Occupation outcomes for graduates of STEM and other fields, by diversity groups

Understanding the multiple barriers underrepresented groups face, and how they interact, is important in supporting current and future STEM graduates.

Gender, race and class are not independent from other aspects of life. Supporting girls and women facing further challenges to STEM participation, including those with disability and those from culturally and linguistically diverse backgrounds, is critical to ensuring a diverse and inclusive STEM workforce that is representative of our society.

The Australian Bureau of Statistics has measured how people with intersectional identities moved through STEM pathways, from their university graduation in 2011 to their occupation in 2021.

 

Data insights

Occupation outcomes by year

This dataset followed people who attained a higher education qualification in 2011. By 2012-13, the proportion of STEM graduates in STEM occupations peaked.

In that year (2012-13), most men STEM graduates were working in STEM occupations (59%). A much lower percentage of women STEM graduates were also working in STEM occupations (33%).

After 2012-13, the proportion of STEM graduates working in STEM occupations decreased.

For both women and men STEM graduates, health-qualified occupations become more prevalent in the years after graduation.

• For women, 4% of 2011 graduates were in health occupations in 2012-13, but this increased to 18% in 2020-21.
• For men, 1% of 2011 graduates were in health occupations in 2012-13, compared to 7% in 2020-21.

This shows that attraction to STEM occupations is low for women STEM graduates. It also shows that while STEM occupations do retain STEM graduates over a 10-year timeline, there is a gradual decrease over time for both genders.

2011 qualifications among underrepresented groups

This analysis shows that diversity groups were underrepresented in the cohort of 2011 graduates.

An example was Aboriginal and/or Torres Strait Islander peoples, who comprised 1.2% of the cohort despite making up about 2.7% of the Australian population in 2011 (ABS 2024). 

This gives important context for the underrepresentation of these groups in STEM and other careers. For example, their outcomes after graduation may be similar to others, but they may be less likely to complete university education.

When discussing the proportions who complete STEM qualifications (as opposed to health or non-STEM qualifications), the rate was consistent across demographic groups in the 2011 cohort:

• 16% of the total graduate cohort completed a STEM qualification.
• When broken down further, most demographic groups had 15% to 18% of people completing a STEM qualification.
• Aboriginal and/or Torres Strait Islander people were lower than this, with 9% of graduates completing a STEM qualification, and the rest opting for health or non-STEM qualifications.

STEM participation, by diversity or location group

Analysis of outcomes for intersectional groups is shown below. It highlights the pervasive impact of gender on STEM occupation outcomes.

For women who graduated with a STEM qualification in 2011 and were born in another country, STEM occupation outcomes 10 years later in 2021 were similar to women who were born in Australia.

• 31% of women born overseas who graduated with a STEM qualification in 2011 were working in a STEM occupation in 2021.
• This is the same proportion as for women born in Australia or not stated (31%).

For men who graduated with a STEM qualification and were born in another country, a lower percentage were working in STEM occupations in 2021 (54%) compared to men born in Australia or not stated (56%).

These proportions and relationships are similar for the language and disability diversity groups. However, there is a slightly larger difference between men graduates with a disability (47% working in STEM occupations) and men graduates with no disability (54% working in STEM occupations).

For Aboriginal and/or Torres Strait Islander people who graduated with a STEM qualification in 2011:

• There was a higher proportion of women working in STEM occupations in 2021 (34%) than non-Indigenous women (30%).
• There was a lower proportion of men working in STEM occupations in 2021 (40%) than non-Indigenous men (56%). This difference of 16 percentage points between Aboriginal and/or Torres Strait Island men and non-Indigenous men is the largest of any group shown in this analysis.

Socio-economic status above uses the quintiles of Socio-economic Indexes for Areas Index of Relative Socio-economic Advantage and Disadvantage (SEIFA IRSAD). Quintile 1 is the group living in relatively most disadvantaged areas and quintile 5 is the group living in the relatively most advantaged areas. 

When looking at people who graduated with a STEM qualification in 2011 by socioeconomic status group in 2021:

• Women in quintile 5 (relative most advantaged) were the least likely group to be working in STEM occupations in 2021 (27%).
• Interestingly, men living in the relatively most advantaged and disadvantaged socioeconomic status areas had the lowest proportions working in STEM occupations.
• Men in quintiles 1 and 5 were the least likely groups to be working in STEM occupations in 2021 (52% and 53% respectively).

When looking at people who graduated with a STEM qualification in 2011 by their remoteness (location) group in 2021:

• Women in outer regional, remote or very remote areas were more likely to be working in STEM occupations in 2021 (35%) than women in inner regional areas (28%) or major cities (31%).
• The opposite is true for men. Men in major cities and inner regional areas were slightly more likely to be working in STEM occupations (56% and 55% respectively) than men in outer regional, remote or very remote areas (53%).

About the data

The occupations on this page use Australian and New Zealand Standard Classification of Occupations (ANZSCO). Our department commissioned the Australian Bureau of Statistics (ABS) to carry out this longitudinal study. 

The ABS looked at the occupational outcomes of the 2011 cohort of university graduates for the following 10 years through to 2021. To do this, the ABS analysed the Person Level Integrated Data Asset (PLIDA, formerly Multi-Agency Data Integration Project or MADIP) which is a linked dataset providing anonymised and aggregated analysis of the following:

• 2011 higher education data
• 2021 Census of Population and Housing (Census) data
• 2010–11 to 2020–21 personal income tax income tax return (PIT ITR) data 

This allowed the ABS to determine demographic and occupation details through the years from 2011 until 2021. For each chosen variable of analysis, the outcomes have been explored using one of these datasets. Coverage of the 2011 graduate cohort in each of the linked datasets differs, because differences between the data collected in those datasets and the PLIDA “spine” (key linking information). For more information, see ABS 2024.

Not all demographics are captured in this analysis (for example, those of diverse gender identity), either because of small numbers or lack of available data. The analysis is done by each demographic group and gender, and it does not reflect multiple demographic indicators (for example, people with disability located in a metro area). It also does not represent those in the VET education pathway, although analysis of these data is possible in the future.  The data is presented by proportion and includes groups of vastly different sizes. This indicates that number changes in smaller groups will show greater impacts (to proportions or percentage) than number changes in larger groups.

We have grouped the qualifications and occupations broadly into STEM, non-STEM and health. Determinations between STEM, health and non-STEM were based on the 2021 Census of Population and Housing, see STEM Equity Monitor methodology for more information.

This analysis covers 10 years of graduates’ careers and builds on the previous output of the STEM Equity Monitor, which showed only the initial 5 years of graduates’ careers. Changes to the size of the 2011 cohort between the current and previous analysis are because of improvements to the PLIDA “spine” (key linking information), and changes to the characteristics of people, such as those recorded as deceased or visiting from overseas in the 2021 Census were removed from the cohort.

Read about the ABS’ Person Level Integrated Data Asset

Read more about our methodology.