Add annotated-claims-analysis.md

annotated-claims-analysis.md

@@ -0,0 +1,130 @@
+**Annotated Claims Analysis**
+
+**Claim A — Caseload composition implies underlying criminal behavior**
+
+**Neutral paraphrase**
+The document argues that because a public defender’s clients disproportionately come from one demographic, this distribution reflects underlying patterns of criminal behavior in the broader population.
+
+**Reasoning error**
+
+* **Selection bias** (indigent defense is not a random sample)
+* **Causal leap** (from who appears in court to why harm occurs)
+
+**Primary sources to check**
+
+* **BJS NCVS (National Crime Victimization Survey):** measures victimization regardless of arrest
+  [https://bjs.ojp.gov/programs/ncvs](https://bjs.ojp.gov/programs/ncvs?utm_source=chatgpt.com)
+* **FBI Crime Data Explorer (UCR/NIBRS):** police-recorded incidents
+  [https://cde.ucr.cjis.gov/](https://cde.ucr.cjis.gov/?utm_source=chatgpt.com)
+
+**Limitations**
+
+* NCVS measures *victim reports*, not arrests or guilt; it excludes homicide and institutional populations.
+* UCR/NIBRS measure *police activity* and reporting practices; they do not capture unreported crime or prosecutorial filtering.
+
+**Claim B — Courtroom behavior reflects stable group traits**
+
+**Neutral paraphrase**
+Observed courtroom interactions are presented as evidence of stable, group-level behavioral characteristics.
+
+**Reasoning error**
+
+* **Anecdote** (personal observations generalized to populations)
+* **Overgeneralization** (within-group variation ignored)
+
+**Primary sources to check**
+
+* **NCVS** (contextualizes exposure to violence and victimization)
+  [https://bjs.ojp.gov/programs/ncvs](https://bjs.ojp.gov/programs/ncvs?utm_source=chatgpt.com)
+* **Peer-reviewed criminology syntheses** (for behavior under stress and institutional settings)
+
+**Limitations**
+
+* Surveys do not measure courtroom demeanor; they contextualize exposure and outcomes.
+* Institutional stress effects cannot be inferred from population surveys alone.
+
+**Claim C — Arrest/incarceration shares equal criminal propensity**
+
+**Neutral paraphrase**
+Disproportionate incarceration is treated as direct evidence of higher criminality rather than as an outcome shaped by system processes.
+
+**Reasoning error**
+
+* **Confusing correlation with causation**
+* **Measurement error** (treating system outputs as behavior)
+
+**Primary sources to check**
+
+* **U.S. Sentencing Commission (USSC): Demographic Differences in Federal Sentencing**
+  [https://www.ussc.gov/research/research-reports/2023-demographic-differences-federal-sentencing](https://www.ussc.gov/research/research-reports/2023-demographic-differences-federal-sentencing?utm_source=chatgpt.com)
+* **BJS Recidivism Patterns Explorer**
+  [https://bjs.ojp.gov/recidivism-patterns-explorer](https://bjs.ojp.gov/recidivism-patterns-explorer?utm_source=chatgpt.com)
+
+**Limitations**
+
+* USSC covers *federal* cases only; state systems differ.
+* Recidivism reflects prior enforcement, supervision conditions, and definitions of “re-arrest,” not just new offending.
+
+**Claim D — Media reporting explains public misunderstanding**
+
+**Neutral paraphrase**
+The document asserts that media practices systematically obscure relevant facts, leading to public misperception.
+
+**Reasoning error**
+
+* **Unfalsified assertion** (no systematic content analysis presented)
+
+**Primary sources to check**
+
+* **BJS/FBI publications** (direct access to raw tables and methods)
+  [https://bjs.ojp.gov/](https://bjs.ojp.gov/?utm_source=chatgpt.com)
+  [https://cde.ucr.cjis.gov/](https://cde.ucr.cjis.gov/?utm_source=chatgpt.com)
+
+**Limitations**
+
+* Official datasets do not analyze media framing; independent media studies are required to test this claim.
+
+**Claim E — Trial outcomes are driven by inherent communication deficits**
+
+**Neutral paraphrase**
+The document suggests that trial outcomes are strongly affected by defendants’ communication abilities framed as inherent traits.
+
+**Reasoning error**
+
+* **Causal leap** (from courtroom performance to inherent attributes)
+* **Overgeneralization**
+
+**Primary sources to check**
+
+* **USSC sentencing analyses** (controls for legally relevant factors)
+  <https://www.ussc.gov/research>
+* **Courtroom procedure literature** (effects of counsel, jury instructions, and evidentiary rules)
+
+**Limitations**
+
+* Sentencing data capture outcomes after legal thresholds; they do not measure language proficiency or demeanor.
+* Trial selection effects (pleas vs trials) strongly shape observed outcomes.
+
+**Claim F — Family structure and welfare use explain outcomes**
+
+**Neutral paraphrase**
+Family circumstances and benefit receipt are presented as primary explanations for justice-system outcomes.
+
+**Reasoning error**
+
+* **Oversimplification** (single-factor explanation)
+* **Ecological fallacy** (group averages applied to individuals)
+
+**Primary sources to check**
+
+* **U.S. Census Bureau: Poverty and family statistics**
+  [https://www.census.gov/library/publications/2025/demo/p60-287.html](https://www.census.gov/library/publications/2025/demo/p60-287.html?utm_source=chatgpt.com)
+* **CDC/NCHS: Teen birth trends**
+  [https://www.cdc.gov/nchs/fastats/teen-births.htm](https://www.cdc.gov/nchs/fastats/teen-births.htm?utm_source=chatgpt.com)
+* **SSA Disability statistics**
+  [https://www.ssa.gov/policy/docs/statcomps/di\_asr/index.html](https://www.ssa.gov/policy/docs/statcomps/di_asr/index.html?utm_source=chatgpt.com)
+
+**Limitations**
+
+* Census/CDC data describe populations, not causality for criminal cases.
+* SSA statistics reflect eligibility rules and medical determinations, not criminal behavior.