Trends in indigenous educational participation and attainment, 1986-96.

This paper provides a summary and overview of indigenous people within the Australian education system. A cohort analysis of changes in educational participation and the level and type of educational qualification over the last three censuses for the indigenous and non-indigenous populations is presented. The main finding is that although there have been some absolute improvements in indigenous educational outcomes over the period 1986 to 1996, relative to the nonindigenous there have been little if any real gains. The lack of improvement relative to the non-indigenous population occurs not only in the proportion of the population with post-secondary qualifications, but also in the proportion of indigenous teenagers staying at school. By any measure, the indigenous population remains severely disadvantaged. Another finding is that, for younger age groups, the nonindigenous population has a higher participation rate in post-secondary education than the indigenous population. This situation is reversed for older age groups.

Introduction

Indigenous education policy in Australia today has evolved alongside an awareness of the need to improve indigenous educational outcomes in order to secure the future prospects of the indigenous population.(1) Over the course of nearly 25 years, a variety of policy reviews has been conducted and countless reports written dealing with indigenous education (Aboriginal Consultative Group, 1975; Department of Employment, Education and Training, 1988; National Review, 1995) and training and employment (Aboriginal and Torres Strait Islander Commission, 1994; Committee of Review, 1985; Royal Commission, 1991). In each review, report or policy recommendation over this period, access, participation and equity have remained primary themes.

Whereas it is relatively easy to develop policy, it is often difficult to measure the impact of specific policies. The Australian Bureau of Statistics (ABS) censuses provide a means to chart some important social indicators of education outcomes. The general level of educational attainment in Australia has increased significantly over the period 1986 to 1996. Although this is certainly good news, the aggregate numbers do not reveal important variations within the overall population. If minority groups fail to keep up with the rate of increase of other Australians, it is likely that they will suffer increasing disadvantage and marginalisation in labour markets. In this context, it is important to examine how indigenous Australians fare as the general level of educational attainment continues to rise. If they do not keep up with the increase in qualifications in the rest of the population, then it is likely that they will remain uncompetitive in the labour market and high rates of indigenous poverty will be perpetuated indefinitely.

This paper focuses on recent trends in indigenous and non-indigenous education using the 1986, 1991 and 1996 Censuses.(2) The statistics presented summarise recent patterns and provide a basis for considering ways to improve educational outcomes and related indigenous economic status. The paper is intended to provide a summary and overview targeted to policy makers.

Inter-censal changes in Australian education, 1986-96 Attendance at educational institutions

The relatively poor educational attainment among indigenous people is partly a function of leaving school at a younger age, on average, than the non-indigenous population (Hunter & Schwab, 1998). Table 1 depicts the relative proportions of the working-age indigenous and non-indigenous populations who left school at different ages. At one extreme, in 1996, 3.4 per cent of indigenous males and 3.1 per cent of indigenous females had never attended school as compared with only 0.6 per cent of non-indigenous males and 0.9 of non-indigenous females. At the other extreme, only 22.0 per cent of indigenous males and 23.4 per cent of indigenous females left school when aged 17 or more years as compared with 38.7 per cent of non-indigenous males and 36.8 per cent of non-indigenous females.

Table 1 Age left school for the working-age population, 1986-96

                                            Indigenous

                                      1986      1991       1996

Male
Still at school                        6.4       6.2        6.9
No schooling                           7.8       5.5        3.4
Left school at or before              23.0      20.6       19.2
  14 years old
Left school at 15 years old           29.8      27.9       26.6
Left school at 16 years old           20.6      21.7       21.9
Left school at 17 years old            8.0      10.8       13.0
Left school at 18 years old            3.2       4.7        6.2
Left school at 19 years or older       1.2       2.7        2.8
Total 15 plus                        100.0     100.0     1000.0
Number (000s)                           66        77        102

Female
Still at school                        6.5       5.8        6.7
No schooling                           7.6       5.3        3.1
Left school at or before              20.0      17.4       16.2
  14 years old
Left school at 15 years old           30.1      27.8       27.0
Left school at 16 years old           22.6      23.9       23.5
Left school at 17 years old            8.8      12.4       14.4
Left school at 18 years old            3.1       4.7        6.2
Left school at 19 years or older       1.1       2.7        2.8
Total 15 plus                        100.0     100.0      100.0
Number (000s)                        71.00     82.00     110.00

                                           Non-indigenous

                                      1986      1991       1996

Male
Still at school                        5.0       5.1        5.2
No schooling                           0.7       0.9        0.6
Left school at or before              21.3      16.5       14.7
  14 years old
Left school at 15 years old           23.2      21.8       20.6
Left school at 16 years old           20.9      20.4       20.2
Left school at 17 years old           15.4      17.2       20.4
Left school at 18 years old            9.6      10.9       13.8
Left school at 19 years or older       3.9       7.2        4.5
Total 15 plus                        100.0     100.0      100.0
Number (000s)                         5727      6158       6507

Female
Still at school                        4.8       4.9        5.0
No schooling                           0.9       1.1        0.9
Left school at or before              22.2      17.0       15.2
  14 years old
Left school at 15 years old           25.4      23.5       21.8
Left school at 16 years old           21.3      20.9       20.3
Left school at 17 years old           15.4      17.7       21.0
Left school at 18 years old            7.7       9.4       12.2
Left school at 19 years or older       2.3       5.5        3.6
Total 15 plus                        100.0     100.0      100.0
Number (000s)                         5880      6355       6767

Note: The working-age population is defined as those aged 15 and over.
Source: Unpublished data from the 1986, 1991 and 1996 Censuses.

One factor underpinning poor secondary school completion rates for indigenous students is the relatively high rates of suspension and expulsion from schools. Although indigenous children and young people comprise 3 per cent of the student population, they make up 12 per cent of school suspensions in New South Wales. Children as young as five years of age are being suspended, excluded, and expelled from schools (National Inquiry, 1997). In addition, Groome and Hamilton (1995, p. 3) find that `Aboriginal students are likely to lose between two and four years of schooling through absenteeism; rates for the total population are less than half these'.

Hunter and Schwab (1998) have shown that there has been some improvement in school retention of indigenous teenagers, with an increasing proportion staying at school longer between 1986 and 1996. Although this has led to marginal relative improvement in the position of indigenous to non-indigenous Australians for all ages (except 18-year-old females), it is disconcerting that the absolute difference between indigenous and non-indigenous teenagers in the percentages at school increased for all age groups over age 15. Long and Frigo (1998) reach similar conclusions.

An implication of these statistics is that indigenous people will continue to be under-represented in post-secondary education. Table 2 shows the proportion of working-age population attending different categories of educational institutions. The proportion of indigenous males attending an educational institution increased from 17.6 to 19.2 per cent between 1986 and 1996. This is against a decrease of 1.1 percentage points to 15.5 per cent for non-indigenous males. Therefore, over the period 1986 to 1996, a higher proportion of indigenous males were attending an educational institution than non-indigenous males and the gap increased between 1986 to 1996. Similarly a higher and increasing proportion of the indigenous female population were attending an educational institution than nonindigenous females over this period. Hunter and Schwab (1998) have shown that the apparent over-representation of the indigenous population attending educational institutions is merely an artifact of their relatively high fertility rates and the substantially higher mortality rates among older indigenous people. Notwithstanding this, it is useful to examine the compositional changes in post-secondary education.

Table 2 Proportion of the working-age population attending
educational institutions, 1986-96

                                          Indigenous

                                     1986    1991    1996

Male
Government secondary school           6.4     6.0     5.6
Non-government secondary              0.9     1.1     1.2
  school
TAFE                                  5.6     6.3     7.3
University or CAE                     3.4     3.2     4.2
Other educational institution         1.4     1.5     1.0
Total                                17.6   18. I    19.2

Female
Government secondary school           6.3     5.6     5.2
Non-government secondary              0.9     0.9     1.2
  school
TAFE                                  7.1     6.3     7.8
University or CAE                     3.5     4.3     5.1
Other educational institution         1.4     1.3     1.0
Total                                19.2    18.5    20.3

                                         Non-indigenous

                                     1986    1991    1996

Male
Government secondary school           3.6     3.5     3.3
Non-government secondary              1.6     1.6     1.8
  school
TAFE                                  5.5     4.6     4.2
University or CAE                     4.7     4.8     5.3
Other educational institution         1.3     1.2     0.9
Total                                16.6    15.7    15.5

Female
Government secondary school           3.4     3.3     3.2
Non-government secondary              1.6     1.6     1.7
  school
TAFE                                  4.9     3.8     4.0
University or CAE                     4.7     5.3     5.9
Other educational institution         1.5     1.5     1.4
Total                                16.1    15.5    16.2

Note: Working-age population is defined as those aged 15 and over.
Source: As for Table 1.

Within the tertiary education sector, there are several significant differences between the indigenous and non-indigenous populations. First, indigenous males and females in 1986, 1991, and 1996 were more likely to be attending a technical and further education (TAFE) institution than their non-indigenous counterparts. For indigenous males, the participation rate at TAFE increased in each intercensal period with an overall increase of 1.7 percentage points to 7.3 per cent between 1986 and 1996. Over the same period, the participation rate of nonindigenous males fell from 5.5 per cent in 1986 to 4.2 per cent in 1996.

Second, for indigenous females, the participation rate at TAFE fell from 7.1 per cent in 1986 to 6.3 per cent in 1991, but then increased to 7.8 per cent in 1996. For non-indigenous females, the participation rate fell from 4.9 per cent in 1986 to 3.8 per cent in 1991 and then showed a small increase to 4.0 per cent in 1996.

Third, participation rates at universities and colleges of advanced education (CAE) increased over the period 1986 to 1996 for both indigenous and nonindigenous males and females. For indigenous males, there was a fall from 3.4 per cent in 1986 to 3.2 per cent in 1991, and then an increase to 4.2 per cent in 1996. For non-indigenous males, although participation increased in both inter-censal periods, the slightly smaller overall increase of 0.6 percentage points meant that there was a small decrease in the difference between indigenous and nonindigenous participation rates.

Fourth, for indigenous females, there was an increase in attendance rates at universities and CAEs of 1.6 percentage points, from 3.5 per cent in 1986 to 5.1 per cent in 1996. For non-indigenous females, there was an increase from 4.7 per cent in 1986 to 5.9 per cent in 1996.

In summary, although the participation rate of indigenous people remains below that of their non-indigenous counterparts, there has been some catch-up in the participation rates at tertiary institutions. For males, the ratio of the indigenous to the non-indigenous participation rate increased from approximately 72 per cent to 79 per cent, and for females the ratio increased from 74 per cent to 86 per cent. However, the different demographic structures of the indigenous and nonindigenous populations mean that such conclusions are potentially misleading. The cohort analysis in the following section examines educational attendance for particular age groups so that demographic distortions of the interpretation are minimised.

Cohort analysis of educational attendance

The 1986, 1991, and 1996 censuses are cross-sectional data sets which, in principle, include the entire Australian population. However, they can be treated as panel data by grouping individuals into cohorts, and treating the averages within these cohorts as individual observations that vary over time. These cohorts are defined such that each individual is a member of only one cohort, which is the same for all time periods. In this paper, cohorts are defined on the basis of year of birth and sex. Cohorts are defined by five-year age groups starting with those 5 to 9 years old in 1986.

Table 3 presents the results of disaggregating the proportion of the population attending university or CAE into age cohorts. Table 4 does the same for the proportion attending TAFE or other non-university or CAE educational institution.

Table 3 Cohort analysis of proportion of population attending
university or CAE, 1986-96

                                 Indigenous

Age at 1986 Census           1986   1991   1996

Male

Aged 5 to 9 years             --     --     2.3
Aged I 0 to 14 years          --     2.1    6.0
Aged 15 to 19 years           1.3    4.8    5.4
Aged 20 to 24 years           3.8    3.5    4.9
Aged 25 to 29 years           4.4    4.2    5.0
Aged 30 to 34 years           4.0    3.4    4.1
Aged 35 to 39 years           4.4    3.4    3.2
Aged 40 to 44 years           3.7    2.7    3.5
Aged 45 to 49 years           3.6    2.0    3.2
Aged 50 to 54 years           6.0    3.3    1.2
Aged 55 to 59 years           2.1    0.0    0.0
Aged 60 to 64 years           0.0    0.0    2.0
Aged 65 years and over        3.9    2.6    4.0
Total 15 plus                 3.4    3.2    4.2

Female

Aged 5 to 9 years             --     --     3.5
Aged I 0 to 14 years          --     3.4    6.8
Aged 15 to 19 years           2.5    5.9    5.7
Aged 20 to 24 years           3.7    4.5    5.5
Aged 25 to 29 years           3.7    4.8    6.0
Aged 30 to 34 years           5.0    5.6    5.9
Aged 35 to 39 years           4.8    4.7    4.9
Aged 40 to 44 years           5.0    4.2    5.3
Aged 45 to 49 years           3.9    3.8    3.3
Aged 50 to 54 years           4.3    2.3    2.0
Aged 55 to 59 years           1.2    0.9    0.0
Aged 60 to 64 years           0.0    0.0    1.5
Aged 65 years and over        1.9    2.2    2.4
Total 15 plus                 3.5    4.3    5.1

                               Non-indigenous

Age at 1986 Census           1986   1991   1996

Male

Aged 5 to 9 years             --     --     9.2
Aged 10 to 14 years           --     9.0   17.3
Aged 15 to 19 years           6.4   15.0    7.7
Aged 20 to 24 years          10.8    6.1    5.5
Aged 25 to 29 years           5.5    4.8    4.5
Aged 30 to 34 years           4.9    4.0    3.5
Aged 35 to 39 years           4.1    2.8    2.5
Aged 40 to 44 years           3.2    1.9    1.7
Aged 45 to 49 years           2.5    1.2    1.2
Aged 50 to 54 years           2.0    0.9    0.9
Aged 55 to 59 years           1.9    0.8    1.1
Aged 60 to 64 years           2.0    1.2    1.6
Aged 65 years and over        3.3    1.6    2.0
Total 15 plus                 4.7    4.8    5.3

Female

Aged 5 to 9 years             --     --    13.0
Aged 10 to 14 years           --    12.1   19.4
Aged 15 to 19 years           7.8   15.7    7.6
Aged 20 to 24 years           9.7    5.7    5.3
Aged 25 to 29 years           4.6    4.6    4.7
Aged 30 to 34 years           4.4    4.4    4.2
Aged 35 to 39 years           4.4    3.8    3.3
Aged 40 to 44 years           3.9    2.8    2.2
Aged 45 to 49 years           3.1    1.8    1.4
Aged 50 to 54 years           2.5    1.2    1.1
Aged 55 to 59 years           2.3    1.0    1.1
Aged 60 to 64 years           2.5    1.3    1.7
Aged 65 years and over        3.5    1.9    2.3
Total 15 plus                 4.7    5.3    5.9

Source: As for Table 1.

As an example of the interpretation of the cohort analysis in Table 3, consider indigenous males aged 25 to 29 in 1986. The proportion of this group attending a university or CAE was 4.4 per cent in 1986. By 1996, this group was aged 35 to 39 years and the proportion attending a university or CAE had increased to 5.0 per cent. This cohort can also be compared with people who were 25 to 29 years old in 1996. The relevant proportion of the group attending a university or CAE for such people, aged 15 to 19 years in 1986, was 5.4 per cent. That is, the proportion of this group attending a university or CAE increased by 1 per cent to 5.4 per cent between 1986 and 1996. In this way, the changing age structure of educational participation can be analysed.

A striking fact is that, although there have been increases in the indigenous youth participation rate at universities and CAEs, the increases for non-indigenous youth have been much larger. For example, in 1986 the participation rate for indigenous females aged 20 to 24 was 3.7 per cent. By 1996, participation had increased to 6.8 per cent, an increase of 3.1 percentage points. This can be contrasted with non-indigenous females aged 20 to 24, whose participation rate was 9.7 per cent in 1986 and had increased to 19.4 per cent in 1996, an increase of 10.4 percentage points.

In 1996, the participation rate at universities and CAEs for indigenous males and females aged 30 years and older was generally higher than for nonindigenous males and females aged 30 and older. This is suggestive of some catch up of educational attainment for older indigenous males and females. This may reflect better government funding for students, full pay scholarships, study leave with pay, and perhaps positive changes in the attitudes, programs and entry procedures of universities and vocational education and training institutions and providers (Schwab, 1996).

Table 4 presents a cohort analysis of participation rates at TAFE and the other non-university tertiary education institutions. For indigenous males and females aged 15 to 24 years in 1996, a lower proportion of the cohort was attending a TAFE or other post-secondary educational institution than non-indigenous males and females aged 15 to 24 years. However, there was a narrowing of the gap in participation rates in such institutions for indigenous compared with nonindigenous people.

Table 4 Cohort analysis of proportion of population attending TAFE or
other non-university/CAE post-secondary educational institution,
1986-96

                                  Indigenous

Age at 1986 Census           1986    1991     1996

Male

Aged 5 to 9 years              --      --      8.8
Aged 10 to 14 years            --     9.1      9.9
Aged 15 to 19 years           8.0     9.4      8.9
Aged 20 to 24 years           8.9     7.8      8.2
Aged 25 to 29 years           7.2     7.2      7.8
Aged 30 to 34 years           7.4     6.7      6.8
Aged 35 to 39 years           7.1     5.8      7.2
Aged 40 to 44 years           6.6     6.3      7.3
Aged 45 to 49 years           7.1     7.2      7.0
Aged 50 to 54 years           2.8     6.3      8.5
Aged 55 to 59 years           0.0     7.1      5.7
Aged 60 to 64 years           2.4     4.2      4.9
Aged 65 years and over        3.9     9.0      4.0
Total 15 plus                 7.0     7.8      8.3

Female

Aged 5 to 9 years              --      --      8.3
Aged 10 to 14 years            --     7.8      8.7
Aged 15 to 19 years           6.6     7.1      8.5
Aged 20 to 24 years           8.6     7.9      9.7
Aged 25 to 29 years           9.1     8.2      9.2
Aged 30 to 34 years          10.1     7.7      8.8
Aged 35 to 39 years           9.6     6.7      8.1
Aged 40 to 44 years           7.8     7.5      7.5
Aged 45 to 49 years           8.5     7.7      9.8
Aged 50 to 54 years           8.3     8.7     10.3
Aged 55 to 59 years           9.7     8.9     11.5
Aged 60 to 64 years           7.5     1.8      6.1
Aged 65 years and over        9.7     8.1      7.8
Total 15 plus                 8.5     7.6      8.8

                                Non-indigenous

Age at 1986 Census           1986    1991     1996

Male

Aged 5 to 9 years              --      --     11.1
Aged 10 to 14 years            --    12.3     10.9
Aged 15 to 19 years          13.0    11.1      6.3
Aged 20 to 24 years          10.5     7.1      5.3
Aged 25 to 29 years           7.3     6.0      4.5
Aged 30 to 34 years           6.1     4.9      3.6
Aged 35 to 39 years           5.2     3.7      2.7
Aged 40 to 44 years           4.3     3.0      2.3
Aged 45 to 49 years           3.8     2.5      2.4
Aged 50 to 54 years           3.7    22.0      2.3
Aged 55 to 59 years           4.0     2.3      2.6
Aged 60 to 64 years           4.9     2.7      2.9
Aged 65 years and over        6.7     4.4      3.4
Total 15 plus                 6.8     5.8      5.1

Female

Aged 5 to 9 years              --      --      8.9
Aged 10 to 14 years            --     8.7      9.4
Aged 15 to 19 years           8.3     8.5      6.5
Aged 20 to 24 years           7.6     6.0      5.7
Aged 25 to 29 years           5.8     5.5      5.5
Aged 30 to 34 years           5.8     5.2      4.8
Aged 35 to 39 years           5.7     4.4      3.7
Aged 40 to 44 years           4.9     3.5      3.0
Aged 45 to 49 years           4.3     2.8      2.6
Aged 50 to 54 years           4.4     2.6      2.8
Aged 55 to 59 years           5.1     2.9      3.4
Aged 60 to 64 years           6.6     3.6      4.1
Aged 65 years and over        8.7     5.4      4.5
Total 15 plus                 6.4     5.3      5.3

Source: As for Table 1.

In 1986, 1991, and 1996, the proportion of indigenous people aged over 24 years attending a TAFE or other post-secondary educational institution was higher than that recorded for the non-indigenous population. The only exception to this observation was for indigenous males aged 65 years and over in 1986.

An important point revealed by Tables 3 and 4 is that the life-cycle profile of participation in post-secondary education differs between the indigenous and the non-indigenous population. The non-indigenous population has a higher participation rate in post-secondary education at a younger age than does the indigenous population, which shows much higher participation rates later in the life-cycle.

A significant component of the Working Nation labour market program(3) of the early 1990s was the provision of formal training (typically at TAFE). This may provide an institutional explanation of the later indigenous starts in TAFE and other post-secondary courses. However, if training programs from Working Nation were heavily weighted towards indigenous youth, as were the labour market programs from that policy (Hunter & Gray, 1998), then one should not place too much emphasis on this as an explanation of the substantial increases in postsecondary participation of older indigenous people.

The human capital investment model of education (Becker, 1975) suggests that the level of education a person decides to attain is an investment decision, where the costs are incurred now and the returns accrue over the rest of the person's working life. The human capital model therefore predicts that investment in human capital is more likely to occur early in a person's life as this will leave the longest period for the returns to that investment to be realised. Thus the delayed educational participation pattern of indigenous Australians presents an apparent paradox to the human capital investment model. One possible explanation is that to indigenous communities the return from education is not the private gain of higher future earnings, but rather a gain which is realised by the entire community in the form of increased cultural capital (Schwab, 1996). Family formation at a younger age may also limit participation in education at a younger age by indigenous Australians, particularly women.

Recent trends in qualifications

Some comparisons between the 1986 and the 1991 Censuses are complicated by significant changes to the vocational education and training and the higher education sectors during that period. For example, in 1988 the Commonwealth Government eliminated the `binary divide' that had separated universities and CAEs for 25 years; suddenly the number of Australian universities increased from 19 to over 50. Comparisons of qualifications during this period can be difficult as a result of these changes. Perhaps more significantly, changes in the system of classification of qualification field and level used by the ABS make comparison between the 1986 and 1991 Censuses highly problematic.(4) In this paper, we therefore limit comparison of qualification field and type to the 1991 and 1996 Censuses.

While there have been absolute increases in the proportion of the indigenous population with a post-secondary qualification, there have also been increases for the non-indigenous population. Both indigenous males and females remain severely disadvantaged in terms of post-secondary educational qualifications. Indigenous males just hold their position relative to the non-indigenous population, whereas indigenous females experience a small relative loss. Table 5 shows the qualification level of the working-age population. Because of the changes in the classification of courses and educational institutions, the ratio of the proportion of the indigenous population with a particular qualification to the proportion of the non-indigenous population with a particular qualification is calculated. As this ratio gets closer to one, the less inequality in educational attainment there is between the indigenous and non-indigenous populations.

Table 5 Highest level of qualification in the working-age population,
1991-96

                           Indigenous      Non-indigenous

                           1991    1996     1991      1996

Male
Higher degree               0.2     0.3      1.9       2.4
Postgraduate diploma        0.2     0.4      1.0       1.3
Bachelor degree             1.2     2.6      8.4      10.1
Undergraduate diploma       1.1     1.3      3.0       3.1
Associate diploma           1.0     2.0      2.1       3.6
Skilled vocational         15.3    16.3     24.6      23.8
Basic vocational            3.0     2.7      3.3       2.6
Total qualification        22.0    25.6     44.2      46.9

Female
Higher degree               0.1     0.4      0.9       1.4
Postgraduate diploma        0.6     0.9      2.0       2.4
Bachelor degree             2.1     4.3      8.3      11.4
Undergraduate diploma       4.4     3.2      8.9       6.8
Associate diploma           1.6     3.6      1.7       3.5
Skilled vocational          2.7     3.3      3.5       4.1
Basic vocational            7.1     6.2      6.5       5.5
Total qualification        18.6    21.9     31.8      35.1

                           Ratio
                           indigenous/
                           non-indigenous

                           1991    1996

Male
Higher degree              0.13    0.13
Postgraduate diploma       0.24    0.31
Bachelor degree            0.14    0.26
Undergraduate diploma      0.36    0.42
Associate diploma          0.46    0.56
Skilled vocational         0.62    0.68
Basic vocational           0.91    1.04
Total qualification        0.50    0.55

Female
Higher degree              0.14    0.29
Postgraduate diploma       0.30    0.38
Bachelor degree            0.26    0.38
Undergraduate diploma      0.49    0.47
Associate diploma          0.95    1.03
Skilled vocational         0.78    0.81
Basic vocational           1.08    1.13
Total qualification        0.58    0.62

Note: The not stated and inadequately described categories
are proportionately allocated to other cells. The working-age
population is defined as those aged 15 and over.

Source: As for Table 1.

For indigenous females, the proportion with any qualification increased from 18.6 per cent in 1991 to 21.9 per cent in 1996. For non-indigenous females, the proportion with some post-secondary qualification also increased by 3.3 percentage points to 35.1 per cent in 1996. For indigenous males the proportion with some qualification increased from 22.0 per cent in 1991 to 25.6 per cent in 1996, an increase of 3.6 percentage points. Within the non-indigenous male population, the proportion with some qualification increased by 2.7 percentage points to 46.9 per cent in 1996.

For indigenous males and females, the degree of inequality in educational attainment, as compared with non-indigenous males and females, increases with qualification level. Between 1991 and 1996, there was a significant narrowing of the gap for all educational levels. Indigenous males were more qualified than their non-indigenous counterparts only for basic vocational qualifications. For indigenous females, there had been complete catch-up in basic vocational qualifications in 1991 and complete catch-up had also occurred in relation to associate diplomas by 1996.

The extent of catch-up in the proportion of the population with a bachelor degree is particularly striking. For indigenous males, the ratio of the proportion of the indigenous to non-indigenous population with a bachelor degree increased from 0.14 in 1991 to 0.26 in 1996. For indigenous females, the ratio increased from 0.26 to 0.38 between 1991 and 1996.

Table 6 reports the field of qualification for 1991 and 1996 Censuses. For each qualification field, indigenous people have a lower proportion of their population with a qualification than does the non-indigenous population (with the exception of males with a miscellaneous qualification). The pattern of increases in educational qualifications by field of qualification for indigenous males and females is very closely correlated with that in the non-indigenous population. For indigenous males, between 1991 and 1996, there was a faster rate of growth in the proportion of the population with qualifications in all fields except for the field of natural and physical sciences. For indigenous females, there was a faster rate of growth in the proportion of the population with qualifications in the fields of education, society and culture, engineering and miscellaneous fields.

Table 6 Field of highest qualification for the working-age population,
1991- 1996

                                   Indigenous     Non-indigenous

                                  1991    1996    1991    1996

Male
Business and administration        1.2     2.7     4.5     5.8
Health                             0.4     0.8     1.8     2.0
Education                          0.9     1.2     2.0     2.2
Society and culture                1.8     2.5     3.5     4.1
Natural and physical sciences      0.4     0.9     2.3     2.9
Engineering                        8.1     8.9    18.7    19.2
Architecture and building          4.2     4.4     6.1     6.2
Agriculture                        2.2     1.3     2.4     1.5
Miscellaneous fields               2.8     2.9     2.9     2.9
Total qualified                   22.0    25.6    44.2    46.9

Female
Business and administration        5.3     6.1     8.0     9.1
Health                             3.8     3.8     7.7     7.6
Education                          2.5     3.2     5.7     6.0
Society and culture                3.3     4.9     4.6     6.0
Natural and physical sciences      0.3     0.7     1.3     1.8
Engineering                        0.5     0.8     1.2     1.4
Architecture and building          0.1     0.1     0.2     0.3
Agriculture                        1.2     0.3     0.9     0.3
Miscellaneous fields               1.5     2.0     2.2     2.6
Total qualified                   18.6    21.9    31.8    35.1

Note: The not stated and inadequately described categories are
proportionately allocated to other cells. The working-age population
is defined as those aged 15 and over.

Source: As for Table 1.

Reflections on indigenous education

The survey of findings presented above provides some insight into patterns of various educational indicators and outcomes. In this section, we will reflect on some of those patterns and explore some of the issues that the empirical data raise for policy makers. Some of the issues are concrete, others more speculative, but all are relevant to the formulation of future education policy.

Problem for changing relative education levels

The implications for labour market success of an absolute improvement but a relative decline in indigenous educational outcomes are shaped by the role that increased educational attainment plays in improving labour market outcomes. If education in itself leads to increases in productivity and employability, then we would expect this to translate into absolute improvements in the labour market outcomes (in terms of employment rates and wage level) of indigenous Australians. Notwithstanding these absolute improvements, we would also expect a decline in employment and wage rates relative to the non-indigenous population.

If there is aggregate excess labour supply (unemployment), as is the case in the Australian labour market, and education improves the productivity of workers, then it is more likely that there may be an absolute, as well as relative, decline in indigenous employment outcomes. However we would still expect absolute improvements in the wage outcomes of indigenous Australians.

Although conventionally it is argued that increased education leads to improved labour market outcomes by increasing individual productivity, an alternative view is that education leads to improved labour market outcomes by providing a signal to employers of a person's innate productivity. Stated in a different way, when employers are deciding whom to employ for a job, they cannot determine precisely what each applicant's actual productivity will be. They therefore have to make some assessment or an educated guess as to the probable productivity of each applicant. If people who have a higher level of educational attainment have higher innate productivity, then employers may use a person's highest level of education as an indicator of potential productivity. In this case, employers are not looking at a person's absolute level of educational attainment, but rather at that person's relative attainment.

If education levels of a given population double, this will have no effect on the probability of any particular individual being employed vis-i-vis any other individual. In addition, it should have no effect on wages. However, if the education level of indigenous Australians increased in absolute terms, but fell relative to the education attainment of non-indigenous Australians, then employers may assume, on the basis of relative educational attainment, that indigenous workers have lower potential productivity. The outcome could well be that indigenous employment rates and wages would fall relative to that of non-indigenous workers and may, in fact, worsen in absolute terms (particularly if there is surplus labour supply). Indeed Hunter and Gray (1998) point to an absolute and relative decline in employment between 1986 and 1996.

Factors underlying poor indigenous education outcomes

A recent analysis of the determinants of educational attainment among young indigenous Australians revealed that a range of variables influence the probability of attending school (Hunter & Schwab, 1998). For example, place of residence affects school attendance only for teenagers in remote communities who are markedly less likely to be in school than their urban counterparts.(5) A range of social environmental factors was found to affect the likelihood of a young person attending school. Both poor quality housing and residence in a household where others have been arrested decrease the probability that a young person will be attending school, whereas the presence of household members with qualifications or who are attending school significantly increases the likelihood of school attendance. Arrest was found to reduce the probability of attending school but was not found to be significantly related to having post-school qualifications. The research also indicated that difficulty with English is also an important predictor of whether or not an individual has educational qualifications.

Effect of geographical location

Indigenous Australians in rural and remote areas have very low levels of educational attainment, compared with both indigenous Australians in urban areas and non-indigenous Australians in rural areas (ABS, 1995a). For example, in 1991, 42 per cent of persons in rural areas, aged 15 and over, had left school before they were 16, as compared with only 36 per cent of persons living in urban areas in the same category (ABS, 1995b). The proportion of men in rural areas with postschool qualifications (27 per cent) was less than that of men in urban areas (32 per cent). For females, there was little difference between women in rural and urban areas with similar proportions having post-school qualifications. This suggests that, given the overall level of labour demand and the level of demand for different qualifications, increases in the educational attainment of indigenous Australians in rural areas should lead to improved chances of finding employment.

Another way to think about this is that, in regional areas, there is surplus labour and the jobs which are available have to be rationed in some way. If increases in educational qualification either lead to a bigger pool of jobs from which a person could potentially be employed, or even just give them a better chance of getting a rationed job, then this may lead to improvements in indigenous labour market outcomes. However if, as seems to be the case, regional economies are declining in importance, then indigenous workers in rural areas are competing in a better educated workforce for a declining pool of jobs. That is, the relative educational deficit of the rural indigenous population, vis-i-vis other rural residents, is even more important than it once was.

Role of targeted assistance

ABSTUDY, the Aboriginal Study Assistance Scheme, includes a range of programs and income support allowances and supplements intended to foster indigenous participation in education beyond secondary schooling, it is one of the most contentious special programs in indigenous affairs and is poorly understood by many in the community; it is sometimes cited as a program that provides an unfair advantage to indigenous students and their families. In May 1997, a number of changes to ABSTUDY, including substantial reductions in funding by the fiscal year 2000-01, were announced and implemented in 1998. Subsequently the Commonwealth Government announced its intention to conduct a review of the program in the context of its plan to introduce an all-encompassing Youth Allowance to replace ABSTUDY (Schwab & Campbell, 1997). In December 1998, with the results of the review still unreleased, another collection of changes to ABSTUDY was announced. Effective from January 2000, these changes signal a continuation of the Commonwealth Government's desire to align all indigenous educational assistance programs with the common Youth Allowance. According to the announced changes, ABSTUDY will remain as a separate named program, but many of the special programs and income support allowances will be tightened up and means tested under guidelines currently in place through the Youth Allowance. The impact of these changes is difficult to predict at this early stage. Indeed there is as yet no research available on the impact of the first round of changes implemented in 1998. It seems likely, however, that since access and eligibility will be reduced for some who would have in the past had access to the program, there will be some corresponding reduction in participation.

ABSTUDY, and its predecessor, have been in place for nearly three decades and have come to symbolise recognition of the special educational disadvantage of indigenous Australians. The program is seen by indigenous people to be a `tried and true' means of enabling access and participation where it would otherwise be difficult (Stanley & Hansen, 1998). At the same time, it has long been clear that indigenous students have had and continue to have unique needs as a result of cultural differences and a history of disadvantage and dispossession. ABSTUDY has provided greater programmatic and administrative flexibility than has been possible under AUSTUDY or the new Youth Allowance scheme to meet the special cultural needs of indigenous students (for example, travel for cultural activities, increased options for bridging and preparatory courses). Consequently the latest round of reductions in ABSTUDY benefits and changes in eligibility criteria must be evaluated carefully; there is among many educational practitioners serious concern that indigenous participation in education at every level could well be impeded.

Where there have been achievements in indigenous access, participation, and outcomes in education, there has often been a corresponding recognition of cultural difference which has assisted these advances (Bourke, Burden, & Moore, 1996). Programs which encourage indigenous participation in education, while simultaneously being responsive to family, community and cultural commitments, have been powerful tools for bringing indigenous students into educational settings. The variety of course delivery modes has been instrumental in allowing individuals to choose the level of engagement, study environment, and method of study most suitable to them at particular stages of their lives. That some of these options are now under threat has led many educators to predict decreases in participation in higher education by Aboriginal and Torres Strait Islander students (Schwab & Campbell, 1997).

Low levels of indigenous employment remain one of the most intractable of contemporary social issues and it is likely that current reforms in industrial relations and labour market programs will, if anything, exacerbate this problem (Hunter, 1997; Taylor & Altman, 1997). Policies aimed at reducing or eliminating targeted educational support programs for indigenous people in favour of mainstream programs risk undermining existing gains in educational participation, and ultimately employment.

Ongoing impediments to indigenous education

There are several reasons for low indigenous attendance at educational institutions. These include disaffection with school, difficulties of attending school arising from poverty, the experience of arrest, high mobility, indigenous inter-group tensions, family pressures particularly in single parent families, high levels of sickness and high death rates among adults and the consequent social obligations (Groome & Hamilton, 1995, p. 4; Hunter & Schwab, 1998; Schwab, 1998).

Though empirical data on this issue are scarce, racism and cultural miscommunication appear to be important factors influencing decisions by some indigenous students to abandon school. A significant number of indigenous students, when asked to reflect on why they had left school, said that they had felt depersonalised and had lost self-esteem under the pressure of racial harassment and `put downs' from both teachers and students (Groome & Hamilton, 1995, p. 45).

Racism from teachers is a more difficult experience to deal with than racism from other students. The types of racism experienced include racial abuse and vilification, negative comments about families and behaviour on the basis of race, prejudicial treatment, negative personal comments about `extra money' and `special benefits' (Groome & Hamilton, 1995, p. 37).

The above factors are clearly a result of a prolonged history of cultural conflict and policies that failed to meet the needs of indigenous students. Indeed a number of submissions to the National Inquiry (1997) into the `stolen generation' drew attention to the relationship between past racist policies and practices in education, which excluded or marginalised indigenous children, and contemporary low secondary school retention rates and low participation rates in tertiary education.

Policy reviews over the past 25 years have identified a range of important areas where attention should be concentrated in attempts to improve outcomes in indigenous education. Prominent among these are: promotion of increased involvement of indigenous families, educators and communities in educational decision making at the local, regional and national levels; increases in the number of indigenous people employed in education and training; ensuring equitable access and participation to education and training; promotion, maintenance and support of the teaching of indigenous studies, cultures and languages; and the provision of community development and training including English literacy and numeracy for indigenous adults. It will continue to be important to track and assess the outcomes of policies aimed at assisting indigenous people as they attempt to overcome the many impediments to their educational success.

The main finding of this paper is that there have been some absolute improvements in indigenous educational outcomes over the period 1986 to 1996. Indigenous post-secondary qualification rates, especially for mature age students, have increased at a faster rate than the rates for non-indigenous Australians. However it is of concern that relative to the non-indigenous there have been little if any real gains. The lack of improvement relative to the non-indigenous population occurs not only in the proportion of the population with post-secondary qualification, but also in the proportion of indigenous teenagers staying at school. By any measure, the indigenous population remains severely disadvantaged.

Keywords 

Aboriginal education      educational economics    human capital
educational attainment    educational policy       Torres Strait
                                                    Islander education

Acknowledgements

We are especially indebted to Dr John Taylor and Professors Jon Altman and Max Neutze and two anonymous referees for their insightful comments. The research in this paper was partially funded by the Ronald Henderson Research Foundation.

Notes

(1) The Australian Bureau of Statistics (ABS) standard for indigenous status is the following: An Aboriginal or Torres Strait Islander is a person of Aboriginal or Torres Strait Islander descent who identifies as an Aboriginal or Torres Strait Islander and is accepted as such by the community in which he or she lives.

(2) The education levels reported do not tell us what happened to the educational attainment of the original population. In order to motivate the inter-censal comparisons, we need to assume that the experience of people who identified as indigenous for the first time in the last census is the same, at least in terms of education, as those who identified in 1986 and 1991 (see Hunter, 1998).

(3) In May 1994, the Federal Government introduced a set of labour market programs targeted at the long-term unemployed. The main features included the provision of formal training (typically at TAFE), a big expansion in labour market programs, case management of the unemployed, a Youth Training Initiative, training wages for all trainees, and direct job creation. In addition, any person who had been on unemployment allowances for over 18 months was offered a full-time job (for at least 12 months) mainly in the private sector. The program was phased out after the change of federal government in March 1996.

(4) The ABS classification system of the field of qualification changed between 1986 and 1991 to the ABS Classification of Qualifications (ABSCQ). According to the ABS, the ABSCQ maintained some degree of comparability with the 1986 Census classification. However, differences in the classification structure coding process used in the 1986 Census may pose practical difficulties when attempting detailed comparison.

(5) Rural areas and urban centres are defined respectively as population clusters of less than 1000 and 1000 or more people. Remote areas are defined as being in a rural area that is more than 100 kilometres from the nearest TAFE (Hunter & Borland, 1997).

References

Aboriginal and Torres Strait Islander Commission. (1994). Review of the Aboriginal Employment Development Policy. Canberra: Author.

Aboriginal Consultative Group. (1975). Education for Aborigines: Report to the Schools Commission. Canberra: Schools Commission.

Australian Bureau of Statistics. (1993). ABS Classification of Qualifications (ABSCQ) (Cat. no. 1262.0). Canberra: Author.

Australian Bureau of Statistics. (1995a). National Aboriginal and Torres Strait Islander Survey 1994 : Detailed findings (Cat. no. 4190.0). Canberra: Author.

Australian Bureau of Statistics. (1995b). Year book, Australia, 1995 (Cat. no. 1301.0). Canberra: Author.

Becker, G. (1975). Human capital: A theoretical and empirical analysis, with special reference to education (2nd ed.). New York: Columbia University Press.

Bourke, C.J., Burden, J., & Moore, S. (1996). Factors affecting performance of Aboriginal and Torres Strait Islander students at Australian universities: A case study. Canberra: Department of Employment, Education, Training and Youth Affairs, Higher Education Division, Evaluations and Investigations Program.

Committee of Review of Aboriginal Employment and Training Programs (M. Miller, Chair). (1985). Report ... Canberra: AGPS.

Department of Employment, Education and Training. (1988). Report of the Aboriginal Education Policy Task Force (P. Hughes, Chair). Canberra: AGPS.

Groome, H. & Hamilton, A. (1995). Meeting the educational needs of Aboriginal adolescents (National Board of Employment, Education and Training, Commissioned Report No. 35). Canberra: AGPS.

Hunter, B. (1997). An indigenous worker's guide to the Workplace Relations and Other Legislation Amendment Act. Journal of Industrial Relations, 39(4), 439-456.

Hunter, B. (1998). Assessing the utility of the 1996 Census on indigenous Australians (Centre for Aboriginal Economic Policy Research Discussion Paper No. 154). Canberra: Australian National University.

Hunter, B. & Borland, J. (1997). The interrelationships between arrest and employment: More evidence on the social determinants of indigenous employment (Centre for Aboriginal Economic Policy Research Discussion Paper No. 136). Canberra: Australian National University.

Hunter, B. & Gray, M.C. (1998). Recent changes in the Australian workforce: A focus on the structure of indigenous employment. Australian Bulletin of Labour, 24(3), 222-242.

Hunter, B. & Schwab R.G. (1998). The determinants of indigenous educational outcomes (Centre for Aboriginal Economic Policy Research Discussion Paper No. 160). Canberra: Australian National University.

Long, M. & Frigo, T. (1998). Have school retention and participation improved for indigenous students? Paper presented at the Australian Population Association Conference, September-October 1998, Brisbane.

National Inquiry into the Separation of Aboriginal and Torres Strait Islander Children from their Families. (1997). Bringing them home: Report. Sydney: Human Rights and Equal Opportunity Commission.

National Review of Education for Aboriginal and Torres Strait Islander Peoples (M. Yunupingu, Chair). (1995). Final Report. Canberra: AGPS.

Royal Commission into Aboriginal Deaths in Custody. (1991). National Report (Vol. 2). Canberra: AGPS.

Schwab, R.G. (1996). Indigenous participation in higher education: Culture, choice and human capital theory (Centre for Aboriginal Economic Policy Research, Discussion Paper No. 122). Canberra: Australian National University.

Schwab, R.G. (1998). Educational failure' and educational `success' in an Aboriginal community (Centre for Aboriginal Economic Policy Research, Discussion Paper No. 161). Canberra: Australian National University.

Schwab, R.G. & Campbell, S.F. (1997). The future shape of ABSTUDY: Practical and policy implications of the recent proposed changes (Centre for Aboriginal Economic Policy Research, Discussion Paper No. 140). Canberra: Australian National University.

Stanley, O. & Hansen, G. (1998). ABSTUDY: An investment for tomorrow's employment. Canberra: Aboriginal and Torres Strait Islander Commission.

Taylor, J. & Altman, J.C. (1997). The job ahead: Escalating economic costs of indigenous employment disparity. Canberra: Aboriginal and Torres Strait Islander Commission.

Dr M. C. Gray is a Post-doctoral Fellow and Dr B. H. Hunter and Dr R. G. Schwab are Fellows at the Centre for Aboriginal Economic Policy Research, The Australian National University.

Correspondence concerning this article should be addressed to Dr Boyd Hunter, Centre for Aboriginal Economic Policy Research, Australian National University, Canberra, Australian Capital Territory 0200. [e-mail]