Abstract
Summary
The primary aim of this study was to identify the factors influencing turnover intention, with particular focus on the adverse changes in social care organizations caused by COVID-19, as well as job satisfaction and burnout. A multistage stratified sampling method and a cross-sectional design were used with a random sample (N = 664). Data were collected through computer-assisted personal interviews. The relationships between key variables were analyzed using Pearson correlation and structural equation modeling.
Findings
The current study revealed that the uncertainty and unpredictability experienced by social workers during the COVID-19 pandemic, along with negative changes in working conditions, significantly reduced their job satisfaction and led to higher levels of burnout, ultimately increasing their intention to leave their workplace. The results obtained from the structural equation model showed that COVID-related negative changes directly affected social workers’ turnover intention Additionally, these changes had an indirect effect through several mediating factors, including increased job demands, emotional exhaustion, and reduced job satisfaction, which in turn contributed to a higher likelihood of turnover intention.
Application
The findings could provide valuable guidance for social care organizations in developing more targeted measures to retain their workforce. It is particularly important to prepare social workers—both during their university and postgraduate studies—for interventions during crisis or emergency restrictions and for managing the challenges that arise in such situations. Furthermore, to improve retention, targeted measures should be implemented to enhance job satisfaction, with a special focus on improving their financial situation and working conditions.
Introduction
The turnover and shortage of social professionals have been a worldwide challenge for social care organizations. Employee turnover in the social care sector has been a subject of concern for many years, as persistently high rates contribute to workforce instability and difficulties in maintaining service quality (Mor Barak et al., 2001; Nilsen et al., 2023). Moreover, turnover in the sector has intensified following the COVID-19 pandemic: nearly 421,000 social service workers in the EU had left the sector at the end of 2020, almost 2 percentage points higher than in 2019 (Federation of European Social Employers, 2021). Within the sector, personal service workers (12.7%) and care workers (11.8%) have the highest percentage of employed people who started working in roles related to direct social care and support services within the last 12 months. These occupations include positions such as home care assistants, nursing aides, residential care workers, and other frontline caregiving roles that involve providing direct assistance to individuals in need of social or healthcare support (Eurostat, 2021). This rate is 3 to 4 percentage points lower than that of the elementary workers, the occupations with the highest turnover, but twice as high as in other human services areas (e.g., among healthcare or teaching professionals).
The negative consequences of high turnover in social care organizations are well documented: not only does it reduce organizational performance (Hom et al., 2017; Park & Shaw, 2013), but it also has a detrimental effect on the quality of services provided (Pharris et al., 2022). The persistently high turnover rate disrupts the connection between clients and their social professionals, hindering the establishment and maintenance of resilient relationships between them (Burstain, 2009; Curry, 2019). The research of Strolin-Goltzman et al. (2010) drew attention to the fact that loss of trust and lack of stability are the most harmful consequences of turnover from the point of view of families and children in care. For children in care, especially young adolescents, it can be traumatic loss if the helping figure they trusted and could rely on quit and thus cause poor child welfare outcomes (Griffiths & Royse, 2017; Healy et al., 2007; Walsh, 2007). According to experience, it typically takes about 2 years for a new frontline social worker to learn what needs to be done in their jobs and to develop the knowledge, skills, abilities, and dispositions to work independently (Curry, 2014; Ellett et al., 2007).
The turnover of social professionals is also a serious problem in Hungary, yet it is an underresearched field. According to the data of The Hungarian Central Statistical Office, 56,919 professionals 1 work in institutions providing basic and specialized social services in Hungary, and the number of vacancies is at least 3,600. Although there are no recent official statistical data on labor flow processes in the sector, empirical research consistently reports that social institutions are struggling with a persistent labor shortage (Rácz, 2015). In 2020, labor shortages were greater than 10% in more than half (59%) of workplaces, with the worst situation in this regard being in specialist care (nursing homes, homes for the disabled, rehabilitation institutions) and child protection (Gyarmati, 2021).
Overview of research literature
Turnover intention, job satisfaction, and burnout in social work
Employee turnover is a complex process involving various motivations, from intentions to leave to actual departure from the job. Intention to quit is a negative psychological response to specific occupational or organizational conditions, which can lead to a consciously considered decision to voluntarily leave their current job (Tham, 2007). The relevant literature considers intention to quit to be the best predictor of actual turnover (Lambert et al., 2001).
The investigation of factors leading to employee intentions to quit has been extensively researched in the international literature. Literature reviews show that organizational conditions are the most important factors influencing employee intentions to quit, such as work climate, organizational culture, remuneration or job satisfaction (Li et al., 2016; Pi et al., 2018). Studies among human service workers have also shown that the main determinants of the intention to leave the workplace are related to work and organization (Mor Barak et al., 2001). Research among professionals working in social care organizations has consistently shown that job satisfaction is highly predictive of intention to leave (Acker, 2010; Brimhall et al., 2014; Lambert et al., 2001). Social workers who are satisfied with their job or work-related conditions are more likely to stay in their current job, while low levels of job satisfaction contribute significantly to high rates of intention to leave (Kalliath & Kalliath, 2015). Some previous research has also shown that among social professionals, remuneration (e.g., salary, benefits) is the most likely factor to determine the level of job satisfaction, and higher job satisfaction related to pay has typically been associated with lower turnover intention (Kim & Kao, 2014; Kim & Stoner, 2008). Supervisor support is positively associated with job satisfaction in multiple studies (Cole et al., 2004; Lambert et al., 2012), and Smith and Shields (2013) even found that this is the strongest determinant of job satisfaction. There is also evidence that among social workers, the sense of success in helping others is closely related to job satisfaction, which buffers turnover intention (Gellis, 2002).
Burnout has been shown to significantly affect turnover intention and actual turnover among social workers (Alarcon, 2011; Kaiser et al., 2020; Kim & Stoner, 2008). Job burnout is a psychological characteristic or symptom in the workplace (Maslach et al., 2001). Based on Cherniss (1980), this is a process in which the attitude and behavior of specialists deteriorates under the influence of stress at work. Burnout is an extremely serious problem among various professions, but especially considerable in human service areas. Several studies have shown that social workers form one of the main risk groups in this regard as they are exposed to a variety of workplace events that could negatively impact on their mental health (Győri et al., 2024). Previously identified factors contributing to the development of burnout include overwhelming high workloads, job demands, low job control, threats of violence, and working with vulnerable people who have experienced trauma or stressful life events (Ertas, 2015). These factors are all familiar challenges in social work. International and domestic research has shown that child protection professionals are at higher risk of burnout than those working in other social services fields (Conrad & Kellar-Guenther, 2006; Győri & Perpék, 2022). Researchers link this to a higher prevalence of compassion fatigue—secondary traumatic stress arising from helping someone who has experienced or suffered trauma—in child protection.
Several studies have associated job satisfaction and burnout in social work. Most of them show that high job satisfaction is negatively associated with burnout (Abu-Bader, 2000; Gómez-García et al., 2021). The relationship between burnout and job satisfaction is especially salient among social workers considering that the average expected working life for social professionals is much shorter than for other helping professionals such as doctors or nurses (Curtis et al., 2009).
Social work context in Hungary
The results of Hungarian research reflect that Hungarian social workers work in an environment in extremely poor working conditions lacking resources (Husz et al., 2020). Despite the fact that material conditions have improved somewhat in the past 10 years thanks to various EU cofinanced projects, there is still a significant infrastructural backlog: for example, 30% of social workers working in villages and 55% of those working in cities do not even have mobile phones provided by their workplace for their own use (Husz et al., 2020). The effectiveness of professional support and assistance provided to clients is also hindered by the significant workload of professionals. Despite the fact that Hungarian law stipulates that a social worker in family support services can provide care for up to 25 families based on a cooperation agreement and an additional 10 families without an agreement, in reality, due to a lack of capacity, many social workers handle a significantly higher caseload. More than a quarter of them care for over 25 families, and it is not uncommon for some to manage up to 50 families or more. The situation is particularly concerning in villages, where for an average of 43 families cared for, there is often only one family helper who performs all social assistance tasks alone (Husz et al., 2020).
Besides, in the case of one-person services, substitution is not provided in case of illness or leave of the specialist concerned, so there is no time and opportunity for prevention (Pataki & Somorjai, 2006). In addition to difficult working conditions, excessive workload, high caseloads, and administrative obligations, social workers and other professionals within the social system receive neither adequate financial nor professional recognition. Their remuneration is well below the national average, and their salaries are significantly lower than those of professionals working in healthcare or other human service sectors. In 2022, the average salary of social workers was HUF 439,000 (EUR 1100) gross, which is only 87% of the average salary in Hungary (HUF 504,340 gross, EUR 1280) (HCSO, 2023). According to a 2016 survey, the social care profession ranks in the last fifth of the prestige order of the 173 occupations surveyed in Hungary (HCSO, 2018).
In addition, in recent years there have been continuous changes in the system of social services, which have posed additional uncertainties for professionals. Such far-reaching changes included changes of maintainers of institutions (Act XXVIII/2020) and the integration of family assistance and child welfare (Act CXXXIII/2015). The underlying reason for the professional–organizational integration of child welfare services and the transformation of the care to make it a two-tier system was that the legislator intended to promote the strengthening of the preventive and developmental character of social assistance activities by cleaning the profile, that is, by separating it from its official function (Gál, 2015). However, qualitative studies show that the coordination of the operation of primary child welfare care and specialist care has been unsuccessful to this date (Balogh et al., 2019).
The coronavirus pandemic has also posed serious challenges to both the entire healthcare system and individual professionals (Rácz, 2020). During the pandemic period, conflicts intensified in several arenas (family, school), the number of clients/recipients of care multiplied, and their needs often became more serious. As in other countries, the social services have been transformed by the COVID-19 pandemic in Hungary as well. In March 2020, basic social and child welfare services, which can be provided without face-to-face contact, have been moved online, with the exception of crisis intervention. 2 Ban on visits was implemented in all inpatient and residential social care institutions, while admission to residential institutions was also announced. 3 The operation of homeless services was restructured to avoid overcrowding. For specialized social and child protection services, a ban on visits, discharge and admission was introduced till June 18, 2020. The staff of the institutions had to maintain online contact with the residents. 4 Despite their workload, before and during the pandemic periods, far less attention was paid to those working in the social sector than in the healthcare and education sectors. Although the physicians and healthcare workers in Hungary received a one-off HUF 500,000 (EUR 1290) allowance in July 2020 for their work during the pandemic, the government recognized the extra burden of workers in state-maintained social institutions, only with a modest (HUF 75,000, i.e., 193 EUR, which was merely15% of the healthcare workers’ bonus) and delayed the payment of the allowance until December 2021. This bonus covered roughly one third of the 80–90,000 workers in the social sector, as it did not cover institutions run by the Church or municipalities. Thus, although they worked on the frontline, social workers were not adequately compensated during the epidemic. The combined effect of these circumstances could lead to a deterioration in the mental health of social workers, a weakening of their professional commitment, reduced efficiency and high turnover.
Study objectives
Research conducted during the pandemic has shown that the increased workload of social workers due to COVID and their high risk of being exposed to infection through their occupation had a significant impact on their well-being at work: perceived stress had increased among those involved in this profession (Banks et al., 2020; Ben-Ezra & Hamama-Raz, 2021) and had often led to health issues (e.g., emotional exhaustion, somatic illnesses) as well (Tayri, 2022). Despite the growing research on the high turnover of social workers, studies examining the correlation between negative changes in social care organizations during the COVID-19 pandemic and social worker turnover have been limited. However, recent research has begun to address this gap. For example, Lushin et al. (2023) explored specific types of burnout, job satisfaction, and turnover among U.S. child welfare case workers during the early waves of the COVID-19, while McFadden et al. (2024) examined perceptions of safe staffing, mental well-being, and intentions to leave the profession among U.K. social workers. Additionally, MacLochlainn et al. (2025) investigated the pandemic's impact on older people's social workers in the United Kingdom using a mixed-methods approach. These studies provide valuable insights into the challenges social workers faced during the pandemic and their potential impact on turnover. Current study aims to build on and complement existing research, by examining the direct and indirect impact of difficulties caused by work and organizational changes during the pandemic on social workers’ turnover intention, addressing the gap in understanding how such factors played out in a European context under distinct regulatory and governance conditions. This focus on a European setting, particularly Hungary, differentiates it from other studies that might be based on North American or other contexts.
Consequently, two main hypotheses were as follow: (a) The negative changes in working conditions due to the COVID-19 pandemic would be associated with social workers’ turnover intention. As the COVID-19 pandemic was a health crisis, the Hungarian government, as part of its emergency governance, imposed a ban on health and social workers dismissals between November 2020 and May 2022. This unique context provides a rare opportunity to examine how such centralized policies impact social workers’ turnover intentions during a crisis. (b) Job satisfaction and burnout would mediate the relationship between COVID-19 pandemic-related negative changes at work and social workers’ turnover intention. In our view, in the already difficult social sector, the health crisis which was managed by applying highly centralized system solutions in Hungary, ignoring the experience and creative solutions of professional employees, has aggravated the difficulties at work and placed additional burdens on employees.
In summary, this research combines a unique geographical and regulatory context with a dual focus on direct and mediated effects, thereby enriching the existing literature on social workers’ experiences during the pandemic.
Method
Sample and procedure
Our data source was a cross-sectional survey conducted between May and November 2022 using CAPI devices. Based on the number of social workers in Hungary, the tolerated margin of error (5%), the confidence level (95%), and expected response rate of at least 80%, a sample of 664 participants was randomly selected from the registries held by the Hungarian Central Statistical Office, reflecting the status of social workers as of 31st December, 2021. The sample was representative for (a) type of specialty of social care (basic social services, family and child welfare services, specialized care, child protection), (b) type of provider (state, municipality, religious organizations, and other nonprofit or for-profit organizations), (c) location (territorial units for statistics, NUTS-2 regions), (d) age, and (e) gender. Study aides delivered and collected the questionnaires to and from the participants. Participants were reassured of the anonymity of the study in writing. Participants who declined participation were automatically replaced to ensure final sample size. Following data cleaning, we reviewed the distribution of the sample and applied correction weighting to counteract slight deviations from the expected distribution and maintain representativeness. The average sample weight was 1, and the total range was 0.653–1.100. The study was approved by the Institutional Ethics Review Board of the HUN-REN Centre for Social Sciences (TK-20/2021).
Measures
Turnover intention
The turnover intention was measured by the Roodt's Turnover Intention Scale (TIS, Roodt, 2004) using its abridged six-item version (TIS-6), which is considered by the literature to be a reliable measurement tool for predicting actual fluctuation (Bothma & Roodt, 2013; Muliawan et al., 2009). Using the items of the scale on a Likert-type scale, respondents can indicate how often and with what certainty they perceive the statements listed. The statements relate to 9 months prior to the survey. Item 1 of the scale represents the idea of frequent termination, Item 2 means the relationship between work and personal professional ambitions, Item 3 refers to frustration at work, Item 4 expresses the idea of another, more suitable job, Item 5 stands for the acceptance of another job with the same salary as that of the current one, and Item 6 serves for expressing the anticipation regarding the next working day. Each item is assessed using a 5-point Likert scale ranging from “strongly disagree” (1) to “strongly agree” (5), for which the Cronbach's alpha was .804. Higher scores indicate a higher likelihood of quitting one's current job.
No Hungarian version of the TIS existed prior to this study, and translation and counter-translation work were required for cultural adaptation purposes. The Hungarian version of TIS was prepared based on international recommendations. The original scale in English was translated into Hungarian by three researchers of the research project for producing the initial translation, who developed a common version based on the three translations. This version was a counter-translation made by a native-speaker bilingual translator, and there was no substantive difference in meaning between the two versions. The final version was piloted involving 30 people. Based on their opinion, we have made a few more minor changes. Subsequently, we included the questionnaire in our sample, along with other questionnaires.
Burnout
Burnout was measured by the Mini version of Oldenburg Burnout Inventory (Demerouti et al., 2003; Halbesleben & Demerouti, 2005). The Hungarian abbreviated version of the measuring tool was developed by Ádám et al. (2020), and the Hungarian version was also validated on an average population (Mészáros et al., 2020). The questionnaire measures burnout on two subscales: five items measured exhaustion, which is physical and emotional tiredness connected to work (e.g., “There are days when I feel tired before I arrive at work.”), and five measured disengagement, which contains a lack of motivation in work and intensive depersonalization (e.g., “Lately, I tend to think less at work and do my job almost mechanically.”). All items are scored on a 4-point Likert-scale, ranging from 1 (“totally disagree”) to 4 (“totally agree”). Both subscales showed acceptable internal consistency (Cronbach's alphas = .822 and .717, respectively).
Job satisfaction
To measure social workers’ satisfaction with their current work, we used the Hungarian version (Győrffy & Girasek, 2014) of brief overall job satisfaction measure. The scale includes single questions to assess the degree of satisfaction with the work itself, coworkers, supervision, reward received for work (material rewards, moral esteem), working conditions, and the workplace in general. Each item using a 5-point Likert scale which ranged from 1 ('strongly disagree”) to 5 ('strongly agree”). Cronbach's alpha for this scale was .84.
COVID-19 pandemic-related work and organizational changes
A six items scale was designed for the purpose of the current study, which measured negative changes in social care organizations following the COVID-19 outbreak. The statements of the measuring tool developed by us were aimed at the difficulties and changes related to professional work and work environment, which can be specifically linked to the outbreak of the pandemic. Participants were asked to indicate whether any of the following occurred in their workplace after the pandemic outbreak and, if so, to what extent it made day-to-day work more difficult: staff shortages or redundancies; changes in work or responsibilities; administrative burden; changes in safety rules and regulations affecting work; financial constraints; exposure to the risk of infection at the workplace (e.g., problems ensuring safe working conditions, lack of appropriate protective equipment or disinfectants, difficulties in maintaining social distancing, etc.). Respondents were able to rate each statement on a 5-point Likert scale ranging from “not at all” (1) to “completely” (5). Cronbach's alpha was .83.
Demographics
A demographic questionnaire included a series of questions on personal characteristics such as age, gender, marital status, and subjective income situation (categorical variables: can cover usual expenses with difficulties, with minor difficulties, they can cover them relatively easily and easily or very easily). In addition, information on each respondent's professional history was collected to identify their respective educational degree or diploma, job position (frontline social worker, middle manager, or head of institution), number of years spent working in the social work profession, and type of specialty of social care, which determines a work context and work content as well (e.g., family and child welfare, child protection, street social work, school social work, care for disabled people/addicts/elderly, etc.).
Data analysis
First, descriptive statistics and correlations were obtained. After this, confirmatory factor analysis (CFA) was used to examine the construct validity of the latent variables. Finally, structural equation modeling (SEM) with maximum likelihood estimation method and 95% bias-corrected bootstrap was applied to test the hypotheses. In the SEM, fit indices were analysed in order to evaluate the final model. When assessing the fit, we expected the relative fit index comparative fit index (CFI) and the Tucker–Lewis index (TLI) providing comparative fit to be above 0.900, and the root mean square error of approximation (RMSEA) and standardized root mean squared residual (SRMR) indicators used for absolute fit to be below 0.080 (Hu & Bentler, 1999). All data analyses were performed using the IBM SPSS version 24, and the IBM AMOS version 26. A value of p < .05 was considered statistically significant.
Results
Characteristics of respondents
The majority of participants made up of female respondents (85.7%), and being in married or living common law relationships (73.6%). The average age of respondents was 44 (SD: 9.4). As far as their level of education is concerned, the vast majority (79.6%) had a higher social education: 48% of them graduated from college while 30% had a university degree. 5 They had been working in the social work profession for an average of 12.7 years (SD: 8.7). Most of the participants (70.6%) worked as a frontline social worker; one-sixth (17.6%) as middle managers and another tenth (11.7%) as heads of institutions. Almost equal proportions of the participants had been employed in institutions providing basic social services (34.9%) and family and child welfare (32.7%), while a quarter (24.1%) of them worked in specialized social care institutions and 8.3% in child protection care (including those working in foster care networks and children's homes). Most of the services they worked at belonged to the public sector (81.9%), and the rest were church-maintained organizations (9.6%), or other nonprofit or private organizations (8.4%). Socioeconomic status was measured by subjective income situation, with 35.9% of participants in a household that can cover usual expenses with difficulties, 36.2% with minor difficulties, and 27.9% can cover them relatively easily and easily or very easily.
Descriptive and bivariate correlations
Table 1 presents descriptive statistics and intercorrelations between the study variables. The average score on the six-item turnover intention scale is 15.2 (SD = 5.1, min 6, max 27). If the total score given for TIS items is divided into two categories applying the cutoff of 18, following the recommendation of Bothma and Roodt (2013), it is evident that 30.0% of participants are characterized by a serious mobility intention, that is, a conscious and deliberate intention to leave the workplace. The average score for the exhaustion dimension of burnout in our sample is 12.8 (SD = 3.1) and the average score for disengagement is 10.9 (SD = 2.7). Based on international standards (Peterson et al., 2008), a significant proportion of the social professionals surveyed, namely 67.5%, reach the burned-out range on the exhaustion dimension (above ≥2.25), while it is lower on the disengagement dimension (55.1% is above ≥2.51). Examining the respondents’ satisfaction with work and workplace, it becomes clear that the factors they are most satisfied with include the organizational culture of their workplace, that is, their colleagues (M = 3.81) and their superior (M = 3.79), while they are not pleased with their working conditions (M = 3.33) as well as the nonmaterial (M = 2.64) and financial recognition of their work (M = 2.17). Respondents reported an average of three types of challenges related to professional work and work conditions related to the COVID-19 pandemic. It was the changes in safety rules and regulations (M = 4.21) and an increase in the administrative burden (M = 4.13) that made their everyday work more difficult the most as well as issues related to providing safe working conditions (M = 3.97) also caused above average difficulties, while changes in responsibilities (M = 3.26), staff reductions (M = 2.97) and financial constraints (M = 2.97) posed challenges to a lesser extent.
Descriptive statistics and intercorrelations.
Note. M = mean; SD = standard deviation.
p < .01.
p < .001.
Significant correlations were found between all the variables studied. Work and organizational changes after the COVID-19 outbreak positively correlated with turnover intention and all two dimensions of burnout, and negatively correlated with job satisfaction. Job satisfaction negatively correlated with burnout and turnover intention. All two dimensions of burnout positively correlated with turnover intention. The results of the correlation analysis indicate that all the variables adopted in the study are significant and influence turnover intention.
The SEM: Direct and indirect effects of the study variables on turnover intention
Before the path analysis, we conducted CFA to test the factorial validity of our measures through maximum likelihood estimation. The CFA was performed on interitem variance-covariance matrices. The results of the CFA, including latent variables, showed that the models fit the data well (CMIN/df = 4.118; SRMR = 0.046; CFI = 0.925; TLI = 0.918; RMSEA = 0.059). These results provide support for the factorial validity of the constructs measured.
SEM revealed direct and indirect effects of study variables on turnover intention. For the purpose of achieving a more parsimonious model, the proposed model was further modified by trimming nonsignificant paths based on the series of chi-square significant tests. Removing the nonsignificant paths did not influence the model fit. As a result, the final model had a good fit to the data (CMIN/df = 4.436; SRMR = 0.052; CFI = 0.947; TLI = 0.939; RMSEA = 0.064). The variables explained 49.9% of the variance in turnover intention among social workers represented in this study.
Figure 1 presents all standardized path coefficients and demonstrates the relationships among constructs in the model. Standardized direct and total indirect effects are shown in Table 2. All of the path coefficients are statistically significant at p < .05. The results show that the relationship between the COVID-19 pandemic-related negative changes in social care organizations and social workers’ turnover intention was significant with β = 0.155, p = .000. That is, after the outbreak of the COVID-19 pandemic, negative changes affecting working conditions directly and significantly increased the turnover intentions of social professionals, thereby hypothesis H1 is confirmed. It also had an indirect effect mediated by job satisfaction on turnover intention (β = −0.135, p = .000). Furthermore, the COVID-19 pandemic-related work and organizational changes indirectly and positively—but weakly—influenced turnover intention (β = 0.071, p = .004) via burnout. Also, both the job satisfaction of social workers and burnout mediated the association between negative changes at workplace during the pandemic and turnover intention, thereby the statistical results support hypothesis H2. These results indicate that work and organizational changes caused by the COVID-19 have also led to turnover intentions while reducing feelings of job satisfaction and increasing burnout.
Path diagram of the results of the structural equation model.
Summary of the structural equation model (standardized direct and indirect effects).
It can also be observed that COVID-19-related changes in the social care organizations directly affected both job satisfaction (β = −0.176, p = .000) and burnout (β = 0.197, p = .004). Reviewing the direct effects of job satisfaction, it can be seen that it had a very strong direct negative effect on turnover intention (β = −0.527, p = .000), and also had a direct adverse effect on burnout (β = −0.338, p = .000). Thus, burnout not only mediated the impact of pandemic-related changes in working conditions to turnover intentions, but also directly increased it (β = 0.280, p = .000).
Regarding demographic and work-related variables in the model, age was significantly associated with burnout (β = −0.167, p = .036), but had no direct impact on turnover intention. Gender was also associated with levels of burnout (β = 0.254, p = .017), but was not associated with turnover intention. Years of experience of social work negatively predicted burnout (β = −0.215, p = .007) and turnover intention (β = −0.101, p = .049). Job position, on the one hand, had a direct positive effect on job satisfaction (β = 0.151, p = .032) and on turnover intention (β = 0.131, p = .035), on the other hand, it also had an indirect influence on the intention to turnover through feelings of job dissatisfaction (β = −0.102, p = .050). Finally, the area of working (family and child welfare) had a direct impact on both turnover intent (β = −0.142, p = .005) and burnout (β = −0.184, p = .002).
Discussion
During the COVID-19 pandemic, like healthcare and other care service workers, social professionals—especially those on the frontline—faced a high risk of infection, as well as mental and physical health issues due to increased workloads, constantly changing social distancing regulations, partial online communication with clients, and other unpredictable work factors (Nyashanu et al., 2020). Despite the growing research on social workers’ turnover, the study of factors influencing turnover in the social sector during the COVID-19 pandemic is a less explored area. This study aimed to examine how the COVID-19-pandemic-related negative changes in the social care organizations affected the turnover intention of Hungarian social workers. Meanwhile, we analyzed and controlled for the role of job satisfaction, burnout and other work-related and demographical characteristics on turnover intention. We used the SEM to determine the effects of work and organizational changes during the COVID-19 on turnover intention of social workers.
The findings of this study indicate that Hungarian social workers demonstrate a serious intention to leave their profession, which aligns with turnover intention levels observed among Canadian social workers (Wilson, 2022) but differs from those reported in China (Su, 2020) and healthcare professionals globally (Quek et al., 2021; Sungbun et al., 2023). These results suggest that while social workers in Hungary face challenges similar to those in other countries, their specific working conditions may contribute to distinct patterns of turnover intention. Moreover, the high prevalence of burnout, particularly exhaustion and disengagement, highlights the strain placed on social professionals, especially those working on the frontline. These findings are consistent with prior studies linking excessive workload and inadequate workplace support to burnout and job dissatisfaction (e.g., Lushin et al., 2023; McFadden et al., 2024).
Our results also suggest that, while Hungarian social workers generally report satisfaction with their colleagues, superiors, and organizational management, dissatisfaction with working conditions and lack of financial and nonmaterial recognition remains a significant concern. This aligns with previous research emphasizing the critical role of fair compensation and professional recognition in employee retention within social care settings (MacLochlainn et al., 2024).
Overall, these findings reinforce the need for targeted policy interventions to address burnout, improve workplace conditions, and enhance professional recognition in the social care sector. Future research should further explore these dynamics across different organizational and national contexts to develop evidence-based retention strategies.
The hypothesized model concerning the effect of COVID-19-related negative changes in the social care organizations on job satisfaction, burnout, and turnover intention were supported by the SEM model. The first contribution of the study is that it verifies the significant impact of difficulties caused by work and organizational changes during the pandemic on social workers’ turnover intention. Its second contribution is a deeper understanding of how the COVID-19 pandemic-related negative changes affect social workers’ turnover intention. The results confirmed a double-mediating mechanism of job satisfaction and burnout in difficulties caused by pandemic that affect turnover intention. Job satisfaction plays a mediating role between work and organizational changes caused by pandemic and turnover intention. Work satisfaction is one of the key employee job attitudes influencing intention to quit (Mor Barak et al., 2001). In crisis situations, unpredictability and uncertainty, together with difficulties affecting working conditions, fundamental organizational problems and modes of operation, determine employee job satisfaction. Therefore, it is not surprising that the changes caused by the pandemic influenced social workers’ intention to turnover through their job satisfaction. Burnout also plays a mediating role between COVID-19-related negative changes at work and turnover intention. The fact that burnout affects social workers’ turnover has been confirmed by previous research (Kim & Stoner, 2008; Nilsen et al., 2023; Zhang et al., 2022). The findings from this study showed that pandemic-related work and organizational changes had a significant positive impact on social workers’ intention to turnover through their burnout level. Furthermore, job satisfaction and burnout partially double-mediate the impact of COVID-19-related changes on turnover intention. The demonstration of this mediating mechanism is the main result of this research.
Several studies have confirmed that there is a significant negative correlation between job satisfaction and burnout (Abu-Bader, 2000). Usually, it is the employee's attitude towards their job that changes first, which then affects their burnout. Job dissatisfaction is considered an antecedent and effective predictor of burnout (Hombrados-Mendieta and Cosano-Rivas, 2011). Therefore, there is an impact path from COVID-19-related negative changes at work to job satisfaction, burnout and turnover intention. Besides the double-mediation of job satisfaction and burnout, difficulties caused by the pandemic still have a significant impact on employees’ intention to turnover. This is an important aspect raised by the study, as it shows how it changes in social care organizations during extreme crisis situations. Specifically, a social worker with higher difficulties brought about by COVID-19 at work experienced relatively higher burnout and lower job satisfaction, and high burnout and low job satisfaction increased the likelihood of turnover intention.
Last but not least, this study verified that some other work-related factors had a significant impact on turnover intention. Results showed that job position had significant direct and indirect effects on turnover intention. In line with previous research findings (Cho & Song, 2017; Wang et al., 2019) we found that frontline job positions characterized by more frequent direct and stressful relationships with clients were one of the important predictors of turnover intention. However, our results also highlighted that job satisfaction mediated the relationship between social workers’ job position and the intention to quit their current jobs. It has also been confirmed that years of experience in the social field is directly and negatively related to turnover intention (Gupta & Shaheen, 2017). Another notable result of our analysis is that child protection as a field of social work had a significant effect on the turnover intent model. Social workers working in child protection were more likely to intend to quit their job compared to the employees involved in family and child welfare used as a reference group. This finding is consistent with previous research (Conrad & Kellar-Guenther, 2006). The sociodemographic variables scrutinized in this study, such as age and gender, had no significant effect on turnover intention, only on burnout. The results of several other studies also suggest that the role of demographic differences in turnover intent is minimal (Kim & Kao, 2014; Pugh, 2016).
Overall, the current study revealed that the uncertainty and unpredictability experienced by social workers during the COVID-19 pandemic and negative changes in working conditions significantly reduced their job satisfaction and led to higher levels of burnout, which contributed to an increase in their intention to leave their workplace. The findings revealed by the study could serve as a starting point for future research focusing on a deeper understanding of the problem of labor turnover in the social services sector. In this research, we primarily focused on the correlation between the COVID pandemic situation and work-related problems. It was a pandemic, and governments intended to use different methods and different forms of cooperation to find the most appropriate solutions, especially in the period before vaccines appeared. In Hungary, the relevant measures were implemented in a centralized way; no data were published, orchestration was considered the most effective method. To date, documents relating to the measures are not public. Due to the nature of the pandemic as crisis, both the health and social sectors were exposed to particularly high risks, still the crisis management concept did not rely on the local, on-site experience of professionals, and they were only the executors of central orders, which often did not lead to results (e.g., purchases of inadequate ventilators).
On the other hand, our results can provide guidance for social care organizations to develop more targeted measures to retain their workforce. It would be of particular importance to prepare social workers—even during their university or postgraduate studies—for interventions during crisis or emergency restrictions and for dealing with the problems arising in such situations.
Limitations of the study
Although there were some significant findings in this study, several limitations still remain. First of all, due to its cross-sectional nature, we cannot draw conclusions on causal relationships. Secondly, the data collected were self-report which is subject to well-established methods biases. Thirdly, the current study only focused on turnover intention, not on actual turnover behavior. Furthermore, this study was designed as a retrospective cross-sectional investigation, where social workers were asked to recall their experiences and the organizational changes they faced during the COVID-19 pandemic. Although we provided detailed reminders—specifying the exact period, contextual circumstances, and relevant organizational frameworks—to anchor their memories, the reliance on retrospective data collection may have introduced recall bias.
Conclusion
The majority of Hungarian social workers were dissatisfied with their work environment, with nonmaterial and material esteem of their work. An extremely high proportion suffered from exhaustion, while a smaller proportion suffered from disillusionment. One-third had strong intentions to quit their jobs. During the COVID-19 pandemic, changes in safety rules, regulations and administrative burdens made their daily work more difficult. The COVID-19-related negative changes in the social care organizations not only did increase the turnover intention of social professionals directly and significantly, but also had indirect effects on turnover intention through burnout and job satisfaction as mediators. In order to retain social workers, targeted measures should be taken to increase their job satisfaction, with special regard to the improvement of their moral esteem, material situation and working conditions.
Footnotes
Ethical Approval
Ethical approval for this project was given by: The HUN-REN Centre for Social Sciences Research Ethics Board (code of ethics: TK-20/2021).
Funding
The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the Ministry of Innovation and Technology of Hungary from the National Research, Development and Innovation Fund (Grant No. FK 138315).
Declarations of Conflict of Interests
The authors declare that there is no conflict of interest.
