To have a clearer picture that guides decision-making adapted to the times of the COVID-19 Pandemic, the Planning and Development Directorate of the Costa Rican Institute of Tourism analyzed two statistical models to visualize trends in tourist visitation in the future.
It is a model of gradual recovery of 5% called TCC (Continuous Growth Trend) and a simple seasonal model whose results are compared with a projection of the situation prior to COVID-19, that is, a projection of what what would have happened if the Pandemic had not existed.
Both models are dynamic and their results must be handled with caution as uncertainty is a constant in current health circumstances. The models, then, are sensitive to a number of factors, among them: advance of vaccination, opening of borders, closures, confinements, fear of traveling, among others. For this reason, they will be reviewed every three months.
Gustavo J. Segura, Rector Minister of Tourism, stated that “tourism recovery and public policy require statistical tools to measure demand, especially given the impossibility of applying the models we used before the Pandemic to make future projections”.
For his part, Rodolfo Lizano, ICT planning and development director, who presented the models to the press representatives on Monday, said that “ICT has always had a commitment to professionalism and continuous improvement of tools for measure tourist activity, which allows the institution to direct the tactics and actions for the promotion of tourism”.
About the models
The explanation of both models requires a pre-COVID-19 base that uses the arrival data between January 2010 and February 2020 and the Holt-Winters estimation model with an adjustment of 0.986, which has allowed estimating the expected trend that it would have followed the series had the epidemic not existed.
TCC Model (Continuous Growth Trend)
It considers the monthly estimates of the pre-COVID-19 estimation model (in non-pandemic conditions and with the Holt-Winters model) and proposes a gradual recovery of 5% per month.
That is, the first estimated month uses 5%, the second month 10%, the third month 15%, and so on. These percentages are applied to the monthly estimates of the pre-Covid-19 estimation model.
This model is related to the expected evolution of the Pandemic and the respective flexibility of mobility restriction measures and the gradual opening of different types of tourism services and companies. With the TCC, you could be closer to the pre-pandemic figures through the first quarter of 2022.
Simple seasonal model
The seasonal variation model with trend is an optimal model for demand patterns that show cyclical behavior and that in turn show a particular trend. It allows determining the forecast when there are periodic fluctuations in the time series, this, generally, as a result of the influence of economic phenomena or, as in this case, of the pandemic.
Certainly, the analysis of the data series between the month of March 2020 and December 2020 indicates that the data series no longer fit the Holt-Winters model, due to the great uncertainty that exists at the present time. Instead, the analysis indicates that for this time series the most suitable model is the simple seasonal model with an adjustment of 0.917.
The projection values obtained with this model are made up of the lower confidence limit, the prediction as such and the upper confidence limit as a reference for the behavior that the data on international arrivals could present. It is recommended to consider the prediction value and the upper confidence limit.
It is important that entrepreneurs, collaborators and travelers remember that throughout this recovery process it is key to continue promoting the application of health protocols and, in particular, the provisions on hand washing, physical distancing, bubble travel and the use of the mask.