Navigating Time: A Comparative Analysis of Senior Tourists’ Intentions and Length of Stay in Italy Pre-, during, and Post-COVID-19

Abstract

The worldwide COVID-19 pandemic led to a rapid global health crisis, which shocked the tourism industry by closing borders and reducing tourism demand. This study aims to analyze the impact that the COVID-19 pandemic has had on senior Italian tourism. If, on the one hand, seniors represent, from a tourism perspective, one of the most promising markets for the near future; on the other hand, they are the population that has most suffered the consequences of the pandemic event. Drawing on national statistical data, this study examines the impact of COVID-19 on the holiday intentions and length of stay of senior Italian tourists in Italy. The proposed methodological approach evaluates both travel intentions and length of stay by comparing the pre-pandemic, pandemic, and post-pandemic periods. It breaks down the decision-making process into two phases (to go on holiday or not, and length of stay) and is based on the estimation of a probit model with sample selection, which provides a more accurate approach to addressing the complex decision-making dynamics surrounding travel during pandemic periods, offering greater precision in modeling the factors influencing travel intention and length of stay. The results show a strong correlation between the decision to travel and the length of stay in the period before the pandemic. However, during the post-pandemic period there is a change in dynamics, where other factors gain importance in determining the length of stay regardless of the decision to travel.

1. Introduction

In recent years, and due to the growing ageing demographic, there has been a growing interest in studying the relationship between tourism and active aging of the population [1]. Given the growing number of seniors, this market should be considered as one of the most important for the tourism industry in the coming decades [2].

According to Eurostat statistics, senior tourism is flourishing within the EU. In Italy, in 2019, the over 55 age group represented the largest group of tourists who spent four or more nights in a domestic destination [3].

However, the pandemic led to a series of travel restrictions. Consequently, people’s freedom and willingness to travel has been significantly reduced [4].

Seniors have been hardest hit by COVID-19 infections and by rather high mortality rates. Zambianchi [5], analyzing the psychological impact of COVID-19 on the travel intentions of senior Italian tourists, highlights the following aspects as relevant constraints: anxiety, fear, and a lack of adequate precautions by accommodation facilities or tourist destinations. The author noted that this group will experience the greatest change in the way they plan their vacations.

Indeed, seniors perceive much more risks, and have more worries and fears related to tourism. These aspects highlight the importance of age when tourism-related destinations and enterprises target specific market segments [6].

The pandemic has led to a combination of different types of crises and generated quite high levels of uncertainty. Thus, as crises often change travel motivations, it can be assumed that travel intentions and motivations may have changed following the pandemic [7]. It should be added that the government’s lockdown policy may have affected travel intentions and, especially, the length of stay (LOS) of tourists, as they did not have the opportunity to stay in the destination longer than they would have without the travel restrictions or quarantine measures.

From this perspective, the behavior patterns of tourism have changed. Tourists tend to progressively choose less crowded destinations, with the possibility of enjoying nature, seeking safety, hygiene, and tranquility, as well as choosing domestic destinations [8,9]. These changes offer new opportunities for tourist destinations.

Nature-based recreation and tourism activities are physical activities performed outdoors or in a natural environment for leisure. These activities are often seen as an essential component of daily life, which benefit the wellbeing of seniors. In particular, such activities benefit senior people by offering them the opportunity to improve their vitality and satisfaction by motivating them to continue living an active life [10].

In the context of the COVID-19 pandemic, these benefits have taken on new meaning, given the impact of social distancing and government restrictions.

In fact, spending time on leisure activities outdoors, close to home, increasing their physical activity, and walking have been tools adopted by many people when facing the period hindered by personal limitations [11]. This has led to an emphasis on the role of outdoor recreational tourism and nature-based tourism during and following the COVID-19 pandemic. Nature-based tourism concerns sustainable tourism and involves tourists who seek to achieve relaxation, discover, learn, and escape into nature [12]. For many people, the pandemic restrictions and the closure of many leisure places meant that nature-based resources have taken on significant importance for carrying out leisure activities [13]. Research carried out during the first lockdown periods shows that nature-based recreation and tourism activities increased significantly, highlighting, in some cases, increases of more than 200% [14].

Furthermore, according to Zhang [15], nature-based tourism can contribute to active aging. Indeed, senior tourists wish to engage in outdoor activities, including trekking, scuba diving, sailing, and other activities [16]. These activities contribute to socialization and, consequently, to the improvement of their wellbeing during the recreational experience. These aspects are emphasized by Bianchi and Drennan [10], who argue that social interactions (particularly reuniting with loved ones) during outdoor recreational activities constitute fundamental aspects of wellbeing.

This social aspect becomes of significant importance when encouraging the inclusion and full enjoyment of the tourism product by all individuals, even and especially if elderly people and those with disabilities, mainly linked to advancing age, are involved.

Indeed, senior tourists with disabilities face several constraints related to their participation in tourism activities [17]. From this perspective, tourism must guarantee the inclusion of people with disabilities in society, satisfying the tourist and recreational needs of all tourists, including seniors, especially those of advanced age and/or with disabilities, aiming to satisfy and improve the quality of their lives. These are key aspects of the United Nations’ Sustainable Development Goals [18].

Inevitably, all these aspects affect the travel intentions (to go on holiday or not) of individuals and the LOS [19]. Driven by the effect of the coronavirus, this study aims to explore the impact of the COVID-19 pandemic on the travel intentions of senior Italian tourists (their decision-making process toward the decision on whether to go on holiday or not) and evaluates its effect on their LOS.

The decision-making process of travelers has been considered to understand whether the travelers have chosen to maintain or cancel their travel plans (go on holiday or not) and, if they have opted for the first solution (to go on holiday), which variables have influenced their LOS in the destination. This approach is consistent with Nicolau and Más [20], who propose a two-stage choice process if the decision to go on vacation and the LOS are nested and are not independent. Furthermore, the comparison with subsequent years, characterized by the easing of the restrictive measures, the diffusion of the vaccine, and the reduction in pandemic cases, will allow us to evaluate the validity of the model on the behavior of senior tourists, even in the absence of the travel restrictions imposed due to the pandemic.

Although there have been several publications on the impact of COVID-19 on the tourism sector [21,22], few studies have addressed the impact on the travel intentions and LOS of seniors [23]. While the effects of the COVID-19 pandemic on the travel intentions and LOS of senior tourists in the years following the pandemic are lacking, to the authors’ knowledge. Some studies highlight the extent to which senior tourists believe that the COVID-19 pandemic will change the way they go on holiday [5]; some risk-focused research points out that senior tourism will continue to be discouraged or limited due to their vulnerability to COVID-19 [24]; and there are even studies that have used personality traits as segmentation variables [25].

To fill this gap, this study aims to provide a key aspect to understanding senior tourism during a period marked by uncertainties and travel constraints. Our primary objectives are to gain an empirical understanding regarding the variations in the travel intentions and LOS among senior Italian tourists. To achieve this, we compare four distinct time periods: the pre-pandemic era (2019), the pandemic period (2020), the onset of COVID-19 vaccine administration, and its subsequent dissemination (2021), extending our analysis through 2022.

Particularly this study aims to answer to the following research questions:

RQ1: How have the travel intentions and LOS evolved among senior Italian tourists in the time periods preceding and following the COVID-19 pandemic? And subsequently:

RQ2: Has there been a substantial rise in the preference for nature-based tourism among senior Italian travelers during these time periods?

RQ3: To what extent has senior tourist behavior undergone significant changes and what are the key factors influencing these shifts in behavior?

These factors may become apparent due to two main drivers: the continued need for social distancing measures in the post-COVID-19 era and the growing emphasis on sustainability goals. These factors could necessitate destinations that accommodate fewer tourists, encourage longer stays and, consequently, boost tourist spending in those destinations [26]. Additionally, this study adds value to the existing literature by quantifying the changes in the LOS brought about by the pandemic for senior Italian tourists. It also offers insights into the senior Italian tourist segment within the national tourism market by distinguishing between “young seniors” (aged 55–64) and “older seniors” (aged 65 and above), even in the years that followed the peak of the pandemic.

2. Literature Review

2.1. Definition of Senior Tourists

The categorization of senior tourists lacks a universally accepted age standard in the tourism literature. Various authors have divergent definitions, ranging from a minimum age of 60 years [27], to over 60 [28] and, even, over 65 [29]. The age range often varies across studies, contingent upon country-specific considerations and differences in retirement systems. In this study, we adopt a specific framework proposed by Campolo et al. [9], delineating senior Italian tourists into two distinct categories:

–

“Younger seniors” (aged 55–64 years);

–

“Older seniors” (aged 65 and above).

This categorization choice aligns with the framework introduced by Campolo et al. [9] and allows for the exploration of age-related nuances within the senior tourist demographic.

2.2. The Effects of COVID-19 on Travel Intentions and LOS of Senior Tourists

The LOS, defined as the total number of overnight stays spent by a certain number of visitors in a destination, represents a very interesting variable in the study of tourism demand [30], especially in periods characterized by limitations and strong competition to attract visitors with greater purchasing power [31,32]. Furthermore, it is one of the most important variables in the decision-making process of a tourist [33], and it is linked to diverse determinants of the whole tourist experience, namely the choice of accommodation, the means of transport, or the existence of constraints, such as daily spending amount, time, and number of participants, etc. [2]. Consistently, the literature on the LOS has focused on understanding its determinants.

There is a relationship between senior tourists’ travel intentions and the LOS [34]. Although the reason for a “holiday” plays a decisive role in senior tourism, other reasons can be sought in relation to visits to friends and relatives (VFR), concerning wellbeing motivations, and concerning nature-based tourism experiences [2,35].

Indeed, as the pandemic continued, and health authorities and organizations promoted the utilization of open spaces, individuals started (re)discovering their local natural assets. This stemmed from a need to take part in socially distanced activities, spend quality time with family, prioritize exercise, and personal wellbeing, and reduce the time spent on indoor activities [36].

Furthermore, the emerging research shows that the perceived risk from the coronavirus produced uncertainty and fear, which has led to an increase in stress and vulnerability [8]. Zheng et al. [37] demonstrate how fear can influence people’s attitudes and behaviors by increasing their motivation for self-protection. The risk of contagion has affected the willingness of people to engage in tourist activities typically carried out in crowded places and within urban contexts [38].

Accordingly, the literature describes the development of other forms of motivation and preferences, more related to risk. Perceived risks induced potential tourists to reschedule their travel plans [39] or to travel anyway, but adopt strategies based on prudent behavior. For example, tourists might prefer short-distance travel, domestic tourism [40], or travel using their own car [41]. The only solution that prompted tourists to book a holiday, during the pandemic, was the possibility of cancelling the reservation for free [22]; in addition to the provision of travel insurance, in case of illness, including related to COVID-19 [21]. However, these measures should have been offered by travel agencies, which represent the most popular booking method for senior tourists [42].

2.3. Socio-Demographic Determinants

The literature describes a positive relationship between age and LOS. Based on demographic projections, seniors are the most representative tourism market in which LOS prevails [31]. Zimmer et al. [43] confirmed that the population aged 60–70 has the highest LOS in the tourism market. However, because of the pandemic crisis, it seems that some additional aspects have negatively affected these preferences. Li et al. [44], investigating the behavioral adaptation of tourists during the pandemic, demonstrate their subordination to cultural, socio-demographic, and psychological aspects. Almeida et al. [31] show that travel-related fears vary by socio-demographic status. Pappas [6], examining the risks associated with the COVID-19 pandemic, highlights how age was the most relevant factor. Accordingly, socio-demographic variables seem to play a decisive role in travel intentions. This is because seniors have experienced the highest percentage of deaths from COVID-19, especially in Italy where the virus spread first in Europe [45].

2.4. Socio-Economic Determinants

According to Losada et al. [46], senior tourists prefer longer stays because they have more free time (due to retirement) and greater purchasing power. From the perspective of tourist demand, LOS is positively associated with disposable income [47] and negatively associated with daily expenditure [33,47,48]. Accordingly, the decrease in income after retirement and the deterioration of health due to aging could affect the travel intentions and the LOS of senior tourists. Furthermore, the pandemic crisis has led to a serious reduction in household disposable income and a progressive increase in prices, as a result it has become more difficult for people to manage their finances. More specifically, an Italian National Institute of Statistics (ISTAT) [49] “survey on Italians” vacation prospects during the pandemic, revealed how Italians perceived the COVID-19 pandemic and the lockdown as factors that significantly affected their income. Particularly, economic concerns prevail among the constraints that affected tourism during the pandemic. This mainly concerned people aged 50–64, who indicated both economic constraints and fears related to the risk of COVID-19 infection as barriers to their travel intentions. While, for over 65 s, health problems preceded economic constraints.

2.5. Determinants Related to Travel Features

The destination attributes. One category of factors influencing the LOS is the tourist destination attributes [50,51]. From this perspective, the determinants linked to the material and immaterial attributes of a destination become relevant, such as the climate, the natural environment, the landscape, and the natural resources [47,52]. Salmasi et al. [33], analyzing the LOS in Italy, found that seaside destinations are preferred for longer holidays and to improve one’s wellbeing. This aspect assumes greater importance in the post-pandemic era; indeed, it emerges how people prefer more spacious destinations, with the possibility to practice outdoor activities [53,54,55].

More specifically, seniors are willing to travel in small towns and rural areas [56]. Consequently, densely populated cities are at a disadvantage in these new models, while nature-based tourism destinations have an advantage [24]. Several factors are linked to this trend. First, the trend can be attributed to the psychotherapeutic benefits associated with nature-based tourism, as highlighted in the study by Buckley and Westaway [57]. This segment of tourism is likely to continue evolving, as one of the most significant post-COVID-19 tourism trends.

Secondly, the appeal of nature destinations has grown during the pandemic and beyond, offering opportunities for relaxation, hiking, wildlife observation, and environmental appreciation. These activities have a notably positive impact on mental health, especially among senior individuals. Additional benefits encompass forming new relationships within the community, boosting self-esteem and confidence, and fostering personal development, as shown in the research by Bianchi and Drennan [10].

Komasi’s study, in 2022 [12], reveals that one of the key motivations for nature-based tourism often centers around environmental factors. Travelers are drawn to secluded, unspoiled destinations, where they can learn about and appreciate nature.

Beyond environmental motivations, cultural interests also play a significant role. Many tourists are intrigued by the prospect of experiencing traditional rural lifestyles, which foster a sense of belonging and attachment to original cultural practices and traditional farming systems. It is observed that individuals often gain insights into culture through their interactions with nature, as noted in the research by Huang and Lin [58]. From this perspective, nature-based tourists can be classified along a continuum, using Mehmetoglu’s [59] approach. “Specialists” travel with the primary motivation of experiencing nature and outdoor activities. In contrast, “generalists” consider nature-based motives as central but not the primary reason for their trip, leading to longer vacations.

Moreover, the definition of “rural resources” varies among authors. For some, it coincides with agricultural resources and the transformation of agricultural products, while for others, it extends to include cultural and artistic resources found in small towns and cities that characterize rural areas, as described by Belletti [60].

In the context of the pandemic, nature-based tourists combine their motivation for nature-based or rural resources with goals related to wellbeing, culture, VFR tourism, and the desire to reconnect with loved ones, as discussed by Hall et al. [24], Fan et al. [61], and Salmasi [33].

The type of accommodation. Among the types of accommodation preferred by senior tourists, Vigolo [62] includes hotels, relatives’ and friends’ houses, second homes, and tourist apartments. Bai et al. [63] underline that about 54% of people over 65 and about 46% of people aged 55–64 choose friends’ and relatives’ house due to the absence of accommodation costs. This allows them to extend their stay with the same budget [1,64]. Conversely, staying in a hotel reduces their LOS [33,65]. This trend was also confirmed during the pandemic, during which tourists tended to choose cheaper accommodation. Indeed, Baños-Pino et al. [23] demonstrate a reduction in expenditure for accommodation in the summer of 2020 compared to the pre-pandemic period.

The travel arrangements (solo traveler or a group of tourists). As the group size increases, the LOS decreases, due to economic constraints [65]. Consequently, solo travelers or couples have a greater LOS than groups [66]. The authors who studied the effects of SARS on Chinese domestic tourism, demonstrated a decrease in group tours and opportunities for socialization. This aspect becomes evident during the COVID-19 pandemic due to the risk of contagion, primarily for senior people [67].

The distance and destination. In general, the literature proves that there is a positive relationship between the distance from the destination of origin and the LOS [52,66]. Silberman [68] justifies this relationship with the travel costs, which are fixed and independent of the number of days spent in the tourist destination. Accordingly, longer stays allow for the costs to be split over a longer period. However, recent studies have revealed that tourists tend to minimize distance and seek lesser-known destinations because of the COVID-19 emergency [15,53]. Therefore, domestic tourism is the preferred travel option [40].

The means of transportation. Generally, air travel is associated with a longer stay [34]. In regard to domestic tourism, the choice to travel by plane can be linked to the presence of low rates offered by low-cost companies. Those who prefer to make use of different options during their journey seem to choose the use of their own car [69]. Public transport is the option most frequently chosen by senior tourists, compared to young seniors. However, the use of public transport decreased during the COVID-19 pandemic, firstly for seniors, coupled with an increase in the willingness to travel by car [41,69].

3. Methodology

3.1. The Model

To model the LOS, different econometric techniques were proposed. Alegre and Pou [65], to analyze the LOS on the Balearic Islands, applied a discrete choice logit model. To study the situation in Turkey, Gokovali et al. [50] used duration models to estimate the LOS. Martínez-Garcia and Raya [64] used a log logistics and Cox survival models to estimate the LOS in Costa Brava. Campolo et al. [9] utilized the zero-truncated negative binomial to investigate the LOS of senior Italian tourists. Nicolau and Más [20] proposed a two-stage choice process if the decision to go on holiday and the LOS are nested and non-independent. In doing this, they applied a random parameter logit model, which accounts for the unobserved heterogeneity of individuals.

Following other studies that deal with the subject/consumer decision-making process in two steps [20,70,71], in this analysis we employed a maximum likelihood probit model with sample selection [72,73] to separate the decision-making process into two stages (travel intention and LOS).

For our analysis, we utilized the Stata command “Heckprobit”, which is known for providing consistent and asymptotically efficient parameter estimates in such models.

This methodology is consistent with the issues raised by the COVID-19 pandemic. Indeed, the perception of the risk of infection places the COVID-19 pandemic into the mix in relation to the decision-making process and, therefore, concerns the travel behavior of tourists [74]; therefore, the perceived risk, fears and anxieties induced by the pandemic will most likely influence the travel intentions of potential tourists.

They have two decisions to make jointly: the proposed model, jointly estimating the travel intention and the LOS, allows the outcome to be estimated in an unbiased manner, avoiding methodological problems due to the possible sample selection.

Even if we are interested in modeling the outcome, there are two dependent variables in the probit sample selection model, because we must also model the sample selection process for each equation in the model.

In the first equation or selection equation (Equation (1)), the dependent variable (holiday) is a binary variable that indicates whether each case in the sample is observed or unobserved; while in the second equation or outcome equation (Equation (2)), the dependent variable (LOS) is a binary variable that indicates whether the subject takes a short or long holiday. This last dependent variable is observed only for subjects that go on holiday in the period analyzed. The error terms used in the determination of the selection and the outcome value may be correlated.

In particular, the dichotomous dependent variable in Equation (1) assumes a value equal to 1 if the subject goes on holiday, becoming a “tourist” in our analysis, and with a value equal to 0 otherwise. While the dichotomous dependent variable in Equation (2) assumes a value equal to 1 if the subject/tourist stays longer (greater than 4-day vacation, or equivalently greater than 3 nights of accommodation), and has a value equal to 0 otherwise. Here, longer stays include trips lasting four days or more [33], and it is constructed considering the days spent on vacation declared by the subject at the time of the interview.

The probit model with sample selection assumes that there exists an underlying relationship. The factors that influence the selection also influence the subsequent outcome of interest.

The model can be expressed as follows:

Selection equation:

$$Holida{y}_i=Z_i \gamma +u_i$$

(1)

Outcome equation:

$$LO{S}_i=X_i\beta +{\epsilon }_i$$

(2)

The following selection rule applies:

$Holida{y}_i$ = 1 if the subject goes on holiday

$Holida{y}_i$ = 0 otherwise

$LO{S}_i$ = 1 if the days spent on vacation $>$ 4

$LO{S}_i$ = 0 if $>$0 days spent on vacation $\le$ 4

$(LO{S}_i, X_i)$ observed when $Holida{y}_i=1$

Where $Z_i$ is a vector of the variables determining the selection process; γ are the parameters used to perform the estimate; $X_i$ is a vector of variables that determines the outcome LOS; β are the parameters used to perform the estimate; $u_i$ and ${\epsilon }_i$ are the error terms of the two equations: $u_i$ ~ N(0,1); ${\epsilon }_i$ ~ N(0,1); corr ($u_i$ ${\epsilon }_i$) = ρ.

The crucial aspect of the probit model with sample selection is that it involves two separate probit models: one for the selection equation and one for the outcome equation. Each equation has its own error term, but both error terms are assumed to be normally distributed with the same mean and variance. The correlation between these two error terms is a key feature of the model, and it captures the interdependence between the decision to participate (selection) and the outcome or behavior being modeled. This correlation parameter ρ, is estimated in the model and provides insights into how the two equations are related. When ρ ≠ 0, standard probit techniques applied to the outcome equation (Equation (2)) result in biased estimates.

We estimate the same model in 2019, 2020, 2021, and 2022, from a comparative perspective. Some of the explanatory variables used in the selection equation provided the demographic and socio-economic information of the subjects: woman is a dummy variable that indicates the gender (1 = yes, 0 = otherwise); age is a categorical variable that indicates the age (1 = 55–64 as reference; 2 = 65–74, 3 = ≥75); worker is a dummy variable about the subject’s working status (1 = yes, 0 = otherwise); education indicates the education level expressed in years of schooling; area is the area of residence, where the subject lives in Italy (1 = north as reference; 2 = centre; 3 = south and islands); married is a dummy variable that identifies whether the subject is married (1 = Yes, 0 = otherwise); N household is the number of household members.

In the outcome equation, further to some of the variables mentioned above (woman and education), we also include information related to the trip: daily expenditure represents the expenditure per person and per day at the destination; N participant is the number of participants in the holiday; travel time indicates the period/season (quarter) in which the journey took place (1 = January–March as reference, 2 = April–June, 3 = July–September, 4 = October–December); moving considers the movement from the area of residence to the area of destination (1 = from north to north as reference, 2 = from north to centre, 3 = from north to south and islands, 4 = from centre to north, 5 = from centre to centre, 6 = from centre to south and islands, 7 = from south and islands to north, 8 = from couth and islands to centre, 9 = from south and islands to south and islands); concerning the purpose of the travel we consider five dummy variables, one for each motivation (cultural trip as reference, holiday, rural, wellness, VFR). Moreover, we consider the way in which the tourist plans his/her vacation: booking accommodation is related to the accommodation reservation mode (1 = no reservations as reference, 2 = reservation made independently, 3 = reservation made through agency); booking transport is related is related to the way in which the transfer is booked (1 = no reservations as reference, 2 = reservation made independently, 3 = reservation made through agency); transport refers to the means of transportation (1 = other as reference, 2 = airplane, 3 = train, 4 = own vehicle); accommodation indicates the type of accommodation (1 = campsite as reference, 2 = similar hotel; 3 = hotel; 4 = rented; 5 = non-rented). Finally, we include a dummy variable, younger, which assumes a value equal to 1 if the age of the tourist is less than 65 years old or its value is 0 otherwise.

The probit selection model requires that there is at least one variable included in the selection equation, but not the outcome equation, to function as an exclusion restriction. In our case, the marital status, the number of household members, and the employment status are included only in the selection equation.

3.2. The Data

For our study, we selected two sub-samples from each round of four repeated cross-sectional surveys from the “Trips and holidays” section, a focus area included in the Household Budget Survey [75], spanning the years 2019 to 2022, covering both the pre-pandemic and post-pandemic periods. The surveys collect information on the tourist flows of residents, providing a set of information on tourist trips: daily expenditure, the duration of the holiday, the characteristics of the trips and the destination, and booking information, etc.

The Household Budget Survey is widely recognized and conducted by the national statistical institute (ISTAT), ensuring reliability due to its large scale and inclusive nature. The “Trips and holidays” focus within the survey is specifically designed to capture accurate insights into travel patterns, ensuring the validity of our data.

We only selected subjects aged 55 years or older, in accordance with the aim of our study.

For the year 2019, the sample is formed by 18,044 subjects (16,915 censored observation or no tourists, and 1129 uncensored subjects or tourists); for the year 2020, it is formed by 24,599 subjects (23,655 censored and 944 uncensored tourists); for the year 2021, it is formed by 27,079 subjects (25,843 censored and 1236 uncensored tourists); and, finally, for the year 2022, the sample is formed by 27,433 subjects (25,916 censored and 1517 uncensored tourists).

Over the four years under analysis, the data reveal a decline in the participation of senior Italian individuals in vacations, declining from 6.26% in 2019 to 3.84% in 2020. However, in 2021 and 2022, these percentages exhibit a gradual increase (4.56% and 5.53%, respectively), although they remain below the levels observed before the pandemic. Furthermore, the proportion of senior Italian citizens opting not to travel due to the health emergency as their primary reason rises from 25% in 2020 to 31% in 2021, coinciding with the partial relaxation of pandemic-related restrictions in Italy. In 2022, 10% of Italian seniors abstain from travel for the very same reason. This result confirms that, beyond the travel restrictions, senior Italian tourists’ travel risk appetite and perceptions influenced their psychological behavior and travel intentions, as also demonstrated by Pocinho et al. [8].

In both surveys, our samples show the same structure according to female gender (54% and 52% in the censored and uncensored sub-samples). The uncensored samples, i.e., tourists, have a higher educational level in both surveys (more than 12 years of schooling) and the percentage of workers is higher in respect to the censored sub-samples. Between the two surveys, the percentage of younger senior tourists increases (from 51% to 54%).

From 2019 to 2020, there is a decrease in the percentage of tourists taking a long vacation (from 50% to 43%), as well as a contraction in the daily expenditure (from EUR 81 to 74 per day).

The average number of participants taking part in the trip remains unchanged.

Different percentages are observed on the quarter preference in which the holiday took place, but this may be due more to travel restrictions than to personal choice.

Leisure trips increased in 2020 to the detriment of the remaining categories; furthermore, there was an increase in the percentage of tourists who booked accommodation directly, without a travel agency, and an increase in the percentage of tourists who did not book transportation. While, contextually, the percentage of tourists who chose to travel using their own car also increased (from 65% to 74%).

Finally, between 2019 and 2020, the percentage of tourists choosing to stay in hotels decreased, while the percentage relating to the “Similar Hotel accommodation” category increased. For descriptive purposes, we report in

Table 1. Sample characteristics: censored and uncensored sample in 2019–2022.

		2019		2020		2021		2022
		Censored Obs 16,915	Uncensored Obs 1129	Censored Obs 23,655	Uncensored Obs 944	Censored Obs 25,843	Uncensored Obs 1236	Censored Obs 25,916	Uncensored Obs 1517
Variable		Mean	Mean	Mean	Mean	Mean	Mean	Mean	Mean
Woman		0.54	0.52	0.54	0.52	0.53	0.52	0.54	0.50
Age:	55–64	0.37	0.51	0.38	0.54	0.40	0.50	0.38	0.51
	65–74	0.33	0.33	0.34	0.36	0.33	0.36	0.34	0.35
	>74	0.3	0.16	0.27	0.11	0.27	0.14	0.28	0.14
Education level (years of schooling)		9.4 (4.21)	12.15 (4.25)	9.88 (4.27)	12.46 (4.09)	10.01 (4.26)	12.57 (4.05)	10.01 (4.21)	12.55 (4.01)
Worker		0.24	0.37	0.25	0.39	0.26	0.39	0.26	0.38
Area of residence: North		0.45	0.54	0.44	0.59	0.43	0.59	0.43	0.57
Centre		0.2	0.25	0.24	0.24	0.23	0.25	0.23	0.25
South and islands		0.35	0.21	0.32	0.17	0.34	0.16	0.34	0.18
Married		0.64	0.72	0.65	0.69	0.64	0.72	0.63	0.74
N° of household components		2.21 (1.02)	2.30 (1.00)	2.22 (1.01)	2.23 (0.95)	2.22 (1.02)	2.22 (0.94)	2.19 (1.01)	2.22 (0.89)
Long trip (>4 days)			0.50		0.43		0.52		0.49
Daily expenditure			80.87 (68.32)		74.11 (62.54)		74.60 (58.76)		84.12 (75.72)
Number of components			1.83 (0.71)		1.82 (0.68)		1.89 (0.73)		1.87 (0.69)
Travel time:		Jan–Mar	0.25		0.19		0.05		0.12
		Apr–Jun	0.25		0.12		0.21		0.27
		Jul–Sep	0.23		0.59		0.53		0.42
		Oct–Dec	0.26		0.1		0.21		0.19
Moving from–to:		North–North	0.35		0.38		0.39		0.35
		North–Centre	0.08		0.1		0.09		0.09
		North–South	0.11		0.11		0.11		0.12
		Centre–North	0.06		0.07		0.07		0.08
Centre–Centre			0.12		0.11		0.12		0.09
Centre–South			0.07		0.06		0.07		0.07
South–South			0.09		0.1		0.08		0.07
South–Centre			0.05		0.04		0.03		0.04
South–North			0.06		0.03		0.04		0.06
Motivations:		Holiday	0.39		0.5		0.53		0.49
		Rural	0.13		0.11		0.12		0.12
		Wellbeing	0.04		0.02		0.02		0.03
		VFR	0.31		0.29		0.26		0.29
		Culture	0.13		0.06		0.08		0.07
Booking accommodation:		Direct res.	0.21		0.35		0.35		0.37
		Agency reservation	0.2		0.07		0.08		0.10
		No reservation	0.59		0.58		0.57		0.52
Booking transport:		Direct res.	0.16		0.1		0.13		0.17
		Agency reservation	0.08		0.04		0.05		0.06
		No reservation	0.76		0.85		0.82		0.76
Means of transportation:		Airplane	0.07		0.04		0.07		0.08
		Train	0.11		0.08		0.07		0.11
		Own vehicle	0.65		0.74		0.73		0.68
		Other	0.17		0.13		0.13		0.12
Accommodation:		Similar hotel	0.07		0.10		0.09		0.09
		Hotel	0.32		0.28		0.29		0.34
		Rented	0.09		0.08		0.10		0.08
Non-rented			0.47		0.48		0.46		0.44
Campsite			0.04		0.05		0.06		0.05

Note: We include the standard deviations in parentheses, only for continuous values.

, the general characteristics of both sub-samples.

4. Results and Discussion

In light of a certain paucity of the studies, this research aimed at detecting the main factors that affect travel intention and LOS of senior Italian tourists and attempts to answer three research questions. In order to answer the research questions in

Table 2. Estimation results: probit model with sample selection for 2019–2022.

	Model A			Model B			Model C			Model D
	2019			2020			2021			2022
	Coef	p	Marginal Effects	Coef	p	Marginal Effects	Coef	p	Marginal Effects	Coef	p	Marginal Effects
SELECTION EQ.: Holiday
Constant	−2.09	***		−2.09	***		−2.10	***		−1.99	***
Woman (1 = yes; 0 = otherwise)	0.02		0.002	0.00		0.000	0.02		0.002	−0.03		−0.003
Age: 55–64 (ref.)
65–74	−0.09	*	−0.012	−0.07	#	−0.007	−0.04		−0.004	−0.11	**	−0.013
≥75	−0.23	***	−0.028	−0.37	***	−0.026	−0.24	***	−0.020	−0.30	***	−0.030
Worker (1 = yes; 0 = otherwise)	0.02		0.000	0.05		0.004	0.04		0.005	0.00		0.000
Education (years of schooling)	0.06	***	0.007	0.05	***	0.004	0.05	***	0.005	0.06	***	0.006
Area: North (ref.)
Centre	0.03		0.004	−0.14	***	−0.012	−0.11	**	−0.011	−0.10	**	−0.012
South and islands	−0.27	***	−0.029	−0.37	***	−0.027	−0.44	***	−0.035	−0.38	***	−0.037
Married (1 = yes; 0 = otherwise)	0.15	***	0.021	0.08	*	0.007	0.20	***	0.020	0.30	***	0.032
N household	−0.02		−0.004	−0.04	*	−0.005	−0.07	*	−0.008	−0.10	***	−0.011
OUTCOME EQ.: LOS
Constant	−2.58	***		−2.93	***		−2.14	***		−0.84
Woman (1 = yes; 0 = otherwise)	0.09		0.007	0.03		0.003	0.07		0.008	0.04		0.010
Education (years of schooling)	0.04	***	0.004	0.03	**	0.003	0.04	*	0.005	−0.01		−0.002
Younger (age < 65 = Yes; 0 = no)	−0.13	#	−0.011	−0.07		−0.006	−0.08		−0.011	−0.24	*	−0.061
Daily expenditure	−0.01	***	−0.001	−0.01	***	0.000	−0.01	**	−0.001	−0.01	***	−0.002
N participants	−0.10	*	−0.009	−0.06		−0.006	−0.09		−0.012	−0.13	*	−0.033
Travel time: ref. Jan–Mar
Apr–Jun	0.28	**	0.018	0.24	*	0.017	−0.17		−0.016	−0.24		−0.055
Jul–Sep	0.55	***	0.045	0.41	***	0.032	0.33		0.044	0.67	***	0.188
Oct–Dec	0.00		0.000	0.08		0.005	−0.34		−0.030	0.04		0.009
Moving from: North–North
North–Centre	0.00		0.000	0.37	***	0.033	0.27		0.034	0.43	**	0.108
North–South	0.51	***	0.044	0.90	***	0.080	0.85	*	0.105	1.13	***	0.287
Centre–North	0.09		0.008	0.14		0.013	0.23		0.029	0.01		0.002
Centre–Centre	−0.38	**	−0.033	−0.32	**	−0.029	−0.08		−0.010	−0.25		−0.064
Centre–South	0.23	*	0.020	−0.03		−0.003	0.14		0.017	0.62	***	0.157
South–South	−0.26	**	−0.022	−0.27	**	−0.024	−0.28	*	−0.035	0.09		0.022
South–Centre	−0.05		−0.004	−0.40	**	−0.036	−0.28		−0.035	0.20		0.051
South–North	0.14		0.012	0.44	*	0.040	0.05		0.007	0.95	***	0.242
Motivation (culture (ref.)
mot_leisure	0.57	***	0.049	0.59	***	0.053	0.45	*	0.056	1.18	***	0.300
mot_rural	0.45	**	0.038	0.46	**	0.041	0.23		0.029	0.87	***	0.222
mot_wellbeing	0.53	**	0.045	0.70	**	0.062	0.19		0.023	0.86	**	0.218
mot_relatives	0.31	**	0.026	0.38	*	0.034	0.02		0.002	0.72	***	0.183
Booking accommodation: ref. no reservation
Direct reservation	0.59	***	0.051	0.30	*	0.026	0.37		0.046	0.36	*	0.093
Agency reservation	0.32	**	0.027	0.41	*	0.037	0.38		0.048	0.19		0.049
Booking transport: ref. no reservation
Direct reservation	0.00		0.000	0.08		0.007	0.42	*	0.052	0.44	*	0.111
Agency reservation	0.23	#	0.020	0.20		0.018	0.69	*	0.085	0.52	*	0.132
Transport: other (ref.)
Airplane	0.08		0.007	0.03		0.002	−0.21		−0.026	−0.03		−0.006
Railway	−0.17		−0.015	−0.14		−0.012	−0.04		−0.004	−0.41	*	−0.103
Own car	−0.14		−0.012	0.05		0.005	0.10		0.012	0.17		0.043
Accommodation: ref (campsite)
Similar hotel	0.16		0.014	0.05		0.004	−0.41		−0.051	−0.68	**	−0.172
Hotel	0.07		0.006	0.27	*	0.024	−0.20		−0.024	−0.25		−0.065
Rented	0.35	*	0.030	0.51	**	0.046	0.32		0.040	0.15		0.038
Non-rented	0.26	#	0.023	0.23		0.020	−0.01		−0.001	−0.36		−0.093
MODEL SUMMARY
Wald χ2	200.52 (p. 0.0000)			143.65 (p. 0.0000)			156.99 (p. 0.0000)			248.66 (p. 0.0000)
Log likelihood	−4464.482			−4239.067			−5309.767			−6231.013
ρ	0.92			0.91			0.85			0.27
LR test of indep. eqns. (ρ = 0):	χ2: 5.93 (p. 0.0149)			χ2: 9.83 (p. 0.0017)			χ2 1.79 (p. 0.1810)			χ2 0.51 (p. 0.4758)

Note: p-value: # 0.10 ≥ p > 0.05; * 0.05 > p > 0.01; ** 0.01 ≥ p > 0.001; *** p ≤ 0.001.

, we use the estimation results from a probit model with sample selection for the years 2019 to 2022. The model includes two equations: one for the decision to go on holiday (Equation (1)) and the other for the length of the trip (LOS, Equation (2)). Table 2 not only provides the coefficient values and their associated significance levels, but also presents the calculated marginal effects.

In reference to the first research question, RQ1, that is how travel intentions and LOS have evolved among senior Italian tourists in the time periods preceding and following the COVID-19 pandemic, the data suggest certain trends in regard to travel intention. It is important to note that during 2019 (the pre-pandemic period), the model’s coefficients may have been influenced by factors unrelated to the pandemic and linked to the circumstance in which senior Italian travelers perceive travel constraints (not only linked to health) as more impactful than other age groups [76].

Indeed, we have seen a typical distribution of travel intentions among different demographic groups. The year 2020, characterized by lockdowns and heightened concerns about the pandemic, saw significant changes in travel intentions. Overall, travel intentions decreased, especially for seniors over 75 years of age, confirming what was established by Karl et al. [76]. This demonstrates how some factors, such as age and marital status, continued to influence travel intentions even during this challenging year. In 2021, when several restrictions and safety measures were still in place, but people were adapting to the “new normal”, we observed that certain factors like age and the number of family members became significant in determining travel intentions. These changes likely reflect evolving attitudes toward travel in response to the ongoing pandemic. In 2022, as the pandemic’s impact began to wane due to vaccination efforts and the reduction and end of restrictions, there may have been a resurgence in travel intentions. Factors like education level remained significant, indicating that socio-demographic characteristics continued to play a role in making travel decisions, as demonstrated by Karl et al. [76].

Overall, the results reveal how the COVID-19 pandemic and its various phases have influenced travel intentions over the 4-year period. While some factors, like age and education, consistently played a role, others, such as lockdowns and restrictions, had more pronounced effects during specific years. This comparative analysis provides valuable insights into the dynamic nature of travel behavior during and after a major global event, like a pandemic.

The results indicate that travel intentions increase for women (although this coefficient is not significant in all models). Additionally, travel intention is consistently negatively associated with age (with significant coefficients in all models). This relationship demonstrates a reduction in travel intentions, in line with the findings by Pappas [6] and Zheng et al. [37]. Furthermore, travel intention is positively correlated with education level, and the coefficients remain constant and highly significant over time, confirming the observations made by Li et al. [41] that travel-related fears vary based on socio-demographic status.

In comparison to the pre-pandemic period, travel intentions decrease for individuals residing in the center, south, and island regions of Italy (statistically significant in all models over time) when compared to those living in northern Italy. Moreover, we observe that the propensity increases for married individuals, while the number of family members negatively influences it (significant starting from 2020). Occupational status and gender, on the other hand, are not significant.

Regarding the intention to go on a longer vacation (more than 4 days), we observe that while travel intentions increase for “younger senior tourists”, the LOS increases for “older senior tourists”. In other words, travel intentions are higher for “younger senior tourists”, but once “older senior tourists” decide to go on holiday, they tend to stay for a longer duration. This finding aligns with the research by Almeida et al. [31], Losada et al. [46], and Zimmer et al. [43]. However, this effect is statistically significant only in the year 2022. This limitation hinders us from making robust assessments related to the pandemic period.

Gender is not statistically significant in predicting the length of stay in all years.

Education has a generally positive and significant effect on the LOS in most years, except for the last wave, where the coefficient turns negative and loses its significance. This shift in the coefficient’s sign may be attributed to a potential positive relationship between education and income.

Mergoupis and Steuer [77] suggest that “both the decision whether to take a holiday or not, and the decision on how much time to spend on a holiday, depend on the wage that can be commanded in the labor market”. In our study, it is reasonable to assume that individuals with higher levels of education may have earned, or are currently earning, more in the labor market compared to those with lower educational attainment. However, it is important to note that we cannot directly verify this assumption as the survey data does not include information on income.

Therefore, while education appears to influence vacation duration, the exact mechanism through which education and income are related in this context remains unverified due to the absence of income data in the surveys.

The number of family participants and daily expenditure have a negative impact on the LOS. This finding is consistent with the research by Alén et al. [2], Salmasi et al. [33], Alegre et al. [48], Barros and Machado [47], Alegre and Pou [65], and Boto-García et al. [66], and is associated with economic constraints.

However, while daily expenditure remains significant over time and exerts a consistent influence on the LOS, the number of family participants is only statistically significant in both the pre-pandemic year (2019) and the year following the easing of the pandemic-related restrictions (2022). This suggests that the impact of group size on the LOS is more pronounced in these specific years, possibly reflecting changing travel behaviors and economic considerations during different periods.

In the year before the pandemic (2019), this influence may have stemmed from factors unrelated to COVID-19. Travel patterns in 2019 were shaped by a range of economic and social variables, with group size potentially playing a substantial role in shaping the duration of vacations.

In 2022, even though the pandemic had not officially ended, the loosening and then ending of the restrictions marked a significant shift. The impact of these changes on travel behaviors and economic considerations could have persisted. People might have continued to prioritize smaller group travel, as stated by Boto-García et al. [66], or other pandemic-related factors, as demonstrated by Wen et al. [67], which could account for why group size remained a significant determinant of the LOS.

Overall, the persistence of the impact of group size on the LOS in these specific years suggests that it goes beyond the immediate effects of the pandemic and reflects broader changes in travel behaviors and economic conditions during those periods, as amply demonstrated by Alegre and Pou [65].

Looking more closely at the variables related to holiday planning, the LOS increases during the spring and summer. The significance and direction of the coefficients vary across different years, possibly due to factors such as seasonality, holiday destination, or changing travel behaviors, including the impact of the COVID-19 pandemic.

In particular, in some years, such as 2019 and 2020, the coefficient for the second quarter April–June is positive and statistically significant, indicating that, during the spring months, travelers tend to have longer stayed compared to the reference quarter. However, in 2021, there is a negative coefficient, suggesting that travel during this period (which includes the early months of the pandemic) is associated with shorter stays. In 2022, the coefficient is again negative, indicating that even after the pandemic, travelers during this period have shorter stays. In most years, such as 2019, 2020, and 2022, the coefficient for this quarter is positive and statistically significant, suggesting that the summer months are associated with longer stays compared to the reference quarter. Only in 2021, when several restrictions and safety measures were still in place, the coefficient is not significant. In 2022, it is again positive and statistically significant, indicating that even after the pandemic, travelers during this summer period have longer stays.

These considerations can also be extended to the quarter July–September, which shows, in all 4 years analyzed, a greater LOS, especially in the “pre” and “post” restriction years. Finally, in the fourth quarter (fall and early winter months), the direction of the coefficients varies across different years, but it is always insignificant.

Some changes occur between the two models with respect to the distance from the destination of origin (area of residence of the tourist) and the tourist destination. In general, the LOS increases when the distance between these two places is longer. Also, in this case, the meaning and direction of these coefficients is not constant over time. In particular, it is positive and significant over time, when the movement occurs from the north to the south and the islands and from the south/islands to the north, only in the years 2020 and 2022.

The first result can be traced back to a preference for seaside destinations [33], or for destinations with favorable climatic conditions, natural environments, landscapes, and natural resources [47,52], where senior tourists can enjoy outdoor activities safely [53].

The move from the south/islands to the north (Italy) represents something new with respect to the literature on the topic. Another novelty is represented by the shift from central to northern Italy, which, until the pre-pandemic period, was negative and not significant. These aspects are probably attributable to travel related to health recovery, since the most important medical facilities are in northern Italy, which could be in line with the findings by Tiago et al. [35] and Alén et al. [2].

Another justification could be linked to reuniting with loved ones. Indeed, Pearl et al. (2022) demonstrated the importance of VFR as a resilience factor in the face of crises (including the pandemic).

Another noteworthy novelty is linked to the movements by senior Italian tourists from central Italy to the south/islands. In 2019, the associated coefficient is positive and statistically significant. In 2020, however, this coefficient becomes negative and loses its statistical significance, only to return to having a positive and significant impact in 2022.

Furthermore, the LOS increases whether tourists book accommodation directly or through a travel agency. Comparing the situation over time, the coefficients maintain the positive trend (except in 2021) although less significant, confirming both the results by Javalgi et al. [42] and partially by Rahman et al. [21] about booking direct. Furthermore, booking with travel agencies could provide greater flexibility of booking dates and present lower perceived risk, contributing to the strengthening of the conclusions by Villacé-Molinero et al. [22].

When considering the booking of a transfer, we observe that this inclination grows stronger when tourists make the reservation. However, in this case, the coefficients only become statistically significant in the last two waves (2021 and 2022), whereas in 2020, only the coefficient associated with “agency reservation” reaches statistical significance (p < 0.10).

The LOS increases for tourists traveling by airplane, corroborating the results by Wang et al. [34], and for those who use their own car. In the latter case, the coefficient is positive only in 2020, and confirms the statements by Li et al. [41], who show a preference for traveling with one’s own car during the COVID-19 pandemic, and this trend persists even after the first year of the pandemic; however, the coefficients are not significant. The only significant coefficient is found in 2022: traveling by train negatively impacts the LOS, which confirms the findings by Li et al. [41] and Zillinger [69].

Our estimation results also reveal distinct patterns between 2019–2020 and 2021–2022. During the first two waves, there is an increase in the LOS for tourists who choose accommodation other than camping (the reference category), indicating a shift in travel behavior. However, in the latter two waves, the opposite trend is noticeable. For instance, the coefficient for “hotel” is positive in 2019 and 2020, but turns negative in the last two waves. Furthermore, it attains statistical significance only in 2020, coinciding with the peak of the pandemic crisis. Furthermore, the positive sign contradicts what has been established by Baños-Pino et al. [23]. However, during the last two waves, it becomes negative and loses its statistical significance. These findings indicate that seniors who select hotels or similar accommodation are less likely to prolong their vacations compared to those who opt for campsites, confirming the results by Salmasi et al. [33] and Alegre and Pou [65], who demonstrate that staying in a hotel reduces the LOS. This difference becomes more pronounced in the post-pandemic period. Similar results emerge when considering accommodation similar to hotels, with the sole significant coefficient being associated with the year 2022. The “rented” category displays a consistent positive trend over time, but is only significant in 2019 and 2020. Lastly, the “non-rented” category, which was weakly significant in 2019 and 2022, undergoes a shift in the direction of the association.

The second research question, RQ2, was addressed to explore whether there was a substantial rise in the preference for nature-based tourism among senior Italian travelers during the time periods preceding and following the COVID-19 pandemic.

To address this question, it is important to clarify that, in this paper, we are focusing on nature-based tourists who consider “nature-based” motivations as a central element of their holiday, while acknowledging the significance of other factors. As elucidated by Hyelin et al. [78,79], nature-based tourism typically encompasses various motivations, including a desire for novelty, relaxation, social interactions, and self-fulfillment. These diverse aspects have a direct impact on overall wellbeing.

From this perspective, it becomes clear that senior Italian tourists who choose rural resource-based vacations, prioritize leisure, wellbeing, or visiting relatives and/or friends, tend to opt for longer vacations. This pattern holds true across multiple time periods and becomes even more pronounced both during the pandemic year (confirming the results of Fergusson et al., [36]) and after the pandemic, except for the year 2021, where we observe a reduction in statistical significance, while consistent trends persist across all motivations except for “leisure”. In this context, the preference for nature-based tourism aligns with recent findings in the literature [53,56]. These results also align with emerging travel models that prioritize nature-based tourism over urban destinations, as evidenced by Hall et al. [24]. This shift presents an opportunity for more remote locations to thrive, while mitigating the negative effects of tourist overcrowding [32,80,81].

The third research question, RQ3, was aimed at evaluating whether there have been changes in tourist behavior over time, particularly in the periods before and after the pandemic. We noticed that before the pandemic there was a correlation between the decision to travel and the LOS, and this correlation persists even during the year of the lockdown. However, following this period, we observe a profound transformation not only in the main decision, whether to go on holiday or not, but also in regard to travel preferences and habits.

Regarding the regression diagnostics, the Wald test is highly significant, indicating a good model fit for all four models (χ² = 200.52, p value = 0.0000; χ² = 143.65, p value = 0.0000; χ² = 156.99; p value = 0.0000 and χ² = 248.66, p value = 0.0000, in model A, B, C, and D, respectively), while the likelihood ratio test indicates that we can reject the null hypothesis that the errors for the selection and outcome equations are uncorrelated (χ² = 5.93; p value = 0.015 and χ² = 9.83; p value = 0.002, in model A and B, respectively). While in model C and D the p value is greater than 0.05 (χ² = 5.93; p value = 0.015 and χ² = 9.83; p value = 0.002, in model A and B, respectively).

These results suggest a shift in the selection dynamics and factors influencing the LOS after the pandemic crisis. The lack of ρ significance during this period may indicate a change in dynamics. In the post-pandemic period, other factors may have become more influential in the decision to go on a long or short trip, regardless of the decision to go on holiday or not. This could be due to changes in traveler preferences, travel restrictions, or other circumstances.

In summary, the change in ρ significance after the pandemic crisis may reflect real variations in the relationships between the sample selection (to go on holiday or not) and the LOS. It is important to carefully examine the context and consider other factors that may have contributed to this variation in the selection dynamics and travel choices between the two periods.

5. Conclusions

This study has provided valuable insights into the dynamic nature of travel decision making and LOS choices in the context of pre-pandemic and post-pandemic periods. The change in correlation between these time frames sheds light on the changing selection dynamics and the factors influencing the trip duration.

In particular, the first result from this study shows a strong correlation between the decision to travel and the LOS before the pandemic. However, this interplay losses significance during the post-pandemic period, which indicates a shift in dynamics, where other factors have gained prominence in determining the trip duration, regardless of the decision to travel. This shift may be attributed to evolving traveler preferences, altered travel restrictions, or other contextual changes.

Our analysis of the vacation preferences of senior Italians travelers over the years 2019 to 2022, which encompassed the pre-pandemic, pandemic, and post-pandemic periods, has revealed several noteworthy trends. These trends shed light on the evolving dynamics of senior Italian tourist behavior in response to the unique circumstances brought about by the COVID-19 pandemic.

Firstly, we observed that Italian seniors with higher levels of education tended to extend their vacations in all years except 2022. This change in 2022 may be attributed to various factors, including shifts in economic conditions and travel motivations.

Secondly, the number of family participants and daily expenditure exerted a consistent negative influence on the length of senior Italian travelers’ vacations. However, the significance of the number of family participants varied, being significant only in the pre-pandemic year 2019 and 2022. This suggests that the impact of group size on vacation duration was more pronounced during these specific years, possibly reflecting changing travel behaviors influenced by the pandemic.

When considering travel times and regions of origin, we noticed fluctuations in the coefficients between the years. These variations could be linked to shifts in travel patterns and preferences influenced by external factors, such as pandemic-related restrictions and economic changes.

In terms of the motivation for travel, Italian seniors motivated by rural experiences, leisure, wellbeing, or VFR, were more likely to take longer vacations, with these effects being more pronounced during the pandemic period. These findings align with emerging travel models emphasizing nature-based tourism and could offer opportunities for less crowded destinations.

Regarding booking preferences, our analysis indicated that booking accommodation and transfers significantly impacted the vacation duration, with variations in significance across the years, possibly reflecting changing travel restrictions and safety concerns.

Finally, the choice of accommodation played a crucial role in Italian seniors’ vacation durations. Notably, the preference for hotel stays shifted over time, with hotels being positively associated with longer vacations in 2019 and 2020, but negatively associated in 2021 and 2022. This suggests evolving travel behavior and economic considerations among Italian seniors in response to the pandemic and its aftermath.

This study is also important from a policymakers’ and tourism managers’ perspective for helping to identify critical factors that influence the senior Italian tourists’ LOS.

5.1. Theoretical Implications

This study explored the impact of the COVID-19 pandemic on the travel intentions of senior Italian tourists (their decision-making process toward the decision to go on holiday or not) and evaluated the effect of the COVID-19 pandemic on their LOS. The study findings investigated this premise and determined that there has been a decrease in the travel intentions of senior Italian tourists, which increases with the age.

In particular, in analyzing the intention to embark on longer vacations (lasting more than 4 days) among senior tourists, a noteworthy trend emerges when considering age differentials. Specifically, our findings reveal that travel intentions exhibit an upward trajectory among what we term “younger senior tourists”. This subgroup, comprised of relatively younger individuals within the senior demographic, demonstrates a heightened inclination or willingness to undertake extended vacations. Intriguingly, the LOS takes a distinctive turn when examining “older senior tourists”. Although travel intentions may not be as pronounced for this group, once the decision to go on holiday is made, these individuals exhibit a proclivity for a prolonged duration of stay. In essence, our results suggest that while travel intentions are higher for “younger senior tourists”, the subsequent actions of “older senior tourists” are characterized by an inclination toward extended stays; thereby, contributing valuable insights into the nuanced dynamics of travel preferences among different age segments within the senior tourist demographic.

This implies several theoretical implications, in particular, the current study has confirmed the importance of socio-demographics and trip characteristics in explaining the LOS, but distinctively it contributes to the extant tourism literature in six main ways.

Firstly, the result that Italian seniors with higher levels of education tend to extend their vacations, except in 2022, suggests a dynamic relationship between education, economic conditions, and travel motivations. This finding could lead to further exploration of how education influences senior Italians’ decision-making processes regarding travel, and how external factors impact these decisions.

Secondly, the negative influence of the number of family participants and daily expenditure on the LOS, with varying significance across the years studied, highlights the complex interplay between group dynamics, economic considerations, and travel behavior. This implies that the impact of group size on the LOS is not constant and may change in response to external factors like the pandemic.

Thirdly, fluctuations in the coefficients related to travel times and regions of origin across different years suggest that external factors, such as pandemic-related restrictions and economic changes, can significantly influence Italian seniors’ travel patterns and preferences. This underscores the need for adaptable travel models that can account for changing circumstances.

Fourthly, the influence of motivations for travel on the LOS, particularly the preference for rural experiences, leisure, wellbeing, and VFR during the pandemic period, aligns with emerging travel models emphasizing nature-based tourism. This finding implies that Italian seniors may increasingly seek less crowded destinations and nature-based experiences, which could have implications for tourism industry strategies.

Fifthly, the impact of booking accommodation and transfers on the LOS, with variations in significance over the years, suggests that changing travel restrictions and safety concerns can influence Italian seniors’ decisions regarding how and where to book their vacations. This highlights the importance of understanding how external factors affect booking preferences.

Sixth, the shifting preference for hotel stays and its association with vacation duration in different ways, across the years, reflects evolving travel behavior and economic considerations among Italian seniors in response to the pandemic and its aftermath. This finding indicates that the choice of accommodation is not static and may be influenced by external factors.

5.2. Managerial Implications

The results of this research can offer several managerial implications linked to the development of the domestic tourist market for senior Italian travelers. Indeed, new products and services could be actively developed to address the needs of this target group, as well as the formulation of promotions and discounts, specially to attract those who are very price sensitive, or who have suffered from economic constraints following the pandemic crisis. Indeed, COVID-19 has provided the tourism sector with opportunities to adapt, innovate, and exploit new avenues of operation. From this perspective, travel insurance could guarantee coverage in the event of illness, including post-coronavirus, when booking a holiday. At the same time, the tourism industry should emphasize safety and health measures, and any tourism activities that make tourists feel safer and reduces travel risks and perceptions.

The preference for nature-based tourism and rural tourism may translate into an opportunity for more remote places to thrive, while reducing the effects of tourist overcrowding. From this perspective, the tourism industry must rethink and restore the direction toward more inclusive, sustainable, and responsible development for the future, to guarantee the opportunity for both national and international senior tourism to enjoy longer stays.

Furthermore, the insights derived from our analysis hold significant managerial implications for the tourism industry, particularly concerning the distinctive preferences of “younger senior tourists” and “older senior tourists”.

For the cohort of “younger senior tourists”, characterized by a heightened inclination for extended vacations, strategic marketing efforts should be tailored to underscore the allure of longer-stay options. Advertisements and promotional materials can emphasize the appeal of extended leisure activities, diverse experiences, and packages curated to align with the preferences of this demographic.

Conversely, for “older senior tourists”, whose travel intentions may not be as pronounced, but tend to result in extended stays, the focus should shift towards promoting the benefits of longer-stay durations. Tourism operators and managers can highlight the enriching aspects of extended vacations, such as opportunities for relaxation, cultural immersion, and in-depth exploration of the destination. Customized travel packages that cater specifically to the interests and needs of “older senior tourists” can be developed.

In crafting overall strategies, it is imperative to gain a deeper understanding of the motivations behind the travel decisions of both groups. Conducting market research and surveys can provide valuable insights into the specific interests, concerns, and preferences of “younger senior” and “older senior” tourists, allowing for more targeted and effective marketing strategies.

Ultimately, personalized services based on age-specific preferences, along with the development of specialized offerings for each senior group, can enhance the overall tourist experience. Recognizing and capitalizing on the nuanced travel behaviors of both “younger senior” and “older senior” tourists could contribute to the success and competitiveness of tourism offerings in the industry.

5.3. Limitations and Future Research Agenda

A limitation of this research is the impossibility to apply and test theory or models related to crisis behavior aspects. This study cannot assess whether senior Italian tourists’ decision making is subject to risk perceptions for instance. Hence, this study could not integrate theory, such as the health belief model (HBM) or protection motivation theory (PMT). Furthermore, this paper focuses only on senior tourism, comparisons with different cohorts could have led to different results. Furthermore, this study was conducted in Italy and analyzed senior Italian tourists; different and equally interesting results could have been obtained by comparing different countries or comparing incoming and outgoing tourism. Furthermore, the results revealed a change in the selection dynamics and factors that influence the LOS between the two time phases (pre-pandemic and post-pandemic years). In particular, the change in ρ significance between the two periods may reflect real variations in the relationships between sample selection (to go on holiday or not) and LOS. Future research could carefully examine the context and consider other factors that may have contributed to this variation in the selection dynamics and travel choices between the two periods.

In particular, the future research agenda could be addressed in the direction of these limitations by proposing different data analysis models or different methodology, comparing different categories of tourists, or using other variables and other predictors.