For Texas school districts
Enter your district, the bond amount and the propositions Texas law requires. See the chance each one passes, what would improve it, and how a community survey changes the picture.
Works for every Texas ISD in the data, from the largest to the smallest.
Live odds for every proposition, a climate finder, and a survey blend, all in one view.
Search 983 districts, with their bond history, community profile and the November 2026 results.
How to read a Baselice style survey next to the model, and what public surveys got right and wrong.
Share of propositions that passed. Since 2019 November elections have been much harder than May.
Technology does best. Athletics and recreation are the weakest, and the gap is wider in November.
Texas school districts
Enter a district, a total bond amount and the propositions Texas law requires. The planner estimates the chance each one passes, adjusts for current conditions, and blends in a community survey if you have one. Built from propositions in districts, 2006 to 2025.
Climate and this adjustment are planner judgment, not fitted values. Use the adjustment for local factors the data cannot see, such as a tax rate election on the same ballot, a rising rate environment or recent district news.
Survey results are blended with the model. Surveys that did not use the real ballot language overstated support by about 30 points in the public cases we found, so they are discounted heavily.
Filled in from precinct returns and Census data when you pick a district. Edit any value.
A defeat in the last four years lowers the odds.
Texas law (Education Code 45.003, as amended by SB 30 in 2019) lets buildings, buses and sites share one proposition, and requires separate propositions for specified purposes such as technology, athletics and recreation, performing arts and staff housing. Every proposition must state that it is a property tax increase. Confirm your final structure with bond counsel.
Chance of passing. The selected date is shaded.
Change in the chance of passing for the highlighted proposition, in percentage points. Once date and project type are set, the dollar amount shows little effect in the data, so a smaller ask helps mainly by changing what voters see, not what the model measures.
Texas results
Look up any district in the data. Pick one to send it straight to the planner.
| District | Outcome | Voting no |
|---|---|---|
| Cypress Fairbanks | All 5 failed | 63 to 73% |
| Irving | All 4 bonds failed | 56 to 60% |
| Keller | All 5 bonds failed | 53 to 65% |
| McKinney | 4 bonds and tax rate failed | 57 to 69% |
| Comal | Both bonds failed | 59% |
| Channelview | 4 bonds and tax rate failed | 60 to 72% |
| Bartlett | 2 bonds and tax rate failed | 73 to 74% |
| Florence | Bond and tax rate failed | 65 to 67% |
| Manor | Both bonds passed | 45 to 48% |
| Northside, San Antonio | All failed | not reported |
| Dallas, May 2026 | All 4 passed, 8.7% turnout | 25 to 28% |
Evidence
The planner uses a regression model fitted to past Texas school bond elections. Here is what it learned, how well it predicts, and what it cannot see.
After 2019, November elections lost ground. May 2026 returned to the earlier pass rate.
Bonds in more Republican districts did pass less often. Most of that gap comes from home ownership, income and age mix. With those accounted for, party lean moves yes share by only about half a point for every 10 points of Republican share.
Positive means voters were friendlier than the model expected. Over 18 election cycles since 2017 the gap rarely passed 3 points, and the last cycle did not predict the next. November 2026 ran about 13 points below, using 18 propositions in 9 districts, with the Irving, McKinney and Channelview vote shares estimated from reported ranges. It is the largest gap in the series.
Tested on districts the model had never seen. Brier score measures probability accuracy, lower is better; guessing the average scores 0.223.
Blue is what the model predicted, orange is what happened.
Five propositions decided after the model was built, scored with history through 2025.
About 12 percent of the variation in yes share is explained by what can be measured from outside a district. The rest is local: trust, the plan, the campaign and recent history.
Districts choose their own election dates, so the date effect is an association. Party lean is included, but it carries almost no weight on its own.
The range shown reflects uncertainty in the model fit. The odds themselves already include the natural spread of results around the expected yes share.
Survey guide
A survey from a firm like Baselice and Associates measures your own electorate. The model measures what happened in similar districts. Used together they are stronger than either one.
| District and date | Survey | Survey | Vote |
|---|---|---|---|
| Houston ISD, Nov 2024 | Hypothetical, no tax | 75% | 42% |
| Houston ISD, Nov 2024 | Actual language | 39% | 42% |
| Round Rock, Nov 2017 | Hypothetical | 76% | 43 to 48% |
| Richardson, Nov 2025 | Initial, all respondents | 46% | 53 to 64% |
| Dripping Springs | Initial | 62% | 71% |
| Arlington, May 2026 | Initial | about 50% | 59% |
| Northside, Nov 2026 | Initial, decided | 61% | lost |
Compiled from public board presentations and news reports; a handful of cases, so treat as guidance and verify figures before relying on them. Actual language surveys land within roughly 10 points of the vote in either direction. Hypothetical wording is the dangerous one.
The survey result is treated as a noisy and possibly biased reading of the electorate. The model gives a prior for the expected yes share and its uncertainty. The survey, with its error of about 9 points plus sampling error, pulls the estimate part of the way. The more respondents and the closer the wording to the real ballot, the more weight it gets. Hypothetical wording is discounted by 28 points before blending.
The error and the discount come from fewer than ten public survey and vote pairs. As more districts publish surveys, they can be refitted. If your district can share a survey and its result, it improves the tool for everyone.
About
Texas districts spend months building a bond package and then learn how voters feel on election night. This tool is meant to move that learning earlier.
A planning aid that estimates the chance a school bond proposition passes, from the date, the category, the community and the history of the district, adjusted for the election climate and any survey.
It runs alongside a survey firm, not instead of one, and it is not a guarantee.
Texas Bond Review Board local bond election results from 2006 to 2025, Texas Legislative Council precinct returns for party lean, Census 2020 and the American Community Survey, and unofficial county and district results for November 2026.
Results for November 2026 are unofficial until canvassed, and the state does not require districts to report them until August 7, 2027.
The model explains about 12 percent of the variation in yes share. The rest is local: the plan, the campaign, trust and timing. Districts choose their own election dates, so the date effect is an association.
Confirm the legal structure of any proposition with bond counsel.
Westfall Educational Consulting, LLC.