What attracts teachers to rural and remote schools? Incentivising teachers’ employment choices in New South Wales

Abstract

Staffing rural and regional schools remains an intractable problem. This study identifies effective incentives for attracting teachers to difficult-to-staff rural and remote schools in New South Wales (NSW), Australia. Compared to their urban counterparts, students in these schools are disadvantaged by teacher staff shortages, inexperience and attrition. The research investigated the ability for existing incentives of the NSW Department of Education, other education systems and other professions to attract professionals to rural and remote appointments using a discrete choice experiment methodology. The findings identify ways of attracting teachers of differing levels of experience and commitment to work in such areas.

Keywords

rural and remote education teaching employment incentives teaching employment conditions teacher choice discrete choice experiment equity in education

Introduction

Like much of Australia, parts of New South Wales (NSW) are characterised by sparse population and isolation. This can lead to educational disadvantage in several ways, including difficulty in attracting and retaining teaching staff (Hudson & Hudson, 2019; White, 2019). This study set out to identify incentives that are most effective in attracting teachers to rural/remote schools. This paper sets out to answer the questions: (i) which incentives are most attractive to questionnaire respondents? and (ii) which incentives appeal to different sub-groups of teachers? The research evaluated the relative value of existing incentives used by the NSW Department of Education (DoE), other education systems and other professions to attract professionals to rural/remote appointments using the discrete choice experiment (DCE) methodology (Aubusson et al., 2014; Burke et al., 2010).

Literature review

The provision of equitable educational opportunities has been prioritised by education systems in Australia for decades (Roberts & Green, 2013; Sharplin, 2009; Yarrow et al., 1999). Whilst teacher shortages in rural and remote locations occur internationally (Cuervo, 2020; Goldhaber et al., 2020), the situation is more pronounced in Australia than in Canada or New Zealand (Sullivan et al., 2018). Various Australian jurisdictions have established programmes to increase the uptake of teaching in rural/remote locations (e.g. NSW Department of Education, 2021a, 2021b, n.d. a; NSW Government, 2018; Queensland Department of Education, 2018). Nevertheless, the problems persist in Australia (Downes et al., 2021). In their review of the literature in Australia, Downes and Roberts (2018) identify among the foremost of these issues are: difficulty in attracting and retaining staff; teacher inexperience; and teaching out of field (see also Weldon, 2016). Problems are yet more acute in certain subject areas, such as mathematics and science (Lonsdale & Ingvarson, 2003). At the same time, poorer educational outcomes persist for students in rural and regional areas (Centre for Education Statistics and Evaluation, 2020; Smith et al., 2018). Policies and structures can play a critical role in ‘ameliorating or exacerbating rural educational disadvantage’ (Sullivan et al., 2018, p. 1). A key question for policymakers is how to attract teachers to rural/remote locations by recognising ‘alterable and inalterable factors’ (McEwan, 1999, p. 854) associated with such schools: alterable factors are those controllable by the education authorities, and inalterable factors are those determined by a school’s location. Similar questions have also arisen about attracting professionals in other sectors (e.g. healthcare) to such regions in Australia (Buykx et al., 2010) and internationally (Buchan et al., 2013; Mbemba et al., 2013).

The literature not only identifies potential incentives attracting teachers to rural/remote schools, but also offers advice that offerings should be bundled and be flexible to meet the career and personal needs of individuals and their contexts (Haslam McKenzie, 2007; Humphreys et al., 2009; Lehmann et al., 2008), rather than a one size fits all approach (Honda & Vio, 2015). For example, in NSW, incentive bundles can be tailored to suit individual and/or location needs up to a specified total value, according to the position and/or location (NSW Department of Premier and Cabinet, 2010).

Engagement with and support for pre-service teachers in rural environments appears to engender positive attitudes in graduates towards accepting rural appointments (Kline et al., 2013). Such teachers will inevitably fill the void created by retiring baby boomers and others – supporting and nurturing pre-service teachers may pay dividends. Hence, preparing pre-service teachers for the circumstances they will encounter in rural schools appears beneficial. This can occur through core capstone units of study covering topics such as multi-grade teaching, isolation and engagement with parents and the community (Jenkins & Cornish, 2015). Similarly, pre-service teachers undertaking a professional experience (PE) in a rural school develop a greater appreciation of rural contexts and are more likely to teach in rural schools (Masinire 2015). Less logistically challenging activities, such as field trips to rural locations, can provide similar opportunities for teachers to gain first-hand experiences. These experiences can address a teachers’ anxiety about the unknown and develop their confidence to apply their skills in a rural or remote teaching context (Sharplin, 2009; Sharplin et al., 2011).

Teachers typically enter the profession for altruistic reasons (Fray & Gore, 2018) and financial incentives alone are insufficient to attract teachers to rural/remote locations (Honda & Vio, 2015). As Roberts and Downes (2020) point out, ‘incentivising the profession solely through external motivations ignores the character of professional practice: the desire to be able to perform the role one trained and prepared for’ (p. 4). Cuervo (2020) notes that the distributive justice approach of allocating more funds and resources to rural schools has also been insufficient to redress urban/rural inequities, and calls for a commensurate distribution of ‘regard’, or respect for rural communities. Kelly and Fogarty (2015) identify incentives focussing on four aspects of rural teachers’ lives: professional development; the distinctiveness of rural teaching; economic burdens; and social isolation. They also indicate that a teachers’ responsiveness to incentives is determined by their personal characteristics, attitudes and knowledge of rural contexts (see also, Durksen & Klassen, 2018). With respect to offering social incentives, schools may seek support from and partnership with the local community in order to offer socially inclusive experiences for teachers (Haslam McKenzie, 2007). However, social incentives are also noted to be less effective in times of constrained economic and employment markets (Haslam McKenzie, 2007).

Methodology

This study was funded by the NSW Department of Education with the aim of investigating which of its existing suite of incentives were effective in attracting teachers to rural/remote regions and to ascertain whether other incentives, not currently offered, might also prove effective. To identify potentially effective incentives, we examined data gathered by the DoE’s Centre for Education Statistics and Evaluation (CESE). We also reviewed the incentives offered by other Australian education systems and non-education services (e.g. health). Relevant scholarly and professional literature was also scanned to understand incentives for attracting teachers and other professionals to rural/remote regions employed internationally. This review informed the selection of the final list of incentives for the DCE (Burke et al., 2010).

Materials

An online survey was developed that included a DCE task in which teachers evaluated chose between schools with differing incentives, attitudinal measures of a teachers’ commitment to the profession and desire to work in rural/remote regions, as well as a range of measures relating to the teacher and the school at which they currently work. The DCE task included several choice sets, with each set containing four hypothetical rural/remote schools for evaluation (see Figure 1 for an example). Each hypothetical school was described in terms of its remoteness using an existing DoE five-level point classification system. Transfer points range from one transfer point for the least difficult schools to staff, up to eight transfer points for the most remote and difficult-to-staff schools (NSW Department of Education, 2021a, n.d. b) and schools may have 1, 2, 4, 6 or 8 transfer points (see NSW Department of Education, 2021c). Whilst the transfer points generally reflect location differences, some schools attract more points because of their difficulty in staffing.

Figure 1.

Example of a choice task.

As the DCE examined incentives offerings from rural/remote locations, the levels varied in the DCE to describe the hypothetical schools were 2, 4, 6 or 8 points (i.e. four-levels). Respondents were provided accompanying information to indicate that the descriptions encountered in the DCE were reflective of the rural/remoteness of their location rather than reflective of being difficulty to staff for other reasons, such as socioeconomic disadvantage. Hence, comparable to other classification systems this generally equates to schools ranging from urban coastal locations (2-points), non-urban coastal or inner regional (4-points), outer regional (6-points) and remote locations (8-points).

In addition, each school was described by three incentives determined by an experimental design. The incentives shown were drawn from a pool of 30 occurring at a low or high level, resulting in a total of sixty incentive levels (see Table 1).

Table 1.

Final Incentives List.

No	Area	Low level	High level
1	EMP	Guaranteed priority transfer after 4 years of service to school of choice	Guaranteed priority transfer after 2 years of service to school of choice
2	EMP	Guaranteed perm. Appointment to R/R location after 2 yrs R/R teaching	Guaranteed permanent appointment to a R/R location
3	FIN	Rental subsidies of 70%	Rental subsidies of 90%
4	FIN	Up to $100/week living allowance not including rental subsidies	Up to $250/week living allowance not including rental subsidies
5	FIN	Additional salary of $2500/year during service at this school	Additional salary of $5000/year during service at this school
6	PRD	Guaranteed 5 days/yr and $2000/year expenses for PL	Guaranteed 10 days/yr and $4000/year expenses for PL
7	PRD	Guaranteed 5 days/yr and $2000/year expenses for PD for AITSL progress	Guaranteed 10 days/yr and $4000/year for expenses for PD for AITSL
8	TRV	Fly-in, fly-out (FIFO) provided 2 times a year	Fly-in, fly-out (FIFO) provided 4 times a year
9	TRV	Travel/accomm. of $5000/year for own/family medical treatment in city	Unlimited travel/accommodation for own/family medical treatment in city
10	REC	Extra 3 days long service leave per year	Extra 6 days long service leave per year
11	REC	Two days per semester of paid study time	Four days per semester of paid study time
12	LIV	Wear and tear on own car of $5000 for appointment	DoE 4WD vehicle provided for the duration of the appointment
13	LIV	Health insurance to the value of $2400/year including dental/ancillary	Free health insurance to the value of $4800/year, including dental/ancillary
14	LIV	Free home communication package for up to $60/month	Free home communication package for up to $90/month
15	LIV	Own-choice technology for personal use, up to $1000 for every 3 years	Own-choice technology for personal use, up to $3000 for every 3 years
16	LIV	Reimbursement of mobile phone plan for up to $60/month	Reimbursement of mobile phone plan for up to $90/month
17	LIV	$200 fees paid for local club/association of choice	$500 fees paid for local club/association of choice
18	EDR	Reimbursement of 50% remaining tertiary tuition	Reimbursement of all remaining tuition fees for past tertiary education
19	EDR	Accommodation and HECS fees of $8000/year for regional teaching degree	Accommodation/HECS of $16,000/year for regional teaching degree
20	FAM	$5000 value of service from recruitment service for partner	$10,000 service from recruitment service for partner
21	FAM	Priority appointment for partner, where partner is a qualified teacher	Priority appointment for partner at the same school
22	FAM	School fees for children boarding at secondary school, up to $10,000/year	School fees for children boarding at secondary school, up to $5000/year
23	MON	Recruitment benefit of $5000 after Term 2	Recruitment benefit of $5000 after Term 1
24	MON	Visa vouchers valued to $500/year	Visa vouchers valued to $1000/year
25	EDF	Up to $15,000 paid study in priority area in R/R school (e.g. STEM)	Up to $30,000 paid study in priority subject in R/R school (e.g. STEM)
26	EDF	Travel and accommodation costs of $1500 for rural/remote practicum	Travel/accommodation $3000 paid for R/R practicum in last semester
27	EDF	Reimbursement of 50% of any tuition fees for current education	Reimbursement of all tuition fees for current education
28	SPP	Support staff (e.g. teacher’s aides) for 5 days per fortnight	Additional support staff (e.g. teacher’s aides) for 5 days per week
29	SPP	Phone help with R&R mentor	Phone help and twice-yearly visits with R&R mentor
30	SPP	24/7 phone help line (EAP) for 10 additional hours per annum	24/7 phone help line (EAP) for 20 additional hours per annum

EMP = employment guarantees; FI = financial (rental/salary); PRD = professional development; TRV = travel; REC = additional leave; LIV = living costs/subsidies; EDR = education expenses (reimbursement); FAM = family (spouse/children); MON = other monetary/financial; EDF = education expenses (current/ future); SPP = support for teaching.

Upon viewing each choice set, teachers were asked to nominate the school that they would most prefer to teach in among the four options or to nominate to remain at their current school. After nominating their most preferred option, the chosen option was removed, and respondents were asked to indicate their most preferred school from the remaining four options. This process continued until a full ranking of all five schools in each set was obtained. Respondents were then presented a new choice set with four different hypothetical schools for evaluation, completing five choice sets in total.

In addition to the choice task, survey respondents were asked to indicate their level of agreement on a 7-point Likert scale to ten attitudinal statements in order to measure respondents’ commitment to the teaching profession (six items) and their intention to work in rural/remote locations (four items). Data from these questions were analysed using a Confirmatory Factor Analysis (CFA). The factor scores and descriptive statistics for each statement are presented in Table 2. These latent variables both have high reliability (Cronbach Alphas for both ‘commitment’ and ‘intention’ above 0.9) and high model fits (Bagozzi & Yi, 1988).

Table 2.

Commitment to Profession and Intention for Rural and Remote Employment.

Commitment to the profession (AVE = .66; CR = .96)	Mean	S.D.	Factor loading
I feel a sense of belonging to the teaching profession	5.48	1.52	0.71
I feel as if the challenges of the teaching profession are my own*	4.23	1.71
Being a teacher has great deal of personal meaning for me	5.92	1.33	0.87
I enjoy talking about teaching with other people	5.59	1.45	0.83
I feel like part of the teaching profession	5.57	1.49	0.80
I have an emotional attachment to being a teacher	5.73	1.43	0.84

Intention to work in rural and remote locations (AVE = .75; CR = .95)	Mean	S.D.	Factor loading
I am likely to seek a teaching position in a rural or remote area within the next 2 years	3.33	2.16	0.89
I am likely to seek information about a teaching position in a rural or remote area in the future	3.66	2.20	0.94
I am likely to talk positively about a school in a rural or remote area as a good place to teach*	5.23	1.71
I would like to teach in a school in a rural or remote area	4.44	2.04	0.76

* item dropped from final CFA

AVE = Average Variance Extracted; CR = Composite Reliability

Latent scores for the two latent constructs, commitment to the teaching profession and intention to work in a rural/remote location, were calculated for each respondent and used in the prediction of latent class membership as described below.

The final section of the survey asked respondents about a range of socio-demographic characteristics (e.g. age, income and education), along with characteristics of the school at which they currently worked and the school at which they had completed their practicum. In addition, respondents were asked to indicate what would attract them to teach in a rural/remote area on a range of measures, including their level of agreement that a remote/rural location offers greater job security, an appealing lifestyle and is less stressful than working in a ‘big city’.

These socio-demographic characteristics, individual measures regarding the attractiveness of remote locations, as well as the latent scores capturing commitment to the profession and attitudes to work remotely, were used to predict latent class membership. Members of each latent class are predicted to be teachers that hold similar preferences for the incentives in the same class, but are distinct from the preferences for incentives exhibited by other classes, as reflected in their choice of schools in the DCE task. That is, combining the DCE task data with individual level measures provided a choice model to estimate two sets of effects simultaneously: a set of effects to describe teachers’ preferences for incentives and schools that are distinct for each latent class (i.e. segment), and a set of estimates to capture which teacher characteristics are significant in predicting which latent class a teacher is most likely to belong.

Participants

A total of 5911 in-service and pre-service teachers completed the online survey. A description of the sample is presented in Table 3. Among the 5911 respondents, 67.1% held permanent positions. The majority of respondents had teaching duties (83.4%). The largest group was ‘classroom teacher’ (62.8%) and over 95% of respondents had taught in the previous five years. Seventy-two per cent of respondents were females. The mean respondent age was 43 years. Most respondents were teaching in larger towns or cities, with 30.6% teaching in Sydney.

Table 3.

Sample Characteristics.

Sociodemographic	%	Current position	%
Gender (Male)	26.5	Permanently employed	67.1
Age 18–34 years	26.9	Casual or temporary position	29.6
Age 35–54 years	50.1	Pre-service (student teacher)	2.1
Age 55 years or above	21.2	Classroom teacher (non-executive)	62.8
Aboriginal/Torres Strait Islander	2.8	Principal or executive	26
Have a partner	77.4	Current school (location, type, size)
Partner is qualified teacher	25.9	Capital city	30.6
Have dependent children	56.0	Large coastal city/town	17.2
Current homeowners	71.6	Large inland country city/town	14.7
Annual household income (Gross)	AUD126,300	Small coastal town	8.0
Teaching qualification/experience	%	Small inland country town	18.6
Postgraduate degree or equivalent	28.6	Remote coastal location	0.9
Graduate diploma or certificate	17.9	Remote inland country location	8.6
Bachelor degree or equivalent	47.5	Primary	44.9
Advanced diploma or diploma	3.1	Central	6.1
Currently completing	1.3	Secondary	44.8
Teaching experience: <4 years	22.5	<160 students	13.6
Teaching experience: 4–7 years	13.8	160–450 students	29.7
Teaching experience: 8–15 years	23.4	451–700 students	23.7
Teaching experience: 16 + years	40.3	701 + students	32.2

Percentages represent proportion of sample overall. Totals may not add up to 100% in cases where ‘other’ option was provided or respondents chose not to provide this information (<5%).

Analysis

First, using a traditional choice modelling approach the results were considered at the aggregate level, examining preferences among incentives while ignoring individual differences in preferences (i.e. assuming that all teachers held identical preferences for each incentive). Several methods and models have been developed to account for differences in individual preferences (e.g. Burke et al., 2020), but also individual differences in choice variability (e.g. Burke & Reitzig, 2007; Fiebig et al., 2010; Islam et al., 2007; Magidson & Vermunt, 2007; Swait & Louviere, 1993). The current study employed a scale-adjusted latent class choice model (SALCM) first developed by Magidson and Vermunt (2007). Scale-adjusted latent class choice model have been applied in a small, diverse range of other fields, including to forecast the attractiveness of various museum offerings (Burke et al., 2010), programmes to enhance forest biodiversity (Thiene et al., 2012) and pedagogical approaches in education (Burke et al., 2015). In the current study, the SALCM was used to identify latent segments that simultaneously differ: (i) in terms of preferences for a given incentive of interest and (ii) in the variability with which overall employment choices are made. The SALCM was estimated using LatentGold 5.1 (Vermunt & Magidson, 2016).

Results

Valuation of Incentives – discrete choice experiment (DCE) model

The general attractiveness of the 60 incentives, or aggregate level responses, was investigated using a traditional choice modelling approach. The resulting aggregate parameter estimates are shown in the first columns of Table 4. The estimates indicate the change in attractiveness of a hypothetical school that occurs when an incentive is offered relative to when that incentive is not offered. The parameter estimates in Table 4 are sorted in terms of their decreasing contribution to the attractiveness of a hypothetical school, on average, across all teacher respondents.

Table 4.

Aggregate and SALCM Preference Parameter Estimates.

			Incentive and school description	Aggregate	Class 1 (38.1%)	Class 2 (23.1%)	Class 3 (38.8%)
EMP	1	H	Guaranteed priority transfer after 2 years of service to school of choice	2.408 (0.065)**	2.581 (0.097)**	2.695 (0.032)**	2.069 (0.052)**
LIV	12	H	DoE 4WD vehicle provided for the duration of the appointment	1.502 (0.068)**	1.325 (0.103)**	1.632 (0.038)**	1.597 (0.052)**
FIN	3	H	Rental subsidies of 90%	1.401 (0.068)**	1.162 (0.103)**	1.418 (0.038)**	1.623 (0.052)**
PRD	6	H	Guaranteed 10 days/yr and $4000/year expenses for PL	1.267 (0.066)**	1.713 (0.093)**	0.542 (0.043)**	1.260 (0.053)**
FIN	4	H	Up to $250/week living allowance not including rental subsidies	1.257 (0.071)**	1.116 (0.112)**	1.322 (0.036)**	1.357 (0.052)**
REC	10	H	Extra 6 days long service leave per year	1.118 (0.075)**	1.168 (0.121)**	1.060 (0.035)**	1.103 (0.052)**
EMP	1	L	Guaranteed priority transfer after 4 years of service to school of choice	1.078 (0.092)**	0.299 (0.167)	1.712 (0.035)**	1.463 (0.052)**
TRV	8	H	Fly-in, fly-out (FIFO) provided 4 times a year	0.906 (0.075)**	1.140 (0.118)**	0.574 (0.046)**	0.874 (0.051)**
TRV	9	H	Unlimited travel/accommodation for own/family medical treatment in city	0.883 (0.064)**	2.103 (0.089)**	−0.244 (0.041)**	0.359 (0.053)**
FIN	3	L	Rental subsidies of 70%	0.872 (0.072)**	0.269 (0.109)*	1.091 (0.041)**	1.330 (0.053)**
PRD	7	H	Guaranteed 10 days/year and $4000/year for expenses for PD for AITSL	0.838 (0.075)**	0.379 (0.119)**	1.027 (0.039)**	1.173 (0.054)**
EMP	2	H	Guaranteed permanent appointment to a R/R location	0.806 (0.071)**	1.062 (0.109)**	0.420 (0.039)**	0.785 (0.052)**
LIV	13	H	Free health insurance to the value of $4800/year, including dental	0.776 (0.080)**	0.330 (0.134)*	0.978 (0.039)**	1.091 (0.053)**
FIN	5	H	Additional salary of $5000/year during service at this school	0.766 (0.078)**	0.538 (0.130)**	0.597 (0.037)**	1.088 (0.052)**
FIN	5	L	Additional salary of $2500/year during service at this school	0.761 (0.107)**	1.412 (0.205)**	2.320 (0.037)**	1.958 (0.051)**
EMP	2	L	Guaranteed perm appointment to R/R location after 2 yrs R/R teaching	0.645 (0.069)**	1.224 (0.104)**	−0.042 (0.043)	0.485 (0.052)**
TRV	9	L	Travel/accommodation of $5000/year for own/family medical treatment in city	0.524 (0.097)**	−0.271 (0.180)	0.922 (0.037)**	1.064 (0.052)**
SPP	28	H	Additional support staff (e.g. teacher’s aides) for 5 days per week	0.485 (0.064)**	1.257 (0.089)**	−0.237 (0.043)**	0.158 (0.053)**
EDR	18	H	Reimbursement of all remaining tuition fees for past tertiary education	0.420 (0.088)**	0.162 (0.157)	0.701 (0.035)**	0.506 (0.052)**
FIN	4	L	Up to $100/week living allowance not including rental subsidies	0.240 (0.094)**	−0.125 (0.173)	0.750 (0.037)**	0.293 (0.052)**
REC	10	L	Extra 3 days long service leave per year	0.231 (0.073)**	0.050 (0.115)	0.753 (0.040)**	0.098 (0.053)
PRD	6	L	Guaranteed 5 days/year and $2000/year expenses for PL	0.213 (0.074)**	0.113 (0.112)	−0.026 (0.045)	0.451 (0.055)**
SPP	28	L	Support staff (e.g. teacher’s aides) for 5 days per fortnight	0.164 (0.071)**	0.494 (0.108)**	−0.326 (0.040)**	0.131 (0.052)*
LIV	14	L	Free home communication package for up to $60/month	0.137 (0.070)**	0.973 (0.104)**	−0.495 (0.044)**	−0.305 (0.052)**
PRD	7	L	Guaranteed 5 days/year and $2000/year expenses for PD for AITSL	0.077 (0.112)**	−0.577 (0.219)**	0.334 (0.037)**	0.563 (0.052)**
MON	23	L	Recruitment benefit of $5000 after Term 2	0.077 (0.079)**	−0.328 (0.130)*	0.524 (0.038)**	0.208 (0.053)**
FAM	20	H	$10,000 service from recruitment service for partner	0.058 (0.094)**	−0.182 (0.170)	−0.113 (0.040)**	0.393 (0.051)**
LIV	13	L	Health insurance to the value of $2400/year, including dental/ancillary	0.053 (0.088)**	−0.681 (0.154)**	0.922 (0.041)**	0.257 (0.053)**
EDR	18	L	Reimbursement of 50% remaining tertiary tuition	0.030 (0.070)**	0.517 (0.104)**	−0.139 (0.043)**	−0.345 (0.053)**
LIV	15	H	Own-choice technology for personal use, up to $3000 for every 3 years	0.018 (0.080)**	0.001 (0.133)	0.052 (0.041)	0.015 (0.051)
LIV	14	H	Free home communication package for up to $90/month	−0.015 (0.074)**	−0.007 (0.116)	0.095 (0.041)*	−0.088 (0.052)
TRV	8	L	Fly-in, fly-out (FIFO) provided 2 times a year	−0.072 (0.078)**	0.060 (0.126)	0.271 (0.039)**	−0.402 (0.053)**
FAM	21	H	Priority appointment for partner at the same school	−0.113 (0.078)**	0.317 (0.127)*	−0.175 (0.041)**	−0.496 (0.053)**
EDF	25	H	Up to $30K paid study in priority subject in R/R school (e.g. STEM)	−0.122 (0.080)**	−0.574 (0.132)**	0.211 (0.038)**	0.122 (0.053)*
LIV	15	L	Own-choice technology for personal use, up to $1000 for every 3 years	−0.127 (0.071)**	0.452 (0.110)**	−0.183 (0.040)**	−0.658 (0.052)**
EDF	26	H	Travel/accommodation $3000 paid for R/R practicum in last semester	−0.194 (0.082)**	0.161 (0.133)	−0.920 (0.046)**	−0.113 (0.053)*
MON	23	H	Recruitment benefit of $5000 after Term 1	−0.272 (0.098)**	−1.364 (0.178)**	0.786 (0.043)**	0.168 (0.054)**
LIV	16	H	Reimbursement of mobile phone plan for up to $90/month	−0.345 (0.076)**	0.179 (0.118)	−0.829 (0.047)**	−0.569 (0.054)**
EDR	19	H	Accommodation/HECS of $16,000/year for regional teaching degree	−0.378 (0.074)**	0.026 (0.116)	−0.469 (0.041)**	−0.718 (0.054)**
MON	24	H	Visa vouchers valued to $1000/year	−0.431 (0.069)**	0.409 (0.103)**	−0.663 (0.043)**	−1.112 (0.052)**
EDF	26	L	Travel and accommodation costs of $1500 for rural/remote practicum	−0.477 (0.075)**	−0.066 (0.119)	−0.714 (0.042)**	−0.737 (0.053)**
REC	11	H	Four days per semester of paid study time	−0.539 (0.080)**	−0.130 (0.129)	−0.660 (0.046)**	−0.866 (0.053)**
FAM	21	L	Priority appointment for partner, where partner is a qualified teacher	−0.562 (0.079)**	−0.587 (0.128)**	−0.360 (0.042)**	−0.656 (0.054)**
FAM	20	L	$5000 value of service from recruitment service for partner	−0.574 (0.098)**	−0.908 (0.177)**	−0.971 (0.046)**	−0.016 (0.051)
FAM	22	L	School fees for children boarding at secondary school, up to $10,000/year	−0.609 (0.071)**	0.392 (0.104)**	−1.478 (0.047)**	−1.072 (0.053)**
LIV	16	L	Reimbursement of mobile phone plan for up to $60/month	−0.610 (0.087)**	−0.649 (0.152)**	−0.428 (0.041)**	−0.680 (0.052)**
EDF	27	L	Reimbursement of 50% of any tuition fees for current education	−0.633 (0.088)**	−0.036 (0.143)	−1.323 (0.055)**	−0.807 (0.053)**
FAM	22	H	School fees for children boarding at secondary school, up to $5000/year	−0.671 (0.105)**	−0.130 (0.190)	−1.276 (0.055)**	−0.840 (0.052)**
LIV	17	H	$500 fees paid for local club/association of choice	−0.794 (0.075)**	−0.092 (0.116)	−0.648 (0.042)**	−1.564 (0.054)**
EDF	27	H	Reimbursement of all tuition fees for current education	−0.813 (0.089)**	−1.297 (0.152)**	−0.381 (0.046)**	−0.596 (0.053)**
EDF	25	L	Up to $15,000 paid study in priority area in R/R school (e.g. STEM)	−0.820 (0.150)**	−1.549 (0.313)**	−0.676 (0.045)**	−0.195 (0.053)**
EDR	19	L	Accommodation and HECS fees of $8000/year for regional teaching degree	−0.857 (0.089)**	−0.474 (0.151)**	−1.376 (0.048)**	−0.923 (0.052)**
LIV	12	L	Wear and tear on own car of $5000 for appointment	−0.977 (0.160)**	−3.489 (0.344)**	0.495 (0.040)**	0.603 (0.052)**
REC	11	L	Two days per semester of paid study time	−0.995 (0.092)**	−1.223 (0.162)**	−0.363 (0.041)**	−1.144 (0.055)**
MON	24	L	Visa vouchers valued to $500/year	−1.129 (0.143)**	−0.882 (0.287)**	−1.105 (0.053)**	−1.384 (0.054)**
SPP	29	L	Phone help with R&R mentor	−1.305 (0.087)**	−0.592 (0.142)**	−1.996 (0.051)**	−1.592 (0.054)**
SPP	29	H	Phone help and twice yearly visits with R&R mentor	−1.307 (0.090)**	−1.312 (0.158)**	−1.273 (0.042)**	−1.323 (0.054)**
LIV	17	L	$200 fees paid for local club/association of choice	−1.312 (0.084)**	−0.866 (0.138)**	−1.130 (0.048)**	−1.855 (0.053)**
SPP	30	H	24/7 phone help line (EAP) for 20 additional hours per annum	−1.402 (0.089)**	−0.879 (0.151)**	−1.771 (0.049)**	−1.693 (0.053)**
SPP	30	L	24/7 phone help line (EAP) for 10 additional hours per annum	−1.556 (0.087)**	−1.293 (0.145)**	−1.417 (0.045)**	−1.894 (0.053)**
			2-point school	−0.079 (0.371)**	−0.408 (0.724)	−0.039 (0.149)	0.217 (0.157)
			4-point school	−1.404 (0.577)**	−1.327 (0.857)	−1.501 (0.214)**	−1.421 (0.517)**
			8-point school	−1.555 (0.566)**	−1.389 (0.824)	−1.994 (0.212)**	−1.458 (0.524)**
			Status quo (propensity to remain at current school)	.9385 (.090)	.3574 ** (.086)	−1.312** (.086)	−3.442** (.024)
			Scale parameter	1.000 (−)	1.684** (0.335)	1.009 (0.903)	1.000 (−)

L/H=low/high level of incentive. Mean parameter estimate listed with standard error in parentheses. **/* denote significant at .05/.01 level. EMP = employment guarantees; FI = financial (rental/salary); PRD = professional development; TRV = travel; REC = additional leave; LIV = living costs/subsidies; EDR = education expenses (reimbursement); FAM = family (spouse/children); MON = other monetary/financial; EDF = education expenses (current/ future); SPP = support for teaching.

The relative value and statistical significance provide insights into which incentives for rural and remote teaching are most to least valued among teachers. The most valued incentive out of the 60 incentives was that in which teachers were offered a guaranteed priority transfer after 2 years of service at a school to a school of their choice (β=2.408; p < .0001). That is, teachers were significantly more likely to choose to work at a school offering this incentive relative to any other incentive, on average. The least attractive incentive was one in which teachers were offered a 24/7 phone help line for personal use, either with an additional 20 hours (β=−1.4022; p < .0001) or additional 10 hours (β=−1.555; p < .0001) of assistance per year. The negative sign and size of the coefficient indicates that, on average, the incentive was the least preferred relative to all other 60 incentives considered in the study.

Overall, the most attractive incentives that emerged from DCE responses were, in descending order of preference: guaranteed priority transfer after 2 years’ service; AUD5000 additional salary pa; a DoE-provided four-wheel drive vehicle; rental subsidies of 90%; and guaranteed priority transfer after 4 years’ service. Significantly, both the high and low levels of the last factor (i.e. priority transfer) occur in the top five incentives; these two incentives (high/low) concern supporting teachers to leave the rural and remote area, rather than to remain.

The DCE model estimates also revealed the impact that the level of remoteness of a school had on teachers’ employment choices. There was no significant difference between the attractiveness of schools offering employment in terms of whether they were in an urban location (i.e. a 2-point school in the DoE classification) compared to a respondent’s current employment location (β=−.0790; p = .8318). However, teachers were significantly less likely to take up employment offers made by schools that were located in inner regional locations with a 4-point classification (β=−1.404; p < .05) or very remote locations with an 8-point classification (β=−1.555; p < .01). An example of an 8-point school – presented to respondents as background information to the DCE task – is a school located in the fictitious town of Goolangong, a town of 450 people, 795 km from Sydney and 217 km from the nearest town centre of Coonamble. In contrast, an example of a 4-point school is one located in the fictitious town of Dunaden, a town of 4100 people, 270 km from Sydney and 60 km from the large regional town of Bathurst. On average, only a guaranteed priority transfer after 2 years’ employment to a school of a teacher’s choosing would be a sufficient incentive for a teacher to work at a very remote school location (i.e. with an 8-point classification).

Scale adjusted latent class choice models (SALCM)

Three latent classes of respondents were identified. The SALCM model results presented in the latter columns of Table 3 denote the preferences among incentives and attractiveness of a school based on its remoteness (as denoted by its point classification) and against the status quo of a teacher to remain at their current school, among teachers predicted to belong to each of the three latent classes. First, the latent classes can be distinguished by their propensity to remain at their current school in lieu of choosing a possible rural or remote school location. Specifically, those in Class 1 were least likely to change schools, with a significant preference to remain at the current school when compared to other offerings on average (β=.357; p < .001). In contrast, those teachers predicted to hold preferences for incentives and schools consistent with the other two latent classes were more likely to switch schools, particularly for those in Class 3 (β=−3.442; p < .001).

Second, the latent classes can be distinguished by their valuation of the incentives evaluated in the DCE task as attributes of the hypothetical schools. The results in Table 3 indicate that the three latent classes/segments agree upon the incentive that is most attractive (i.e. guaranteed transfer), but preferences for all other incentives varied between the latent classes. For example, members of Class 1 placed significant value on the offering of unlimited travel and accommodation expenses paid for medical treatment in city locations, for themselves and their dependent children (β=2.103; p < .001) more so than Class 3 (β=.359; p < .001) relative to all other incentives on average; on average, Class 2 valued this incentive significantly less so than all other incentives (β=−.244; p < .001).

The earlier DCE modelling identified guaranteed priority transfer after 2 years’ service, AUD5000 additional salary p.a., a DoE-provided four-wheel drive vehicle, rental subsidies of 90% and guaranteed priority transfer after 4 years’ service as being the top five preferred incentives averaged across the entire sample of teachers. The pattern of preference for these five incentives was consistent for respondents in Classes 2 and 3. For Class 1, guaranteed priority transfers was still the most preferred option, but travel and accommodation expenses of AUD5000 per year to seek medical treatment for self and dependents was the second choice. Also attractive to Class 1 teachers was guaranteed travel and accommodation expenses for professional learning activities.

To further aid interpretation, an overall mean value of the incentives based on their categorisation in Table 1 (e.g. the low and high levels of ‘EMP1’ and ‘EMP2’ all relate to incentives offering employment guarantees) were calculated using the mean parameter estimates presented in Table 4. The summary of these results is presented in Table 5, along with a rank ordering across incentive areas at the aggregate level and within each class. The incentives that were most valued at the aggregate level were the four incentive levels relating to employment guarantees, with variations relating to the number of years of service to obtain the guaranteed transfer and whether the transfer is to a school of the teacher’s choosing or a remote location. These incentives were also the top ranked areas for Class 1 respondents and second highest ranked area for Classes 2 and 3 among the 11 categories. Respondents in Classes 2 and 3 both showed their strongest preference, on average, for the six incentives offering financial rewards in the form of rental subsides and salary bonuses (‘FIN3’ through ‘FIN5’).

Table 5.

Mean Valuation and Ranking of Incentives by Category.

Incentivised support for	Aggregate	Class 1	Class 2	Class 3
EMP: Employment guarantees (4)	1.234 [1]	1.292 [1]	1.196 [2]	1.201 [2]
FIN: Financial (including rental subsidies/additional salary) (6)	0.883 [2]	0.729 [3]	1.250 [1]	1.275 [1]
PRD: Profession Development/Learning (4)	0.599 [3]	0.407 [4]	0.469 [3]	0.862 [3]
TRV: Travel (4)	0.560 [4]	0.758 [2]	0.381 [4]	0.474 [4]
REC: Additional leave (4)	−0.046 [5]	−0.034 [6]	0.198 [5]	−0.202 [6]
LIV: Living costs/subsidies (12)	−0.141 [6]	−0.210 [8]	0.038 [6]	−0.180 [5]
EDR: Education expenses (reimbursement) (4)	−0.196 [7]	0.058 [5]	−0.321 [8]	−0.370 [7]
FAM: Family (spouse/children) (6)	−0.412 [8]	−0.183 [7]	−0.729 [10]	−0.448 [9]
MON: Monetary/Financial (other) (4)	−0.439 [9]	−0.541 [10]	−0.115 [7]	−0.530 [10]
EDF: Education expenses (current/future) (6)	−0.510 [10]	−0.560 [11]	−0.634 [9]	−0.388 [8]
SPP: Support (outside/inside classroom) (6)	−0.820 [11]	−0.388 [9]	−1.170 [11]	−1.036 [11]

Note: ( ) indicates number of incentives in category; [ ] indicates ranking of incentive category.

The results can also be considered in terms of the set of incentives that offer relatively less value and appear in the bottom rows of Table 5. The six incentives that focused on offering teachers opportunities to be supported inside or outside their classroom through mentorship or provision of a teachers’ aides (‘SPP28’ through ‘SPP30’) offered the least value to teachers as being attractive enough to choose a position in a rural or remote school relative to other categories of incentives. The results in Table 5 also show that the outcome aggregated for the entire sample is largely driven by the rejection of these incentives by teachers predicted to belong in Classes 2 and 3, as compared to those in Class 1.

The results also show similar differences across the classes in other areas. For example, Class 1 is seen to value incentives that are inclusive of their spouse or children (ranked 7th highest) as compared to those in Class 2 (ranked 10th highest). On the other hand, respondents belonging to Class 2 and 3 instead place relatively greater value on the set of incentives that subsidize living expenses offered for the duration of the appointment, including those relating to a vehicle, health insurance, home technology set-up and mobile communications. Class 3 is also characterised by placing greater value relative to the other two classes (and particularly Class 1) on incentives that support current and future education opportunities, as well as reimbursement of past tertiary fees.

Profile of teacher segments with different incentive preferences

The SALCM model also provides a prediction about which of the three latent classes (or segments) a teacher is most likely to belong to, based on a range of individual characteristics, including socio-demographics, experience, current school location, location of professional experience (PE), as well as psychographic measures reflective of their commitment to the profession and intentions to work remotely. The parameter estimates for these predictions are presented in Table 6, with each indicating the predicted change in probability of belonging to a particular segment for a teacher described by a given characteristic relative to a teacher without the same characteristic. For example, the results show that with a couple with children are significantly more likely to belong to latent Class 1 (β = .213; p < .001) rather than latent Class 3 (β=−.263; p < .001); in contrast, a teacher living in a single person household is more likely to hold similar preferences to other teachers predicted to belong to latent Class 3 (β = .118; p < .001) rather than latent Class 1 (β = −.137; p < .001). Similarly, teachers who are currently teaching in an inland country location – including a remote, small or large town – are significantly more likely to hold preferences for incentives and schools of latent Class 3 as described in Table 2. In contrast, teachers currently working in a capital city or large coastal city/town are significantly more likely to be members of a segment of teachers with preferences described by latent Class 1.

Table 6.

Drivers of Latent Class Membership.

	LC 1 (38%) Experienced urbans		LC 2 (23%): Mid-career coastals		LC 3 (39%): Responsive rural ECTs
Covariate	Est	S.E.	Est	S.E.	Est	S.E.
Intercept	0.322	(0.011)**	−1.983	(0.014)**	0.718	(0.010)**
Annual Household Income (Gross)	0.051	(0.009)**	0.021	(0.010)**	−0.072	(0.006)**
Current college debt	−0.523	(0.012)**	−0.153	(0.016)**	0.676	(0.014)**
Home owner (no mortgage)	0.337	(0.021)**	−0.069	(0.031)*	−0.269	(0.027)**
Home owner (with mortgage)	0.200	(0.014)**	0.169	(0.018)**	−0.369	(0.015)**
Currently renting	−0.500	(0.026)**	−0.062	(0.028)*	0.563	(0.018)**
Partner is qualified teacher	0.088	(0.022)**	−0.019	(0.030)	−0.069	(0.024)**
Partner is employed full-time	0.194	(0.014)**	0.051	(0.019)**	−0.244	(0.015)**
Do not have a partner	−0.175	(0.024)**	−0.062	(0.029)*	0.237	(0.020)**
Couple family with children	0.213	(0.014)**	0.050	(0.019)**	−0.263	(0.015)**
Couple family with no children	0.026	(0.023)	−0.025	(0.030)	−0.001	(0.023)
Single person household	−0.137	(0.029)**	0.019	(0.034)	0.118	(0.025)**
Indigenous	−0.050	(0.011)**	−0.05	(0.014)**	0.100	(0.011)**
Age (years)	0.120	(0.036)**	−0.023	(0.044)	−0.097	(0.033)**
Gender (Male)	−0.081	(0.021)**	0.088	(0.026)**	−0.055	(0.012)**
Location of tertiary studies
Capital city	0.193	(0.015)**	−0.064	(0.020)**	−0.129	(0.015)**
Large coastal city/town	−0.057	(0.020)**	0.067	(0.024)**	−0.010	(0.019)
Small coastal town	−0.038	(0.033)	0.099	(0.038)**	−0.060	(0.031)
Small inland country town	−0.082	(0.023)**	−0.126	(0.029)**	0.208	(0.020)**
Practicum – capital city	0.193	(0.015)**	−0.064	(0.020)**	−0.129	(0.015)**
Practicum – coastal town	−0.036	(0.020)**	0.059	(0.024)**	−0.023	(0.019)
Practicum – inland country town	−0.055	(0.023)**	−0.045	(0.029)**	0.100	(0.020)**
Postgraduate degree	−0.063	(0.020)**	0.025	(0.025)	0.038	(0.019)*
Graduate dip/certif	0.106	(0.024)**	0.038	(0.031)	−0.143	(0.026)**
Bachelor degree	−0.038	(0.015)*	0.019	(0.020)	0.019	(0.015)
Teaching experience (years)	0.013	(0.032)**	−0.001	(0.041)	−0.012	(0.029)**
Permanently employed	0.355	(0.012)**	0.237	(0.016)**	−0.593	(0.014)**
Classroom teacher (non-executive)	−0.205	(0.014)**	0.026	(0.017)	0.178	(0.013)**
Size of school (students enrolled)	0.047	(0.035)	0.020	(0.044)	−0.067	(0.036)
School level (K-12)	0.000	(0.023)	0.045	(0.027)	−0.045	(0.020)*
Current school location
Capital city	0.168	(0.018)**	0.012	(0.024)	−0.180	(0.020)**
Large coastal city/town	0.068	(0.025)**	0.030	(0.032)	−0.098	(0.026)**
Large inland country city/town	−0.082	(0.029)**	−0.007	(0.035)	0.089	(0.026)**
Small coastal town	0.043	(0.036)	0.074	(0.045)	−0.117	(0.041)**
Small inland country town	−0.057	(0.025)*	−0.088	(0.033)**	0.145	(0.023)**
Remote coastal location	−0.001	(0.109)	−0.007	(0.151)	0.008	(0.104)
Remote inland country	−0.138	(0.042)**	−0.007	(0.046)	0.145	(0.032)**
Commitment to profession	0.219	(0.020)**	−0.100	(0.028)**	−0.119	(0.022)**
Intention to work remotely	−0.750	(0.029)**	−0.157	(0.029)**	0.907	(0.017)**
Attractiveness to R/R locations
Greater job security	−0.092	(0.020)**	−0.119	(0.023)**	0.211	(0.015)**
Appealing lifestyle	−0.142	(0.015)**	−0.006	(0.019)	0.148	(0.013)**
Greater community connection	−0.036	(0.018)*	−0.087	(0.022)**	0.123	(0.015)**
Better housing	−0.086	(0.017)**	−0.019	(0.020)	0.104	(0.014)**
Quicker career advancement	−0.117	(0.018)**	0.019	(0.021)	0.098	(0.015)**
Ability to earn more money	−0.067	(0.017)**	0.050	(0.020)*	0.017	(0.015)
Quality of life	−0.104	(0.014)**	0.088	(0.017)**	0.017	(0.013)
Enjoyable social life	0.171	(0.020)**	0.019	(0.024)	−0.189	(0.018)**
Not as stressful	0.189	(0.022)**	−0.044	(0.028)	−0.146	(0.019)**

Mean parameter estimate listed with standard error in parentheses. **/* denote significant at .05/.01 level.

Based on the three latent class membership probabilities calculated for each individual teacher in the sample, around 38% of teachers are predicted to hold preferences for incentives and schools that match those of latent Class 1, whilst 23% and 39% of teachers are predicted to belong to latent Class 2 and 3, respectively.

Taken together, the results presented in Table 6 indicate that members of Class 1 as compared to the sample aggregate profile was more likely to be older, have more teaching experience and currently working in a capital or coastal city location in a permanent and/or executive position. Members of this class have the lowest intention to work remotely, but are also those teachers who are more committed to the profession. We label members of Class 1 as ‘Experienced Urbans’.

Class 3, we label ‘Responsive Rural early career teachers (ECTs)’, with members of this class most likely to consist of teachers who are single, younger, who have been teachers for a shorter period, currently working in schools located in inland locations, but in a non-permanent position. In contrast to teachers in Class 1, members of Class 3 have higher intentions to work remotely (i.e. more responsive to such opportunities), but are less committed to the profession.

Class 2 appears to be medial concerning characteristics such as age, experience, income and current school with respect to transfer points. They are also medial regarding interest in teaching in rural/remote schools. They comprise a higher proportion living in a small coastal town and their tertiary studies were more likely undertaken in a coastal location rather than in a capital city or country town. They are similar to Class 1 in terms of being permanently employed, but more aligned with Class 3 in being classroom teachers. They are more likely to be teaching Years 11 or 12. They are slightly older than Class 3, but younger than those in Class 1, as reflected in their experience and tertiary debt levels. They are a mixture of homeowners and renters. Class 2, we label ‘Mid-career Coastals’.

Discussion

The main purpose of this study was to understand the valuation of incentives that can be made to teachers to make offers of employment for work at a rural and remote location more attractive. Several considerations, however, emerge from the research findings that indicate the valuation of incentives is contingent on a number of factors, including the aspect of teaching in a rural or remote location that the incentive addresses and individual differences relating to the teachers’ background including their opportunity for relocation. In the remaining section, we highlight the results in terms of the valuation of incentives that emerge in terms of whether they improve the experience of teaching or living in a rural and remote location, or offer opportunities for furthering a teacher’s career. The valuation of the incentives is discussed in terms of differences arising from the career and life-stage that an individual teacher identifies with, and related attractiveness to considering employment and life in a rural and remote areas.

Incentives to support teaching experiences are less valued

Across the 30 types of incentives (with a total of 60 high and low incentive combinations in total) considered, several directly related to supporting a teachers’ professional experience in teaching at a rural and remote location. However, the findings indicate teachers were overwhelmingly less attracted to employers offering assistance via face-to-face mentoring or help over the phone, versus other incentives. Teachers were more receptive to additional support staff (e.g. Aboriginal teachers’ aides), although this appeared among the median list of ranked incentives. This contrasts with other work that highlights the value in mentoring, particularly for early-career teachers (e.g. Buchanan et al., 2013), but consistent with other studies indicating mixed experiences and challenges for both mentors and mentees (Hobson et al., 2009).

Incentives to support short-term travel are moderately valued

A second area of incentives referred to facilitating the expense and time associated with travelling to and from rural and remote locations. To provide some perspective, our study examined locations that were over 1000 km from any capital city, often requiring multiple modes of transport to access and complicated by the challenging road conditions and/or climate. Incentives that were most popular in this regard appeared to be the provision of an all-terrain/4WD vehicle during their appointment, provision of time off and travel support to attend professional learning, additional long service leave and unlimited travel for medical treatment in city locations. The tyranny of distance is also valuable to overcome for those accepting relocation away from family and friends (Sharplin, 2002). In addition, associated benefits will also accrue to students of these teachers (Harvey and Clark, 2018; Watson et al., 2016).

Incentives for professional development are valued over supporting further education

As with our participants, the literature highlights the importance of supporting professional development, particularly attendance at activities away from the rural/remote school location, and to other incentives around working conditions such as consideration of class sizes and provision of support staff (e.g. Aubusson et al., 2014; Burke et al., 2015). Less prominent in our findings was importance of providing pre-service teachers with PEs in rural and remote areas (Hudson & Hudson, 2008; Kline et al., 2013). Similarly, programmes to incentivise further study in specific areas, such as STEM, appear to offer relatively less value than programmes than reimburse the costs from previous tertiary studies.

Incentives of financial support (including rental subsidies) are more valued, particularly by Early Career Teachers

A fourth area of incentives more valued by teachers relates to improving conditions and living expenses for the duration of their employment in a rural or remote location, such as the offer to improve home or mobile communications. However, we found that teachers were overwhelmingly more attracted to employment opportunities that offered incentives with direct economic benefits, such as salary supplementation, monetary benefits towards living allowances and heavily subsidised rental arrangements. Related to this issue is that most teachers receptive to rural and remote postings were found to be those most likely to be at the beginning of their careers, have lower household incomes and potentially looking to secure their financial position by working in hard-to-staff locations. The value in offering rental subsidies is further supported by the finding that workers most likely to be incentivised and to take up rural and remote positions did not already own their own home and/or were living in shared households.

Incentives offering opportunities beyond the current placement are most valued

The most significant area of offering incentives for teachers to accept employment relates to conditions regarding the service period and point systems. The results indicate that securing permanency and transfer conditions are significant considerations. Of interest to Experienced Urbans (i.e. members of Class 1) is whether incentives are extended to be inclusive of partners if working in rural/remote locations.

These findings, however, also raise concerns about incentivising rural and remote teaching over the longer term, particularly via the accrual of transfer points and guaranteed placements to a school of the teacher’s choosing, possibly at a non-rural or less remote location. These incentives create a challenge for employing bodies in that the same mechanism that attracts teachers to rural/remote schools, also serves to propel them back to more populous and more popular locations (Reid et al., 2010). This means that rural/remote schools are likely to be staffed by the most inexperienced personnel, including executive staff, and who require considerable support (HardwickFranco, 2018; Fray & Gore, 2018). This further exacerbates the complex and chronic problems experienced within such schools and those encountered by their students, including limited opportunity, and lowering of aspirations and expectations.

Like Honda and Vio (2015), Haslam McKenzie (2007) and other research cited above, we found that while both social and financial incentives are significant motivators to attract teachers to rural/remote schools, each in itself offers limited attraction. Moreover, we can confirm that no single suite of incentives will attract everyone, nor on their own are enough to move teachers from their current positions. This indicates that the bundling and tailoring of incentives is required, rather than presuming that one, or a limited number of approaches will be sufficient (Humphreys et al., 2009; Lehmann et al., 2008; NSW Department of Premier and Cabinet, 2010).

The latent class analysis identified three sets of teacher segments, labelled as Experienced Urbans (latent class 1), Mid-Career Coastals (class 2) and Responsive Rural ECTs (class 3). The identification of these teacher groups provides insights into the challenges presented in attracting teachers that are experienced, committed to the profession, but responsive and capable of doing so given their life stage. For example, Experienced Urbans are more likely to be those teachers who are more established in their careers and ‘life stage’ than either of the other classes, indicated by higher levels of income and home ownership, with partners working full-time, living as a couple/family with children. Whilst this group was, on average, the most committed to the profession, they were less likely to be interested in teaching in a rural/remote school and the majority of incentives offered would not be valued enough for them to select employment options for a position at a rural/remote school. At the other extreme, Responsive Rural ECT teachers are described as those teachers early in their careers, and pre-service teachers, and were entirely responsive to the incentives offered, but teachers who expressed lower commitment to the profession. However, they favoured those incentives and use of the transfer-points that allowed them to further their careers. There was some familiarity with rural and remote regions amongst these teachers, including having spent time on internships and/or in their childhood in these regions. These respondents tended to be single and renting, and therefore potentially more open to relocation. Finally, teachers identified as Mid-Career Coastals are more likely to be early to mid-career teachers sitting between the two other classes in terms of their experience, age, home ownership and life stage. Similarly, Mid-Career Coastals appear medial with respect to their level of attraction to work in rural and remote locations and their commitment to the profession relative to teachers belonging to the other segments.

In turn, the latent segment profiles reveal a significant challenge of incentivising teachers to work in rural and remote regions. Whilst the most experienced and most committed teachers may be more valuable to schools and the students of rural and remote locations, they are the least open to these opportunities, and any range of incentives, even those of greater value, appear insufficient to compensate such teachers to start a new life – including for their families – away from their current urban locations. In contrast, it is the least experienced, but less committed teachers who appear most willing to change.

Conclusions and recommendations

It is apparent in the associated profiles of these clusters, that stages in career and stages in life are interrelated factors in determining a teacher’s propensity to seek opportunities in rural/remote locations. This has strong implications in the strategic targeting of particular segments identified in the data. Those teachers more willing to teach in rural/remote areas are typically less experienced and less committed to the teaching profession than others, and less likely to have secured, or pursued, a permanent position. They also appear to have lower levels of teaching qualifications. These two findings appear to be mutually at odds. It is possible that some of those with high non-teaching credentials are career-changers. Many of these more experienced teachers appear to be ‘settled in the city’, with corresponding family and financial (mortgage) commitments holding them there. Teachers with children in senior high school or at university might be concerned about more limited education opportunities rurally, which raises further questions about perceptions of opportunities in rural versus urban education. Such teachers appear to be difficult to entice to the country. The preponderance of science teachers in this group, in the context of a broader shortage of science teachers, no doubt exacerbates attraction problems.

There emerged a loosely inverse relationship between commitment to the profession and to teaching rurally. It may be that age brings with it a greater commitment to the profession. Conversely, it may be that those with less commitment have by then left the profession, raising the mean level of commitment of those remaining. Alternatively, lower professional commitment might be a cultural attribute of younger people, at any point in time, or a trait of the current rising generation. This has broader implications for the teaching profession.

Experienced Urban teachers (Class 1) appear unlikely to be attracted by the same set of incentives as other teacher segments already living in regional areas and/or those with less experience in the profession. Given the higher level of commitment that Class 1 teachers profess to the profession, it would be desirable to consider and explore other strategies for attracting this cohort. One strategy might be to challenge and change perceptions that this group may harbour about lifestyle in rural and remote regions. Marketing strategies highlighting clean air, extra space, accommodation savings, lower stress, reduced commuting time and greater community connections, might be attractive. Additionally, incentives of a more ideological nature, such as the ability to make a difference in students’ lives, could be promoted.

Several other suggestions arising from the research emerge may require consideration. For example, one suggestion is that teachers who are close to retirement, but living in the main cities, might be enticed to share their work experience by spending an additional few years in rural and remote regions. Such teachers may accept such opportunities as their responsibilities to dependent children diminish. A new position of ‘leading teacher (rural)’ or similar, might be created to attract the teachers currently in executive roles. Such a role might entail some teaching responsibilities, along with a leadership role in teaching and administration, and/or be shared by more than one school (Schuck et al., 2016). Indeed, since the release of the report from Schuck and colleagues, instructional leaders in schools have assumed some of these responsibilities in areas, including literacy and numeracy, which are priority areas of focus from 2022 (NSW Government, 2021). Other suggestions include trial periods in which teachers could spend a term in a rural and remote school to gain familiarity with the experience. Also, more opportunities for pre-service teachers to undertake PE in rural/remote settings could be offered to acquaint teachers with such settings and therefore be more aligned with the experiences of responsive rural ECTs (i.e. class 3) who were found to be more responsive to incentives and more attracted to working in a rural/remote school. We offer the recommendation of rural PE cautiously; however, as decisions to teach in such areas are determined by other factors, such whether an individual has grown up in the country (see Gereluk et al., (2020).

The NSW Department of Education (n.d. a) has launched a policy featuring people, practice, participation and partnerships as key areas (p. 7). This resonates with Cuervo’s (2020) call for both material and socio-cultural means to address the problem. White and Downey (2021) refer to ‘place-attentive strategies’ which focus on connections between place, people and power (p. 12). A focus on place (Corbett & Gereluk, 2020) might benefit from being informed by Indigenous connections to Country. While there are no doubt circumstances peculiar to NSW, we believe that some of these incentives might be applicable to other contexts, and are worthy of trial.

Rural and remote students are disadvantaged compared to their urban counterparts (Sullivan et al., 2018). Such disadvantage is exacerbated by staff shortages, inexperience and attrition. More pressingly, we predict that the condition for rural and remote schools is likely to worsen in the near future, with the imminent retirement of many baby boomers from teaching, as well as problems with recruitment (Baker, 2021). COVID-19 has served to hone teachers’ digital skills, but has also highlighted the existing ‘digital divide’ fracturing in part along rural-urban lines (Kormos & Wisdom, 2021). The pandemic has also raised rural home prices, as part of a remote working awakening, nullifying one former advantage of ‘the bush’.

As noted earlier, many teachers enter the profession for altruistic reasons. As suggested by Rice et al. (2017), while some problems exist in staffing schools in low socio-economic contexts in the city, the problem seems greater still in rural locations. For many teachers, it appears that the prospect of rural teaching presents (forgive the Sydney analogy) a bridge too far, as it is further removed from the support of friends and family, alongside perceptions of more limited educational and other professional (e.g. health) services. Redressing associated disadvantages for teachers and learners, particularly by increasing the supply of the former, should be prioritised as a matter of social justice and equity.

Footnotes

Acknowledgements

We would like to acknowledge Professor Sandy Schuck, who led this project, and our colleagues Prof. Peter Aubusson and Dr. Edward Wei.

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the NSW Department of Education, Australia.

ORCID iDs

Paul F. Burke

John Buchanan

References

Aubusson

Burke

Schuck

Kearney

Frischknecht

(2014). Teachers choosing rich tasks: The moderating impact of technology on student learning, enjoyment and preparation. Educational Researcher, 43(5), 219–229. https://doi.org/10.3102/0013189X14537115

Bagozzi

R. P.

(1988). On the evaluation of structural equation models. The Journal of the Arkansas Medical Society, 16(1), 74–94. https://doi.org/10.1007/BF02723327

Baker

(2021, June 23). NSW schools struggle to find teachers as supply collapses. Sydney Morning Herald. https://www.smh.com.au/national/nsw/nsw-schools-struggle-to-find-teachers-as-supply-collapses-20210621-p582w5.html

Buchanan

Prescott

Schuck

Aubusson

Burke

Louviere

(2013). Teacher retention and attrition: Views of early career teachers. Australian Journal of Teacher Education, 38(3), 124–141. https://doi.org/10.14221/ajte.2013v38n3.9

Buchan

Couper

I. D.

Tangcharoensathien

Thepannya

Jaskiewicz

Perfilieva

Dolea

(2013). Early implementation of WHO recommendations for the retention of health workers in remote and rural areas. Bulletin of the World Health Organization, 91(11), 834–840. https://doi.org/10.2471/BLT.13.119008

Burke

P. F.

Aubusson

P. J.

Schuck

S. R.

Buchanan

J. D.

Prescott

A. E.

(2015). How do early career teachers value different types of support? A scale-adjusted latent class choice model. Teaching and Teacher Education, 47(April), 241–253. https://doi.org/10.1016/j.tate.2015.01.005

Burke

P. F.

Burton

Huybers

Islam

Louviere

J. J.

Wise

(2010). The scale-adjusted latent class model: Application to museum visitation. Tourism Analysis, 15(2), 147–165. https://doi.org/10.3727/108354210X12724863327605

Burke

P. F.

Eckert

Sethi

(2020). A multiattribute benefits-based choice model with multiple mediators: New insights for positioning. Journal of Marketing Research, 57(1), 35–54. https://doi.org/10.1177/0022243719881618

Burke

P. F.

Reitzig

(2007). Measuring patent assessment quality – Analyzing the degree and kind of (in)consistency in patent offices’ decision making. Research Policy, 36(9), 1404–1430. https://doi.org/10.1016/j.respol.2007.06.003

10.

Buykx

Humphreys

Wakerman

Pashen

(2010). Systematic review of effective retention incentives for health workers in rural and remote areas: Towards evidence‐based policy. Australian Journal of Rural Health, 18(3), 102–109. https://doi.org/10.1111/j.1440-1584.2010.01139.x

11.

Centre for Education Statistics and Evaluation (2020). Evaluation of the rural and remote education blueprint – final report. NSW Department of Education. https://education.nsw.gov.au/about-us/educational-data/cese/publications/cese-evaluations/rural-and-remote-education-blueprint-final-report

12.

Corbett

Gereluk

(Eds.), (2020). Rural teacher education: Connecting land and people. Springer.

13.

Cuervo

(2020). A social justice approach to rural school staffing: The need for a politics of distribution and recognition to solve a perennial problem. Journal of Pedagogy, 11(1), 127–146. https://doi.org/10.2478/jped-2020-0007

14.

Downes

Roberts

(2018). Revisiting the schoolhouse: A literature review on staffing rural, remote and isolated schools in Australia 2004–2016. Australian and International Journal of Rural Education, 28(1), 31–54. https://journal.spera.asn.au/index.php/AIJRE/article/view/112

15.

Downes

Roberts

Dean

(2021). Researching the schoolhouse: Rethinking research on the staffing of rural, remote and isolated schools in Australia (2000–2019). Rural Education and Communities Research Group. University of Canberra. https://researchsystem.canberra.edu.au/ws/portalfiles/portal/44551806/Downes_et_al_2021_Researching_the_schoolhouse.pdf

16.

Durksen

T. L.

Klassen

R. M.

(2018). The development of a situational judgement test of personal attributes for quality teaching in rural and remote Australia. Australian Educational Researcher, 45(2), 255–276. https://doi.org/10.1007/s13384-017-0248-5

17.

Fiebig

D. G.

Keane

M. P.

Louviere

Wasi

(2010). The generalized multinomial logit model: Accounting for scale and coefficient heterogeneity. Marketing Science, 29(30), 393–421. https://doi.org/10.1287/mksc.1090.0508

18.

Fray

Gore

(2018). Why people choose teaching: A scoping review of empirical studies, 2007–2016. Teaching and Teacher Education, 75(10), 153–163. https://doi.org/10.1016/j.tate.2018.06.009

19.

Gereluk

Dressler

Eaton

S. E.

Becker

(2020). “Growing our own teachers”: Rural individuals becoming certified teachers. In Corbett

Gereluk

(Eds.), Rural teacher education: Connecting land and people (pp. 137–153). Springer Singapore.

20.

Goldhaber

Strunk

K. O.

Brown

Naito

Wolff

(2020). Teacher staffing challenges in California: Examining the uniqueness of rural school districts. AERA Open, 6(3), 1–16. https://doi.org/10.1177/2332858420951833

21.

Hardwick-Franco

K. G

(2018). Educational leadership is different in the country; What support does the rural principal need?. International Journal of Leadership in Education, 22(3), 301–315. https://doi.org/10.1080/13603124.2018.1450997

22.

Haslam McKenzie

(2007). Attracting and retaining skilled and professional staff in remote locations. Desert Knowledge Cooperative Research Centre. http://www.nintione.com.au/resource/DKCRC-Report-21-Staff-Attraction-and-Retention.pdf

23.

Hobson

A. J.

Ashby

Malderez

Tomlinson

P. D.

(2009). Mentoring beginning teachers: What we know and what we don't. Teaching and Teacher Education, 25(1), 207–216. https://doi.org/10.1016/j.tate.2008.09.001

24.

Honda

Vio

(2015). Incentives for non-physician health professionals to work in the rural and remote areas of Mozambique – a discrete choice experiment for eliciting job preferences. Human Resources for Health, 13(1), 23. https://doi.org/10.1186/s12960–-015-0015-5

25.

Hudson

(2008). Changing preservice teachers' attitudes for teaching in rural schools. Australian Journal of Teacher Education, 33(4), 67–77. https://doi.org/10.14221/ajte.2008v33n4.6

26.

Hudson

S. M.

Hudson

P. B.

(2019). ‘Please help me find teachers for my rural and remote school’: A model for teaching readiness. Australian and International Journal of Rural Education, 29(3), 24–38. https://journal.spera.asn.au/index.php/AIJRE/article/view/233

27.

Humphreys

Wakerman

Kuipers

Wells

Russell

Siegloff

Homer

(2009). Improving workforce retention: Developing an integrated logic model to maximise sustainability of small rural and remote health care services. Australian Primary Health Care Research Institute, Australian National University. https://rsph.anu.edu.au/files/full_report_10797.pdf

28.

Islam

Louviere

J. J.

Burke

P. F.

(2007). Modeling the effects of including/excluding attributes in choice experiments on systematic and random components. International Journal of Research in Marketing, 24(4), 289–300. https://doi.org/10.1016/j.ijresmar.2007.04.002

29.

Jenkins

Cornish

(2015). Preparing pre-service teachers for rural appointments. Australian & International Journal of Rural Education, 25(2), 14–27. https://journal.spera.asn.au/index.php/AIJRE/article/view/13

30.

Kelly

Fogarty

(2015). An integrated approach to attracting and retaining teachers in rural and remote parts of Australia. Journal of Economic and Social Policy, 17(2), 1–19. https://search.informit.org/doi/10.3316/ielapa.279405534432440

31.

Kline

White

Lock

(2013). The rural practicum: Preparing a quality teacher workforce for rural and regional Australia. Journal of Research in Rural Education, 28(3), 1–13. https://jrre.psu.edu/sites/default/files/2019-08/28-3.pdf

32.

Kormos

Wisdom

(2021). Rural schools and the digital divide: Technology in the learning experience. Theory and Practice in Rural Education, 11(1). 25–39. https://doi.org/10.3776/tpre.2021.v11n1p25-39

33.

Lehmann

Dieleman

Martineau

(2008). Staffing remote rural areas in middle-and low-income countries: A literature review of attraction and retention. BMC Health Services Research, 8(1), 19–10. https://doi.org/10.1186/1472-6963-8-19

34.

Lonsdale

Ingvarson

(2003). Initiatives to address teacher shortage. Australian Council for Educational Research. https://research.acer.edu.au/cgi/viewcontent.cgi?article=1003&context=workforce

35.

Magidson

Vermunt

J. K.

(2007). Removing the scale factor confound in multinomial logit choice models to obtain better estimates of preference. In Proceedings of the Sawtooth Software Conference (pp. 139–154). Sawtooth Software. https://sawtoothsoftware.com/resources/technical-papers/conferences/sawtooth-software-conference-2007.

36.

Masinire

(2015). Recruiting and retaining teachers in rural schools in South Africa: Insights from a rural teaching experience programme. Australian & International Journal of Rural Education, 25(1), 2–14. https://journal.spera.asn.au/index.php/AIJRE/article/view/91

37.

Mbemba

Gagnon

M.P.

Paré

Côté

(2013). Interventions for supporting nurse retention in rural and remote areas: An umbrella review. Human Resources for Health, 11(1), 44. https://doi.org/10.1186/1478-4491-11-44

38.

McEwan

P. J.

(1999). Recruitment of rural teachers in developing countries: An economic analysis. Teaching and Teacher Education, 15(8), 849–859. https://doi.org/10.1016/S0742-051X(99)00025-6

39.

NSW Department of Education (2021a). Benefits and incentives. https://education.nsw.gov.au/teach-nsw/find-teaching-jobs/choose-rural/benefits-and-incentives

40.

NSW Department of Education (2021b). Rural experience program guidelines. https://education.nsw.gov.au/about-us/careers-at-education/scholarships-and-programs/rural-experience-program/rural-experience-guidelines

41.

NSW Department of Education (2021c). Your career journey: Teacher promotional positions and transfer. https://education.nsw.gov.au/about-us/careers-at-education/why-work-at-education/your-career-journey#Teacher5

42.

NSW Department of Education (n.d. a). Rural and remote education strategy (2021-2024): Every student, every school, everywhere. https://education.nsw.gov.au/content/dam/main-education/en/home/teaching-and-learning/curriculum/rural-and-distance-education/rural-and-remote-education/rural-and-remote-education-strategy-2021-2024.pdf

43.

NSW Department of Education (n.d.b). Rural and remote public schools. https://teach.nsw.edu.au/exploreteaching/high-demand-rural-locations/rural-and-remote-school-list

44.

NSW Department of Premier and Cabinet (2010). Guidelines for implementation of the rural and remote incentive scheme. https://arp.nsw.gov.au/assets/ars/cd4f98ab0d/Guidelines_for_Implementation_of_the_Rural_and_Remote_Incentive_Scheme.pdf

45.

NSW Government (2018). Rural and remote human resources strategy. https://web.archive.org/web/20190228185347/https://teach.nsw.edu.au/__data/assets/pdf_file/0018/550062/Rural-and-Remote-HR-Strategy_v5.pdf

46.

NSW Government (2021). Literacy and numeracy priorities. https://education.nsw.gov.au/teaching-and-learning/curriculum/literacy-and-numeracy/priorities

47.

Queensland Department of Education (2018). Advancing rural and remote education in Queensland state schools. https://education.qld.gov.au/schools-and-educators/other-education/Documents/advancing-rural-remote-teacher-factsheet.pdf

48.

Reid

Green

Cooper

Hastings

Lock

White

(2010). Regenerating rural social space? Teacher education for rural—regional sustainability. Australian Journal of Education, 54(3), 262–276. https://doi.org/10.1177/000494411005400304

49.

Rice

Richardson

P. W.

Watt

H. M.

(20172020). Hard-to-staff Australian schools: How can we ensure that all Australian students have access to quality teachers?. In Bentley

Savage

G. C.

(Eds), Educating Australia: Challenges for the decade ahead (pp. 279–295). Melbourne University Publishing.Roberts, P. & Downes, N. Incentivising the profession: A national mapping of incentives to attract and retain staff in Australian non-metropolitan communities. Rural Education and Communities Research Group. University of Canberra https://researchsystem.canberra.edu.au/ws/portalfiles/portal/52676939/Incentivising_the_Profession.pdf

50.

Roberts

Downes

(2020). Incentivising the Profession: A national mapping of incentives to attract and retain staff in Australian non-metropolitan communities. Australia: Rural Education and Communities research group, University of Canberra. Centre for Sustainable Communities Monograph Series, 2.

51.

Roberts

Green

(2013). Researching rural places on social justice and rural education. Qualitative Inquiry, 19(10), 765–774. https://doi.org/10.1177/1077800413503795

52.

Schuck

Burke

Buchanan

Aubusson

Wei

Graham

Lusty

(2016). Attracting teachers to rural and remote areas: A report to the NSW Department of education. New South Wales Department of Education.

53.

Sharplin

(2002). Rural retreat or outback hell: expectations of rural and remote teaching. Issues in Educational Research, 12(1), 49–63.

54.

Sharplin

(2009). Getting them out there: A rural education field trip. In Lyons

Choi

J.-Y.

McPhan

(Eds.), Proceedings of International Symposium for Innovation in Rural Education (pp. 215–224). University of New England. http://simerr.une.edu.au/ISFIRE/pages/ISFIRE_proceedings.pdf.

55.

Sharplin

O’Neill

Chapman

(2011). Coping strategies for adaptation to new teacher appointments: Intervention for retention. Teaching and Teacher Education, 27(1), 136–146. https://doi.org/10.1016/j.tate.2010.07.010

56.

Smith

Parr

Muhidin

(2018). Mapping schools’ NAPLAN results: A spatial inequality of school outcomes in Australia. Geographical Research, 57(2), 133–150. https://doi.org/10.1111/1745-5871.12317

57.

Sullivan

McConney

Perry

(2018). A comparison of rural educational disadvantage in Australia, Canada, and New Zealand using OECD’s PISA. SAGE Open, 8(4), 1–12. https://doi.org/10.1177/2158244018805791

58.

Swait

Louviere

(1993). The role of the scale parameter in the estimation and comparison of multinomial logit models. Journal of Marketing Research, 30(3), 305–314. https://doi.org/10.1177/002224379303000303

59.

Thiene

Meyerhoff

De Salvo

(2012). Scale and taste heterogeneity for forest biodiversity: models of serial nonparticipation and their effects. Journal of Forest Economics, 18(4), 355–369. https://doi.org/10.1016/j.jfe.2012.06.005

60.

Vermunt

J. K.

Magidson

(2016). Technical guide for Latent GOLD 5.1: Basic, advanced, and syntax. Statistical Innovations Inc. https://www.statisticalinnovations.com/wp-content/uploads/LGtechnical.pdf.

61.

Watson

S. L.

Watson

W. R.

Richardson

Loizzo

(2016). Instructor’s use of social presence, teaching presence, and attitudinal dissonance: A case study of an attitudinal change MOOC. The International Review of Research in Open and Distributed Learning, 17(3), 54–74. https://doi.org/10.19173/irrodl.v17i3.2379

62.

Weldon

P. R.

(2016). Out-of-field teaching in Australian secondary schools. Australian Council for Educational Research https://research.acer.edu.au/cgi/viewcontent.cgi?article=1005&context=policyinsights

63.

White

(2019). Recruiting, retaining and supporting early career teachers for rural schools. In Sullivan

Johnson

Simons

(Eds.), Attracting and Keeping the Best Teachers (pp. 143–159). Springer Singapore. https://doi.org/10.1007/978-981-13-8621-3_8

64.

White

Downey

(2021). International trends and patterns in innovation in rural education. In White

Downey

(Eds.), Rural education across the world: Models of innovative practice and impact (pp. 3–21). Springer Singapore. https://doi.org/10.1007/978-981-33-6116-4_1

65.

Yarrow

Ballantyne

Hansford

Herschell

Millwater

(1999). Teaching in rural and remote schools: A literature review. Teaching and Teacher Education, 15(1), 1–13. https://doi.org/10.1016/S0742-051X(98)00036-5