Accounting Behavior in Panic Buying After the Aceh Flood: Perceived Justice as a Mediator of Social and Risk Factors

Abstract

This study examines the role of accounting factors, particularly accountability and transparency, in explaining panic buying behavior after the Aceh flood, while investigating the mediating role of perceived justice and the influence of social and risk factors. Using a quantitative approach, data were collected from 205 respondents in East Aceh, Langsa City, and Aceh Tamiang and analyzed using Structural Equation Modeling–Partial Least Squares (SEM-PLS) in SmartPLS. The findings indicate that social influence is the strongest predictor of panic buying, followed by perceived risk. Accountability does not significantly affect perceived justice or panic buying, whereas transparency positively influences perceived justice. Perceived justice negatively affects panic buying and mediates the effect of transparency on panic buying.

Keywords: panic buying; accounting behavior; perceived justice; social influence; risk perception.

Introduction

Panic buying refers to excessive purchasing behavior that occurs under uncertain conditions, such as natural disasters, supply shortages, or sudden price changes (Yuen et al., 2021). This behavior often appears after floods when communities face uncertainty regarding the availability of essential goods and potential supply disruptions. Previous studies indicate that information uncertainty and perceived scarcity significantly influence panic buying behavior after disasters (Dulam et al., 2021). Aceh experienced a major flood disaster in November 2025 that disrupted community activities and increased uncertainty regarding essential goods availability. According to the Aceh Disaster Management Agency, as of November 27, 2025, floods affected 16 districts/cities, including East Aceh, Langsa City, and Aceh Tamiang, impacting 33,817 households or 119,988 people, with 6,998 households or 20,759 people displaced. The disaster was triggered by high rainfall intensity, strong winds, and unstable environmental conditions, causing flooding, landslides, and disruptions to local economic activities. (Aceh Disaster Management Agency, 2025). The selected research areas represent communities directly affected by the disaster. East Aceh recorded 7,972 affected households or 29,706 people, while Langsa City experienced flooding in residential areas, affecting 110 houses with water levels reaching 20–40 cm in several locations. Aceh Tamiang was also included due to its vulnerability to recurring flood impacts and disruptions to community activities (Aceh Disaster Management Agency, 2025). These conditions provide a relevant context for examining panic buying behavior in disaster-affected communities. This study examines the role of accounting factors, particularly accountability and transparency, in explaining panic buying behavior after the Aceh flood, while investigating the mediating role of perceived justice and the influence of social and risk factors. The study contributes by integrating an accounting perspective with post-disaster consumer behavior to understand how institutional trust and social conditions shape purchasing responses in disaster-affected communities. Using a quantitative approach, data were collected via a Likert-scale questionnaire from 205 respondents in East Aceh, Langsa City, and Aceh Tamiang, and analyzed using Structural Equation Modeling with Partial Least Squares (SEM-PLS) in SmartPLS. The findings indicate that social influence is the strongest predictor of panic buying, followed by perceived risk, while accountability does not significantly affect perceived justice or panic buying. Transparency positively influences perceived justice, and perceived justice negatively affects panic buying, indicating that perceived justice mediates the effect of transparency on panic buying. These findings highlight the importance of transparent communication, perceived fairness, and social risk management by governments and business actors to reduce uncertainty and maintain market stability after disasters.

In this situation, accounting practices have an important role in supporting accurate, accountable, and transparent information governance related to inventory management, distribution, and pricing of goods. According to, market management needs to be carried out by prioritizing the principles of transparency, justice, and social responsibility so that market stability is maintained. Accounting not only functions as a system for recording financial transactions, but also as a means of providing information that can be accounted for to stakeholders (Ismail, 2022). Accounting information that is presented in an accountable and transparent manner allows the public to obtain an objective overview of inventory conditions, distribution mechanisms, and price changes. According to (Nazarenko et al., 2024) accounting practices that emphasize reporting transparency and fairness in resource management can increase public confidence in economic activities. With clear information about the price of goods, stock availability, and distribution policies, it is hoped that the public will not be easily affected by panic that triggers panic buying behavior (Jazemi et al., 2024). In addition, the government, through trade and distribution policies, plays a strategic role in maintaining price stability and ensuring the availability of goods during an emergency. Law Number 7 of 2014 concerning Trade regulates the control of the distribution of goods under certain conditions, including when natural disasters occur, in order to prevent hoarding practices that are detrimental to the community. In this context, information transparency supported by an accounting system is an important foundation for the government and business actors in conveying information about prices, distribution, and availability of goods to the public (Peran et al., 2025). Thus, in post-flood conditions, implementing accounting practices that uphold accountability, transparency, and fairness is expected to increase public trust in the market and suppress panic buying. Conditions in the field show that the uncertainty of information regarding the price and availability of goods is still one of the main causes of the emergence of panic buying behavior after the flood. In Langsa City, Aceh Tamiang Regency, and East Aceh Regency, some business actors have not fully implemented accounting practices that are able to produce inventory and price information openly to the public. Information about the availability of goods is often not conveyed clearly, which raises distrust among consumers. According to (Y. Chen et al., 2020) these conditions can worsen the crisis and increase people's tendency to make excessive purchases. In the context of flood disasters, people are more encouraged to make decisions to buy goods quickly without considering their actual needs or aspects of equitable distribution. On the other hand, business actors often experience limitations in conveying transparent information about the condition of the supply of goods, so consumer anxiety is increasing. The gap between ideal and factual conditions creates various problems in the local economy. Without good information management, this condition has the potential to worsen market instability and increase inequality in the distribution of goods. Behavioral panic buying not only has an impact on consumer welfare, but it also causes losses for business actors due to disruption of distribution mechanisms and inventory management. According to research by (Dulam et al., 2021) and (Wu et al., 2023), panic buying can cause significant disruptions to supply chains, increase price uncertainty, and trigger inflation, especially in basic commodities that are difficult to obtain after a disaster. Another impact of panic buying is the emergence of social inequality in the distribution of goods (Yuen et al., 2022a). People with higher purchasing power tend to be able to control the supply of goods faster, while low-income groups have difficulty obtaining necessities. This condition magnifies existing social gaps and has the potential to cause conflicts in society. Therefore, research on the role of accounting through increasing accountability, transparency, and perception of fairness in reducing panic buying behavior is important to be carried out as an effort to minimize the negative impact of flood disasters on people's social and economic conditions. Various previous studies have examined the phenomenon of panic buying in the context of natural disasters and market uncertainty. Research by (Billore & Anisimova, 2021) It shows that unclear policies and limited information about the distribution of goods during crises exacerbate panic buying behavior. Meanwhile, research (Yuen et al., 2021) It found that the application of transparency in the delivery of information about the distribution of goods after a disaster was able to reduce the level of public anxiety so that panic buying behavior could be reduced. However, there is still very little research linking accounting practices to consumer behavior in emergencies, particularly through accountability mechanisms, transparency, and the perception of fairness as a mediating variable. This research aims to fill this gap by examining how accounting practices can affect panic buying behavior through increasing accountability, transparency, and the perception of justice in the post-flood community in Aceh. Previous studies on panic buying have primarily focused on psychological and social determinants, such as uncertainty, perceived scarcity, information gaps, risk perception, and social influence. (Yuen et al., 2022) and (Dulam et al., 2021) demonstrated that these factors contribute to excessive purchasing behavior during crises; however, they provided limited explanation of how governance-related information influences consumer responses. Meanwhile, disaster governance studies have largely examined transparency, accountability, and public trust from an institutional perspective. At the same time, behavioral accounting research has focused on accountability, fairness, and information disclosure within organizational contexts. Therefore, the relationship among governance mechanisms, perceived justice, and consumer behavior following disasters remains underexplored. This study addresses this gap by developing a perceived-justice mediation model that examines how accountability and transparency influence panic buying behavior after the Aceh flood. The study extends previous research by introducing perceived justice as a mechanism linking governance information to consumer behavior, providing empirical evidence from disaster-affected communities in Aceh, and applying PLS-SEM to examine the direct and indirect relationships among governance, fairness perceptions, and panic buying behavior. The urgency of this research is even stronger when looking at the impact of panic buying after the flood, which can worsen local economic conditions, disrupt market stability, and create injustice in the distribution of goods. The results of the study are expected to be able to provide recommendations on the importance of implementing accountable and transparent accounting practices in post-disaster market management so as to create a fairer distribution system and be able to increase public trust. In addition, the results of the research are also expected to be considered by the government and business actors in formulating policies that are more responsive to the needs of communities in disaster-prone areas. Based on this description, research on Accounting Behavior in Panic Buying After the Aceh Flood: Perceived Justice as a Mediator of Social and Risk Factors is important to be carried out. This research is expected to make a theoretical contribution to the development of behavioral accounting literature as well as a practical contribution for the government and business actors in building more accountable, transparent

Methods

Research Approach

This study employed a quantitative cross-sectional research design using Structural Equation Modeling–Partial Least Squares (SEM-PLS) to examine the hypothesized relationships among accountability, transparency, perceived justice, social effect, risk perception, and panic buying behavior in post-disaster communities in Aceh. Data were collected from January to March 2026 in Aceh Timur, Langsa City, and Aceh Tamiang using purposive sampling. Respondents were selected based on the criteria of being at least 17 years old, residing in flood-affected areas, purchasing essential goods after the disaster, and being involved in household purchasing decisions. After excluding incomplete and ineligible responses, 205 valid responses were obtained. Data were collected through direct surveys and online questionnaires, while the instrument was reviewed before the main survey to ensure clarity and contextual suitability. SEM-PLS was selected because the model involved latent constructs, direct and indirect relationships, and mediation analysis. Therefore, the study examined structural associations rather than causal relationships. The adequacy of the sample size was evaluated using an a priori statistical power analysis with G*Power 3.1 based on multiple linear regression analysis. The calculation used an alpha level of 0.05, statistical power of 0.80, medium effect size (f² = 0.15), and five predictors for the most complex endogenous construct, namely panic buying behavior. The analysis indicated that a minimum sample size of 92 respondents was required. Since this study obtained 205 valid responses, the sample size exceeded the minimum requirement and was considered adequate for estimating the proposed structural relationships and mediation effects. The 10-times rule commonly used in PLS-SEM was considered only as supplementary guidance, while the statistical power analysis served as the primary justification for sample adequacy. Accountability and transparency are positioned in this study as exogenous governance-related variables that influence panic buying behavior through perceived justice as a mediating mechanism. In addition, social effect and risk perception are conceptualized as exogenous behavioral predictors because both factors represent important psychological and social mechanisms underlying panic buying behavior in post-disaster situations. The inclusion of these constructs strengthens the structural model by capturing both governance-related and behavioral determinants of consumer responses after disasters. The analysis was conducted using SmartPLS through two main stages: measurement model evaluation (outer model) and structural model assessment (inner model). The outer model was evaluated based on convergent validity, discriminant validity, composite reliability, and Cronbach’s alpha. The inner model was assessed using bootstrapping procedures to examine path coefficients, coefficient of determination (R²), predictive relevance (Q²), effect size (f²), and mediation effects.

Conceptual Model

The conceptual model of this study was developed to explain the influence of accounting dimensions on panic buying behavior following flood disasters in Aceh. Specifically, the model examines how accountability and transparency influence panic buying both directly and indirectly through perception of justice. From the perspective of accounting, accountability reflects the responsibility of institutions, authorities, and market actors in ensuring fair distribution, truthful information disclosure, and ethical management during disaster situations. Transparency refers to the openness and clarity of information regarding the availability of goods, distribution mechanisms, and disaster-related policies. Both variables are expected to shape public perceptions of justice, particularly regarding fairness in access to essential goods during crisis conditions. Perception of justice is conceptualized as a mediating mechanism because individuals who perceive fair treatment and transparent governance are less likely to engage in irrational purchasing behavior. Conversely, uncertainty, distrust, and perceived injustice may intensify panic-driven consumption behavior. In addition, this study incorporates social effect and perception of risk as exogenous behavioral predictors of panic buying behavior. Social effect represents the influence of family, social interactions, community behavior, and collective purchasing tendencies that may encourage individuals to engage in excessive purchasing during crisis situations. Meanwhile, perception of risk reflects individuals’ subjective assessment of disaster severity, uncertainty, and potential scarcity of essential goods, which may increase the tendency to conduct panic buying. Unlike governance-related factors that explain how accountability and transparency shape perceived justice, social effect and risk perception represent behavioral mechanisms that directly influence consumer responses during post-disaster conditions. Therefore, both constructs are modeled as direct predictors of panic buying behavior and included in the structural model to enhance the explanation of consumer behavior after disasters. The conceptual framework of this study is presented in Figure 1. Based on the conceptual framework, this study proposes that accountability and transparency significantly influence perception of justice and panic buying behavior. Furthermore, perception of justice is expected to mediate the relationship between accounting practices and panic buying behavior in post-disaster contexts. The inclusion of social effect and perception of risk as control variables enhances the explanatory power and contextual relevance of the research model.

Population and Sample/Informants

The population of this study consists of communities affected by post-flood conditions in three regions of Aceh Province, namely Aceh Timur, Kota Langsa, and Aceh Tamiang. These regions were selected because they experienced flood impacts that disrupted the distribution of necessities, increased uncertainty regarding product availability, and potentially triggered panic buying behavior among the community. In addition, the three regions represent strategic areas with relatively high levels of economic and social interaction during post-disaster recovery conditions, making them relevant for examining consumer behavior from the perspective of accounting. This study employed a purposive sampling technique, in which respondents were selected based on specific criteria aligned with the objectives of the research. The use of purposive sampling is considered appropriate because not all members of the population possess relevant experiences related to panic buying behavior following flood disasters. The respondent criteria in this study include: (1) individuals aged at least 17 years old, (2) domiciled in one of the research areas, (3) having purchased necessities after the flood disaster, (4) knowing price changes, product scarcity, or purchasing restrictions following the flood, and (5) acting as decision makers or policy holders regarding personal or household expenditures. These criteria were established to ensure that respondents possess adequate understanding and direct experience related to the phenomenon being investigated. A total of 205 respondents participated in this study, consisting of 68 respondents from Aceh Timur, 74 respondents from Kota Langsa, and 63 respondents from Aceh Tamiang. The distribution of respondents across the three regions was intended to provide proportional representation of communities affected by post-disaster economic conditions and to strengthen the generalizability of the findings within the research context. The suggestions frequently utilized in Structural Equation Modeling–Partial Least Squares (SEM-PLS) are used to determine the sample size. The "10-times rule" states that the minimum sample size must be at least ten times the number of structural routes that point to a latent construct in the model. (Jhantasana, 2023). Panic buying is the endogenous construct in this study with the greatest number of incoming channels. It is directly impacted by perceptions of risk, social effect, accountability, and justice. Therefore, the total sample obtained in this study would be significantly larger than the minimum required sample size based on these criteria. Furthermore, (Kock & Hadaya, 2018) explain that SEM-PLS analysis generally requires sample sizes ranging from 100 to 200 observations for models with moderate complexity, mediation relationships, and predictive-oriented analysis. Since this study includes multiple latent variables, mediation effects, and control variables, the use of 205 respondents is considered statistically adequate and methodologically appropriate. This sample size provides sufficient statistical power to estimate path coefficients, evaluate mediation effects, and enhance the robustness and predictive accuracy of the SEM-PLS model. Therefore, the sample size of 205 respondents is considered logical, representative, and adequate for testing the conceptual framework developed in this study.

Research Location

This research was conducted in three flood-affected regions in Aceh Province, namely Aceh Timur, Kota Langsa, and Aceh Tamiang. These locations were selected purposively based on several academic and contextual considerations related to the objectives of the study. First, the three regions frequently experience flood disasters that disrupt economic activities, supply distribution systems, and access to essential goods. Such conditions create social uncertainty and potentially trigger panic buying behavior among communities during post-disaster periods. Second, the selected regions represent interconnected socio-economic areas within the eastern part of Aceh Province, characterized by relatively high community mobility and active market interactions. These characteristics make the regions highly relevant for examining how accountability, transparency, and perception of justice influence consumer behavior in crises. The existence of traditional markets, retail distribution channels, and varying levels of public access to information also provides a suitable setting for analyzing panic buying behavior from the perspective of accounting. Third, the selection of these locations was intended to capture variations in community experiences and behavioral responses following flood disasters. Although all three regions share similar disaster vulnerabilities, each area possesses distinct demographic, economic, and social characteristics that may influence perceptions of fairness, transparency, and risk during emergency situations. Therefore, conducting the study across multiple locations enhances the contextual richness and external validity of the research findings. Overall, the research locations were considered appropriate because they provide empirical relevance for investigating the relationship between accounting practices and panic buying behavior in post-disaster contexts within Aceh Province.

Instrumentation or Tools

A structured questionnaire was administered to respondents in flood-affected areas of Aceh using a five-point Likert scale ranging from 1 (“strongly disagree”) to 5 (“strongly agree”). The measurement items were adapted from previous studies on panic buying, crisis behavior, governance, and fairness perceptions. Specifically, panic buying items were adapted from (Noda & Teramoto, 2024), risk perception from (Priolo et al., 2025), social effect from (Mnif et al., 2025), perceived justice from (Solano-Silva et al., 2025) and accountability and transparency from (Cheong, 2024) . All items were modified to reflect the Aceh flood context, particularly regarding essential goods availability, price information, distribution processes, and consumer responses during the recovery period. The questionnaire was developed in Indonesian to ensure contextual suitability for local respondents. Before the main survey, the instrument was reviewed to assess item clarity, relevance, and contextual appropriateness. The review focused on ensuring that each indicator represented the intended construct and that the wording was understandable within the post-flood community context. Based on the review process, several wording adjustments were made to improve contextual relevance, particularly by adapting general crisis- related statements into statements related to essential goods availability, price changes, distribution conditions, and purchasing experiences after the Aceh flood. A separate pilot test was not conducted before the main survey. This limitation is acknowledged because the instrument was adapted from previously validated measurement scales and was adjusted through contextual review before data collection. Future research may conduct pilot testing with a larger preliminary sample to further evaluate measurement stability and refine the instrument before wider application. The accountability variable was measured using four indicators that reflect institutional responsibility and ethical governance during disaster situations. These indicators include: (1) responsibility in ensuring the availability of essential goods, (2) fairness in distribution practices, (3) responsiveness of authorities in handling post-disaster economic problems, and (4) consistency in implementing policies related to public welfare during crisis conditions. The transparency variable was also measured using four indicators reflecting openness and clarity of information delivery. The indicators include: (1) clarity of information regarding the availability of goods, (2) openness related to price changes after the flood, (3) accessibility of information concerning distribution policies, and (4) accuracy and timeliness of information provided to the public during post- disaster conditions. Perception of justice was measured through four indicators representing respondents’ perceptions of fairness in economic and social treatment during post-disaster recovery. These indicators include: (1) fairness in access to essential goods, (2) fairness in product distribution, (3) equal treatment among community members, and (4) fairness in government or market policies implemented after the flood disaster. The panic buying variable was measured using four indicators representing excessive and irrational purchasing behavior during emergency situations. These indicators include: (1) purchasing goods in excessive quantities, (2) buying products due to fear of scarcity, (3) purchasing behavior driven by panic and uncertainty, and (4) stockpiling essential goods for anticipated future shortages. In addition, this study incorporates two control variables, namely social effect and perception of risk. Social effect was measured using four indicators consisting of: (1) influence of family behavior, (2) influence of friends or surrounding communities, (3) tendency to follow collective purchasing behavior, and (4) social pressure encouraging the purchase of essential goods during crises. Meanwhile, perception of risk was measured using four indicators reflecting respondents’ subjective assessments of disaster-related uncertainty and scarcity threats. These indicators include: (1) fear of product shortages, (2) concern regarding future uncertainty after the flood, (3) perception of increasing economic vulnerability, and (4) perceived risk of difficulties in accessing essential goods during post-disaster conditions. The instrument items were formulated based on the conceptual framework of the study and adjusted to the socio- economic context of flood-affected communities in Aceh. Before data collection, the questionnaire was reviewed to ensure clarity, contextual relevance, and suitability for SEM- PLS analysis.

Data Collection Procedures

Communities impacted by post-flood conditions in Aceh Timur, Kota Langsa, and Aceh Tamiang were the focus of the survey method used for data collection in this study. Direct field surveys and online surveys using digital forms were the two methods used to administer the study. The goal of combining these strategies was to guarantee greater representation of populations inside the research areas, enhance response rates, and expand respondent accessibility. Respondents who satisfied the predefined sampling criteria were given printed questionnaires to complete in order to conduct direct surveys. This method allowed researchers to speak with respondents face-to-face, explain questionnaire items when needed, and make sure respondents understood the study's context. Residential neighborhoods, neighborhood markets, and community settings impacted by flood-related economic disruptions were the main locations for field studies. In addition, online data collection was conducted through digital questionnaire forms distributed via social media platforms and communication applications commonly used by the community. The use of online surveys provided flexibility for respondents to complete the questionnaire independently and efficiently, particularly for individuals with limited availability for face-to-face interactions. The digital survey approach also facilitated faster data compilation and reduced administrative limitations associated with manual data collection. Before participating in the survey, respondents were informed about the objectives of the study and assured that their responses would remain confidential and used solely for academic purposes. Only respondents who fulfilled the established criteria were included in the final dataset. The combination of offline and online survey procedures was considered appropriate for obtaining comprehensive and reliable data regarding panic buying behavior and accounting perceptions in post-disaster contexts.

Data Analysis Technique

The data in this study were analyzed using the Structural Equation Modeling–Partial Least Squares (SEM-PLS) approach with the assistance of SmartPLS software. SEM-PLS was selected because this study examines complex relationships involving direct effects, indirect effects, mediation relationships, and control variables within a predictive and exploratory research framework. In addition, SEM-PLS is considered appropriate for behavioral and social science research because it does not require strict assumptions of multivariate normality and is suitable for moderate sample sizes. (Hair & Alamer, 2022) Measurement model evaluation (outer model) and structural model evaluation (inner model) comprised the two primary phases of the analytical process. The purpose of the outer model assessment was to examine the research instruments' validity and reliability. Average Variance Extracted (AVE) and outer loading were used to assess convergent validity; satisfactory validity is indicated by AVE values above 0.50 and outer loading values above 0.70. Cronbach's Alpha and Composite Reliability scores were used to evaluate reliability; threshold values greater than 0.70 indicated adequate internal consistency. Additionally, to make sure that each construct in the model was empirically different from the others, discriminant validity was assessed using the Fornell- Larcker Criterion and Heterotrait-Monotrait Ratio (HTMT). After the measurement model fulfilled the required criteria, the structural model evaluation was conducted to test the proposed relationships among variables. This stage included assessment of collinearity through Variance Inflation Factor (VIF), coefficient of determination (R²), predictive relevance (Q²), and effect size (f²). The significance of path relationships was tested using the bootstrapping procedure to obtain path coefficients, t-statistics, and p-values. The structural equations used in this study are formulated as follows: Direct Effect Model 𝑃𝐵= 𝛽0 + 𝛽1𝐴𝐶𝐶+ 𝛽2𝑇𝑅𝑃+ 𝛽3𝑃𝐽+ 𝛽4𝑆𝐸+ 𝛽5𝑃𝑅+ 𝜀1 𝑃𝐽= 𝛽6𝐴𝐶𝐶+ 𝛽7𝑇𝑅𝑃+ 𝜀2 Equation (1) explains that panic buying behavior is influenced by accountability, transparency, perception of justice, social effect, and perception of risk. Equation (2) explains that perception of justice is influenced by accountability and transparency. Indirect Effect Model (Mediation Effect) 𝐼𝐸𝐴𝐶𝐶→𝑃𝐵= (𝛽6 × 𝛽3) 𝐼𝐸𝑇𝑅𝑃→𝑃𝐵= (𝛽7 × 𝛽3) The indirect effect model explains the mediating role of perception of justice in the relationship between accountability, transparency, and panic buying behavior. Total Effect Model 𝑇𝐸𝐴𝐶𝐶→𝑃𝐵= 𝛽1 + (𝛽6 × 𝛽3) 𝑇𝐸𝑇𝑅𝑃→𝑃𝐵= 𝛽2 + (𝛽7 × 𝛽3) Where: PB = Panic Buying ACC = Accountability TRP = Transparency PJ = Perception of Justice SE = Social Effect PR = Perception of Risk IE = Indirect Effect TE = Total Effect β₁–β₇ = Path coefficients indicating the direction and strength of relationships among variables ε₁–ε₂ = Error terms The mediation effect was tested using the bootstrapping procedure in Smart PLS to determine the significance of indirect effects. In addition, social effect and perception of risk were incorporated as control variables to strengthen the explanatory power and robustness of the model in explaining panic buying behavior during post-disaster conditions.

Ethical Approval

To guarantee the safety, confidentiality, and voluntary involvement of every participant in the research process, this study was carried out in compliance with ethical research principles. Respondents were made aware of the study's goals, the questionnaire's scope, and the fact that the data would only be used for academic and research reasons before data collection. Respondents were free to decline or stop participating at any time without facing any repercussions because participation in this study was completely voluntary. To maintain confidentiality and anonymity, the questionnaire did not request personally sensitive information that could directly identify respondents. All responses were treated confidentially and analyzed collectively to ensure that individual identities remained protected. In addition, respondents were informed that the study focused on perceptions and experiences related to post-disaster economic conditions and panic buying behavior, without causing physical, psychological, or social harm to participants. Considering that this study involved community perceptions and behavioral responses in post-flood situations, the researchers ensured that the data collection procedures were conducted respectfully and sensitively toward disaster-affected communities. The survey process was designed to minimize discomfort and avoid coercion during respondent participation. Furthermore, this research adhered to general ethical standards in social science research, including principles of informed consent, confidentiality, voluntary participation, and responsible data management. Therefore, the implementation of this study was considered ethically appropriate for examining panic buying behavior and accounting practices in post- disaster contexts.

Result and Discussion

Descriptive Statistics

Descriptive statistical analysis was carried out to give a summary of respondents' reactions to each indicator utilized in this study prior to the SEM-PLS analysis. The descriptive statistics include skewness, excess kurtosis, standard deviation, mean, median, minimum, and maximum values.

These statistical metrics are crucial for comprehending the data's distribution properties and determining whether the dataset is suitable for additional structural investigation. The mean values describe the average response of respondents to each indicator, while the standard deviation reflects the variability of responses. In addition, skewness and kurtosis values were examined to identify the distribution pattern of the data. According to (Pavlov et al., 2020), skewness and kurtosis values within an acceptable range indicate that the data distribution does not exhibit severe abnormality, thereby supporting the robustness of SEM-PLS estimation. Based on Table 1, the mean values of all indicators range from 2.55 to 3.98, indicating moderate respondent perceptions across all constructs. The highest mean value was observed in indicator TRA1 (3.98), suggesting that respondents generally perceived transparency-related information positively during post-flood conditions. Meanwhile, the lowest mean value was found in indicator SE2 (2.55), indicating relatively lower agreement regarding certain dimensions of social influence on panic buying behavior. The standard deviation values ranged from 1.11 to 1.35, indicating moderate variability in respondents’ answers. This suggests that respondents exhibited relatively diverse perceptions and experiences related to accountability, transparency, justice perception, social influence, risk perception, and panic buying behavior following flood disasters. Furthermore, the skewness values ranged between -0.91 and 0.38, while excess kurtosis values ranged between -1.09 and 0.11. These results indicate that the data distribution does not exhibit severe non-normality problems. Most indicators showed slightly negative skewness, implying that respondents tended to provide relatively higher agreement responses toward several indicators, particularly those related to transparency and perception of justice.

Table 1. Descriptive Statistics of Research Indicators

Source: Primary Data.

Measurement Model Assessment (Outer Model)

The validity and reliability of the constructs included in this study were assessed using the measurement model. The outer model assessment in SEM-PLS analysis is to confirm that each indicator can measure its corresponding latent components with accuracy. Through the analysis of outer loading values, Cronbach's Alpha, rho A, Composite Reliability (CR), and Average Variance Extracted (AVE), convergent validity and internal consistency reliability are assessed.

Convergent Validity

Convergent validity was evaluated by examining the outer loading values of each indicator toward its respective construct. According to (Fithri et al., 2024), outer loading values above 0.70 indicate that the indicators adequately represent the latent constructs. Based on Table 2, all indicators demonstrated acceptable convergent validity because the majority of outer loading values exceeded the recommended threshold of 0.70. The Accountability construct exhibited strong loading values ranging from 0.775 to 0.876, indicating that all indicators adequately represent accountability dimensions in post- disaster conditions. Similarly, the Panic Buying construct showed very high loading values, particularly PB3 (0.898) and PB4 (0.894), suggesting that these indicators strongly capture panic buying behavior among respondents. The Perceived Justice, Perceived Risk, and Social Effect constructs demonstrated satisfactory outer loading values above 0.70, confirming the adequacy of their indicators in representing the respective constructs. The TRA1 indicator within the Transparency construct showed a lower loading value of 0.642. An assessment of the measurement model indicated that removing TRA1 did not substantially improve the construct reliability and validity, as the Transparency construct already achieved acceptable values of Composite

Table 2. Outer Loading Values

Source: Primary Data.

Reliability and Average Variance Extracted (AVE)

. Therefore, TRA1 was retained due to its theoretical relevance in capturing the accessibility and availability of transparent information during post-disaster situations, particularly regarding public access to information on essential goods, distribution processes, and market conditions. Nevertheless, the relatively lower loading value of TRA1 is acknowledged as a measurement limitation, indicating that this indicator may require further refinement in future studies. Overall, the convergent validity results confirm that all indicators adequately represent their intended constructs. Reliability and Average Variance Extracted (AVE) Reliability assessment was conducted using Cronbach’s Alpha, rho_A, and Composite Reliability values, while convergent validity at the construct level was evaluated through Average Variance Extracted (AVE). According to (Li & Lay, 2024), values above 0.70 for reliability measures and above

Cronbach's Alpha rho_A Composite Reliability Average Variance Extracted (AVE) Accountability 0.866 0.876 0.909 0.714 Panic Buying 0.895 0.896 0.927 0.761 Perception of Justice 0.806 0.817 0.873 0.632 Perception of Risk 0.873 0.895 0.912 0.723 Social Effect 0.88 0.881 0.917 0.735 Transparancy 0.8 0.814 0.871 0.631 0.50 for AVE indicate satisfactory measurement quality. Based on Table 3, all constructs achieved Cronbach’s Alpha values above 0.70, indicating satisfactory internal consistency reliability. The Panic Buying construct obtained the highest Cronbach’s Alpha value (0.895), suggesting strong consistency among its indicators. Similarly, all rho_A values exceeded 0.70, further confirming the reliability of the constructs. The Composite Reliability values ranged from 0.871 to 0.927, exceeding the recommended threshold and indicating strong reliability across all constructs. The highest Composite Reliability value was found in the Panic Buying construct (0.927), demonstrating that the indicators consistently measure the same latent variable. Additionally, every AVE number, which ranged from 0.631 to 0.761, was higher than the 0.50 criterion. These findings support sufficient convergent validity at the concept level by showing that each construct accounts for more than 50% of the variation of its indicators. All of the constructs utilized in this study have satisfactory reliability and convergent validity, according to the reliability and AVE results, which supports the measurement model's eligibility for additional SEM-PLS structural analysis.

Table 3. Reliability and AVE Results

Source: Primary Data.

Structural Model Assessment (Inner Model)

After confirming the adequacy of the measurement model, the next stage involved evaluating the structural model (inner model). The structural model assessment aims to examine the predictive capability of the model and the strength of relationships among latent constructs. This assessment includes collinearity testing, coefficient of determination (R²), predictive relevance (Q²), and effect size (f²).

Collinearity Test

The collinearity test was conducted using the Variance Inflation Factor (VIF) to identify potential multicollinearity issues among indicators and constructs in the model. According to Hair et al. (2021), VIF values below 5 indicate that collinearity does not represent a serious problem in SEM- PLS analysis. Based on Table 4, all VIF values were below the threshold value of 5, ranging from 1.246 to 3.269. The highest VIF value was observed for PB4 (3.269), while the lowest value was found for TRA1 (1.246). These findings indicate that multicollinearity was not present in the model and that each indicator contributed uniquely to its respective construct. Therefore, the structural model is considered free from serious collinearity issues and suitable for further analysis.

Table 4. Collinearity Assessment (VIF)

Source: Primary Data.

Coefficient of Determination (R²)

The coefficient of determination (R²) was used to evaluate the explanatory power of the model toward endogenous constructs. According to Hair et al. (2021), R² values of 0.75, 0.50, and 0.25 indicate substantial, moderate, and weak explanatory power, respectively According to Table 5, the Panic Buying construct had an R2 value of 0.668, meaning that 66.8% of the variance in panic buying behavior was explained by accountability, transparency, perception of justice, social effect, and perception of risk taken together. This outcome indicates that the suggested model has significant predictive potential in predicting panic buying behavior during post-flood settings and shows a comparatively great explanatory capability. Accountability and transparency accounted for 31.9% of the variance in perception of justice, according to the Perception of Justice construct, which had an R2 value of 0.319. This result suggests that accounting aspects significantly influence public perceptions of justice in post- disaster settings, even if it is lower than the Panic Buying construct.

Table 5. Coefficient of Determination (R²)

Source: Primary Data.

Predictive Relevance (Q²)

To determine if the structural model can predict endogenous constructs, predictive relevance (Q²) was evaluated using the blindfolding process. Q² values greater than zero suggest sufficient predictive significance, according to (Russo & Stol, 2021). Table 6 shows that the Q² value for panic buying was 0.495, demonstrating the model's excellent predictive usefulness in understanding panic buying behavior. In contrast, the Perception of Justice Q² value was 0.197, indicating a modest level of predictive relevance. The structural model exhibits good predictive capability because all endogenous constructs achieved Q² values larger than zero. These findings show that the model has sufficient predictive accuracy in predicting respondent behavior and views in post-disaster scenarios in addition to understanding links among dimensions.

Table 6. Predictive Relevance (Q²)

Source: Primary Data.

Effect Size (f²)

The effect size (f²) assessment was conducted to examine the magnitude of the contribution of each exogenous construct Indicator VIF ACC1 2.489 ACC2 2.653 ACC3 2.067 ACC4 1.679 PB1 2.356 PB2 2.159 PB3 3.16 PB4 3.269 POJ1 1.967 POJ2 1.811 POJ3 1.849 POJ4 1.582 PR1 1.959 PR2 2.025 PR3 2.543 PR4 2.53 SE1 2.627 SE2 2.433 SE3 1.955 SE4 2.173 TRA1 1.246 TRA2 2.091 TRA3 2.618 TRA4 1.973 R Square R Square Adjusted Panic Buying 0.668 0.661 Perception of Justice 0.319 0.312 SSO SSE Q² (=1-SSE/SSO) Accountability 820 820 Panic Buying 820 414.485 0.495 Perception of Justice 820 658.431 0.197 Perception of Risk 820 820 Social Effect 820 820 Transparancy 820 820 toward endogenous variables. According to (Messner, 2023), f² values of 0.02, 0.15, and 0.35 indicate small, medium, and large effects, respectively. Based on Table 7, Social Effect demonstrated the strongest contribution toward Panic Buying behavior with an f² value of 1.352, indicating a very large effect size. This finding suggests that social influence plays a dominant role in encouraging panic buying behavior during post-flood situations. In disaster contexts, individuals are highly influenced by surrounding community behavior, social panic, and collective purchasing actions. Furthermore, Transparency showed a large effect on Perception of Justice with an f² value of 0.456. This result indicates that transparent information dissemination and clarity regarding disaster-related policies significantly shape public perceptions of fairness and justice. Meanwhile, Perception of Justice and Perception of Risk exhibited small effect sizes on Panic Buying behavior, indicating that these constructs contribute modestly but meaningfully to explaining panic buying tendencies. On the other hand, Accountability demonstrated negligible effect sizes toward both Panic Buying and Perception of Justice, suggesting that accountability alone may not strongly influence behavioral responses unless supported by transparency and broader social dynamics. Overall, the effect size analysis confirms that social and informational factors play substantial roles in shaping panic buying behavior in post-disaster conditions, while accounting dimensions contribute indirectly through justice perception mechanisms.

Table 7. Effect Size (f²)

Source: Primary Data.

Structural Model and Hypothesis Testing

The final SEM-PLS structural model was assessed to look at the relationships between constructs after the measurement and structural models were found to be adequate. The structural model estimation provides information regarding path coefficients, directions of relationships, explanatory power, and the significance of direct and indirect effects among variables. Figure 2 illustrates the overall structural relationships among accountability, transparency, perception of justice, perception of risk, social effect, and panic buying behavior. The model indicates that social effect has the strongest influence on panic buying behavior, while transparency

Hypothesized Path β t p 2.5% CI 97.5% CI Decision Accountability → Panic Buying 0.071 1.779 0.076 0.001 0.148 Not supported Accountability → Perception of Justice −0.048 0.589 0.556 −0.181 0.107 Not supported Perception of Justice → Panic Buying −0.118 2.018 0.044 −0.237 −0.011 Supported Perception of Risk → Panic Buying 0.165 2.874 0.004 0.065 0.281 Supported Social Effect → Panic Buying 0.771 19.295 <0.001 0.691 0.850 Supported Transparency → Perception of Justice 0.575 8.079 <0.001 0.448 0.707 Supported Note : β = path coefficient; CI = bootstrap confidence interval. substantially contributes to perception of justice. In addition, perception of justice demonstrates a negative relationship with panic buying behavior, suggesting that stronger perceptions of fairness may reduce excessive purchasing behavior during post-disaster situations. The bootstrapping procedure in SmartPLS was conducted to test the hypothesized relationships using a two-tailed significance approach with a 5% significance level. A relationship was considered statistically significant when the t-statistic exceeded 1.96, and the 95% bootstrap confidence interval did not include zero. The analysis reported path coefficients, t-statistics, p-values, and bootstrap confidence interval bounds (2.5% and 97.5%) to evaluate the significance and stability of the estimated relationships. The accountability–panic buying relationship was interpreted cautiously because it did not reach the conventional significance threshold (β = 0.071, t = 1.779, p = 0.076, 95% CI [0.001, 0.148]). Similarly, the indirect effect of transparency on panic buying through perceived justice was statistically significant, as indicated by its bootstrap confidence interval, which did not include zero (β = −0.068, t = 1.995, p = 0.047, 95% CI [−0.140, −0.006]). These results provide a more rigorous assessment of the hypothesized relationships by considering the bootstrap confidence intervals alongside the p-values rather than relying solely on statistical significance thresholds. Based on Table 8, accountability had no statistically significant direct effect on panic buying behavior (β = 0.071, t = 1.779, p = 0.076), and therefore H1 was not supported. The estimated effect was positive but did not reach the conventional 5% significance level. Accordingly, accountability alone does not provide sufficient evidence to explain variations in panic-buying behavior among communities affected by post-flood conditions. This finding suggests that the presence of accountability mechanisms may not be sufficient to influence purchasing behavior when communities continue to experience uncertainty, perceived scarcity, and information constraints during the post-disaster recovery period. Accountability also had no statistically significant effect on perceptions of justice (β = −0.048, t = 0.589, p = 0.556; 95% CI [−0.181, 0.107]), resulting in the rejection of H2. The confidence interval includes zero, further indicating that the observed relationship is not statistically distinguishable from zero. This finding suggests that accountability practices may not automatically translate into perceptions of justice in the management and distribution of post-disaster resources. In contrast, perceived justice had a significant negative effect on panic buying behavior (β = −0.118, t = 2.018, p = 0.044; 95% CI [−0.237, −0.011]), supporting H3. Because the bootstrap confidence interval does not include zero, the negative relationship is statistically supported. This finding indicates that stronger perceptions of fairness in resource distribution, equitable treatment, and aid management are associated with lower tendencies toward panic buying. Perceived justice therefore appears to function as an important psychological mechanism that can reduce excessive purchasing behavior under post-disaster conditions. Perceived risk had a positive and statistically significant effect on panic buying behavior (β = 0.165, t = 2.874, p = 0.004; 95% CI [0.065, 0.281]), supporting H4. The positive coefficient indicates that greater perceptions of uncertainty, potential scarcity, and disaster-related threats are associated with stronger tendencies toward panic buying. This result suggests that risk perception may motivate individuals to engage in precautionary purchasing as a form of self-protection when facing crisis-related uncertainty. Social effect demonstrated the strongest positive and statistically significant effect on panic buying behavior (β = 0.771, t = 19.295, p < 0.001; 95% CI [0.691, 0.850]), supporting H5. The magnitude of the coefficient indicates that social influences constitute a dominant predictor of panic- buying behavior in the model. This finding suggests that purchasing decisions during crisis conditions are shaped not only by individual perceptions but also by social pressure, observed behavior, collective anxiety, and the tendency to imitate the actions of others. Thus, collective social dynamics are particularly important in explaining panic buying following disasters. Finally, transparency had a positive and statistically significant effect on perceptions of justice (β = 0.575, t = 8.079, p < 0.001; 95% CI [0.448, 0.707]), supporting H6. The confidence interval does not include zero, confirming the statistical significance of the relationship. This finding indicates that greater openness of information, public access to information concerning policies and aid distribution, and clearer communication from relevant authorities are associated with stronger perceptions of justice. Nevertheless, the interpretation of this construct should remain attentive to the relative contribution of individual indicators and their measurement quality.

Table 8. Direct Effect Results

β = path coefficient; CI = bootstrap confidence interval. Source: Primary Data.

Mediation Analysis

The purpose of the mediation analysis was to determine whether the links between accountability, transparency, and Specific Indirect Path β t p 2.5% CI 97.5% CI Decision Accountability → Perception of Justice → Panic Buying 0.006 0.500 0.618 −0.015 0.030 Not supported Transparency → Perception of Justice → Panic Buying −0.068 1.995 0.047 −0.140 −0.006 Supported Note : β = specific indirect effect; CI = bootstrap confidence interval. Total Effect Path β t p 2.5% CI 97.5% CI Decision Accountability → Panic Buying 0.077* 1.847 0.065 0.001 0.158 Not supported Accountability → Perception of Justice −0.048 0.589 0.556 −0.181 0.107 Not supported Perception of Justice → Panic Buying −0.118 2.018 0.044 −0.237 −0.011 Supported Perception of Risk → Panic Buying 0.165 2.874 0.004 0.065 0.281 Supported Social Effect → Panic Buying 0.771 19.295 <0.001 0.691 0.850 Supported Transparency → Panic Buying −0.068 1.995 0.047 −0.140 −0.006 Supported Transparency → Perception of Justice 0.575 8.079 <0.001 0.448 0.707 Supported Note : β = total effect coefficient; CI = bootstrap confidence interval. panic buying behavior are mediated by perceptions of justice. The mediation analysis examined whether perceptions of justice mediated the relationships between accountability, transparency, and panic buying behavior. As reported in Table 9, the indirect effect of accountability on panic buying through perceptions of justice was not statistically significant (β = 0.006, t = 0.500, p = 0.618; 95% CI [−0.015, 0.030]). Because the bootstrap confidence interval includes zero, the mediation effect is not statistically supported. Thus, perceptions of justice do not significantly mediate the relationship between accountability and panic buying behavior. In contrast, perceptions of justice significantly mediated the relationship between transparency and panic buying behavior (β = −0.068, t = 1.995, p = 0.047; 95% CI [−0.140, −0.006]). The bootstrap confidence interval does not include zero, providing statistical evidence for the indirect effect. The negative indirect effect indicates that greater transparency is associated with stronger perceptions of justice, which in turn are associated with lower panic-buying behavior. Therefore, perceptions of justice function as a statistically significant mediating mechanism through which transparency is associated with reduced panic buying.

Table 9. Indirect Effect Results

β = specific indirect effect; CI = bootstrap confidence interval. Source: Primary Data.

Total Effect Analysis

The total effect analysis was conducted to evaluate the combined direct and indirect effects among constructs in the structural model. Table 10 presents the total effects of the structural relationships. Accountability had a positive but statistically insignificant total effect on panic buying (β = 0.077, t = 1.847, p = 0.065). Therefore, the total effect does not provide sufficient statistical evidence to support a significant relationship between accountability and panic buying at the 5% significance level. Perceived justice maintained a significant negative total effect on panic buying (β = −0.118, t = 2.018, p = 0.044; 95% CI [−0.237, −0.011]). Perceived risk also had a significant positive total effect on panic buying (β = 0.165, t = 2.874, p = 0.004; 95% CI [0.065, 0.281]). Social effect exhibited the strongest total effect on panic buying (β = 0.771, t = 19.295, p < 0.001; 95% CI [0.691, 0.850]). These results reinforce the importance of social influence, perceived risk, and perceived justice in explaining panic-buying behavior under post- disaster conditions. Transparency had a significant negative total effect on panic buying (β = −0.068, t = 1.995, p = 0.047; 95% CI [−0.140, −0.006]). This result is consistent with the significant indirect effect reported in Table 9, indicating that the relationship between transparency and panic buying operates through perceptions of justice. Transparency also had a significant positive total effect on perceptions of justice (β = 0.575, t = 8.079, p < 0.001; 95% CI [0.448, 0.707]), confirming the role of transparency in strengthening perceptions of fairness in post-disaster resource management. The findings of the study show that panic buying after the Aceh floods is more influenced by direct social and psychological pressures than by institutional accountability. Social influence emerged as the strongest predictor of panic buying behavior, while transparency influenced panic buying indirectly through the perception of fairness. In contrast, accountability did not show a significant influence on either the perception of fairness or on panic buying. The insignificance of the influence of accountability does not mean that accountability does not have a role in organizational governance. These results show that administrative and procedural accountability cannot necessarily be felt directly by the community when facing emergency conditions. In post-disaster situations, consumers tend to make decisions based on conditions that can be observed directly, such as the availability of goods, price stability, distribution mechanisms, and the purchasing behavior of the surrounding community. Thus, even if the organization has implemented a good accountability mechanism, the benefits will not affect consumer behavior if it is not accompanied by open information and policy results that can be felt by the public. In contrast, transparency has been shown to have a more important role through increased perceptions of justice. Clear information about stock availability, prices, distribution schedules, and purchasing restriction policies can reduce the uncertainty faced by the public during the crisis. The information disclosure provides confidence that the distribution process of goods is carried out fairly so as to reduce public concerns about potential shortages. With the increasing perception that distribution is done fairly, the urge to overbuy becomes lower. These results show that transparency is an important instrument in accounting and governance systems because it can increase public trust in the management of resources during emergencies. The dominance of social influence suggests that panic buying is a behavior that is highly influenced by social interaction. Consumers tend to use other people's purchasing behavior as an indicator of the possibility of a shortage of goods in the future. When the community witnesses relatives, neighbors, or community members making large purchases, this condition triggers the formation of herd behavior that encourages other individuals to take similar actions. This phenomenon explains why social influence has a greater impact than institutional accountability, because information from the social environment is received directly and quickly than information about organizational accountability mechanisms. In addition to social influences, risk perception has also been shown to increase people's tendency to make panic purchases. Consumers attribute flooding to the possibility of supply chain disruptions, limited availability of goods, and potential future price increases. This condition encourages people to make purchases in quantities that exceed normal needs as a form of anticipation of uncertainty. However, the results show that social influence has greater power than risk perception, so it can be indicated that risk perception increases when individuals observe excessive buying behavior by others. The findings of this study strengthen the understanding of the role of accounting in managing crisis conditions. Accountability, transparency, and the perception of justice do not work in isolation in influencing people's behavior. Accountability serves as the basis for organizational accountability in managing resources and the distribution of goods, transparency is a means to communicate this information to the public, while the perception of justice is the result of public evaluation of the implementation of the policies implemented. Therefore, the effectiveness of an accounting system in reducing panic buying depends not only on the existence of accountability mechanisms, but also on the ability of the organization to convey information openly so that the public feels that there is fair treatment. Practically, the government, disaster management agencies, and business actors need to provide accurate, consistent, and easily accessible information on stock conditions, prices, distribution mechanisms, and purchasing restriction policies during the emergency response period. In addition, given the strong social influence on panic buying, public communication strategies need to involve community leaders, community leaders, and trusted media to reduce the spread of inaccurate information and prevent the formation of excessive perceptions of scarcity. Thus, transparent information management and the development of a perception of justice in the community are important factors in reducing panic purchasing behavior after floods. Overall, the results showed that post-flood panic buying in Aceh was more influenced by social influences and risk perception, while accounting contributed through governance mechanisms that were manifested in transparency and the perception of justice. Accountability remains an important element in organizational management, but its effectiveness in influencing people's behavior depends on the extent to which it is transparently communicated and results in the perception that the policies implemented have been implemented fairly.

Table 10. Total Effect Results

β = total effect coefficient; CI = bootstrap confidence interval. Source: Primary Data.

Discussion of Key Relationships

Accountability and Panic Buying Behavior

The findings revealed that accountability did not significantly influence panic buying behavior among post- flood communities in Aceh (β = 0.071, p = 0.076). This indicates that accountability mechanisms alone were insufficient to directly reduce excessive purchasing behavior during disaster situations. In emergency contexts, communities tend to prioritize immediate survival concerns, scarcity perceptions, and emotional responses rather than evaluating formal accountability systems implemented by institutions or authorities. These findings partially differ from previous studies emphasizing the importance of accountable governance in strengthening public trust during crises (Lang et al., 2025). For instance, studies in disaster governance literature argue that accountability can improve institutional legitimacy and public confidence in resource distribution systems. However, this study suggests that accountability may function as a more abstract governance mechanism that is less immediately visible to communities experiencing urgent economic uncertainty and supply disruptions. The insignificant relationship may also reflect the contextual characteristics of post-disaster communities in Aceh, where behavioral reactions are more strongly shaped by collective social dynamics and perceived scarcity rather than formal administrative accountability. This finding highlights that accountability mechanisms require complementary dimensions, such as transparency and effective communication, to produce meaningful behavioral impacts during disaster situations.

Accountability and Perception of Justice

The results further demonstrated that accountability did not significantly influence perception of justice (β = -0.048, p = 0.556). This finding suggests that respondents did not directly associate institutional accountability with fairness perceptions in the context of post-flood disaster management. One possible explanation is that communities evaluate fairness primarily through tangible experiences, such as equal access to goods, visible aid distribution, and transparent public communication, rather than through formal accountability procedures. In disaster conditions, fairness is often perceived through practical outcomes rather than administrative mechanisms. This finding contrasts with Islamic governance theories emphasizing accountability as central principles for achieving justice and social welfare (Hafidz et al., 2025). However, the results indicate that accountability values may not automatically translate into perceived fairness unless they are operationalized through visible and transparent actions that directly affect community welfare.

Perception of Justice and Panic Buying Behavior

The study found that perception of justice negatively and significantly influenced panic buying behavior (β = -0.118, p = 0.044). This finding indicates that stronger perceptions of fairness reduce the tendency of individuals to engage in excessive purchasing behavior during post-disaster situations. This result supports behavioral theories suggesting that perceived fairness reduces uncertainty, anxiety, and defensive consumption behavior (Wen & Chi, 2013). When communities believe that resources are distributed fairly and that access to essential goods is maintained equitably, they are less likely to engage in panic-driven purchasing. The findings also align with Islamic economic principles emphasizing distributive justice and social balance as mechanisms for maintaining economic stability (Khalidin et al., 2024). Fairness perceptions may strengthen collective trust and reduce fear-driven competition over scarce resources. Therefore, justice perception emerges as an important psychological mechanism for mitigating irrational consumer behavior during disaster conditions. Compared with previous studies on panic buying during crises such as COVID-19, this study reinforces the argument that emotional stability and trust in distribution systems are essential factors in reducing panic purchasing behavior (Omar et al., 2021). However, this research extends prior literature by emphasizing the role of accounting dimensions and justice perception within flood-disaster contexts in Aceh.

Perception of Risk and Panic Buying Behavior

The results showed that panic buying behavior was positively and significantly influenced by perception of risk (β = 0.165, p = 0.004). This finding suggests that people are more prone to participate in panic buying behavior when they perceive greater degrees of uncertainty, scarcity, and disaster- related concerns. This finding is consistent with protection motivation theory and risk perception literature, which explain that individuals respond to perceived threats through preventive and self- protective actions, including stockpiling essential goods (Garbe et al., 2020). During post-flood situations, fear of shortages, price increases, and supply disruptions likely intensified consumers’ motivations to purchase goods excessively. The findings also support prior studies conducted during pandemic and disaster situations, where perceived uncertainty became a major driver of irrational purchasing behavior (Ghouri et al., 2024). However, this study contributes additional insight by demonstrating that risk perception remains highly influential even in localized disaster contexts, such as flood-affected communities in Aceh. These results indicate that disaster communication strategies should focus not only on logistical responses but also on reducing public uncertainty and fear through accurate, timely, and transparent information dissemination.

Social Effect and Panic Buying Behavior

One of the most significant findings of this study is that social effect had the strongest positive influence on panic buying behavior (β = 0.771, p = 0.000). This indicates that panic buying behavior is heavily shaped by social influence, observational learning, and collective behavioral responses within communities. This finding supports social contagion theory, which argues that individuals often imitate the actions of others during uncertain situations (Prentice et al., 2021) and (Vedadi et al., 2021). In post-disaster environments, observing neighbors, relatives, or community members purchasing large quantities of goods may trigger similar behavior among other individuals, creating a chain reaction of panic consumption. The exceptionally large effect size further confirms that panic buying is not merely an individual economic decision but also a socially constructed behavior influenced by group dynamics and collective fear. This result aligns with previous studies conducted during global crises, including the COVID-19 pandemic, where herd behavior and social influence significantly contributed to stockpiling behavior. Within the Acehnese social context, which is characterized by strong communal relationships and collective interactions, social influence may become even more dominant in shaping behavioral responses. Consequently, community-based intervention strategies and social communication mechanisms are essential for mitigating panic buying during disaster conditions.

Transparency and Perception of Justice

The findings demonstrated that transparency significantly and positively influenced perception of justice (β = 0.575, p = 0.000). This indicates that transparent communication, openness of information, and clarity regarding disaster- related policies substantially strengthen public perceptions of fairness. This finding is highly consistent with accounting principles emphasizing transparency (tabligh) as a fundamental mechanism for maintaining trust and social accountability (Shalhoob, 2025) and (Mahyudin et al., 2025). Transparent dissemination of information regarding aid distribution, price stabilization, and availability of goods may reduce suspicion, uncertainty, and perceived inequality among affected communities. The results also support governance literature suggesting that transparency enhances procedural fairness and public confidence in institutional decision-making (Wang & Guan, 2022). In disaster situations, transparent communication becomes particularly important because uncertainty and misinformation can intensify social panic and irrational consumer behavior. By showing that transparency not only enhances governance quality but also indirectly reduces panic buying behavior through the mediating role of justice perception, this study expands on earlier findings.

Mediation Role of Perception of Justice

The results of the mediation study showed that while perception of justice did not mediate the association between accountability and panic buying behavior, it did significantly mediate the relationship between transparency and panic buying behavior (β = -0.068, p = 0.047). This finding suggests that transparency influences panic buying behavior indirectly by strengthening public perceptions of fairness (Mao et al., 2022) and (T. Chen et al., 2022). When communities receive clear, open, and reliable information during disaster conditions, they are more likely to perceive equitable treatment, which subsequently reduces anxiety- driven purchasing behavior. The mediation result highlights the importance of psychological and social mechanisms in disaster-related consumer behavior. Transparency alone may not directly suppress panic buying; however, it can foster trust and fairness perceptions that ultimately stabilize community purchasing patterns. From the perspective of accounting, this finding reinforces the argument that transparency and justice are interconnected ethical principles that contribute to social stability and economic balance during crisis situations.

Theoretical and Practical Implications

Theoretically, this study contributes to the literature on panic buying behavior by integrating accounting dimensions, behavioral economics, and disaster management perspectives within a SEM-PLS framework. The findings demonstrate that social and psychological factors play more dominant roles in shaping panic buying behavior than formal accountability mechanisms alone. Practically, the results suggest that policymakers and disaster management authorities should prioritize transparent communication strategies, fair distribution systems, and community-based interventions to reduce panic buying behavior during post-disaster situations. Efforts to reduce misinformation, strengthen social trust, and improve public perceptions of fairness may help stabilize consumer behavior and minimize supply chain disruptions. Furthermore, the dominance of social influence indicates that governments and local institutions should collaborate with community leaders, religious figures, and social networks to disseminate accurate information and discourage collective panic behavior during emergency situations.

Limitations and Cautions

Several limitations should be considered when interpreting the findings of this study. First, this research employed a cross- sectional design, meaning that the data were collected at a single point in time during the post-flood period in Aceh. Consequently, the study captures respondents’ perceptions and behavioral tendencies only within a specific temporal context. Behavioral responses such as panic buying may change over time as disaster conditions stabilize, government interventions improve, and market availability returns to normal. Therefore, causal interpretations should be approached cautiously despite the structural relationships identified through SEM-PLS analysis. Second, the study focused exclusively on three flood- affected areas in Aceh, namely Aceh Timur, Kota Langsa, and Aceh Tamiang. Although these regions represent areas significantly impacted by flooding, the findings may not fully generalize to other disaster-prone regions with different socio- economic characteristics, cultural dynamics, infrastructure conditions, or governance systems. Community responses to disasters can vary substantially depending on local economic resilience, accessibility to goods, and social structures. Third, this study relied on self-reported survey data, which may be subject to common method bias, response bias, and social desirability bias. Respondents may have underestimated or overestimated their purchasing behavior, perceptions of fairness, or perceptions of risk due to emotional, social, or psychological considerations. Although procedural measures were implemented to ensure anonymity and reduce response bias, the possibility of subjective interpretation remains unavoidable in behavioral research. Fourth, the study primarily examined social effect and perception of risk as control variables, while other potentially influential factors were not incorporated into the model. Variables such as income level, education, religiosity, trust in government, emotional distress, media exposure, and supply chain accessibility may also influence panic buying behavior during disaster situations. The exclusion of these variables may limit the comprehensiveness of the explanatory model. Additionally, the transparency construct demonstrated one indicator (TRA1) with a relatively lower loading value compared to other indicators, although it remained within acceptable thresholds for SEM-PLS analysis. This suggests that future studies may further refine and strengthen the operationalization of transparency indicators to improve construct measurement quality. Another important caution relates to the interpretation of accountability within disaster contexts. The findings indicate that accountability did not significantly influence panic buying behavior or perception of justice. However, this does not imply that accountability lacks importance in disaster governance. Rather, the results suggest that accountability may operate indirectly, institutionally, or over longer time horizons that are not immediately perceived by affected communities during emergency conditions. Finally, because this study was conducted within the socio- cultural context of Aceh, which is characterized by strong communal values and social norms, caution should be exercised when generalizing the findings to regions with different cultural orientations. Social influence and collective behavioral patterns may function differently across societies with varying levels of collectivism and social cohesion. Despite these limitations, the study provides important empirical insights into the relationship between accounting dimensions, justice perception, social influence, and panic buying behavior in post-disaster settings. The findings contribute to the growing literature on behavioral responses during crises and provide a foundation for future disaster governance and consumer behavior research.

Recommendations for Future Research

Future research is encouraged to expand the scope and depth of investigation regarding panic buying behavior within disaster and crisis contexts. First, subsequent studies should consider employing longitudinal research designs to observe changes in consumer behavior over different disaster phases, including pre-disaster preparedness, emergency response periods, and post-disaster recovery stages. A longitudinal approach would provide a more comprehensive understanding of how perceptions of risk, fairness, and social influence evolve over time and influence purchasing behavior dynamically. Second, future studies may broaden the geographical scope by including additional disaster-prone regions across Indonesia or other countries with diverse socio-economic and cultural characteristics. Comparative studies between regions, urban and rural communities, or different disaster types (e.g., floods, earthquakes, pandemics, or droughts) would enrich understanding regarding contextual differences in panic buying behavior and governance responses. Third, researchers are encouraged to incorporate additional explanatory variables that may strengthen the predictive capability of the model. Variables such as trust in government, media exposure, emotional anxiety, religiosity, financial literacy, digital information consumption, perceived scarcity, and supply chain resilience may provide deeper insights into behavioral responses during crises. Integrating psychological and socio-economic dimensions could produce a more holistic behavioral framework. Fourth, future studies should further explore the role of Islamic values and ethical governance principles in shaping community responses during disasters. Variables derived from Islamic behavioral economics, such as maqashid sharia, social solidarity, trustworthiness, moderation, and ethical consumption behavior, may provide valuable theoretical extensions to the current model. Additionally, future researchers may examine moderating variables that potentially strengthen or weaken the relationships identified in this study. Demographic factors such as age, income level, educational background, and gender may influence how individuals perceive fairness, risk, and social pressure during emergency situations. Moderation analysis could provide more nuanced insights into vulnerable population groups. Methodologically, future studies are also encouraged to combine quantitative and qualitative approaches through mixed-method designs. In-depth interviews, focus group discussions, and ethnographic observations may help uncover emotional, cultural, and social mechanisms underlying panic buying behavior that cannot be fully captured through survey-based analysis alone. Furthermore, future research could investigate the influence of digital ecosystems and social media platforms on panic buying behavior during disaster situations. The rapid dissemination of information, rumors, and viral purchasing trends through digital networks may significantly intensify collective panic and irrational consumption patterns. Understanding the role of digital communication may contribute to more effective crisis communication strategies. Finally, future studies should examine the effectiveness of policy interventions aimed at mitigating panic buying behavior, such as price control mechanisms, transparent supply chain management, rationing systems, and community- based communication strategies. Evaluating practical interventions would provide valuable policy recommendations for governments, humanitarian agencies, and disaster management institutions in managing public behavior during crisis conditions. Overall, future research has substantial opportunities to deepen theoretical development and practical understanding of panic buying behavior by integrating behavioral economics, accounting, disaster governance, and socio-cultural perspectives within broader and more diverse research settings.

Conclusion

This study shows that panic buying among post-flood communities in Aceh was driven primarily by social and psychological factors. Social effect was the strongest predictor of panic buying, while perceived risk significantly increased panic buying behavior. In contrast, perceived justice significantly reduced panic buying. Transparency did not reduce panic buying directly but operated indirectly by strengthening perceived justice. This finding indicates that clear and accessible information regarding prices, product availability, and distribution mechanisms can enhance perceptions of fairness and subsequently reduce excessive purchasing behavior. Accountability, however, was not significantly related to either perceived justice or panic buying, suggesting that formal responsibility alone may be insufficient unless it is translated into visible and transparent practices. The main contribution of this study is the demonstration that accounting principles influence post-disaster consumer behavior through different mechanisms. Transparency and justice have practical behavioral relevance, whereas accountability requires stronger public visibility and implementation. Governments, disaster management agencies, and business actors should therefore prioritize timely market information, fair distribution, and community-based communication to reduce uncertainty, limit social contagion, and maintain market stability during disasters.

Author Contributions

All authors actively contribute to the research and development process of this manuscript. The contributions of each author are described as follows: The first author is responsible for conceptualizing the research, identifying the research problem, developing the conceptual framework, preparing the research methodology, designing the research instrument, collecting data, processing and analyzing data using the Structural Equation Modeling– Partial Least Squares (SEM-PLS) approach, interpretation of the research results, and preparation of the initial draft of the manuscript. The first author is also responsible for coordinating the overall research and communication process during the manuscript revision process. The second author contributes to the development of theoretical foundations, search and evaluation of relevant literature, preparation of theoretical arguments, analysis of the relationship between research constructs, and provides substantive input to the interpretation of research results and strengthening the discussion section. The second author also plays a role in refining the manuscript to comply with the standards of scientific publications. The third author contributes to the validation of research methodology, evaluation of statistical analysis approaches, checking the consistency of structural model test results, interpretation of quantitative analysis results, and providing input on technical

and academic aspects in manuscript preparation. The fourth author contributes to the manuscript editing process, critical review of the content of the article, refinement of the presentation of research results, checking the suitability between research objectives, methods, and conclusions, and helping to ensure the final quality of the manuscript before the publication submission process. All authors have conducted a critical review of the final version of the manuscript, given approval to the content of the article to be published, and are responsible for all aspects of this research and publication.

Acknowledgements

The authors express their appreciation and gratitude to all respondents from East Aceh, Langsa City, and Aceh Tamiang who have participated in this study and are willing to share their experiences and views regarding post-flood conditions. The contribution of the respondents is invaluable in providing empirical data that supports the implementation of this research. The authors also expressed their gratitude to individuals, institutions, and parties who provided support during the research process but did not meet the criteria as authors. Special appreciation was given to research assistants, local community leaders, and field support parties who assisted in the coordination process, access to research sites, distribution of research instruments, and data collection. In addition, the authors appreciate the contributions of academic peers and colleagues who provide input, methodological suggestions, and critical review during the process of research development and manuscript drafting. This support helps to improve the quality of analysis and presentation of research results. The authors also expressed their gratitude to the Faculty of Islamic Economics and Business, Langsa State Islamic Religious Institute, for the academic support, research facilities, and scientific environment that supported the implementation of this research. No external financial support was received for this research, and none of the parties mentioned in this section had a role in the authorship, final analysis, or publication decisions of the manuscript.

References

Aceh Disaster Management Agency. (2025). Report on the Flood Situation in Aceh – November 27, 2025 [Disaster Report]. Aceh Disaster Management Agency (BPBA). https://bpba.acehprov.go.id/berita/kategori/bencana/banjir-di-16- kabupaten-kota-provinsi-aceh-per-27-november-20759-jiwa- mengungsi-satu-orang-hilang-terseret-arus-banjir

Billore, S., & Anisimova, T. (2021). Panic buying research: A systematic literature review and future research agenda. International Journal of Consumer Studies, 45(4), 777–804. https://doi.org/10.1111/ijcs.12669

Chen, T., Jin, Y., Yang, J., & Cong, G. (2022). Identifying emergence process of group panic buying behavior under the COVID-19 pandemic. Journal of Retailing and Consumer Services, 67, 102970–102970. https://doi.org/10.1016/j.jretconser.2022.102970 Chen, Y., Rajabifard, A., Sabri, S., Potts, K. E., Laylavi, F., Xie, Y., & Zhang, Y. (2020). A discussion of irrational stockpiling behaviour during crisis. Journal of Safety Science and Resilience, 1(1), 57–58. https://doi.org/10.1016/j.jnlssr.2020.06.003

Cheong, B. C. (2024). Transparency and accountability in AI systems: Safeguarding wellbeing in the age of algorithmic decision-making. Frontiers in Human Dynamics. https://doi.org/10.3389/fhumd.2024.1421273

Dulam, R., Furuta, K., & Kanno, T. (2021). Consumer Panic Buying: Realizing Its Consequences and Repercussions on the Supply Chain. Sustainability. https://doi.org/10.3390/su13084370

Fithri, P., Hasan, A., Syafrizal, S., & Games, D. (2024). Validation Studies a Questionnaire Developed to Measure Incubator Business Technology Performance using PLS-SEM Approach. Andalasian International Journal of Applied Science, Engineering and Technology. https://doi.org/10.25077/aijaset.v4i1.132

Garbe, L., Rau, R., & Toppe, T. (2020). Influence of perceived threat of Covid-19 and HEXACO personality traits on toilet paper stockpiling. PLoS ONE, 15. https://doi.org/10.1371/journal.pone.0234232

Ghouri, M. W. A., Wang, G., Hussain, M. A., Li, Z., & Chin, T. (2024). Provoking Buying Behaviors Amid Crises: Unfolding the Underlying Mechanisms of Psychological Impairments. International Journal of Mental Health Promotion. https://doi.org/10.32604/ijmhp.2024.044759 Hafidz, J., Saktiawan, M. D., Wiranto, A. P., Noviyansyah, A., & Osman, A. K. (2025). Contemporary Legal Accountability Reform in Public Procurement: A Framework Integrating Ethical Norms and Anti- Corruption Mechanisms. MILRev: Metro Islamic Law Review. https://doi.org/10.32332/milrev.v4i1.10664

Hair, J., & Alamer, A. (2022). Partial Least Squares Structural Equation Modeling (PLS-SEM) in second language and education research: Guidelines using an applied example. Research Methods in Applied Linguistics. https://doi.org/10.1016/j.rmal.2022.100027

Ismail, S. (2022). Perception of the Malaysian Federal Government accountants of the usefulness of financial information under an accrual accounting system: A preliminary assessment. Meditari Accountancy Research. https://doi.org/10.1108/medar-04-2020-0845

Jazemi, R., Farahani, S., Otieno, W., & Jang, J. (2024). Review on Panic Buying Behavior during Pandemics: Influencing Factors, Stockpiling, and Intervention Strategies. Behavioral Sciences, 14. https://doi.org/10.3390/bs14030222

Jhantasana, C. (2023). Should A Rule of Thumb be used to Calculate PLS-SEM Sample Size. Asia Social Issues. https://doi.org/10.48048/asi.2023.254658

Khalidin, B., Musa, A., Fardesi, M., & Ulfia, N. (2024). Islamic Economics towards the Sustainability of Economic Development. INTERNATIONAL JOURNAL OF SOCIAL SCIENCE HUMANITY & MANAGEMENT RESEARCH. https://doi.org/10.58806/ijsshmr.2024.v3i11n16

Kock, N., & Hadaya, P. (2018). Minimum sample size estimation in PLS‐SEM: The inverse square root and gamma‐exponential methods. Information Systems Journal, 28, 227–261. https://doi.org/10.1111/isj.12131

Lang, S., Bian, Q., & Ben. (2025). Crisis Governance and Public Trust: Revealing the Chain Mediation Effect of Social Fairness and Well-Being. Journal of Asian Governance. https://doi.org/10.63851/001c.140948

Li, J., & Lay, Y. F. (2024). Examining the reliability and validity of self-efficacy beliefs, stress, perceived teachers’ support and academic burnout scales using the PLS-SEM approach. Humanities and Social Sciences Letters. https://doi.org/10.18488/73.v13i1.3987

Mahyudin, M., Sugianto, S., & Majid, M. Abd. (2025). Implementation of Islamic Accounting in Supporting Islamic Economic Principles: A Systematic Literature Review (SLR). JAMPARING: Jurnal Akuntansi Manajemen Pariwisata Dan Pembelajaran Konseling. https://doi.org/10.57235/jamparing.v3i2.6675

Mao, Q., Hou, J.-X., & Xie, P. (2022). Dynamic Impact of the Perceived Value of Public on Panic Buying Behavior during COVID-19. Sustainability. https://doi.org/10.3390/su14094874

Messner, W. (2023). Hypothesis Testing and Machine Learning: Interpreting Variable Effects in Deep Artificial Neural Networks using Cohen’s f2. ArXiv, abs/2302.01407. https://doi.org/10.1016/j.asoc.2023.110729

Mnif, M., Chikh, S., Watelain, E., & Jarraya, M. (2025). Social and Emotional Effect on Human Behavior. Journal of Motor Behavior, 57, 737–755. https://doi.org/10.1080/00222895.2025.2550375

Nazarenko, O., Zhou, K.-P., Nazarenko, I., & Harkusha, S. (2024). Impact of transparent economy on high-tech enterprises in the modern business environment of China. Problems and Perspectives in Management. https://doi.org/10.21511/ppm.22(3).2024.37

Noda, S., & Teramoto, K. (2024). A dynamic model of rational “panic buying.” Quantitative Economics. https://doi.org/10.3982/qe2126

Omar, N., Nazri, M., Ali, M., & Alam, S. (2021). The panic buying behavior of consumers during the COVID-19 pandemic: Examining the influences of uncertainty, perceptions of severity, perceptions of scarcity, and anxiety. Journal of Retailing and Consumer Services, 62, 102600–102600. https://doi.org/10.1016/j.jretconser.2021.102600

Pavlov, G., Pavlov, G., Maydeu-Olivares, A., Maydeu-Olivares, A., & Shi, D. (2020). Using the Standardized Root Mean Squared Residual (SRMR) to Assess Exact Fit in Structural Equation Models. Educational and Psychological Measurement, 81, 110–130. https://doi.org/10.1177/0013164420926231 Peran, A., Syariah, A., Transparansi, U., Keberlanjutan, D., Lembaga, E., Syariah, P., Maal, B., Tamwil, W., Khazanah, M., Rafly, P. M., Ramadani, D., Sari, B.

P., Azwari, P. C., Syariah, E., & Palembang, M. (2025). Analisis Peran Akuntansi Syariah Untuk Transparansi Dan Keberlanjutan Ekonomi Lembaga Pembiayaan Syariah Baitul Maal Wa Tamwil (BMT) Mitra Khazanah Palembang. AKUNTANSI 45. https://doi.org/10.30640/akuntansi45.v6i1.4281

Prentice, C., Quach, S., & Thaichon, P. (2021). Antecedents and consequences of panic buying: The case of COVID‐19. International Journal of Consumer Studies. https://doi.org/10.1111/ijcs.12649

Priolo, G., Vignoli, M., & Nielsen, K. (2025). Risk perception and safety behaviors in high-risk workers: A systematic literature review. Safety Science. https://doi.org/10.1016/j.ssci.2025.106811

Russo, D., & Stol, K.-J. (2021). PLS-SEM for Software Engineering Research. ACM Computing Surveys (CSUR), 54, 1–38. https://doi.org/10.1145/3447580

Shalhoob, H. (2025). ESG Disclosure and Financial Performance: Survey Evidence from Accounting and Islamic Finance. Sustainability. https://doi.org/10.3390/su17041582

Solano-Silva, D., Guadalupe, C., & Sam-Castañeda, E. (2025). Ni justo ni legítimo: The role of social status and neoliberal context on perceived social justice in Latin America and its political consequences. The British Journal of Social Psychology, 64. https://doi.org/10.1111/bjso.12894

Vedadi, A., Warkentin, M., & Dennis, A. (2021). Herd behavior in information security decision-making. Inf. Manag., 58, 103526. https://doi.org/10.1016/j.im.2021.103526

Wang, Q., & Guan, Z. (2022). Can Sunlight Disperse Mistrust? A Meta-Analysis of the Effect of Transparency on Citizens’ Trust in Government. Journal of Public Administration Research and Theory. https://doi.org/10.1093/jopart/muac040

Wen, B., & Chi, C. (2013). Examine the cognitive and affective antecedents to service recovery satisfaction: A field study of delayed airline passengers. International Journal of Contemporary Hospitality Management, 25, 306–327. https://doi.org/10.1108/09596111311310991

Wu, S., Shen, Y., Geng, Y., Chen, T., & Xi, L. (2023). Consumer Panic Buying Behavior and Supply Distribution Strategy in a Multiregional Network after a Sudden Disaster. Systems, 11(2), 110. https://doi.org/10.3390/systems11020110

Yuen, K. F., Leong, J. Z. E., Wong, Y., & Wang, X. (2021). Panic buying during COVID-19: Survival psychology and needs perspectives in deprived environments. International Journal of Disaster Risk Reduction, 62, 102421–102421. https://doi.org/10.1016/j.ijdrr.2021.102421

Yuen, K. F., Tan, L., Wong, Y., & Wang, X. (2022a). Social determinants of panic buying behaviour amidst COVID-19 pandemic: The role of perceived scarcity and anticipated regret. Journal of Retailing and Consumer Services, 66, 102948–102948. https://doi.org/10.1016/j.jretconser.2022.102948

Yuen, K. F., Tan, L., Wong, Y., & Wang, X. (2022b). Social determinants of panic buying behaviour amidst COVID-19 pandemic: The role of perceived scarcity and anticipated regret. Journal of Retailing and Consumer Services, 66, 102948–102948. https://doi.org/10.1016/j.jretconser.2022.102948