St. Joseph Public Schools — MI

Bond: $98.5M · May 7, 2024 · Failed 3,464 No / 2,554 Yes (~58/42, 910-vote margin) · NCES district 2632850 Stated purpose: 50% infrastructure (roofs, boilers, athletic facilities, classrooms), 50% classroom/learning-environment improvements; included controversial $11.7M Gard Street early-childhood center acquisition Contacts: Jenny Fee, Superintendent (now succeeded by Kristen Bawks for Nov 2025 follow-up) · Brenda Graham, CFO · Mike King, Director of Facilities · (269) 926-3100 · sjschools.org Sources: Herald-Palladium · WSJM “back to the drawing board” · WNDU “Vote No St. Joe” opposition · WNDU Nov 2025 follow-up bond

1. Snapshot

Suburb-Small district on Lake Michigan in Berrien County (south of Benton Harbor, north of Niles), 2,930 students across 5 schools (St. Joseph HS, Upton MS, Brown/Clarke/Lincoln Elementaries). SAIPE poverty 4.4% — one of the lowest in the failed-bond cohort. Demographics 73% White / 8% Black / 8% Hispanic / 6% Asian / 6% Multiracial — the most diverse in this MI batch. Per-pupil expenditure $12,078 (FY2020) — at the bottom of the peer band.

2. Why this was a hard sell — community context (ACS)

Metric St. Joseph Note
Median household income $94,787 Comfortable
Median home value $260,200 Above MI median
Bachelor’s+ 49.1% High — top quartile nationally
Graduate degree 19.0% High
Professional occupation 56.1% Majority white-collar — highest in this batch
Owner-occupied 77.9% Moderately high
Non-English household 8.6% Highest non-English in this MI batch
Gini index 0.444 Moderate

This is a high-information, college-educated, professional-class community — a profile that should support school bonds. The fact that an organized opposition group (Vote No St. Joe) formed and the bond lost 58/42 in a community with 49% bachelor’s+ attainment is the diagnostic: in a sophisticated electorate, specifics matter more than enthusiasm. The bond lost on substance, not on tax fatigue.

3. Peer comparison

Top-10 peers via MCP (default weights + 0.20 plantOps emphasis). FMX-customer status now resolved against opted-in local benchmarking server.

# Peer State Enrollment Per-pupil SAIPE poverty Similarity FMX customer
1 Harper Creek Community Schools MI 2,778 $12,403 11.1% 0.955
2 Yorktown Community Schools IN 2,793 $10,979 9.5% 0.940
3 Bangor Township Schools MI 2,691 $10,649 16.5% 0.931
4 Muscle Shoals City AL 2,979 $12,469 14.2% 0.921
5 Templeton Unified CA 2,215 $12,758 10.9% 0.914
6 Clarkston School District WA 2,495 $13,669 17.7% 0.910
7 Harrison County KY 2,982 $12,680 19.0% 0.910
8 Lakeshore School District (Berrien) MI 2,635 $11,808 7.1% 0.910
9 Pendleton SD 16 OR 2,929 $13,382 12.4% 0.909
10 Jasper City AL 2,950 $12,284 22.5% 0.909

Confirmed FMX customers in top 15 (1): Marysville Public Schools (MI).

4. The gap story (what the data would have shown voters)

St. Joseph’s facilities investment is near the peer median — and that’s the problem. The bond pitched a fix for “buckets under leaking ceilings” while the data showed the district was already spending in line with peers, just at the bottom of the band. The opposition spotted this and won the substance argument.

The peer data essentially says: St. Joseph is adequately resourced compared to nearby peers. That’s exactly what the “Vote No St. Joe” group used.

FMX peer operational benchmarks (live)

Pulled live from fmxFacilities for each confirmed FMX-customer peer above. These are the actual operational profiles St. Joseph is being measured against — and what the next campaign can cite directly.

FMX peer Bldgs Total sqft Portfolio age Resolution rate Cost/sqft WO/1K sqft HVAC % of WOs
Marysville Public Schools (MI, 90% similarity) 8 91.7% 16.2%

Most peers above have partial snapshots — they’re FMX customers but several operational fields (sqft, cost-per-sqft, portfolio age) haven’t been backfilled yet. The presence of FMX-customer peers at high similarity is still the load-bearing outreach signal: comparable districts are on the platform; the FMX team should validate whether their data layer is mature enough to cite.

5. Bond history (Ballotpedia + news)

The 3-prior-bond stack is the structural issue: a community already servicing debt from three approved bonds is the opposite of Lake Fenton or Saginaw Twp (no current school debt). Stacked-bond fatigue is real and was the opposition’s strongest argument.

6. What voters / opposition actually said

This is the strongest, most-named opposition in this entire batch:

Espino’s “needs vs wants” framing is the textbook how-to-defeat-a-bond playbook. The district had no counter-narrative grounded in per-building condition data.

7. What we could have told them

  1. “Upton Middle School roof, flooring, heating — even the opposition agrees these are real. The bond was 100% these things plus $11.7M for a preschool building. If we’d split the ballot — Upton + roofs as Question 1, Gard Street ECC as Question 2 — Question 1 would have passed and the opposition would have lost their strongest argument.” Single most important strategic finding for the Nov 2025 follow-up.
  2. “Our plant ops spend is $1,101/pupil — middle of our peer set. The bond isn’t ‘catch up to peers,’ it’s ‘replace systems that have hit end of life.’ That’s a building-specific argument — and we needed building-specific data to make it.” Diagnose the campaign framing.
  3. “0 nurses across 5 schools. Yorktown IN, a 2,793-student same-locale peer, has 4. Of the 4 nearby MI districts most like us, we tie with Bangor Township for the worst nurse staffing.” Concrete, peer-named.
  4. “Suspension rate 2.0% — the lowest of any district in our peer set. Expulsions: 0. The ‘safety’ framing of the bond didn’t match our climate data, and voters could feel the disconnect. The next ask needs to be about building condition, not ‘school safety.’” Calibrate the messaging.
  5. For the Nov 2025 $60M follow-up: Espino’s “needs vs wants” frame is still live. The follow-up bond should explicitly publish the needs/wants ratio: e.g., “$45M roofs/boilers/HVAC, $10M security, $5M classroom modernization, $0 new construction.” Pre-empt the opposition’s strongest argument by adopting their language.

8. FMX outreach hook

St. Joseph now has 1 confirmed FMX-customer peers in the top 15 peer set. The outreach team has live proof points — these are not “likely customers, validate later,” they are named, opted-in, and their operational data is queryable today:

Opener for the call: “Your top-similarity peers include Marysville Public Schools (MI) — both already running FMX. They publish work-order resolution rates, HVAC burden, and per-building cost data your bond campaign couldn’t cite. We can have your portfolio benchmarked against them inside 60 days, in time for your next ballot ask.”

Lead the call against the contact named in the spreadsheet (Director of Operations / CFO / Superintendent as applicable). Validate the named FMX peers above against the internal customer list before outbound — these were resolved via the MCP unredacted endpoint on the local server.