Keywords

1 Introduction

Corruption is widely considered to be one of the greatest imminent threats globally to democracy, economic wealth, and well-being (Poznan Declaration, 2014; United Nations Office of Drugs and Crime, 2004). Corruption refers to dishonest or illegal conduct by those in power (Merriam-Webster, 2018) and costs people their lives, their freedom and health, and traps millions in poverty (Transparency International, 2016). In this chapter corruption is defined as the “abuse of publicly entrusted power for private gain”, a definition adopted by the international anti-corruption agency, Transparency International (2016). Research suggests that if we identify where corruption blocks good governance we may be able to push back its corrosive effects (e.g., Clausen, Kraay, & Nyiri, 2011; De Graaf, 2007; Tavits, 2008). We propose that corruption may undermine ethical leadership in organisations. Ethical leadership is inspired by utilitarian consequentialism (as opposed to egoistical consequentialism), and is necessary for effective occupational safety and health (OSH) management at work, particularly to create a corporate climate of concern for worker psychological health and well-being (or Psychosocial Safety Climate, PSC). Ethical leadership is characterised by honest, trustworthy, principled, fair and balanced decision making, and care for people and the broader society (Brown & Treviño, 2006, p. 579). The purpose of this research is to provide insights for theory development and to build multilevel evidence about how corrupt values in society may infiltrate workplace governance in relation to the health and safety of employees, and affect work conditions and worker health.

Establishing, maintaining and growing an organisation requires on-going interaction with the nation’s public institutions. Bribery, kickbacks, nepotism, cronyism, and enacting regulations unfairly, are everyday examples of corruption in public institutions (Poznan Declaration, 2014). In society, every transaction between public officers and organisations is a point where corruption may take place. Kickbacks may be paid to local government officials for awarding building or service contracts. In relation to worker health, bribes may be paid to government work health and safety officials to stave off prosecution for occupational safety and health breaches. Corruption may occur in isolated instancesFootnote 1 but in aggregate, for a nation, may represent a corruption or erosion of the social values of care and concern for others.

Crucially, research shows that corruption and the ethical quality of public institutions (a proxy for corruption, i.e., Job, 2005) impact on national economic performance such as per capita income or gross domestic product (GDP) and on national well-being (Holmberg & Rothstein, 2011; Tavits, 2008). Worker health and well-being is a subset of national well-being, and workers are the main cog in workplace productivity and the national economy. Given the profound effects of societal corruption and because governments, public policy, and organisations work together for the benefit of citizenry (Okulicz-Kozaryn, Holmes, & Avery, 2014), it is important to consider how a corruption of public values in public institutions infiltrates other work organisations to affect the corporate climate, work conditions and the health of the workers within those organisations.

We propose a potential path of influence, that how the corruption of public values influences work conditions, health and well-being, is through organisational governance, and the cultivation of a climate for worker psychological health, referred to as PSC. The kind of ethical stance required for PSC and effective and humane governance at work, we argue, is in opposition to that driving corrupt behaviours. We draw on ethical leadership theory (Brown & Treviño, 2006; Brown, Treviño, & Harrison, 2005) and define ethical leadership as “the demonstration of normatively appropriate conduct through personal actions and interpersonal relationships, and the promotion of such conduct to followers through two-way communication, reinforcement, and decision-making” (Brown et al., 2005, p. 120). Many workers of the world operate in a capitalist economy, within workplaces where there are managers and followers. Since what managers do has consequences for others, we draw on philosophical principles of consequentialism to discuss the morality required to be an ethical leader, because manager actions have consequences for how work is designed, whether it is healthy or too demanding, and the nature of social relations, whether they are supportive or harassing. Senior managers of occupational safety and health within organisations hold considerable social power to set the tone for psychological health, referred to as PSC, by determining what policies, practices, and procedures are enacted. PSC refers to a particular organisational climate that arises through policies, practices and procedures that address worker psychological health and safety (Dollard & Bakker, 2010). Our study explores the potential general influence of a corruption of public values on (ethical) PSC practices (not explicitly corrupt acts) of the most senior managers of occupational safety and health across nearly 19,000 organisations. We use data linkage, and large social survey databases from 31 European countries to explore these relationships and their cross-level effects on worker reports of work conditions, health and well-being. PSC at the country-level indicates the responses from the most senior OSH manager across these organisations aggregated to the country level.

We particularly focus on psychosocial working conditions within organisation, because of the rising rates of work stress, bullying, and violence in the workplace and the challenge for policymakers, stakeholders, and organisations to determine the root causes of psychosocial risks in order to manage them (Stolk, Staetsky, Hassan, & Kim, 2012). As far as is known the empirical link between a corruption of public values and work stress (used generally here to refer to the link between psychosocial work conditions and poor worker health) has not yet been examined. We theorise that corruption of public values gives rise to PSC (itself a consequence of ethical leadership) that in turn engenders particular working conditions. Our research is innovative in this respect as although the management literature has explored how corruption builds within organisations (Ashforth & Anand, 2003; Kammeyer-Mueller, Simon, & Rich, 2012), it is largely silent on the issue of how corrupt societal values infiltrate organisations and affect organisational leaders (our focus), work conditions, health and well-being. At a national level previous research showed workplace protection from unions (union density) was positively related to PSC that in turn correlated with worker health and gross domestic product (Dollard & Neser, 2013). We extend this research by considering if corruption of social values (evident by corruption as a proxy) is also precursor to PSC, and consider its effect on work conditions, health and well-being.Footnote 2

1.1 Psychosocial Safety Climate and Ethical Leadership

Psychosocial Safety Climate reflects management values and concern for worker psychological health and well-being, and is largely influenced by senior management (Dollard & Bakker, 2010). Psychosocial safety climate is developed through a leadership style that supports workers’ basic psychological needs (Williams et al., 2014), such as the need for autonomy, competence, and relatedness (Deci & Ryan, 2000). Work stressors, such as high demands, low control, and low support, threaten the fulfilment of these needs. Therefore to protect worker health, managers need to consider how work is designed and how the social-relational aspects are managed in order to achieve both organisational and worker health objectives. Since there are competing interests at work, such as productivity imperatives versus worker health, leaders are guided by ethics and values when making decisions. Peak institutions such as the World Health Organization (2010) hold that ethics and values are at the heart of decisions to create healthy workplaces.

Ethical leadership refers to leading in a manner that respects human rights and dignity (Ciulla, 2004), and concerns how leaders use their social power in decisions and actions (Resick, Hanges, Dickson, & Mitchelson, 2006). Ethics concerns what is “morally good and bad, right and wrong” (Singer, 1985). There are different philosophical positions on ethical leadership. Consequentialism is an approach that relates the value of an action to the value of its consequences (Burnes & By, 2012). Scholars have argued that a utilitarian analysis is useful within organisations because leaders are largely judged by the consequences of their actions rather than their intentions (Burnes & By, 2012). Utilitarian ethics derives from Jeremy Bentham’s principle of utility, which refers to the value of actions for the greater good, producing the greatest happiness for the greatest number (Burnes & By, 2012, p. 245). By contrast, individual (egoistic) consequentialism that derives from Thomas Hobbes, views that actions are ethically right if they maximise benefits to the actor. Under utilitarian consequentialism, the leader acts in the best interests of the majority whereas under egoistic consequentialism, as in the case of corruption, the leader acts for self-interest (Burnes & By, 2012). We refer to this self-interest as an example of a corruption of public (pro-social) values.

As a guiding philosophy of ethical leadership, utilitarian consequentialism is crucial in the role of senior occupational safety and health leadership, since their major goal is to pursue human rights and dignity for the majority of workers within a health and safety framework. In a high PSC organisation we expect strong adherence to utilitarianism ethical principles at the highest echelons of the organisation. In other words ethical leadership through utilitarian consequentialism, creating the best circumstances for most should create a good PSC.

In this study we considered PSC in terms of utilitarian consequentialism (organisational intervention in the form of prevention yields the greatest effect) and operationalised PSC in terms of the extent to which the most senior organisation occupational safety and health managers: (a) implemented workplace procedures for psychosocial risks, and (b) involved workers in taking measures to deal with psychosocial risk factors—since worker involvement is an important requirement for an effective occupational safety and health management system (Stolk et al., 2012).

1.2 A Corruption of Societal Values Related to PSC

The everyday assumption that power itself corrupts does not have a strong empirical basis (Sturm & Antonakis, 2015). Rather, the cultural, social, political economy, and legal origins of countries seem important (e.g., LaPorta, Lopez-De-Silanes, & Shleifer, 2008). Within societies, corruption over time becomes normalised and built into political regimes (Cini & Drapalova, 2013). The ethics of corrupt behaviour is egoistic consequentialism; it is a misuse of power for private or personal gain rather than for the majority interests. We argue that corruption in public institutions opposes or undermines the utilitarian ethical leadership of organisational occupational safety and health including the development of PSC. We expect that powerful societal factors endorse or enable corrupt or ethical occupational safety and health leadership that can lead to a strong PSC.

Prilleltensky (2000) and colleagues are critical of current approaches to ethics in organisational leadership (e.g., Block, 1993; Covey, 1989) because they fail to address the “fundamental contradictions between profits and values”, glossing over the conflict between managers and workers in capitalist societies. They make the case that values are held in social contexts and that when matters of ethical despair emerge, tensions will arise among values, interests, and power. We therefore, expect that societal factors endorse or enable corrupt or ethical occupational safety and health leadership in aggregate (i.e., utilitarian consequentialism). The contradiction between profits and values is particularly salient for occupational safety and health leadership in their interaction with the external world and within their organisations; at once they are concerned with organisational goals and the health and safety of workers. Maintaining safety may compromise profits. Figure 3.1 portrays a number of societal influences on the corruption of public values and PSC, outlined next.

Fig. 3.1
figure 1

Effects of corruption of public values at work (Between—country effects)

Full size image

Externally, powerful social actors such as corporations, public institutions and worker unions interact to determine prevailing labour market policies (i.e., labour regulations, industrial relations) and welfare state and policies (i.e., social policies) that affect work conditions, worker health and well-being (Benach, Muntaner, & Santana, 2007). National policies help to shape the quality of work (Dragano, Siegrist, & Wahrendorf, 2011), such as hours of work, competition policies, and how much job control one has. National legislation and guidance for managing psychosocial risks are important for work-related psychosocial risk management (Leka, Jain, Zwetsloot, & Cox, 2010). However in a corrupt society where self-interest trumps concern for others, we expect on average that work-related policy development and enforcement for worker health to be less than in other societies, and the enactment of work-related regulations and policies to be weaker (e.g., in some public services bribes may be paid to office bearers to turn a blind eye to infringements). Where corrupt public institutions prevail, trust in public institutions would be low.

Welfare state regimes typify the role of the state in social and economic well-being, and are classified by Epsing-Anderson (1990) as social democratic, conservative and liberal—and they reflect and instil guiding values and principles of the society (e.g., equality of opportunity, equitable distribution of wealth and power). Countries characterised as social democratic regimes (see Epsing-Anderson, 1990) (e.g., Sweden, Denmark, Norway, and Finland), for instance, value egalitarianism, transparency, accountability, and participation in decision-making between social power actors, which would likely lead to less corruption than in other regimes (Resick et al., 2006).

National cultural dimensions also give rise to values that are also related to corruption; high power distance gives rise to a lack of trust between groups; high uncertainty avoidance gives rise to a preference for bureaucratic structures that tend to encourage unethical behaviour; low individualism or collectivist societies have multiple competing values with no single standard (Davis & Ruhe, 2003). But societal values also influence expectations of the ethical dimensions required for organisational leadership (Resick et al., 2011). Researchers found that country level endorsement of ethical dimensions for leadership (e.g., character/integrity, altruism, collective motivation, and encouragement) closely mapped to the Corruption Perception Index (Transparency International, 2011) with the highest levels of endorsement for ethical leadership in Nordic European societies (Resick et al., 2006).

Within the organisation itself, an occupational safety and health leader’s power is vulnerable because of potential value conflicts with other interests in the organisation that may favour profits (e.g., CEO aims to maximise profits for shareholders; personal interests such as CEO salaries influence ethical decisions) over worker health leading to conflicts, that may be overt or covert, between functional areas of the organisation (e.g., the business and finance vs. human resource vs. occupational safety and health branches). To retain their own position, power and privileges at the corporate level (in the upper echelons), occupational safety and health leaders may reduce their efforts that would otherwise be in the interests of workers (such as to make pro-worker health changes in the organisation). Over time, corrupt values and practices may become normalised (Ashforth & Anand, 2003). For instance the theory of planned behaviour suggests that normative influences or “the way things are done” including corrupt behaviours, would influence subjective norms: if powerful significant others disapprove or do not prioritise tending to the psychological needs of others, the occupational safety and health leader would be less likely to implement PSC (Ajzen, 1991). Moreover corrupt practices are likely to be associated with procedural injustice, itself a recognised stressor (Siegrist, 1996); remedies by occupational safety and health leaders to address injustice and act against corruption may elicit threats in terms of their own job security and other rewards. In sum, in societies with corrupt public values we expect that the institutions and social conditions that support utilitarian ethical leadership that underpin PSC in organisations would be compromised. We propose:

Hypothesis 1

Corrupt public values (with societal corruption as a proxy) negatively relate to corporate PSC (see Fig. 3.1).

1.3 Psychosocial Safety Climate Related to Work Conditions, Health and Well-Being

Psychosocial safety climate theory is a multilevel perspective on work stress. Under PSC theory, understanding the driving values of management enables us to predict work conditions, the health and well-being of workers (Dollard & Bakker, 2010). An important challenge in understanding work stress is to correctly identify its root cause. According to the hierarchy of controls logic, identification of more distal causes, will yield a more effective control strategy (Dollard & McTernan, 2011). The dominant theories of work stress focus on job design causes and propose that work stress arises when high job demands are combined with low levels of control (i.e., Job Demand-Control theory, Karasek, 1979), and support (Job Demand-Control-Support (DCS) theory, Johnson & Hall, 1988).

Job demands are commonly operationalised in terms of qualitative demands, such as emotional demands (Karasek et al., 1998). For employees, job control refers to being able to control the method, order, and timing of job tasks, the skills used, being able to learn new things (skill discretion) and influence decisions (decision authority) (Karasek, 1979). Social support “refers to overall levels of helpful social interactions available on the job from both co-workers and supervisors” (Karasek & Theorell, 1990, p. 69). In the current study we operationalised demands, control and support accordingly.

Psychosocial Safety Climate theory proposes that PSC is a precursor to these job design “causes” of work stress and therefore to DCS theory. In high PSC organisations, with the psychological health of workers in mind, the upper echelon managers (potentially influenced by occupational health and safety managers) work to meet the basic psychological needs or employees, such as by bolstering job control and social support, and making job demands manageable.

There is empirical evidence for the role of PSC as a “cause of the causes”. Researchers found that PSC reported by one group of nurses predicted emotional demands, job control and supervisor support reported by other nurses working in the same work unit 24 months later (Dollard et al., 2012). We propose:

Hypothesis 2

PSC negatively relates to emotional demands and social support and positively relates to job control.

Fundamentally the theory of PSC concerns a relationship between PSC and worker health and well-being; via management political will, in high PSC contexts, quality job design and policies, practices and procedures will increase worker health and well-being. Moreover social exchange theory predicts that if workers perceive that management show concern towards them, in exchange, employees will reciprocate with increased feelings of satisfaction with the job (Blau, 1964). Evidence shows that PSC predicts work engagement which like job satisfaction reflects positive affect towards work (Idris, Dollard, & Tuckey, 2015).

Hypothesis 3

PSC positively relates to worker health and job satisfaction.

1.4 Corrupt Public Values, PSC, Work Conditions and Worker Health and Well-Being

Connecting the political, social and economic literatures with management approaches, conceptually, we anticipate flow-on effects of societal corruption to worker health and well-being. Political and economic research has linked societal corruption to distal [e.g., happiness (Hudson, 2006), life satisfaction (Helliwell, 2003; Tay, Herian, & Diener, 2014)] but not to focal measures of worker well-being, such as worker health and job satisfaction (Fisher, 2010). Yet worker health and job satisfaction are crucial for work performance (Judge, Thoresen, Bono, & Patten, 2001), and national productivity.

We argue that potential gate keepers of the process linking corrupt public values to work conditions, worker health and well-being, are the organisation’s most senior occupational health and safety managers and their enactment of PSC (there are likely others too such as CEOs). In this chapter we investigate whether corruption is a “cause of the cause of the causes” of worker health and well-being through the actions of the most senior occupational safety and health managers.

Researchers have anticipated multilevel (Ashkanasy, 2011)—external, organisational (Dollard, Osborne, & Manning, 2013), and individual (Judge, Heller, & Mount, 2002)—causes of worker health and well-being. Increasingly researchers have investigated external factors such as economic (i.e., GDP) factors and have found that these predicted job satisfaction ahead of company (i.e., career prospects) and personal (i.e., depression) factors, after ruling out other macro factors (i.e., social progress) (Augner, 2015). Tay and Harter (2013) found that economic (i.e., GDP) and labor market (i.e., lower unemployment rates) forces influenced job satisfaction. Prior research has empirically connected national level union density (an external factor that likely improves national labour and social policy) to PSC and in turn worker health (Dollard & Neser, 2013); here we investigate additional non-economic sources, such as whether corrupt public values operate over and above these influences. We argue that PSC arises from an ethical and values based leadership position and promotes quality work conditions and worker health and well-being. In support of this logic, researchers found that ethical leadership was associated with employee psychological safety and the exercise of voice, considered important for improving workplace conditions, worker health and productivity (Walumbwa & Schaubroeck, 2009).

Hypothesis 4

Corrupt public values indirectly relate to work conditions via PSC, and,

Hypothesis 5

Corrupt public values negatively relate to worker health and wellbeing via PSC.

2 Method

2.1 Study Databases

We used secondary data and combined multiple data sets (as identified below) to assess the theoretical framework. Matching data bases was possible at the country level. There were 31 European countries across the matched data sets; 27 EU Member States (Austria; Belgium; Bulgaria; Cyprus; Czech Republic; Denmark; Estonia; Finland; France; Germany; Greece; Hungary; Ireland; Italy; Latvia; Lithuania; Luxembourg; Malta; Netherlands; Poland; Portugal; Romania; Slovakia; Slovenia; Spain; Sweden; UK; Switzerland) plus Croatia, Turkey, and Norway. Ethics approval was not sought beyond what is already documented by the data owners. The data were de-identified.

2.2 Measures

Corrupt Public Values. As a proxy we used the Corruption Perception Index (CPI) 2009 (Transparency International, 2011). We reversed the 10-point scale to 0 = very clean, 10 = highly corrupt. The CPI is considered the most comprehensive international indicator of corruption and has been validated in previous research (DiRienzo, Das, Cort, & Burbridge, 2007). The index comprises data gathered across business people and risk analysts and assesses the degree of corruption among public officials and politicians, including committing offenses, embezzling public funds, and accepting bribes in public procurement (Lancaster & Montinola, 1997).

Psychosocial Safety Climate was measured using data from the European Survey on New and Emerging Risks—Psychosocial Risks (ESENER) 2009. The most senior occupational safety and health manager from 18,782 organisations was interviewed. This measure comprised five questions with a definition read out if necessary: “Does your establishment have a procedure to deal with”: “work-related stress?” (MM 250); “bullying or harassment?” (MM 251); “work-related violence?” (MM 252); “What about the role of employees: Have they been consulted regarding measures to deal with psychosocial risks?” (MM 266); and “Are employees encouraged to participate actively in the implementation and evaluation of the measures?” (MM 267), (reversed to No = 0, Yes = 1, not an issue or don’t know = system missing).

Using our own data sets we found that this five item measure correlated with the psychometrically valid PSC-12 scale (Hall, Dollard, & Coward, 2010), cross-sectionally, r = .45, p < .01, n = 175 and r = .47, p < .01, n = 1037, and longitudinally r = .30, p < .001, n = 1043, and r = .36, p < .001, n = 175. The test retest was .52, p < .001, n = 175. In this study the PSC reliability was .74.

Individual Level Factors, work conditions, worker health and well-being were assessed using the European Working Conditions Survey (EWCS) 2010—employee sample. Data were from 35,187 employees interviewed face-to-face at home. Participants were aged 15 or older and in employment (worked for pay or profit for at least an hour in the week preceding the interview). We excluded participants who were not in employed work (e.g., self-employed) (q6 = 3) or did not work as an employee (hh2d = 1), yielding 32,203 employees for the study.

Emotional Demands were assessed with two items, “Your job requires you to hide your feelings” (q51p) and “You get emotionally involved in your work” (q51 m), with responses reverse scored to 1 = never, 5 = always, and adjusted for items; high scores reflect high demands (intercorrelation was r = .21, p < .01).

Job Control was assessed using a focused, multi-faceted measure, canvasing decision latitude and skill discretion, but not job complexity (see De Jonge, Dollard, Dormann, Le Blanc, & Houtman, 2000) using ten items, e.g.: “Does your job involve learning new things?” (q49f); “Are you able to choose or change your methods of work?” (q50b); and, “You can influence decisions that are important for your work” (q50o). Response formats varied, and were reverse-scored to 1 = never, 5 = always, or 1 = no, 2 = yes, with high scores indicating high control. As response formats varied, we standardised the items before adding them together. Alpha was .81.

Social Support was assessed with two items, “Your colleagues help and support you” (q51a) and “Your manager helps and supports you” (q51b) with responses reverse-scored, to 1 = never, 5 = always and adjusted for number of items; higher scores indicate higher support (inter-correlation was r = .55, p < .01).

Work Affects Health Positively was measured with a single item, “Does work affect your health or not?”; responses were recoded to 0 = mainly negatively, 1 = no, 2 = yes mainly positively.

Subjective General Health was used to validate the work affects health measure. It was assessed with a single item, “How is your health in general?”, with scores on a five-point scale, reversed to 1 = very bad, to 5 = very good.

Job Satisfaction was assessed using a global job satisfaction item, “On the whole (how satisfied are you) with working conditions in your main paid job?” (q68). Items were reverse-scored to 1 = not at all satisfied, 2 = not very satisfied, 3 = satisfied, 4 = very satisfied. A global index of job satisfaction (single-item measure) is a valid measure of general job satisfaction (Wanous, Reichers, & Hudy, 1997).

  • Level 1 Controls

Demographics. We assessed gender (1 = male, 2 = female); age in years; highest level of education coded according to the ISCED classification (0 = pre-primary education to 6 = second stage of tertiary education); and hours of work usually worked in the main job. We used education as a proxy for socioeconomic status, because around one-third of the data was missing for the income measure. These demographics were used as controls at the individual level because they are related to well-being (e.g., Clark, 1997).

  • Level 2 Controls

Union Density refers to the percentage of the workforce that are union members; national levels were assessed using combined data from the OECD (2009), Hall-Jones (2017), and the Eurofound (2011) for Croatia.

Gross Domestic Product (GDP 2009). This was assessed using total purchasing power parity (PPP) converted to GDP per capita (CIA World Fact Book, 2011). For all 31 countries, total PPP was converted to GDP per capita greater than international $10,000. We included national wealth (i.e. GDP) as a control variable as it relates to well-being (Augner, 2015).

Income Inequality (GINI). The GINI coefficient assesses income inequality with values from 0 = equality to 1 = inequality. It is regularly used in sociological and economic research. We used the GINI from the period 2005–2009 (CIA World Fact Book, 2011). We included it as a control since it is related to well-being (Clark, Frijters, & Shields, 2008).

Human Development Index 2010 is a United Nations summary measure of achievement across three dimensions; a long and healthy life, a decent standard of living, and education (United Nations Development Program, 2015).

Organisational Size (EWCS 2010) was determined by the question, how many people work at your workplace and ranged from 1 (works alone) to 8 (500 and over).

  • Level 1 Participants (EWCS 2010).

Participants were 35,187 employees, 54.5% male and 45.5% female, aged 15–91 (M = 39.96, SD = 11.41). Average weekly hours usually worked in the main paid job was 38 (SD = 10.71); education levels ranged from pre-primary education to the second stage of tertiary education, with most reporting upper secondary education (40%). Monthly net earnings ranged from 22,000 to 27,000 Euro (missing n = 9430).

3 Statistical Analyses

We used hierarchical linear modelling (HLM) and HLM 7.00 software (Raudenbush, Bryk, Cheong, Congdon, & du Toit, 2007) for hypothesis testing since our data were nested, individuals within countries (N = 31). We created two data files for HLM. The Level 2 data file comprised variables from several data sets matched at the country level (e.g., corruption, PSC (aggregated), GDP, GINI). There were two Level 1 data files; for Hypothesis 1 to assess PSC we used ESENER data at the individual level and we used the employer weight (emp_wei2) in analyses; for all other hypotheses, the Level 1 file was from the EWCS and data were weighted for design, stratification and population (w5_all_e) in the HLM analysis. The two data files were linked via the country variable. The measures GDP, GINI, and population were log transformed at Level 2 to correct for skew.

We ran preliminary baseline random coefficient models to establish whether there was sufficient between-country variance in the measures to warrant prediction (Intra-class coefficients ICC1 are reported below). All hypotheses were assessed with the predictor at the country level, and the dependent measure at the individual level. Following Zhang, Zyphur, and Preacher (2009), we grand-mean centred the variables at Level 2, and we group-mean centred variables at Level 1 (gender was not centred). Each path was assessed with the control measures in the model.

To test the mediation hypotheses, all paths of the mediation hypotheses required the assessment of between-country effects. In multilevel mediation, Zhang et al. (2009) argue that it is important to consider the variance created by between-groups (i.e., country) and within-group effects (i.e., individual influences) in the criterion measure. For mediation, we assessed the components of the mediated paths in two steps, first, path a (X, antecedent · M, mediator) and second, path b (M · Y, dependent; with X in the model). Note that each Level 2 · Level 1 relationship is a between-country effect. For example to test Hypothesis 5, that corruption relates to worker health via PSC; path a was tested regressing Level 1 M (PSC) on Level 2 X (corruption), controlling for Level 2 controls; path b regressed Level 1 Y (worker health) on Level 2 M (PSC), with Level 1 and Level 2 controls included (see measures). For all paths the strategy was to include controls but these were removed when not significant to increase the degrees of freedom. A significant relationship between X and Y is not required for mediation analysis, particularly in this study when the antecedent, corruption, is distal from the dependent variable (Shrout & Bolger, 2002).

Finally, we used the Monte Carlo Method for Assessing Mediation (MCMAM), a parametric bootstrapping approach thought to be superior to Sobel tests (MacKinnon, Lockwood, & Williams, 2004), to test the significance of the between-groups mediation effect using confidence intervals, because the indirect effect usually has a non-normal distribution (Bauer, Preacher, & Gil, 2006; Selig & Preacher, 2008). We used 20,000 repetitions and a 95% confidence interval to determine the upper and lower levels of the interval; the mediation is significant if the confidence interval does not contain zero.

4 Results

4.1 Descriptive Statistics

Means, standard deviations, and inter-correlations of variables are shown in Table 3.1. At the country level (below the diagonal), corruption was negatively related to union density, PSC, job control, and the health and well-being measures (see Fig. 3.1). At the individual level, the work conditions, health and job satisfaction variables were related as expected.

Table 3.1 Means, standard deviations, and inter-correlations
Full size table

We turn now to multilevel hypothesis testing. We first assessed the baseline random coefficients model (null model) to assess the ICCs and found between-country effects explained sufficient variance to warrant HLM modelling; the percentage of variance due to country factors for PSC was 24.4% which provided a rationale for aggregating it to the country level. The ICC was 5% for emotional demands, 7.8% for job control, 6.6% for social support, 3.5% for work positively affects health, and 7.5% for subjective general health, and 8% for job satisfaction, all providing support for the notion that there was sufficient between-country variance to explain.

Hypothesis 1 proposed that corruption negatively relates to PSC. We found that corruption alone accounted for 17% of the country variance in PSC reported by managers. We entered the control variables from Table 3.1 that showed significant relations to PSC. Corruption was significant: B = −.27, SE = .06, t = −4.28, p = .001 (see Table 3.2 and Fig. 3.1 reporting t-values). Hypothesis 1 was supported.

Table 3.2 HLM random intercept model for psychosocial safety climate; Path a
Full size table

Hypothesis 2 proposed that PSC relates, negatively to emotional demands, positively to job control, and positively to social support. Hypothesis 2 was supported for all work conditions. For example, PSC was significantly negatively related to emotional demands, B = −.15, SE = .06, p = .03 (see Table 3.3).

Table 3.3 HLM random intercept and slope models for outcome measures; Path b
Full size table

Hypothesis 3 proposed that PSC is positively related to well-being and health. This was supported for job satisfaction, and the health measures (work positive health and subjective general health) (see Table 3.3).

Hypothesis 4 proposed that corruption is indirectly related to work conditions through PSC. Path a was determined as for Hypothesis 1. The indirect effect of corruption on emotional demands via PSC was significant: 95% CI [0.01, 0.08] (see Table 3.3, final column). It was also significant for social support, but was not supported for job control, since only corruption was significant as a main effect in the final model. Hypothesis 4 was mostly supported.

Hypothesis 5 proposed that corruption is indirectly related to worker well-being and health through PSC. As shown in Table 3.3, for path b, PSC was significantly related to the health measures even with corruption in the model. The indirect effect of corruption on worker health was significant; for example, for work affects health positively, 95% CI [−.03, −.01]. For job satisfaction, union density persisted as a main effect. Hypothesis 5 was supported only in relation to the health measures.

5 Discussion

The research is timely as it responds to urgent calls to root out corruption in society (United Nations, 2018). Our study investigated the corruption of public values because of its profound social, political, economic, health, and ethical consequences. First, we extended investigations of societal corruption in public institutions, by exploring its impact in other organisations (both public and private), after accounting for several other national level variables. Second, we considered ethical leadership as the theoretical mechanism linking corruption of public values to PSC; through the principle of utilitarian consequentialism, ethical leadership results in high PSC. Third, the investigation extended thinking in this area and considered societal factors that impinge on workplace leadership, work climate and conditions, and worker well-being. And, fourth, the study involved comprehensive data linkage from multiple sources, including experts (corruption data), leaders (PSC), and workers (e.g., work conditions, health and well-being), and used multilevel analysis parsing national and individual effects.

There are three main findings from our research. First, our research showed that corruption of public values at the country level was associated with PSC. We found that over and above GDP, income inequality, population size, union density, the human development index, and organisational size, that corruption of public values (proxy to corruption or the quality of public institutions), was important for the consequential development of PSC in organisations. We demonstrated this in a cross-level analysis, with these national societal factors as antecedents to individual reports of PSC by the most senior occupational safety and health leaders representing nearly 19,000 organisations from 31 countries.

In this chapter the case was made that societal factors endorse or enable ethical occupational safety and health leadership. Since occupational safety and health systems are usually supported by public institutions, in a society with corrupt public values we expect that work-related policy development and enforcement and the enactment of work-related regulations and policies to be weaker. Corruption in public institutions is a form of egoistic behaviour that opposes or undermines ethical utilitarian leadership required for the enactment of PSC. Integrity, a core ingredient of ethical leadership involving the ability for morally correct behaviour despite external pressures (or lack of ethical standards) to act otherwise (Resick et al., 2006), may be under threat in corrupt societies. The role of occupational safety and health manager is vulnerable because of the “fundamental contradictions between profits and values” in the upper echelons of organisational decision making. Since organisations reflect external values and ethics, these contradictions within organisations may be more salient in corrupt societies, and in capitalist societies.

Second, we found that work conditions (emotional demands and social support) and worker health could be explained by variations in national levels of a corruption of public values through PSC. Third, corruption had a direct effect on worker job control, and union density and corruption together accounted for job satisfaction.

Since our analysis was cross-sectional we cannot rule out a complex of factors that have a common cause (social, cultural, legal), that similarly affect public values and PSC leadership. Although we cannot confirm a causal link between corruption of public values and PSC, at the very least we can be clear—ethical leadership actions taken inside the workplace to establish PSC covary with factors external to the organisation, both the corruption of public values and union density. Also we can not rule out reverse causation—for instance if workers are exploited in terms of their health and poor work conditions this could create corrupt actions from workers in the long run.

Our results showing that union density accounted for job satisfaction does not concur with Augner (2015) who found GDP to be the most important macroeconomic predictor of job satisfaction. Fundamentally our results support alternative views growing in society that it is important to consider national well-being and quality of life aspects in non-economic terms, in terms different to the GDP (Constanza et al., 2014). In supplementary analyses we also considered national political orientation (i.e., left–right leaning) (Okulicz-Kozaryn et al., 2014) and found that our results held even with these variables in the model.

5.1 Theoretical Implications

The theoretical development in this chapter concerning national level influences on PSC, work conditions and worker health was supported empirically and justifies future consideration of societal factors when explaining PSC, worker health and well-being. The proposition linking corruption in society to work conditions and worker health because of its effect on ethical leadership for PSC seems plausible and warrants future theory building in the foundational aspects of PSC.

We find reason to extend PSC theory to include macro social-contextual factors beyond union density, in this case societal corruption. National political power relations between power actors (corporations, institutions, and unions), along with national culture and welfare state regimes give rise to societal values and cultural practices (Gupta & Hanges, 2004). In turn, these values influence corruption and expectations of the ethical dimensions required of organisational leaders (Resick et al., 2006, 2011; Van Muijen & Koopman, 1994). By linking a corruption of public values empirically to PSC we establish the possibility that specific societal values such as transparency, accountability, participation in decision-making, and equitable distribution of power likely prevalent in non-corrupt societies give rise to the ethics required to establish PSC in organisations. Scholars highlight integrity, as a core ingredient of ethical leadership, that concerns the ability for morally correct behaviour despite external pressures to act otherwise (Resick et al., 2006); in corrupt societies integrity may be a point of resistance to prevailing social forces.

The specific work condition, job control, however, we found was better explained by the direct effects of corruption. In a corrupt society power is annexed by those who are corrupt yielding less and less power for others. Workers, at the lower rungs, therefore may feel a low sense of control over their work because of a lack of power in decision making processes in high corruption societies. Importantly this relationship can be seen in some additional analyses we undertook at the individual level in the European Social Survey (ESS) (2012) of 21,853 workers in paid employment from 23 countries. Corruption was assessed in the ESS using the proxy, trust in public institutions (parliament, legal system, police, politicians) (aggregated to the country level this was related r = −.92 to the Transparency International CPI). Control was assessed in terms of control over how daily work is organised, and whether employees were allowed to influence policy decisions about activities of the organisation. Higher levels of perceived trust in public institutions (lower corruption) was associated with higher levels of job control, r = .28, p < .001 (European Social Survey, 2012)

Taken together, a corruption of public values may infiltrate workplaces, firstly, as we predicted via influences on upper echelon processes, culminating in varying levels of PSC. Secondly corruption may influence workplaces via disempowering processes, culminating in lower job control; since control is a basic human need, low control may in turn lead to negative effects on health and well-being.

The results suggest that dominant theories of work stress, such as the Job Demands Control Support model should be contextualised to consider, beyond job design factors, work organisation, and external factors. Considerable variance in PSC (24%), the work environment, health and well-being measures exists due to national level factors. Here we explained some of this in terms of national levels of corruption and union density previously (Dollard & Neser, 2013).

Much theoretical work on distal well-being (e.g., life satisfaction and happiness) emerging from the fields of politics, economics, sociology and psychology already incorporates societal factors such as corruption in societal institutions, inequality, political orientation, and trust in society (De Graaf, 2007; Pinto, Leana, & Pil, 2008). The theoretical innovation here was to connect corrupt public values (i.e., corruption) to the work domain, to worker health and well-being (via PSC), which in turn could have implications for population worker health and productivity.

5.2 Practical Implications

We heeded the call from researchers and social commentators to identify the effects of corrupt public values in order to break its corrosive effects. In so doing, this research uncovered the workplace as a fruitful site for resisting corruption. Societal corruption likely affects ethical leadership in organisations (Resick, Mitchelson, Dickson, & Hanges, 2009). In this study we see that it affects occupational safety and health leadership for PSC, which ideally encompasses a worker-centred leadership ethos that values worker health and well-being, and rails against prioritising profits over worker well-being (Dollard & Bakker, 2010; Dollard & Karasek, 2010; Hall et al., 2010).

We discovered that whether or not the most senior occupational safety and health managers implemented PSC depended on prevailing levels of societal corruption; when corruption was high, leaders were less likely to implement PSC. We found that corrupt public values effects were evident in the health of workers and their work conditions. With this new knowledge policy makers can now consider how to make processes more transparent to ensure that interactions between organisations and public institutions are not corrupted by unethical behaviours. Workplaces can offer a site for resistance, because with this new evidence, management, occupational safety and health, and human resource personnel can work together to take action to address corruption within and between organisations. Although our study focused on the most senior occupational safety and health managers, we expect that PSC is transmitted within organisations via other power actors in organisations too including the CEO and senior management group and leadership at the work team or unit level. The limited capacity and competence of occupational safety and health personnel is well documented so customised ethical leadership development may be required to assist occupational safety and health personnel at all levels with implications for PSC (Frick, 2004; Mirza & Isha, 2017; Tuckey, Li, & Chen, 2017).

The Poznan Declaration (2014) posits that “it seems likely that dysfunctional governmental and anti-social market behaviours have their roots in the value systems of decision-makers at various levels” (p. 4). It proposes a societal approach for tackling corruption at an international level, currently endorsed by nearly 70 universities, that advocates teaching professionals in all university faculties how to think and act critically and ethically, and educating all professionals in the values of transparency, democracy, equality, legality, objectivity, integrity, freedom of opinion and information, and human-centred leadership. This comprehensive approach could be specifically championed and adopted by management departments worldwide to assist organisations and society to counteract corruption in society. Transparency International (2016) has also outlined a comprehensive plan to tackle corruption worldwide which is worthwhile considering.

5.3 Limitations and Future Research

Although variables follow a temporal sequence, with corruption, union density, and PSC assessed in 2009, work conditions and health in 2012, the data was lagged, not longitudinal, limiting our capacity to determine the direction of causation. As we have seen in prior studies, PSC is related to future specific corrupt practices such as bullying (Bond, Tuckey, & Dollard, 2010; Dollard, Dormann, Tuckey, & Escartin, 2017). Therefore a fruitful line of future research would be to look at the relations between corruption and leadership for PSC, and other unethical behaviours like bullying and harassment, within and between organisations and public institutions (Hutchinson, Vickers, Wilkes, & Jackson, 2009; Vickers, 2014).

Another limitation was the power of the study, restricted at the upper level, because we only had complete data from 31 countries. Since our upper level sample was small it did not make much sense to analyse long mediational paths such as Corruption  ⟶ PSC ⟶ working conditions ⟶  health. Several problems arose from using secondary data. First, operationalisation of the variables was limited to availability. Nevertheless for PSC we were able to show in ancillary analyses that the PSC scale was reliable, and related to a psychometrically valid measure of PSC. Second, the response formats in secondary data are predetermined. For the PSC items we also assessed whether the results varied when responses to PSC items such as “work stress is not an issue” were coded as “no”, and the results were virtually identical.

Our analysis enabled the explanation that societal corruption affects individual level work conditions and health via organisational PSC, where all paths were modelled  as between-country level effects. The level of analysis at which the antecedent is measured (in this case country) must be carried through in the mediation process (Zhang et al., 2009). Therefore the PSC measure had to be considered at the country level for path b in the mediation process. An issue arising is the ecological fallacy problem (Shively, 1969). The ecological fallacy is to infer that observations found at the national or population level apply at the organisational or individual level. As there may be great variability in the observations within the population, we cannot predict organisational or individual level results within countries based on knowing about the national levels. We cannot say, for instance, that all organisations in Sweden have high levels of PSC. We can only say that on average the levels of PSC are high in Sweden, and at the national level, this may have caused good worker health, on average. Nevertheless we expect some concordance between country levels of PSC and the levels of PSC within organisations.

Future research could present analyses at four levels: societal, organisational, job design (e.g., occupational categorisations), and individual. Developing a tool to assess national level legislation for psychological health and safety (Potter, O’Keeffe, Leka, Webber, & Dollard, 2019), and assessing the influence of the legislation directly on organisational PSC (see Potter, Dollard, Owen, O’Keeffe, Bailey, & Leka, 2017) would also assist in clarifying the impact of what the state does, rather than just assess the quality of state institutions.

More qualitative research is needed to untangle different scenarios about how corrupt behaviours influence leadership decision-making for worker health in organisations.

Our proxy for corrupt public values was corruption in the public sector; how corruption within corporate culture affects employee health and business decisions is important for future research (e.g., the American International Group collapse) (Ferrell & Fraedrich, 2014). Generalisability of results requires additional testing in other regions, for example, the Asia Pacific.

Finally without establishing measurement equivalence of PSC across countries (Vandenberg & Lance, 2000), it is wise to exercise caution when making comparison across countries using PSC scores.

6 Conclusion

We found that a crucial link between a corruption of public values in society and work conditions (emotional demands, social support, work positive for health) and worker subjective health was due to whether the most senior occupational safety and health leaders were able to implement PSC, the policies, practices, and procedures for the protection of worker psychological health and well-being. Levels of job control were directly affected by corrupt public values levels. By exposing the insidious effects of corrupt public values for workers and proposing that ethical leadership for worker psychological health might be a mechanism, we hope that societal attention will focus on the workplace as a site of resistance, and that managerial, occupational safety and health, and human resource efforts will be mobilised to counteract and prevent corruption stemming from the state.

  • Key Messages

  • Corruption of public values may infiltrate workplaces via influences on upper echelon ethical leadership (e.g. PSC).

  • We found that a corruption of public values at the country level was negatively  associated with PSC.

  • Work conditions (emotional demands and social support) and worker health could be explained by variations in national levels of a corruption of public values through PSC.

  • Corruption had a direct effect on worker job control perhaps through disempowering processes.

  • Union density and corruption together accounted for job satisfaction.