Urban Tourism in the Global South: Seasonality, Host Professionalisation, Digital Reputation and Location in the Short-Term Rental Markets of Mexico City and Santiago de Chile

This repository contains the replication package and processed results for the chapter: "Urban Tourism in the Global South: Seasonality, Host Professionalisation, Digital Reputation and Location in the Short-Term Rental Markets of Mexico City and Santiago de Chile"Authors: Patricio Torres Luque, Pablo Esteban Torres Luque, and Nazario Pescador Peredo (2026). Forthcoming in the ICMTT 2026 proceedings (Springer). Data OverviewThe scripts use the publicly available Inside Airbnb listings and reviews files for Mexico City and Santiago de Chile, scraped on 30–31 December 2025. Monthly activity is measured with reviews posted between December 2024 and November 2025. Listing-level occupancy is Inside Airbnb's estimate for the 365 days preceding the scrape. The analytical sample comprises listings with at least one review in that window and an available review score: 19,111 in Mexico City and 12,420 in Santiago de Chile. Files Included replication_audit_v3.py: Python script (version 3) reproducing the sample sizes, monthly review indicators and sensitivity analyses, OLS models with HC3, host- and unit-clustered standard errors, standardised coefficients, ANOVA, global and residual Moran's I, and Figure 1 (requires pandas, statsmodels, scipy, esda, libpysal, matplotlib). Together with the version 2 script, it reproduces all results reported in the chapter. audit_results_v3.xlsx: Output tables from the version 3 script. fig1_seasonality-2.png: Figure 1, monthly share of reviews (PNG, 300 dpi). optionA_str_cdmx_scl.py and optionA_results.xlsx: Version 2 script and outputs (descriptive statistics in Table 2, unit means and LISA statistics in Table 4). STR_processed_README-2.txt: Variable definitions and execution instructions. Key Methodological Notes Monthly activity is proxied by guest review dates (418,453 reviews in Mexico City; 229,750 in Santiago). Reviews are an imperfect proxy of completed stays. Occupancy is Inside Airbnb's review-based estimated_occupancy_l365d, a modelled estimate capped at 255 nights, not observed bookings. Host professionalisation is measured as multi-listing hosts (calculated_host_listings_count > 1). Raw data are not redistributed; they are available from Inside Airbnb (http://insideairbnb.com/get-the-data).