Modelling Tourist Arrivals in South Africa To Assess The Impact of the COVID-19 Pandemic on the Tourism Sector

The paper examines tourism flows from all foreign countries to South Africa (SA) from 2009 to February 2020 using time series models. The resultant model is used to forecast and assess the impact of COVID-19 on tourist arrivals in South Africa by comparing with actual tourist arrivals after February 2020. Monthly data on tourist arrivals to South Africa from the Overseas, the Southern African Development Commission (SADC), other African countries and those who did not specify area of origin were considered. The Box and Jenkins methodology identified the ARIMA(1,1,0)(1,1,0)12 as the best model as confirmed by the Akaike Information Criterion (AIC) and the Bayesian Information Criterion (BIC). The model has good forecasting power as evidenced by its Mean Absolute Percentage Error (MAPE). Therefore, the fitted model’s forecasts may be used in the absence of COVID19.The COVID-19 adversely affected forecasting of future values as forecasted values differ greatly with actual observations recorded since COVID-19 induced lockdown measures in March of 2020. This study helps to have an overview of the total contribution being realized from tourism receipts that can be inferred through tourist arrivals before and after the COVID-19 pandemic and how best the tourism sector may be rejuvenated. The study highlights are that the number of tourist arrivals to South Africa follows an ARIMA(1,1,0)(1,1,0)12 model. Prior to the COVID-19 pandemic, there was a persistent upward trend and seasonality in tourist arrivals to South Africa since the 1994 democratic elections. South Africa is currently losing more than 90% of its monthly tourist arrivals because of the pandemic, and shows little sign of an imminent recovery, hence a devastating impact on the tourism industry.