Introduction

Insufficient physical activity (PA) is the fourth leading risk for mortality worldwide [1] and a major risk factor for overweight and obesity, type 2 diabetes, coronary heart disease, certain cancers and mental ill-health [2]. Despite this, 70% of Australians aged 15 years and over in 2004/05 were classified as either being ‘sedentary’ or having ‘low exercise levels’ (considering exercise frequency, intensity and duration) [3]. Recent research has demonstrated an association between PA levels and environmental factors such as urban density, land use and access to public transport and recreational facilities, including bicycle and foot paths [4], [5]. Policy interventions are essential in addressing inadequate PA levels, as strategies such as education and treatment programs alone are unlikely to result in significant sustained changes in PA [6], [7]. Consistent with a socio-ecological model for health promotion [8], multi-level and multi-sector approaches addressing urban planning, transportation, sports/recreation and settings such as schools and workplaces are required to address obesogenic environments and inadequate PA levels [9].

Understanding the policy setting is essential for the development of coherent and feasible regulatory approaches for creating environments more conducive to PA. The policy literature provides a number of theories and frameworks for explaining and predicting policy processes and outcomes. Two policy science factors were considered to be of particular relevance to this study: governance systems and the influence of stakeholders in the policy-making process. Australia has a federal governance system, comprising a federal parliament, eight state/territory parliaments and over 500 local councils [10]. This structure provides the basis for managing policy and program implementation [11]. The varying influence of stakeholders (both individually and together as networks) and the interactions between competing interests have a major influence on the policy process and outcomes [12]. For example, business interest groups typically have greater influence over government decision-making than other stakeholders due to their contribution to the economy [13].

There has been little empirical research analysing the perceptions of those involved in government policy-making in relation to: 1) promising policy interventions for creating environments more conducive to PA, or 2) barriers and facilitators to regulatory change in this area. Greater understanding and clarity of these factors is an important step in identifying which regulatory options are more likely to be enacted. This paper is part of a wider study investigating policy relating to food and physical activity environments, as well as overarching approaches. We have previously investigated this at the local government level in Australia [14], [15] and at the state/territory level in relation to food environments [16]. The current paper examines findings relating to the PA-policy environment at the Australian state/territory government level. The paper is informed by the following research questions:

1. What do those involved in policy-making at the Australian state government level consider to be promising regulatory interventions for creating environments more conducive to PA?
2. Which of the proposed interventions are most or least supported?
3. What are the perceived barriers and facilitators to implementing the interventions?

Methods

Recruitment and participants

A purposive sampling technique was used to ensure reach across: 1) the diverse range of sectors that influence the nutrition/PA environments either directly or indirectly, and 2) all Australian states/territories. Participants were selected by the research team on the basis of having expertise in the development of government policy or expertise relating to food/nutrition or PA. Potential participants were identified from several sources: 1) the study investigators who had considerable experience in the Australian policy arena; 2) high-level officers from government and non-government organizations (NGOs); 3) study participants, who were asked to identify peers in other sectors/states; and finally 4) senior government officers who were asked to nominate a suitable participant within their sector. In total, 45 interviews were conducted with 47 senior representatives (all non-elected bureaucrats, for example, directors, managers or senior policy officers) from government departments, relevant NGOs and statutory authorities (see [16]). Participants were judged by the investigators or the referring people to have extensive policy knowledge related to PA and/or nutrition. A written invitation was followed by a telephone call; with an 84% response rate. For the purposes of this paper, five interviews were excluded from the analysis, as PA was not addressed. In total, data from 40 interviews with 42 participants were included in the analysis. The 42 participants were from nine policy areas (see Table 1). Ethics approval was granted by Deakin (EC 232–2007) and Monash (2007–00–2150) Universities. Participants provided written informed consent prior to participation.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Summary of interviews by sector and by state/territory jurisdiction.

https://doi.org/10.1371/journal.pone.0042831.t001

Data collection and management

In-depth interviews (mean length 61 minutes) were conducted between April 2009 and June 2010. All interviews were on a one-on-one basis, excluding three which were with two participants. All interviews, excluding one, were conducted by the first author (JS). Interviews were conducted in-person for participants in Victoria and via telephone for inter-state participants. Participants did not have access to the questions prior to the interview. All interviews were digitally recorded excluding one where the participant did not consent to recording. Interviews were transcribed verbatim and managed in N-Vivo 8 (QSR International, Melbourne). The interviews comprised three components:

1. A semi-structured component to elicit participants' suggestions for regulatory interventions to create environments conducive to PA;
2. A structured component to identify participants' support for eleven preselected interventions relating to walkability/cycling, public transport, facilities for PA and opportunities for PA (see Table 2). In order to maximise the opportunity for discussion that was not limited to the preselected interventions, participants were questioned about their ideas prior to being presented with the list of preselected interventions. These eleven interventions were based on an extensive literature review of interventions to create environments more conducive to PA and prevent obesity. Interventions were included if they could be reflected in regulatory form and if they were within state government powers. To ensure informed discussion, participants were only questioned about interventions applicable to their department or area of professional expertise (e.g., planning department representatives commented on interventions relating to urban planning).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Participants' responses to the preselected interventions.

https://doi.org/10.1371/journal.pone.0042831.t002

1. A semi-structured component was again used to identify participants' perceptions of the barriers, facilitators and feasibility relating to the interventions discussed.

In this project, regulatory intervention was defined as any intervention by government that was either legislated or enforced through some other means, for example, conditions attached to funding.

Results

Discussion

This study aimed to identify potential regulatory interventions for creating environments more conducive to PA, and barriers/facilitators to their implementation at the state/territory government level in Australia. The regulatory interventions frequently suggested by participants fell into two main areas: urban planning and provision of infrastructure to promote active-travel, and discouraging the use of private motorised vehicles. In response to the set of preselected interventions for which the states/territories had authority, interventions promoting active transport over private vehicle use and facilities for PA generally received the strongest support, for example, mandating that school PA facilities are available to the public outside of school hours. The least supported interventions included subsiding cost of public transport, subsiding cost of bicycle-related equipment and mandating minimal public transport infrastructure to reduce public transport ‘deserts’. These were perceived by participants as not economically viable or unlikely to be effective in increasing PA levels.

Only one of the interventions frequently suggested by participants to create healthier PA-environments, also appeared on the list of interventions preselected by the research team. This was: prioritising active transport over motorised vehicles. This agreement between participants' ideas and a preselected intervention suggests that the ‘prioritisation of active transport over motorised vehicles’ is an especially promising intervention. The lack of concordance between participants and the research team on other interventions may reflect the heterogeneity of the states/territories of Australia in terms of policy need, or alternatively, suggests that there is no clear agreement on the best policy options to address inadequate PA levels. The main barriers to implementing policy to create healthier PA environments were: the power of the ‘road lobby’, weaknesses in the planning system and the cost of potential interventions; whereas the need for evidence, collaboration across sectors and synergies with climate change/environmental agendas emerged as facilitators. Douglas et al [19] suggest that rising car use reflects and reinforces the physical and social environments that have been created. They suggest that cars are the “new tobacco” and note similarities in the tobacco and car lobbies in terms of car/tobacco use being seen as an individual and aspirational choice. Douglas et al. suggest that well-coordinated advocacy in order to build public awareness and engage with policy makers over time is essential to counteract the car lobby and thus increase active transport and reduce car use. Furthermore, policy makers in Canada believed that systemic factors were the main barriers to creating a healthy built environment for youth, specifically economic (governments having mandates but insufficient resources); communities inheriting built environments designed to facilitate car use; governance (distinct and competing mandates); and cultural factors (aspirations and acceptability and safety of active transport) [20].

The results from this present study show some similarities with the findings of a series of studies conducted in nine European countries which analysed stakeholder opinions (including representatives from government, NGOs, food/fitness industry and public health specialists) on policy options for addressing obesity [21]. Participants were asked to compare the performance of several policy options. Cross-national results showed that of the options relating to PA environments, stakeholders scored ‘improving communal sporting facilities’ most highly, whereas ‘changing planning and transport policies’ was less popular [21]. The less popular options were considered costly, with very long timeframes for implementation [21]. Our participants also cited cost as a barrier; however, in contrast to the European study, our participants commonly suggested planning and transport policy initiatives to promote PA. This may reflect different local environmental contexts, for example, European countries typically have higher density cities with greater mixed use developments and more active transportation. The differences in findings highlight the importance of conducting analyses at the local level.

This study highlights the pivotal role of the planning and transport sectors in implementing policy initiatives to create environments more conducive to PA. In general, most of the interventions suggested and supported by participants fell within the jurisdiction of the planning and transport sectors. Participants cited two issues in the planning system as barriers to achieving PA-promoting policy: that planning schemes do not explicitly address health outcomes and that planning schemes are open to interpretation by the local government. In Australia, state and local governments share responsibility for planning schemes between them. Although the exact delineation of powers varies among the states, state governments set overall planning policy and general planning standards, with local governments responsible for more specific policy and the consideration of individual planning applications (see for example the Victorian Planning and Environment Act 1987 [22]). Consequently, many of the interventions suggested and supported by state government participants will depend upon local governments for implementation, making their feasibility dependent upon both levels of government. The sharing of roles makes the creation of well-implemented planning policy difficult, for example, even if the state guiding policy is health supportive (or at least neutral in being neither supportive nor damaging) there is sufficient scope for councils to prefer developments that do not promote health. Equally, there is the capacity for local government to implement policy initiatives to promote healthier PA environments. For example, in Victoria Australia, the Port Phillip City Council's Municipal Strategic Statement states that it will ‘create an integrated and sustainable transport network which supports the use of public transport, cycling and walking above private car travel’ [23]. The devolution of responsibilities for planning in Australia (whereby state and local governments share responsibility for planning schemes, rather than a centralized ‘top-down’ approach), highlights the necessity for a whole-of-government approach in creating environments more conducive to PA. Potential for inconsistent approaches between state and local governments suggests the need for more clearly defined roles and responsibilities.

The creation of a built environment which is more conducive to PA was highly supported by participants. In particular, participants thought that planning controls should be strengthened to encourage development around existing infrastructure (for example, public transport) and to discourage continued urban sprawl via increased housing density. The importance of changes of this type is well-documented in the literature and has been incorporated in some jurisdictions with varying success. In Victoria, for example, the Melbourne 2030 plan (released in 2002) specified certain areas of Melbourne as ‘Activity Centres’ – areas where there was existing infrastructure and development should be encouraged [24]. In Melbourne as elsewhere, these policies are some of the obesity prevention-related policies with the highest level of public resistance. Local communities have strongly resisted medium to high-density developments, despite the available infrastructure. As a result, local governments have been reticent to promote development of the sort recommended by state government policy and supported by this study's participants.

The regulatory picture is further complicated by the government's funding of transport. In 2010–11 the federal government provided $935.3 million in funding for local roads [25]. This funding was complemented by funding for the National Transport Plan which amounts to $36.2 billion over 6 years with $9 billion in 2011–12 [26]. Both road and rail projects are funded under the National Transport Plan, although rail projects are focused in rural and regional areas and not in the areas of urban sprawl identified as ‘public transport deserts’ by participants. Given that potential PA-enabling policy levers, including funding, sit across all three governance tiers, the importance of vertical coordination between tiers of governance is highlighted. In accordance with this, our findings at the local government level found that funding provided from the state and federal tiers of government influenced policy enactment at the local level [14]. In this study, participants at state government level frequently linked PA regulatory actions to the local government level, however there was less frequent discussion of linking up to the federal level. This is in contrast to the food component of this study, whereby state level participants frequently made reference to the federal government but less frequent reference to the local government [16].

Participants believed that collaboration across sectors was required to facilitate PA-promoting policies. The imperative for horizontal coordination across government is clear given that many of the PA-promoting policy levers sit within the jurisdiction of transport and planning. Furthermore, it is clearly in the interest of other sectors to increase population rates of PA, for example, the transport and environmental sectors will benefit from reduced traffic congestion [27] and carbon emissions [28], [29] associated with less private vehicle use, respectively. These sectors could contribute to a whole-of-government policy approach by investing in education and research interventions to promote PA and investigate the health and environmental benefits that would accrue. Many governments recognise the importance of ‘whole-of-government’ or ‘joined-up approaches’ for addressing preventive health issues including obesity [30], [31], however few successful examples exist. A promising model however, is that of the Health-in-All policies approach introduced by the South Australian state government in 2007 [32]. This approach uses governance and accountability structures and a ‘health lens’ analysis process as mechanisms to achieve improved population health outcomes [32]. Early evaluations indicate successful engagement of other sectors in achieving health outcomes [33].

The varying influence of different stakeholders and the interactions between competing interests is widely recognised in policy-making theory and practice [34]–[36]. Participants in our study cited the ‘road lobby’ as a barrier to implementing policy in relation to active transport initiatives and de-incentivising private vehicle use. The constituents of the ‘road lobby’ were not clearly defined by participants. Indeed, beyond government, its associated statutory agencies and state automobile associations, it is hard to identify who this lobby is. It is likely that government itself is its own most significant ‘lobby’ in favour of roads, which are considered essential for national and regional economic growth. For example, in Victoria, Australia, the freight and logistics industry alone contributes 14.7% to Gross State Product [37], with congestion estimated to be costing the economy between $1.3–2.6 billion/year [38]. Given the Australian government's commitment to growing business productivity [39], it could be assumed that protecting and building the roads network is associated with economic productivity. In order to achieve a population shift to more active forms of transport it is likely that the conventional approach to transport planning will need to change, that is, prioritising cycling/walking over motorised vehicles, as is evident in several Northern European cities.

Participants perceived the potential cost of interventions to be a barrier to change in this study, with several interventions being unsupported by participants on the grounds that they were not economically viable, e.g., provision of public transport infrastructure requirements to reduce public transport ‘deserts’ or interventions which required retrospectively changing the composition of the city landscape (e.g., establishing walking paths or areas of open space). Financial considerations were also one of several barriers identified by Canadian policy makers when examining barriers to optimising investments in the built environment [20]. Australian cities have low population densities and urban sprawl, therefore increasing car-dependency and making public transport and infrastructure provision challenging and costly. It is interesting that participants perceived greater public investment in public transport as economically unviable; whereas, the conventional mindset in Australia accepts enormous investment in roads. This suggests that active transport appears to be less of a priority than motorised vehicles and roads. In order to achieve increased population levels of PA, it is likely that a better balance between health and environmental agendas and economic agendas needs to occur. These findings suggest that it will be important for the public health community to engage in more detailed discussions with government, particularly regarding the downstream health costs associated with not improving our PA environment.

One of the difficulties with implementing regulatory interventions to create environments more conducive to PA is that policy options require serious long-term investment both politically and financially. This makes policy uptake challenging. One of the challenges is the impact of policy cycles on long-term planning and investment, where government is unlikely to see the benefits of policy implementation within their time in office. For example, for many of the potential PA-promoting policy solutions, it is unlikely there will be immediate measurable effects, such as those within urban and transport planning which require changes in the city landscape and shifts in social norms. These findings highlight the importance of having greater dialogue between the public health community and government regarding the most optimal way to frame existing and potential evidence for such long-term issues involving environmental change.

In addition to collaboration across sectors, participants in our study spoke about a further two strategies to improve uptake PA-related policy. One solution suggested by participants to facilitate the implementation of PA-promoting policy was maximising opportunities associated with climate change/environmental agendas. Several participants spoke specifically about the synergies between PA-promoting policy and climate mitigation policy, highlighting the importance of recognising synergies or co-benefits in policy-making and using these synergies to leverage support with other sectors. The potential for policy co-benefits provides an additional rationale for policy-makers. PA appears to receive less attention than food in the obesity debate [40], [41]; a notion also discussed by some participants in this study. Given the current political focus on climate mitigation, climate change and environmental policy offers an excellent opportunity for the public health community to collaborate across sectors to strengthen advocacy for the uptake of policies to create environments more conducive to PA.

Another solution offered by participants to facilitate the implementation of PA-promoting policy was the generation of evidence. The generation of evidence is a challenge in relation to policy interventions for creating environments more conducive to PA, as many of the solutions (suggested by our participants and others), such as introducing congestion zones and the provision of showers/lockers in workplaces, have indirect associations with PA, that is, potential effect on PA is mediated via other factors. Showing an association between policies and positive behavior change such as increased PA levels is difficult. Furthermore, evidence is however, only one factor in the complex policy-making process [34], [42]. Evidence is used by government in different ways [43] depending on the ideologies of those who influence policy [44]. As Russell et al. highlight, ‘the selection, evaluation and implementation of research evidence are important in the policy-making process, [but] they do not equate to that process’ [45]. It is also important that evidence-producers make evidence accessible (timely, appropriate language) to policy developers [46].

The strengths of this study include the involvement of those involved directly in the policy-making process, and the breadth of stakeholders interviewed. To date, this project has investigated the local and state government tiers; once the federal government data collection phase is complete, further analysis will be undertaken into how policies and actions between governance tiers could be more closely aligned. One of the limitations of the study was pre-selected interventions presented to participants (prioritising active transport over motor vehicles) lacked specificity and may have been better classified as a policy direction rather than specific approach to create environments more conducive to PA. While this study was undertaken in the Australian context, we believe that some of the findings are more widely transferable as several factors, such as car-dependency, are common in other countries. However, uniquely local characteristics which may affect the feasibility of policy options, such as Australia's low population-density and high urban sprawl, may limit the transferability of our findings. As this research is exploratory in nature, the findings are intended to highlight areas for further analysis, and indicate potential barriers/facilitators for policy change – these will need to be explored in different contexts to assess their relevance.

In order to develop coherent and feasible policy approaches, it is necessary to understand the barriers and facilitators to policy change. This study gives an ‘inside view’ from those involved in the policy-making process at the state/territory level of government in Australia. Similar to our findings in the food component of this study (see [16]), the implementation of many of the interventions suggested by state government participants, are largely the responsibility of the federal government and local governments. This finding is not surprising with respect to urban planning in particular, given the shared responsibility between state and local governments as discussed previously.

Acknowledgments

We would like to thank the participants of this study for providing us with valuable insight.