1{2 "validation_errors": [],3 "doc_id": "jof_12846",4 "ontology_version": "1.0",5 "prompt_version": "1.0",6 "model": "claude-opus-5-5",7 "effort": "medium",8 "annotated_at": "2026-11-03T23:11:25+00:00",9 "source_sha256": "b5dfbca07d3c9237c18da695ac6d632cf4c1c19b892632c4427053c330357407",10 "parser": "pymupdf-jof-heuristics-1.3",11 "payload_chars": 29549,12 "usage": {13 "input": 10292,14 "output": 3314,15 "cache_read": 3968,16 "cache_write": 017 },18 "cost_usd": 0.108242,19 "request_id": "req_011CggqePv1tCxVYwG3KSGDU",20 "annotation": {21 "genre": "empirical",22 "subfield": "corporate_finance",23 "subfield_secondary": "law and finance / credit access and financial development",24 "primary_method": "difference_in_differences",25 "identification": {26 "estimand": "Effect of France's 2006 collateral reform, which allowed nonpossessory pledges of physical assets, on the borrowing, real outcomes, start-up formation and survival, and credit-access inequality of private French firms.",27 "strategy": "Matched difference-in-differences comparing high- versus low-fixed-assets-intensity firms before and after 2006. Firms are propensity-score matched on predetermined traits, with firm and industry-year fixed effects. Triple-difference splits by size, age and financial constraints; hazard models for bankruptcy; spatial and Gini analyses; Wurgler-style allocation-efficiency regressions.",28 "key_assumptions": [29 "Absent the reform, high- and low-fixed-assets firms would have followed parallel trends.",30 "Pre-reform fixed-asset intensity is a valid proxy for exposure to newly pledgeable assets.",31 "No contemporaneous shocks differentially affect high-fixed-assets firms; other law changes are discussed in the institutional section."32 ],33 "threats_addressed": [34 "Pre-reform differences between treated and control firms (matching; Table I balance)",35 "Pre-existing differential trends (parallel trends tests)",36 "Generic shocks to asset-intensive firms (placebo regressions on other Napoleonic-Code countries)",37 "Non-collateral channels (short-term unsecured debt used as a falsification outcome)",38 "Rural effect driven by industry or size composition (spatial distribution checks)"39 ],40 "causal_language_strength": 4,41 "uses_causal_claim": true42 },43 "abstract_moves": {44 "question": {45 "present": false,46 "sentence_index": -147 },48 "gap": {49 "present": false,50 "sentence_index": -151 },52 "method": {53 "present": true,54 "sentence_index": 155 },56 "results": {57 "present": true,58 "sentence_index": 159 },60 "magnitude": {61 "present": false,62 "sentence_index": -163 },64 "contribution": {65 "present": false,66 "sentence_index": -167 }68 },69 "introduction": {70 "moves": [71 {72 "paragraph_index": 0,73 "move": "other"74 },75 {76 "paragraph_index": 4,77 "move": "motivation"78 },79 {80 "paragraph_index": 4,81 "move": "literature_gap"82 },83 {84 "paragraph_index": 4,85 "move": "preview_of_approach"86 },87 {88 "paragraph_index": 5,89 "move": "preview_of_results"90 },91 {92 "paragraph_index": 5,93 "move": "identification_defense"94 },95 {96 "paragraph_index": 6,97 "move": "preview_of_results"98 },99 {100 "paragraph_index": 7,101 "move": "preview_of_approach"102 },103 {104 "paragraph_index": 7,105 "move": "preview_of_results"106 },107 {108 "paragraph_index": 8,109 "move": "related_literature_positioning"110 },111 {112 "paragraph_index": 8,113 "move": "preview_of_results"114 },115 {116 "paragraph_index": 8,117 "move": "policy_or_broader_implication"118 },119 {120 "paragraph_index": 9,121 "move": "preview_of_results"122 },123 {124 "paragraph_index": 9,125 "move": "contribution_statement"126 },127 {128 "paragraph_index": 9,129 "move": "related_literature_positioning"130 }131 ],132 "contribution_paragraph_index": 9,133 "contribution_types": [134 "new_evidence"135 ],136 "roadmap_present": false,137 "literature_positioning_location": "inside_introduction"138 },139 "section_roles": [140 {141 "idx": 0,142 "role": "introduction"143 },144 {145 "idx": 1,146 "role": "other"147 },148 {149 "idx": 2,150 "role": "data"151 },152 {153 "idx": 3,154 "role": "results"155 },156 {157 "idx": 4,158 "role": "results"159 },160 {161 "idx": 5,162 "role": "results"163 },164 {165 "idx": 6,166 "role": "results"167 },168 {169 "idx": 7,170 "role": "robustness"171 },172 {173 "idx": 8,174 "role": "conclusion"175 }176 ],177 "results_presentation": {178 "economic_significance_discussed": true,179 "statistical_significance_language": "unclear",180 "magnitude_statements_in_intro": true181 },182 "robustness": {183 "location": "mixed",184 "types": [185 "placebo",186 "falsification",187 "subsample",188 "alternative_specification",189 "other"190 ]191 },192 "tables": {193 "convention_notes": "Table notes write out the estimating equation (firm and industry-year fixed effects, Post×Treated), define outcome columns and panels, and describe the matched sample. Hazard-model specifications are given in the notes where used. Descriptive tables report means before and after the reform plus differences. Standard-error treatment is not visible in the excerpts.",194 "units_stated_in_notes": true,195 "standard_errors_description": "unclear"196 },197 "figures": {198 "uncertainty_shown": "no"199 },200 "conclusion": {201 "bounded_claims": false,202 "policy_implications": true,203 "open_questions": false,204 "restates_contribution": true205 },206 "hedging": {207 "level": "moderate",208 "typical_devices": [209 "'suggests' / 'consistent with' for interpretive claims",210 "'may lead' for extrapolations",211 "confident 'we show/find' for core estimates"212 ]213 },214 "claim_evidence": [215 {216 "claim": "The reform raised long-term leverage and cut zero-debt incidence mainly for high-fixed-assets private firms; short-term debt was unaffected.",217 "evidence_type": "table",218 "evidence_ref": "Table II; Figures 2-4"219 },220 {221 "claim": "Treated firms increased investment, employment, sales and profitability and had lower profit volatility and bankruptcy hazard.",222 "evidence_type": "table",223 "evidence_ref": "Table III"224 },225 {226 "claim": "Small, young and financially constrained firms gained the most credit access.",227 "evidence_type": "table",228 "evidence_ref": "Tables IV-V"229 },230 {231 "claim": "Rural firms gained most, and departmental credit Gini indices fell sharply, especially in rural departments.",232 "evidence_type": "figure",233 "evidence_ref": "Figures 5-6"234 },235 {236 "claim": "Start-ups had higher at-inception leverage, located farther from large cities, and had lower exit and bankruptcy rates after the reform.",237 "evidence_type": "table",238 "evidence_ref": "Tables VI-VII; Figure 7"239 },240 {241 "claim": "Cross-industry capital allocation efficiency improved, more so in fixed-asset-heavy and externally dependent sectors.",242 "evidence_type": "table",243 "evidence_ref": "Tables VIII-IX"244 },245 {246 "claim": "Treated and control firms show parallel pre-reform trends.",247 "evidence_type": "table",248 "evidence_ref": "Table X"249 }250 ],251 "structural_summary": "Opens with a political epigraph, then motivates the topic with the Napoleonic possessory-ownership regime and an understanding gap. The introduction previews results in sequence: credit, real outcomes, democratization by firm type and geography, then macro allocation efficiency, closing with law-and-finance positioning. The body describes the institutional setting and identification logic, then data and the matched DID design. Results sections cover the average firm, distributional and spatial effects, start-ups, and capital allocation efficiency. A dedicated robustness section follows, and the conclusion restates findings and draws policy implications.",252 "writing_profile_keywords": [253 "natural policy experiment",254 "matched difference-in-differences",255 "law and finance",256 "collateral",257 "credit democratization",258 "spatial analysis",259 "intro magnitude previews",260 "policy-oriented"261 ],262 "confidence": 0.8,263 "notes": "Abstract text includes author footnotes, which are ignored for sentence indexing. Introduction paragraphs 0-3 are epigraph fragments. Section I is labeled theory by rule but is an institutional and identification discussion, so it is coded 'other'. Standard-error and significance conventions are not visible in the excerpts."264 },265 "deterministic_checks": {266 "roadmap_regex": false,267 "contribution_regex_in_intro": true,268 "magnitude_in_abstract": false,269 "abstract_n_sentences": 16,270 "intro_n_paragraphs": 11,271 "intro_n_words": 1282,272 "causal_words_in_abstract": 0,273 "hedge_words_in_abstract": 0,274 "stars_mentioned": false,275 "clustered_se_mentioned": false,276 "internet_appendix_mentioned": true,277 "section_roles_rule": {278 "0": "introduction",279 "1": "theory",280 "2": "data",281 "3": "other",282 "4": "other",283 "5": "other",284 "6": "other",285 "7": "robustness",286 "8": "conclusion"287 }288 },289 "schema_valid": true,290 "code_version": "87c73cd-dirty"291}