Feasibility research memo

Research analysis of North Bay parcel-tax and bond election history and implications for Harmony USD. Based on official California Elections Data Archive school-district election reports (2018–2024), CDIAC data, and official county election materials.

Bottom line

What the North Bay record shows

The backbone source is the official California Elections Data Archive school-district election series (Sacramento State / California Secretary of State), the only comprehensive statewide repository of local election results. 2025 special-election context was supplemented with official county measure pages and CDIAC's 2025 bond-and-tax report.

Sonoma County

Marin County

Napa and Mendocino: the retrieved source set confirmed the Napa/Sonoma cross-county Calistoga 2022 bond and no Mendocino school district measures in the 2024 CEDA report. County-level odd-year PDF extraction for Napa and Mendocino was not completed; those counties should be treated as confirmed-minimum findings.

Patterns that matter

1 The threshold gap is the central fact. In this North Bay record, bonds routinely passed in the high-50s and low-60s because the legal threshold was 55%, while parcel taxes needed two-thirds and therefore often failed even when they posted what would have been a healthy bond number. Harmony's 2024 bond passed at 56.6%. Fort Ross's 2024 bond passed at 55.6%. Rincon Valley's 2024 bond passed at 59.5%. By contrast, Petaluma Joint Union High's 2024 parcel tax drew 63.5% in Sonoma County's much larger share of the district and still failed. The lesson is not that voters were anti-school; it is that the parcel-tax bar is simply much higher.

2 Renewals and replacements are materially easier than brand-new operating taxes. Cotati-Rohnert Park's 2024 Measure A renewed an existing $89 parcel tax and passed at 75.0%. Kenwood's 2022 Measure A renewed an existing $52 parcel tax and passed. Shoreline's 2020 Measure L renewed the existing $212 parcel tax and passed. Nicasio's 2024 Measure J maintained its parcel tax and passed at 74.3%. By contrast, Petaluma Joint Union High's first 2024 parcel tax failed, and its rerun at a higher $129 rate in November 2025 failed again. A district trying to win its first operating tax should assume it does not enjoy the renewal advantage.

3 Marin is a different parcel-tax market from Sonoma. In the 2024 presidential election, Marin gave Kamala Harris 80.6% while Mendocino gave her 61.3%. Marin's 2024 school parcel-tax results were correspondingly stronger: Nicasio at 74.3%, Bolinas-Stinson at 72.1%, Sausalito-Marin City at 70.1%. The most Democratic and civically habituated districts were also the ones that passed parcel taxes most comfortably, even at higher dollar amounts or more complex structures.

4 Districts can recover from a failed bond by resizing, but that logic is less forgiving for parcel taxes. Tamalpais Union High's March 2024 bond failed at 53.8%; its smaller November 2024 bond passed at 65.8% — a strong example of right-sizing working for capital asks. But parcel taxes seem less forgiving: Petaluma Joint Union High failed in March 2024 at $89 and failed again in November 2025 at $129. The North Bay record suggests you can sometimes save a bond by narrowing and resubmitting; do not assume the same strategy works for an operating tax.

5 Tax design: flat per-parcel dominated; one square-foot success in a very favorable environment. In the retrieved North Bay source set, flat per-parcel taxes dominated Sonoma. One important counterexample: Sausalito-Marin City's 2024 Measure G used a building-square-foot formula plus a lower rate for unimproved parcels and passed with 70.1%. There is no strong North Bay K-12 acreage-tax sample. The defensible takeaway is that clarity matters more than structural cleverness. Sausalito-Marin City's success cannot be extrapolated to Harmony's different electorate.

6 Timing: no universal rule. Sonoma parcel taxes passed in November 2020 (general) and March 2024 (presidential primary). Waugh failed in November 2024 (general). Petaluma Joint Union High failed in both March 2024 (presidential primary) and November 2025 (special election). Nicasio passed in a May 2024 special election. The best inference: off-cycle timing can help when the district already has a disciplined, tax-tolerant electorate. It does not rescue a measure that lacks the coalition to hit two-thirds.

What this implies for Harmony USD

Harmony's strongest positive signal is that voters backed the 2024 bond above the legal threshold. Its strongest negative signal is that the margin was modest: 56.6% is a bond win, but it is not a parcel-tax launchpad. A parcel tax would need to convert a meaningful number of voters who either voted "no" on the bond or skipped that election. This is a coalition-building campaign, not a turnout-only campaign.

The local analogues cut both ways. Harmony is in a part of Sonoma County where school measures can pass: West Sonoma County Union High, Fort Ross, Sebastopol Union, Shoreline, and Cotati-Rohnert Park all demonstrate that Sonoma voters approve school parcel taxes under the right conditions. But the same county provides warning signs: Petaluma Joint Union High's 2024 parcel tax and Waugh's 2024 parcel tax show that support in the low-60s is not enough, and Harmony's own 2024 bond result sits much closer to those danger zones than to the clear parcel-tax wins.

Recommended structure and campaign approach

Tax structure

The recommended structure is a per-sqft rate on improved parcels combined with a flat rate on unimproved/vacant parcels. This is not a hybrid or a compromise — it is the legally cleanest and politically most-vetted approach available to California school districts under current law.

Two North Bay precedents make the case directly:

• Sausalito-Marin City Measure G (November 2024, passed 70.1%): "$0.15 per building square foot, up to 23,000 sqft; $25 per unimproved parcel; 8-year term." Improved parcels pay per-sqft (capped). Unimproved parcels pay a flat rate. These two rates apply to entirely different parcel classes.
• Alameda USD Measure A (March 2020, upheld by Court of Appeal in Traiman v. Alameda USD, 2023): "$0.265 per building square foot, not to exceed $7,999 per parcel; $299 per vacant parcel." The Court of Appeal unanimously upheld this structure as satisfying §50079's uniformity requirement. California Supreme Court denied review in October 2023.

The legal foundation was clarified by AB 2954 (2018), which amended Gov. Code §50079 to expressly authorize taxing unimproved property at a lower rate than improved property. This is the only class-based rate differentiation the statute currently authorizes — and the improved/unimproved split uses it exactly. Every element of this structure has been tested and approved. There is no comparable appellate authority for any other variable structure. See the legal basis for the full analysis.

Every successful Sonoma K-12 parcel tax in the sample used a flat-only structure ($48–$89/parcel). That track record reflects what those districts chose to put on the ballot, not what Sonoma voters would or wouldn't accept. No Sonoma K-12 measure has tried the improved/unimproved split and failed. The failed Sonoma measures (Petaluma JUH twice, Waugh) were also flat — so simplicity alone does not explain the failures or the successes.

The practical case for the split structure at Harmony's revenue target is strong. At $300,000/year with roughly 2,300 taxable parcels, a flat-only tax implies about $130/parcel average — above every successful Sonoma flat measure. The split structure achieves the same revenue more progressively: a 1,000-sqft home pays a modest per-sqft amount; a 50,000-sqft warehouse pays more but is capped; a vacant parcel pays a small flat amount. Most voters in western Sonoma County are residential property owners with modest-sized homes. Asking them to pay proportionally less than large commercial or agricultural landowners is a political advantage, not a liability. "Clarity" for voters means knowing roughly what they'll owe — and "$0.10 per square foot of your home, not to exceed $200" is clear.

Amount

The Example Tax Tables show the revenue landscape across rate and cap combinations for the improved/unimproved split structure. For a $300,000/year target, representative parameters that work include roughly $0.08–$0.12/sqft on improved parcels with a $150–$250 per-parcel cap, plus a $25–$50 flat rate on unimproved parcels. The median residential homeowner in the district pays a modest amount; large commercial and agricultural structures are bounded by the cap.

For reference, the comparable Sausalito structure ($0.15/sqft, $23k cap, $25 unimproved) implied an average improved-parcel bill around $150–$200 in a higher-value district. Harmony's lower average improvement sizes mean a $0.10/sqft rate with a $200 cap produces a median residential bill around $80–$130 — well within the range voters have approved elsewhere.

Duration

The local winners range from five to eight years; some Marin votes in unusually favorable districts ran to ten years. For Harmony, a five- to eight-year term looks more politically defensible. A senior exemption and independent citizen oversight are not optional — they are standard features of successful local parcel taxes.

Annual increase

The measure should include a 3% annual increase, compounding from the base-year rate. Two of the four successful Marin parcel-tax measures in the 2024 sample — Nicasio (74.3%) and Bolinas-Stinson (72.1%) — both included 3% annual increases and passed comfortably. The structure is simple to communicate: voters know what the rate will be in each year of the measure's life, and 3% is a recognizable, modest number.

The practical case for an escalator is straightforward. School operating costs — primarily salaries — tend to increase at or above general CPI. A flat-rate measure that raises $300,000 in year 1 will raise the same nominal dollar amount in year 8, covering materially less real expenditure by the end of the term. A 3% compounding increase turns $300,000 in year 1 into approximately $380,000 by year 8 — a 27% cumulative increase that better tracks cost growth and avoids the need to return to voters for a renewal at a higher base rate.

A fixed 3% annual increase is also simpler to draft than a CPI-indexed formula, which requires specifying the index, the publication date, a floor, a cap, a fallback, and rounding rules — each a potential drafting ambiguity. For a first-time measure in a community that hasn't established trust in the tax yet, simplicity in the language is an asset. See the legal basis for escalators for the governing statutory framework and drafting requirements.

Message

The message should be operating cuts, not abstractions. The strongest parcel-tax ballot language in the North Bay sample emphasized: attracting and retaining excellent teachers, protecting smaller class sizes, preserving core academics, keeping libraries or arts and music, and supporting counseling. The most persuasive case is not "the district needs help" — it is "without this measure, here are the specific teachers, supports, class-size protections, and family-facing services that will be cut." That is especially important because the last measure was a facilities bond; voters need to understand this is about protecting what happens inside the classroom.

Timing

November is the better target than an off-cycle special election for Harmony's situation. The local evidence shows that special-election parcel-tax success is most common where the electorate is already highly favorable. Harmony does not have proof of that yet — it has proof of a 56.6% bond electorate. Because the district needs to grow the coalition rather than merely preserve a tax-tolerant core, a November general election is the safer bet.

Polling standard

The district should assume that polling can be wrong by a lot and should use a much stricter go/no-go standard than it evidently used on the bond. Since parcel taxes die when they sit in the low-60s, Harmony should not be comfortable placing a measure on the ballot unless credible likely-voter polling is already well into the high-60s before opposition messaging. Clear parcel-tax passes tended to be around 68.5% and above; 63%–64% measures failed.

Open questions and limitations

The Sonoma and Marin conclusions are high-confidence: they come directly from the official 2018, 2020, 2022, and 2024 statewide school-district election reports and, for 2025 context, official county/CDIAC materials. Napa and Mendocino are thinner. The Napa/Sonoma cross-county Calistoga 2022 bond was confirmed; the absence of Mendocino school district measures in the 2024 CEDA report was confirmed. But a complete extraction of every county-level odd-year statement-of-vote PDF for Napa and Mendocino was not done — additional pre-2024 or odd-year items may be missing from the inventory above.

A district-by-district demographic model for Harmony itself was not completed. The strongest evidence here is electoral rather than demographic: Harmony's own 2024 bond performance, the nearby Sonoma parcel-tax record, and the very different Marin parcel-tax market. That is enough for a board-level strategic judgment, but not a substitute for a precinct-level likely-voter model.