Go to top of page

2 The changing APS workforce

The APS workforce has been undergoing substantial change in its size and composition. Wider demographic trends, notably the ageing of the population and labour force, will also impinge significantly on the APS as both an employer and a service provider, because of the ageing of its client base.

The following analysis of key workforce characteristics is drawn from APSED, complemented by the research commissioned by MAC on superannuation.

Size and shape of the APS 4

The APS workforce comprises some 123,494 employees (112,123 ongoing and 11,371 non-ongoing), the majority (75%) working in the ten largest agencies.5 One-third work in Canberra (with almost three-quarters in the ACT, New South Wales and Victoria combined).

Staffing numbers have reduced significantly over recent decades, as shown in Figure 1 below.6

Figure 1: APS staff, 1987 to 2002

Source: APSED

The main reductions in the period from the mid to late 1980s came from the transfer of functions and staff into other Commonwealth entities, and from corporatisation and privatisation. The period between 1996 and 1999 saw a rapid decline in absolute numbers (from 143,236 to 113,627), reflecting a time of significant functional cuts, efficiency improvements and market testing leading to contracting out of functions.

More recently, the long-term decline in the size of the APS plateaued out and then reversed somewhat, with an increase of 9% over the last three years (from 113,627 at end June 1999 to 123,494 at June 2002).

Age profile

The APS workforce is ageing and, on average, is now some four years older than a decade ago. The median age is now 41 years (43 for males and 39 for females), compared to 37 (39 and 34 years respectively) ten years ago.

Figure 2: Age distribution of ongoing staff, 1992 and 2002

Source: APSED

The ageing trend is consistent across all states and territories (although the Northern Territory has a slightly younger median age of 40 and Queensland a slightly older one of 42).

There are significant variations in the age profile between agencies.

Figure 3: Ongoing staff in MAC agencies by age group, June 2002

Source: APSED

Factors contributing to the age profile of the APS include:

  • the relative stability of the workforce
  • the large number of 'baby-boomers' recruited in the late 1960s and the 1970s as reflected in the rising share and numbers of staff aged 45-54: with 30% of employees aged 45-54 at 30 June 2002 (33,225 employees) compared with 19% (27,347 employees) a decade ago
  • a general decline in the proportion of staff aged under 25 over much of the past decade, but for a slight turnaround in recent years: the level falling from 14,029 to 4,192 staff (a drop of 70%) between 30 June 1992 and 30 June 1999, before rising slightly between 1 July 1999 and 30 June 2001. There was a marginal decline in the proportion of ongoing staff aged under 25 in 2001-02, from 4.7% to 4.6%.

The median age has increased across all classifications over the last decade. However, the declining share of ongoing engagements amongst young people, who tend to come in at lower classifications, means that ageing of the APS is accentuated at the lower classifications.

There is a greater concentration of older employees at the higher classification levels. In June 2002, just over two-thirds (69%) of SES employees and 46% of Executive Level employees were aged 45 years or over-up from 59% and 35% respectively ten years ago.

Figure 4: Classification of ongoing staff by age group, June 2002

Source: APSED

Looking ahead, the APS will continue to age as shown in Figure 5. Projections for 2012 show an increase in the age profile. The highest proportion is expected to be in the 45-49 age group, however the peak will be less pronounced than it is now.

Figure 5: Projected age profile 2012, and actual age profile 2002

Source: APSED, Australian Government Actuary projections. These projections assume that the age distributions of new entrants and those leaving are based on engagement and separation age profiles for the past two years. The projections have been derived assuming that the exit rates observed over the two years to 30 June 2002 continue into the future.

As illustrated in Figure 6, the APS tends to be somewhat older than the wider workforce, but has a smaller share of both younger workers and employees over 55.

Figure 6: Age profile of APS and employed persons, 2002

Source: APSED, ABS Labour Force


Trends in engagements (that is, recruitment from outside the APS) contribute to the significant shifts occurring in the size and nature of the Service.

Engagements reached a low point of 5,688 in 1996-97, then increased each year until falling back slightly in 2001-02. Recent engagements have largely been influenced by recruitment in the Department of Defence, the Australian Taxation Office and Centrelink.

The median age of recruitment in the APS has risen slightly for men from 32 years to 34 years since a decade ago, while for women it remains at 30 years.

Behind the changes lie some important trends:

  • Engagements of employees under 20 years have dropped substantially, from 8% in 1991-92 to 2% in 2001-02.
  • This reduction reflects the substantial fall in engagements at APS 1/2 levels from 53% of all engagements in 1991-92 to 17% in 2001-02 (the figure was 85% in 1981-82).
  • Engagements of employees aged 20-24 years have increased in number since 1991-92 (2117 that year) but fell slightly in 2001-02 (from 2724 in 2000-01 to 2311).
  • There has been a significant increase in recruitment of people aged 25 years and more, from 6987 (71%) in 1991-92 to 9593 (79%) in 2001-02.
  • Lateral recruitment has increased, with engagements at APS 5 and above more than doubling over the last decade. Excluding the ATO and Centrelink, the proportion of engagements to middle management levels has increased from 18% in 1996-97 to 34% in 2001-02.
  • Base grade recruitment is most common these days at the APS 3/4 levels, with the latter now accounting for 28% of all engagements. While there was a decline in 2001-02, this primarily reflects continued increases in lateral recruitment at higher levels. Up until 2001-02, the share of engagements at APS 3/4 has increased each year since 1992-93.

A 'typical' new starter for 2001-02 would be aged 32 and engaged at the APS 4 level, and is more likely to be a woman than a man.

Reliable data on formal qualifications and previous employment is not available for the majority of new engagements. Of all employees engaged during 2001-02, including those engaged as part of a formal graduate program, and for whom data is available, 62% had tertiary qualifications and 41% were working in the private sector prior to entering the APS.

Although base grade recruitment continues to be most common at the APS 3/4 levels, there was a marked decline in the proportion of engagements at that level (from 51% in 2000-01 to 42% in 2001-02). This is the first time since 1992-93 that this group has not recorded an increase in engagements. The proportion of engagements at the APS 1/2 classification levels has dropped dramatically over the last 10 years, from 53% of 17 engagements in 1991-92, to only 17% in 2001-02, reflecting, among other things, the rise in skill requirements and qualifications of entrants in the Service. The number of engagements at this level, however, increased in absolute and proportional terms between 2000-01 and 2001-02 from 13% to 17% representing a modest reversal of the strong long-term decline.


As shown in Figure 7, after peaking in the post-1996 downsizing, the incidence of separations declined in 2000-01, before rising slightly during 2001-02, and has been overtaken by the incidence of engagements since 1999-00. However, the incidence of separations still remains higher than during the last period of APS growth in the early 1990s, largely due to increased incidence of resignations by ongoing staff aged 50-54.

Figure 7: Ongoing engagements and separations as a proportion of total ongoing staff, 1991-92 to 2001-02

Source: APSED

From the mid to late 1990s, retrenchments were the major form of separation, but this has been declining over the last three years (see Figure 8). In 2001-02, they accounted for just 25% of all separations and resignations have resumed their role in accounting for the majority of separations.

Figure 8: Selected types of separation as aproportion of total separations, 1991-92 to 2001-02

Source: APSED

It is interesting to note that there was a relative decline in the incidence of age retirements and terminations associated with the period of significant downsizing and higher use of retrenchments in the Service.

Age retirement has been increasing slowly in more recent years but not to levels that might have been expected, given the ageing profile of the workforce. A contributing factor is the increase in resignations of employees in the 50-54 age group. The number of terminations has picked up over recent years (from a low base) and, in 2001-02, accounted for 694 or 8% of all separations.

Rates and methods of separation vary quite markedly between older and younger employees, as illustrated in Figure 9. Separation rates are highest for employees in their early 30s and mid-50s.

Figure 9: Selected separations by age group, 2001-02

Source: APSED

Note: The drop in resignations is artificial as, from age 55, employees retire rather than resign.

Separation rates for employees under 25 have increased over the last decade, plateauing in recent years as shown in Figure 10. Separation rates for this age group are highest at the APS 1 level, but decrease markedly at the APS 4 level and above.

Figure 10: Separations as a proportion of total ongoing staff for selected age groups, 1991-92 to 2001-02

Source: APSED

The decline in separation rates for mature-aged workers (45 and over) slowed in 2001-02. The main separation types for this group in 2001-02 were retrenchment (35%), resignation (34%) and age retirement (21%).

Figure 11 illustrates the impact of retrenchments on all age groups over the past five years with the impact on those aged 55 and over particularly noticeable.

Figure 11: Retrenchments as a proportion of total ongoing staff by age group, 1991-92 to 2001-02

Source: APSED


The figure below shows the age profile of the 121,587 APS employees who were PSS and CSS members as at 30 June 2002.

Figure 12: Age profile-APS combined PSS and CSS membership, 30 June 2002

Source: CSS/PSS data

The PSS now dominates the numbers, with that scheme having 79% (96,530) of overall PSS/CSS members employed in the APS and the CSS having 21% (25,057) as at 30 June 2002. As may be expected, the age profile of the PSS is of a young scheme whereas the CSS is that of a scheme that has been closed for more than a decade and whose membership is ageing.

In the PSS, APS membership numbers peak at age 31 (3,500 members), with ages 28 to 43 each having more than 3,000 members. Members dwindle from this to about 1,000 members at age 56, with very few members being above age 60. In the CSS, the age distribution is indicative of a 'baby boomer' profile, exaggerated by the fact that the scheme has been closed for a considerable time. The age with the highest number of CSS members is 52 and about three-quarters of all current CSS members are aged 40 to 55. There is a sharp decline in members between age 54 and 55 (from 1,318 to 732) dropping to below 400 members at age 59 and with very few members working past age 60.

CSS data shows that, during 2001-02, of the 1,418 APS members of the scheme who were aged 54 at their last birthday, a total of 748 exited the scheme while 670 remained at the end of the year. The number of CSS members who resigned at age 54 during the year was 593, or 41.8% of all members.


Major shifts have occurred in classification arrangements and profiles.

Functional change, multi skilling and technological advances were drivers for significant streamlining and simplification of classification structures in the 1990s. There is now also greater flexibility and variation in how agencies gear their classification structures.

Figure 13 depicts the changing relative size of different classification groups over the last decade.7

Figure 13: Ongoing staff by classification group, 1992 and 2002

Source: APSED

Key changes have been the significant reduction in the use of APS 1/2 classifications, and substantial growth at the APS 4 level, which now accounts for 26% of employees.

The other main levels to show relative growth have been amongst the middle management grades, particularly the APS 6 and Executive Levels.

Senior Executive Service

The past decade involved periods of both growth and decline in the overall numbers of SES employees.8 Growth through the early 1990s saw the number of SES employees peak at 1,775 in June 1994. Consistent with wider reductions across the APS, the SES then reduced to a low of 1,541 in June 1998 after which it has gradually risen to 1,763 as at 30 June 2002. The SES is now about the same size as it was in June 1992, but now constitutes a higher share of all employees.

Band 1 executives make up the largest group within the SES (74%), with Band 2s comprising 20% and Band 3s about 6%.

The proportion of women in the SES has increased steadily over the last decade. In 1992 women represented 14% of the SES group, doubling to 28% at 30 June 2002.

The majority (76%) of senior executives are aged between 40 and 54, with 11% aged 39 years and under and 14% aged 55 or over.

Promotions from within the APS represent the most significant component of movement into the SES. The filling of vacancies by internal recruitment (promotion from within or between APS agencies) has fluctuated-from a low of 71% in 1992-93 to a high of 86% in 1997-98-and figures for the past two years run at around 83%. Meanwhile, external recruitment (the engagement of employees from outside the APS) has represented between 14% (2000-01) and 25% (1992-93) of all movements into the SES.

Recruitment from outside the agency (either from outside the APS or from other agencies) has ranged from a low of 25% (1997-98 and 2000-01) to a high of 44% (1992-93). Over the past year, this figure has been at the middle of that range, although it can vary considerably between agencies.

Yearly departures from the SES range from a low of 121 in 1992-93 to a high of 244 in 1999-2000. The reasons for departures vary and can include: Machinery of Government changes moving employees out of Public Service Act employment; resignation (for a wide range of reasons); retirement or retirement with an incentive. The most significant single category of departures is where SES employees accept an incentive to retire.

Under the provisions of s.37 of the Public Service Act 1999, an agency head may give notice to an employee that they will become entitled to a payment of a specified amount if they retire from the APS (this is the analogue of retrenchment for non-SES employment categories). In 2001-02, 30% of SES separations were in this category.

Women in the APS

The proportion of female employees in the APS has increased since the abolition of the marriage bar in 1966 until, in 2001, women made up just over half of ongoing staff for the first time. Over this period women have predominated in lower level classifications. At June 2002, women comprised 56% of all ongoing staff in the APS classification levels.

Over time, and as numbers of women in the service have increased, women have increased their representation in more senior classifications. As Figure 14 makes clear, there has been strong growth at these levels. Women now make up 37% of Executive level officers and 28% of those in the SES. Over the past decade, women have had a consistently lower length of service before reaching Executive level and the SES. In 2001-02 the median length of service for employees reaching SES Band 1 was 13 years for women and 15 years for men.

Figure 14: Change in number of women at selected levels, weighted and indexed, 1993 to 2002

Source: APSED

On current trends, women's increasing presence in the APS, and particularly at more senior levels, is likely to be sustained. As shown in Figure 15, women's rate of lateral recruitment (engagements) and rate of promotion to APS 5-6 levels and above is higher than their representation at those levels. Moreover, as the number of female graduates has increased over time in Australia overall, the representation of women in APS graduate programs has increased correspondingly, with the result that women have outnumbered men in graduate programs since the early 1990s.

Figure 15: Ongoing staff: engagement, promotion and representation rates for women, 2001-02

Source: APSED

Retention rates of male and female APS employees, including graduates, are much the same up until fifteen years of service, but start to diverge after longer periods of service-with men having higher rates of retention-so that for example, of those with 30 or more years of service, some 83% are males. This pattern probably reflects adjustments in employment patterns following the removal of the marriage bar in 1966. Though women's representation in the APS began to increase immediately following the bar's removal, that representation rose more sharply during the 1970s, as Figure 16 indicates. As a result of these historical factors, men have had substantially longer median length of service before separating than women over the past decade (for example, eleven years for men and six years for women in 2001-02).

Following the removal of the marriage bar, the APS introduced over time a number of flexibilities and leading edge practices.

  • Paid maternity leave was introduced on a legislative basis in 1973.
  • Flexible working hours have been offered since the mid 1970s.
  • A variety of flexible leave arrangements has been put in place through agency agreements (e.g. more generous carers' leave and purchased leave).
  • Access to permanent part-time work has increased (particularly since the Workplace Relations Act removed the capacity for restrictive award provisions).
  • A range of dependent care arrangements has been established through agency agreements (e.g. school holiday programs and elderly care referral assistance).

These practices have contributed to the strong performance of the APS on women's employment shown in Figure 16, and are prominent in employee responses to surveys undertaken for the project.

Figure 16: Women as a proportion of ongoing and non-ongoing staff, 1906 to 2002

Source: APSED

Note: Breaks in the graph are due to the unavailablity of data for some years

Separation rates from the APS have been much the same for male and female employees over the past four years. However, female APS employees are more likely than males to resign from APS employment (60% of total separations for women in 2001-02 compared to 49% for men) and to resign at a younger age (the median age for women who resigned in 2001-02 was 33 years compared to 39 years for men). The younger resignation age for women is likely, in part, to be related to child bearing, as the resignation rate tapers off after the 35-39 age group. However, the value placed by women graduates on the flexible employment conditions in the APS suggests that there is a strong inclination to combine work and family among this group. There is also a very high rate of return from paid maternity leave in the Service. Of the 1594 women who took maternity leave in 2000-01, only 206 (12.9%) had separated by 30 June 2002.

The next most common reason for separating from the APS, retrenchment, was more common for men (28% of all separations for men) than for women (22%). Again, the median age of retrenched men (52 years) was significantly older than for retrenched women (43 years).

Persistent differences between workforce characteristics of male and female APS employees could indicate different priorities affecting the choices of male and female employees. In chapter 5, survey data gathered for the project is used to examine the nature and extent of any attitudinal differences between male and female employees for both the graduate and the mature-aged groups.

Employee characteristics

Changes in the size and nature of the APS have affected the representation rates of other employee groups.

Indigenous employment is significantly higher in the APS (2.4%) than for employed persons overall (1.2%) and roughly equates to the share of Indigenous people in the population as a whole. While Indigenous employment is higher than a decade ago (1.9% at June 1992), it has decreased slightly in recent years due to relatively high rates of separation amongst Indigenous staff (3.1% of separations during 2001-02) and reduced numbers of employees at the APS 1/2 levels. There is some growth in representation at more senior levels and a relatively high representation amongst graduate recruits compared to Indigenous representation overall.

Over the last decade, there has been a consistent decline in the employment of people with a disability, both in absolute numbers and as a share of all staff. The APS now employs a lower proportion of people with a disability than equivalent occupations outside the APS (3.6% compared to 9.7%). People with a disability are over-represented in separations, in particular retrenchments and invalidity retirements.

The proportion of employees identifying as being from a non-English speaking background was 9.6% at June 2002, which is slightly lower than in APS equivalent occupations in industry (12.8%).

Employment arrangements

Recent changes in the APS employment framework (in particular the Public Service Act 1999) have provided agencies with greater flexibility to meet labour and skill requirements in what is now a rapidly changing operating environment (e.g. greater contestability and the need for greater responsiveness, flexibility and efficiency in resource management). In this context, changes in the way labour is employed could be expected.

Non-ongoing employment

There has been a general decline in the use of non-ongoing employees over the last ten years, both in actual numbers and as a proportion of total staff, before an increase in 2001-02 (at June 2002, non-ongoing staff represented 9.2% of all staff, up from 8.9% at June 2001, but lower than the 9.4% at June 2000).

Key features of the use of non-ongoing employment as at June 2002 include:

  • The proportion of those employed on a non-ongoing basis generally declines with age-from 29% of those aged under 25 to around 6% at the 50-54 age category- but then increases to around 11% for employees aged 55 and over.
  • Non-ongoing employment is concentrated in the lower classification levels: making up 57% of all staff at the APS 1 level, 31% of staff at APS 2 level and 12% at APS 3 level but less than 10% for all other classifications. Interestingly, a relatively high proportion of SES Band 3 employees (12%) were non-ongoing in 2002.
  • Women have consistently comprised the majority of non-ongoing APS employees and, in June 2002, accounted for almost two-thirds of these employees.
  • Fixed-term arrangements are much more common amongst non-ongoing employees than employment for a specified task or on an irregular and intermittent basis.
  • There is a wide variation between agencies in the employment of non-ongoing staff, reflecting different functional requirements. For example, at 30 June 2002, the non-ongoing proportion varied from 42% in AIATSIS to 1% in ANAO.

Significant use is also made by many agencies of contract labour and agency hirees not engaged under the Public Service Act.

Part-time employment

Part-time work has increased significantly in the APS over the last decade, reflecting the removal of artificial restrictions and preferences amongst agencies and their staff.

At 30 June 2002, 2.8% of men were employed part-time, a marked increase from 0.6% at 30 June 1992. At 30 June 2002, 15.6% of women were employed part-time, compared to 5.8% at 30 June 1992. Use of part-time employment is concentrated in the category of women aged 30-44, with much lower incidence for men and for both men and women aged 45 and over. Part-time work is mostly concentrated at the APS 1-4 level for both men and women.

There is a greater incidence of part-time work among non-ongoing employees (21%-which gives a rate of 11% for total staff).

While part-time work has been growing in the APS, its incidence is well below that for equivalent occupations in the broader public sector (31%) and private sector (33%).9

Career patterns

Notions of career and working patterns in the public service have also been undergoing changes.

Length of service

The median length of service of ongoing employees has increased slightly, from seven years in 1992 to nine at June 2002, partly reflecting the gradually increasing proportion of staff aged 55 and over and shorter service amongst younger employees. The median length of service for employees aged 50-54 years has increased over the last 10 years, from 12 years to 15 years. As would be expected, increases in length of service correlate with the increases in age and classification.

Although there has been some variability, men have tended to have a longer median length of service than women over the past decade.


The median length of service of APS employees before separating has been increasing (to be 11 years for men and six years for women in 2001-02), as reflected in increased retention rates. For example, Figure 17 below shows a trend decline in the proportion of people separating within 5 years of joining the APS (allowing for the cyclical movements during the downsizing that occurred in the mid-1990s).

Figure 17: Ongoing Staff: Retention by year of engagement 1987 to 2001

Source: APSED

Retention rates for graduates show somewhat different patterns and are considered later in this section.


The rate of progression through the classification structure has not changed markedly over the last decade. For example, the median length of service before reaching Executive Level and the SES has remained relatively consistent over the last decade. In 2001-02 the median length of service for employees reaching Executive Level 1 was 6 years while the median length of service for employees reaching SES Band 1 was 14 years.

However, two qualifications to this observation need to be noted:

  • Over recent years, employees entering the Service as part of a graduate recruitment program have been advancing more quickly through the classification levels than their counterparts of previous years. As a consequence, the median length of service for graduate employees reaching Executive Levels has decreased.
  • Over the past decade, women have had a consistently lower length of service before reaching Executive Level and the SES. In 2001-02 the median length of service for employees reaching SES Band 1 was 13 years for women and 15 years for men.

Mobility rates10

As illustrated in Figure 18, overall rates of mobility (including promotions and transfers between agencies) remained fairly constant through the past decade, declining sharply in the late 1990s, recovering slightly and then falling again. Similar trends were evident for both males and females, although women have had a consistently higher mobility rate than men.

The declining mobility reflects in part a greater attention by agencies to internal career and skills development. The increased use of broadbanding could also be a factor.

The mobility rate for all movements between agencies was 1.5% in 2001-02. Rates of mobility were higher for younger employees, with those aged 20-24 years at 1.7%, 25-29 years at 2.5% and 30-34 years at 2.0%. The mobility rate declined for employees aged 35 years and above.

Figure 18: Mobility rates between agencies, 1992-93 to 2001-02

Source: APSED

Patterns of mobility show some difference between promotions and transfers, as shown in Figure 19.

Figure 19: Promotion and transfer rates between agencies, 1992-93 to 2001-02

Source: APSED

The promotion rate between agencies has reduced slightly over the last couple of years.

Women have a higher rate of promotion between agencies, although this difference between men and women has declined over the last decade.

In 2001-02, some 92% of all promotions in the APS were within agencies, a slight increase compared with recent years. However, the proportion of such internal promotions varies substantially between agencies. Those agencies with high internal promotion rates include the Department of Foreign Affairs and Trade, the Department of Defence and the Australian Bureau of Statistics.

The transfer rate between agencies was fairly steady through most of the 1990s. After a sharp decline in the late 1990s, the rate showed a small rise, then returned to its low point in 2001-02.

There are differences in the transfer rate between agencies for employees at different classifications. The transfer rate is low for employees at lower classification levels but increases for employees at higher classification levels. For example, SES has the highest rate of lateral mobility, whereas APS 1-2 have the lowest. Overall, the transfer rates for each classification have declined for the last decade.

Figure 20: Transfer rates between agencies by classification group, 1992-93 to 2001-02

Source: APSED

Women have a consistently higher transfer rate than men although, as with promotions, the difference between males and females has declined over recent years. Women at each classification level have a slightly higher transfer rate than men, except at the SES level where women's transfer rate is much higher.

Graduate recruitment and retention

Trainee recruitment (which includes graduates and those in traineeships and cadetships) increased significantly over the decade from 539 in 1991-92 to 1573 in 2000-01 and then declined again during 2001-02. As shown in Figure 21, there has been considerable fluctuation in the number of trainees engaged in the APS over the decade.

Figure 21: Engagements of ongoing trainees, 1986-87 to 2001-02

Source: APSED

During 2001-02, 372 people were reported by agencies as being engaged as part of graduate programs, representing a marked decline from 936 graduate engagements in 2000-01.11 The largest recruitment of graduates occurred during 1999 (1,214). This was due mainly to an unusually large intake by the ATO (47% of total graduates for the year). The smallest recruitment of graduates over this period was in 1988-89 (367).

Most agencies displayed variation in the number of graduates recruited each year. Much of the overall variation in graduate numbers is due to the recruitment policies of the Australian Taxation Office (ATO). The ATO recruited 396 graduates in 1985-86 and 578 graduates in 1999-00, but between 1992 and 1994 only recruited 19 graduates. As ATO recruits such a high proportion of graduates overall, any change in its recruitment levels greatly affects the total number of graduates.

Agencies recruiting a large proportion of the total graduates in 2001-02 include Defence (36%), Australian Bureau of Statistics (19%), Health and Ageing (9%), and Agriculture, Fisheries and Forestry-Australia (9%).

The number of female graduates has increased over time and they have consistently outnumbered males since the early 1990s.

The APS Commission has recently undertaken an analysis of graduate recruitment from 1986 to 2001.12 The following material is drawn from that analysis.

Between 1986 and 2001, people recruited as part of a graduate program made up 6% of total ongoing engagements. Overall, the proportion of graduate engagements to that of total ongoing engagements increased between 1989 and 1998 (Figure 22). However, in the last few years this proportion has decreased.

Figure 22: Engagements as part of a graduate program as a proportion of total ongoing engagements

Source: APSED

Since 1994, individuals recruited as part of a graduate program have made up an increasing proportion of total ongoing engagements for employees under the age of 25 (Figure 23), although this has decreased in the last couple of years.

Figure 23: Engagements as a part of a graduate program as a proportion of total ongoing engagements in the APS-for under 25 year olds

Source: APSED

The median age of these graduates13 was 23 years for women and 24 for men and has not changed much across the years.

The majority of graduates (88%) had a highest qualification level of a bachelor's degree-with 12% having post-graduate qualifications.

Graduates with qualifications in the business/administration/economic fields have generally been declining since 1986. This is likely to be related to the increasing number of female graduates since this time. Between 1994 and 1999 there was a large increase in graduates from a legal background, however this has tapered off in recent years. Graduates with an arts and humanities background have fluctuated over time. There has been little variation in the proportion of graduates with a science background.

Most graduate recruits (55%) advanced to an APS 4 level at the end of their graduate program year. They were likely to achieve their first promotion faster than ongoing APS employees engaged at the APS 3 or 4 levels. Comparison of the career progression of graduates from the 1986, 1991 and 1996 cohorts shows a much higher progression after five years of the 1996 graduates to Executive Levels than the earlier groups. In addition, after ten years a much higher proportion of graduates from the 1991 cohort had reached an Executive Level than the 1986 cohort.

Most graduate recruits (65%) do not move from their original agency before separating. Of those who did move, 15% did so in their first year, and a further 26% in their second year. In 2001 the median length of service at the original agency before moving to another agency was 2.3 years.

Of those graduate recruits in the cohorts studied who were still in the APS at June 2001, 72% had worked in one agency, 13% had worked in two agencies, 8% had worked in three agencies and 7% had moved across four or more agencies.

Of graduates recruited since 1986, some 57% had separated by June 2001. For those who had separated, the median length of service was 5.2 years. Almost 10% of graduates separated before they had completed their graduate year.

The highest proportion of graduates (22%) was at an APS 6 level when they separated.

As illustrated in Figure 24, retention rates for the 1986 and 1987 cohorts were relatively low. Retention rates then increased steadily reaching a peak in the 1989 cohort. Since the 1990 cohort, the retention rate of graduates has been declining. For example, 69% of the 1989 cohort was still in the APS seven years later; in contrast, only 50% of the 1994 cohort was still in the APS after seven years. However, this decline is from a cohort with relatively high retention rates, and retention rates for those who have remained for more than five years are similar to those of the mid 1980s.

Figure 24: Graduates: Retention by year of engagement, 1986 to 2000

Source: APSED

Graduate recruits in the study remained in their original agency for a median length of four years.

There is considerable variation in retention by agency. In 1996, the eight largest graduate employers in the APS recruited 319 graduates. Of this group, 196 (61%) were still in the APS at June 2001, and a subgroup of 147 (46%) were employed in their original agency. Retention rates varied from almost 100% in the Department of Foreign Affairs and Trade (29 of 30 in the 1996 cohort still employed in 2001) to 25% in the Australian National Audit Office (5 of 20 in the 1996 cohort still employed in 2001). However, in relation to the latter, 55% of the 1996 cohort remained in the APS.

Comparison with anecdotal evidence from the private sector suggests that, in the short term (up to two years), the APS and private sectors have a similar retention rate. However, in the medium to long term it appears that the APS has a higher retention rate overall. This may be a result of APS graduates moving to another agency, while still staying in the APS, while private sector graduates who move do so to a completely different employer.

Workers' compensation

Ageing represents an issue for workplace health and safety and consequently workers' compensation. The issues related to ageing are changes in:

  • employment age profile, with an ageing workforce
  • claims age profile, as there is a higher risk of injury and illness associated with age
  • claims cost, as there are higher annual claims costs associated with claims for mature-aged employees.

Comcare claims data suggests that for the Commonwealth jurisdiction there is an increased rate of injury after age 49. Injury profiles also change: the incidence of claims for 'other diseases' such as hearing loss, heart attack and respiratory diseases for workers 50 years or older is more than twice the rate of younger workers. There is also an increase in average total claim cost until the age of 50, (and a decrease thereafter, probably as a result of age 65 incapacity cessation provisions).

On the positive side, in recent times absolute numbers of claims have fallen for all age groups, but especially for older workers, suggesting that there is scope for further improvement. Comcare is preparing a campaign on injury prevention and management for older workers as part of its broader Leadership and Accountability Strategy for the APS.

Implications for agencies

Building ongoing capability needs to take account of these changing workforce dynamics.

These trends present a Service-wide picture. There are significant variations between agencies, particularly in relation to age profile. Agencies also vary in the capabilities required, and the degree to which these need to be developed in-house or can be bought in. Agencies need to understand their own demographics, business requirements and skill needs and to structure their workforce planning strategies accordingly.

Wider demographic trends

Wider demographic trends will also have significant implications for the APS as an employer.

As outlined in the Intergenerational Report 2002-03, Australia, like other OECD countries, is experiencing an ageing of its population and labour force, driven by declining fertility and mortality rates.14 Looking ahead, employment growth is expected to slow, reflecting lower labour force growth from lower population growth and declines in overall labour force participation.

While the population of those of labour force age is projected to increase by just 14%, the number of people aged 55 to 64 is projected to increase by more than 50% over the next two decades. This is expected to be the fastest growing group of labour force age.

In addition, there will be a significant reduction in the number of new entrants coming into the labour market. Access Economics has estimated, for example, that whereas the working age population has recently been growing annually by 170,000, for the entire decade of 2020s it will grow by just 125,000.15

Agencies will face increased competition for new entrants to the labour market from the private sector and internationally, an increasing challenge to retain skilled people in a tightening labour market and the need for further expansion in lateral recruitment. The participation of mature-aged workers will need to be maximised, by encouraging older workers to remain in the workforce longer, even if on a part-time or part-year basis, including scope for more phased retirement.

Agencies providing services to the public will also have a clientele with an increasingly ageing profile. This will have implications for the services provided, as well as how they are best delivered. Agencies will need to ensure that their staff training, service standards and systems are attuned to their changing client base. Attention to the style of servicing desired by clients (e.g. preferences for dealing with staff from similar age groups) will need to be considered as part of wider improvements in client service strategies.

As well as needing to respond to these challenges as an employer and service provider, the APS also has a key ongoing responsibility to contribute to the broader economic and social policy objectives of the government of the day in response to the ageing of the population. Optimising the participation of mature-aged workers will be crucial in this regard.

4 The analysis is of ongoing staff unless otherwise specified. Decimal point figures have been rounded to whole numbers except where their order is such that they require presentation to the first decimal point.

5 Centrelink; Australian Taxation Office; Defence; Family and Community Services; Australian Customs Service; Immigration and Multicultural and Indigenous Affairs; Health and Ageing, Australian Bureau of Statistics; Agriculture, Fisheries and Forestry-Australia; and Environment and Heritage.

6 The upper line represents unadjusted figures (the total number of APS staff in the particular year); the lower line abstracts from changes that occurred in the functions performed in the APS over the period by showing the level of staff employed in those functions that remained in the APS at 30 June 2002.

7 This Figure uses the eight level classification structure in place since 1998 as a basis for comparison. Note that, within this structure, flexibility currently exists for agencies to set their own rates of pay and to broadband classifications or create an agency-specific structure.

8 SES data includes employees who are not in SES positions, but who receive a similar level of remuneration.

9 The composition of APS equivalent occupations reflects as much as possible the current structure of the APS. It is based on a sub-set of occupations from the ABS's Australian Standard Classification of Occupations that is similar to the type of work done in the APS.

10 Mobility rates have been calculated excluding movements from the then Department of Employment, Education, Training and Youth Affairs (DEETYA) to the Department of Social Security (DSS) prior to the establishment of Centrelink. This movement was not an Administrative Arrangement Order (AAO), but the later move from DSS to Centrelink was. Mobility rate is calculated as the number of ongoing promotions and transfers between agencies as a proportion of total ongoing staff. Movements due to AAOs are excluded.

11 This decline is mainly due to the fact that the ATO did not recruit graduates in 2001-02.

12 APS Commission, Graduate Trends (forthcoming).

13 For the purposes of the research paper referred to, a graduate is defined as someone who is engaged at the Graduate APS classification, then advances to APS 3 at the end of their training year (many agencies promote to higher levels immediately after such advancement). The definition excludes a small number of persons engaged following an agency advertisement for graduates but at a classification other than a Graduate APS

14 Intergenerational Report 2002-03. Budget Paper No. 5 May 2002.

15 Access Economics, Population Ageing and the Economy, Commonwealth Department of Health and Aged Care, January 2001, page 3.