Covid, work reorganisation and trust: the importance of employment relations

ABSTRACT

The world met the COVID-19 pandemic with various national strategies to handle the crisis. In Norway, upon the advice of health authorities, the government imposed a societal lockdown on 12 March 2020. This had major consequences for people’s work lives and organisations’ management and daily operations. Many employees were temporarily dismissed, while others suddenly had increased workloads. Working from home became the new normal for those who could perform their tasks online. We used survey data from April 2020 to investigate the impact of the changed work conditions on trust relations in the workplace. We asked whether the informants perceived the changes as consonant with the collective agreements. Trust in management increased significantly among those who were least affected but declined among those who experienced major changes at work. However, this distinction became less marked when we considered adherence to collective agreements and employee influence. This implies that management that ‘plays by the book’ can maintain a high level of trust amidst organisational change.

KEYWORDS:

• COVID-19 pandemic
• managing
• work conditions
• employee influence
• trust

Introduction

The COVID-19 pandemic caused a major upheaval in work life and work – life relations. From one day to the next, most employees faced changes in their work situations that also affected their private lives. Employees set up home offices and experienced changes in work hours, pay and the tasks required of them. Infection control measures became the new authority for employees and employers. In other words, infection control measures introduced new employment standards.

Under normal circumstances, Norwegian organisations follow procedures that involve employees in organisational restructuring and employment changes. Employees’ rights are set out in the Work Environment Act, which limits employers’ ability to hire, fire and reorganise employees. Moreover, the basic agreements between employers and employees include collective agreements that guarantee employees the right to exercise it through their representatives. These regulations are key elements in the high trust that characterises Norwegian work life (Falkum 2020).

Hence, the way employers handled the pandemic crisis could have had not only a strong impact on employees’ work lives but also profound effects on their levels of trust in their employers and in the latter’s governance of enterprises and workplaces. However, few studies have addressed the effects of perceptions of organisational trust and justice (for example, Kougiannou et al. 2015; Newman et al. 2019; Searle et al. 2011; Timming 2012).

In work situations, collaborative relations between employees and management can foster trust between them. The pandemic caused sudden changes in the employment relationship for most employees to varying degrees, and exposed their vulnerabilities vis-à-vis employers. In these and similar situations, a key task of collective institutions and work – life regulation is to protect individual employees against the power of employers and ensure a just and equitable outcome for all employees. Such institutions are therefore important guarantors of trust in the workplace. A key empirical question, however, is whether employers actively used such institutions to maintain collective involvement in decision-making in the workplace in the first phase of the pandemic. A report by the European Foundation for the Improvement of Living and Working Conditions (Eurofound 2021) indicates that there were problems in industrial relations (IR) during the early phases of the pandemic and lockdown due to the sudden loss of predictability, the increased uncertainty felt by employers and employees and the shift to online dialogue. Because IR ‘has been shown to play an important stabilising role, especially in times of crisis’, the Eurofound report expresses concern about the potential negative impact of the crisis on actors, institutions and processes at all levels and fears that the pandemic crisis may have sidelined unions and employers’ associations in decision-making (Allinger and Adam 2021). Unilateral decision-making undoubtedly has the potential to reduce trust in organisations. A study in Germany demonstrated that codetermination made a difference in terms of achieving higher employment stability for the incumbent workforce during the pandemic (Fackler et al. 2021). How lockdown affected industries and sectors varied widely between countries (Markey 2020).

In the study reported in this article, we investigated how changes in work conditions in response to the pandemic crisis affected employees’ trust in management and perceptions of organisational justice. Work reorganisation includes (i) more or less work than before, (ii) working from home and (iii) changes in work tasks. We paid particular attention to how employee influence and the maintenance or disruption of employment standards and collective agreements influence the relations between new work conditions, trust and organisational justice. We used the results of a representative survey of employees in Norway conducted in April 2020, one month after the implementation of the lockdown. The data provided information about changes in workload and work tasks, dismissals and work-from-home arrangements resulting from the pandemic; employees’ views on whether normal employment standards were complied with; and employees’ evaluations of the trust relations in their respective work organisations.

In the next section, we define the key concepts of trust and justice and present the theoretical perspectives and the analytical model that underpinned the present study. In this section, we briefly describe how pandemic policies changed employees’ work conditions. We then describe the data obtained and the regression models used, report and discuss the study results and findings and conclude the article.

Theoretical perspectives and analytical approach

Trust and organisational justice

Following Newman et al. (2019, 37), we defined trust as ‘the confidence that one party to the employment exchange (management) will not exploit the other’s (the employee’s) vulnerabilities‘. This definition views trust between actors as being embedded in a larger system or network and based on a long tradition in sociology and the economy of making risk and vulnerability elements of trust (Korczynski 2000). This definition also acknowledges that trust is a social relation that may differ over time and across different categories and subgroups of employees (Blunsdon and Reed 2003; Lysgaard 1961; Svensson 2018).

Fox (1974, 13 and 66) thematised institutionalised trust and considered trust embedded in people’s roles and relationships connected to the level of discretion permitted by their jobs. He defined trust as ‘[having] faith or confidence in [another]’ and claimed that it is a key factor in organisational well-being. Morgan and Zeffane (2003) investigated how different forms of organisational change affect employees’ trust in management. Their findings revealed that any change negatively affects trust, with the most adverse changes being those concerning workplace structure. However, both direct and indirect employee involvement mitigate the negative relationship between organisational change and trust in management. Thus, Morgan and Zeffane (2003, 70) concluded that ‘institutional trust is crucial to maintain high-trust relations’ and that ‘a fall in trust emerges from the failure of the system to circumscribe the worst effects of change’. Theoretically, they viewed employees’ trust in management as always influenced by employees’ situations and shaped by experience and context (Morgan and Zeffane 2003, 58). Blunsdon and Reed (2003) put forward a similar argument, saying that conditions and context have an independent impact on trust relations.

The organisational sciences generally study justice descriptively and focus on employees’ perceptions or evaluations of fairness (Cropanzano and Molina 2015, 379). Cropanzano and Molina’s 2015 review of the literature on organisational justice suggested that in most cases, individuals make fairness judgements after an event has occurred. To judge whether an event is unfair, people make use of a comparative referent. When the event is negative and inconsistent with the reference point, people are apt to decide that an injustice has occurred (Cropanzano and Molina 2015, 379).

Organisational justice and trust are intertwined. Oskarsson, Öberg and Svensson (2009) showed that fair and properly enforced institutions in the workplace have a positive correlation with trust. Workplaces that adhere to formalised rules for dismissal and conflict resolution also enjoy higher levels of organisational citizenship, better behaviour and supervisory trust. The aforementioned researchers demonstrated the role of formal regulations and procedures in maintaining trust.

The concept of an effort bargain

In the present study, we used the concept of effort bargain (Korczynski 2022) as a frame for analysing the translation of labour power into labour during the pandemic. The pandemic was a critical incident that disclosed implicit understandings of what was to be exchanged between employers and employees. Thus, we considered managers’ compliance with collective agreements and employee influence in the translation process on the one hand and the impact of these on perceptions of organisational justice and trust relations in the workplace on the other.

Korczynski (2022) criticises the growing application of psychology to the study of employment relations and human resource management (HRM) and puts forward the concept of the social contract of work to return productively to the original focus of employment relations. The social contract of work is defined as workers’ implicit, collective and socially embedded understanding of the effort bargain at work (Korczynski 2022, 1). The term refers not directly to social relation per se but to one party’s understanding of a social relation––that is, to workers’ understanding of the effort bargain (Korczynski 2022, 10). The employment relationship is suffused with implicit understandings of what is to be exchanged, which the workers become aware of as they experience, engage in and reflect upon events at work (Korczynski 2022, 8). Korczynski (2022) argues that research must be designed to access critical incidents that can be understood as events at work that throw the effort bargain into light and make it explicit to the parties involved, which is latent in the normal social order. It cannot be assumed that there is an implicit understanding of the effort bargain shared by workers and management; rather, the differences can become apparent through the playing out of critical incidents, which are key issues to be uncovered by research (Korczynski 2022, 10).

The effort bargain at work concerns the link between the translation of labour power into labour on the one hand and the return to workers for their labour on the other. In terms of the translation of labour power into labour, a wide range of elements, such as various aspects of work organisation, from the terms of the division of labour to the form of control, can be crucial (Korczynski 2022). To study changes in trust and employee experiences and anticipations of justice in the introduction of new work conditions and employee influence as a moderator, Godard (2014) connected trust relations to economics, law and labour unions. This IR approach, as it emerged in the 1980s, stands in contrast to the more psychologically oriented HRM approach (Godard 2014; Legge 2004). The IR perspective includes collective values established and institutionalised by employers’ associations and labour unions’ mutual agreements, and national labour laws.

According to Korczynski (2022), further research on the concept of the social contract of work must ask questions different from those linked to the psychological contract concept (for instance, questions about causes of contract breach that extend beyond surface-level issues) to consider wider structures of political economy. He calls for questions that move beyond individualising modes of knowledge creation, such as the questionnaire survey, to consider collective understandings and actions, such as in strong national IR traditions. Korczynski acknowledges the use of a (modified) version of the psychological contract concept as there is a conceptual space for complementarity between a modified version of the psychological contract concept (with the individual as the unit of analysis) and the concept of the social contract of work. A modified version of the concept of psychological contract and the concept of the employment contracts can be thought of as complementary, in order to be utilised depending on whether the research questions concern collective understandings of the effort negotiation or individual understandings of the effort negotiation (Korczynski 2022, 11). The concepts of psychological and employment contracts connects to the concept of institutional trust, in theory as well as practice in work places as outlined in the previous paragraph.

A survey that represents work life participants may catch collective and representative values, interpretations and understandings of changing realities and individual anticipations, experiences and consequences.

Analytical approach

The pandemic situation was unique in that it came as an exogenous shock that affected the whole work sphere in various ways. The Norwegian government’s decision to implement a national lockdown on 12 March 2020 had major implications for work life and labour markets as employers had to redistribute resources, make employees work at home and close businesses or change opening hours. The Norwegian strategy for combating the virus led to four main responses that cut across educational groups and the private and public sectors: (i) more work than before; (ii) less work than before; (iii) working from home; and (iv) changes in work tasks. We know that employees’ ratings for vertical trust, organisational justice and reliance-based trust significantly increased in the post-lockdown period compared with the ratings a month earlier (Drange, Falkum, and Wathne 2020). However, the theoretical argument gives reason to believe that trust relations vary according to the new work conditions that employees have been exposed to and to employees’ perceptions of codetermination during the implementation of those conditions.

First, changes in work hours and the distribution of such changes can affect the perception of organisational trust and justice as they affect employees’ private lives, pay levels and ability to plan. Salaried workers, for instance, do not necessarily get paid for the extra work they put in. Employees who lost their jobs or had their work hours reduced during the lockdown received unemployment benefits, but the rate was 80% of annual salaries below NOK 300’, and 62,4% of annual salaries of NOK 600’ (Mortensen 2021). The number of people applying for unemployment benefits was highest in the first weeks after 12 March 2020: 372 000, about 17% of the total Norwegian workforce of about 2.2 million employees (Bakken and Vidal-Gil 2020).

Second, employees are subject to employers’ ‘right to manage’, defined as an ‘administrative prerogative’ from the early 1900s. This means that the employer decides what tasks shall be carried out, by whom and when. Employees typically resist the ‘administrative prerogative’ when economic and industrial strategies involve major changes in workplaces (Author 2), such as those that occurred during the pandemic. Moreover, employees who provide services to clients likely have a strong attachment to their primary work tasks as they work not only for their employers but also for their clients and, ultimately, society. Local health managers moved skilled employees to special COVID-19 test centres to arrest the spread of the virus and care for COVID-19 patients. Some of these professionals and their unions claimed that such secondments would cause harm to vulnerable clients. This is an example of how disagreements between managerial and professional staff can lower the trust between them. Conversely, employees moved to more important tasks or tasks with higher levels of responsibility could report higher levels of trust and justice.

Third, more than half of the Norwegian workforce began working from home on 12 March 2020. The abrupt change and the uncertainty of the situation meant that employers and managers had very little time to prepare routines and infrastructure for digital work arenas. Management had to produce ad hoc solutions to coordinate the internal and external organisation of business. For employees, their homes suddenly had several functions. The homes of those with children were both kindergartens/schools and offices, and the employees had to juggle their roles as parents, freshly minted kindergarten/school teachers and employees. This meant that employers had to give employees much flexibility concerning not only where they worked but also when and how they worked. The situation gave management little ability to control labour. Thus, working from home is a major change in relations between employees and their managers and colleagues. Some home workers might regard this as increased autonomy, while others might be stressed by the lack of direct leadership, which they had become used to.

Changes in work hours and tasks and working from home were elements in industries’ and sectors’ pandemic strategies, which had different policies. However, employers were responsible for the implementation and execution of the lockdown policies in their respective establishments. This situation permitted employers to act fast, which could shift power from employees to management. Under these circumstances, the management’s decision to involve or not involve its social partners in decision-making can affect the levels of organisational trust and fairness. Moreover, the new employment policies defined some employees as critical for society and others as not. The ‘critical’ status also conveyed other conditions and rights both inside and outside the workplace. This is an example of changes in the power balance that plays out within the organisation, but it has connections to the wider political economy in which the employer operates (Korczynski 2022). Such distinctions, such as between critical and non-critical personnel, may challenge the legitimacy of the collaborative system and social trust in work life. Against this backdrop, we analysed the association between post-lockdown changes in employees’ work situations and levels of organisational justice and trust in management. The key questions were whether employees perceived their managers’ handling of the pandemic as being in line with the collective agreements and established standards of IR and how much influence employees felt they had on the processes. To answer Korczynski (2022) calls for research questions that move beyond individualising modes of knowledge creation and consider wider structures of political economy, we included employees’ evaluations of adherence to collective and representative values under changing realities to link the individual and collective levels. We analysed the results of a survey that considered employee influence rather than individual-level attributes to contribute to an understanding of institutional-level matters during the pandemic (Godard 2014, 7). In the aforementioned way, we tried to further our knowledge of the nature of the employment relationship as a subordinate relationship under conflict-of-interest conditions and the possible implications of this for the understanding of problems with trust (Godard 2014, 7).

Data

We used survey data to conduct our study. In April 2020, we distributed copies of a questionnaire to a representative sample of the working population to measure the effects of the lockdown on work – life relations and work conditions. We engaged the services of TNS Kantar, a professional agency, to conduct sampling and data collection. The sampling procedure required the respondents to be in employment and between the ages of 18 and 67.

The survey had a gross sample of 3,195 respondents drawn from a panel of about 60,000 Norwegian citizens. The response rate was 38%. However, not all the respondents were eligible to participate in the study due to issues such as self-employment or non-response. Eventually, we had 1,150 respondents. The attrition analyses displayed significant differences between the respondents and non-respondents in age, business sector and education. The respondents had a higher mean age than the original sample (+3 years), had higher education (56% vs 52%) and more often worked in the private rather than public sector (65% vs 63%). There were no significant differences in gender, size of the work organisation or management positions. Because we included variables with response bias in our models, we did not use sample weights to adjust for this marginal observed selection bias.

Dependent variables

We used three scales to measure vertical trust, reliable trust and organisational justice in our analyses. This provided a wider framework for assessing organisational justice and trust relations.

The scale that was used to measure vertical trust was taken from Copenhagen Psychosocial Questionnaire III (Burr et al. 2019). It measured the trust between management and employees. The scale consisted of four items: (i) Does the management trust the employees to do their work well? (ii) Can the employees trust the information that comes from the management? (iii) Does the management withhold important information from the employees (Reversed) and (iv) Are the employees able to express their views and feelings? The answers were assessed based on a 5-point Likert scale ranging from 0 (to a very limited extent) to 4 (to a very large extent). The Cronbach’s alpha value was 0.82.

The scale that was used to measure organisational justice also came from COPSOQ III (Burr et al. 2019). Its four items were (i) Are conflicts resolved in a fair way? (ii) Are the employees recognised when they have done a good job? (iii) Are all the employees’ suggestions treated seriously by the management? and (iv) Is work distributed fairly? The answers were assessed based on a 5-point Likert scale ranging from 0 (to a very limited extent) to 4 (to a very large extent). The Cronbach’s alpha value was 0.88.

The reliable trust scale was taken from Gillespie’s (2015) Behavioral Trust Inventory. It measured willingness to rely on the manager’s work-related skills, abilities and knowledge (reliance) and consisted of five items that measured how willing the respondents were to (i) rely on their leader’s task-related skills and abilities, (ii) depend on their leader to handle an important issue on their behalf, (iii) rely on their leader to represent their work accurately to others, (iv) depend on their leader to back them up in difficult situations and (v) rely on their leader’s work-related judgements. The answers were assessed based on a 5-point Likert scale ranging from 0 (not at all willing) to 4 (completely willing). The Cronbach’s alpha value was 0.94.

We translated all the items into Norwegian. We used the mean item scores in the analyses; hence, all the items had similar weights, and the scores for the scales fell within the same range as the scores for the original items.

Independent variables

The independent variables in our analysis were new working conditions, compliance with collective agreements and employee influence.

The variable of new working conditions identified those who, in the post-lockdown period, (i) had an increased workload, (ii) had different tasks, (iii) had a reduced workload and (iv) worked from home.

The variable of compliance with collective agreements and employment standards measured the responses to five survey items: To what extent do you trust the managers in times of crisis to (i) maintain practices that are in line with the Work Environment Act, (ii) maintain practices that are in line with the collective agreements, (iii) involve union officials in decision-making processes, (iv) re-employ employees according to the current rules and regulations and (v) order employees to undertake additional work according to the current rules and regulations. The answers were assessed based on a 5-point Likert scale ranging from 0 (not at all) to 4 (to a very large extent).

The variable of employee influence measured the responses to six survey items regarding how great an influence the respondents would say they had on the following: (i) decisions concerning their work, (ii) how they did their work, (iii) whom they worked with, (iv) the quality of their work, (v) the amount of work assigned to them and (vi) how quickly they worked. The answers were assessed based on a 5-point scale ranging from 0 (no influence) to 4 (considerable influence).

The six items were very similar albeit not identical to COPSOQ III’s (Burr et al. 2019) ‘influence at work’ scale.

Control variables

We included control variables that potentially confounded the relationship between new work conditions and perceptions of trust and organisational justice. The control variables were gender, age, education, business sector and management position. We chose these control variables because they correlated with the likelihood that employees experienced changes in their work conditions and with the types of changes these were (cf. Table 2; see also Ingelsrud 2021). Because these variables were also associated with different positions in the labour market and areas of employment, with women having jobs different from those of men and young employees usually having less influential positions than older workers, these variables could also influence how employees would evaluate trust and organisational justice. Managers are likely to display higher levels of perceived trust and organisational justice than other employees because their positions give them the power to make decisions. A manager is also more likely to be able to work from home than other employees. Thus, not controlling for management position could give a positive bias to the relationship between working from home and perceived trust and organisational justice.

Table 1. Descriptive statistics on key variables.

	Mean	Min	Max	N
Vertical trust (scale)	3,03	0	4	1150
Organizational justice (scale)	2,63	0	4	1150
Reliance-based trust (scale)	2,89	0	4	1150
Home office^a	0,57	0	1	1150
More hours^a	0,16	0	1	1150
Less hours^a	0,24	0	1	1150
Changed tasks^a	0,31	0	1	1150
Employment standards (index)	2,84	0	4	1150
Employee influence (index)	2,79	0	4	1150

^aThese are not mutually exclusive statuses. The share of workers who experienced neither change in work hours, work tasks nor home office is 0.22.

Table 2. Descriptive statistics on key variables across gender, education, and business industry.

	Home office	More hours	Less hours	Changed tasks	Empl. Stand-ards	Empl. influence	N
Gender
Male	50%	14%	23%	21%	2,77	2,86	536
Female	63%	19%	26%	40%	2,89	2,74	614
Education
Elementary education	20%	7%	22%	27%	3,06	2,81	41
High school, general degree	50%	14%	34%	35%	2,92	2,65	110
High school, trade certificate	25%	9%	25%	27%	2,74	2,77	181
Vocational education	33%	15%	24%	24%	2,68	2,74	130
short tertiary education (<4 years)	66%	18%	24%	30%	2,87	2,79	385
Long tertiary education (≥4 years)	83%	22%	20%	37%	2,86	2,89	303
Business industry
Public administration	91%	14%	19%	32%	3,04	2,85	118
Education	86%	32%	23%	74%	2,86	2,78	111
Health- and social services	32%	23%	14%	47%	2,77	2,50	124
Armed forces, police and justice	62%	4%	42%	38%	2,88	2,47	26
Industry and construction	42%	12%	17%	11%	2,86	2,97	247
Retail	29%	15%	33%	33%	2,79	2,64	171
Transportation	38%	4%	58%	38%	2,80	2,45	26
Liberal occupations	85%	10%	33%	37%	2,75	3,04	67
Hospitality	58%	17%	83%	42%	2,88	3,11	12
ICT and finance	82%	16%	22%	18%	2,83	2,71	68
Oil, gas and energy	61%	16%	17%	16%	2,77	2,90	128
Other	69%	17%	33%	25%	2,83	2,84	52

Bold font: p ≤ 0,05.

The variable of gender separated men and women, with men being the reference category. We included age as a continuous variable and in the second degree to allow for a curvilinear function. The variable of education distinguished between elementary education, high school education, vocational training, short tertiary education (<4 years) and long tertiary education (≥4 years). The variable of business sector distinguished between (i) public administration (the reference category), (ii) education, health and social services, (iii) the armed forces, police and justice, (iv) industry and construction, (v) retail, (vi) transport, (vii) creative occupations (entertainment, media, sports, research and analysis), (viii) hospitality, (ix) information and communication technology (ICT) and finance, (x) oil, gas and energy and (xi) other. Finally, the variable of management position distinguished between employees who had human resource responsibilities and those who did not (the reference category).

Modelling strategy

In Figure 1, we sketch out our analytical model, which includes compliance with collective agreements as a crisis-moderating element in the Norwegian work life model that can foster predictability and confidence in the ability to handle crises when they arise.

Figure 1. Research model.

The model displays direct associations between new work conditions and trust on the one hand and compliance with employment standards, employee influence and trust on the other. We also expected to find an indirect relationship between new work conditions and trust relations in the workplace through employers’ compliance with the ‘rules of the game’ for labour.

We develop our analyses in four models. The first model displays the association between new work conditions on the one hand and vertical trust, organisational justice and reliance-based trust on the other. The reference categories are those who did not experience a change in work hours due to the lockdown, those who did not work from home and those who did not experience changes in their work tasks due to the lockdown. Models 2 and 3 adjust for employee influence and compliance with employment standards, respectively, and model 4 controls for both. Models 1–4 also include the full set of control variables.

We used ordinary least squares (OLS) regression on our dependent variables. We chose OLS regression because of the linear properties of our dependent variables, which scored responses on a continuous scale from 0 to 4. OLS regression provides the best linear unbiased estimates, provided that the model assumptions hold (Allison 1999). We tested the model assumptions; performed controls of normality of residuals, homoscedastic residuals, multicollinearity, linearity and influential cases; and tested the model specifications. The results of these tests displayed variance inflation factors below the threshold (<10) and textbook examples of normalised residuals. We obtained satisfactory test results for the linearity assumption on the continuous variables. We observed some issues for the model specification for vertical trust and, to a minor degree, for organisational justice, but not for reliance-based trust. The test for homoscedastic standard errors showed violations of the OLS regression assumptions in all three models. We used robust standard errors to correct for this violation. The robust standard errors did not alter the significance of our findings. The tests revealed that we had 39 influential cases in our sample. The omission of these cases made the association between working from home and the dependent variables negative, although this was significant only for reliance-based trust and organisational justice. Our influential cases displayed strong positive correlations between working from home and the dependent variables, which strongly biased the sample association. The influential cases did not affect any of the other estimates.

Findings

Descriptive statistics

Table 1 presents the descriptive statistics of our key variables.

First, the scores for vertical trust, organisational justice and reliance-based trust were high: 3.03, 2.63 and 2.89, respectively. Second, regarding new work conditions, our study showed that the lockdown led to major changes in work practices. As shown in the table, 57% reported that they chose or were told to work from home; 16% reported that they had more to do than before; 24% reported that they had less to do than before; and 31% reported changes in their work tasks due to the pandemic. The changes in workload varied with the business industry (cf. Table 2), which was thus an important control variable in the analyses. Most of those who had more to do than before were in the health and care services and education sectors, while those who had less to do than before were in the hospitality and transportation sectors. Third, the mean scores for employment standards and employee influence were 2.84 and 2.79, respectively.

Who was most likely to have new work conditions after the lockdown?

Table 2 shows that women were subjected to changes in work conditions to a greater extent than men across all four indicators, but the difference was not statistically significant for women working fewer hours. Women had higher scores for employment standards, which was likely because more women worked in the public sector and more regulated industries, but they had lower scores for employee influence.

Those with a tertiary degree had a significantly higher likelihood of working from home, working more hours and reporting changes in their work tasks. Employees with tertiary education were more likely than those with trade or vocational training to report that their employers complied with the employment standards.

The distribution of the new work conditions and compliance with the employment standards and employee influence across areas of employment showed how the impact of the lockdown strategies varied between employees. First and foremost, about nine in 10 employees in the public administration, education (at all levels, from kindergarten to university) and creative occupations worked from home, as did 82% of the employees in the ICT and finance sectors. In the education sector, 32% of the employees, and in the health and social services sector, 23%, worked more hours after the lockdown. This was undoubtedly due to the extra work involved in home-schooling children and the extra burden placed on the health services by the pandemic. A high proportion of employees in these sectors also reported changes in their work tasks. A significant proportion of employees in the public administration, armed forces, police and justice services, retail, transport and creative occupations reported changes in their work tasks, with the figure ranging from 32% to 42%. Employees in the industry and construction sectors and in the oil, gas and energy sectors were least affected by change. Interestingly, we found no significant variation in the evaluation of employers’ compliance with employment standards across such industries, while we found a significant variation in employee influence. However, this variation may not be a result of the pandemic situation.

New work conditions, employee influence and compliance with employment standards

Table 3 shows the linear regression results for the three dependent variables. The analyses investigated the relationships between new work conditions, management compliance with rules and regulations, trust and organisational justice.

Table 3. Regression table.

	Model 1		Model 2		Model 3		Model 4
	B	se	b	se	b	se	b	se
Dependent var: Vertical trust
Home office	0.075	(0.051)	−0.021	(0.047)	0.046	(0.040)	0.002	(0.039)
More hours	−0.103	(0.059)	−0.069	(0.055)	−0.113*	(0.046)	−0.096*	(0.045)
Less hours	−0.188***	(0.052)	−0.153**	(0.048)	−0.084*	(0.041)	−0.078	(0.040)
Changed tasks	−0.075	(0.049)	−0.055	(0.045)	−0.090*	(0.038)	−0.079*	(0.038)
Employment standards			0.352***	(0.025)			0.171***	(0.022)
Employee influence					0.511***	(0.019)	0.456***	(0.020)
Constant	3.101***	(0.377)	2.416***	(0.352)	1.793***	(0.301)	1.601***	(0.295)
Observations	1150		1150		1150		1150
R-squared	0.036		0.178		0.404		0.433
Dependent var: Organizational justice
Home office	0.128*	(0.057)	−0.015	(0.050)	0.095*	(0.046)	0.008	(0.043)
More hours	−0.119	(0.067)	−0.069	(0.058)	−0.130*	(0.054)	−0.095	(0.050)
Less hours	−0.171**	(0.058)	−0.118*	(0.051)	−0.057	(0.047)	−0.044	(0.044)
Changed tasks	−0.058	(0.055)	−0.029	(0.048)	−0.074	(0.044)	−0.053	(0.041)
Employment standards			0.516***	(0.027)			0.339***	(0.025)
Employee influence					0.558***	(0.022)	0.449***	(0.022)
Constant	2.367***	(0.427)	1.361***	(0.374)	0.938**	(0.347)	0.558	(0.323)
Observations	1150		1150		1150		1150
R-squared	0.033		0.272		0.377		0.466
Dependent var: Reliance-based trust
Home office	0.072	(0.067)	−0.067	(0.061)	0.037	(0.056)	−0.042	(0.054)
More hours	−0.098	(0.078)	−0.050	(0.071)	−0.111	(0.065)	−0.079	(0.063)
Less hours	−0.211**	(0.068)	−0.160*	(0.062)	−0.090	(0.057)	−0.078	(0.055)
Changed tasks	0.016	(0.064)	0.044	(0.059)	−0.002	(0.054)	0.018	(0.052)
Employment standards			0.504***	(0.033)			0.308***	(0.031)
Employee influence					0.595***	(0.027)	0.496***	(0.028)
Constant	2.811***	(0.497)	1.829***	(0.457)	1.287**	(0.423)	0.941*	(0.407)
Observations	1150		1150		1150		1150
R-squared	0.022		0.191		0.313		0.368

* p < 0.05, ** p < 0.01, *** p < 0.001.

All models include full set of controls.

Regarding vertical trust, model 1 displays a positive association with trust for those working from home and negative associations with trust for those working longer hours, those working fewer hours and those who experienced changes in their work tasks. However, the association is only statistically significant for those working fewer hours. The introduction of employee influence in model 2 changes the estimates slightly. The introduction of compliance with employment standards in model 3 also makes the association between vertical trust on the one hand and working longer hours and changed work tasks on the other statistically significant. In model 4, with joint control for compliance with employment standards and employee influence, the association between working fewer hours and vertical trust, however, becomes weaker and is no longer statistically significant. Both the index for employment standards and that for employee influence show strong and significant associations with vertical trust. This implies that compliance with employment standards and employee influence partially explain the association between new work conditions and vertical trust, supporting the model outlined on page X

Model 1 shows that the perception that there is organisational justice, changes with new work conditions. Those working from home reported a higher level of organisational justice than those who were not. The association ceases to be statistically significant once we adjust for employee influence in model 2. Those working from home scored higher on employee influence than those who remained in the workplace after the lockdown. It is employee influence and not working from home per se that creates the association in model 1. Those who worked fewer hours reported lower levels of organisational justice than those who did not experience changes in their work hours after the lockdown. This association ceases to be statistically significant once we adjust for compliance with employment standards in model 3.

The negative association between fewer work hours and employment standards once again attenuates the association between fewer work hours and the dependent variable. Taken together, once we compare employees with similar scores for employment standards and employee influence, working from home or working fewer hours after the lockdown does not correlate with lower scores for organisational justice.

We found no direct association between new work conditions and reliance-based trust, except for those working fewer hours. The association was −0.211 points. The association attenuates and is no longer significant in model 3 after adjustment for compliance with employment standards.

Discussion and conclusion

The pandemic has paved the way for a new public narrative regarding work where employment relations have become increasingly important. This illustrates the importance of employment relations as a legitimate field of study (Hodder and Lucio 2021, 430). In the present study, we granted Korczynski’s (2022) request and asked questions aligned with the social contract concept of work. We investigated how changes in work conditions in response to the pandemic crisis affected employee trust and determined if employee influence made a difference. Unlike Korczynski (2022), who does not consider surveys capable of portraying collective understandings and actions, we assume that it is possible to use surveys to obtain data about the social contract of work, but this requires questions about the practice of institutional arrangements. By combining questions from COPSOQ III (Burr et al. 2019), an instrument for psychosocial conditions, with questions about practicing collective agreements in the translation of labour power into labour, we analysed employment relations at an institutional level. A psychologically oriented alternative to our research approach would be to keep the data collection at the individual level and, for example, to explain changes in employees’ trust in management with differences in personality types. At a time where numbers crush words both in political contexts and in academia, it is not advisable to allow psychologists to obtain a monopoly on quantitative methods and thus decide how these are to be carried out. For example, there is good reason to question the enormous reuse of certain questionnaires, such as COPSOQ III (Burr et al. 2019), without combining them with other questions. Do we ask questions that capture the data necessary to understand the development being investigated? We need to look critically at the value of validated question batteries (Godard 2014). Combining established questions with new questions prevents echo chambers.

The pandemic forced us to reflect on the scope of just transitions––that is, who is affected by a transition and who is covered by policy responses to it (Dimitris et al. 2021, 5). The answers to these questions may result in new forms of division of labour. Dimitris, Krause and Morena (2021) argue that transitions mandated by public policies are more easily recognisable and thus legitimise the demand for justice. However, many transitions are the results of corporate pressures, routinely connected with enabling but less visible public policies. Unjust transitions result in opposition to any kind of structural change (Dimitris et al. 2021).

Our analyses showed that those whose work hours were reduced after the lockdown evaluated trust relations in the workplace more negatively than those whose work hours were not reduced. Employers’ compliance with IR (collective agreements/employment standards) and employee influence, however, explain this association. This suggests that it is not the reduction of work hours per se that is associated with lower trust but the fact that those whose work hours are reduced experience violations of IR standards or a lack of influence more often than those whose work hours are not reduced. Working more hours and changes in work tasks displayed negative associations with vertical trust in the workplace. These correlations became stronger after we adjusted the analyses for employers’ compliance and employee influence. This suggests that those who work more hours and those whose work tasks are changed have more varied experiences of how employers comply with standards and involve employees than those who do not experience such changes. Once these factors are considered, those who work more hours and those whose work tasks are changed, as groups, also perceive lower levels of vertical trust. This finding shows that a trust relation between actors is not just an individual phenomenon but is embedded in a larger system or network (Korczynski 2000) that differs over time and across different categories and subgroups of employees (Blunsdon and Reed 2003; Lysgaard 1961; Svensson 2018).

In other words, it seems that it is the collective standards that help increase direct influence, suggesting that there is a connection between participation at different organisational levels – collective agreements and bargaining as well as individual employee participation – IR as well as HR. This is in line with the findings of other studies that find union representation and employee participation to be mutually supportive in the Norwegian work life (Drange, Falkum, and Wathne 2020), and to reflect that both collective and individual relations at the work places may connect to the concept of institutional trust in the work life in general as well as in the single work places.

The results of the present study indicate that employees who experienced the greatest changes in their work situations developed higher degrees of uncertainty and manifested less trust in their employers. It is likely that employees who compared themselves to employees who still had full-time work and less insecurity during the pandemic perceived injustice in the workplace (Cropanzano and Molina 2015). This is in line with Oskarsson, Öberg and Svensson’s (2009) finding that workplaces that adhere to formalised rules also enjoy higher levels of supervisory trust.

The main conclusion of the present study is that employees who experience major changes in their work relations and situations and whose employers do not comply with employment standards score lower on the trust variables than those who experience less changes in their work relations and situations. This is in line with the finding of previous studies that management’s decision to involve or not involve the employees in decision-making can affect the levels of organisational trust and fairness (Blunsdon and Reed 2003; Morgan and Zeffane 2003). Trust, understood as the confidence that the management will not exploit the employees’ vulnerabilities, represented an implicit understanding of the employment relationship before the pandemic. This understanding became explicit through the pandemic when employers put in place new employment standards for the pandemic policy. The materialisation of a new division of labour, where some employees worked from home while others were required to be physically present at the workplace, made the social contract of work physically visible. For example, place of work is an ambiguous matter as working from home or at the workplace may be considered an advantage or a disadvantage. Our study showed that employees working from home could develop higher levels of trust in management as this type of work requires employers to allow more employee flexibility and autonomy. This finding supports Fox’s (1974 theory of trust as embedded in people’s roles and relationships connected to the level of discretion permitted by their jobs. The important aspect is whether employers involve employees in the decision-making processes, or the extent to which the employer complies with employment standards during the transition period. The new work conditions make power relations more visible.

When compliance with collective agreements functions as a crisis-moderating element and creates predictability and trust, the ability to handle a crisis increases. Workplaces that ‘play by the book of IR’ handle social contracts and effort bargains more like matters reliant on mutual trust relations between the parties involved than those that do not. At a higher level, these findings indicate that the way employers adhered to the government’s pandemic guidelines could have challenged the ability of national IR to ensure fair outcomes for all employees.

The national pandemic lockdown strategy was a game-changing occurrence for all work lives as it dictated a new and unprecedented distribution of labour. First and foremost, it changed employment relationships rather than labour institutions and established collaborative structures at both the national and local levels. In the present study, we investigated the associations between changed employment relationships, compliance with collective agreements and trust. The findings suggest that compliance with collective agreements is key to maintaining organisational trust under challenging circumstances.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This work was supported by the Norwegian Research Council grant: 314486 “Digital Prism and the Nordic Model of Workplace Democracy under Pressure.” We are grateful to the anonymous reviewer who helped us clarify some points and encouraged us to develop others.

Notes on contributors

Ida Drange

Ida Drange is a Research Professor at the Work Research Institute/Oslo Metropolitan University. Her research interests cover collective bargaining systems and wage formation, occupational regulation, and labour market inclusion of foreign trained professionals.

Eivind Falkum

Eivind Falkum is Research Professor Emeritus at the Work Research Institute/Oslo Metropolitan University in Norway. His research has focused industrial relations, labor relations, union codetermination, employee participation, the growth of HRM under the new liberalism, industrial development, labor markets and organization theory. He holds a Doctor philos. degree about power and opposition in the Norwegian work life.

Christin Thea Wathne

Christin Thea Wathne is a Research Director and Research Professor at Work Research Institute/ Oslo Metropolitan University in Norway. Her research interests include industrial relations, leadership and management, New Public Management, organizational development, organizational learning, professions, social identity and working environment and mastering.

Appendix: Scales and items

Table

Variable	N	Mean	Min	Max
Vertical trust	1150	3,03	0	4
Does the management trust the employees to do their work well?	1150	3,27	0	4
Can the employees trust the information that comes from the management?	1146	3,2	0	4
Does the management withhold important information from the employees? R	1139	2,96	0	4
Are the employees able to express their views and feelings?	1140	2,68	0	4
Organizational justice	1150	2,63	0	4
Are conflicts resolved in a fair way?	1140	2,74	0	4
Are employees appreciated when they have done a good job?	1144	2,59	0	4
Are all suggestions from employees treated seriously by the management?	1139	2,52	0	4
Is the work distributed fairly?	1141	2,66	0	4
Reliance-based trust	1150	2,89	0	4
Rely on your leader’s task-related skills and abilities	1143	3,03	0	4
Depend on your leader to handle an important issue on your behalf	1128	2,84	0	4
Rely on your leader to represent your work accurately to others	1131	2,75	0	4
Depend on your leader to back you up in difficult situations	1129	2,95	0	4
Rely on your leader’s work-related judgements	1137	2,89	0	4
Employment standards	1150	2,84	0	4
(i) maintain a practice that is in line with the Work environment Act	1129	2,99	0	4
(ii) maintain a practice that is in line with collective agreements	1065	2,94	0	4
(iii) involve union officials in decision-making processes	1038	2,59	0	4
(iv) re-employ employees in line with current rules and regulations	1080	2,85	0	4
(v) order employees to undertake additional work in line with current rules and regulation	1035	2,82	0	4
Employee influence	1150	2,79	0	4
i) decisions concerning your work	1131	2,73	0	4
ii) on how you do your work	1139	3,18	0	4
iii) on who you work with	1107	2,12	0	4
iv) on the quality of your work	1140	3,42	0	4
v) on the amount of work assigned to you	1134	2,48	0	4
vi) how quickly you work	1141	2,8	0	4