National trends in the age-standardized suicide rate for males have increased from 16.2 per 100,000 population in 2011 to 18.6 per 100,000 in 2020; however, these estimates include the entire male population, whether employed or not (ABS 2022). Since employed people are known to have lower suicide rates than unemployed people, our results are plausibly consistent with them. That is, the higher overall rate (16.2 to 18.6, 2011 to 2020) seen in the overall male population (which includes unemployed males as well as those in high-risk jobs) should be higher than that of employed men, excluding construction workers (average of 13.2 over 2001-2019), and lower than that of construction workers who, although employed, have an average of 26.6/100 000 over the period 2001-2019. With this in mind, our results showing lower suicide rates among male employees in general and construction workers in particular are promising, notwithstanding the limitations below.
Although NCIS data provides the most accurate and comprehensive information on suicide deaths available in Australia, there are limitations that should be acknowledged. First, some of the occupations listed by the Coroner are ambiguous or insufficient for coding purposes. This results in 2% missing data, which is a minimal loss in this study. It is also important to note that coronary processes can be long, with some cases remaining “open” for years. Therefore, suicide cases may be underestimated in recent years. There are also variations in coronary processes that exist in state and territory jurisdictions, which may contribute to some inconsistencies and undercounting. In particular, inconsistencies in determining intent may lead to significant undercounting in some jurisdictions.24. Suicides are estimated to be underestimated by 11-16% due to an inability to judge intent (as in single occupant car crashes)24. We also point out that we have classified construction workers according to the ANZSCO occupations listed in Supplementary File 1. Not all of these occupational codes distinguish between sectors, and the category “construction worker” is likely to differ. include some workers employed in smaller sectors such as mining and energy, which could lead to measurement error in the estimates. That said, given that construction is quite a large industry, the misclassification of a small number of mining and energy workers as construction workers is unlikely to affect the results (see the sensitivity analysis).
We note that we compared the suicide rates of construction workers to those of other people in the labor force, as rates of mental illness and suicide are generally higher among the unemployed and those not in the labor force. labor force (NILF). This is known as the “healthy worker effect”25. Comparing rates in specific occupational groups to the general population would thus be compared to a high benchmark pulled up by unemployment and the NILF, thus biasing relative risk estimates downward for groups with relatively high rates in the active population. By using other employed workers, we mitigated this bias.
Implications of the results
Assuming that the observed trends are real, we can only speculate as to what might explain them, as this analysis does not directly address this question. Employed men might experience lower suicide rates compared to an increase in all men because they have a lower risk of suicide in the first place (healthy worker effect) and have higher rates of mental illness low, which will reduce the risk of suicide compared to those who are unemployed or inactive (NILF)26. In addition, employed men may be better placed to benefit from expanded State and Commonwealth suicide prevention programs27 that have demonstrated some impact on suicidal behavior in the general population and mental health services28.29.
Construction workers could experience a greater drop in suicide rates due to other factors combined with the potential impacts described in the previous paragraph. First, construction workers started from around twice the rate of other working men in the early 2000s, leaving more room for improvement. The fact that the observed trend is quite gradual and steadily declining over 19 annual estimates suggests that it is more than a regression to the mean or some other statistical aberration. The state of economic activity in the construction sector showed substantial improvement during the study period30, and may have played some role in influencing suicidal tendencies. Construction “starts” and “completions” increased significantly from 2012 to 2019, until the onset of the COVID-19 pandemic. Where there is an increasing demand for labor in building and construction, as implied by increased activity in the sector, working conditions, including exposure to psychosocial risks such than job insecurity, are improved31; this could in turn reduce the risk of suicide for construction workers32. In the absence of readily available unemployment rates in the construction sector, we can also consider national unemployment rates. In the early 2000s, unemployment was between 6 and 7% in Australia, and since the mid-2000s it has fluctuated between 4 and 6%33. Thus, trends in the sector’s economic activity and unemployment rates may have contributed to the decline in the number of suicides among construction workers.
There have also been numerous industry-wide suicide prevention initiatives in the Australian construction industry, which may have contributed to lower suicide rates. Since the 2003 Cole Royal Commission report identified suicide as a possible reason for 41% of all deaths among building and construction workers in Queensland34the Australian construction industry has demonstrated a strong commitment to suicide prevention within its workforce13.33. Programs such as MATES under construction13 and the similar Incolink Blue Hats program have implemented multimodal approaches to reduce mental health stigma, increase suicide prevention knowledge, and promote help-seeking and help-giving behaviors.12,13,35. Other programs have targeted high-risk workers in the sector, such as young apprentices in rural and regional areas11.12. Evaluation of these programs suggests reduced stigma, increased awareness of self-reported suicide, and positive attitude change toward help-seeking11,13,36,37,38,39. However, these programs have operated state by state rather than nationwide. MATES has had the greatest reach in Queensland and also operates in Western Australia, South Australia, New South Wales and the Northern Territory. It is estimated that >300,000 workers participated in MATES, and more than 23,000 of these workers volunteered to be trained as on-site “connectors”, their role being to connect workers in distress to help them)40. Given the scope of construction-specific programs and targeted interventions, it is possible that this has contributed to the prevention of suicide among construction workers.
Despite the apparent decline in suicide rates among construction workers, they remain at high risk compared to other workers. Our results suggest that the combined effects of all relevant intervention efforts – from the population level to the sector-specific level – can have a positive impact. The Australian construction industry’s continued efforts to prioritize the mental well-being of employees, promote helping behaviors and seek ways to reduce work-related stressors could likely contribute to the continued decline in the suicide rate in the area.