voter intent
The chart on the right is a drift index. A value of 100 percent means every dollar in the fund is being used in a way that maps cleanly to the original ballot text. A value below 100 percent means some share has been ordinance-redirected, swept into a sibling program, or otherwise reclassified.
Drift index — voter intent vs. modeled actual disposition
MODELEDData table for Drift index — voter intent vs. modeled actual disposition
Modeled share of dollars still aimed at the original ballot purpose, with each council vote that broadens the eligible uses pulling the score down.
| Year | On voter intent (%) | Drift (%) | Note |
|---|---|---|---|
| 2019 | 98 | 2 | |
| 2020 | 92 | 8 | |
| 2021 | 88 | 12 | |
| 2022 | 81 | 19 | Classification shift |
| 2023 | 79 | 21 | |
| 2025 | 76 | 24 |
WEEKLY MEMO · AFFORDABLE HOUSING BOND
No memo for this fund yet this week
The weekly run hasn’t produced a published memo for this fund yet.