1[2 {"id":"C1","text":"Exactly one reference, R3, is unverifiable; R1, R2, R4 and R5 pass all three checks (n_verified = 4).","results_keys":["unverifiable_references","n_verified","reference_checks"],"status":"supported","caveat":"Judged only against the provided extracts."},3 {"id":"C2","text":"R3 fails the title check, the year check (2024 vs 2026) and the finding check: the extract reports ~0.5% at the event year rising gradually to 3.4–3.7% after four to five years, not an immediate jump.","results_keys":["reference_checks"],"status":"supported","caveat":"None of the three failures depends on a judgement call."},4 {"id":"C3","text":"The R1, R2, R4 and R5 cited magnitudes, uncertainty statements and populations match printed cells (Table 6 5 km; Table 2 ATE column; Appendix Table 2; Table 3 col (4)).","results_keys":["reference_checks"],"status":"supported","caveat":"R1 Table A19 prints SE 0.218 for the same estimate; R4's SE 0.091 for the own-wage elasticity looks like a duplicated cell in the source (see data_issues)."},5 {"id":"C4","text":"A naive number-presence baseline yields the same unverifiable set but would not detect R3's timing misattribution on its own.","results_keys":["baseline_unverifiable_references","baseline_n_verified"],"status":"supported","caveat":"The baseline flags R3 only via title/year."},6 {"id":"C5","text":"The source extracts contain internal inconsistencies (R1, R4, R5) that do not change any verdict.","results_keys":["data_issues"],"status":"supported","caveat":"Cannot be resolved from the extracts alone."}7]8