1{2 "record_id": "6e4be5daf59355440452370cc31893fbd8320793",3 "eligible": false,4 "exclusion_reason": "Repeat-sales difference-in-differences model (Tables 3-6): dependent variable is the change in log price between two sales of the same property; regressors are a time-bracket indicator (B) and its interactions with property type, rural/coastal location, flood-defence construction (B*D) and defence design variables. Property characteristics are differenced out and enter only as interactions; no hedonic price-level regression with characteristic coefficients. Flagged for human review: the flood-defence treatment effect (urban +12.6% to +16.7%, rural -0.8% to -5.0%) is an environmental (flood_risk) effect but not a hedonic coefficient.",5 "study": {6 "title": "Assessing the Economic Benefits of Flood Defenses: A Repeat-Sales Approach",7 "year": 2018,8 "doi": "10.1111/risa.13136",9 "country": "United Kingdom",10 "region_city": "England",11 "sample_start": 1995,12 "sample_end": 2014,13 "property_type": "mixed",14 "price_type": "transaction",15 "data_source": "HM Land Registry transactions (over 12 million) merged with GIS data on 1,666 flood defences (Environment Agency)",16 "document_type": "journal",17 "estimators": [18 "other"19 ],20 "spatial_method": "repeat-sales DiD with county fixed effects",21 "time_fe": true,22 "location_fe": true,23 "se_type": "robust",24 "n_obs_total": 7217966,25 "ai_methods": [],26 "notes": "Repeat-sales (Table 3: 7,217,966 sale pairs, 1,824 treated; Table 3 column 2 floodplain sample 1,135,690). Robust SEs in parentheses; * 10%, ** 5%, *** 1%."27 },28 "models": [],29 "coefficients": []30}