Keywords

1 Introduction

Airbnb has experienced remarkable growth since 2018 and became a dominant player in the peer-to-peer (P2P) tourist accommodation market. By 2017, Airbnb featured 4 million accommodations—more than the collective capacity of the world's top five hotel chains (Hartmans, 2017). As a disruptive innovation, Airbnb revolutionized the accommodation market, particularly in urban centers, and triggered negative externalities (Bugalski, 2020).

However, in March 2020, the trajectory of Airbnb and the entire global tourism industry faced an abrupt disruption due to the transmission of COVID-19, as declared by the World Health Organization (WHO, 2022). Shortly thereafter, Brian Chesky, the CEO and co-founder of Airbnb, acknowledged the devastating impact of the preceding months in a letter to the platform’s hosts (Airbnb, 2020), outlining measures such as allowing reservation cancellations, full fee refunds, and various host assistance programs.

Given the evolving context, we ask the following question: Has the pandemic redefined the P2P market for tourist accommodation? We focus on supply, demand, and prices. We begin by outlining a microeconomic model of consumer choice, the empirical methodology and the pertinent data to answer our questions. Later, we compare recent results with our perceived trends using economic arguments, specifically the overall decrease in demand, the impacts of the pandemic's dual supply and demand shocks on hosts’ prices and income. Finally, the conclusions are exposed.

2 Conceptual Framework

Based on the quasi-linear utility derived from homothetic preferences (Varian, 1983) and on the Hotelling model of horizontally differentiated competition (Hotelling, 1929), a consumer’s dilemma can be seen as the choice among various alternatives: an Airbnb, a hotel room, and the outside option (staying home). Individual preferences are distributed among a continuum in [(0, V1), (0, V2)] where v > 0, with a joint cumulative distribution function F. ui1, being the utility of Airbnb, will increase with i's value for social factors such as befriending the host or living like a local. If i values professionality more, then ui2 > ui1. Prices p1 and p2 are equal for all consumers.

$$\underset{{D}_{i1},{D}_{i2}}{{\text{max}}}\left\{{u}_{i1}-{p}_{1}, {u}_{i2}-{p}_{2}, 0\right\} | {D}_{i1}+{D}_{i2}\le 1, {(D}_{i1},{D}_{i2})\in {\left\{\mathrm{0,1}\right\}}^{2}$$

where Di1, Di2 are dichotomic variables that equal 1 when the decision is to go to the option 1, 2 respectively (if both are 0, the outside option is chosen). xi1 and xi2 vary by consumer according to certain cumulative distribution function (xi1, xi2)~F. Therefore,

$$Pr.\left({x}_{i1}>{x}_{i2}\cap {x}_{i1}>0\right)=\int \int \left[1 -\mathrm{ F}\left({{\text{x}}}_{{\text{i}}2}, {{\text{x}}}_{{\text{i}}1}\right)\right]{\text{f}}\left({{\text{x}}}_{{\text{i}}1}\right) {{\text{dx}}}_{{\text{i}}1}{{\text{dx}}}_{{\text{i}}2}$$

where f(xi1) is the probability density function (PDF) of xi1, as the share of potential consumers in the market that book Airbnb. A shock, such as the pandemic, negatively affects xi1 and xi2 for all, but does not change the outside option. So what we expect is a decrease of the total share of Airbnb and hotels, but their relative share varies ambiguously if the effect varies by option. As for the hosts, they may initially enjoy a markup in pricing due to this differentiation but now will lose part of it due to the fall in demand. And if the pandemic were to affect the marginal cost of supply, total occupancy could further contract.

We use data the unofficial database InsideAirbnb (2022), extensively utilized in previous P2P accommodation research (Dann et al., 2019), which offers metrics such as city-specific listings, guest reviews, and host pricing. Our sample features Amsterdam, Austin, Barcelona, Bordeaux, Dublin, Florence, Los Angeles, London, Lyon, Madrid, New York, Paris, Sydney, Toronto, and Vienna selected based on the completeness of monthly data between February 2019 and 2021. We measured supply by counting the number of listings, thereby quantifying the availability of accommodations in each city. For approximating occupancy, we employed the number of reviews as a lower-bound, reflecting confirmed guest activity. Assuming the pandemic did not significantly alter the behavior of guests refraining from leaving reviews, we can reliably employ this variable as a proxy for occupancy in our study. Additionally, we calculated the average prices for each city. To obtain the global average across the 15 cities, we weighed them proportionally according to their number of listings.

We suggest the following hypotheses: the pandemic might have caused (1) decrease in the number of stays; (2) a transformation in the type of accommodations offered (a shift towards entire homes due to isolation measures, reinforcing prior trends against the “shared” nature of P2P); (3) a reduction in supply and/or pricing; and (4) resilience of the P2P accommodation market because it has shown adaptability to shocks in the past.

3 Results

To answer our questions, we compare results to those of prior studies. Airbnb’s growth resulted in vast literature, with review articles (Dann et al., 2019; Dolnicar, 2019; Guttentag, 2019; Hati et al., 2021; Ozdemir & Turker, 2019; Prayag & Ozanne, 2018; Sainaghi, 2020) that substantiate critical findings (Benítez-Aurioles, 2020).

As reference points, we tracked articles through Web of Science, Scopus and Google Scholar published in 2020 and 2021 containing the keywords “COVID” and “Airbnb”, removed speculative and marginally related contributions and selected 15 papers with quantitative analyses and arguments on the pandemic's impact on the P2P accommodation market: Benítez-Aurioles (2021a; 2021b); Boros and Kovalcsik (2021); Chen et al. (2020); Dolnicar and Zare (2020); Gossen and Reck (2021); Gyódi (2021); Jang et al. (2021); Kadi et al. (2020); Llaneza and Raya (2021); Liang et al. (2021); Martínez (2021); Trojanek et al. (2021); Yiu and Cheung (2021); and Zhu and Liu (2021).

To verify the fall in occupancy of the pandemic, we look at stays. Figure 1 shows the impact on demand and supply. Reviews, as a proxy for occupancy, fall in all cities since Q1-2020. Because guests have 14 days after leaving to review, March might show more activity; but there is a marked decline afterwards. The number of listings also supports fast supply adjustments to demand changes. Pre-pandemic, there was only a fall in supply after the summer season. This affirms supply-side adaptability to market conditions, suggesting a demand drop led to the withdrawal of accommodations with reduced expected occupancy.

Fig. 1
A double-line graph compares the evolution of the total number of listings and reviews on Airbnb in the set of 15 selected cities versus the time from February 19 to February 21. Both lines first rise until February 2020, then decline with fluctuations.

Source Compiled from InsideAirbnb (2022)

Evolution of the total number of listings (rigth) and reviews (left) on Airbnb in the set of 15 selected cities (*), in thousands. February 2019–February 2021. (*) Amsterdam, Austin, Barcelona, Bordeaux, Dublin, Florence, Los Angeles, London, Lyon, Madrid, New York, Paris, Sydney, Toronto, and Vienna.

Full size image

Additionally, Eurostat data offers a valuable point of comparison. Table 1 displays overnight stays in EU countries from 2019 to 2020, revealing a substantial decline in tourist demand due to the pandemic. Overall, it fell by 46.7% in the European Union with inter-country variety (from 20.6% in Germany to 72.1% in Malta). Liang et al. (2021), in 12 major cities worldwide across Europe, America, Asia, and Oceania, confirm a significant decrease in Airbnb reservations during the pandemic, particularly among foreign tourists, supporting that the pandemic led to a notable fall in occupancy rates.

Table 1 Overnight stays (thousands) in tourist accommodation based on data from digital platforms. Entire homes (E), shared rooms (S), percentage of E over the total (% E), percentage increase of the total between 2019 and 2020 (% ∆ total)
Full size table

As for prices, Fig. 2 shows a drop in the P2P tourist accommodation market assignable to a decline in both supply and demand during the summer of 2020. However, (1) these may reflect actual exchange prices, and (2) the weighted average says nothing about the characteristics of the remaining post-pandemic accommodations. In sum, initial descriptive analysis suggests a reduction in both sides of the transaction, which result in a fall in volume but have an uncertain theoretical impact on price (although our data leans more to a downward movement).

Fig. 2
A line graph of the evolution of the average price for accommodation in the set of 15 selected cities versus the time from February 19 to February 21. The line first rises until February 2020, then declines with fluctuations.

Source Compiled from InsideAirbnb (2022)

Evolution of the average price for accommodation, weighted by number of ads, in the set of 15 selected cities (*), in dollars. February 2019–February 2021. (*) Amsterdam, Austin, Barcelona, Bordeaux, Dublin, Florence, Los Angeles, London, Lyon, Madrid, New York, Paris, Sydney, Toronto, and Vienna.

Full size image

To address whether the pandemic caused a drop in prices or if the supply decrease outweighed the demand decline, quantitative estimates like Benítez-Aurioles (2021b) identified statistically significant price decreases during the pandemic in 22 cities; Chen et al. (2020) reported host income losses in Sidney of nearly 90% comparing January to August 2020. Jang et al. (2021) revealed spatially diverse revenue losses in Florida destinations. These suggest that the demand drop was more substantial than that of supply.

For the question of whether the percentage of entire homes increased, we observed that there was a growing prevalence of entire home accommodations in the P2P market before the pandemic (Benítez-Aurioles, 2021a; Ke, 2017), but this amplified the trend. In most European Union countries except Malta and Estonia, overnight stays in entire homes increased from 2019 to 2020, often exceeding 95% of the total, which suggests that private and shared room rentals may lose significance within the P2P market for tourist accommodation due to health concerns related to interactions with hosts.

Assessing the adaptability and resilience of the P2P accommodation market, we remember it has lower entry and exit barriers due to small fixed costs. Temporary closures due to low tourist demand are common in the hotel industry, and seasonality in tourist demand within the P2P market is evidenced before the pandemic (Benítez-Aurioles, 2021b). Llaneza and Raya (2021), for Barcelona, identify (1) professionals adjusting prices and minimum stays and (2) non-professional hosts who do not react during the pandemic. Li et al. (2019) already showed this ability of professional hosts to stabilize as a strategic response to maximize profit. Farmaki et al.’s (2020) semi-structured interviews with P2P accommodation hosts in a series of Mediterranean countries found there were pessimistic hosts considering leaving the platforms and optimistic ones who think the pandemic could enhance the P2P sector. All of these findings support that hosts adapt according to the market incentives of each changing circumstance.

4 Conclusions

The COVID-19 pandemic has given us a chance to confirm if the trends in the P2P tourist accommodation market pre-crisis are persevering—particularly those from its rapid expansion. Given the reduced pandemic era demand, we have drawn conclusions answering our hypotheses about the changes in this industry. First, the data supports the hypothesis of a significant decrease in the number of stays in the P2P market. As revealed in our analysis, there has been a pronounced and sustained reduction in guest activity, reflected in a decline in the number of reviews. Second, prior trends indicating a shift towards entire homes in the P2P market have been reinforced by the pandemic. The substantial increase in the percentage of overnight stays in entire homes across European Union countries underscores this, primarily attributed to health concerns and the minimization of social contact with hosts. Third, the simultaneous shocks to supply and demand during the pandemic have led to a decrease in prices and income for hosts. However, professional hosts, who manage multiple accommodations, have demonstrated their adaptability by employing revenue management strategies, including lower prices and extended stay options, effectively transforming short-term rentals into a medium-term market. This strategic response, aimed at stabilizing demand, suggests that the P2P market can effectively adapt to changing circumstances. Lastly, and in the same line, we can say that despite the challenges posed by the pandemic, the P2P accommodation market has displayed resilience, evidenced by the adaptability of hosts, the market's continued operation, and the confirmation of previous observations during a period of significant demand contraction.

In terms of implications, our study distinguishes itself by analyzing multiple destinations rather than focusing on a single one. This information can be harnessed by policymakers to bolster risk management strategies, including the development of support mechanisms for P2P hosts in times of tourism crises. Moreover, aligning these insights with more robust regulations for the P2P accommodation sector can yield substantial benefits. Enhanced regulatory frameworks empower governments not only to aid but also to implement more effective taxation and oversight measures, curbing excessive activity. As for the sector itself, the growing demand for “isolated” stays has reinforced the trend toward an emphasis on entire-home accommodations. Thus, hosts who remain in the market should position their offerings as viable competitors to traditional hotels.

Future research has the opportunity to expand on these findings by confirming how effective host strategies are in recovering from the crisis's impacts. We could delve into host responses to pandemics in various geographical regions, incorporating guest perspectives and social aspects. Another avenue for research could involve comparing these findings with those from the traditional hospitality industry. This comparative analysis could help us draw implications for future responses to health emergencies. Furthermore, exploring the changing regulatory environment and its impact on the P2P market could offer valuable insights into the industry's direction in the aftermath of the pandemic.