Review of Australian Government Health Workforce Programs

4.3 Reform of the ASGC-RA rural classification system

Page last updated: 24 May 2013

Prior to the commencement of this Review, submissions were tendered to the Senate Community Affairs Committee Inquiry into the Factors Affecting the Supply of Health Services and Medical Professionals in Rural Areas which described specific issues relating to use of the Australian Standard Geographic Classification - Remoteness Areas (ASGC-RA) classification system.

The Senate committee concluded that while the ASGC-RA is a useful tool to determine remoteness, better outcomes may be achieved if it were overlaid with other measures rather than as the sole determinant of incentive payments. The committee recommended that the current classification systems used for workforce incentive purposes be replaced with a scheme that takes account of regularly updated geographical, population, workforce, professional and social data to classify areas where recruitment and retention incentives are required.93

Of particular concern amongst stakeholders is the creation of unintended perverse incentives within workforce programs which rely on the ASGC-RA for eligibility, and perceived disadvantages for rural communities which must compete with larger centres within the same classification band to attract and retain health practitioners.

The Department provided advice to the Senate Inquiry that the effectiveness of the ASGC-RA as a basis for workforce incentives would be considered during this Review and there would be opportunities to gather more information about stakeholder concerns at that time. During the consultation phase of the Review, the participation of stakeholders enabled a better understanding of the issues at play, and reform of the current rural classification system has been identified as a major priority.


The ASGC-RA is a geographic classification system that was developed in 2001 by the Australian Bureau of Statistics (ABS), as a statistical geography structure which allows quantitative comparisons between 'city' and 'country' Australia. The purpose of the structure was to classify data from census Collection Districts (CDs) into broad geographical categories, called Remoteness Areas (RAs). The RA categories are defined in terms of ‘remoteness’ - the physical distance of a location from the nearest Urban Centre (access to goods and services) based on population size.94

The Department adopted the ASGC-RA system to support the introduction of scaled incentive programs for GPs under the Rural Health Workforce Strategy.

The use of ASGC-RA replaced the earlier Rural, Remote and Metropolitan Areas (RRMA) and Accessibility/Remoteness Index of Australia (ARIA) classification systems. The use of the ASGC-RA is not unique to DoHA. It is used extensively by Centrelink, the Department of Families, Housing, Community Services and Indigenous Affairs, (FaHCSIA) and the Department of Education, Employment and Workplace Relations.

The Government, with advice from DoHA, considered a number of alternatives before concluding that ASGC-RA was the most efficient mechanism, based on the reasoning that:

  • ASGC-RA was independently updated by the ABS after each population census
  • ASGC-RA could be easily used to monitor the performance of programs
  • Other agencies (such as DHS) are able to build payment systems and update as necessary
  • ASGC-RA presented fewer anomalies than other models

Since the introduction of the rural workforce incentive programs in 2010, concerns have been raised by key stakeholder groups that the classification system, which categorises communities into remoteness areas, is disadvantaging some small rural communities across Australia, in particular those that are within inner and outer regional Australia (RA2–3).

Twenty three communities were identified by DoHA from stakeholder feedback as experiencing particular issues under the ASGC-RA classification system. These communities are either in close proximity to each other or in the same RA classification band but with different population sizes or communities. These communities include locations such as Mt Isa, Cowra, Ballarat, Roxby Downs and Kalgoorlie.

In late 2010, DoHA engaged GISCA, the National Key Centre for Social Applications of Geographic Information Systems at the University of Adelaide, to investigate and provide advice in reference to a number of communities that were classified within the same category as larger, better serviced, rural communities. GISCA staff are recognised as experts in this field.

The review by GISCA allowed DoHA to clarify and test the concerns raised by some stakeholders regarding the effectiveness of ASGC-RA as the basis of funding for rural health programs.

GISCA completed the review in early 2011. The review found that overall the ASGC-RA classification system was functioning reasonably well.

However, the review did identify inconsistencies and anomalies in the ASGC-RA system, particularly where some smaller communities had been unfairly classified in the same category as neighbouring larger communities. Such identified boundary issues are not uncommon to other geographical classification systems.

A number of alternative classification systems are currently in use.

Box 4.9: Alternative classification systems in use

RRMA is based on 1991 Statistical Local Area (SLA) boundaries and 1991 population census data. Therefore, it is no longer an accurate model for determining need due to significant population changes and urban expansion. Following the announcement of the RHWS in the 2009-10 Budget and the corresponding use of ASGC-RA in 2009, 32 targeted rural health programs moved to the ASGC-RA.

In February 2010, a further 28 general departmental programs were identified as still using old remoteness classifications (e.g. RRMA). Since then:

  • five programs have ceased;
  • one program has moved to the ASGC-RA;
  • a decision on six aged care programs was deferred until after the Productivity Commission aged care review that reported in August 2011. The government response in April 2012 noted that ARIA would continue to be the system of choice, at this time, noting that further investigation into other classification systems would be undertaken;
  • a decision has been made that a remoteness classification is not needed for six pharmacy programs;
  • eight programs have not changed – e.g. because budget funding has not been forthcoming; and
  • information has not yet been obtained on the status of two programs.

However, RRMA continues to be used for determining eligibility for a small number of programs, particularly under the Practice Incentives Program. Variants of ARIA are also used by some program areas. The use of multiple classification systems across the portfolio has been raised as a significant concern by stakeholders and needs to be addressed.

An update of RRMA was considered in 2004, but did not proceed apparently due to the large number of areas that would be consequently reclassified as ineligible. Further, while a key concern about RRMA is that it uses 1991 Census data, simply updating RRMA with more recent census data would not remedy methodological flaws that have led to distorted incentives. Major issues included:

RRMA measures distances using a straight-line, compared to ASGC-RA’s use of road distance.

RRMA classifies all capital cities as RRMA 1 (ineligible for incentives), including Darwin.

It should also be noted that RRMA faced technological obsolescence as the ABS has now moved to the Australian Statistical Geography Standard (ASGS). The ASGC boundaries and codes were published for the final time on 14 July 2011. For one year from July 2011 the ASGC and the ASGS operated in tandem. From July 2012, the ASGS is the sole ABS statistical geography. Remoteness Areas, Section of State and Urban Centres and Localities are now part of the ASGS, but are built from Statistical Areas Level 1 (SA1s) rather than Census Collection Districts. The ASGS has ceased using Statistical Local Areas (SLAs) which are essential to the RRMA classification.

[end of box]

As noted above, the ABS will progressively replace the current ASGC with the new Australian Statistical Geography Standard (ASGS) as its geographical framework. The framework will entail a new range of statistical areas.

The smallest geographic unit of the ASGS will be the ‘mesh block’, which comprises around 30-60 households. This is smaller than the current ASGC’s Census Collection District and may therefore reduce the number of anomalies generated by a remoteness classification. It is expected that the new ASGS-RA geography will be available in mid-2013.

The new ABS ASGS geography structure will be more stable over time and better represent the service areas of general practitioners. The improved precision of the ASGS provides an opportunity to develop a reliable, flexible and credible classification system capable of measuring where population need for medical services have not been met.

The ASGS is based on 2011 census data and will therefore ensure the use of remoteness area classifications are based on the latest population statistics.

The transition to the ASGS may therefore remove some of the perceived inconsistencies of the remoteness area classifications produced under the ASGC-RA. It will not however, alter the core methodology of the current ASGC-RA system or resolve the major concerns of stakeholders.

It should also be noted that even if the proposal for a pooled regional incentives scheme were adopted (discussed under Rural Recruitment and Retention strategies) there would still be a need to refine the rural classification system. The need to determine rurality in a consistent way is not limited under the Health Workforce Fund to incentive based programs. Measures such as the AGPT and STP with distribution and educational objectives also depend on a reliable classification scheme to meet these objectives.

Rural Classification System Working Group

The findings of the Review have been informed by a key workshop of major stakeholder groups that was conducted in Canberra on 6 November 2012. This working group aimed to develop common understandings of the issues related to the use of rural classifications and work through possible options for system enhancement. The following organisations participated:

  • Rural Doctors Association of Australia (RDAA)
  • Australian Medical Association (AMA)
  • Royal Australian College of General Practitioners (RACGP)
  • Rural Health Workforce Australia (RHWA)
  • General Practice Registrars Australia (GPRA)
  • National Rural Health Alliance (NRHA)
  • Australian Medicare Local Alliance (AML Alliance)
  • Rural Health Research – Monash University School of Rural Health

There was general consensus across the working group that the current ASGC-RA classification system fails to categorise towns effectively in relation to health workforce requirements. The working group proposed that a classification system should be a structured system with some flexibility to enable it to be adapted for individual program guidelines. The system needs to be one that all stakeholders can understand. There was also strong support for the idea that there should be flexibility at the local level in any new system. The key tenet was a mechanism which permits greater consideration of the workforce situation on the ground, rather than the current approach derived from uniform statistical analysis.

It was broadly agreed by all parties that it is important that a classification system use reliable and up-to-date data, as population growth changes quite rapidly in some areas which affects community health needs. It was noted that the ASGC-RA system can be updated every five years with ABS population census data.

While the ASGC-RA system provides a useful platform as a rural classification measure to determine eligibility for health workforce initiatives, the system needs to become more defined to avoid unintended negative consequences and to allow more targeted program investment.

The working group identified that, should a decision be made to reform the classification system, it would be important that an announcement and transition dates be communicated to stakeholders as soon as possible. The transition period should allow stakeholders adequate time to adjust particularly in respect of reporting requirements.

Reform options

Key principles

As part of the review process, consideration has been given to how the rural classification system can be used effectively by a range of programs, addressing stakeholder concerns about the current system, while being cost-effective and efficient to implement.

There is no ‘natural’ classification which differentiates ‘rural’ and ‘remote’ communities from urban centres. Any ‘rural-urban’ classification used to guide resource allocation must be fit-for-purpose and the best use of a system may vary between initiatives.

The key concept is to have a system that is sufficiently well defined to be used with a degree of flexibility between different programs, which may vary in the purposes for which rurality is measured. Adoption of a more defined system, which better enables recognition of the differences between locations in the same RA band, will enable program eligibility requirements to be designed to be fit for purpose and better understood by participants.

Following the completion of stakeholder consultations a number of commonly agreed key principles have been defined that should be a feature of any new rural classification system for health workforce programs:

  • Objective and evidence-based – meets the needs of programs and recipients
  • Easy to interpret
  • Regularly updated (preferably by an independent source such as the Australian Bureau of Statistics)
  • Is not subject to arbitrary amendments
  • Remains stable over time (i.e. will not alter in accordance with short-term fluctuations in service)
  • Allows for discrimination between large and small towns in less remote areas
  • Maintains the current mechanism of scaling for remoteness to provide greater incentives to communities of highest need
  • Measures both remoteness and rurality, to allow differentiation between locations of a similar size which may vary greatly in accessibility.

While these key principles will be an important aid in the reform process it is further recognised that there is no perfect system to measure rurality and that anomalies will occur regardless of the model that is implemented. These issues will vary based on how any revised classification system is applied to different programs. In particular, boundary issues will be inherent to any system. The appropriate management of these is to minimise ‘within-group’ variance and to maximise ‘between-group’ variance.

In some cases risks can be mitigated by designing revised programs with flexibility at the regional and local levels, nevertheless, the potential challenges and uncertainties of using a rurality defining system need to be recognised before final decisions are reached on implementation of any new model. Consideration must be given to the impact of applying any new system to a wide range of diverse programs which do not necessarily reflect the purpose for which the classification was developed.95

While the ideal reform of the rural classification system would lead to adoption of a single measure of rurality across all programs within DoHA, the logistics, cost and lack of stakeholder support present risks which at present, outweigh the benefits of pursuing this end. From a technical perspective, DHS requires at least 12 months (and a significant financial investment) to incorporate changes of this nature in to their system. Given the number of programs in the portfolio which do not presently use ASGC-RA, transition arrangements including stakeholder education and support would entail additional funding requirements and increased workload for program areas. However, this does not preclude the recommendation that all new programs commencing from 2013 onwards have ASGC-RA as the preferred classification system for measuring rurality.

Discussion at the rural classification system working group meeting centred on the use of the Humphreys’/Monash model.

Box 4.10: Humphreys’/Monash model

The “Monash model” is based on research by Professor John Humphreys and Dr Matthew McGrail of the Monash University School of Rural Health. The model is an attempt to design a new multi-layered series of classification zones to be used for rural incentive initiatives using an evidence-based approach. ASGC-RA remains the basis for the Monash model.

The model comprises geographical data, population data and data from the Medicine in Australia: Balancing Employment and Life Study (MABEL) to form a 13 category system.96 However, it has been refined to six categories for ease of application by the Commonwealth and stakeholders. The categories are:

  • RA1 (usually ineligible for most programs)
  • RA 2–5 and populations greater than 50,000 people
  • RA 2–5 and populations between 15,000 and 49,999 people
  • RA 2–5 and populations between 5,000-14,999 people
  • RA 2–3 and populations between 0-5,000 people
  • RA 4–5 and populations between 0-5,000 people

These categories are based on the principle that it is vital to maximise “between group” differences and minimise “within group” differences in order to fairly measure access to health services and the need for differing incentive levels as part of program design. Analysis has shown that the adoption of a new six-level rurality classification measures is statistically equivalent to the full 13-level classification.97

Australian and international literature indicates that 24 hour on-call/after-hours care is one of the biggest barriers within the primary health care setting. Other professional factors that affect workforce distribution include hours worked, type of procedures, on-call arrangements and ability to have time off. Non-professional factors include spouse support and schooling arrangements. These are known as the six sentinel indicators, which were mapped under MABEL. However, the initial study demonstrated that while geographical remoteness was statistically associated with all six indicators, population size provided a more sensitive measure in directing where recruitment and retention incentives should be provided.98

[end of box]

The “Monash model” has appeal on a number of levels and was generally well supported in discussions at the Rural Classification System Working Group and during other review consultation forums. The use of “sentinel indicators” (listed above) was considered to be the major strength of this model, as opposed to modelling based in Medicare data which does not take into account these factors.99

While MABEL data has driven the development of the “Monash” model it does not need to be used on an ongoing basis. The analysis of the MABEL data led the Monash research team to the position that town size is an excellent proxy for effectively measuring the six “sentinel indicators”.

The Review has identified a number of advantages and disadvantages to the use of this system, as follows:


  • Incorporates statistical measurements linked to health service data as opposed to being merely a blunt geographic tool. Represents a more evidence-based approach to determining the best classification system.
  • Addresses the disadvantage small towns and localities experience under the current RA classification.
  • Provides the ability to link higher incentives to smaller locations with demonstrably greater health needs.


  • Doctors could be perversely incentivised to locate just outside town boundaries where they will receive a higher incentive than if they were based inside the town.
  • Towns that are accessible to and are serviced by larger towns will be treated the same as similar sized towns in more remote locations that experience workforce shortages e.g. Alice Springs, Mt Isa and Port Hedland.
  • Similar size towns will receive the same incentive irrespective of remoteness or access to services. For example, Coffs Harbour, NSW is in the same category as Mt Isa, QLD and Margaret River, WA is in the same category as Newman, WA.

The lack of a defined measurement of remoteness was raised as a significant draw-back with this model during the ASGC-RA Workforce Group discussions.

In general, stakeholders were of the view that remote areas need to continue to have a separate classification in any new system, recognising the particular challenges these areas face due to the tyranny of distance. This is the case even for larger towns in remote areas such as Mt Isa and Alice Springs. Although these areas may have a reasonable core base of health service providers, the overall view was that they still face more difficult challenges in recruitment and retention than less remote towns with similar population levels.

The modified Monash model (proposed system)

As discussed above, there is significant appeal to the Monash model, particularly in its use of evidence to provide a more accurate measure of service needs in different communities. The major concern is the way in which this model appears to discount the impact of remoteness upon communities, with subsequent consequences for health workforce recruitment and retention.

A “modified Monash model” has therefore been developed and is put forward for consideration by Government.

Box 4.11: The modified Monash model

This new model continues to use town size as the key classification determinant, as per the original Monash proposal, but recognises the different health service issues caused by remoteness, leaving the classifications for RAs 4 and 5 unchanged.

The seven categories under the proposal are:

  1. RA1;
  2. RA2 and RA3 with population > 50,000;
  3. RA2 and RA3 with population 15,000 to 50,000;
  4. RA2 and RA3 with population 5,000 to 15,000;
  5. RA2 and RA3 with population < 5,000;
  6. RA4;
  7. RA5.

The model currently applies the proposed incentive structure to towns based on the ABS 2006 Urban Centre/Locality (UCL) classification and the 2006 ASGC remoteness area (RA) classification. The list of towns that fall into each category is based on the 2011 UCL classification and the 2006 RA classification. The 2011 ASGS-RA classification has been released in December 2012 and this should be incorporated into the final model.

There are a number of strengths of the proposed classification, based on a revision of the “Monash” model, including;

The proposed classification addresses the disadvantage small towns in RA2 and RA3 experience relative to larger towns.

By retaining separate categories for RA4 and RA5, the proposed system recognises disadvantages experienced by medium-large towns in remote areas (e.g. Mt Isa, QLD).

The proposed classification addresses anomalies under the current ASGC-RA system (e.g Cherbourg, QLD).

The proposed classification addresses boundary issues along the RA2–3 boundary. Towns of similar size and location that sit on either side of the boundary no longer receive different incentives (e.g. Orange & Forbes, NSW).

The proposed classification introduces more categories in RA2 and RA3, recognising the diversity of towns in RA2–3.

To incorporate town size into the classification system, defined town boundaries will need to be settled on. The ABS UCL classification is the most logical definition to achieve this but there are some challenges with this approach, as discussed below.

[end of box]

Analysis of the proposed model

A key consideration in determining the appropriateness of this proposed classification is how well it targets communities most in need of primary care services.

To illustrate this issue the Review has examined how the proposed revision to the RA system might impact on incentive payments administered under GPRIP, which is the main financial incentive initiative current using the ASGC-RA system. It is recognised that the RA system is used for a range of other government programs, which may not experience some of the issues related to how the RA system impacts on incentive payments.

In the absence of complete data on the availability and usage of health services within communities, 2010-11 Medicare data on GPRIP eligible services were compared with town populations weighted by age specific GP access rates. This was used to produce a relative GP Medicare service level for every town. The following chart illustrates the distribution of relative GP Medicare service levels in each town under the proposed classification.

Figure 4.6: Relative MBS service levels by town, 2010-11100

Figure 4.6: Relative MBS service levels by town, 2010-11 D

Source: Data analysis, Portfolio Strategies Division, Department of Health and Ageing, 2012

Figure 4.6 shows that median GP service levels in 2010-11 were progressively lower as the incentive class increased from 2 to 7. This suggests that the proposed incentive classification appropriately provides a basis through which program eligibility and payment levels can be adjusted to enable higher incentives (or other forms of support) to be directed towards towns with relatively higher need.

There are also weaknesses in this model, as exist in any classification system. The model is inherently more complex than continued use of the current ASGC-RA system and would essentially involve the Health portfolio designing and maintaining its own overlay classification methodology. This has impacts in terms of keeping data up to date (with ABS census updates) as well as consistency with other Commonwealth portfolios.

The key weaknesses in the model that have been identified are as follows:

  • Similar sized towns in RA2–3 are treated the same irrespective of their proximity to larger towns. For example, small towns within a short distance of Ballarat, VIC are treated the same as Charters Towers, QLD on the edge of RA3.
  • Exacerbates existing disparities along the borders between RA1–2 and RA3–4. Small towns on the edge of RA1 (e.g. Kurri Kurri, NSW) and medium-large towns on the edge of RA3 (e.g. Kalgoorlie, WA) would receive considerably less in incentives than nearby towns in RA2 and RA4 respectively.
  • The Government must maintain and update the classification, define towns/catchment areas and establish business rules to implement in the DHS system, which diverts resources otherwise available for programs to benefit health workers.

Different programs may need to use the modified ASGC-RA system in more flexible ways. Rural training programs (where program eligibility rules determine where students can undertake placements) may require a slightly different application of the system in comparison with incentive programs. Depending on how the proposed new system is used, further work and consultation with stakeholders may be required in order to minimise potential anomalies in the system and ensure it is as fair as possible.

For example, if the GPRIP Program is maintained in its existing form issues will arise from the new classification such as:

  • Population cut points, where particular townships are either just under or just over the level required to generate different incentive amounts;
  • Boundary issues that arise because towns and localities along category boundaries receive different incentives to nearby towns;
  • Using town size as a determinant means greater use of the ABS Urban Centre/Locality (UCL) boundaries. UCLs are not designed to capture health service flows or catchment areas and cover only a small portion of the overall landmass of Australia. Rules would have to be established for the treatment of doctors practising outside UCL boundaries.

Creating buffer zones (such as a set radius e.g. 20km around UCL boundaries) or catchment areas (based on the distances people travel to work) may be a logical solution to address issues with UCL boundaries. While there are a number of possible options for designing buffer zones, each of these has limitations and further analysis is required before this type of approach could be applied to program rules.

While it is acknowledged that additional work is required to fine tune the proposed rural classification model, it does appears to offer significant advantages in comparison to continued use of the ASGC-RA classification without refinement. If the Government agrees to implement this model, it should do so in consultation with key stakeholders, to ensure important users understand the new systems and work with Government to resolve the boundary and other issues before the new system is implemented. An implementation working group could be established for this.

This model builds on the evidence base provided by the Monash University research team but maintains the measures of remoteness considered desirable by stakeholders. This proposal would provide a system that can be used with a degree of flexibility within individual programs, allowing greater targeting of different types of government investment. Changes to program delivery models, as discussed elsewhere in this review, may help to mitigate some of the risks. How the system is used by each program becomes the key consideration, rather than the geographic system itself.

This principle of flexibility may go some way to addressing local inequities which prevent successful recruitment and retention of medical practitioners, however, broader lifestyle concerns including expectations of on-call roster participation, educational opportunities for offspring and employment for spouses cannot be overcome merely by increasing the difference in incentives across the RA bands.

Incentives are only a powerful tool insofar as they form part of the total experience of a medical professional within the community.

As highlighted by the RDAA in recent media articles,101 doctors are currently departing Mt Isa (RA4), potentially for other more ‘desirable’ practice locations such as Cairns or Townsville (RA3). While the “modified Monash model” if implemented, would not change the classification of these locations, there are fundamental differences in the locations themselves which would continue to exist regardless of the size of any incentive. Government investments would be best directed in ways which will generate meaningful incentives at the regional level, rather than attempting a catch-all fix for every comparable location across Australia.

Other options for systems to measure rurality

In considering this issue a number of other proposals have been examined which would combine continued use of the ASGC-RA system with other measures designed to allow greater definition of rural areas, as a means of determining need.

Box 4.12: Other options for systems to measure rurality


A proposal has been considered to combine ASGC-RA with a measure of socioeconomic status. The best available measure is the ABS Socioeconomic Indexes for Areas (SEIFA). This proposal would involve a sub-classification with the RA boundary in accordance with relative socioeconomic disadvantage.

Under the proposal, towns would be classified by ASGC-RA and sub-classified by relative levels of socioeconomic disadvantage, e.g.

  • RA1
  • RA1 and relatively high levels of socioeconomic disadvantage
  • RA2
  • RA2 and relatively high levels of socioeconomic disadvantage
  • RA3
  • RA3 and relatively high levels of socioeconomic disadvantage
  • RA4
  • RA4 and relatively high levels of socioeconomic disadvantage
  • RA5
  • RA5 and relatively high levels of socioeconomic disadvantage

The SEIFA classification would target towns most in need and addresses many problem areas. However, the model also presents some problems, including:

  • Large towns generally contain highly disadvantaged and highly advantaged populations making a single SEIFA score meaningless.
  • Current SEIFA scores are based on out-dated 2006 census data. 2011 census-SEIFA data will not be released until late March 2013.
  • DoHA would have to determine what it considers to be a high level of disadvantage. This determination will disadvantage towns that fall below the cut-off point.
  • SEIFA has not previously been used for this purpose. There may be unknown drawbacks to its use.

On balance, the complexity associated with use of the SEIFA system combined with ASGS-RA, and the challenges of implementing it across various programs lead to the recommendation that this approach is not suitable for health workforce programs.

Survey-based data systems

Other proposals recommend combining ASGS-RA and/or town size with a measure of doctor service levels in a town or measures of difficulty recruiting doctors to a town.

These types of models revolve around ongoing surveys and data systems designed to more definitively measure health service needs.

Such a classification has the potential to direct incentives to where they are needed most but presents challenges in implementation and system maintenance.

This type of classification would:

  • be highly volatile as service levels and recruitment activity fluctuates (particularly in small towns);
  • create the potential for perverse incentives for doctors to underservice or for practices to leave vacancies unfilled;
  • be vulnerable to manipulation to the extent medical practices control service levels and recruitment activity; and
  • potentially disadvantage towns with high demand for services relative to towns with less demand.

While this type of approach may be beneficial at the local level, it would be challenging to implement nationally and is not recommended.

[end of box]


Immediately following the Review, further development of the “modified Monash model” classification and data systems will be required. In the medium term, the transition of Health Workforce Programs and potentially other Commonwealth initiatives should be overseen by an Implementation Working Group, the composition of which will be determined in consultation with the initial stakeholder working group convened in Canberra in November 2012.

Use of the reformed classification system

The following are a list of circumstances in which the new rural classification system could be used:

  1. To inform funding allocations to regions under a novel regional incentive model, allowing remoteness and town population related issues to be accommodated within the funding formulas.
  2. To support an alternative to a new incentive model, where GPRIP is retained but incentives are better targeted on the basis of need.
  3. To inform funding allocations and define eligibility requirements under rural training programs such as the RCTS and the STP.
  4. To define eligibility requirements and funding for locum schemes.
  5. As a potential mechanism to allow more targeted investments in scholarships such as RAMUS, with preference given to recipients from smaller, more remote locations or who complete their education or training in such locations.
Recommendation numberRecommendationAffected programsTimeframe
Recommendation 4.20The ASGC-RA system should be substantially adapted to the needs of health workforce programs to more appropriately recognise differing access to health services within broad geographic regions and within communities.

A modification to the “Monash model” is recommended as the approach most likely to provide positive enhancements to current systems. This ‘modified Monash model’ would retain the ability to provide greater definition between locations in the same ASGC-RA bands (RA2 and 3) while recognising the need to allow for remoteness as a key factor (retaining RA4 and 5).

The geographic classification components of the revised system should be based on the Australian Statistical Geography Standard (ASGS), as the ABS will soon replace the use of ASGC with this enhanced system.

Further work on the implementation of this model will be required before it can be used within individual programs. The model is not appropriate for application inflexibly across programs. Each initiative may need to adjust its guidelines to use the revised system in the most effective way.

The Department should commence discussions with stakeholders on a revised model based on the core principles outlined in the Report. This should include discussions across the portfolio around the implications for other program areas and the potential for broader application of the model outside workforce initiatives. An implementation working group should be established.

GPRIP and other incentive programs, rural training programsShort term – further development of the classification model and data systems will be required immediately following this review.

Medium term – health workforce programs, and potentially other Commonwealth initiatives, will need to transition to the enhanced system.

93 Senate Community Affairs Reference Committee, The factors affecting the supply of health services and medical professionals in rural areas, August 2012, Recommendation 8, p. xv

94 Doctor Connect website, Department of Health and Ageing, 2012,

95 Humphreys and McGrail – correspondence with DoHA, December 2012.

96 Presentation to the review of Australian Government Health Workforce Programs, Humphreys et al. 2012

97 Humphreys et. al, “Who should receive recruitment and retention incentives? Improving targeting or rural doctors using medical workforce data”, Australian Journal of Rural Health 20, 2012, pp. 3-10.

98 ibid.

99 ibid.

100 The proportion of towns with GP billing activity appears to be misleadingly low in RA 1 because there are a number of towns in RA 1 (e.g. Otford, NSW) with very little GP activity within the town itself, but whose population can readily access the nearest major city for GP services. Median service level is represented in the chart by the line where the light blue and dark blue boxes meet.

101 Cronin, S. “A right, rural mess: sort out classification”, Medical Observer, National. 22 March 2013