Tourist Attitudes to the COVID-19 Pandemic and Their Influence on Sustainable Tourism Behaviour: Evidence from Cáceres, a UNESCO World Heritage City

Abstract

Research on post-COVID tourism behaviour has expanded rapidly, yet inland UNESCO World Heritage cities remain underexamined, particularly in Mediterranean contexts. This study examines whether the pandemic produced durable changes in tourist behaviour and in willingness to pay for sustainable services in Cáceres, Spain. A structured face-to-face survey was administered to 421 visitors in March 2023, after public-health restrictions had been lifted. The analysis covered self-reported behavioural change, perceived impacts on different destination types, perceived effects on local sustainability objectives and changes in willingness to pay (WTP) for sustainable services. Descriptive statistics were complemented by an exploratory binary logistic regression predicting increased WTP. Because the model includes only sociodemographic predictors and shows modest fit, it is used to describe associations rather than to predict. Reported behavioural change was limited: mean scores for crowd avoidance, health–safety preferences, shorter stays and substitution towards rural and nature tourism ranged from 1.73 to 1.91 on a five-point scale. Respondents nevertheless perceived substantial spatial effects of the pandemic, particularly on natural parks (92.6%) and rural destinations (84.1%). Most believed that the pandemic had accelerated sustainability efforts mainly through greater institutional and business awareness (54.9%). WTP proved relatively stable, with 62.7% reporting no change and 26.1% an increase. Women and respondents with university education showed higher odds of reporting increased WTP. Because constructs such as institutional trust and pro-environmental values were not measured directly, these attitudes are interpreted—rather than demonstrated—as reflecting governance-related confidence and value orientations more than lingering health concerns. This governance-and-values reading is the study’s main interpretive contribution and requires confirmation with direct measures of the underlying constructs.

1. Introduction

The COVID-19 pandemic was the deepest disruption in the recent history of tourism. Travel bans, social-distancing rules and capacity restrictions abruptly interrupted mobility and exposed the systemic vulnerability of the sector (Nhamo et al., 2020; Škare et al., 2021). Beyond the short-term economic shock, the crisis was quickly framed as a potential turning point for sustainability, justice and destination governance: it temporarily slowed overtourism and forced destinations to reconsider the social and environmental foundations of recovery (Higgins-Desbiolles, 2020; Sigala, 2020; Zenker & Kock, 2020).

As restrictions were progressively lifted, a central question emerged: did tourists internalise durable behavioural changes, or did they rapidly return to pre-pandemic routines? Early studies documented higher risk perception, greater interest in low-density spaces and a preference for nearby or nature-based tourism (Bae & Chang, 2021; Z. Li et al., 2020). Later evidence, however, suggested a progressive re-normalisation of travel behaviour once sanitary confidence improved, especially in European domestic markets (Gössling & Schweiggart, 2022; Pappas, 2021). At the same time, scholars warned that support for sustainability might weaken under post-pandemic inflation and economic uncertainty, widening the well-known attitude–behaviour gap in tourism (Juvan & Dolnicar, 2014; Pulido-Fernández & López-Sánchez, 2016).

These debates are especially relevant for medium-sized cultural destinations, where heritage value, urban form and governance capacity shape recovery trajectories differently from coastal or rural destinations. Cáceres is one of Spain’s UNESCO World Heritage Cities and an inland destination strongly dependent on cultural, monumental and gastronomic tourism (Garrod & Fyall, 2000). Its scale makes it suitable for exploring destination resilience (Alfaro-Navarro & Andrés-Martínez, 2023), because visitor concentration, walkability and public management can be observed more clearly than in metropolitan tourism systems (Biggs, 2011; Cheer & Lew, 2017). Inland medium-sized heritage cities are analytically distinct because their recovery depends less on volume growth than on the management of a finite historic core. In such a core, flows, walkability and conservation pressures concentrate in a small, highly visible area, which makes the link between visitor demand and public management easier to observe than in large metropolitan or dispersed coastal systems. The tourism profile of Cáceres supports this case selection. It is the most visited destination in the Spanish region of Extremadura, with an annual volume of around 305,000 to 342,000 tourists between 2016 and 2018 and a city population below 100,000; the average monthly tourist-to-resident ratio remained between roughly 26% and 30%, indicating an absence of overtourism (Jurado-Rivas & Sánchez-Rivero, 2019). Crucially for the interpretation of post-pandemic attitudes, Cáceres is a predominantly domestic destination: in the years before the pandemic, foreign visitors represented less than 20% of total arrivals (Jurado-Rivas & Sánchez-Rivero, 2019), a structural feature consistent with the heavily domestic composition of the present sample.

The case of Cáceres contributes to the literature in four specific ways:

• It focuses on an inland UNESCO World Heritage city in the Iberian Peninsula, namely a medium-sized cultural destination with a predominantly domestic visitor profile.
• Fieldwork was conducted in March 2023, after public-health restrictions had been lifted, making it possible to observe post-pandemic attitudes beyond the immediate effects of regulation.
• The analysis is based on 421 face-to-face surveys, with a sampling error of ±4.8% at the 95% confidence level under a simple random sampling assumption.
• It places willingness to pay (WTP) for sustainable services at the core of the analysis in order to identify which visitor profiles are most likely to accept sustainability-related price premiums.

On this basis, the article pursues four objectives: (O1) to measure the intensity of self-perceived behavioural change attributable to the pandemic; (O2) to identify which destination types tourists perceive as most affected by COVID-19; (O3) to analyse how tourists think the pandemic influenced sustainability objectives; and (O4) to evaluate whether the pandemic altered tourists’ WTP for more sustainable services.

Beyond its empirical focus, the study seeks to clarify whether post-pandemic support for sustainability in heritage tourism should be read primarily as a residual effect of health-related risk aversion or, instead, as an expression of broader value orientations and confidence in destination governance. This distinction matters for how heritage cities design post-pandemic management strategies. If sustainable preferences are mainly fear-driven, they may fade quickly once normality returns; if they are linked to trust, values and perceived institutional credibility, they may provide a more durable basis for destination transformation.

The remainder of the article is organised as follows. Section 2 develops the theoretical framework and presents the hypotheses; Section 3 explains the methodology; Section 4 reports the results; Section 5 discusses the findings; Section 6 presents the conclusions and policy implications; and Section 7 outlines the main limitations and avenues for future research.

2. Theoretical Framework

2.1. Pandemic Shock, Resilience, and Heritage-City Recovery

The pandemic triggered a collapse in international tourism that was unprecedented in scale and duration, with 2020 marking a systemic break in mobility patterns (UNWTO, 2021). This disruption renewed the relevance of resilience thinking in tourism research. Rather than treating resilience as mere recovery to the status quo, recent scholarship understands it as the capacity of destinations to adapt, reconfigure governance arrangements and, where necessary, move towards a more sustainable trajectory (Fotiadis et al., 2021; Hall et al., 2020; Ritchie, 2004). For heritage cities, resilience depends not only on visitor demand but also on the capacity to manage public space, regulate flows and preserve cultural capital while remaining economically viable (Garrod & Fyall, 2000; Mandić et al., 2025). This reading is consistent with early assessments of COVID-19 as a systemic shock capable of reshaping tourism governance and destination development pathways (Gössling et al., 2020).

This perspective is especially relevant in inland heritage destinations, whose recovery has received less attention than that of coastal or nature-based tourism. In such places, resilience tends to manifest less through volume growth than through stability, quality of visitation and the institutional capacity to align conservation with economic activity (Cheer & Lew, 2017). Research on World Heritage Sites emphasises the particular resilience of cultural tourism destinations (Wei & Chen, 2025): symbolic value, conservation frameworks and established demand can accelerate recovery once mobility restrictions are lifted. The first question, therefore, is whether post-pandemic tourists still display behaviours associated with health fear, or whether heritage-city tourism has already returned to its pre-crisis patterns.

2.2. Post-Pandemic Behavioural Change: Risk Perception and a TPB Lens

Research on crises and tourism consistently shows that perceived risk shapes destination choice, trip planning and avoidance behaviour (Sönmez & Graefe, 1998; Quintal et al., 2010). During the acute phase of COVID-19, tourists were expected to prefer open-air settings, lower density, shorter stays and touchless services (Bae & Chang, 2021; Z. Li et al., 2020; Luo & Lam, 2020). The Theory of Planned Behaviour (TPB) helps to organise these expectations, since it links attitudes, subjective norms and perceived behavioural control to behavioural intentions (Ajzen, 1991). In a pandemic context, stronger beliefs about contagion and greater social approval of cautious travel should raise the intention to avoid crowded destinations or to demand additional health guarantees.

Yet the TPB also helps to explain why these adaptations may weaken over time. Once the perceived probability and consequences of contagion decline, attitudes and social norms become less restrictive, and hedonic and cultural motivations regain salience. Evidence from China, Hong Kong and South Korea points to a stronger persistence of risk-sensitive behaviour under prolonged restrictions (Bae & Chang, 2021; Luo & Lam, 2020; Wen et al., 2021). In European domestic markets, by contrast, the reactivation of mobility often coincided with rapid behavioural normalisation (Gössling & Schweiggart, 2022; Wilkinson & Coles, 2024). In the Spanish context, post-COVID debates on urban cultural tourism have likewise stressed the need to rethink city-tourism models beyond simple demand recovery (Martín-Azami & Nogués-Pedregal, 2021).

2.3. Sustainability, Willingness to Pay, and Pro-Environmental Commitment: VBN, Governance, and Hypotheses

The sustainability debate after COVID-19 developed along two competing lines. One body of work read the crisis as an opportunity to rethink tourism around ecological justice, slower mobilities and reduced dependence on mass tourism (Brouder, 2020; Higgins-Desbiolles, 2020; Ioannides & Gyimóthy, 2020; Stankov et al., 2020). A second, more cautious strand argued that economic stress could undermine sustainable consumption, particularly where sustainability implies price premiums (Pulido-Fernández & López-Sánchez, 2016). WTP is a useful indicator in this respect, because it connects declared attitudes to economic commitment. Recent tourism research has further shown that WTP depends on segmentation, governance credibility and the perceived transparency of sustainability claims (Ezeh & Dube, 2024; Tölkes, 2018).

To interpret this relationship, the Value-Belief-Norm (VBN) model is especially useful (Van Riper & Kyle, 2014). VBN holds that pro-environmental behaviour is activated when values shape ecological beliefs and personal norms, producing a felt moral obligation to act (Stern et al., 1999; Wynveen et al., 2015; Liu et al., 2018; Jaratmetakul & Pommi, 2026). In tourism and hospitality settings, VBN has been combined with the TPB to explain why some travellers adopt green choices even when these require effort or additional expenditure (Han, 2015). For post-pandemic heritage tourism, this implies that sustainable behaviour is unlikely to stem from health fear alone; it emerges instead from the interaction between personal norms, destination governance, market trust and the concrete design of sustainability offers. Research in Asian contexts has documented similar resilience patterns, showing that active policy frameworks are key drivers of sustainable tourism behaviour after COVID-19 (Srisawat et al., 2023).

Despite these advances, three gaps remain. First, much of the post-pandemic evidence has focused either on broad recovery trends or on nature-based and coastal destinations, leaving inland heritage cities comparatively underexamined. Second, studies on willingness to pay for sustainability have often treated pro-environmental commitment as a relatively stable consumer disposition, without considering how pandemic-related disruption may have altered the relationship between perceived risk, trust in governance and economic support for sustainability. Third, although the post-COVID debate has produced important normative reflections on destination transformation, fewer empirical studies have examined how visitors themselves interpret the links between pandemic experience, sustainability objectives and destination management in heritage settings. By addressing these gaps, the present study links post-pandemic behavioural normalisation to a governance-and-values interpretation of sustainable tourism attitudes in an inland UNESCO World Heritage city. Resilience thinking, the TPB and the VBN model can be read as a single sequence rather than as three separate perspectives. Resilience defines the destination-level context in which heritage cities adapt after a shock; the TPB explains why risk-driven adaptations weaken as perceived contagion declines; and VBN explains why support for sustainability may nonetheless persist, through values and personal norms that operate independently of health fear. Within this framework, governance translates these individual dispositions into trusted and visible destination practices. It must be stated explicitly, however, that the framework is used here as an interpretive lens, not as a measurement model. The core TPB constructs (attitudes, subjective norms and perceived behavioural control) and the core VBN constructs (biospheric values, ecological beliefs, personal norms and environmental self-identity), together with institutional trust and governance credibility, were not operationalised in the questionnaire. The hypotheses below are therefore framed as descriptive, exploratory expectations about observable behaviour and stated attitudes, and the theoretical constructs are mobilised only to interpret—not to test—the resulting patterns.

On the basis of this theoretical framework, four hypotheses are proposed:

• H1: The self-perceived impact of the pandemic on current tourist behaviour is low, indicating a return to pre-pandemic preferences among heritage tourists.
• H2: Tourists perceive that natural and rural spaces were the destination types most structurally altered by post-pandemic flows.
• H3: Tourists perceive that the pandemic accelerated sustainability objectives mainly through greater institutional and corporate awareness.
• H4: WTP for sustainable services remained resilient after the pandemic and was not weakened by the recovery context.

3. Methodology

3.1. Research Design and Case Study

This study adopts a quantitative, cross-sectional design based on a survey of visitors to Cáceres, Spain. The destination was selected because it is a consolidated inland heritage city with intense cultural use, manageable visitor concentrations and an explicit local discourse linking heritage management to sustainability. Data were collected through structured face-to-face questionnaires using paper-assisted personal interviewing (PAPI). This article concentrates on Block VI of the instrument, labelled “Tourist attitude in pandemic situations”, and uses the sociodemographic information from Block I to model the predictors of increased WTP.

3.2. Why March 2023? Fieldwork Timing, Intercept Sampling, and Bias Control

March 2023 was selected for three methodological reasons. First, it fell within a period in which public-health restrictions had been fully lifted: mask mandates, COVID passports and capacity limits had all disappeared, so the study could measure durable effects without contamination from still-active regulations. Second, the period coincided with a demanding macroeconomic context in Spain, with post-pandemic inflation above 6%, which provided a stringent scenario for testing whether WTP for sustainability remained resilient. Third, tourism flows in Cáceres had largely recovered to levels comparable with those observed before the pandemic (Jurado-Rivas & Sánchez-Rivero, 2019, 2022), making it possible to assess attitudes once emergency conditions had subsided.

Systematic intercept sampling was carried out at three high-flow points in the historic centre: Plaza Mayor (35% of completed questionnaires), Museo de Cáceres (40%) and Arco de la Estrella (25%). To reduce temporal bias, interviews were spread between 10:00 and 19:00 and balanced across weekdays (40%) and weekends (60%), reflecting the different composition of urban visitors across the week. Interviewers rotated across points and time slots, avoided approaching large, organised groups, and recorded the time and location of each questionnaire to monitor coverage. About 15% of the visitors approached declined to participate; these cases were recorded but excluded from the final sample. After data cleaning, the final valid sample comprised 421 questionnaires.

Under a simple random sampling assumption, a sample of this size would correspond to an approximate margin of error of ±4.8% at the 95% confidence level. Because the study relied on systematic intercept sampling rather than a strict probability design, this figure should be read only as a rough indicator of sample-size adequacy. Three sources of bias should therefore be kept in mind when reading the results. First, intercept sampling at high-flow points, combined with a refusal rate of roughly 15%, introduces potential self-selection and coverage bias, so the sample cannot be treated as strictly representative of all visitors. Second, the realised sample is heavily skewed towards domestic (94.3%) and university-educated (52.7%) visitors; although the strong domestic share is consistent with the destination’s structural profile (foreign arrivals below 20% before the pandemic; Jurado-Rivas & Sánchez-Rivero, 2019), the educational profile is comparatively high and may inflate stated support for sustainability. Third, because P14 and P17 are self-reported attitudinal measures, they are exposed to social-desirability bias, particularly for sustainability-related willingness to pay; the analyses below are interpreted with this caution in mind.

3.3. Measurement Instrument and Coding

Block VI comprised four analytical components (P14–P17), as summarised in

Table 1. Summary of the measurement instrument used in Block VI (P14–P17).

Block	Items	Scale	Coding	Reliability
P14	Five items: uncrowded sites; stricter health measures; shorter stay; COVID passport; shift to rural/nature tourism	Likert scale (1–5)	Mean score by item	Cronbach’s alpha = 0.82
P15	Four destination types: coastal, rural, World Heritage Cities, and Natural Parks	Multiple response	Percentage of cases	_ *
P16	Perceived effect of the pandemic on sustainability objectives in Cáceres	Single response (4 options)	Frequency and percentage	_
P17 **	Self-reported change in willingness to pay for sustainable services	Trichotomous response	Descriptive table; binary recoding for logit (1 = more willing, 0 = otherwise)	_

* Indicates that internal consistency reliability was not applicable because the variable was measured using a single item or a multiple-response question. ** P17 was recoded into a binary dependent variable for the logistic model: 1 = “more willing to pay”; 0 = all other valid responses. Source: Own elaboration.

. P14 measured self-perceived behavioural change through five Likert items ranging from 1 (strongly disagree) to 5 (strongly agree); the internal consistency of this battery was acceptable (Cronbach’s alpha = 0.82). P15 asked respondents to identify the destination types perceived as most affected by pandemic-related changes in tourist flows. P16 used a single-choice item to capture expectations about the effect of the pandemic on sustainability objectives. P17 measured the self-reported change in WTP for sustainable tourism services through a trichotomous response scale (same, more willing, less willing). The instrument builds on the survey programme that the authors have applied to the same destination in earlier work. The sociodemographic items (Block I) and the overall questionnaire architecture replicate those used in previous studies of willingness to pay for sustainability in Cáceres (Jurado-Rivas & Sánchez-Rivero, 2019, 2022), which ensures continuity and comparability of the measurement framework. Block VI, by contrast, was developed specifically for this study to capture pandemic-related attitudes; its items were drawn from the post-COVID tourism literature reviewed in Section 2 and refined by the research team. Two measurement choices should be made explicit. First, internal consistency was assessed for the P14 battery (Cronbach’s alpha = 0.82), but the block was not subjected to formal pilot testing or confirmatory factor analysis; P15–P17 are single-item or multiple-response measures, for which dimensionality testing does not apply. These points are revisited in the Limitations. Second, unlike the authors’ prior contingent-valuation work, which elicited actual monetary premiums for specific services—a hotel room, a restaurant, a museum ticket and a taxi ride (Jurado-Rivas & Sánchez-Rivero, 2019, 2022)—P17 captures only a coarser, self-reported directional change in willingness to pay (more, the same or less). It is therefore an attitudinal indicator of economic support for sustainability rather than an estimate of any specific amount, and the results should be read accordingly.

3.4. Analytical Strategy

The analysis combines descriptive statistics with an exploratory binary logistic regression. Descriptive indicators were used to assess mean scores, frequency distributions and multiple responses. To identify the sociodemographic predictors of increased WTP, a binary logit model was estimated by maximum likelihood. The dependent variable took the value 1 when respondents declared that they were more willing to pay after the pandemic and 0 otherwise. The predictors were sex, age, education and origin, together with a sex × education interaction term (the full coding and recoding of these variables is provided in Appendix A,

Table A1. Coding and recoding of independent variables for the logistic model.

Variable	Code	Category	Use in the Model
SEX	1	Woman	sex_woman = 1
	2	Man	Reference (sex_woman = 0)
AGE	1	18–25 years	Joint reference (<40 years)
	2	26–40 years	Joint reference (<40 years)
	3	41–55 years	Dummy: age_41_55
	4	56–65 years	Dummy: age_over55
	5	>66 years	Dummy: age_over55
	99	DK/NA	Excluded (1 case)
EDU	1	No formal education	Reference (non-university)
	2	Primary education	Reference (non-university)
	3	Secondary education	Reference (non-university)
	4	Higher education/University	edu_univ = 1
	99	DK/NA	Excluded (8 cases)
ORIGIN	1	Spain	origin_nac = 1
	2	Foreign	Reference (origin_nac = 0)

Final valid n for the logistic regression = 398 after excluding 1 missing value in AGE and 8 missing values in EDU. Source: Own elaboration.

).

Given the modest number of predictors and the absence of psychographic variables, the model is treated as exploratory and descriptive rather than predictive in a strong classificatory sense. It achieved an area under the ROC curve (AUC) of 0.653 (95% CI: 0.591–0.715), indicating modest discriminatory ability. The model is therefore used to identify broad associations in the sample rather than to generate robust individual-level predictions.

4. Results

4.1. Sociodemographic Profile of the Sample

As shown in

Table 2. Sociodemographic profile of respondents (n = 421).

Variable	Category	Frequency	%
Sex	Women	232	55.1
	Men	189	44.9
Age *	18–25 years	38	9.0
	26–40 years	121	28.7
	41–55 years	161	38.2
	56–65 years	74	17.6
	>66 years	26	6.2
Education	No formal education	13	3.1
	Primary education	57	13.5
	Secondary education	121	28.7
	University education	222	52.7
Origin	National	397	94.3
	International	24	5.7

Source: Own elaboration. * One respondent did not report age (n = 420 for this variable).

, the sample is balanced by sex, with a slight female majority (55.1%). It is also comparatively mature and highly educated: the largest age group is 41–55 years (38.2%), and university education accounts for 52.7% of the total sample (n = 421) and 53.7% of valid responses (n = 413). The profile of visitors is overwhelmingly domestic (94.3%), which matters when interpreting the low persistence of sanitary concern in the post-pandemic period.

4.2. P14: Changes in Tourist Behaviour Due to COVID-19

Self-reported behavioural change was very limited across the P14 items (

Table 3. Intensity of self-reported post-COVID changes in tourist behaviour (1 = strongly disagree; 5 = strongly agree).

Code	Evaluated Behavioural Factor	Mean	SD	Valid n
P14-2	Visit sites that are not crowded, even within a World Heritage city	1.91	1.09	416
P14-1	Prefer sites that fully comply with health–safety measures even if incidence is low	1.85	1.02	416
P14-4	Identify tourism services that require a COVID passport	1.77	1.10	414
P14-3	Shorten the duration of the visit	1.74	1.06	414
P14-5	Replace cultural tourism with rural or nature tourism	1.73	1.07	416

Source: Own elaboration.

). Mean scores range from 1.73 to 1.91 on a five-point scale, which indicates that visitors do not see the pandemic as a decisive factor in their current behaviour. The highest mean corresponds to a preference for visiting less crowded sites (1.91), yet even this value remains clearly below the midpoint. The pattern is therefore one of rapid normalisation rather than persistent avoidance.

The full distribution confirms this interpretation. For P14-1 (“Prefer sites that fully comply with health–safety measures even if incidence is low”), 45.1% of respondents strongly disagreed and 34.0% disagreed that the pandemic had changed their requirements, whereas only 3.1% agreed and 3.8% strongly agreed. By spring 2023, then, residual health-related caution appeared limited and insufficient to shape tourist behaviour in any systematic way.

4.3. P15: Destinations Perceived as Most Affected by the Pandemic

The multiple-response question on destination types points to a clear spatial narrative of post-pandemic change (

Table 4. Destinations perceived as most affected by pandemic-driven flow changes (multiple response, n = 417) **.

Destination Type	Frequency	% of Cases *
Natural Parks	390	92.6
Rural Destinations	354	84.1
World Heritage Cities	326	77.4
Coastal Destinations	311	73.9

* Percentages refer to cases; frequencies exceed n because respondents could select more than one destination type. Source: Own elaboration. ** Four respondents did not answer this question (valid n = 417).

). Natural Parks (92.6%) and Rural Destinations (84.1%) were perceived as the most affected by demand shifts, which suggests a strong association between the pandemic and the search for low-density, open-air environments. World Heritage Cities also recorded a high proportion (77.4%), though lower than the nature-oriented categories, pointing to a perception of relative resilience rather than immunity.

4.4. P16: Perceived Effect of the Pandemic on Sustainability Objectives

As shown in

Table 5. Perceived effect of the pandemic on sustainable tourism objectives in Cáceres (single response, n = 419).

Response Option	Frequency	%
It will make the objectives be met sooner because companies and public administrations will be more aware of sustainability (P16-1)	230	54.9
It will not influence whether the city’s sustainability objectives are met sooner or later (P16-4)	96	22.9
It will make the objectives be met later due to the paralysis or slowdown of tourist activity (P16-3)	57	13.6
It will make the objectives be met sooner because tourists will be more aware of sustainability (P16-2)	36	8.6

Source: Own elaboration.

, the modal response was optimistic but clearly institutional in orientation: 54.9% of respondents believed that the pandemic had accelerated sustainability objectives mainly because firms and public administrations had become more aware of the issue. Only 8.6% attributed faster progress to stronger environmental awareness among tourists themselves. This pattern suggests that respondents associated progress on sustainability primarily with governance and business action rather than with a substantial shift in tourists’ own environmental values.

4.5. P17: Change in Willingness to Pay for Sustainable Services

P17 offers the most direct indication of relatively resilient support for sustainability (

Table 6. Self-reported post-pandemic changes in willingness to pay (WTP) for sustainable services (n = 421).

Response/Economic Attitude	Frequency	Valid %	Total %
No, I would pay the same (maintain WTP)	255	62.7	60.6
Yes, more willing to pay (increase WTP)	110	27.0	26.1
Yes, less willing to pay (decrease WTP)	42	10.3	10.0
Do not know/no answer	14	-	3.3

Source: Own elaboration.

). Because it records self-reported attitudes rather than observed purchasing behaviour, it should be read as an attitudinal signal and may be affected by social-desirability bias. Almost two thirds of respondents maintained their pre-pandemic WTP, and more than one quarter declared a higher WTP despite the inflationary context, while only one tenth reported a lower WTP. Nearly nine out of ten valid responses therefore indicate either stability or an increase in stated willingness to pay for sustainable services.

Importantly, this result refers to stated rather than observed behaviour and should therefore be interpreted as an attitudinal indicator of economic support for sustainability rather than as direct evidence of actual purchasing behaviour.

4.6. Binary Logistic Regression: Predictors of Increased WTP

The exploratory binary logit model confirms that the most relevant predictors of increased WTP are female sex and university education (

Table 7. Binary logistic regression predicting increased post-pandemic WTP (valid n = 398).

Predictor	B	OR	95% CI	p
Sex (woman = 1)	0.920	2.51	[1.54, 4.08]	<0.001
Age 41–55 years	−0.204	0.82	[0.48, 1.39]	0.453
Age > 55 years	0.050	1.05	[0.58, 1.91]	0.869
University education	0.642	1.90	[1.17, 3.09]	0.010
National origin	0.212	1.24	[0.42, 3.60]	0.697
Sex × Education	−0.158	0.854	[0.31, 2.34]	0.758
Constant	−2.088	0.12	[0.04, 0.41]	0.001

Model fit: −2 log-likelihood = 438.25; likelihood ratio chi-square (df = 6) = 23.10, p < 0.001; Cox and Snell R² = 0.056; Nagelkerke R² = 0.082; AUC = 0.653 (95% CI [0.591, 0.715]). The model is interpreted as exploratory because its discriminatory capacity is only marginally acceptable. Source: Own elaboration.

). Women had 2.51 times higher odds of reporting increased WTP than men, and respondents with university education had 1.90 times higher odds of doing so. Age and national origin were not significant. The sex × education interaction was also non-significant (p = 0.758), so the two effects operate additively on the log-odds scale rather than conditionally. Because the interaction is not significant, the data provide no evidence of a synergistic effect, and no combined or multiplicative odds ratio for university-educated women should be inferred; sex and education are best reported as two independent associations of modest size. In line with the limited overall fit of the model (Nagelkerke R² = 0.082; AUC = 0.653), they should be read as broad tendencies in the sample rather than as the contours of a clearly delimited high-WTP segment.

5. Discussion

The results support the four hypotheses, although their contribution lies less in identifying unexpected behavioural patterns than in clarifying how post-pandemic sustainability attitudes take shape in an inland heritage destination once sanitary restrictions have disappeared. First, self-perceived behavioural change was minimal (H1): all P14 means remained below 2.0 on the five-point scale (Table 3), which suggests that the behavioural legacy of COVID-19 had largely faded by the time of fieldwork. Second, natural parks (92.6%) and rural destinations (84.1%) were perceived as the spaces most affected by pandemic-related demand reallocation, followed by World Heritage Cities (77.4%) and coastal destinations (73.9%) (H2). Third, 54.9% of respondents perceived the pandemic as having accelerated sustainability mainly through greater awareness among firms and public institutions, whereas only 8.6% attributed that effect primarily to tourists themselves (H3).

Finally, stated WTP appeared relatively resilient: 62.7% maintained their previous willingness to pay and 26.1% increased it despite the inflationary context (H4). These findings suggest that the post-pandemic sustainability debate in heritage tourism should not be reduced to lingering health-related caution; it is better interpreted through the interaction of destination governance, perceived credibility and value-based support for a sustainable transition.

These results nuance part of the international literature. In Asian contexts, where restrictions lasted longer and risk governance was more visible in everyday life, studies reported more persistent behavioural adaptations. Bae and Chang (2021) observed sustained support for “untact” tourism in South Korea, and Luo and Lam (2020) found significant travel anxiety around travel-bubble destinations in Hong Kong. The Cáceres case, by contrast, suggests a rapid dissipation of health fear once legal restrictions disappear and domestic mobility resumes. This contrast may be partly explained by the overwhelmingly domestic composition of the sample, by the shorter duration of restrictive measures in the observed phase and by the symbolic familiarity of a heritage city visited mainly for culture, leisure and gastronomy rather than for high-contact nightlife or long-haul travel.

The contribution of the Cáceres case is not simply to confirm that education is associated with stronger support for sustainability, a pattern already documented in previous research, but to show how that support is interpreted and articulated in a specific destination context. In an inland heritage city, where the visitor experience is closely tied to symbolic value, walkability, public-space management and conservation credibility, sustainable attitudes appear to depend not only on individual predispositions but also on the perceived capacity of institutions and local actors to translate sustainability into tangible and trustworthy practices. This contextualised reading matters for heritage destination management, because it suggests that post-pandemic sustainability is not a generic market preference but a destination-specific relationship between values, trust and governance.

These findings should, however, be read in light of the sample composition. Because 94.3% of respondents were domestic visitors, the rapid normalisation observed in Cáceres may partly reflect the behaviour of domestic post-pandemic travel rather than a pattern that generalises across all UNESCO heritage destinations. Caution is therefore needed when extrapolating these results to cities that depend more heavily on international or long-haul demand.

The main theoretical contribution is to read the findings through a combination of the TPB and VBN, as a way of interpreting—rather than demonstrating—why a decline in perceived health risk need not weaken support for sustainability. From a TPB perspective, lower perceived contagion risk reduces the salience of the restrictive attitudes and subjective norms associated with sanitary caution. From a VBN perspective, biospheric values and personal norms can continue to support environmentally responsible choices even after the health emergency has faded. Relatively resilient WTP may therefore reflect a shift from risk-avoidance motives to value-based and governance-mediated ones, a reading consistent with Wilkinson and Coles (2024). It also resonates with Asian studies in which institutional trust and visible public coordination reinforced support for sustainability-oriented tourism transitions after COVID-19 (Srisawat et al., 2023). These mechanisms are inferred from the observed pattern of behavioural normalisation and relatively stable WTP, not from direct measurement. The present data establish the descriptive associations reported in Section 4, whereas attributing those patterns to values, personal norms, institutional trust or governance credibility remains a theoretical interpretation that awaits dedicated measurement.

At the same time, the profile of increased WTP raises an equity issue. Because higher WTP is concentrated among university-educated women, the transition to sustainable tourism cannot rely solely on premium products aimed at affluent, culturally advantaged niches. Otherwise, destinations risk forms of tourism-related green gentrification, in which sustainability becomes associated with more expensive experiences, symbolic distinction and selective accessibility (Gould & Lewis, 2017; Nikšić Radić & Dragičević, 2025; Athira & Prathapan, 2026). For Cáceres and similar destinations, the policy challenge is therefore twofold: to create transparent green premiums that finance conservation and low-impact services, and to design accessible tiers of sustainable consumption so that the ecological transition does not become socially exclusionary.

More broadly, the study suggests that post-pandemic transition in heritage destinations may depend less on exceptional crisis effects than on whether destinations can institutionalise sustainability in credible, socially intelligible and economically acceptable ways. The key issue is not whether COVID-19 changed tourists in a generic sense, but whether the crisis opened a window in which heritage destinations can convert temporary disruption into more durable governance legitimacy, selective willingness to pay and support for conservation-oriented management. This shifts the discussion from a narrow behavioural reading of post-pandemic tourism towards destination governance and sustainable transformation.

6. Conclusions

6.1. Main Conclusions

6.1.1. Theoretical Contributions

The article contributes theoretically by integrating resilience thinking with the behavioural logics of the TPB and VBN in order to interpret post-pandemic sustainability attitudes in a heritage tourism context. More specifically, once the acute phase of the health emergency has passed, sustainable tourism attitudes are better understood not as a direct continuation of sanitary risk aversion but as the outcome of an interaction between declining health-related fear, persistent value orientations and confidence in destination governance. The article thus proposes a governance-and-values reading of post-pandemic sustainable behaviour in inland heritage destinations, extending discussions that have tended to emphasise either crisis-induced behavioural change or generic pro-environmental predispositions in isolation.

6.1.2. Empirical Contributions

Empirically, the study shows that tourists in Cáceres had largely normalised their behaviour by March 2023: preferences for uncrowded places, stricter health measures, shorter stays and substitution towards rural tourism all remained very low. At the same time, respondents perceived that the strongest pandemic-related reallocation of demand had affected natural and rural destinations, whereas World Heritage Cities were seen as comparatively more resilient. Most importantly, stated WTP for sustainable services appeared relatively stable, with 62.7% of respondents reporting no change and 26.1% a higher WTP. Although the logistic model is exploratory and should be read cautiously, it suggests that women and respondents with university education had higher odds of reporting increased WTP. The empirical contribution therefore lies not only in identifying these patterns but in showing how they combine in a specific inland UNESCO heritage destination, where post-pandemic sustainability appears to be associated more with institutional credibility and destination-specific governance than with lingering health concerns.

6.1.3. Public Policy Contributions

From a policy perspective, the findings suggest several practical directions. These should be read as preliminary implications rather than tested prescriptions: specific instruments such as eco-fees, earmarked green surcharges or certification schemes were not evaluated in the survey, and the directions below follow from the descriptive pattern of relatively stable stated WTP and require dedicated testing before implementation. First, municipal eco-fees or earmarked green surcharges could be introduced, provided they are transparently linked to heritage conservation and environmental management. Second, sustainable mobility bundles could connect the historic centre with low-impact transport and visitor-dispersion strategies. Third, transparent communication about sustainability-related surcharges may strengthen visitor trust. Fourth, heritage-capacity management could be reinforced through timed access, visitor-flow monitoring and measures to protect residential liveability. Finally, green certification schemes may help visitors identify credible providers and reduce uncertainty about sustainability claims.

Overall, the Cáceres case suggests that the enduring legacy of COVID-19 in inland heritage tourism lies less in persistent health-related fear than in the opportunity to convert institutional credibility and relatively stable stated WTP into a more sustainable model of destination management.

These implications are especially relevant for inland heritage destinations, where sustainability policies must simultaneously preserve symbolic value, maintain residential liveability and ensure that conservation-oriented governance remains socially and economically legitimate.

7. Limitations and Future Research

This study has six main limitations. First, its cross-sectional design captures attitudes at a single moment and does not allow the temporal stability of post-pandemic normalisation to be verified. Second, WTP is self-reported rather than observed through actual purchasing behaviour, so responses may be affected by social desirability bias. Future studies should employ methods such as discrete-choice experiments, certainty calibration techniques, or the crosswise model to mitigate this limitation. Third, the predominance of domestic tourism limits the transferability of the findings to outbound markets in which COVID-19 restrictions lasted longer or were socially more salient. Fourth, the explanatory power of the logistic model is modest because psychographic variables identified by J. Li et al. (2024), including biospheric values, personal norms, environmental self-identity, and trust in governance, were not included in the specification. Fifth, fieldwork was concentrated in a single period (March 2023), so the results may be affected by seasonal bias: the early-spring visitor profile of an inland heritage city need not coincide with that of peak summer or holiday periods, and the observed normalisation cannot be assumed to hold across the whole year. This temporal concentration compounds the composition of the sample, which is strongly domestic (94.3%) and highly educated (52.7% with a university degree); both features are likely to shape the findings on sustainability and willingness to pay, so the estimates should be read as characteristic of this profile rather than as generalisable to all visitors. Sixth, the measurement instrument has limitations of its own: the Block VI battery was not pilot-tested and its dimensionality was not assessed through factor analysis (only the internal consistency of P14 was examined), and willingness to pay was captured as a coarse directional self-report rather than through the monetary contingent-valuation measures used in the authors’ earlier work (Jurado-Rivas & Sánchez-Rivero, 2019, 2022). Future applications would benefit from a validated, multi-item instrument and from elicitation of actual price premiums.

Future research should therefore advance in three directions. A first avenue for future research would be longitudinal design replicating the survey in 2027–2028 to test whether the normalisation observed in March 2023 remains stable. A second line would compare several European UNESCO World Heritage Cities in order to identify how destination scale, governance and source markets condition sustainable tourism trajectories. A third line would combine stated preferences with behavioural experiments and the psychographic constructs emphasised by J. Li et al. (2024), especially biospheric values, personal norms, environmental self-identity, and trust in governance. Such designs would refine the interpretation of post-COVID WTP and clarify whether sustainable tourism in heritage cities can expand beyond highly educated, higher-income market segments.