Determinants of Hotel Room Rates in Stockholm : A Hedonic Pricing Approach

Using Lancaster’s characteristics approach to consumer theory the study investigates the impact of a variety of characteristics on the rates charged for hotel rooms in Stockholm. In particular, it examines how a number of these characteristics explain variations in room rates between weekdays and weekends. Data for 105 observations (weekday sample) and 110 observations (weekend sample) were extracted from 49 hotels in and around the Swedish capital of Stockholm from the Internet-based hotel booking site hotels.com. Based on semilog regression analysis three models for the whole sample, weekday sample and weekend sample were estimated. The results show that the provision of breakfast, distance from the city center, availability of minibar in a room, provision of free cancellation option, average room size, number of guest rooms (hotel size) and hotel star rating have significant effects on both weekday and weekend room rates. By contrast, the presence of bathroom and the provision of free parking space turned out to be insignificant in both models; whereas the provision of a 24-hour room service and association with hotel chain appear to be significant only in the weekday model. The impact of the star variable is also found to be more pronounced during the weekdays than weekends. These results are expected to shed light on which hotel characteristics consumers have to pay extra, and enhance hotel managers’ strategic pricing. The study also highlights the critical role of a proper definition and measurement of the hotel star rating in hedonic pricing.Many hedonic studies have regressed the number of stars, which are ordinal measures of quality that indicate the relative ranking of hotels, in their cardinal sense. As a result, they have failed to measure the impact of each rating independently. To solve this recurrent problem, the study attempts to provide another method of quantifying the star variable whereby one can accurately measure the star rating and capture the impact of each rating independently by creating “one less dummy variable than there are alternatives”. It is hoped that this will inspire further research and analysis