1# RECORD 87f031f0ad9dc17cd7151e97fc7ad56a0e7b728f2# TITLE: Spatial Variability of the 'Airbnb Effect': A Spatially Explicit Analysis of Airbnb's Impact on Housing Prices in Sydney3# YEAR: 2021.0 VENUE: ISPRS Int. J. Geo Inf. DOI: 10.31235/osf.io/cw2a94# AUTHORS: William Thackway; M. Ng; C. Lee; Vivien Shi; C. Pettit5# PAGES: 17 CANDIDATE TABLE PAGES (1-based): [10, 11, 12]6# ABSTRACT: Over the last decade, the emergence and significant growth of home sharing platforms such as Airbnb has coincided with rising housing unaffordability in many global cities. It is in this context that we look to empirically assess the impact of Airbnb on housing prices in Sydney - one of the least affordable cities in the world. Employing a hedonic property valuation model, our results indicate that Airbnb’s overall effect is positive. A 1% increase in Airbnb density is associated with approximately a 2% increase in property sales price. However, recognising that Airbnb’s effect is geographically uneven and given the fragmented nature of Sydney’s housing market, we also employ a GWR to account for the spatial variation in Airbnb activity. The findings confirm that Airbnb’s influence on housing prices is varied across the city. Sydney’s northern beaches and parts of western Sydney experience a statistically significant value uplift attributable to Airbnb activity. However, traditional tourist locations focused around Sydney’s CBD and the eastern suburbs experience insignificant or negative property price impacts. The results highlight the need for policymakers to consider local Airbnb and housing market contexts when deciding the appropriate level and design of Airbnb regulation.789===== PAGE 1 (score 0, numeric-paren 0, table-words 0, var-words 0) =====1011Spatial Variability of the ‘Airbnb Effect’: A Spatially Explicit12Analysis of Airbnb’s Impact on Housing Prices in Sydney13Author:14Thackway, WT; Ng, MKM; Lee, CL; Shi, V; Pettit, CJ15Publication details:16ISPRS International Journal of Geo Information17v. 1118Chapter No. 1192220-9964 (ISSN)20Publication Date:212022-01-0122Publisher DOI:23https://doi.org/10.3390/ijgi1101006524License:25https://creativecommons.org/licenses/by/4.0/26Link to license to see what you are allowed to do with this resource.27Downloaded from http://hdl.handle.net/1959.4/unsworks_78888 in https://28unsworks.unsw.edu.au on 2026-10-20293031===== PAGE 10 (score 44, numeric-paren 7, table-words 6, var-words 17) =====3233ISPRS Int. J. Geo-Inf. 2022, 11, 65349 of 1635Table 2. Hedonic Modelling Results.36(1)37(2)38(3)39(4)40Model41Baseline42Baseline OLS43Control OLS44IV45Dep. Variable46log(sales)47Airbnb Density480.0844 ***490.0301 ***500.0201 ***51(Standard Error)52(0.0020)53(0.0019)54(0.0024)55Observations5636,2645736,2645836,2645936,26460(Intercept)6112.7496212.7006312.94641365Bedrooms660.085670.106680.113690.11270Baths710.112720.096730.079740.08075Parking760.036770.046780.040790.04080Dist. to City Centre81−0.00082−0.00083−0.00084−0.00085Dist. to Coast86−0.00087−0.00088−0.00089−0.00090Dist. to Commercial Zones91−0.00092−0.00093−0.00094−0.00095Dist. to Swim Place96−0.00097−0.00098−0.00099−0.000100Dist. to Industrial Zones101−0.000102−0.000103−0.000104−0.000105Within 400 m of Railway Station1060.1441070.0821080.0301090.030110Within 800 m of Railway Station1110.1261120.0771130.2411140.025115Within 1600 m of Railway Station1160.0731170.0441180.0041190.037120Crime Rate121−0.227122−0.200123−0.015124−0.010125Median Family Income1260.0001270.0001280.0001290.000130Within 100 m of Main Road1310.006132−0.016133−0.023134−0.022135Within 100 m of Railway Line136−0.065137−0.047138−0.061139−0.064140Adjusted R21410.7461420.7961430.8681440.871145RSS1462593.3721472028.7091481312.7131491288.169150AICc1517287.198152−1615.774153−17,289.324154−17,973.784155Notes: Robust Standard Errors used. GOF statistics evaluated are: Adjusted R2 (adjusted coefficient of deter-156mination), RSS (Residual Sum of Squares), AICc (Corrected Akaike Information Criterion). Significance levels:157*** p < 0.01.158Upon implementing the TSLS regression with an IV specification, a positive and statis-159tically significant coefficient for Airbnb density is evident. This confirms the robustness160of our baseline results in which Airbnb activities do have a positive impact on housing161prices. Specifically, the IV model implies a 1% increase in Airbnb density is associated with162a 2.01% (column (4)) increase in house prices. The magnitude of Airbnb density in Model163(3) declines slightly compared with Model (2). This is in line with expectations, as the164purpose of the IV specification was to remove any correlation of the instrumental variable165with the error term and capture the effects of Airbnb more accurately. Ref. [22] provides166a detailed discussion of the validity of the IV term constructed through the ‘shift-share’167approach, concluding that it can be assumed exogenous from time-varying, municipality168level shocks. The exogeneity of the IV term with unobserved shocks contained in the error169term is expected to provide the most accurate estimate of the Airbnb density variable.170Therefore, the IV Airbnb density coefficient is the preferred measure of Airbnb’s impact on171housing prices in Sydney.172For the average postcode in terms of Airbnb activity (3.074), the IV Airbnb density coef-173ficient implies a house price increase of 2.26%, lower than the result of 3.7% in Barcelona [23].174However, with 50% more tourists annually in a city with roughly a third of Sydney’s pop-175ulation, it must be noted that Barcelona’s tourist market is far more substantial than the176Sydney market. Similarly, our estimate of a 2.01% increase in housing prices associated177with a 1% increase in Airbnb density is lower than the estimate of 3.5% in Lisbon and178Porto [24], however, there is a similar disparity in Sydney and Portugal’s tourist markets. It179must also be observed that these figures are not directly comparable, given the different180spatial granularity used in other studies performed at an aggregate level. Nonetheless, our181results are roughly aligned when considering the relative tourist markets of comparable182studies.183184185===== PAGE 11 (score 18, numeric-paren 0, table-words 6, var-words 4) =====186187ISPRS Int. J. Geo-Inf. 2022, 11, 6518810 of 16189In each stage of hedonic modelling the explanatory variable is statistically significant190and presents a clear associational impact of Airbnb activity increasing property prices.191Through the use of individual property sales data, the magnitude of Airbnb’s impact is able192to be evaluated for the first time in the context of residential property pricing. In general,193this does show that Airbnb activity leads to a higher level of housing prices. However, the194application of an OLS model to estimate Airbnb’s effect on the housing market ignores195both the spatial variation of Airbnb activity and housing submarkets within Sydney. A196GWR is therefore employed to capture the spatial nuances of Airbnb’s relationship with197the Sydney housing market.1984.2. GWR Results199Table 3 compares the goodness of fit (GOF) of the spatially explicit model against the200global models, and by all metrics, the GWR outperforms the OLS models. The adjusted201R2, which measures the proportion of total variation captured by the model, is 0.911202for the GWR compared to 0.871 for the best performing OLS model. Furthermore, the203GWR reduces the residual sum of squares of the IV model by a third from 1288.169 to204886.834, indicating that there is less unexplained variance in the spatially explicit model.205Finally, the corrected Akaike Information Loss (AICc) reduces from −17,973.784 with the IV206model to −26,824.740 for the GWR model, demonstrating that the GWR model minimises207information loss in the model.208Table 3. Model Performance Comparison.209Model210Adj R2211RSS212AICc213Baseline OLS2140.7962152028.709216−1615.774217Control OLS2180.8682191312.713220−17,289.324221IV2220.8712231288.169224−17,973.784225GWR2260.911227886.834228−26,824.737229Table 4 summarises the distribution of coefficient parameter estimates for the GWR.230It must be noted that GWR is foremost a descriptive spatial tool rather than a detailed231method for causal inference [29]. Therefore, the magnitudes of coefficient estimates are not232the primary concern of the GWR analysis, rather the variation of the directional association233(i.e., positive/negative) of the Airbnb density term and property prices. The median Airbnb234Density coefficient of 0.0127 corroborates the hedonic modelling results that Airbnb has235the predominant impact of increasing property prices. However, the 1st quantile estimate236of −0.0432 implies that for a significant portion of Sydney properties, Airbnb has a negative237influence on prices. Therefore, while Airbnb’s overall effect on the property market is238positive, the GWR results demonstrate that this impact is not homogenous, and in some239cases, the negative externalities produced by Airbnb may also reduce property prices.240Figure 2 visualises the Airbnb Density parameter estimates throughout Sydney. The241volatility of coefficient estimates may be related to bandwidth selection issues, due to242the application of a single bandwidth for all variables. This is a known limitation of243the GWR method, that could be potentially overcome in future work by applying an244MGWR [42]. It must also be noted that the magnitude of coefficient estimates may not245be directly comparable across Sydney areas, due to significant differences in underlying246Airbnb densities. Nevertheless, examining the direction of Airbnb density coefficients can247provide valuable insights into the geographical variation of Airbnb’s relationship with248property prices throughout Sydney.249250251===== PAGE 12 (score 28, numeric-paren 0, table-words 2, var-words 19) =====252253ISPRS Int. J. Geo-Inf. 2022, 11, 6525411 of 16255Table 4. GWR Model Results.256(1)257(2)258(3)259(4)260(5)261Min262Q1263Median264Q3265Max266Airbnb Density267−0.1217268−0.04322690.01272700.0312710.1456272Observations27336,26427436,26427536,26427636,26427736,264278(Intercept)27910.70728012.63528113.27828213.72828319.971284Bedrooms2850.0032860.0822870.1022880.1322890.213290Baths291−0.0152920.0492930.0772940.1022950.177296Parking2970.0012980.0292990.0353000.0483010.1302Dist. to City Centre30303040305030603070308Dist. to Coast30903100311031203130314Dist. to Commercial Zones31503160317031803190320Dist. to Swim Place32103220323032403250326Dist. to Industrial Zones32703280329033003310332Within 400 m of Railway Station333−0.487334−0.0263350.0363360.1213371.081338Within 800 m of Railway Station339−0.485340−0.0063410.0423420.13431.293344Within 1600 m of Railway Station345−0.591346−0.0223470.0043480.0273490.529350Crime Rate351−2.75352−0.291353−0.0933540.0423552.479356Median Family Income35703580359036003610362Within 100 m of Main Road363−0.107364−0.045365−0.022366−0.0053670.051368Within 100 m of Railway Line369−0.173370−0.098371−0.0473720.0163732.366374 375(a) Sydney 376(b) Zoom into inner Sydney 377Figure 2. Spatial distribution of GWR parameter estimates for Airbnb density variable with 190 near-378est neighbours.379Spatial clustering of positive coefficient estimates is evident towards Manly and along380the northern beaches, where there is substantial Airbnb activity. This may be expected,381given the conceptual framework of Airbnb reducing the supply of long-term rentals and in382turn increasing housing prices. There is also a significant collection of positive coefficients,383ranging up to 0.146, towards the western centre of Paramatta. The emergence of Paramatta384as Sydney’s second CBD over the last decade may explain business motivated Airbnb usage385in these areas [2]. In these areas, where housing prices are lower and the tourist market is386less developed, an increase in Airbnb may be seen as a positive sign for homeowners in387signalling growth and greater potential rental revenue. Furthermore, ref. [43] documented388an increasing prominence of ‘investor landlords’ in Western Sydney, for whom Airbnb389presents a direct opportunity to increase rental yield. While Airbnb density here compared390to the more touristy central and coastal areas is lower and the impact on housing prices391will therefore be less, this nonetheless demonstrates that Airbnb can affect housing prices392outside of traditional tourist hotspots.393394395===== PAGE 13 (score 9, numeric-paren 0, table-words 0, var-words 9) =====396397ISPRS Int. J. Geo-Inf. 2022, 11, 6539812 of 16399Surprisingly, towards Sydney’s more traditional tourist areas of the CBD and eastern400suburbs, the Airbnb density coefficients are insignificant or negative. The insignificant401coefficients in the city centre may be explained by the significant pre-existing hotel and402holiday rentals in these areas, meaning the entrance of Airbnb has not disrupted the housing403market as significantly as non-traditional tourist areas. The negative coefficients evident404in the eastern suburbs and lower north shore raise the question of whether in high-price405areas, there exists a peak threshold of Airbnb activity, whereafter the over-saturation of the406tourist market is seen as a detractor for local residents. In these cases, negative externalities407associated with Airbnb activity such as noise, congestion, and access to amenities for locals,408may outweigh the increased rental revenue potential for homeowners [34]. Notably, it409has been shown in Sydney that high- and low-end housing submarkets follow different410pricing mechanisms [17,18]. As such, it is not surprising to document a dissimilarity effect411of Airbnb activities on housing prices, highlighting the importance of a disaggregated412study in gauging the effect of Airbnb.413It must be noted that upon applying this rationale, it might be expected that Manly414and the northern beaches would also experience negative Airbnb coefficients. While this415demonstrates the potential limitation of GWR in identifying specific Airbnb property416market hot/cold spots, it serves to highlight the spatial complexity of the Airbnb effect417across the city. Subsequently, one may assume that Airbnb’s impact on affordability is also418uneven due to heterogeneous effects. This is an important consideration in the context of419affordability policies and highlights the need for policymakers to consider local tourist and420housing market conditions when creating Airbnb regulations.421Clearly, Airbnb’s effect within a city is highly contextual to the city’s local tourist422and housing market dynamics. Therefore, understanding how Airbnb’s housing market423varies spatially in other global cities across different continents may help shed light on its424potential heterogeneous effect in different housing submarkets. Future work could also425examine the variable impacts of Airbnb using case studies of housing submarkets within a426city, as well as the temporal evolution of the Airbnb effect over the last decade. Finally, with427the global COVID-19 pandemic greatly altering Airbnb activity in major cities, analysing428market dependence on pandemic development may provide a new and updated lens into429Airbnb’s relationship with the housing market. Ultimately, through accounting for spatial430variation in modelling, our results demonstrate that the relationship between Airbnb and431housing prices is complex and varies significantly across and within cities such as Sydney.4325. Conclusions433The rapid expansion of home-sharing platforms, such as Airbnb, that has been border-434less and largely unregulated has led many to question what their effect on local communities435and housing has been. In large tourist cities such as Sydney, where the growth of Airbnb436has coincided with rising rental and housing prices, its relationship to the housing market437has been a subject of considerable attention for both academics and government regulators.438However, as policymakers and local governments have begun to apply a range of regula-439tory measures in response to these concerns, the lack of quality empirical evidence to gauge440the causal impact of platforms such as Airbnb on housing prices has become evident.441This study examines empirically the extent to which Airbnb impacts housing prices442in Sydney for the first time. Several stages of OLS modelling were performed to control443for external shocks and address potential collinearity concerns. The OLS modelling was444performed from a residential property valuation perspective, including property charac-445teristics, mobility variables, and proximity to amenity measures for the first time in an446Airbnb housing market study. Our results indicate that a 1% increase in Airbnb density is447associated with a 2.01% increase in property prices. This corroborates existing research on448the overall impact of Airbnb on the housing market.449Despite this, the existence of housing submarkets and documented spatial unevenness450of Airbnb activity throughout Sydney imply that a global model will be insufficient in451accounting for the spatial nuances of Airbnb’s property market impact. Therefore, a GWR452