Corpus › Gil & Sequera (2020)

The professionalization of Airbnb in Madrid: far from a collaborative economy

Javier Gil, Jorge Sequera · 2020 · Current Issues in Tourism · Published online 7 May 2020 (file name suggests the 2022 print issue)· doi:10.1080/13683500.2020.1757628

journal article · language: en · peer reviewed: yes · Tourism studies / urban sociology · approach: quantitative · source: pdf/gil_2022_airbnb_madrid.pdf

Airbnbsharing economytouristificationhousingprofessionalizationMadrid

Analysis mode

document only

Verdict

mostly supported

Rubric score

3.3/5

Claim support index ?

67/100

core claims (thesis/main/conclusion): 70

InsideAirbnb data for Madrid show a supply dominated by entire apartments, 'commercial' (high-availability) listings and multi-listing hosts, concentrated in the Central District — i.e. professionalised 'airbnbisation' rather than a sharing economy.

Thesis (C1) Airbnb in Madrid is dominated by professional actors specialised in renting entire apartments as short-term rentals, mainly in the Central District, whose activity does not comply with sharing-economy principles. adequate

The cross-sectional picture of professionalised, centrally concentrated Airbnb supply in Madrid is robust: nearly every reported share reconciles with the others. Weaknesses are the retrospective growth series (survivorship and back-dating bias), the blanket equation of multi-listing with professional non-resident hosting, an arithmetic slip ('nearly 10 points'), and an impact claim in the abstract that the paper itself concedes has not been studied.

Claims (13)

351211

Issues (6)

132

Numeric checks (7)

Summary

Opening with 'Claudia' — the pseudonym of Friendly Rentals (Wyndham) with 211 Madrid listings — the paper argues that Madrid's city centre undergoes 'airbnbisation' driven by professional short-term-rental operators. Using one InsideAirbnb scrape (18,332 active listings; date not stated) it analyses spatial distribution, accommodation type, p2p vs commercial listings (183-day availability rule), multi-listing hosts and listing accumulation, plus a reconstructed growth series based on host registration dates and a comparison with April 2017 data. The Central District holds 47.78% of listings; entire apartments are 63% citywide and 76.3% in the centre; commercial listings 75.7% (80.07% centre); multi-listing hosts control 59.8% of listings (68.85% centre); hosts with 6+ listings (4.3% of hosts) control 32.1% of listings, and their share grew since 2017. The paper concludes the model resembles traditional accommodation, harms the city's economic sustainability, and recommends banning commercial listings plus a territorial cap.

Research questions

Is Madrid's Airbnb supply dominated by entire apartments, commercial listings and multi-listing hosts?
Is the supply disproportionately concentrated in the Central District, with over-representation of professional listings there?
Has Airbnb's growth reduced or reinforced professionalisation?

Stated contribution

An in-depth temporal and territorial examination of professionalisation, concentration and 'monopolisation' of Airbnb in Madrid.

How the argument is built

Sharing-economy principle (idle capacity) → Airbnb's self-narrative ('Belong Anywhere') → literature on touristification and rent gaps → five indicators → results by district and over time → principles under which sharing-economy benefits could materialise → policy recommendation.

Argument map

Claims (boxes) coloured by how well the paper supports them; concepts as ellipses. Arrows point from premise to what it supports. Thick black border = thesis; blue border = conclusion/recommendation. Click a node to jump to it. Layout engine: graphviz.

strongadequatepartialweakunsupportednot applicable supports / generalizesexplains / enables / illustratesdefines / motivates / qualifies line width = strength

show evidence items actual size (scroll)

Relations as a table (16)

From	Type	To	Strength	Note
K2	defines	C11
K3	defines	C4
K4	defines	C1
C2	supports	C1	strong
C3	supports	C1	moderate
C4	supports	C1	strong
C5	supports	C1	strong
C6	qualifies	C5	moderate	Interpretation of multi-listing as 'professional' turns C5 into evidence of professionalisation.
C7	supports	C1	moderate
C8	supports	C1	weak
C9	supports	C1	moderate
C13	supports	C10	moderate	The impact claim is borrowed from literature, not derived from Madrid data.
C1	supports	C10	weak
C11	defines	C1	moderate	The benchmark against which Madrid's supply is judged.
C1	motivates	C12	moderate
C10	motivates	C12	moderate

Claims (13)

Filter:

C1 adequate thesisinterpretive non trivialpp. 1, 14-15

Airbnb in Madrid is dominated by professional actors specialised in renting entire apartments as short-term rentals, mainly in the Central District, whose activity does not comply with sharing-economy principles.

Assessment: The cross-sectional indicators are internally consistent (district-weighted shares reconcile to city totals within rounding) and jointly support professional dominance of the listing stock. 'Dominated' rests on listing shares; 'professional' rests on proxies (availability, multi-listing) that are imperfect but convergent.

Evidence: E2 E3 E4 E5 E7 · Issues: I2 I3

C2 strong mainempirical descriptive non trivialpp. 7, 13-14

Supply is over-concentrated in the Central District (47.78% of listings; over 50% in prior years), a district with ~5% of population and <1% of the city's territory.

Assessment: Count and share reconcile (8,759/18,332 = 47.78%). Note the conclusion says 'over 50% ... has been concentrated' while the current figure is below 50%.

Evidence: E1 · Issues: I5

C4 adequate mainempirical comparative non trivialpp. 9-10

Commercial listings are 75.7% of supply and 80.07% in the Central District, 'nearly 10 points greater than the Madrid average'.

Assessment: Shares reconcile (weighted 75.67%). The '10 points' statement is wrong: the gap to the city average is 4.4 points (8.4 points to the rest of the city). The commercial classification inherits the limits of the availability proxy.

Evidence: E3 · Issues: I1 I2

C5 strong mainempirical descriptive non trivialpp. 11-13

Multi-listing hosts control 59.8% of listings (68.85% in the centre); fewer than 5% of hosts (6+ listings) control over 30% of listings; 65 hosts control 2,753 listings.

Assessment: Band shares sum to 100%, 2,753/18,332 = 15.02%, and district-by-band shares reproduce the 47.78% and 68.85% totals exactly — strong internal consistency.

Evidence: E4 E5 E8 · Issues: I3

C7 adequate mainempirical comparative non trivialp. 13

Professionalisation is increasing over time: since April 2017 hosts with 6-20 and >20 listings gained 2.92 and 6.61 points of market share while single-listing hosts lost 4.84 points.

Assessment: The reported relative changes are arithmetically consistent with the point changes (e.g. 6.61/8.41 = 78.6%). Comparability of the two scrapes (filters, active-listing definition) is assumed rather than documented; 'growth' refers to growth in share, not listings.

Evidence: E6 · Issues: I4

C11 partial mainnormative contestedp. 15

Sharing-economy benefits can only materialise if supply is based on locals hosting in their own home, in private/shared rooms, and evenly distributed; concentrated supply makes the claimed benefits 'impossible'.

Assessment: A reasonable set of criteria derived from the idle-capacity definition, but stated as necessary conditions without argument (e.g. why an entire home rented while the resident travels would not qualify; why concentration makes benefits 'impossible' rather than smaller).

No evidence cited in the paper for this claim.

C10 unsupported conclusionempirical causal non trivialp. 1 (abstract), p. 3

This model generates negative impacts on the economic sustainability of the city and its inhabitants.

Assessment: No outcome data on rents, housing supply, displacement or local economy are analysed. The conclusion itself says a study of these impacts 'would be necessary', contradicting the abstract's assertion.

Evidence: E10 · Issues: I6

C6 weak intermediatedefinitional contestedp. 6

All multi-listing hosts are professionals not hosting in their own residence, who have probably extracted their units from the residential market.

“in all cases their activity represents professionalized hosts, since they are not carrying out the activity in their own residence”

Assessment: Stipulated rather than shown, and contradicted by the paper's own data: 44.5% of listings of hosts with 2-5 listings are private rooms, many of which may be rooms in the host's own home. 'Probably extracted from the residential market' rests on other cities' literature.

Evidence: E10 · Issues: I3

C3 strong supportingempirical comparative non trivialpp. 6-10

Entire apartments make up 63% of supply, 76.3% in the Central District vs 51% elsewhere, indicating specialisation of the centre in entire-apartment rentals.

Assessment: District figures reconcile with the city total (0.478 x 76.3 + 0.522 x 51 = 63.1).

Evidence: E2

C8 weak supportingempirical comparative non trivialpp. 10-12

Airbnb's growth has not reduced — and has reproduced — the over-representation of entire apartments, commercial listings and multi-listing hosts in the Central District.

Assessment: Based on a retrospective series built from host registration dates of listings still active at scrape time, classified with current attributes. Survivorship (exited listings, likely disproportionately occasional p2p ones, are missing) and back-dating (a host's later listings are assigned to the host's sign-up year) bias the historical composition. Only the 2017 comparison (C7) is a genuine two-point measurement.

Evidence: E9 · Issues: I4

C9 adequate supportingempirical comparative non trivialp. 13

Large professional hosts specialise in the Central District and in entire apartments.

Assessment: 64.5% and 58.4% of 6-20 and >20 band listings are central vs 37.1% for single hosts; entire-apartment share 71-76% vs 56-60%. The relationship is not monotonic (>20 less central than 6-20; 2-5 fewer entire apartments than singles), which the text glosses over.

Evidence: E7

C12 not applicable recommendationnormative contestedp. 16

Madrid should ban commercial listings (as San Francisco did), add a territorial criterion, and fund inspection and enforcement.

Assessment: Policy recommendation consistent with the diagnosis; no assessment of enforcement feasibility or trade-offs is provided.

No evidence cited in the paper for this claim.

C13 adequate backgroundbackground common knowledgepp. 4-5

STRs yield higher rents than long-term letting, so dwellings are extracted from the residential market, raising rents and displacing tenants (literature premise).

Assessment: Well-established in the cited literature; used as a premise, not tested here.

Evidence: E10

Evidence & data

Design

Quantitative. Descriptive cross-sectional analysis of one InsideAirbnb snapshot with a retrospective growth reconstruction and a comparison with an April 2017 snapshot from earlier work.

Scope: Madrid, 2010s to c. 2019.

Methods

Shares by district, accommodation type, p2p/commercial rule, host listing-count band
Retrospective growth series using host registration date
Comparison of market shares with April 2017

Transparency notes

Scrape date is never stated; no table of counts by band (only shares and graphs); growth method is described and its survivorship limitation footnoted.

Limitations acknowledged by the authors

Growth series counts only listings active at scrape time and uses host registration date (fn. 4).
p2p rule is described as the most lenient possible.

Data sources

ID	Source	Type	Description	Coverage	Loc.
D1	InsideAirbnb Madrid	scraped platform data	18,332 active listings at the time of collection (scrape date not reported; context implies 2019).	City of Madrid, 21 districts collapsed to Central vs rest.	pp. 5-6
D2	Gil & Sequera (2018) April 2017 data	secondary literature	Earlier listing-accumulation shares used for the 2017-2019 comparison.		p. 13
D3	Press and company sources	media	El País (housing-stock ratio in the Central District), Airbnb press material, Wyndham website.		pp. 3-4, 7

Evidence items (10)

ID	Kind	Description	Loc.	Source	Used by
E1	statistic	18,332 active listings; Central District 8,759 (47.78%); >50% in the centre in previous years; Central District listings equal 8.77-22.5% of housing stock (El País).	p. 7, Graphs 1-2, Map 1	D1	C2
E2	statistic	Entire apartments 63.0% citywide; 76.3% Central vs 51% rest.	pp. 7-10, Graphs 3-4	D1	C1 C3
E3	statistic	Commercial listings 75.7% citywide; 80.07% Central; 71.64% rest.	pp. 10-11, Graphs 5-6	D1	C1 C4
E4	statistic	Multi-listing hosts control 59.8% of listings; 68.85% Central; 51.57% rest.	p. 11, Graphs 7-8	D1	C1 C5
E5	statistic	Host bands: 1 listing 74.50% of hosts / 40.17% of listings; 2-5: 21.19% / 27.7%; 6-20: 3.65% / 17.11%; >20: 0.66% / 15.02% (65 hosts, 2,753 listings).	pp. 12-13, Graph 9	D1	C1 C5
E6	statistic	Change vs April 2017: 6-20 band +2.92 pp (+20.58%), >20 band +6.61 pp (+78.6%), single-listing -4.84 pp (-10.75%).	p. 13	D2	C7
E7	statistic	Central-District share by host band: 37.05% (1), 47.28% (2-5), 64.46% (6-20), 58.41% (>20); entire-apartment share by band: 59.87%, 55.52%, 71.01%, 76.44%.	p. 13, Graphs 10-11	D1	C1 C9
E8	case vignette	'Claudia' = Friendly Rentals Madrid (Wyndham), 211 listings (138 central); other pseudonymous multi-hosts.	p. 3	D3	C5
E9	figure	Retrospective growth series of listings by type, listing class and host type, built from host registration dates of currently active listings. ⚠ Survivorship bias and back-dating of all of a host's listings to the host's join date.	pp. 7-14, Graphs 1, 4, 6, 8	D1	C8
E10	literature	Literature on STR conversion, rent gaps, rent increases and displacement in other cities.	pp. 4-5		C6 C10 C13

Concepts & definitions

ID	Term	Definition	Origin	Analyst note
K1	Airbnbisation p. 1, 5	Transformation of residential housing units into Airbnb listings.	borrowed Yrigoy (2017)
K2	Sharing-economy principle (idle capacity) p. 2	A resource qualifies for the sharing economy only if temporarily disused; all positive impacts derive from this.	borrowed Benkler; Botsman & Rogers; Frenken et al.; et al.
K3	Commercial vs p2p listing p. 6	Available > 182 days/year = commercial; < 183 = p2p.	operational	Same rule as Gil (2021); calendar availability is a noisy proxy for non-residential use.
K4	Professionalisation pp. 4-6	Operationalised via entire-apartment share, commercial share and listing accumulation by multi-listing hosts.	operational

Issues (6)

I6 major internal inconsistency Impact claim in abstract contradicted by conclusionp. 1; p. 16

The abstract concludes the model 'generates negative impacts on the economic sustainability of the city and its inhabitants', but no impact is measured and the conclusion states that a study of impacts on rents, supply and displacement 'would be necessary'.

Impact on conclusions: The impact conclusion should be treated as a hypothesis.

Affects: C10

I2 moderate measurement Availability-based classificationp. 6

Same 182-day calendar-availability rule as in Gil (2021): booked nights are not 'available', calendar defaults vary, so the rule is a noisy proxy for residential use. No sensitivity analysis.

Impact on conclusions: Magnitude uncertain; the citywide direction is plausible.

Affects: C1 C4

I3 moderate conceptual Multi-listing ≠ professional / non-resident in all casespp. 6, 13

The paper asserts that all multi-listing hosts are professionals not hosting in their own residence, yet 44.5% of listings in the 2-5 band are private rooms. Hosts with several rooms in their own home would be misclassified.

Impact on conclusions: Concentration findings stand; 'professional' label overstated for the 2-5 band.

Affects: C1 C5 C6

I4 moderate measurement Retrospective growth series biasedp. 5, fn. 4; p. 13

Growth is reconstructed from host registration dates for listings active at scrape time (acknowledged in fn. 4). This omits exited listings and back-dates listings added later by long-standing hosts, and it classifies past listings using current availability. The 2017-vs-now comparison relies on an earlier paper's scrape whose comparability is not documented.

Impact on conclusions: Temporal claims are weaker than the cross-sectional ones.

Affects: C7 C8

I1 minor arithmetic Misstated gap ('nearly 10 points')p. 9

Central commercial share 80.07% vs city average 75.7% is a 4.4-point gap, not 'nearly 10 points' (the gap to the rest of the city is 8.4).

Impact on conclusions: Presentational; direction unchanged.

Affects: C4

I5 minor transparency Scrape date not reported; '>50%' in conclusionpp. 7, 13

The date of the main dataset is never given. The conclusion says over 50% of supply 'has been' in the Central District, while the current figure is 47.78%.

Impact on conclusions: Minor.

Affects: C2

Numeric verification checks (7)

Scope internal = numbers printed in the document; cross paper = other corpus documents; external / replication = outside sources or re-analysis of original data (web mode only).

ID	Status	Scope	What was checked	Computation	Result	Loc.	Claims
V1	pass	internal	Central District share.	8,759 / 18,332	47.78% — matches.	p. 7	C2
V2	pass	internal	District-weighted reconstruction of city-wide shares.	0.4778 x Central + 0.5222 x Rest for entire apts (76.3/51), commercial (80.07/71.64), multi-host (68.85/51.57)	63.09%, 75.67%, 59.83% vs reported 63.0%, 75.7%, 59.8% — consistent.	pp. 7-11	C3 C4 C5
V3	fail	internal	'Nearly 10 points greater than the Madrid average'.	80.07 − 75.7	4.37 points (8.43 vs rest of city).	p. 9	C4
V4	pass	internal	Host-band shares sum and >20 band listings.	74.50+21.19+3.65+0.66; 40.17+27.7+17.11+15.02; 2,753/18,332	100.00%; 100.00%; 15.02% — matches. Implied number of hosts ≈ 65/0.0066 ≈ 9,850.	pp. 12-13	C5
V5	pass	internal	Band-level district and accommodation shares reproduce totals.	Σ band listing share x band central share; Σ band share x band entire share	Central = 47.78% (reported 47.78%); multi-host share of central = 68.85% (reported 68.85%); entire apts = 63.06% (reported 63.0%).	p. 13	C5 C9
V6	pass	internal	2017 comparison: point changes vs relative growth.	2.92/(17.11−2.92); 6.61/(15.02−6.61); 4.84/(40.17+4.84)	20.58%, 78.6%, 10.75% — all match. Implied 2-5 band change: −4.69 points (not reported).	p. 13	C7
V7	pass	internal	'Less than 5% of hosts control over 30% of listings'.	3.65+0.66; 17.11+15.02	4.31% of hosts, 32.13% of listings — matches.	p. 13	C5

Overall assessment

Evaluation frame: Descriptive quantitative analysis of scraped platform data. Judged on reconciliation of reported figures, validity of proxies for 'professional' and 'commercial', soundness of the temporal reconstruction, and whether impact conclusions are matched by impact evidence.

Dimension	Score	Rationale
conceptual clarity	4	Clear operational definitions tied to an explicit sharing-economy benchmark.
logical coherence	3	Cross-sectional argument is tight; the abstract's impact conclusion contradicts the conclusion's call for future impact studies.
evidential support	4	Cross-sectional indicators are numerous and internally consistent; temporal and impact claims are weaker.
methodological rigor	3	Standard InsideAirbnb approach; retrospective growth method is biased; no sensitivity checks.
transparency	3	Methods described and a limitation footnoted, but the scrape date and underlying counts by band are missing.
proportionality	3	Main conclusion proportionate; the 'negative impacts' and 'monopolization' language overreach.

Strengths

Exceptional internal numerical consistency across districts and host bands.
Clear, reusable indicator framework.
Genuine two-point comparison with April 2017 data.
Acknowledges survivorship limitation of the growth series.

Weaknesses

Scrape date missing.
Growth series methodologically weak.
'Professional' label applied to all multi-listing hosts.
Impact on city sustainability asserted but not studied.

Robust claims

C2 C3 C5

Fragile claims

C6 C8 C10

What would strengthen the work

State scrape date and filtering; publish band counts.
Use consecutive scrapes rather than registration dates for growth.
Split 2-5 band by room type to isolate home-sharing multi-room hosts.
Drop or qualify the impact claim in the abstract.

Links to the rest of the corpus

X5 method drift moderate Shifting thresholds for 'commercial' / 'non-residential' listings

Availability >182 days defines commercial listings in Gil (2021) and Gil & Sequera (2020); Gil (2021) also uses >90 days to argue non-residence; Gil (2023) uses >60 days to estimate homes 'extracted from the rental market' and displacement.

Implication: Lower thresholds mechanically raise displacement estimates; sensitivity analysis is never reported.

Papers: Gil (2021)this paperGil (2023)

Premises taken from other corpus papers

Gil (2019): Distinction between peer-to-peer hosts and professional multi-listing hosts. robust

Provenance

Analyst: Claude (Anthropic), LLM-generated audit · 2026-11-08 · access: document only · text: text/gil_2022_airbnb_madrid.txt · schema v1.1

No external sources were consulted (offline run). Figures are assumed to be correctly reported; claims[].external, citation checks and external/replication checks were not performed.

PDF page 1 is the publisher cover page; article page n = PDF page n+1.