1{"text": "{\"genre\":\"empirical\",\"subfield\":\"corporate_finance\",\"subfield_secondary\":\"macro_finance\",\"primary_method\":\"reduced_form_regression\",\"identification\":{\"estimand\":\"Effect of city-level local government debt on investment and financing constraints of private versus state-owned manufacturing firms\",\"strategy\":\"Exploits geographic segmentation of China's credit market. Uses city-level and firm-level panel regressions, heterogeneity by firm distance to city border, and Rajan-Zingales external-dependence interactions with city-year, industry-year and industry-city fixed effects. Adds cash-flow sensitivity tests with exogenous and endogenous (switching regression) sample splits.\",\"key_assumptions\":[\"Credit markets are geographically segmented across cities\",\"Industry external financial dependence is technologically determined and comparable across cities\",\"Cash-flow sensitivity of investment indicates financing constraints\",\"Fixed effects absorb confounding local demand shocks\"],\"threats_addressed\":[\"reverse causality\",\"omitted variables\",\"endogeneity of industry financing needs to local conditions\",\"misclassification from exogenous sample splits\"],\"causal_language_strength\":4,\"uses_causal_claim\":true},\"abstract_moves\":{\"question\":{\"present\":false,\"sentence_index\":-1},\"gap\":{\"present\":false,\"sentence_index\":-1},\"method\":{\"present\":true,\"sentence_index\":1},\"results\":{\"present\":true,\"sentence_index\":0},\"magnitude\":{\"present\":false,\"sentence_index\":-1},\"contribution\":{\"present\":true,\"sentence_index\":1}},\"introduction\":{\"moves\":[{\"paragraph_index\":0,\"move\":\"motivation\"},{\"paragraph_index\":0,\"move\":\"preview_of_results\"},{\"paragraph_index\":0,\"move\":\"identification_defense\"},{\"paragraph_index\":0,\"move\":\"mechanism_or_theory_sketch\"},{\"paragraph_index\":1,\"move\":\"preview_of_approach\"},{\"paragraph_index\":1,\"move\":\"preview_of_results\"},{\"paragraph_index\":2,\"move\":\"preview_of_approach\"},{\"paragraph_index\":2,\"move\":\"preview_of_results\"},{\"paragraph_index\":3,\"move\":\"related_literature_positioning\"},{\"paragraph_index\":3,\"move\":\"literature_gap\"},{\"paragraph_index\":4,\"move\":\"related_literature_positioning\"},{\"paragraph_index\":4,\"move\":\"contribution_statement\"}],\"contribution_paragraph_index\":4,\"contribution_types\":[\"new_evidence\",\"new_data_or_measure\",\"new_identification\"],\"roadmap_present\":false,\"literature_positioning_location\":\"inside_introduction\"},\"section_roles\":[{\"idx\":0,\"role\":\"introduction\"},{\"idx\":1,\"role\":\"data\"},{\"idx\":2,\"role\":\"empirical_design\"},{\"idx\":3,\"role\":\"results\"},{\"idx\":4,\"role\":\"results\"},{\"idx\":5,\"role\":\"results\"},{\"idx\":6,\"role\":\"results\"},{\"idx\":7,\"role\":\"results\"},{\"idx\":8,\"role\":\"conclusion\"}],\"results_presentation\":{\"economic_significance_discussed\":true,\"statistical_significance_language\":\"unclear\",\"magnitude_statements_in_intro\":false},\"robustness\":{\"location\":\"mixed\",\"types\":[\"alternative_specification\",\"alternative_measure\",\"other\"]},\"tables\":{\"convention_notes\":\"Table notes define the dependent variable and how it is computed, then list the explanatory variables with their abbreviations. A separate variable-definition table gives data sources (ASIF, ATS). Regression tables move from city-level to firm-level to industry-level specifications. The switching regression table reports separate selection and outcome panels.\",\"units_stated_in_notes\":true,\"standard_errors_description\":\"unclear\"},\"figures\":{\"uncertainty_shown\":\"unclear\"},\"conclusion\":{\"bounded_claims\":true,\"policy_implications\":true,\"open_questions\":false,\"restates_contribution\":true},\"hedging\":{\"level\":\"moderate\",\"typical_devices\":[\"'our estimates suggest'\",\"'is likely to'\",\"'could undercut'\",\"'consistent with'\",\"calling correlational evidence 'preliminary'\",\"calling estimates 'conservative'\"]},\"claim_evidence\":[{\"claim\":\"Cities with more local public debt show lower manufacturing investment\",\"evidence_type\":\"table\",\"evidence_ref\":\"Table III, Table V\"},{\"claim\":\"Local debt is associated with lower capital productivity in city manufacturing\",\"evidence_type\":\"table\",\"evidence_ref\":\"Table IV\"},{\"claim\":\"Firms near city borders, with access to banks in other cities, are less affected by crowding-out\",\"evidence_type\":\"table\",\"evidence_ref\":\"Table VII\"},{\"claim\":\"Local debt reduces investment more for financially dependent private firms but not SOEs\",\"evidence_type\":\"table\",\"evidence_ref\":\"Table VIII, Figure 3\"},{\"claim\":\"Local debt raises investment–cash flow sensitivity for private and small firms\",\"evidence_type\":\"table\",\"evidence_ref\":\"Table IX, Table X, Figure 4\"},{\"claim\":\"In a switching regression, debt affects cash-flow sensitivity only for constrained firms; private and small firms are more often constrained\",\"evidence_type\":\"table\",\"evidence_ref\":\"Table XII\"},{\"claim\":\"The purpose-built local debt data line up with official NAO aggregates\",\"evidence_type\":\"figure\",\"evidence_ref\":\"Figure 1\"},{\"claim\":\"China's credit market is geographically segmented\",\"evidence_type\":\"figure\",\"evidence_ref\":\"Figure 2\"}],\"structural_summary\":\"The paper opens with the post-2008 surge in Chinese local debt and states the crowding-out finding upfront. It argues that geographic credit-market segmentation makes China a good test setting, then previews three escalating identification approaches before positioning itself against the public debt–growth and Chinese stimulus literatures. The body proceeds in order: novel debt data, documentation of segmentation, preliminary city-level and firm-level correlations, border-distance heterogeneity, Rajan-Zingales industry interactions, and cash-flow sensitivity tests with exogenous and then endogenous splits. Each section motivates itself by the weakness of the previous approach. The conclusion revisits the stimulus and warns of costs from capital misallocation.\",\"writing_profile_keywords\":[\"institutional setting as identification\",\"novel hand-collected data\",\"layered identification strategies\",\"heterogeneity tests\",\"financing constraints\",\"state-owned vs private firms\",\"policy relevance\"],\"confidence\":0.8,\"notes\":\"Standard error treatment and significance conventions are not visible in the excerpted table notes. No explicit roadmap paragraph appears in the excerpts. The intro excerpt omits the description of the first approach, which is likely city-level regressions.\"}", "json": {"genre": "empirical", "subfield": "corporate_finance", "subfield_secondary": "macro_finance", "primary_method": "reduced_form_regression", "identification": {"estimand": "Effect of city-level local government debt on investment and financing constraints of private versus state-owned manufacturing firms", "strategy": "Exploits geographic segmentation of China's credit market. Uses city-level and firm-level panel regressions, heterogeneity by firm distance to city border, and Rajan-Zingales external-dependence interactions with city-year, industry-year and industry-city fixed effects. Adds cash-flow sensitivity tests with exogenous and endogenous (switching regression) sample splits.", "key_assumptions": ["Credit markets are geographically segmented across cities", "Industry external financial dependence is technologically determined and comparable across cities", "Cash-flow sensitivity of investment indicates financing constraints", "Fixed effects absorb confounding local demand shocks"], "threats_addressed": ["reverse causality", "omitted variables", "endogeneity of industry financing needs to local conditions", "misclassification from exogenous sample splits"], "causal_language_strength": 4, "uses_causal_claim": true}, "abstract_moves": {"question": {"present": false, "sentence_index": -1}, "gap": {"present": false, "sentence_index": -1}, "method": {"present": true, "sentence_index": 1}, "results": {"present": true, "sentence_index": 0}, "magnitude": {"present": false, "sentence_index": -1}, "contribution": {"present": true, "sentence_index": 1}}, "introduction": {"moves": [{"paragraph_index": 0, "move": "motivation"}, {"paragraph_index": 0, "move": "preview_of_results"}, {"paragraph_index": 0, "move": "identification_defense"}, {"paragraph_index": 0, "move": "mechanism_or_theory_sketch"}, {"paragraph_index": 1, "move": "preview_of_approach"}, {"paragraph_index": 1, "move": "preview_of_results"}, {"paragraph_index": 2, "move": "preview_of_approach"}, {"paragraph_index": 2, "move": "preview_of_results"}, {"paragraph_index": 3, "move": "related_literature_positioning"}, {"paragraph_index": 3, "move": "literature_gap"}, {"paragraph_index": 4, "move": "related_literature_positioning"}, {"paragraph_index": 4, "move": "contribution_statement"}], "contribution_paragraph_index": 4, "contribution_types": ["new_evidence", "new_data_or_measure", "new_identification"], "roadmap_present": false, "literature_positioning_location": "inside_introduction"}, "section_roles": [{"idx": 0, "role": "introduction"}, {"idx": 1, "role": "data"}, {"idx": 2, "role": "empirical_design"}, {"idx": 3, "role": "results"}, {"idx": 4, "role": "results"}, {"idx": 5, "role": "results"}, {"idx": 6, "role": "results"}, {"idx": 7, "role": "results"}, {"idx": 8, "role": "conclusion"}], "results_presentation": {"economic_significance_discussed": true, "statistical_significance_language": "unclear", "magnitude_statements_in_intro": false}, "robustness": {"location": "mixed", "types": ["alternative_specification", "alternative_measure", "other"]}, "tables": {"convention_notes": "Table notes define the dependent variable and how it is computed, then list the explanatory variables with their abbreviations. A separate variable-definition table gives data sources (ASIF, ATS). Regression tables move from city-level to firm-level to industry-level specifications. The switching regression table reports separate selection and outcome panels.", "units_stated_in_notes": true, "standard_errors_description": "unclear"}, "figures": {"uncertainty_shown": "unclear"}, "conclusion": {"bounded_claims": true, "policy_implications": true, "open_questions": false, "restates_contribution": true}, "hedging": {"level": "moderate", "typical_devices": ["'our estimates suggest'", "'is likely to'", "'could undercut'", "'consistent with'", "calling correlational evidence 'preliminary'", "calling estimates 'conservative'"]}, "claim_evidence": [{"claim": "Cities with more local public debt show lower manufacturing investment", "evidence_type": "table", "evidence_ref": "Table III, Table V"}, {"claim": "Local debt is associated with lower capital productivity in city manufacturing", "evidence_type": "table", "evidence_ref": "Table IV"}, {"claim": "Firms near city borders, with access to banks in other cities, are less affected by crowding-out", "evidence_type": "table", "evidence_ref": "Table VII"}, {"claim": "Local debt reduces investment more for financially dependent private firms but not SOEs", "evidence_type": "table", "evidence_ref": "Table VIII, Figure 3"}, {"claim": "Local debt raises investment–cash flow sensitivity for private and small firms", "evidence_type": "table", "evidence_ref": "Table IX, Table X, Figure 4"}, {"claim": "In a switching regression, debt affects cash-flow sensitivity only for constrained firms; private and small firms are more often constrained", "evidence_type": "table", "evidence_ref": "Table XII"}, {"claim": "The purpose-built local debt data line up with official NAO aggregates", "evidence_type": "figure", "evidence_ref": "Figure 1"}, {"claim": "China's credit market is geographically segmented", "evidence_type": "figure", "evidence_ref": "Figure 2"}], "structural_summary": "The paper opens with the post-2008 surge in Chinese local debt and states the crowding-out finding upfront. It argues that geographic credit-market segmentation makes China a good test setting, then previews three escalating identification approaches before positioning itself against the public debt–growth and Chinese stimulus literatures. The body proceeds in order: novel debt data, documentation of segmentation, preliminary city-level and firm-level correlations, border-distance heterogeneity, Rajan-Zingales industry interactions, and cash-flow sensitivity tests with exogenous and then endogenous splits. Each section motivates itself by the weakness of the previous approach. The conclusion revisits the stimulus and warns of costs from capital misallocation.", "writing_profile_keywords": ["institutional setting as identification", "novel hand-collected data", "layered identification strategies", "heterogeneity tests", "financing constraints", "state-owned vs private firms", "policy relevance"], "confidence": 0.8, "notes": "Standard error treatment and significance conventions are not visible in the excerpted table notes. No explicit roadmap paragraph appears in the excerpts. The intro excerpt omits the description of the first approach, which is likely city-level regressions."}, "model": "claude-opus-5-5", "stop_reason": "end_turn", "latency_s": 19.44, "usage": {"input": 7392, "output": 2284, "cache_read": 3968, "cache_write": 0}, "cost_usd": 0.076042, "request_id": "req_011CggqzcB1oPyYyFVaknH6d", "cached": false, "ts": "2026-11-03T23:15:51+00:00", "effort": "medium"}