Covid-19 and touristic demand
A mathematical modeling approach
Travelling has for sure an important role in everyone’s life. It helps us in broadening our minds and satisfying human fundamental needs of exploring and discovery. That’s probably why tourism industry has been able to overcome a significant number of crises that set upon it through the years, such as terroristic attacks, epidemics and natural disasters. And also in the present situation, while the entire economic system is facing tough challenges, turism industry is the one that will be likely to suffer the consequences of the Covid-19 pandemic longer. Even though, provided the volatility and novelty of the current situation, making estimates is harder than ever, the present article aims at unpacking travelling anxiety and its consequences on touristic demand in its essential components, taking inspiration from precedent crises research and published consumer sentiment analyses.
Similar events research
The first tricky part in tackling the problem of understanding the Covid-19 impact is to understand which comparable events in the past can be taken as a reference to build a forecasting model. In order to do that, it might be helpful to try to list the main distinctive features that characterize the present outbreak:
- Crisis type: Covid-19 represents most of all a health crisis, even though due to its catastrofic consequences it is becoming an economic one as well
- Geographic boundaries: started from localized focuses, the spread of the disease is currently global and fast evolving, also due to the virus high infection rate
- Time boundaries: this crisis is prolongued in time and it is currently not clear when it will finish and if it might be cyclical, depending on specific factors
Provided the peculiarities above, we have identified two major events happened in the past years that present similar features to the present outbreak, and that had in the past important impacts on travel anxiety and touristic demand: the 9/11 terroristic attack of 2001 and the SARS epidemic of 2003. Let’s take few lines to briefly describe the two events, in order to better understand similarities and dissimilarities with Covid-19 crisis:
9/11 (2001) The 9/11 event has been an unprecedent terrorist-related event. Even though it had major consequences on the US travel market, due to the media attention received and to the extraordinary type of event, it affected international travel demand worldwide. This event radically changed security and flying policies in travelling, with the introduction of new strict regulations.
SARS (2003) The SARS (Severe Acute Respiratory Syndrome) has been the first pandemic of the 21st century. Even though it represented itself a health crisis, it was strongly connected to the travel industry, since “travelers belonged to those primarily affected in the early stages of the outbreak, travelers became vectors of the disease, and finally, travel and tourism themselves became the victims”. This epidemic had its core between March 2003 and June 2003 and Asian tourism industry had been the one most impacted from caused travel anxiety, facing a substantial decrease in travel demand.
Those two events and the related impacts on airline demand are widely studied in the paper Comparing the Impact of the September 11th Terrorist Attacks on International Airline Demand from H. Ito and D. Lee. As studied in the article, both events produced a decrease in RPKs produced around the 20%-35%, reaching a the minimum in a three month range, and taking between 6 to 9 months to reach again the pre-crisis levels.
In their article, Ito & Lee study the impact of travel anxiety on airline demand, by modelling it as a linear function of some key parameters, among which shock and fear effects are taken into account, togheter with other important variables representing the supply influence, the economic trend and the demand seasonality. In the following we will explore in more details the shock and fear effects, by undersanding how they can be modeled to analyze the present situation.
A mathematical model : φ(shock, fear)
As anticipated above, in the analysis presented by Ito & Lee, by analysing 9/11 and SARS crises impact on air travel demand, shock and fear factors emerge as the ones that most of all characterize the crises negative influence on touristic demand. These two components can be seen as acting in two distinct phases in time: the first one is related to the short term consequences of the crisis and is directly linked to its duration, while the latter is more related to the long lasting marks that the crisis will leave on consumers and their perception of travelling risks.
Let’s see in more details how can we interpret these two components:
Shock. This term represents the immediate effect on bookings slow down or cancellations that are registered. This can be seen as a quantitative translation of the resistence to travel connected to the concern of having logistic or health problems due to the presence of Covid-19. In more details, this term takes into account, for example, the fear of having your travel cancelled due to a virus relapse, being blocked in a foreign country with no possibility to come back home because of frontiers closing or even worse, getting infected by the virus and bring it back to your country, being in turn the cause of a new focus. This component can be modeled as a direct function of the pandemic presence, i.e. daily contagions.
Fear. This second term is more related to the long lasting resistance to travel that has been instilled by this unprecedent situation. This relates with travel anxiety and perceived unhealthiness of specific touristic venues as a result of the current media impact. Let’s think for example to the Diamond Princess case or quarantined hotels due to the virus presence. Some travel situation naturally facilitate the creation of crowds and queues in relatevely closed ambients and, in a situation where social distancing seems to be the only solution to survive and fight the virus, those situations will last in our minds as potentially risky ones for a long time. In order to quantitatively assess how much this second component is radicated yet in the travellers minds and how the travelling products can be changed in order to survive to this new reality, consumer sentiment surveys can help, together with looking ahead to countries which are slowly approaching the after-Covid new normal, such as China.
Shock effect and willingness to travel
As anticipated, the shock effect can be seen as a direct function of the daily contagions movements in time. In order to give a high level interpretation of this fact, we can focus on Italy situation, by comparing Google queries for travels (viaggi in Italian) with the daily contagions trend in the last 90 days.
Even though Google trends might be also impacted by impossibility to travel due to governments travel bans, provided the generality of the keyword analysed, we can assume that the extracted trends are mostly related to the awareness and consideration phases of the travel shopping funnel, which takes place approximately on average 45 days before the travel date. As we can see from the graph above, the two compared quantities are as expected negatively correlated, meaning that the higher the number of contagions, the lower the travel searches (and willingness to travel in the short term). By understanding the interplay between these two quantities it is possible to quantitatively measure how long the shock effect will last in the time and how intense it will be, once a forecast on future contagions has been obtained.
Regarding this last point, it has to be said that it is of couse hard to make predictions on how the disease will evolve in the future and how long it will last. Actually, first of all, this is an experts field and can’t be blindly deducted from the few contagions data collected to date. Secondly, this is impacted by many different factors, such as social distancing measures, number of samples collected and tests done or the discovery of a vaccine for the disease. However, just to have a high level perception of how the contagions curve could evolve following an approach purely based on data recorded to date, a skewed-normal fitting forecast applied for Italy, France, Germany, Spain and US cases is provided below.
Fear effect and consumer sentiment
The fear effect being more related to subconscious connections that our minds are now creating between Covid-19 and travel experience, it is hard to assess that quantitatively and in a stage when the shock component is still active and predominant. From an ex-post analysis on the 9/11 and SARS crises, Ito & Lee estimated in their article a negative impact out touristic demand of approximately 15% due to the fear effect. But how can we know if it is correct to assume a similar impact will occur also for the present crisis? Of course, there is no aswer yet to this question. However, various and extensive analyses on Consumer Sentiment are currently developed and periodically published by many firms and can help in having at least an idea about how travel risk perception is changing in consumers’ minds. Among those, I personally suggest the weekly BCG Consumer Sentiment snapshots, for their completeness and readability. In the first two snapshots, some figures about willingness to travel are presented, which outline the increasing travelling concern due to perceived contagion risk in travelling situations.
Actually, almost 38% of the respondants see travel and taking a cruise or domestic flight as risky activities. They resulted to be also concerned about other travel related activities, such as visiting theme parks, casinos, hotels and taking a bus or a train. The analysis shows furthermore that respondants expect to spend on average -26% on activities that require travel and engagement in a group environment. These are for sure first insights that tell us that for sure we will see a raising and persistent component of a fear factor which will keep slowing down touristic demand in the long run. However, we postpone a quantitative estimate of such a contribution to a point in time when the shock effect will be vanishing, letting us better isolating the intrinsic changes this outbrek has left in our travelling behaviour.
Hope this read could have helped some of you in providing a methodology to understand how the present outbreak could impact our future willingness to travel and why. Any comment or feedback is highly appreciated!
Let’s see at the next read.
