Source-linked AI summary
Airbnb in tourist cities: comparing spatial patterns of hotels and peer-to-peer accommodation
Javier Gutierrez, Juan Carlos Garcia-Palomares, Gustavo Romanillos, Maria Henar Salas-Olmedo
TL;DR
Airbnb’s rapid expansion has raised questions about how peer-to-peer accommodation is distributed within tourist cities. This study compares Airbnb listings with hotels, sightseeing spots, and resident populations in Barcelona using geolocated data and spatial autocorrelation. It finds a strong centre-periphery relationship, with Airbnb extending around the main hotel axis and aligning more closely with tourist attractions, while expansion adds pressure in several residential areas.
Problem
The study addresses limited evidence on where Airbnb accommodation is located in mass-tourism cities and how its distribution relates to hotels, attractions, and residents.
Method
The authors analyze Inside Airbnb listings, hotel registry records, and geolocated Panoramio photographs using census-section spatial analysis and 1-kilometre distance-weighted autocorrelation.
Results
Airbnb and hotels show close spatial associations and clear centre-periphery patterns, with Airbnb concentrated in central areas and extending beyond the main hotel axis.
Takeaways & Limitations
Airbnb is more closely aligned with major sightseeing concentrations than hotels and adds residential areas to zones experiencing tourism pressure.
Takeaways & Limitations
The study is limited to Barcelona and uses a 1-kilometre spatial-interaction radius.
Abstract
from arXiv · showhide
In recent years, what has become known as collaborative consumption has undergone rapid expansion through peer-to-peer (P2P) platforms. In the field of tourism, a particularly notable example is that of Airbnb. This article analyses the spatial patterns of Airbnb in Barcelona and compares them with hotels and sightseeing spots. New sources of data, such as Airbnb listings and geolocated photographs are used. Analysis of bivariate spatial autocorrelation reveals a close spatial relationship between Airbnb and hotels, with a marked centre-periphery pattern, although Airbnb predominates around the main hotel axis and hotels predominate in some peripheral areas of the city. Another interesting finding is that Airbnb capitalises more on the advantages of proximity to the main tourist attractions of the city than does the hotel sector. Finally, it was possible to detect those parts of the city that have seen the greatest increase in pressure from tourism related to Airbnb's recent expansion.
1. Introduction
Collaborative consumption has expanded through online peer-to-peer platforms, with Airbnb becoming a major tourism accommodation example. The study addresses limited research on Airbnb’s spatial distribution by comparing listings with hotels, attractions, and tourism pressure in Barcelona.
- Sharing economy and P2P accommodation: Collaborative consumption enables peer-to-peer access to goods and services through community-based online platforms.It emphasizes using rather than owning and can reduce users’ costs and resource waste.
- Sharing economy and P2P accommodation: Airbnb connects hosts with spare space to guests seeking accommodation and has expanded rapidly worldwide.The platform mainly offers entire homes or apartments and private rooms.
- Sharing economy and P2P accommodation: Airbnb’s appeal combines cost savings, household amenities, and potential local experiences, despite disadvantages in service quality, professionalism, reputation, and security.Reviews, ratings, profiles, messaging, and customer support are described as mechanisms that foster trust.
- Research gap and study aim: Airbnb may disperse tourism spending beyond hotel districts while also concentrating in central areas where existing buildings facilitate expansion.Historic centres may be easier for Airbnb to enter than for hotels because Airbnb does not require whole buildings and comparable zoning permits.
- Research gap and study aim: Prior research largely examined Airbnb as disruptive innovation or a competitor, while not examining the spatial distribution of its listings.This study compares Airbnb and hotel densities with sightseeing locations using GIS and spatial statistics.
- Research gap and study aim: The study also relates accommodation locations to tourist attractions and resident populations to examine locational advantages and tourism pressure.Crowding is framed through carrying capacity and sustainability thresholds.
2. Study Area: Barcelona
The study focuses on Barcelona, a compact, densely populated historic city experiencing substantial mass tourism and a large concentration of tourist accommodation.
- Study area: Barcelona’s municipality covers 10,130 hectares and contains 1,604,000 inhabitants.Its population density exceeds 158.3 inhabitants per hectare.
- Study area: The study area is the municipality of Barcelona, a historic city described as facing serious mass-tourism problems.The municipality is characterized as relatively compact and densely populated.
- Tourism context: Barcelona was the fifth European city by international tourist numbers in 2014 and ranked among the world’s twenty-five favourite city destinations.Its popularity increased considerably after hosting the 1992 Olympic Games.
3. Data
The study combines geolocated Airbnb listings, hotel registry data, and geolocated tourist photographs to compare accommodation distributions and sightseeing concentrations in Barcelona.
- Accommodation data: Airbnb data came from Inside Airbnb and included listing locations, 365-day availability calendars, and reviews for each listing.The source compiles public information from the Airbnb website and covers more than 30 cities.
- Accommodation data: October 2015 Airbnb records contained coordinates and listing characteristics including room type, activity, availability, and host-level information.Coordinates were mapped in ArcGIS, revealing a clear concentration of listings in the city centre.
- Accommodation data: Hotel data came from the Catalonia Tourism Registry and included room and bed counts plus postal addresses geolocated with ArcGIS address matching.The registry was compiled and updated weekly by the regional government.
- Sightseeing data: Geolocated Panoramio photographs supplied geographic coordinates and related metadata for identifying sightseeing hotspots.Photograph coordinates were used to create GIS point layers.
- Spatial data visualization: Figure 2 maps hotel and Airbnb offers alongside the density of photographs taken by tourists and residents.The figure links accommodation locations with a spatial indicator of photographed activity.
4. Methodology
The methodology aggregates accommodation and photograph data by census tract, maps their densities, and applies univariate and bivariate spatial autocorrelation to identify clusters and relationships.
- Data preparation: Census-tract aggregation produced density maps and descriptive statistics for hotels, Airbnb, and tourist photographs.These measures assessed accommodation intensity and concentration.
- Data preparation: Accommodation distributions were normalized so the relative predominance of hotels and Airbnb could be compared across census sections.Normalization addressed different variable ranges.
- Spatial analysis: Univariate spatial autocorrelation identified clusters in accommodation and photograph distributions, while bivariate analysis examined relationships between accommodation types and photographs.The workflow also calculated accommodation rates per 1,000 inhabitants to analyze tourism pressure on residents.
- Spatial analysis: Global Moran’s I measured spatial autocorrelation, and Anselin Local Moran’s I identified local accommodation patterns and cluster types.LISA distinguishes High-High, Low-Low, and outlier configurations.
- Spatial weights: Spatial interaction was modeled within a 1-kilometre radius, corresponding to a typical 15-minute walk, with weights inversely proportional to distance.This distance rule defines which observations contribute to the spatial statistics.
- Spatial analysis: GeoDa computed univariate and bivariate Global and Anselin Local Moran’s I statistics.The software provided the spatial-analysis tools used in the study.
5. Results
Barcelona’s hotels and Airbnb accommodations share a statistically significant centre–periphery structure, but Airbnb spreads more broadly through central residential districts and aligns more closely with tourist attractions. Airbnb also extends tourism pressure beyond the main hotel axis into residential areas.
- Distribution of tourist accommodations: Hotels concentrate along the Ramblas–Paseo de Gracia axis and selected business and coastal areas, while Airbnb is less spatially concentrated.Airbnb averages 48 accommodations per census section versus 69 for hotels; maximum values approach 600 for Airbnb and exceed 2,000 for hotels.
- Distribution of tourist accommodations: Normalised densities place Airbnb in a broader concentric ring around Plaza de Cataluña, including residential districts where its relative presence exceeds hotels.These areas include El Raval, La Barceloneta, La Ribera, the Gothic Quarter, and the area around Sagrada Familia.
- Spatial clustering: Both accommodation types show strong positive spatial autocorrelation, with high–high clusters in the centre and low–low clusters on the periphery.Airbnb’s high–high clusters extend across central residential sections, whereas hotel clusters concentrate along the main hotel axis.
- Sightseeing spots and accommodation: Tourist photographs cluster around Las Ramblas, the Gothic Quarter, Sagrada Familia, and Paseo de Gracia, with weaker spatial autocorrelation than hotels or Airbnb.The photograph distribution identifies two high–high zones: the historic centre and an area centred on Sagrada Familia.
- Hotels versus Airbnb: The hotel–Airbnb relationship forms a clear centre–periphery pattern, with Airbnb-dominant central residential outliers and hotel-dominant outliers in some peripheral areas.High–high clusters occur along the Ramblas–Plaza de Cataluña–Paseo de Gracia axis; low–low clusters prevail on the periphery.
- Sightseeing spots and accommodation: Bivariate spatial autocorrelation is stronger for Airbnb than hotels in relation to tourist photographs, indicating closer proximity to sightseeing areas.Airbnb high–high sections extend into the Ensanche around Sagrada Familia, where hotels often form low–high patterns.
- Pressure from tourism on residential areas: Hotels generate more than 500 accommodations per 1,000 inhabitants in several sections along the main hotel axis, while Airbnb pressure declines more gradually toward the periphery.Airbnb exceeds 100 places per 1,000 inhabitants in some sections and reaches almost 400; pressure also extends into central residential areas.
6. Conclusions
The study identifies a clear centre-periphery structure in Barcelona’s Airbnb and hotel accommodation, with Airbnb extending into central residential areas and differing from hotels in its relationship to attractions. Airbnb’s expansion also adds pressure to residential areas and contributes to conflicts with residents and local businesses.
- Spatial patterns: Airbnb accommodation in Barcelona follows a clear centre-periphery pattern and extends beyond the main hotel axis into central residential districts.Its spatial distribution has greater positive spatial autocorrelation and more regular HH-to-LL clustering than hotels.
- Spatial patterns: Airbnb and hotels are closely associated spatially, but their differences concentrate in outlier areas: Airbnb prevails centrally while hotels predominate peripherally.The bivariate pattern moves from hotel-dominated central HH clusters through Airbnb-dominated LH areas toward peripheral LL clusters, with some hotel-dominated HL outliers.
- Tourist attractions: Airbnb benefits more than hotels from proximity to Barcelona’s most visited attractions.The accommodation-attraction maps show similar distributions, but Airbnb has more HH sections and hotels have more outliers, including peripheral hotel locations.
- Tourism pressure: Airbnb’s expansion adds residential areas to existing tourism-pressure zones and contributes to conflicts with residents in several central neighbourhoods.The study links these conflicts to entire homes being removed from the ordinary rental market, increased rents, displacement of local residents, and businesses becoming more tourist-oriented.
- Policy response: Barcelona is attempting to control Airbnb’s expansion through inspections, tax enforcement, and fines of up to 90,000 euros.These measures are described as reducing Airbnb’s competitive advantage over traditional accommodation and its prospects for further expansion.