Change points in Australian suicide rates: Trends from 2000 to 2022

Abstract

Objective:

There is concern that suicides in Australia have been increasing, especially among young women. Our aim was to describe the long-term trends in suicide rates, identifying any change points among all suicides and among age-sex population groups.

Method:

We extracted annual Australian suicide data from 2000 to 2022 from the General Record of Incidence of Mortality. We used join point regression to model change points in age-standardised suicide rates over time for the total population, males, females and for eight age-sex strata.

Results:

Age-standardised population suicide rates declined by 4.4% per year (95% confidence interval: −6.1 to −2.6) between 2000 and 2005 and then rose by 2.0% per year until 2018 (95% confidence interval: 1.4 to 2.5). No change was observed after this. Similar patterns were observed for males and females. When disaggregated by age and sex, a more complex picture emerged. Suicides declined in the early part of the century for some groups but not others. Most age-sex groups experienced an increase in suicides from the mid-to-late 2000s. Suicides declined between 2020 and 2022 for 15- to 24-year-old men by 16.0% per year (95% confidence interval: −24.0 to −7.4). In all other age-sex groups, suicides either plateaued or continued to rise.

Conclusion:

The broad trends in suicides observed in the total population and in males and females mask more complex patterns occurring in some age and sex groups. Understanding these long-term patterns is critical to informing interventions to reduce suicide.

Keywords

Suicide trends join point regression Australia

Introduction

In Australia, the number of suicides has been increasing since the mid-2000s (Spittal et al., 2024) with the increase among young women (<25 years) being of particular concern (Stefanac et al., 2019). Annual counts on their own, however, are hard to interpret because these do not account for changes in the population size, which has also increased in Australia over this period, or changes in the age structure of the population. One example of how the changing age structure could influence suicide counts is when the average age of the population increases, as is currently happening in Australia (Australian Bureau of Statistics, 2020). Suicide risk is associated with older age, especially among men, so any increase in the number of suicides may reflect more people aging into a risky age group. These two issues can be addressed by analysing age-standardised rates as this adjusts for changing population size by holding the age structure constant across time. Analysis of age-standardised rates shows clearer evidence that suicides have been increasing in Australia (Australian Institute of Health and Welfare, 2024), but does not elucidate precisely when suicide rates changed and whether this is uniform for all age and sex groups.

The aim of this research was to describe long-term trends in suicide rates in Australia. We analysed Australian annual age-standardised suicide rates to identify the points in time when suicide rates changed and for which population groups. We focused on the period from the year 2000 onwards.

Methods

Datasets

In Australia, death certificates are coded to the International Statistical Classification of Diseases and Related Health Problems (version 10) by the Australian Bureau of Statistics. These deaths are aggregated into counts by year, age and sex for different causes by the Australian Institute of Health and Welfare. The data, referred to as the General Record of Incidence of Mortality (GRIM), are freely available for download from the Institute’s website (Australian Institute of Health and Welfare, 2024). We downloaded the GRIM data for suicide (deaths coded X60-X84 or the equivilent codes for earlier versions). For all suicides between 2000 and 2022, we extracted annual counts by age and sex for people aged 15 and older as well as the corresponding population sizes in the same years. We grouped age into four categories: 15–24 years, 25–44 years, 45–64 years and ⩾65 years. From the same source, we obtained the Australian Standard Population in 2001.

Analysis

We modelled suicide trends from 2000 to 2022 using join point regression (Muggeo, 2003). This technique models time trends in the data, seeking to identify points when the time trend changes (referred to as change points). If there are no change points, the time trend is a straight line through the data. Where there are one or more change points, these are the points where the trend line changes.

We estimated join point models on the age-standardised rates for all suicides, male suicides, female suicides and suicides in the eight age-sex strata. Age-standardised rates were calculated using direct standardisation to the 2001 Australian standard population aged 15 and older (Australian Bureau of Statistics, 2013). The standard population was specified in 5-year bands from 15 to 19 years through to 80 to 84 years and then ⩾85 years. For analyses by age groups (15–24, 25–44, 45–64, ⩾65 years), rates were age-standardised to the 2001 population in the corresponding age group also using 5-year bands. The model was fitted using linear regression with the log age-standardised rate as the outcome. The results from all models were transformed so that the slopes could be interpreted as the annual percent change (APC). To arrive at a final model in each group, we trialled regressions from zero to seven join points and chose the best fitting model according to the Bayesian Information Criterion (Muggeo and Adelfio, 2011).

Data management and analysis was performed in R (V4.4.2). Join point regression and model comparison was conducted with the Segmented Package (v2.0.1) (Muggeo and Muggeo, 2017).

Results

Overall, there were 60,768 suicides in people aged 15 and older in Australia between 2000 and 2022. Of these, 46,464 were among men and 14,304 among women. Table 1 shows the number of suicides and the suicide rates by sex and by the age-sex strata in beginning, middle and end periods of the study. The age-standardised rates were 15.6, 13.1 and 15.3 per 100,000 for the population aged 15 and older in 2000, 2011 and 2022, respectively.

Table 1.

Number of suicides and age-standardised suicide rate at the beginning, middle and end of the study for people aged 15 and over.

	Number of suicides			Rate per 100,000
	2000	2011	2022	2000	2011	2022
All people	2,358	2,373	3,236	15.6	13.1	15.3
Males	1,856	1,802	2,447	25.1	20.3	23.5
Males 15–24 years	264	242	250	20.0	15.1	15.0
Males 25–44 years	920	702	916	31.9	22.1	25.1
Males 45–64 years	438	578	830	20.1	21.1	27.4
Males ⩾65 years	234	280	451	23.5	20.0	21.8
Females	502	571	789	6.6	6.3	7.4
Females 15–24 years	75	93	118	5.9	6.1	7.6
Females 25–44 years	238	209	285	8.2	6.6	7.7
Females 45–64 years	118	185	243	5.5	6.6	7.8
Females ⩾65 years	71	84	143	5.4	4.9	5.9

All suicides

The suicide rate fluctuated over time with change points identified in 2005 and 2018 (Figure 1 and Table 2). Between 2000 and 2005, rates declined by 4.4% per year (95% confidence interval [CI] = [−6.1, −2.6]), from 15.6 to 13.0 suicides per 100,000. The age rates then increased by 2.0% per year until 2018 (95% CI = [1.4, 2.5]) to 15.8 suicides per 100,000. From 2018 onwards, there was no evidence of a change in the rates (APC = −2.3, 95% CI = [−5.6, 1.1]).

Figure 1.

Age-standardised suicide rates per 100,000, aged 15 and over.

Table 2.

Annual percent change for people aged 15 and over.

Period	Annual percentage change	95% confidence interval
All suicides
2000–2005	−4.4	−6.1 to −2.6
2005–2018	2.0	1.4 to 2.5
2018–2022	−2.3	−5.6 to 1.1
Male suicides
2000–2006	−4.3	−6.2 to −2.4
2006–2019	1.7	1.0 to 2.3
2019–2022	−2.4	−6.0 to 1.4
Males 15–24 years
2000–2009	−4.4	−5.9 to −2.9
2009–2020	4.6	3.0 to 6.2
2020–2022	−16.0	−24.0 to −7.4
Males 25–44 years
2000–2006	−5.6	−8.6 to −2.6
2006–2022	0.7	−0.01 to 1.3
Males 45–64 years
2000–2022	1.7	1.2 to 2.2
Males ⩾65 years
2000–2005	−3.0	−6.0 to −0.7
2005–2022	0.6	0.1 to 1.0
Female suicides
2000–2004	−5.0	−7.3 to −2.6
2004–2017	3.1	2.4 to 3.7
2017–2022	−1.7	−3.5 to 0.2
Females 15–24 years
2000–2002	−15.0	−25.0 to −2.0
2002–2009	1.0	2.6 to 4.8
2009–2011	16.0	−11.0 to 53.0
2011–2022	1.4	−0.5 to 3.2
Females 25–44 years
2000–2003	−3.1	−7.9 to 1.9
2003–2004	−21.0	−29.0 to −12.0
2004–2007	6.8	1.6 to 12.0
2007–2011	−0.5	−3.6 to 2.8
2011–2014	5.8	0.6 to 11.0
2014–2022	0.1	−1.0 to 1.2
Females 45–64 years
2000–2017	2.7	1.8 to 3.7
2017–2022	−3.3	−9.2 to 3.1
Females ⩾65 years
2000–2010	−0.8	−2.6 to 1.1
2010–2017	4.7	1.4 to 8.0
2017–2022	−2.5	−6.3 to 1.5

Male suicides

Among males, the pattern was similar to that described for the total population. Change points were identified in 2006 and 2019 (Figure 2(a)). In the 2000 to 2006 period, rates declined by 4.3% per year (95% CI = [−6.2, −2.4]), from 25.1 to 20.1 suicides per 100,000. In the 2006 to 2019 period, rates increased by 1.7% per year (95% CI = [1.0, 2.3]) to 25.2 suicides per 100,000. In the final period, there was no evidence of a change in rates (APC = −2.4, 95% CI = [−6.0, 1.4]).

Figure 2.

Males, age-standardised suicide rates per 100,000, aged 15 and over.

In the 15–24 male age group, change points were identified in 2009 and 2020 (Figure 2(b)). In the 2000 to 2009 period, rates declined by 4.4% per year (95% CI = [−5.9, −2.9]), from 20.0 to 13.2 suicides per 100,000, but then increased over the 2009 to 2020 period by 4.6% per year (95% CI = [3.0, 6.2]) to 21.3 suicides per 100,000. In the 2020 to 2022 period, rates declined sharply by 16.0% per year (95% CI = [−24.0, −7.4]) to 15.0 suicides per 100,000.

Among males aged 25–44, in contrast to all males and the 15- to 24-year age group, just one change point was identified, in 2006 (Figure 2(c)). Between 2000 and 2006, rates declined by 5.6% per year (95% CI = [−8.6, −2.6]) from 31.9 to 21.4 suicides per 100,000. After this, there was no evidence of an change in rates (APC = 0.7, 95% CI = [−0.01, 1.3]).

In the 45–64 male age group, no change points were identified (Figure 2(d)). Suicide rates increased by 1.7% per year over the whole period (95% CI = [1.2, 2.2]), from 20.1 to 27.4 suicides per 100,000.

Finally, for males aged ⩾65 years, one change point in 2005 was identified (Figure 2(e)). From 2000 to 2005, rates declined by 3% per year (95% CI = [−6.0, −0.7]) from 23.5 to 18.3 suicides per 100,000. Rates then increased by 0.6% per year after 2005 (95% CI = [0.1, 1.0]), to 21.8 suicides per 100,000.

Female suicides

The pattern for all females was similar to that observed for all males, with two change points identified as occurring in 2004 and 2017 (Figure 3(a)). In the first period, 2000 to 2004, rates declined by 5.0% per year (95% CI = [−7.3, −2.6]), from 6.6 to 5.4 suicides per 100,000. Between 2004 and 2017, rates increased by 3.1% per year (95% CI = [2.4, 3.7]) to 8.3 suicides per 100,000. In the 2017 to 2021 period, there was no evidence of a change in rates (APC = −1.7, 95% CI = [−3.5, 0.2]).

Figure 3.

Females, age-standardised suicide rates per 100,000, aged 15 and over.

In the 15−24 female age group, three change points were identified: 2002, 2009 and 2011 (Figure 3(b)). One of these periods (2000–2002) was associated with a decline in rates, one was associated with an increase (2002–2009), and in the remaining periods, there was no evidence of a change in trend (2009–2011 and 2011–2022). Between 2000 and 2002, rates declined by 15.0% per year (95% CI = [−25.0, −2.0]) from 5.9 to 4.3 suicides per 100,000. Between 2002 and 2009, rates increased by 1.0% per year (95% CI = [2.6, 4.8]) to 4.4 suicides per 100,000. There was no evidence of a change in rates in the final two periods (2009–2011: APC = 16.0, 95% CI = [−11.0, 53.0]; 2011–2020: APC = = 1.4, 95% CI = [−0.4, 3.2]).

In the 25–44 female age group, five change points were identified: 2003, 2004, 2007, 2011 and 2014 (Figure 3(c)). Only one period (2003–2004) was associated with a decline in rates and two periods were associated with an increase (2004–2007 and 2011–2014). There was no evidence of a change in trend in the other periods (2000–2003, 2007–2011, 2014–2022). In the 2003 to 2004 period, rates declined by 21% (95% CI = [−29.0, −12.0]), from 7.3 to 5.7 per 100,000. In the 2004 to 2007 period, rates increased by 6.8% per year (95% CI = [1.6, 12.0]) to 6.7 per 100,000 and in the 2011 to 2014 period, rates increased by 5.8% per year (95% CI = [0.6, 11.0]) from 6.6 to 8.0 per 100,000.

In the 45–64 female age group, one change point was identified in 2017 (Figure 3(d)). From 2000 to 2017, rates increased by 2.7% per year (95% CI = [1.8, 3.7]), from 5.5 per 100,000 to 10.3 per 100,000. There was no evidence of a change in rates after this (APC = −3.3, 95% CI = [−9.2, 3.1]).

Finally, in the ⩾ 65 female age group, two change points were identified in 2010 and 2017 (Figure 3(e)). Only the 2010 to 2017 period was associated with a change in rates over time. In this period, rates increased by 4.7% per year (95% CI = [1.4, 8.0]), from 4.6 to 6.5 per 100,000.

Discussion

We examined trends in Australian age-standardised suicide rates in people aged 15 and older by sex and age group. Our primary interest was in identifying any periods in time between 2000 and 2022 where the suicide rate changed. For people aged 15 years and older, we identified two change points in the suicide rate. The first change point occurred in 2005, characterised by a sharp decline in the suicide rate from 2000 to 2005 and then a long-running increase until around 2018 (the second change point). When considering sex differences, males had a higher suicide rate than females, but similar directional trends and change points were observed in overall male and female rates across the study period. More heterogeneity was observed when the data were broken down by age group and sex.

The results supported a general decline in suicide rates from study beginning (2000) until around 2005, reflected by significant annual percentage reductions in the overall male rate (2000–2006), female rate (2000–2004) and in the combined rate (2000–2005). These overall effects were repeated in several sex and age-groups. For males, annual declines across a similar period were observed for those aged 15–24 years (2000–2009), 25–44 years (2000–2006) and ⩾65 years (2000–2005). Annual declines across a similar period were observed for females aged 15–24 (2000–2002) and 25–44 (2003–2004).

The modelling also indicated a long running increase in the suicide rate from around 2005 until around 2018. Significant annual increases were apparent in the male overall model (2006–2019) and the female overall model (2004–2017), as well as in the combined model (2005–2018). The overall effects were reflected across many sex and age group categories, with a long-running upward shift in the suicide rate apparent in three of the male groups and three of the female groups. However, there was notable variability in when the long-term increase occurred exactly. For males, long-term increases occurred among those aged 15–24 (2009–2020), 45–64 years (2000–2022) and ⩾65 years (2005–2022). For females, annual increases occurred among those aged 25–44 years over the 2004–2014 period, which was characterised by a period of increasing rates (2004–2007), then a plateauing (2007–2011) followed by another period of increasing rates (2011–2014). Stable long-term increases were also observed for females aged 45–64 years (2000–2017) and ⩾65 years (2010–2017).

Overall models also indicated a recent steadying of rates for all people (2018–2022), for male suicide rates (2019–2022) and female suicide rates (2017–2022). That is, rates appear to have remained unchanged over these periods. In the smaller age-sex groups, a large downwards trend was observed only among males aged 15–24 years (2020–2022).

Trends in suicide rates can shift with societal changes. Illustratively, past spikes in the Australian suicide rate have coincided with Great Depression in the 1930s (Bastiampillai et al., 2020), greater public access to sedatives in the 1960s (Oliver and Hetzel, 1972) and the Global Financial Crisis (GFC) in 2007 (Milner et al., 2014). Relatedly, national suicide rates are often sensitive to national unemployment rates (Milner et al., 2013; Stack and Haas, 1984). Here, the minimum change point observed (in 2005) co-occurred with the end of a long-running decline in the Australian unemployment rate (Karanassou and Sala, 2010). Likewise, the upward shift in the suicide rate observed after the 2005 change point is consistent with economic and social impacts of the GFC, including a substantial rise in unemployment in Australia (Waring and Lewer, 2013). The extended rise could also be associated with mental health impacts from electronic device use, an issue on the rise since the mid-2000s (Spittal et al., 2024). Notably, the overall rates clearly hide heterogeneity in the timing and strength of trends across male and female age groups, underlying the complexity of aligning societal events with suicide rates.

The rising rates we observed between 2005 and 2018 stand in contrast to international trends. A similar study to ours using the World Health Organization (WHO) and Global Burden of Disease data found that the worldwide trends could be characterised by a 2.2% reduction per year between 2000 and 2009, a larger 3.0% reduction per year between 2009 and 2016 and then a levelling off until 2019 (Ilic and Ilic, 2022). Similarly, the most recent Global Burden of Disease study shows a reduction in age-standardised suicide rates between 1990 and 2021 (GBD 2021 Suicide Collaborators, 2025). Internationally, many countries observed either a steadying of suicide rates from early 2020 or in some cases a decline in rates during this period (Gerstner et al., 2022; Pirkis et al., 2022; Pirkis et al., 2021). This pattern coincides with the onset of the COVID-19 pandemic. The exception was Japan, which experienced an increase in suicides for most age and sex groups (Spittal, 2022).

Our study has strengths and limitations. Strengths include the use of comprehensive national data over a 23-year period, the ability to stratify the analysis by age and sex groups and use of a modern epidemiological method to identify change points. Limitations include a change in the process used by the Australian Bureau of Statistics to capture suicides in 2008 (Australian Bureau of Statistics, 2023) due to concerns that suicides were previously being undercounted (De Leo, 2007). In this revised process, additional time is allowed for intent to be determined, with recent years treated as preliminary counts and subject to revision. Some of the increase in suicides during this period may be attributable to the Bureau’s change in methodology. Relatedly, because 2021 and 2022 are preliminary counts, the count of the number of suicides identified in these years may increase as the data are revised. If this happens, the steadying of rates we observed in recent years may instead become an increase. Although the Australian Bureau of Statistics codes causes of death to a consistent national standard, it is possible that some suicides may be misclassified as unintentional or undetermined intent. It is unknown how many suicides are misclassified, although the number of deaths that are reassigned to suicide during the revision process is generally small – between 5 and 25 deaths per year (Australian Bureau of Statistics, 2023) – suggesting that the overall number of misclassified suicides is unlikely to change the overall patterns observed here. No reliable long-term data are available to identify suicide trends among Aboriginal and Torres Strait Islander groups nor are there reliable denominator data for these groups. These groups have a higher suicide rate than most other groups in the Australian population and may have different trends over time. Finally, because we have analysed aggregate data there is a risk of the ecological fallacy. Our speculative explanations for changes in rates, e.g., the occurrence of the GFC (Chang et al., 2013) may not hold for individuals; i.e., we do not know whether those who died by suicide were directly impacted by the financial crisis.

In summary, our analysis of Australian suicide trends from 2000 to 2022 revealed turning points around 2005, when the overall suicide rate was at its lowest, and around 2018, when the overall suicide rate was at its highest. Aspects of this pattern were observed for some, but not all age-sex groups. This highlights the importance of ongoing monitoring of suicide rates in age and sex groups as the overall pattern may be different to the pattern in these groups.

Footnotes

Acknowledgements

None.

Author Contributions

M.J.S. was responsible for the study concept and design. L.R. acquired the data and did the statistical analyses. All authors interpreted the data. L.R. and M.J.S. drafted the manuscript. All authors revised and critically analysed the manuscript for important intellectual content. All authors had full access to the data used in the study. M.J.S. had final responsibility for the decision to submit for publication.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship and/or publication of this article: The research is funded by an Investigator Grant to Matthew Spittal (GNT2025205) by the National Health and Medical Research Council.

Ethical Approval and Informed Consent Statements

Ethical approval for this study was obtained from the University of Melbourne’s Low and Negligible Risk Committee (ID 2023-27242-41349-3).

Consent to Participate

Consent to participate was waived by the Low and Negligible Risk Committee as it was not practicable or possible to obtain consent.

Consent for Publication

Not applicable. This study does not contain any data from any individual person.

ORCID iDs

Angela Clapperton

Matthew J Spittal

Data availability

Data are available to download from the Australian Institute of Health and Welfare ().

References

Australian Bureau of Statistics (2013) National, State and Territory Population: Standard Population Data Cube. Available at: www.abs.gov.au/statistics/people/population/national-state-and-territory-population/sep-2024/31010DO003_200106.xlsx (accessed 9 April 2025).

Australian Bureau of Statistics (2020) Twenty Years of Population Change. Available at: www.abs.gov.au/articles/twenty-years-population-change (accessed 7 January 2025).

Australian Bureau of Statistics (2023) Technical Note: Causes of Death Revisions Methodology. Available at: www.abs.gov.au/methodologies/causes-death-australia-methodology/2022#technical-note-causes-of-death-revisions-methodology (accessed 28 January 2025).

Australian Institute of Health and Welfare (2024) General Record of Incidence of Mortality (GRIM) Books 2022: Intentional Self-harm (Suicide). Available at: www.aihw.gov.au/reports/life-expectancy-deaths/grim-books/contents/grim-excel-workbooks (accessed 7 May 2024).

Bastiampillai

Allison

Looi

, et al. (2020) Why are Australia’s suicide rates returning to the hundred-year average, despite suicide prevention initiatives? Reframing the problem from the perspective of Durkheim. The Australian and New Zealand Journal of Psychiatry 54: 12–14.

Chang

S-S

Stuckler

Yip

, et al. (2013) Impact of 2008 global economic crisis on suicide: Time trend study in 54 countries. BMJ 347: f5239.

GBD 2021 Suicide Collaborators (2025) Global, regional, and national burden of suicide, 1990-2021: A systematic analysis for the Global burden of disease study 2021. The Lancet Public Health 10: e189–e202.

De Leo

(2007) Suicide mortality data need revision. Medical Journal of Australia 186: 157–158.

Gerstner

Narváez

Leske

, et al. (2022) Police-reported suicides during the first 16 months of the COVID-19 pandemic in Ecuador: A time-series analysis of trends and risk factors until June 2021. Lancet Regional Health Americas 14: 100324.

10.

Ilic

(2022) Worldwide suicide mortality trends (2000-2019): A joinpoint regression analysis. World Journal of Psychiatry 12: 1044.

11.

Karanassou

Sala

(2010) Labour market dynamics in Australia: What drives unemployment? Economic Record 86: 185–209.

12.

Milner

Morrell

LaMontagne

(2014) Economically inactive, unemployed and employed suicides in Australia by age and sex over a 10-year period: What was the impact of the 2007 economic recession? International Journal of Epidemiology 43: 1500–1507.

13.

Milner

Page

LaMontagne

(2013) Duration of unemployment and suicide in Australia over the period 1985–2006: An ecological investigation by sex and age during rising versus declining national unemployment rates. Journal of Epidemiology and Community Health 67: 237–244.

14.

Muggeo

(2003) Estimating regression models with unknown break-points. Statistics in Medicine 22: 3055–3071.

15.

Muggeo

Adelfio

(2011) Efficient change point detection for genomic sequences of continuous measurements. Bioinformatics 27: 161–166.

16.

Muggeo

MVM

(2017) Package ‘segmented’. Biometrika 58: 516.

17.

Oliver

Hetzel

(1972) Rise and fall of suicide rates in Australia: Relation to sedative availability. Medical Journal of Australia 2: 919–923.

18.

Pirkis

Gunnell

Shin

, et al. (2022) Suicide numbers during the first 9-15 months of the COVID-19 pandemic compared with pre-existing trends: An interrupted time series analysis in 33 countries. Eclinicalmedicine 51: 101573.

19.

Pirkis

John

Shin

, et al. (2021) Suicide trends in the early months of the COVID-19 pandemic: An interrupted time-series analysis of preliminary data from 21 countries. The Lancet Psychiatry 8: 579–588.

20.

Spittal

(2022) COVID-19 and suicide: Evidence from Japan. Lancet Regional Health-western Pacific 27: 100578.

21.

Spittal

Mitchell

Clapperton

, et al. (2024) Age, period and cohort analysis of suicide trends in Australia, 1907–2020. The Lancet Regional Health 51: 101171.

22.

Stack

Haas

(1984) The effect of unemployment duration on national suicide rates: A time series analysis, 1948–1982. Sociological Focus 17: 17–29.

23.

Stefanac

Hetrick

Hulbert

, et al. (2019) Are young female suicides increasing? A comparison of sex-specific rates and characteristics of youth suicides in Australia over 2004–2014. BMC Public Health 19: 1389.

24.

Waring

Lewer

(2013) The global financial crisis, employment relations and the labour market in Singapore and Australia. Asia Pacific Business Review 19: 217–229.