TL;DR
This study reveals how "preferred" badges and positive reviews on accommodation-sharing platforms significantly increase landlords' likelihood of raising prices during peak seasons. Interestingly, while professional landlords are more likely to adjust prices, their percentage of price increase is actually dampened, suggesting a strategic pricing approach that leaves less room for further hikes.
What Problem Does This Paper Address?
The sharing economy, particularly platforms like Airbnb, has transformed accommodation rentals. However, a key challenge remains information asymmetry: renters often lack detailed knowledge about properties, leading to trust issues and potentially hindering smooth transactions. To combat this, platforms use signals like "preferred" badges (platform-certified) and customer reviews (user-generated) to build trust and inform renters. While previous research has looked at these signals individually, this paper investigates their combined effect on landlords' pricing strategies, specifically during peak demand periods like festivals and weekends, and how landlord professionalism influences these dynamics.
Methods
The researchers employed a quantitative approach using binary Logit models to predict the probability of landlords adjusting prices during peak seasons. They then used Ordinary Least Squares (OLS) regression models to estimate the magnitude (percentage change) of these price adjustments. To explore the moderating role of landlord professionalism, they incorporated interaction terms between the signals and landlord professionalism in their models. This hierarchical approach allowed them to disentangle the main effects of signals from their moderated effects.
Datasets
The study utilized a comprehensive dataset of 11,795 observations from 58.com, a leading Chinese accommodation-sharing platform. Data was collected from 36 cities and spanned 61 days (April 1 to May 31, 2022). This dataset was enriched with publicly available information like consumer price index (CPI) and provincial tourism revenue.
Key Insights and Ideas
- Signals Matter, Especially on Weekends: Both platform-certified signals ("Preferred House") and user-generated signals ("favorable rate") significantly increase the likelihood of landlords raising prices. Interestingly, these signals have a stronger impact on price adjustment probability during weekend peak seasons compared to festival peak seasons. This suggests landlords are more agile in responding to the less predictable weekend demand.
- Professionalism's Double-Edged Sword: While professional landlords (those managing more properties) are more likely to adjust prices during peak seasons (supporting H3a-d), their percentage of price increase is actually less pronounced. This implies they might already set higher baseline prices, leaving less room for further significant hikes.
- User-Generated Signals Dampened by Professionalism: Increased landlord professionalism weakens the impact of user-generated signals on the percentage of price adjustments. This is likely because professional landlords may prioritize consistent pricing strategies over rapid adjustments based on reviews, especially if they already have high base prices.
- Platform Signals Remain Consistent: Unlike user-generated signals, landlord professionalism did not significantly moderate the effect of platform-certified signals on the percentage of price adjustments. This suggests the influence of "Preferred House" status on price adjustments is relatively consistent across different levels of landlord professionalism.
Key Results
- "Preferred Houses" are 28.7 times more likely to see price hikes during weekend peak seasons compared to non-preferred houses.
- Each percentage increase in positive reviews ("favorable rate") increases the probability of landlords raising prices by a factor of 2.397 during festival peak seasons and 2.737 during weekend peak seasons.
- Hypotheses H1a, H1b, H1c, H1d, H2a, H2b, H2c, and H2d were supported, indicating positive effects of both signal types on price adjustment probability and percentage.
- Hypotheses H3a, H3c, and H3d were supported, showing that landlord professionalism positively moderates the effect of signals on the probability of price adjustments.
- Hypotheses H4c and H4d were supported, indicating that landlord professionalism negatively moderates the effect of user-generated signals on the percentage of price adjustments.
- Hypotheses H3b, H4a, and H4b were not supported, suggesting nuances in how professionalism moderates signal effects depending on the peak season and signal type.
Limitations and Future Work
The study acknowledges its limitations, including:
- Geographic Scope: The data is from China, which may not be generalizable to other markets with different characteristics.
- Signal Types: Only platform certification and user-generated reviews were considered; other signals (e.g., amenities, location) could also play a role.
- Future Research: The authors suggest exploring the impact of other signals, conducting cross-country comparisons, and investigating how factors like property size and amenities influence pricing. They also highlight the need for more sophisticated dynamic pricing tools and targeted marketing strategies based on demand dynamics.
Summary updated