Contents - Previous - Next

This is the old United Nations University website. Visit the new site at

7. Education and spatial disadvantage

S.R. Walker

It has been asserted by a number of writers that remote areas are disadvantaged with respect to education provision and levels of attainment. Fyfield (1970), for example, remarks that "anyone who has been associated with both country and metropolitan schools knows how unequal they are," pointing out that a lower proportion of non-metropolitan high school students go on to attend university and that retention rates are lower. Radford (1962) noted that in Victoria, of all boys leaving metropolitan high schools, 14 per cent became university students, whereas only 5 per cent of those leaving nonmetropolitan high schools attended university. More recently, the Henderson Report (1975) also identified differences in levels of attainment according to place of residence. In rural areas, the retention rate is lower than in metropolitan centres, and the percentages of those who have never attended school or attended only primary school are higher.

Many of the reasons forwarded for these regional differences are related to remoteness or other environmental factors. Fyfield, for example, suggests that country children (or those living in remote areas) are disadvantaged as a direct result of their distance from the metropolitan centres that provide the cultural opportunities and range of occupations thought to be desirable in encouraging positive attitudes towards further education, In more remote areas, school facilities may be limited (the argument holds), and the teachers are younger, less experienced, discouraged from additional training, and have a high turnaround. The Henderson Report (1975) also suggested that jobs and careers that require a high level of education and that may be more attractive are concentrated in the metropolitan centres, and this contributes to the migration of the more educated to the larger cities. Indeed, it is this ruralurban drift that has frequently been the principal source of concern in the discussion of the implications of such spatial disparities, whether with respect to service provision or service outputs (the levels of educational attainment). (See also table 5.20, where Burnley summarizes reasons for leaving Alice Springs, based on a field study in August 1980.)

Such data as are provided by those studies suggest that arid zone towns in Australia, being remote from metropolitan centres, are likely to be characterized by relatively low levels of educational attainment. However, this effect of remoteness can only be supposition without knowledge of the effects of other factors also known to be of importance (socio-economic status and cultural background, for instance). This study will examine, on an area basis, the extent to which a measure of educational attainment is related to remoteness, controlling for the effect of examples of these variables.

Even if differences in educational attainment are found between the remote arid zone towns and other places, this still begs the question of whether in fact improving the provision of educational services will affect the outcome (educational attainment). Jencks (1972) has argued that school quality has little effect on educational attainment -that cultural attitudes, taste for schooling, and family background are more important in explaining variation in educational attainment. Equal expenditures on education do not necessarily mean equal outcomes. Similarly, Rich (1979) has also argued that a failure in previous studies of urban service distributions has been the focus on service inputs rather that on outcomes.

That remote areas may be disadvantaged with respect to outcomes has been suggested by a number of studies, but it is less clear whether such outcomes are the result of differences in service provision. Fyfield (1970) remarks, in the case of Australia, "There is no reason to doubt that the Education Department seeks to follow a policy of impartiality in basic provisions, though one could point to inequalities in money spent on items regarded as extras" (p. 99). However, there is little evidence to suggest that such variations in funding disadvantage remote areas. For example, Coates, Johnson, and Knox (1977) studied differences in local government expenditure per school child in Britain, but the implications of their analysis suggest that factors other than remoteness or less residentially desirable locations were of primary importance.

There is, therefore, a need to consider these issues in the case of Australian arid zone urban areas: to what extent is remoteness or the physical environment a determinant either of educational provision or of outcomes? To what extent is there an association between such inputs and outcomes? Clearly the answers to those questions have a number of policy implications, not the least of which is whether increased funding on education in remote areas will assist in preventing the out-migration of young people from such centres. At the time of writing, a university is planned for the Northern Territory (to include an institution in Alice Springs). This may perhaps be considered as part of a policy of dispersal of higher education facilities, which has been practiced in recent years in Australia (with mixed results). Whether such expenditure is justifiable is beyond the scope of this report, but the analysis may be able to ascertain to a limited extent the effects improving access to educational facilities. (The analysis is limited by reason of the nature of the variables available for inclusion in the study.)

TABLE 7.1. Educational Attainment and the Provision of School Facilities: Correlations with Other Variables Selected

Variable Education variables
Enrollment provision╣ Teacher provision▓ Teacher provision│ Level of provision4
Road distance to nearest state capital -.219* -.026 -.089 -.128
Road distance to the nearest of Sydney or Melbourne -.203* -.068 -.115 -.196*
Road distance to nearest town of the 118 surveyed -.209* -.039 -.076 -.238*
Population recorded in the census 1976: size of local        
government area -.428* -.183* -.137 -.286*
Percentage change in population 1971-1976 -.092 -.025 .045 -.364*
Percentage of households with incomes>=A$15,000 (1976) -.297* -.002 -.085 -.520*

1 Ratio of children 5-17 years to total school enrollments (1976); higher values indicate "better" provision,
2 Ratio of teachers resident to total population, 1976.
3 Ratio of teachers resident to persons aged 5-17 years, 1976.
4 Percentage of population over 15 years with no qualifications.

* Significant at the .05 level.

Analysis: The Relationship between Remoteness and Urban Education

Firstly, if Fyfield's assumption is correct (that a policy of impartiality in service provision may be expected, then one might expect that remote areas are not disadvantaged with respect to indices of education provision. To investigate this proposition, three measures of provision were selected (table 7.1) and related to three measures of remoteness. Pearson correlation coefficients were calculated to ascertain the strength of the relationship between the variables. Despite the similarity of results among the three distance variables and between the two teacher/child ratios, all coefficients are included for comparison.

With respect to the provision of teachers, the data suggest that remote urban centres are not necessarily more disadvantaged than towns closer to metropolitan areas. Certain arid zone towns, namely Alice Springs and Tennant Creek, are relatively well provided in this regard, with levels of pro" vision comparable to those of other Australian towns (table 7.2). (The university centre of Armidale scores highest on these indicators.) These are also many centres that score worse than Mt. Isa or Broken Hill-Queanbeyan, Dubbo, Cessnock, and Bega, for example. These New South Wales urban areas indeed suggest that proximity to another area which is residentially attractive, as much as distance, can deter teachers from residing in a particular town.

On the other hand, there is a significant correlation between the remoteness variables and the provision of school places (per child of school age) Regarding this criterion, the more remote areas do appear to be relatively disadvantaged (table 7.11. This can be seen as substantiating earlier findings concerning retention rates of country schools: however, what the authors of these previous studies failed to note was a possible explanation for such differences in retention rates. That there are fewer enrollments in remote areas than expected (from the numbers of school children which the census records as being resident in such centres) suggests the effect of boarding schools on the market demand for local school places. This is further supported by the significant correlation between household income and local enrollment provision: incomes tend to be higher in towns less well provided. Boarding school may provide an appropriate alternative to local school for parents financially able to meet the costs. A number of parents surveyed in Alice Springs also viewed these schools as providing better educational opportunities and future prospects for their children. However, improvements to the local high school would not necessarily lead to a change in such parental preferences. Both income and attitudes will affect the retention rates of children in such local schools.

TABLE 7.2. Comparisons between Arid Zone Towns and Other Urban Areas: Education and Socio-economic Status, 1976

Variable 118 centres total1 Alice Springs Katherine Tennant Creek Kalgoorlie Mt. Isa Broken Hill
x (s)
Ratio of children 5-17 years to total school enrollments (1976); higher values indicate "better" provision 123 (37.6) 106 118 118 64 92 86
Ratio of teachers resident to total population 1976 18 (5.1) 24 19 22 19 16 13
Ratio of teachers resident to persons aged 5-17 years, 1976 70 (22.4) 92 73 86 74 60 55
Percentage of population over 15 years withno qualifications 68 (4.4) 57* 57* 60 63 65 72
Percentage of households with incomes>=A$15,000 (1976) 18 (4.7) 32* 25 23 22 30* 19
Percentage of the labour-force unemployed,1976 census 5 (1.6) 3 5 4 3 5 6
Ratio of females in the labour force to womenover 15 yrs 38 (5.3) 53* 54* 44 42 48 27
Percentage change in population 1971-1976 5 (11.0) 27* 24 25 -8 .. -7

* Values more than two standard deviations from the mean.
1 Mean and standard deviation, 118 urban centres in the study.

The provision of services is also dependent on the stage of development of an urban area, and this is exemplified in the strong correlation between population change and service provision (table 7.1). That is, in centres of faster growth- regardless of their location-education services tend to be underprovided. This is the result in part of a lag effect (or, conversely, an oversupply in towns of declining population numbers! and in part of the numbers of young school-leavers in these towns who are of school age but are seeking employment rather than remain at school. Many in this group will have come to the town from elsewhere, so that the levels of provision in themselves may be misleading. This effect occurs mainly in coastal areas, as for instance in Shellharbour, Port Macquarie, and the Gold Coast, which all have relatively low levels of education service provision -all lower scores than the arid zone towns.

If one were to conclude from this brief study of some serviceprovision indicators that remoteness posed some disadvantage with respect to the facilities available, one might also consider that educational outcomes would also be so spatially distributed. Radford (1962), and others since, have indicated that relatively fewer students from country schools enter university: this might suggest that the remote arid zone towns are disadvantaged in terms of the proportion of qualified adults. To test this, correlation coefficients were again calculated for selected variables (table 7.1).

Far from the central areas of Australia being educationally disadvantaged regarding this criterion, these towns (with the exception of Broken Hill) tend to be relatively well-provided with qualified persons (tables 7.1 and 7.2). The further the distance from the main metropolitan centres, the higher the percentage of adults who have some tertiary qualification. Alice Springs, Katherine, and Darwin all score significantly above average on this index: of the 118 urban areas surveyed, these three were characterized by the highest proportions of qualified persons. The explanation for this relates in part to the functions of these urban areas, requiring specialist skills such as research or engineering (the Pine Gap station at Alice Springs, for example, employs highly qualified personnel on short-term contracts from the United States). Household incomes in these towns are concomitantly exceedingly high. It is readily apparent that educational qualifications in an area might depend more on factors other than the provision of school facilities or locational disadvantage.

School Facilities and Outcomes

To state that remote areas may be educationally disadvantaged because of the level of service provided implies linked effects. The first is that output of services will change the educational outcomes in an urban area, thereby achieving the desired equitable standard of educational attainment. The second is that improvements in the quantity of services provided will result in improved quality of service at the same time. These two related effects are to be tested as far as possible within the limitations of the data. The former will involve a statistical analysis to investigate the extent to which service provision does influence a measure of educational attainment in an urban centre, and the latter will draw upon residents' attitudes to assess the quality of service provided. The measure of educational attainment used is "qualifications" achieved.

TABLE 7.3. Determinants of the Level of Educational Qualifications: Step-wise Multiple Regression

Variable r R Overall F Significance
Percentage of households with incomes >= A$15,000 (1976) - .517 .517 41.87 <.0001
Percentage change in population 1971-1976 - .354 .581 29.12 <.0001
Percentage of employed females in retailing or finance, 1976 .149 .598 21.00 <.0001
Population recorded in the census 1976: size of local government area -.280 .605 16.18 <.0001
Ratio of teachers resident to persons aged 5-17 years, 1976 - .063 .608 13.00 <.0001
Road distance to nearest state capital - .124 .609 10.83 <.0001
Road distance to nearest town of the 118 surveyed - .232 .613 9.37 <.0001
Ratio of children 5-17 yrs to total school enrollments (1976), higher values indicate "better" provision .223 .614 8.17 <.0001
Road distance to the nearest of Sydney or Melbourne -.198 .615 7.22 <.0001
Retail sales per capita, 1973-1974 .136 .615 6.45 <.0001
Percentage of households with incomes <= A$6,000 (1976) .368 .615 5.81 <.0001

A step-wise regression model was employed to measure the effect of a number of variables on the level of educational qualifications in an area (table 7.3). While one of the serviceprovision variables was found to be significantly correlated with educational qualification, it is evident from the analysis that qualifications were much more related to income and population characteristics than to the level of service provision (table 7.3). Four variables entered into the regression equation as being discriminant before the indicator of service provision: the proportion of households with high incomes, population change, population size of town, and employment in retail or finance activities. Together, these are also more important determinants of the level of qualifications in a centre than are the measures of remoteness; once the effect of these variables has been accounted for, it can be seen that the remoteness measures contribute little to the explanation.

The data suggest, therefore, that educational outcomes need not necessarily be dependent on inputs, and further, that distance from the metropolitan centres need not be a disadvantage in itself. Higher qualified persons are attracted to towns of larger size, rapid expansion, and financial rewards. If a town in an arid area can provide these opportunities, then highly qualified persons will be attracted there (perhaps for only a short-term period) regardless of the remote and arid location. In the Gillen district survey of Alice Springs, most of the qualified persons stated that they moved to the town for such job opportunities or financial gain. Other professional workers are also attracted there to provide social services or assistance to disadvantaged groups; the presence of legal aid or employment centres, the Institute for Aboriginal Development, including a number of graduate linguistic specialists, the Department of Aboriginal Affairs, and research workers all contribute to the high level of education attainment apparent in Alice Springs (table 7.2).

Hence, despite the apparent association between one measure of service provision (Enrollment provision 7.1) and qualifications, once the effects of social and population variables are controlled for, such provision explains very little of the variation in the level-of-education index. Table 7.4 again shows that income, population size of town, and the importance of a town as a retail centre explain more of the variance in school-enrollment provision than does the level of educational achievement. Similarly with respect to teacher provision: educational need of an area is of less importance than population size of town or its retail centrality (table 7.5). However, it is interesting to note that in both cases, it is the smaller towns that tend to be advantaged with respect to these measures of service provision, rather than the larger. This may mean, therefore, that certain remote centres may be at an advantage, despite -or even because of-their remoteness.

However, even if towns in remote areas are not entirely disadvantaged in terms of the quantity of facilities provided, they may be at some disadvantage in qualitative terms. To assess the quality of service is by no means easy, although one possible approach is to ascertain residents' opinions of the service they are receiving. This approach was adopted here: Alice Springs respondents in the Gillen survey were asked both an open-ended question about disadvantages of living in Alice, and a specific question on their assessment of the adequacy of educational services.

Overall, residents" attitudes towards local education facilities were favourable. Ratings of "good," both with respect to school teaching and facilities, were common. About 22 per cent of the respondents made unfavourable comments concerning school provision; the most frequently mentioned items are shown in table 7.6. Concern was generally expressed not about schools in general but the high school provision. Parents' complaints in this respect were not so much related to the school facilities (although gymnasium and theatre additions were suggested) but to the perceived quality of the teaching. Residents suggested that this was due to the rapid turnover of teachers (one child reportedly having experienced six teachers in the space of a single year), which in turn was the result of the difficulty of attracting "good" teachers. Residents advanced two reasons for this difficulty: the shortage of accommodation in the town, and being in a different educational authority area from Adelaide (teachers would reportedly have to resign from their South Australian post to take up an appointment in Alice Springs). These, then, are the kinds of problems which the local schools face: not so much the quantity of facilities provided but the problem of attracting good teaching staff. Most of the respondents were unconcerned about the lack of higher education facilities in Alice Springs, although a few individual adults would have liked the addition of some further education facility (table 7.6). Among those who had children, it seemed that there were three main reasons for this: some residents did not envisage that their children would attend university at all; of those whose aspirations for their children extended to a university education, most intended to move out of the town anyway before their children were of the appropriate age; and finally, there were those who would envisage attending a university in Alice Springs only if it were of a good standard. Again, the problem of qualitative rather than quantitative provision is of importance. It is not, of course, impossible to establish a successful higher education institution in a remote area, as the University of New England has admirably shown, but it is clear that more than just the construction of a building would be required to attract young people.

TABLE 7.4. The Relationship between School-enrollment Provision and Selected Socio-economic Indicators

Variable r R Overall F Significance
Population recorded in the census 1976: size of local        
government area -.428 .428 25.84 <.0001
Percentage of households with incomes <=A$6,000 (1976) .420 .542 23.7 <.0001
Retail sales per capita, 1973-1974 .276 .594 20.59 <.0001
Percentage of the population Aboriginal or Islander, 1976 .099 .627 18.09 <.0001
Road distance to the nearest of Sydney or Melbourne - .203 .652 16.53 <.0001
Road distance to the nearest state capital - .219 .661 14.27 <.0001
Percentage of employed females in retailing or finance, 1976 .070 .671 12.78 <.0001
Percentage change in population 1971 - 1976 - .092 .673 11.18 <.0001
Road distance to nearest town of the 118 surveyed - .209 .674 9.89 <.0001
Percentage of households with incomes>=A$15,000 (1976) -.297 .675 8.86 <.0001
Percentage of population over 15 yrs with no qualifications .222 .675 7.98 <.0001

TABLE 7.5. The Relationship between Teacher/Child Ratios and Selected Socio-economic Indicators

Variable r R Overall F Significance
Percentage of employed females in retailing or finance, 1976 - .342 .342 15.24 <.001
Percentage of households with incomes <= A$6,000 (1976) .231 .404 11.12 <.001
Retail sales per capita, 1973- 1974 .062 .434 8.76 <.001
Percentage of population over 15 yrs with no qualifications - .063 .449 7.06 <.001
Road distance to the nearest of Sydney or Melbourne - .115 .454 5.76 <.001
Percentage change in population 1971-1976 .041 459 4 90 <.001
Population recorded in the census, 1976: size of local government area - .141 .465 4.30 <.001
Percentage of households with incomes >= A$15,000 (1976) - .088 .471 3.85 .001
Percentage of the population Aboriginal or Islander, 1976 - .030 .473 3.42 .001
Road distance to nearest state capital - .092 .474 3.08 .002
Road distance to nearest town of the 118 surveyed - .079 .474 2.77 .003

TABLE 7.6. Unfavourable Attitudes towards Educational Facilities in Alice Springs: A Sample of Residents in Gillen

Items Percentage of respondents mentioning item╣
School teaching:  
Poor standard in high school 8
Turnover of teaching staff 5
Difficulty of attracting good teachers  
-because of shortage of accommodation 8
-because of other reasons 5
Lack of discipline in the schools 9
Teachers not conscientious 4
School facilities or equipment 5
Lack of higher education or adult  
education classes 4
Total percentage of respondents  
mentioning an unfavourable item  
in response to open-ended questions 22

1 Some respondents referred to more than one item.

From the evidence, it is not clear that improvement in educational facilities in Australia's arid lands will in itself prevent the rapid population turnover or improve the retention rate at high schools. It is true that reasons given for moving out of Alice Springs did include mention of better opportunities for children, as reported by Burnley in chapter 5 (table 5.20), but generally people intended to move out for reasons other than education-for job changes or advancement, family reasons, or environmental amenity. Many residents are migrating to Alice for employment or financial opportunities and leaving for the same reason.

This study has also questioned the assumption that equal expenditure on education necessarily means equal outcomes. Even if the provision of educational facilities is improved in remote arid zone towns, there may be little improvement in the level of educational attainment in these towns (Jencks' argument, 1972), or little reduction in the inequality evident in the areas. This is not to deny the importance of adequate education facilities but only to point to the importance of other factors in determining the level of attainment and equal outcomes. This study has suggested that social, demographic, cultural, and political factors are probably of more importance than the physical environment or geographic locale per se.

Contents - Previous - Next