The effect of compensation and religiosity on managers’ CSR decision

This study aims to examine the relationship between compensation schemes and religiosity and managers’ investment decisions in CSR. We argue that compensation scheme and managers’ level of religiosity associate with managers’ choice whether to invest in valueincreasing or value-decreasing CSR. Data were collected through a laboratory experiment which involved 100 participants. The Chi-square test results show that managers who work under a performance-based compensation tend to choose a value-increasing CSR. Meanwhile, managers who work under nonperformance-based compensation tend to choose a value-decreasing CSR. However, this study failed to prove the relationship between manager’s religiosity level and CSR investment decision made by managers. A small variance of our participants’ religiosity level may lead to this finding. This research contributes to the CSR and behavioral finance literature by providing an understanding of how compensation and religiosity can direct managers’ investment decisions on CSR. Our results imply the importance of designing an appropriate compensation scheme in an organization in order to direct managers to make an optimal CSR decision.


Introduction
Social responsibility describes corporate actions that are not limited to meeting the company's economic goals and other goals set by law, but includes those for social purposes (McWilliams & Siegel, 2000). Companies that carry out CSR activities are often regarded as 'doing good' (Wood, 2010) because the activities do not solely provide benefits for the company, but are expected to provide benefits to the broad stakeholders. A company involvement in CSR activities is often based on an expectation that the CSR activities will bring a positive impact to the company performance, through an improvement of company's reputation (Wang & Qian, 2011;Brammer & Millington, 2005;Godfrey, 2005;Porter & Kramer, 2006;Saiia, Carroll, & Buchholtz, 2003), increased consumer satisfaction (Mohr & Webb, 2005;Mohr, Webb, & Harris, 2001), and increased employee productivity (Balakrishnan, Sprinkle, & Williamson, 2011). Hence, corporate involvement in CSR activities tends to be strategic (Saiia et al., 2003;Werbel & Wortman, 2000), and seen by managers as an investment to improve company performance. Therefore, instead of considering it as expenditure, fund allocation to CSR activities are seen as investment that will generate benefits in the long run.
In the context of Indonesia, very few companies in Indonesia undertake CSR in following their companies' vision and mission (Fauzi, 2014). Creating a positive image is the motive for most companies to involve in CSR practices (Fauzi, 2014;Gunawan, 2007). This implies that most companies in Indonesia also adopt a strategic view of CSR. Literature about CSR practice in Indonesia is dominated by an investigation on the relationship between CSR and company's performance (Kartadjumena & Rodgers, 2019;Afiff & Anantadjaya, 2013) as well as an exploration of the motives underlie companies' engagement in CSR. However, studies on the motivations underlie companies' involvement in CSR are mostly explorative in nature. Hence, an empirical evidence on the factors that may motivate companies to engage in CSR which provides a causeeffect relationship would be significant to provide a deeper understanding about CSR practice in Indonesia.
CSR decisions are individual manager decisions that may include those of what activities to do and how much funds to be invested. Because CSR decisions are on manager discretion, the decisions may vary among managers across companies. A manager may choose to invest in CSR that can increase the value for shareholders (value-increasing CSR), however, for certain reasons, such as the altruistic nature of a manager or manager's desire to get personal benefits from the CSR activities, a manager may make a decision that can potentially reduce the company value (value-decreasing CSR).
Considering that CSR has an ethical dimension, managers' CSR related decisions appears to have ethical nuance. O'Fallon & Butterfield (2005) argue that manager's ethical decisions are influenced by organizational and individual factors simultaneously. Accordingly, this study predicts that CSR decision made by managers is also influenced by both factors. This study considers compensation and religiosity as organizational and personal factors that might influence managers' CSR decisions for several reasons. First, compensation is one of the most widely used tools to direct managers' decisions and actions to be in line with the interests of shareholders. Compensation is an organizational factor that has been shown to play a role in influencing managers' decisions and behavior (Malmendier & Tate, 2015;Baxamusa, 2012;Malmendier & Tate, 2005a;2005b). Second, we argue that religiosity contains moral or ethical values that is in line with CSR in which moral or ethical values are one of the dimensions. Moral values often underlie individual decisions and actions. That is why investigating the role religiosity play to influence managers' CSR decisions is of important. Particularly in the context of Indonesia, the investigation becomes more important

The effect of compensation and religiosity on managers' CSR decision
Poppy Dian Indira Kusuma, Wiwiek Rabiatul Adawiyah, Bambang Agus Pramuka, Zahrotush Sholikhah | 269 | because Indonesia is a country where most of the population believes in God (Suryadinata, Nurvidya, & Ananta, 2003). Thus, we expect that religious values would greatly influence the decision and behavior of its population.
Although compensation is generally designed to produce an alignment effect (i.e. an effect that directs managers' decisions and behavior to be in line with the interests of shareholders), however, some researchers prove otherwise. Studies examining the effect of compensation on managers' decision and behavior still show a mixed result. Some researchers provided evidences of the alignment effect of compensation (Lei et al., 2014;Baxamusa, 2012;Broussard, Buchenroth, & Pilotte, 2004), but some others demonstrated an entrenchment effect where compensation actually motivates managers to take action which is not in line with the interests of shareholders (Baxamusa, 2012;Aggarwal & Samwick, 2006). We argue that this mixed result might be due to two reasons, the compensation effect on managerial decision and behavior might be contextual in nature. In a certain context, compensation may result in an alignment effect, but in other context it can give an entrenchment effect. In addition to the contextual nature of compensation, managers' decision and behavior may not only be affected by compensation, but also influenced by personal factors which are embedded within the managers. One of the personal factors that may influence managers' decisions is religiosity. Religion teaches ethical and moral values which may play a direct role in providing life guidance for humans. At the organizational level, religiosity may also play an indirect role in corporate governance. Therefore, discussing CSR from a religious perspective should provide a better understanding of CSR. Several studies have tried to understand CSR from ethical and moral dimensions (Hollingworth & Valentine, 2015;Arli & Tjiptono, 2014;Deng, 2012;Reed, Aquino, & Levy, 2007). However, religion related dimension is often ignored (Raimi et al., 2012).
Religiosity is a belief in God which makes a person committed to follow the principles and rules which are believed to originate from God (McDaniel & Burnett, 1990). The relationship between religion and business ethics has been the subject of significant recent conceptual and empirical works (Epstein, 2002;1998;Calkins, 2000;Weaver & Agle, 2002). Previous research has proven the relationship between religion and business (Parboteeah, Hoegl, & Cullen, 2008;Vitell, Paolillo, & Singh, 2005). A growing body of empirical studies has also examined the relationship between religiosity and managerial decisions in a business context (Longenecker, McKinney, & Moore, 2004). Meanwhile, the relationship between manager's religiosity level and CSR decisions has also been widely discussed in the literature (Ibrahim, Howard, & Angelidis, 2008;Parboteeah et al. 2008;Angelidis & Ibrahim, 2004;Conroy & Emerson, 2004;Giacalone & Jurkiewicz, 2003;Weaver & Agle 2002;Trevino, Hartman, & Brown. 2000). Although the relationship between the religiosity level of managers and CSR decisions seems obvious, empirical evidence still shows an inconclusive result (Van der Duijn Schouten, Grafland, & Kaptein, 2014).
Even though compensation can be in the form of other than money, but money commonly takes the biggest portion of the compensation. The role of money and religiosity in determining ethical decisions has been proven by several researchers (Singhapakdi et al., 2010;Vitell et al., 2005;. A love for money and religiosity often consider to be two factors that may having a contradictory effect on human behavior. Through experimental testing, this study suspects that managers who work under different compensation schemes and with different level of religiosity will have different CSR decisions. Specifically, the purpose of this study is to examine how managers with certain compensation schemes (i.e. performance based vs nonperformance based) will choose between value-increasing CSR or valuedecreasing CSR. In addition, the same examination | 270 | will also apply to different levels of religiosity (high vs low). A value-increasing (a value-decreasing) CSR refers to an investment in CSR in an optimal (a nonoptimal) amount. A value-increasing CSR is expected to increase value for shareholders while a valuedecreasing is assumed to reduce the value for shareholders.
Our study contributes to CSR literature in the form of an understanding of the role of compensation and religiosity in motivating managers to choose the optimal CSR investment. Our study also provides an insight to the practice about the importance of planning and designing an appropriate compensation scheme in a company. We structured our paper as follows. The next section contains a review of the literature and hypotheses development. This is followed by an explanation of the methods used in this study and a discussion of the results. This paper ends with conclusions and implications.

Hypotheses Development
Compensation is often used by companies as a strategic tool to control manager's actions, including an action of making CSR investment decision. While it is obvious that compensation has an influence on manager decisions, however, studies that link manager/CEO compensation and CSR still shows a mixed result (Jian & Lee, 2015). Some studies have shown that CEO compensation is negatively related to CSR (Coombs & Gilley, 2005;Russo & Harrison, 2005). Some others show that CSR-related performance can increase CEO compensation (Berrone & Gomez-mejia, 2009). Although finance literature has largely discussed the relationship between manager compensation and managers' investment decisions (Baxamusa, 2012;Aggarwal & Samwick, 2006), however, research linking manager compensation with manager CSR decisions is still lacking. Jian & Lee (2015) argue that managers may differ in their CSR decisions. The decisions include that relates to how much fund a CEO will allocate to CSR. Some managers may allocate fund for CSR activities at optimal level, while some others may not. In general, Jian & Lee (2015) examine the effect of CSR investment decisions made by CEOs on the amount of compensation to be received by them. Results on this study indicates that CEOc will be rewarded when investing in CSR at an optimal level. Otherwise, they will get a small amount of compensation.
In contrast to Jian & Lee (2015), we attempt to examine the association between managers' compensation and manager decision regarding investment on CSR. Instead of examining the effect of manager investment decision regarding CSR on manager compensation, as in Jian & Lee (2015), we investigate how compensation scheme may influence managers' investment decision on CSR. Evidence on how compensation may affect managers' CSR decision in the context of Indonesia is still lacking. A study by Kusuma, Sholihin, & Agritansia (2018) is one of the exceptions, which shows that compensation scheme has an influence on managers' ethical judgment regarding overinvestment on CSR. Based on the work of Kusuma et al. (2018), our study predicts that different compensation scheme will produce different CSR investment decisions as well. We assume that manager will invest if doing so will bring profit to the company and/or personal benefit to manager.
Investing in CSR has been viewed as either increasing value for shareholders or spending money in unprofitable project (Jian & Lee, 2015). An increased value for shareholders, based on the first view, may be due to some positive outcomes resulting from investment in CSR. Among those positive outcomes may include an increase in company reputation (Wang & Qian, 2011;Porter & Kramer, 2006;Brammer & Millington, 2005;Godfrey, 2005;Saiia et al., 2003), corporate relational wealth (Fombrun Gardberg, & Barnett, 2000;Godfrey, 2005;Griffin, 2004), access to important resources (Fombrun et al., 2000), customer satisfaction (Mohr  Mohr et al., 2001), relief from government (Shane & Spicer, 1983;Spicer, 1978), potential employees intention to join (Harjoto & Jo, 2015;Turban & Cable, 2003;Albinger & Freeman, 2000;Greening & Turban, 2000), and employees motivation and performance (Balakrishnan et al., 2011). This first view is in line with stakeholder theory which suggests that CSR can reduce conflicts of interest between stakeholders and can build good relationships with stakeholders. According to the second view, managers may invest in CSR to get personal benefits such as: to increase self-reputation (Barnea & Rubin, 2010;Hemingway & Maclagan, 2004). To improve reputation, managers will focus more on short-term profits and make nonoptimal investments to CSR.
Those two views show that managers as decision makers may not always make decisions at an optimal level. For certain reasons, such as: personal benefits or self-reputation, managers can distort their investment decisions by choosing investment on a value-decreasing CSR. We argue that managers' expectation regarding compensation can be a factor which leads to different decisions among managers. Thus, compensation schemes may play a role in differing managers' decision. Based on an egocentrism argument, we argue that compensation scheme can create an egocentric decision as compensation scheme may create a different amount of compensation received by managers.
An argument in the equity theory (Adams, 1965) suggests that people normally want to have their compensation comparable to others'. In other word, they want a fair distribution of compensation. However, literature of distributive justice concluded that the judgment of distributive justice is not only constructed based on equity norm, but is also formed based on egoism and egocentrism (Burrus & Mattern, 2010). Egocentrism perspective argues that people egocentrically have a care to themselves more than others. Thus, this perspective presumes that judgment on a distributive justice is not formed in a comparative way.
Based on an egocentrism perspective, we predict that compensation schemes can influence managers' decisions whether to invest in value-increasing CSR or value-decreasing CSR. A value-decreasing CSR in this study is a proxy for an excessive fund allocation for CSR that requires a transfer of wealth from shareholders to other stakeholders. Meanwhile, a value-increasing CSR is a proxy of an optimal fund allocation for CSR that is expected to bring benefit to company in the long run. We predict that managers with performance-based compensation (i.e. a compensation scheme that directly relates manager compensation to company performance) will tend to choose investments in CSR that will increase value for shareholders. Managers under this compensation scheme are more likely to pay more attention to the company performance because an increase in shareholder wealth will have a positive impact on the compensation that the manager will receive.
In contrast, a manager working under nonperformance-based compensation (i.e. a compensation scheme that is not directly related to company performance) will be less motivated to choose a value-increasing CSR because an increase and decrease in company performance has little effect on the amount of compensation he will receive. Furthermore, choosing a value-decreasing CSR may actually bring personal benefits to the manager. By investing a large amount of fund to CSR, managers' reputation, in the eyes of stakeholders, is more likely to increase. At the same time, however, such investment will harm the wealth of shareholders. Therefore, the first hypothesis of this study is formulated as follows: H 1 : compensation schemes relate to managers' CSR investment choices. Specifically, managers working under performance-based compensation are more likely to choose value-increasing CSR, compared to those under nonperformance-based compensation.

| 272 |
Some researchers argue that religiosity has a strong impact on work values (Parboteeah et al., 2008) which can further influence attitudes and behavior (Ramasamy et al., 2010). According to Allport & Ross (1967), one's religious orientation shows how a person lives religious values in his life. A religious person will involve his religious values to make decisions in his life. Hunt & Vitell (1993) explain that religiosity can influence one's business decisions, particularly those that involve ethical issues. Given that CSR has concerns about ethical issues (Carroll, 1979), religiosity is a personal factor that might influence managers' investment decisions on CSR.
Although concern for social and environmental matter is a corporate responsibility, however, decision related to CSR is an individual decision of manager. Personal factors of manager may affect manager's investment decisions on CSR. Religiosity has been proven to have a positive impact on a person's ethical attitudes and behavior. For example, in the context of marketing, Cyril De Run et al., (2010) show that religion is a tool to filter messages coming to someone, which can further influence the outcome of marketing communication. Adi & Adawiyah (2016) show that religiosity has a capability to moderate the relationship between environment orientation and environmental marketing practices. Another example is the work of Chitwood, Michael, & Carl (2008) that demonstrates the role of religiosity to reduce the behavior of drug abusing. Ideally, one's understanding of religion will have a positive impact on one's ethical decisions and behavior. However, empirical evidence shows an anomaly, especially in the Indonesian context (Arli & Tjiptono, 2013). With the majority of the population trusting God (Suryadinata et al., 2003), various unethical behaviors still occur in Indonesia, such as: buying pirated films (McGuire, 2009) and corruption (Henderson & Kuncoro, 2004).
Understanding CSR from a religious perspective should further encourage companies to engage in CSR activities and reporting (Raimi et al., 2014). However, whether a manger will engage in value-increasing or value-decreasing CSR has been an open question that requires further investigation. According to Ferrel & Gresham (1985), individual factors that contain moral philosophy also influence one's ethical decision-making process. Religiosity is one personal factor that has strong moral values or philosophies. Religious ideology discusses a lot about one's judgment of right and wrong. Therefore, a different level of religiosity can lead to different ethical decisions (Clark & Dawson, 1996).
Based on personality and values theory, religion contains personal values (Fry & Slocum 2008) which shape one's personality or identity (Chusmir & Koberg, 1988). One's personality and values can then be used to predict their attitudes and behavior within the organization (Rokeach, 1986). In addition, Pichon, Boccato, & Saroglou (2007) argue that religion can directly influence managers' CSR related behavior because religion partly teaches about the importance of economic improvement and sustainable development. Religiosity within a manager will encourage the manager to embrace their social responsibilities. One aspect of social responsibility is generating profit as stated in 3P principles (i.e. profit, people, and planet). Increasing company value is closely related to achieving corporate profit. Therefore, we predict that high religiosity level of managers will direct their decisions towards valueincreasing CSR. This conjecture is formulated in the following second hypothesis.
H 2 : religiosity level of managers is associated with managers' CSR investment choices. Specifically, the higher the level of managers' religiosity the more likely managers will choose a value-increasing CSR.

Method, Data, and Analysis
A Laboratory experiment was chosen to be a method to collect the data. The use of experiment is appropriate for this study to establish a high degree of casual-effect relationship between compensation, religion, and managers' CSR decision. A

Poppy Dian Indira Kusuma, Wiwiek Rabiatul Adawiyah, Bambang Agus Pramuka, Zahrotush Sholikhah
| 273 | laboratory experiment was conducted involving 110 participants. The involvement of students in this experiment was voluntary. Participants were playing a role as managers. Of the 110 participants, 5 participants did not pass the manipulation check, 2 participants did not give a complete response, and 3 participants were excluded from the analysis because they show the same religiosity scores as average religiosity score of all participants. This results in a final sample size of 100 participants. Participants were undergraduate students from economics and business school and were in the fifth semester or up wards. Fifth-semester students or upwards were chosen since they have taken investment related courses, so it is reasonable to assume that they have sufficient investment knowledge. Liyanarachchi & Milne (2005) provide evidence that students can serve as surrogates for practitioners in making decisions. The selection of students as participants in this study is acceptable for several reasons. First, even though students may not have sufficient investment and managerial experiences, they have adequate knowledge about investing and managing. In addition, the natural characteristics of students will be more easily to be manipulated through an experiment (Trinugroho & Sembel, 2011) and will provide a more genuine response. Finally, students will become professionals in the future and according to Reiss & Mitra (1998), their behavior may mirror the future behavior of professionals. Previous studies also found to use students as surrogates for managers in various decision-making tasks are Kusuma et al. (2018), Madein & Sholihin (2015), Trinugroho &Sembel (2011), andRutledge &Karim (1999).
Acting as managers of a fictitious company, participants were asked to make a decision related to the allocation of funds for the 'Go Green Project,' a corporate social responsibility activity that will be carried out for the next year. A set of scenarios was given to each participant. The scenario illustrates the consequences will raise if the managers allocates fund for CSR at the optimal level and if they allocate it at non-optimal level. An allocation of fund at an optimal level is assumed to be increasing the value for shareholders. Thus, an optimal allocation of fund is a proxy for value-increasing CSR in this study. Conversely, the proxy for value-decreasing CSR is a non-optimal allocation of funds which is assumed to destroy shareholders' value.
We believe that defining the optimal allocation of fund to CSR is challenging and is not easy for companies since there is no limit to the funds that can be allocated to their stakeholders. Barnea & Rubin (2010) argue that the relationship of the amount of fund allocated to CSR and the firm value is non-monotonic. They argue that a small expenditure for CSR should positively affect firm value through various outcomes, such as: an increase in employees' productivity or an avoidance of reputational-related cost and fines. However, at some point, an additional amount of fund of CSR must decrease the wealth of shareholders. The optimal and non-optimal allocation of funds for CSR in this study is defined based on Kusuma et al. (2018). In view of the government requirement and based on a panel of CSR experts, they argued that on average, companies in Indonesia allocate fund for social responsibility by 2 per cent of the net profit. Any allocation exceeds 5 per cent of the net profit is considered as a non-optimal allocation, which can potentially reduce the value of shareholders.
Each participant was randomly assigned into a group under performance-based or nonperformance-based compensation scheme. Through the scenario, participants know the compensation scheme they get and the consequences that may raise as a result of their decisions. Participants are then asked to determine the percentage of fund allocation for the company's CSR activities. Participants choose between two options, i.e. whether to allocate a maximum profit of 2 per cent (value-increasing CSR) or more than 2% (value-decreasing CSR). Participants who choose value-increasing CSR | 274 | (value-decreasing CSR) are coded as 1 (0). The scenario was adapted from Kusuma et al. (2018), with a modification to suit the context of this study. The modification is made by involving experts and practitioners to produce valid and reliable instruments.
The independent variables of this study are the compensation scheme and religiosity. The compensation scheme variable is manipulated into 2 types: performance-based and nonperformancebased compensation scheme. Under a performancebased compensation, the compensation component will consist of basic salary and a bonus that is calculated based on the manager's ability to exceed the company's profit targets. In contrast, a participant under nonperformance-based compensation scheme will get basic salary and benefits that is in accordance with their position. Participants who get performance-based (nonperformance-based) compensation are coded as 1 (0).
The religiosity variable was measured using a scale developed by Allport & Ross (1967). The mean value of the response scores of all participants was used to group participants into group of high and low level of religiosity. Participants with a score above (below) the mean value were grouped in participants with high (low) level of religiosity. A participant who has score equal to the mean score was excluded from the analysis. Participants who were included as having a high (low) level of religiosity were coded as 1 (0).
Manipulation checks were carried out by asking the participants to respond to two questions using a 5 point-scale (from 1: strongly disagree to 5: strongly agree). The questions are asking if the amount of fund allocation for CSR will affect company's profit and if company's profit will affect the amount of compensation the manager will receive.

Results
Participants of this study were 100 students, consisting of 51% female and 49% male. As many as 94% of participants are Muslim and the rest are Protestant and Catholic. Chi-square test is used to examine the relationship between compensation schemes/religiosity level and manager decision regarding CSR investment. Table 1 shows the distribution of participants into the compensation scheme and CSR investment choice groups. A total of 51 participants received nonperformance-based compensation and 49 participants were under performance-based compensation. Of the 51 participants who received nonperformance-based compensation, 20 participants decided to choose value-increasing CSR and 31 participants chose value-decreasing CSR. Meanwhile, out of 49 participants with performance-based compensation, 30 participants chose to invest in value-increasing CSR and 19 people chose to invest in value-decreasing CSR.
Chi-square test results, as shown in Table 2, show a significance value of Pearson Chi-square of 0.028 or less than 0.05. This result indicates that compensation scheme is significantly associated with manager's decision in choosing CSR investments. The difference in investment choice made by managers under performance-based and non-perfor-  Table 3 shows a distribution of participants based on their religiosity level and their CSR investment choices. Out of 100 participants, 54 participants had a high religiosity level and 46 participants had a low religiosity level. Out of 54 participants categorized as having a high level of religiosity, 26 participants decided to invest in value-increasing CSR and 28 participants chose to invest in valuedecreasing CSR. Meanwhile, of the 46 participants who had a low level of religiosity, 24 participants chose to invest in value-increasing CSR and 22 participants chose to invest in value-decreasing CSR.
A result of Chi-square test, as shown in Table  4, demonstrates an insignificant value of Pearson Chi-square (0.688 or more than 0.05). This result indicates that the religiosity level of managers does not significantly correlate with managers' decisions in choosing CSR investments. Thus, hypothesis 2 of this study is not supported. From Table 3 it can be seen that managers with high religiosity levels tend to choose CSR value-decreasing, while managers with low religiosity levels tend choose value-increasing CSR.

Discussion
Previous studies provide evidence that compensation, in term of its amount (Jian & Lee, 2015) and its scheme (Kusuma et al. 2018;Hobson, Mellon, & Stevens, 2011), can influence managers' decisions. Unfortunately, the effect of compensation on man-  This study shows that different compensation schemes can have different effects on managers' investment decision on CSR. In the context of this study, performance-based compensation tends to have an alignment effect. This means that managers who are compensated under performance-based scheme tend to choose to invest in CSR to increase shareholder value since an increase in company performance will increase their compensation. Conversely, managers under nonperformance-based compensation tend to choose to invest in value-decreasing CSR. Under nonperformance-based compensation, managers will not gain benefit from investing in value-increasing CSR. Moreover, by investing more funds in CSR, which may do harm than good for the shareholders, managers may be appreciated as socially responsible managers. In other word, it can improve their self-reputation at the expense of shareholders.
These findings are in line with the argument of egocentrism that egocentric managers tend to make decisions that would bring benefits to them. In this study, compensation is proved to be one of the factors that lead to an egocentric behavior. These findings also lend support to Kusuma et al. (2018) who argue that managers under different compensation scheme have different exchange relationships with their companies. Further, the different exchange relationships result in different decisions. This result provides an additional empirical evidence on the findings of previous studies that also seek to link CSR to compensation (Kusuma et al., 2018;Jian & Lee, 2015;Berrone & Gomez-mejia, 2009;Coombs & Gilley, 2005;Russo & Harrison, 2005).
In addition to the compensation scheme, this study also examines the effect of managers' religiosity level on their investment decision on CSR. However, the result show that managers' religiosity is not associated with manager's CSR investment decisions. The small variation in participants' reli-giosity level may contribute to this finding. A variance of 12.7 percent of our data indicates a small variance. This small variation may be due the limitation of religiosity measurement we used in this study. In measuring the religiosity level, an average score of religiosity of all participants served as a base to group our participants into high and low religiosity level. By using this method, when the average of religiosity score of all participants is too high (too low), an individual participant with a high score of religiosity has an unexpected chance to be assigned to a low (high) religiosity group. In other word, this method may not reflect a true religiosity level of the participants.

Conclusion
Given the above results, our study provides significant contribution to CSR and finance literature in providing explanations on the relationship between compensation, religiosity, and managers' investment decision on CSR. Our analysis produces two main results. First, compensation schemes have a relationship with managers' investment decision on CSR. Managers who work under performancebased compensation are more likely to choose an investment on CSR which creates value for shareholders. Meanwhile, those who are under nonperformance-based compensation tend to invest in value-decreasing CSR. Second, managers' religiosity does not associate with managers' investment decision on CSR. In term of practical contribution, our study provides an insight for companies to the importance of designing a suitable compensation for managers because a different compensation scheme has different effect to managers' decision.
This study is not without limitations. The use of laboratory experiment in this study may one of the limitations. A laboratory experiment often involves simplification on certain reality. For this reason, it would be worthwhile to consider field experiment for the future research where an experiment can be carried out in a more natural setting.

| 277 |
Another limitation is related to the method we used to measure the religiosity. The measurement of religiosity level in our study was based on the average score of religiosity of all participants. This method contains weaknesses because a small variation in the level of participant's religiosity could not be anticipated. To that end, the next researcher may consider measuring the religiosity level of participant before the experiment in order to get participants with varying degrees of religiosity. In addition, researchers vary in their views of the use of student as professional surrogate. Even though we agree with the view that with careful selection, stu-dents are sufficient to represent professional in an experiment, we do not deny that the use of student as surrogate for professional may also introduce weaknesses. Lack of experience could be one of the weaknesses that may lead to a biased response. For that reason, future studies could consider professional managers as participants in the experiment. Finally, future studies may consider a qualitative research method as an alternative method as it is also suggested by Winschel & Stawinoga (2019). A qualitative research method may serve to provide richer information.