THE LEVY FEED

The Levy Feed began in my COID 352 (Design for Impact) course, grew into a data-analytics capstone (COID 269), and became a live product at levyfeed.com. Along the way it won the College of Innovation & Design 352 Pitch Competition, became a finalist in the Idaho Entrepreneur Challenge, was presented at Hackfort 2026, and went through Boise State's Venture College incubator as a mentored cohort. Solo founder — I ran the research, the data analysis, the interaction design, the prototype build, and the pitch. Advisor: Professor Anthony Saba. Concluded May 2026.

In 2025, Nampa's supplemental levy passed by just 465 votes (~52%). Post-election research told a different story than "voters don't like schools": 82% of NO voters reported feeling uninformed, not opposed, and 65% of NO votes came from residents without children in the schools. Budget data is technically public but practically inaccessible — buried in board packets, formatted as dense PDFs, explained in plain language once a year (in October, when the district wants money).

And there's a structural twist almost no one sees: under Idaho's HB 574 (Idaho Code § 34-914), districts are prohibited from advertising the HB 292 state tax relief on the official ballot. Voters see the gross "sticker price," never the "sale price." For Nampa's median home the ballot showed $427/year — the actual cost was $152, so the ballot legally overstated the real cost by 64%.

Twice, the research proved me wrong. (1) I assumed clearer charts would fix disengagement. Comparative empathy mapping killed that — people scan, they don't study, and information becomes actionable based on trust and identity, not data clarity. That drove the pivot from data tables to an image-first feed.

(2) I assumed transparency would predict votes. The data complicated it: across 15 districts, measure type dominated outcomes — the two highest-transparency districts both failed (general-obligation bonds need a 66.67% supermajority) while the lowest-transparency district passed its simple-majority levy. That sharpened the thesis: The Levy Feed isn't a tool that predicts whether transparency wins votes — it's a tool that measures whether voters can engage informedly with what they're being asked to decide.

Qualitative: 5 semi-structured interviews, comparative empathy mapping (NO vs. YES voters), a Draw-It exercise, and low-fi prototype testing. The gap between groups wasn't information volume — it was relationship; even YES voters couldn't defend the budget to a skeptical neighbor.

Quantitative (the data capstone): a 1,700+ row, four-source dataset — Idaho SDE finance reports, Secretary of State election results, U.S. Census ACS data pulled via a custom Python API script, and a primary-research instrument I built myself, the Transparency Index (four pillars × five criteria, scored 0–2, 40 points per district, 15 districts audited blind to outcome; Nampa scored lowest at 18/40, Vallivue highest at 36/40). I also turned the HB 574 problem into math — modeling any homeowner's true net cost from the homeowner exemption (§ 63-602G) and the gross/net rates — and coined a metric, the Information Friction Coefficient (180% for Nampa), for the gap between ballot price and real cost.

“I'd actually vote yes if I knew where the money was really going, but they only talk to me in October when they want something.”

A year-round, image-first transparency platform (levyfeed.com) with five public surfaces — Feed (social-style stories all year, not just at campaign time), Calculator ("See Where Your Tax Dollars Go" — enter a home value and ZIP, see the ballot price, the relief HB 574 hides, and your actual cost), Impact (personalized dollar breakdown), Stories ("Meet Mrs. Chen — your levy helped fund her science-lab supplies"), and Q&A (district-answered questions plus a candid "Where We're Struggling" section).

There's a district-facing side too — The Signal, an admin dashboard scoped to the Nampa finance office's review needs: anonymized, aggregated engagement metrics, never individual user data — the privacy guarantee is structural, because trust is the product.

Process arc: POV → "How might we make engaging with school budgets feel like scrolling social media?" → five ideation concepts → V1 (text-heavy; validated the feed concept but too much reading) → V2 (image-first pivot). Under the hood: a Python data pipeline (Census API; choropleth via TIGER/Line shapefiles → Mapshaper → SVG) with analysis in Excel, prototyped in Figma and built on Replit. A sibling product, The Tuition Feed, applies the same model to higher-education budgets.

Because The Levy Feed is independent third-party research, it can legally publish the net-cost calculation the district is prohibited from advertising — a district can point voters to it without violating HB 574. The math is documented in a full technical appendix (homeowner exemption, gross/net rate derivation, a Pearson-correlation framework, and the Information Friction metric).

Live at levyfeed.com with a working calculator on real Nampa data. 352 Pitch Competition — winner. Idaho Entrepreneur Challenge — finalist. Hackfort 2026 — presenter. BSU Venture College — mentored cohort. Nampa School District — presented to district stakeholders, putting the transparency recommendations in front of the people who can act on them.

Concluded, May 2026 — The Levy Feed wrapped with a presentation to Nampa School District stakeholders, the culmination of the research and prototype work. The Data Analysis capstone (COID 269) built the 1,700+ row, four-source dataset and the Transparency Index framework, which documents an approach any Idaho district could adapt ahead of the 2027 cycle.