Professor Brett Inder
Professor Graham Meadows
Dr Kathryn Cornwell
Department of Econometrics and Business Statistics, Monash University
Project completion year
It is well recognised that there are complex links between mental illness and labour market experiences. For example, episodes of unemployment, workplace transitions and highly stressful work environments are just some of the ways in which work can affect mental health.
The focus of this research is on effects working in the opposite direction: understanding the ways in which mental illness may affect a person’s ability to engage with employment and their workplace productivity. An understanding of the mechanisms by which mental illness affects workplace performance and of the magnitude of these effects can provide important insight into the potential value that can come from workplace programs targeting prevention of onset or recurrence and relapse of episodes of depression or anxiety disorders.
This research involves conducting an up-to-date estimation of the impact of mental illness on all aspects of labour market engagement, using the 2007 National Survey of Mental Health and Well Being. Specific interest is in the impact of mental illness on absences from work, occupational effectiveness, occupational skill category and labour income.
Data analysed for this study is from the 2007 National Survey of Mental Health and Well Being of Adults. The research involves building statistical models that measure the impact of having a mental disorder on employment outcomes in the 12 months prior to the interview. Models are also built to explain the impact of mental illness on weekly hours worked, days out of role, occupational skill level and income and loss of work hours due to time spent in care giving.
The statistical analysis showed a number of important results, summarised briefly here:
Labour force participation
Non-participants in the labour force are neither employed nor looking for work (unemployed). The results show that those experiencing a mental illness are much more likely to not be participating in the labour force. The only exception is for those with a substance use disorder, who participate at much the same rate as the overall population.
Depending on the precise way in which mental illness is captured in the models, the results show that the average person suffering a mental illness is between 5 and 7 per cent more likely to not be employed than an individual with no mental illness. This represents an unemployment rate almost double that of a person without an illness. The impact is strongest for those with at least one affective disorder (more than 8 per cent), and weaker (but still significant) for anxiety disorders. The impact on employment is also stronger for males than females, and strongest for working age adults in the age range 35-54, with comparatively small effects for those younger and older.
Affective disorders and substance use disorders are more than twice as likely amongst those from the poorest 10 per cent of the population compared to the richest 10 per cent. There is also some tendency for anxiety disorders to decline as one moves up the income/wealth scale, but the variation is much stronger for affective and substance use disorders.
Mental illness has a significant impact on earning capacity, with those experiencing an illness earning almost 5 per cent less than a person with no illness, given otherwise identical levels of education, age and socio-economic status. The income effect of illness is stronger for more severe illness (severity is measured using Kessler scores), but does not appear to vary much between types of illness.
Days of absence
The modelling suggests that after controlling for other factors, mental illness leads to an additional 2.1 days of absence from work per month. The effect is even stronger for an older worker (aged 45-64), and relatively small for younger workers, and does not vary significantly between males and females.
We estimate that the average person experiencing a mental illness will work 2.0 hours less per week than someone without an illness. The detrimental impact of illness on hours worked is much stronger for males than females, and more than twice as strong among those who are aged 35-64, with virtually no impact on hours worked for those aged under 35.
The employed in the survey are placed in broad occupational categories. Mental health conditions vary widely between different categories. For example, substance use disorders are most common among technicians and trades workers and least common among professionals. Anxiety disorders are far more common among administrative and sales roles, and least among technicians and trades workers. There is less variation in affective disorders across occupations, although the highest prevalence is among professionals.
There is no clear evidence of mental illness leading to an occupational skill level gap. In other words, it is not clear that experiencing an illness causes an individual to be more likely to work in occupations that are less skilled than their qualifications and other characteristics would predict.
Industry of employment
There is wide variation in incidence of illness by industry type. For example, substance use disorders are most prevalent in mining, construction, accommodation and food services. Anxiety disorders are most prevalent in the IT, media, financial and insurance industries. Affective disorders have very low prevalence in mining, manufacturing and the arts and recreation services sector, but are most prevalent in accommodation and food services industry, as well as in the professional and technical sector and essential services industry (electricity, gas, water, waste).
Caring for a person with a mental illness
Care giving has a substantial impact on number of hours worked. On average, the reduction in hours worked for a caregiver of someone experiencing a mental illness is greater than the reduction in hours worked for a person who is actually experiencing a mental illness. If care giving is of a day-to-day practical nature (for example, household support or driving to medical appointments), then it has a negative effect on the chances of employment, while care provided through financial support is positively associated with employment.
The total labour market cost of mental illness in Australia in terms of individuals’ forgone earnings or employers’ loss in productivity is estimated to be almost $49 billion per annum. This estimate excludes a number of costs that cannot be easily measured with available data, thus represents a lower bound on the costs.
Concentrating on just the effects relevant to those who are employed, the estimated average cost of illness in terms of reduced hours worked and days of absence from work is $2,100 per employee per annum, averaged across all employees.