Abstract
This article reports results from a survey of academic social scientists in Australian universities on reported levels of research utilisation by non-academic users. Using the scale of research utilisation we examine levels of research impact and explore a range of variables to understand factors influencing the use of academic social science research by policy-makers and practitioners. The results indicate that research uptake is enhanced through mechanisms that improve research transfer and the intensity of interactions between academic researchers and users. Our findings provide insights into how the impact of social science research can be improved and draws attention to factors that need to be considered in efforts to evaluate and enhance the impact of academic social research.
There has been increasing pressure on universities to demonstrate the impact of the research they produce. This has involved demonstrating impact through quality measures such as citation counts and journal impact factors. National examples of these assessment schemes include the Research Excellence Framework (REF) in the UK, the Excellence in Research for Australia initiative, the Performance Based Research Fund in New Zealand and other similar initiatives operating in parts of Europe and the US (Geuna and Martin, 2003; Lewis and Ross, 2011). Recently, university research assessment exercises such as the REF have expanded to encompass broader measures of research impact to include the economic, social, cultural and environmental benefits arising from university research.
When it comes to the social sciences, the issue of measuring research impact has been regarded as especially challenging, with many analysts claiming that such assessment needs to involve both quantitative and qualitative indicators (Donovan, 2011; Smith et al., 2011). Pinpointing the policy and practice impacts of academic social research is difficult, because the uptake of social research knowledge rarely follows a linear path from academic knowledge producers to end-users 1 in the fields of policy and practice (Molas-Gallart et al., 2000; Weiss, 1980). Greater collaboration between academic research producers and the users or consumers of research is increasingly seen as a way of improving this knowledge flow, but even in such collaborative contexts an improved research uptake is not guaranteed. The conclusion from most studies is that closer collaboration is insufficient to ensure that academic social research has a discernible impact on policy or practice, and that a range of variables influences the uptake and use of academic social research by non-academic end-users (Belkhodja et al., 2007; Bogenschneider and Corbett, 2010; Cherney and McGee, 2011; Huberman, 1990; Landry et al., 2001a; 2001b; Oh and Rich, 1996; Weiss and Bucuvalas, 1980).
The consideration of social research impact, and the benefits of closer synergies between academic knowledge producers and end-users in generating research uptake, raises important issues concerning the value and orientation of social research and its contribution to evidence-based policy (Head, 2010). If one takes the position that social science research offers tangible social benefits (beyond the pursuit of knowledge as a value in its own right), identifying specific factors that influence research utilisation becomes a key concern. In this article we examine these issues from the perspective of social science knowledge producers – academic researchers in the social science disciplines. While such a study has some limitations (which will be addressed in the discussion), a full understanding of the process of knowledge transfer and uptake requires an examination of the experiences and perspectives of both knowledge producers and users. Using survey data from a study of university-based social science researchers in Australia who have engaged in research collaborations, the article principally aims to examine factors that influence perceptions of the policy uptake of social research. We use a similar approach to that adopted in the study by Landry et al. (2001a, 2001b) and draw on the Knott and Wildavsky (1980) research-use scale to examine factors that appear to influence reported levels of research impact. We do not examine knowledge utilisation rates between specific social science disciplines (see Cherney et al., 2013, for a comparison across sociology, education, psychology, economics and political science). While we recognise that different scientific disciplines may generate diverse forms of utilisation, given that the various contexts in which knowledge is produced and processed vary across research domains, our aim here is to generate broader generalisations about knowledge utilisation in the social sciences. Hence, we have examined research utilisation across the social sciences as a grouping. We have done this because there is a dearth of empirical studies on research utilisation that move beyond specific qualitative case studies, noting that such studies potentially lack external validity or statistical reliability (Landry et al., 2001a, 2001b; Rich and Oh, 2000).
The article is organised as follows. First, the explanatory model underpinning this study (i.e. the ladder of utilisation) will be discussed. Second, the article outlines the data collection methods used for a recent survey administered to Australian academics as part of our current study. Third, the various dependent and independent measures are described and we outline why – contrary to other studies – we have adopted an index to measure research use for our dependent variable. Fourth, key results from our sample of social researchers are provided, focusing on reported levels of research utilisation and variables that appear to influence knowledge transfer and application. Finally, the article discusses the results and the limitations of the study and concludes with observations about the assessment of knowledge utilisation and translation in the social sciences.
Literature review
Measuring research utilisation
There is no accepted model for measuring research utilisation (Lester, 1993; Oh and Rich, 1996; Smith et al., 2011). A common distinction made in the literature is to differentiate between conceptual, instrumental, symbolic and political use of research (Nutley et al., 2007). Conceptual impact refers to research being used to change understanding about an issue or choice, instrumental use constitutes direct impact on policy or practice, symbolic use of research refers to the tactic of policy-makers using research to delay action, while political use is regarded as more strategic in nature, such as when research is adopted to legitimise or justify a pre-existing position (Monaghan, 2011; Nutley et al., 2007; Weiss, 1979). While these dimensions provide insight into the various ways in which social research can be adopted or can have an impact on policy decision-making, they do not adequately capture the types of activities that characterise different stages of knowledge utilisation – that is, they overlook the breadth of social research usages, ranging from practices that encompass the transmission of ideas through to actual applications in practice (Cherney and McGee, 2011; Henry and Mark, 2003). This is where more detailed scales of research use (or staged models) have become helpful in understanding that utilisation is related to various decision-making processes, particularly concerning the actions of the producers and consumers of social research.
In this study we replicated a modified version of the Knott and Wildavsky (1980) research use (RU) scale, similar to that adopted in Landry et al. (2001a, 2001b). A number of knowledge utilisation scales are available in the literature (e.g. Estabrooks, 1999; Larsen, 1982). However, we have adopted a similar scale to that of Knott and Wildavsky (1980) because it has been frequently cited in the literature, has been used to measure research use among government officials and academics, and has been shown to be reliable (Cherney and McGee, 2011; de Goede et al., 2011; Landry et al., 2001a, 2001b; Lester, 1993). Conceptually, the RU scale can be referred to as a ‘ladder of utilisation’ and Table 1 provides the descriptions for each stage of research use (or rung of the ladder), as presented in our questionnaire to Australian social scientists. The benefit of this scale is that it operationalises research use as a cumulative process that may progress through a number of stages: transmission, cognition, reference, effort, influence and application. The scale is cumulative in the sense that cognition builds on transmission, reference builds on cognition, effort on reference, influence on effort, and application on influence. The RU scale has been criticised as perpetuating a linear understanding of research utilisation because it implies that each stage, or ‘rung of the ladder’, must be sequentially navigated to generate increasing forms of research uptake. Scholars have argued that this ignores the non-linear, or indirect, pathways through which research influences policy decisions (e.g., see Davies and Nutley, 2008). However, the RU scale does recognise the fact that the research utilisation process varies between a range of activities spanning knowledge transfer, translation and uptake. The scale also provides opportunities to measure the significance of factors that distinguish different levels of research utilisation (Cherney and McGee, 2011; Knott and Wildavsky, 1980; Lester, 1993).
Research utilisation scale.
Independent variables influencing research use
Just as there is no agreed conceptual model relating to research utilisation, there is no definitive list of variables developed to help predict knowledge use (Lester, 1993). Most studies have categorised variables under broad headings relating to supply-side and demand-pull factors, as well as dissemination and interaction variables. Supply-side factors include research outputs and the context in which the researcher works. These can include the types of research outputs produced by academics (e.g. qualitative or quantitative studies), whether research is focused on non-academic users, the importance of internal or external funding sources, and the institutional drivers that influence the initiation of collaborations with external partners and end-users (Bogenschneider and Corbett, 2010; Cherney et al., 2012a, 2012b). Barriers that researchers encounter within their own institutional settings such as academic reward systems (e.g. incentives to publish in ‘A star’ journals) also influence the production and supply of research to external agencies (Jacobson et al., 2004). Demand-pull factors relate to the end-user context, including whether end-users consider research to be pertinent, whether research coincides with end-users’ needs, the priority users place on the quality of the research, and the feasibility of adopting research recommendations. Added to this are organisational processes such as the value policy-makers and practitioners place on research evidence and their level of skills to interpret and apply research knowledge. Such factors influence the demand for academic research within end-user organisations (Belkhodja et al., 2007; Ouimet et al., 2009). Dissemination variables are concerned with efforts to adapt and tailor research products (e.g. reports) for users and to develop strategies focused on communicating research results (Huberman, 1990). The more researchers invest in adaptation and dissemination, the more likely research-based knowledge will be adopted. Adaptation includes efforts to make reports more readable and easier to understand, efforts to make conclusions and recommendations more specific or more operational, efforts to focus on variables amenable to interventions by users, and efforts to make reports appealing (Cherney and McGee, 2011). Dissemination efforts include strategies aimed at communicating research to targeted end-users, such as when researchers use different social media to communicate their research messages, hold meetings to discuss the scope and results of their projects with specific users or partners, and target particular forums, for example reporting on their research to government committees. Finally, interaction variables focus on the intensity of the relationship between knowledge producers and potential users or beneficiaries of research. These types of factors include informal personal contacts and networks between researchers and potential end-users, participation in committees, or experience with research partnerships, such as the number of research partnerships in which an academic has engaged. The argument is that the more intense these linkages are, the more likely research uptake will occur (Huberman, 1990; Lomas, 2000).
The current study
The data presented here are drawn from an Australian Research Council funded project with nine industry partners. 2 The project involved four phases: (1) a targeted survey of Australian social scientists; (2) a targeted survey of policy personnel; (3) interviews with a selection of academic respondents; and (4) interviews with a selection of policy personnel. Results reported here are based on phase 1 data. The survey administered to academic social scientists was partially based on existing questions and scales (e.g. Bogenschneider and Corbett, 2010; Landry et al., 2001a, 2001b). New questions were also developed to capture additional data relating to the benefits and problems of engaging in research collaborations.
The survey was first piloted among Fellows of the Academy of the Social Sciences in Australia (ASSA) in September–October 2010. It is estimated that nearly 500 members were sent the survey and 81 surveys were completed, with a response rate of about 17%. There were no significant changes to the survey following the pilot other than minor editing of some lead-in questions to make them clearer. No scales were changed. We have reported the combined results from the same questions used in both the pilot and main survey. A database was established of Australian academics who had secured at least one Australian Research Council (ARC) grant (what are termed Discovery or Linkage grants) 3 between 2001 and 2010 within the field of social and behavioural science. The selection of relevant disciplines was based upon the ‘field of research’ codes used by the ARC to categorise the funded projects, and comprised codes relating to anthropology, criminology and law enforcement, human geography, political science, policy and administration, demography, social work, sociology, other studies in human society, psychology, education and economics. Using this database, a web link to the survey was sent via email to 1950 academic researchers between November 2010 and February 2011. The same reminder email was sent twice during this period and the survey closed in May 2011. A total of 612 completed surveys were received, which constitutes a response rate of 32%. When the main academic survey was combined with the ASSA pilot, the final total included 693 responses. The response rate achieved is indicative of the difficulty of encouraging time-poor academics to participate in projects where they themselves are the subjects of the research. It may also be noted that web-based surveys often suffer from low response rates (Sue, 2007).
The reason for targeting academics who had secured research grants was to ensure the project captured experienced academics who were likely to have had a history of research collaborations, since one aim was to understand the impact and dynamics of such partnerships. Studies have also shown that seniority and the number of external competitive research grants are key factors influencing higher levels of engagement with non-academic end-users and increased levels of research impact (Cherney and McGee, 2011; Landry et al., 2001a, 2001b).
Participants’ background
Most respondents to the survey were drawn from senior academic positions, with the data skewed towards academics at level D and above, that is, Associate Professors/Readers and Professors. As Figure 1 indicates, over 40% of the sample was at the level of Professor, followed by Levels D and C. The dominance of senior academics is a result of the recruitment strategy used to generate the sample, given that it targeted academics who had an established profile (e.g. ASSA members) and a history of securing national competitive grants, the award of which tends to be strongly influenced by track record (i.e. ARC grants). This does raise the issue of whether our sample is biased and its effect on reported levels of research use, which we address in the data analysis with the inclusion of particular control variables in our model.
Professional profile.
Respondents were mainly drawn from academics who occupied teaching-and-research positions compared to research-only roles (65% compared to 35% respectively). Respondents were asked their disciplinary background, and these responses are outlined in Figure 2.
Major research discipline identified.
Dependent variable
Knowledge utilisation was measured using a validated version of the Knott and Wildavsky (1980) RU scale, similar to that adopted in the study by Landry et al. (2001a; 2001b). As indicated previously, the scale is based on six stages namely: transmission, cognition, reference, effort, influence, and application. For each of these six echelons (or stages) of the research utilisation ladder, respondents were asked to estimate what had become of their research using a 5-point scale ranging from 1 (never), 2 (rarely), 3 (sometimes), 4 (usually), to 5 (always) utilised in some way.
Previous researchers (Cherney and McGee, 2011; Landry et al., 2001a, 2001b) have used this scale cumulatively (with each stage building upon the next) and assigned a value of 1 when respondents replied always, usually, or sometimes, and with all other responses assigned the value of 0, which means they ‘failed’ to move up the scale. There are two ways that this cumulative approach can be analysed. The first is to run a separate logistic regression for each stage of research utilisation as Landry et al. (2001b) did in their study. Hence respondents who pass all six stages would be represented in each stage or regression model (see Figure 3). This is particularly problematic with our sample, because a large proportion (66%) of the sample reported they passed all six stages. Hence the question arises whether such a method would really be determining what predicts movement from one stage to the next, or whether progression across each stage is masked by the dominant group. In order to address this criticism, a second approach would be to create an ordinal variable with seven levels, including in each level only those individuals who ‘passed’ that level. Thus respondents in each level would be unique. Table 2 presents the number of respondents categorised in each level/echelon according to such progression criteria. For instance, 4% of the sample passed the transmission stage but did not progress further. However, an ordinal logistic regression analysis is not possible due to our small sample size and the number of cases in each level.
Number of academic researchers climbing the echelons of the ladder of knowledge utilisation – progression is subject to passing previous echelons.
Proportion of academic respondents at each stage of the research utilisation scale.
The next possible option would be to examine whether these stages are in fact exclusive. Does a negative experience (‘failure’) in one stage preclude academic researchers from progressing to other stages? Or should these stages comprise an index? Descriptive statistics, as presented in Figure 4, illustrate that failure in one stage does not preclude academic researchers from passing subsequent stages. This is an important consideration owing to the criticism that the RU scale perpetuates a linear conceptualisation of research utilisation; indeed, the data in Figure 4 indicates that one does not necessarily have to traverse in sequence each rung of the research utilisation ladder to reach the ultimate stage, that is, ‘application’ of research findings by users – hence the RU ladder is not necessarily cumulative. A factor analysis of the items (or stages), revealed a 1-factor solution and a Cronbach’s alpha coefficient of 0.91 (see Table 3). The results indicate that these items are measuring one construct and that the index seems to be reliable. Unlike the previous studies outlined above, in this study it was decided to use the items as an index to measure research use given the results of the factor analysis. This was also decided given the criticism made of the RU scale relating to assumptions about the process of knowledge diffusion. A mean index score was calculated for all six stages. The mean score for the research utilisation index is presented in Table 4.
Number of academic researchers passing each stage of research utilisation. Failure in one stage does not preclude passing subsequent stages.
Internal reliability coefficients (Cronbach’s alpha) for variables.
Means and standard deviations a academic research.
Standard deviations only reported for continuous measures.
Independent variables
A number of indices were created and included in our model as independent variables. The items used in each index were determined by factor analyses, with each index comprising a 1-factor solution. The Cronbach’s alpha coefficients for these independent variables are presented in Table 3 and detailed descriptions of index compositions can be found in Cherney et al. (2012a, 2012b).
Descriptive statistics for each independent variable are presented in Table 4. Academic researchers indicated that academic funding (i.e. national competitive grants such as ARC grants and internal university funds) were more important than funding from government and non-government agencies in ensuring their research is conducted. Academic researchers indicated that the ‘useability’ of the research is given a higher priority by end-users, compared to other features, such as the quality or feasibility of research. A very high level of importance is accorded by academic researchers to the use of refereed publications as a method through which to disseminate their research, followed by the importance of tailoring research to meet the needs of end-users. Table 4 also illustrates a high level of agreement among academic researchers concerning the fact that they encounter barriers in the transfer and uptake of their research. The average number of research partners with whom researchers engaged was nine research partners per researcher. The number of grants received by these academic researchers varies between 0 and 67, with the average researcher having received 9 grants.
Control variables
A number of control variables were included in the model to control for the personal characteristics of the respondents, namely disciplinary background, home institution and position type. There were five main disciplines as depicted in Figure 2 and the remaining disciplines were categorised as other. University affiliation comprised researchers affiliated with Go8 universities and those who were not (coded as 1 and 0 respectively). As presented in Table 4, the sample was equally represented in both groups. Position type (academic level) comprised two groups, Levels D and E academics and Levels A–C academics (coded as 1 and 0 respectively). As mentioned above, position type comprised teaching-and-research or research-only positions.
Data analysis
Given that our dependent variable is approximately continuous, a multiple linear regression model was used to estimate the associations between research utilisation (our dependent variable) and a number of explanatory variables such as benefits and barriers associated with engaging in research with policy-makers and practitioners. As a preliminary check, we examined the correlations between all variables in the model. They ranged between .001 and .74, suggesting that multicollinearity was unlikely to be a problem (the correlation matrix was too large to depict in an Appendix). This was confirmed by a relatively low value of the mean Variance Inflation Factor (VIF) of 1.72, with the individual variables’ VIFs ranging from 1.15 to 3.32. The four highest correlations were between (1) importance of meetings and dissemination activities with end-users and importance of tailoring research when end-users are the focus (0.74); (2) problems relating to the orientation of research partnerships and ‘consequences’ of investing in research partnerships (0.73); (3) importance of using contacts, seminars and reports to present research to policy-makers and practitioners and importance of meetings and dissemination activities with end-users (0.71); and (4) importance of using contacts, seminars and reports to present research to policy-makers and importance of tailoring research when end-users are the focus (0.68). All four correlations were statistically significant.
Regression results
The regression results are presented in Table 5. The results indicate that 13 variables were significantly related to the reported utilisation of academic research. The more the research is targeted to end-users, the more likely academic researchers report research utilisation. The more beneficial the collaboration to academic researchers, the more likely they report utilisation. Academic researchers reported that, when end-users felt that research was of high quality and useable, it was more likely to lead to utilisation. However, academic researchers indicated that the more policy-makers or practitioners prioritised the ‘feasibility’ of research the less likely were they to perceive that end-users would use academic research. The importance of tailoring research for end-users is positively and significantly associated with reported levels of research use. The importance of meetings and dissemination activities with end-users and using media to disseminate findings is associated with reported levels of research uptake. Using interactions such as contacts, seminars and reports to present research to end-users is also associated with reported levels of research impact. As the number of grants increases so does the likelihood of research utilisation. In addition, it appears that the adoption of quantitative approaches increases the likelihood of research uptake for those in our sample. However, despite our sample indicating that academic funding was important, there was a negative relationship found between types of funding sources and reported research utilisation by end-users. Even with the inclusion of our control variables (position type and level, university affiliation and disciplinary background) the model remained stable, strengthening the overall generalisability of our results.
Multiple linear regression equations predicting utilisation of academic research.
Standard errors in parentheses.
p < 0.10, ** p < 0.05, *** p < 0.01.
Discussion
We recognise there are limitations in asking academics to comment on the impact of the research they produce and levels of knowledge utilisation. It can be argued that respondents were required to make retrospective judgements of processes they may not have directly observed. Such self-report data can also reflect social desirability biases in that academics may tend to inflate the relevance of their research. There is considerable merit in the argument that tracing the impact of research evidence needs to be complemented by more detailed qualitative case study work. The survey data only gives insight into broad patterns of knowledge utilisation as reported by our academic sample – hence providing one possible perspective. We propose, however, that while our sample is biased towards senior academics, the overall seniority and experience of our survey respondents places them in a strong position to make judgements about research utilisation by end-users, compared to academics with less experience of working with external partners.
What do our results indicate about the transmission and uptake of research-based knowledge in the social sciences? From our data analysis it is clear that active efforts by academic social researchers to ‘push out’ or disseminate research products to potential beneficiaries is central to promoting utilisation by non-academic end-users. Academic researchers in our sample were mindful that, in order for their research to have an impact on policy or practice, there is a requirement to directly engage with users through meetings and dissemination processes, and that it is essential to tailor research projects and findings to their needs. While ‘producer-push’ models of research translation have often been criticised as overly rational and linear (Nutley et al., 2007), our results demonstrate that academics who wish to see their research utilised by policy-makers and practitioners have to commit to strategies that help diffuse their research to end-users. To do so academics need to find supporters within end-user organisations who are willing to help push research up to senior levels and act as change agents to help promote the use of academic research by relevant personnel (Cherney, 2013). According to our sample, academics need to step outside their comfort zone of traditional forms of research dissemination and begin engaging in alternative forms of dissemination. Hence, it is not surprising we found a relationship between reported levels of research utilisation and the importance accorded by our respondents to the media for presenting their research (see Table 5).
Despite accusations that academic social researchers do not understand the needs of policy-makers or practitioners, our respondents were aware that such users do have different priorities and perceptions when it comes to judging the relevance and use of academic social research. This awareness may be linked to the fact that our respondents were relatively experienced in engaging with multiple external research partners, making them mindful of these contextual issues. Likewise, as has occurred in many countries, academic researchers in Australian universities are under increasing pressure to demonstrate the value and impact of the research they produce and are therefore alert to issues relating to research impact. This awareness was potentially increasing at the time when the survey was completed, because the first round of the Australian university research assessment exercise, the ERA, had been conducted in 2010, and Australian universities were preparing for another round in 2012. University research assessment exercises such as the ERA have generally focused on quality indicators such as journal rankings, journal impact factors and citation counts. However, these indicators can deter academics from investing in alternative outlets for practitioner audiences beyond the standard peer-reviewed academic journals (Cherney, 2013).
There are a number of noteworthy results that arose from the regression analysis. One was that the quality of research was still seen as important in influencing research utilisation among end-users (i.e. policy-makers or practitioners), but there was recognition that quality is not the only factor, with judgements about ‘feasibility’ also seen as influencing (in a negative way) levels of research use. The ‘feasibility’ variable comprised three items: research recommendations are seen as economically and politically feasible and research findings support a current position or practice (see Cherney et al., 2012b). When feasibility is seen as a strong priority, our respondents reported that academic research is less likely to be used by policy-makers. Another result was the relationship found between methodological approaches and levels of research utilisation. While not strong, there was a positive relationship between whether respondents reported they used quantitative approaches and subsequent utilisation. Landry et al. (2001a, 2001b) found a similar result in relation to particular social science disciplines, as did Cherney and McGee (2011) and Cherney et al. (2013). It is possible that the perceived effect of quantitative methods in influencing levels of utilisation may reflect the preferences and bias of users for particular types of research. The perceived neutrality of quantitative research vis-a-vis the interpretative and contestable nature of qualitative research may be attractive to policy-makers and practitioners seeking digestible summaries of research outcomes, compared to the nuanced and detailed nature of results typically associated with qualitative methodologies. This association requires further investigation of the preferences of policy-makers and practitioners for certain types of research methods and products.
The results in Table 5 showed a negative relationship between the importance of particular types of sources to fund the undertaking of research and the likelihood of research use. This negative relationship was more significant when it came to academic funding sources (e.g. internal university funds and research council grants), compared to other types of funding from government and private sector agencies. This supports a similar finding by Landry et al. (2001b). This is perhaps the result of the fact that there could be relatively low expectations relating to the utilisation of research that is funded through sources mainly supporting academically oriented projects in the social sciences. However, with the increasing pressure on university research funding councils such as the ARC to demonstrate that their funded projects are utilised beyond the boundaries of academia, the influence of funding source on research uptake is a topic that requires further investigation. It may be the case that the different types of contractual relationships that underpin particular funding sources potentially influence forms of research translation and uptake.
Conclusion
There is little doubt that efforts to measure the impact of university-funded research have made academics more mindful of how they translate and transfer their research to users. Efforts by academic researchers to engage with potential users of research clearly help to generate improved levels of uptake. While our results largely pertain to academics in the social sciences, there is little doubt they provide broader lessons for other research fields, particularly in understanding how the contexts and activities of academic researchers, and their relations with research users in fields of policy and practice, shape levels of knowledge translation and uptake. It must be recognised, though, that the decisions and choices of end-users to draw on research-based knowledge are largely beyond the control of academic researchers. However, given the overwhelming finding that dissemination and forms of interactions between academic researchers and end-users have a major bearing on research utilisation, the development of improved incentives to invest in these types of activities would clearly produce benefits in enhancing knowledge transfer and uptake. While this study only provides insights into the perspectives of academic social researchers in Australia, it does highlight more broadly how the impact of social science research can be improved.
Footnotes
Funding
This project was finically supported through the Australian Research Council Linkage project LP100100380. This project has received cash and in-kind support from the following industry partners: Australian Productivity Commission; Australian Bureau of Statistics; Queensland Health; Queensland Department of Communities; Queensland Department of Employment; Queensland Department of Premier and Cabinet; Victorian Department of Planning and Community Development; Victorian Department of Education and Early Childhood; and the Victorian Department of Human Services.