Does public service motivation truly predict dishonesty? Behavioural evidence from the private and public sectors

The number of studies on the relationship between public service motivation and ethics is increasing. Although research consistently finds evidence for a positive public service motivation–ethical intentions relationship, research on public service motivation and unethical behaviour remains scarce and showcases mixed findings. Based on insights from person–environment fit theory, we argue that the sector in which an individual is employed may help explain these mixed findings. We test this assumption by employing an incentivized dice game among private and public sector workers (n  =  576). Using a well-powered research design we find neither support for the public service motivation–ethical behaviour relationship, nor for the sector as an intervening variable. We discuss the implications of these interesting null findings, which challenge the common assumption that public servants are morally superior to private sector workers. Points for practitioners (Human resource) managers of both private and public sector organizations need to be aware of the fact that employees engage in behavioural dishonesty (i.e. cheating). Contrary to the often-heard assumption, we find no evidence that public service motivation serves as a buffer for this specific type of unethical behaviour. This means that stimulating public service motivation or recruiting public-service motivated individuals does not help to reduce the risk of behavioural dishonesty. Rather, (human resource) managers can invest in other measures to limit this type of unethical behaviour such as ethical leadership and an ethical climate.


Introduction
Following the highest standard of ethical conduct is inherent to the mission of public organizations (Belle and Cantarelli, 2017).Nevertheless, public servants across the world are frequently confronted with and engage in unethical behaviour (Meyer-Sahling et al., 2019).Our understanding of what causes unethical behaviour and how it can be reduced, therefore, remains of utmost importance (Belle and Cantarelli, 2017).A concept that has been associated with high levels of morality is that of public service motivation (PSM) (Olsen et al., 2019;Wright et al., 2016) or 'an individual's orientation to delivering services to people with a purpose to do good for others and society' (Perry and Hondeghem, 2008: vii).Public service motivation and ethics are assumed to be related because they reflect similar public values and promote the public interest against behaviour driven by self-interest (Ripoll and Schott, 2023) A growing body of empirical research indeed shows that PSM is positively associated with ethical intentions and attitudes and negatively associated with unethical intentions (Meyer-Sahling et al., 2019;Stazyk and Davis, 2015;Van Roekel and Schott, 2022;Wright et al., 2016).However, findings are mixed when the relationship between public service motivation and actual (un)ethical behaviour is studied.Although Olsen et al. (2019) find public service motivation to be positively linked to behavioural dishonesty, which presents one aspect of unethical behaviour (Lasthuizen et al., 2011;Treviño et al., 2014), Gans-Morse et al. (2022) find no relationship between those two concepts.This is interesting because both studies use the very same measurement instruments and rely on data collected among university students.However, a noticeable difference between the studies is their sample selection.Although Olsen et al. (2019) focused on students who were likely to enter public sector careers, Gans-Morse et al. ( 2022) recruited a random sample stratified by university departments.This makes us wonder whether sectoral differences may play a role in the public service motivation-behavioural dishonesty relationship.
This study aims to explain the mixed findings.We use insights from personenvironment fit theory (Kristof- Brown et al., 2005) and argue that the sector employees work inor the departments students are socialized bymay influence the relationship between public service motivation and behavioural dishonesty.Public organizations are traditionally associated with values that have a strong link with ethics and morality, such as lawfulness and incorruptibility (Van der Wal et al., 2008).Public service-motivated individuals working in the public sector may, therefore, experience a better fit with their working environment and are consequently more likely to show behaviour reflecting their public service motivation; that is, behavioural honesty.Typical private sector values are, in contrast, profitability and innovativeness (Van der Wal et al., 2008).Individuals with high levels of public service motivation working in the private sector may not automatically engage in moral conduct, because the environment they are working in emphasizes conflicting values.Based on this line of reasoning, we aim to answer the following research question: does the sector individuals work in help explain the mixed findings regarding the relationship between public service motivation and behavioural dishonesty?
The results of our well-powered study show that: There is no significant negative relationship between public service motivation and behavioural dishonesty.This relationship is the same for people working in the private and the public sectors.
These findings are relevant as they not only contribute to the limited research on (un) ethical behaviour in the field of public administration in general (Bélle and Cantarelli, 2017) and the public service motivation-ethics relationship in particular (Van Roekel and Schott, 2022), but also to the discussion on the distinctiveness of public and private sector employees.Public organizations and their employees 'operate under greater public scrutiny and are subject to unique public expectations for fairness, openness, accountability, and honesty' (Rainey, 2009: 86).However, we know very little when it comes to the question of whether they also act more ethically (Sulitzeanu-Kenan et al., 2022).This study contributes to these gaps in our knowledge by studying the relationship between public service motivation and ethics using data collected among public and private sector employees (n = 576).We assess ethics using a behavioural approach by asking participants to play a simple dice game (Gans-Morse et al., 2022;Olsen et al., 2019), which presents a good predictor for real-world dishonest behaviour (Hanna and Wang, 2017;Jiang, 2013;Potters and Stoop, 2016).
To unfold these contributions, the article proceeds as follows.First, we will develop the theoretical framework and discuss why: Public service motivation is expected to be negatively associated with behavioural dishonesty.Sectoral differences are expected to influence this relationship.
Following a description of the methodology applied in the study, we will test our two hypotheses through a series of ordinary least squares (OLS) regression models.Finally, we will elaborate on the main contributions of this study, its limitations and implications for research.

Introducing public service motivation
Forty years ago, Rainey (1982) laid the foundation for the concept of public service motivation by asking a large sample of public and private managers to indicate their willingness to engage 'in meaningful public service' (288).Since then, research has increased immensely as shown by the growing number of meta-analyses on, for instance, the relationship between public service motivation and organizational correlates (Harari et al., 2017) and public service motivation and job satisfaction (Homberg et al., 2015).Scholars commonly agree that public service motivation is a multidimensional concept (e.g.Kim et al., 2013;Neumann and Schott, 2021;Olsen et al., 2019;Schott et al., 2018) consisting of four dimensions: 'self-sacrifice' (SS), 'compassion' (COM), 'attraction to public participation' (APP) and 'commitment to the public values' (CPV).
The origin of public service motivation lies in public institutions.By means of social processes, such as socialization and social learning, public values and institutional logics are transmitted and individuals are expected to 'acquire a new social identity as members of the institution' (Perry and Vandenabeele, 2008: 60).This helps explain why higher levels of public service motivation have consistently been found among public sector workers if compared to private sector employees (Lyons et al., 2006).

Introducing unethical behaviour and cheating
Unethical behaviour is a complex and multifaceted concept, which is difficult to define and measure (Lasthuizen et al., 2011).We define ethics as the collection of norms and values that function as standards to assess the integrity of a persons' conduct (Benjamin, 1990).Following Jones's (1991) seminal article, an ethical decision is a decision that is 'both legal and morally acceptable to a larger community.Conversely, an unethical decision is either illegal or morally unacceptable to the larger community' (367).Applying this definition to behaviour, unethical behaviour is either illegal or morally unacceptable to the larger community.This approach towards unethical behaviour is a contextual one; that is, common values and laws determine which type of behaviour is unethical and which type is acceptable and morally good.
Different typologies of unethical behaviour exist.For example, Lasthuizen et al. (2011) identify 10 different dimensions of unethical behaviour in their seminal article: (a) corruption in the form of bribing; (b) corruption in the form of favouritism; (c) fraud and theft of resources; (d) conflict of (private or public) interest through gifts; (e) conflict of (private and public) interest through sideline activities; (f) improper use of authority to achieve organizational goals; (g) misuse and manipulation of information, such as cheating an concealing information; (h) indecent treatment of colleagues or citizens and customers; (i) waste and abuse of organizational resources; and (j) misconduct in private time that harms people's trust in the (public organization).Every dimension contains more or less serious form of unethical behaviour (e.g.type (h): 'indecent treatment of colleagues', sexual harassment or violence as well as bullying and gossiping).In addition, this typology is an encompassing one as it does not only entail 'unethical conduct related to the performance of duties or directed at the organization and its members, but also misconduct during private time' (Lasthuizen et al., 2011: 389).Treviño et al. (2014) differentiate between three related types of (un)ethical behaviour in organizations, which describe a continuum ranging from unethical to very ethical behaviour: (a) 'unethical behavior that is contrary to accepted moral norms in society (e.g., lying, cheating, stealing)'; (b) 'routine ethical behavior that meets the minimum moral standards of society (e.g., honesty, treating people with respect)'; and (c) 'extraordinary ethical behavior that goes beyond society's moral minima (e.g., charitable giving, whistleblowing)' (636)(637).In this study we focus on the first aspect of Treviño et al.'s (2014) and Lasthuizen et al.'s (2011) seventh category of unethical behaviour by studying cheating, which presents a unethical behaviour that seeks to create an unfair advantage and enhances benefits for the actor (Mitchell et al., 2018;Pascual-Ezama et al., 2020). 1 This is in line with recent research on the relationship between public service motivation and unethical behaviour (Gans-Morse et al., 2022;Olsen et al., 2019).However, it also means that we only study a small aspect of unethical behaviour.

Public service motivation and behavioural dishonesty
After having introduced the main concepts of this study, we elaborate on their relationship.Research on the relationship between public service motivation and ethics (e.g.Brewer and Selden, 1998;Ripoll, 2022;Wright et al., 2016) relies on one main argument, namely that a positive relationship between these two concepts can be expected because 'they reflect similar public values and promote the public interest against behavior driven by self-interest' (Ripoll and Schott, 2023: 4).Put differently, whereas ethics can be seen as the collection of norms and values that function as standards to assess the integrity of a person's conduct (Benjamin, 1990), public service motivation refers to a motivation to act according to these norms and values.
Van Roekel and Schott (2022) discuss this relationship in more detail by making use of insights from identity theory.A key assumption of this theory is that individuals engage in specific behaviours because they want to signal to others and themselves who they are (Schott and Ritz, 2018;Stets and Burke, 2003).Regarding public service motivation, this means that the higher they score on public service motivation, the more likely they are to engage in ethical behaviour because this type of behaviour reflects who they are and what they find important.Consistent with this line of reasoning, public service motivation has frequently been found to be positively associated with ethical attitudes, such as a higher willingness to report unethical behaviour (Mayer-Sahling et al., 2019;Wright et al., 2016), whistleblowing intentions (Brewer and Selden, 1998) and with ethical obligations rooted in virtue and integrity (Stazyk and Davis, 2015).
Interestingly, studies focusing on the relationship between public service motivation and actual ethical behaviour: (a) are more limited; (b) exclusively focus on cheating behaviour, which presents only a small aspect of ethical behaviour (Lasthuizen et al., 2011;Treviño et al., 2014); and (c) report mixed findings.In particular, Gans-Morse et al. ( 2022) do not find a relationship between public service motivation and behavioural dishonesty (i.e.cheating), whereas Olsen et al. (2019) do find a positive relationship on the basis of very similar research designs (i.e. both studies use the very same measurement instruments and rely on data collected among university students).
The rationale for the expected negative relationship between public service motivation and cheating behaviour can be found in the public value '(un)fairness', which is central to both public service motivation and cheating behaviour.Cheating behaviour is a type of unethical behaviour that seeks to create an unfair advantage and enhance benefits for the actor (Mitchell et al., 2018).If people cheat, they intentionally ignore rules to gain an advantage.Public service-motivated individuals, in contrast, care for the value of fairness as it presents an important means to achieve outcomes that go beyond organizational interest and self-interest (Quratulain et al., 2019).This theoretical linkage has also been demonstrated by the constitutive items developed in the frequently used PSM measurement scale (Kim et al., 2013).Items such as 'I think equal opportunities for citizens are very important' and 'I get very upset when I see other people being treated unfairly' highlight the importance of equity and fairness to public service-motivated individuals.We therefore assume that public service motivation can serve as a buffer for cheating behaviour and put forward our first hypothesis: 2 H1: Public service motivation is negatively associated with cheating behaviour.

Public service motivation and behavioural dishonesty in the public and private sectors
This leaves us with the question of why the findings of Olsen et al. (2019) and Gans-Morse et al. ( 2022) are inconsistent.A noticeable difference between these two studies is that the former focused on students who studied public administration and indicated that their likelihood of getting a job in public administration is higher than 50%.The latter study, in contrast, recruited a random sample stratified by university departments without controlling for sector ambitions.
Using insights from person-environment fit (P-E fit) theory (Kristof- Brown et al., 2005) we argue that the fit (or misfit) between a public service-motivated individual and his/her environment may influence the relationship between public service motivation and behaviour dishonesty.A core argument of the P-E theory is that an incompatibility (or misfit) between the values of an individual and his/her work environment can cause negative effects whereas a fit (or compatibility) is beneficial for individuals and the environment.
Research on the so-called 'dark sides' of public service motivation draws heavily on this theory (Schott and Ritz, 2018).Scholars commonly argue and show that if public service-motivated individuals are prevented from acting upon their desire to provide meaningful public servicesif there is a misfit between them and their working environmentthis is likely to negatively affect their physical well-being (Liu et al., 2015), their intentions to stay in the current job (Quratulain and Khan, 2015) and their job satisfaction (Van Loon et al., 2015).Moreover, individuals are expected to be most successful and satisfied in organizations sharing their common values and personalities (O'Reilly et al., 1991).In line with this, research shows that employees who experienced a so-called 'public service motivation fit'that is, the extent to which they perceive that their values are congruent with those of the public sector organization they work forwere more satisfied and less inclined to leave their job and the organization they work for than employees who do not experience such a fit (Wright and Pandey, 2008).
Research provides evidence that dominant public sector values are likely to present a better fit with public service motivation than dominant private sector values.Put differently, the 'predominance hypothesis' stating that one set of values is becoming predominant in all organizations (Schultz, 2004) because of 'sector blurring' is not supported.Although public organizations are still associated with values that have a strong link with ethics and morality, such as lawfulness and incorruptibility, the values of profitability and innovativeness dominate private organizations ( Van der Wal et al., 2008).Or, using the words of Van der Wal and Huberts (2008), 'value solidity seems the dominant feature in both sectors' (264).
Based on these findings, we argue that public service-motivated individuals working in the public sector may experience a better fit between their public service motivationrelated values and their working environment than individuals working in the private sector.Consequently, they are less likely to engage in unethical behavioursuch as behavioural dishonestythat conflicts with their public service motivation if compared to employees who do not experience such a fit.In contrast, highly public servicemotivated employees working in the private sector may not automatically engage in moral conduct, because this type of behaviour is less facilitated by their working environment.Put differently, a mismatch between their public service motivation-related values and their working environment can be assumed, which may weaken the negative relationship between public service motivation and behavioural dishonesty.To test this line of reasoning, we put to the test the following hypothesis: H2: The relationship between public service motivation and cheating is moderated by the sector.

Data collection
To test our hypotheses, we used a survey based on a representative sample of UK citizens (n = 576), collected via an online panel company called Prolific.Ethical approval was obtained via the Faculty Ethical Review Committee of Utrecht University.The preregistration, data, syntax and materials can be found on the Open Science Framework (OSF): https://doi.org/10.17605/OSF.IO/D5A6C.

Sample and procedure
One potential concern when using panel data to test our hypotheses is the possibility of self-selection bias.To reduce the potential impact of non-probability sampling on our study, Prolific implements several measures.First, Prolific sources participants from multiple panels simultaneously, including panels with one-time and recurring participants.Additionally, we took care to ensure that the invitations we sent to participants did not reveal any information about the survey or specific tasks involved, minimizing the risk of self-selection related to the task itself.
Our sample for this survey was drawn from both the public and private sectors, as it was part of a larger study on the ethical behaviours of managers in the public sector.We used stratified sampling to ensure a representative sample based on age, sex and education levels.Forty percent of the participants were employed in the public sector and the remaining 60% were employed in the private sector.The sample demonstrated relatively low levels of risk-taking, with an average score of 2.09 (SD = 0.60) on the risk-taking index (1-5).The majority of the sample (54%) was female, and the average age was 39.9 years (SD = 9.85).None of the participants was removed from our dataset.A power calculation shows that based on a small effect size of f = 0.05, two predictors (one interaction and one dependent variable) and a power of 0.95, at least n = 312 participants are required to detect an effect (Faul et al., 2007).
We used the Qualtrics platform to present the survey to the participants.The survey itself consisted of a number of successive steps.First, participants signed informed consent and proceeded to read the instructions.Next, the participants played a dice game (see Dice task), followed by a public service motivation measure and control questions (see Measures).

Dice task
Our dependent variable is cheating rate (see Measurement).Following Olsen et al. (2019), we measured (dis)honesty among individuals by investigating their cheating behaviour in a simple dice game.Participants were asked to roll a fair die 20 times in total.The participants were instructed to write down the outcome of the dice throws (Jiang, 2013;Potters and Stoop, 2016).The participants are reimbursed for the number they write down as an outcome of their throws.Hence, there is some incentive for participants to write down a higher number than the dice shows.Cheating behaviour is deduced from the a-priori known distribution of a fair dice, which can be compared to the scores the participant entered.The respondents were reimbursed for their participation based on a fixed component (1 GBP for participation/1.13EUR) and a variable component of 0.01 GBP based on the die throws of the participants.The fixed component is based on minimum wage (1 GBP for 6 minutes' work).The theoretical minimum variable earnings are 0.20 GBP (0.23 EUR) and the maximum is 1.2 GBP (1.36 EUR).The average variable component of our participants was 0.74 GBP (0.84 EUR).
The primary advantage of this behavioural measure above self-reported data on, for instance, rule breaking or lying is that it is less biased due to social desirability and strategic behaviour (Olsen et al., 2019).Analogous to Jiang (2013) and Erat and Gneezy (2011), this study focuses on 'black selfish lies' which benefit the self at the cost of the other.This paradigm has frequently and reliably been used in experimental lab settings, survey research and in the field (Jiang, 2013;Pascual-Ezama et al., 2020;Potters and Stoop, 2016) as well as with a range of participants, such as students, practitioners and children (Bucciol and Piovesan, 2011) and public employees (Hanna and Wang, 2017).In public administration, Olsen et al. (2019) and Gans-Morse et al. (2022) used this game to infer cheating behaviour of students in Denmark and of public servants and students in Russia and Ukraine, respectively.Together, these studies suggest that cheating behaviour in this type of dice game presents a good predictor for real-world dishonest behaviour.

Measures
Cheat rate.In order to assess dishonesty, we use an individual-level estimate called cheat rate, which is identical to the approach by Barfort et al. (2019).The cheat rate can be formalized as follows.In our experiment we use a sample n, indexed by i.All participants throw the die k times, indexed by k.As mentioned earlier, our experiment has n = 576 and k = 20.In each round, the probability of winning is 1 6 .In our experiment, we did not observe whether participants wrote down their scores honestly.Rather, the analysis is based on the self-reported scores by the participants.If Yik indicates if a respondent i reported to win in round k, we can denote Yi = Σ K k = 1 1Yik as the total number of times that participant i guessed correctly.Following Barfort et al. (2019), we assume that participant i will be dishonest in some rounds.Furthermore, we assume that in case participants do not report honestly, they will do so in their own favour; that is, reporting a win instead of a loss.To test if the cheat rate differs across sectors, we use the dice throws to create an individual-level cheat rate.In a generalized and unbiased form, the estimated cheat rate can be reordered and written as follows: 3 Public service motivation.Our independent variable is public service motivation.We used Kim et al.'s (2013) 16-item validated measure of public service motivation.The reliability of the unidimensional public service motivation measure was 0.91, which is excellent.Since public service motivation is known to stimulate the intention to act ethically (Mayer-Sahling, 2018; Van Roekel and Schott, 2022), (un)ethical behaviour is measured first by using the dice game.Only then are participants presented with the survey questions.
Sector.In the second hypothesis, we test whether the sector (public vs private) has a moderating role in the relationship between public service motivation and cheating behaviour.
Sector was measured by a single-item question: 'Which of the following best describes the sector you primarily work in?' with either public or private as answering categories.
Risk Taking Index.As cheating involves a number of risks, such as the risk of being caught, we measured risk propensity using the standardized 6-item Risk Taking Index (RTI) (Nicholson et al., 2005).The RTI aims to measure risk-taking tendencies in multiple domains such as recreational risks, health risks, career risks, financial risks, safety risks and social risks.The reliability of the RTI measure was 0.65, which is rather low.
Gender and age.Both gender and age were used as statistical controls.

Results
In this study, we examine the relationship between public service motivation and cheating in a sample of 576 individuals.In line with earlier studies, the cheat rate in our sample is 0.345 (Gans-Morse et al., 2022;Olsen et al., 2019;Wright and Christensen, 2021).Our participants scored an average of 3.98 (SD = 0.53) on the public service motivation measurement (with a scale ranging from 1 to 5).Before conducting statistical analyses to test our hypothesis, we checked the sample for straightliners and 'speeders'.No participants were removed from the sample.The descriptives and correlations of all variables are presented in Table 1 (see online supplemental material).
Based on an independent t-test, we conclude that there are no statistically significant differences between the public (M = 0.354, SD = 0.114) and private (M = 0.359, SD = 0.110) sectors when it comes to the cheat rate t(472.86)= −0.56,p = 0.57 (Figure 1, see online supplemental material).
Our first hypothesis is: public service motivation is negatively associated with cheating behaviour.We used an OLS-regression to test the effect of public service motivation on the cheat rate (Table 2, Model I, see online).Against our expectations, public service motivation does not influence cheating behaviour in a dice game, f (1519) = 0.9409, p = 0.332, R 2 = 0.01.
The second hypothesis is: H2: The relationship between public service motivation and cheating is moderated by the sector; for employees working in the public sector the relationship is stronger than for employees working in the private sector.This hypothesis is tested using the PROCESS macro for R (Hayes, 2017).The results show that there is no interaction effect present between sector (public/private) and public service motivation in our data, f(517) = 0.0012, p = 0.912 (see the table in the online supplemental material, Model II).Moreover, the addition of controls does not alter the model.Hence, we reject hypothesis 2. The relationship between public service motivation and cheating behaviour is not moderated by sector.

Discussion and conclusion
The main aim of this study was to empirically investigate whether the sector in which individuals are working helps explain the mixed findings between public service motivation and a specific type of unethical behaviour, behavioural dishonesty (i.e.cheating).We find neither support for public service motivation to significantly reduce behavioural dishonesty (H1), nor support for our hypothesis that the relationship between public service motivation and behavioural dishonesty is influenced by a fit with the working environment; that is, that the relationship between public service motivation and ethics is stronger for employees working in the public sector (H2).
Based on the effects found in previous work (Olsen et al., 2019), we did an a-priori power analysis.The results of this analysis indicated that a sample size of at least 312 participants was required to detect an effect (see also Method section).The sample size of our study is 576.We are therefore confident that the null finding regarding both hypotheses is not the result of a lack of statistical power.A possible explanation for the null finding regarding H2 is that working in the public sector presents no guarantee that public service-motivated individuals can fulfil their public service motives (Wright and Pandey, 2008).In other words, public sector employees in our sample may simply not have experienced a fit with their working context in terms of public service motivation-related values.Future research should test our assumption that the private sector employees perceive a better fit with their working context empirically when analysing the relationship between public service motivation, behavioural dishonesty and sectoral differences.
Although our study primarily rests on null findings, we contribute to the literature in several ways.First, we contribute to the generalizability of the findings by Gans-Morse et al. (2022) and Wright and Christensen (2021) who studied the relationship between public service motivation and behavioural dishonesty among university students.We, in contrast, studied the very same relationship among private and public sector workers, thereby including non-academics who also represent an essential proportion of public sector employees.Together, these findings suggest that Olsen et al.'s (2019) findings may be an outlier.This raises important questions regarding the suitability of the dice game as a proxy for unethical behaviour in the context of public service motivation research.Public service motivation is 'an individual's orientation to delivering services to people with a purpose to do good for others and society' (Perry and Hondeghem, 2008: vii).Cheating in a dice game, however, does not provide a threat to society.There are no victims involved as there are in real life.Building on this point, cheating behaviour is likely to vary if the cheater can identify those who suffer from the consequences of their cheating.Moreover, in the dice game, the decision to cheat may vary with the outcomes of a stochastic variable (the pips of a die) which presupposes an opportunistic view on cheating behaviour.That is, cheating of individuals is conditional on how attractive the reward is.Literature on cheating behaviour of students shows that factors other than the opportunity to cheat play an important role, such as expectations of success and the extent of the reward (Whitley, 1998).Further research could focus on how variations of the stakes affect cheating.
Second, we indirectly contribute to the literature on the distinctiveness of public servants.We expect more from public servants when it comes to morality (Rainey, 2009: 86).In particular, we expect them to set aside their personal interest because they see it as their duty to serve the community.In line with this assumption, research consistently finds higher levels of public service motivation among public servants if compared to private sector workers (e.g.Bélle and Ongaro, 2014;Lyons et al., 2006).However, do public servants actually behave more ethically than their counterparts working in the private sector?As far as we know, empirical evidence on this question is very limited (for an exception, see the recent work by Sulitzeanu-Kenan et al., 2022).Moreover, daily news uncovers major and minor acts of dishonesty by public servants (Belle and Cantarelli, 2017), and large-scale survey evidence shows that unethical behaviour is common in developing as well as developed countries across the world (Meyer-Sahling et al., 2019).In this study, we find no substantial differences in behavioural dishonesty between public and private sector workers.This questions the general assumption of the moral superiority of public servants.
As with any study, our study is not without some limitations.First, we used a crosssectional survey design to capture the effect of public service motivation on cheating behaviour.Surveys are efficient and reliable in sampling a large group of respondents.In our case, we sampled 576 public and private sector managers.It is, however, impossible to draw causal inferences based on our data.Our survey design to test cheating behaviours does, however, mimic earlier work of others (e.g.Gans-Morse et al., 2022;Wright and Christensen, 2021).The choice for a survey-based dice game may be a two-edged sword, meaning that it improves measurement invariance across studies.At the same time, it may also provoke the risk of biased findings because of the survey method.Moving forward, alternative and more diverse methods of studying cheating behaviour in different domains of professional behaviour are much needed.
Second, the dice game has some important limitations.As mentioned, the dice game does not fully capture the true nature of cheating behaviour beyond its own setting.Moreover, the dice game is a low-stakes game, whereas behaviour scholars are interested in large(r)-scale dishonesty or corruption.The participants in our dice task certainly cheatedand to a comparable level to those in earlier studies.However, we did not test the assumption that participants may consider the financial externalities of their lying behaviour during the experiment and hence refrain from lying empirically.(i.e.Erat and Gneezy, 2012;Jiang, 2013).We hope that future research will overcome this limitation by making this mechanism more obvious to the participants.
The third limitation is a conceptual one.Following previous research, we focused on only one small aspect of unethical behaviour, namely cheating (Gans-Morse et al., 2022;Olsen et al., 2019).However, unethical behaviour is more than that (Lasthuizen et al., 2011;Treviño et al., 2014).Future research should therefore study the relationship between public service motivation and other aspects of unethical behaviour.It may be interesting to focus on more severe and clearer types of unethical behaviour.What counts as unethical behaviour is not always self-evident, particularly when it comes to borderline cases of ethical misconduct (Chubin, 1985).People may, for example, cheat for the benefit of others (rather than for their own benefit).This example demonstrates that cheating is a very 'grey form' of unethical behaviour as opposed to corruption, fraud and discrimination.This raises the question of whether we did not find a relationship between public service motivation and cheating simply because cheating is not 'unethical enough'.We hope that research will study the relationship between public service motivation and more severe forms of unethical behaviour in the future.
To conclude: we provide additional evidence on the relationship between behavioural dishonesty (i.e.cheating) measured by an incentivized dice game.Whereas previous studies relied on data collected among students, we used data from actual employees and tested the assumption that the sector moderates the public service motivation-unethical behaviour relationship.We find neither support for the hypothesis that highly public service-motivated individuals cheat less nor for the argument that this relationship is stronger for individuals working in the public sectors, while assuming a person-environment fit for these individuals.Our findings challenge the common assumption that public servants act morally superior to private sector workers.In addition, our findings suggest that the dice game may not be suitable to capture unethical behaviour relevant to highly public service-motivated individuals regardless of the sector people are working in.An important challenge for future research therefore is to develop measurement instruments that capture cheating or dishonest behaviour reliably and in a valid way.One promising way forward is to study the relationship between public service motivation and real-world (un)ethical behaviour that more strongly incorporates the societal aspects.