Abstract
Trust in government and its antecedents may differ for different levels of government. This article examines the little-studied Australian state of Western Australia (WA). Drawing on a representative online panel survey, and testing three groups of hypotheses, we look at perceptions of ethical government, performance as measured by satisfaction, trust in government, political participation, and perceptions of transparency, and the interrelationship between these variables. Trust in state government is higher than in federal government in WA across a number of measures. Perceptions of performance and ethical government are also higher at state level, while respondents are more likely to participate in politics at state level. Perceptions of transparency are associated with trust in government at both state and federal level, as well as being related to perceptions of ethical government and performance. However, transparency is not related to political participation at either state or federal level.
Trust in government is seen as somewhat of a panacea in a still growing body of research in public administration, associated with such things as willingness to pay taxes, accept government directives—and in the case of the Covid-19 pandemic—act in accordance with government directives, including adopting pharmaceutical and non-pharmaceutical medical interventions (Goldfinch and Taplin 2022). How is this relevant to local and state governments? The question we pose in this field note is this: how does trust in government and its claimed antecedents of perceptions of ethical behavior, political participation, and perceptions of government performance, differ for different levels of government, particularly in federal systems? While an existing body of research suggests local and state governments do differ in these (Berner and Smith 2004; Biddle, Gray, and McAllister 2024; Jedwab and Kincaid 2018; Jensen and Piatak 2023; Pears and Sydnor 2022; Weinschenk and Helpap 2015), there are subnational governments that also perhaps deserve further study to see if this holds more widely. So we address the concerns of this journal and its focus on state and local government, while broadening the field of research to the little-studied Australian state of Western Australia (WA).
We add novelty by studying this geographically large and comparatively wealthy, but somewhat research-neglected, state of (perhaps) intrinsic interest. We provide novelty by testing variables on both federal and state government. We also add novelty to the broader public administration literature by testing a wide range of antecedents of trust—seldom all found in one article—adding to the cumulative body of research. Drawing on a representative online panel survey, and testing three groups of hypotheses, we look at perceptions of ethical government, performance as measured by satisfaction, trust in government, political participation, and perceptions of transparency, and the interrelationship between these variables, with this focus suggested by a body of research (c.f. Etzioni 2010). Importantly, we provide novel findings of interest to the broader literature on trust, transparency, and public administration generally.
We find trust in state government is higher than in federal government in WA across several measures. Perceptions of performance and ethical government are also stronger, while respondents are more likely to participate in politics at state level. Perceptions of transparency are associated with trust in government at both state and federal level, as well as being related to perceptions of ethical government and performance. However, participation is not associated with perceptions of ethical government, government performance, or trust. This is despite a body of existing research suggesting otherwise (c.f. Langella et al. 2023). Transparency has a direct relationship to trust in government at state level, but an indirect one at federal level, mediated by other antecedents. This finding underlines the importance of studying different levels of government to unpack these complex relationships. Of note, we find no link between political participation and transparency, despite a considerable body of official and academic literature claiming otherwise, and assertions of such in Australian statute (c.f. Langella et al. 2023). Hence this note provides a “falsifying case” of interest to the trust, transparency, and participation literature in political science and public administration.
This note proceeds with a brief outline of Australia’s federal system, particularly as it pertains to Western Australia. We then develop our hypotheses, before describing our methods and reporting our results. We discuss the importance of our findings and implications for practice and further research in our discussion and conclusion.
Western Australia and Federalism in Australia
The Commonwealth of Australia (Commonwealth and federal government are used interchangeably) was established in 1901 by the federation of the six self-governing British colonies of the Australian continent, after a series of constitutional conventions (Galligan 1995; Mein Smith, Hempenstall, and Goldfinch 2008). These became the states of the new country, with their powers set out in the Australian Constitution of 1901, as interpreted by convention and by the purposely-established High Court. Federation had received prior majority support in referendums in each of the colonies: Queensland, New South Wales, Victoria, Tasmania, South Australia, and Western Australia (Galligan 1995).
The Commonwealth is the senior partner in the relationship (Fenna 2018; Wilkinson, McKenzie, and Bolleter 2022). Although most powers are shared with greater powers ostensibly ascribed to the states, the federal government has monopoly powers in defence, foreign trade, foreign affairs, immigration, customs and excise, and it dominates social services. Moreover, its dominance of revenue generation and spending give it power over aspects of other policy, with federal control of income tax since 1942 and the goods and services tax adopted in 2000, seeing it raise around 80% of revenue collected by all levels of government, while providing over half of state revenues (Sanders 2024). Specific Purpose Payments are made to the states by the Commonwealth in particular areas of policy, albeit with most revenue delivered through untied general revenue assistance (Galligan 2012). Fiscal equalization sees a presumption of equal per capita public spending on citizens across states and territories, taking into account differences in ability to raise revenue and the relative costs of service provision, giving the federal government a central role in income redistribution. States play a key role in criminal justice, education, health and other social services, and infrastructure, and there are differences in legal frameworks, and labor market regulation, amongst other things. Services however, are often delivered in cooperation between state and federal governments (French 2012). High Court decisions and various intergovernmental agreements saw a gradual extension of federal power after 1901 (Fenna 2018).
Relations between different levels of government have been mostly harmonious with some exceptions, notably the federal income tax laws unsuccessfully challenged by four states in the High Court in 1942, with a further largely unsuccessful challenge in 1957. Policy and legal harmonization and other matters are often settled by negotiation and subsequent intergovernmental agreements, notably with the Council of Australian Governments established in 1992 with committees in a number of cross-jurisdictional and policy areas of national significance, and with representation from federal, state, territory and local government (French 2012). In 2020 this was replaced by the National Cabinet, which plays a similar role (Albanese 2024).
Of all states, WA’s relationship with federation is possibly the most contested (Lecours and Béland 2019; Zimmermann 2012). Larger than Western Europe and separated by deserts in the south and national parks in the north from much of the rest of Australia, its capital Perth is more than 3,000 km from the federal capital Canberra and closer to some Indonesian cities than to the populous capitals of the eastern states. WA developed in relative isolation through the nineteenth and into the twentieth century. Indeed, until the completion of the Trans-Australian railway in 1917, sea routes saw the eastern states better connected to New Zealand than Western Australia (Mein Smith, Hempenstall, and Goldfinch 2008). Western Australia had received responsible government from Britain only in 1890, and was late to joining the federation, to the extent it is not mentioned in the preamble to the Commonwealth of Australia Constitution Act 1901 enacted by Westminster. Sentiment quickly turned against federation, with 66% of votes cast in favor of secession in a 1933 referendum held by the Western Australian government. However, appeals to the British parliament to allow secession were rejected, with the British parliament citing no powers of secession in the Australian constitution and restricted British powers of amendment (Zimmermann 2012).
State identity in WA remains relatively strong. A 2021 poll found over half of WA respondents identified “first and foremost” as Western Australian rather than Australian, in marked contrast to other states where 78%–94% saw themselves as firstly Australian (Cowan and Carling 2021). An inchoate secessionist sentiment continues, particularly with the recent WA resource boom seeing it contributing more in tax to the Australian federal coffers than it received back in funding. Indeed a 2020 poll found over a quarter of West Australians supported secession (Law 2020). This was in the context of a dramatic revival—or at least a re-focus—of the importance of federalism in Australia in the wake of the Covid-19 pandemic (Biddle, Gray, and McAllister 2024; Fenna and Goldfinch 2023). WA closed its borders to interstate and intercountry travel and imposed strict lockdowns. This closure was despite requests otherwise from the federal government. As the pandemic manifested, the six Australian states and two territories pursued similar policies, albeit with variations in terms of length and severity of lock-downs, and differing approaches to monitoring and enforcement. In the 2021 WA state elections, voters enthusiastically endorsed these policies, with the incumbent Labor government dramatically increasing its majority, and reducing the Liberal party to just two seats. The 2025 election saw a swing against Labor, but it maintained a massive majority of 46 seats out of a possible 59, with the Liberal party increasing to 7 seats to regain its leading position in the opposition alliance, and the country-based National party increasing to 6 from 4. In sum, these key differences to other Australian states, and the contribution this study makes—particularly as a falsifying case—to the understanding of state and local governments, provide further justification for WA’s dedicated study.
Hypotheses
Subnational governments are often responsible for delivering the bulk of public services, and are physically and perhaps psychologically closer to citizens, allowing citizens to sometimes better engage, develop familiarity with, and form more accurate perceptions of personalities, performance and processes (Dardanelli et al. 2018; Pears and Sydnor 2022; Lecours 2021; Scott et al. 2021; Wilkinson et al. 2023). Some studies suggest citizens may have more positive perceptions and higher levels of trust in state and local government vis-a-vis central, albeit with mixed findings (Cole and Kincaid 2000; Jensen and Piatak 2023; Pears and Sydnor 2022; Weinschenk and Helpap 2015). Trust in this case is the belief that one’s vulnerabilities will not be systematically exploited by another to whom one ascribes such trust. Greater degrees of political participation are also seen at subnational level, although findings differ (Berner and Smith 2004; Goldfinch, Gauld, and Herbison 2009; Lecours 2021; Muthomi 2024). As noted, federal and state governments are established in the Australian constitution. Australia also has local governments which are creatures of the states. 1 Descriptive results for local government are reported, but are not tested in hypotheses, due to reasons of economy for a short note, and we focus on state and federal (see Table 1). This discussion gives us Hypothesis H1, subdivided into H1a, H1b, H1c, and H1d:
H1. State government in Western Australia is perceived more positively than federal government with respect to trust (H1a), performance (H1b), ethical behaviour (H1c), while political participation is higher at state level than federal level (H1d).
Transparency is “generally defined as . . . enabling the public to gain information about the operations and structure” of government processes, decisions, and outcomes (Etzioni 2010, 389) and provides somewhat of a panacea for problems of government, particularly in official and consultancy literature (Hood and Heald 2006; Krah and Mertens 2020). A body of research argues by shining a light on actions, processes and outcomes, transparency encourages better and more ethical behavior and results from government actors, while building understanding and engagement from the citizenry. Hence it is associated with more ethical and less corrupt government (Chen and Ganapati 2023); better government performance and citizen satisfaction (Hood and Heald 2006); higher trust in government (Mabillard 2022; Porumbescu 2015); and higher levels of political participation (Langella et al. 2023); albeit with findings contested (Grimmelikhuijsen et al. 2013; Schnell 2018). The link between transparency and participation is explicit in the Australian Freedom of Information Act 1982, with open government claimed to lead to increased “public participation in Government processes, with a view to promoting better-informed decision-making [and] increasing scrutiny, discussion, comment, and review of the Government’s activities” (section 3(1)(2)). A similar provision obtains in the West Australian Freedom of Information Act 1992, which has as an objective the enabling of “the public to participate more effectively in governing the State” (section 3(1)). This gives us hypothesis H2, subdivided into H2a, H2b, H2c, and H2d:
H2. Higher perceived levels of transparency are associated with higher perceived levels of ethical behaviour of government actors (H2a), higher levels of satisfaction with government performance (H2b), higher levels of political participation (H2c), and greater trust in government (H2d), at both state and federal level.
The relationship between perceived transparency and trust is likely to be mediated by perceived ethical behavior, government performance, and political participation (Park and Blenkinsopp 2011). This gives us hypothesis H3:
H3. The relationship between perceived trans-parency and trust is mediated by perceived ethical behaviour, government performance, and political participation at both state and federal level.
Descriptive Statistics.
Note: NA = not available/applicable.
Scales 1–4, with 4 being the highest.
Four participation items not measured at both Federal and State levels were: follow a political party on social media (23%), followed a politician on social media (15%) searched for information (38%) and contacted a government official (31%; coded 0/1).
Methods and Research Design
Survey
A commercial company, Dynata, delivered surveys to a Western Australian online panel, generating representative quotas based on age, gender, and region. A broadly representative sample of 504 was used in the final analysis, consistent with other studies, and sufficient to generate statistically significant results (Goldfinch, Taplin, and Gauld 2021). Of this, 53% were female; and 81%, 15%, and 4% resided in cities, towns, or rural regions, respectively. There was a higher-than-expected proportion of older respondents with 36% over 64 years old. Thirty-six percent had completed a university education. Ethics approval was obtained from Curtin University Human Research Ethics Office (HRE2020-0358). Informed consent was obtained from all subjects and/or their legal guardian(s).
Variables
Trust in government was measured by “Federal Government (in Canberra) in Australia is generally trustworthy” and “State Government in Australia is generally trustworthy.” Trust in the prime minister/premier, members of parliament, and public servants, at both federal and state levels, were also measured, with similar measures of local government (Table 1). Responses were on a four-point scale from strongly agree to strongly disagree. Questions were derived from Goldfinch, Taplin, and Gauld (2021).
Satisfaction with government performance was measured by the questions “I am satisfied with the services provided by my state government” and “I am satisfied with the services provided by the federal government,” with responses on a four-point scale from “Strongly agree” to “Strongly disagree.” Questions were derived from Christensen and Lægreid (2005).
Perceptions of ethical government behavior were measured by asking: “My state government treats people fairly and honestly” and “The federal government [in Canberra] treats people fairly and honestly,” with responses on a four-point scale from “Strongly agree” to “strongly disagree.” Questions were adapted from Goldfinch, Yamamoto, and Aoyagi (2023).
Political participation was measured both for state and federal level, looking at voting at state and federal level, and contact with federal and state ministers and parliamentarians (Table 1). Voting is compulsory in Australia, although about 90% of those enrolled actually vote and prosecutions for non-compliance are rare. Four other measures of political participation were also collected, not specifically referring to state or federal level. These included whether the respondent followed politicians on social media, followed political parties on social media, had actively searched for information on parliament or the government, and whether they had contacted a government official. These are reported in Table 1. Questions were derived from Putnam et al., (1993) Putnam (2000) and Goldfinch, Yamamoto, and Aoyagi (2023). These questions were used to provide alternative and/or supplementary measurements of political participation consistent with the literature (Goldfinch, Yamamoto, and Aoyagi 2023). For regressions predicting trust in federal/state government, political participation was measured with the number of items performed at the federal/state level out of voting, contacting a minister and contacting a member of parliament (Table 1). Including the four participation measures not measured specifically at the federal/state level (see Table 1) did not change results.
Perceptions of transparency were measured by the following questions: “Government provides clear information on its processes, decisions and services” and “Government websites make it easy to find the information you want,” with answers on a four-point Likert scale from “strongly agree” to “strongly disagree.” Results were averaged (Cronbach’s α = 0.81). Questions were developed by the authors.
Our control variables were age, education, income, gender, residence, ethnicity, and voting behavior. We use 0/1 dummy variables to capture residence (town and rural, relative to city), political support at federal level (relative to voting [center-right] Liberal/National coalition/alliance) and ethnicity (relative to Caucasian).
Statistical Treatment
Comparisons between state and federal (Hypothesis H1) were tested by paired t-tests. For other hypotheses we report correlations between variables and full multiple regression results predicting trust from all other variables, including controls. We test for mediating effects (H3) by examining whether significant relationships between transparency and trust are explained by significant relationships between transparency and mediators, and between mediators and trust (Baron and Kenny, 1986). To ensure common method variance did not impact results, we also repeated regressions with trust in public servants as an independent variable. While trust in public servants is strongly related to trust in government (p < .001) for both state and federal government, our results remain substantially unaffected by the inclusion of this variable into regressions. As noted above, our sample contains a higher proportion of older respondents but our results are not impacted by the age of the respondents. 2
Results
Descriptive results are reported in Table 1. Trust in government (H1a), performance (H1b), ethics (H1c), and participation (H1d) are all significantly (p < .001) higher at the state level compared to the federal level (Table 1). Thus Hypothesis 1 is supported. With respect to more specific measures of trust, we find higher levels for state versus federal (p < .001) for trust in the leader (prime minister/premier at the federal/state levels) and trust in members of parliament, but an insignificant (p = .155) difference in the level of trust in public servants at federal and state levels. Regressions are discussed further below. In terms of local government, Table 1 reports descriptive statistics, not tested in the hypotheses. 3
Table 1 shows results for the three items for political participation, with contacting ministers (p < .001) and members of parliament (p = .002) both significantly more common at the state level. While exceptions to compulsory voting apply, voting is similarly high (p = .100) in both federal and state elections. Other types of participation are also reported in Table 1, which were measured but not specified at a federal/state level.
Correlations between variables (Table 2) show transparency is positively associated with perceptions of ethical government (H2a), performance (H2b), and trust in government (H2d). Hence hypotheses H2a, H2b, and H2d are supported. This is true for both state (top right of Table 2) and federal governments (lower left of Table 2). While many correlations between variables are similar for state and federal governments (e.g., the correlation between performance and ethics are both 0.72), when they differ, they tend to be higher for federal government. For example, the correlation between trust and performance is r = 0.71 for federal government but only r = 0.60 for state government.
Correlations Between Variables (Federal Below and State Above the Diagonal).
Note: Significant (p < .05) correlations in bold.
Hypothesis H2c is not supported, with correlations between participation and trust insignificant (p > .05) and negative (r = −0.04 federal; r = −0.05 state), contrary to the hypothesized direction. Note this contrasts with hypothesis H1d: we find support for the hypothesis that political participation is higher for state government than federal government (H1d), but no support for the hypothesis that political participation is correlated with trust in government, for either state or federal government (H2c).
After including all variables in the regression to predict trust in government (Table 3), transparency is found to be insignificantly related to trust in federal government (p = .115) and significantly related to trust in state government (p = .002). Thus the relationship between transparency and trust is fully mediated by the other variables at federal level, but only partially mediated at the state level. Results (not shown) confirm transparency (and performance, and ethics) are significantly (p < .001) related to trust, if each are included separately in regressions predicting trust along with only the control variables. That is, transparency is significantly correlated with trust after adjustment for control variables (for both state and federal government). Thus, we find strong support for hypothesis H3 at the federal level, but weaker support at the state level. For state government there is a direct relationship between transparency and trust that cannot be explained through the effects of ethics and performance. Participation does not mediate the relationship between transparency and trust as it is not significantly related to any variable.
Multiple Regressions Predicting Trust in Federal and State Government.
p < .05. **p < .01. ***p < .001.
While political participation was significantly higher for state vis-a-vis federal government (H1c), political participation (at state and federal level) was not significantly related to trust or other variables (Tables 2 and 3). This result also holds when other measures of political participation are included (results available from the authors on request). That is, we find no evidence that any of our political participation measures were related to trust in government. The most noteworthy relationship we found was that transparency was significantly higher amongst respondents who followed a politician (p = .004) or a political party (p = .001) on social media. However, following a politician on social media has a relatively small correlation with trust in government (r = 0.03 federal; r = 0.10 state) compared to the other variables correlated with trust (Table 2), and is insignificantly related to trust if used as participation in regressions (Table 3).
It is unclear from Table 3 whether relationships (hypotheses 2 and 3) are significantly different between federal and state regressions. To address this, we conducted a regression predicting the difference in trust (state—federal) using the differences in independent variables. We found significant results for performance (p < .001) and ethics (p < .001). Thus the degree to which trust in state government exceeds trust in federal government depends significantly on the corresponding differences in perceptions of performance and ethical behavior. 4
Discussion and Conclusion
We contribute to the study of federal systems and subnational governments, trust in government, and public administration more broadly, by examining the little-studied Australian state of Western Australia. Its citizens exhibit differences in perceptions of the state’s place in the Australian federation with stronger degrees of state identity and skepticism toward the federation, compared to other states (Cowan and Carling 2021). Differences (and similarities) regarding trust in government and its antecedents are also suggested in previous broader Australian studies (Biddle, Gray, and McAllister 2024; Goldfinch, Taplin, and Gauld 2021), but these lack focus without dedicated studies of a WA sample. This leaves a research gap that is addressed by this study. Moreover, examining a broader range of antecedents of trust in this article than found in some previous studies, increases our contribution to the understanding of comparative federalism, trust, and political science and public administration broadly. Our novel finding that transparency is not related to political participation despite a body of literature claiming otherwise, has interest to the broader literature.
We find higher levels of trust in government, more positive perceptions of ethical behavior and performance measured by satisfaction, and higher levels of political participation in WA’s state government, vis-a-vis federal. These findings may not be surprising but were certainly not inevitable. Physical closeness may facilitate greater local political participation, particularly in a jurisdiction so physically isolated as WA, but an increasingly online world could have lessened this distance. Moreover the difference between trust in government at state vis-a-vis federal is greater if measured at political leadership level (premier at state, prime minister at federal), suggesting offices and personalities do matter, and that our sample perhaps distinguishes between political leaders and other government actors. Members of parliament in both jurisdictions received comparatively lower levels of trust (albeit higher at state), with public servants lower still. Moreover, there are statistically insignificant differences found between trust of state and federal public servants. This contrasts with some studies finding various measures of trust in government are nearly interchangeable, and suggests we are unraveling something more than just a simple state-versus-federal dynamic (c.f. Manning and Guerrero 2013). Indeed, we may be capturing perceptions of trust in actual personalities rather than simply in the office itself. This might be a function of the small political culture of Western Australia.
We show transparency—somewhat of a panacea in a body of (particularly official) literature but contested in academic work—is associated with trust in government, and perceptions of ethical government and performance, at both state and federal level. This relationship differs for state and federal government however. The relationship between trust and transparency is a direct one at state level, underlining its importance. But it is an indirect one at federal level, mediated by other factors, suggesting the importance of further study of subnational governments to better understand this phenomenon.
One of our key findings, with interest to public administration broadly, is that transparency is not associated with political participation at either state or federal level. This holds across our different measures of participation. It contrasts with such a link assumed in academic, government, quasi-government and consultancy literature, found in some studies, and indeed asserted in Australian and Western Australian statute. Nor are trust in government and our other antecedents associated with political participation, again contra some other studies (c.f. Langella et al. 2023). Trust is significantly correlated with performance and ethical government for both state and federal governments, confirming their importance as antecedents. These findings are a contribution to the broader literature on trust and state and local government.
Are there policy/practitioner implications? This study highlights the importance of state government in terms of delivering services and encouraging participation. In the Australian federal system where services are often delivered jointly, a local focus for even these services might encourage a greater engagement and higher level of satisfaction. Moreover, if encouraging political participation is a genuine concern, our findings show it is a complex phenomenon, and simply making information available through transparency is neither an easy nor necessarily successful way to encourage it. Moreover, the intuitively appealing link between transparency and participation unravels if transparency acts to suggest participation is less needed. Perceiving governments as transparent may make us see them as more trustworthy, competent and ethical, and suggest we are happy to leave them to govern. There may be a “latent transparency” where citizens value that transparency exists, even if they do not seek information (Grimmelikhuijsen, Piotrowski, and Van Ryzin 2020); and notably we do measure information search as a participation variable. Transparency in this sense is a signifier of good things. And it might go some way to develop trust and belief in government performance and ethical behavior. In sum, this note highlights the importance of understanding subnational jurisdictions which for many citizens provide the bulk of their direct experience of government and even perhaps a central identity, with their unique characteristics washed out in multi-country studies and an overarching focus on central governments.
This study opens avenues for further research. Our differing findings on the relationship between trust and transparency suggest it might be worth examining whether a correlation between transparency and trust represents a causal effect, or whether this effect is direct or indirect through perceptions of performance and ethics, or even whether correlations are simply spurious and driven by other factors. For example, we might hypothesize that transparency drives perceptions of performance and ethics, which in turn drive trust. But it might be that perceptions of performance drive both trust and perceptions of transparency. Again, we reiterate the need to examine different levels of government to unravel the complexities of these relationships. Moreover, some issues might be better addressed with qualitative research, and perhaps time-series panel data. For example, further studies of why some actors at state level are better trusted would benefit from qualitative interviews to better understand and contextualize results, and unpack causal relationships. Further research on the differences between citizen preferences toward, and perceptions of actual transparency, how this differs between levels of government, and how this affects perceptions of other aspects of government, would also be useful. Finally, comparing Australian sub-national governments with others in federal and non-federal countries would further illuminate this discussion.
Footnotes
Data Availability Statement included at the end of the article.
Data Availability Statement
Data available from authors on reasonable request.
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) received no financial support for the research, authorship, and/or publication of this article.
