Project Red Team - audit memo
Regional Sports and Events Complex — Economic Impact (free-data path) :root { --ink: #1a1d21; --ink-soft: #4a5158; --ink-faint: #8b939b; --p
Decision Status: At least one check failed. The headline claims in this document should not be repeated in a…
At least one check failed. The headline claims in this document should not be repeated in a public decision until the findings below are answered. This is a finding about what the document shows, not a finding that its numbers are wrong.
What Can Be Said: The claim inventory records what this document asserts
- The claim inventory records what this document asserts.
- Matched support and contradictions may be used as review questions.
- Source safety and detector limits travel with every count.
- Contradicted findings rest on text the document contains.
What Cannot Be Said: This audit does not re-run the study model
- This audit does not re-run the study model.
- A detector miss is not proof that the submitted study omitted support.
- The audited headline claims should not be repeated as cleared by this run.
Why: Whether the tax figure is a net gain to the public purse
- Whether the tax figure is a net gain to the public purse
- Whether the study says which year's model and which year's dollars it used
- Whether the tax figure is gross collections or money this jurisdiction keeps
What Would Change The Answer: Point the audit at public service costs, abatements,…
- Point the audit at public service costs, abatements, incentives, or other cost-side fiscal evidence in the full report.
- Point the audit at model year or data year in the full report.
- Point the audit at gross-vs-net tax treatment and retention/abatement assumptions in the full report.
Next Best Action: Use the missing-evidence list as the work plan for human review
The audit object's own summary line
not release-safe: at least one check failed and the headline claims should not travel yet.
The Checks This Audit Ran, Worst Finding First
- Whether the tax figure is a net gain to the public purseDo not release fiscal benefit language as net benefit.
- Whether the study says which year's model and which year's dollars it usedDo not release year-sensitive claims without vintage disclosure.
- Whether the tax figure is gross collections or money this jurisdiction keepsTax claims should not be used as fiscal benefit until retention is shown.
- Whether the survey behind the spending figure carries the weight put on itCarry the response weakness beside every survey-derived result.
- Whether the headline is new activity or activity that already existsClaim type is at least machine-labelable from the text.
- Whether the visitor money is new to the area or already circulating in itThe split is addressed; a reviewer should still check the arithmetic.
- Whether the area modeled is the area the board is deciding forA geography is present for reviewer checks.
- Whether the return ratio has a defined cost sideThe ratio is reviewable from the text cues.
- How much of the submitted document the audit could read and matchOur table detector matched no tabulation in this file, so figures stated in prose could not be checked against one. The file's text was read in full and nothing was rebuilt to get it, so this is a limit on our detection: it does not establish that the document tabulates nothing, and it does not establish that we read past a table that is there.
Claims A Reader May Quote Only With Support
These are the figures a reader is most likely to repeat back in a public meeting. Each one is quotable only with the support named beside it and the findings in this memo attached to it.
| the submitted file's own wording, character for character | Output, a measure of activity not of wealth: $4.0M |
| the submitted file's own wording, character for character | Jobs: 35 jobs |
| the submitted file's own wording, character for character | Output, a measure of activity not of wealth: $5.4M |
| the submitted file's own wording, character for character | Output, a measure of activity not of wealth: $8.0M |
| the submitted file's own wording, character for character | Output, a measure of activity not of wealth: $340,912 |
| the submitted file's own wording, character for character | Output, a measure of activity not of wealth: $2.1M |
| the submitted file's own wording, character for character | Output, a measure of activity not of wealth: $5.4M |
| the submitted file's own wording, character for character | Output, a measure of activity not of wealth: $58.8M |
| What is claimed | Figure as stated | Where it was found | The text it was read in |
|---|---|---|---|
| Output, a measure of activity not of wealth | $4.0M | Stated in the sentence quoted | Regional Sports and Events Complex — Economic Impact (free-data path) Whether a $4.0M county contribution is recovered through the revenue the facility generates. |
| Jobs | 35 jobs | Stated in the sentence quoted | Where the wages land, by band and by household — which is not a share of anyone's profit and not a distribution to anyone's members Residents of the host community, and whoever speaks for them Examined — 35 jobs placed across 5 income bands, averaging 74% of the regional median wage. |
| Output, a measure of activity not of wealth | $5.4M | Stated in the sentence quoted | Summary of findings $5.4M Annual Operations — annual output every year at full operation 35.0 Annual Operations — jobs supported recurring while operating; direct, indirect and induced $8.0M Construction — one-time output total over the build period 26.1 Construction — job-years temporary, not permanent positions $340... |
| Output, a measure of activity not of wealth | $8.0M | Stated in the sentence quoted | Summary of findings $5.4M Annual Operations — annual output every year at full operation 35.0 Annual Operations — jobs supported recurring while operating; direct, indirect and induced $8.0M Construction — one-time output total over the build period 26.1 Construction — job-years temporary, not permanent positions $340... |
| Output, a measure of activity not of wealth | $340,912 | Stated in the sentence quoted | Summary of findings $5.4M Annual Operations — annual output every year at full operation 35.0 Annual Operations — jobs supported recurring while operating; direct, indirect and induced $8.0M Construction — one-time output total over the build period 26.1 Construction — job-years temporary, not permanent positions $340... |
| Output, a measure of activity not of wealth | $2.1M | Stated beside a table or figure | 28% of its demand is spending the region would have seen anyway, so $2.1M of gross output was removed before any figure above was computed, and $5.4M is what remains. |
| Output, a measure of activity not of wealth | $5.4M | Stated beside a table or figure | 28% of its demand is spending the region would have seen anyway, so $2.1M of gross output was removed before any figure above was computed, and $5.4M is what remains. |
| Output, a measure of activity not of wealth | $58.8M | Stated in the sentence quoted | Over the 2026–2037 horizon the project generates $58.8M in total regional output undiscounted, or $47.0M in present value at 4.0%. |
Quote safety. The submitted file was read as text, so this is that file's own wording, character for character. It may be quoted as what the file we were given says. What that is not is a statement that the file we were given is the published document; the name and SHA-256 of what was read are recorded in the source section.
114 further claims are inventoried in the reviewer pack.
The Three Failed Checks This Memo Leads With, On Support Our Detector Did Not Match
1 Whether the tax figure is a net gain to the public purse
Fiscal revenue is detected, but no cost-side or net-fiscal treatment was detected.
Basis: our detector did not match the support in the document as submitted. That is a miss, not an absence.
2 Whether the study says which year's model and which year's dollars it used
RIMS II is named, but no model or data year was detected.
Basis: our detector did not match the support in the document as submitted. That is a miss, not an absence.
3 Whether the tax figure is gross collections or money this jurisdiction keeps
Tax claims are detected, but gross-vs-net treatment was not detected.
Basis: our detector did not match the support in the document as submitted. That is a miss, not an absence.
A check is one of the 9 tests this audit runs against a submitted document. An evidence item is one disclosure a check looked for and did not match, so 3 failed checks can account for 3 evidence items.
Every count above and below is produced by a text detector reading the document. Its measured limits are printed under each table of findings and set out in full below them, and they apply to all of these counts.
Not Found In The Document As Submitted
Everything in this section is a miss by our detector, not an absence in the study. The detector reads for known phrasings and studies phrase things in ways no list anticipates. Read each line as "we did not find it", and close it by pointing us at the passage.
| Subject | What the audit found | Why it matters | Weight |
|---|---|---|---|
| Whether the tax figure is a net gain to the public purse | Our detector did not match this in the document as submitted: Public service costs, abatements, incentives, or other cost-side fiscal evidence. What the detector read: tax or fiscal revenue appears without a detected public cost-side ledger. | Gross tax revenue may be mistaken for net fiscal benefit. | blocking |
| Whether the study says which year's model and which year's dollars it used | Our detector did not match this in the document as submitted: Model year or data year. What the detector read: RIMS II is named, but no model/data year was detected. | Model structure may not match the spending period. | blocking |
| Whether the tax figure is gross collections or money this jurisdiction keeps | Our detector did not match this in the document as submitted: Gross-vs-net tax treatment and retention/abatement assumptions. What the detector read: tax revenue appears without gross/net or tax-retention language. | A gross tax collection can be mistaken for fiscal benefit to the decision maker. | blocking |
Counted by detector: 76% recall on held-out studies, about 4 in 10 flags not what was sought, and per-document recall observed across 67% to 87% with overlapping intervals. Full limits are stated above.
Where These Checks Have Already Missed A Real Disclosure
These are misses this detector has actually made on published economic impact studies. Each was found afterwards by hand, in text this same extraction had already recovered. They are printed beside the rows above so a reader can see what a not-found row is worth before acting on one. The studies are unnamed because the error in each case was ours.
| Check | What it reads for | What a published study wrote instead | What that made this pack say |
|---|---|---|---|
| Whether the tax figure is gross collections or money this jurisdiction keeps | Gross-versus-net or retention language in the same sentence as a tax figure. | “fiscal benefits (in the form of gross tax receipts) for Bexar County”, with the labelling clause in one sentence and the dollar figure in the next. | This check accepts the words “gross tax” and failed the document anyway, because only the sentence carrying the dollar figure falls inside the window it reads. |
| Whether the tax figure is a net gain to the public purse | Cost-side or net-fiscal language near a tax claim. | A table headed “Detailed Net Fiscal Impacts by Level of Gov.”, with a net position by level of government tabulated under it. | This pack listed the cost side as not found. The net fiscal impacts were tabulated under a heading the tax-claim window does not reach. |
Contradicted By The Study's Own Disclosures
Everything in this section rests on text the study does contain. These are not misses: the document's own disclosure works against the claim attached to it, so they survive the detector's recall limit and a reader cannot close them by finding a passage we overlooked.
| Subject | What the audit found | Why it matters | Weight |
|---|---|---|---|
| Whether the survey behind the spending figure carries the weight put on it | Survey coverage appears weak: 7 of 39 responses were detected. No quotation or table reference was recorded for this finding, so it may be acted on as our reading of the document and may not be quoted as though a passage had been reproduced here. What was matched is stated with the finding; the run did not keep the text it was matched in. | Primary inputs may be directionally biased or too sparse for the claimed precision. | caution |
Counted by detector: 76% recall on held-out studies, about 4 in 10 flags not what was sought, and per-document recall observed across 67% to 87% with overlapping intervals. Full limits are stated above.
What Every Count In This Memo Is Worth
Every count in this memo comes from a text detector, not from a person reading the study. Measured on three studies it had not seen, it found 76% of the passages it looks for, and about 4 in 10 of the sentences it flagged turned out not to be what it was looking for. Recall on individual documents ran 67% to 87%, and those ranges overlap, so we do not claim it reads one firm's writing better than another's. Treat every count here as approximate.
A second, pooled precision figure of 82% exists and is not the one to use here: it mixes an exhaustive positive-find pass into the representative sample, so it describes a set of sentences no count in this pack was drawn from. It is quoted in this pack only so that a reader who meets it elsewhere knows which of the two figures applies.
A count in this pack may be set against another study's count only where three things hold: both were extracted on the same basis, the detector's measured error is printed beside both, and the ordering survives resampling. Where the ordering does not survive, we publish the band and the reason for it instead of a rank.
Source Safety
audit-safe: HTML; plain-text extraction; 1,986 text lines; 41 tables; 100% text extraction quality; HTML table markup supplied; 100% source readiness
This source arrived as markup and its text was read in full. 41 tables arrived already structured, so the audit read them as the document laid them out.
Five separate readings are being reported here and the summary line above carries three: text extraction quality, table reconstruction coverage, and the source readiness the two combine into. All three measure the file. Fourth, claim detection coverage, which measures our detector instead and is not a percentage of this document: what is measured is 76% recall and about 4 in 10 flags not what was sought, taken on held-out studies, and it is stated with the rest of the run limits rather than here. A detection percentage for this document would need a count of what the study contains, which is the number this audit does not assert about any document.
Fifth, quote safety: the file was read as text, so document text reproduced in this pack is the submitted file's own wording, character for character. This is the only one of the five about what may be repeated rather than about how much was legible, and every table in this pack that reproduces document text says beside each row what that row's text is.
A source's state is not support for a number. Whether this study could reach a source (route ready), whether the source answered when something asked it (endpoint answered), and how much of a submitted file could be read are facts about a configuration, a publisher and a document. A figure is supported only where a reading pulled into the evidence record carries it, with its own claim, vintage and consequence.
Reading Limits On The Submitted File
This section is about our reading of the file you sent us, not about the study's contents. A thin extraction lowers what any finding below is worth, in both directions.
The audit reported no limits on reading the submitted file.
What Human Review Would Change
This audit reads a document. It does not re-run the study's model, and it cannot tell a defensible number from an indefensible one. What it can do is tell a reader exactly which passages to go and find. Every line in the not-found column closes the moment somebody points at the passage.
3 evidence items, across 3 checks that failed, hold this document at not release-safe. 3 of those items close if a reader finds the passage the audit missed. Every one of them closes that way.
| Subject | What the audit found | What would close it |
|---|---|---|
| Whether the tax figure is a net gain to the public purse | Our detector did not match this in the document as submitted: Public service costs, abatements, incentives, or other cost-side fiscal evidence. What the detector read: tax or fiscal revenue appears without a detected public cost-side ledger. | Point the audit at public service costs, abatements, incentives, or other cost-side fiscal evidence in the full report. |
| Whether the study says which year's model and which year's dollars it used | Our detector did not match this in the document as submitted: Model year or data year. What the detector read: RIMS II is named, but no model/data year was detected. | Point the audit at model year or data year in the full report. |
| Whether the tax figure is gross collections or money this jurisdiction keeps | Our detector did not match this in the document as submitted: Gross-vs-net tax treatment and retention/abatement assumptions. What the detector read: tax revenue appears without gross/net or tax-retention language. | Point the audit at gross-vs-net tax treatment and retention/abatement assumptions in the full report. |
What This Audit Can Still Be Used For
What this audit matched is usable now. The contradicted findings rest on text the document contains, the claims inventory records what the document asserts, and the source-safety reading is a measurement of the file itself. What the detector did not match settles nothing in either direction. A miss is a miss: it does not show that the document lacks a disclosure, and it does not show that it carries one.
Audit reading: release status not release-safe; 122 claims inventoried; 3 evidence items not found; 1 finding contradicted by the document; 0 reading limits on the file; 130 evidence observations.
This memo audits the document as submitted. It does not prove the underlying model is correct; it shows what the document claims, what support was detected, and what evidence was not found. A finding here is a reason to ask a question, not a finding of error.