Behavioural and Institutional Triggers of Corruption: A Governance and Risk Management Perspective

The term “corruption” originates from the Latin word “corruption” and generally denotes the erosion of trust within an organisation or society. Nonetheless, various interpretations of corruption exist, with T​r​a​n​s​p​a​r​e​n​c​y​ ​I​n​t​e​r​n​a​t​i​o​n​a​l​ ​(​2​0​2​0​b​) providing the most widely accepted definition: “the abuse of entrusted power for private gain.” Corruption not only undermines confidence but also has detrimental effects on democratic values, economic progress, exacerbates inequality, fosters poverty and societal rifts, erodes trust, and precipitates environmental crises (T​r​a​n​s​p​a​r​e​n​c​y​ ​I​n​t​e​r​n​a​t​i​o​n​a​l​,​ ​2​0​2​0​a). Additionally, corruption is known to induce adverse behaviours and impede societal advancement (T​a​n​z​i​ ​&​ ​D​a​v​o​o​d​i​,​ ​1​9​9​7). It manifests in two primary forms. The first is when officials engage in “petty corruption” during interactions with the public. This type of corruption arises due to the delegation of power to ordinary citizens seeking to fulfil their needs (T​r​a​n​s​p​a​r​e​n​c​y​ ​I​n​t​e​r​n​a​t​i​o​n​a​l​,​ ​2​0​2​0​c). The second is a significant unresolved legal concern highlighted by Transparency International as “grand corruption,” which they define as “the abuse of power at a high level that benefits a few at the expense of many and causes serious and widespread harm to individuals and society and often goes unpunished” (T​r​a​n​s​p​a​r​e​n​c​y​ ​I​n​t​e​r​n​a​t​i​o​n​a​l​,​ ​2​0​1​6).

From a corporate governance and risk management perspective, corruption represents a failure of oversight, internal controls, and ethical culture rather than solely a legal or developmental issue. Understanding the behavioural and institutional triggers that enable corrupt conduct is therefore essential for boards, regulators, insurers, and risk managers seeking to design effective governance, compliance, and enterprise risk management frameworks (G​r​i​m​a​ ​e​t​ ​a​l​.​,​ ​2​0​2​1).

This study aims to highlight the underlying causes of corruption, examine the factors driving its occurrence, and propose recommendations for researchers to address deficiencies in corruption prevention. The prevention of corruption is paramount to the sustained viability of international development and global economic growth (A​l​l​d​r​i​d​g​e​,​ ​2​0​1​6). Given the escalating prevalence of corruption globally, this study is of significant importance in raising awareness, delineating the perpetrators and victims of corruption, elucidating its triggers, discerning its timing and location, unravelling the reasons behind its occurrence, and tracing its evolution.

This research is pivotal for the economic, sustainable, and social milieu, as the ramifications of corruption are so profound and insidious that they can precipitate a pernicious cycle. The Organisation for Economic Co-operation and Development (OECD) notes that, due to the clandestine nature of corruption, it is seldom investigated, as victims often do not seek reparation for their losses. Consequently, the OECD posits that combating such malfeasance may hinge on obtaining credible information from individuals who encounter corruption (A​s​i​a​n​ ​D​e​v​e​l​o​p​m​e​n​t​ ​B​a​n​k​ ​&​ ​O​E​C​D​,​ ​2​0​0​3).

Indeed, corruption, among other pretexting crimes, was documented in the second money laundering legislation of the European Union. Hence, efforts to prosecute bribery and secure convictions to stem corruption are ongoing. The OECD is steadfast in establishing the inherent nexus between corruption and money laundering. Following the OECD Declaration, the Financial Action Task Force contends that any criminal activity involving the acquisition or utilisation of funds inherently involves corruption (A​l​l​d​r​i​d​g​e​,​ ​2​0​1​6). However, understanding the triggers is imperative to curtail such clandestine activities as corruption.

The impetus to investigate the triggers of corruption stems from the absence of explicit legal norms and regulations, as well as the exploitation of governmental structures and bureaucracy. The absence of such regulations perpetuates corrupt practices. Conversely, corruption is a social phenomenon entrenched in intricate rules and closely monitored norms and conventions (B​l​u​n​d​o​ ​&​ ​O​l​i​v​i​e​r​ ​d​e​ ​S​a​r​d​a​n​,​ ​2​0​0​6). Citizen engagement, lack of press freedom, competition within the market and political arena, economic prosperity and liberties, trade openness and globalisation, political turbulence, ethnic and religious factors, urbanisation, organisational dynamics, and governmental frameworks all serve as triggers for corruption, facilitating its surreptitious occurrence and giving rise to maladies such as collusion, embezzlement, theft, fraud, extortion, bribery, nepotism, and commercial interference.

This study endeavours to uncover and delineate the unconscious and automatic triggers that intermittently manifest. These norms and rules exert an influence extending beyond the realm of corruption. Therefore, a deeper examination of individual and communal behaviours is warranted to grasp the underlying realities that must be acknowledged before effective mitigation strategies can be devised. It reframes corruption as a failure in behavioural and institutional risk management rather than solely a legal or ethical breach. By integrating behavioural risk theory, corporate governance, and enterprise risk management perspectives, the study identifies emotional, psychological, environmental, and structural triggers that undermine internal controls and oversight mechanisms. Unlike corruption perception indices that measure prevalence, this research provides a governance-oriented framework for identifying, assessing, and mitigating corruption risk within organisations. The findings contribute to the governance and risk management literature by offering empirically grounded insights relevant to boards, regulators, insurers, and risk professionals concerned with behavioural risk, control effectiveness, and ethical culture.

In doing this, the study seeks to explore and ascertain the following Research questions (RQ):

RQ1: What are the perceived triggers of corruption?

RQ2: How do participants rank the causes of corruption?

RQ3: Do these perceptions change with the different demographics?

These research questions are examined to inform governance structures, behavioural risk oversight, and organisational control frameworks. By addressing these objectives, the study aims to provide a comprehensive understanding of the multifaceted nature of corruption, its underlying drivers, and the potential influence of demographic factors on individuals’ perceptions of corruption. This research will contribute to the development of targeted interventions and policies that effectively combat corruption across diverse socio-cultural contexts.

The following theoretical propositions to guide this study were identified (Y​i​n​,​ ​2​0​0​9):

P1 (Behavioural Risk): Psychological and emotional factors, including moral disengagement and emotional intelligence, significantly influence individuals’ propensity to engage in corrupt behaviour.

P2 (Institutional Environment): Weak governance structures, poor regulatory enforcement, and tolerance of unethical norms increase corruption risk.

P3 (Social Norms and Power): Socialisation, power asymmetries, and group norms mediate the acceptance and normalisation of corrupt conduct.

P4 (Demographics): Demographic characteristics influence perceptions of corruption triggers but do not independently explain corruption behaviour.

B​a​n​d​u​r​a​ ​(​1​9​8​6​) further elaborates on the distinction between moral norms and self-regulation. Self-regulation operates by triggering various social and psychological mechanisms to disengage from inhumane behaviour. However, the presence of certain norms may undermine personal control, allowing individuals to justify different behaviours even when they adhere to similar moral standards. This complexity of human responsibility is depicted in Figure 1, where moral disengagement offers individuals numerous options in morally challenging situations. Additionally, reprehensible behaviour may be accompanied by deceptive language that persuades the victim to perceive it as socially acceptable.

After committing a morally harmful act, individuals often feel compelled to justify their wrongdoing. If accepted, this justification can harm both the personal and social aspects of the individual by rationalising the behaviour or suggesting that there was no choice. This rationalisation process is particularly evident in military contexts, where dedicated warriors may morally justify their actions as the defence of cherished principles and the preservation of humanity and the Earth (B​a​n​d​u​r​a​,​ ​1​9​9​9).

Emotions also play a significant role in corrupt behaviour. A​k​ı​n​ ​(​2​0​1​9​) argues that offering a bribe can evoke feelings of terror and tension, particularly due to the risk of exposure and legal repercussions. These emotions may diminish if the other party responds positively and participates in the corrupt act. However, residual fear and tension may persist, especially if the other party fails to uphold their end of the bargain. Additionally, anger can arise when individuals are coerced into paying bribes, stemming from a sense of injustice in the transaction (P​a​v​i​a​ ​e​t​ ​a​l​.​,​ ​2​0​2​1).

Sociological factors, as noted by S​u​t​h​e​r​l​a​n​d​ ​(​1​9​8​3​), suggest that social connections and group dynamics can influence behaviour by shaping individuals’ perceptions of acceptable norms. Cultural influences also play a significant role in shaping attitudes toward corruption. According to J​u​l​i​á​n​ ​&​ ​B​o​n​a​v​í​a​ ​(​2​0​2​0​), corrupt practices are often embedded within socio-cultural contexts, where group worldviews influence emotions, values, and priorities. Moreover, culture can either promote or deter corruption, with progressive cultures fostering traits such as secularism, justice, and community engagement, while stagnant cultures may perpetuate negative behaviours (H​a​r​r​i​s​o​n​,​ ​2​0​0​0).

Power dynamics also play a crucial role in corruption triggers. B​e​n​d​a​h​a​n​ ​e​t​ ​a​l​.​ ​(​2​0​1​5​) suggest that individuals who gain power may become more resistant to social norms and more prone to engaging in corrupt behaviour. Social conditions, including peer pressure and societal expectations, can further exacerbate tendencies toward corruption. For instance, individuals may conform to corrupt norms if they perceive their environment as accepting or tolerant of such behaviour (J​u​l​i​á​n​ ​&​ ​B​o​n​a​v​í​a​,​ ​2​0​2​0).

Environmental influences also shape corrupt behaviour. M​o​r​g​a​n​ ​(​2​0​2​0​) suggests that environments characterised by fairness and democracy tend to discourage corruption, while anarchy and corruption breed further misconduct. Furthermore, the prevalence of corruption in an environment can desensitise individuals to unethical behaviour and reduce feelings of guilt or shame associated with corrupt actions (W​u​ ​&​ ​Z​h​u​,​ ​2​0​1​6).

The internet has emerged as a powerful tool in the fight against corruption, offering access to information and fostering transparency. However, P​a​r​i​s​e​r​ ​(​2​0​1​1​) warns of the “filter bubble”, where individuals are exposed only to information that aligns with their existing beliefs and preferences. This selective exposure can reinforce biases and limit exposure to diverse perspectives, potentially impacting individuals’ attitudes and behaviours, including their susceptibility to corruption (see Figure 1).

The threefold filter bubble triggers distinct emotions and interactions to particular conduct. This framework linked the individual’s psychological state to others who eventually make up their social bubble, all driven by an algorithm. This framework enhances attitude stability and offers assurance and security for the individual inside the group (G​e​s​c​h​k​e​ ​e​t​ ​a​l​.​,​ ​2​0​1​9).

Demographic Indicators: Views on corruption often depend on demographic factors, including norms, traditions, politics, economics, history, customs, culture, sex, age, and educational levels (B​o​r​o​š​a​k​ ​&​ ​Š​u​m​a​h​,​ ​2​0​1​9).

Religion: Religion can influence attitudes toward corruption by fostering a sense of community and solidarity, shaping individualistic attitudes and behaviours in line with group standards (D​a​r​-​N​i​m​r​o​d​ ​&​ ​H​e​i​n​e​,​ ​2​0​1​1).

Genetics: Human actions and traits can influence perceptions and evaluations of behaviours, with genetic explanations for criminal behaviour potentially impacting how offenders are viewed and treated within the criminal justice system (L​e​b​o​w​i​t​z​ ​e​t​ ​a​l​.​,​ ​2​0​1​9).

Gender: There is a correlation between gender and corruption, with patriarchal countries exhibiting higher levels of corruption. Studies suggest that increased female representation in politics correlates with lower levels of corruption (B​a​u​h​r​ ​&​ ​C​h​a​r​r​o​n​,​ ​2​0​2​0; B​a​u​h​r​ ​&​ ​C​h​a​r​r​o​n​,​ ​2​0​2​1).

Education is viewed as a positive influence on individuals, fostering intellectual qualities and virtuous character traits. However, shortcomings in education systems may lead to corruption, as individuals lack the moral urge despite possessing knowledge (K​i​d​d​,​ ​2​0​1​9).

Region: Corruption spreads across borders due to factors such as cross-border trade integration and the absence of anti-corruption laws. Changes in institutional frameworks in one country can affect the corruption levels in neighbouring countries (B​e​c​k​e​r​ ​e​t​ ​a​l​.​,​ ​2​0​0​9).

These factors interact, shaping individuals’ attitudes and behaviours toward corruption. For instance, education may impact how individuals perceive corrupt behaviour, while gender dynamics and cultural norms influence corruption levels within societies. Understanding these complex interactions is crucial for developing effective anti-corruption strategies and promoting ethical behaviour across diverse socio-cultural contexts.

Corruption is commonly defined as the misuse of entrusted power for private gain and encompasses practices such as bribery, nepotism, embezzlement, and abuse of authority (N​y​e​,​ ​1​9​6​7; W​e​l​l​s​,​ ​2​0​1​7). While definitions vary across cultural and historical contexts, contemporary scholarship converges on the understanding that corruption undermines public interest, institutional legitimacy, and governance structures (A​l​b​i​s​u​ ​A​r​d​i​g​ó​ ​&​ ​H​o​u​g​h​,​ ​2​0​1​8; P​h​i​l​p​,​ ​1​9​9​7). International organisations, including Transparency International, the OECD, the IMF, and the World Bank, have adopted mainly Nye’s conceptualisation, emphasising the exploitation of public authority for personal benefit (T​r​a​n​s​p​a​r​e​n​c​y​ ​I​n​t​e​r​n​a​t​i​o​n​a​l​,​ ​2​0​2​0​b; W​o​r​l​d​ ​B​a​n​k​ ​G​r​o​u​p​,​ ​2​0​2​0).

The literature identifies corruption as a pervasive and multifaceted global phenomenon with extensive economic, social, political, environmental, and humanitarian consequences. Economically, corruption distorts markets, discourages investment, misallocates resources, and exacerbates inequality (A​k​h​m​e​t​o​v​ ​e​t​ ​a​l​.​,​ ​2​0​1​8; H​i​n​d​r​i​k​s​ ​e​t​ ​a​l​.​,​ ​1​9​9​9). While some scholars argue that corruption may temporarily alleviate bureaucratic inefficiencies in weak institutional environments (“greasing the wheels”), empirical evidence predominantly supports the “sand-the-wheels” hypothesis, demonstrating that corruption undermines long-term economic growth, efficiency, and trust (H​u​n​t​i​n​g​t​o​n​,​ ​1​9​6​8; M​é​o​n​ ​&​ ​W​e​i​l​l​,​ ​2​0​1​0; U​r​b​i​n​a​,​ ​2​0​2​0).

Socially and politically, corruption erodes public trust, weakens the rule of law, facilitates impunity, and disproportionately harms vulnerable populations by restricting access to healthcare, education, justice, and basic services (T​a​n​z​i​ ​&​ ​D​a​v​o​o​d​i​,​ ​1​9​9​7; U​n​i​t​e​d​ ​N​a​t​i​o​n​s​ ​O​f​f​i​c​e​ ​o​n​ ​D​r​u​g​s​ ​&​ ​C​r​i​m​e​,​ ​2​0​1​8). It is closely linked to organised crime, political instability, and weakened judicial systems, thereby perpetuating cycles of inequality, poverty, and institutional dysfunction (C​h​ê​n​e​,​ ​2​0​0​8; G​a​r​c​í​a​-​S​a​y​á​n​,​ ​2​0​1​8; T​r​a​n​s​p​a​r​e​n​c​y​ ​I​n​t​e​r​n​a​t​i​o​n​a​l​,​ ​2​0​1​9​a).

Corruption also poses a significant threat to environmental sustainability and the integrity of infrastructure. Weak regulatory enforcement enables illegal logging, pollution, overfishing, unsafe construction practices, and biodiversity loss, often with fatal consequences (M​c​D​o​n​n​e​l​l​,​ ​2​0​1​7; O​’​H​i​g​g​i​n​s​,​ ​2​0​0​6; S​i​n​h​a​ ​e​t​ ​a​l​.​,​ ​2​0​1​9). These effects directly undermine progress towards the United Nations’ 2030 Sustainable Development Goals, impeding economic, social, and environmental sustainability (W​o​r​l​d​ ​E​c​o​n​o​m​i​c​ ​F​o​r​u​m​,​ ​2​0​1​5).

The literature identifies multiple structural and contextual factors that trigger corruption, including bureaucratic inefficiency, weak legal systems, limited press freedom, political instability, natural resource endowments, low wages, poverty, and deficient governance structures (D​i​m​a​n​t​ ​&​ ​T​o​s​a​t​o​,​ ​2​0​1​8; T​a​n​z​i​,​ ​1​9​9​8). Emerging research further highlights the role of education, digitalisation, eGovernment, globalisation, contagion effects, and migration in shaping corruption dynamics (D​i​m​a​n​t​ ​e​t​ ​a​l​.​,​ ​2​0​1​5; S​h​a​b​b​i​r​ ​e​t​ ​a​l​.​,​ ​2​0​2​1; Y​o​u​ ​&​ ​N​i​e​,​ ​2​0​1​7). In addition to the structural and institutional factors proposed in the literature, the present study also draws on the interaction between psychological, social, environmental, and economic triggers of corruption, highlighting the role of emotional intelligence and behavioural drivers in precipitating corrupt conduct.

Cultural antecedents provide additional explanatory power. Drawing on H​o​f​s​t​e​d​e​ ​(​1​9​8​0​)’s framework, high power distance, uncertainty avoidance, and masculinity are associated with higher perceived corruption, whereas individualism correlates with lower corruption levels, a pattern reflected in Corruption Perception Index rankings (D​a​v​i​s​ ​&​ ​R​u​h​e​,​ ​2​0​0​3; T​r​a​n​s​p​a​r​e​n​c​y​ ​I​n​t​e​r​n​a​t​i​o​n​a​l​,​ ​2​0​2​1​a).

Measurement tools such as the Corruption Perception Index, Worldwide Governance Indicators, and Global Corruption Barometer are widely employed but face limitations in capturing implementation effectiveness and behavioural drivers (P​e​r​s​s​o​n​ ​&​ ​T​a​b​e​l​l​i​n​i​,​ ​2​0​0​3; T​r​a​n​s​p​a​r​e​n​c​y​ ​I​n​t​e​r​n​a​t​i​o​n​a​l​,​ ​2​0​1​9​b). Despite extensive anti-corruption initiatives—including whistle-blower protections, institutional reforms, and international agreements such as the OECD Anti-Bribery Convention evidence regarding their sustained effectiveness remains mixed (M​i​n​,​ ​2​0​1​9; O​E​C​D​,​ ​1​9​9​7; O​E​C​D​,​ ​2​0​0​9).

Behavioural frameworks such as the Fraud Triangle (C​r​e​s​s​e​y​,​ ​1​9​5​3; N​a​t​i​o​n​a​l​ ​W​h​i​s​t​l​e​b​l​o​w​e​r​ ​C​e​n​t​e​r​,​ ​2​0​2​0; S​u​t​h​e​r​l​a​n​d​,​ ​1​9​4​9), Neutralisation Theory (S​y​k​e​s​ ​&​ ​M​a​t​z​a​,​ ​1​9​5​7), and and cognitive psychological models (U​4​ ​A​n​t​i​-​C​o​r​r​u​p​t​i​o​n​ ​R​e​s​o​u​r​c​e​ ​C​e​n​t​r​e​,​ ​2​0​1​8) offer micro-level explanations, emphasising motivation, opportunity, rationalisation, and social acceptability as key drivers of corrupt behaviour. However, the literature highlights persistent gaps, particularly in psychological, organisational, and gender-based analyses of corruption, underscoring the need for more granular, interdisciplinary research (J​u​l​i​á​n​ ​&​ ​B​o​n​a​v​í​a​,​ ​2​0​2​0).

This study employed a mixed-methods approach, utilising both primary and secondary sources to obtain impartial facts regarding the inducements of corruption. Primary data were gathered through the administration of questionnaires to a diverse sample of participants aged 18 and above, drawn from both local and global populations accessed via social media and academic channels. This method enabled direct collection of insights from individuals, providing firsthand perspectives on the factors contributing to corruption. Complementing the primary data, secondary sources were utilised to deepen the understanding of corruption, its motivational drivers, and underlying causes. These secondary sources encompassed a wide range of materials, including journal papers, books, websites, newspaper articles, worldwide surveys, collected statistics, empirical studies, and data derived from various scholarly sources such as papers, periodicals, dissertations, and reports. The incorporation of secondary data facilitated a comprehensive analysis, drawing on the research techniques and findings of other scholars. This approach enabled the researcher to synthesise existing knowledge, validate findings, and identify additional potential sources of corruption, thus enriching the overall understanding of the phenomenon.

The sample size was determined using a sample size calculator, assuming a world population of 7.9 billion, a 5% margin of error, and a 95% confidence interval. The calculated optimal sample size required to generalise findings to the entire population was determined to be 385 individuals. However, 454 valid questionnaire responses were collected, surpassing the required sample size. Therefore, the findings can be confidently generalised to the entire population with the specified levels of confidence.

Non-probability purposive sampling was utilised to select the sample, as the survey was posted on social networking sites such as Facebook and LinkedIn, targeting specific contacts. Additionally, non-probability snowball sampling was employed, as participants were encouraged to repost the survey on their social networks and prompt their contacts to respond, thereby increasing participation.

To drive traffic to the questionnaire, the D​i​l​l​m​a​n​ ​e​t​ ​a​l​.​ ​(​2​0​0​8​) strategy was employed. This approach utilises social exchange theory as a framework to enhance response rates. Various subgroups on Facebook and LinkedIn were targeted, and, based on the groups’ objectives, assistance was sought to complete the survey, as the study aligns with the groups’ agendas. Continued efforts will be made to attract attention to groups focused on reducing corruption and supporting future academic endeavours. This strategic approach establishes purpose and attracts attention to the questionnaire, effectively managing it on social media platforms to enhance engagement. Information about the survey was provided in posted messages, emphasising how participation could benefit respondents, which fostered a positive regard towards the survey and accelerated participation.

Given the non-probability and social-media-based sampling approach, the findings should be interpreted as perception-based and exploratory, rather than representative of national populations.

We created an online questionnaire in Google Forms and administered it from January to March 2023. This questionnaire was broken into four sections, as per Table 1:

Moreover, respondents were asked one open-ended question in which they could express their viewpoints, experiences, and knowledge about what corruption means to them and why it occurs.

In the questionnaire, a Likert scale was used to assess respondents’ attitudes, beliefs, and behaviours regarding assertions about the adverse psychological effects, societal pressure, poor behaviour, and a corrupted environment. The respondents had the option to choose whether they (1) strongly agree, (2) somewhat agree, (3) neither agree nor disagree, (4) somewhat disagree, or (5) strongly disagree with the statements provided in the questionnaire.

This was done by joining various Facebook groups; including ‘Corruption without justice in the military’, ‘Gender and Women studies’, ‘Citizen Journalism & Discussion Community’, ‘Supporting Journalism Students and Graduates’, National Anti-Corruption Corruption Commission’, ‘History and World affairs’, ‘Women for Women (Malta)’ and others, allowing varied individuals to access this questionnaire and obtain a broad-spectrum study on what causes corruption.

Data collected from online surveys were analysed using SPSS 27. Various statistical methods were employed to compare the study’s findings with the literature review, including exploratory factor analysis with Cronbach’s alpha, the Friedman test, and multiple linear regression.

Construct validity is a crucial consideration for inferring about unobservable or latent variables, and factor analysis is a valuable technique for validating and measuring psychological constructs. Decisions regarding how many variables to retain are pivotal for several reasons. Given evidence of resilience across options for other decisions, such as the choice of factor-analytic technique and rotation method, factor retention decisions are of greater significance. Exploratory factor analysis involves categorising related variables into dimensions (H​a​y​t​o​n​ ​e​t​ ​a​l​.​,​ ​2​0​0​4). In this study, the number of assertions was reduced from 33 to 5 uncorrelated variables. Subsequently, Cronbach’s alpha was calculated to assess the internal consistency of the research findings. Cronbach’s alpha is a statistic used to demonstrate the fitness for purpose of tests and scales constructed or adapted for research projects (T​a​b​e​r​,​ ​2​0​1​8).

Nonparametric tests, specifically the Friedman and Kruskal–Wallis tests, were utilised. The Friedman test facilitates comparisons of mean rankings across groups, aiding the evaluation of differences among variables and thus demonstrating overall statistical significance between the mean ranks of associated elements.

Multiple linear regression was employed to examine how established components differ in their outcomes as a function of other demographic variables. Demographics were utilised as independent variables to understand their impact and relationship with the factors. The Bonferroni adjustment was applied to control for multiple comparisons between demographics and factors, addressing the research question.

Additionally, the Kruskal-Wallis H and Mann-Whitney U tests were utilised to assess whether the distribution of scores for each component varied by demographic factor. The Kruskal-Wallis H test, also known as the ‘one-way ANOVA on ranks,’ is a nonparametric rank-based test used to evaluate statistically significant differences between two or more groups of an independent variable on a continuous or ordinal dependent variable. It serves as a nonparametric alternative to one-way ANOVA and extends the Mann-Whitney U test to compare more than two independent groups.

We then conducted a case-study analysis utilising the theoretical framework proposed by B​r​a​u​n​ ​&​ ​C​l​a​r​k​e​ ​(​2​0​0​6​). We developed the qualitative thematic analysis by assembling multiple sources of information and conducting a range of quantitative analyses, as inspired by Stake. This type of qualitative thematic analysis approach is known as a collaborative qualitative thematic analysis (Y​i​n​,​ ​2​0​1​3). This technique was chosen because it aims to develop a better understanding of the concerns about the triggers of corruption. This aided the study through various research designs chosen by different writers on different propositions and literature, and this information was utilised to explain, describe, and investigate various phenomena related to corruption (Y​i​n​,​ ​2​0​1​3). To extract coding from the case studies and develop distinct themes, we utilised NVivo. The themes were then assessed to understand the sub-factors within each theme and their relationships with one another. A similar approach to the open analysis technique has been duplicated.

According to B​r​a​u​n​ ​&​ ​C​l​a​r​k​e​ ​(​2​0​0​6​), a theme indicates a predictable response or meaning within the data set. It captures something essential about the data in connection with the research subject. In this part, we went through many procedures in NVivo to generate themes for the case studies mentioned above. The initial stage involved becoming acquainted with the facts, such as jotting down thoughts. The second stage involved developing preliminary codes by compiling data relevant to each code. The next stage was to identify topics by grouping codes into potential themes. The fourth stage involved reviewing the themes to ensure they aligned with the coded extracts and the complete dataset, creating a thematic map of the investigation. The fifth stage involves identifying and naming themes and building an overarching narrative by developing precise definitions and titles for each subject. For each subject, the analysis will be presented below.

The significance of a theme is assessed not by quantitative metrics, but rather by whether it captures something noteworthy about the larger research subject. The thematic synthesis of prior empirical studies was conducted to understand the various techniques employed by the researchers in their investigations of the propositions presented above. Using NVivo, we can display the themes for this study, along with the sub-themes that provide greater detail.

The main themes identified were Psychology, Sociology, Environment, Internet, and Demographics. The analytical process behind these themes involved a transition from description to data analysis, revealing patterns in semantic content and facilitating interpretation. An attempt was made to theorise the importance and implications of these patterns in relation to previous literature.

Table 2 contains the proposition titles and thematic synthesis of prior empirical studies.

The following graphic represents an exploratory chart with many psychological variables (see Figure 2: Psychology Explore Diagram). The psychological element of this study is influenced by factors such as behaviour, emotional intelligence, and emotions, as identified in this subject. The actor indicated in the literature suffers from these qualities at the cost of corruption, and the causes of D​i​m​a​n​t​ ​&​ ​T​o​s​a​t​o​ ​(​2​0​1​8​)’s research of corruption.

The thematic analyses reveal that the identified factors are connected to the literature on societal issues, including impunity and the lack of impartial justice, and to their adverse impact on the individual. The basis of the factors, social norms, cultural influences, social dominance and power, has been established (see Figure 3). These characteristics are the root of how corruption triggers start.

The environment is essential because it provides a further basis for how other themes, including the Internet, have evolved. Since the Internet is an unseen instrument, it nevertheless demonstrates a unique environment in which the possibility of manipulating and harassing victims to repress corruption is increased. The theme of the environment includes economic factors and corrupt societies. This subject is associated with psychology and sociology because the environment holds people accountable for their psychologically composed behaviours and is part of a society operating within an environment (see Figure 4).

The last theme for the case studies is demography, which comprises Religion, Genetics, Gender, Education, and Region. This collection of factors is the theme of descriptions of the background and possible exposure to corruption of the individual. The literature shows that individuals are exposed to specific cultivation practices. According to studies, the history of past generations may affect individuals’ likelihood of engaging in particular criminal behaviours. Studies have revealed that gender influences people’s increased exposure to corruption, and researchers claim that the person is controlled through the confined schooling system and exposed to more illicit activity if he/she live in a high-risk nation.

In Table 3, we show the distribution of responses, with 454 respondents split by gender: the majority were female (59%) and male (40%). We were interested in understanding respondents’ gender to analyse how gender characteristics affect the survey results.

Table 4 shows that the majority of the respondents were aged between 26 to 35 years old (35%) and these were followed by 36 to 45 years old (22%), 46 to 55 years old (17%), 18 to 35 years old (15%), 56 to 65 years old (8%), with only 65 years old and over with only 3%. This shows the majority of the respondents are Millennials and Gen X (B​e​r​e​s​f​o​r​d​ ​R​e​s​e​a​r​c​h​,​ ​2​0​2​1).

The majority of the participants attained an education with 178 respondents (39%) were post graduates, 146 respondents (32%) obtained a degree, 65 respondents (14%) had their matriculation, whilst only 42 respondents (9%) had secondary education, 21 respondents (5%) had vocational education and only 2 participants (0%) had only primary education as shown in Table 5. We were interested in the educational level of those supporting the survey.

The distribution of marital status among the respondents was dominated by those who were never married, with 214 respondents (47%) and 206 respondents (45%) married, as shown in Table 6.

The respondents’ data were entered into SPSS (Version 27), and statistical analysis was performed, particularly an exploratory factor analysis, which is a means of evaluating the theoretical knowledge of the psychological factor analysis presented in the research question (P1) variables. The data and the authors’ interpretation determine the final number of factors.

Because the items were measured on an ordinal scale, we utilised the median (Md) as a measure of central tendency and the inter-quartile range as a measure of dispersion. Where a set of items could be clustered into a construct (or theme), we used Cronbach’s alpha to examine the internal consistency reliability of the measures. We computed the mean (M) as a measure of central tendency and the standard deviation as a measure of spread after combining the items into a single Likert scale.

We utilised Equamax via principal component extraction and Kaiser Normalisation for exploratory factor analysis. The Kaiser-Meyer-Olkin statistic, which is a measure of sample adequacy for the suitability of applying factor analysis, was within the acceptable range (above 0.7), with a value of 0.751. This bolstered the continuation of factor analysis, and hence the analysis proceeded.

Factor analysis yielded the best 5-factor solution for 21 statements, which together explained 52.974% of the variance. Therefore, variables that show low association with multiple factors simultaneously are neglected in the analysis. Table 1 shows which statements are grouped under each of the five factors. Factor 1, which has now been termed “Positive Emotions”, explained 16.926% of the variance and comprised 4 items. Factor 2, which has now been termed “Environment” explained 13.990% of the total variance and comprised 4 items, Factor 3, which has now been termed “Causes” explained 9.099% of the total variance and comprised 6 items, Factor 4, which has now been termed “Negative Emotions” explained 7.130% of the total variance and comprised 4 items and Factor 5, which has now been termed “Economic” explained 5.828% of the total variance and comprised 3 items (H​a​i​r​ ​e​t​ ​a​l​.​,​ ​1​9​9​8) (Table 7).

Factor 1 – Positive Emotions:

We compared these findings to the literature; B​a​n​d​u​r​a​ ​(​1​9​9​9​) describes how moral standards change. The human person is motivated by self-worth and a desire to act against immoral behaviour. Nonetheless, the results reveal that while the majority of respondents are neutral, the remainder are contradictory. This maintains the level of complication in performing self-regulation. When respondents were asked to reply to this question, they did so by naming the standards they use. Those who accept certain good feelings regarding illegal conduct relinquish personal control over their emotions. As a result, when respondents responded to this question, they went through their Moral Agency distribution. They imagined how they would feel if it were them, as stated by Anthony Bandura – the mechanism of moral disengagement process (B​a​n​d​u​r​a​,​ ​1​9​8​6).

Factor 2 – Environment:

When there is a possibility of corruption, the majority of respondents believe the environment influences it. As a result, as studies stated, the environment is a perceived trigger of corruption on the assumption that a corrupted environment is a disadvantage to the human being and that all other emotions linked to performing bribery, extortion, and so on are dependent on the expectation of the environment they live in (J​u​l​i​á​n​ ​&​ ​B​o​n​a​v​í​a​,​ ​2​0​2​0; M​o​r​g​a​n​,​ ​2​0​2​0; N​o​r​t​h​,​ ​1​9​9​1; W​u​ ​&​ ​Z​h​u​,​ ​2​0​1​6).

Factor 3 – Causes:

The exploratory factor analyses demonstrate a link between the tools that induce corruption. Favouritism, low institutional quality, restricted press freedom, privately hosted parties for authorised officials, and the contagion effect were among the tools proposed by D​i​m​a​n​t​ ​&​ ​T​o​s​a​t​o​ ​(​2​0​1​8​). These factors demonstrate that, as the researchers expected, they are corruption triggers.

Factor 4 – Negative Emotions:

Factors 1 and 4 are related but at different extremes of the spectrum. When the mean rank of positive and negative emotions is compared, they are highly similar. According to Bandura’s research, the same mechanism of moral disengagement applied to negative emotional attachment when the respondent answered the question (B​a​n​d​u​r​a​,​ ​1​9​8​6).

Factor 5 – Economic:

Corruption has a significant impact on economic management, transparency, and the size of the economy (A​k​h​m​e​t​o​v​ ​e​t​ ​a​l​.​,​ ​2​0​1​8). The exploratory factor analysis verified that these variables are sources of corruption triggers and identified them as an economic component.

Cronbach’s alpha coefficients for this scale ranged from 0.561 to 0.782. As a result, the author may infer that this scale is trustworthy when used in a statistical study (Table 8).

The author then calculated the corruption triggers from these 5 criteria and 30 statements and used multiple linear regression to investigate how this measure changes with demographic factors.

Cronbach’s alpha is valid and consistent at 0.70; therefore, factors 2 through 5 show that the internal consistency common range is low, whilst the table above shows that factor 1 – Positive Emotions, its variables are the most closely related as a group. Nonetheless. Factors from 1 to 5 show that there is moderate reliability (H​i​n​t​o​n​ ​e​t​ ​a​l​.​,​ ​2​0​0​4) (Figure 5).

While some factors exhibit Cronbach’s alpha values below the conventional 0.70 threshold, this is consistent with exploratory behavioural research involving complex psychological and institutional constructs. As noted in prior governance and behavioural risk studies, alpha values above 0.50 are acceptable in early-stage scale development, particularly when constructs capture heterogeneous emotional and environmental dimensions of risk.

The mean across all exhibited criteria shows how respondents related to and responded to the survey. The mean was calculated as seen in Table 9. The factor with the highest mean score is Environment as seen in Table 8 (n = 454), where the majority of respondents agreed with the statements about bribery becoming a social norm, peer pressure triggering corruption, corruption induced by low levels of respect for authorities, and immunizing the negative emotions associated with bad behaviour in a highly corrupt environment (such as anarchy environment). At the same time, the majority of respondents neither agreed nor disagreed with the assertions for components 1, 4, and 5 (Positive Emotions, Negative Emotions, and Causes). Whereas for Factor 3 – Economic, respondents disagreed with the majority of the statements regarding inefficient bureaucracy and non-transparency increasing favouritism, limited press freedom triggering corruption, frequent interactions of authorised officials in private settings increasing the possibility of corrupt practices, and the contagion effect on neighbouring countries. Nonetheless, as shown in the subsequent multiple linear regression analysis, the link between demographics and the Economic component is not significant (p > 0.05).

Except for a few variables, none of the predictors was significant in the linear regression. As a consequence, the Mann-Whitney and Kruskal-Wallis tests were used to determine if the distribution of scores for each component varied by demographic factor. The author then calculated the triggers of corruption from these 5 variables and 33 statements, and ran multiple linear regression to examine how this measure varied across demographic characteristics such as gender, age, education, status, and nationality.

When the author tested for multiple linear regression with gender, the findings revealed no significant association between the variables and gender (p > 0.05), as shown in Table 10.

When the author evaluated for multiple linear regression with age, the findings reveal that no significant association exists between the variables (Positive Emotions, Causes, and Economic) (p > 0.05). At the same time, there is a significant connection between the variables Environment (Beta = 0.145; T = 2.757; p < 0.05) and Negative Emotions, where (Beta = 0.095; T = 1.977; p < 0.05) as shown in Table 11.

When the author tested a multiple linear regression with education, the findings revealed no significant relationships among the variables (Positive Emotions, Environment, Negative Emotions, and Economic) (p-value > 0.05). While there is a statistically significant link between the Causes factor (Beta = 0.118; T = 2.228; p < 0.05) as shown in Table 12.

When the author tested a multiple linear regression with status, the findings revealed that no significant relationships exist among the variables (Positive Emotions, Causes, and Economic) (p > 0.05). At the same time, there is a significant connection between the variables Environment (Beta = 0.152; T = 2.889; p < 0.05) and Negative Emotions, where (Beta = 0.102; T = 2.122; p < 0.05) as shown in Table 13.

When the author tested for a multiple linear regression with nationality, the findings revealed that no significant relationships existed among the variables (Positive Emotions, Causes, and Economic) (p > 0.05). At the same time, there is a significant connection between the variables Environment (Beta = 0.101; T = 1.917; p < 0.05) and Negative Emotions, where (Beta = 0.106; T = 2.193; p < 0.05) as shown in Table 14.

In conclusion, the demographic variables of nationality, status, and age are comparable in relevance to the factors of Environment and Negative Emotions. On the other hand, the education factor is significant for the Causes factor.

To examine how respondents ranked the emotions associated with engaging in corrupt behaviour, a Friedman two-way analysis of variance by ranks was conducted. The null hypothesis that all emotions were ranked equally was rejected, indicating statistically significant differences among the ranked emotions (χ² = 409.015, df = 7, p < 0.001).

As shown in Table 15, shame ranked highest among emotions associated with corrupt behaviour, with a mean rank of 5.34. This was followed closely by joy, which recorded a mean rank of 5.27. Other emotions, including fear, stress, anger, fulfilment, happiness, and self-worth, exhibited lower mean ranks. The results indicate that respondents did not uniformly perceive emotional responses to corruption.

These findings demonstrate that respondents differentiate between emotional experiences associated with corrupt acts. Further interpretation of these rankings is presented in the Discussion section.

Qualitative data from the open-ended survey question were analysed using thematic analysis following B​r​a​u​n​ ​&​ ​C​l​a​r​k​e​ ​(​2​0​0​6​)’s framework. A word-frequency query was first conducted to identify commonly occurring terms, which informed the initial coding process. Subsequently, responses were coded and grouped into themes using NVivo.

The analysis identified five dominant themes reported by respondents: behaviour, culture, society, greed, and power. These themes accounted for the largest proportion of coded references across the dataset. Less frequently cited themes included gender, internet-related factors, education, environmental conditions, and structural or bureaucratic factors.

Cluster analysis revealed similarities between themes, indicating that behaviour was closely associated with societal norms, values, and culture, while greed was frequently linked to power and inequality. Environmental and internet-related themes appeared more isolated, with fewer overlapping references. Table 16 illustrates the thematic clustering based on similarity.

These qualitative results complement the quantitative findings by identifying recurring patterns in respondents’ descriptions of corruption triggers. Interpretation of these themes and their implications is addressed in the Discussion section.

The analyses carried out by the Authors demonstrate that the themes reflect her writings and her suggestion that they are the major causes of corruption, as seen from a thematic perspective. These themes, however, also cover good and adverse emotions, the environment, causes, and economic factors, as identified in the exploratory factor analysis. The technical SPSS results indicated that the Cronbach’s alpha is reliable and that the five themes identified in the exploratory factor analysis are important in demographic studies. The Friedman Test analysed emotions to understand how respondents ranked these feelings, which also featured in theme analyses of corrupt acts.

This study examined corruption through a governance and risk management lens, focusing on behavioural, emotional, environmental, and institutional triggers. Using a mixed-methods approach, five primary categories of corruption triggers were identified: positive emotions, environmental factors, structural causes, negative emotions, and economic factors. These findings reinforce the argument that corruption is not solely a legal or compliance failure but a complex behavioural risk phenomenon embedded within organisational and societal contexts.

Across both quantitative and qualitative analyses, environmental conditions emerged as the most influential perceived trigger of corruption. Respondents consistently associated corrupt behaviour with contextual factors such as social norms, peer pressure, tolerance of unethical conduct, and weak respect for authority. This aligns with governance literature emphasising the role of institutional environments in shaping ethical behaviour and control effectiveness.

A key contribution of this study lies in its examination of emotional responses associated with corrupt behaviour. The Friedman test results revealed that respondents perceived individuals engaging in corruption as experiencing both negative emotions, particularly shame, and positive emotions, notably joy. This dual emotional response highlights the internal conflict inherent in corrupt acts.

These findings provide empirical support for Bandura’s theory of moral disengagement, which posits that individuals can rationalise unethical behaviour while maintaining a positive self-concept. The coexistence of positive and negative emotions suggests that corrupt behaviour is often accompanied by emotional justification mechanisms that reduce psychological discomfort and weaken self-regulation. This behavioural insight helps explain why corruption can persist even in environments with formal rules and sanctions.

The exploratory factor analysis confirmed that structural and institutional factors constitute a distinct source of corruption risk. Factors such as inefficient bureaucracy, lack of accountability, limited press freedom, and preferential treatment of elites were empirically validated as significant triggers. These findings are consistent with D​i​m​a​n​t​ ​&​ ​T​o​s​a​t​o​ ​(​2​0​1​8​)’s framework, which emphasises the role of institutional weaknesses and contagion effects in facilitating corruption.

Economic factors were also identified as a separate dimension, although respondents demonstrated more neutral or sceptical attitudes towards purely economic explanations. This suggests that while financial incentives play a role, corruption is more strongly perceived as a function of environmental tolerance and governance failures than of economic necessity alone.

The qualitative analysis further reinforced these conclusions by highlighting the importance of power, greed, and cultural acceptance in normalising corrupt practices. Together, these findings indicate that corruption risk emerges from the interaction between individual behaviour and institutional context rather than from isolated economic pressures.

The analysis of demographic variables revealed that age, status, and nationality were significantly associated with environmental and emotional factors, while education was significantly associated with structural causes. In contrast, gender did not demonstrate a statistically significant influence across the identified factors.

These results suggest that demographic characteristics shape how individuals perceive corruption risk rather than directly determining corrupt behaviour. The influence of nationality and status on environmental and emotional dimensions underscores the importance of socio-institutional context in shaping ethical perceptions. Education’s association with structural causes indicates that higher levels of education may increase awareness of governance failures and institutional drivers of corruption.

The findings of this study have important implications for corporate governance, risk management, and regulatory oversight. First, corruption should be recognised as a behavioural risk that extends beyond compliance and legal enforcement. Boards and risk committees should incorporate behavioural indicators, cultural diagnostics, and environmental risk assessments into enterprise risk management frameworks.

Second, internal control systems and ethics programmes should address emotional rationalisation and social norms that enable corrupt behaviour. Traditional rule-based controls may be insufficient if organisational cultures tolerate unethical practices or fail to reinforce accountability.

Third, for insurers and regulators, understanding behavioural corruption triggers is essential for assessing governance quality, moral hazard, and systemic risk exposure. Integrating behavioural risk insights into governance assessments can enhance prevention strategies and strengthen institutional resilience.

By reframing corruption as a behavioural and institutional risk phenomenon, this study contributes to the governance and risk management literature. It advances existing research by empirically linking emotional responses, environmental conditions, and structural weaknesses to corruption risk, offering a more nuanced understanding of why corrupt behaviour persists.

Future research could extend this framework through cross-country comparative studies, longitudinal designs, or sector-specific analyses. Further investigation into digital environments, information asymmetries, and behavioural nudges may also yield valuable insights into corruption-prevention strategies.

The cost of corruption goes far beyond the sums spent on bribes, cash being stolen, and taxes avoided. Corruption stifles progress. It exacerbates inequality, diminishes government legitimacy and erodes public faith in democracy. It influences everything from school learning results to climate change mitigation efforts, and it is a key barrier to meeting the Sustainable Development Goals. Despite this, billions of people live in severely corrupt cultures, with more than two-thirds of nations receiving scores of 50 or fewer on Transparency International’s 2020 Corruption Perceptions Index (CPI) (T​r​a​n​s​p​a​r​e​n​c​y​ ​I​n​t​e​r​n​a​t​i​o​n​a​l​,​ ​2​0​2​1​b).

Anti-corruption activities are sometimes characterised as a battle to combat, root out and even destroy corruption. While such rhetoric may be effective in calling attention to the issue, it is critical to acknowledge corruption’s historical endurance. Corruption is not a newly discovered ‘illness’ that can be treated with the right mix of medications; understanding its causes may help plan how to reduce it and raise awareness. It is extremely tough to combat corruption, and it takes time. The institutions, laws, customs, and norms of the world’s least corrupt countries evolved over decades, if not centuries. The process is neither linear nor complete. Because there is no one road to reform, it is necessary to examine ‘best fit’ rather than ‘best practice’ methods, because the viability of change is decided in part by whether an enabling environment exists.

Fighting corruption, on the other hand, is both obvious and necessary; to do so, we must first understand it. The identification process should be comprehensive, context-specific, and practical. The variables discussed in this study give the audience the tools they need to combat corruption. These major factors of corrupt behaviour demonstrate how an individual is most likely to engage in illegal activities and how society accepts them. It is also known that governmental bodies, organisations, and authorities seek to develop anti-corruption strategies based on such accepted behaviour. The Authors of this study point out that entities can use quantitative and qualitative analysis results in Education when establishing the learning structure for students, Governing Authorities when formulating policies and regulations within societies, Institutions and Organisations when designing and establishing duties, training, and education at the place of work.

These structural and behavioural dimensions of corruption are further illustrated by Malta’s recent experience. As shown in Table 17, Malta’s CPI scores over the five-year period reflect a gradual decline, coinciding with several high-profile national corruption scandals. These events highlight how institutional weaknesses, governance failures, and socially tolerated behaviours can collectively undermine anti-corruption efforts, reinforcing the argument that corruption risk must be addressed through comprehensive governance, behavioural oversight, and systemic reform rather than legal compliance alone.

Overall, the findings demonstrate that corruption risk cannot be mitigated through legal compliance alone but requires integration into corporate governance frameworks, enterprise risk management systems, and behavioural risk oversight.

While the Authors examined psychological, environmental, and behavioural triggers of corruption within the socio-economic and technological context prevailing at that time (G​a​l​e​a​ ​T​a​y​l​o​r​,​ ​2​0​2​1), the rapid acceleration of AI, digital governance systems, and algorithmic decision-making since 2021 necessitates renewed scholarly attention. Emerging technologies have fundamentally altered both the opportunities for corruption and the mechanisms through which corrupt behaviour is rationalised, concealed, or normalised.

Advances in AI-driven automation, data analytics, and platform governance have introduced new corruption risks, including algorithmic opacity, automated bias, manipulation of digital procurement systems, and the concentration of decision-making power within technologically sophisticated actors. While digitalisation and e-government initiatives were previously identified as potential mitigators of corruption through increased transparency (D​i​m​a​n​t​ ​&​ ​T​o​s​a​t​o​,​ ​2​0​1​8; S​h​a​b​b​i​r​ ​e​t​ ​a​l​.​,​ ​2​0​2​1), the contemporary deployment of AI systems introduces asymmetries in information, accountability, and oversight that may inadvertently facilitate new forms of institutional and behavioural corruption.

From a behavioural perspective, the psychological triggers identified in the author’s earlier research, such as emotional rationalisation, environmental normalisation, and perceived opportunity, remain highly relevant in the AI-enabled governance landscape. However, these triggers are now mediated through technological systems that can obscure responsibility, diffuse accountability, and reinforce neutralisation strategies, thereby weakening ethical restraint and increasing tolerance for unethical conduct. The perceived legitimacy of algorithmic outcomes may further enable individuals to externalise blame and justify corrupt decisions as system-driven rather than personally enacted.

Consequently, while technological innovation holds significant promise for preventing corruption, the post-2021 digital environment underscores the need to integrate behavioural, institutional, and technological perspectives. Future anti-corruption frameworks must therefore address not only traditional governance failures but also the ethical, psychological, and regulatory implications of AI-driven decision-making systems. This evolving context reinforces the need to revisit earlier models of corruption triggers and adapt them to the realities of a rapidly transforming digital and technological landscape.

Ultimately, corruption risk cannot be effectively mitigated solely through legal compliance. This study demonstrates that behavioural, emotional, and institutional factors must be addressed through integrated governance, risk management, and cultural oversight mechanisms. By recognising corruption as a behavioural risk phenomenon, organisations can move towards more resilient, ethically grounded, and sustainable governance systems.