Tourism Spending in Response to Stimulus Checks: Insights From Mental Accounting

Abstract

The tourism sector closely monitors fiscal stimulus policies during economic downturns, as these policies are generally assumed to increase consumer spending to aid recovery. However, previous retrospective analyses fail to capture the intricacies of the individual decision-making process. The varying contexts in which stimulus measures are applied further disrupt predictable consumers’ behavior, challenging the rational utility-maximization framework for household budget allocations. To address these gaps, a shift towards behavioral economics is essential for a nuanced understanding of how stimulus funds are allocated by recipients to tourism products. Through a mental accounting approach, this study suggests that factors at policy- (size and source), firm- (product presentation), and individual-level (tourism’s importance and contribution to quality of life) jointly influence tourism spending derived from stimulus checks. This study underscores the significance of behavioral economics principles in shaping effective fiscal stimulus policies and marketing strategies within the tourism sector.

Keywords

fiscal stimulus income tourism spending mental accounting quality of life health risk perception

Highlights

The study found that mental accounting determines the stimulus expenditure of tourists.

Dirichlet and Beta regressions explain stimulus allocation patterns.

Rigidity and malleability affect mental accounts and budget reallocation.

Introduction

While free markets fuel the prosperity of the tourism industry, governmental interventions remain critical, especially during challenging times (C. M. Hall et al., 2020). This is evident in past incidents such as the 2003 SARS outbreak (Kuo et al., 2008), the 2008 global financial recession (Meng, 2014), and the recent COVID-19 pandemic (Fang et al., 2022), where government aid plans have played a vital role in mitigating adverse impacts and building the resilience of the tourism industry (Higgins-Desbiolles, 2020). Among various measures, fiscal stimulus payments have gained notable attention. These measures, including direct transfers and tax rebates, are commonly referred to as stimulus checks (Fang et al., 2022). They aim to boost consumption by providing consumers with additional disposable income (Baker et al., 2023).

Given the well-established positive correlation between income and tourism demand, such policies hold significant potential for tourism recovery. However, when assessing their specific impacts, an important question arises: As a distinct source of income, can stimulus checks effectively translate into increased tourism spending? To answer this, we must figure out how stimulus funds are processed into tourism consumption.

Studies on how subsidies impact tourism demand (e.g., Chow et al., 2021) rarely distinguish between subsidies and regular income. However, emerging evidence suggests that an effortless income source can significantly increase the income elasticity of leisure activities (Dardis et al., 1994), travel frequency and duration (C.-M. Chen & Chen, 2018), and expenditures on domestic and international traveling (Crouch et al., 2007). This deficiency indicates that conventional economic paradigms, which treat money as fungible, may not fully capture the impact of stimulus checks with a windfall nature on tourism spending behaviors.

Mental accounting, a theory in behavioral economics, recognizes the non-fungibility of money and postulates that individuals segregate their finances into “sub-accounts” for various expenses (Shefrin & Thaler, 1988). With this framework, the study aims to explore how individuals allocate and reallocate stimulus funds for tourism through the mental accounting mechanism. This endeavor is intended to deepen our understanding of the intricate relationship between income changes and tourism spending, with a particular focus on unique income sources like stimulus checks. Additionally, this study aims to provide actionable insights for future public and marketing measures by identifying strategies to enhance the allocation of stimulus funds toward tourism, thus supporting the tourism industry amidst adverse conditions.

The findings indicate that sizable and bonus-framing stimulus checks increase tourism spending. Meanwhile, the perceived contribution of tourism to quality of life is significantly related to the allocation of stimulus funds toward tourism, while the health risk perception is not. Notably, as the importance of tourism increases or expenses become more ambiguous, tourism spending from stimulus funds rises due to the rigidity and malleability of tourism mental accounts. These results contribute to understanding the underlying behavioral mechanism that shapes the response of tourism spending to stimulus checks. Practically, they inform more effective policy design and marketing strategies by suggesting that (1) positively framed, large-sized fiscal measures, and (2) targeting frequent travelers, developing multi-purpose tourism products, and offering relaxation and therapeutic experiences can direct more stimulus funds toward tourism.

The structure of this paper is as follows. Section 2 provides an extensive literature review, followed by an introduction to the mental accounting framework in Section 3. Section 4 outlines the research design, data collection procedures, and models. Section 5 presents the findings derived from rigorous data analysis. Finally, Section 6 concludes the paper.

Literature Review

Stimulus checks, as a cash-based stabilization tool, aim to boost consumption by increasing consumers’ disposable income, thereby facilitating economic recovery (Baker et al., 2023). However, proponents of the Permanent Income Hypothesis (PIH) and Life-Cycle Hypothesis (LCH) question their effectiveness in stimulating spending. They argue that consumption decisions are primarily determined by individuals’ permanent or expected lifetime income rather than by temporary fluctuations in income (R. E. Hall & Mishkin, 1982).

Despite these theoretical reservations, empirical evidence suggests that fiscal stimulus payments have elicited stronger consumption responses than previously anticipated. On the one hand, the excess sensitivity is explained by financial factors, such as the relaxation of liquidity constraints and the reduction of precautionary saving motives facilitated by stimulus funds (Flavin, 1981). On the other hand, it relates to cognitive bias. Individuals often exhibit temporal myopia when differentiating between permanent and temporary income changes (Shapiro & Slemrod, 2003). Stimulus checks, perceived as windfall rather than regular income, result in a higher propensity to spend (Sahm et al., 2012). That challenges the conventional economic theory assuming that money is fungible and its source does not affect consumption (Neumann et al., 2007).

The departure from classical economic rationality has been noted in the tourism literature. For example, changes in household head salary or non-salary income were found to have a more pronounced effect on leisure expenditures than changes in other income streams (Dardis et al., 1994).Another study noted that retirees supported by their children or labor insurance benefits tend to shorten travel duration compared to those relying on personal savings (C.-M. Chen & Chen, 2018). More recently, Boto-García et al. (2024) proved that lottery winnings lead to increased annual trips and higher expenditure. These findings clearly violate the fungibility assumption in the traditional tourism spending analysis.

Also, the fiscal policy literature reveals an excess sensitivity of tourism expenditure to windfall incomes compared to other expenditure categories. For instance, Souleles (2002) found that during the Reagan tax cuts, most of the spending response occur in non-durable goods, with disproportionate spending on summer trips. A hypothetical experiment modeled after realistic government stimulus schemes revealed a significantly higher propensity to spend on domestic and overseas travel than other discretionary expenditures, such as home renovations, leisure activities, and charitable donations (Crouch et al., 2007). By contrast, a report on the 2021 Hong Kong Consumption Voucher Scheme (HKCV) indicated that while the program increases spending on travel, the growth is negligible compared to groceries, restaurants, and services (Geng et al., 2022). These studies collectively suggest that stimulus-induced income shocks can shift spending patterns across categories, underscoring the need to examine how tourism competes with other discretionary expenditures.

Recent research highlights the necessity of incorporating inter-category trade-offs when analyzing tourism demand in response to income fluctuations (e.g., Bronner & de Hoog, 2016; Crouch et al., 2007). As Dolnicar et al. (2008, p. 45) argue,

An exclusive focus on the absolute level of disposable income as an explanatory variable in (for instance) tourism demand models results in misleading conclusions about the effects of income changes if it is not taken into account that households are faced with competing spending options.

This perspective is especially critical for the impact evaluation of stimulus payments on tourism spending, as allocation strategies can significantly influence the proportion of funds directed toward tourism, which, in turn, inform public policy in the tourism industry (Crouch et al., 2007).

Mental accounting theory, emphasizing monetary non-fungibility and cross-category trade-offs, is well-suited to explore the relationship between stimulus checks and tourism spending. The following section will focus on developing hypotheses within this framework.

Theoretical Frameworks

Shefrin and Thaler (1988) proposed that individuals tend to partition financial resources mentally into separate categories, known as “mental accounts.” The concept diverges from the traditional economic model of utility-maximization and provides insight into how people establish rules to regulate spending within designated budgetary domains. Recent studies have applied the mental accounting principle to analyze tourism expenditure patterns (e.g., Guan et al., 2023; D. Kim & Jang, 2017). As a decision-making process, mental accounting involves three interconnected steps, based on which we develop hypotheses as follows:

Stage 1: Categorization

The first stage—Categorization—entails the psychological segregation of funds into distinct mental accounts based on their sources and amounts. Specifically, bonus-framed income, which is unexpected and unearned, is often allocated to a separate or pendent account. This framing tends to induce higher spending and hedonic purchases (Ha et al., 2006), as the positive emotions associated with such funds alleviate self-control anxiety (Epley et al., 2006) and guilt for indulgent consumption (Kivetz, 1999). In contrast, rebate-framed income emphasizes the recovery of losses and thus motivates savings or spending on utilitarian purchases (Lin et al., 2022). Similarly, tourism studies have found that bonus-framing has a more positive effect on tourists’ subsequent purchases (Ma & Li, 2023) and attraction membership subscriptions (Byun & Jang, 2015) than discounts. The impact is also observed in fiscal stimulus policies. The 2001 U.S. tax rebates, framed as a “return to the taxpayer,” were often saved or spent on durable goods (Shapiro & Slemrod, 2003), while China’s public coupon program, described as government subsidies, was effective in stimulating tourism demand (Liu et al., 2021). Therefore, we hypothesize:

H1: Bonus-framed stimulus checks boost tourism spending more than rebate-framed ones.

Income size also affects how individuals mentally categorize funds. Smaller incomes, perceived as trivial to overall wealth, are readily spent on discretionary items as they require little justification. Larger incomes are seen as substantial additions to wealth, leading to more deliberate financial decision-making (Fagereng et al., 2021).

The availability of larger-sized subsidies makes luxurious products, such as trips, more affordable. Additionally, these subsidies evoke a sense of relief and relaxation among individuals. The positive emotions encourage individuals to categorize such windfalls as strategic funds (Kivetz, 1999) and to allocate them toward carefully planned, experiential, or self-rewarding expenses that may have been deferred during difficult times (E. E. K. Kim et al., 2022). Following this logic, it can be postulated:

H2: Larger stimulus checks boost tourism spending share more than smaller ones.

Stage 2: Mental Budgeting

The second stage is Budgeting, where individuals strategically allocate funds across multiple accounts and set spending thresholds based on cost-benefit analysis. Essentially, consumers implicitly budget for goods or experiences that fulfill psychological needs while minimizing perceived costs (Thaler, 1999). During economic downturns, while priorities usually shift towards necessities, individuals recognize the value of certain pleasurable activities that can enhance life quality without detracting from essential needs (Kivetz, 1999). These activities are regarded as “investments” in personal happiness or well-being that reduce the pain of paying and promote expenditure on rewarding experiences.

Tourism is widely recognized for its significant contribution to enhancing the quality of life (Gilbert & Abdullah, 2004). Y. Chen et al. (2013) suggested that vacations positively impact on short-term quality of life under specific contexts. Notably, under adverse economic conditions, it was observed that some people prioritize summer holidays to pursue happiness (Bronner & de Hoog, 2016). In times of health crises, tourism products prove especially beneficial for individuals with physical or psychological distress by providing an escape from their difficulties (S. Kim & Gal, 2014). Building on these insights, a recent study further demonstrated that the quality of life index significantly predicts tourism demand (Berbekova et al., 2025). Therefore, we expect that:

H3: Tourism′s contribution to the quality of life is positively related to tourism spending share.

During pandemic-induced recessions, health crises present psychological challenges to consuming tourism products. The perceived health risks can trigger negative emotions such as fear, anxiety, and worry, as observed in previous outbreaks like SARS, Ebola, and COVID-19 (Peng et al., 2022; Reisinger & Mavondo, 2005). In such situations, tourists are more likely to exhibit risk aversion due to the high level of uncertainty and risk involved (Lee et al., 2012). They may be motivated to cancel or postpone travel plans to minimize negative consequences (Bronner & de Hoog, 2016). Therefore, we assume:

H4: Health risk perception is negatively related to tourism spending share.

Stage 3: Mental Adjustments

The final stage is Adjustments, where individuals reassess and reallocate the budgets of their mental accounts based on incurred costs or surplus funds. On the one hand, mental accounts can be independent, with unambiguous demarcation of expenses and funds that is resistant to modification (Heath & Soll, 1996). Such rigidity helps facilitate self-regulation, ensuring that important needs are met without diverting to lesser priorities.

While tourism is commonly viewed as a discretionary expenditure that may be sacrificed for essential needs in difficult times, some research has proved that tourism demand is resilient among “loss-averse travelers” (Bronner & de Hoog, 2016). Namely, some individuals place considerable importance on tourism, and they endeavor to maintain traveling even under financial strain (Smeral, 2010). For those people, the mental accounts for tourism are hard to break to give way for other purchases, thus ensuring the funds for tourism consumption. As such, it can be posited:

H5: The importance of tourism is positively related to tourism spending share.

On the other hand, contrary to the initial assumption of precise categorization and demarcation in mental accounts, later research suggests that monetary transactions can be vague and confusing. In such cases, slight variations in account naming, allocation, or organization can influence consumption decisions (Thaler, 1999), resulting in outcome extensibility and money fungibility (Shafir & Thaler, 2006). When encountering an attractive but ambiguous purchase, consumers may transfer the expense to a surplus account to justify spending more (Cheema & Soman, 2006). For example, tourists might find spending easier when their trip combines business and leisure (Kivetz, 1999). That suggests the following hypothesis:

H6: The expense ambiguity is positively related to tourism spending share.

In conclusion, tourism spending is not a random process but a deliberate choice (Brida & Tokarchuk, 2015). Building upon the aforementioned arguments, we propose the research framework illustrated in Figure 1.

Methodologies

Research Design

Rationale for hypothetical scenario design

Although analyses of real-world fiscal stimulus programs offer valuable insights into their actual impact on tourism spending, they have inherent limitations for this research. First, existing programs vary significantly in timing, location, and scale. These variations make it challenging to isolate the specific impacts of variables of interest. In contrast, hypothetical treatments enable precise control over variables such as stimulus amount and framing, enhancing internal validity and facilitating comparisons of alternative policy designs.

Second, while some studies have used small amounts of real money to elicit behavioral responses (e.g., Heilman et al., 2002; Milkman & Beshears, 2009), this approach is insufficient to capture the full impact of substantial windfalls like large-scale fiscal payments because the size of the stimulus, as discussed earlier, plays a critical role in shaping consumer behaviors. Replicating these programs in experimental settings is practically infeasible. In response to these challenges, researchers have utilized structured expenditure simulations. For instance, Crouch et al. (2007) and Dolnicar et al. (2008) operationalized this approach by presenting participants with a hypothetical AUD 2,000 (approximately US$2,570 based on the exchange rate at the time of the survey) bonus to explore the trade-offs between tourism and other discretionary expenditures. Given this, hypothetical experiments were adopted in this research.

To enhance the credibility of hypothetical scenarios in experiments, we selected Shenzhen, a major city in China, as the research site. First, as of this writing, Shenzhen has not launched any universal cash payment policies. The absence of such programs minimizes confounding variables that could distort consumption patterns due to substitution or inter-temporal effects from past stimulus practices. Second, Shenzhen shares cultural and geographical proximity with Hong Kong, where multiple cash handouts were distributed during 2021–2023. The similarity ensures that Shenzhen residents can relate to and engage realistically with our hypothetical scenarios. Meanwhile, choosing Shenzhen facilitates future comparative studies between actual responses to stimulus checks in Hong Kong and those predicted through our hypothetical experiments in Shenzhen.

Additionally, we crafted the notification posters for the hypothetical programs by adopting formats typically utilized in local public coupon distribution efforts to enhance their realism (see the supplemental questionnaire).

Experiment design and variables definitions

The experiment design follows the three-step mental accounting framework developed in the previous section. One point should be made clear: We merged the first two steps—Categorization and Budgeting—into a single measurement. This is because the two, while conceptually distinct, are closely linked in practice and can be challenging to disentangle behaviorally. Asking participants about the two steps separately runs the risk of prompting unnatural responses. A more streamlined approach is to have participants reveal their combined categorization and budgeting heuristics through an allocation decision. By distributing the provided stimulus across spending categories, their choices demonstrate how they implicitly segregate funds mentally and set budgetary boundaries concurrently within those designated domains. Therefore, the three-step mental accounting process is examined in two studies.

Study 1: The allocation task employs a between-subjects factorial experimental design. Participants are randomly assigned to one of eight groups based on the combinations of three factors: stimulus check size (CN ¥5,000 vs. CN ¥1,000, approximately US$720 vs. US$140), source-framing (rebate vs. bonus), and economic context (pandemic-induced recession vs. non-pandemic-induced recession). They need to distribute the hypothetical stimulus funds across given spending categories.

Study 2: The trade-off task employs a within-subjects experimental design. Participants are presented with two scenarios representing surplus and non-surplus conditions in tourism mental accounts and are asked to make consumption decisions accordingly. Specifically, in the surplus condition, where only the tourism mental account has a balance, participants decide whether to use funds budgeted for tourism to cover unexpected expenses in other categories. In the non-surplus condition, where funds in the tourism mental account are depleted, participants are asked if they are willing to pay for ambiguous expenses that combine tourism with other categories.

The variables utilized in each experiment are presented in Table 1.

Table 1.

List of Experiments and Variables

Study 1: AllocationDependent Variable: Proportion of Funds Allocated to Tourism (%)
Variables	Definition, Measurement, and Value
Source	The source of stimulus checks: rebate or bonus. “Rebate-sourced” checks are described as tax rebates, while “bonus-sourced” checks are labeled as government support funded by fiscal surplus.
Size	The amount of stimulus checks: CN¥5,000 (large, approx. US$720) or CN¥1,000 (small, approx. US$140). The large amount is roughly equivalent to one-third of the average monthly wage in Shenzhen (CN¥13,730 in 2022, approxi. US$1,970).
Context	Economic contexts of fiscal stimulus: non-pandemic-induced recession or pandemic-induced recession. The latter is specifically set against the backdrop of the recent Covid-19 pandemic.
QOL	Tourism’s contribution to quality of life, measured via a five-point Likert scale across four dimensions: physical relaxation, mental health, social connection, and environmental involvement. The scale is developed by integrating and simplifying dimensions from existing literature (Dolnicar et al., 2012; Uysal et al., 2016).
HRP	Health risk perception, assessed by individual evaluation of outbreak threats, including vulnerability and severity based on prior research (Duncan et al., 2009; Yıldırım & Güler, 2022). Specifically, vulnerability pertains to the likelihood of disease contraction, while severity refers to its potential consequences.
Study 2: Trade-OffsDependent Variable: Consumption Intention (Yes or No)
Variables	Definition, Measurement, and Value
Importance	(Surplus Scenario of Tourism Mental Accounts) The relative importance of tourism expenditure against other spending categories. Participants rank the priority of all spending categories on a three-level scale. The relative importance score is calculated as the difference between tourism’s rank and another category’s (e.g., tourism ranked 3 vs. education ranked 2 yields a score of 1), producing a five-level scale (-2 to 2). In this scenario, participants decide whether to reallocate the surplus funds in tourism mental accounts to cover unexpected expenses in other categories (e.g., purchasing professional books).
Ambiguity	(Non-Surplus Scenario of Tourism Mental Accounts) The ambiguous nature of expenses combining tourism with non-tourism categories, such as a day tour (pure tourism) versus a study tour (blending education and tourism). In this scenario, participants choose between pure tourism and ambiguous expenses to assess how ambiguity affects fund reallocation. This design follows the methodology used by Cheema and Soman (2006). They classified food delivery as a clear food expense, a music concert as a clear entertainment expense, and dinner at a restaurant with live entertainment as an ambiguous option.

Note. Currency conversions are based on the exchange rate of 6.986 (1 USD = 6.986 CNY) as of the end of 2022. Source: International Monetary Fund (IMF), https://data.imf.org.

Controls for order and learning effects

In Study 1, the between-subjects design inherently controls for cross-condition contamination, as each participant is exposed to only one experimental condition without sequential task interference. Additionally, the order of given spending categories is randomized across participants to eliminate potential priming effects from category sequencing.

In Study 2, both scenarios and expenditure options are presented in randomized order to prevent participants from developing response patterns across scenarios.

Manipulation checks are conducted on separate cohorts who do not participate in formal surveys, ensuring that participants’ interpretation of key constructs is not confounded by prior experimental exposure.

Data Collection

Pilot survey

In order to establish a comprehensive choice set for the subsequent experiments, a pilot survey was conducted in September 2023. Employing a convenience sampling method, we interviewed 26 Shenzhen residents, aged between 18 and 72, to assess their spending plans under a hypothetical financial incentive scenario of CN¥3,000 (approximately US$430). This preliminary survey yielded five primary expenditure categories (i.e., Utilitarian, Education, Sports & Health, Leisure & Entertainment, and Tourism see the supplemental Table S1 for details), allowing the subsequent experiments to encompass the main expenditure categories typically considered by most individuals (Crouch et al., 2007).

Manipulation checks

To validate experimental manipulations, we recruited 74 participants through convenience sampling (distributed via WeChat groups) and administered the questionnaire through Qualtrics in September 2023. Participants were presented with hypothetical fiscal stimulus programs depicted in posters and were asked to report their perceptions regarding the size, source, and context of these programs. Chi-square contingency table analyses were conducted to assess whether participants’ perceptions aligned with the objective information provided in the experiments. Results showed significant correspondence between the perceived and objective information for all three manipulated factors: the size (c² = 89.47, p < .001), source (c² = 79.18, p < .001), and context (c² = 69.26, p < .001) of the fiscal stimulus programs. That indicated that participants accurately discerned the characteristics of the scenarios as intended by the experimental design.

Then, we tested whether participants consistently perceived expense ambiguity with our definition. Participants rated three pairs of spending options (Tourism vs. Tourism combined with Education/Sport & Health/Leisure & Entertainment) on a 7-point scale, where 1 indicated a pure tourism product and 7 indicated a pure other product. For instance, participants rated “a day tour in an ancient city” and “a study tour for history learning” between Tourism and Education. Scores closer to 1 indicated a distinctly tourism-related expense, while scores further from 1 reflected greater ambiguity in expense classification. Paired t tests showed significant differences in the ratings for each pair (Tourism vs. Education: t = -10.60, p < .001; Tourism vs. Sport & Health: t = -2.95, p < .05; Tourism vs. Leisure & Entertainment: t = -2.27, p < .05). This supported the definition that combinations of tourism and other categories constitute ambiguous expenses.

Formal survey

The formal survey collected data from September to December 2023 through a market research firm Kantar (980 responses). The investigation employed a stratified sampling method based on the Shenzhen population structure to ensure sample representativeness. Invalid responses (roughly 21% of total responses) were excluded for (1) failing attention checks, (2) more than 10% incomplete responses, or (3) implausibly short completion times (< 2.7 mins, one-third of the median duration). Additionally, a complementary on-site survey (23 participants) was conducted to enhance subgroup balance. After these adjustments, the formal survey yielded 797 valid responses.

The demographic profile of the survey sample aligns closely with Shenzhen’s population characteristics (see the supplemental Table S2 for details). The sample comprises 54.8% males and 45.2% females, closely mirroring Shenzhen’s census data (55.04% male and 44.96% female). The age distribution shows a higher concentration of young and middle-aged adults (18–49 years old, 85.2% of the sample) than the city’s overall age structure (15–49 years old, 69.7% of the population). Education levels in the sample show a higher concentration of individuals with higher education: 82.4% have a college degree or above, compared to 32.12% in Shenzhen’s census. The sample has a slightly higher employment rate (72.1%) than the city’s overall employment rate (67.18%) and shows a higher proportion of married individuals (75.3%) compared to the city’s overall married population (64.14%). Income levels in the sample are relatively concentrated, with 46.2% earning between CN¥8,000 and CN¥18,000 (approximately US$1,140 and US$2,570) per month, close to the city’s average monthly wage of CN¥13,730 (approximately US$1,970) (Shenzhen Municipal Bureau of Statistics, 2023). Overall, the sample is representative of Shenzhen’s population in terms of gender, income, and employment, but is skewed toward younger, more educated, and family-oriented individuals. This divergence may stem from the survey’s age inclusion criterion (18+ years) and limited accessibility to elderly populations through online surveys.

Modeling Approach

The model for Study 1

Study 1 examines the proportional distribution of stimulus funds across various spending categories, focusing on the percentage allocated to tourism expenditure. As the dependent variable represents proportions, it exhibits compositional properties constrained by (1) non-negativity, as component shares must be within the interval [0, 1]; and (2) closure, as the components sum to unity, making the response variable as a vector of several interdependent fractional values. Such characteristics constrain the data to a simplex S^D rather than the Euclidean space R^D. Therefore, linear regression is unsuitable for analyzing compositional data due to potential violations of two crucial assumptions: normal distribution and constant variance (Ferrer-Rosell et al., 2015). Recent studies propose interpretable alternatives: Beta regression for single proportions and Dirichlet regression for multi-category proportions (Hanretty, 2021).

In Beta regression, the conditional model for the mean $μ$ and the dispersion parameters $ϕ$ is usually assumed to be linear on the logit transformed scale:

y = B e t a (μ, ϕ)

(1)

η = \log (\frac{μ}{1 - μ}) = X β

(2)

where $η$ is known as the linear predictor, $β$ is a vector of parameters to be estimated and X is the matrix of covariate values.

The Dirichlet distribution, as the generalization of the beta distribution, is mathematically represented as

f (y; α) = \frac{1}{B (α)} \prod_{i = 1}^{C} y_{i}^{α_{i - 1}}, for y_{i} \geq 0, \sum_{i = 1}^{C} y_{i} = 1

(3)

where:

$y = (y_{1}, y_{2}, . . ., y_{c})$ is a C-dimensional vector denoting the proportions of c components; $α = (α_{1}, α_{2}, . . . α_{c})$ is the vector of shape parameters that determine the distribution’s shape; $B (α)$ is the normalization constant (i.e., the Beta function) of the Dirichlet distribution and can be expressed as $\prod_{i = 1}^{C} Γ (α_{i}) / Γ (\sum_{i = 1}^{C} α_{i})$ , where $Γ (\cdot)$ is the gamma function.

It is important to clarify that Dirichlet regression requires a strict interval between (0, 1) for the response variables. However, in certain cases, participants can allocate none or all of the stimulus checks to a specific category. The presence of 0 and 1 poses challenges in the equations. Specifically, when $\log (y_{i} =) = - \infty$ is calculated, it results in an undefined value of –∞. To address this issue, we employed a modification of the transformation proposed by Smithson and Verkuilen (2006), considering N as the total number of observations,

y^{*} = \frac{y (N - 1) + 1 / K}{N}

(4)

As N → ∞, the compression vanishes. Larger data sets are less affected by this transformation.

The model for Study 2

Study 2 is to explore the interaction between tourism and other mental accounts. This study involves assessing the rigidity and malleability of mental accounts as measured by the discrete outcome variable of consumption intention (i.e., whether the individual intends to consume within a given scenario and option). Given the dichotomous nature of the dependent variable, a binary logistic regression analysis would be an appropriate statistical method to model the relationship. A regression model can be established

y_{i j} = \log (\frac{P}{1 - P}) = β_{0} + β_{1} x_{i j} + β_{2} z_{i j}

(5)

where:

For each participant i and option j, x_ij denotes the relative importance in the tourism surplus scenario or expense ambiguity in the tourism non-surplus scenario. z_ij refers to the dummy variable of the economic context. $β_{0}$ , $β_{1}$ , $β_{2}$ represent the intercept term and coefficients assigned to the variables x_ij and z_ij, respectively.

It is important to note that each participant is given repeated options across different pairs of tourism and other expenditure categories to eliminate biases caused by specific preferences toward certain categories. As a result, the observations from the same individual may exhibit internal correlations across the different options. Given the repetitive measurement structure, using the maximum likelihood estimation (MLE) is inappropriate since it assumes independence of observations. Instead, a quasi-likelihood estimation method, known as the generalized estimating equations (GEE), is more suitable in this case.

Results and Discussion

Results and Discussion of Study 1

Construct validation of QOL and HRP

Before assessing the performance of models, it is crucial to validate the constructs of QOL and HRP. Exploratory Factor Analysis (EFA) revealed that the QOL items loaded onto two factors: Intrinsic Health (i.e., physical and mental health) and Extrinsic Connectivity (i.e., social and environmental engagement), explaining 47.1% of the variance. The HRP items loaded onto two factors: Vulnerability and Severity, explaining 59% of the variance. The factor loadings were significant for both constructs. The Confirmatory Factor Analysis (CFA) results further validated these structures, with acceptable model fit indices for QOL (χ² = 7.033, df = 1, p = 0.008) and HRP (c² = 1.792, df = 1, p = 0.181). Reliability analyses showed good internal consistency for both scales (QOL: Cronbach’s α = 0.768; HRP: Cronbach’s α = 0.672). These results confirm the validity and reliability of the QOL and HRP constructs for further analysis.

Model performance assessment

To examine the impact of external policy attributes (H1, H2) of stimulus checks on tourist expenditure, we initially utilized Dirichlet regression with the alternative parameterization via the “DirichletReg” package in R, designating “Utilitarian” as the baseline category. Subsequently, we narrowed our focus solely on tourism expenditure proportions and conducted Beta regression to investigate the role of internal individual attributes (H3, H4) in tourism spending, while also validating H1 and H2 from the Dirichlet regression. The results are summarized in Table 2.

Table 2.

Coefficient Estimates of the Dirichlet and Beta Regressions

	Coefficient Estimates
	Dirichlet Regression(Reference Category: Utilitarian)				Beta Regression
Variables	S&H	E	L&E	T	T
Intercept	–1.88^***	–1.12^***	–1.06^***	–1.13^***	–2.58^***
Source (Reference: Rebate)
Bonus	.32^***	.20^*	.28^***	.34^***	.31^***
Size (Reference: Small)
Large	.29^***	.22^**	.10	.34^***	.23^**
Context (Reference: Non-pandemic)
Pandemic	.10	.07	.11	.08	–1.28^**
Gender (Reference: Male)
Female	–.06	–.06	–.07	–.17^*	–.07
Age (Reference: 18–29)
30–39	–.52^***	–.56^***	–.79^***	–.64^***	–.39^***
40–49	–.14	–.13	–.62^***	–.25	–.11
50–59	.28	.32	–.12	–.17	–.27
60 or above	.33	.13	–.05	–.29	.38
Family (Reference: Single)
Married without children	–.01	.11	.11	–.04	–.01
Married with children	.11	.51^***	.17	.27^*	.16
Single parent family	–.45	.08	–.10	–.16	–.02
Other	–.07	–.09	.64	–.09	–.07
Income (Reference: below CN¥5,000, approx. US$720)
CN¥5,000–8,000	.11	–.07	.08	–.10	–.19
CN¥8,000–1,8000	.44^**	–.13	.35^*	.33^*	.09
Above CN¥1,8000	.36^*	–.03	.21	.04	–.20
Education (Reference: High school or below)
College	.29	.04	.06	.01	–.08
Master’s degree or above	.02	.17	–.09	–.22	–.31
Occupation (Reference: Employed)
Self-employed	.12	.10	–.04	.08	.08
Retired	.20	–.28	–.13	.35	.55^*
Unemployed	–.56	–.67	–.43	–.75	–.45
Students	.27	.10	–.10	.39	.38
Other	–.29	–.72^***	–.42^*	–.30	–.26
QOL					.28^***
Context * QOL					.29^*
HRP					.06
Model summary	Links: Logit (Means) and Log (Precision)Parametrization: alternativeAIC: -7457 BIC: –7022pseudo-R² = 0.325Log-likelihood/d.f.: 38224/93Constant: 1.25, p < .001				AIC: –1569BIC: –1442 Log-likelihood/d.f.:812/27pseudo-R² = 0.14Φ coeff. = 2.52, p < .001

Note. N = 797, S&H = Sports & Health, E = Education, T = Tourism, L&E = Leisure & Entertainment, QOL = tourism’s contribution to quality of life, HRP = health risk perception. Currency conversions are based on the exchange rate of 6.986 (1 USD = 6.986 CNY) as of the end of 2022. Source: International Monetary Fund (IMF), https://data.imf.org.

p < .05, **p < .01, ***p < .001.

To assess the model’s performance, we calculated the pseudo-R² based on model likelihoods (Smithson & Verkuilen, 2006). The measure is defined as

R^{2} = 1 - {[\frac{L ({\hat{θ}}_{0})}{L ({\hat{θ}}_{1})}]}^{\frac{2}{n}}

(6)

where $L ({\hat{θ}}_{0})$ and $L ({\hat{θ}}_{1})$ are the likelihoods of the constant and the covariate models, respectively; n is the sample size. The model achieved a pseudo-R² of .32, showing a 32.5% improvement in the log-likelihood compared to the constant model. The likelihood ratio test further validated the enhanced fit (c² = 307.68, p < .05).

The residual plot (see the supplemental Figure S1) shows a generally random distribution of residuals with no clear patterns, indicating a good model fit and limited systematic bias in capturing spending behavior across categories. Additionally, the Q-Q plot (see the supplemental Figure S2) reveals that while most data points align closely with the theoretical normal distribution line, deviations in the tails suggest the presence of heavy-tailed residuals, implying a higher frequency of extreme values than expected. Given this, we further conducted a robustness test to validate the reliability of the model estimates.

In the robustness test, a log-ratio analysis was applied. The initial step in the log-ratio method involves replacing the zero values. Following Ferrer-Rosell et al. (2015), we used the same zero replacement strategy suggested by Martín-Fernández et al. (2003), namely replacing $x_{i d} = 0$ with:

x_{i d}^{'} = k δ_{i d} with 0 < k < 1

(7)

where $δ_{i d}$ is the smallest detectable proportion for individual i and component d. Next, non-zero $x_{i d}$ values need to be reduced in order to preserve the unit sum and the ratios among non-zero components:

x_{i d}^{'} = x_{i d} (1 - \sum_{x_{i d} = 0} x_{i d}^{'})

(8)

Subsequently, the additive log-ratio transformation (ALR) was employed, as utilized by previous studies (Ferrer-Rosell et al., 2015; Fry et al., 1996).

\begin{array}{l} y_{i d} = \ln (x_{i d} / x_{i D}) = \ln (x_{i d}) - \ln (x_{i D}) with d = 1, 2, 3, . . . D - 1 \end{array}

(9)

After transforming the data, a multi-factor ANOVA for each category was conducted. The results, along with a comparison to the Dirichlet regression, are presented in Table 3.

Table 3.

Significance Comparison of Main Explanatory Variables Between Dirichlet Regression and Log-Ratio Analysis (Reference Category: Utilitarian)

	Dirichlet Regression		Log-Ratio Analysis
	Variables		Variables
Category	Source (Bonus)	Size (Large)	Source (Bonus)	Size (Large)
S&H	√	√	√	√
E	√	√	×	○
L&E	√	×	√	×
T	√	√	√	√

Note. S&H = Sports & Health; E = Education; L&E = Leisure & Entertainment; T = Tourism.

√ indicates that the coefficients are significant at the 0.001, 0.01, or 0.05 level. ○ indicates that the coefficients are significant at the 0.1 level. × indicates that the coefficients are insignificant at all levels.

Table 3 shows consistent significance between the two methods, except for Education. In the Dirichlet regression, both the size and source of stimulus checks show significance in determining funds allocation to Education. However, in the log-ratio analysis, stimulus size’s significance diminishes, and the source factor loses significance in Education. The differing significance may be attributed to the fact that the log-ratio approach emphasizes components with higher percentages and variation, whereas Education exhibits relatively lower magnitude and variability, making its subtle response to the independent variable less detectable. To assess the Dirichlet model’s sensitivity, we systematically excluded recession context and demographic variables. Interestingly, the significance of the two main independent variables was consistent with the initial Dirichlet regression, even in the Education category. Therefore, we consider the results of the Dirichlet regression to be robust.

Analysis and Discussion

The results reveal significant correlations between the size and/or source of stimulus funds and the allocation proportions in four categories (see Figure 2). Notably, the significance in the Tourism category is particularly robust, as confirmed by consistent results across several alternative models, supporting H1 and H2.

Specifically, the bonus source correlates positively with the proportion of stimulus funds allocated to the four categories. It echoes the mental accounting perspective of money non-fungibility, with bonus framing provoking positive emotions that are more likely to facilitate discretionary spending (Levav & McGraw, 2009). In contrast, the rebate-framing emphasizes the return of loss, which can result in hedonic avoidance by creating “physical or psychic distance” from the windfall (Duhachek, 2005), thus leading individuals to consume utilitarian goods subsequently. This finding suggests that individuals tend to match their cognitive and emotional associations with the source of funds to their spending patterns.

Regarding the size of stimulus checks, larger funds (approximately a third of average monthly income in this study) are found to significantly increase expenditures in Sports & Health, Education, and Tourism, whereas their impact on Leisure & Entertainment remains insignificant. This finding aligns with previous research indicating that the marginal propensity to consume increases with the payment size (Fuster et al., 2021; Souleles, 1999); however, it exhibits nuanced category-specific variations rather than following a uniform pattern. Tourism, as well as Sports & Health, and Education, often involves substantial upfront costs. Larger funds facilitate the fulfillment of such aspirations, making them more responsive to larger stimulus amounts. In contrast, Leisure & Entertainment, ranging from free park visits to high-end concerts, provides flexible options for varying budgets, thereby reducing the sensitivity to stimulus size differences.

Overall, the Dirichlet regression results suggest that stimulus checks can potentially offset the shifts in consumption patterns induced by recessions. During economic downturns, consumers tend to adjust their spending behaviors to conserve economic resources and ensure financial prudence (Koos et al., 2017). Typically, they prioritize essential expenditures while deprioritizing positional and discretionary items (Sarmento et al., 2019). The bonus framing of the stimulus checks, however, seems to facilitate a mental decoupling from these austerity norms. Upon receiving substantial windfalls, they use the stimulus funds to fulfill postponed desires while maintaining essential spending through regular income. Thus, the increase in tourism spending likely reflects the release of pent-up demand and the framing-induced permission to indulge. These findings highlight that stimulus checks, especially those framed as bonuses and substantial size, can restore discretionary spending affected by economic downturns and support economic recovery.

The results exhibit significant individual-level differences. Compared to younger individuals aged 18–29, participants aged 30–39 tend to allocate less to the four categories. Similarly, middle-aged respondents between 40–49 years old direct less toward Leisure & Entertainment than the baseline. The marital and family status also plays a role, as married individuals with child(ren) allocate a larger share to Education and Tourism than single people. Income level is another significant factor, with the middle-income group spending more on Sports & Health, Leisure & Entertainment, and Tourism than lower-income groups. However, the high-income group only differs in allocating more to Sports & Health, showing no other significant differences. Interestingly, occupation and educational background generally do not impact spending patterns, except those in “other” unspecified occupations who directed more funds toward Education and Leisure & Entertainment. Respondents in this “other” category are not part of the typical population, such as graduates preparing for further study.

The Beta regression results suggest that the contribution of tourism to quality of life (QOL) significantly impacts tourism spending out of stimulus checks (β = .28, p < .001), supporting H3. That aligns with previous studies that underscore the role of QOL in shaping traveling decisions under adverse conditions, which found QOL to be the primary driver of crisis resilience in tourist demand (Bronner & de Hoog, 2016). Furthermore, the pandemic context seems to amplify the significance of QOL, as evidenced by the significantly positive interaction between the economic context and QOL (β = .29, p < .05). The finding echoes the accelerating paths of tourism demand under pandemic conditions, where psychological and physical pressures enhanced the travel desire and escape motive (E. E. K. Kim et al., 2022), resulting in “revenge travel.”

The context of economic recession does not significantly influence the spending allocation across categories. However, when we specifically examined tourism spending and considered psychological factors, the pandemic context does tend to reduce tourism spending (β = −1.28, p < .05). Interestingly, the observed reduction in tourism spending does not appear to be significantly correlated with health risk perception (HRP), as there is no substantial correlation between tourism spending and HRP (β = .06, p > .05), rejecting H4.

A plausible explanation for the discrepancy lies in the strategic timing of stimulus interventions. According to Steel and Harris (2020), fiscal stimulus programs were typically launched during the stable phase or post-lockdown when the pandemic has been controlled to an acceptable level. The well-timed interventions may have mitigated the fear and anxiety associated with health crises, thus reducing the negative impact on travel decisions. Meanwhile, during pandemics, innovations in tourism products, such as contactless services (Rathjens et al., 2025), nature-based tourism (Qiu et al., 2021), and virtual tourism (Lu et al., 2022), have been implemented to minimize health risks and make travel safer. Furthermore, according to Boto-García and Baños-Pino (2023), individuals’ travel habits and preferences, shaped by past travel patterns, exhibit a resilience to crises that can outweigh the perceived health risks. The resilience is particularly evident among females, parents with children, educated and wealthy individuals, and residents of densely populated urban areas, which align closely with the demographics represented in our sample.

Regarding the significance of the pandemic context observed in the Beta regression, we posit that this might be attributed to external restrictions imposed to curb the virus’s spread, which likely hinder the convenience of tourism activities and reduce the accessibility to tourism destinations (Moreno-Luna et al., 2021). Additionally, concerns about service quality, potentially exacerbated by workforce shortages (Jones et al., 2024), may play a role. These factors warrant further examination in future research.

Results and Discussion of Study 2

Table 4 presents results from Study 2. Specifically, the statistically significant coefficients (Importance 1: β = −.46, p < .001; Importance 2: β = −.21, p < .01) underscore the role of perceived importance in shaping the rigidity of mental accounts. When the alternative option is considered less important than tourism activities, the mental account designated for tourism exhibits a higher level of inflexibility because of greater self-control imposed to ensure the funds exclusively used for tourism purposes, thereby supporting H5. The finding aligns with the concept of goal-derived categories (Ratneshwar et al., 1996), where the subjective valuation of a goal directly influences the degree of effort and commitment directed toward its realization. The high perceived importance of tourism over alternative expenditures acts as a psychological anchor, reinforcing the boundaries of the mental account designated for tourism-related expenses. The phenomenon underscores the capacity of mental accounting to serve as a powerful self-regulatory tool, enabling consumers to prioritize and safeguard resources for their cherished goals (Levin, 1998).

Table 4.

Coefficient Estimates of the Binary Logistic Regression with GEE

Variable	Coefficient	SE	OR	Significance	Wald Chi-Square
Tourism Surplus Scenario (Independent, QIC = 4253.360; QICC = 4249.770)
Intercept	.72^***	.10	2.05	< .001	49.31
^aImportance = 1 (Reference: 3)	–.46^***	.10	.63	< .001	20.37
^aImportance = 2 (Reference: 3)	–.21^**	.09	.81	.02	5.37
Context (Reference: Pandemic)	–.14	.10	.87	.19	1.75
Tourism Non-Surplus Scenario (Unstructured, QIC = 6078.893; QICC = 6079.959)
Intercept	–.17^***	.05	.84	<.001	13.83
^bAmbiguity (Reference: 1)	–.79^***	.08	.46	<.001	104.18
Context (Reference: Pandemic)	–.03	.04	.97	.47	.48

Note. N = 797; ^aImportance rated as: 1 = Tourism less important, 2 = Equal importance, 3 = Tourism more important, compared to the paired expenditure category. ^bAmbiguity scored as: 1 = Ambiguous expense, indicating a mix of tourism and paired categories, 2 = Unambiguous expenses, clearly delineating pure tourism products/services.

p < .01. ***p < .001.

The ambiguity of expenses is also significant (β = −.79, p < .001), supporting H6. It indicates that expenses blending tourism with other purposes can enhance the intention to consume compared to pure tourism-related products and services. The finding resonates with the conclusion of Ratneshwar et al. (1996), who observed heightened cross-category consideration in ambiguous goal settings. With ambiguous expense descriptions, individuals can circumvent their budgetary rules and rationalize the allocation of expenses to surplus accounts, engaging in what can be deemed “creative bookkeeping” (Cheema & Soman, 2006). This cognitive flexibility is exemplified by reasons like “that business meeting is really important” to couple an excursion with a more justifiable goal (Rajagopal & Rha, 2009). This mechanism demonstrates the malleability of mental accounting in response to situational cues and reveals a subtle cognitive bias that allows consumers to reconcile their spending with their internal budgetary rules. Consequently, mental accounting, as a self-regulatory mechanism, may lose its effectiveness sometimes as consumers can exploit the cognitive bias to bypass the constraints of mental accounting and achieve their desired spending goals.

Conclusion

The study uses the mental accounting framework to explore the decision-making process of spending cash-based stimuli across different levels. The findings reveal that policy-level factors (i.e., size and source), along with firm-level considerations (i.e., product presentation), and individual-level perceptions (i.e., tourism’s importance and contribution to quality of life) jointly influence the allocation of stimulus funds toward tourism. Building on these findings, the research contributes to the existing literature in three ways.

Theoretically, it offers a novel perspective on the relationship between tourism spending and income by moving beyond traditional economic assumptions of money fungibility and rational consumer behavior. The study examines the interaction between external framing and internal cognitive processes by emphasizing fiscal stimuli as a unique income and including inter-category trade-offs. It demonstrates that mental accounting principles effectively explain the allocation of windfall income toward tourism.

Practically, the study bridges fiscal policies, behavioral economics, and tourist spending behavior, paving the way for more effective industry response and policy formulation. From a marketing perspective, the findings suggest that strategies such as targeting frequent tourists, developing multi-purpose tourism products, and offering relaxation and therapeutic experiences can stimulate funds allocated to tourism, especially during pandemics. Regarding policy designs, fiscal stimulus measures with large amounts and positive framing may effectively increase tourism expenditure. As tourism becomes an integral part of people’s lifestyles (Nieto García et al., 2020), its perceived importance and contribution to quality of life will continue to increase. In this context, cash-based economic stimulus policies hold the potential to boost tourism spending under exceptional circumstances.

Methodologically, while we followed the hypothetical scenario-based experimental approach employed by Crouch et al. (2007) and Dolnicar et al. (2008), our study extends their work in two ways. First, we contextualize experiments within specific fiscal stimulus policies implemented during different types of recessions and allow for the variations of policy design factors. This approach provides a more nuanced evaluation of how such policies impact tourism spending under varying conditions. Second, we examine the dynamics of fund allocation over time rather than focusing solely on a one-time budget. That reflects real-world spending decisions, which frequently require adjustments as circumstances evolve.

While contextualized within a single city, this study’s findings hold broader implications. As China’s largest immigrant hub, Shenzhen is a microcosm of domestic tourism convergence, reflecting diverse tourism spending behaviors. Additionally, the city’s population mirrors the demographics of many emerging economies characterized by a younger population structure (Rai & Garg, 2024). Given Shenzhen’s representativeness of the domestic market and the resonance of emerging markets, our findings offer insights into tourism demand prediction, especially in scenarios where China or similar markets undergo positive income shocks. These insights are particularly valuable for major international tourist destinations that cater to such source markets, enabling them to respond swiftly and effectively.

However, it is important to acknowledge several limitations that could be addressed in future research. First, although the findings from a specific research site could extrapolate to similar contexts, they do not fully capture the impact of cultural differences, such as distinct saving and spending paradigms. Including cross-cultural perspectives in future research would enhance the generalizability of these findings across diverse populations. Second, the current framework focuses on behavioral mechanisms without including macroeconomic factors (e.g., liquidity constraints, precautionary savings motives, and consumer sentiment). A dual-process model that combines behavioral heuristics with rational optimization could provide more nuanced predictions. Last, methodologies based on hypothetical scenarios may oversimplify the complexities of real-world fiscal dynamics, potentially leading to discrepancies between anticipated and actual consumer responses to fiscal stimulus programs. Future studies could enhance validity by tracking tourism spending on comparable stimulus initiatives to gather revealed preferences data.

Supplemental Material

sj-docx-1-jht-10.1177_10963480251338215 – Supplemental material for Tourism Spending in Response to Stimulus Checks: Insights From Mental Accounting

Supplemental material, sj-docx-1-jht-10.1177_10963480251338215 for Tourism Spending in Response to Stimulus Checks: Insights From Mental Accounting by Linlin Nie and Haiyan Song in Journal of Hospitality & Tourism Research

Footnotes

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

ORCID iDs

Linlin Nie

Haiyan Song

Supplemental Material

Supplemental material for this article is available online.

Author Biographies

Linlin Nie (e-mail: linlin.nie@connect.polyu.hk) is a PhD candidate at the School of Hotel and Tourism Management, The Hong Kong Polytechnic University, Hong Kong SAR, PR China.

Haiyan Song (e-mail: haiyan.song@polyu.edu.hk) is the Mr and Mrs Chan Chak Fu Professor in International Tourism, Director of Research Centre for Digital Transformation of Tourism at the School of Hotel and Tourism Mangement, The Hong Kong Polytechnic University, Hong Kong SAR, PR, China.

References

Baker

S. R.

Farrokhnia

R. A.

Meyer

Pagel

Yannelis

(2023). Income, liquidity, and the consumption response to the 2020 economic stimulus payments. Review of Finance, 27(6), 2271–2304. https://doi.org/10.1093/rof/rfad010

Berbekova

Assaf

A. G.

Uysal

(2025). Interdisciplinary approach to tourism demand modeling: Quality of life indicators. Journal of Hospitality & Tourism Research, 49(3), 635–639. https://doi.org/10.1177/10963480241229238

Boto-García

Baños-Pino

J. F.

(2023). Deep habits and travel resilience after COVID-19. Tourism Economics, 29(2), 543–550. https://doi.org/10.1177/13548166211052139

Boto-García

Muñiz-Fernández

Pérez

(2024). Windfall money and outbound tourism: A natural experiment from lottery winnings. Tourism Economics, 30(5), 1257–1280. https://doi.org/10.1177/13548166231203042

Brida

J. G.

Tokarchuk

(2015). Keeping mental budgets: Visitors’ spending at a Christmas market. Tourism Economics, 21(1), 67–82. https://doi.org/10.5367/te.2014.0437

Bronner

de Hoog

(2016). Crisis resistance of tourist demand: The importance of quality of life. Journal of Travel Research, 55(2), 190–204. https://doi.org/10.1177/0047287514541006

Byun

Jang

(2015). Effective promotions for membership subscriptions and renewals to tourist attractions: Discount vs. Bonus. Tourism Management, 50, 194–203.

Cheema

Soman

(2006). Malleable mental accounting: The effect of flexibility on the justification of attractive spending and consumption decisions. Journal of Consumer Psychology, 16(1), 33–44. https://doi.org/10.1207/s15327663jcp1601_6

Chen

C.-M.

Chen

Y.-C.

(2018). Income source effect on retiree’s tourism behavior. Annals of Tourism Research, 68, 4–6. https://doi.org/10.1016/j.annals.2017.10.002

10.

Chen

Lehto

X. Y.

Cai

(2013). Vacation and well-being: A study of Chinese Tourists. Annals of Tourism Research, 42, 284–310. https://doi.org/10.1016/j.annals.2013.02.003

11.

Chow

C. K. W.

Tsui

W. H. K.

(2021). Airport subsidies and domestic inbound tourism in China. Annals of Tourism Research, 90, Article 103275. https://doi.org/10.1016/j.annals.2021.103275

12.

Crouch

G. I.

Oppewal

Huybers

Dolnicar

Louviere

J. J.

Devinney

(2007). Discretionary expenditure and tourism consumption: Insights from a choice experiment. Journal of Travel Research, 45(3), 247–258. https://doi.org/10.1177/0047287506295912

13.

Dardis

Soberon-Ferrer

Patro

(1994). Analysis of leisure expenditures in the United States. Journal of Leisure Research, 26(4), 309–321. https://doi.org/10.1080/00222216.1994.11969964

14.

Dolnicar

Crouch

G. I.

Devinney

Huybers

Louviere

J. J.

Oppewal

(2008). Tourism and discretionary income allocation. Heterogeneity among households. Tourism Management, 29(1), 44–52. https://doi.org/10.1016/j.tourman.2007.04.008

15.

Dolnicar

Yanamandram

Cliff

(2012). The contribution of vacations to quality of life. Annals of Tourism Research, 39(1), 59–83. https://doi.org/10.1016/j.annals.2011.04.015

16.

Duhachek

(2005). Coping: A multidimensional, hierarchical framework of responses to stressful consumption episodes. Journal of Consumer Research, 32(1), 41–53. https://doi.org/10.1086/426612

17.

Duncan

L. A.

Schaller

Park

J. H.

(2009). Perceived vulnerability to disease: Development and validation of a 15-item self-report instrument. Personality and Individual Differences, 47(6), 541–546. https://doi.org/10.1016/j.paid.2009.05.001

18.

Epley

Mak

Idson

L. C.

(2006). Bonus of rebate?: The impact of income framing on spending and saving. Journal of Behavioral Decision Making, 19(3), 213–227. https://doi.org/10.1002/bdm.519

19.

Fagereng

Holm

M. B.

Natvik

G. J.

(2021). MPC heterogeneity and household balance sheets. American Economic Journal: Macroeconomics, 13(4), 1–54. https://doi.org/10.1257/mac.20190211

20.

Fang

Zhu

Jiang

(2022). Estimated effects of economic policies for COVID-19 on the leisure and recreation industry under public health interventions. Current Issues in Tourism, 25(1), 1–13. https://doi.org/10.1080/13683500.2021.1970118

21.

Ferrer-Rosell

Coenders

Martínez-Garcia

(2015). Determinants in tourist expenditure composition—The role of airline types. Tourism Economics, 21(1), 9–32. https://doi.org/10.5367/te.2014.0434

22.

Flavin

M. A.

(1981). The adjustment of consumption to changing expectations about future income. Journal of Political Economy, 89(5), 974–1009. https://doi.org/10.1086/261016

23.

Fry

J. M.

Fry

T. R.

McLaren

K. R.

(1996). The stochastic specification of demand share equations: Restricting budget shares to the unit simplex. Journal of Econometrics, 73(2), 377–385. https://doi.org/10.1016/S0304-4076(95)01727-5

24.

Fuster

Kaplan

Zafar

(2021). What would you do with $500? Spending responses to gains, losses, news, and loans. The Review of Economic Studies, 88(4), 1760–1795. https://doi.org/10.1093/restud/rdaa076

25.

Geng

Shi

C. M.

Song

Z. M.

(2022). Evaluating Hong Kong consumption voucher scheme. Department of Economics, The Chinese University of Hong Kong. https://cz-jrc.econ.cuhk.edu.hk/research/evaluating-hong-kong-consumption-voucher-scheme/

26.

Gilbert

Abdullah

(2004). Holidaytaking and the sense of well-being. Annals of Tourism Research, 31(1), 103–121. https://doi.org/10.1016/j.annals.2003.06.001

27.

Guan

(2023). Impulsive shopping overseas: Do sunk cost, information confusion, and anticipated regret have a say? Journal of Hospitality & Tourism Research, 47(3), 549–573. https://doi.org/10.1177/10963480211024450

28.

Hyun

Pae

J. H.

(2006). Consumers’ “Mental accounting” in response to unexpected price savings at the point of sale. Marketing Intelligence & Planning, 24(4), 406–416. https://doi.org/10.1108/02634500610672134

29.

Hall

C. M.

Scott

Gössling

(2020). Pandemics, transformations and tourism: Be careful what you wish for. Tourism Geographies, 22(3), 577–598. https://doi.org/10.1080/14616688.2020.1759131

30.

Hall

R. E.

Mishkin

F. S.

(1982). The sensitivity of consumption to transitory income: Estimates from panel data on households. Econometrica, 50(2), 461–481. https://doi.org/10.2307/1912638

31.

Hanretty

(2021). Forecasting multiparty by-elections using dirichlet regression. International Journal of Forecasting, 37(4), 1666–1676. https://doi.org/10.1016/j.ijforecast.2021.03.007

32.

Heath

Soll

J. B.

(1996). Mental budgeting and consumer decisions. Journal of Consumer Research, 23(1), 40–52. https://doi.org/10.1086/209465

33.

Heilman

C. M.

Nakamoto

Rao

A. G.

(2002). Pleasant surprises: Consumer response to unexpected in-store coupons. Journal of Marketing Research, 39(2), 242–252. https://doi.org/10.1509/jmkr.39.2.242.19081

34.

Higgins-Desbiolles

(2020). Socialising tourism for social and ecological justice after COVID-19. Tourism Geographies, 22(3), 610–623. https://doi.org/10.1080/14616688.2020.1757748

35.

Jones

Manoharan

Jiang

Singal

(2024). A framework of facilitators and barriers to career adaptability: Migrant hotel workers during COVID-19. Journal of Hospitality & Tourism Research, 48(3), 483–500. https://doi.org/10.1177/10963480221118820

36.

Kim

Jang

(2017). Symbolic consumption in upscale cafés: Examining Korean Gen Y consumers’ materialism, conformity, conspicuous tendencies, and functional qualities. Journal of Hospitality & Tourism Research, 41(2), 154–179. https://doi.org/10.1177/1096348014525633

37.

Kim

E. E. K.

Seo

Choi

(2022). Compensatory travel post COVID-19: Cognitive and emotional effects of risk perception. Journal of Travel Research, 61(8), 1895–1909. https://doi.org/10.1177/00472875211048930

38.

Kim

Gal

(2014). From compensatory consumption to adaptive consumption: The role of self-acceptance in resolving self-deficits. The Journal of Consumer Research, 41(2), 526–542. https://doi.org/10.1086/676681

39.

Kivetz

(1999). Advances in research on mental accounting and reason-based choice. Marketing Letters, 10(3), 249–266. https://doi.org/10.1023/A:1008066718905

40.

Koos

Vihalemm

Keller

(2017). Coping with crises: Consumption and social resilience on markets. International Journal of Consumer Studies, 41(4), 363–370. https://doi.org/10.1111/ijcs.12374

41.

Kuo

H.-I.

Chen

C.-C.

Tseng

W.-C.

L.-F.

Huang

B.-W.

(2008). Assessing impacts of SARS and avian flu on international tourism demand to Asia. Tourism Management, 29(5), 917–928. https://doi.org/10.1016/j.tourman.2007.10.006

42.

Lee

C.-K.

Song

H.-J.

Bendle

L. J.

Kim

M.-J.

Han

(2012). The impact of non-pharmaceutical interventions for 2009 H1N1 influenza on travel intentions: A model of goal-directed behavior. Tourism Management, 33(1), 89–99. https://doi.org/10.1016/j.tourman.2011.02.006

43.

Levav

McGraw

A. P.

(2009). Emotional accounting: How feelings about money influence consumer choice. Journal of Marketing Research, 46(1), 66–80. https://doi.org/10.1509/jmkr.46.1.66

44.

Levin

(1998). Are assets fungible?: Testing the behavioral theory of life-cycle savings. Journal of Economic Behavior & Organization, 36(1), 59–83. https://doi.org/10.1016/S0167-2681(98)00070-5

45.

Lin

S. C.

Reich

Kreps

T. A.

(2022). Feeling good or feeling right: Sustaining negative emotion after exposure to human suffering. Journal of Marketing Research, 60(3), 543–563. https://doi.org/10.1177/00222437221126917

46.

Liu

Shen

Chen

(2021). Stimulating consumption at low budget: Evidence from a large-scale policy experiment amid the COVID-19 pandemic. Management Science, 67(12), 7291–7307. https://doi.org/10.1287/mnsc.2021.4119

47.

Xiao

Wang

Zhang

Zhou

(2022). The potential of virtual tourism in the recovery of the tourism industry during the COVID-19 pandemic. Current Issues in Tourism, 25(3), 441–457. https://doi.org/10.1080/13683500.2021.1959526

48.

(2023). Tourist subsequent responses to promotion framing. Annals of Tourism Research, 103, Article 103658. https://doi.org/10.1016/j.annals.2023.103658

49.

Martín-Fernández

J. A.

Barceló-Vidal

Pawlowsky-Glahn

(2003). Dealing with zeros and missing values in compositional data sets using nonparametric imputation. Mathematical Geology, 35, 253–278. https://doi.org/10.1023/A:1023866030544

50.

Meng

(2014). Is a tourism subsidy the best response to the global financial crisis? A short-run CGE simulation for Singapore. Asia Pacific Journal of Tourism Research, 19(3), 325–341. https://doi.org/10.1080/10941665.2012.742916

51.

Milkman

K. L.

Beshears

(2009). Mental accounting and small windfalls: Evidence from an online grocer. Journal of Economic Behavior & Organization, 71(2), 384–394. https://doi.org/10.1016/j.jebo.2009.04.007

52.

Moreno-Luna

Robina-Ramírez

Sánchez

M. S.-O.

Castro-Serrano

(2021). Tourism and sustainability in times of COVID-19: The case of Spain. International Journal of Environmental Research and Public Health, 18(4), Article 1859. https://doi.org/10.3390/ijerph18041859

53.

Neumann

J. V.

Morgenstern

Kuhn

H. W.

Rubinstein

(2007). Theory of games and economic behavior (60th anniversary ed.). Princeton University Press. https://doi.org/10.1515/9781400829460

54.

Nieto García

Muñoz-Gallego

P. A.

Viglia

González-Benito

Ó.

(2020). Be social! The impact of self-presentation on peer-to-peer accommodation revenue. Journal of Travel Research, 59(7), 1268–1281. https://doi.org/10.1177/0047287519878520

55.

Peng

M. Y.-P.

Khalid

Usman

Khan

M. A. S.

Ali

(2022). Fear of COVID-19 and hotel frontline employees’ sense of work alienation: Intervening and interactional analysis. Journal of Hospitality & Tourism Research, 48(5), 821–833. https://doi.org/10.1177/10963480221112054

56.

Qiu

Sha

Scott

(2021). Restoration of visitors through nature-based tourism: A systematic review, conceptual framework, and future research directions. International Journal of Environmental Research and Public Health, 18(5), 1–19. https://doi.org/10.3390/ijerph18052299

57.

Rai

Garg

(2024). Demographic transition and inflation. Economic Systems, 48(4), Article 101214. https://doi.org/10.1016/j.ecosys.2024.101214

58.

Rajagopal

Rha

J.-Y.

(2009). The mental accounting of time. Journal of Economic Psychology, 30(5), 772–781. https://doi.org/10.1016/j.joep.2009.06.008

59.

Rathjens

Gunden

Zhang

Jain

Law

(2025). A systematic review of contactless technologies research in hospitality and tourism during the COVID-pandemic. Journal of Hospitality & Tourism Research, 49(2), 351–365. https://doi.org/10.1177/10963480231191192

60.

Ratneshwar

Pechmann

Shocker

A. D.

(1996). Goal-derived categories and the antecedents of across-category consideration. Journal of Consumer Research, 23(3), 240–250. https://doi.org/10.1086/209480

61.

Reisinger

Mavondo

(2005). Travel anxiety and intentions to travel internationally: Implications of travel risk perception. Journal of Travel Research, 43(3), 212–225. https://doi.org/10.1177/0047287504272017

62.

Sahm

C. R.

Shapiro

M. D.

Slemrod

(2012). Check in the mail or more in the paycheck: Does the effectiveness of fiscal stimulus depend on how it is delivered? American Economic Journal. Economic Policy, 4(3), 216–250. https://doi.org/10.1257/pol.4.3.216

63.

Sarmento

Marques

Galan -Ladero

(2019). Consumption dynamics during recession and recovery: A learning journey. Journal of Retailing and Consumer Services, 50, 226–234. https://doi.org/10.1016/j.jretconser.2019.04.021

64.

Shafir

Thaler

R. H.

(2006). Invest now, drink later, spend never: On the mental accounting of delayed consumption. Journal of Economic Psychology, 27(5), 694–712. https://doi.org/10.1016/j.joep.2006.05.008

65.

Shapiro

M. D.

Slemrod

(2003). Consumer response to tax rebates. American Economic Review, 93(1), 381–396. https://doi.org/10.1257/000282803321455368

66.

Shefrin

H. M.

Thaler

R. H.

(1988). The behavioral life-cycle hypothesis. Economic Inquiry, 26(4), 609–643. https://doi.org/10.1111/j.1465-7295.1988.tb01520.x

67.

Shenzhen Municipal Bureau of Statistics. (2023). Shenzhen statistical yearbook. https://tjj.sz.gov.cn/zwgk/zfxxgkml/tjsj/tjnj/content/post_11182604.html

68.

Smeral

(2010). Impacts of the world recession and economic crisis on tourism: Forecasts and potential risks. Journal of Travel Research, 49(1), 31–38. https://doi.org/10.1177/0047287509353192

69.

Smithson

Verkuilen

(2006). A better lemon squeezer? Maximum-likelihood regression with beta-distributed dependent variables. Psychological Methods, 11(1), 54. https://doi.org/10.1037/1082-989X.11.1.54

70.

Souleles

N. S.

(1999). The response of household consumption to income tax refunds. The American Economic Review, 89(4), 947–958. https://doi.org/10.1257/aer.89.4.947

71.

Souleles

N. S.

(2002). Consumer response to the reagan tax cuts. Journal of Public Economics, 85(1), 99–120. https://doi.org/10.1016/S0047-2727(01)00113-X

72.

Steel

Harris

(2020). Covid-19 economic recovery: Fiscal stimulus choices for lower-income countries. Institute for Fiscal Studies. https://media.odi.org/documents/fiscalstimulus_covid_final.pdf

73.

Thaler

R. H.

(1999). Mental accounting matters. Journal of Behavioral Decision Making, 12(3), 183–206. https://doi.org/10.1002/(SICI)1099-0771(199909)12:3<183::AID-BDM318>3.0.CO;2-F

74.

Uysal

Sirgy

M. J.

Woo

Kim

(2016). Quality of Life (QOL) and well-being research in tourism. Tourism Management, 53, 244–261. https://doi.org/10.1016/j.tourman.2015.07.013

75.

Yıldırım

Güler

(2022). Factor analysis of the COVID-19 perceived risk scale: A preliminary study. Death Studies, 46(5), 1065–1072. https://doi.org/10.1080/07481187.2020.1784311

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

1.70 MB