Skip to content
Bay EvidenceInvestigative research · Bay Area public systems
Send a tip
Public accountability record

Corrections & Method Advisories

Bay Evidence separates actual publication errors from recurring interpretation risks. Corrections repair the record; advisories preserve analytical cautions that apply across multiple stories and datasets.

Corrections ledger

Material errors and revisions

A correction is used when Bay Evidence published a materially inaccurate fact, characterization, attribution, citation, or presentation. Routine copy edits that do not change meaning are not entered here. Substantive updates should carry an update note on the affected work.

COR-0052026-09-23

Citizens Transparency Institute field-guide download

Correction

New signups and downloads were paused after document-integrity review found that the public PDF was structurally malformed and omitted substantive content on two detail pages.

Editorial note

The withdrawn file should not be treated as an authoritative or complete guide. Bay Evidence retained the underlying source work for repair and will restore the download only after content, rendering, link, accessibility, and data-protection checks pass.

Inspect the supporting record →
COR-0042026-09-13

Beyond Visibility evidence status

Correction

The public presentation now identifies the work as a descriptive evidence brief under a method advisory. It does not treat visible street reductions as demonstrated longitudinal housing retention.

Editorial note

Point-in-Time counts are cross-sectional snapshots. The brief does not contain person-level retention data, cannot attribute category changes to specific interventions, and should not be cited as evidence of durable housing exits. The downloadable July PDF retains its original subtitle while the interpretive framing is under revision.

Inspect the supporting record →
COR-0012026-07-10

City of Grants Pass v. Johnson

Correction

The U.S. Supreme Court issued the decision on June 28, 2024.

Editorial note

The holding should be characterized narrowly: enforcement of generally applicable laws regulating camping on public property does not constitute cruel and unusual punishment under the Eighth Amendment. The decision did not establish that every encampment operation is lawful under every constitutional, statutory, disability, property, due-process, or state-law doctrine.

Inspect the supporting record ↗
COR-0022026-07-10

San Francisco Proposition C

Correction

Proposition C should not be characterized as a real-estate transfer tax.

Editorial note

The 2018 measure used tax mechanisms directed at large businesses to raise dedicated homelessness funding. Contemporaneous political reporting should be read alongside authoritative descriptions of the measure's actual tax structure.

Inspect the supporting record ↗
COR-0032026-08-10

Research Report No. 1 · Version 1.1

Correction

The report was reissued to remove Appendix A, which publicly cataloged candidate bibliography entries that were excluded because they could not be verified as written.

Editorial note

The authoritative bibliography now contains only sources retained for publication. The bibliographic-verification rule is a standing methodology control, and Eric-Michael Wilson II is identified as the report author. The removed appendix was an editorial audit trail, not evidence supporting the report's findings.

Inspect the supporting record →
Standing methodological advisories

Interpretation rules that should travel with the evidence

ADV-001

CASPEH methodology

The California Statewide Study of People Experiencing Homelessness is representative cross-sectional research, not a longitudinal cohort study.

Interpretation of behavioral-health prevalence must retain the study's duration-bias warning: cross-sectional sampling is more likely to capture people whose homelessness episodes are long than people whose episodes are brief.

ADV-002

Point-in-Time count comparability

Do not blend PIT figures across years without explicit year labels and methodology notes.

Changes in counting conditions, timing, definitions, geography, outreach, or enumeration methods can compromise direct longitudinal comparison. A reported percentage change is not automatically a clean population change.

ADV-003

Observed homelessness versus housing outcomes

A decrease in tents, observed street homelessness, or a PIT estimate is not automatically an increase in permanent housing placements.

People can enter shelter, move to another geography, shift into vehicles, become institutionalized, disengage from services, or disappear from a measurement system. Durable housing outcomes require direct placement and retention evidence.

ADV-004

Appropriation, expenditure, and outcome

An appropriation is not an expenditure; an expenditure is not a service count; a service count is not a durable outcome.

Fiscal accountability should preserve each stage separately so that budget authority, actual spending, people served, housing exits, retention, and returns to homelessness are not collapsed into one success claim.

Request a correction

Identify the exact page and statement, explain what is wrong, and include the strongest available source. Bay Evidence will distinguish disagreement from factual error and will not remove accurate material merely because it is inconvenient to a subject of coverage.

Email a correction request →Read editorial standards →