In recent decades, management by objectives (MbO) – introduced by the economist Peter Drucker in 1954 – has become more and more prevalent in companies (Ahlers 2010). There is good reason for this: MbO shows a clear positive impact on company productivity (Rodgers and Hunter 1991). MbO (or Management by Results) is appropriate for and fostering a complex-dynamic organisation and leadership approach. For employees such goal-oriented leadership practices result in higher self-regulatory demands in terms of, for example, planning and decision making (Höge 2011). As a result, self-motivation and shared responsibility increase. For a better understanding of the impact on the employees, it is important to reflect the ambivalent forces. Ordonez et al. (2009) criticize the overstatement of goals and show several examples of negative side effects. Moreover, this kind of leadership also has a social impact in companies. Different approaches described such social dynamics (see for instance Kotter 2000). In the current paper we focus on an approach by Peters (2011) to explain these dynamics. The approach of indirect control by Peters (2011) goes one step further than other approaches by describing the implications of the social dynamics for occupational health management. This is what we will focus on: Research has shown that employees will often try to meet their work goals by all available means, even when this may involve risks to their health (Kratzer and Dunkel 2013; Krause et al. 2012). However, they do not do this just for their own benefit, but also to avoid letting down their colleagues. In addition, MbO settings operate with benchmarking and key performance indicators. This can lead employees to compare their performance and compete with others. We would argue that those negative consequences of goal-oriented leadership could be explained by self-endangering work-behaviors. Higher demands and requirement due to new working conditions do not necessary lead to worse health outcome. There are conflicting assumptions about it (e. g. Badura et al. 2012). The specific behaviors employees use to cope with the demands seem to be detrimental (Kaur et al. 2010). The present study focuses on the concept of self-endangering work behaviors as one possible maladaptive coping mechanism under goal-oriented and performance-oriented leadership practices and as a possible mediator between work demands and negative outcomes.

Our first aim is to analyze whether self-endangering work behaviors might partially explain the association between work demands and exhaustion in the context of goal- and performance-oriented leadership; and second, to determine which kinds of self-endangering work behavior are relevant in this context.

1 Indirect control

The usual practice is for managers and staff to agree on specific goals that employees have to reach within a certain timeframe. Success (or failure) in achieving these goals then has consequences for the employees (e. g., in terms of financial rewards or job security). In line with the philosopher Klaus Peters (2011), we call this kind of leadership practice indirect control.

Indirect control changes work conditions for employees. It requires a major degree of autonomy, high self-regulatory demands, and great flexibility. Managers or other superiors no longer give clear directives on how to do a task, and employees are free to decide on how exactly they will achieve their goals. They can plan and act autonomously. They “only” have to attain the set goal. Although more autonomy is usually associated with better subjective well-being (Deci and Ryan 2000), this is not always the case, and employees might also be subject to too much autonomy (Kubicek et al. 2014). Indirect control granting employees high autonomy is also accompanied by a strong exposure to market requirements and competition (Kratzer and Dunkel 2013). Organizational problems and pressures become the employees’ individual problems (Kratzer and Dunkel 2013), making their working conditions similar to those of the self-employed. Indirect control settings make economic aspects of one’s work more important: As well as doing one’s job correctly, one also needs to take an economic perspective and both calculate and prove one’s profitability for the company. This often requires achieving success on key performance indicators showing that one is better than internal or external competitors are. Some characteristics make work under indirect control especially taxing. It is often accompanied by such demands as goal spirals, invisible work, and obstructive process instructions that hinder personal productivity (Krause and Dorsemagen 2017).

1.1 Work demands under indirect control

Goal spirals.

Goals are often described as being dynamic and rising from year to year. If you attain a high goal in one year, your goal for the next year is even higher. As a result, your own success may become a threat to you (Chevalier and Kaluza 2015).

Invisible work.

Invisible work is tasks that have to be done in any case to reach the goal but cannot be debited to any project. A good example is administrative routines. These sometimes time-consuming tasks are often not factored into indirect control settings. If a substantial part of their work seems to be invisible, employees might perceive this as not being appreciated by their managers or the company (Krause and Dorsemagen 2017).

Obstructive process instructions

Even under indirect control settings, the former leadership strategies of controlling, standardizing, and providing guidelines often still continue. Employees have to meet goals while simultaneously complying with tight regulations and guidelines (Krause and Dorsemagen 2017).

1.2 Self-endangering work behaviors

What do employees do when confronted with such work demands? They try to cope. Dettmers et al. (2016) have described several self-endangering work behaviors that employees exhibit under such circumstances. They conceptualize these self-endangering work behaviors as a maladaptive active coping strategy. “Whereas self-endangering work behavior is directed toward goal attainment and may even succeed in achieving that, this success is built on detrimental behaviors that may impair well-being and health” (Dettmers et al. 2016, p. 43).

Such self-endangering work behavior in employees can take many forms. Krause et al. (2015a) have described eight different forms here: extension of working hours, intensification of working hours, sickness presenteeism, faking, substance abuse to recuperate, substance abuse to perform, reduction of quality, and bypassing safety standards. The current study focuses on the first four forms.

Extension of working hours.

This behavior describes extending working time to the disadvantage of private and family time as well as of recovery and leisure time. Another aspect of this behavior is being permanently reachable for work issues (Dettmers et al. 2016).

Intensification of working hours.

This means increasing the intensity and pace of one’s work while simultaneously decreasing social interactions and taking fewer breaks (Korunka and Kubicek 2013).

Sickness presenteeism.

There are two different conceptualizations of this form: first, the behavior of going to work despite illness (Aronsson et al. 2000); second, a drop in the productivity of an organization due to employees failing to attain the common performance level through their poor health (Pauly et al. 2008). The present study focuses on the first conceptualization and therefore on the behavioral component of sickness presenteeism (Baeriswyl et al. 2016).

Faking.

Faking describes the behavior of presenting wrong information or hiding relevant information on ones’ performance or goal progress in order to reduce pressure in the short term (Krause and Dorsemagen 2017).

1.3 Consequences of self-endangering work behaviors

Recent decades have seen a marked increase in mental illnesses in the workforce that has led to more work absences (Schuler et al. 2016). How can this be explained? We argue that new forms of leadership might play a role. Recent studies have shown that MbO might be a precursor of lower subjective well-being and more stress. In the long run, it might even lead to reduced performance (e. g., Krause et al. 2012). We argue that these negative consequences of goal-oriented leadership can be explained by self-endangering work behaviors. First studies have shown a mediating effect of self-endangering work behaviors in the association between working conditions and subjective well-being (e. g., Dettmers et al. 2016).

The current study addresses two research questions in this context.

Research question 1:

Do self-endangering work behaviors partially explain the association between work demands and exhaustion?

Research question 2:

Which self-endangering work behaviors may partially explain the association between work demands and exhaustion?

2 Method

2.1 Design

The online survey was conducted in spring 2016 with a convenience sample recruited mainly through an alumni network of Zurich University of Applied Sciences (ZHAW).

2.2 Sample

All members of the sample were working at least partially in an indirect control setting. This was operationalized with responses to the question: “I work in a management by objectives work setting” on a 5-point scale ranging from 1 (does not apply) to 5 (applies completely). Only participants who answered with 3 or higher were included in the sample. This reduced the original sample from 838 to 670 participants. Due to missing values, data from only 607 participants were entered into the structural equation analyses.

A total of 35.9% of the participants were female. In terms of age, 35.9% were 35 years old or younger, 22.6% were between 36 and 45 years old, 27.4% were between 46 and 55 years old, and 14.1% were older than 55 years. The majority of the sample worked full-time (76.8%) with only 4.2% working less than 20 h a week. More than one-half (55.8%) held a managerial function. Participants worked in very different industrial sectors, such as health and social services (18.5%), finance and insurance services (14.3%), and manufactory industry/energy supply (14.1%). Due to the recruitment strategy over the alumni network of the ZHAW, 94.9% held a university degree. This is not representative for Swiss population.

2.3 Measures

Work demands.

These were assessed with the three scales goal spirals, invisible work, and obstructive process instructions (Schraner 2015). Goal spirals were assessed with two items (e. g., “My working environment is characterized by constantly rising performance and profit goals.”); invisible work, with three items (e. g., “A substantial part of my everyday work is not considered in the evaluation of my performance.”); and obstructive process instructions, with two items (e. g., “Standardized processes are hindering my everyday work.”). All items were rated on 5‑point scales ranging from 1 (does not apply) to 5 (applies completely).

Self-endangering work behavior.

Four different kind of self-endangering work behaviors were assessed: extension of working hours, intensification of working hours, sickness presenteeism, and faking (Krause et al. 2015a). All items started by asking “In the last three months, how often has it occurred that …” They were rated on 5‑point scales ranging from 1 (very seldom/never) to 5 (very often). Extension of working hours was assessed with six items (e. g., “In the last three months, how often have you refrained from compensating leisure activities in favor of your work?”); intensification of working hours, with three items (e. g., “In the last three months, how often have you worked at a pace that you could not maintain in the long run?”); sickness presenteeism, with four items (e. g., “In the last three months, how often have you worked a full day/a full shift despite sickness?”); and faking, with three items (e. g., “In the last three months, how often have you whitewashed information [e. g., in reporting] to reduce pressure in the short term?”).

Exhaustion.

We measured exhaustion with the validated seven-item subscale from Demerouti’s (1999) burnout scale. A sample item reads: “After work I usually feel weary and drawn.” Responses were given on a 4-point scale ranging from 1 (absolutely inapplicable) to 4 (absolutely applicable).

Table 1 reports means, standard deviations, Cronbach’s alpha, and intercorrelations.

Table 1 Means (M), standard deviations (SD), correlations, and Cronbach’s alpha
Full size table

2.4 Data analyses

Data analyses were conducted in lavaan (Rosseel 2012). This is an R package for analyzing structural equitation models. The research questions were addressed with mediation analyses. Fig. 1 presents the proposed model. One model was estimated with self-endangering work behaviors as a latent factor consisting of extension of working hours, intensification of working hours, sickness presenteeism, and faking. Four further separate models were estimated for each of the self-endangering work behaviors. These four separate models included the self-endangering work behaviors as latent variables composed of between two and six items each. All models included work demands as a latent factor consisting of the mean scores on goal spirals, invisible work, and obstructive procedural guidelines. Exhaustion was also included as a latent factor composed of seven items. We tested for significance with bootstrapped standard errors.

Fig. 1
figure 1

Tested mediation model for self-endangering work behavior (as latent factor) in the association between work stress and exhaustion

Full size image

3 Results

3.1 Self-endangering work behaviors as latent factor

Fig. 2 depicts the results of the mediation analyses. First, work demands correlated positively with self-endangering work behaviors. Second, self-endangering work behaviors were associated positively with exhaustion. Third, self-endangering work behavior partially mediated the association between work demands and exhaustion. Indicators for the model fit are reported in the caption to Fig. 2. The model fit was acceptable.

Fig. 2
figure 2

Self-endangering work behaviors (as latent factor) as a mediator in the association between work demands and exhaustion. Direct effect: 0.119 (90% CI [0.000, 0.217]), indirect effect: 0.265 (90% CI [0.179, 0.379]), total effect: 0.383 (90% CI [0.267, 0.502]); goodness-of-fit indices: χ2 = 295.690, df = 74, p = 0.000, χ2/df = 3.996, RMSEA = 0.070 (90% CI [0.062, 0.079]), CFI = 0.912, SRMR = 0.047

Full size image

3.2 Separate models for all self-endangering work behaviors

Figs. 3, 4, 5 and 6 depict the mediation findings for the separate analyses with one figure covering each self-endangering work behavior. Estimates for the direct, indirect, and total effect are listed in the caption to the respective figure. All self-endangering work behaviors except sickness presenteeism showed a mediation effect in the association between work demands and exhaustion. In other words, the self-endangering work behaviors extension of working hours, intensification of working hours, and faking partially explained the association between work demands and exhaustion. The goodness-of-fit indices (reported in the caption to the respective figure) showed an acceptable fit for the models extension of working hours and faking and a good fit for intensification of working hours and sickness presenteeism.

Fig. 3
figure 3

Extension of working hours as a mediator in the association between work demands and exhaustion. Direct effect: 0.320 (90% CI [0.199, 0.439]), indirect effect: 0.056 (90% CI [0.030, 0.097]), total effect: 0.376 (90% CI [0.251, 0.491]); goodness-of-fit indices: χ2 = 380.510, df = 101, p = 0.000, χ2/df = 3.767, RMSEA = 0.068 (90% CI [0.060, 0.075]), CFI = 0.912, SRMR = 0.060

Full size image
Fig. 4
figure 4

Intensification of working hours as a mediator in the association between work demands and exhaustion. Direct effect: 0.209 (90% CI [0.111, 0.294]), indirect effect: 0.182 (90% CI [0.131, 0.254]), total effect: 0.391 (90% CI [0.257, 0.500]); goodness-of-fit indices: χ2 = 170.162, df = 62, p = 0.000, χ2/df = 2.754, RMSEA = 0.054 (90% CI [0.044, 0.063]), CFI = 0.968, SRMR = 0.036

Full size image
Fig. 5
figure 5

Sickness presenteeism as a mediator in the association between work demands and exhaustion; direct effect: 0.348 (90% CI [0.224, 0.487]), indirect effect 0.031 (90% CI [−0.008, 0.059]), total effect: 0.379 (90% CI [0.247, 0.502]); goodness-of-fit indices: χ2 = 154.250, df = 74, p = 0.000, χ2/df = 2.084, RMSEA = 0.042 (90% CI [0.033, 0.052]), CFI = 0.976, SRMR = 0.033

Full size image
Fig. 6
figure 6

Faking as a mediator in the association between work demands and exhaustion; direct effect: 0.280 (90% CI [0.153, 0.429]), indirect effect: 0.117 (90% CI [0.040, 0.186]), total effect: 0.396 (90% CI [0.266, 0.516]); goodness-of-fit indices: χ2 = 204.204, df = 62, p = 0.000, χ2/df = 3.294, RMSEA = 0.061 (90% CI [0.052, 0.071]), CFI = 0.936, SRMR = 0.044

Full size image

4 Discussion

Our study shows that self-endangering work behavior partially explains the association between work demands and exhaustion. If employees report higher work demands, they also report higher scores on self-endangering work behaviors such as extension of working hours, intensification of working hours, sickness presenteeism, and faking. The results support our assumption that self-endangering work behaviors might serve as a maladaptive coping mechanism and partially explain the association between specific work demands and exhaustion. The work demands examined here (goal spirals, invisible work, and obstructive process instructions) are typical for indirect control settings. The separate analyses revealed which self-endangering work behaviors are involved. These revealed that all investigated self-endangering work behaviors except sickness presenteeism mediate between work demands and exhaustion.

We argue that self-endangering work behaviors might explain the increases in health problems under indirect control conditions. Indirect control is not a problem per se; what seems to be potentially detrimental is the specific work demands that go along with it.

4.1 Strengths and limitations

Being cross-sectional, the present study was unable to investigate the respective association longitudinally, and we cannot draw any conclusions on causality. The suggested effects might well take the opposite direction. We cannot rule out the assumption that exhaustion leads to higher self-endangering work behavior. However, it would seem implausible to conceptualize work demands as an effect of exhaustion and/or self-endangering work behaviors.

Another limitation is the common method source of our data. We cannot rule out the possibility that some associations in our results were due at least in part to common method bias. Nonetheless, several studies have shown that a common method source is not an unconditional problem (e. g., Meade et al. 2007). Furthermore, we do not see any other efficient way of measuring subjective constructs such as exhaustion and self-endangering work behaviors and work demands that could serve as an alternative to asking participants.

Another limitation is the self-selection of our study participants. The majority of the sample has an academic background. Therefore, the results cannot be generalized to the entire Swiss working population. Furthermore, we cannot rule out the possibility that people who experience very high work demands would not have taken part in our study because of their lack of time.

4.2 Conclusion and transfer

Indirect control is not harmful per se. It grants high autonomy to employees and offers them opportunities for growth and development. This might provide many beneficial outcomes for employees (Krause et al. 2015b). However, how can such indirect control simultaneously increase employee exhaustion? Whether indirect control is beneficial or harmful depends on how companies implement it. In beneficial indirect control settings, employees can negotiate their goals. Furthermore, these goals can be adjusted if circumstances change. In addition, companies offer support in goal pursuit if needed. And last but not least, effort is rewarded and appreciated. In contrast, harmful indirect control settings do not value effort but only success. In indirect control settings, leadership functions not only through goal setting but also by providing a general framework and by formulating guidelines with which to comply. These rules define specific aspects such as how and what counts as direct work and which tasks are designed for internal posting. If these regulations are very strict, they might become an obstacle in everyday work and hinder the goal pursuit process. On top of reaching their goals, employees also have to deal with obstructive process instructions and time-consuming controls. Harmful indirect control settings are also characterized by goal spirals and by a competitive climate.

As pointed out in the introduction, indirect control imposes higher self-regulatory demands on employees. It also increases individual responsibility for the success of one’s work outcomes, and employees have to bear the consequences of failure just like the self-employed. We argue that indirect control not only impacts on the individuals but also triggers social processes. Key performance indicators and benchmarking foster comparisons between colleagues and teams – especially if these indicators are transparent on the individual level. When employees work extra hard to meet their goals, they do not do this just for themselves, but also in order to avoid leaving their colleagues in the lurch. Particularly when teams aspire to meet common goals, colleagues pay attention to how much their colleagues are having to struggle. This leads to high expectations toward oneself and toward one’s colleagues.

Employees feel guilty about staying at home even when they are sick: They know that if they do not work due to illness, their colleagues will have even more to do. As a result, they still come to work even when they are not feeling well.

Employees feel guilty about not working at weekends: they are particularly disturbed when they know that their colleagues are having to do so in order to meet the team goals.

Hence, on one hand, indirect control increases the pressure on individuals to reach their goals. On the other hand, it evokes social processes that trigger even more self-endangering behavior. Therefore, we would argue that under conditions with indirect control but high work demands and low work resources, detrimental social processes may impact on employees by fostering self-endangering work behavior. In the long run, this may well lead to employee exhaustion.

4.3 Implications for occupational health management

The tasks and role of occupational health management change continuously. Employees do not wait for their occupational health manager to introduce measures to reduce working hours or to set the maximum number of working hours per day. Employees would not see such measures as an aid but as more of a hindrance in pursuing their goals. They are not obliged to extend their working hours or work more intensely, but they want to do so. They either want to attain their goals or they feel a pressure to attain them. They do not harm their health consciously, but accept the risk. We would advise management to focus on reducing any detrimental work conditions that might arise under indirect control. For instance, abolishing goal spirals would greatly reduce the pressure on employees. In line with Keupp (2016) and Iwers (2017), we wish to emphasize the importance of breaking the cycle of gains and focusing on beneficial working conditions and self-care in order to deliver sustainable work place health promotion.