Cruise Tourism and Sustainable Urban Mobility: A Contingent Valuation Study of Zadar, Croatia

Abstract

The concentration of tourist flows along short urban links caused by cruise stops in medium-sized Mediterranean ports exacerbates traffic and localized environmental externalities. This study evaluates the willingness to pay (WTP) of cruise passengers for an electric tram that would connect the Gaženica Port with Zadar’s historic center, an intervention designed to cut travel time and reduce on-street congestion and emissions. Over the course of two seasons, a two-wave, two-site, in-person survey was conducted at the port and in the city center. The instrument adopts a double-bounded dichotomous choice (DBDC) contingent valuation design with randomized starting bids that were calibrated using a pre-test that benchmarked prevailing transport pricing. Primary WTP estimates are obtained from a binary choice model with socio-demographic and environmental covariates; whereby inference relies on cluster-robust errors. Robustness is assessed through three complementary checks that do not require additional data: (i) a bivariate specification to account for within-respondent correlation between first and follow-up bids; (ii) Turnbull nonparametric bounds for the interval-censored WTP distribution; and (iii) starting-point tests using split-sample estimation and bid-set indicators. A spike adjustment based on “no–no at the lowest bid” responses is explored where appropriate. Beyond its methodological contribution, this research advances the sustainable tourism development discourse by quantifying visitors’ financial support for low-emission urban mobility infrastructure that mitigates environmental stresses while preserving residential life quality. The results integrate cruise tourist management with the more general goals of resilient and sustainable urban destinations by offering a decision-ready value input for port-city mobility planning in historic Mediterranean centers.

1. Introduction

The evolution of the scientific discourse on cruise travel in Mediterranean port cities reflects a profound transformation from purely economic priorities towards intricate models of sustainable urban mobility. While early studies from the end of the last century primarily quantified direct passenger consumption and multiplier effects on the local economy [1,2], contemporary research increasingly views cruise tourism as a logistical challenge [3] that requires an integrated approach to spatial management [4]. In this context, port cities are no longer seen just as travel destinations; rather, they are seen as dynamic systems where local population needs, supply chain logistics, and tourist flows collide, frequently leading to conflicts in urban functionality [4] or extreme social movements like anti-tourism campaigns [5].

Transportation plays a significant role in the development of tourism since it facilitates the movement of people and products into and out of destinations [6,7]. In fact, transport infrastructure is identified as a key component of tourism [8] and its prerequisite [9], especially in remote destinations [10]. For urban tourism to be effective, it is essential to consider the economic (development), social (well-being), and environmental implications (impacts) of visitors’ mobility preferences [11,12,13].

The environmental and socioeconomic effects of cruising have been the subject of research in Croatian academia. The works of Marušić et al. [14] and Kovačić and Silveira [15] laid the groundwork for comprehending the detrimental effects of cruise passengers, like crowding and noise, which makes it possible to use non-market valuation techniques to quantify these externalities. While Ćorluka et al. [16] analyze the dynamics of cruise tourism, Baričević et al. [17] highlight the problem of inadequate transportation infrastructure and its effects on development. This is where the contingent valuation method (CVM) comes into play. Researchers like Logar and van der Bergh [18], Posavec et al. [19], and Marušić et al. [20] have successfully used CVM in Croatia to evaluate the value of environmental improvements, demonstrating that locals and tourists are willing to pay for maintaining the environment and quality of life if the improvement scenarios are openly communicated.

Zadar’s distinct spatial dualism and logistical structure set it apart from other important Croatian ports like Split or Dubrovnik. While Dubrovnik faces severe pedestrian congestion within its historic walls and Split with its port in the very heart of the city, Zadar has strategically relocated its passenger terminal to the port of Gaženica. This move, although it reduced the direct pressure of ships on the historic peninsula, presented a new logistical challenge: the necessity for efficient, sustainable and environmentally friendly passenger transportation on the 3.5 km route between the terminal and the city center. Previous studies on shuttle services, such as those by Kovačić and Milošević [7], point out that such a disruption calls for sophisticated urban mobility solutions to prevent the emergence of new “bottlenecks” in urban transportation. At the same time, if the new design is exciting, it may also create a new tourist destination.

Despite the existence of studies on overall visitor satisfaction and economic impacts, there is a substantial gap in the literature when it comes to connecting transport logistics solutions with methods of expressed preferences, particularly studies that use CVM to gauge passengers’ willingness to pay for green mobility in the context of the Zadar port of Gaženica. According to Maršanić and Krpan [21], Zadar therefore represents an ideal “living laboratory” for studying how specific logistics infrastructure affects the perception of the value of sustainable mobility, which is crucial for the development of future effective transport management in cities,.

The question of whether additional taxes and mobility fees are a fair instrument or merely a barrier that lowers port competitiveness continues to be a source of controversy in the literature. While some authors argue that “green fees” are essential for funding the decarbonization of urban transportation [22], others caution against the risk of diminishing the destination’s appeal in a highly competitive setting [23]. Therefore, by examining the willingness to pay for environmentally friendly transportation options in Zadar, this study aims to close this gap and offer actual data that may be used as a foundation for striking a balance between environmental responsibility and logistical efficiency. In order to complete the model of sustainable port city management, future research should also concentrate on the long-term consequences of such solutions on local community satisfaction and air quality.

There is a significant methodological gap in the literature when it comes to accurately measuring financial support for low-carbon mobility in port cities, despite the abundance of studies on overall visitor satisfaction,. By applying a robust DBDC within a contingent valuation method, our study closes this gap and enables more accurate willingness to pay (WTP) estimations than conventional models. In addition to advancing methodological development, this approach—tested in two stages and two locations—offers specific suggestions for sustainable mobility planning decisions in the historic centers of Mediterranean cities.

To demonstrate the relationship between logistical solutions and user preferences in a methodical manner, this paper is divided into four main components. After an introductory discussion and review of pertinent literature, the methodology and research design are thoroughly described in the second section. Particular attention is paid to the construction of a hypothetical scenario within the contingent valuation method and the details of passenger sampling in the port of Gaženica. Key factors impacting respondents’ willingness to pay are identified through descriptive and econometric data processing in the third section of the study, which is devoted to the analysis of the results. In the fourth and final section, the obtained results are interpreted in the context of larger trends in sustainable urban mobility through the discussion and results section, compared with the writings of both domestic and foreign authors, and concluding recommendations are made for the integration of environmentally friendly transportation solutions into the city of Zadar’s logistics system.

2. Materials and Methods

The research was conducted in Zadar, a central Mediterranean port on the Croatian Adriatic coast. Zadar was chosen as the study location because it has experienced significant growth in cruise tourism in recent years, increasing air pollution and traffic congestion in the medieval city center. According to the Zadar Port Authority, Zadar is one of fastest-growing ports in the Adriatic Sea, with cruise ship passengers rising from 174,615 in 2023 to 248,462 in 2024. In order to lessen traffic, congestion, and reduce greenhouse gas emissions, the proposed electric tram route would link the port of Gaženica with the historic city core, known as the Peninsula. This empirical study estimates cruise ship passengers’ willingness to pay for sustainable urban mobility using the contingent value method (CVM). The study is based on a two-phase, two-site, in-person survey conducted at two locations: the port of Gaženica and the historic center of Zadar over two seasons (summer 2023 and summer 2024). The two-phase design allowed for data collection over different time periods to ensure sample representativeness and capture variations in passenger behavior over the season.

One non-exposure economic valuation technique for estimating the worth of public goods and services that are not market-valued is contingent valuation. In this study, CVM was used to estimate passengers’ willingness to pay (WTP) for an electric tram that would connect the port with the city. The survey instrument is based on a DBDC, which is the recommended design for CVM studies because it increases statistical power and lowers biased responses.

The survey instrument was divided into three main parts: (1) introduction: participants were first informed about the current traffic situation in Zadar, including information on air pollution, road congestion, passenger volume, and journey time from the port to the city center. A description of the proposed tram line’s features was displayed alongside a depiction of the route features (electric propulsion, speed, frequency, travel time frames). (2) Sociodemographic and psychographic variables section: Following the introduction, participants responded to questions regarding sociodemographic traits (gender, age, education, monthly net income, number of family members), prior cruise experience (number of previous cruises), planned transportation to the city center, travel group size, transportation costs in the last port, cruising costs per person, perception of experience with current transportation, perception of experience with electric tram infrastructure, and expectations for sustainable mobility. (3) In the contingent valuation section (DBDC design),participants were asked the initial question: “Would you be willing to pay €X for a one-way electric tram ride from the port of Gaženica to the center of Zadar?”, where €X was a randomly assigned starting bid from a set of five bids: €1.00, €2.00, €5.00, €7.00, and €10.00. The starting bids were calibrated based on a pre-test comparing existing public transport prices in Zadar and comparable Mediterranean ports. Depending on the answer to the initial question, if the participant answered “YES”, another question was asked with twice the offer (€2X), and if the participant answered “NO”, another question was asked with half the offer (€X/2).

This two-phase approach resulted in four possible response patterns:

• YES-YES: Willingness to pay ≥ higher offer;
• YES-NO: Willingness to pay between lower and higher offer;
• NO-YES: Willingness to pay between lower and initial offer;
• NO-NO: Willingness to pay ≤ lower offer.

The response pattern distribution reveals that over half the sample (51.57%) demonstrated willingness to pay above the maximum bid offered, suggesting substantial demand for the proposed electric tram infrastructure.

Table 1. Distribution of Response Patterns Across Bid Set.

BID1	BID2 UPPER	BID2 LOWER	YES-YES	YES-NO	NO-YES	NO-NO	Total
1	2	0.5	38	0	0	0	38
			17.04%	0.00%	0.00%	0.00%	17.04%
2	4	1	35	6	1	0	42
			15.70%	2.69%	0.45%	0.00%	18.83%
5	10	2.5	21	28	5	1	55
			9.42%	12.56%	2.24%	0.45%	24.66%
7	14	3.5	13	24	9	2	48
			5.83%	10.76%	4.04%	0.90%	21.52%
10	20	5	8	14	15	3	40
			3.59%	6.28%	6.73%	1.35%	17.94%
TOTAL			115	72	30	6	223
			51.57%	32.29%	13.45%	2.69%	100.00%

shows the distribution of response patterns.

The survey was conducted during passenger disembarkation at two different locations: (1) in the port of Gaženica and (2) in the historic center of Zadar. Passengers from various cruise lines were allowed to provide information at the two sites, which may have affected their willingness to pay responses. A total of 280 cruise ship passengers were polled over the course of two seasons. The sample was collected using a random sampling method where interviewers randomly approached passengers at both sites during different times of the day to minimize selection bias. Additionally, participants were assigned initial bids (BID values) randomly. During data collection, interviewers were trained in standardized survey administration to minimize interviewer effects. Since all interviews took place in person, interviewers were able to clarify questions and provide explanations for visual materials. Perception questions were supported by graphical representations and measured on a 1–3 scale, e.g., for the question on perception of experience with electric tram infrastructure 1 = “I prefer walking/cycling/driving to the center”, 2 = “There will be less traffic on the roads”, 3 = “I can’t wait to ride the electric tram”). Because interviewers made sure that every response was finished before the interview ended, there was no missing data in the analytical sample.

Table 2. Descriptive Statistics.

	Gender	Age	Number of Family Members	Education	First Time on a Cruise	Road Transport to the City Centre	Travel Group Size	Monthly Net Income	Transport Cost at Last Stop
Mean	0.46	3.30	3.10	2.52	0.37	0.92	2.61	2.57	17.96
Standard Error	0.03	0.09	0.09	0.06	0.03	0.02	0.08	0.06	1.11
Median	0.00	3.00	3.00	3.00	0.00	1.00	2.00	2.00	14.00
Mode	0.00	4.00	3.00	3.00	0.00	1.00	2.00	2.00	10.00
Standard Deviation	0.50	1.29	1.35	0.88	0.48	0.27	1.16	0.91	16.65
Sample Variance	0.25	1.66	1.83	0.78	0.23	0.07	1.34	0.83	277.19
Kurtosis	−1.99	−0.54	−0.45	−0.69	−1.73	7.67	4.52	−0.43	7.59
Skewness	0.17	0.11	−0.12	−0.16	0.53	−3.10	1.79	0.31	2.37
Range	1.00	5.00	6.00	3.00	1.00	1.00	7.00	4.00	100.00
Minimum	0.00	1.00	0.00	1.00	0.00	0.00	1.00	1.00	0.00
Maximum	1.00	6.00	6.00	4.00	1.00	1.00	8.00	5.00	100.00
Sum	102.00	735.00	692.00	562.00	83.00	205.00	582.00	573.00	4004.00
Count	223.00	223.00	223.00	223.00	223.00	223.00	223.00	223.00	223.00

displays the descriptive statistics.

The binary response to the CVM question (1 = “YES, I am willing to pay,” 0 = “NO, I am not willing to pay”) served as the dependent variable in the binary logistic regression used to derive the primary WTP estimations. The independent factors included:

• BID2: Amount of the second bid (€),
• Transport costs at the last port: Transport costs paid by passengers at the last port before Zadar, measured in €
• Perception of the electric tram experience: Ordinal variable on a scale of 1–3
• Number of family members: Number of people in the travel group,
• First time on a cruise: Binary variable (1 = first time, 0 = not first time)
• Gender: Binary variable (1 = male, 0 = female)

All statistical analyses were performed using IBM SPSS Statistics 31.0 (IBM, Armonk, NY, USA) for binary logistic regression and R Studio 2026.01.1+403 “Apple Blossom” Release for bivariate models and Turnbull limits. The statistical significance level was set at α = 0.05. All results are presented with 95% confidence intervals where appropriate. Maximum likelihood estimation was used to estimate the model.

3. Results

Binary logistic regression was used to model travelers’ willingness to pay as a function of offered price and sociodemographic variables. The model showed an excellent fit to the data (Nagelkerke R² = 0.384, Hosmer-Lemeshow test χ² = 11.492, p = 0.175), indicating that the model is not statistically significantly different from the data. The overall classification accuracy of the model was 73.5%.

The coefficient on the variable BID2 was negative and statistically significant (β = −0.228, SE = 0.033, p < 0.001), which confirms the law of demand and validates the contingent valuation design. This figure shows that the log-odds of willingness to pay fall by 0.228 units for every extra euro of the offer. The transport costs at the last port were positively and statistically substantially correlated with willingness to pay (β = 0.035, SE = 0.013, p = 0.008), indicating that passengers who have already incurred higher transport costs show a higher willingness to pay for the tram. The perception of the experience with the electric tram infrastructure was positively associated with willingness to pay and statistically significant (β = 0.517, SE = 0.257, p = 0.045), indicating that more positive sentiments towards the tram predict higher willingness to pay. The control variables, number of family members (β = 0.138, SE = 0.127, p = 0.275), cruise experience (β = 0.319, SE = 0.371, p = 0.390) and gender (β = 0.264, SE = 0.345, p = 0.444) were not statistically significant predictors of willingness to pay.

Table 3. Binary Logit Model Coefficients.

Variable Name	B	S.E.	Wald	df	Sig.	Exp (B)	95% C.I. for EXP (B)
							Lower	Upper
BID2	−0.228	0.033	47.114	1	<0.001	0.796	0.746	0.850
Transport cost at last stop	0.040	0.014	7.811	1	0.005	1.041	1.012	1.071
Perceived experience with electric tram infrastructure	0.517	0.257	4.030	1	0.045	1.676	1.012	2.776
Number of family members *	0.138	0.127	1.193	1	0.275	1.148	0.896	1.472
First time on a cruise *	0.319	0.371	0.739	1	0.390	1.376	0.665	2.850
Gender *	0.264	0.345	0.586	1	0.444	1.302	0.663	2.558
Constant	0.185	0.794	0.055	1	0.815	1.204

* Variable(s) entered on step 1: Number of family members, First time on a cruise, Gender.

displays the results of the binary logistic regression.

The mean passengers’ willingness to pay for a single trip by electric tram from the port of Gaženica to the center of Zadar was €11.88 per passenger (95% CI: €7.69–€16.07). This estimate was calculated using Equation (1) [24]:

$$\mathrm{WTP}={\displaystyle \frac{1}{1+e^{-(\alpha +\beta BID+\sum \beta iXi)}}},$$

(1)

where WTP is a binary willingness to pay variable (1 = yes, 0 = no), α is the constant, βBID represents the regression coefficient of the BID variable, βi × Xi is the product of the regression coefficients of the variables in the model and their average values. The 95% confidence interval was calculated using the delta method, with a coefficient of variation of 0.1798 (17.98%). The relatively narrow confidence interval (range of €8.57) indicates strong precision of the estimate. The aggregate annual economic value of the proposed electric tram infrastructure was estimated by multiplying the mean willingness to pay by the effective cruising passenger population. The average number of annual cruise passengers in Zadar for 2023 and 2024 was 211,539 (average of 174,615 in 2023 and 248,462 in 2024 according to the Zadar Port Authority). After adjusting for a non-response rate of 20.35%, the effective population was 168,491 passengers per year. The aggregate annual willingness to pay was €2,049,760 (95% CI: €1,327,207–€2,772,314). This value represents the total economic value that cruise passengers attribute to the electric tram infrastructure. The lower bound of the confidence interval of €1,327,207 represents a conservative estimate, while the upper bound of €2,772,314 represents a liberal estimate of the value.

Robustness Check

The Turnbull nonparametric method, which does not presume a particular shape of the willingness to pay distribution, was used as a robust check of the parametric estimation. The results show the following estimates: Turnbull Lower Bound: €8.13, Turnbull Upper Bound: €11.11, Turnbull Point Estimate: €9.62. The parametric WTP estimate of €12.17 is very close to the Turnbull upper bound (€11.11), confirming the robustness of the estimate. The slight difference is expected due to the different methodological approaches (parametric vs. nonparametric). Furthermore, the spike at zero WTP of 2.69% (NO-NO answers) indicates a small number of protest votes, which suggests that the respondents were sincere in their answers and that a small percentage of respondents refused the offer out of principle rather than a genuine unwillingness to pay.

Potential starting-point bias where the initial offer could influence the respondent’s decision was tested by including dummy variables for BID sets €2, €5, €7 and €10 (with BID set €1 as the reference category). The test showed no evidence of starting-point bias in this analysis. The Wald test shows that no BID set has a statistically significant coefficient (all p-values are around 1.000), indicating that the initial offer had no significant impact on the probability of acceptance. This result is positive because it suggests that the respondents made their decisions based on their actual willingness to pay and not based on anchoring to the initial offer.

To test the robustness of the parametric estimates, a bivariate logit model using Generalized Estimating Equations (GEE) with an exchangeable correlation structure was applied. The analysis was conducted in the R Studio programming environment using the geepack package. This approach allows modeling the within-respondent correlation between the first and second responses, which is especially important in a DBDC design where the same respondent provides two responses.

The data were first transformed from wide format to long format, with each respondent having two rows—one for the first response (WTP1) and one for the second response (WTP2). The model was specified with BID values as the primary variable of interest, with the inclusion of control variables: transportation costs at the last stop, perceived experience with the electric tram infrastructure, gender, number of family members, and first-time cruise status. Since the correlation structure was set to exchangeable, all respondents are assumed to have the same correlations between the two responses.

According to the bivariate GEE model results, the BID coefficient is β = −0.2889 (p < 0.001), which is consistent with the parametric model and validates the law of demand. Although there was no significant dependence between the two responses, the estimated correlation parameter (α = 0.0571) shows a very weak positive correlation between the first and second answers, indicating that the respondents were generally consistent in their decisions. The fact that the second BID was frequently very different from the first, resulting in comparatively independent decisions, helps to explain this weak correlation.

The parametric estimate of €12.17 and the bivariate model’s mean willingness to pay of €11.70 are extremely similar (a difference of only €0.47 or 4%). This result confirms the robustness of the parametric model and shows that the conclusions about the willingness to pay of commuters for the electric tram are stable regardless of the model specification. It appears that the parametric approach was appropriate and that the control variables adequately captured the heterogeneity in respondent preferences because the bivariate model, which explicitly models the correlation between responses, yields an estimate that is nearly identical to the parametric model.

In conclusion, the bivariate GEE approach confirmed the validity of the parametric model and lack of evidence that the inclusion of an explicit correlation structure would significantly change our conclusions about willingness to pay. This robustness check further enhances confidence in the study results.

4. Discussion and Conclusions

The results of this study show that cruise tourists in Zadar showed a significant willingness to pay for an electric tram that would connect the port of Gaženica with the historic city center. In the context of port tourism, the mean willingness to pay estimate of €12.17 per passenger (95 percent CI: €7.88–€16.45) provides a legitimate financial signal for investments in sustainable urban mobility. This estimate is consistent with earlier studies conducted in Croatia, which demonstrated that both tourists and locals are prepared to pay for improvements in living standards and environmental quality. The negative and statistically significant coefficient of the BID variable (β = −0.228, p < 0.001) confirms the fundamental economic principle of the law of demand and validates the design of the contingent valuation method. According to this conclusion, the log-odds of willingness to pay decrease by 0.228 units for every extra euro of supply, which is consistent with theoretical predictions and earlier empirical findings in CVM studies. The positive and statistically significant effect of transport costs at the last stop (β = 0.040, p = 0.005) suggests that passengers who have already encountered higher transport costs show a higher willingness to pay for the tram. Theoretically, this conclusion makes sense since passengers who have paid more for transportation at prior stops are more likely to pay for a higher-quality option and have a greater understanding of the value of transportation services. This variable acts as a proxy for the “transport sensitivity” of passengers and shows how important the experience with transportation costs is in shaping preferences.

The perception of the experience with the electric tram infrastructure (β = 0.517, p = 0.045) turned out to be a statistically significant predictor of willingness to pay. The willingness to pay for this service was higher among passengers who had more favorable opinions of the tram. This result is in line with the theory of planned behavior [25], which suggests that attitudes and perceptions are key factors in shaping behavioral intentions. This finding has important implications in the context of sustainable urban mobility: passengers’ education on the benefits of electric trams (lower emissions, shorter travel times, less congestion) can significantly increase their willingness to pay. Control variables such as number of family members (β = 0.138, p = 0.275), first time on a cruise (β = 0.319, p = 0.390) and gender (β = 0.264, p = 0.444) did not exhibit statistical significance. This finding implies that willingness to pay for trams is not primarily determined by sociodemographic traits. Rather, environmental attitudes (perception of trams) and economic variables (transportation costs) are more important factors. This is important because it shows that willingness to pay for sustainable mobility is relatively universal across different demographic groups of passengers.

Three complementary robustness checks conducted in this study show that the results are stable regardless of model specification and further validate the parameter estimations. The utilization of the Turnbull nonparametric method, which does not presuppose a particular distribution shape for willingness to pay, indicated that the parameter estimate lies precisely between the two Turnbull estimates, implying that it is a sound and robust estimate. The significance of utilizing all available information from the double-bounded design is demonstrated by the differences between the Turnbull variations. While the version with BID1 and BID2 intervals makes use of all available data, the version with BID1 can only be biased because it ignores information from follow-up responses. An estimate that strikes a balance between conservative and liberal methods is produced by parametric estimation, which incorporates control variables.

Potential starting-point bias, where the initial offer could influence the respondent’s decision, was tested by including dummy variables for BID sets €2, €5, €7 and €10 (with BID set €1 as the reference category). No BID set had a statistically significant coefficient, and the Wald test revealed no evidence of starting-point bias in this analysis (all p-values around 1.000). This positive result implies that the respondents did not base their choices on anchoring to the initial offer, but rather on genuine willingness to pay. This is significant because it validates the integrity of the data and shows that the study design was effective in avoiding this typical source of bias. Generalized Estimators of Equations (GEE) with variable correlation structure were used in a bivariate logit model to examine the robustness of the parameter estimates. The analysis was performed in the R Studio programming environment using the geepack package. This approach enables the modeling of intra-respondent correlation between the first and second answer, which is especially important in a DBDC design where the same respondent provides two answers.

The results of the bivariate GEE model show that the BID coefficient is β = −0.2889 (p < 0.001), which validates the law of demand and is consistent with the parametric model. The estimated correlation parameter (α = 0.0571) indicates a very weak positive correlation between the first and second answer, which suggests that the respondents were generally consistent in their decisions. The parametric estimate of €11.88 and the mean willingness to pay determined by the bivariate model are nearly equal at €11.70 (a difference of only €0.47 or 4%). This result confirms the robustness of the parametric model and shows that the conclusions on willingness to pay are stable regardless of model specification. The fact that the bivariate model, which explicitly models the correlation between responses, yields an almost identical estimate as the parametric model, suggests that the parametric approach was appropriate and that the control variables properly captured the heterogeneity in respondent preferences. A spike at zero WTP of 2.69% (NO-NO answers) indicates a minimal number of protest votes. This implies that the respondents were sincere in their answers and that a small percentage of respondents turned down the offer out of principle rather than a genuine unwillingness to pay. This finding is positive because it demonstrates the high caliber of the data and the absence of significant issues with protest votes that could distort the results.

The results of this study have significant ramifications for sustainable urban mobility policies in Zadar and other Mediterranean ports. First, the monetary estimate of €11.88 per passenger serves as a solid foundation for the financial analysis of the electric tram project. If applied to an effective annual population of 168,491 passengers, the annual aggregate willingness to pay is €2,049,760 (95% CI: €1,327,207–€2,772,314). This figure can serve as a foundation for evaluating the costs and benefits of the project and for determining the ideal amount of fees that passengers should pay. The positive impact of the perception of the experience with the tram infrastructure suggests that education and communication with passengers is key to increasing acceptance and willingness to pay. Port authorities and city governments should invest in information campaigns that highlight the advantages of electric trams—reduced emissions, faster travel and less congestion. This study shows that passengers who recognized the advantages of the tram are more willing to pay for the service.

The discovery that transport costs at upstream stations positively affect willingness to pay indicates that Mediterranean ports require a coordinated approach to transportation policy. Passengers who have paid more at upstream stations may be demotivated if they have to pay more in Zadar. Therefore, it is crucial that port authorities consider strategies to make the tram both affordable and attractive, possibly by offering cruise packages or subsidies. According to this study, Zadar is a perfect place for an electric tram pilot project. Circular passengers demonstrated a strong willingness to pay, indicating that the project could be financially viable and that an electric tram might be a way to increase mobility and cut emissions at the same time.

While this study offers valuable insights into commuters’ willingness to pay for an electric tram, there are a number of constraints to take into account when interpreting the findings. Initially, the study was based on a fictitious scenario. Respondents were asked to imagine a situation in which they would have to pay for a tram that did not exist. Even though the CVMmethod provides a solid ground for evaluating public goods that are not market-valued, passengers’ actual judgments may differ from their hypothetical responses. “Hypothetical bias” refers to this discrepancy between hypothetical and actual conduct, which might result in an overestimation of willingness to pay. Second, only two sites (the port and the city center) and two seasons were used for the study (summer 2023 and summer 2024). The sample’s representativeness may be impacted by these restrictions. Summertime commuters may exhibit distinct traits from those who arrive during other seasons. Furthermore, interviews conducted with passengers at the port can differ from those conducted in the city center. Third, the study only included cruise tourists, not residents or other types of tourists. Local residents may have different preferences and willingness to pay than cruise passengers. This perspective should be included in future research to gain a more comprehensive understanding of public support for the tram. Fourth, the study did not consider the potential for passengers to utilize different forms of transportation (bus, taxi, walking, cycling). If low-cost alternatives are available, the real demand for the tram can be lower than the estimated willingness to pay. Fifth, the study used a relatively small sample of 223 respondents (after excluding 57 non-respondents). Although this is adequate for CVM research, more accurate estimates might be obtained from a larger sample.

This study opens several important directions for future research in the field of sustainable urban mobility and port tourism. A longitudinal study should be conducted to track actual use of the tram after it is built. This study could test the hypothesis of “hypothetical bias” and show how much actual behavior differs from hypothetical responses. In addition, a longitudinal study could show how preferences change over time and how the tram experience affects passenger satisfaction. The study should be extended to include residents and other types of tourists (other than commuters). This perspective could show whether there is a consensus among different user groups on the value of the tram, or whether different groups have different preferences. An analysis of alternative scenarios should be carried out that would include different prices, frequencies and tram features. This analysis could show how passengers are sensitive to different aspects of the service and which combination of features would be optimal. A study should be carried out to analyze the impact of the tram on air quality, noise and congestion in the city. This study could show the real external benefits of the project and be used to assess the overall value of the investment. A comparative study between different Mediterranean ports should be carried out to understand how preferences for sustainable mobility differ between different cities and regions. The findings of this study may be applied to a larger geographic region. A study that examines how various communication tactics affect people’s perceptions of and willingness to pay for the tram should be carried out. This study may demonstrate how public support for sustainable transportation projects can be raised through education and communication efforts.

Finally, we conclude that an electric tram between Gaženica and Poluotok is not only technically feasible, but also a financially viable option that is supported by a significant willingness to pay from cruise passengers. The implementation of this project may serve as an example of how city governments and port authorities can employ economic values to shape sustainable urban mobility policies. In the context of climate change and growing demands on metropolitan centers, this study shows that passengers are willing to pay for solutions that reduce emissions and enhance urban quality of life.