The Effort-Reward Model and Its Effect on Burnout Among Nurses in Ecuador

Abstract

Burnout has harmful consequences for individuals and organizations. The study of its antecedents can help us to manage and prevent it. This research aims to explore the role of the effort-reward imbalance (ERI) model as well as the mediation of the working experience in the burnout processes. For this purpose, we have conducted a study in 629 employees from two hospitals in the city of Guayaquil (Ecuador). For this study, the Spanish version of the Maslach Burnout Inventory was applied, as well as the ERI Questionnaire, along with other socio-demographical and occupational variables. A statistical analysis was performed with the obtained data, using structural equation models (SEMs). Results showed that employee effort has a stronger and statistically significant direct effect on emotional burnout, whereas the perception of the obtained reward also had this effect but indirectly in a negative sense, with job experience as a mediating variable.

Introduction

The study of burnout in jobs in direct contact with the public is well-known in the scientific literature (Van Mol et al., 2015; Giorgi et al., 2017; Salvagioni et al., 2017). In addition, it is known that burnout occurs as a response to chronic emotional and interpersonal stressors of the working activity. It consists of three components (Maslach and Jackson, 1981; Maslach et al., 2001): tiredness or emotional exhaustion (which represents the basic dimension referred to the feeling of lacking emotional and physical resources to confront the perceived stressors), depersonalization (which represents the interpersonal dimension and expresses excessive detachment, cynicism and insensitivity toward users, and other aspects of work) and diminished personal accomplishment (which represents the self-assessment dimension referred to the feelings of incompetence and lack of personal fulfilment and productivity at work).

Thus, the causes of burnout must be sought not only in objective conditions (Garcia et al., 2015) but mainly, in the individual’s interaction with those conditions, mediating by other psychosocial variables, such as personality, emotional regulations, as well as the perceptions, evaluations, and expectations of people (Salami and Ajitoni, 2016; Bang and Reio, 2017; Hwang et al., 2019; Ortega-Jiménez et al., 2021).

Several authors agree on the adverse job characteristics that are associated with burnout processes, especially in nursing employees (Dall’Ora et al., 2020; Almodibeg and Smith, 2021; Boateng et al., 2021). In addition, the negative consequences that this exhaustion process has both on organizations, affecting the quality of the service they provide (Leon-Perez et al., 2016), and on their lives, developing mental (Ding et al., 2014) and psychosomatic disorders (Guerrero-Barona et al., 2020).

Nevertheless, the influence of other work-related and organizational variables in burnout processes has been less studied. Some studies showed that the work overload can cause in the employees a perception of inability to satisfy the demands required by the staff or the organization (Peiró et al., 2001; Portoghese et al., 2014).

From this perspective, the so-called effort-reward imbalance (ERI) theory, proposed by Siegrist (1996), establishes that work stress would be generated from the lack of correspondence between a high level of effort that work requires and low rewards, both material and psychosocial, that the employee perceives. It is clear that the overload and work rhythms, the lack of support or the absence of fair rewards are some of the new psychosocial risks at work that are going to affect the well-being of workers (Leka et al., 2017; Alvarado and Bretones, 2018). This lack of reciprocity between the high cost and the low perceived benefit will cause strong negative consequences for the organizational results as well as personal consequences for the workers (Devonish, 2018). Among the organizational consequences, we can mention the intention to quit (Leineweber et al., 2021), perception of injustice at work (Topa et al., 2016; Juvani et al., 2018) or burnout processes (Oren and Littman-Ovadia, 2013; Backhaus et al., 2018; Jachens et al., 2019).

Among the personal consequences of this imbalance, we can mention: long-term implications for employee health (Siegrist and Li, 2016), such as development of coronary diseases (Gilbert-Ouimet et al., 2014; Dragano et al., 2017), higher levels of anxiety (Eddy et al., 2018) and depression and mental illnesses in general (Rugulies et al., 2017; Koutsimani et al., 2019; Diekmann et al., 2020), among others.

Also, other authors have observed that those workers who experienced ERI presented high levels of emotional exhaustion (Bakker et al., 2000; Basińska and Wilczek-Rużyczka, 2013; Tian et al., 2021).

Therefore, we propose the following hypothesis in this study:

• H1: There is a positive relation between high levels of work effort and burnout.

• H2: There is a negative relation between high levels of work reward and burnout.

However, in this relationship between ERI and burnout, we have to take into account the existence of other mediating or moderating organizational variables. Some of these studied variables had an occupational character, such as home-work interference (Gorgievski et al., 2019) or working shifts (Xie et al., 2011).

An example of this variable could be the job experience in the occupation. In any case, the results obtained in several studies about the relations between seniority and burnout are contradictory. Thus, some authors found that the number of years of stressing work, especially in emotional works, increased the experienced stress and in consequence, the emotional exhaustion (Makara-Studzińska et al., 2020). Nevertheless, other authors pointed out a reverse relation in such way that the younger workers and therefore, with less experience, that start their professional careers, have less realistic believes that make them more exposed to occupational exhaustion.

Therefore, we propose the following hypothesis in this study:

• H3: The years of experience will have a significative mediating function in the relation between the ERI model and burnout. From this hypothesis, two sub hypotheses would arise:

• H3a: The years of experience will have a mediating function by increasing in a significative and positive way the relation between effort and burnout.

• H3b: The years of experience will have a mediating function by increasing in a significative and negative way the relation between reward and burnout.

Materials and Methods

Participants

In order to achieve our objective, we carried out a study in nurses from two hospitals of the city of Guayaquil (Ecuador). The participants were selected by non-probabilistic and accidental sampling method, being the inclusion criterion to be working as a nurse, with more than 1 year of job experience in the category and who was working on the day of data collection.

Finally, a total of 629 surveys were conducted. However, 70 questionnaires (11.1%) were subsequently rejected in the data analysis for different reasons (incomplete data, choice of several answers on the same item, and non-response). The final study sample consisted of 559 nurses from two hospitals (45.9 and 54.1%, respectively).

The age of participants ranged between 19 and 70 years old, with an average age of 42.06 years old and a standard deviation (SD) of 12.41. On the other hand, the seniority of employees participating in the study ranged between 1 and 42 years, being the average work experience 12.37 years old and a SD of 11.18. Distribution by gender shows predominance of the female sex in the study sample since 84.2% are women, compared to 15.8% who are men. Regarding the duration of the contractual relationship, the majority of the participants had a permanent contract with the organization (81%) compared to 19% who had a temporary contract.

Variables and Instruments

For this research, the following measurement instruments were used:

In order to measure the burnout variable, we used the Spanish validation by Olivares-Faúndez et al. (2014) of the Maslach Burnout Inventory – Human Service Survey (MBI-HSS) (Maslach and Jackson, 1981). This 25-item scale is divided into three subscales (emotional exhaustion, personal accomplishment, and depersonalization) and measures how often people feel emotionally overloaded or exhausted because of their work. The inventory uses a scale ranging from 1 (never) to 7 (every day). The reliability index of this instrument, measured through the Cronbach’s alpha coefficient, was 0.81 in our study. This questionnaire is one of the most widely used in studies on this syndrome (Wheeler et al., 2011; Cañadas-de la Fuente et al., 2014).

We also applied the version adapted to Spanish by Juárez et al. (2015) of the ERI Questionnaire (Siegrist et al., 2004). In this study, we have only taken into account two of its dimensions: effort, which has 6 questions and records information about the employee’s assessment of his/her work situation in relation to the effort made at work; and reward, which has 11 questions that measure the professional reward. These 17 questions were formulated on a Likert-type scale with a standardized measurement format ranging from 1 (it does not affect me at all) to 4 (it affects me a lot). The Cronbach’s alpha index obtained in this study was 0.84, whereas each dimension had an alpha coefficient of 0.79 (Effort) and 0.75 (Reward). The ERI Questionnaire is one of the most widely used (Hanson et al., 2000; Rantanen et al., 2013; Siegrist et al., 2014).

Finally, we considered the following socio-demographical variables: gender, age, and job experience. The mediating factor job experience was measured in terms of the number of years held in the profession as nurse.

Procedure

Concerning the procedure, we initially requested authorization from the Ethics Committee of the hospitals (#CE-0718), arranging a particular place and date for the collection of information.

On the date of the information collection, participants were informed about the purpose of the research, trying not to transmit information that could prompt responses to the questions included in the questionnaires.

Every potential participant agreed to collaborate. Instruments were administered in the above-mentioned order by self-application.

Data Analysis

The collected data were processed and analyzed using SPSS^© 20.0 (IBM Company, Chicago, IL, United States) and AMOS 20.0 (SPSS Inc., Chicago, IL, United States). The statistical program SPSS version 20 (IBM, Chicago, IL, United States) was applied to calculate the descriptive analysis (frequencies, mean scores, and SDs). The relationships between variables were analyzed through correlation analyses and regression model using structural equation models (SEMs) and standardized fit indexes. Conventional levels of acceptable model fit [GFI, AGFI, NFI, and CFI, values over 0.85; root mean square error approximation (RMSEA) values >0.05] were taken into consideration (Browne and Cudeck, 1993; Chen et al., 2008).

The fit of the SEMs is determined by the fit indices that show the degree of concordance between the data predicted by the model and the observed data. The fit indices contribute to measure if the model fits well enough to provide a useful approximation to reality and a reasonable explanation of the data trends.

In addition, given that the value of χ² is affected by the sample size (Barrett, 2007), we used the fit index CMIN/DF, which is less sensible and on the condition that the value was <3 (Hair et al., 2009).

In general, the SEMs allow us to test all the supposed direct and indirect effects among several variables simultaneously. These models are a superior alternative to others for proving mediations such as simple regression analysis, especially in cases where there are multiple indicators for each construct (Iacobucci, 2008).

Declaration

This study was conducted in accordance with the Helsinki Declaration and the Good Practice Guide. The protocol was approved by the management of the participating hospitals. The study instructions were given to participants in written and oral form. Confidentiality of personal data and anonymity of subjects were protected. For this purpose, all data were coded and only researchers were allowed to access them.

Results

Table 1 shows means, SD, and matrix of correlations of the study variables:

TABLE 1.

Means, standard deviations, and correlations of variables in the model.

	M	SD	1	2	3	4	5	6
1. Effort	11.17	3.72
2. Reward	22.40	7.88	0.66**
3. Emotional exhaustion	12.83	10.34	0.61**	0.46**
4. Gender (woman = 2)	1.84	0.36	0.13	0.13	0.09*
5. Age (years)	42.06	12.41	–0.10	–0.13	−0.19**	0.11**
6. Experience (years)	12.37	11.18	–0.11	−0.18*	−0.13**	0.03	0.77**
7. Contract (permanent = 2)	1.81	0.39	–0.12	–0.11	−0.10*	0.03	0.37	0.35**

*p < 0.05; **p < 0.01; ***p < 0.00.

From the results in Table 1 we observed the strong and significative correlations among the variables effort, reward and emotional exhaustion, although we were not able to observe correlation relationships with socio-demographical gender and age variables, with the exception of years of experience.

In addition, to check whether our data had a normal distribution, we applied two goodness-of-fit tests (Kolmogorov–Smirnov test and Shapiro–Wilk test) which obtained very good scores above 0.05 (K–S = 0.90 p = 0.00 and W = 0.11 p = 0.00, respectively).

In order to analyze the relationships between the different study variables, we designed a model (see Figure 1), which was analyzed using structural equation modeling (SEM) techniques. Specifically, we analyzed the effects of the two independent variables observed (Effort and Reward) on the dependent variable “Emotional Exhaustion” and the influence of the organizational variable “Job experience” as a mediating variable. In addition, standardized coefficients and the coefficient of determination (R2) were included in the model.

FIGURE 1.

Effort-reward model. **p < 0.01; ***p < 0.001.

Figure 1 shows that the variable “Effort” is the one that presents the strongest and the most statistically significant direct effect on “Emotional Exhaustion” (β = 0.53, p < 000), which would support our Hypothesis 1. However, the variable “Reward” did not have a statistically significant direct effect on “Emotional Exhaustion,” therefore we could not support our Hypothesis 2.

Regarding the Hypothesis 3 in which we pointed out the mediating effect of the years of experience in these variables, we could only partially prove the data obtained (Hypothesis 3b). After observing the results, we can appreciate that the effect the variable “Reward” produces on the dependent variable may be indirect and, in that case, be mediated by “Job experience.” All these variables manage to explain 35% of “Emotional Exhaustion” variability.

However, fit in SEMs is determined by the indexes that express the degree of concordance between the data predicted by the model and the observed data. Fit indexes help measure whether the model fits well enough to provide a useful approximation of reality and a reasonable explanation of data trends. In our case, Table 2 contains the main goodness of fit indexes for the evaluation of our model.

TABLE 2.

Fit index of the effort-reward model.

χ 2	CMIN/DF	GFI	AGFI	NFI	CFI	RMSEA	PRATIO
0.855 df = 1	0.85	0.99	0.99	0.99	1.00	0.00	0.17

The value of χ² was not significant (p = 0.26). This figure shows a good fit and allows us to accept the model; however, given the instability that χ², we complement this information with other goodness of fit indexes such as the CMIN/DF index, which is less sensitive to sample size issues. The value of this index is below the levels mentioned by the authors, which reinforces the idea that the model is acceptable. GFI and AGFI are close to 1, which demonstrates an excellent overall fit of the model. Moreover, the incremental fit indicators NFI and CFI present a very good fit, coming close to unity. The RMSEA (0.00) is well below the threshold established by the authors, which provides evidence in favor of the acceptance of the model. To sum up, the proposed model obtained good fit indexes.

Discussion

Our study allowed us to have a more comprehensive view of the importance of the ERI model in the burnout processes, especially in the variable “Reward” and how this relation is measured by other occupational variables. Yet, although the importance of both variables has been proved, we have also looked into the relationship that job experience has on professional burnout.

The proposed model, as well as the analysis of goodness and relationship between variables, showed that burnout is affected by their overstrain and by the low rewards perceived from their performance, although in the latter case, this relationship is mediated by work experience.

Thus, in general terms, we can affirm that job experience plays an important role in workers’ chronic stress. These results would be in line with the findings found by other authors (Brewer and Shapard, 2004; Bretones and González, 2011; Kruczek et al., 2020) who showed that occupational factors had a predictive role in employees’ exhaustion.

Practical Implications

The findings of our study recover the value of job experience as a personal resource that shows a mediating effect on the perception of the reward received from their performance and its effect on professional burnout.

Thus, employees with greater job experience will better deal with situations of emotional distress as a reaction against the lack of reciprocity between overstrain and the benefits obtained from that performance (Feuerhahn et al., 2012) because they have developed better coping skills and with them they can minimize the level of professional burnout and increase subjective wellbeing (Leonova et al., 2019).

From these results, we could confirm that high effort and low rewards are associated with professional burnout; in the first case directly and positively although in the second case, mediated by job experience, indirectly and negatively. Therefore, it is necessary for researchers to show greater interest in the study of mediating factors in the relationship of the effort-reward model and burnout, in order to specify the evaluation processes as well as to suggest proposals for the vocational training curriculum and the intervention strategies. All these studies will enable us to create healthier working environments (Jáimez and Bretones, 2011) as well as the development of their workers and the prevention of psychosocial risks at work.

Limitations and Recommendations of the Study

The present study presents some limitations to taken into account for future research. First, the cross-sectional design of the study. Besides, we believe that it would be advisable to conduct new studies as well as in other samples and countries about the work factors that mediate in the relationship between the effort-reward model and burnout, in order to specify the evaluation processes as well as the intervention strategies. Our study has focused on nursing professionals, so it would be advisable to carry out research in other positions (health and non-health) in order to confirm our model. It would also be recommendable to perform studies in other countries and cultural contexts that would allow us to generalize the results found.

Data Availability Statement

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

Author Contributions

LA, JR, and FB: conceptualization and writing draft preparation. FB and JR: formal analysis and data curation. LA: project administration. All authors have read and agreed to the published version of the manuscript.

Conflict of Interest

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Publisher’s Note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.