Seminarios 2018

 

12 de abril

Paolo Figini; Booking in the Rain: the impact of weather forecasts on sea and sun Destinations.

A particularity of tourism is the dependence on factors beyond the control of supply and demand, as weather conditions. The availability of accurate weather forecasts increases the overall market efficiency, although not everyones welfare might be gaining but, on the contrary, if weather forecasts are biased and tourists are not aware of the distortion, the equilibrium is welfare worsening for both firms and tourists. It has recently been argued that private weather forecasters systematically bias forecasts for commercial purposes, with negative effects on the tourism market. Media have often echoed the complaints of tourism operators, amplifying the alleged cost that hotels have to pay for a strategy that has also been defined as meteo- terrorism. Within this framework, this paper tackles three main research questions. Is there any systematic bias in the accuracy of weather forecasts and, in case, is this bias larger for private than for public forecasters? Do (biased) weather forecasts impact on the tourists behaviour and, as a consequence, on the pricing strategy of the hotel sector? What is the estimated economic impact of wrong weather forecasts on the hotel sector in a seasonal sea sun destination? Data have been collected through a scraper in the period June September 2015, monitoring Rimini, an important Italian seaside destination. Each day the scraper collected information on: i) the weather forecasts published by three commercial websites / apps and one public provider; ii) the actual weather as recorded in the official and public archive; iii) the prices published on an important booking engine by the population of hotels of Rimini for standard holiday types. The lead period considered is of 15 days, for both weather forecasts and prices. The whole dataset includes more than 1 million observations related to 874 hotels. Results show that weather forecasters are not systematically biased and, surprisingly, the commercial providers are more accurate than the public service. Moreover, by regressing the posted price on the weather forecast posted in the same day we found that, controlling for a set of variables, that in the case of a popular private forecaster there is a strong impact of weather forecasts on prices: the worse the forecast, the lower the price. Finally, we provide a rough estimate of the total cost of bad forecasts, by distinguishing the impact of bad weather from the one of bad forecasts. The joint reading of these results allows us to conclude that, on average, weather forecasts are not biased and that, as a consequence, the variability of prices when the forecasts are bad is the rational response of the tourism supply when tourists are less willing to pay for a holiday break that is perceived of lower quality.