Economic changes and crises make people more concerned about their jobs (De Witte, 2005). Among the various definitions of job insecurity, we highlight the perception of the inability to maintain the desired work situation (Greenhalgh and Rosenblatt, 1984) and the feeling of threat that employees feel of becoming unemployed (De Witte, 2005). There seems to be a consensus in the literature that job insecurity is subjectively perceived by employees. Such implies that the same event can be interpreted differently by several people. For example, in an organization that is in the process of being laid off, there may be employees who feel more or less threatened by the situation depending on their perception of job insecurity.

Some authors defend that job insecurity is a multidimensional concept. Borg and Elizur (1992), for instance, argue that, on one hand, there is a cognitive dimension of job insecurity associated with the belief that employees are likely to be dismissed in the future (e.g., “I will be fired”), and on the other hand, an affective dimension associated with the emotional burden (e.g., “I am concerned about being fired”). Conversely, Hellgren et al. (1999) defend that job insecurity comprises a quantitative labor insecurity dimension, concerning the continuity of employment (e.g., “will I continue to perform this function in the future”) and a qualitative labor insecurity dimension, regarding the continuity of different work aspects (e.g., “will I keep my salary” or “will I continue to work the same hours”), with the latter not directly implying dismissal. Hence, as Vander Elst et al. (2014, 2016) suggest, job insecurity can be defined as the employees’ feelings (e.g., of fear or worry) that their job is at risk, associated with the undesired possibility of losing their current job in the future. Shoss (2017) stresses that the increasing attention given to this construct is aligned with the technological, economical, and political changes observed over the past few decades that have left many employees feeling insecure about the future of their jobs.

Job insecurity is a stressor that causes job dissatisfaction (Rosenblatt et al., 1999), burnout (Ismail, 2015), and higher levels of anxiety and heart disease (Burchell, 1994) in employees. Similarly, De Witte (1999) demonstrated that people with greater levels of job insecurity also exhibit lower levels of mental health, equivalent to levels observed in unemployed people, which indicates that feeling insecure in the workplace can be as detrimental as being unemployed. The negative impact of job insecurity (for the individual or household) on mental health during COVID-19 has been highlighted in recent literature, namely in United States young adults (18–26 years; Ganson et al., 2021). Moreover, job insecurity also negatively affects other important variables, such as organizational commitment (Buitendach and De Witte, 2005) and performance (Schreurs et al., 2012).

In a normal context, that is, outside the pandemic situation, the pinpointed ways to mitigate the adverse impact of job insecurity include, but are not limited to, employing Human Resources (HR) practices to help reduce the unpredictability of the future in the workplace. Parker et al. (1997) demonstrated that the use of explicit communication with employees about organizational changes leads to a lower perception of job insecurity. Greenberg and Lind (2000) take a step further by showing that merely communicating is not enough. Employees must also be encouraged to participate in the organizational decisions, so that they can feel a greater control over their future at work, especially in a context of rapid changes and/or crisis, as observed in the outbreak of COVID-19. It is also interesting the fact that employees with permanent contracts are more negatively affected by job insecurity than employees on temporary contracts (De Cuyper and De Witte, 2009). This suggests that employees with greater work stability are those who are most affected by the fear of losing it. In other words, even though their contracts protect them against it, the fear of losing their job is higher when working conditions are more stable.

Nowadays, employees around the world are experiencing growing uncertainty about their future employment due to the pandemic situation. As demonstrated, living under the chronic threat regarding the continuity of their job has adverse consequences. Considering the current reality, it is necessary to explore other mitigating factors of this process besides HR practices, which is the main goal of the study. In the same line of thought, another objective is to analyze potential protective factors for job insecurity, such as organizational resources (e.g., perceived organizational support) and job resources (e.g., job autonomy), as suggested by previous studies (e.g., Jiang et al., 2021), from the theoretical perspective of job demands and resources.

The perception of organizational support is a type of organizational resource that might ease the negative effects of job insecurity (POS; Morgan, 2018). POS is conceptualized as the extent to which employees believe that their organization and leaders value their contributions and care about their well-being (Eisenberger et al., 1986; Shore and Wayne, 1993; Lee and Peccei, 2007), thus fulfilling their socio-emotional needs (e.g., self-esteem and emotional support) in the workplace (Rhoades and Eisenberger, 2002; Kim et al., 2016).

Van Woerkom et al. (2016) showed that POS can buffer the negative effects of job demands, which in their study were the workload and emotional demands facets. Additionally, Rhoades and Eisenberger (2002) found that job security can be considered a facet of POS, in situations where the employers wish to keep an employee in the organization. As Morgan (2018) showed, POS is an optimal choice as a resource due to its well-established relationship with job (in)security and its similarities with other job resources.

The central idea of the Job Demands-Resources (JD-R) model (Demerouti et al., 2001) is that working conditions, which are specific to every occupation and/or labor function, can generally be classified as either job demands or job resources (Bakker and Demerouti, 2007). Hence, the JD-R model can be applied to various occupational settings in order to analyze the consequences of specific work environments on employees’ well-being and work outcomes. Job demands can be described as the physical, psychological, social, or organizational aspects of the job that require continuous physical or mental effort that are associated with certain physiological and psychological costs (Demerouti et al., 2001). It can also be defined as the physical, psychological, social, or organizational aspects of the job that can positively contribute to the achievement of work goals, reduction of job demands and its related costs, or stimulation of personal growth and development (Demerouti et al., 2001), such as job autonomy and POS. According to Demerouti et al. (2001), the main pathogenic health indicator within the JD-R model is burnout. In contrast, as a salutogenic health indicator, Schaufeli et al. (2002) introduced the concept of work engagement. The JD-R model comprises two causal – essentially independent – processes, namely the health impairment process and the motivational process. Job resources play a motivational role by stimulating work engagement and positive organizational outcomes, such as performance or organizational commitment (Bakker and Demerouti, 2017; Lesener et al., 2019).

The JD-R model is an uncommonly used theoretical framework to examine job insecurity (Mauno et al., 2007), but nonetheless represents a promising framework to study the effects of job insecurity on individual and organizational outcomes (Schaufeli, 2016). Although the JD-R model has been expanded and can be used to explain a wide variety of organizational phenomena, research has yet to fully attempt to integrate the job insecurity literature in this model (Morgan, 2018). Previous studies have shown that the contextual factors, like job insecurity in the workplace, can be considered as challenges for employees (Lu et al., 2014). In fact, the resources employed as an attempt to deal with the job demands associated with job insecurity can be physical, such as a supplementary income to address the economic vulnerability, or psychological, such as fostering employees’ POS to offset their decreased trust in the organization (Morgan, 2018). Based on this perspective, employees with high levels of job insecurity would highly benefit from the resources made available to them.

Work engagement, based on the Theory of Self-determination (Deci and Ryan, 1985), is a positive and satisfactory state of mind in relation to work, which is characterized by vigor, dedication, and absorption. It implies a sense of accomplishment that also involves a positive cognitive state and persists over time, revealing a motivational and social nature (Schaufeli et al., 2002).

Studies carried out in the organizational context indicate that employees with a higher level of work engagement are highly motivated and achieve better outcomes at work (Bakker et al., 2008). As a result, work engagement can show opposite effects of burnout, as per example, higher job satisfaction, more organizational commitment and organizational citizenship behaviors, and better job performance (Bakker et al., 2008; Nahrgang et al., 2011).

A study developed by Galanti et al. (2021), highlighted some job demands and resources that may affect negative (work stress) and positive (work engagement and job productivity) outcomes of employees’ remote work. The observed results showed that work engagement could decrease based on social isolation and family-work conflict. Job demands of remote work can significantly decrease productivity and work engagement. Managers, HR officers, and workers engaged in remote activities should consider family work conflict, social isolation, and distracting work environments as potential obstacles and job autonomy and self-leadership as potential enablers of working from home (WFH) engagement.

A non-probabilistic method was used to select a convenience sample. A total of 207 Portuguese employees participated, of which 64.7% (n = 134) were women, aged between 20 and 65 years, with an average of 44.68 years (SD = 9.92). Of the total, 35.7% of the respondents had a master’s degree (n = 74; 35.7%). The majority of the participants were employed in organizations in the tertiary sector (n = 183; 88.4%). The type of organizations were public institutions (n = 58; 28.0%), private companies (n = 18; 8.7%), national private companies (n = 66; 31.9%), multinational companies with headquarters in Portugal (n = 5; 2.4%), and multinational with headquarters abroad (n = 13; 11.1%).

During the first lockdown in Portugal that started in March 2020, the majority of the participants changed their labor situation to telework (n = 110; 53.1%), was given a layoff (n = 11; 5.3%) or was dismissed (n = 2; 1.0%). Of the remaining participants, 31.4% did not suffer any change in their employment situation (n = 65) and 9.2% (n = 19) started working under other conditions (e.g., on leave for family assistance, working 2 weeks in telework and 2 weeks on-site, or given a partial layoff). Regarding the working schedules during this period, 65.2% (n = 135) of the participants reported no changes.

The questionnaire was organized in two main sections. The first, related to sociodemographic and socio-academic characteristics, asked participants for information regarding their gender, age, and education level. The second section integrated the following four self-report measures:

The Job Insecurity Scale (JIS; De Witte, 2005) is an 8-item measure that varies on a five-point Likert scale between 1 – “totally disagree” and 5 – “totally agree.” The higher the score, the greater the insecurity felt by employees about their work. Both the original scale and the one implemented in this study presented a Cronbach α greater than 0.8. This measure reflets two dimensions: quantitative labor insecurity, linked to the concern about loss of function or employment; and qualitative labor insecurity, associated with the concern about negative changes in function or employment. In this study, a single factor version was used, based on the psychometric evaluation of the scale across five European countries (Vander Elst et al., 2014). Accordingly, the obtained adjustment values allowed for the use of the one-dimensional structure (χ2/df = 10.107; p < 0.00; CFI = 0.755, IFI = 0.757; RMSEA = 0.255; SRMR = 0.115).

The Job Resources Questionnaire (Lee et al., 2017) is composed of 14 items measured on a five-point Likert scale, ranging from 1 – “never” to 5 – “always.” The original scale presented between acceptable and good internal consistency across all dimensions, namely job autonomy (α = 0.85), performance feedback (α = 0.88), and technology resources (α = 0.67). In this study, the Cronbach α coefficient values showed a greater variation between dimensions: job autonomy (α = 0.91), performance feedback (α = 0.88), and technology resources (α = 0.52). As the observed value for technology resources was below acceptable, this facet of job resources was removed from the following analysis. Considering the remaining two facets (i.e., job autonomy and performance feedback), the overall scale presented a good reliability value (α = 0.91). In this study, the adjustment value for this structure was good (χ2/df = 3.190, p < 0.00; CFI = 0.908, IFI = 0.909; RMSEA = 0.103; SRMR = 0.064).

The Perceived Organizational Support Scale (Eisenberger et al., 1986) is an 8-item measure scored in a seven-point Likert scale, varying from 1 – “strongly disagree” to 7 – “strongly agree.” The Portuguese version (Santos and Gonçalves, 2010) was used in this study. Both the original scale and the Portuguese version presented an internal consistency above 0.78, which was an improvement over the 0.7 validity reported by Nunnally (1978). In this study, the adjustment value of this structure was acceptable (χ2/df = 9.349, p < 0.00; CFI = 0.808, IFI = 0.809; RMSEA = 0.244; SRMR = 0.144).

The Utrecht Work Engagement Scale (UWES-3; Schaufeli et al., 2019) was adopted in this study, based on face validity, theoretical reasoning, and earlier feedback from respondents. As such, three items were used, one for each dimension of work engagement: vigor (“at my work, I feel bursting with energy”), dedication (“I am enthusiastic about my job”), and absorption (“I am immersed in my work”). Regarding the reliability of the scale, the Cronbach α coefficient for this study was acceptable (α = 0.76).

An online questionnaire was developed through the Survey Monkey platform and shared through social media. The questionnaire was active between May and June 2020, with the application of the instrument lasting an average of 15 min.

The study was approved by the Ethics Committee of the Faculty of Psychology and Educational Sciences of the University of Porto (2020/07-10b). Respondents were informed of the anonymous and confidential nature of their participation and data. Their participation was voluntary and there were no rewards given, monetary or otherwise.

The data was analyzed using the IBM SPSS program (version 26.0) and JASP (version 0.14). The conducted analysis were (a) descriptive statistics, including computing means, standard deviations, skewness, and kurtosis (when appropriate); (b) Pearson correlations; and (c) mediation. The confidence intervals were calculated using the bias corrected bootstrap given that the sampling distribution of the indirect effect was asymmetric (Biesanz et al., 2010).

To reduce the influence of common method bias in our study (Podsakoff et al., 2003) we used two different approaches. The first approach focused on designing instruments that emphasized anonymity and confidentiality of responses and that used different instructions to create psychological separation between sets of variables. The second approach consisted in applying statistical control by performing a principal component analysis with varimax rotation with all variables, which allowed us to examine the variance of the common method. It is recommended that the first factor explain less than 50% to ensure that common method variance does not represent a problem in the study. The Harman test was successfully performed in all studies to ensure the dimensionality of all variables and resistance to the effects of the common method (Podsakoff et al., 2003).