# Austin-area MUD tax, debt and financing-fee evidence

Research snapshot: 30 October 2026. Geographic core: Travis, Williamson and Hays counties. This is a reproducible source cohort and homeowner directory, not a certification that every legally created or nonlevying MUD has been identified.

## Prior work and what this adds

Austin Local Team's October 2026 explainer already analyzed the Comptroller's 2025 rates. Its 155 count can be reproduced as positive-rate county Detail rows whose names contain the literal abbreviation MUD. That counts seven cross-county duplicates and omits the fully spelled Crosswinds Municipal Utility District. It does not establish 155 independent districts.

The HOA Book already combines TCEQ, Comptroller and BRB data and publishes debt per GIS parcel. This investigation must not claim to be the first such join, or to invent debt per parcel. Parcels are not occupied homes.

The additional work here is a transparent district-ID crosswalk, reconciliation of three counties' tax lists, preservation of disagreements and missing values, direct GO/revenue debt decomposition, and issue-level measurement of professional fee concentration. The accompanying case studies examine the economic claims direct debt does not capture.

Prior-work links:

- https://austin.localteam.house/news/austin-mud-tax-rate-buyer-notice
- https://thehoabook.com/texas/water-districts/williamson-county-mud-51/

## Deliverables and which to use

`austin_area_mud_tax_debt_ledger_2025.csv` and `.json` preserve the **174-unit analysis cohort** selected from county Detail rows in the Comptroller's 2025 Working workbook, version 28 January 2026. The 174 records comprise 173 district-name units and Brushy Creek Defined Area. Of these, 149 have reported positive rates; 25 are source-starred unreported rates, represented as null rather than zero.

`austin_area_mud_lookup_expanded.csv` and `.json` are the **177-record homeowner directory**. They preserve the original fields and add explicit `lookup_preferred_2025_*` fields. Display those preferred rate fields for homeowners. Use `directory_record_id` as the nonnull unique key, `in_analysis_cohort` to preserve analytical scope, and `lookup_preferred_2025_rate_source_url` for the displayed rate's provenance. Three county-verified supplements are outside the 174-unit core. Two existing records receive corrected or newly available preferred rates without changing their raw state-source fields.

`county_lookup_corrections_and_supplements_2025.csv` and `.json` contain those five exceptional records alone. `county_supplemental_mud_directory_2025.csv` contains the three additional records alone.

`county_rate_audit_2025.csv` and `.json` preserve 235 source rows: 65 Travis table records, 151 records across two overlapping Williamson lists, and 19 Hays MUD-name records. These are source-row counts, not district counts. The audit separately retains two historical Lake Pointe defined-area placeholders and Lakeside WCID2A, which the short Williamson list places among MUDs.

`brb_firm_fee_summary_fy2025.csv` is the preferred concentration table for the report. It uses an explicitly dated fiscal year. `brb_professional_fee_rows_fy2025.csv` provides the 117 supporting role records. `brb_issuer_fee_rows_fy2025.csv` preserves all 1,080 matched fee records, including payer flags and comments. `brb_issuance_summary_fy2025.csv` contains 57 distinct issues, counting par once per issuer/issue/closing-date key. An absent role is null, not free service.

The older five-year-chart tables, `brb_firm_fee_rows.csv` and `brb_firm_fee_summary.csv`, are a robustness check. BRB calls these the past five fiscal years on an FY2025 issuer page; the chart CSV has no year field. The exact first year was not independently verified, so do not relabel them FY2021-2025.

`census_diagnostics.json`, `county_audit_diagnostics.json` and `fy2025_fee_diagnostics.json` contain reproducible control totals and scope notes. `census_source_register.json` records the underlying public sources. `tceq_area_candidate_inventory.csv` retains 274 area-related registry records, including inactive records and known county/type inconsistencies. It is a discovery inventory, not a count of current legal MUDs. `adjacent_water_and_limited_district_tax_rows_2025.csv` keeps neighboring district forms separate from the MUD-name cohort.

## County reconciliation and the consequential discrepancies

Among 149 state-source reported-rate units, 147 are corroborated by county rate lists, one conflicts, and Vista MUD is absent from the reviewed county tables. A further previously unreported state record receives a county rate.

| Record | Preserved state snapshot | Preferred homeowner rate | Basis |
|---|---:|---:|---|
| South Fork Ranch MUD #1 | $0.5775 total; $0.1625 M&O, $0.415 I&S | $1.20 total, all M&O | Williamson full 2025 summary and the district's own information page agree |
| Williamson County MUD #44 | Unreported | $1.20 total, all M&O | Williamson full 2025 summary |
| Williamson County MUD #39 | Not in county Detail cohort | $0.95 total, all M&O | Both Williamson lists; the Comptroller Summary includes ID 246-305-04 and this rate, but no Detail row |
| Collins MUD #1 | Not in cohort | 2025 unverified; 2026 $1.00, all M&O | Travis says newly levying in 2026; 2025 zeros have no worksheet |
| New Sweden MUD #1 | Not in cohort | 2025 unverified; 2026 $0.95, all M&O | Travis says newly levying in 2026; 2025 zeros have no worksheet |

Wells Branch's 2025 rate is $0.24, corroborated by both county tables and the official October 2025 newsletter. Travis separately lists $0.2518 for 2026. A district webpage mixes years; it should not overwrite the documented historical rate.

The frozen core's calculated regional levy is $321,884,768. Replacing only South Fork's rate while holding its state taxable base of $55,810,217 constant raises the estimate by $347,418.60 to $322,232,186.60. Estimated I&S falls $231,612.40 and M&O rises $579,031.00. This is a rate sensitivity using the same base, not actual collections and not a fully corrected official levy. The original source-based aggregate may describe approximate scale if this disagreement is disclosed. Do not call its maximum or median a definitive regional statistic.

County sources:

- https://www.traviscountytx.gov/tax-rates
- https://www.wilcotx.gov/DocumentCenter/View/16987
- https://www.wilcotx.gov/DocumentCenter/View/17149/2025-Truth-in-Taxation-Summary?bidId=
- https://hayscad.com/wp-content/uploads/2025/12/2025-ALL-HAYS-CAD-JURISDICTIONS.pdf
- https://districtdirectory.org/williamson-county/south-fork-ranch-municipal-utility-district/
- https://wellsbranchmud.com/word-archives?catid=12&id=3940&m=0&task=download.send

## Fiscal and statistical definitions

BRB direct debt is outstanding principal at 31 August 2025. It is not lifetime principal plus interest, unissued authorized bonds, unreimbursed developer costs, or all contractual and overlapping claims on household taxes. The 2025 adopted tax rates can fund fiscal 2026 budgets, so the dates do not coincide.

136 of the 174 units match BRB issuer records. Their known direct debt is $2,893,195,000: $2,767,750,000 tax supported and $125,445,000 revenue supported. The 131 matched positive-rate units have the same total; the five matched source-unreported-rate units all report zero direct debt. The 131 matched positive-rate units cover $319,964,550 of the $321,884,768 regional calculated levy, or 99.4034%. They cover $62,440,755,884 of $63,372,434,763 in summed taxable-base entries, or 98.5298%.

Brushy Creek Defined Area overlaps its parent district. Summed taxable-base entries are not the value of distinct property. Separate BRB issuer principal can be summed, but do not add a consolidated parent-audit debt total to those amounts. Cross-county district IDs are deduplicated for debt; county portions of taxable value are preserved.

The workbook supplies only one taxable-value field, not separate M&O and I&S taxable bases. Estimated allocation = taxable value times the component tax rate divided by 100. Source calculated levy is not actual collections. Different exemptions and source rounding mean the component estimates need not exactly sum to the rounded levy. The original cohort allocation is approximately 66.4% I&S dollars, before the stated South Fork sensitivity.

The eight apparent source-zero-debt/positive-I&S rows are leads, not proof of a contract-financing mechanism. South Fork is a demonstrated rate-data conflict and should be removed from any asserted example count. Master-district contracts require reading the agreements and audits, as the case studies do.

## Financing-firm concentration: exact scope

FY2025 matched records cover 43 issuers and 57 bond issues totaling $340,825,000 in par, exactly equal to the sum of FY2025 issued debt on the matched BRB issuer pages. Reported BRB-classified BondCounsel fees total $5,833,446.15. The leading three firms receive 84.4025%. FinancialAdvisor fees total $6,502,086.28; the leading three receive 87.9108%.

The two roles together equal $12,335,532.43, or 3.6193% of that par. This is two reported fee categories, not all issuance cost, general district operating costs, profits, a measure of competitive pricing, or evidence that concentration caused higher charges. Names are preserved as reported rather than merged into corporate parents.

The Crosswinds $96,000 Barrett record is classified BondCounsel but its source comment says General Counsel. Excluding that record makes the leading-three counsel share about 85.8147%, leaving the concentration conclusion intact. State the classification caveat.

Do not sum all fee rows and label the result net issuer cash cost. Some are marked underwriter paid and some comments say a fee was included in another fee. The source rows preserve that information for further reconciliation.

The five-year displayed-chart robustness check covers 93 issuers with records among 136 checked: top-three counsel 79.1974%; top-three financial advisers 90.2362%. Missing chart is unknown, not zero.

## Reproduction

Using the saved sources and the primary Python runtime with openpyxl:

1. Run `build_census.py` to reconstruct the frozen state cohort, crosswalk, registry inventory, issuer-level decomposition and displayed five-year fee summaries.
2. Run `reconcile_county_rates.py` to parse saved county extracts, audit rates, and rebuild the preferred-rate directory and sensitivity.
3. Run `analyze_fy2025_fees.py` to reconstruct the dated issue-level concentration and check par against the issuer pages.

`fetch_brb_details.py` documents retrieval of BRB issuer pages and only actually linked chart URLs. Source files and downloads are retained alongside these scripts. All name normalization and exceptions are explicit; no fuzzy name match is silently accepted. Unknowns remain null.