The mental health of fly-in fly-out workers before and during COVID-19: a comparison study The mental health of fly-in fly-out workers before and during COVID-19: a comparison study
Gilbert, Jessica M.; Fruhen, Laura S.; Burton, Cindy T.; Parker, Sharon K.
AUSTRALIAN JOURNAL OF PSYCHOLOGY 2023, VOL. 75, NO. 1, 2170280 https://doi.org/10.1080/00049530.2023.2170280 The mental health of fly-in fly-out workers before and during COVID-19: a comparison study a b a a Jessica M. Gilbert , Laura S. Fruhen , Cindy T. Burton and Sharon K. Parker a b Centre for Transformative Work Design, Future of Work Institute, Curtin University, Perth, Australia; School of Psychological Science, Psychology at Work Lab, University of Western Australia, Perth, Australia ABSTRACT ARTICLE HISTORY Received 15 July 2022 Objectives: This study gives an overview of the impact of FIFO work on workers’ mental health Accepted 13 January 2023 before and during COVID-19, using three comparison samples as well as norm data. It provides a timely update on FIFO workers' mental health and how it has been impacted during COVID- KEYWORDS Remote work; wellbeing; Method: Comparisons are conducted with three participant samples, namely two FIFO worker burnout; psychological samples (one before and one during the Covid pandemic) and a purposefully sampled bench- distress; FIFO mark sample, and Australian population norm data on mental health. Constructs included in surveys were psychological distress, burnout, suicide intention, as well as social, psychological, and emotional wellbeing. Results: FIFO workers were found to have worse mental health than the matched benchmark sample, and the Australian norm samples pre-COVID-19. Differences between FIFO workers and the matched benchmark sample persisted for psychological distress and burnout after controlling for demographic factors. Mental ill-health and poor well-being were higher during the COVID-19 pandemic than before. Conclusions: FIFO workers need to be considered an at-risk group for adverse mental health outcomes, and this is even more so the case during COVID-19. Findings are attributable to the experience of FIFO work as well as the demographic character of the workforce. KEY POINTS What is already known about this topic: (1) Research findings on FIFO workers' mental health are mixed. (2) A comprehensive comparison of FIFO worker mental health with the wider Australian population on a range of mental health indicators is needed to provide clarity on this issue. (3) Impacts of COVID-19 on FIFO worker mental health have been anecdotally reported but have to date not been empirically tested. What this topic adds: (1) This study shows that FIFO workers had worse mental health compared to non-FIFO workers before COVID-19 in 2018. (2) It documents differences in FIFO workers’ mental health before (2018) and during Covid (2020). (3) The study’s findings clearly identify FIFO workers as an at-risk group for mental health. The mental health and wellbeing of fly-in fly-out work- are an at risk group for higher levels of psychological ers (FIFO workers) in Australia is a topic of concern distress, loneliness, and suicide risk (Education and amongst researchers, practitioners, policymakers, and Health Standing Committee, 2015; Infrastructure the broader community (Parker et al., 2018). FIFO work- Planning and Natural Resources Committee, 2015). places involve “work in relatively remote locations Yet these inquiries and research report that findings where food and lodging accommodation is provided are mixed, and more research is needed to provide for workers at the work site, but not for their families. clarity around the state of FIFO workers' mental health Schedules are established whereby employees spend to guide practical intervention (Fruhen et al., 2022). In a fixed number of days working at the site, followed by addition, FIFO work has also undergone some key a fixed number of days at home” (Storey, 2001, p. 135). challenges during COVID-19 with travel restrictions, Multiple government inquiries have concluded that quarantine needs and border closures having FIFO workers operate in a unique environment and impacted workers and their usual fluctuations CONTACT Jessica M. Gilbert Jess.Gilbert@curtin.edu.au © 2023 The Author(s). Published by Informa UK Limited, trading as Taylor & Francis Group. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. 2 J. M. GILBERT ET AL. between site and home (Gilbert et al., 2020). These warrants attention and initiatives towards addressing challenges’ impact on FIFO workers' mental health is this issue are hampered by the long-standing debate yet to be investigated. To that end, this study provides around the mixed findings. a timely update on FIFO workers' mental health before Further, research from around the globe has identi- and during COVID-19. fied that COVID-19, while predominantly a physical Across academic studies, findings are mixed regard- health crisis, has had a significant impact on mental ing FIFO workers’ mental health relative to other work- health (e.g., Biddle et al., 2020). As has been the case ers (Fruhen et al., 2022). Out of studies that directly for many others, the working conditions of FIFO workers compare the mental health and wellbeing of FIFO changed significantly in response to the public health workers with workers in other forms of employment, measures that needed to be implemented (Gilbert et al., many have shown that the mental health and well- 2020). Of note, for FIFO workers these measures meant being of FIFO workers is worse relative to others (e.g., often increased social isolation on site, less favourable Bowers et al., 2018; Considine et al., 2017; Henry et al., rosters, and extended periods away from family for 2013; Lester et al., 2015; Sellenger & Oosthuizen, 2017). many (Gilbert et al., 2020; Trinca, 2020). Concerningly, Other studies identify no difference between these the resulting changes in response to COVID-19 affected groups’ mental health (Bradbury, 2011; Clifford, 2009). issues that have previously been documented as being Further, comparisons with those in other non-mining connected with worse mental health outcomes in FIFO forms of employment, or in a mining job but living workers, such as longer on site rosters, inability to visit residentially show that FIFO work is associated with family during time off, or lack of opportunity to socialise better mental health (Bradbury, 2011; Joyce et al., with others on site (Albrecht & Anglim, 2018; Dorow & 2013; Miller et al., 2019; Velander et al., 2010). These Jean, 2021; Gardner et al., 2018; Parker et al., 2018). mixed results have been attributed to the varying Consequently, the COVID-19 public health measures, quality in the research designs of studies in this area, while necessary and designed to protect workers, may in particular, the lack of research that uses a matched have had unintended negative side effects for worker benchmark sample on key demographics or similar mental health and these effects may persist beyond occupations (Parker et al., 2018). Further, a commonly these measures being in place. Establishing the extent identified confounding factor related to the research to which FIFO workers' mental health may have been findings on FIFO worker mental health is the demo- affected by these measures is key. Political decision- graphic makeup of this workforce (i.e., male, particular makers are debating how living with COVID-19 can be age group, educational background; Considine et al., managed (Karp, 2021) understanding how the measures 2017; Parker et al., 2018). This possible confounding put in place may affect FIFO workers can support effect highlights the need for research that controls for informed trade-off decision-making and means mental these attributes or matches samples to discern the health can be strategically considered. These findings impact of FIFO work on worker mental health indepen- can help understand what FIFO workers have been dent of demographic characteristics. In addition, through, can inform actions and levels of support research that captures the full spectrum of FIFO work- needed during the return to a new normal, and can ers' mental health is lacking, with many studies focus- shape responses to future crises. Thus, research into the ing on one or a couple of mental health aspects. impact of COVID-19 on FIFO workers, is therefore impor- Mental health is not merely the absence of mental tant and, to our knowledge, is lacking in the literature. illness but rather a state of wellbeing (World Health To address the issues outlined above, the present Organisation, 2013) and as such requires the absence study aims to generate insights into the state of FIFO of negative indicators of mental health and presence workers’ mental health before and during COVID-19. of positive indicators. Accordingly, in this study, we The present study considers not just mental ill- consider negative mental health indicators (i.e., psy- health, but also positive mental health and wellbeing chological distress and suicidal thoughts) as well as outcome indicators (in line with Greenspoon & positive ones (i.e., social, emotional, and psychological Saklofske, 2001; World Health Organisation, 2013) wellbeing) to capture FIFO workers' mental health and controls for key demographic variables where more holistically. In short, a representative and con- feasible to determine the role of these demographics textualised study providing a comprehensive analysis for FIFO worker mental health. In what follows, we of FIFO worker mental health that gives more definitive present the results of a series of comparisons, via answers regarding the state of FIFO workers’ mental which the study provides comprehensive insights health and wellbeing is needed. Without such a study, into FIFO workers’ mental health before and during it is unclear to what extent FIFO workers’ mental health the COVID-19 pandemic. AUSTRALIAN JOURNAL OF PSYCHOLOGY 3 Method used from the National Health Survey 2017–2018 in comparisons (Australian Bureau of Statistics, 2018b). Participants Data from three samples were collected for this study, FIFO worker sample (2018) namely a FIFO worker sample in 2018, a purposefully sampled benchmark sample for comparison in 2018, The responses from the FIFO worker sample were col- and a further FIFO sample in 2020 (the first wave of lected between November 2017 and February 2018. The COVID-19; see Table 1 for an overview of all samples). 2018 FIFO worker sample (N = 3108) consisted mainly of In addition to the collected data, data on psychological men (82.8%) and had an average age of around 41 (M = distress (K10) of the wider Australian population is 40.85; SD = 10.59) years of age. The most frequent Table 1. Overview of demographic characteristics. Sample FIFO worker Benchmark FIFO worker 2020 2018 sample 2018 sample COVID-19 sample Characteristic (N = 3108) (N = 326) (N = 362) Gender Male 82.8% 77.3% 81.6% Female 17.1% 22.7% 18.4% Other 0.1% 0% 0% Age <24 3.3% 0.6% 2.0% 25–34 29.7% 9.4% 18.3% 35–44 29.4% 20.8% 25.2% 45–54 25.2% 28.9% 22.6% 55+ 12.4% 40.3% 31.9% M(SD) 40.85 (10.59) 50.17 (11.31) 43.79 (10.82) Aboriginal/Torres Strait Islander Yes 2.9% 0% 2.9% No 94.2% 98.4% 90.9% Prefer not to say 2.9% 1.6% 6.1% Marital status Single, never married 15.6% 13.9% 14.6% Married/domestic partnership 74.6% 71.9% 75.1% Widowed, divorced, separated 9.8% 14.2% 10.4% Children 0 39.1% 38.7% 40.8% 1 13.0% 12.3% 16.5% 2 27.3% 30.6% 23.6% 3 13.4% 12.3% 12.0% 4 4.2% 3.9% 5.5% 5 1.4% 1.9% 0.3% 6 or more 1.4% 0.3% 1.3% Age youngest child 0–2 months 8.3% 3.2% 4.9% 1 up to 3 years 15.7% 7.9% 13.7% 3 up to 5 years 13.1% 4.7% 13.1% 6 up to 8 years 8.8% 7.9% 8.2% 8 up to 2 years 13.7% 12.1% 9.8% 12 up to 8 years 16.0% 17.9% 18.6% Over 18 24.3% 46.3% 31.7% Highest level of education Primary school 0.2% 0.0% 0.0% Secondary school 22.3% 16.5% 12.6% Apprentice 13.5% 4.5% 19.4% Tafe, College 27.8% 20.0% 20.1% University undergraduate degree 18.6% 30.0% 23.0% Postgraduate degree 9.2% 21.9% 12.0% Other training courses 8.4% 7.1% 12.9% Profession Administrative 2.8% 14.8% 3.3% Managerial 20.1% 32.6% 23.8% Professional/Technical 25.1% 28.1% 25.7% Operator 18.8% 2.9% 16.0% Technician or Trade/Maintainers 21.8% 3.9% 21.9% Camps and catering 1.3% 1.3% 0.0% Logistics and supply chain 2.4% 1.6% 3.3% Other 7.6% 14.8% 6.1% 4 J. M. GILBERT ET AL. highest levels of education were TAFE/college (27.8%), 2018b) was used as norm data for Kessler-10 data on completion of secondary school (22.3%), and a university psychological distress (depression and anxiety) in the undergraduate degree (18.6%). wider Australian population before COVID-19. The The 2018 FIFO worker sample was highly represen- 2017–2018 sample was nationally representative of the tative of the broader Western Australian resource sec- male Australian adult population (N = 8,658 completed tor population. With regard to industry according to the K10 measure). For the comparison with the FIFO the Australian Bureau of Statistics (Australian Bureau of worker samples (as a majority male workforce), male Statistics, 2018a), proportions were similar in terms of respondents were specifically included in the com- mining (WA resources 94,400 people, 84%; FIFO sam- parison to the population K10 mean using the mean ple 2577 people, 82.9%) versus oil and gas (WA sum scores of psychological distress (K10, anxiety and resources 17,900 people, 19%; FIFO sample 531 peo- depression) for the 2018 FIFO sample and the existing ple, 17.1%). The FIFO worker sample was also repre- data sets. We also include data on psychological dis- sentative of WA resources sector workers (Australian tress reported by Rahman et al. (2020; N = 587) that Bureau of Statistics, 2018a) with respect to gender (WA was collected from the wider Australian population in resources male 81.6%, female 18.4%; FIFO sample male 2020 to compare with the FIFO workers in 2020. 82.8%, female 17.1%), and age (around 80% being from 25 to 54 years old). Procedure This study was approved by university ethics commit- Benchmark sample (2018) tees at the University of Western Australia and Curtin Responses from a benchmark sample (N = 326) of University. Ethics board approval number University of non-FIFO workers were collected through purposeful Western Australia: RA/4/1/9262; Ethics board approval sampling. Using quotas on key demographics number Curtin University: HRE2018–0449. ensured that the benchmark samples, like the FIFO The 2018 FIFO worker responses were collected via sample, consisted of males (77.3%) from Western an online and pen and paper survey. The survey was Australia (78.0%) and matched the FIFO sample in advertised via industry bodies, companies, unions, and terms of marital status. Despite our efforts, the mental health organisations. To generate the compari- benchmark sample differed from the FIFO worker son sample that matched key attributes of the FIFO sample in its age (M = 50.7 years; SD = 11.31), educa- worker sample, a data collection company was engaged. tional level (over 50% had completed a university The 2020 FIFO worker responses were collected by undergraduate or postgraduate degree) and their online survey, advertised through industry bodies, com- professional roles (with a higher representation of panies, and social media. In all surveys, participation was administrative and managerial roles). voluntary and confidential. Only high-quality responses were included in both samples, with cases retained if: multiple careless response checks were passed and FIFO worker sample COVID-19 (2020) there was a response time of at least 2 s per item The 2020 FIFO sample included n = 362 FIFO workers. (Ward & Meade, 2018) and at least 70% of the survey The data were collected between May and was completed (Dittman et al., 2016). October 2020. The sample was mostly male (81.6%), and mostly aged between 35 and 54 years. Most parti- Measures cipants reported being married or in a domestic part- nership (75%), with at least one dependent child (59%). Psychological distress, including feelings of depression, Participants reported to be mostly engaged in mining restlessness, fatigue, worthlessness, and anxiety, was (61%), oil and gas (22%), and construction (8%) indus- measured via the Kessler-10 (K10; Kessler et al., 2002). tries. The reported highest level of education com- Responses to items were on a 5-point scale (ranging pleted was most commonly TAFE or traineeship from 1-None of the time, to 5-All of the time). The items (39.5%), and university undergraduate degree (23%). were reliable with a Cronbach’s α = .92 or higher. A summed score across all items was computed. Wellbeing was measured using a shortened ver- Additional comparison data used sion (nine items) of the Mental Health Continuum In addition to the data collected from the samples listed (Lamers et al., 2011) which assesses social wellbeing above, a pre-existing data set from the 2017–2018 (i.e., social integration, contribution, coherence, National Health Survey (Australian Bureau of Statistics, actualisation and acceptance); emotional wellbeing AUSTRALIAN JOURNAL OF PSYCHOLOGY 5 (i.e., positive emotions); and psychological wellbeing compare the 2018 FIFO worker sample with the bench- (i.e., self-acceptance, growth, purpose). Respondents mark sample from 2018, and the 2018 FIFO worker rated the frequency of every wellbeing aspect in the sample with the 2020 FIFO worker sample using non- past month on a 6-point Likert scale (ranging from parametric statistical analyses. In each of these compar- 1-Never to 6-Every day). We shortened the scale to isons the sample sizes were unequal, the data and its reduce the burden on participants by selecting the residuals were not normally distributed, and the var- highest loading items in each dimension. Cronbach’s iances were not equal between samples. Accordingly, alphas for each wellbeing dimension were high we used Welch’s t-tests to compare the means of the (Cronbach’s α = .81 or higher). Mean scores were 2018 FIFO worker sample with the 2018 benchmark computed for each subscale. sample, and then the 2018 FIFO worker sample with Burnout is a state of mental exhaustion due to pro- the 2020 FIFO worker sample. This approach allowed us longed exposure to work-related stressors (Taris et al., to document the impact of FIFO work per se. We further 1999). Two items from the Maslach Burnout Inventory conduct hierarchical regression analyses with boot- exhaustion subscale were used (“I feel emotionally strapping (1000 bootstrap samples, 95% confidence drained from my work”; and “I feel used up at the intervals; see Hayes, 2017); entering control variables end of the work day”) as per Dollard and Bakker in step 1 and a dummy coded FIFO work variable (2010). Responses were on a 7-point scale from (FIFO work = 1; benchmark sample = 0) in step 2 for 1-Never to 7-Every day. Mean scores were computed the 2018 FIFO worker sample with the 2018 benchmark for the two items (Cronbach’s alpha was .87 and over). sample. The regression analysis provides insights into Suicide intention was measured via three items from the extent to which FIFO vs other work explain addi- the Self-Injurious Thoughts and Behaviours Interview tional variance beyond demographic characteristics (Nock et al., 2007). The items ask about thoughts and (age, gender, education, and professional role) in the plans about suicide on an 8-point agreement scale mental health and wellbeing outcomes (a dummy from 1-Strongly disagree to 8-Strongly agree coded variable was also used for comparing scores (Cronbach’s α = .61). A mean score was computed. from 2018 and 2020). Third, we performed a one Control variables included were age, gender, educa- sample t-test to compare the FIFO worker sample tion, and professional role, given the variation between from 2018 means to the Australian norm group the FIFO workers and the benchmark sample on these means from the Australian Bureau of Statistics demographics. Education was coded into two dummy (Australian Bureau of Statistics, 2018b, 2019). An over- coded variables: one for higher education (university view of correlations is provided in the Appendix. undergraduate and postgraduate) and one for college (apprentice, TAFE, college, other training courses). The Results professional role was dummy coded to represent whether the worker was an operator/technician vs Table 2 provides an overview of the K10 scores, respec- administrative, managerial and professional role. tively, for each of the samples and existing data sets that were included in this study (please see Appendix for an overview of correlations).Psychological distress scores Analysis (K10) given via percentages representing different cate- First, we provide a comprehensive descriptive overview gories of psychological distress for 2018 samples show of the psychological distress scores for each of the that 21.8% and 10.8% of the FIFO worker sample had high samples and data sets reported on, by clustering parti- and very high psychological distress, respectively, com- cipants into groups that represent low, moderate, high, pared to approximately 8.0% and 3.7% of the Australian and very high psychological distress scores following norm sample and 12.7% and 4.6% of the benchmark guidelines by Andrews and Slade (2001). Second, we group before COVID-19. A total of 68% of the Australian Table 2. Overview of FIFO workers and comparison samples from 2020 to 2018 with levels of psychological distress (K10). 2020 COVID FIFO 2018 FIFO 2018 matched bench- 2018 Australian men 2020 COVID Australian sample % sample % (K10) sample % (K10) mark group % (K10) norm data % (K10) (Rahman et al., 2020; K10) Low 27.4 37.1 55.84 65.9 37.5 Moderate 31.7 30.3 26.95 22.1 29.3 High 28.3 21.8 12.66 8.4 20.3 Very High 12.7 10.8 4.55 3.6 13.0 High + Very high 40.9 32.6 17.21 12.0 33.3 combined 6 J. M. GILBERT ET AL. norm group and 55.84% of the matched benchmark Consistent across these mental health aspects, FIFO work- sample reported low psychological distress, whereas ers were shown to have worse mental health and well- only 37.1% of the FIFO workers included in this study being than the benchmark group. Further, the statistical reported low levels of psychological distress. The 2020 comparison using regression analyses (see Table 4) FIFO worker sample recorded an increased number of showed significant differences persisting for two of the workers experiencing high or very high levels of psycho- mental health indicators after controlling for demo- logical distress (40.9%) as well as fewer FIFO workers graphic characteristics in 2018. First, for psychological experiencing low levels of psychological distress (27.4%). distress, when added in Step 2 following the demo- In data collected from the wider Australian population graphic variables, FIFO work accounted for an additional (Rahman et al., 2020) fewer people fell into the high to 0.3% of the variance in psychological distress, with very high psychological distress category (37.3%) and unstandardised regression coefficient B = 1.440 (95% CI a larger group responded with low levels of psychological .574, 2.206), p = .001.Second, FIFO work added in step 2 distress compared to the FIFO workers (37.5%). accounted for an additional 1% of the variance in burn- To investigate the difference in mental health of FIFO out, with unstandardised regression coefficient B = .654 workers and other workers before COVID-19, FIFO work- (95% CI .436, .867), p = .001. While the amount of addi- ers (2018 sample) were first compared to the 2018 bench- tionally explained variance was small, FIFO work signifi - mark sample. An overview of the state of mental health in cant explained psychological distress and burnout over FIFO workers and the benchmark sample is provided in and above demographic variables. No statistically signifi - Table 2. Results of the Welch’s test indicate significant cant differences for wellbeing or suicidal thoughts were differences in all mental health aspects between FIFO indicated after controlling for demographics (see workers and the benchmark sample (see Table 3). Table 5). Table 3. Comparison of mental health and wellbeing between FIFO and benchmark samples (2018). Welch’s t-test Construct Group M SD df F p Psychological distress (K10) FIFO 19.36 7.14 Between 1 68.25 .000 Benchmark 16.30 6.07 Within 398.18 Burnout FIFO 3.88 1.33 Between 1 73.58 .000 Benchmark 3.00 1.72 Within 372.49 Emotional wellbeing FIFO 4.47 1.12 Between 1 8.48 .004 Benchmark 4.65 1.09 Within 375.98 Social wellbeing FIFO 3.38 1.33 Between 1 24.24 .000 Benchmark 3.74 1.24 Within 381.78 Psychological wellbeing FIFO 4.17 1.19 Between 1 8.77 .003 Benchmark 4.35 1.03 Within 395.27 Suicide intention FIFO 1.77 1.37 Between 1 6.22 .013 Benchmark 1.57 1.23 Within 325.81 Table 4. Regression analyses using the 2018 FIFO worker sample and the benchmark sample from 2018. Psychological K10 Burnout Emotional WB Social WB WB Suicide intention B 95% CI B 95% CI B 95% CI B 95% CI B 95% CI B 95% CI Step 1 Age −.13** −.15, −.11 −.03** −.03, −.02 .01** .00, .01 .01** .01, .01 .01* .00, .01 −.01* −.01, −.00 Gender −.44 −1.13, 1.13 −.19* −.34, −.02 −.13* −.24, −.02 −.15* −.27, −.03 −.07 −.17, .05 .07 −.07, .20 Prof role −.94* −1.51, −.41 −.04 −.17, .09 .09 −.00, .18 .22** .12, .32 .17* .08, .26 −.05 −.18, .07 Education: TAFE −.37 −1.07, .27 .01 −.15, .16 .05 −.05, .16 .03 −.09, .14 .03 −.08, .14 −.13 −.26, .03 Education: −1.70** −2.49, −.91 −.21* −.40, −.02 .17* .05, .31 .35** .22, .48 .18* .04, .30 −.32* −.18, .07 university R .05 .03 .01 .04 .02 .01 Step 2 Age −.12** −.14, −.09 −.02** −.03, −.02 .01* .00, .01 .01** .00, .01 .00* .00, .01 −.01* −.01, −.00 Gender −.48 −1.16, .20 −.21* −.36, −.05 −.13* −.24, −.02 −.15* −.27, −.03 −.06 −.17, .05 .07 −.07, .20 Prof role −.86 −1.41, −.34 −.01 −.14, .13 .08 −.01, .17 .21** .11, .31 .16* .07, .25 −.05 −.17, .07 Education: TAFE −.38* −1.09, .27 .00 −.15, .15 .05 −.05, .16 .03 −.09, .14 .03 −.08, .14 −.13 −.26, .03 Education: −1.58** −2.36, −.80 −.16 −.35, .03 .17* .04, .30 .34** .20, .47 .17* .04, .30 −.31* −.48, −.14 university FIFO work 1.44** .57, 2.21 .65** .44, .87 −.08 −.21, .05 −.16 −.31, .01 −.06 −.19, .06 .08 −.09, .24 R .06 .04 .01 .04 .02 .01 ∆R .00 .01 .00 .00 .00 .00 Note: CI=confidence intervals, **p ≤ .01, *p < .05; FIFO variable coded 1 = FIFO, 0 = Not FIFO; gender coded 0=female 1 = male; prof role coded 1 = managerial and professional roles, 0 = frontline roles (e.g., operators, administrators, drivers, cleaners, traders, caterers)s. AUSTRALIAN JOURNAL OF PSYCHOLOGY 7 Table 5. Comparison of FIFO worker samples and Australian population norm psychological distress (before COVID-19, data from 2018). One sample t-test Group M SD/SE df T p-value Before COVID-19 (data from 2018) K10 FIFO 19.36 SD = 7.14 Between 1 Norm 15.50 SE = 0.13 Within 3041 29.77 <.000 K10 Men FIFO 19.25 SD = 7.12 Between 1 Norm 15.10 SE = 0.14 Within 2521 29.26 <.000 K10 Women FIFO 19.91 SD = 7.27 Between 1 Norm 15.90 SE = 0.10 Within 515 12.54 <.000 Table 6. Comparison of mental health and wellbeing between FIFO worker samples in 2018 (before COVID-19) and 2020 (during COVID-19). Welch’s t-test Group M SD df F p Before and during COVID-19 (2018 & 2020) Psychological distress (K10) FIFO 2018 19.36 7.20 Between 1 9.04 .003 FIFO 2020 20.87 4.8 Within 286.65 Social Wellbeing FIFO 2018 3.38 1.33 Between 1 .026 .873 FIFO 2020 3.35 1.27 Within 300.37 Emotional Wellbeing FIFO 2018 4.47 1.12 Between 1 12.21 <.001 FIFO 2020 4.18 1.21 Within 289.55 Psychological Wellbeing FIFO 2018 4.17 1.19 Between 1 2.28 .132 FIFO 2020 4.06 1.12 Within 293.11 Suicide intention FIFO 2018 1 3.55 .061 FIFO 2020 276.87 Next, one sample t-tests comparing the FIFO worker < .000) in 2020 compared to 2018. After controlling for sample with the average score of the Australian norm demographic variables in hierarchical regression analyses data (both from 2018) showed that the scores for the (see Table 7), the differences between FIFO workers in FIFO sample on psychological distress were significantly 2018 and 2020 remained significant for psychological higher than for the norm group (t (3041) = 29.77 p = .000) distress B = 1.07 (95% CI 0.55, 1.56), p = .001, and emo- before COVID-19 (see Table 4). tional wellbeing B = −0.17 (95% CI −.25, −.08), p = .001. To address our second aim of identifying the impact of Additionally, after controlling for demographics signifi - COVID-19 on FIFO worker mental health, we compare cantly higher suicidal intention was indicated in 2020 FIFO worker mental health reported in 2018 with scores compared to 2018 in FIFO workers B = 0.16 (95% CI .01, from 2020. First, results from Welch’s t-tests reported in .23), p = .05. No difference was found for social or psycho- Table 6 indicate a significantly higher level in psychologi- logical wellbeing, across the two time points. Comparing cal distress (F(1, 286.65) = 9.04; p = .003) as well as signifi - FIFO workers’ levels of psychological distress to a sample cantly lower emotional wellbeing (F(1, 289.55) = 12.21; p pulled from the wider Australian population during Table 7. Regression analyses using the 2018 and 2020 FIFO worker samples. K10 Emotional WB Social WB Psychological WB Suicide intention B 95% CI B 95% CI B 95% CI B 95% CI B 95% CI Step 1 Age −.10** −.13/-.08 .004 .00/.01 .01** .00/.01 .00 .00/.01 −.01 −.01/.00 Gender −.46 −1.13/.19 −.12* −.22/-.01 −.19* −.31/-.06 −.07 −.18/.04 .05 −.10/.19 Prof role −.87** −1.45/-.27 .08 −.02/.17 .23** .13/.33 .17** .07/.28 −.03 −.16/.10 Education: TAFE −.35 −.99/.29 .02 −.07/.14 .04 −.07/.16 .05 −.06/.16 −.13 −.28/.02 Education: university −1.56** −2.35/-.77 .18* .05/.31 .35** .20/.49 .20* .07/.35 −.33** −.52/-.14 R .035** .010** .033** .015** .009** Step 2 Age −.11** −.13/-.09 .01* .00/.01 .01** .00/.01 .00 .00/.01 −.01* −.01/.00 Gender −.42 −1.12/.23 −.12* −.23/-.02 −.19* −.31/-.06 −.07 −.19/.03 .05 −.09/.19 Prof role −.84* −1.43/-.26 .08 −.02/.17 .23** .13/.33 .17** .07/.28 −.02 −.15/.10 Education: TAFE −.44 −1.10/.19 .04 −.06/.15 .05 −.07/.17 .05 −.05/.16 −.14 −.29/.01 Education: university −1.69** −2.49/-.92 .20* .08/.33 .35** .21/.50 .22* .08/.36 −.35** −.53/-.15 Year 1.07** .55/1.56 −.17** −.25/-.08 −.04 −.12/.05 −.09* −.17/.00 .16* .01/.23 R .040** .016** .033 .016 .011* ∆R .005 .006 .000 .001 .002 Note: CI=confidence intervals, **p ≤ .01, *p < .05; gender coded 0 = female 1 = male; prof role coded 1 = managerial and professional roles, 0 = frontline roles (e.g., operators, administrators, drivers, cleaners, traders, caterers); Year coded 0 = 2018 1 = 2020. 8 J. M. GILBERT ET AL. COVID-19 using a one sample t-test showed FIFO workers other issues (Laplonge, 2016). So, a conclusion to that had worse mental health (t (250) = 2.61, p = .01). end may be limited in its utility, if not counterproduc- tive. Instead, we suggest that it will be more beneficial to recognise that FIFO workers, as they are in terms of Discussion demographic makeup and work conditions, are an at- The results of this study provide insights into the sub- risk group for mental health and to direct our focus jective mental health and wellbeing of Australian- onto the specific aspects of FIFO work that could be based FIFO workers before and during COVID-19. The changed to protect FIFO workers. Thus, we propose to study captured the wide spectrum of mental health focus on protective factors such as boosting support, ranging from mental-ill health and suicidal thoughts to managing work design (including workload), and mental health and wellbeing (Greenspoon & Saklofske, humanising work culture by considering justice sys- 2001; World Health Organisation, 2013). tems at work, and supporting social connection We identify three key insights from this study. First, (Gilbert, 2019). In addition, the perceived stigma to the results show that FIFO workers are at a greater risk of seeking mental health services in case of mental health mental ill-health (i.e., depression, anxiety, burnout, and issues (Tynan et al., 2016) needs to be considered given suicide intention) and lower wellbeing in comparison to our study’s findings of FIFO workers’ heightened levels other groups. Second, our results identify that the worse of mental-ill health and lower levels of wellbeing. mental health of FIFO workers compared to other work- Some methodological issues need to be considered ers was only partially attributable to their demographics. when interpreting our findings. First, the results show Differences in mental health persisted for some mental a static picture of mental health and wellbeing in FIFO ill-health measures (i.e., psychological distress and burn- workers and others. Research grounded in the stability out) when demographic variables were controlled for. and change model byOrmel and Schaufeli (1991) shows Overall, this illustrates that FIFO workers are an at-risk that around 50%–60% of variation in psychological dis- group. They are so only in part because of their age, tress is the result of a stable factor, with the remainder gender, level of education, and professional roles, all of being subject to fluctuation based on state level psy- which have been previously noted as risk factors chological distress, which is likely affected by life events (Education and Health Standing Committee, 2015), and and other temporal aspects (Breslin et al., 2006). Our in part because of the nature of the FIFO experience cross-sectional data cannot account for such fluctua - itself (before and during COVID-19). Third, the results tions. Second, the results indicate differences between show that mental health of FIFO workers was worse FIFO workers across years and compared to other during COVID-19 compared to before (i.e., regarding groups, allowing some conclusion of causality. psychological distress, emotional well-being and suici- However, some limitations in asserting causality need dal intention) and that their mental health was worse to be recognised. For example, it is unclear whether at- than that of the wider population. Overall, these results risk individuals self-select into FIFO work arrangements. illustrate the nuanced nature in which FIFO work may be Further, there were indeed some differences in the affecting workers, a consideration that remains relevant, demographics between the three samples included in and is in fact amplified in its relevance during COVID-19. this study. Key to our findings is that we controlled for The findings of this study can move the debate and such differences in our analyses. Finally, other factors research literature around FIFO worker mental health were not controlled for in our study like differences in beyond being stuck at deciding “if” and move the con- weather or economic changes. Irrespective of these versation towards the “why” and how to change things issues, it needs to be recognised that the differences for the better. that are reported in our study raise an issue that war- Our results demonstrate that mental health out- rants attention and identifies FIFO workers as an at-risk comes for workers are accounted for by being group for mental ill-health and wellbeing engaged in FIFO work beyond demographics on In conclusion, our study shows a refined picture of many mental health outcomes. Yet, they also show the extent to which FIFO workers are an at-risk group that demographics are relevant to understanding in terms of their mental health. It further shows that FIFO workers' mental health. We identify that the this is even more so the case since COVID-19. The demographic make-up of the FIFO workforce is not results illustrate the importance for FIFO workers and easily changed, for example, through selective hiring employers to pay attention to mental health. Industry, government, and other relevant stakeholders should practices and doing so may in fact not be a feasible or ensure that the support options they can provide suit desirable, due to enduring masculine work culture and AUSTRALIAN JOURNAL OF PSYCHOLOGY 9 the constraints of FIFO work. Given the higher levels of for children [PhD]. Curtin University, https://espace.curtin. edu.au/ mental-ill health and poorer wellbeing in FIFO workers Breslin, F. C., Hepburn, C. G., Ibrahim, S., & Cole, D. (2006). during COVID-19 offering support and designing work Understanding stability and change in psychological dis- to address these issues is now even more important. tress and sense of coherence: A four-year prospective study 1. Journal of Applied Social Psychology, 36(1), 1–21. https://doi.org/10.1111/j.0021-9029.2006.00001.x Disclosure statement Clifford, S. (2009). The effects of Fly-in/Fly-out commute arrangements and extended working hours on the stress, No potential conflict of interest was reported by the author(s) lifestyle, relationship and health characteristics of Western Australian mining employees and their partners: Preliminary report of research findings [PhD thessis]. University of Funding Western Australia, Perth. http://www.web.uwa.edu.au/__ data/assets/pdf_file/0003/405426/FIFO_Report.pdf This work was supported by Mental Health Commission WA Considine, R., Tynan, R., James, C., Wiggers, J., Lewin, T., grant number MHC 17/2944 (2017 – 2018). Inder, K., Perkins, D., Handley, T., & KellyHandley, T. (2017). The contribution of individual, social and work characteristics to employee mental health in a coal mining ORCID industry population. Plos One, 12(1), e0168445. https://doi. Jessica M. Gilbert http://orcid.org/0000-0003-0458-1546 org/10.1371/journal.pone.0168445 Laura S. Fruhen http://orcid.org/0000-0002-9599-3468 Dittman, C., Henriquez, A., & Roxburgh, N. (2016). When a Cindy T. Burton http://orcid.org/0000-0001-5811-2660 non-resident worker is a non-resident parent: Sharon K. Parker http://orcid.org/0000-0002-0978-1873 Investigating the family impact of fly-in, fly-out work prac- tices in Australia. Journal of Child and Family Studies, 25(9), 2778–2796. https://doi.org/10.1007/s10826-016-0437-2 Data availability statement Dollard, M. F., & Bakker, A. B. (2010). Psychosocial safety climate as a precursor to conducive work environments, psychologi- The data are not publicly available due to their containing cal health problems, and employee engagement. Journal of information that could compromise the privacy of research Occupational and Organizational Psychology, 83(3), 579–599. participants. https://doi.org/10.1348/096317909X470690 Dorow, S., & Jean, S. (2021). Managing liminal time in the fly-in fly-out work camp. Human Relations, 0(0), 0018726721989792. References https://doi.org/10.1177/0018726721989792 Education and Health Standing Committee. (2015). The impact Albrecht, S. L., & Anglim, J. (2018). Employee engagement of FIFO work practices on mental health. Legislative and emotional exhaustion of fly-in-fly-out workers: A diary Assembly Parliament of Western Australia. http://www.par study. Australian Journal of Psychology, 70(1), 66–75. liament.wa.gov.au/parliament/commit.nsf/Report+Lookup https://doi.org/10.1111/ajpy.12155 +by+Com+ID/2E970A7A4934026448257E67002BF9D1/ Andrews, G., & Slade, T. (2001). Interpreting scores on the $file/20150617±+Final+Report+w+signature+for+website. Kessler psychological distress scale (K10). Australian and pdf New Zealand Journal of Public Health, 25(6), 494–497. Fruhen, L. S., Gilbert, J. C., & PArker, S. K. (2022). Fly-in-fly-out https://doi.org/10.1111/j.1467-842X.2001.tb00310.x work: A review of the impact of an extreme form of Australian Bureau of Statistics. (2018a). Labour force, work-related travel on mental health. Organizational Australia, detailed - electronic delivery, February (Cat. Psychology Review. 20413866221134938. no. 6291.0.55.001). In. Gardner, B., Alfrey, K. -L., Vandelanotte, C., & Rebar, A. L. Australian Bureau of Statistics. (2018b). National health survey: (2018). Mental health and well-being concerns of fly-in First results, 2017–18 (Cat. No 4364055001DO007_20172018). fly-out workers and their partners in Australia: In. A qualitative study. BMJ Open, 8(3), e019516. https://doi. Australian Bureau of Statistics. (2019). Microdata: National org/10.1136/bmjopen-2017-019516 health survey, 2017-18 (Cat. no. 4324.0.55.001). https:// Gilbert, J. M. (2019). Between a rock and a hard place: www.abs.gov.au/AUSSTATS/abs@.nsf/Lookup/4324.0.55. Exploring ways to optimise wellbeing for people who work 001Main+Features12017-18?OpenDocument fly-in, fly-out in the Australian mining industry [Doctoral Biddle, N., Edwards, B., Gray, M., & Sollis, K. (2020). Hardship, thesis]. Curtin University. http://hdl.handle.net/20.500. distress, and resilience: The initial impacts of COVID-19 in 11937/77509 Australia. https://csrm.cass.anu.edu.au/sites/default/files/ docs/2020/6/The_initial_impacts_of_COVID-19_in_ Gilbert, J. M., Fruhen, L., & Sharon, K. P. (2020). FIFO worker Australia_2020_4.pdf mental health and wellbeing: The impact of COVID-19 (sum- Bowers, J., Lo, J., Miller, P., Mawren, D., & Jones, B. (2018). mary of findings) . https://www.transformativeworkdesign. Psychological distress in remote mining and construction com/fifo-mental-health-survey workers in Australia. The Medical Journal of Australia, 208 Greenspoon, P. J., & Saklofske, D. H. (2001). Toward an inte- (9), 391–397. https://doi.org/10.5694/mja17.00950 gration of subjective well-being and psychopathology. Bradbury, G. S. (2011). Children and the fly-in/fly-out lifestyle: Social Indicators Research, 54(1), 81–108. https://doi.org/ Employment-related paternal absence and the implications 10.1023/A:1007219227883 10 J. M. GILBERT ET AL. Hayes, A. F. (2017). Introduction to mediation, moderation, and model. Journal of Personality and Social Psychology, 60(2), conditional process analysis: A regression-based approach. 288–299. https://doi.org/10.1037/0022-35188.8.131.528 Guilford publications. Parker, S., Fruhen, L., Burton, C., McQuade, S., Loveny, J., Henry, P., Hamilton, K., Watson, S., & Macdonald, N. (2013). Griffin, M., Page, A., Chikritzhs, T., Crock, S., Jorritsma, K. & FIFO/DIDO mental health. https://www.aasw.asn.au/docu Esmond, J. (2018). Impact of FIFO work arrangements on the ment/item/5136 mental health and wellbeing of FIFO workers. https://www. Infrastructure Planning and Natural Resources Committee. mhc.wa.gov.au/media/2547/impact-of-fifo-work- (2015). Inquiry into fly-in, fly-out and other long distance arrangement-on-the-mental-health-and-wellbeing-of-fifo- commuting work practices in regional Queensland. https:// workers-full-report.pdf www.parliament.qld.gov.au/documents/committees/ Rahman, M. A., Hoque, N., Alif, S. M., Salehin, M., Islam, S. M. S., IPNRC/2015/FIFO/02-rpt-009-09Oct2015.pdf Banik, B., Sharif, A., Nazim, N. B., Sultana, F., & Cross, W. Joyce, S. J., Tomlin, S. M., Somerford, P. J., & (2020). Factors associated with psychological distress, fear Weeramanthri, T. S. (2013). Health behaviours and out- and coping strategies during the COVID-19 pandemic in comes associated with fly-in fly-out and shift workers in Australia. Globalization and Health, 16(1), 1–15. https://doi. Western Australia. Internal Medicine Journal, 43(4), 440–444. org/10.1186/s12992-020-00624-w https://doi.org/10.1111/j.1445-5994.2012.02885.x Sellenger, M., & Oosthuizen, J. (2017). Quantitative analysis of Karp, P. (2021). Conspiracy, Covid and the coalition: Why are mental wellbeing of fly-in fly-out construction project sup- more of its MPs appealing to the alt-right? The Guardian. port service workers. Journal of Preventive Medicine & https://www.theguardian.com/australia-news/2021/dec/ Healthcare, 1(1). https://www.jscimedcentral.com/ 11/conspiracy-covid-and-the-coalition-why-are-more-of- PreventiveMedicine/preventivemedicine-1-1001. its-mps-appealing-to-the-alt-right php#Citation Kessler, R. C., Andrews, G., Colpe, L. J., Hiripi, E., Mroczek, D. K., Storey, K. (2001). Fly-in/fly-out and fly-over: Mining and regio- Normand, S. L., Walters, E.E., & Zaslavsky, A. M. (2002). nal development in Western Australia. The Australian Short screening scales to monitor population prevalences Geographer, 32(2), 133–148. https://doi.org/10.1080/ and trends in non-specific psychological distress. 00049180120066616 Psychological Medicine, 32(6), 959–976. https://doi.org/10. Taris, T. W., Schreurs, P. J. G., & Schaufeli, W. B. (1999). Construct 1017/s0033291702006074 validity of the maslach burnout inventory-general survey: A Lamers, S. M. A., Westerhof, G. J., Bohlmeijer, E. T., Ten two-sample examination of its factor structure and Klooster, P. M., & Keyes, C. L. M. (2011). Evaluating the correlates. Work & Stress, 13(3), 223–237. https://doi.org/10. psychometric properties of the mental health 1080/026783799296039 continuum-short form (MHC-SF). Journal of Clinical Trinca, H. (2020). When FIFO proves the tyranny of Psychology, 67(1), 99–110. https://doi.org/10.1002/jclp.20741 distance. The Australian. https://www.theaustralian.com. Laplonge, D. (2016). A toolkit for women: The mis(sed) man- au/business/the-deal-magazine/long-distance-call-for agement of gender in resource industries. Journal of -mental-health-assistance-for-fifo-workers-during-covid- Management Development, 35(6), 802–813. https://doi. shutdown/news-story/8940eeeb944039ece3e6a372 org/10.1108/JMD-07-2014-0078 6b962230 Lester, L., Waters, S., Spears, B., Epstein, M., Watson, J., & Tynan, R. J., Considine, R., Rich, J. L., Skehan, J., Wiggers, J., Wenden, E. (2015). Parenting adolescents: Developing Lewin, T. J., James, C., Inder, K., Baker, A.L., Kay-Lambkin, F. strategies for FIFO parents. Journal of Child and Family and Perkins, D., & Kelly, B. J. (2016). Help-seeking for Studies, 24(12), 3757–3766. https://doi.org/10.1007/ mental health problems by employees in the Australian s10826-015-0183-x mining industry. BMC Health Services Research, 16(1), Miller, P., Brook, L., Stomski, N. J., Ditchburn, G., & Morrison, P. 498–510. https://doi.org/10.1186/s12913-016-1755-1 (2019). Depression, suicide risk, and workplace bullying: Velander, F., Schineanu, A., Liang, W., & Midford, R. (2010). A comparative study of fly-in, fly-out and residential Digging for gold and coming up blue: A health survey in resource workers in Australia. Australian Health Review, the mining industry. Journal of Health, Safety and 44(2), 248. https://doi.org/10.1071/AH18155 Environment, 26(5), 389–401. https://researchoutput.csu. Nock, M. K., Holmberg, E. B., Photos, V. I., & Michel, B. D. edu.au/ws/portalfiles/portal/8800896/28161+Schineanu.pdf (2007). Self-injurious thoughts and behaviors interview: Ward, M. K., & Meade, A. W. (2018). Applying social psychol- Development, reliability, and validity in an adolescent ogy to prevent careless responding during online surveys. sample. Psychological Assessment, 19(3), 309–317. https:// Applied Psychology, 67(2), 231–263. https://doi.org/10. doi.org/10.1037/1040-35184.108.40.2069 1111/apps.12118 Ormel, J., & Schaufeli, W. B. (1991). Stability and change in World Health Organisation. (2013). Comprehensive mental psychological distress and their relationship with health action plan. http://www.who.int/mental_health/ self-esteem and locus of control: A dynamic equilibrium action_plan_2013/en/ AUSTRALIAN JOURNAL OF PSYCHOLOGY 11 Appendix: Correlation table 1 2 3 4 5 6 7 8 9 10 2018 FIFO worker sample 1. Age 2. Gender .17** 3. Professional Role .14** .11** 4. Education: TAFE .18** .17** −.07** 5. Education: university −.18** −.17** .07** −1.00** 6. K10 −.14** −.04 −.08** .09** −.09** 7. Burnout −.11** −.07** −.01 .03 −.03 .65** 8. Emotional wellbeing .02 −.05** .05* −.08** .08** −.66** −.48** 9. Social wellbeing .03 −.07** .08** −.15** .15** −.48** −.40** .61** 10. Psychological wellbeing .01 −.04* .08** −.10** .10** −.58** −.41** .73** .61** 11. Suicide intention −.02 .02 −.03 .09** −.09** .35** .19** −.32** −.19** −.28** 2018 Benchmark sample 1. Age 2. Gender .19** 3. Professional Role .06 .12* 4. Education: TAFE .14* .05 −.10 5. Education: university −.14* −.05 .10 −1.00** 6. K10 −.35** −0.10 .01 −.05 .05 7. Burnout −.28** .00 .19** −.11* .11* .61** 8. Emotional wellbeing .20** .05 .03 .11 −.10 −.62** −.38** 9. Social wellbeing .12* 0.09 .07 .00 −.00 −.37** −.28** .58** 10. Psychological wellbeing .18** .06 .14* .04 −.04 −.50** −.30** .73** .58** 11. Suicide intention −.06 .06 .00 −.05 .05 .25** .14* −.32** −.15* −.29** 2020 FIFO worker sample 1. Age 2. Gender .15* 3. Professional Role .16* .05 4. Education: TAFE .14* .26** .01 5. Education: university −.14* −.26** −.01 −1.00** 6. K10 −.16 −.06 −.11 .00 .00 8. Emotional wellbeing .09 0.1 .07 −.05 .05 −.72** 9. Social wellbeing .12 .05 .12 −.03 .03 −.49** .59** 10. Social wellbeing .01 .07 .16* −.14* .14* −.67** .72** .61** 11. Suicide intention −.05 .01 −.17** −.03 .03 .25** −.24** −.13* −.31** Note: **p ≤ .01, *p < .05; gender coded 0 = female 1 = male; prof role coded 1 = managerial and professional roles, 0 = frontline roles (e.g., operators, administrators, drivers, cleaners, traders, caterers).
Australian Journal of Psychology
Taylor & Francis