1# RECORD eb2663068d73d32e83378e9ddd14f5caa58f11e62# TITLE: Location, location and professionalization: a multilevel hedonic analysis of Airbnb listing prices and revenue3# YEAR: 2019.0 VENUE: Regional Studies, Regional Science DOI: 10.1080/21681376.2019.15926994# AUTHORS: Robbin Deboosere; D. Kerrigan; D. Wachsmuth; A. El-geneidy5# PAGES: 15 CANDIDATE TABLE PAGES (1-based): [6, 9]6# ABSTRACT: ABSTRACT Hedonic modelling techniques have frequently been used to examine real estate valuation, and they have recently started to be applied to short-term rental valuation. Relying on a web-scraped data set of all Airbnb transactions in New York City (NYC) between August 2014 and September 2016, this paper presents the first hedonic regression model of Airbnb to take into account neighbourhood effects and to predict both average price per night and revenue generated by each listing. The model demonstrates that locational factors – above all, transit accessibility to jobs – and neighbourhood variation have a large impact on both price per night and monthly revenue, and further reveals how professionalization of the short-term rental market is driving more revenue to a narrower segment of hosts. Further, the findings suggest that Airbnb hosts earn a significant premium by converting long-term housing in accessible residential neighbourhoods into de facto Airbnb hotels. This premium incentivizes landlords and hosts with properties in accessible neighbourhoods to replace long-term tenants with short-term guests, forcing those in search of housing to less accessible neighbourhoods.789===== PAGE 1 (score 4, numeric-paren 0, table-words 0, var-words 4) =====1011Full Terms & Conditions of access and use can be found at12https://www.tandfonline.com/action/journalInformation?journalCode=rsrs2013Regional Studies, Regional Science14ISSN: (Print) 2168-1376 (Online) Journal homepage: https://www.tandfonline.com/loi/rsrs2015Location, location and professionalization:16a multilevel hedonic analysis of Airbnb listing17prices and revenue18Robbin Deboosere, Danielle Jane Kerrigan, David Wachsmuth & Ahmed El-19Geneidy20To cite this article: Robbin Deboosere, Danielle Jane Kerrigan, David Wachsmuth & Ahmed21El-Geneidy (2019) Location, location and professionalization: a multilevel hedonic analysis22of Airbnb listing prices and revenue, Regional Studies, Regional Science, 6:1, 143-156, DOI:2310.1080/21681376.2019.159269924To link to this article: https://doi.org/10.1080/21681376.2019.159269925© 2019 The Author(s). Published by Informa26UK Limited, trading as Taylor & Francis27Group28Published online: 03 Apr 2019.29Submit your article to this journal 30Article views: 359831View related articles 32View Crossmark data33Citing articles: 15 View citing articles 343536===== PAGE 6 (score 30, numeric-paren 0, table-words 5, var-words 16) =====3738generated revenue in at least two separate months allows this model to be more accurate. Finally,39to ensure the analysis only includes listings actively reserved during this period, listings with no40user reviews were discarded. The resulting population for this model was 386,153. This popu-41lation is equal to the number of unique listings multiplied by the number of months that the list-42ing had revenue. Using a geographic information system (GIS) the x,y-coordinates of each listing43were mapped for further spatial analysis.44The model has two separate dependent variables and 20 independent variables, presented in45Table 1. The first dependent variable is the average price per night, which for the study popu-46lation was US$153.16. This is the same dependent variable used by Gibbs et al. (2018),47Wang and Nicolau (2017) and Teubner et al. (2017). The second dependent variable is the aver-48age revenue per month, which was US$2287.19. There are three categories of independent vari-49ables: structural, host, and location and neighbourhood. All data on an individual listing’s50structural and host variables come from the AirDNA data set. Structural variables included51are the number of bedrooms (mean = 1.1), the number of bathrooms (mean = 1.1), the maximum52number of guests (mean = 3.0), a dummy variable that equals 1 if the listing is an entire53Table 1. Model variables.54Variable55Description56Meana57SD58Price59Average price per night (US$)60153.1661113.2562Revenue63Revenue per month (US$)642287.19652352.4566Structural variables67Entire apartment68Dummy variable if the entire apartment is rented690.6109700.487571Shared room72Dummy variable if the room is shared730.0274740.163375Bedrooms76Number of bedrooms771.1112780.660679Bathrooms80Number of bathrooms811.0850820.326283Guests84Maximum number of guests852.9785861.728187Host variables88Superhost89Dummy variable if the host is a superhost900.0868910.281692Multihost – 2–1093Dummy variable if the host has between two and 1094listings950.3655960.481697Multihost – >1098Dummy variable if the host has more than 10 listings990.01651000.1272101Number of photos102Number of photographs for a listing10314.201510410.8734105Instant booking106Dummy variable for is an instant booking is available1070.12921080.3355109Reviews110Number of reviews11131.592611236.1101113Overall rating114Overall star rating of listing1154.57381160.3934117Location and neighbourhood variables118Subway119Dummy variable if the listing is within 800 m of a120subway stop1210.90381220.2948123Transit accessibility124Accessibility to jobs by transit in 30 min (100,000125jobs)12613.18121278.1446128Dist Times Square129Network distance to Times Square (km)1306.43431313.9046132Diversity133Land-use diversity, measured by an entropy index1340.95571350.2867136Population density137Population density in the census tract (1000 residents/138km2)13929.103714015.4984141Median household142income143Median household income in the census tract (US144$1000)14574.142714636.4930147Percentage white148Percentage of residents who identify as white in the149census tract15047.504515126.6700152Listing density153Number of Airbnb listings in the census tract154322.2075155261.7637156Note: aThe mean of a dummy variable represents the percentage of all listings that have a value of 1 for that157dummy variable.158Location, location and professionalization: a multilevel hedonic analysis of Airbnb listing prices and revenue159147160REGIONAL STUDIES, REGIONAL SCIENCE161162163===== PAGE 7 (score 21, numeric-paren 0, table-words 0, var-words 21) =====164165apartment, and a dummy variable that equals 1 if the listing is a shared room. These structural166variables were included based on previous studies (Gibbs et al., 2018; Hill, 2015; Lee et al.,1672015; Teubner et al., 2017; Wang & Nicolau, 2017).168The host variables used were: a dummy variable equal to 1 if the host has superhost status; a169dummy variable equal to 1 if the host has between two and 10 listings; a dummy variable equal to1701 if the host has more than 10 listings; and a dummy variable equal to 1 if the instant book feature171was available. Additionally, the model used variables for the number of photographs included172with each listing (mean = 14.2), the number of reviews (mean = 31.6) and the overall rating of173the listing (mean = 4.6). While Teubner et al. (2017) looked at details such as the host’s gender,174age and if they were smiling in their profile photograph, they found no significant price effects175based on these details. Accordingly, such variables were not included in this study, and host vari-176ables, in general, were chosen based on previous studies (Gibbs et al., 2018; Hill, 2015; Lee et al.,1772015; Wang & Nicolau, 2017).178Multiple location and neighbourhood variables were used. These included: a dummy vari-179able equal to 1 if a listing is located within 800 m of a subway station (approximately a 10-min180walk); the number of jobs that can be reached within 30 min by transit from each listing (‘job181accessibility’ per 100,000 jobs; mean = 13.2); the network distance to Times Square (mean =1826.4 km); the land-use diversity measured by an entropy index (mean = 0.96); the population183density per census tract (1000 residents/km2; mean = 29.1); the median household income184(in US$1000s; mean = 74.1); and the percentage of residents who identify as non-Hispanic185White (mean = 47.5). Accessibility to jobs is used as a proxy for accessibility to tourist ame-186nities based on the idea that the majority of Airbnb guests are tourists. Ideally there would187be a direct variable for accessibility to tourist amenities such as restaurants, entertainment188venues, local landmarks and museums. As this information is not widely available, accessibility189to jobs is a proxy as most typical tourist venues and attractions are employers and contribute to190jobs. Further, Ashworth and Page (2011) note the fact that tourist landscapes coexist with191employment, dispersed and grouped into zones such as the ‘entertainment district’. This192‘employment-based attraction’ is especially true as tourism in major urban centres tends to193be based around attractions such as cultural venues, including museums and theatres, as well194as shopping streets and restaurant and bar districts (Bull & Church, 2001). In NYC, Midtown195Manhattan is one of the preeminent centres of employment, and also the location of many196tourist amenities. Therefore, high access to jobs in NYC reflects a high degree of access to197Midtown Manhattan and other tourist amenities, but also reflects other parts of the city198that would be attractive to tourists. Transit accessibility is used as tourists are less likely to199have access to the use of a personal vehicle. Cervero and Kockelman (1997) argue that200land-use diversity is likely representative of other attributes including density and walkability.201Studies have shown that higher degrees of walkability (albeit based on ‘walk score’ data) are202related to higher property values (Pivo & Fisher, 2011; Rauterkus & Miller, 2011).203All the distances involved were calculated as network distances, based on the nearest census204block group centroid. The distance from the closest block group centroid was used instead of the205distances from the listing x,y-coordinates to reduce computational complexity. The relevant sha-206pefiles used in the calculations were provided by the City of New York’s Planning Department.207Data regarding accessibility to jobs were gathered from the Access Across America project. This208data set measures the number of jobs accessible within 30 min per block and was aggregated to209the census block group to be on the same scale as the other location and neighbourhood variables.210Owing to the overwhelming number of jobs in Midtown Manhattan, this measure acts as an211approximate proxy of the centrality of any individual block group. Land-use diversity – a212proxy for neighbourhood walkability – was measured through an entropy index based on the213148214Robbin Deboosere et al.215REGIONAL STUDIES, REGIONAL SCIENCE216217218===== PAGE 9 (score 98, numeric-paren 60, table-words 3, var-words 13) =====219220Extra bathrooms command a larger price and revenue premium than extra bedrooms. One221additional bedroom is associated with a price increase of 13.2% and a 11.0% increase in monthly222revenue. By contrast, each additional bathroom is associated with a 14.2% increase in listing price223per night and 14.8% higher revenue per month. The higher performance associated with224additional bathrooms could reflect the extent to which they indicate more luxurious listings or225serve as a proxy for a larger home at a given number of bedrooms. An extra place for a guest pre-226dicts a 6.4% higher listing price and a 7.7% increase in monthly revenue.227Table 2. Model results.228Price229Revenue230Coefficient231p232Confidence233intervala234Coefficient235p236Confidence237intervala238Time variables239January240–0.1063241***242–0.1085243–0.1040244–0.5841245***246–0.5958247–0.5725248February249–0.1406250***251–0.1431252–0.1380253–0.3920254***255–0.4049256–0.3790257March258–0.0874259***260–0.0898261–0.0850262–0.0919263***264–0.1041265–0.0797266April267–0.0296268***269–0.0319270–0.02732710.0879272***2730.07622740.0997275May2760.0019277–0.00032780.00422790.2917280***2810.28032820.3031283June2840.0402285***2860.03792870.04242880.2706289***2900.25912910.2822292July2930.0215294***2950.01932960.02382970.1971298***2990.18563000.2086301August3020.0121303***3040.00993050.01423060.2288307***3080.21803090.2396310September3110.0413312***3130.03923140.04343150.3691316***3170.35853180.3797319October3200.0346321***3220.03223230.03693240.3041325***3260.29233270.3159328November329–0.0069330***331–0.0093332–0.0046333–0.0621334***335–0.0742336–0.0501337Structural variables338Entire apartment3390.5115340***3410.50523420.51783430.5121344***3450.49853460.5256347Shared room348–0.3670349***350–0.3815351–0.3525352–0.4235353***354–0.4549355–0.3922356Bedrooms3570.1321358***3590.12703600.13723610.1097362***3630.09883640.1206365Bathrooms3660.1421367***3680.13403690.15023700.1481371***3720.13073730.1656374Guests3750.0637376***3770.06153780.06593790.0768380***3810.07203820.0815383Host variables384Superhost3850.0565386***3870.04603880.06703890.1507390***3910.12893920.1726393Multihost – 2–10394–0.0037395–0.00923960.00183970.0660398***3990.05424000.0778401Multihost – > 10402–0.0924403***404–0.1135405–0.07134060.0887407***4080.04464090.1328410Number of photos4110.0036412***4130.00334140.00384150.0019416***4170.00134180.0024419Instant booking420–0.0177421***422–0.0256423–0.00994240.1424425***4260.12564270.1591428Reviews429–0.0005430***431–0.0006432–0.00044330.0084434***4350.00824360.0086437Overall rating4380.0826439***4400.07714410.08804420.0433443***4440.03114450.0555446Location and neighbourhood variables447Subway448–0.0201449***450–0.0338451–0.0063452–0.0350453***454–0.0593455–0.0107456Transit accessibility4570.1220458***4590.10904600.13504610.1320462***4630.11204640.1520465Dist Times Square466–0.0147467***468–0.0168469–0.0126470–0.0245471***472–0.0277473–0.0213474Diversity4750.0093476–0.00554770.02404780.0065479–0.02114800.0341481Population density482–0.0011483***484–0.0016485–0.0006486–0.0011487***488–0.0017489–0.0005490Median household491income4920.0012493***4940.00094950.00154960.0009497***4980.00064990.0013500Percentage white5010.0019502***5030.00155040.00225050.0010506***5070.00055080.0014509Listing density5100.0001511***5120.00015130.00025140.0001515***5160.00015170.0002518Constant5193.3377520***521–0.0338522–0.00635235.6364524***5255.53235265.7404527Notes: Dependent variables: ln(price per night) and ln(monthly revenue).528a95% confidence interval.529*95% significance level; **99% significance level; ***99.9% significance level.530150531Robbin Deboosere et al.532REGIONAL STUDIES, REGIONAL SCIENCE533534535===== PAGE 10 (score 14, numeric-paren 0, table-words 1, var-words 10) =====536537A set of characteristics that collectively suggest a listing is a commercial operation, as opposed538to a part-time rental, are associated with substantially higher monthly revenue than the baseline,539but not consistently higher nightly rates. Commercial Airbnb operators, in other words, appear to540set pricing more aggressively in order to maximize occupancy and, hence, revenue. The first of541these characteristics is ‘superhost’ status, which Airbnb confers on hosts based on a high fre-542quency of hosting, positive reviews from guests and a low cancellation rate. According to the543model, this superhost status allows hosts to charge a slight premium (5.6% more per night)544and results in a substantial increase of 15.1% in monthly revenue. The differential between545these two figures implies that superhosts conduct considerably more bookings per month than546other hosts. The significant impact of superhost status on both price and revenue supports a simi-547lar finding by Wang and Nicolau (2017), but contradicts Teubner et al. (2017) who did not548identify superhost status as having statistical significance.549Airbnb allows hosts to activate an instant booking feature, which allows prospective guests to550bypass the usual host approval process and have their requested booking confirmed immediately.551Listings with the instant booking feature activated have 1.8% lower average nightly prices, but552have much larger monthly average revenue of 14.2%. Much like superhost status, instant booking553is characteristic of high-volume listings, which provide a more hotel-like experience than less554commercial listings whose hosts prefer to approve booking requests manually.555The final factor associated with commercial short-term rental operations that predicts higher556revenue is the number of listings a given host controls. Hosts with between two and 10 listings do557not charge a significantly different price per night from hosts with a single listing, but they earn5586.6% more in monthly revenue, reflecting higher occupancy rates. By contrast, hosts managing559more than 10 properties charge a substantially lower price per night than single-listing hosts560(9.2% less), but earn substantially more revenue (8.9%). Noting the same result, Li et al.561(2015) suggest that this pattern could reflect learning effects among high-volume hosts, especially562responses to change in demand (either seasonal or event based).563Listings with more photographs charge marginally higher prices and earn marginally more564revenue. One additional review is associated with a 0.05% decrease in price, but a 0.8% increase565in revenue. This additional revenue is likely due to the fact that lower priced listings are booked566more frequently and therefore they will have more reviews (Gibbs et al., 2018). One additional567star in average rating is associated with a price increase of 8.3% and a revenue increase of 4.3%.568Proximity to mass transit is a strong predictor of listing price and revenue, but only when it569facilitates access to the central city. In fact, controlling for accessibility, listings in proximity to570subway stations charge less per night (2.0% less) and have less monthly revenue (3.5% less).571However, listings with high accessibility (defined here as access to jobs via transit) charge an aver-572age of 12.2% more per night and earn an average of 13.2% more monthly revenue. This finding573Table 2. Continued.574Random-effects parameters575Price576Revenue577Estimate578Confidence579intervala580Estimate581Confidence582intervala583Borough level584sd(Constant)5850.06325860.03325870.12055880.06955890.03345900.1450591Census tract level592sd(Constant)5930.08915940.08265950.09605960.07755970.06935980.0868599Listing level600sd(Constant)6010.26416020.26246030.26596040.49586050.49136060.5002607sd(Residual)6080.14366090.14336100.14406110.73736120.73556130.7390614Note: a95% confidence interval.615Location, location and professionalization: a multilevel hedonic analysis of Airbnb listing prices and revenue616151617REGIONAL STUDIES, REGIONAL SCIENCE618