Using Longitudinal Trajectories of Working Hours to Search for Quiet Quitters: Characterizing Their Imprints

Abstract

The aim of this study is to provide an academic basis for understanding the phenomenon of quiet quitters and begin to detail the characteristics that distinguish them. The defining behavioural characteristic of quiet quitters is that they reduced the hours they worked over time, especially over the pandemic period. A sample of more than 2500 employees in Australia who had been working full-time toward the end of 2019, before the pandemic, and working full-time toward the end of 2022, after many pandemic constraints had been lifted, was analysed using multinomial regression. There were many variables that distinguished between the trajectories of hours worked between 2019 and 2022. Two groups of employees had dramatically or substantially reduced their working hours and displayed nuances in their characterisation, suggesting that they were quiet quitters. The quiet quitters appear to have experienced powerful imprinting during the time of pandemic constraints, and that imprinting may be working against prior occupational norms. The group most like prototypical quiet quitters are likely to leave their job soon, and many of them are confident they will find a job at least as good as the one they now have, with more flexibility. Having a group of employees with a new approach to work may require revisiting many approaches to management.

1. Introduction

After many of the constraints brought in for the COVID-19 pandemic were lifted, the demand for labour bounced back to unusually strong levels, but the supply of labour had declined, a drop often attributed to a decrease in the labour force participation rate (e.g., for the United States of America/USA, per Lee et al. 2024). That is, the COVID-19 pandemic changed the nature of employment so dramatically that phenomena such as “The Great Resignation” arose, with employees leaving work at relatively high rates (Baranes and Brown 2023). The Great Resignation and associated retirements may have had an impact by reducing the number of potential workers available for the labour market. However, economic analyses suggest that more than half of the decline in aggregate hours worked occurred because workers reduced their hours (Lee et al. 2024), often to the minimum that their roles would allow, which became known in the popular press as quiet quitting (The Economist 2023).

That is, during the pandemic and its associated interventions, many employees changed how they approached work. But who are these quiet quitters and why did they emerge?

There is very little academic research on the phenomena of quiet quitting specifically nor the characteristics of quiet quitters. Among the suggested reasons for the parallel phenomenon of the Great Resignation was the perception that some employees do not want to continue working for bad bosses in bad work environments and instead want to take more control of their lives (Lambert 2023). Similarly, in the popular press, the reasons why employees reduce their working hours was said to be to have more of a balance between work and non-work and/or protecting their own health, while not leaving their jobs (The Economist 2023), thereby still being able to pay their bills.

Some clues as to why the phenomenon of quiet quitters emerged may come from the work-oriented activities that occurred during the main pandemic-intervention period. During the main pandemic responses, many organisations brought in “mandatory” teleworking, especially during lockdowns, that forced many workers to work from home (Andrade and Lousã 2021). The pandemic and its associated activities and events, along with exposure to more flexible work practices such as working from home, may have motivated some people to reevaluate their work and life priorities. Employees stopped placing excessive value on work, with the result that some chose to work fewer hours and seek a better work–life balance (Lee et al. 2024). During the height of the pandemic interventions, workers searched for more meaningful work, a better work–life balance, and sought jobs where they would feel valued and respected (Aroles et al. 2021; Stamos and Kotsopoulos 2024). These descriptions suggest what was happening at that time, but not necessarily the mechanisms of how this new group of employees known as quiet quitters emerged.

At a fundamental level, the key driver for this change and the size and presence of quiet quitting as a phenomenon appears to be the COVID-19 pandemic and its associated hassles, drama, and fear of ill health or death for either oneself or friends and family. That is, the pandemic and events associated with the pandemic acted as a “sensitive moment” that can imprint on workers and shape them for years, if not decades, despite later changes in the work environment (Simsek et al. 2015).

1.1. The Emergence of Quiet Quitting Due to Imprinting

Imprinting that could lead to the development of quiet quitters occurs when, during a period of susceptibility (Marquis and Tilcsik 2013), an employee develops characteristics that reflect key features of the environment. Imprinting has the three essential features of a brief sensitive period, a powerfully impacting environment, and the persistence of characteristics over time (Simsek et al. 2015).

To create the imprint of quiet quitting, there needs to be a sensitive period, such as COVID-19 pandemic activities, where employees are significantly more malleable and imprintable (per Stinchcombe 1965). The COVID-19 pandemic activities often led to substantial changes in everyday life while under the threat of illness and death. Efforts to reduce viral transmission such as reducing contact and occasionally enacting isolation, while many workers were still trying to work, led to more flexible practices such as remote meetings, enabling many jobs to be more flexible. Many workers no longer had to commute as much to work, saving the employee substantial amounts of time, effort, and cost.

The strength of the impact of the environment is substantial during this sensitive period. The anxiety and cognitive unfreezing experienced during such sensitive periods opens individuals up to stimuli from the environment (Schein 1971). Consequently, employees look to others for cues on how to behave and are more likely to adopt new behaviours, cognitive models, and norms (Marquis and Tilcsik 2013).

To summarise, the defining behavioural characteristic of quiet quitters is that they have reduced the hours they work. That reduction in work hours occurred after having been imprinted during the sensitive, imprint-malleable period of pandemic activity and responses. The anxiety and cognitive unfreezing experienced during this sensitive period opened some individuals up to stimuli from the environment (applying Schein 1971), and they were more likely to adopt new approaches to work. In the case of the pandemic and quiet quitters, their new approaches to work, and perhaps to their lives, could include protecting their health and wellbeing—a key stimuli of the pandemic. These new wellbeing-protecting imprints adopted by some employees led them to being labelled as quiet quitters.

1.2. The Specific Drivers and Constraints Contributing to the Emergence of Quiet Quitting

A variety of issues may have influenced which employees were more susceptible and/or which employees adopted or did not adopt these new work norms. Stress and anxiety, along with job satisfaction, particularly their satisfaction with the flexibility of their work, were often seen as key drivers of this new imprint. At the same time, many employees had already been pushing back against the scope creep of work roles and against the intensification of their work, especially because of potential health impacts. On the other hand, employees may have felt constrained by social forces if they had children or were married because they may put up with poor work to, for example, enable the stability of schools and friendship circles for their family. On top of these drivers and constraints, any new imprints would have to contend with, or overlap, pre-existing imprints such as those that may have come in to being during educational experiences or professional and occupational enculturation.

To improve our understanding of quiet quitting, considering patterns of a wide range of characteristics may be useful (Sinclair et al. 2005). The specific issues that may be in these patterns of characteristics are briefly reviewed in this section.

Health, work-based attitudes, and perceptions such as the employee’s levels of stress, anxiety, and job satisfaction may be starting points for characterising quiet quitting. Quiet quitting employees are often seen to be trying to maintain a personal work–life balance or rejecting the notion that their lives should be dominated by work, as well as trying to protect or maintain their health and wellbeing (Mahand and Caldwell 2023). The transition to working from home or other changed forms of work during the stricter pandemic policy period brought about ‘technostress’ that also impacted attitudes such as satisfaction (Stamos and Kotsopoulos 2024). Furthermore, employees’ job dissatisfaction may have existed before the pandemic but has now been able to be expressed (Formica and Sfodera 2022).

More specifically, the dissatisfaction at work of quiet quitters may be regarding the lack of flexibility of work practices, particularly in the context of many employees experiencing flexible work arrangements such as working from home during the height of the pandemic period (Lee et al. 2024). During this time, some organisations used social aspects of sustainable human resources practices to retain employees (Ullah et al. 2021) and the resulting employee-oriented work flexibility can enhance the work–life balance and its associated outcomes (Ajzen and Taskin 2021).

Conversely, up to and during the height of the pandemic period, workers were often forced to adopt ways of working with a loss of boundaries between work and family (Andrade and Lousã 2021), which can have negative consequences, particularly on their health (Andrade and Neves 2024). Together, these points suggest that health, stress, anxiety, and attitudes, such as job satisfaction and satisfaction with work flexibility, may have impacted the formation of quiet quitters.

Beyond these perceptions and attitudes, there are other behaviours that may be helpful in characterising quiet quitters. The employee turnover field proposes that there are many types of stayers and leavers in organisations. Reluctant stayers want to leave the organisation but feel that they are constrained by social forces, whether from inside and/or outside the organisation, such as children still living at home (Li et al. 2016). Perhaps quiet quitters are a subset of, or transition stage for, reluctant stayers before they find a better alternative and leave. The nature of the relationships between possible social constraints and movement is unclear though, with assertions that some women embracing traditional gender roles would be more likely to give up work or follow relocating husbands (Hom et al. 2012). The reluctance to move due to social constraints such as family, particularly those with children, may therefore constrain the emergence of quiet quitters.

At a more specific level, focusing on the imprinting mechanism that creates quiet quitters suggests that new imprints have to overcome the potentially imprint-weakening effects of other, often overlapping, imprints. That is, earlier sensitive periods may have meant that employees were exposed to imprints (Simsek et al. 2015), such as some educational experiences, bosses, and leaders, as well as occupational, sectoral, and professional enculturation. Therefore, employees who have been through multiple sensitive periods may have multiple overlapping imprints with their different generations of characteristics (Marquis and Tilcsik 2013).

Indicators of those other imprints that may prevent the occurrence of new imprints such as quiet quitting could include educational and professional enculturation within occupations. Educational and professional imprinting associated with the worker’s occupation may be quite strong, particularly during receptive times such as when at university or when passing through accreditation processes that are often conducted by professional associations. The potential differential impact of occupations on the creation of quiet quitters may also be reflected by how many occupations, except those in health services, were often forced to stay home during the pandemic (Formica and Sfodera 2022). There are also other characteristics that vary by occupation, such as which sector they are in, whether they have a supervisory role, and the strength and activity of trade unions or professional associations, that may also prevent the creation of a new imprint. The obstacle of pre-existing imprints can therefore be represented by indicators such as occupation, including whether have a supervisory role, sector, and union or association membership.

Further, as the second part of the name implies, if quiet quitters are thinking of quitting, they first must search for job alternatives and then obtain those alternative jobs. If they have not been searching or do not have the movement capital to be a contender for other jobs, then that could be another hurdle before turnover occurs (Hom et al. 2012). Indicators of the movement capital needed to be a contender for other jobs often anchor around levels of education, although there have been some unusual links between education, gender, and the broader move toward reducing hours at work. For example, across the economy of the USA, the decline in the hours worked has been larger for men than women and larger still for those men with a bachelor’s degree than those with less education (Lee et al. 2024).

Conflating that possible effect though is the fact that younger people in the workforce tend to more educated (Lee et al. 2024). Age has also arisen as an issue in the popular press with concerns that younger workers are disproportionately represented among quiet quitters, despite examinations of that concern suggesting that younger workers consider themselves lucky to be in work (The Economist 2023). Together, these issues suggest that applying for other jobs, education, gender, and age, as well as possible education by gender interactions, may need to be examined to clarify if quiet quitters are able to quit and move to other jobs.

In summary, a variety of issues may have determined which employees were more susceptible and/or which employees were imprinted with quiet quitting. A further complication is that to explore the relationships between these new ways of working and outcomes such as work–life balance and health, longitudinal data are needed (Andrade and Lousã 2021). Therefore, the current study will try to use research design factors, such as focusing on the context where quiet quitters should be easiest to spot (full-time workers from before to after the height of pandemic interventions), to investigate the key determinants and characteristics of quiet quitters as a group.

1.3. The Current Study

The COVID-19 pandemic may have prompted greater reflection on the relative value of time spent at work as opposed to time spent with family, on hobbies, or enjoying one’s personal life. The dramatic and wide-ranging changes that occurred during the height of the pandemic may represent a sensitive period during which workers sometimes changed how they worked. Assessing the decline in hours worked may be complicated for part-time workers because of likely pressures that would be difficult to account for, such as family, financial, or a changing demand for hours to be worked. Reducing hours worked for part-time workers would have a direct impact on income earned, which would act as a part-time worker-specific constraint on becoming quiet quitters and would be a topic of focus for future research.

Therefore, a clearer test of the nature of quiet quitting would be to study full-time employees, where they would still obtain the same overall income irrespective of the hours worked (as long as they remained employed). The delineation of quiet quitters is then by focusing on whether there is a reduction in hours worked across the duration of the main pandemic impacts.

Consequently, this study will delineate the main trajectories of hours worked by full-time employees over time, specifically from before the pandemic set in to a time after when most, if not all, pandemic restrictions had been lifted and many workers were returning to a new normal. Once those trajectories of hours worked have been specified, analyses will determine which characteristics discriminate between the trajectories involving a decline in hours worked relative to the other trajectories of hours worked. Summarising the review above, the characteristics hypothesised to discriminate between the trajectories of hours worked before and after the imprinting period of the pandemic interventions include the following: stress, anxiety, health, job satisfaction, satisfaction with work flexibility, having children, occupation, sector, having a supervisory role, membership of a trade union or professional association, and whether the employees apply for other jobs, their education level, gender, age, and any education by gender interaction.

Looking to the future, the possible continuation of any new norms and behaviours such as quiet quitting will depend on the potency of the COVID-19 situation (used as a control variable in the analyses) as a sensitive period for imprinting, and it may be too early to know just yet how long these imprints will persist. Therefore, beyond the drivers and constraints presented in these hypotheses, it would be informative to also know more about quiet quitters, even if only in descriptive terms. Consequently, after the above hypothesised variables have been tested, more variables are analysed to provide a fuller descriptive characterisation of the quiet quitters. For now, though, workers will tend to be less receptive to changes while they remain outside of a sensitive period, and thus the imprint is likely to linger on.

2. Materials and Methods

The data for this study came from the Household, Income and Labour Dynamics in Australia (HILDA) Survey, a large survey in Australia that aims to be nationally representative, excluding prisons and the military. A variety of processes, including adding top-up cases and recontact procedures are used to monitor and maintain the survey and data’s representativeness (as detailed in Summerfield et al. 2023). The HILDA Survey has the University of Melbourne Ethics approval number 1647030.

The sample used in this study comprised respondents who were effectively full-time employees at both Time 1 and Time 2. That is, the respondents had usually worked more than 35 h a week in 2019 (Time 1) and had also been usually working more than 35 h per week in 2022 (Time 2). In terms of the timing relative to COVID-19, most of the surveys for HILDA are conducted in the second half of the calendar year. That is, the 2019 surveys were mostly completed three to six months prior to when Australia began to implement COVID-19 policies (such as with respect to international travel). Similarly, the 2022 surveys were mostly conducted nine to fifteen months after most COVID-19 constraints were lifted, allowing some time for workers to adapt to their, possibly new, work routines.

Further parameters were used in order to best delineate possible quiet quitters. The sample was further trimmed to those employees with only one job and who were not in a nursing home or similar, excluding those with time off work from a workers’ compensation claim in the last year, nor those who were independent contractors and freelancers, and retaining those who were aged greater than 24 years and less than 59 years, in order to exclude those actively contemplating retirement or still at a more formative stage of their career. The final, core sample had 2693 full-time employees matched longitudinally from Time 1 (2019) to Time 2 (2022).

2.1. Measures

The measures detailed below are grouped in terms of the variables used to select the sample and those variables in the multinomial regression, followed by a set of variables used in descriptive post hoc analyses. One of the criteria used to delimit the sample, but not used in later analyses, is that all casual workers at either Time 1 or Time 2 were excluded.

To derive the target variables, participants were asked, at both Time 1 and Time 2, “Including any paid or unpaid overtime, how many hours per week do you usually work in all your jobs? This includes any work performed at the workplace and at home. Don’t include time on-call”. However, if they indicated that their working hours varied, the participants were asked to estimate their average hours worked per week over the last four weeks. The answers were combined to derive an indicator of the number of hours usually worked. Those with more than 35 h a week for both Time 1 and Time 2 were selected as being full-time.

A sample selection and predictor variable was age. Age at Time 1 was obtained by asking respondents to indicate their “Age last birthday at the 30th June 2019”. Age at Time 2 was similarly asked relative to 30 June 2022. A further categorical variable was created from age at Time 1 by grouping ages up to and including 27 (1), 28 to 34 inclusive (2), 35 to 41 inclusive (3), and 42 years and over (4) to reflect generational considerations and career stages and to ensure that there were enough respondents in each grouping.

Further Predictor Variables

Work-oriented job stress or anxiety at Time 1. Two items asked about job-related stress or job-related anxiety (per Summerfield et al. 2023). The stress item stated “[M]y job is more stressful than I had ever imagined”. The anxiety item reflected the worries and fears of the employee: “I fear that the amount of stress in my job will make me physically ill”. Each of these two items was rated from strongly disagree (1) to strongly agree (7). For the job stress question, the responses were grouped as 1–3 (less stressed), 4 (neutral), and 5–7 (stressed). The job anxiety responses were grouped 1–3 (less anxious), 4 (neutral), and 5–7 (anxious).

Job satisfaction. Participants indicated from 0 (totally dissatisfied) to 10 (totally satisfied) for the question “All things considered, how satisfied are you with your job?” For the analyses below, the responses were grouped into less satisfied (0–6), moderately satisfied (7–8), and very satisfied (9–10).

Satisfaction with the flexibility of the job. Participants were asked “in your job, how satisfied or dissatisfied are you with: [T]he flexibility available to balance work and non-work commitments?” Responses were from 0 (totally dissatisfied) to 10 (totally satisfied). For the analyses below, the responses were grouped into less satisfied (0–6), moderately satisfied (7), and very satisfied (8–10).

Global general health. At Time 1, global general health was assessed using the SF-36’s global question (Ware et al. 2000): “In general, would you say your health is” with the response options of excellent, very good, good, fair, and poor. The responses were coded such that very good and good were combined, and that fair and poor were combined.

Resident children at Time 1. The respondents’ answers across three related questions were used to determine whether they had children residing with them, coded as 1 for yes and 2 for no. The three prompting questions asked about children who lived with them at least half of the time or were at a non-private dwelling, such as boarding school, but would usually otherwise be with them, or resident step/foster/grandchildren with no resident natural/adopted parent.

Occupational group. Participants were asked “What kind of work do you do in this [your main] job? That is, what is your occupation called and what are the main tasks and duties you undertake in this job? Please describe fully.” The 2006 Australian and New Zealand Standard Classification of Occupations (ANZSCO) classification system was applied to respondents’ text responses and the occupation families were grouped as follows: (0) Managers, (1) Professionals, (2) Technicians/Trades, (3) Community service, clerical, and administration, sales, (4) Machine operators/drivers, and (5) Labourers.

Employer’s sector. Participants were asked “which one of these categories best describes your employer?” Responses were coded as (1) Private for profit (which included ‘Other commercial” employers), (2) Government, whether a government business enterprise or commercial statutory authority or other government organisation, such as a public service department, local councils, schools, and universities, and (3) Private sector “not-for-profit” organisation.

Having a supervisory role at Time 1. Respondents were asked “[A]s part of your job, do you normally supervise the work of other employees?” The answers were coded as yes (1) and no (2).

Union/association member. Respondents were asked yes/no questions of, first, whether they belonged to a trade union, and then, if they had answered no, whether they belonged to any other union or employee association where that employee association serves their work-related interests (excluding associations that accord members professional status in their field)? The responses to these two questions were then combined such that anyone indicating yes to either question was coded as yes, and those indicating no to both questions (or the few people indicating do not know) were coded as no.

Applied for a job with a new employer in last two years at Time 2. Respondents were asked “[H]ave you applied for a job with a new employer at any time in the past two years? The answers were coded as yes (1) and no (2). Although this question was not asked at Time 1, it does cover the two-year time period prior to its assessment.

Education. At Time 1, participants were asked the highest year of school they had completed and whether, excluding hobby or recreation courses, they had completed further qualifications after leaving school. Across the responses, the highest level of education was coded and then grouped as follows: Postgraduate (PG) degree such as a Master’s degree or doctorate (Graduate and PG degrees), Baccalaureate/Bachelor’s degree or Graduate Diploma, Certificate 3 or 4, Diploma or Advanced Diploma (Less than or equal to/LTE Diploma), or Year 12 of school or less.

Sex. Respondents indicated their sex, coded 1 for male and 2 for female. The responses given were identical at Time 1 and Time 2.

Education by sex interaction. To reflect the specific concerns of some economists about males with less than university education reducing their contributions to the workforce, an interaction variable was created from the sex and education variables, but with a simpler grouping of education level as having either a university degree (or higher) versus not having as high an education as a full three or four year university degree. There were four resulting combinations, although due to the overlap with its above component variables, not all these combinations played a direct part in the multinomial regression. The four combinations were as follows: Males with a sub-bachelor degree, males with greater than or equal to a bachelor’s degree, females with sub-bachelor degree, and females with greater than or equal to a bachelor’s degree.

COVID-19 impact—a control variable. Respondents answered two questions about the impact of COVID-19 on their life. The first question asked “[T]hinking about your life today, how much has your life changed because of the coronavirus pandemic?” This first COVID-19 impact question was answered from ‘to a great extent’, ‘to a moderate extent’, ‘a little’, to ‘not at all’. The second COVID-19 impact question then read “[A]nd has that change been for”, with response options of better, worse, or neither better nor worse. Combining these two variables across each other led to 10 combinations of responses as follows: the three degrees of extent by the whether better, worse, or neither, along with the initial response option of not at all. For the analyses below, the 10 combinations were grouped as ‘Greatly Worse’ (1), ‘Moderately worse’ (2), all of the combinations with neither and the not at all respondents were coded as ‘Neither’ (3), ‘Little better’ (4), or (5) ‘Moderately to greatly better’.

2.2. Extra Variables Used in the Post Hoc Analyses for Further Descriptive Characterization

Tenure with current employer at Time 2. Respondents indicated how long they had worked for their current employer in either years or, if less than one year, in weeks. The responses were then combined into a variable representing their tenure with their current employer in a decimalized number of years.

Time in occupation at Time 2. Respondents indicated how long they had worked in their current occupation, including time spent in the same occupation with previous employers or in previous businesses in either years or, if less than one year, in weeks. The responses were then combined into a variable representing their time in their current occupation in a decimalized number of years.

Likelihood of quitting in next 12 months at Time 2. Participants were prompted to think about their employment prospects over the next 12 months. Then, respondents were asked to indicate what they thought: “is the per cent chance that you will leave your job voluntarily (that is, quit or retire) during the next 12 months?”

Likelihood of losing job in next 12 months at Time 2. Participants were asked to indicate “What do you think is the per cent chance that you will lose your job during the next 12 months? (That is, get retrenched or fired or not have your contract renewed)”.

Likelihood of getting a job as good at Time 2. An indication of their personal view of their movement capital was obtained by asking participants to think about their chances of getting a job as good as their current job. That is, participants were asked “If you were to lose your job during the next 12 months, what is the per cent chance that the job you eventually find and accept would be at least as good as your current job, in terms of wages and benefits?” A small number of cases (n = 14) indicated that they would not seek another job, confirming that most of those on the brink of retirement had been excluded from the sample and were excluded from the post hoc analyses.

Job search in last 4 weeks at Time 2. Participants were asked “At any time in the last 4 weeks have you actually looked for a new job?” Respondents indicated yes or no.

Marital status. Participants were asked at Time 1 “which of these best describes your current marital status? And by “married” we mean in a registered marriage.” The responses were grouped as follows: (1) Married (in a registered marriage), (2) Separated, but not divorced, divorced, or widowed, (3) never married but living with someone in a relationship, and (4) never married and not living with someone in a relationship.

Language other than English spoken at home (LOTE) at Time 2. Participants were asked “Do you speak a language other than English in this home?” Respondents indicated yes or no.

Physical functioning, general health, and mental health at Time 2. The respondents’ scores for these three facets of health were measured using the relevant scales from the SF-36. The scores used were the 0–100 scale scores generated by applying the transforms in Ware et al. (2000).

3. Results

The sample had 2693 full-time employees matched longitudinally from Time 1 (2019) to Time 2 (2022). The main indicator of quiet quitters is their decline in hours worked over time. Bear in mind that all of the sample were full-time at both Time 1 and Time 2. To derive the key target variable, the usual number of hours worked per week at Time 1 and Time 2 were obtained, and then the longitudinal combinations were mapped as trajectories.

Given that there is some variability in what may be considered as an employee’s minimum hours, a range band was used where the employee usually worked 35–40 h per week. That is, a category representing the minimum hours that most full-time jobs require were those employees usually working in the range from 35 to 40 h per week, with most in the range of 37.5 to 40 h. Further bands of hours worked were constructed for 41 to 50 h and for those greater than 50 h per week. The trajectories of hours worked over time were then developed by cross tabulating the bands of hours worked at Time 1 with the bands of hours worked at Time 2. The resulting nine combinations ranged from those who had stayed at the minimum hours required through to those who had stayed working greater than 50 h per week. The nine permutations were then condensed down to five groups to reflect similarities in possible representations of quiet quitters (and to have sufficient numbers in each group to be able to analyse).

Those with the largest decline in working hours over time (n = 157) were those working greater than 50 h per week at Time 1 who had reduced their hours by more than 20% (n = 110) or down to the minimum band (n = 47). Those with the next largest decline in hours worked were those in the 41 to 50 h band, who had reduced their hours to the minimum required band (n = 311). The third grouping had the most consistent working hours trajectory and were those who had worked the minimum hours required at both Time 1 and Time 2 (n = 1223), which was used as the reference group for comparison with the other longitudinal working hours groupings in the multinomial regression analyses below.

The remaining two bands increased their hours worked over time. The fourth grouping were those who increased their hours worked from near the minimum required to 41 to 50 h per week (n = 266) or stayed working 41 to 50 h per week over time (n = 480). The final grouping comprises those who ended up working more than 50 h per week (n = 260), consisting of those who started near the minimum (n = 34), started on 41 to 50 h per week (n = 105), or stayed working more than 50 h over time (n = 121).

The variables analysed in the multinomial regression are summarised for each of the target variable’s categories in

Table 1. Descriptive statistics for the regression variables by hours-worked trajectory from Time 1 to Time 2 (n = 2693).

Categorical Variables (n)	Dramatic Reduction in Hours	Slight Drop to Minimum	* Stayed Near Minimum	Increased to/Stayed 40–50 h	Ended Up Doing >50 h
Stressful Job
- Stressed	74	124	306	241	109
- Neutral	26	75	272	159	47
- Less stressed	56	112	643	345	104
Anxious Job—Anxious	44	66	150	124	63
- Neutral	30	38	151	87	49
- Less anxious	82	207	920	534	148
Global general health
- Excellent	9	47	164	90	28
- Very good to good	132	244	943	588	198
- Fair to poor	15	20	114	67	34
Job Satisfaction
- Less satisfied	40	59	152	101	45
- Moderately satisfied	90	189	670	414	157
- Very satisfied	26	63	399	230	58
Flexibility Satisfaction
- Less satisfied	72	96	251	200	103
- Moderately satisfied	30	59	198	139	42
- Very satisfied	54	156	772	406	115
Resident children
- Yes	83	135	598	402	156
- No	73	176	623	343	104
Occupation
- Managers	46	81	158	179	78
- Professionals	45	102	403	239	75
- Technicians/trades	25	43	177	102	35
- Community service, clerical, and admin sales	22	60	367	142	27
- Machine operators, drivers	15	14	62	60	36
- Labourers	3	11	54	23	9
Sector
- Private, for-profit	100	200	703	504	191
- Government	45	92	421	185	57
- Private Not-for-profit	11	19	97	56	12
Supervisory role
- Yes	102	207	564	472	173
- No	54	104	657	273	87
Union/Association Member
- Yes	51	64	309	190	79
- No	105	247	912	555	181
Applied for new job in past 2 years
- Yes	59	124	338	205	74
- No	97	187	883	540	186
Highest Education
- Postgraduate	19	38	121	103	32
- Baccalaureate or Graduate diploma	52	107	394	245	76
- Tertiary certificate or Diploma	53	101	430	253	106
- High school	32	65	276	144	46
Sex
- Male	102	197	663	500	186
- Female	54	114	558	245	74
Education by Sex
- Male sub-bachelors	67	126	421	301	128
- Male GE Bachelors	35	71	242	199	58
- Female sub-bachelors	18	40	285	96	24
- Female GE bachelors	36	74	273	149	50
Age
- LE 27	21	42	207	109	29
- 28–34	36	97	301	167	59
- 35–41	37	44	238	167	52
- 42+	62	128	475	302	120
COVID-19 Impact
- Greatly worse	13	19	64	57	18
- Moderately worse	11	27	116	76	40
- Neither	61	147	614	361	130
- Little better	47	80	289	177	63
- Moderate to greatly better	24	38	138	74	9
Continuous Variable	Mean (Standard Deviation)
Age last birthday as at 30 June 2019	38.59 (9.529)	38.15 (9.480)	38.22 (9.639)	38.77 (9.271)	40.06 (9.428)

Note: * The middle group are the comparison group for the multinomial regression analyses. GE: Greater than or equal to. LE: Less than or equal to.

. Note that the continuous variable of age is not significantly different across the hours-worked trajectories and thus the grouped age variable was used in the regression instead.

Multinomial regression analyses were performed across the changes in the hours worked categories with stressful jobs, anxious jobs, overall job satisfaction, satisfaction with the flexibility of work, the impact of COVID, occupation, global general health, age, sector, education, education and gender interaction, applying for other jobs, has resident children, being a union/association member, and whether they hold a supervisory role as predictors. All of the predictors, except whether they had applied for a job in the last two years, were measured at Time 1. The analyses were performed using SPSS 29. The longitudinal analyses were conducted following Tabachnick and Fidell (2007) and the best practice guidelines from Osborne (2015), with consideration for the large sample size.

There were 262 missing cases, which were determined to be missing completely at random/MCAR (Little’s MCAR, p = 0.758). A new dataset was created with 20 multiple imputations to impute the values for the cases with missing data. The results were very stable with the same variables standing out, but with some having their significance levels change slightly. The more stable results of the pooled multinomial regression are presented in

Table 2. Characteristics distinguishing between categories of longitudinal changes in hours worked from pooled analyses (n = 2955).

Change in Hours Worked Category [Ref: Stayed at Minimum]	Substantial Drop		Dropped to Minimum		Min Increased Hours to >40		Ended Doing >50 h
Predictors [Reference Category]	B (SE)	Odds Ratio	B (SE)	Odds Ratio	B (SE)	Odds Ratio	B (SE)	Odds Ratio
Stressful Job [Ref: Less stressed]
- Stressed	0.255 (0.249)	1.291	0.591 (0.180)	1.806 *	0.251 (0.134)	1.286 †	0.258 (0.200)	1.294
- Neutral	−0.328 (0.276)	0.720	0.389 (0.180)	1.476 *	0.029 (0.134)	1.029	−0.258 (0.215)	0.772
Anxious Job [Ref: Less anxious]
- Anxious	0.625 (0.270)	1.868 *	0.188 (0.204)	1.207	0.152 (0.161)	1.164	0.593 (0.222)	1.809 *
- Neutral	0.487 (0.266)	1.627 †	−0.244 (0.216)	0.783	−0.119 (0.158)	0.887	0.445 (0.220)	1.561 *
Global general health [Ref: Fair to poor]
- Excellent	−0.403 (0.465)	0.668	0.755 (0.300)	2.127 *	0.014 (0.214)	1.014	−0.089 (0.309)	0.915
- Good to very good	0.254 (0.307)	1.290	0.490 (0.256)	1.632 †	0.062 (0.166)	1.064	−0.200 (0.225)	0.819
Job Satisfaction [Ref: Very satisfied]
- Less satisfied	0.351 (0.315)	1.421	0.332 (0.236)	1.394	−0.164 (0.177)	0.849	−0.082 (0.262)	0.921
- Moderately satisfied	0.304 (0.241)	1.356	0.325 (0.162)	1.384 *	−0.077 (0.111)	0.926	0.226 (0.177)	1.253
Flexibility satisfaction [Ref: Very satisfied]
- Less satisfied	1.105 (0.223)	3.020 **	0.424 (0.170)	1.528 *	0.439 (0.129)	1.551 **	0.814 (0.179)	2.256 **
- Moderately satisfied	0.598 (0.248)	1.818 *	0.260 (0.179)	1.297	0.300 (0.132)	1.349 *	0.239 (0.203)	1.270
Resident children [Ref:—No]—Yes	0.020 (0.188)	1.020	−0.382 (0.139)	0.683 *	0.014 (0.103)	1.014	0.183 (0.152)	1.201
Occupation [Ref: Labourers]
- Managers	1.647 (0.642)	5.190 *	0.766 (0.367)	2.151 *	0.673 (0.268)	1.961 *	0.885 (0.385)	2.423 *
- Professionals	0.960 (0.647)	2.612	0.300 (0.368)	1.350	0.316 (0.264)	1.372	0.244 (0.391)	1.277
- Technicians/trades	1.023 (0.640)	2.783	0.018 (0.371)	1.018	0.041 (0.266)	1.042	−0.062 (0.388)	0.940
- Community service, clerical and admin., sales	0.472 (0.644)	1.604	0.057 (0.360)	1.059	0.002 (0.259)	1.002	−0.484 (0.398)	0.616
- Machine operators, drivers	1.388 (0.670)	4.009 *	0.212 (0.423)	1.237	0.592 (0.295)	1.807 *	1.142 (0.401)	3.132 *
Sector [Ref: Not-for-profit]
- Private for-profit	0.186 (0.347)	1.204	0.475 (0.276)	1.608 †	0.149 (0.179)	1.161	0.671 (0.310)	1.957 *
- Government	−0.160 (0.360)	0.852	0.219 (0.285)	1.245	−0.411 (0.188)	0.663 *	−0.242 (0.327)	0.785
Supervisory [Ref: No]—Yes	0.529 (0.189)	1.697 *	0.639 (0.137)	1.895 **	0.526 (0.099)	1.692 **	0.671 (0.154)	1.955 **
Union/association member [Ref:—No]—Yes	0.410 (0.206)	1.506 *	−0.192 (0.168)	0.825	0.158 (0.117)	1.171	0.345 (0.169)	1.412 *
Applied for job in last 2 years [Ref: No]—Yes	0.380 (0.183)	1.462 *	0.535 (0.135)	1.708 **	0.007 (0.106)	1.007	0.119 (0.155)	1.126
Education [Ref: High school]
- Postgraduate	0.526 (0.424)	1.693	0.402 (0.306)	1.495	0.733 (0.225)	2.081 *	1.088 (0.362)	2.969 *
- Baccalaureate or Graduate Diploma	0.355 (0.372)	1.426	0.410 (0.261)	1.507	0.399 (0.194)	1.490 *	0.734 (0.319)	2.083 *
- Tertiary Certificate or Diploma	−0.003 (0.244)	0.997	0.036 (0.178)	1.036	0.114 (0.132)	1.121	0.393 (0.197)	1.481 *
Education by Sex Interaction
- Male Sub-bachelors	0.897 (0.295)	2.452 *	0.814 (0.204)	2.258 **	0.778 (0.150)	2.177 **	1.036 (0.245)	2.819 **
- Male GE bachelors	0.201 (0.264)	1.223	0.103 (0.194)	1.109	0.318 (0.142)	1.375 *	0.097 (0.223)	1.102
Age band [Ref: 42+]
- LE 27	−0.032 (0.297)	0.969	−0.488 (0.216)	1.614 *	0.019 (0.155)	1.019	0.043 (0.233)	1.044
- 28–34	−0.135 (0.233)	0.874	0.012 (0.162)	1.012	−0.082 (0.127)	0.921	−0.204 (0.187)	0.816
- 35–41	0.079 (0.225)	1.083	−0.483 (0.189)	0.617 *	0.011 (0.126)	1.011	−0.192 (0.188)	0.825
COVID impact [Ref: Moderately+ better]
- Greatly worse	0.022 (0.382)	1.022	0.279 (0.321)	1.322	0.547 (0.234)	1.728 *	1.409 (0.413)	4.094 **
- Moderately worse	−0.871 (0.394)	0.418	0.003 (0.281)	1.003	0.183 (0.207)	1.201	1.503 (0.369)	4.494 **
- Neither	−0.578 (0.263)	0.561	0.058 (0.211)	1.060	0.144 (0.160)	1.155	1.015 (0.335)	2.759 *
- Little better	−0.047 (0.280)	0.954	0.205 (0.232)	1.228	0.132 (0.176)	1.141	0.989 (0.352)	2.688 *
Intercept	−4.599 ** (0.864)		−4.139 ** (0.593)		−1.921 ** (0.405)		−4.793 ** (0.670)

Note: ^† p < 0.10, * p < 0.05, ** p < 0.01. Ref. = reference/comparison category. Reference categories were set to 0. The interaction does not have an explicit Reference group given its comparisons with component variables. GE: Greater than or equal to. LE: Less than or equal to.

.

The multinomial regression had -2LL = 6653.905, a significant improvement over the base model’s -2LL of 7245.936 (χ2 (132) = 592.032, p < 0.001), showing that the predictors, as a group, distinguished across the categories of the target variable. The overall multinomial model had so-called pseudo R-square measures, by type, of Cox and Snell = 0.197, Nagelkerke = 0.211, and McFadden = 0.081. The logit estimates and odds ratios of the predictors are summarised in Table 2. The comparison group for the overall regression was the most stable group in terms of hours worked and those employees who had stayed at or near the minimum hours at both Time 1 and Time 2.

The post hoc analyses in

Table 3. Post hoc descriptive results by category of change in hours worked.

Categorical Variables (n)	Dramatic Reduction in Hours	Slight drop to Minimum	Stayed Near Minimum	Increased to/Stayed 40–50 h	Ended Up Doing >50 h
Works from home at all, Time 2
- Yes	77	147	527	358	125
- No	79	164	694	387	135
Taken paid annual leave in last year, Time 2
- Yes	136	280	1116	673	227
- No	20	31	105	71	33
LOTE at Time 2
- Yes	15	29	158	60	17
- No	141	282	1063	685	243
Marital Status at Time 1
- Married	86	151	609	396	155
- Separated, Divorced, Widowed	19	22	110	80	28
- Partnered	25	71	254	147	39
- Single	26	67	248	122	38
Sector at Time 2
- Private, for-profit	100	192	693	508	189
- Government	44	96	424	176	48
- Private not-for-profit	12	23	104	61	23
Job search in last 4 weeks at Time 2
- Yes	28	48	182	113	53
- No	128	263	1039	632	207
Chances of voluntarily quitting, Time 2
0%	69	136	566	341	114
1–10%	25	63	268	173	44
22–32%	18	46	145	88	30
33–50%	23	44	123	92	43
>50%	21	22	119	51	28
Chances lose job, next 12 months, Time 2
0–10%	120	257	1019	629	208
11–30%	20	33	123	79	26
31+%	16	21	79	37	26
Find as good a job, Time 2
0–29%	29	31	157	104	33
30–50%	16	63	288	149	41
51–99%	68	143	534	341	108
100%	43	74	242	151	78
Continuous Variables	Mean (Standard Deviation)
Tenure with current employer at Time 2 *	7.19 (8.08)	6.96 a (7.14)	8.00 (7.60)	9.05 b (8.55)	9.06 b (8.84)
Time in occupation at Time 2 *	8.41 a (9.02)	9.18 (8.75)	10.04 (8.86)	10.73 b (9.73)	11.08 b (9.57)
General health scale at Time 2	68.61 (18.222)	68.38 (17.742)	69.13 (17.999)	68.00 (18.310)	68.07 (18.553)
Mental health scale at Time 2	72.62 (16.671)	72.27 (16.126)	72.85 (16.685)	72.77 (15.884)	70.98 (17.584)
Physical functioning scale at Time 2	90.68 (15.114)	91.28 (16.513)	92.27 (13.486)	92.14 (14.247)	90.33 (16.719)

Note: * The continuous variables where the ANOVA was significant, with a relative to b denoting the cells that the post hoc tests indicated were different from each other. Note that the analyses here are for descriptive purposes only and that the sample sizes may differ by variable. LOTE: Language other than English spoken at home.

are for descriptive purposes only to help inform the characterisation of the categories of employees grouped by their trajectories in hours worked pre- and post-pandemic constraints. Note that for the last few rows, there were no significant differences for any of the continuously scaled Time 2 health variables. The distinguishing characteristics per group have been summarised in

Table 4. The relatively distinguishing issues and descriptive characteristics across changes in the work hours categories.

Quiet Quitters (c. 5.79%)	Getting By/Counterculture (c. 11.55%)	Functionary/Necessity (c. 45.34%)	Qualified, Extended Hours (c. 27.66%)	Long Hours (c. 9.65%)
Distinguishing Variables
Supervisory	Supervisory	Non-supervisory	Supervisory	Supervisory
Anxious	Stressful	Less anxious		Anxious
Very dissatisfied with flexibility	Dissatisfied with flexibility	Satisfied with flexibility	Dissatisfied with flexibility	Dissatisfied with flexibility
Applying for jobs	Applying for jobs		More likely be in government	More likely be in private for-profit
	Males with sub-bachelors	Less qualified	More qualified, especially for ≥ Bachelors	More likely have post high school education
More likely be managers or machine operators/drivers	More likely to be managers	Slightly more likely to be female; Less likely to be managers	Slightly more likely to be male; More likely be managers or machine operators/ drivers	More likely be managers or machine operators/drivers
Slightly more likely to be in union/ association		Slightly less likely to be in union/association		Slightly more likely to be in union/association
	More of young; healthier; no children; moderately satisfied with job	COVID-19 impacted life for better	COVID-19 impacted life greatly worse	COVID-19 impacted life for worse
Descriptive Characteristics
Less time in occupation	Less tenure with employer		Longer tenure with employer and longer time in occupation	Longer tenure with employer and longer time in occupation
More employees very likely to quit in next year	Fewer employees very likely to quit in next year		Fewer employees very likely to quit in next year	More employees very likely to quit in next year
More likely to lose job soon			Less likely to lose job soon	More likely to lose job soon
Curvilinear: Some feel very unlikely get a job as good, many feel very likely to get good job				Many feel very likely get as good a job
Slightly more likely to have searched for job in last 4 weeks		Slightly less likely to have searched for job in last 4 weeks		More likely to have searched for job in last 4 weeks
		More likely to be LOTE		Less likely to be LOTE
More likely to be separated, divorced, or widowed	Less likely to be separated, divorced, or widowed			Less likely to be single, less likely to be partnered yet unmarried
Less likely to have taken annual leave in last year		More likely to have taken annual leave in last year		Less likely to have taken annual leave in last year

Note: These qualitative summaries are slight emphases that were somewhat distinguishing—and are not absolutes. LOTE: Speak a language other than English at home.

.

4. Discussion

This study investigated a set of trajectories of the hours worked by full-time employees over time, specifically from before the pandemic set in to a time after most pandemic-era restrictions had been lifted and many workers were returning to a new normal. The analyses then outlined characteristics that distinguished each trajectory from other trajectories of hours worked.

The following text will discuss the patterns of results across the variables hypothesised to characterise different approaches to work, such as quiet quitting. The characteristics hypothesised to discriminate between the trajectories of hours worked before and after the imprinting period of the pandemic interventions include stress, anxiety, health, job satisfaction, satisfaction with work flexibility, having children, educational and professional enculturation within occupations and sectors, having a supervisory role, membership of a trade union or professional association, and whether the employees search for and apply for other jobs, their education level, gender, age, and any education and gender interactions.

There was no single variable that characterised how the different work approaches may have emerged. There were patterns of variables characterising the trajectories of hours worked. There were two trajectories of reduced hours worked that may be enacting variations of quiet quitting. The most clear-cut trajectory of reduced hours may indicate a new imprint of prototypical quiet quitters. The second trajectory of reduced hours worked may be a form of quiet quitting that has a pattern of characteristics which suggests that they are an emerging counterculture, differentiated from other groups by having a higher proportion of younger workers, but consisting of workers from all ages.

Commentary that had characterised quiet quitters in terms of a single issue, such as job satisfaction or age, was not supported. Job satisfaction was not a distinguishing characteristic, although the employees’ satisfaction with the flexibility of their work was a distinguishing characteristic. However, dissatisfaction with flexibility characterised all of the groups except the comparison group, not just those who had reduced their working hours.

Consequently, the characterisation of the groups needs to move toward considering patterns of distinguishing issues, not just univariate approaches. The first two categories seem to be most akin to prototypical quiet quitters. For the interim, the labels applied to the groupings are as follows: the prototypical quiet quitters (n = 156; 5.79%), those who are getting by or may represent the beginning of a counter-culture (n = 311; 11.55%), the comparison group where the pattern of characteristics may indicate a functional approach top their job (Functionary/Necessity, n = 1221; 45.34%), the qualified working extended hours (n = 745; 27.66%), and those working long hours (n = 260; 9.65%).

The quiet quitters were characterised by anxiety and stress, being very dissatisfied with the flexibility of their work practices and thus applying for jobs. Their descriptive characteristics included job searching and a strong indication that they were likely to quit in the next year, although a subset of them were concerned that if they lost their job, they might not be able to secure a job as good as their current job in a timely manner. The quiet quitters had spent less time in their occupation, which may mean that they had been subject to less enculturation and/or that their training may have been under modern, rather than older, norms.

The getting by/counterculture category included slightly more males with sub-bachelor education and slightly more younger people, who were also healthier and less likely to have children. Oddly, that category was more moderately satisfied, perhaps, as The Economist (2023) suggested, thankful that they had a (full-time) job at all, which may be why they were less likely to quit in the next year, even though they had less tenure with their employer. However, they were keeping a watch for job opportunities and occasionally applying for jobs.

The third group, which acted as the comparison group in the multinomial regression, seemed to be employees who were working because they had to, were more likely to have children residents with them at home, were more likely to speak a language other than English (LOTE) at home, and were staying where they were, and it appears that the COVID-19 responses may have made their work more flexible. However, for many of the characteristics, they were the lowest scoring. For example, one of the most consistently distinguishing variables across categories was whether the person had a supervisory role, indicating that the comparison group were less likely to be supervisory and all four other categories were more likely to be supervisory.

The fourth and fifth categories increased their working hours over time and are consequently less of a focus for an investigation of quiet quitting. However, there are some interesting messages across the categories, including one or both of the categories of employees who increased their hours or maintained long hours.

Although education is often considered an indicator of movement capital (e.g., Hom et al. 2012), the most qualified three groups were, from the most qualified, the qualified workers working extended hours, the longer hours category, and the quiet quitters. The most qualified category (the qualified workers working extended hours) was not as likely to be moving soon, as indicated by them not being as likely to be searching for or applying for jobs. They have longer tenure with their employer, are more likely to work for governments, and have spent a longer time in their occupation. Perhaps their qualifications and government roles make them feel that they like where they are or reflect their roles as technocrats, although there is some tension building, given that COVID-19 impacted their life for the worse and that they are dissatisfied with their flexibility at work. Similarly, it may just be that their workplaces, which were more likely to be governments, encouraged and were supportive of further, including postgraduate, education.

For the more prototypical quiet quitter group, there was an unusual pattern in the data regarding one of the indicators of movement capital—how likely they thought they were to obtain a job as good as their current job if they left their work. Analyses of this descriptive variable suggested a curvilinear relationship within quiet quitters where some felt the chances were very unlikely that they would get a job as good, but, at the other end of the scale, there were many quiet quitters who felt that it was very likely that they could get a job that was just as good. That is, some of the quiet quitters appeared to be reluctant stayers and could quit as soon as key constraints were perceived as being lifted, and many were confident that they could get a job just as good as their current job.

These subgroups within the prototypical quiet quitter group may reflect how some organisations enacted new working practices akin to Taylorist principles (Aroles et al. 2021) across expanded roles (Andrade and Neves 2024), whereas other more employee-oriented flexible working practices may have been more conducive to balancing work and life demands (Ajzen and Taskin 2021), to which many employees have positive responses (Stamos and Kotsopoulos 2024). The differences between these organisations and their consequent impact on employees may explain part of the complexity of the drivers of the phenomena of quiet quitting.

In contrast, those in the fifth category, those who had stayed working long hours, felt that it was very likely that they would get as good a job if they left. However, in looking at the pattern of results, those in the fifth group may also have been working the long hours because it helped them to feel that they were more likely to keep their job.

An indication of the control that the employees felt they had about staying or leaving, often used to consider whether they are reluctant or enthusiastic stayers or leavers in turnover research (e.g., per Li et al. 2016), was their perception of how likely they were to be fired in the near future. In a descriptive sense, the quiet quitters and those continuing to work long hours felt that it was more likely they would lose their job soon, whereas the more qualified workers working long hours in the fourth category felt less likely to lose their job soon.

In a similar spirit, part of the pattern across characteristics repeatedly showed that the quiet quitters and those continuing to work long hours have several warning signs of their possible imminent departure. Some indicators such as applying for jobs were obvious, but others such as not taking annual leave in the last year may have been because the quiet quitters and those working long hours have found a good pace, do not need to take annual leave, and/or are not using annual leave so that they are paid out more when leaving the organisation. When looking at these results from more of a configural point of view, it may be that those working long hours felt they may lose their job soon and that may be why they are working the long hours and not taking annual leave and applying for jobs.

Conversely, relative to the turnover literature (e.g., Li et al. 2016), three of the groups—all except those working out of necessity and the qualified working extended hours—appeared to include reluctant stayers, not just the quiet quitters. Social forces such as resident children, job opportunities, their movement capital, and perceptions of the flexibility of their work may be part of the interplay among characteristics that led to their chosen trajectory of hours worked over time. Thus, the person-centred approach of considering configurations of characteristics (Sinclair et al. 2005) was informative and may assist further in understanding the phenomenon of quiet quitting.

The large number of significant variables distinguishing between the trajectories of hours worked over time in the multinomial regression confirms the utility of the configural approach and indicates that a univariate approach to understanding quiet quitting is inappropriate. In turn, the configurations that emerged are likely to have been brought about by a complex set of interacting pressures, such as those associated with imprinting.

4.1. Imprinting as a Likely Mechanism of Quiet Quitting

The findings from this study indicate that it is not just the COVID-19-related imprinting that is developing new imprints such as quiet quitting, but also the clashes between old and new imprints that determine how widespread any new imprints may be. Further, perhaps a key driver of the potential creation of new imprints is the anxiety associated with sensitive periods.

The two categories of workers who reduced their working hours over time appear to have done so in an effort to resolve some of the anxiety associated with the pandemic. Both categories with reduced hours were middling on whether COVID-19 had had a negative or positive impact on their lives (at Time 2), in contrast to the necessity group (for better) or the two groups that had increased or maintained higher hours (where they indicated that COVID-19 had impacted them for the worse).

During the worst of the COVID-19 period, the two groups reducing their hours may have been more susceptible and developed characteristics that reflect the calls for flexibility from their environment (per Marquis and Tilcsik 2013). To some extent, the two groups reducing their hours may have been resolving some of the anxiety of that time by changing their mindset and behaviours. That is, sensitive periods may often entail anxiety and cognitive unfreezing (per Schein 1971), which makes susceptible employees adopt a new imprint in order to reduce some of that anxiety.

Perhaps both of the groups with reduced working hours are quiet quitters, albeit different forms of quiet quitters—the prototypical quiet quitter and those acting as a counterculture to the mainstream norms. At this time, it may be too early to know whether these quiet quitting imprints will persist. One set of issues that may act to weaken these new imprints is if there are older imprints that counter the new imprints (Simsek et al. 2015).

Educational and professional enculturation within occupations may be an alternative set of imprinting that could be working against the setting of the new quiet quitting imprints. Occupational norms, which may include educational norms, may be acting to block the quiet quitting imprints. For example, while the quiet quitters in Table 4 are slightly more likely to be in a trade union or association that acts to protect their interests at work, so are those continuing to work more than 50 h per week. The strength of the occupational enculturation and norms may be more powerful than any effects of the like of trade unions, with the quiet quitters having less time in the occupation, whereas those continuing to work long hours had longer times in the occupation, with any occupation-related education therefore being of an earlier time and having been reinforced for longer.

4.2. Practical Implications

The emergence of quiet quitters is an indication of the importance of hours worked, which is an issue that has not been receiving as much attention with managers and human resources professionals as some other issues. Although employee turnover may be important, especially if it is high for your organisation, at a broader level, decreases in the hours worked after the period of pandemic lockdowns may underpin key issues such as the low unemployment rate and tight labour market. Furthermore, the decline in hours worked had more of an impact than the so-called Great Resignation (Lee et al. 2024). Thus, the broad lesson for policy makers and practitioners is that we need to pay more attention to the hours worked of employees.

There are also a variety of implications arising from the more specific findings of this study. For example, of the two groups that may be enacting a form of quiet quitting, one may be an indication of a new imprint, whereas the other may be an emerging counterculture, popularised by some younger workers, but consisting of workers from all ages.

Both of the quiet quitting groups, and the group working long hours, may be disproportionately thinking of leaving in the next year. To prevent these employees leaving, managers will need to enhance the flexibility of work practices and address organisation-specific drivers of anxiety where possible and feasible. Further, the group working long hours are taking a toll from those longer hours, which will have health implications in the long run.

A possible key limitation of the results reported here, and an important consideration in seeking to understand the pattern of results across studies, is that this study was conducted in Australia. The Australian context has a relatively supportive welfare system, with relatively universal health care (OECD 2024). Other limitations such as common method variance are less likely to be a problem with this study due to the longitudinal nature of the data and the substantial period of time between surveys. Further, the dataset used above is a subset of a dataset that uses a variety of processes to maintain a strong degree of representativeness of the sample (see details in Summerfield et al. 2023).

5. Conclusions

In trying to better understand the phenomenon of quiet quitting this study investigated a set of trajectories of the hours worked by full-time employees over time, especially those who reduced their working hours. The large number of significant variables distinguishing between the trajectories of hours worked longitudinally in the multinomial regression confirmed the utility of the configural approach.

There were one if not two groups of quiet quitters. One group may have been more prototypical quiet quitters, whereas the other group may be part of a growing movement acting as a counterculture to old norms of working long hours. That second group of workers reducing their working hours includes higher proportions of younger people, which may have been overly focused upon by some of the popular press about quiet quitting, despite their group having workers with a range of ages. The prototypical quiet quitters are considering actual quitting and have reduced their hours worked, suggesting that the moniker applies and has now been academically confirmed.

The prototypical quiet quitters are characterised by a pattern highlighting drivers, such as anxiety and stress, and a dissatisfaction with work flexibility, who are also applying for jobs. Descriptively, the prototypical quiet quitters are also more likely to be searching for jobs, are more likely to quit in the next year despite concerns that they may or may not get a job as good as their current job, and were also more likely to feel that they may lose their job soon. Their prototypical quiet quitting imprint may also have taken hold because they have had less time in their occupation, suggesting that they had weaker pre-existing imprints from occupational enculturation to be overcome before forming their new imprint.

Having a potentially sizable group of employees with a new approach to work means that a lot of management and human resources practices may have to be rethought and revised. Managers will have to consider that there are now new groups of employees who may not react like previous employees, who may now need to be specifically accounted for.

The COVID-19-related imprinting may have developed new imprints such as quiet quitting, but there are also going to be pressures between the old and new imprints that will influence how long the quiet quitting imprint lasts. More broadly, the emergence of quiet quitters is an indication of the importance of hours worked as an issue for human resources professionals to be aware of and be ready to address.