Job Quality and Construction Workers’ Mental Health: Life Course Perspective

Pirzadeh, Payam; Lingard, Helen; Zhang, Rita Peihua

doi:10.1061/(ASCE)CO.1943-7862.0002397

Open access

Technical Papers

Sep 27, 2022

Job Quality and Construction Workers’ Mental Health: Life Course Perspective

Authors: Payam Pirzadeh, Ph.D. https://orcid.org/0000-0001-6511-0946 [email protected], Helen Lingard, Ph.D. https://orcid.org/0000-0003-3645-8390 [email protected], and Rita Peihua Zhang, Ph.D. https://orcid.org/0000-0001-9907-6491 [email protected]Author Affiliations

Publication: Journal of Construction Engineering and Management

Volume 148, Issue 12

https://doi.org/10.1061/(ASCE)CO.1943-7862.0002397

PDF

Abstract

Psychosocial job quality has been proven to be linked to workers’ mental health. Drawing on a life course perspective, this study sought to identify, compare, and contrast the psychosocial characteristics of job quality that are related to mental health in three age groups of manual/nonmanagerial construction workers, i.e., young workers, middle-aged workers, and older workers. Data were extracted from the national and longitudinal Household, Income, and Labour Dynamics in Australia (HILDA) Survey data set. The study used 15 waves of data from the HILDA survey with 6,352 responses from 1,768 participants. Longitudinal random-intercept regression models were used to examine the association between each of five aspects of job quality (i.e., job demands and complexity, job control, perceived job security, effort-reward fairness, and job intensity) and mental health. Overall, the research results showed that construction workers’ mental health declined when experiencing adverse job conditions and the magnitude of decline increased as the number of job adversities increased. Specifically, workers of the midage group experienced more accelerated decline in mental health compared with the other two groups when experiencing two adverse job conditions. Age-related differences were also identified in the way that individual job quality aspects are related to mental health. Although low job security and perceived unfairness of effort and reward were significant predictors of mental ill-health in all age groups, job demand and complexity and high job intensity were predictors of mental ill-health in midage and older construction workers but were not significant contributors to mental ill-health among younger workers. The findings highlight the need to develop targeted approaches to protecting and promoting the mental health of construction workers in different age groups.

Introduction

Work and Mental Health

Work-related mental health impacts are a global concern due to the significant adverse impact on employees and costs associated with lost productivity (World Health Organisation 2021). In Australia, work-related mental health conditions account for about 6% of workers’ compensation claims, incurring a cost of approximately AUD 543 million per year (Safe Work Australia 2021).

Construction is a high-risk industry for mental ill-health (Boschman et al. 2013). Moreover, untreated mental health disorders contribute to suicide among manual (nonmanagerial) construction workers (Heller et al. 2007; Milner et al. 2017b). The incidence of suicide in the construction industry is reported to be high, particularly among low-skilled workers (Office for National Statistics 2017; Milner et al. 2014), and there are concerns that this could be elevated by the effects of the COVID-19 pandemic (King and Lamontagne 2021).

Work Environment as a Contributing Factor to Poor Mental Health

Workers’ mental health has been linked to the extent that psychosocial hazards are present in the work environment (Stansfeld et al. 1999; Broom et al. 2006; Stansfeld and Candy 2006). Psychosocial hazards are features of the design and management of work that increase the risk of work-related stress and have the potential to cause psychological or physical harm (WorkSafe Victoria 2021).

Psychosocial risk factors that contribute to mental ill-health in construction include long and inflexible work hours, excessive workloads and time pressure, role conflict, role ambiguity, a lack of autonomy or control, and insufficient recovery opportunity (Lingard 2004; Yip and Rowlinson 2009; Enshassi et al. 2016; Zhang and Bowen 2021). Job insecurity is also reported to adversely affect the mental health of construction workers (Turner and Lingard 2016) and has been identified as a precursor to suicide among male construction workers (Milner et al. 2017b). In addition, workers in construction have reported exposure to interpersonal conflict, bullying, sexism, racism, discrimination and the marginalization of minority groups (George and Loosemore 2019; Galea et al. 2018; Bowen et al. 2013), which also contribute to poor mental health outcomes (Goldenhar et al. 1998; Rospenda et al. 2009).

Job Quality and Mental Health

Work can enhance health by providing a financial income and a sense of purpose and facilitating individuals’ social connections within a community (Butterworth et al. 2011a). However, the impact of work on health depends on the quality of the jobs people perform. Job quality has been defined as “the sets of work features which foster the well-being of the worker” (Green 2007, p. 9) and is understood to be a multidimensional construct reflecting the presence or absence of psychosocial risk factors in a particular job.

Butterworth et al. (2011a) reported psychosocial job adversities (i.e., low job security, low job control, high job demands and complexity, and a perception of unfair pay) to be associated with poor mental health and physical health. Butterworth et al. (2011b) found that individuals in better quality jobs have significantly better mental health compared with those who are in the poor-quality jobs. Specifically, a linear relationship between the number of adverse conditions and mental health was noticed, i.e., each additional adversity caused further decline in mental health (Butterworth et al. 2011b). Being in a poor-quality job may even be worse for mental health than having no job at all (Broom et al. 2006). For example, when unemployed people transition into low-quality jobs, the negative impacts on their mental health are significantly worse than among those who remained unemployed (Leach et al. 2010b).

Knowledge Gap and Research Aim

Recent reviews have begun to explore the relationships between psychosocial risk factors and construction employees’ mental health (Sun et al. 2020; Chan et al. 2020). However, empirical research to understand the way that the quality of construction jobs is related to mental health is lacking. This understanding will enable jobs to be designed in ways that minimize harm. Further, little research has been conducted to understand how psychosocial risk factors are experienced by different age groups of the workforce, although studies in the construction industry provide some evidence of age-related differences (Yang et al. 2017; Peng and Chan 2019). This knowledge gap has led this study to examine how the experience of psychosocial risk factors vary across the life course to develop targeted prevention strategies. Also, previous studies of stress and mental health in the construction industry have focused on managerial or professional samples, with less attention being paid to the factors affecting manual workers. This study will contribute to the body of knowledge by investigating determinants of mental health among the nonmanagerial/manual workforce.

This study examined the way that psychosocial job quality factors are related to mental health among manual construction workers in different age groups, i.e., young workers, middle-aged workers and older workers. Research objectives were to (1) identify job quality characteristics that predict mental health by age group; (2) compare and contrast the determinants of mental health among workers in different age groups; and (3) consider the implications of observed similarities/differences for the development of strategies to facilitate healthy aging in the construction industry.

The theoretical background of the research is briefly described; then, the research methods and the results are presented and discussed.

Theoretical Background

The job demands-control (JDC) theory describes two key dimensions of the psychosocial work environment, i.e., job demands and job decision latitude (Karasek 1979). Job demands are psychological stressors present in the work environment, and job decision latitude reflects the extent to which someone has control over their work (Karasek 1979). According to the JDC model, jobs can be divided into four types: (1) high-strain jobs, characterized by high demands and low decision latitude, are considered to be the most problematic in relation to mental strain and health outcomes; (2) active jobs, involving simultaneously high demands and decision latitude, are associated with average levels of mental strain and are likely to facilitate employee development; (3) passive jobs, characterized by low demands and low decision latitude, are associated with average levels of mental strain but can be demotivating; and (4) low-strain jobs, combining low demands and high decision latitude, involve lower than average levels of mental strain and lower risk of ill health (Karasek 1979). The JDC theory is supported by empirical evidence (Häusser et al. 2010). Drawing on the JDC model, aspects of job quality that are frequently examined by researchers include job demands and complexity, and job control (Butterworth et al. 2011a; Leach et al. 2010a; Stansfeld and Candy 2006).

The effort-reward imbalance (ERI) theory posits that an imbalance between (high) effort and (low) reward causes a strain reaction that negatively affects workers’ health (Siegrist 1996). This model assumes that work is underpinned by a norm of reciprocity whereby efforts are rewarded commensurately with rewards, which can be in the form of money, esteem, career opportunities, or job security (Siegrist et al. 2004). However, sometimes workers’ efforts exceed perceived rewards and this imbalance between high costs and low gains can lead to a state of sustained strain reactions in the autonomic nervous system and ultimately produce emotional distress (Siegrist 1996). This is particularly evident when workers have few alternatives in the labor market, for example because they are low in skill or have restricted mobility (Siegrist et al. 2004). The ERI model is supported by empirical research (van Vegchel et al. 2005). Based on the ERI model, fair pay and job security have been previously examined as aspects of job quality linked to mental health (Butterworth et al. 2011a, b; Milner et al. 2014; Strazdins et al. 2004).

The combined effects of the JDC and ERI models on workers’ health are reportedly stronger than their separate effects (De Jonge et al. 2000; Rydstedt et al. 2007). For this reason, the present study includes components of both the JDC and the ERI models.

Age-Related Work and Health Effects

Young people are at a risky life stage for mental ill-health, with 75% of lifetime mental disorder cases starting by the age of 24 years (Kessler et al. 2005). Young workers experience high exposure to workplace hazards (Safe Work Australia 2015) and suffer from high risk of psychological ill-health when transitioning into work, particularly when they enter jobs characterized by low control, high demands, low security, and unfair pay (Milner et al. 2017a). Young workers are often engaged in precarious forms of employment characterized by inequality, low income/status, and job insecurity (McDonald et al. 2007; Louie et al. 2006). LaMontagne et al. (2013) reported that younger workers (aged between 15 and 25) consistently experience lower job control than their older counterparts. Young construction workers are also reported to experience high levels of psychological distress (Pidd et al. 2017), which is attributed to bullying and harassment in the workplace (McCormack et al. 2013; Ross et al. 2022). Critically, Australian construction apprentices are two and a half times more likely to die by suicide than other young men of their age (Mates in Construction 2016).

However, job quality also affects the mental health and physical functioning of older workers (Welsh et al. 2016). As people age, they naturally experience a reduction in physical work ability, which has implications for jobs that involve physically demanding work (Kenny et al. 2008) and can increase the risk of workplace injury (Truxillo et al. 2012). An increasing incidence of bodily pain among older construction workers has been linked to higher levels of depression, anxiety, and stress severity (Turner and Lingard 2020).

Aging is also associated with cognitive decline characterized by a reduction in processing speed, reasoning, memory, and executive functions (Deary et al. 2009). Older workers experience stress in jobs that involve intensive short-term processing of information (Kanfer and Ackerman 2004) and are more adversely affected by requirements to work under time pressure and by poor employment conditions than younger workers (De Zwart et al. 1999). In Europe, older construction workers have been found to experience high levels of burnout (Oude Hengel et al. 2012) and, in Hong Kong, older construction workers’ mental health is negatively correlated with psychological job demands and positively related to decision autonomy and social support (Peng and Chan 2020).

The concept of successful aging describes a situation in which subjective and objective outcomes that are valued by both workers and their organizations are achieved (Olson and Shultz 2019). The timing of retirement is one such outcome that is influenced by individual as well as job factors (Cloostermans et al. 2015). In considering how to support successful aging, it is helpful to consider the role played by job quality and design (Zacher et al. 2019). Thus, the present study focused on the relationship between job quality and mental health among manual construction workers in the Australian construction industry.

Method

Data and Sampling

Data were extracted from the Household, Income, and Labour Dynamics in Australia (HILDA) Survey data set (Melbourne Institute of Applied Economic and Social Research 2022). HILDA is a national and longitudinal study of Australian households. Data are collected in annual waves from over 13,000 individuals within over 7,000 Australian households, and includes social, demographic, health, and economic factors. This study used data from the latest 15 waves of the survey (years 2005–2019). The HILDA data set was searched to identify participants working in the construction industry [based on the Australian and New Zealand Standard Industrial Classification (ANZSIC) (ABS 2006b) industry classification code]. Manual workers (technicians and trades workers, machinery operators and drivers, and laborers) were identified based on Australian and New Zealand Standard Classification of Occupations (ANZSCO) (ABS 2006a) occupation classification codes. Cases with missing data on items of job quality and other covariates were excluded. The data set analyzed in the final (most restricted) model included 6,352 responses from 1,768 individuals.

Measures

Mental health was measured using a subscale of the SF-36 General Health Survey, which has been previously validated (Butterworth and Crosier 2004). The mental health scale includes five items and measures how often in the last 4 weeks respondents experienced positive emotions (e.g., feeling happy or calm) and negative emotions, i.e., symptoms of anxiety and depression (e.g., feeling nervous, depressed, or down). In the HILDA data set, the mental health scale is transformed, based on the procedure suggested by Ware et al. (2000), to reflect a 1–100 range. A higher score indicates better mental health. Studies have indicated reasonable validity for the SF-36 scale (Butterworth and Crosier 2004; Ware et al. 1994).

Job quality was measured using five components. Four components were selected from a previously validated measure (e.g., Butterworth et al. 2011b; Leach et al. 2010a). These were (1) job demands and complexity (four items, e.g., My job is complex and difficult); (2) job control (three items, e.g., I have a lot of freedom to decide how I do my own work); (3) perceived job security (three items, e.g., I have a secure future in my job); and (4) effort-reward fairness (a single item, i.e., I get paid fairly for the things I do in my job). An additional component was included measuring work intensity (Dinh et al. 2017). This component comprised three items, e.g., I do not have enough time to do everything at my job.

Responses to each of the job quality items were recorded on a 7-point Likert response format ranging from 1 (strongly disagree) to 7 (strongly agree). The scores for negative items were reversed during data preparation. The scores were averaged across all items for each component to reflect a single score for each aspect of job quality.

Control Variables

Demographic variables were also incorporated into the analysis as control variables as follows:

•

Age: Participants’ age at the time of their first survey response was included in the analysis as a control variable. In addition, to compare relationships between age groups, participants were divided into three age groups based on their age at their first response: younger than 24; between 25 and 45; and older than 45. These have been referred to as the exploration, establishment, and maintenance stages of a working life span (Schmitt and Unger 2019).

•

Household structure: Following Milner et al. (2017a), participants’ household structure was categorized into couple; couple with children; single parent; single person; or other.

•

Education: The highest education level of each participant was captured as a categorical variable reflecting (1) Year 11 and below; (2) Year 12; (3) Certificate III or IV; (4) advanced diploma or diploma; (5) bachelor or honors; (6) graduate diploma or graduate certificate; or (7) postgraduate study: master’s or doctorate.

•

Household income: the gross income band of participants’ households for the previous financial year in each wave was captured using the following 11 categories: (1) negative or zero; (2) $1–$9,999 AUD; (3) $10,000–$19,999 AUD; and so on up to (10) $100,000–$124,999 AUD; and (11) $125,000 AUD or more.

Other control variables included gender, years worked in current occupation, hours usually worked per week, having a disability or long-term health condition, having experienced adverse life events in the last 12 months (e.g., death of spouse, child, close relative, or friend; serious personal injury or illness; victim of violence or crime; separation from spouse, and so on), whether participants were caring for another person due to a health condition or because they are elderly or had a disability requiring care, whether participants had parenting responsibility for any children aged 17 years or less, and whether participants had experienced unemployment or not being in the labor force at some point during the survey waves.

Data Preparation and Statistical Analysis

Following the approach used by Butterworth et al. (2011a), Welsh et al. (2016), and Milner et al. (2017a), the scores for job quality components were dichotomized at the values closest to lowest quartile. That is, for each job quality component, the range of scores reported by the participants in the sample was considered, and the lower quartile score (the point that 25% of the job quality component scores were below it) was calculated. A new binary variable was generated for each component with a value of 1 indicating that a participant’s score for the corresponding job quality component was within the lowest quartile of the range of scores in the sample. Thus, it was possible to identify the presence or absence of the adverse job conditions for each participant. The number of adverse job conditions was summed for each participant to reflect the total number of adverse job conditions they experience. The maximum value for this variable was five.

Four job categories were identified: jobs with no adverse conditions; jobs with one adverse condition; jobs with two adverse conditions; and jobs with three or more adverse conditions. The number of participants reporting four or five adverse conditions was small, and poorest quality jobs were deemed to be those in which at least three adverse conditions were present.

Descriptive analysis was performed to explore the distribution of each variable within and over the survey waves. The lowest number of participants (355 participants) was in the first wave (i.e., year 2005) and the highest (604 participants) was in Wave 12 (i.e., year 2016). The sample mostly included male participants. In terms of age distribution, earlier waves included more midage participants, whereas late waves included a higher number of young participants. However, the overall number of responses from young and midage participants for all the survey waves were similar (i.e., 3,130 responses from young participants and 3,269 responses from midage participants). The majority of participants worked full time. The distribution of job adversity categories remained similar over the survey waves, with most of the participants reporting fewer than three job adversities. The median percentage of participants in each job quality category for the whole sample was as follows: no job adversity: 31.27%, one job adversity: 34.92%, two job adversities: 23.92%, and three or more job adversities: 10.29%. Sample characteristics for the first and last waves included in the analysis are presented in Table 1.

Table 1. Sample characteristics for the first and last waves

Variables	First wave (2005)	Last wave (2019)
Gender
Male	346 (97.5%)	573 (96.6%)
Female	9 (2.5%)	20 (3.4%)
Age
Younger than 25 years	98 (27.6%)	307 (51.8%)
Between 25 and 45 years	182 (51.3%)	230 (38.8%)
Older than 45 years	75 (21.1%)	56 (9.4%)
Household structure
Couple without children	119 (33.5%)	198 (33.4%)
Couple with children	121 (34.1%)	192 (32.4%)
Single parent	7 (2%)	12 (2%)
Single person	46 (13%)	84 (14.2%)
Other	62 (17.5%)	107 (18%)
Education
Year 11 and below	103 (29%)	135 (22.8%)
Year 12	45 (12.7%)	91 (15.3%)
Certificate III or IV	168 (47.3%)	239 (40.3%)
Advanced diploma or diploma	24 (6.8%)	31 (5.2%)
Bachelor or honors	14 (3.9%)	29 (4.9%)
Graduate diploma/certificate	1 (0.3%)	10 (1.7%)
Postgraduate	—	3 (0.5%)
Missing data	—	1 (0.2%)
Hours usually worked per week
35 h or more	330 (93%)	521 (87.9%)
34 h or less	25 (7%)	72 (12.1%)
Long-term health condition
Yes	51 (14.4%)	81 (13.7%)
No	304 (85.6%)	512 (86.3%)
Job quality
No adverse conditions	116 (32.7%)	158 (26.7%)
1 adverse condition	124 (34.9%)	207 (34.9%)
2 adverse conditions	74 (20.9%)	149 (25.1%)
3 or more adverse conditions	41 (11.5%)	79 (13.3%)
Mental health scale
Mean	77.40	74.87
Standard deviation	15.11	16.02
Missing data	2	4

Pearson product-moment correlations were performed for each wave to explore the bivariate relationships between the variables. Table 2 provides the results at baseline (i.e., year 2005). Inspecting the correlation over years 2005–2019 revealed no concerns about collinearity between the predictors, i.e., there was no high correlation between main independent variables, and the variance inflation factor (VIF) values associated with regression coefficients were well below 10 (the largest value was 4.40). The distribution of education remained consistent over the waves of the survey, with most of the participants indicating Certificate IV or below as their highest level of education.

Table 2. Bivariate correlations between selected variables for survey responses by construction workers at baseline

Variables	MH	JobAdv	Age	HC	F	Hrs	Exp	Edu	AdvEv	Caring	Incom	CopC	SingPar	SingPer	Other	UnEmpl	ParResp
Mental health (MH)	1	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—
Job adversities (JobAdv)	$- 0.3218$ ^b	1	—	—	—	—	—	—	—	—	—	—	—	—	—	—	—
Age group (Age)	0.0670	$- 0.0670$	1	—	—	—	—	—	—	—	—	—	—	—	—	—	—
Long-term health condition (HC)	$- 0.1713$ ^b	0.0668	0.1309^b	1	—	—	—	—	—	—	—	—	—	—	—	—	—
Female (F)	$- 0.0054$	0.0359	$- 0.0028$	0.0923^a	1	—	—	—	—	—	—	—	—	—	—	—	—
Work hours per week (Hrs)	$- 0.0047$	0.0332	0.0394	$- 0.0029$	$- 0.0677$	1	—	—	—	—	—	—	—	—	—	—	—
Years worked in current occupation (Exp)	0.1280^a	$- 0.1407$ ^a	0.6199^b	0.0594	$- 0.1005$	0.0188	1	—	—	—	—	—	—	—	—	—	—
Education (Edu)	0.0770	$- 0.0356$	0.2567^b	0.0696	0.0978^a	0.1148^b	0.0862	1	—	—	—	—	—	—	—	—	—
Adverse life event (AdvEv)	$- 0.1246$ ^a	0.0868	0.0024	0.0946	$- 0.0231$	$- 0.0261$	0.0659	0.0154	1	—	—	—	—	—	—	—	—
Caring responsibility (Caring)	$- 0.0309$	$- 0.0725$	0.0369	0.0447	$- 0.0256$	0.0561	0.0047	0.0763	$- 0.0150$	1	—	—	—	—	—	—	—
Household income band (Incom)	0.0974	$- 0.1014$	0.0146	0.0464	0.0115	0.1493^b	0.0299	0.0773	$- 0.0812$	$- 0.0106$	1	—	—	—	—	—	—
Couple with children (CopC)	0.0501	$- 0.0637$	0.0675	$- 0.0523$	0.0165	0.1755^b	0.0731	0.1799^b	$- 0.0658$	0.0360	0.0520	1	—	—	—	—	—
Single parent (SingPar)	0.0329	$- 0.0161$	0.1166^a	$- 0.0265$	0.1618^b	0.0040	0.0442	$- 0.0190$	$- 0.0094$	0.0637	$- 0.0519$	$- 0.1059$ ^a	1	—	—	—	—
Single person (SingPer)	$- 0.0519$	0.0154	0.0064	$- 0.0368$	$- 0.0684$	$- 0.0244$	$- 0.0398$	$- 0.0540$	0.0068	$- 0.0715$	$- 0.3648$ ^a	$- 0.2962$ ^b	$- 0.0684$	1	—	—	—
Household structure—other (Other)	$- 0.0258$	0.0674	$- 0.5216$ ^b	$- 0.0650$	$- 0.0028$	$- 0.2035$ ^b	$- 0.3744$ ^b	$- 0.2849$ ^b	0.0153	$- 0.0788$	0.1237^b	$- 0.3267$ ^b	$- 0.0754$	$- 0.2109$ ^b	1	—	—
Unemployment (UnEmpl)	$- 0.0234$	0.1870^b	0.0896	0.1053^a	0.0926^a	$- 0.1409$ ^b	0.1100^a	$- 0.0485$	0.1800^a	$- 0.0881$	$- 0.1158$ ^a	$- 0.2348$ ^b	$- 0.0221$	0.0465	0.0588	1	—
Parenting responsibility (ParResp)	$- 0.0296$	$- 0.0760$	0.1156^a	$- 0.0074$	0.0733	0.1958^b	0.0557	0.1698^b	$- 0.0142$	0.0392	0.0246	0.7992^b	0.0935	$- 0.1303$ ^a	$- 0.3497$ ^b	$- 0.2327$ ^b	1

a

Correlation is significant at 0.05 level (two-tailed).

b

Correlation is significant at 0.01 level (two-tailed).

The data set comprised repeated measures for individuals along the survey waves. The data included variables at both person level (demographics) and wave level (job quality aspects, work hours, and so on that could change for individuals each year along the survey waves). To account for this nested structure, longitudinal random-intercept regression models were used to examine the association between the five aspects of job quality and mental health. These models comprised two levels, i.e., responses were clustered within individuals and allowed for correlated responses for each individual over the waves (i.e., dependence of observations for each individual). The models fitted a fixed regression slope for mental health, reflecting the average within-person changes of mental health scores over the survey waves while allowing the intercept to vary between respondents to reflect the different initial mental health states of individuals. A benefit of using this type of model is that a balanced data set is not required. Consequently, attrition does not cause an issue for model fit and all the available data can be used even if individuals only participate in some survey waves. The statistical analysis was conducted using Stata version 17.

Results

Number of Adverse Job Conditions and Mental Health

Initially, three random intercept models were fitted to examine the association between the total number of adverse job conditions and mental health in different age groups. The models included variables indicating the three job adversity categories, i.e., jobs with one, two, and three or more adverse conditions (jobs with no adverse conditions served as the reference group). The covariates described previously were included in the models. Mental health was the dependent variable. The results indicate a statistically significant decline in mental health for all age groups as the number of adverse job conditions increased. However, the decline in mental health of younger workers (

< 25 years

of age) who reported one adverse job condition was not statistically significant compared with jobs with no adverse conditions. In addition, the magnitude of decline in mental health when working in a job with two adverse conditions was greater for midaged [

coefficient = - 4.42

, 95% confidence interval (CI) =

- 5.60

to

- 3.25

, and

p < 0.001

] and older workers (

coefficient = - 4.21

,

95 % CI = - 6.24

to

- 2.18

, and

p < 0.001

) compared with younger workers (

coefficient = - 2.36

,

95 % CI = - 3.85

to

- 0.88

, and

p = 0.001

).

This initial finding was systematically investigated in further random-intercept models, which were fitted in steps as presented in Fig. 1. The coefficient values and their corresponding standard errors (in parentheses) for the models are listed in Table 3.

Fig. 1. Models fitted to examine the association between job quality and mental health in different age groups.

Table 3. Age and job quality predictors of mental health

Variables	Model A	Model B	Model C	Model D
Age group
Age group 1 ( $< 25$ )	(Ref)	(Ref)	(Ref)	(Ref)
Age group 2 (25–45)	0.82^a (0.64^b)	0.6 (0.75)	1.78 (0.92)	1.52 (0.93)
Age group 3 ( $> 45$ )	3.49 (0.9)^**	2.35 (1.06)^*	3.65 (1.27)^**	3.85 (1.27)^**
Job categories based on number of adverse conditions
No adverse conditions	(Ref)	(Ref)	(Ref)	(Ref)
1 adverse condition	$- 1.46$ (0.35)^**	$- 1.45$ (0.37)^**	$- 0.83$ (0.61)	$- 0.83$ (0.61)
2 adverse conditions	$- 3.95$ (0.41)^**	$- 3.70$ (0.43)^**	$- 2.06$ (0.7)^**	$- 2.04$ (0.7)^**
3–5 adverse conditions	$- 6.93$ (0.55)^**	$- 6.38$ (0.6)^**	$- 5.17$ (0.9)^**	$- 5.15$ (0.9)^**
Job categories by age groups (interaction terms)
1 adverse condition × Age group 2	—	—	$- 0.75$ (0.81)	$- 0.74$ (0.81)
2 adverse conditions × Age group 2	—	—	$- 2.66$ (0.94)^**	$- 2.64$ (0.94)^**
3–5 adverse conditions × Age group 2	—	—	$- 2.22$ (1.29)	$- 2.15$ (1.29)
1 adverse condition × Age group 3	—	—	$- 1.45$ (1.11)	$- 1.44$ (1.11)
2 adverse conditions × Age group 3	—	—	$- 2.56$ (1.32)	$- 2.58$ (1.32)
3–5 adverse conditions × Age group 3	—	—	$- 1.47$ (1.92)	$- 1.45$ (1.92)
Years worked in current occupation	—	$- 0.01$ (0.02)	0.004 (0.02)	0.002 (0.02)
Average work hours per week	—	0.002 (0.02)	0.003 (0.02)	$- 0.005$ (0.02)
Gender
Male	—	(Ref)	(Ref)	(Ref)
Female	—	$- 4.75$ (1.82)^**	$- 4.69$ (1.82)^**	$- 4.34$ (1.83)^*
Education	—	0.15 (0.25)	0.17 (0.25)	0.15 (0.25)
Disability/long-term health condition	—	$- 2.98$ (0.48)^**	$- 2.95$ (0.48)^**	$- 2.89$ (0.48)^**
Experience of adverse life event	—	$- 1.74$ (0.31)^**	$- 1.73$ (0.31)^**	$- 1.73$ (0.31)^**
Caring for another person	—	$- 0.57$ (1.00)	$- 0.61$ (1.00)	$- 0.59$ (1.00)
Parenting responsibility	—	$- 1.3$ (0.59)^*	$- 1.32$ (0.59)^*	$- 1.36$ (0.59)^*
Household income band	—	$- 0.01$ (0.11)	$- 0.01$ (0.11)	$- 0.11$ (0.11)
Household structure
Couple	—	(Ref)	(Ref)	(Ref)
Couple with children	—	$- 0.59$ (0.64)	$- 0.57$ (0.64)	$- 0.57$ (0.64)
Single parent	—	$- 4.97$ (1.43)^**	$- 5.02$ (1.43)^**	$- 4.96$ (1.44)^**
Single person	—	$- 3.93$ (0.67)^**	$- 3.97$ (0.67)^**	$- 3.90$ (0.67)^**
Other	—	$- 1.51$ (0.67)^*	$- 1.54$ (0.67)^*	$- 1.44$ (0.67)^*

Any experience of unemployment/NILF	—	—	—	$- 1.57$ (0.65)^*
Wave	—	$- 0.08$ (0.04)	$- 0.08$ (0.04)	$- 0.08$ (0.04)

Note: ^* $p < 0.05$ ; and ^** $p < 0.01$ . Ref = reference group; and NILF = not in the labor force.

a

Coefficient value.

b

Standard error.

Model A included variables reflecting age groups (younger than 25 years old served as the reference group) and job categories (jobs with no adverse conditions was the reference group). The model indicated a significant decline in mental health among workers who reported that they experienced adverse job conditions compared with those who experienced no adverse conditions. Moreover, the magnitude of mental health decline increased with the number of adverse job conditions. The poorest jobs (three or more adverse conditions) were associated with an average 6.93-point decline in mental health scores. A four- or five-point difference on the mental health scale is considered to be “clinically relevant” (Butterworth et al. 2011b).

Controlling for covariates in Model B slightly reduced the magnitude of the coefficients; however, the relationships identified in Model A remained significant. Apart from job quality, the factors of being a female, having a long-term disability or health condition, experience of adverse life events in the last 12 months, having parenting responsibilities, being a single parent, and being a single person were all significantly associated with poorer mental health. The inclusion of survey wave numbers did not indicate any significant systematic change in mental health scores over time.

Model C included interaction terms between job quality categories and age groups to further examine the combined effect of age and total adverse job conditions on mental health. The significant associations previously identified in Model B remained significant in Model C, except for the decline in mental health for jobs with one adverse condition. Furthermore, Model C indicated that the decline in mental health experienced by midage workers who experience two adverse job conditions was significantly larger than the decline in mental health of younger workers in the same category of jobs. This was evident by the statistically significant (

p = 0.005

) interaction term between Age group 2 and jobs with two adverse conditions. In addition, the interaction between jobs with two adverse conditions and Age group 3 was borderline significant (

p = 0.052

), suggesting a weaker but similar effect for older workers.

Model D included an additional variable to control for any experience of unemployment or not being in the labor force during the survey period. Previous research has indicated that periods of unemployment can negatively impact individuals’ mental health (Paul and Moser 2009; Backhans and Hemmingsson 2012). Controlling for this potential effect in Model D slightly changed the coefficients, but the associations previously identified in Model C remained significant. In addition, any experience of unemployment or not being in the labor force was negatively and significantly associated with mental health.

As a sensitivity test, another model was fitted that regressed mental health on all the variables included in Model D and included random effects for both slope and intercept. Thus, by allowing a separate slope for each individual, this model accounted for the differences between individuals in terms of their mental health score changes over the survey waves. The significant effects in this model were identical with those found in Model D, and the coefficient differences were minor. Comparison of the goodness of fit between this model and Model D, using the Akaike information criterion and the Bayesian information criterion, indicated only a small improvement when including a random effect for slope in the model. Therefore, it was decided to proceed with Model D.

Model D indicated that workers in all age groups whose jobs have two or more adverse conditions have lower mental health scores than workers whose jobs have no adverse conditions. Moreover, the inclusion of interaction terms (between job categories and age groups) indicated that the mean decline in mental health for midage workers who worked in jobs with two adverse conditions was greater than younger workers who worked in similar jobs, and the difference in means was statistically significant (

p = 0.005

). A similar effect was observed for older workers, but the coefficient was borderline significant (i.e.,

p = 0.051

).

Fig. 2 illustrates the moderating effect of age on the association between job quality and mental health based on the estimated coefficients in Model D. The mean mental health scores and the 95% confidence intervals were estimated for each category of jobs. The slope of lines connecting the mean values indicate the mean rate of decline in mental health scores for each age group as the number of adverse job conditions increased.

Fig. 2. Moderating effect of age on the association between job quality and mental health.

The rate of decline in mental health scores was relatively steady for older workers across different adverse conditions, and the rate of decline was slightly slower when moving to jobs with three or more adverse conditions. Midage workers’ mental health had a sharp decline when working in jobs with two adverse conditions and the rate of decline remained similar when moving to jobs with three or more adverse conditions. Younger workers experienced a slower decline in mental health scores compared with other groups when working in jobs with one adverse condition or two adverse conditions, but experienced the greatest decline among all the age groups when moving to jobs with three or more adverse conditions.

Job Quality Components and Mental Health

Models were fitted for each age group regressing mental health on the five job quality components (Table 4). The association between experiencing different types of adverse job conditions and mental health was explored separately for each age group (jobs with no adverse conditions served as the reference category). The models were adjusted for the covariates included in Model D.

Table 4. Relationship between job quality variables and mental health

Variables	Model E: Age group 1 ( $< 25$ )	Model F: Age group 2 (25–45)	Model G: Age group 3 ( $> 45$ )
Adverse components of job quality
High job demand and complexity	$- 0.6$ ^a (0.7^b)	$- 1.9$ (0.66)^**	$- 3.05$ (1.21)^*
Low job control	$- 0.22$ (0.67)	$- 0.92$ (0.67)	0.01 (1.17)
Low job security	$- 4.11$ (0.62)^**	$- 3.56$ (0.52)^**	$- 2.9$ (0.85)^**
Unfairness of effort and reward	$- 1.87$ (0.55)^**	$- 1.76$ (0.46)^**	$- 1.91$ (0.83)^*
High job intensity	0.23 (1.07)	$- 2.7$ (0.94)^**	$- 3.87$ (1.5)^*
Age (at baseline)	0.03 (0.10)	0.17 (0.08)^*	0.34 (0.15)^*
Years worked in current occupation	$- 0.03$ (0.09)	$- 0.01$ (0.03)	$- 0.01$ (0.03)
Average work hours per week	$- 0.04$ (0.03)	$- 0.01$ (0.02)	0.08 (0.04)
Gender
Male	(Ref)	(Ref)	(Ref)
Female	$- 3.10$ (3.32)	$- 3.37$ (2.50)	$- 4.61$ (4.15)
Education	$- 0.17$ (0.42)	0.76 (0.38)	0.27 (0.63)
Disability/long-term health condition	$- 2.24$ (1.01)^*	$- 3.88$ (0.65)^**	$- 1.58$ (0.92)
Experience of adverse life event	$- 2.33$ (0.55)^**	$- 1.63$ (0.43)^**	$- 0.86$ (0.69)
Caring for another person	$- 0.5$ (2.06)	$- 0.19$ (1.32)	$- 3.24$ (2.05)
Parenting responsibility	$- 0.11$ (1.43)	$- 1.08$ (0.71)	0.74 (1.56)
Household income band	$- 0.26$ (0.19)	0.15 (0.16)	$- 0.42$ (0.24)
Household structure
Couple	(Ref)	(Ref)	(Ref)
Couple with children	$- 1.52$ (1.51)	$- 0.51$ (0.76)	$- 1.96$ (1.89)
Single parent	2.34 (4.79)	$- 4.64$ (1.8)^*	$- 7.00$ (2.57)^**
Single person	$- 3.54$ (1.09)^**	$- 5.33$ (1.00)^**	$- 0.16$ (1.69)
Other	$- 0.24$ (0.90)	$- 8.14$ (1.55)^**	$- 10.84$ (4.11)^**

Any experience of unemployment/NILF	$- 0.52$ (1.01)	$- 3.40$ (1.04)^**	0.08 (1.65)
Wave	$- 0.16$ (0.10)	$- 0.03$ (0.06)	0.14 (0.10)

Note: Age group 1 included 2,406 responses from 731 individuals; Age group 2 included 2,939 responses from 759 individuals; and Age group 3 included 1,007 responses from 278 individuals. ^* $p < 0.05$ ; and ^** $p < 0.01$ . Ref = reference group.

a

Coefficient value.

b

Standard error.

Model E indicated that low job security and unfairness of effort and reward (as perceived by participants) were significantly negatively associated with mental health scores in younger workers. Models F and G indicate that for midage and older workers, in addition to low job security and unfairness of effort and reward, high job demand and complexity and high job intensity were also significantly negatively associated with lower mental health scores.

Comparing the magnitude of the coefficients suggested a larger impact of low job security on younger workers’ mental health (

coefficient = - 4.11

,

95 % CI = - 5.34

to

- 2.88

, and

p < 0.001

) compared with other age groups. For older workers, the impacts of job demand and complexity (

coefficient = - 3.05

,

95 % CI = - 5.43

to

- 0.68

, and

p = 0.012

) and job intensity (

coefficient = - 3.87

,

95 % CI = - 6.82

to

- 0.92

, and

p = 0.01

) on mental health were larger compared with midage workers. For midage workers, any experience of unemployment or not being in the labor force was negatively and significantly associated with mental health.

Discussion

Job Quality and the Mental Health of Manual Construction Workers

The results indicated that the experience of adverse job conditions is negatively related to the mental health of manual construction workers when a range of covariates is controlled for. Moreover, the strength of the relationship increased with the number of adverse job conditions. The negative relationship was the strongest for those who experience three or more adverse job conditions. The relative coefficient values also suggest that, with the exception of being female and/or a single parent, experiencing three or more adverse job conditions is the strongest single factor contributing to diminished mental health. This finding is consistent with previous European studies that showed job characteristics are more strongly related to construction workers’ mental health than individual characteristic (Alavinia et al. 2007; Stattin and Järvholm 2005).

Previous research has shown that psychosocial work characteristics are related to poor mental health in professional/managerial construction workers (Love et al. 2010; Leung et al. 2011; Bowen et al. 2014). Our results show the same is true for manual workers, providing important evidence to suggest that focusing attention on the design and provision of good-quality work has considerable potential to help to improve mental health among the frontline construction workforce.

The analysis also revealed a significant interaction effect between age and the experience of job adversities. Specifically, the decline in mental health scores was accelerated in the midage group when the number of job adversities increased from one to two. The finding is consistent with research into the experience of midage workers by Strazdins et al. (2004), who report increased odds ratios for depression, anxiety, physical health problems, and poor self-rated health when high job strain is experienced at the same time as high perceived job insecurity. Strazdins et al. (2004) concluded that psychosocial risk factors in the workplace have synergistic effects, and when excessive job demands are combined with job insecurity, a threshold level is reached beyond which psychological harm occurs. The present study indicates that the threshold or tipping point for mental ill-health may be lower for midaged workers compared with those in younger or older cohorts. Reasons for this finding are unclear but previous studies have reported that job-related burnout is higher among workers aged 34–44 than for older and younger workers (Yang et al. 2017).

Researchers have argued that midlife is a complex stage of development, in which many people report they are generally satisfied with their lives, but at the same time experiencing life as stressful and sometimes feeling anxious or depressed (Arnett 2018). Sources of stress in midlife relate to financial issues, work issues, physical health, and relationships with children, spouses, or partners (Arnett 2018). Midage workers are more likely than younger or older workers to be juggling work with family demands and to have more significant financial responsibilities (Evandrou and Glaser 2004).

Further research is needed to better understand the reasons why midage construction workers appear to be more negatively impacted when they experience two adverse job conditions than younger and older workers. However, this finding has important implications because negative experiences of work stress in midlife are predictive of serious and complex health problems in later life (Nilsen et al. 2014). The results of the present study highlight the importance of taking a life course perspective to managing work-related mental health risk because exposure to factors that contribute to declining health may be experienced before a decline is observed.

Relationship between Individual Job Quality Factors and Mental Health

The results revealed that low job security and perceived unfairness of effort and reward were significant predictors of reduced mental health in all age groups, supporting the ERI model. Over the last decades, there has been a significant growth in precarious employment, resulting in short-term contracts, casual and part-time jobs, and increased outsourcing of work (Quinlan et al. 2001). Flexible forms of employment have benefited employers but generated widespread perceptions of job insecurity, which have negative impacts on workers’ health (Strazdins et al. 2004; Sverke et al. 2002).

Precarious employment is common for project-based construction workers and affects health in various ways (Turner and Lingard 2016). It can increase workers’ exposure to hazardous working conditions and contribute to material deprivation relating to income, wealth and savings; housing quality; superannuation; and others (Benach et al. 2016). The findings of the current study suggest that the potentially harmful effects of the industry’s flexible employment policies and practices should be carefully considered to minimize adverse mental health outcomes for workers at all stages in their working lives.

The findings also support the ERI theory of work health in that perceived effort-reward imbalances were linked to reduced mental health in all age groups. Relatively few studies could be found in which the ERI model was used to understand the factors impacting mental health in blue-collar samples. The finding that the mental health of manual construction workers is predicted by a perceived imbalance between effort and reward has important implications for the prevention of mental ill-health. However, more work is needed to better understand the way that job demands and rewards are experienced by manual construction workers in order to inform the design of work that is perceived to be fairer in its effort-reward balance.

For example, in a qualitative study of UK-based bus drivers, effort was found to have facets that were not captured in Siegrist’s original conceptualization, with workload and fatigue playing a particularly important role in shaping job strain and mental health (Tse et al. 2007). In the present study, perceived unfairness of effort and reward was measured using a single item (“I get paid fairly for the things I do in my job”), which is unlikely to capture the full gamut of job-related costs and gains experienced by manual construction workers.

In the present study, job control was not a significant determinant of mental health for any age group. This finding is inconsistent with previous analyses of population data contained in the HILDA data set (Bentley et al. 2015). The finding also does not support the JDC model, which posits that job control is key dimension of psychosocial work environment that protects individuals’ health. It is not clear why job control was not significantly related to mental ill-health in the construction worker sample. However, this finding does suggest there is value in conducting industry-specific studies. Calnan et al. (2004) observed that the components of the JDC and ERI models make distinct contributions to explaining the relationship between job characteristics and health in different occupational groups, reflecting the fact that what is important in some occupations may be less salient in others.

However, the absence of a significant relationship between job control and mental health in the present study should not be taken to mean that job control is not important to the mental health of manual construction workers. It is possible that the way in which control is conceptualized in the HILDA survey may not reflect the work context experienced by manual construction workers, in which autonomy is constrained by design specifications, construction technologies, project timelines, and resources, including the tools and equipment provided to workers. It is unclear whether other forms of control, for example in relation to work pace and rest breaks, may be more salient to manual/nonmanagerial construction workers, particularly in older age groups (as also discussed subsequently).

Age-related differences were also found in the job quality components that predicted mental health. Most notably, high job demand and complexity and high job intensity were linked to reduced mental health among midage and older construction workers but were not significant predictors of mental health among younger workers. This is consistent with previous research showing that compared with younger workers, older workers are more adversely affected by psychosocial job demands, including working under time pressure (De Zwart et al. 1999). Working under time pressure is a feature of working in construction jobs in which completion of work to a predetermined timeline is an essential determinant of project success (Serrador and Turner 2015). Project-based workers’ psychological health is reportedly impacted by complexity, intensity of work, and time pressure (Zika-Viktorsson et al. 2006; Tüchsen et al. 2005). Knowing that complex and demanding and/or high intensity work negatively impacts the mental health of midage and older construction workers suggests that targeted strategies are needed to ensure a good fit between workers’ resources and the job demands they experience.

A good person–job fit is a prerequisite for successful aging (Kooij 2015), and workers are known to engage in adaptive behaviors to trade-off the impacts of resource depletion associated with aging and maintain an effective fit with their job demands (Baltes and Baltes 1990). Thus, older workers can (1) actively choose the life and work goals that they will pursue to avoid stressful situations (selection); (2) seek to attain, develop, and use particular resources to help them achieve their goals (optimization); and (3) apply strategies to make up for resource losses or the limitations they may experience that are associated with aging (compensation) (Schmitt and Unger 2019). However, manual construction workers’ ability to use such strategies will be limited unless organizational support is also provided.

It is recommended that age-sensitive assessments of the psychosocial risks inherent in a job be undertaken and supports for older workers be provided where possible (Varianou-Mikellidou et al. 2019). Although job control was not found to be a significant predictor of mental health in the present study, the extant literature recommends the implementation of strategies to increase older workers’ control over the way they work (for example, with respect to the level and type of work engagement, work hours and timing of work, the pace of work, and the timing and frequency of rest breaks) (Costa and Sartori 2007; Drake et al. 2017; Griffiths et al. 2009). The potential for these forms of job control to have a positive impact should be explored in future studies.

Conclusions

This research revealed that the mental health of manual construction workers in Australia is influenced by various psychosocial job quality factors. These findings are consistent with previous research that has identified exposure to an adverse psychosocial work environment as being an important determinant of social inequalities in health (LaMontagne et al. 2013). Importantly, the study makes a unique contribution to the body of knowledge by showing that the experience of psychosocial job characteristics is age-sensitive in two ways. First, midage construction workers experience more accelerated decline in mental health compared with other age groups when the number of job adversities increases from one to two, highlighting the need to better understand the mental health experience and circumstances of midage construction workers. Second, facets of job quality that are predictive of mental health vary among workers in different age groups. The findings suggest that a one-size-fits-all solution to protect workers’ mental health is unlikely to be effective and highlight the importance of understanding the association between job quality and health using a life course perspective in which prevention and support strategies are age-sensitive and targeted. However, the life course perspective does not imply that organizational support should be withheld until mental health challenges manifest. Appropriate organizational support strategies that address key features of job quality will help to enable successful aging among manual construction workers.

The findings of the current study in relation to job control are inconclusive, but it is recommended that job control is not dismissed as an important characteristic of job quality on this basis. It appears that the content of the job control measure identified from the HILDA survey data does not reflect the specific work context of the manual construction workers. There is a need to consider the facets of control that may be appropriate and applicable to the work of manual construction workers and to consider the way that job control opportunities can be increased, particularly to alleviate some of the job demands that are negatively related to mental health among midage and older construction workers.

With the understanding of the associations between specific psychosocial job conditions and mental health for each age group of construction workers, it is important for construction organizations to eliminate such health risks through appropriate job design. Job quality should be considered as an occupational health and safety (OHS) factor that needs to be included in the organizational OHS agenda and be proactively managed. OHS risk management should go beyond issues relating to physical safety and health but include issues of mental health. Providing quality jobs that enable construction workers to maintain their physical and psychological health and well-being is critical for the construction industry to develop a sustainable workforce, which has been defined as one whose members have “the positive energy, capabilities, vitality, and resources to meet current and future organizational performance demands while sustaining their economic and mental health on and off the job” (Kossek et al. 2014, p. 299).

Limitations and Future Research

A number of limitations to the present study should be noted. First, the analysis was limited to the data captured in the HILDA data set. A comprehensive list of job quality characteristics was not included in the analysis. Further, the way that some of the variables were operationalized and measured in the HILDA survey could have had an impact on the results. In particular, it is recommended that the nature of job control, work effort, and rewards be unpacked further in qualitative studies involving manual construction workers.

Second, the analysis did not consider gender differences in experiences or relationships. This is a limitation because previous studies have indicated that female workers in the construction industry experience psychosocial risk factors differently from their male counterparts (Loosemore and Waters 2004; Goldenhar et al. 1998). Unfortunately, the number of female manual construction workers in the present sample precluded an analysis by gender. However, further research is needed to better understand how female construction workers are impacted by job quality characteristics across the course of their working lives.

Data Availability Statement

Some or all data, models, or code used during the study were provided by a third party. Direct request for these materials may be made to the provider as indicated in the Acknowledgments.

Acknowledgments

This paper uses record data from the Household, Income, and Labour Dynamics in Australia Survey. The HILDA project was initiated and is funded by the Australian Government Department of Social Services (DSS) and is managed by the Melbourne Institute of Applied Economic and Social Research (Melbourne Institute). The findings and views reported, however, are those of the authors and should not be attributed to either the DSS or the Melbourne Institute.

References

ABS (Australian Bureau of Statistics). 2006a. Australian and New Zealand standard classification of occupations (ANZSCO). 1st edition. Canberra, Australia: ABS.

Abstract

Introduction

Work and Mental Health

Work Environment as a Contributing Factor to Poor Mental Health

Job Quality and Mental Health

Knowledge Gap and Research Aim

Theoretical Background

Age-Related Work and Health Effects

Method

Data and Sampling

Measures

Control Variables

Data Preparation and Statistical Analysis

Results

Number of Adverse Job Conditions and Mental Health

Job Quality Components and Mental Health

Discussion

Job Quality and the Mental Health of Manual Construction Workers

Relationship between Individual Job Quality Factors and Mental Health

Conclusions

Limitations and Future Research

Data Availability Statement

Acknowledgments

References

Information

Published In

Copyright

History

ASCE Technical Topics:

Authors

Affiliations

Metrics

Citations

Download citation

Cited by

Figures

Other

Share

Copy the content Link

Share with email

Share

Request Username

Create a new account

Change Password

Password Changed Successfully

Verify Phone

Congrats!