Source-linked AI summary
Who Benefits from the "Sharing" Economy of Airbnb?
Giovanni Quattrone, Davide Proserpio, Daniele Quercia, Licia Capra, Mirco Musolesi
TL;DR
Municipalities lack empirical evidence for regulating Airbnb, despite the platform’s rapid growth and contested effects. The paper crawls London listings, matches them with hotel and census data, and finds changing spatial and temporal patterns that motivate responsive, data-informed regulation.
Problem
Municipalities lack empirical evidence about Airbnb adoption, beneficiaries, and socio-economic effects for designing regulation.
Method
The study analyzes crawled London Airbnb data alongside hotel, census, Foursquare, and Ordnance Survey measures.
Results
Airbnb listings cover a broader area than hotels, with rooms especially widespread; Airbnb houses’ overlap with hotel areas increased by 7% from 2012 to 2015.
Takeaways & Limitations
The findings support evidence-informed, algorithmic regulation that uses data analysis to make rules responsive to changing real-time demands.
Takeaways & Limitations
The study is limited to London and uses cross-sectional socio-economic data, making causal mechanisms difficult to study.
Abstract
from arXiv · showhide
Sharing economy platforms have become extremely popular in the last few years, and they have changed the way in which we commute, travel, and borrow among many other activities. Despite their popularity among consumers, such companies are poorly regulated. For example, Airbnb, one of the most successful examples of sharing economy platform, is often criticized by regulators and policy makers. While, in theory, municipalities should regulate the emergence of Airbnb through evidence-based policy making, in practice, they engage in a false dichotomy: some municipalities allow the business without imposing any regulation, while others ban it altogether. That is because there is no evidence upon which to draft policies. Here we propose to gather evidence from the Web. After crawling Airbnb data for the entire city of London, we find out where and when Airbnb listings are offered and, by matching such listing information with census and hotel data, we determine the socio-economic conditions of the areas that actually benefit from the hospitality platform. The reality is more nuanced than one would expect, and it has changed over the years. Airbnb demand and offering have changed over time, and traditional regulations have not been able to respond to those changes. That is why, finally, we rely on our data analysis to envision regulations that are responsive to real-time demands, contributing to the emerging idea of "algorithmic regulation".
1. INTRODUCTION
The paper addresses poorly understood regulation of Airbnb by empirically studying its adoption and socio-economic beneficiaries in London. It analyzes crawled Airbnb data and concludes with data-based regulatory recommendations.
- Sharing-economy platforms use information technology to let users share underutilized goods and services across sectors including accommodation.
- Airbnb’s rapid growth has prompted competing claims about extra income, resource utilization, economic activity, and negative externalities.
- Municipal regulation has often applied existing laws without fully understanding the benefits and drawbacks of new marketplaces.
- The study fills an empirical gap by analyzing Airbnb adoption and socio-economic conditions in London, a diverse city with enthusiastic adoption.
- The authors crawl London Airbnb data from 2012 to 2015, compare benefiting and non-benefiting census areas, and propose five regulatory recommendations.
2. RELATED WORK
Prior work discusses how sharing-economy platforms should be regulated, but this paper argues that empirical evidence about adoption and beneficiaries is missing. It responds with a socio-economic investigation of Airbnb adoption.
- Existing research discusses regulation and policies intended to let sharing-economy platforms operate legally while protecting providers, users, and third parties.
- The paper identifies a lack of empirical evidence about what the sharing economy is, how it is adopted, and who benefits from it.
- The study addresses this gap by empirically investigating Airbnb adoption to support evidence-informed policy making.
3. OVERVIEW
The overview maps Airbnb houses, Airbnb rooms, and hotels across London and quantifies their spatial overlap. Airbnb rooms spread broadly across the city, while Airbnb properties generally cover more area than hotels.
- Hotels cluster mainly in central London and near Heathrow, while Airbnb houses extend into adjacent areas and rooms cover much of London, including suburbs.
- The study compares Airbnb and hotel adoption using fuzzy-set overlap measures computed separately for rooms and houses from 2012 to 2015.
- Airbnb properties, especially rooms, tend to appear in hotel areas, whereas hotels generally do not appear in Airbnb areas.
- Airbnb houses’ overlap with hotel areas increased by 7% from 2012 to 2015, while Airbnb listings overall covered a broader city area than hotels.
- The overview frames the analysis around socio-economic characteristics, listing-type differences, temporal evolution, and customer destinations.
4. DATASETS AND METRICS
The study combines detailed London Airbnb records with socio-economic and neighborhood measures to analyze platform offering and demand. Airbnb offering is normalized by area, while reviews proxy demand.
- The analysis requires detailed Airbnb property records alongside socio-economic data and derived neighborhood metrics.
- Researchers periodically collected consumer-facing London Airbnb information since mid-2012, including host, listing, location, price, availability, and review attributes.
- The dataset contains 14,639 hosts, 17,825 listings, and 220,075 guest reviews collected from March 2012 to June 2015.
- Airbnb offering per area is measured as listings divided by area in square kilometers, with population and dwelling normalizations producing comparable results.
- Airbnb demand per area is measured as reviews divided by area, using reviews as a proxy because guests review stays more than 70% of the time.
4.2 Socio-economic Conditions
The study uses census and deprivation data to characterize London wards socio-economically, adding specialized measures of diversity, creativity, and housing conditions.
- Census Data: 2011 UK census data provide demographic, housing, population-density, education, green-space, ownership, sales, and house-price measures for London wards.
- Cultural Metrics: Ethnic diversity is measured with a Gini-Simpson index calculated across five census ethnicity categories.
- Cultural Metrics: The Bohemian Index measures the fraction of people employed in arts, entertainment, and recreation.
- Deprivation Measures: The analysis also considers a Melting Pot Index and the UK Index of Multiple Deprivation.
- Deprivation Measures: IMD includes seven domains, while this study uses its income and living-environment indexes.
4.3 Attractiveness
Area attractiveness is represented through transport accessibility, Foursquare activity, and the density of selected Ordnance Survey points of interest.
- Attractiveness Measures: Transport accessibility is a census-based metric in which higher values indicate greater public-transport accessibility.
- Foursquare: 26,344,115 Foursquare check-ins collected between 04/03/2014 and 08/04/2014 measure area attractiveness through check-ins per square kilometer.
- Ordnance Survey: The study uses Ordnance Survey data collected in July 2015, comprising 513,786 London metropolitan-area points of interest.
- Ordnance Survey: Attraction density counts eating-and-drinking, attractions, retail, sports, and entertainment points of interest normalized by area size.
4.4 Hotel Data
Hotel presence is measured from Ordnance Survey accommodation points of interest and normalized by the size of each London area.
- Hotel Offering: Hotel offering counts hotels, motels, country houses, and inns within each area.
- Hotel Offering: The hotel_offering metric normalizes these accommodation points of interest by area size.
- Metrics: Table 2 lists the metrics computed for subsequent analyses.
5. METHOD
The method analyzes Airbnb across 625 London wards, aligning socio-economic and platform measures despite a census-data timing gap and testing relationships with OLS models.
- Overview: The study measures Airbnb offerings across city areas and examines their relationship with neighborhood socio-economic conditions.
- Unit of Analysis: Greater London is divided into 625 wards, which serve as the spatial units of analysis.
- Unit of Analysis: Ward-level aggregation averages LSOA deprivation scores, with little information loss because within-ward scores are consistent.
- Temporal Alignment: 2011 census indicators are compared with datasets from 2014/2015 despite a four-year gap, assuming they remain unchanged within that window.
- Regression Approach: OLS regressions relate Airbnb offering, demand, or hotel offering to the remaining socio-economic metrics for each area.
- Regression Approach: The models test residuals for geographic spatial autocorrelation using Moran’s test.
- Data Preparation: Log transformations and z-score standardization address skewed variables and differing scales, enabling comparison of β scores.
6. RESULTS
The results link Airbnb offering and demand to neighborhood socio-economic conditions, while showing that offering evolved temporally and demand remained concentrated near tourism centers. Airbnb listings extend beyond hotel coverage, but flexibility across the city is not fully realized.
- 6.1 Preliminary Analysis: Airbnb listings and demand correlate with attractive, accessible areas populated by young, employed, foreign-born residents, rather than suburban areas dominated by owned houses and flats.These cross-correlations motivate regression analysis because correlated predictors may not retain the same significance levels.
- 6.2 RQ1. Socio-economic Conditions: Distance from the center negatively predicts Airbnb offering, which remains associated with attractive, well-to-do areas and young, tech-savvy residents.The distance measure uses the shortest Euclidean distance from each ward to London’s ten geographic centers.
- 6.4 RQ3. Temporal Adoption: Airbnb offering shifted over time: central geography dominated 2012, while later adoption increasingly involved low-income renters and hosts who did not tend to own their properties.By 2014 and 2015, low income and the number of rented houses were the strongest predictors; the income correlation became increasingly negative.
- 6.5 RQ4. Where Do Airbnb Customers Actually Go?: Airbnb demand showed no comparable temporal evolution from 2012 to 2015 and remained highest in touristic, central, dense areas with many FourSquare check-ins.Reviews serve as a demand proxy because the review-to-stay completion rate exceeds 70%.
- 6.5 RQ4. Where Do Airbnb Customers Actually Go?: Airbnb properties, especially rooms, overlap hotel areas but cover a broader city area; many listings far from tourist areas are not rented out.Thus, the platform’s theoretical flexibility in distributing travelers across diverse areas is not fully exploited in practice.
7. DISCUSSION
The paper proposes evidence-informed, adaptable regulation for short-term rentals, using transferable sharing rights, differentiated rules, data sharing, and ongoing evaluation. Its recommendations reflect Airbnb’s uneven geographic and socio-economic effects across London.
- Municipalities should regulate short-term rentals by considering how, where, when, and what to regulate.
- How: Transferable sharing rights should use real-time market demand and municipal policies to address externalities while capturing opportunities such as decentralized economic activity.Policies may vary across neighborhoods within the same city.
- Where & When: Initial geographic conditions strongly influence which areas ultimately benefit from Airbnb, while Airbnb guests can support local economies through spending in hosting communities.
- Where & When: Airbnb listings cover a wider geographic area than hotels and can consequently distribute tourists across the city.
- Where & When: Sharing rights should account for four factors: adoption consequences, local economic development, sustainable tourism, and avoiding short-term-rental hotspots.
- What: Rooms and entire apartments should face different sharing-right terms because they are associated with different socio-economic conditions and social consequences.Rooms tend to cluster in low-income, highly educated areas with many non-UK-born residents, whereas houses tend to be in wealthy areas; central neighborhoods may experience weakened social fabric.
- Set, Enforce, and Refine: Municipalities should support data sharing and continually evaluate short-term-rental impacts so regulations can be enforced and refined.Data can help identify anomalous behavior and assess effects such as growing demand on public services.
- Limitations: The study is limited to London and uses cross-sectional socio-economic data, making broader generalization and causal analysis difficult.
8. CONCLUSION
The study uses London as a living laboratory to examine Airbnb in a relatively unregulated context and derives five recommendations from changing demand and geographic patterns. It contributes to algorithmic regulation, while calling for broader evidence-informed legal frameworks.
- The work advocates evidence-informed policy making, using short-term-rental data analysis to offer regulatory recommendations.
- The study analyzes Airbnb in London as a living laboratory and develops five recommendations for regulation.
- Changing tourism-area demand over time makes traditional regulations unlikely to respond adequately to short-term-rental dynamics.
- The paper contributes to algorithmic regulation, in which large-scale data analysis produces rules responsive to real-time demands.The approach might also apply to civic issues beyond the sharing economy.
- Future work should develop comprehensive evidence-informed legal frameworks that accommodate sharing-economy activity and visitors while preserving residents’ sense of home.