Employee perceptions of corporate social responsibility and organizational corruption: empirical evidence from Kosovo

Purpose – The purpose of this paper is to investigate the relationship between perceived corporate social responsibility (PCSR) and perceived organizational corruption (POC) in a developing country context. The research suggests mediating roles of perceived ethical leadership (PEL) and job satisfaction (JS) in the PCSR–POC relationship. Design/methodology/approach – Data were collected through survey questionnaires. The sample consisted of 434 employees working in private and public organizations in Kosovo. Regression analysis was conducted by using a serial mediationmodel. Findings – Applying a social learning framework, this study provides evidence that employees’ perceptions of corporate social responsibility (CSR) are positively related to perceived ethical leadership and job satisfaction and negatively related to perceived organizational corruption. Furthermore, results confirm that the relationship between CSR perception and organizational corruption perception is mediated by ethical leadership perception and job satisfaction. Practical implications – Armed with the findings, organizations can adopt CSR practices to positively influence employee behaviors and attitudes. From these results, it is possible to better comprehend the role of CSR in dealing with relevant aspects such as corruption at the organizational level, especially in developing and emergingmarkets. Social implications – The findings of this research indicate that employees in socially responsible organizations perceive less organizational corruption. Adopting a more ethical and responsible management approach might represent a promising solution to fight the corruption phenomenon inside and even outside organizations. These results should serve as reflection for bothmanagers andpublic authorities. Originality/value – With regard to CSR, previous studies have investigated different employee outcomes but never considered the potential impact on corruption at the organizational level. Furthermore, this study extends the literature by conceptualizing perceived ethical leadership and job satisfaction as mediators between perceived CSR and organizational corruption perception, in a developing country context where the concept of CSR is still less investigated.


Introduction
Corruption is receiving significant prominence in public and scholarly debates. It represents the misuse of public power or organizational resources for private or personal gain and is commonly understood to be strongly undesirable by any community (Lange, 2008). The wide-ranging negative effects of corruption are legion, such as unequal resource allocations, constrained economic growth, decreased trust in government and reduced Festim Tafolli is a PhD student at the Department of Marketing and International Management at Alpen-Adria-Universitä t Klagenfurt, Klagenfurt, Austria. Sonja Grabner-Krä uter is an associate professor of marketing at Alpen-Adria-Universitä t Klagenfurt, Klagenfurt, Austria. legitimacy of market economy and democracy, distorting markets and opportunities and rising poverty and discrimination (Osuji, 2011;Lourenço et al., 2018). Several studies confirm that corruption is accompanied by economic stagnation and social decline (Lewicka-Strzalecka, 2006). Certainly, it is not an easy task to find an answer on how to effectively comprehend and then prevent corruption. The causes and effects of corruption at the national and organizational levels are issues that interest scholars, managers and policymakers. Over the past decades, a variety of organizational corruption reports have been published, suggesting a wide range of organizational corruption prevention and control schemes (Ashforth et al., 2008). Some are compliance-based, focusing on extrinsic rules and benefits, such as institutional guidelines, code of conduct and regulatory and legal penalties. Others are based on integrity and focus on intrinsic motivation, such as organizational values, ethical leadership, job satisfaction and self-controls (Lange, 2008). This paper investigates deeper into the intrinsic motivations analyzing if and how organizational corruption is related to corporate social responsibility (CSR), which represents a factor that affects managers, employees and the ethical climate within an organization (Peterson, 2002). Organizational research from a micro perspective shows that work behaviors and attitudes play key roles in turning CSR into favorable organizational outcomes (Valentine and Fleischman, 2008;Kim et al., 2010). For example, researchers suggest that CSR enhanced the commitment of employees to the organization with further increase to its performance (Ali et al., 2010). It is therefore crucial to investigate why and how CSR may contribute to employee outcomes in the place of work. A better understanding of these micro mechanisms through which CSR benefits the organization also has practical implications for organization managers, as they can learn how to produce beneficial employee results while the organization participates in CSR actions.
In this context, organizational researchers have started to investigate CSR's effect on attitudinal and behavioral outcomes of employees within the organization (Aguilera et al., 2007). In particular, CSR has been found to enhance employee organizational commitment and organizational trust (Farooq et al., 2014), ethical leadership (Gao and He, 2018), job satisfaction (Dhanesh, 2014), employee engagement (Glavas and Piderit, 2009), work meaningfulness (Glavas and Kelley, 2014), organizational citizenship behavior (Zhang et al., 2014) and to decrease employees' emigration intent (Grabner-Krä uter et al., 2020). It should be noted, however, that the underlying psychological mechanisms linking PCSR and employee outcomes are not sufficiently investigated and understood. Some of the recent studies have made improvements in the area by using social identity theory and social exchange theory (De Roeck et al., 2014;De Roeck and Maon, 2018). Notwithstanding these empirical findings, the theoretical mechanisms that underlie the effects of perceived (PCSR) on perceived organizational corruption (POC) remain elusive. In other words, knowledge regarding if and how CSR can encourage employees toward ethical and positive behavior and favor an environment less prone to corruption is in its infancy.
In the current study we theorize a serial mediation model and test the mediating roles of ethical leadership and job satisfaction in the PCSR-POC relationship. We expect that an organization that adopts and follows a social and responsible strategy is related to higher levels of ethical leadership perception and thus to higher job satisfaction, reducing employee's deviant behavior and corruption in the workplace. Drawing on findings from recent micro-CSR research (for an overview, see Aryati et al., 2018;Gorsira et al., 2018;Manzoor et al., 2019), we suggest that CSR activities may be positively associated to perceived ethical leadership, job satisfaction and negatively related to employees' organizational corruption perception ( Figure 1).
This research contributes to the literature in several ways. First, we investigate the perception of CSR by employees in a developing country context and its potential impact on decreasing deviant and corrupt behavior in the workplace with a unique data set that includes survey data from 434 Kosovar employees. Second, we corroborate previous findings on positive relationships between employees' CSR perceptions and ethical leadership perception and job satisfaction in emerging markets (Gharleghi et al., 2018). Third, to the best of our knowledge, this is the first study to examine the association between CSR perception of employees and their perception of organizational corruption. Fourth, beyond examining direct effects, our study will contribute to the existing literature by exploring the underlying mechanism through which employees' CSR perceptions are linked to ethical leadership and job satisfaction and how this impacts the perception of organizational corruption. Guided by a social learning perspective (Bandura, 1971), according to which an individual's attitudes and behaviors are modified as the person observes and attends to stimuli in the social environment, we propose that perceptions of CSR activities are related to perceived ethical leadership and job satisfaction and eventually lower employees' perception of organizational corruption. From a practical point of view, providing evidence of positive associations between employee perceptions of CSR and valued outcomes can foster the emergence of a clearer and stronger business case for CSR in public and private firms in Kosovo, what would support the legitimization of managers' CSR investments on economic grounds (see also Grabner-Krä uter et al., 2020).
The organization of the paper is as follows. In Section 2, we first discuss the conceptualization of corruption before we move to several peculiarities of corruption at the organizational level in developing countries. Furthermore, we provide information on the phenomenon of corruption in Kosovo. In Section 3, we shortly discuss different theoretical perspectives and develop our hypotheses on the relationships between employees' perception of CSR and selected outcomes. Section 4 describes the method of the empirical study, and Section 5 reports on relevant findings of the research study. In Section 6, based on the findings, academic and managerial implications are addressed, as are the limitations of this study. The paper ends with a short conclusion given in Section 7.

Corruption in developing countries
Among the various challenges the world is experiencing, corruption continues to play a crucial role, influencing the development of the public and private sectors, in both developed and developing countries (Schonfelder et al., 2016;Suleimenova et al., 2018). The broad scientific literature, particularly in the political context, defines corruption as the abuse of public power for personal benefit (Aguilera and Vadera, 2008). Based on these definitions, corruption refers to different unethical activities, such as extortion, influence peddling, forgery, fraud, embezzlement and bribery. Corruption represents a phenomenon that goes beyond public officials affecting individuals, businesses and supply chains in general (De Sousa Monteiro, 2019). Anyway, as suggested by Collins et al. (2009), at the Notes: PCSR -perceived corporate social responsibility; PEL -perceived ethical leadership; POC -perceived organizational corruption; JS -job satisfaction end to make progress in comprehending and managing corruption, it is necessary to focus on a narrow part of it in a specific context. Mashali (2012) points to the definition of the Swedish International Development Cooperation Agency, which identifies three levels of corruption: grand corruption (political corruption), petty corruption (bureaucratic corruption) and state capture (comprehends the whole state apparatus). Petty corruption refers to relatively smaller bribes where people are engaged in dealing with bureaucratic transactions such as tax returns, licensing requirements, application for government benefits, registration of a new organization, passport, employees' safety standards, environmental hazards, etc. (Ariane et al., 2007). Grand corruption is more difficult to identify and to measure and usually is unknown to the general public. In each department, ministry or organization, corruption at the high level is classified as grand corruption because it can affect all or a great part of employees, in both public and private sector organizations. Vice versa, corruption that affects some of the employees, at lower level, is placed in the category of petty corruption. Nevertheless, corruption by managers and employees is mostly ignored and we should be worried about.
The presence of corruption, mostly in emerging markets, is justified because it frequently represents the only way to survive in specific contexts, where the ability to perform a corrupt act is seen as pride and a skill even if condemned by the surrounding society. The need to be corrupted sometimes is due to the weakness of the state to provide adequate services. For instance, in Russia officers accept bribe payments by migrants because their salary is lower than that of migrants from factory work. In addition, patients give unofficial payments toward doctors because they consider it as respect for his/her efforts and expertise and because doctors are not properly compensated (Reeves and Clean, 2013). Furthermore, the situation in Ukraine is even worse because public officials cannot survive with their regular wages (Polese, 2008). In Romania, property owners are proud of their connections with judges and their ability to pay bribes and reaffirm their power as owners (Zerilli, 2005).
Despite the different perceptions of corruption by countries, several research studies suggest that the phenomenon of corruption is condemned and frustrates the majority of citizens in developing countries where deviance behaviors are present and largely developed (World Economic Forum, 2016). In developing countries, the difficulty is that the punish/reward approach is necessary but not enough because corruption is not a problem of individual deviance but of the whole system where everyone is corrupt, it is a problem of context (Osburg, 2018). The case of Kosovo, a new and developing country represents an appropriate example to investigate the problem of corruption trying to understand the possible causes and effects at the organizational level.

Study context -Kosovo
Kosovo is one of the poorest countries in Europe, with a per capita gross domestic product (GDP) of about US$ 4,442 in 2019. This makes the per capita income on average about one-tenth of European Union (EU) levels (Sørheim, 2014). Kosovo's rating as one of the poorest countries in Europe is strongly related to its high unemployment levels and extensive corruption. As stated by Sørheim (2014), corruption is a crucial factor slowing down economic growth and deterring international firms from investing in Kosovo. Consequently, reducing corruption is of vital importance and can be seen as one of the main criteria for Kosovo's further European integration. In Kosovo (and South-Eastern Europe, in general), the concept of CSR is relatively new (Grabner-Krä uter et al., 2020). To increase CSR awareness in Kosovo, and also to incorporate CSR best practices into their own companies, several Kosovar and international business leaders established the Kosovo CSR Network in April 2011 (Kosovo CSR Network, 2017). Even though at present a relatively small organization, the Network aspires to grow and become a stimulating network to attract new businesses in promoting CSR across Kosovo. Its mission is "to embrace responsibility for the firm's activities and to promote a positive effect through our activities on the environment, customers, employees, societies, stakeholders and the general public" (Kosovo CSR Network, 2017).
Although corruption represents a global phenomenon, it has different forms and dimensions, in line with the nature of institutions and of the socio-economic conditions. Corruption is particularly salient in emerging economies (Hardi and Buti, 2012) and Kosovo is no exception. Kosovo is in a region where the corruption phenomenon hinders the development of democratic processes and the creation of a society with transparent and accountable institutions. Kosovo, according to the Corruption Perception Index of Transparency International Kosovo, is one of the more corrupt countries worldwide (with a score of 36 on a scale from 0 to 100, where 0 is highly corrupt and 100 is very clean) (Transparency International, 2019). With the term corruption in our case, we refer to the petty corruption that influences the daily lives of citizens in their interactions with public administration and in their work environment.
According to The World Bank's report (2019) "Doing Business" that compares business regulations for domestic firms in 190 economies, Kosovo shows a few improvements (Rank: 44 and Ease of Doing Business Score: 74.15 compared to the Regional Average -Europe and Central Asia: 72.34). However, a lack of accountability and transparency in Kosovo's public administration results in wide-ranging corruption that negatively influences the investment climate. The customs, judiciary, procurement and public utilities are the most influenced by corruption. Notwithstanding abuse of office, extortion, active and passive bribery and money laundering are prohibited by Kosovo's Criminal Code, the practices of offering bribery and gifts are common and largely diffused (Business Anti-Corruption Portal, 2017). In Table 1 some key indicators at the country and regional level for corruption are summarized. The most challenging problem is government contracts: 51% of the firms were expected to give gifts to secure government contract. The value of gift expected to secure a government contract was 4.3% of the contract value which is higher than the same in Europe and Central Asia (1.6%) or the average of all countries (1.9%). In addition, the percentage of firms expected to give gifts to secure government contract was 38.9% in Kosovo, compared to 22.4% in Europe and Central Asia and 30.0% in all countries worldwide (The World Bank, 2013).

Corruption at the organizational and the individual level
The main part of the literature analyzes corruption at the national level, neglecting the situation within organizations (Dong et al., 2012). According to Collins et al. (2009), the involvement of organizations in corruption is essentially driven by decisions of executives and, consequently, by decisions of individuals. However, while a number of studies on the national and organizational levels offer explanations for differences in corruption among countries and between organizations, they do not offer insight into why certain individualsgiven specific country and organizational conditionsengage in corruption, while others do not (Gorsira et al., 2016). Recent studies about the reasons why individuals engage in corruption demonstrate that both personal characteristics and social influences are relevant (Dong et al., 2012). There are different disciplines, mainly criminology, psychology and economics that spotlight the reasons why individuals engage in corruption, but none seem to give definitive and concrete results (Prabowo, 2014).
Concerning the problem of corruption, both at organizational and individual levels, different studies investigate the idea that corruption exists because of "bad barrels" such as corrupt organizations that encourage unethical behaviors, or because of "bad apples", which refers to the personal characteristics of individuals, i.e. some people are just born and brought up "badly" and are prone to do bad things. Usually, individuals that are more ambitious from a career point of view are more oriented to deviant behaviors (Reiss and Mitra, 1998). Furthermore, findings suggest that more unethical acts are done by males compared to females, by younger people compared to older, by less religiously committed compared to those who are more, by individuals who are more attached to the organization because they are more prone to commit bad acts with the purpose of improving their organization (Jha and Panda, 2017). While the argument related to "bad apples" gives attention to the individual attributes, the "bad barrel" argument proposes that corruption is linked to factors inside the organization, comprehending the organizational culture, leadership and organizational ethical climate (Pertiwi, 2018). The organizational climate is influenced by leadership. Several studies found that when organizational leadership is perceived positively, less counterproductive employees' behavior takes place (Mayer et al., 2009). Vice versa, when leaders become morally disengaged, as a result, employees' ethical performance is negatively affected (Bonner et al., 2014). Integrating "bad apples and bad barrels", several scholars point out that unethical behavior or corruption depend on the individual's reflection which, in turn, is a result of his/her responses to situational factors.

Employees' corporate social responsibility perception
According to one of the most widely used definitions, CSR refers to an organization's "actions and policies that take into account stakeholders' expectations and the triple bottom line of economic, social, and environmental performance" (Aguinis and Glavas, 2012). An employee's perception of CSR reflects how the employee views the total amount of CSR activities their organizations undertake (El Akremi et al., 2018;Grabner-Kräuter et al., 2020). It captures how employees perceive their firm's CSR activities, rather than objective CSR engagement from a firm's perspective as it is analyzed e.g. by investment research firms (Ng et al., 2019). "How employees perceive the CSR of their employer has more direct and stronger implications for employees' subsequent reactions than actual firm behaviors of which employees may or may not be aware" (Rupp et al., 2013, p. 897). Although both objective and perceived CSR are relevant and perceived CSR is built upon objective CSR, perceived CSR reliably outperforms objective CSR as a predictor of numerous work outcomes such as job satisfaction, organizational engagement, organizational citizenship behavior, and therefore can be considered as the most appropriate independent variable to examine employees' responses to CSR practices (De Roeck and Maon, 2018;Gond et al., 2017).
Concerning employee responses to CSR, recent reviews of micro-CSR studies (Wang et al., 2020;Gond et al., 2017;De Roeck and Maon, 2018;El Akremi et al., 2018) suggest that CSR perceptions positively relate to a number of workplace outcomes. Prior micro-CSR research suggests significant effects of perceived CSR on various job-related outcomes, for example, job satisfaction (Valentine and Fleischman, 2008), organizational identification and commitment (Wang et al., 2016) work engagement, creativity, compassion and organizational citizenship behavior (Rupp and Mallory, 2015;Hur et al., 2017). However, most studies have focused on relationships between CSR perceptions and positive employee outcomes, neglecting their impact on negative employee outcomes (deviance, violence, sabotage, bribery) (Gond et al., 2017). In this regard, the intent of this work is to investigate possible effects of perceived CSR on negative employee behaviors, such as corruption, within organizations.

Perceived corporate social responsibility and organizational corruption perception
Perceived organizational corruption, as we define it, represents the corruption perceived by employees, which comprises both the corruption engaged in by employees for their own personal gain (also called occupational corruption) and engaged in by the organization in relation to other stakeholders (Crisan et al., 2018). Several researchers argue that a good predictor of deviant and corrupt employees' behaviors is the organizational environment (Taneja et al., 2011). Social learning theory (Bandura, 1971) holds that the ethical conduct of an employee is affected by the employee's internal aspects and the organization's ethical atmosphere. Appelbaum et al. (2007) affirmed that the features of the workplace environment are a good predictor of workplace violence, as intense form of deviance, rather than individual characteristics of personality. Employees are less likely to engage in deviant and corrupt behaviors in a workplace environment with a good ethical climate. Peterson (2002) suggests that the factors that influence an organization's ethical environment include rules, laws, professional codes, personal morality, friendships and social responsibility.
Examining the status of CSR in Nepal, Adhikari (2012) found that employees who were aware of their company's CSR initiatives had more positive and ethical company-related associations. A key aspect of corporate social responsibility is business ethics where CSR fulfillment tends to promote an ethical and responsible environment (Llewellyn, 1998). The theory of social exchange suggests that employees' commitment to the organization depends on their perception of the organization's values and benefits (Wayne et al., 1997). In this respect, it is reasonable to assume that the extent to which an organization engages in CSR will be related to corporate corruption prevention. Organizations that adhere to their CSR policies are at the lowest risk of corruption in the organization. Drawing on these arguments and findings, we propose our first hypothesis as follows: H1. Higher perceived CSR is associated with lower perceived organizational corruption.

Perceived corporate social responsibility and ethical leadership perception
CSR represents an indicator of moral and ethical issues concerning organizational decisionmaking and behavior (Branco and Rodrigues, 2006). Consequently, CSR influences employees' attitudes toward the supervisors, who are moral agents who are obliged to execute policies within all CSR dimensions (Wood, 1991). Previous research shows that, especially in a developing country context, social responsibility must start from the top (Baba-Abugre, 2014). To make CSR strategies operative, organizations should have leaders whose objectives and passions are focused toward responsible behaviors within organizations (Baba-Abugre, 2014). In general, CSR is socially desirable, reflecting the prosocial values of decision-makers (top managers) (Gao and He, 2018). Thus, CSR activities should increase employee perceptions of ethical leadership (Gao and He, 2018). According to Rupp et al. (2018), CSR activities are among the essential contextual factors that allow employees to make inferences about supervisors' intentions and characteristics. Employees also expect that supervisors, who plan CSR activities, are ethical or act according to ethical principles when evaluating the reasons for the organization to engage in CSR activities. The notion that immoral supervisors would lead to an organization fulfilling its social responsibilities is perceived by workforce as self-contradictory (Hansen et al., 2016). In this context, employees assume that CSR practices are derived from ethical values or ethical behavior management. CSR practices may be some of the antecedents at this point that allow the supervisor to be perceived as an ethical leader. In line with this reasoning, we propose the following hypothesis:

Perceived ethical leadership and organizational corruption perception
Ethical leadership represents an effective means in preventing deviant behavior and, at the same time, the supervisor can create a positive organizational atmosphere. In line with social learning theory (Bandura, 1971), the organizational leader acts as a model explaining and learning the effect of ethical leadership to subordinates as well as observing others' conduct with the authority to reprove or to reward them (Aryati et al., 2018). Different studies have examined the relationship between ethical leadership and different forms of unethical behavior, including employee misconduct, deviant behavior and bullying (Mayer et al., 2012). Those results have even showed that improved ethical leadership can minimize the deviant behavior of employees in the workplace. When supervisors perform ethically and responsibly, employees will also be more engaged in their work and less in deviant behavior, such as bribery. In addition, employees need to feel that they are morally supported in their actions by management and the entire organization (Appelbaum et al., 2007). Several studies provide evidence on the negative relationship between ethical leadership and corruption. De Graaf and Huberts (2008) revealed that weak leadership represents one of the most relevant causes of corruption. Based on these considerations, the following hypothesis is suggested: H3. Higher perceived ethical leadership is associated with lower perceived organizational corruption.

Perceived ethical leadership and job satisfaction
Job satisfaction represents an overall evaluative judgment regarding one's job (Weiss, 2002). It is related to several factors such as the sector, kind of work, wage/salary, professional opportunities, supervision, benefits, work values and ethics, organizational practices and relationships with coworkers (Misener et al., 1996). Studies provide evidence that job dissatisfaction leads to numerous negative effects such us problems of grievances, absenteeism, low morale and high turnover, unethical actions and misconduct in the workplace (Gangadhraiah et al., 1990). On the other hand, leadership behavior represents one of the most relevant factors that affect employee job satisfaction (Gangadhraiah et al., 1990). Consistently, a study by Vitell and Singhapakdi (2008) demonstrates that ethics institutionalization positively influences organizational commitment and job satisfaction. Therefore, we propose the following hypothesis: H4. Higher perceived ethical leadership is associated with higher job satisfaction.

Perceived corporate social responsibility and job satisfaction
Employees represent one of the most significant stakeholder groups within an organization (Rodrigo and Arenas, 2008). Employees' perceptions of CSR may increase their positive attitudes and feelings towards their work, i.e. their job satisfaction. Most of the researches have been carried out in the past about the relationship among employee CSR perception and their work satisfaction (Valentine and Barnett, 2003) and others found that a higher perceived CSR determine a higher job satisfaction (Valentine and Fleischman, 2008). The available literature on CSR and JS establishes a positive relationship, but this relationship in a developing country still needs to be explored. Therefore, it can be hypothesized as follows: H5. Higher perceived CSR is associated with higher job satisfaction.

Job satisfaction and organizational corruption perception
The effects of job satisfaction are multifaceted; low job satisfaction is one of the strongest causes of deviant workplace behaviors, such as abusing privileges, sharing confidential information and bribery (Omar et al., 2011;Garduno, 2019;Marcus and Wagner, 2007). Meta-analysis research by Hershcovis et al. (2007) stated that individuals who are dissatisfied with their jobs are likely to put less effort into their work or to act in destructive ways toward their organization. In a recent cross-sectional survey conducted by Nasir and Bashir (2012) in Pakistan among 100 employees in different government sectors, the workplace deviance ratio is very high due to the low rate of employee satisfaction. It was reported that job satisfaction contributes inversely to 80.2% of the workplace deviance variance. They also observed that most deviant behaviors in the workplace are explained by low levels of employees' job satisfaction (Nasir and Bashir, 2012). Drawing on these findings, we propose that the following: H6. Higher job satisfaction is associated with lower perceived organizational corruption.
Taken together, we finally propose a serial mediation hypothesis in which employees' perception of CSR activities are expected to foster perceived ethical leadership. Higher perceived ethical leadership then is posited to enhance JS, which in turn is expected to diminish organizational corruption perception.
H7. Perceived ethical leadership and job satisfaction mediate the relationship between perceived CSR and perceived organizational corruption.

Materials and methods
To gain a better understanding of the correlates of corruption perception, we developed a large-scale study involving organization employees and public officials who habitually interact professionally with colleagues of the other sectors. Our study was conducted in Kosovo. We used a self-administrated questionnaire to collect the data. For such purpose, we adopted and adapted established scales from literature to cover all relevant measurement constructs and used a seven-point Likert scale ranging from 1 (strongly disagree) to 7 (strongly agree) expressing respondents' degree of agreement to each question. (Appendixconstructs and questions for item wording).

Sample
The sample consists of 434 employees working in different public and private organizations in the main cities of Kosovo. The questions were presented in a randomized order, to counter order effects. A data quality check was performed on reliability of answers and on straight lining and, based on this, 28 respondents were disqualified because of meagre response quality. The Albanian version of the questionnaire was created strictly following the backtranslation method (Brislin, 1980). Before distributing the questionnaire, a pre-test with 67 employees was conducted. These respondents were not included in the final sample.
Organizational corruption, concretely bribery, is more emphasized in public organization where managers and employees try to gain personal benefits through public positions and institutions. In this way, the target was to collect data from public institutions in different sectors.

Measures
All measures were adapted to the work context (see Appendix -Constructs and questions). To measure POC, we used 8 items from an instrument developed by Crisan et al. (2018), (e.g. "Employees take bribes to perform their duties"). Employee Perceived (PCSR) was measured using Iqbal's et al. (2018) scale that comprises 16-items (a = 0.879) (e.g. "My organization provides quality working environment for employees"). To measure perceived ethical leadership (PEL), a 10-item scale was adopted from Pelletier and Bligh (2006), (e.g. "The top leadership of this organization is concerned with ethical practice "). Finally, to measure job satisfaction (JS), we relied on the 3-item scale developed by Valentine and Fleischman (2008), (e.g. "All in all, I am satisfied with my job"). In our analyses, we controlled for several demographic variables. Gender (1 male and 2 females) and sector (1 public and 2 private) were measured on a dichotomous scale and the variables age, education and salary/wage on an ordinal scale.

Description of sample characteristics
In total, we were able to obtain responses from 434 employees. Among those respondents, 250 (57.6%) worked in public sector industries, such as education, health and public administration; the 184 private-sector employees who responded (42.4%) worked in production, services, health and education. The number of male respondents (59.0%) was higher than the female respondents (41.0%). The majority of the respondents are between 30 and 39 years of age (33.6%), 27.0% are between 40 and 49 years of age, and 25.1% are younger than 30 years of age. Almost 70% of the respondents indicate that their monthly family income is equal to or lower than e600 (Table 2).

Descriptive statistics and correlations
We calculated construct mean values for all multi-item scales and used them in the further analyses. All reverse coded items were recoded before aggregating scale items to their mean. Table 3  Past studies suggest a link between various demographic factors, perceptions of ethicality and employees' work-related outcomes (Forte, 2004;Melgar et al., 2010;Wimbush et al., 1997). Therefore, we included the control variables sector, age, gender, salary/wage and education in our regression analyses.

Reliability analysis
The Appendix at the end of the paper shows that the internal consistency reliability values for all measures lie above the recommended threshold of 0.70 for Cronbach's alpha. The highest reliability was obtained for perceived CSR (0.94) and the lowest reliability was obtained for JS (0.73). The lower reliability value for JS might reflect the fact that this scale comprised only three items (Appendixconstructs and questions).

Regression analysis
The results of the path analysis-based mediation analysis to test the serial mediation hypothesis show a good statistical performance of the estimated models. All models are significant and provide an adequate fit based on R 2 values. We report the model summaries of the three OLS regression models in Table 4.

Mediation analysis
Direct effects. We can report on positive and significant coefficients of perceived CSR on PEL (0.899, p < 0.001), as well as on JS (0.397, p < 0.001). As both direct effects are significant and show the proposed direction, we can accept both H2 and H5. The Table 3 Correlation analyses -Pearson correlation matrix  coefficient of PEL in the regression model with JS as the outcome is also positive and significant (0.389, p < 0.001) and thus, H4 is also supported. In addition, PEL shows a significantly negative effect (À0.297, p < 0.01) on POC, which supports H3. Finally, JS shows a significantly negative effect (À0.306, p < 0.001) on POC, which supports H6. The coefficient of the direct effect of perceived CSR on POC is not significant (À0.060, p = 0.507) at the 5% level in our estimated model. As this nonsignificant coefficient is a first indicator of a full serial mediation, we have to further explore the indirect effects, as well as the total effect of perceived CSR on POC. Figure 2 shows the direct effects considered in the research model.
Indirect effects. The total indirect effect of perceived CSR on POC is the sum of all indirect effects. It indicates the extent to which the long-way and all shortcut mediations together explain the effect of perceived CSR on POC. The total indirect effect of perceived CSR on POC is negative (À0.495) and significant, as the bootstrapped 95% confidence interval (À0.678; À0.327) does not include zero (Demming et al., 2017). The estimation results illustrated in Table 5 indicate that the long-way mediation of perceived CSR on POC via PEL and JS (path a1db2) is significantly negative (effect: À0.107; 95% CI: À0.182 to À0.049). Further, the indirect effect (path a2b2) of perceived CSR on POC via JS is also significantly negative (effect: À0.121; 95% CI: À0.200 to À0.059). Finally, the indirect effect of perceived CSR on POC via PEL is significant (effect: À0.267; 95% CI: À0.438 to À0.101) as the bootstrapped 95% confidence interval does not include zero.
Considering both the significant negative path of the long-way mediation and the nonsignificant direct effect of perceived CSR on intention to emigrate, our results fully support the serial mediation hypothesis: Perceived CSR leads to increased PEL, which leads to higher JS, which ultimately contributes to lower POC. Therefore, we can report that  Notes: PCSRperceived corporate social responsibility; PELperceived ethical leadership; POCperceived organizational corruption; JSjob satisfaction. Level of significance: ÃÃÃ p < 0.001 serial mediation via the mediator's PEL and JS in our model (H7) is supported. Finally, the findings show that the total effect of perceived CSR on POC is significantly negative (À0.555, p < 0.001). This negative significant result is in line with H1, which is therefore supported. See Table 5 for detailed regression results.

Discussion
Despite the growing importance of CSR in developing countries, until now, employeefocused CSR research has received little attention in a developing country context (De Roeck et al., 2014;Jamali and Karam, 2018;Grabner-Krä uter et al., 2020). By taking a micro-level analysis perspective, the purpose of this research was to investigate associations among employees' perceptions of CSR, ethical leadership, job satisfaction and perceptions of organizational corruption, as well as to highlight the underlying mechanisms that might explain these relationships. In line with prior micro-CSR research (Aryati et al., 2018;Gorsira et al., 2018;Iqbal et al., 2018;Sarfraz et al., 2018;Choi et al., 2018), our findings suggest that perceived CSR is significantly related to ethical leadership and job satisfaction. Furthermore, our findings identify CSR perceptions of employees as a novel associated factor of organizational corruption perception and suggest that the relationship between perceived CSR and perceptions of corruption is fully mediated by ethical leadership and job satisfaction. In a nutshell, it appears that organizational corruption is lower in organizations that are highly engaged in activities related to CSR. As a whole, these findings offer several theoretical insights and practical implications.

Theoretical implications
In line with prior micro-CSR research (for example, Glavas and Kelley, 2014;De Roeck and Maon, 2018), our study demonstrates that perceived CSR is positively related to favorable work-related outcomes and thus, corroborates previous findings in a hitherto largely neglected developing country context. This study also supports findings by Elçi et al. (2013) who found that ethical leadership and job satisfaction are positively interrelated. From a theoretical perspective, our results are consistent with Yukl's (2013) assertion that leaders influence their employees' attitudes and behaviors. Ethical leadership is closely related to multiple work outcomes of employees; thus, organization leaders represent a relevant factor in influencing their employees' ethical behavior (Brown et al., 2005).
Furthermore, our results add to previous research showing that improved ethical leadership can reduce deviant behaviors of employees in the workplace. For instance, our findings corroborate the results of Downe et al. (2016) who found that the actions of leaders are important in promoting good conduct and fostering an ethical culture. Our results also complement those of Trevino and Weaver (2001), who found that when employees perceive general organizational justice and ethics program follow-through, there is less unethical behavior and greater willingness to report problems. More specifically, our study demonstrates that employees' positive perceptions of CSR activities can diminish organizational corruption perception. In this way, our results are in line with Appelbaum et al. (2007), who found that employees are less likely to engage in deviant behaviors in a workplace environment with an ethical and responsible climate. Moreover, the results add to the findings of De Graaf and Huberts (2008) which revealed that weak leadership represents one of the most relevant causes of corruption.
Finally, this study included job satisfaction as a mediator variable in the relationships among perceived CSR, ethical leadership and organizational corruption perception. Based on our results, it appears that job satisfaction is affected by perceptions of CSR and ethical leadership, and that job satisfaction negatively influences organizational corruption perception. The results are in line with other empirical studies showing that job satisfaction plays a prominent role in workplace deviance (Marcus and Wagner, 2007;Nasir and Bashir, 2012). This suggests that it is favorable to create working conditions that promote satisfaction among employees because higher satisfaction among employees might help reduce the phenomenon of corruption at both the individual and organizational level.

Practical implications
In terms of practical implications, our research suggests that managers and supervisors are well advised to invest time and resources in promoting perceived CSR, ethical leadership and job satisfaction. These outcomes are usually associated with positive organizational results (Martin and Cullen, 2006), as a plethora of research evidence demonstrates. Our results suggest that establishing CSR practices is strongly associated with ethical leadership, which then impacts on the extent of employees' deviant and corrupt behaviors in the workplace. When managers or supervisors act responsibly and act in compliance with ethical standards, employees might as well refrain from committing unethical and corrupt actions. Furthermore, this study demonstrates that perceived CSR also is positively related to job satisfaction and thus points to another means of discouraging employees from corrupt actions that damage the organization.
The results of the study provide further evidence of positive associations among employee perceptions of CSR and valued outcomes, supporting the development of a clearer and stronger business case for CSR in private and public organizations in Kosovo. There is a favorable opportunity that as organizations become aware that they can gain economic benefits from well-managed CSR practices, they become gradually more motivated to invest supplementary resources to enlarge their CSR practices (Grabner-Krä uter et al., 2020). This "virtuous cycle" among CSR and economic benefits (Orlitzky, 2005) could encourage Kosovar organizations to enhance their CSR investments over time, and the prospective result would be greater social value and macro-level societal impact created by private and public organizations' CSR actions.
Furthermore, the findings of the present study confirm previous recommendations that organizations have to communicate their CSR initiatives to positively affect work-related outcomes (De Roeck et al., 2014). Hence, organizations are encouraged to share information about their CSR activities especially with their employees, as most relevant internal stakeholders and include elements of social responsibility in their communication strategy. As illustrated in a recent study by Rupp et al. (2018), organizations should make efforts to make CSR initiatives known (both formally and informally) among their employees, for instance, through email notifications, staff meetings and social media to enhance employees' CSR perceptions. CSR programs could be co-developed by specialists in human resource management to provide better working environments and organizations with short and long-term performance goals. In this way, it is possible to start fighting the phenomenon of corruption at the organizational level, i.e. from the workplace and then to extend the policies and measures also to the social and national level.

Limitations and directions for future research
As with any empirical work, this research has several shortcomings that should be kept in mind when interpreting the results and that need to be considered in future studies. A first limitation is that our analysis was conducted in a single country (i.e. Kosovo), thus restricting the generalizability of the results. To ensure that our results are not context-dependent, further research should be performed in other national and work contexts. While the essence of CSR might be a universal value regardless of peculiarities in Eastern European society, for sure, there are cultural differences in CSR comprehension. Data were collected from organizations in Kosovo and, therefore, the possibility of such cultural impacts cannot be excluded. Additional studies that analyze these cultural issues and their effects are needed. Furthermore, the cross-sectional data used in the analysis cannot support causal hypotheses, nor can such data guarantee the direction of the hypothesized causal links in the model.
As other researchers have pointed out, qualitative researches are relevant to improve our comprehension about human experiences and social interactions. The use of qualitative methods, including myths, narratives, stories, is expected to provide interesting insights into individual perceptions of CSR activities and their behavioral consequences.
Another limitation is that our study only reports the evaluations offered by employees, not actual corporate behaviors. Furthermore, same-source bias or general method variance bias could be another limiting factor because we used self-report measures obtained from the same sample in our analyses. In addition, because all estimation procedures implemented in PROCESS are based on observed variables, the usual problems associated with measurement error in predictors and outcomes in linear models can arise (Hayes, 2018). Using structural equation modelling and estimating latent variable measurement models could solve some of these problems.

Conclusion
Despite the abovementioned limitations, the current findings add to the CSR literature by improving our understanding of CSR dynamics at the individual and organizational level in a developing country context. Our findings extend CSR research by identifying CSR perceptions of employees as a novel associated factor of organizational corruption perception. This study provides evidence that perceived CSR is positively related to perceptions of ethical leadership and job satisfaction, which then is negatively associated with organizational corruption perception. In this way, our research helps to deepen the comprehension of micro-foundations of CSR by elucidating the underlying mechanism by which perceived CSR might be associated with perceived organizational corruption in a developing country context.
In developing and emerging markets, acts of corruption and other types of deviant behavior often arise as a consequence of dissatisfaction and frustration, going beyond the effects of rules and laws imposed by the public institutions. The phenomenon of corruption is very complex as it depends on a number of factors. However, the results of this study showed the potential role of CSR in fostering an ethical and benevolent environment that discourages corrupt behaviors. Based on these results, the government on one hand and organizations on the other hand are well-advised to develop policies that favor CSR engagement both at national and organizational levels, especially in developing countries where corruption is widespread and deeply embedded. Corruption needs to be tackled at institutional and individual levels simultaneously; the state, public and private organizations and each individual citizen have to be involved in the fight against corruption.