THE HIGH COURT OF AUSTRALIA AND PUBLIC OPINION TOWARD NATIVE TITLE LAND RIGHTS By William Martin Myers A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Political Science 2012 ABSTRACT THE HIGH COURT OF AUSTRALIA AND PUBLIC OPINION TOWARD NATIVE TITLE LAND RIGHTS By William Martin Myers Under what circumstances can courts influence the public? This question is of much greater importance for courts than for other political institutions because they have no formal mechanisms to enforce their decisions and instead depend on their ability to convince the other branches and the public of the wisdom of those judgments. Empirical studies of the influence of court decisions on public opinion have been developed in, but limited to the United States when, in fact, this relationship should be observed in courts around the world. To investigate whether this is the case, this dissertation is the first to provide a systematic, theory driven, empirical evaluation of judicial decisions on public opinion outside of the United States and the first to do so in Australia. Furthermore, the psychological foundations of influence, which suggests persuasion may depend on individual levels of sophistication, has not been assessed in previous studies because of survey artifacts, but is overcome and investigated here. Lastly, the extent to which contextual effects, related to individual surroundings, act as an alternative or rival influence to court decisions on public opinion is assessed. Copyright by WILLIAM MARTIN MYERS 2012 This dissertation is dedicated to my grandfather, my namesake, William Martin Myers. iv ACKNOWLEDGMENTS This dissertation project owes a great many debts, personally and professionally. I want to thank the Department of Political Science at Michigan State University for taking a chance on a history major from Ann Arbor that did not know exactly what he wanted to study, but knew it had to be political science. I have many faculty members to thank for providing encouragement; pushing me to be my best and for never letting me take the easy route. Christopher Achen, thank you for helping me decide this was for me. Ken Sylvester, we spent way too much time pouring over historical agricultural censuses, but the experience I gained coding that data still proves to be valuable. William Jacoby, you helped cultivate a deep love and interest in political behavior; without your guidance I would never have truly appreciated the glory of The American Voter. Carlos Pereira, your infectious love of political institutions brought me a perspective that I fear I would have missed otherwise. Nathan Monroe, your Congress seminar was incredible and if I ever get to teach a graduate course that will be my model. My committee, Ryan Black, Eric Chang and Christopher Smith, deserves special recognition for supporting my interests and for allowing this dissertation to slowly unfold. Thank you giving me the freedom to explore. My greatest professional debt is owed to my Chair, Reginald Sheehan. Kindred spirit? Jedi master? You have given me more than I ever could have imagined. Those of us, your students, who have had the fortune to know you and have you in our lives are the better for it. I could not have done this without your guidance and faith. Thank you. Graduate school is a grind and it helps to have a cheering section. I could never have done it without the support of my mom, Bonnie Brown, or my brother, Michael v Myers. I love you both so much. I had the great fortune of having an incredible group of graduate school friends. We learned, we laughed, we drank and we helped each other become political scientists. Thank you Judd Thornton, Bob Lupton, Davia Downey, Kris Dunn, and Shane Singh. Prashant Patel, my best friend, thank you for helping me believe that I could do this. Thank you, Jennifer Eisinger, for helping me cross the finish line. vi TABLE OF CONTENTS List of Tables……………………………………………………………………………..ix List of Figures…………………………………………………………………………….xi Chapter 1 Introduction……………………………………………………………………1 1.1 Challenges to Research……………………………………………………….4 1.2 Research Questions…………………………………………………………...6 1.3 Organization of the Dissertation……………………………………………...8 Chapter 2 Placing Native Title in Context………………………………………………12 2.1 Introduction………………………………………………………………….12 2.2 Aboriginal Australians and the Constitution: From Federation to the Racial Discrimination Act…………………………………………………………..14 2.3 Land Rights in Australia, the United States, Canada and New Zealand…….16 2.4 The Murray Islands Annexation…………………………………………….20 2.5 Establishing Native Title: Mabo v Queensland (No. 2)……………………..23 2.6 The Mason Court in Context………………………………………………...29 2.7 Shock and Surprise………………………………………………………….32 2.8 Wik Peoples v Queensland: Facts, Ruling and Reaction……………………40 2.9 Extinguishing Native Title: Ward, Anderson and Yorta Yorta……………..44 2.10 Counting Claims: Native Title from 1992 to 2012………………………...49 2.11 Conclusion…………………………………………………………………50 2.12 Tables and Figures…………………………………………………………52 Chapter 3 Probing the Conditionality of Support……………………………………….53 3.1 Introduction………………………………………………………………….53 3.2 The High Court of Australia and Native Title………………………………55 3.2.1 The Native Title Cases…………………………………………...58 3.3 (Under What Circumstances) Can Courts Influence the Public?....................65 3.3.1 Aggregate Level Opinion: The Positive Response Hypothesis…………………………………………………….….65 3.3.2 Aggregate Level Opinion: Data and Modeling Strategies……….68 3.3.3 Aggregate Level Opinion: Results……………………………….71 3.3.4 Individual Level Opinion: The Conditional Response Hypothesis………………………………………………………..73 3.3.5 Individual Level Opinion: Data and Modeling Strategies……….77 3.3.6 Individual Level Opinion: Results……………………………….81 3.4 Conclusion…………………………………………………………………..87 3.5 Tables and Figures…………………………………………………………..90 Chapter 4 Political Sophistication and Attitude Extremity……………………………...95 4.1 Introduction………………………………………………………………….95 4.2 The Psychological Foundations of Influence………………………………..97 4.2.1 The Role of Heuristics………………………………………….100 vii 4.2.2 The Need for Cognition: Political Sophistication………………103 4.3 Data and Modeling Strategies: Crystallization and Volatility Hypotheses…………………………………………………………………109 4.4 Results: Crystallization Hypothesis………………………………………..115 4.5 Results: Volatility Hypothesis……………………………………………..118 4.6 Discussion: Crystallization and Volatility Hypotheses…………………….121 4.7 Attitude Extremity…………………………………………………………122 4.8 Data and Modeling Strategy: Extremity and Moderation Hypotheses…….126 4.9 Results: Extremity and Moderation Hypotheses…………………………...130 4.10 Discussion: Extremity and Moderation Hypotheses……………………...133 4.11 Conclusion………………………………………………………………..134 4.12 Tables and Figures………………………………………………………..138 Chapter 5 Context and Place…………………………………………………………...152 5.1 Introduction………………………………………………………………...152 5.2 Perceiving the Threat of Native Title………………………………………153 5.3 Politicized Places…………………………………………………………..157 5.4 Contextual Proximity to a Case……………………………………………159 5.5 Data, Measurement and Model…………………………………………….161 5.6 Results……………………………………………………………………...165 5.7 Conclusion…………………………………………………………………168 5.8 Tables and Figures…………………………………………………………170 Chapter 6 Conclusion…………………………………………………………………...173 6.1 A Final Word on Judicial Retreat………………………………………….177 Appendices……………………………………………………………………………...179 Appendix A……………………………………………………………………..180 Appendix B……………………………………………………………………..185 References………………………………………………………………………………186 viii LIST OF TABLES Table 2.1 Native Title Claims by State and Territory, 1992 to 2012……………………52 Table 3.1 Aggregate Level Support for Transfer of Land Rights to Aborigines, 19902007………………………………………………………………………………………90 Table 4.1 High Sophisticates: Likelihood of Support for the Transfer of Land Rights to Aborigines, Pre- and Post-Mabo v. Queensland………………………………………..138 Table 4.2 High Sophisticates: Likelihood of Support for the Transfer of Land Rights to Aborigines, Pre- and Post-Wik Peoples v. Queensland………………………………...139 Table 4.3 High Sophisticates: Likelihood of Support for the Transfer of Land Rights to Aborigines, Pre- and Post-Ward, Anderson and Yorta Yorta (Trio)…………………...140 Table 4.4 Low Sophisticates: Likelihood of Support for the Transfer of Land Rights to Aborigines, Pre- and Post-Mabo v. Queensland………………………………………..141 Table 4.5 Low Sophisticates: Likelihood of Support for the Transfer of Land Rights to Aborigines, Pre- and Post-Wik Peoples v. Queensland……………………………..…142 Table 4.6 Low Sophisticates: Likelihood of Support for the Transfer of Land Rights to Aborigines, Pre- and Post-Ward, Anderson and Yorta Yorta (Trio)…………………...143 Table 4.7 High Sophisticates: Likelihood of an Extreme Attitude for the Transfer of Land Rights to Aborigines, Pre- and Post-Wik Peoples v. Queensland………………..144 Table 4.8 High Sophisticates: Likelihood of an Extreme Attitude for the Transfer of Land Rights to Aborigines, Pre- and Post- Ward, Anderson and Yorta Yorta (Trio)….145 Table 4.9 Low Sophisticates: Likelihood of an Extreme Attitude for the Transfer of Land Rights to Aborigines, Pre- and Post-Wik Peoples v. Queensland………………..146 Table 4.10 Low Sophisticates: Likelihood of an Extreme Attitude for the Transfer of Land Rights to Aborigines, Pre- and Post- Ward, Anderson and Yorta Yorta (Trio)….147 Table 5.1 Fixed-Effects Logistic Regression with Random-Intercepts………………..170 Table 1A Summary Statistics………………………………………………………….180 Table 1A Summary Statistics, Cont……………………………………………………181 Table 2A Likelihood of Support for the Transfer of Land Rights to Aborigines, Pre- and Post-Mabo v. Queensland………………………………………………………………182 ix Table 3A Likelihood of Support for the Transfer of Land Rights to Aborigines, Pre- and Post-Wik Peoples v. Queensland……………………………………………………….183 Table 4A Likelihood of Support for the Transfer of Land Rights to Aborigines, Pre- and Post-Ward, Anderson and Yorta Yorta (Trio)………………………………………….184 Table 1B Summary Statistics…………………………………………………………..185 x LIST OF FIGURES Figure 3.1 Percent Supporting the Transfer of Land Rights to Aborigines, 19902007………………………………………………………………………………………91 Figure 3.2 Likelihood of Support for the Transfer of Land Rights to Aborigines, Pre- and Post-Mabo v. Queensland………………………………………………………………..92 Figure 3.3 Likelihood of Support for the Transfer of Land Rights to Aborigines, Pre- and Post-Wik Peoples v. Queensland…………...……………………………………………93 Figure 3.4 Likelihood of Support for the Transfer of Land Rights to Aborigines, Pre- and Post-Ward, Anderson and Yorta Yorta (Trio)…………………………………………...94 Figure 4.1 Mean Support for the Transfer of Land Rights Between High and Low Sophisticates, 1990-2004……………………………………………………………….148 Figure 4.2 Mean Difference in Support for the Transfer of Land Rights Between High and Low Sophisticates, 1990-2004……………………………………………………..149 Figure 4.3 Mean Level of Attitude Extremity in Support for the Transfer of Land Rights Between High and Low Sophisticates, 1990-2004……………………………………..150 Figure 4.4 Mean Difference in Level of Attitude Extremity in Support for the Transfer of Land Rights Between High and Low Sophisticates, 1990-2004……………………….151 Figure 5.1 National Salience of Native Title…………………………………………..171 Figure 5.2 Predicted Random-Intercepts by State-Year……………………………….172 xi Chapter 1 Introduction The question of whether courts can convince the public of the wisdom of their decisions is of much greater importance than for other political institutions. As Alexander Hamilton declared in Federalist No. 78 that the “judiciary . . . has no influence over either the sword or the purse; no direction either of the strength or of the wealth of the society; and can take no active resolution whatever. It may truly be said to have neither force nor will, but merely judgment” (Hamilton 1966, 464). Later Mr. Justice Felix 1 Frankfurter, in his dissenting opinion in Baker v. Carr asserted, “The Court’s authority—possessed of neither the purse nor the sword—ultimately rests on sustained public confidence in its moral sanction. Such feeling much be nourished by the Court’s complete detachment, in fact and in appearance, from political entanglements and by abstention from injecting itself into the clash of political forces in political settlements.” These statements reflect the underlying reality of not just the Supreme Court of the United States, but democratic courts around the world: an institution with no formal mechanisms to guarantee enforcement of its decisions naturally depends on its ability to convince the other branches and the public of the wisdom of those judgments. The inherent limitation of the judiciary has been referred to as the ‘implementation problem’ (Rosenberg 2008). Unlike the popularly elected branches, courts cannot quickly respond 1 Baker v. Carr (1962) 369 U.S. 186. 1 to public disapproval as the judicial norm of stare decisis constrains decision-making (Bailey and Maltzman 2011) and prevents a court from reversing itself even if it were persuaded to do so by an unreceptive public. As De Tocqueville observed, “The power of the Supreme Court Justices is immense, but it is power springing from opinion. They are all-powerful so long as the people consent to obey the law; they can do nothing when the people scorn it” (De Tocqueville 2000). Without a natural constituency or effective enforcement powers courts are forced to rely on the rule of law, the reputation of the court and judges as impartial and fair, and the perception that their decisions as legitimate. Together these form the core of the judiciary’s symbolic powers, which are leveraged to influence the public (Gibson et al 1998; Gibson and Caldeira 2009). Gibson and Caldeira (2009) argue that courts are the beneficiaries of positivity bias whereby any exposure, both positive and negative, causes people to pay attention to courts and become exposed to the legitimizing myths surrounding courts and the rule of law. The lesson imparted is that the judiciary is different from the popularly elected branches and therefore more legitimate and trustworthy. Polls of the public have consistently shown the courts to be the most trusted branch of government (Hibbing and Theiss-Morse 1995). It may be in a court’s best interest to try to sway public opinion towards their position than to rely on rival political institutions. The courting of the public by the judiciary works to ensure the cooperation of the elected branches in the implementation process through an indirect mechanism: the electoral connection. Vanberg (2001, 347) suggests, “In democracies where a high court enjoys a high degree of public support, a legislative decision not to comply with a judicial ruling may result in a negative public backlash.” The fear of a public backlash may lead 2 to increased respect for both the courts as institutions and their decisions (Vanberg 2000). In this sense, the electoral connection (Mayhew 1974) can serve as an indirect mechanism for courts to ensure implementation of their decisions. Of course, the establishment of this mechanism assumes that either the public can monitor legislative responses to court actions reliably and effectively (Vanberg 2001) or the courts themselves can reach out to the public directly by selectively promoting their decisions (Staton 2006). Furthermore, in order to protect its reputation and build legitimacy within the public, courts must take into account the public’s mood when making decisions (McGuire and Stimson 2004; McGuire et al 2009; Mishler and Sheehan 1993) so as not to appear too far out of step and risk damaging that vital relationship. This does not relieve the courts of their obligations to make good law, even if their decisions prove to be unpopular (Bickel 1986) such as in contested presidential elections (Gibson et al 2003). Courts in democratic regimes are thus obliged to behave as “republican schoolmasters” who engage with and educate the public about the law through their opinions (Lerner 1967). It is through this dialogue that courts simultaneously fulfill their judicial function as well as engage the public as teachers of the law in a democratic society. If courts cannot affect public opinion through their decisions then they risk exacerbating an already weakened institutional position. However, the courts and the justices that fill their ranks have their own goals (Segal and Spaeth 2002) and sometimes this tension forces courts to make difficult and politically unpopular decisions. It is during these times that courts draw on their reputation and legitimacy or political capital (Bartels and Mutz 2009; Choper 1980; 3 Gibson et al 2003; Mondak 1992; 1994; Mondak and Smithey 1997) to coax the public to follow its lead. Though the effect of court decisions on public opinion has been relatively confined to the study of the Supreme Court in the United States (Franklin and Kosaki 1989; Hanley et al 2012; Hoekstra and Segal 1996; Johnson and Martin 1998; Marshall 1989; Persily et al 2008; Stoutenborough et al 2006) there is no a priori reason to suggest that other similarly situated courts around the world should or would behave any differently. If after all it is true that justices cannot help, but to try to influence the public in democratic regimes (Lerner 1967, 180) evidence of such a phenomenon should be seen in courts and countries around the world. 1.1 Challenges to Research Surprisingly though, scholars and researchers have been reluctant to explore whether courts in other countries have a similar relationship with their own citizens. This is the case for two reasons. First, the vast majority of judicial politics research has been focused on the United States and the Supreme Court of the United States, in particular, “despite its (potentially) decreasing importance, and continue[s] (with limited exceptions) to ignore courts abroad, despite their increasing prominence” (Epstein 1999, 1). Despite several prominent collections of studies providing useful backgrounds, descriptions and platforms for country-based and comparative judicial politics research (Jackson and Tate 1992; Jacob et al 1996; Russell and O’Brien 2001; Tate and Vallinder 1995) as well as newly created multi-country judicial decision-making databases (Haynie et al 2007), the judicial politics subfield is still slowly embracing the world outside of the United States. 4 This has had a deleterious affect on the generation of general theories of judicial behavior and the place of courts in democratic regimes. Too often the United States and its Supreme Court are viewed as unique and court scholars reject theoretical propositions found to be significant in the American setting as not relevant in other judicial systems despite common democratic and legal heritages. Secondly, an explanation for the absolute scarcity of studies on whether courts can influence public opinion in other countries is related to the common problems that have plagued researchers in the United States: survey research design. Caldeira (1991, 305) catalogues a number of these hurdles, which have served to generally dissuade researchers in the United States and abroad from wading too deep into these questions. An initial issue facing prospective researchers is that surveys rarely “pose questions to the public on cases before the court” (Caldeira 1991, 305). This may be the most important hurdle to conducting this research, as questions about the court cases must be asked in order for any kind of analysis to take place. Even in the American context, Marshall’s exhaustive study of Supreme Court influence over public opinion found only 18 matches in 45 years of predecision and postdecision surveys (Marshall 1989), which suggests just how high the barrier to conducting this research can be. To complicate matters further, question wording looms as a pressing concern. In many cases “the wording often does not match the issues at controversy” (Caldeira 1991, 305) or the question wording changes from one survey to the next thwarting before and after or longitudinal comparisons. The most prominent examples deal with the issue of 5 2 abortion. Additionally, the lags or amount of time between when surveys are conducted and the decision increases the possibility that alternative factors may influence opinion change. Lastly, “most issues before the court simply do not excite much public interest” (Caldeira 1991, 305). For example, Epstein and Segal (2000, 74) show that of the 6114 cases decided by the Supreme Court of the United States between 1946 and 1995 only 3 914 appeared on the front page of the New York Times , which indicates that most cases are not salient enough to garner national attention. 1.2 Research Questions The question that motivates this dissertation is under what conditions can courts influence the public? Initially, theory asserted that because courts were distinct institutions they were uniquely able to confer legitimacy on policy questions brought before them (Dahl 1957). The public was then thought to positively respond to and accept these judicial decisions as essentially correct. Exhaustive studies were carried out across a multitude of issue areas, but none uncovered evidence of court rulings gathering public support (Marshall 1989). If courts are able to influence the public through its decisions then the process is more complex than originally assumed. Later research uncovered that the public does indeed respond to decisions, but the sorting that takes place among social groups is masked in the aggregation of overall public opinion (Franklin and Kosaki 2 See Luks and Salamone (2008, 87-89) on the effects of question wording on survey responses towards the issue of abortion and Epstein et al (1994, 600) on how differences between judicial decisions and question wording on surveys impact the distribution of opinion. 3 Approximately 15% of Supreme Court decisions are considered salient or important enough to garner front-page national interest and attention. 6 1989). The sorting process that social groups undertake is restricted to instances when courts make a landmark ruling. Subsequent rulings in the same area do not cause social groups to sort because they have already moved in response to the first landmark decision (Johnson and Martin 1998). These theories, and their development, have solely taken place in the United States. One of the main contributions of this dissertation is to examine these theories of judicial influence on public opinion in a context outside of the United States and in Australia. Surprisingly, very little research has been conducted in countries outside of the United States. Fletcher and Howe (2000) carried out a descriptive examination of several decisions made by the Supreme Court of Canada on public opinion. Goot and Rowse (2007) report the public reaction to the High Court of Australia’s landmark decision on native title rights. However, neither of these studies moves beyond basic descriptions of public reaction and certainly do not engage in any kind of theoretical or empirical testing. This dissertation is the first to provide a systematic, theory driven, empirical evaluation of judicial decisions on public opinion outside of the United States and the first to do so in Australia. Furthermore, the psychological foundations of judicial influence on public opinion place a great deal of emphasis on the differences between low and high sophisticated individuals (Chaiken 1980; Petty and Cacioppo 1981). Typically, though, this distinction has been disregarded because the early surveys used to explore public opinion did not include measures that are used to differentiate between levels of sophistication. Later research did not capitalize on measures being included in surveys in order to stay consistent with previous research (Hanley et 2012). This dissertation will 7 explore the differences between low and high sophisticates and their responses to court decisions. Recognizing these differences allows for additional aspects of the psychological models of persuasion to be investigated. For example, past research suggests that public attitudes may polarize in response to judicial decisions, but it is unclear what drives attitudes toward the extremes. Theory suggests that high sophisticates will exhibit attitude extremity (Taber and Lodge 2006) in response to subsequent judicial decisions, which may clarify why polarization is observed. Lastly, research on court influence of local opinion, where cases originate, has uncovered that geographic proximity affects opinion change (Hoekstra 2003; Hoekstra and Segal 1996). This dissertation develops a theory based on geographic proximity and the politicization of those environments (Hopkins 2010) that helps people connect their everyday experiences to judicial decisions. The sociodemographic makeup of those environments will help or hinder judicial influence of opinion. 1.3 Organization of the Dissertation The following chapter, Chapter 2, provides the necessary historical background to understand the importance and ramifications of the High Court’s native title decisions. The discussion begins at federation in 1901, continues to the 1967 referendum and concludes with the adoption of the Racial Discrimination Act in 1975. I then discuss the legal history of land rights for Indigenous peoples in Australia and compare it to developments in other settler countries with Indigenous populations. The legislative action that led Eddie Mabo to assert a right to native title and launch a legal challenge is discussed in detail. Thorough descriptions of the background and facts of each case, 8 judicial opinions and reactions to each of the cases are taken in turn. A total of five cases were selected and represent the core of the Court’s jurisprudence in this issue area over the span of ten years. The first case was immediately deemed to be a “judicial revolution” (Russell 2005, 5) and recognized a claim of native title to the island of Mer by Australian Aborigines as part of the common law. The second case arrived four short years later and aimed to clarify native title rights in regards to pastoral leases, but, most importantly, brought native title to mainland Australia. Six years later a trio of cases was heard by the High Court in the later part of 2002 that would substantially define the potential and limits of native title. These decisions had the effect of defining native title as a bundle of rights as compared to exclusive possession by bringing greater clarity to the rules by which native title could be claimed or extinguished. These cases represent the most important rulings issued by the High Court in the area of native title. The chapter concludes with an assessment of native title claims in Australia and their differential impact across the states and territories. Studies of the relationship between American public opinion and decisions of the U.S. Supreme Court have provided support for the theory that salient court decisions can and do have an impact on public opinion. However, these theories have not been tested in other societal, political or legal contexts. This first empirical chapter, Chapter 3, addresses a critical theoretical question: Is the United States Supreme Court unique in its ability to persuade the mass public through its decisions? The similarities between the Australia High Court and the United States Supreme Court, as well as, the cultural and political similarities provides a comparative context well suited for exploring the generalizability of court influence on public opinion. The empirical analysis begins with 9 an aggregate level time-series analysis of the Court’s ability to influence overall public opinion consistent with a theory of legitimation. Next, an individual level analysis is undertaken where the validity of Johnson and Martin’s conditional response hypothesis is assessed. The results show that the Australian High Court was able to affect the structure of opinion after the initial ruling, but not with subsequent rulings, which is consistent with conditional response theory. Testing these theories in a foreign political and legal context provides confirmatory evidence that a general characteristic of not only the Supreme Court of the United States or the High Court of Australia, but of high courts in democracies around the world is being observed. The second empirical chapter, Chapter 4, is focused on the psychological foundations that underlie how attitude change or persuasion takes places in response to judicial decisions. While previous research on court influence of public opinion relies on the Elaboration Likelihood Model, a leading social-psychological model of persuasion, the studies often give short shrift to the critical distinction between levels of political sophistication emphasized in this literature. This is the result of wanting to be consistent with earlier research that suffered from the absence of survey questions associated with political sophistication such as political interest, education level and media consumption. Separate empirical analyses are conducted based on level of political sophistication, which indicates that high and low sophisticates respond differently to landmark and subsequent judicial decisions. The next section empirically assesses how differing levels of political sophistication leads to attitude extremity or moderation in response to both initial and subsequent decisions. 10 The third and final empirical chapter, Chapter 5, considers how contextual factors particular to individual environments affect the court’s ability to influence public opinion. In particular, the geographic proximity to the origin of the cases may serve as an impediment to judicial influence of public opinion. Individuals who are closer to where the case originated may be more likely to identify with the people and places involved and will be more likely to have a personal stake in the outcome, which reduces the likelihood of judicial influence. While geographic proximity may provide an important glimpse into the role of place in moderating the influence of court decisions on public opinion, there is good reason to explore and probe further into whether it is the place itself or its context that tempers judicial influence. An important and growing body of empirical research suggests that when an issue becomes nationally, individuals will become more aware of the politicized context of the environment in which they live and draw conclusions about that issue from their experiences. An empirical analysis is undertaken to assess whether the presence of salient sociodemographic changes affect judicial influence on public opinion. The concluding chapter, Chapter 6, summarizes the main findings and discusses their meaning in relation to the High Court’s role in the Australian political system. The discussion focuses particularly on the implications of the court as a political actor in a system and legal tradition that eschews an active judicial role for the courts. The implications for the court and its standing with the public is further elaborated upon. 11 Chapter 2 Placing Native Title in Context “Until we give back to the black man just a bit of what was his and give it back without provisos, without strings to snatch it back without anything but complete generosity of spirit in concession for the evil we have done to him – until we do that, we shall remain what we have been so far: a people without integrity, not a nation, but a community of 1 thieves.” 2.1 Introduction The previous chapter introduced the general research question that is investigated throughout this dissertation: under what conditions can courts influence the public? The three sets of cases, five in total, used to investigate this larger question represent the core of the High Court’s jurisprudence in the area of native title land rights for Aboriginal Australians. Native title refers to Indigenous people, in a settler country, having claim to and ownership over their traditional lands. Countries like the United States, Canada and New Zealand have long accepted the principle of native title and have, to varying degrees, over time, fully integrated and acknowledged the rights of Indigenous populations to their land. Australia, however, had historically denied not only native title to the Aboriginal population, but basic rights as well. Why would decisions involving property and land law be of such significance to attract public attention let alone 1 Quote attributed to Xavier Herbert, a white Australian author, displayed at the entrance to Oyster Cove, a sacred Aboriginal site in Tasmania. 12 influence public opinion? The issue of native title may seem an unusual choice to investigate the influence of court decisions on public opinion since other studies have focused on issues like abortion, the death penalty and gay rights. In Australia, few issues can be considered as fundamentally important as native title. The first of these cases, Mabo (No. 2), was simply groundbreaking. The High Court discarded two hundred years of legal doctrine to declare native title existed on the island of Mer. It is not hyperbole to suggest that the case touched off a national crisis. The second case, Wik, sent similar shockwaves through Australia by bringing native title to the mainland. The decision seemingly touched off a constitutional crisis with political leaders questioning the very legitimacy of the High Court. The third set of cases, three in total, redefined native title as bundle of rights instead of in terms of exclusive possession of land. The purpose of this chapter is to provide the necessary historical background to understand the importance and ramifications of the High Court’s native title decisions. The discussion begins at federation in 1901, continues to the 1967 referendum and concludes with the adoption of the Racial Discrimination Act in 1975. I then discuss the legal history of land rights for Indigenous peoples in Australia and compare it to developments in other settler countries with Indigenous populations. The legislative action that led Eddie Mabo to assert a right to native title and launch a legal challenge is discussed in detail. Thorough descriptions of the background and facts of each case, judicial opinions and reactions to each of the cases are taken in turn. The chapter concludes with an assessment of native title claims in Australia and their differential impact across the states and territories. The historical descriptions, accounts of the cases 13 and reaction that follow provide critical contextual information relevant to understanding the challenges the High Court faced in influencing the public attitudes on the issue of native title – the topics of Chapters 3, 4 and 5. 2.2 Aboriginal Australians and the Constitution: From Federation to the Racial Discrimination Act When the Australian Federation was formed in 1901, the new constitution contained two provisions that effectively removed Aboriginal Australians from political life in Australia. One provision, Section 127, said: “In reckoning the numbers of the people of the Commonwealth, or of a State or other part of the Commonwealth, aboriginal natives shall not be counted.” The removal of Aboriginal Australians from official tabulations of the population was meant to prevent states such as Queensland and Western Australia from using their large Aboriginal populations to gain additional seats in Parliament or 2 extra funds from Commonwealth grants (Smith 1980). The other provision, Section 51(xxvi), enabled the Parliament to make laws for “[t]he people of any race, other than the aboriginal race.” Aborigines, unable to be counted as citizens, were similarly denied 3 the franchise and unable to exercise even the most basic democratic rights and responsibilities. 2 The irony of the Australian constitution being modeled closely after the United States constitution, which had its own debate about how to limit the effect of the black population on the allocation of seats in the House of Representatives, is not lost here. 3 Section 41 of the Constitution provided Commonwealth voting rights to those who previously held voting rights in one of the states or territories. The Commonwealth Franchise Act of 1902 stated that no Aboriginal shall be entitled to vote unless they were previously registered to vote in a state or territory before ratification. Aborigines gained 14 It was not until the 1967 national referendum that voters agreed to remove two clauses from the Constitution discriminated against Aborigines: Section 51(xxvi) and Section 127. The referendum garnered over ninety percent support among the voters, “an outstanding expression of sentiment about Aborigines” (Goot and Rowse 2007, 27). The practical effect would allow the Commonwealth to begin implement policies beneficial to Aboriginal Australians, which were thought to speed their assimilation into Australian society (Goot and Rowse 2007). The Commonwealth could now enact legislation curbing discriminatory practices, financial and welfare assistance as well as the preservation of cultural heritage. The most prominent piece of legislation arising from the newly empowered Whitlam Labor government was the Racial Discrimination Act 1975 (RDA). The statute rose out of Australia’s commitment to the United Nations’ International Convention on the Elimination of all Forms of Racial Discrimination, which calls on members to eliminate racial discrimination and promote understanding among all races. The RDA effectively made racial discrimination unlawful in Australia and nullified inconsistent legislation from states and territories. The law reached into areas such as employment (section 15), land or housing (section 12), provision of goods and services (section 13), access to places and facilities (section 11), joining a trade union (section 14) and job advertisements (section 16). Racial discrimination thus occurs when someone is treated less fairly compared to someone else in a similar situation because of his or her race, color or national or ethnic origin. the right to vote in Commonwealth elections regardless of their voting rights at the state or territory level in 1962 when the Commonwealth Electoral Act was passed. 15 The validity of the RDA was challenged when Queensland Premier Joh BjelkePetersen sought to block a group of Aborigines from buying a large tract of land in northern Queensland. John Koowarta, one of the Australian Aborigines involved in the land sale, complained to the Human Rights and Equal Opportunity Commission, which upheld his complaint against the government of Queensland. Queensland appealed the decision to the High Court on the grounds that the Commonwealth did not have the power to enact the Racial Discrimination Act under Section 51(xxvi). The High Court affirmed the Commonwealth’s power to enact the legislation under the external affairs power in Section 51(xxix), which enables the Commonwealth to engage in and make 4 treaties and to generally conduct foreign affairs. However, the lands were never sold to Koowarta and his group because Premier Bjelke-Petersen, after the case, declared the land in question to be a national park. Nevertheless, the constitutionality of the Racial Discrimination Act would prove to be a formidable ally for Aborigines in future court cases in which the principle concern was land rights. 2.3 Land Rights in Australia, the United States, Canada and New Zealand Native title or land ownership by indigenous peoples was generally accepted to not exist in Australia. The land of Australia was terra nullius or an empty land belonging to no th one when European settlement occurred in the late 18 century. This view was backed up by significant English legal authority and was endorsed by the Judicial Committee of 4 Koowarta v Bjelke-Petersen (1982) HCA 27 16 5 the Privy Council in 1889. However, the issue or question of native title had only been addressed directly once by the Australian courts, by the Supreme Court of the Northern 6 Territory, in the Gove Land Rights case. In the late 1960’s, Nabalco, a mining consortium of Swiss and Australian companies, secured a twelve-year bauxite mining lease from the Commonwealth in the Gove Peninsula in the northeast part of the Northern Territory. The Yolngu people, living in Yirrkala on the Gove Peninsula, petitioned the Commonwealth to prevent the mining leases from entering operation until there had been appropriate consultation with Aboriginal leaders. The Yolngu were unsuccessful in their attempts to have a House of Representatives Committee hear their claims. As a result, the Yolngu turned to the courts and sued the mining consortium, Nabalco, and the Commonwealth on the basis that they maintained native title rights over the land. Justice Richard Blackburn of the Supreme Court of the Northern Territory decided against the Yolngu on the grounds that native title was not part of the common law in Australia and even if it had existed it would likely have been extinguished by now. For the native title to exist under common law it would have to be expressly recognized under statutory provision. Additionally, claiming a connection to the land today does prove that the Yolngu held the same links to the same areas of land as their ancestors did. Justice Blackburn went on to explain that pre-existing interests or claims to land were not recognized unless they were rights grounded in a strict property rights perspective, which given the culturally bounded communal relationship Aborigines have with the land it was 5 6 Cooper v Stuart (1889) 14 App Cas 286 Milirrpum v Nabalco (1971) 17 FLR 141 17 impossible to argue ownership (Mercer 1997, 7). This reasoning led some to label the judgment an example of ‘judicial racism’ because of its dismissal of communal property rights in favor of individual property rights (Cunneen 1992). However, Justice Blackburn did admit a long period of Aboriginal occupation of the area and an elaborate system of native laws and customs, but in light of the Privy Council’s ruling in Cooper v Stuart, he was compelled to deny recognition of native title based on legal precedent, not historical fact. In other words, irrespective of the individual or communal nature of the claim, the British assertion of sovereignty over Australia extinguished all such claims (Patapan 2000, 110). The legal status of indigenous land rights in Australia differed drastically in comparison to the United States, Canada and New Zealand. Both the Congress and the Supreme Court of the United States have consistently recognized Indian tribes as political bodies with powers of self-government, which included sovereignty over their members and territories. Since the beginning of white settlement, numerous treaties were concluded between U.S. governments and Indian nations, and the Constitution grants Congress the power to regulate commerce with Indian tribes (Patapan 2000, 111). In Canada the Royal Proclamation of 1763 recognized Indian ownership of lands, which led to numerous treaties between the Crown and Aboriginal peoples (Russell 2005, 273). In 1973 the Supreme Court of Canada, for the first time, acknowledged that aboriginal title to land existed prior to colonization and was not derived wholly from 7 statutory law. The case, Calder v Attorney-General of British Columbia , arose when Frank Calder and the Nisga’a Nationl Tribal Council brought action against the 7 Calder v Attorney-General of British Columbia (1973) SCR 313 18 government of British Columbia seeking a declaration of aboriginal title to lands in the province that had not been lawfully extinguished. The Supreme Court found that native title continued to exist even in a settled colony, but was split over whether the particular claim was valid. In particular, three of the Canadian justices criticized the propositions of law outlined by Justice Blackburn in the Gove Land Rights case as ‘wholly wrong’ (Patapan 2000, 111). The Canadian Supreme Court went on to recognize a fiduciary duty 8 owed by the Crown to the first nations in the Guerin case . These obligations were given greater recognition when the Canadian Constitution was repatriated and the Charter of Rights and Freedoms became entrenched, which had the effect of making native title rights constitutional rights (Russell 2005, 274). Similar developments took place in New Zealand. The sovereignty of the Maoris had been recognized when the Treaty of Waitangi was signed in 1840 and in 1975, the Treaty was given force when it was implemented in the Treaty of Waitangi Act (Patapan 2000, 111; Russell 2005, 272-273). Since the 1987 landmark decision by the Supreme 9 Court of New Zealand in New Zealand Maori Council the principles of the Treaty were to be applied in the interpretation of legislation. The difference between the Aboriginal experiences in the United States, Canada and New Zealand compared to Australia can be easily traced back to the early recognition of the rights of native populations. In all of the countries above, except Australia, white settlers entered into treaties with native populations, which resulted in long-standing legal obligations and responsibilities governing interactions between the groups. Perhaps if the 8 9 Guerin v The Queen (1984) 2 SCR 335 New Zealand Maori Council v Attorney-General (1987) 1 NZLR 641 19 early Australian settlers had entered into similar agreements with the Aboriginal population then the courts and the common law would have been compelled under the recognition doctrine to accept claims of native title. In fact, Justice Blackburn made just this assertion in the Gove Land Rights case; however, history did not unfold in this manner and the Aboriginal people of Australia were seemingly left without a legal remedy for the injustice they suffered under and since colonization. 2.4 The Murray Islands Annexation In 1982, Eddie Mabo, along with James Rice and Reverend David Passi, filed a statement of claim that sought a ruling from the High Court on the question of whether the island of Mer, the largest of three islands making up the eastern Torres Strait Murray Island group, was subject to native title. The Torres Strait Islands are a group of 274 small islands, of which 17 are inhabited, that separates the 170-kilometre-wide strait between Australia and Papua New Guinea (Russell 2005, 19). The statement of claim was a direct reaction to Queensland’s plan to repeal the 1971 Torres Strait Islanders Act and remove the reserve status of the islands, which would have the effect of treating the islands “as lands totally at the disposal of the Queensland government” (Russell 2005, 194). The Act would effectively allow Queensland to relocate islanders at will as well as restrict fishing rights and offered the people of the Murray Islands a limited form of land title, a “deed of grant in trust” (Mercer 1997, 10). Mabo and the other plaintiffs argued that the offer made by the government was inappropriate because Queensland had never extinguished native title despite the 1879 annexation of the Murray Islands. In fact, the Meriam people, numbering around 400, had had uninterrupted use and ownership of their land. 20 Legal maneuvering, and the delays that come from it, allowed the government of Queensland to pass the Queensland Coast Islands Declaratory Act 1985. This act was meant to counter and clarify that Queensland’s intentions in 1879 were not only to acquire sovereignty over the Murray Islands, but also to extinguish any pre-existing Meriam islander land rights (Mercer 1997, 10). In other words, the annexation merely asserted sovereignty, but did not explicitly state the government’s intentions to abolish the rights of the native islander populations; the act was intended to retrospectively abolish any and all native title rights. This action opened the door to native title. As a result of the Queensland Coast Islands Declaratory Act 1985, Mabo, along with his previous plaintiffs, challenged the validity of the Act and eventually found relief at the High Court. The crux of their case was that the act was invalid because it was contrary to the Racial Discrimination Act 1975. Mabo argued that the Coast Islands Declaratory Act had the effect of removing or limiting rights that are held only by a certain race or ethnic group, which the RDA contravened. Coupled with the assertion that acts of Parliament are superior to those of state governments from Section 109 of the Constitution, the RDA would therefore provide remedy and invalidate the Coast Islands Declaratory Act. Queensland, on the other hand, predictably, argued that the Coast Islands Declaratory Act was valid that had the effect of extinguishing any rights that the plaintiffs may have possessed or may have survived annexation. Interestingly, the case proceeded on the basis of an assumption: the existence of native title rights. In and of itself, this turn of events was remarkable. The opinion of Justice Blackburn in the Gove Land Rights case was quite emphatic in its assertion that native title land rights could not exist in Australia. Both parties agreed the case should 21 proceed on this assumption though the issue of native title would not be addressed in the case. In fact, the High Court agreed that the Coast Islands Declaratory Act did effectively extinguish native title rights, if native title did exist, but the question turned on the validity of the Act. But why would the High Court and the government of Queensland agree to the assumption of native title? The islanders made a persuasive case. Unlike the Yolngu in the Gove Land Rights case, the Meriam people had lived and continued to live, for centuries, in settled, named villages along the coastlines on the islands (Mercer 1997, 10). They cultivated inland gardens, which were associated with particular individuals; the rights were passed on patrilineally and protected by strict religious rules known as Malo’s Law (Mercer 1997, 10; Russell 2005, 199). Not only had the Meriam continued to inhabit their ancestral lands, but they had maintained ancestral laws governing those lands. Perhaps, even more importantly, from a strict property rights point of view, the Murray Island group were the only Torres Strait islands not covered by deeds of grant in trust (Mercer 1997, 10), which would be interpreted as extinguishment of native title. By the standards of Justice Blackburn as well as other common law courts around the world, the Meriam people could convincingly demonstrate a claim to native title. Still, this case was about the validity of an act of the government of Queensland that, if valid, would extinguish native title regardless of whether its claimants could make a strong case. The High Court rendered its judgment in December 1988 in Mabo v Queensland 10 (No 1) . The judgment was four to one in favor of Eddie Mabo. The majority, consisting of Justices Brennan, Toohey, Gaudron and Deane found that if native title 10 Mabo v Queensland (No. 1) (1988) 166 C.L.R. 186 22 rights did exist it should be treated as a broader human right to own and inherit property. The Coast Islands Declaratory Act had the effect of arbitrarily depriving the Meriam people of their rights to hold property. The intent of the Act was to eradicate property rights grounded in Meriam law while treating rights to property under Queensland law quite differently (Mercer 1997, 10). Therefore, the Act had no force because it contravened the RDA. Mabo (No. 1) was a narrow victory for the islanders. Queensland’s effort at “a blanket extinguishment of native title rights had failed by one vote” (Russell 2005, 212). The opinions of the justices, however, did acknowledge that native title, though not proved to exist in this case, was vulnerable to the authority of the Commonwealth, which could “set aside or qualify the RDA in order to extinguish land rights” (Russell 2005, 213). The question of whether native title survived colonization and remained intact at present was yet to be determined. 2.5 Establishing Native Title: Mabo v Queensland (No. 2) On June 3, 1992 the High Court of Australia announced their ruling in Mabo v. 11 Queensland (No. 2) with a six to one majority in favor of recognizing native title as 12 part of the common law. The court rejected the idea that “Australia was a terra nullius when the white man arrived, an idea that had been the foundation of Australian law and policy for nearly two centuries, Mabo appeared to be a judicial revolution” (Russell 2005, 5). The High Court ruled that the Meriam people of Murray Island, living in permanent 11 12 Mabo v. Queensland (No. 2) (1992), 175 C.L.R. 1 Literally translated as land of no one (Russell 2005, 5). 23 communities with social and political organization, had continuously and exclusively inhabited the Island and its surrounding islands and reefs. The High Court recognized that a form of indigenous title to land might continue to exist in areas where indigenous people still occupied and could display a continuing association with their traditional land. However, where there was possible conflict with non-indigenous interests it would be the rights of the native title holders that would yield. How did the High Court arrive at such a conclusion? The major lines of contention across the opinions were centered around, first, whether native title could survive annexation by the Crown; second, whether native title could be extinguished without consent; third, the issue of compensation and finally, bringing native title into the common law. The majority consisted of Justices Brennan, McHugh, Toohey, Deane and Gaudron, along with Chief Justice Mason with Justice Dawson as the lone dissenter. However, as is the case with seriatim opinion writing, the majority really consisted of three distinct sets of opinions or factions: Justice Brennan was joined by Justice McHugh and Chief Justice Mason, Justices Deane and Gaudron wrote jointly, and Justice Toohey authored his own opinion. The key point between the majority opinions and Dawson, the lone dissenter, concerned whether land rights could survive the British settlement, colonization and annexation of indigenous territories. The justices, in the majority, found that according to English law, native rights to land could and did survive the establishment of British sovereignty in Australia (Russell 2005, 250). The core of this argument lies in the distinction between the Crown’s ownership of the land and radical title of the Crown. Racial title refers to sovereign political authority, but is not equivalent to having absolute 24 ownership. Justice Brennan’s opinion cites numerous sources that distinguish between 13 land being held by the Crown and being owned by the Crown. 14 equating sovereignty and ownership of law” It was this “fallacy of that explains why it was assumed that native title was automatically extinguished when Britain exerted sovereign authority over Australia. In fact, all three of the majority opinions cited strong scholarly backing for the notion that under the constitution of the British Empire, declared by its highest judicial authorities, sovereignty did not necessarily involve acquisition, ownership or title to all land in the territory (Russell 2005, 252). Justice Dawson, like Justice Blackburn in the Gove Land Rights case, relied much more heavily on recognition doctrine. For Justice Dawson, land rights of indigenous peoples stemmed only from recognition from imperial or colonial authorities; there was no legal obligation to recognize indigenous land ownership. Recognition had to be explicit in terms of a treaty or policies that allowed indigenous occupation of land because the notion of property rights did not exist prior to British rule (Russell 2005, 252). In other words, because British colonial authorities did not recognize or accept preexisting indigenous land rights at the time of settlement there was no reason to suggest land rights existed among the indigenous population. Now that the majority had found that native title could survive annexation, the question became whether actions by the Crown and eventually the Commonwealth government had the effect of extinguishing those land rights. Furthermore, if native title had been extinguished would the indigenous population be entitled to compensation? 13 14 Mabo v. Queensland (No. 2) (1992), 175 C.L.R. 1 at 45 Mabo v. Queensland (No. 2) (1992), 175 C.L.R. 1 at 51 25 Justice Brennan’s judgment succinctly asserted that native title could be extinguished without consent and without compensation. His reasoning distinguished native title from other property rights under the common law in that native title could be extinguished by any government action, such as granting a lease or erecting public works, that is inconsistent with indigenous occupation and use of the land (Russell 2005, 258). According to this line of reasoning, “Aborigines were dispossessed of their land parcel by parcel, to make way for expanding colonial settlement. Their dispossession underwrote 15 the development of the nation”. Justice Brennan does admit that this process has cost many Aborigines their claims to native title. However, in the Murray Islands, “the Crown has alienated only part of the land and has not acquired for itself the beneficial ownership of any substantial part.” He goes on to suggest that “there may be other areas of Australia where native title has not been extinguished and where an Aboriginal people, 16 maintaining their identity and their customs, are entitled to enjoy their native title”. Justice Brennan’s opinion, joined by Chief Justice Mason and Justice McHugh, did make one important caveat in regards to the question of compensation and that dealt with the Racial Discrimination Act 1975. In Mabo (No. 1), the High Court had not yet determined what the Murray Islander’s land rights were, but there was a view towards a fundamental or universal right to own property and protection against being arbitrarily deprived of that property. The RDA entitled just compensation to native title holders who had lost their land through arbitrary dispossession since the act came into force in 1975, but would do nothing for dispossession that occurred prior (Russell 2005, 260). 15 16 Mabo v. Queensland (No. 2) (1992), 175 C.L.R. 1 at 68-69 Mabo v. Queensland (No. 2) (1992), 175 C.L.R. 1 at 69 26 Justices Deane and Gaudron, on the other hand, considered the massive process of coercive dispossession to be a violation of the native title holders’ rights (Russell 2005, 262). The justices lent much more weight to the normal way of acquiring land that occurred through the operation of treaties and agreements in other settler countries, such as New Zealand, Canada and the United States (Russell 2005, 261). They suggest that if native title rights have been “wrongfully extinguished without clear and ambiguous statutory authorization” then “proceedings for compensatory damages” should be pursued. 17 The fact that common law native title was denied by administrators and judges sanctioned the erroneous belief that Australia was terra nullius and implicated the court as being complicit in the dispossession of Aboriginal Australians (Patapan 2000, 126). Patapan argues that this act of contrition by the justices serves a number of purposes because it identifies who was responsible for the injustice, acknowledges gains made by the price of injustice and the proper response to injustice, which calls for shame and atonement (2000, 126). Justices Deane and Gaudron conclude this section of their opinion by asserting, “the acts and events by which that dispossession in legal theory was carried into practical effect constitute the darkest aspect of the history of this nation. The nation as a whole must remain diminished unless and until there is an acknowledgment 18 of, and retreat from, those past injustices”. Though the majority clearly disagreed on the issue of compensation they had yet to define native title as a legal concept within the common law. Was native title a title to land ownership? Or was it a distinct and separate concept from traditional property 17 18 Mabo v. Queensland (No. 2) (1992), 175 C.L.R. 1 at 119 Mabo v. Queensland (No. 2) (1992), 175 C.L.R. 1 at 109 27 rights? Russell (2005, 265) suggests that native title is a bridge or legal connector that may not recognize title to land ownership, but does recognize the traditional connection of native peoples to their land. This traditional connection is defined in by its communal nature and an inherent recognition that the holder of native title cannot be an individual, but an indigenous community (Russell 2005, 266). The majority of the justices recognized that common law native title could not be easily integrated or assimilated with other forms of property and, as a result, stressed its distinctive form and limits (Russell 2005, 267). Both the opinions by Brennan and Deane and Gaudron stressed that owners 19 of native title could not trade, sell or surrender the land unless it was to the Crown, which is a right also claimed in the United States, New Zealand and Canada (Russell 2005, 268). The crucial and defining feature of native title, for the majority, was an identifiable indigenous community living on land governed by its own system of law and maintaining that connection since colonization. Of central importance was the recognition that an indigenous people does not lose its title by modifying its laws and customs, but its title would be extinguished if its members stopped living on their traditional lands (Russell 2005, 269). According to Justice Brennan, once native title is 20 extinguished then ownership is taken over by the Crown and can never be restored. Justices Deane, Gaudron and Toohey were even more emphatic on this point: continuous 21 physical occupancy was required in order to avoid extinguishment of native title. 19 20 21 Mabo v. Queensland (No. 2) (1992), 175 C.L.R. 1 at 70 and 110, respectively. Mabo v. Queensland (No. 2) (1992), 175 C.L.R. 1 at 60 Mabo v. Queensland (No. 2) (1992), 175 C.L.R. 1 at 110 28 These conditions were the distinguishing factor between the native title claim made by the Yolngu people in the Gove Land Rights case and the claim made by the Murray Islanders. Claiming native title would prove to be much more difficult given the stringent requirements needed to prove that native title had not been extinguished. For those Aboriginal groups that had been pushed from their lands with the advance of frontier settlement there would be no land and no ability to receive compensation for this coerced dispossession. Mabo (No. 2) was a victory for Eddie Mabo and the Murray Islanders, but it was a much more limited and even hollow victory for other Aboriginal peoples on mainland Australia. 2.6 The Mason Court in Context The High Court’s decision in the Mabo (No. 2) case was unusual because it was discovering a principle of justice embedded in the common law that had, for centuries, lain dormant. Why now? Why, in this case, was the majority willing to infuse the common law with contemporary international norms regarding native peoples? Why did this court shed its traditional, conservative role in favor of an activist political institution? To answer these questions, we must consider the historical role of the High Court and how that role would evolve in the growing global expansion of judicial power. The Australian constitution and the High Court were modeled closely after its counterparts in the United States because it was the most prominent example of a successful federalist system, which was a key factor in the debates leading up to independence (Patapan 2000; Solomon 1999; Williams 2001). Though the court has the power of judicial review it has never been considered a powerful political institution 29 (Soloman 1999; Williams 2001). We can partially attribute this to the fact that, historically, the court has viewed its constitutional role as circumscribed. Indeed, the judicial orthodoxy of the High Court views its role as one of ‘policing boundaries of state and federal powers and not passing judgment on the substantive value or wisdom of state or federal law’ (Pierce 2006, 71). Explanations abound for the court’s lack of 22 assertiveness including the absence of a bill of rights , prevalence of federalism disputes 23 or even the influence of the Privy Council . Perhaps the most important explanation is the entrenchment of the principle of parliamentary sovereignty into the judicial orthodoxy. Since Australia inherited its common law system from Britain, the premise that parliament exercises unlimited legislative authority that cannot be challenged by a judicial authority (Dicey 1982) became ingrained in the orthodoxy of the Australian judiciary (Pierce 2006, 63). There is a clear expectation that the court is expected to defer to parliament because the courts are subordinate to parliament’s will (Pierce 2006, 64). Judicial deference to the commonwealth asks judges to “refrain from second guessing the merits of parliamentary acts” (Pierce 2006, 65) and having a “greater faith in the elected branches” (Pierce 2006, 66) than the judiciary as best able to promote justice. This traditional view of the High Court and its role is distinctly at odds with its 22 Without an individual rights doctrine in the Australian constitution, litigants do not have the legal foundation for arguing civil rights and liberties cases before the High Court (Pierce 2006, 71). 23 Until 1986, the Judicial Committee of the Privy Council in the United Kingdom could hear appeals from the High Court of Australia. Smyth (2005) demonstrates that the likelihood of dissenting opinions increased after the break from the Privy Council indicating decision-making was influenced by its presence. 30 24 jurisprudence under the leadership of Chief Justice Anthony Mason , but also must be seen within a global political context in which several Anglo-derived judiciaries were undergoing or had undergone institutional transformations (see Tate and Vallinder 1997). In 1982, Canada enacted the Charter of Rights and Freedoms, which constitutionally protected a number of individual rights and, as a result, fundamentally altered the Supreme Court’s role in Canadian politics (Manfredi 2001). In 1990, New Zealand’s parliament enacted the Bill of Rights Act, which, like the Canadian Charter, granted a number of rights and freedoms to its citizens. As a result of these reforms, both the Supreme Courts of Canada and New Zealand were fundamentally transformed by their new powers to enforce rights protections. In 1988, a Constitutional Commission, with the support of Chief Justice Mason, recommended that Australia adopt an entrenched bill of rights similar to Canada, but failed when put to popular referendum (Galligan and Russell 1995, 95). Undeterred, the High Court began to find a number of implied rights in the Constitution. These included 25 prohibiting discrimination against citizens of another state , regulation of political 26 27 advertising , freedom of communication and native title rights for Australian 28 Aborigines . Chief Justice Mason remarked that in Australia, “it is the function of the Courts to keep the principles of law up to date” (Galligan and Russell 1995, 96). Far from being benign, this statement clearly signaled the High Court’s willingness to reject 24 25 26 27 28 Chief Justice Mason held the position from 1987 through 1995. Street v. Queensland Bar Association (1989) 168 C.L.R. 461 Australian Capital Television Pty Ltd v. Commonwealth (1992) 177 C.L.R. 106 Nationwide News Pty Ltd v. Wills (1992) 177 C.L.R. 1 Mabo v. Queensland (No. 2) (1992), 175 C.L.R. 1 31 its traditional role and bring Australian jurisprudence in line with its contemporaries. The difference between the role transformations in Canada and New Zealand compared to Australia is that the source of change lay outside of the judiciary with the legislatures in each country adopting new rights measures and vesting courts with new responsibilities (Pierce 2006, 13). However, in Australia, no new powers were granted to the High Court and no bill of rights or legislation expanded judicial review; the transformation was internally driven (Pierce 2006, 13). “Rather, a group of entrepreneurial High Court judges embarked, on their own volition, to redefine their institution’s mission” (Pierce 2006, 13). Though the High Court was originally modeled after the United States Supreme Court it has historically shied away entering into the kinds of political frays axiomatic of its counterpart. This changed under the Mason Court. 2.7 Shock and Surprise Far from being a case, or set of cases, about simple land rights, the High Court’s Mabo decisions “reveal a profound engagement with the major themes of civilization and culture, race and identity, justice and history” (Patapan 2000, 144). The High Court’s decision challenged the Australian people to confront the reality that their country was not just a society formed by British settlers that non-British immigrants and native peoples were allowed join, but instead is a society that flourished based on the use of force and the oppression of ancient societies with many modern day descendants (Russell 2005, 6). Prime Minister Keating acknowledged this debt at a widely reported speech at Redfern, Sydney, to launch the International Year of Indigenous Peoples some six 32 29 months after the Mabo (No. 2) decision. He asked white Australians to imagine what the world would look like from the Aboriginal perspective and also spoke of the ‘bizarre conceit that this continent had no owners prior to the settlement of Europeans’. He emphasized that the Mabo decision had established a solid foundation for justice. Despite the best intentions of the High Court to right an historical injustice the wisdom of the decision was fiercely debated and the reaction amongst the media, the public and powerful interests was swift. Native title had become a ‘national crisis’ (Short 2007, 861). Despite the fact that the decision recognizing native title was extremely limited, the reaction was shock and surprise (Mercer 1997, 12; Russell 2005, 280). The country’s major newspapers regularly featured hysterical reactions to hypothetical and outrageous Aboriginal land claims over much of Australia. Headlines from major newspapers and television included: ‘the decision has the potential to destroy our society’, ‘80% of Western Australia could be claimed’, and ‘Many mining projects are at risk’ (Russell 2005, 280). The Mabo decision had put the aboriginal problem front and center in the country’s political life (Russell 2005, 277). Former minister for indigenous affairs, Robert Tickner (2001, 94) commented, “The reporting of the native title debate was…abysmal. It reached its lowest point when the front page of the Sydney Sunday paper seriously reported a Mabo land claim over Sydney Opera House, which was without legal foundation of any kind”. The initial public response seemed measured despite the media sensationalism. In fact, contrary to most if not all court cases, the general public was aware of and seemed 29 Reported from The Australian, December 11, 1992. 33 to have a clear grasp of the decision. Goot and Rowse (2007, 104-5) report that several polls indicated awareness of the High Court’s Mabo decision just weeks after the decision had reached about 80 percent while a poll taken almost two years later in 1994 remained at this 80 percent threshold. In June 1993, a year after the decision, an AMR:Quantum 30 poll asked respondents what they believed the Mabo judgment was ‘all about’. The most frequently mentioned explanations were positive or neutral ‘Aboriginal land rights’ (26 percent); ‘Aborigines reclaiming land they owned before European settlement’ (18 percent); ‘restores land rights/native title to Aborigines’ (16 percent); ‘acknowledged Aboriginal title to traditional land they have occupied continuously’ (12 percent); ‘Eddie Mabo’s success in court/proved title to Murray Island’ (10 percent); ‘compensation for Aborigines/loss of land’ (5 percent) (Goot and Rowse 2007, 100). While the public was aware of the decision and its contents the question of support for native title was another matter entirely given the transformative potential of the case. An 31 October 1993 poll by ABG McNair asked the following: The High Court of Australia recently decided that Aboriginal people own land which they have continually associated with since before European settlement. This does not affect any privately owned land. How strongly do you agree or disagree with the High Court decision that Aboriginal people should own their traditional land? 30 AMR:Quantum, ‘National Opinion Survey on Aboriginal Issues June 1993’, Prepared for the Australian Mining Industry Council and the Chamber of Mines and Energy of Western Australia. 31 AGB McNair, ‘Aust Backs Black Title’, Bulletin, October 26, 1993, p. 15. 34 Expressed in these terms, explaining that privately owned land was not affected and claimants would have to demonstrate their ongoing association with the land, the High Court’s Mabo decision won 55 percent support and 24 percent against (Goot and Rowse 2007, 106). Unfortunately, the High Court’s decision was not always or usually expressed in these terms. The decision, and its application to mainland Australia, galvanized powerful economic and political interests that had major stakes in lands still used and occupied by Aborigines: the mining industry. In the wake of the High Court’s decision, the discussions “have almost exclusively been conducted with a view to establishing the effect of native title upon the mining industry” (Coombs 1994, 111). Mining interests represented by the Australian Mining Industry Council (AMIC) were particularly interested in protecting and validating titles and leases that could be at risk in light of Mabo (Russell 2005, 291). The group also wanted to ensure that any compensation due to native title holders, as a result of titles or leases granted after the RDA, would be paid by the government rather than industry (Russell 2005, 291). Indeed, mining interests were setting off the proverbial economic alarm bells about how native title threatened new mining projects (Russell 2005, 291). The power and the influence of the mining industry in Australia cannot be underestimated as it contributes approximately 8.8 percent of Australia’s gross domestic 32 product and employs 3.8 percent of the country’s workforce. 32 When the front-page Australian Bureau of Statistics. 2000. The Australian Mining Industry: From Settlement to 2000. Cat. No. 8414.0. 35 33 headline of the prestigious Financial Review read ‘Mabo: Mining on Hold’ and asserted that native title ‘throws into doubt the ability of new mining projects to get underway without time-consuming delays’, the potential economic costs of native title came into full view. The AMIC was particularly aggressive in asserting the economic costs of native title. Several full-page advertisements were placed by the AMIC in major news papers asserting that Mabo had created a crisis for the mining industry, more native title claims would arise and investment would be frightened off (Mercer 1997, 14). Hugh Morgan, the executive director of Western Mining, associated with the AMIC, suggested that ‘the free, prosperous and dynamic nation that our forebears built … is irremediably tainted’ (Russell 2005, 282). The mining industry had a clear interest in shaping the public debate surrounding native title. To accomplish this goal, some have suggested (Goot 1993a; Goot and Rowse 2007) that the mining industry and those with mining interests systematically and insidiously manipulated the polls measuring public attitudes toward native title. The public opinion polls commissioned by the mining industry (AMIC) accounted for more than one-third of the questions that journalists, newspapers and television had access (Goot and Rowse 2007, 102). Polls conducted by the Roy Morgan Research Centre, whose executive chairman was also chairman of a mining company in Western Australia, accounted for another quarter (Goot and Rowse 2007, 102). All told over 60 percent of all polls conducted in regards to the High Court’s decision were produced at the initiative of organizations with direct links to mining interests. No polls were paid for or conducted by Aborigines or those whose fortunes were tied to Aboriginal interests (Goot 33 Peter Gill. ‘Mabo: Mining on Hold.’ Financial Review, 4 June 1993. 36 1993a, 134). Goot (1993a, 135; see also Goot and Rowse 2007, 103) makes three key observations about the questions in the Morgan polls. First, the phrase ‘may be’ was inserted into questions in order to suggest that the High Court itself was uncertain whether Aborigines were the ‘traditional owners of various land areas’. Second, the questions reference land not in terms of native title rights, but as ‘granted’ or ‘given’; from January 1993 through January 1994, no question acknowledged the High Court’s declaration of Aboriginal land rights. Third, no question listed any of the conditions attached by the High Court to native title. No words or explanations on the requirement that claimants establish a continuing association with the land, that private ownership extinguished native title or that native title rights were more limited than traditional property rights. These polls led to attention grabbing newspaper headlines: ‘Voters say no to Mabo’, ‘Mabo worries 86% of people: survey’, ‘Mabo Doubts Remain High’ (Goot and Rowse 2007, 104-5). Another set of surveys conducted on behalf of the mining industry by AMR:Quantum 34 revolved around a number of perceived effects of the Mabo decision. These survey results proved to be particularly important because they brought the reality and the cost of native title from an abstract concept into concrete detail. The questions asked: Whether you would be very concerned, somewhat concerned or not at all concerned if the effect of this Mabo decision were to: put at risk the existing property titles of other Australians; discourage mining investment in Australia; delay or prevent 34 AMR:Quantum, ‘National Opinion Survey on Aboriginal Issues: Wave III [sic]’, Prepared for the Australian Mining Industry Council and the Chamber of Mines and Energy of Western Australia. 37 economic developments; reduce or prevent employment opportunities in Australia; result in the control of some publicly owned natural resources by a minority group; result in large areas of Australia being claimed by Aboriginal people. The results of these questions indicated that approximately 80 percent of those interviewed would be ‘concerned’ or at least ‘somewhat concerned’ if the High Court’s judgment adversely impacted existing property titles, mining investment, economic development or employment opportunities, or if it resulted in large areas being claimed by Aborigines or some natural resources falling under control of some ‘minority group’ (Goot 1993a, 145; Goot and Rowse 2007, 116). The surveys did not report, however, was whether respondents actually thought any of these outcomes likely or even probable. Given that the High Court’s ruling explicitly stated that conventional property title would have the effect of extinguishing native title asking about the level of concern regarding Mabo affecting existing property was a canard. The only value to asking such a question is for the purposes of propaganda (Goot 1993a, 145). The mining industry was not alone in their efforts to influence the political discourse surrounding native title; state governments accustomed to plenary control over land and resource development asserted their displeasure with the High Court’s decision. Historically, the state and territory governments of Western Australia, Queensland and the Northern Territory have opposed indigenous land rights (Russell 2005, 284). Outside of the Northern Territory, Western Australia and Queensland have the largest Aboriginal populations as well as the two largest state-based mining industries (Goot 1993b, 201). Lest we forget, it was the overreach of Queensland Premier Bjelke-Petersen’s 38 Queensland Coast Island Declaratory Act 1985 that precipitated Mabo (No.1) and eventually Mabo (No.2). However, with Bjelke-Petersen resigning from office in 1987, the anti-land rights mantle was to be assumed by another Liberal Party state leader. The most vocal politician that emerged from the states was Richard Court, the Liberal Party premier of Western Australia. Court appeared to harness what Russell (2005, 282) refers to as ‘settler nationalism’, which grew out of the High Court’s recognition and legitimation of Aboriginal Australians place in the current political and social community. Understood in these terms, the High Court cast “a dark shadow over the beginning of British settlement in Australia and the conditions of its success” (Russell 2005, 282). This nationalist resentment led Court to issue a series of radical proposals with the purpose of refuting the High Court’s challenge to the traditional conception of Australian society. Court spearheaded the call for a national referendum that would overturn the Mabo decision (Russell 2005, 298). He also led a Western Australia Liberal conference that proposed giving the Senate control over appointments to the High Court (Russell 2005, 298). When those measures failed to generate enough political support, Court threatened that the High Court’s ruling could lead to his state seceding from the federation (Goot 1993b, 194). What were the effects of the efforts of both the mining industry and state governments, such as Western Australia, on the opinions of Australians towards the issue of native title? No systematic empirical inquiry has been undertaken to explore and tease out whether public opinion was responsive to these influences. However, outspoken Catholic nun, writer and academic, Veronica Brady asserted that the anti-Mabo campaign 39 orchestrated by the Court government stirred anti-Aboriginal feelings into a frenzy (Goot 1993b, 194-5). Similarly, Henry Reynolds, an academic and historian, who was instrumental in influencing Eddie Mabo’s decision to bring his case before the High Court, wrote that the ‘anti-land rights campaigns waged by conservative politicians and the mining industry’ have ‘held public opinion back a generation’ (Goot 1993b, 194). 2.8 Wik Peoples v Queensland: Facts, Ruling and Reaction The Wik people were one of the first Indigenous people to assert native title rights in the courts after the Mabo decision. In June of 1993, a year after Mabo (No. 2), they applied to the Federal Court for a declaration that they are the traditional owners of 28,000 square kilometers of land and waters on the western side of Cape York Peninsula in the northern part of Queensland (Russell 2005, 316-7). Part of the land that the Wik claimed native title was the Holyrod River Holding, an area of 2830 square kilometers, that had been subjected to a pastoral lease first issued in 1945, but later surrendered in 1973 only to renewed in 1975 for thirty years (Russell 2005, 317). The use of the Holroyd River land by pastoralists was slight. No pastoralists lived on the land and no boundary fencing or buildings were constructed (Russell 2005, 317). The Thayorre people, also of Cape York, whose native title claim partly overlapped with the Wik’s claim, also joined the legal action. This land had come under the Michellton Pastoral Lease, which was first issued in 1915, was forfeited in 1918, reissued and finally surrendered in 1922 (Russell 2005, 317). Notably, no pastoralist had entered the land and its traditional inhabitants continued to occupy it. Since 1922, the land formerly subject to the Michellton leases has been reserved for the benefit of its Aboriginal occupants (Russell 2005, 317). The 40 common thread of these claims centers on the issue of pastoral leases and whether they had the effect of extinguishment on native title. The effect of pastoral leases on native title may sound technical, but the answer to the question had huge consequences for Australia. An analysis of land tenure in Australia in 1980-1 showed that 52.6 percent of Australia’s total land area and 76.4 percent of land held for private use was under leasehold tenure (Holmes 1991, 44). It is reasonable to ask why such a large percentage of the country was tangled up in these lease agreements. Pastoral leases involve the grazing of livestock, mostly sheep and cattle, which is increasing problematic in Australia’s mostly arid interior. Australian legislators created the pastoral lease in order to give pastoralists access to large tracts of land, but stopped short of giving them exclusive ownership or possession of these land areas (Russell 2005, 320). These tracts of land were not small parcels, but could be as ‘as 35 extensive as many a county in England and bigger than some nations’. There was no evidence that the legislators intended the holders of these limited-term leases to be in exclusive possession of the pastoral lands (Russell 2005, 320). Aboriginal peoples continued to live on these rangelands and many pastoral leases had reservation clauses reserving Aboriginal occupancy and use of the lands (Russell 2005, 321). Notwithstanding the plain intention of legislators to give and protect Aboriginal access to these lands, it is another question altogether whether native title survived the leasehold process. If pastoral leases could not extinguish native title then huge swaths of Australia could be swallowed up by Aboriginal land claims. It is for these reasons that the case brought by the Wik and Thayorre would be referred to as the “Mabo of the mainland” 35 Wik Peoples v Queensland (1996), 187 C.L.R. 1 at 217. 41 (Russell 2005, 316). The composition of the High Court, now four years after Mabo (No. 2), had changed modestly. Chief Justice Mason retired in April 1995 and was succeeded as Chief Justice by Justice Gerard Brennan whose vacancy was filled by William Gummow. In November 1995, Justice William Deane retired and was replaced by Michael Kirby. How these new justices would vote and whether the High Court would continue to support its holding in Mabo (No. 2) after the reaction the decision received was in real doubt. Just four years after issuing a blockbuster ruling on native, the High Court, on December 23, 1996, announced a further ruling on native title and the extinguishing 36 effects of pastoral leases in Wik Peoples v. Queensland . In a four to three majority decision, the Court ruled that pastoral leases did not confer a right of exclusive possession and hence did not necessarily extinguish all native title rights that may otherwise have survived. The common thread between the majority opinions was the emphasis placed on the distinctive nature of land tenure in Australia. The word ‘lease’ in the colonial and post-colonial statutes, under which the Holroyd and Michellton leases were issued, had to be understood in terms of the legislation’s economic and social objectives (Russell 2005, 320). The majority could not conclude based on the historical background of pastoral lease agreements, noted above, that lease agreements necessarily implied exclusive possession (Russell 2005, 321). The judgment essentially recognized that a legal regime of coexistence between pastoral leaseholders and native titleholders was possible because a pastoral lease did not give its holder exclusive possession of the land (Russell 2005, 319-20). However, in the event of any 36 Wik Peoples v. Queensland (1996), 187 C.L.R. 1. 42 inconsistency between native title rights and the rights of the pastoralists, the rights of the pastoralists would prevail. As Justice Toohey explained in his judgment, the ‘traditions, customs and practices of the aboriginal claiming the right’ are inconsistent with the rights conferred on the grantee of the pastoral lease, ‘to the extent of any inconsistency the latter 37 prevail.’ Despite the fact that Wik effectively affirmed the core ruling in Mabo (No. 2) and opened the possibility of native title claims to lands that had or have held pastoral leases, the justices balked on what was the most important question of the case to the Wik and Thayorre peoples. The majority did not uphold the native title claims of either the Wik or the Thayorre. Instead, their judgment simply allowed for the possibility that their claim to native title was valid. Both would still have to prove the validity of their claim another day (Russell 2005, 319). The position of the four majority justices was extremely limited in scope; yet recognizing the possibility of native title on mainland Australia ignited a political furor that would lead to rhetorical attacks on the legitimacy of the High Court. Conservative politicians led the backlash against the High Court’s Wik decision from both the state and national levels. Queensland’s Premier, Rob Borbridge, predictably, described the Wik decision ‘as an embarrassment’ (Patapan 2000, 141). He rallied state premiers and called on the Commonwealth government to extinguish native title or at the very least create a sunset clause or terminal date preventing future claims on native title (Russell 2005, 322). Deputy Prime Minister and leader of the National Party, Tim Fischer, criticized the judgment as an example of judicial activism (Patapan 2000, 141) and suggested the need to review how justices to the High Court were appointed 37 Wik Peoples v. Queensland (1996), 187 C.L.R. 75. 43 (Russell 2005, 323). Prime Minister John Howard echoed his deputy’s sentiments regarding judicial activism, asserted that “the laws governing Australians ought to be determined by the Australian Parliament and by nobody else (Williams 2001, 184). Even Commonwealth Attorney General Daryl Williams broke with conventional norms and refused to defend publicly the judiciary against political attacks: “The judiciary should accept the position that it no longer expects the attorney-general to defend its reputation and make that position known publicly” (Williams 2001, 185). The High Court found itself under fire for another native title decision and “it certainly did not seem prepared for the vehemence and stridency…that did not hold back from criticizing individual judges, questioning the place of judicial independence and separation of powers” (Patapan 2000, 145). Perhaps the High Court had gone out on too far of a limb by recognizing the possibility of native title on mainland Australia. The native title decisions represent the kind of political controversy that is left untouched by politicians for a reason: the volatility of emotions associated with the issue. The attacks on the legitimacy of the High Court certainly sent a strong message that the justices “political mandate to be the pace-makers on the rights of Indigenous people has run out” (Russell 2005, 381). The question going forward is whether the High Court will continue to grow native title rights or restrict them even more. 2.9 Extinguishing Native Title: Ward, Anderson and Yorta Yorta Six years after the High Court’s Wik decision, a trio of cases was heard by the High Court in the later part of 2002 that would substantially define the potential and limits of native 44 title. Since the Wik case, three new justices joined the High Court: Hayne replaced Dawson, Callinan replaced Toohey and Gleeson replaced Brennan. All of these appointments were made under the conservative government and leadership of Prime Minister John Howard whose government had a clear and stated interest in appointing justices that were conservative in their judicial approach and by extension conservative in their application and interpretation of native title. The personnel changes portend discouraging implications for Aboriginal Australians seeking an expansion of native title rights. More specifically, a ‘bundle of rights approach’ had been developing in Australia’s lower courts as compared to understanding native title as full and exclusive ownership of land (Russell 2005, 377). These cases would determine and clarify the precise nature of native title claims going forward. Two of those cases were decided on 38 the same day in August 2002: Wilson v Anderson 39 and Western Australia v Ward . The Anderson case was another case about extinguishment of native title by pastoral lease in the western part of New South Wales. In 1901, the Western Lands Act was passed by the state government of New South Wales and conferred exclusive possession of to a pastoral lease over the area. In effect, a perpetual lease was granted for pastoral use, which differed significantly from the Wik case that held pastoral leases do not confer exclusive possession and therefore could not extinguish native title. The High Court, in a six-to-one majority, found that that pastoral lease here was significantly different from traditional pastoral leases because of it was a perpetual lease. The lease was in fact closer to a traditional freehold land grant, which would and did confer 38 39 Wilson v Anderson (2002) HCA 29 Western Australia v Ward (2002) HCA 28 45 exclusive possession. The result was total extinguishment of native title. The Ward case was more complex and had larger implications. The case arose out of an application for a determination of native title by the Mirriuwung, Gajerrong and Balangarra peoples in relation to a land area covering 8000 square kilometers in the East Kimberly region of Western Australia and extended into the northwest corner of the Northern Territory (Russell 2005, 377). Several major economic development projects had taken place in the region, including the Argyle Diamond Mine and the Ord River Irrigation project, which was designed to facilitate citrus fruit farming and general agriculture in the region (Russell 2005, 377). The issue before the High Court was the extent to which these projects extinguished native title. The majority, by five-to-two, found that native title was not necessarily extinguished by all of the projects and leases authorized on their lands. In effect, like the Wik decision, the majority endorsed a regime of coexistence between native title holders and others granted rights by the state (Russell 2005, 377). Native title holders may be entitled to use and access of the land, but they do not retain exclusive possession or control of those lands. Other interests may be granted leases or permits to occupy or extract resources from lands that have been recognized as holding native title. For practical purposes, this prevents Aboriginal Australians from developing the land in nontraditional ways and would certainly prevent claims on minerals or petroleum found on those lands since there is no evidence of Aboriginal laws or customs relating to those substances (Russell 2005, 377-8). The Ward decision clarified the nature and conditions of extinguishment by lease and development as a less than fatal blow to native title claims, but significantly limited the scope of what native title meant for successful 46 claimants. Brennan (2003, 218) calls the High Court’s decisions in Anderson and Ward as a “dilution” of native title and as far as native title holders are concerned, the right to native title expressed in Mabo (No. 2) has “completely lost its bite”. The final case of the trio was handed down in December of 2002. The case, Yorta 40 Yorta v Victoria , involved the Yorta Yorta, whose historical lands were in northern Victoria and southern New South Wales (Russell 2005, 378). The land had seen substantial development of agriculture, forestry, commercial fishing and substantial white settlement; the Yorta Yorta experienced a great deal of forced dispossession and many still continued to live on or near their land (Russell 2005, 378). The question at issue was clarification on what actually constituted a traditional connection to the land that was a critical piece of proving native title in Mabo (No. 2). As the case made its way to the High Court, Justice Olney, of the Federal Court, declared that any rights the Yorta Yorta 41 had to their land had been ‘washed away by the tides of history’ . The High Court, five-to-two, upheld the Federal Court ruling. The Yorta Yorta’s claims to native title had been extinguished long ago. The majority rejected the Yorta Yorta’s argument that changing and adapting the laws and customs governing their relations to their lands never severed their connection to those lands (Russell 2005, 379). Chief Justice Gleeson and Justices Gummow and Hayne asserted that the Yorta Yorta could not adapt their traditional laws and customs to their changing circumstances 40 41 Members of the Yorta Yorta Aboriginal Community v Victoria (2002) HCA 58 Members of the Yorta Yorta Aboriginal Community v The State of Victoria (1998) FCA 1606. 47 42 because, after the white man moved in, they lost any independent power to do so. Essentially, the justices were asserting that in order to show a continuing connection to one’s traditional land, Aboriginal customs and laws must not have changed or evolved since white settlement in 1788. To admit to an evolution in customs or laws would admit 43 to acceptance of a change in sovereignty and an abdication of native title. Russell (2005, 378) calls this the ‘frozen rights approach’ because as Brennan (2002, 213) explains, “freezing social structures and the essential state of traditional law and custom as at 1788 makes proof of native title extremely difficult for Indigenous groups across Australia.” Six years after Wik the High Court delivered three significant native title decisions during the last half of 2002. The trio of cases Ward, Anderson, and Yorta Yorta sought to bring clarity to some basic issues concerning native title. These judgments concerned the extinguishing effects of perpetual pastoral leases, clarified and reconciled how native title can survive competing claims and use of the land as well as how exactly an Indigenous group can demonstrate the requisite connection to land through traditional law and custom, which leads to recognition of native title. There is no question that native title rights became must more restricted as a result of these decisions; the threshold alone for proving continuing connection to the land will prove daunting at the least for future claims. These three decisions, collectively, shifted the concept of native title away from exclusive possession and control to a bundle or collection of rights where Aborigines 42 43 Members of the Yorta Yorta Aboriginal Community v Victoria (2002) HCA 58 at 44 Members of the Yorta Yorta Aboriginal Community v Victoria (2002) HCA 58 at 46 and 87 48 must coexist with other interests on their land. It is no wonder that the headline for The 44 Australian’s article on the Yorta Yorta decision asked, ‘Is this the end of native title?’ 2.10 Counting Claims: Native Title from 1992 to 2012 In May of 2012, the National Native Title Tribunal (NNTT) released a report detailing the outcomes and achievements of the native title system twenty years after the Mabo (No. 2) decision that ushered in a dramatic change to Australia’s social and legal landscape (NNTT, 2012). It is surely fitting to conclude this chapter with a full accounting of native title claims across the states and territories of Australia since it was feared that the decision would cause a great calamity. After Eddie Mabo’s success at the High Court, it was a further two years before the first native title claim was determined to be valid. Since 1994, approximately seventeen percent of the Australian land mass (approximately 1.4 million square kilometers) can be classified under the possession of Aboriginal Australians as a result of native title claims (NNTT, 2012). Some twelve 45 hundred claims have been lodged, but only one hundred and thirty-nine native title claims have been successful (NNTT, 2012). However, the amount of land involved and successful native title claims vary dramatically across the states and territories. [Insert Table 2.1 about here] 44 Stuart Rintoul, “Is this the end of native title?” The Australian, December 13, 2002, p. 1. 45 This number includes Indigenous Land-Use Agreements, which is a distinct concept from native title. ILUA’s are voluntary agreements between indigenous and other groups about the use and management of land. 49 Table 2.1 displays the number of successful claims or determinations of native title from 1992 to 2012 across all states and territories. First, note the variation in the number of successful determinations and total claims of native title across the states. Western Australia, Queensland and the Northern Territory appear to be where most of the claims are centered. This is no accident. These three states have the largest concentration and numbers of Aboriginal Australians. Perhaps, even more surprisingly, is the success rate of claims in these areas, each reaching into the ninetieth percentile. Queensland, at the center of so many of native title cases, has received and successfully determined more native title claims than any of the other states and territories, but the claims tend to be, on average, of modest size. The native title land claims in Western Australia, on the other hand, have been substantial, averaging over thirty-two thousand square kilometers per determination. The area covered by native title claims in Western Australia exceeds sixty-three percent of the total for the whole country. New South Wales and Victoria have had the opposite experience with native title. This is the result of the nation’s two largest cities, Melbourne and Sydney, being located in these states and the resulting development and expansion that surround urban areas. Still several Aboriginal groups have made a number of successful determinations despite the significant challenges required to establish valid claims. 2.11 Conclusion This chapter presented the historical background and individual characteristics of each of the High Court’s most pivotal native title decisions. Without a doubt, Mabo (No. 2) was a landmark decision that did not quickly and quietly disappear. It sparked a massive 50 reaction that seemingly turned the country upside down. Wik, was fundamentally important because of the specific issue of pastoral leases, but also because it brought the possibility of native title to mainland Australia. The trio of cases in 2002 sought to bring clarity and finality to native title challenges by fundamentally limiting its value in favor of a bundling of rights approach that favored coexistence with other interests over justice. The next chapter explores if and whether the High Court’s native title decisions did impact public opinion. Analyses are conducted on both the aggregate and individual levels in order to assess whether theories describing how the United States Supreme Court influences the American public can be applied to the High Court of Australia and the Australian public. 51 2.12 Tables and Figures TABLE 2.1 Native Title Claims by State and Territory, 1992 to 2012 Area Number of Average Total Success Covered by State or Territory Successful Land Per Native Claims Rate Claims Claim^ Title# Western Australia 904832 28 29 97% 32315 Queensland 114165 62 65 95% 1841 Victoria 26185 4 7 57% 6546 New South Wales 1987 2 37 5% 994 Northern Territory 70904 29 31 94% 2445 South Australia 310829 14 14 100% 22202 Totals 1428902 139 183 Note: Data from NNTT (2012) #Units are square kilometers ^Units are square kilometers and rounded to nearest whole number 52 76% 10280 Chapter 3 Probing the Conditionality of Support 3.1 Introduction Hamilton’s assertion in Federalist No. 78 that the Supreme Court of the United States would hardly pose a threat to the freedoms and liberties of the citizens of the varied states or to the elected branches of government was based on an astute understanding of the role and limits of courts generally, not just the proposed Supreme Court of the Constitution (Hamilton 1961, 464). The underlying reality facing the Supreme Court of the United States and democratic courts around the world is that they are institutions with no formal mechanisms to guarantee the enforcement of its decisions. Courts become naturally dependent on their ability to convince the other branches and the public of the wisdom of those judgments resulting in what has been termed the ‘implementation problem’ (Rosenberg 2008). Unlike the popularly elected branches, courts cannot quickly respond to public disapproval as the judicial norm of stare decisis constrains decision-making (Bailey and Maltzman 2011) and prevents a court from reversing itself even if it were persuaded to do so by an unreceptive public. Without a natural constituency or effective enforcement powers courts are forced to rely on the rule of law, the reputation of the court and judges as impartial and fair, and the perception of their decisions as legitimate in order to influence the public (Gibson et al 1998; Gibson and Caldeira 2009). In other words, it may be in a court’s best interest to try to sway public opinion towards their position than to rely on rival political institutions. 53 Furthermore, in order to protect its reputation and build legitimacy within the public, courts must take into account the public’s mood when making decisions (McGuire and Stimson 2004; McGuire et al 2009; Mishler and Sheehan 1993) so as not to appear too far out of step and risk damaging that vital relationship. This does not relieve the courts of their obligations to make good law, even if their decisions prove to be unpopular (Bickel 1986) such as in contested presidential elections (Gibson et al 2003). Courts in democratic regimes are thus obliged to behave as “republican schoolmasters” who engage with and educate the public about the law through their opinions (Lerner 1967). It is through this dialogue that courts simultaneously fulfill their judicial function as well as engage the public as teachers of the law in a democratic society. If courts cannot affect public opinion through their decisions then they risk exacerbating an already weakened institutional position. However, the courts and the justices that fill their ranks have their own goals (Segal and Spaeth 2002) and sometimes this tension forces courts to make difficult and politically unpopular decisions. It is during these times that courts draw on their reputation and legitimacy or political capital (Bartels and Mutz 2009; Choper 1980; Gibson et al 2003; Mondak 1992; 1994; Mondak and Smithey 1997) to coax the public to follow its lead. Though the effect of court decisions on public opinion has been relatively 1 confined to the study of the Supreme Court in the United States (Franklin and Kosaki 1989; Hanley et al 2012; Hoekstra and Segal 1996; Johnson and Martin 1998; Marshall 1989; Persily et al 2008; Stoutenborough et al 2006) there is no a priori reason to suggest 1 For important exceptions see Fletcher and Howe (2000) and Matthews (2005) as prominent examples from Canada. 54 that other similarly situated courts around the world should or would behave differently. If after all it is true that justices cannot help, but to try to influence the public in democratic regimes (Lerner 1967: 180) evidence of such a phenomenon should be observed outside of the United States. The effect or influence of decisions delivered by the High Court of Australia on public opinion is tested with this proposition in mind. The similarities between the two judicial institutions as well as the cultural and political similarities between the two nations provides a comparative context well suited for exploring the generalizability of court influence on public opinion. The empirical analysis begins with an aggregate level analysis of the Court’s ability to influence overall public opinion consistent with a theory of legitimation first conceived of by Dahl (1957). Next, an individual level analysis is undertaken and the validity of Johnson and Martin’s (1998) conditional response hypothesis is assessed. Testing these theories in a foreign political and legal context provide confirmatory evidence that a general characteristic of not only the Supreme Court of the United States or the High Court of Australia, but of high courts in democracies around the world is observed. 3.2 The High Court of Australia and Native Title The Australian constitution and the High Court were modeled closely after its counterparts in the United States because it was the most prominent example of a successful federalist system, which was a key factor in the debates leading up to independence (Patapan 2000; Solomon 1999; Williams 2001). Though the court has the power of judicial review it has never been considered a powerful political institution 55 (Soloman 1999; Williams 2001). Partially attributable to this fact is that, historically, the court has viewed its constitutional role as circumscribed. Indeed, the judicial orthodoxy of the High Court views its role as one of ‘policing boundaries of state and federal powers and not passing judgment on the substantive value or wisdom of state or federal law’ (Pierce 2006, 71). Explanations abound for the court’s lack of assertiveness 2 including the absence of a bill of rights , prevalence of federalism disputes or even the 3 influence of the Privy Council . Perhaps the most important explanation is the entrenchment of the principle of parliamentary sovereignty into the judicial orthodoxy. Since Australia inherited its common law system from Britain, the premise that parliament exercises unlimited legislative authority that cannot be challenged by a judicial authority (Dicey 1982) became ingrained in the orthodoxy of the Australian judiciary (Pierce 2006, 63). There is a clear expectation that the court is expected to defer to parliament because the courts are subordinate to parliament’s will (Pierce 2006, 64). Judicial deference to the commonwealth asks judges to “refrain from second guessing the merits of parliamentary acts” (Pierce 2006, 65) and having a “greater faith in the elected branches” (Pierce 2006, 66) than the judiciary as best able to promote justice. This traditional view of the High Court and its role is distinctly at odds with its 2 Without an individual rights doctrine in the Australian constitution, litigants do not have the legal foundation for arguing civil rights and liberties cases before the High Court (Pierce 2006, 71). 3 Until 1986, the Judicial Committee of the Privy Council in the United Kingdom could hear appeals from the High Court of Australia. Smyth (2005) demonstrates that the likelihood of dissenting opinions increased after the break from the Privy Council indicating decision-making was influenced by its presence. 56 4 jurisprudence under the leadership of Chief Justice Anthony Mason , but also must be seen within a global political context in which several Anglo-derived judiciaries were undergoing or had undergone institutional transformations. In 1982, Canada enacted the Charter of Rights and Freedoms, which constitutionally protected a number of individual rights and, as a result, fundamentally altered the Supreme Court’s role in Canadian politics (Manfredi 2001). In 1990, New Zealand’s parliament enacted the Bill of Rights Act, which, like the Canadian Charter, granted a number of rights and freedoms to its citizens. As a result of these reforms, both the Supreme Courts of Canada and New Zealand were fundamentally transformed by their new powers to enforce rights protections. In 1988, a Constitutional Commission, with the support of Chief Justice Mason, recommended that Australia adopt an entrenched bill of rights similar to Canada, but failed when put to popular referendum (Galligan and Russell 1995, 95). Undeterred, the High Court began to find a number of implied rights in the Constitution. These included 5 prohibiting discrimination against citizens of another state , regulation of political 6 7 advertising , freedom of communication and native title rights for Australian 8 Aborigines . Chief Justice Mason remarked that in Australia, “it is the function of the Courts to keep the principles of law up to date” (Galligan and Russell 1995, 96). Far from being benign, this statement clearly signaled the High Court’s willingness to reject 4 5 6 7 8 Chief Justice Mason held the position from 1987 through 1995. Street v. Queensland Bar Association (1989) 168 C.L.R. 461 Australian Capital Television Pty Ltd v. Commonwealth (1992) 177 C.L.R. 106 Nationwide News Pty Ltd v. Wills (1992) 177 C.L.R. 1 Mabo v. Queensland (No. 2) (1992), 175 C.L.R. 1 57 its traditional role and bring Australian jurisprudence in line with its contemporaries. The difference between the role transformations in Canada and New Zealand compared to Australia is that the source of change lay outside of the judiciary with the legislatures in each country adopting new rights measures and vesting courts with new responsibilities (Pierce 2006, 13). However, in Australia, no new powers were granted to the High Court and no bill of rights or legislation expanded judicial review; the transformation was internally driven (Pierce 2006, 13). “Rather, a group of entrepreneurial High Court judges embarked, on their own volition, to redefine their institution’s mission” (Pierce 2006, 13). Though the High Court was originally modeled after the United States Supreme Court it has historically shied away entering into the kinds of political frays axiomatic of its counterpart. This changed under the Mason Court. Certainly one of the primary examples of the High Court’s change in orientation was the native title issue and as such provides a unique context for assessing whether its decisions affected the public. Would the public in Australia respond like the American public? Would the public, unaccustomed to a political court, accept its decisions en masse or would they polarize along political and social cleavages as the American public does? 3.2.1 The Native Title Cases On June 3, 1992 the High Court of Australia announced their ruling in Mabo v. 9 Queensland (No. 2) with a six to one majority in favor of recognizing native title as part 9 Mabo v. Queensland (No. 2) (1992), 175 C.L.R. 1 58 10 of the common law. The court rejected the idea that “Australia was a terra nullius when the white man arrived, an idea that had been the foundation of Australian law and policy for nearly two centuries, Mabo appeared to be a judicial revolution” (Russell 2005, 5). The High Court ruled that the Meriam people of Murray Island, living in permanent communities with social and political organization, had continuously and exclusively inhabited the Island and its surrounding islands and reefs. The court recognized that a form of indigenous title to land might continue to exist in areas where indigenous people still occupied and could display a continuing association with their traditional land. However, where there was possible conflict with non-indigenous interests it would be the rights of the native titleholders that would yield. Despite the fact that the decision recognizing native title was extremely limited, the reaction was shock and surprise (Russell 2005, 280). Headlines from major newspapers and television included: ‘the decision has the potential to destroy our society’, ‘80% of Western Australia could be claimed’, and ‘Many mining projects are at risk’ (Russell 2005, 280). The Mabo decision had put the aboriginal problem front and center in the country’s political life (Russell 2005, 277). Former minister for indigenous affairs, Robert Tickner (2001, 94) commented, “The reporting of the native title debate was…abysmal. It reached its lowest point when the front page of the Sydney Sunday paper seriously reported a Mabo land claim over Sydney Opera House, which was without legal foundation of any kind”. A year after the Mabo decision twelve percent of those surveyed in a poll thought that their homes could be threatened by the native title decision (Goot 1993a). Goot and Rowse (2007, 104-5) report that several polls indicated 10 Literally translated as land of no one (Russell 2005, 5). 59 awareness of the High Court’s Mabo decision just weeks after the decision had reached about 80% while a poll taken almost two years later in 1994 remained at this 80% threshold. Native title had become a ‘national crisis’ (Short 2007, 861) that was facilitated by the High Court’s best intentions to right an historical injustice. Prime Minister Keating commented a year after the Mabo decision that “I am quite sure the High Court didn’t understand the huge ramifications of this issue” (Horrigan 2003, 207). The Wik people were one of the first Indigenous people to assert native title rights in the courts after the Mabo decision. In June of 1993, a year after Mabo (No. 2), they applied to the Federal Court for a declaration that they are the traditional owners of 28,000 square kilometers of land and waters on the western side of Cape York Peninsula in the northern part of Queensland (Russell 2005, 316-7). The Thayorre people, also of Cape York, whose native title claim partly overlapped with the Wik’s claim, also joined the legal action. The common thread of these claims centers on the issue of pastoral leases and whether they had the effect of extinguishment on native title. The effect of pastoral leases on native title may sound technical, but the answer to the question had huge consequences for Australia. An analysis of land tenure in Australia in 1980-1 showed that 52.6 percent of Australia’s total land area and 76.4 percent of land held for private use was under leasehold tenure (Holmes 1991, 44). If pastoral leases could not extinguish native title then huge swaths of Australia could be swallowed up by Aboriginal land claims. It is for these reasons that the case brought by the Wik and Thayorre would be referred to as the “Mabo of the mainland” (Russell 2005, 316). Just four years after issuing a blockbuster ruling on native, the High Court, on 60 December 23, 1996, announced a further ruling on native title and the extinguishing 11 effects of pastoral leases in Wik Peoples v. Queensland . In a four to three majority decision, the Court ruled that pastoral leases did not confer a right of exclusive possession and hence did not necessarily extinguish all native title rights that may otherwise have survived. The judgment essentially recognized that a legal regime of coexistence between pastoral leaseholders and native titleholders was possible because a pastoral lease did not give its holder exclusive possession of the land (Russell 2005, 31920). However, in the event of any inconsistency between native title rights and the rights of the pastoralists, the rights of the pastoralists would prevail. Despite the fact that Wik effectively affirmed the core ruling in Mabo (No. 2) and opened the possibility of native title claims to lands that had or have held pastoral leases, the justices balked on what was the most important question of the case to the Wik and Thayorre peoples. The majority did not uphold the native title claims of either the Wik or the Thayorre. Instead, their judgment simply allowed for the possibility that their claim to native title was valid. The position of the four majority justices was extremely limited in scope; yet recognizing the possibility of native title on mainland Australia ignited a political furor that would lead to rhetorical attacks on the legitimacy of the High Court. Conservative politicians led the backlash against the High Court’s Wik decision from both the state and national levels. Queensland’s Premier, Rob Borbridge, predictably, described the Wik decision ‘as an embarrassment’ (Patapan 2000, 141). He rallied state premiers and called on the Commonwealth government to extinguish native title or at the very least create a sunset clause or terminal date preventing future claims on 11 Wik Peoples v. Queensland (1996), 187 C.L.R. 1. 61 native title (Russell 2005, 322). Deputy Prime Minister and leader of the National Party, Tim Fischer, criticized the judgment as an example of judicial activism (Patapan 2000, 141) and suggested the need to review how justices to the High Court were appointed (Russell 2005, 323). The High Court found itself under fire for another native title decision and “it certainly did not seem prepared for the vehemence and stridency…that did not hold back from criticizing individual judges, questioning the place of judicial independence and separation of powers” (Patapan 2000, 145). Six years after the High Court’s Wik decision, a trio of cases was heard by the High Court in the later part of 2002 that would substantially define the potential and limits of native title. More specifically, a ‘bundle of rights approach’ had been developing in Australia’s lower courts as compared to understanding native title as full and exclusive ownership of land (Russell 2005, 377). These cases would determine and clarify the precise nature of native title claims going forward. The Anderson case was another case about extinguishment of native title by pastoral lease in the western part of New South Wales. The High Court, in a six-to-one majority, found that that pastoral lease here was significantly different from traditional pastoral leases because of it was a perpetual lease. The lease was in fact closer to a traditional freehold land grant, which would and did confer exclusive possession. The result was total extinguishment of native title. The Ward case arose out of an application for a determination of native title by the Mirriuwung, Gajerrong and Balangarra peoples in relation to a land area covering 8000 square kilometers in the East Kimberly region of Western Australia and extended into the northwest corner of the Northern Territory (Russell 2005, 377). Several major economic 62 development projects had taken place in the region, including the Argyle Diamond Mine and the Ord River Irrigation project, which was designed to facilitate citrus fruit farming and general agriculture in the region (Russell 2005, 377). The issue before the High Court was the extent to which these projects extinguished native title. The majority, by five-to-two, found that native title was not necessarily extinguished by all of the projects and leases authorized on their lands. In effect, like the Wik decision, the majority endorsed a regime of coexistence between native title holders and others granted rights by the state (Russell 2005, 377). Native title holders may be entitled to use and access of the land, but they do not retain exclusive possession or control of those lands. The Ward decision clarified the nature and conditions of extinguishment by lease and development as a less than fatal blow to native title claims, but significantly limited the scope of what native title meant for successful claimants. Brennan (2003, 218) calls the High Court’s decisions in Anderson and Ward as a “dilution” of native title and as far as native title holders are concerned, the right to native title expressed in Mabo (No. 2) has “completely lost its bite”. The final case of the trio was handed down in December of 2002. The case, Yorta 12 Yorta v Victoria , involved the Yorta Yorta, whose historical lands were in northern Victoria and southern New South Wales (Russell 2005, 378). The land had seen substantial development of agriculture, forestry, commercial fishing and substantial white settlement; the Yorta Yorta experienced a great deal of forced dispossession and many still continued to live on or near their land (Russell 2005, 378). The question at issue was clarification on what actually constituted a traditional connection to the land that was a 12 Members of the Yorta Yorta Aboriginal Community v Victoria (2002) HCA 58 63 critical piece of proving native title in Mabo (No. 2). The High Court, five-to-two, found the Yorta Yorta’s claims to native title had been extinguished long ago. The majority rejected the Yorta Yorta’s argument that changing and adapting the laws and customs governing their relations to their lands never severed their connection to those lands (Russell 2005, 379). Essentially, the justices asserted that in order to show a continuing connection to one’s traditional land, Aboriginal customs and laws must not have changed or evolved since white settlement in 1788. To admit to an evolution in customs or laws would admit to acceptance of a change in sovereignty 13 and an abdication of native title. Patapan (2000, 144) writes that “the High Court’s native title decisions reveal a profound engagement with the major themes of civilization and culture, race and identity, justice and history…In this regime the Court becomes the locus for public ethical deliberation and assumes the heavy burden of moral guardian and wise counselor to guide the nation”. Far from its usual position as adjacent to or removed from the political fray (Williams 2001, 173) and much closer to Lerner’s (1967) conception of “republican schoolmasters,” the High Court became embroiled in a case and an issue that forced the Commonwealth and indeed many Australians to come to terms with their colonial past and their unequal present. As a result, Australia’s High Court became the target of unprecedented attack by political leaders and interest groups, across the ideological spectrum, after the native title decisions. “The case—indeed, the whole issue of ‘native title’—became known by his name. To have a view on Mabo was to hold an opinion about the High Court’s judgment” (Goot and Rowse 2007, 98-99). In short, if ever there 13 Members of the Yorta Yorta Aboriginal Community v Victoria (2002) HCA 58 at 46 and 87 64 was an opportunity for a court to be able to influence the public through its decisions it would be in this instance and as a result of the High Court’s decisions involving native title land rights. 3.3 (Under What Circumstances) Can Courts Influence the Public? 3.3.1 Aggregate Level Opinion: The Positive Response Hypothesis The germination of a theoretical connection between court decisions and public opinion is often credited to Robert Dahl and his study of the United States Supreme Court as a key part of national policy making (1957). More specifically, Dahl’s seminal work argued that the main role of the United States Supreme Court is as a legitimizer of public policy. Dahl’s theory can be generalized further by claiming that it is the main function of all national high courts to lend (or deny) its legitimacy to public policies that it upholds (or strikes down). Institutional features such as the regular turnover of personnel and the appointment process work to guarantee that the court’s decisional output is rarely out of step with the dominant lawmaking majorities that fill the elected branches as well as the mass public. Dahl’s main assumption was that court decisions are powerful because they have the unique ability to confer legitimacy on public policy and shape mass attitudes on even the most controversial of issues. An aspect of Dahl’s theory is that courts must first be viewed favorably and trustworthy before they are able to transfer legitimacy to policies. Certainly research on the Supreme Court of the United States has established it as a trustworthy and legitimate institution (Caldeira 1986; Caldeira and Gibson 1992; Gibson and Caldeira 2009; Gibson et al 2003) that exhibits consistently higher levels of support compared with other 65 institutions (Hibbing and Theiss-Morse 1995). Additionally, studies of the legitimacy of eighteen European national high courts (Gibson et al 1998) suggests that courts work to build legitimacy within their publics with the oldest courts (longest track records) having the most institutional support while those individuals most familiar with the courts are the most likely to support them. The High Court of Australia, like its counterparts around the world, enjoys a large reservoir of public support especially in terms of their role as final arbiters of constitutional issues (Galligan and Russell 1995, 99). Indeed, the Australian judiciary “is acknowledged to be of outstanding quality and has enjoyed the public’s confidence” (Evans and Williams 2008, 295). This research suggest that courts may, as a general characteristic associated with their function and institutional role (Adamany and Grossman 1983), be seen as a legitimate institution that is the beneficiary of a positivity bias 14 where the default predisposition toward a court is a positive one. In other words, because courts are consistently viewed as trustworthy and credible they may be uniquely situated to transfer their legitimacy to policies consistent with Dahl’s theoretical assertion. This legitimating function has been consistently demonstrated in a range of experimental and quasi-experimental studies (Clawson et al 2001; Grosskopf and Mondak 1998; Hoekstra 1995; 2000; 2003; Hoekstra and Segal 1996; Mondak 1990; 1992; 1994) and has established that in at least in some circumstances the U.S. Supreme Court can elevate public support for an issue simply by making a decision. This effect, 14 Positivity bias or positivity offset is a phenomenon, most familiar to political psychologists, where individuals instinctively evaluate an object in a positive light until information to the contrary is provided (Cacioppo and Berntson 1994; Cacioppo, Gardner, and Berntson 1999). Gibson and Caldeira (2009) label this phenomenon ‘positivity theory.’ 66 though, does not operate uniformly for all decisions or for all people. In fact, these studies indicate that the Supreme Court’s power to confer legitimacy depends in large part on the salience of the issue, but as Franklin and Kosaki (1995, 370) have noted, “a modest amount of coverage can enlighten a substantial fraction of the public”. In other words, a court’s capacity to confer legitimacy is contingent on the public’s awareness of the decision. Still an expectation arises out of Dahl’s theorizing: large aggregate shifts in public opinion should follow as a court hands down a decision. This is after all the essence of a court conferring legitimacy on a policy. Public opinion should shift in the direction of the court’s decision if courts really do have the ability to influence the public on important policy questions. Yet in one of the most thoroughly exhaustive empirical investigations into whether court decisions can influence public opinion, Marshall (1989) found little to no evidence of aggregate shifts in opinion of the kind expected under Dahl’s theory. In fact, the average shift in opinion was less than one percent and in the few instances in which there was an observable shift, opinion was just as likely to move against the direction of the court’s decision. Support for large aggregate shifts in public opinion following a court decision, a positive response, has found very little support within the United States, but it is far from clear whether we should expect the same in 15 Australia. The longstanding politicized nature of the U.S. Supreme Court stands in stark contrast to the largely apolitical history of the High Court of Australia. While judicial decisions may not be expected to quickly gather support, based on observations 15 Myers and Sheehan (2012) did uncover evidence of the High Court’s ability to legitimate policy and influence elite opinion in regards to several decisions involving industrial relations. 67 in the United States, whether or not this is the case in other countries cannot be dismissed or assumed to be the case without an empirical assessment. Either the courts are unable to affect opinion change or the process of persuasion is more complex than simple aggregate shifts in overall mass opinion. Positive response hypothesis: Each judicial decision will positively shift public opinion in the direction of the High Court’s position. 3.3.2 Aggregate Level Opinion: Data and Modeling Strategies 16 In order to test the positive response hypothesis, annual public opinion (1990-2007) on the transfer of land rights to Aborigines is modeled. The dependent variable is a question 17 asked in surveys from the Australian Election Studies of national adults over this time period. The question reads: The statements below indicate some of the changes that have been happening in Australia over the years. For each one, please say whether you think the change has gone too far, not far enough, or is it about right? Transfer of land rights 18 to Aborigines. The percentage of adults responding that the transfer of land rights to 16 The research design and statistical analysis used here is based on Stoutenborough et al’s (2006) assessment of aggregate level support for legal same-sex relations in the United States in regards to the Supreme Court’s decisions in Bowers v. Hardwick (1986) and Lawrence v. Texas (2003), in particular. 17 The author gratefully acknowledges the assistance of Dr. Nathan Harris, Fellow at RegNet, of the Australian National University College of Asia and the Pacific, for helping to obtain the data used in this and the following analyses. 18 Given the paucity of available surveys the strategy followed was first employed by Stoutenborough et al (2006) in their study of aggregate support for same-sex relations and interpolate values for missing years. In those years in which survey data is not 68 Aborigines has not gone far enough/not nearly far enough is used. This percentage is plotted over time in Figure 3.1. The primary independent variables are the three sets of High Court decisions (five cases, in total) that directly addressed the issue of native title or the transfer of land rights to Aborigines. The effect of each decision is measured by creating simple dichotomous variables coded zero for years in which the decision was not issued, and one for the year in which the decision was issued. [Insert Figure 3.1 about here] A number of societal conditions that might influence public opinion are included 19 as controls. For example, as the average number of newspaper stories on native title increases public support for the Court’s native title rulings is expected to decrease. Given the alarmist rhetoric that permeated media accounts from the outset (Goot and Rowse 2007; Russell 2005; Short 2007), it would be of no surprise if the decisions filtered through the media resulted in a public distressed by the prospect of native title. This is consistent with the role of the news media as a filter disseminating information about the importance of Court decisions and providing potentially negative, neutral or positive available the values were interpolated based on previous and future values. These years included 1991-1995, 1997, 1999, 2000, 2002, 2003, 2005 and 2006. 19 The author used Lexis-Nexus Academic to search for newspaper stories that featured the phrase “native title”. The number of stories featuring the search term were recorded for each month and then averaged for the year. The Sydney Morning Herald was the newspaper source selected, as it is the oldest continuously published broadsheet newspaper in Australia. The average number of stories per year is 12.7 with a minimum of 0 (before the Court’s Mabo decision) and a maximum of 67.2 (the year after the Court’s Wik decision). Additionally, a Modified Dickey Fuller T test was carried out to determine if the variable contained a unit root. The null of a unit root was rejected at a critical value of 1% confirming the variable is stationary. 69 perspectives on the decision (Davis 1994; Hoekstra 2003; Mondak 1994; Mondak and Smithey 1997). As the news media is likely to focus on controversy and conflict (Iyengar 1991) controversial issues like native title should receive more coverage, which suggests the public should be more aware of these decisions (see also Franklin and Kosaki 1995). The extent to which the news media covers decisions and the content of that coverage can play a significant role in how the public responds (Hoekstra 2003). Research in the Australian setting has repeatedly and over time uncovered a deep reservoir of hostility, resentment and racism directed toward Aboriginal Australians. Much of this negativity is sustained through negative stereotypes that suggest Aborigines are dependent on welfare, drunkards and unwilling to assimilate (Larsen 1981; Pedersen et al 2000). Though the most recent research suggests that these negative attitudes may actually be the result of collective guilt about past and present wrongs to Aboriginal 20 Australians (Pedersen et al 2004), old-fashioned and modern racism persists throughout Australian society (Mellor 2003; Pedersen and Walker 1997; Walker 1994; Western 1969). In order to capture this dynamic, a measure of the percentage of Aborigines that 21 make up the population of Australia is included. 20 Australians have tended to exaggerate Recent examples of historical injustices committed against Australian Aborigines include the child removal policy that operated from 1869 to 1969 as well as the more general White Australia Policy, which finally ended with the adoption of the Racial Discrimination Act of 1975. 21 Population statistics were gathered from the Australian Bureau of Statistics (2008), which conducts a census every five years. Population data from 1986 to 2006 was used. Aboriginal Australians made up 1.46% of the population in 1986 or numbered 227,645 out of 15,602,156 and have steadily increased in population size so much so that in 2006 they made up 2.29% of the population or numbered 454,795 out of 19,853,069. A modified Dickey Fuller t-test was carried out to determine if the variable contained a unit root. The test failed to reject the null of a unit root. An augmented Dickey-Fuller test 70 the relative population size of the Aboriginal population by almost an order of magnitude. 22 For example, Goot (1993a, 143) presents the results of a poll that indicated the mean estimate of the Aboriginal population in 1993 to be 13 percent when the actual population size was closer to 1.5 percent. If the majority of Australians think there are significant numbers of Aborigines then granting native title land rights may be seen as a threat to property and resources. 23 The expectation is that as the percentage of Aborigines increases so will hostility and reluctance towards the transfer of land rights. Finally, because the overall model is relatively simple, some factors that shape opinion may not be accounted for in the model. In order to capture the potential impact of unobserved forces, the annual measure of opinion lagged one observation is included as an independent variable. Descriptive statistics for variables used in this model can be found in the Appendix A, Table 1A. 3.3.3 Aggregate Level Opinion: Results Table 3.1 displays the estimates from a Prais-Winsten time-series analysis with 24 bootstrapped standard errors performed on the aggregate level of support for the was then conducted and revealed the series was stationary up to one lag. Therefore a lagged value of the percentage of the Aboriginal population was included in the model. 22 This phenomenon is a familiar one in the United States especially regarding the majority white population’s assessment of minority population size (see Wong 2007). 23 The argument advanced has its roots in theories of racial threat that claim the presence of an outgroup in sufficient numbers will generate competition for scarce resources and thus generate hostility, which in this context will result in less support for native title land rights (Blalock 1967; Key 1949). 24 Bootstrapped standard errors were used in order to account for distortions that may arise from our relatively small sample size. Bootstrapping essentially allows for random resampling of observations with replacement from the original data, in this case 500 71 25 transfer of land rights to Aborigines. Based on the model fit statistics, this model performs very well, with 97 percent of the variation in the dependent variable accounted for; yet, only the lagged value of support reached statistical significance. Neither the percentage of the Aboriginal population, nor its lagged value or the average number of media stories proved to be useful in accounting for the public’s support of native title over the period. Speculation suggests that a general measure of the Aboriginal population may not accurately reflect the potential threat of the group, but given the prospect of population increases potentially leading to more claims of land rights the lack of statistical significance is surprising. [Insert Table 3.1 about here] Media coverage of native title was expected to decrease support for the transfer of land rights and the model does not support this expectation. On the contrary, media coverage of native title stories did not lead the public to be more supportive of the iterations were used, which helps to bolster the validity of the model’s estimates (for a general discussion see Mooney 1996; Mooney and Duval 1993). 25 An initial diagnostic test, an Augmented Dickey-Fuller test for a unit root revealed that the series was stationary up to one lag. Yet, because the dependent variable contained interpolated values, the Breusch-Godfrey test for autocorrelation was used and the null hypothesis of no autocorrelation could not be rejected. The Durbin-Watson d test was not conducted because the model contains a lagged value of the dependent variable, which violates the assumptions of that test. Instead the Breusch-Godfrey test is used in place of the Durbin-Watson h test because it is more statistically powerful (Gujarati 2003, 471). As a result of the series containing autocorrelation, Prais-Winsten GLS regression is employed because under autocorrelation the GLS estimator is BLUE and using the Prais-Winsten transformation performs a transformation that avoids the loss of one observation, which is important given the relatively small sample size of 18. As a further diagnostic check, a Modified Dickey-Fuller T test for a unit root in which the series has been transformed by a GLS regression was conducted. The null hypothesis of a unit root is rejected using the test’s tau statistic for 1 lag at a critical value of 1%. 72 transfer of land rights. Instead of coverage of the issue working against the Court’s decisions, the volume of media coverage appears not to have impacted the public’s views of native title. Additionally, the lagged dependent variable was statistically significant and in the positive direction, which indicates that support amongst the public was able to build on itself over time. This is consistent with the fact that the overall levels of public support for native title increased from 1990 to 2007 by approximately 9%, which may indicate the Court’s decisions have found some long-term legitimation effect. Finally and of considerable importance is the performance of the key independent variables in the model, the dummy variables representing the High Court’s native title decisions. None of the case decision variables are statistically significant, which indicates that the Australian public did not positively respond to the Court’s native title decisions. 26 Contrary to the positive response hypothesis, the model does not support Dahl’s (1957) assertion that a court decision necessarily attracts public support. This finding is consistent with previous tests of this hypothesis in the United States (Marshall 1989) and why later analyses explored individual level opinion to see whether individual level shifts were masked in the aggregate. Whether a similar pattern exists in the Australian native title cases is explored next. 3.3.4 Individual Level Opinion: The Conditional Response Hypothesis 26 Alternatively, the case dummy variables were coded as zero before it was delivered and one for the year it was delivered and every year thereafter in order to investigate whether the cases lead to a break in the slope in addition to the already assumed break in the intercept of the series. The analysis above was replicated and similar results were observed. 73 Another critical turn in the study of the influence of court decisions on public opinion was Franklin and Kosaki’s (1989) study that uncovered a much more nuanced and complicated reaction to the Supreme Court’s abortion decision, Roe v. Wade. In this study, Franklin and Kosaki did not find an aggregate shift in opinion toward easier access to abortion. Instead they uncovered a positive shift towards increased access to abortions for reasons like rape or medication complications, but a much more divisive reaction in attitudes toward discretionary abortions. In the area of discretionary abortions, Franklin and Kosaki found that groups who were initially supportive became more so while those who were initially opposed also became more so. In effect, (between-group) attitudes toward abortion polarized as a result of Roe and by only looking at data in the aggregate one would be led to believe that the court’s decision had no affect on public opinion. Johnson and Martin (1998) uncover similar patterns in their examination of public opinion on the issue of abortion and capital punishment. However, their study prescribes a more careful set of circumstances under which courts an affect public opinion. In both the areas of abortion and capital punishment, courts can only influence public opinion the first time they rule on a controversial issue while subsequent rulings will not affect mass opinion because attitudes will have already been fairly well established. If courts can influence public opinion it is thus conditional in nature; the issue must be of national interest and the ruling must be new or of a landmark nature. Additional evidence supporting the conditional nature of court influence also appears for the issue of gay rights (Stoutenborough et al 2006). Johnson and Martin’s (1998) theory of conditional influence of public opinion has its roots in the social-psychological elaboration likelihood model that is used to explain 74 attitude change and persuasion. First, this model assumes that the public generally views the courts as highly credible or trustworthy institutions, which implies that decisions are seen as a credible information source (Johnson and Martin 1998, 300). Previous research has established that high courts in the United States (Caldeira 1986; Gibson and Caldeira 2009), throughout Europe (Gibson et al 1998) and Australia (Evans and Williams 2008; Galligan and Russell 1995) are viewed under such favorable conditions. The information provided by a decision serves to crystallize opinion as either for or against a given issue. Even though the public may have well-defined attitudes about an issue there will still be an effect because of the legitimacy of the institution. Second, this model theorizes that once an individual forms an opinion, further elaboration or additional thought will not change that opinion. Chaiken’s (1980) experimental findings, consistent with the elaboration likelihood paradigm, show that attitudes persist or resist change when individuals are presented with the same information more than once. The endurance or strength of these attitudes is related to the salience of information and cues at the time of attitude formation (Petty and Cacioppo 1981). When a court issues a landmark decision in a salient issue area, individual attitudes will form and become resistant to change even in the face of subsequent court decisions. In other words, once an issue is elaborated on to the extent of opinion formation, further elaboration, even when information is in conflict with the formed opinion will not cause a change in that opinion (Petty and Cacioppo 1981). The intuition of the public’s conditional response to court decisions assumes that courts are viewed as legitimate and credible institutions, which suggests individual opinions will form in response to landmark decision judicial decisions. Subsequent 75 rulings in the same issue area will not have an effect on opinions as they have already formed as a response to the initial decision. Far from the initial theoretical observations of Dahl, where courts were expected to have a large influence on how the public saw issues, empirical investigations have uncovered a more nuanced and complex set of circumstances under which court decisions influence public opinion. The likelihood of court decisions influencing the mass public is conditioned on whether the case is of significant importance and only the first time a decision is handed down. Conditional response hypothesis: When the High Court makes its first major (landmark) decision in a salient issue area, increased cleavages will occur in between-group attitudes about an issue, and there will be an overall change in public opinion toward the issue. Subsequent response hypothesis: When the High Court decides subsequent cases within the same issue area, little if any change will occur in the structure of group attitudes and there will be no overall change in public opinion toward that issue. Each hypothesis is tested separately. The conditional response hypothesis involves cases in which the Court makes its first landmark decision. For the issue of native title, the landmark decision is Mabo. Another highly visible case addressing the issue of native title is Wik, which will serve to evaluate the subsequent response hypothesis. As a further robustness check, an additional analysis will use the so-called 76 trio of cases, Ward, Anderson and Yorta Yorta that were decided in the same term, in order to provide another test of the subsequent response hypothesis. 3.3.5 Individual Level Opinion: Data and Modeling Strategies The analysis of aggregate level opinion on the transfer of land rights to Aborigines suggests that the influence of the Court’s decisions on opinion is nonexistent as none of the decision dummy variables were statistically significant. Previous analyses in the United States have shown that the movement of opinion of individuals can be masked through the aggregation process (Franklin and Kosaki 1989). In order to uncover whether the Court’s native title decisions did influence individual level opinion the same modeling strategy first used by Franklin and Kosaki (1989) and later by Johnson and Martin (1998) and Stoutenborough et al (2006) is employed here. In each of these studies, the authors examined individual level data obtained from national surveys of adults conducted before and after court decisions and found significant changes in 27 opinion following those decisions. Two surveys, one before the decision and one after, are pooled together so that between group attitudes, predecision and postdecision, can be empirically distinguished as a result of the High Court’s ruling. Specifically, a dummy variable representing the decision is interacted with all of the independent variables, which effectively creates two sets of coefficients: predecision !1,k coefficients and postdecision !2,k coefficients. This 27 Ideally, one would utilize panel data where the same respondents were surveyed before and after a decision, but for these cases, panel data is unavailable. As a result, this analysis follows the example of Franklin and Kosaki (1989), Johnson and Martin (1998) and Stoutenborough et al (2006) and uses two independently drawn random samples of adults from prior to and after a court decision to assess opinion change. 77 interaction allows attitudes that have changed as a result of a decision to be distinguished from one another. This postdecision dummy variable equals 0 before the decision and 1 afterward, which interacts with the !2,k coefficients (the postdecision individual characteristics). The coefficients in this model are then able to vary between the predecision and postdecision samples (Franklin and Kosaki 1989, 755-56; Johnson and Martin 1998, 302). When the postdecision dummy equals 1 the model is “unconstrained” as all of the model coefficients and interactions are present. The equation estimated for the unconstrained model is the following: y = " 0 + "1d1 + (#1,1 + #2,1d1 )x1 + ...+ (#1,k + #2,k d1 )xk + $ (1) The second equation estimated constrains the postdecision dummy to 0, which suppresses ! the interaction effect with the !2,k coefficients. This second equation is referred to as the constrained model. y = " 0 + #1,1 x1 + ...+ #1,k xk + $ (2) The constrained model is only used and the results reported for use in a likelihood ratio ! test, which serves as the direct test of the conditional and subsequent response hypotheses. The results of the likelihood ratio tests will help to explain whether an overall change occurs in attitudes toward the transfer of land rights after the High Court’s decision. Three sets of models, one for each court decision addressing the issue of native 78 title is reported. The first model tests the influence of the Mabo decision and directly tests the conditional response hypotheses. The second and third models test the subsequent response hypothesis and the influence of the Wik and Trio cases, respectively. The data for these analyses comes from five Australian Election Studies conducted 28 in 1990, 1996, 1998, 2001 and 2004. 29 The dependent variable , support for the transfer of land rights to Aborigines, consistent with the aggregate level analysis, is binary with those supporting the transfer of land rights coded 1 while those against coded 30 0. Models are estimated using maximum likelihood estimation. The principal independent variable in each model is the dichotomous case decision variable (Mabo, Wik, and the Trio cases Anderson, Ward and Yorta Yorta). The expectation, consistent with the conditional response and subsequent response hypotheses, is for the Mabo case decision variable to be statistically significant while the Wik and Trio case decision variables will not be. A number of independent variables are included that previous research in the persuasion and attitude change literature has shown to be important in terms of attitude 31 formation and duration. Include are measures of media consumption (television and newspaper consumption) and education level (post-secondary degree or diploma 28 The Australian Election Study conducted in 1993 omitted a question on the transfer of land rights to Aborigines. 29 Respondents indicating the transfer of land rights to Aborigines has not gone far enough or not nearly far enough were coded 1 as supporting native title while those indicating the transfer of land rights to Aborigines had gone much too far, too far or about right were coded 0 as against. 30 As an alternative, the dependent variable was modeled as containing five discrete ordinal categories and similar results were obtained as those presented here. 31 Durable and resistant attitudes are more likely to form after the Court’s landmark Mabo decision and not change in response to future decisions. 79 32 holders). The intuition for including these variables is straightforward: individuals that consume media will be more likely exposed to political information such as court decisions (Franklin and Kosaki 1995; Hoekstra 2003) and the higher ones level of education the more able those individuals are able to process, store and generate arguments related to political information (Converse 1964; Zaller and Feldman 1992). The public attitudes exhibited in these groups should be the most likely to accept the Court’s decisions and, in this case, support native title. Also included is a measure of respondent occupation, specifically whether one works in the farming or mining industries. Berkson (1978) sampled a variety of occupational groups and found high levels of awareness of U.S. Supreme Court decisions when those decisions related to their occupations. Likewise, because all of the native title decisions had a strong potential to impact the farming and mining industries (Goot and Rowse 2007; Russell 2005; Short 2007) through the transfer of land to Aborigines, a measure is included to control for this group and do not expect them to be supportive of the Court’s native title decisions. Research on the effect of court decisions on local opinion revealed contextual effects to be an additional consideration in investigating opinion change within the mass public. Hoekstra (2003, 95-99; see also Hoekstra and Segal 1996), in her study of Supreme Court decisions on local opinion, argues that support for a decision is moderated by the geographic proximity of respondents to the place of the case’s origin. Those individuals that live in closer geographic proximity to where a case originated may have a 32 Previous tests of the conditional response hypothesis (Johnson and Martin 1998; Stoutenborough et al 2006) have not included measures of media consumption because those data sources did not include these types of questions. 80 greater motivation to pay attention to and spend time thinking about the issue and its possible implications (Petty and Cacioppo 1986; Fiske and Taylor 2008; Krosnick et al 1993). Therefore, individuals who are closest to the case, in terms of geographic proximity, will be less likely to be influenced by a court decision as opposed to those individuals who are further removed geographically. To capture this effect, a dummy variable is included and indicates whether a respondent lives in the state in which the case originated. For the Mabo and Wik cases, respondents from Queensland are not expected to be influenced by the High Court’s decisions, while respondents from Western Australia, Victoria and New South Wales, are also not expected to be influenced by the High Court’s Trio decisions. Across all three models, a number of demographic control variables are used including the gender, marital status and union membership of respondents. There are no a priori expectations for these variables, but are included in order to properly fit each of our models. Lastly, measures of party identification are included. The Labor Party, a central left-wing party, has historical ties to the Aboriginal people and would expect their identifiers to support the Court’s rulings and the transfer of land rights. The Liberal and National Parties, center right-wing parties, usually as part of a coalition, have historically opposed special treatment towards Aborigines and their identifiers are not expected to support native title. Descriptive statistics for variables in all models can be found in the Appendix A, Table 1A. 3.3.6 Individual Level Opinion: Results Recall that the conditional response hypothesis posits that some groups of individuals 81 will respond differently to court decisions, but that the overall structure of opinion will only change after the first landmark decision. Figure 3.2 presents the constrained and 33 unconstrained models for public attitudes before and after Mabo. Two models were estimated, a constrained model with each of the independent variables included and an unconstrained model where each of the independent variables is interacted with the Mabo decision variable. This allows for an assessment of how respondents with different characteristics responded to the Court’s decision in Mabo. For each parameter, a positive coefficient indicates an increased likelihood to support the transfer of land rights to Aborigines. [Insert Figure 3.2 about here] Consistent with the conditional response hypothesis, the postdecision dummy— Mabo v. Queensland, the first landmark ruling in the area of native title—had a statistically significant impact on the likelihood of supporting the transfer of land rights to Aborigines. Even controlling for average respondent characteristics, those surveyed following the Mabo decision were 12.5% less likely to support native title than compared 34 to those before the High Court’s decision. Clearly, the High Court miscalculated the overall public’s willingness to address the land rights debate. The unconstrained model reveals a much more nuanced examination of who was responding to the decision in Mabo and how. Two reactions occurred as a result of the 33 34 The full table results can be found in Table 2A of Appendix A. This estimate and later estimates, from each model, were computed by estimating marginal effects at the means. 82 Court’s Mabo decision: some groups became more likely to support the transfer of land rights while other groups became less likely to support the transfer after the decision. In a word, polarization occurred across public attitudes. Those living in Queensland where the case originated and identifiers with the Liberal-National coalition were less likely to support native title, on average, by 8% and 16%, respectively, before the Mabo decision. While the highly educated and females were more likely to support native title before the decision by 8% and 4%. However, after the decision each group became more conflicted and less distinct from other groups. On the other side, high media consumers and Labor party identifiers were, on average, more likely to support native title by 1.4% and 8%, respectively, after the Mabo decision. The High Court’s influence appears to be strongest amongst those who were most likely to hear a great deal about the decision as well as those who were most politically predisposed to support Aboriginal land rights. The data and models presented here cannot tell us definitively if this is a case of people believing an idea is sound in principle, but not reality; however, the uncertainty of the implications of native title were still to be determined and rightly made the public uneasy. The reactions of all of these groups of respondents, especially the political partisans, suggest that the Court’s decision caused them to reevaluate their views. The overall impact of Mabo was to make nearly all groups less supportive of the transfer of land rights to Aborigines, but different groups responded in different ways. The likelihood-ratio test (p < .075) establishes that opinions differed significantly between the two surveys and that some groups of respondents did respond to the decision in different ways. Thus, these results suggest support for the conditional response hypothesis and provide the first empirical evidence that Australia’s High Court, through 83 its decisions, can and does affect public opinion. Figure 3.3 displays the unconstrained model for public attitudes towards the 35 transfer of land rights before and after the Court’s Wik decision. The Wik decision was the first major case after Mabo where a group of Aborigines sought to claim the transfer of land rights through native title and had particular significance since its outcome directly spoke to whether pastoral leases could extinguish land claims. Unlike the Court’s Mabo decision, the Wik decision, as evinced by the decision variable, is not statistically significant and did not have an affect on public attitudes about the transfer of land rights. The overall impact of the Wik decision may have been important in bringing native title to the mainland, but it did not affect the overall structure of opinion. The likelihood-ratio test (p < .222) establishes that opinions did not differ significantly between the two surveys, which is consistent with expectations. Thus, support is found for the subsequent response hypothesis. Only the first landmark decision in an issue area causes public attitudes to respond. This provides further confirmatory evidence of Johnson and Martin’s (1998) conditional response hypothesis from outside of the United States. [Insert Figure 3.3 about here] Though overall public attitudes may not have polarized in reaction to Wik, different groups did respond differently before and after the decision. In particular, those individuals living in Queensland (9%) and married individuals (6%) were both less likely 35 The full table results can be found in Table 3A of Appendix A. 84 to support native title before the High Court’s Wik decision. While the highly educated (11%), high media consumers (1.4%), Labor party identifiers (5%) and females (4%) were all more likely to support native title before Wik was handed down. Once again a wide variety of different groups became conflicted in their opinions and less distinct from other groups in the wake of the High Court’s decision. Conversely, individuals who worked in either the farming or mining industries became 16% less likely to support native title after Wik, which suggests the decision galvanized how individuals involved in these closely affected industries viewed the case’s outcome. Yet, the reaction is surprising given that the decision essentially sided with their interests. The group most responsive to the High Court were identifiers of the Liberal and National coalition. Before Wik, members of this group were 14% less likely to support native title compared to non-group identifiers. After Wik, the same Liberal-National coalition identifiers became 6.5% more likely to support native title. The switch by the Liberal and National identifiers may be explained by the fact that their coalition gained control of the government during this period, which offers a bit of protection against overzealous implementation of the Court’s native title policies (Russell 2005). Figure 3.4 displays the unconstrained model for public attitudes towards the transfer of land rights to Aborigines in response to a trio of court decisions in 2002 – Yorta Yorta, Ward, and Anderson. 36 These decisions had the effect of defining native title as a bundle of rights as compared to exclusive possession by bringing greater clarity to the rules by which native title could be claimed or extinguished. The Trio represents the most important cases involving native title since the Court’s Wik decision. Consistent 36 The full table results can be found in Table 4A of Appendix A. 85 with the results in Figure 3.3 and the Wik decision, the Court’s Trio decisions did not have a statistically significant affect on public attitudes towards the transfer of land rights. The likelihood-ratio test (p < .701) establishes that the overall structure of opinion did not differ significantly between the two survey periods. This further test is consistent with the subsequent response hypothesis, which suggests that overall public attitudes will only change in response to the first landmark decision and not subsequently. The fact that attitudes did not change in response to Wik or the Trio cases provides strong and consistent evidence for Johnson and Martin’s (1998) conditional response hypothesis. [Insert Figure 3.4 about here] Once again different groups did respond to the decision in distinct ways. Prior to the Trio decisions, both Liberal-National coalition identifiers (21%) and married individuals (5%) were less likely to support native title. While those living in Victoria (11%), the highly educated (14%), high media consumers (3%) and members of unions (5%) were all more likely to support native title before the Trio decisions. A familiar pattern once again emerges: a number of groups had distinct opinions regarding native title before a High Court decision, but after a decision the groups all become less distinct from each other. In sum, these results indicate that the first time the Court rules on salient landmark issues like native title land rights, it affects group attitudes and the overall structure of public opinion toward these issues. The empirical results are consistent with both the 86 conditional response and subsequent response hypotheses, which suggests that court influence of the public, though potentially powerful, is limited and conditional in nature. It appears that in life and in court decisions, first impressions may be the most important. 3.4 Conclusion The High Court of Australia was modeled after its counterpart in the United States, but contrary to the American experience the High Court has never been viewed by the public or political actors as being a political institution. Part of this explanation lays in the particularities of the Australian political context such as the absence of a bill of rights and the traditional judicial orthodoxy of the High Court, which stresses deference to Parliament. Given that the High Court is historically viewed by Australians as not being ideologically driven, not engaged in making policy decisions and more importantly not the same type of institution as the United States Supreme Court, theories of court influence might not have as much explanatory value in Australia. But the rise of the Mason Court’s implied rights jurisprudence created an opportunity for the High Court to interject itself into the popular political consciousness with its rulings in the controversial issue area of native title land rights for Aboriginal Australians. Would the Australian public respond in a similar manner as the American public? The first empirical test was of the positive response hypothesis, which suggested that court decisions should result in increased public support of the issue and the High Court’s position, in particular. The issue of native title was a blockbuster landmark decision that was salient and important for every Australian. The results, consistent with previous studies in the United States, did not find support for the positive response 87 hypothesis. The initial indication being that if the High Court can influence the public with its decisions, then like the United States Supreme Court, it is more probably not as easy as initially thought. Next, the conditional and subsequent response hypotheses were tested. These suggest that court influence of the public is much more nuanced and less straight forward than simply issuing a decision. Consistent with Johnson and Martin’s (1998) conditional and subsequent response hypotheses, the empirical evidence demonstrated that the High Court’s decisions affected the overall structure of public opinion in the first landmark decision, Mabo, but not in a subsequent decision, Wik. In order to increase the validity of these results, an additional set of decisions, the Trio cases, were tested and once again found support for the conditional and subsequent response hypotheses. These findings are remarkably consistent with the conditional response theory of court influence on public opinion. Overall, these findings indicate that the High Court of Australia can and did influence public attitudes toward the transfer of land rights to Aborigines. However, the effect was conditional in nature and only in the first instance of a landmark decision. At times the public will react to a landmark decision that has been brought to the forefront of political discourse, but at other times such as when a subsequent ruling is handed down, the impact is minimal. More broadly, these results indicate that judicial influence of public opinion may not be restricted to a particular class of cases or a particular set of issues, but may be a general characteristic of public responsiveness to highly salient, landmark decisions and to courts in particular. Given that judges and courts around the world are surrounded by 88 symbols of fairness, objectivity and impartiality it is no wonder that the public responds to their most important decisions. The next chapter considers the role that political sophistication plays in moderating the influential effects of court decisions on opinion. The first half of the chapter make a distinction between levels of political sophistication and considers whether there is a difference between how high and low sophisticates act in response to the High Court of Australia’s native title decisions. The second half of the chapter engages in with the question of polarization or attitude extremity in terms of the public response to controversial landmark rulings. The differences between levels of political sophistication, once again, are used to determine the extent of attitude extremity among both high and low sophisticates. 89 3.5 Tables and Figures Table 3.1 Aggregate Level Support for Transfer of Land Rights to Aborigines, 1990-2007 Parameter High Court Decisions Mabo v. Queensland Wik Peoples v. Queensland Trio Cases Coef. Bootstrap Std. Error Prob. 1.18 -.93 1.59 Number of Observations Adj-R2 RMSE Rho .339 .952 (.43) (39.88) .018 .909 -5.41 Constant (.08) (36.92) 1.01 4.54 1 Year Lagged Opinion 1 Year Lagged Aboriginal Population .892 .542 .415 .08 -2.23 Societal Conditions Average Media Stories Per Year Aboriginal Population % (8.72) (1.53) (1.96) (13.69) .693 17 .97 .73 .01 Note: Coefficients are Prais-Winsten regression coefficients. Bootstrapped standard errors are in parenthesis. The dependent variable is the percentage of adults responding that the transfer of land rights to Aborigines has not gone far enough/not nearly far enough. Source: Australian Election Studies 1990-2007 90 FIGURE 3.1 Percent Supporting the Transfer of Land Rights to Aborigines, 1990-2007 Note: Data compiled by authors from the Australian Election Studies. Percentage responding that the transfer of land rights to Aborigines had not gone far enough/not nearly far enough. 91 FIGURE 3.2 Likelihood of Support for the Transfer of Land Rights to Aborigines, Pre- and Post-Mabo v. Queensland Note: Data compiled by authors from the Australian Election Studies, 1990 and 1996. The dependent variable is coded 1 if respondents answered the transfer of land rights to Aborigines has not gone far enough/not nearly far enough and 0 otherwise. Error bars represent 95% confidence intervals. 92 FIGURE 3.3 Likelihood of Support for the Transfer of Land Rights to Aborigines, Pre- and Post-Wik Peoples v. Queensland Note: Data compiled by authors from the Australian Election Studies, 1996 and 1998. The dependent variable is coded 1 if respondents answered the transfer of land rights to Aborigines has not gone far enough/not nearly far enough and 0 otherwise. Error bars represent 95% confidence intervals. 93 FIGURE 3.4 Likelihood of Support for the Transfer of Land Rights to Aborigines, Pre- and Post-Ward, Anderson and Yorta Yorta (Trio) Note: Data compiled by authors from the Australian Election Studies, 2001 and 2004. The dependent variable is coded 1 if respondents answered the transfer of land rights to Aborigines has not gone far enough/not nearly far enough and 0 otherwise. Error bars represent 95% confidence intervals. 94 Chapter 4 Political Sophistication and Attitude Extremity 4.1 Introduction Democratic courts around the world are faced with the reality of an institutional arrangement where the enforcement of their decisions depends on the actions of other political actors and institutions. Despite many courts having the power of judicial review and the legitimacy associated with the rule of law, courts have no formal mechanisms to compel compliance. Instead courts rely on informal mechanisms such as the reputation of courts and judges as impartial and fair as well as the perception of their decisions as legitimate (Gibson et al 1998; Gibson and Caldeira 2009) in order to garner public support for their judgments. In this manner, how the public responds to court decisions can have a significant effect on the willingness of government actors and officials to comply with a decision. As Vanberg (2001, 347) suggests, “In democracies where a high court enjoys a high degree of public support, a legislative decision not to comply with a judicial ruling may result in a negative public backlash.” The fear of a public backlash may lead to increased respect for both the courts as institutions and their decisions (Vanberg 2000). In this sense, the electoral connection (Mayhew 1974) can serve as an indirect mechanism for courts to ensure implementation of their decisions. It is of little surprise then that studies of court influence on public opinion have focused on decisions that captured a great deal of public and media attention since it is those landmark cases where courts can bring about profound political conflicts and shift attention towards the issues and policies (Flemming et al 1997) made salient by the decisions. Johnson and Martin’s (1998) study of public reactions to landmark and 95 subsequent decisions in the same issue area revealed that court influence over public opinion is conditional in nature and only in response to an initial landmark ruling (Stoutenborough et al 2006). In addition, the reaction of the public to these salient landmark decisions has followed a consistent and similar pattern of a polarization in group attitudes regardless of issue area (Johnson and Martin 1998; Stoutenborough et al 2006; Persily et al 2008). The underlying process of how the public responds to these judicial decisions is thus of important interest. Do the decisions themselves engender polarized responses or are public reactions more reflective of how those decisions are processed? In order to fully address this question it is valuable to assess how studies of judicial influence on public opinion explain how and why that change occurs. The leading model in social psychology of persuasion, the Elaboration Likelihood Model or ELM (Petty and Cacioppo 1986), assumes that people are motivated to hold correct attitudes, but the amount of effort they are willing to expend on gathering and processing issue-relevant information varies decidedly by individual factors (Petty and Cacioppo 1986) that can be best summed up by the term ‘political sophistication’. Unfortunately, studies of court influence of public opinion have not distinguished between high and low sophisticates, which is the one of main benefits of using the Elaboration Likelihood Model (Kerkhof 1999). The purpose of this chapter is to fully elucidate the differences between high and low sophisticates and their differential responses to landmark judicial decisions such as the High Court of Australia’s rulings in the area of native title land rights. The first half of this chapter is concerned with fully explaining the Elaboration Likelihood Model, its application to the study of court influence of public opinion and the 96 importance of distinguishing between the sophistication levels of the public. A distinction is made between substantive issue processing and heuristic cues with respect to attitude change or persuasion among both high and low sophisticates. Several statistical models are employed to uncover whether high and low sophisticates act similarly or differently in response to the High Court of Australia’s native title decisions. The second half of this chapter is engaged with the question of polarization or attitude extremity. One of the key findings of previous studies involving judicial decisions and public opinion is the appearance of attitude polarization in response to controversial landmark rulings (Franklin and Kosaki 1989; Johnson and Martin 1998; see also Persily et al 2008). It is argued that the term polarization is misappropriated and instead should be considered as attitude extremity. Once again theories from social and political psychology indicate that differences between levels of sophistication may be responsible for attitude extremity. Several statistical models are used to determine the extent of attitude extremity among both high and low sophisticates. This chapter concludes with a discussion of the distinctions made between low and high sophisticates in terms of their respective responses to the High Court’s native title decisions and if the decisions led to extreme attitude positions. An assessment is made on the utility of distinguishing between levels of sophistication when considering whether judicial decisions influence public opinion. 4.2 The Psychological Foundations of Influence A commonality between most studies of court influence on public opinion is the socialpsychological theory used to describe and explain the underlying cognitive process. This 97 theory, the Elaboration Likelihood Model (ELM), has become a major theoretical and 1 foundational framework for studying persuasion. For conceptual clarity, it is important to state at the outset how terms such as ‘influence’ and ‘persuasion’ will be defined and used. In accordance with research norms in psychology, influence and persuasion will be defined similarly as the formation or change of attitudes through information processing in response to a message about an attitude object (Bohner et al 2008; Bohner and Dickel 2011, 403). The process of persuasion is dependent on a variety of individual factors (Petty and Cacioppo 1986, 128) that are related to how likely individuals are to think about a particular issue and form an attitude. These factors are directly related to the motivation and abilities needed to gather, assess and process information necessary for attitude formation and resistance to influence. “Any variable that increases the likelihood of thinking increases the likelihood of engaging” (Petty 1994, 2) in attitude elaboration. The degree to which individuals think through information is important because the strength of an attitude and its resistance to change is a function of the cognitive resources used to 1 Within social psychology two competing models, the Elaboration Likelihood Model (central/peripheral) and the Heuristic Systemic Model (heuristic/systemic), have dominated the theoretical landscape for the last thirty odd years. Both are termed dualprocess models because they each recognize that persuasion can occur through one of two processes. These processes are based in either message arguments and related information (central/systemtic) or heuristic cues (peripheral/heuristic). Differences between types of persuasive influence have consequences for the formation and durability of attitudes. Persuasion occurring via the central/systemic route has been shown to be more impactful while persuasion via the peripheral/heuristic route more volatile. The literature views these models as in theoretical competition, but their differences are based more in emphasis on which route is more important or likely to be accessed by individuals as opposed to divergent theoretical expectations (Kruglanski and Thompson 1999). As such, we view these models as complementary to one another and view the resulting research as mutually enforcing. Why the ELM model has become the model of choice for research on court decisions and public opinion as opposed to the HSM is beyond the scope of this dissertation project. 98 form or fully elaborate an attitude (Bizer et al 2006). Practically speaking, this means that attitudes will vary in their resistance to change depending on how fully formed 2 (elaborated) they are. Consider an individual that has been exposed to issue-relevant information, but does not spend a lot of time or energy thinking about the issue. Contrast this with another individual that has the motivation and ability to think about the issue and process related information. The latter will develop fully formed attitudes that are resistant to change while the attitudes of the former will be less developed and more susceptible to instability and persuasion. Conditional response theory suggests that court influence of public opinion is at its greatest ebb when attitudes form in response to an initial landmark decision (Johnson and Martin 1998). If attitudes are fully formed in response to a landmark decision then subsequent decisions should only serve to reinforce that attitude and the overall structure of public opinion will be relatively stable. If, however, an attitude is not fully elaborated then subsequent decisions will have the same effect on attitudes as a landmark decision and the structure of public opinion will become more volatile. If this is indeed the case, it becomes important to be able to distinguish between those in the public that are most likely to develop fully elaborated attitudes as opposed to those whose attitudes are still open to influence. Cacioppo and Petty (1982) identify a number of factors that lead to an increase in the likelihood of thinking about an issue, which leads to the development of 2 The ELM describes a continuum of elaboration likelihood bounded at one end by the total absence of thought about available issue-relevant information and at the other end by complete elaboration of all the relevant information (Petty 1994, 1). 99 fully elaborated attitudes. The most important deal with source characteristics or 3 heuristic cues and whether the message recipient is high (low) on the need for cognition. 4.2.1 The Role of Heuristics It is widely known that individuals use heuristics as substitutes for detailed knowledge in order to make sense of politics (Huckfeldt et al 1999, 891). Whether this is because people are cognitive misers (Fiske and Taylor 2008) or because these shortcuts may be low cost and numerous (Popkin 1991, 213), the use of heuristics by the mass public is pervasive. These short cuts or heuristics can take a wide variety of forms, such as the idea that experts are more likely to have accurate information or that opinions advocated by a credible source are likely to be correct (for example see Lupia and McCubbins 1998). Citizens that employ heuristic cues are still able to form sensible opinions about issues, which they have spent little time thinking about (Sniderman, Brody and Tetlock 1991). Especially in the context of elites providing opinion leadership, the idea that a person or institution with high credibility can lead the opinions of an unknowledgeable public via heuristic processing of elite endorsements (Gilens and Marukawa 2002) is especially relevant for the study of court influence on public opinion. In this sense, less informed individuals avoid active processing of information and rely on convenient shortcuts that do not require effortful thinking about an issue or issues. People, thus, are likely to be influenced when they hear the views of a trusted institution. 3 Need for cognition refers to and reflects the extent to which people engage in and enjoy effortful cognitive activities (Cacioppo and Petty 1982). 100 Research on court influence of public opinion (Franklin and Kosaki 1989; Johnson and Martin 1998; Stoutenborough et al 2006) all make the same assumption about the role of courts in the persuasion process: the heuristic cue of courts as credible 4 and trustworthy institutions is fixed or constant for the mass public. This assumption has proven to be a sound reflection of public sentiment. Gibson and Caldeira (2009) argue that courts are the beneficiaries of a positivity bias whereby any exposure, both positive and negative, causes people to pay attention to courts and become exposed to the legitimizing myths surrounding courts and the rule of law. Public assessments of national high courts in the United States and Europe have consistently found courts to be widely seen as trustworthy and legitimate institutions (Caldeira 1986; Caldeira and Gibson 1992; Gibson and Caldeira 2009), especially compared to other political institutions (Hibbing and Theiss-Morse 1995). Notably, Bartels and Mutz (2009) suggest that heuristic cues operate concurrently on individuals and simultaneously influence the attitudes they hold. Drawing on an experiment embedded in a national survey, they find that court endorsements have a similar impact on individuals engaged in heuristic processing as well as those engaged in more substantive issue processing. By demonstrating that heuristic cues operate equally or consistently across individual abilities and needs for cognition, Bartels and Mutz 4 Being able to independently untangle the effect of source cues from substantive arguments would be ideal, but as is too often the case, large cross-national surveys nor news organization polls pair questions regarding confidence or trust in courts with questions related to court decisions in the same surveys. 101 (2009) provide empirical evidence that is consistent with the standard assumption that 5 courts are generally viewed favorably by the mass public. Yet, the scholarship in social psychology suggests that heuristic cues provided by courts signaling their endorsement of a particular public policy should have less of a persuasive impact on individuals who are less reliant on heuristic shortcuts. This is because heuristic processing is thought to predominate when motivation or cognitive capacity is low (Chaiken 1980). Experimental research evidence shows that heuristic cues exert no detectable persuasive impact when effortful information-relevant processing occurs (Chaiken 1980; Petty and Cacioppo 1984). In fact, Petty et al (1995) argue that processing information causes individuals to access their prior attitude-relevant beliefs and knowledge much more than the processing of heuristic cues. In other words, full attitude elaboration is unlikely amongst those that rely on heuristics compared to those that rely on issue-relevant information to form attitudes. Instead, heuristic cues are more likely to serve as the main conduit for persuasion or attitude change among those individuals least inclined to fully obtain and process attitude-relevant information. The prevailing assumption in research on court influence on public opinion is that heuristic cues operate uniformly across all individuals regardless of their ability or motivation to fully elaborate their attitudes. This consistent effect of a court endorsement across individuals is corroborated through a survey experiment embedded in a large national survey (Bartels and Mutz 2009) and stands in contrast to a large body of social psychology studies that essentially separate the influence of heuristic cues to individuals 5 However, it should be noted that Bartels and Mutz (2009) were analyzing public opinion and court endorsements on issues (death penalty and flag burning) that were not salient at the time of the survey or were current issues before the Supreme Court. 102 who are unable or unwilling to engage with substantive issue-relevant information (Chaiken 1980; Eagly and Chaiken 1995; Petty and Cacioppo 1984a; Petty et al 1995). The discrepancy lies between the role heuristics play in the attitude elaboration process for those individuals fully capable and willing to process substantive information compared to those individuals who are unwilling or unable to engage in effortful cognitive activity. In other words, the difference between individuals and the attitudes they form lies not in the presence of heuristic shortcuts, but the extent to which substantive issue argumentation enters into the attitude elaboration process. Observationally speaking, the differences between individuals will be seen in the use of attitude-relevant information to fully elaborate attitudes; one group will rely on substantive argumentation while the other will not. For these purposes, the ability to directly account for the persuasive impact of heuristics is not as significant as investigating the distinctive impact of and need for cognition in the persuasion process. 4.2.2 The Need for Cognition: Political Sophistication Cacioppo and Petty (1984, 306) define the need for cognition as “an individual’s tendency to engage in and enjoy effortful cognitive endeavors.” Cacioppo and Petty (1982) describe those with high need for cognition as people who will seek out and reflect upon information in order to evaluate people or situations in their world and will genuinely enjoy the process of doing so. Individuals who are high in need for cognition have been found to be more knowledgeable about events and political figures than people who were low on the need for cognition (Cacioppio, Petty, Kao and Rodriguez 1986; Tidwell, Sadowski and Pate 2000). The ELM distinguishes between those with high 103 (low) need for cognition depending on their abilities and motivations in processing information relevant to forming or changing attitudes. Importantly, the language of low versus high need for cognition in the persuasion literature mirrors many of the distinctions made in political psychology to separate the politically unaware from the politically aware (Kerkhoff 1999, 138) such as the uneducated versus the educated (e.g. Phelan, Link, Stueve and Moore 1995), novice versus expert (e.g., Fiske, Kinder and Larter 1983) and uninvolved versus involved (e.g. Judd, Krosnick and Milburn 1981). The politically unaware or low sophistication respondents are typically compared with the students in the low-involvement condition in Petty and Cacioppo’s classic comprehensive exam experiment. While the politically aware or high sophistication respondents are usually compared to the students in the high-involvement conditions in the Petty and Cacioppo experiments. McGraw and Hubbard (1996, 150) referring to Petty and Cacioppo state that the politically aware, whom they call sophisticates, “are both motivated and able to carefully scrutinize a politician’s explanation.” Though the terminology in the literature has varied, particular cognitive behavior, as described above, and exemplified by ability, opportunity and motivation to engage the political world is more properly defined as political sophistication (Luskin 1990). The politically sophisticated are an intriguing group because they are more apt to be interested in politics, more resistant to persuasive appeals (Chaiken and Baldwin 1981), less susceptible to agenda setting and priming by the media (Iyengar and Kinder 1987) and more easily persuaded by reasoned argument and less by symbolic displays (Chaiken 1980; Petty and Cacioppo 1979; 1984). 104 Luskin (1990, 335) argues that “to become highly sophisticated, we must encounter a certain quantity of political information, be intellectually able enough to retain and organize large portions of the information we encounter, and have reason enough to make the effort.” Consistent with Luskin’s (1990) theoretical statement political sophistication will be conceptualized with regard to three variables: interest in 6 politics, education and media consumption. It is important to consider all three variables as contributing to one concept as opposed to distinct items; it is the combination of ability (education), opportunity (media consumption) and motivation (political interest) that simultaneously engages individuals into the political world. This conceptualization is consistent with Franklin and Kozaki’s (1995) study of media attention and Supreme Court decisions, which found that individuals who were more engaged in politics and were better educated were more likely to learn about a Supreme Court case even when information was relatively high. Interest is motivation. Individuals that are more interested in politics notice more of the political information that they encounter and are more likely to think more seriously about the political information they notice (Chaiken 1980; Petty and Cacioppo 1979). As such, individuals with high motivation or interest in politics will be more likely to seek out and consume more political information than their respective counterparts. Individuals high in the need for cognition create more opportunities for themselves to be exposed to and access all sorts of information about politics (Bizer et al 6 Alternatively, political sophistication can be defined in terms of political knowledge. For instance, Delli Carpini and Keeter (1996) show that levels of political knowledge are closely associated with attitude formation and stability, support for democratic norms and political participation. However, it is important to note that the need for cognition is not necessarily the same thing as an individual’s cognitive ability or the amount of knowledge that they hold (Cacioppo, Petty, Feinstein and Jarvis 1996). 105 2000). The news media, both through print or television, provides the most accessible source of political information and as a result draws the politically sophisticated. “People who like to evaluate should be attracted to all sorts of information about politics because they likely hold attitudes about various political figures and issues” (Bizer et al 2004, 20). The increased ability to generate counterarguments to persuasive messages is generally associated with level of education (Converse 1964; Zaller and Feldman 1992). Arguably, the more education an individual has completed, the greater the store of political information is available to generate arguments. Less well-educated individuals will likely have been less exposed to political information less frequently and will be less able to understand both sides of a given issue. When individuals are unaware of competing considerations or alternatives, there is less information to counter or moderate the (often times one-sided) argument given. Unfortunately, the application of the ELM to studying court influence on public opinion has neglected to account for the critical role that political sophistication plays in explaining persuasion even though making the distinction between politically aware and politically unaware citizens is the principal benefit of using the ELM framework (Kerkhof 1999). This sin of omission is an artifact of problems associated with survey design. Many questions relating to political interest or media consumption do not appear in early cross-national surveys and rarely appear in polls conducted by news 7 organizations and other private firms. 7 For example, the General Social Survey (GSS) did not begin asking questions about media consumption until 1975, which for those researching issues related to abortion, for consistency, precludes them from entering into later analyses as the Roe decision was handed down in 1973. Additionally, the GSS has rarely asked questions about political interest or related to political knowledge. This has effectively precluded researchers from 106 The differences in political sophistication are a critical distinction to make because attitudes vary widely in their durability and impact (Lavine et al 1998; Petty and Krosnick 1995) based on the amount of processing that goes into their formation. Attitudes that are high on the elaboration likelihood continuum are formed by individuals that are motivated, interested and have high cognitive abilities will produce more fully formed attitudes, which are more stable over time, resistant to change or persuasion and are more predictive of behavior (Bizer et al 2006). Whereas, those individuals who fall on the low end of the elaboration likelihood continuum will form attitudes that are highly volatile, inconsistent and seem to fluctuate over time in a seemingly random way (see Converse 1964). As a result, there will be observable differences between respondents on the high side of political sophistication versus the low side in response to court decisions based on their propensity to fully or partially elaborate their attitudes. Johnson and Martin’s (1998) conditional response hypothesis argues that the structure of opinion will change only after an initial landmark decision, but not after subsequent decisions. A critical part of this theory is drawn from Chaiken’s (1980) work on the persistence of attitude change. In particular, Chaiken was interested in the likelihood of attitude change when the same information is presented more than once. In the context of an experiment, Chaiken found that subjects were all less likely to be persuaded the second time information was presented. Additionally, those subjects presented with information about an issue of high consequence (high salience) were even less likely to change attitudes. including any aspect of political sophistication in their analyses of court influence on public opinion in order to be consistent with past research (e.g. Franklin and Kosaki 1989). 107 Chaiken’s findings are consistent with expectations of the Elaboration Likelihood Model, which demonstrates that once an issue is elaborated to the extent that an attitude is formed, further elaborations, even if it is in conflict with their opinion, will tend not to change that opinion (Petty and Cacioppo 1981; see also Chaiken and Baldwin 1981). Petty and Cacioppo (1981, 263) find that the intensity of cognitive processes influences the likelihood for attitude change: “In studies where issue relevant cognitive activity was likely to be intense … the attitude changes produced have been found to be relatively enduring.” In other words, the politically sophisticated, due to their motivation and ability, will be more likely to form stable attitudes in response to court decisions, which will not change even in the face of subsequent information such as additional court decisions; the attitudes of political sophisticates will crystallize. The expectations for the least politically sophisticated are much different. The least politically sophisticated may be exposed to heuristics relating to a decision, which will allow for modest attitude elaboration, but these attitudes will not be able to resist additional persuasive messages. In the face of subsequent court decisions regarding the same issue, the least politically sophisticated will be much more likely than their politically sophisticated counterparts to accept a subsequent decision as new information and update their attitude accordingly. The highly politically sophisticated are expected to behave in accordance with Johnson and Martin’s (1998) conditional and subsequent response hypotheses, but the least politically sophisticated will be just as likely to respond to an initial decision as to subsequent ones. 108 Crystallization hypothesis: The overall structure of opinion among high sophisticates will be affected by an initial landmark court decision, but will not be affected by subsequent decisions within the same issue area. Volatility hypothesis: The overall structure of opinion among low sophisticates will be affected by an initial landmark court decision and will also be affected by subsequent decisions within the same issue area. 4.3 Data and Modeling Strategies: Crystallization and Volatility Hypotheses Previous analyses of the conditional and subsequent response hypotheses suggest that the overall structure of public attitudes responded to the High Court’s landmark Mabo decision and, as expected, was not responsive to the subsequent Wik and Trio (Anderson, Ward, and Yorta Yorta) decisions. Analyses of court influence on public opinion in the United States has not yet explored whether differing levels of political sophistication among the public leads to similar or divergent outcomes from the conditional and subsequent response hypotheses. This is surprising given the theoretical underpinnings of the ELM explicitly calls for a distinction to be made between individuals based on their need for cognition or level of political sophistication. As described in the previous section, political sophistication is a combination of ability, opportunity and motivation; an individual must have all three to be regarded as a high sophisticate or having political sophistication. These concepts are operationalized using measures of education, media consumption and political interest. To be considered in the high sophisticate category, an 109 8 individual must have a bachelor or post-graduate degree, be a high media consumer of television and newspapers and some to a good deal of political interest. The low sophisticate category consists of all other survey respondents that did not meet the above 9 combination of requirements. The percentage of high sophisticates compared to all respondents ranges from 12% for the Mabo case to 14% for Wik and 16% for the Trio of cases. [Insert Figure 4.1 about here] Figure 4.1 reports the mean support for the transfer of land rights to Aborigines between both groups of high and low sophisticates. The lowest level of support for both groups was in 1996 just before the High Court’s Wik decision, while the highest level of support, again for both groups, was in 2004 after the High Court’s Trio of decisions. It is of interest to note that the mean level of support between the low and high sophisticate groups seem to be mirroring one another with support decreasing initially then rebounding resembling a slingshot trajectory of opinion. However, when the difference of means between high and low sophisticates is calculated the apparent mirroring of the two groups disappears. Figure 4.2 shows the mean difference of support for the transfer of land rights between high and low sophisticates. As the figure shows, the gap of 8 A high media consumer is defined as an individual who scored at least a 6 out of 8 on a combined television and newspaper consumption scale. In other words, an individual had to have said they often or sometimes consumed political information from both television and newspapers. 9 For example, individuals that had bachelor or post-graduate degrees, but did not have an interest in politics were coded as low sophisticates because though they have the ability to engage in politics they do not have the motivation to do so. 110 support between high and low sophisticates increased over time, which indicates high sophisticates were not only distinctive from their low sophisticate counterparts, but also much more supportive of land rights in general. The difference though is that the high sophisticate group was much more supportive of the transferring land rights to Aborigines in the beginning of the series and over the course of the entire time period. [Insert Figure 4.2 about here] In order to uncover whether the High Court’s native title decisions crystallized or introduced volatility into the structure of opinion among the most and least politically sophisticated, respectively, several analyses will be conducted utilizing the same modeling strategy introduced first by Franklin and Kosaki (1989) and employed in the previous chapter to test the conditional and subsequent response hypotheses. Two surveys, one conducted before the decision and one conducted after, are pooled together in order to empirically distinguish between group attitudes, predecision and postdecision, as a result of the High Court’s rulings in the area of native title land rights. A dummy variable representing the decision is interacted with all of the independent variables, which creates two sets of coefficients: predecision !1,k coefficients and postdecision !2,k coefficients. The interaction allows for a distinction to be made between attitudes before and after a decision is handed down. The postdecision dummy variable equals 0 before the decision and 1 afterward, which interacts with the !2,k coefficients (the postdecision individual characteristics). The coefficients in the 111 model are then able to vary between the predecision and postdecision samples (Franklin and Kosaki 1989, 755-56; Johnson and Martin 1998, 302). When the postdecision dummy equals 1 the model is “unconstrained” as all of the model coefficients and interactions are present. The equation we estimate for the unconstrained model is the following: y = " 0 + "1d1 + (#1,1 + #2,1d1 )x1 + ...+ (#1,k + #2,k d1 )xk + $ (1) The second estimated equation constrains the postdecision dummy to 0 and suppresses ! the interaction effect with the !2,k coefficients. This second equation is the constrained model. y = " 0 + #1,1 x1 + ...+ #1,k xk + $ (2) The results of the constrained model are only used and reported for use in a likelihood ! ratio test, which serves as the direct test of the crystallization and volatility hypotheses. The results of the likelihood ratio tests help to explain whether an overall change occurs in attitudes toward the transfer of land rights after the High Court’s decision. Three sets of models are reported for each level of political sophistication, one for each court decision (Mabo, Wik, and the Trio cases) addressing the issue of native title, for a total of six models. In order to provide support for the crystallization hypothesis, the likelihood ratio test for high sophisticates is expected to be statistically significant for the Mabo decision, but not statistically significant for the subsequent decisions. For the volatility 112 hypothesis to find support, the likelihood ratio tests for the low sophisticates is expected to be statistically significant for each of the High Court’s native title rulings. The data for the analyses comes from five Australian Election Studies conducted in 10 1990, 1996, 1998, 2001 and 2004. The dependent variable, support for the transfer of land rights to Aborigines, ranges on a 1 to 5 scale that the transfer of land rights had gone much too far, too far, about right, not far enough, and not nearly far enough. Higher values indicate agreement with the High Court that more needs to be done regarding the transfer of land rights to Aborigines. Given that the dependent variable consists of discrete ordinal numbers (1, 2, 3, 4 or 5), the assumption that the distances between the categories are equal should be avoided. 11 An ordered probit model is utilized because it takes into consideration the ordinal nature of the data by treating each observation only as a rank ordering (McKelvey and Zavonia 1975). These models use maximum likelihood estimation. For each model, a positive coefficient indicates an increased likelihood that a respondent believes the transfer of land rights to Aborigines have not gone far enough, while a negative sign indicates that a respondent believes that the transfer of land rights has gone too far. The principal independent variable in each model is the dichotomous case decision variable (Mabo, Wik, and the Trio cases Anderson, Ward and Yorta Yorta). The expectations vary depending on the level of political sophistication. For the high sophisticate group and consistent with the crystallization hypothesis, the Mabo case 10 The Australian Election Study conducted in 1993 omitted a question on the transfer of land rights to Aborigines. 11 McKelvey and Zavonia (1975, 117) provide strong evidence that using the linear regression model to analyze an ordinal dependent variable can lead to incorrect conclusions. Also see Long (1997). 113 decision variable is expected to be statistically significant while the Wik and Trio case decision variables should not be. For the low sophisticate group and consistent with the volatility hypothesis, each of the case decision variables is expected to be statistically significant. Across all models, a number of demographic control variables are included, such as the gender, marital status and union membership of respondents. There are no a priori expectations for these variables, but include them in order to properly fit each of the models. Lastly, measures of party identification are included. The Labor Party, a central left-wing party, has historical ties to the Aboriginal people and would expect their identifiers to support the Court’s rulings and the transfer of land rights regardless of sophistication level. The Liberal and National Parties, center right-wing parties, usually as part of a coalition, have historically opposed special treatment towards Aborigines and we would not expect their identifiers to support native title, again, regardless of level of sophistication. These expectations are consistent with findings in the previous chapter and these groups’ responses to the High Court’s native title decisions. Finally, a measure of respondent occupation, specifically whether one works in the farming or mining industries is included in each model. Berkson (1978) sampled a variety of occupational groups and found high levels of awareness of U.S. Supreme Court decisions when those decisions related to their occupations. Given that all of the native title decisions had a strong potential to harm the farming and mining industries (Goot and Rowse 2007; Russell 2005; Short 2007) through the transfer of land to Aborigines, a measure to control for this group is included in the analyses and do not expect them to be supportive of the Court’s native title decisions regardless of sophistication level. 114 4.4 Results: Crystallization Hypothesis Recall that the crystallization hypothesis posits that the overall structure of opinion of those defined as high sophisticates will change only in response to the High Court’s first landmark decision, Mabo, but will not respond to subsequent judicial decisions in the same issue area. The high sophisticates essentially act in accordance with the expectations of the conditional and subsequent response hypotheses where court influence occurs only in response to an initial decision and crystallizes even in the face of subsequent decisions. Tables 4.1, 4.2 and 4.3 present the constrained and unconstrained models of public attitudes of the high sophisticates before and after the High Court’s native title decisions in Mabo, Wik and the Trio of cases (Anderson, Ward and Yorta Yorta). Two models are estimated for each High Court decision, a constrained model with each of the independent variables included and an unconstrained model where each of the independent variables is interacted with the case decision variable. The unconstrained model allows for a comparison to be made between respondents with different characteristics as a result of the High Court’s decisions. For each parameter, a positive coefficient indicates an increased likelihood to support the transfer of land rights to Aborigines. [Insert Tables 4.1, 4.2 and 4.3 about here] The unconstrained models in Tables 4.1 (Mabo), 4.2 (Wik) and 4.3 (Trio) reveal much more nuanced reactions among particular groups of high sophisticates to the High 115 Court’s landmark Mabo and subsequent native title decisions. In particular, union members were 13.5% more likely to support the transfer of land rights prior to the Mabo 12 decision, but became 16.6% less likely after the decision. This example highlights particular uneasiness about what the decision might mean for union members whose jobs depend either directly or indirectly on the mining and agricultural industries in Australia. In Tables 4.2 and 4.3, attitudes of party identifiers followed a consistent pattern of homogeneous within group attitudes prior to a decision, but would become more heterogeneous after a decision suggesting the High Court’s decisions may have caused a reevaluation of attitudes. This relationship is observed after both the Wik and Trio decisions among the high sophisticates that identify with the Liberal-National party coalition with likelihood of supporting land right transfers being less likely before a decision and then moderating afterwards. Labor Party identifiers followed a similar pattern before and after the Wik decision being likely to support the transfer of land rights and then moderating. What is not observed is a polarizing reaction to the High Court’s decisions occurring across public attitudes, which might be expected if a consensus opinion on land rights transfers developed around the High Court’s decisions. In other words, the High Court’s decisions did not lead groups of high sophisticates to become more consistently supportive or oppositional towards native title. Contrary to expectations of the crystallization hypothesis, the post-decision dummy for the Mabo decision—the High Court’s first landmark ruling in the area of native title—did not have a statistically significant impact on the likelihood of high sophisticates supporting the transfer of land rights to Aborigines. As expected, the post-decision case 12 This estimate and later estimates, from each model, were computed by estimating marginal effects at the means. 116 dummy variables for the High Court’s subsequent native title decisions—Wik and the Trio of cases—did not have a statistically significant impact on the likelihood of high sophisticates supporting land rights transfers. Yet, the results of the likelihood ratio tests paint a picture much closer to the expectations of the crystallization hypothesis. The overall structure of public opinion among high sophisticates did change from before to after the High Court’s Mabo decision (p < 0.0511), but because the Mabo case dummy variable was not statistically significant the credit for the change in attitudes cannot be exclusively attributed to the High Court’s ruling. Still the expectation of the crystallization hypothesis was for the overall structure of opinion to change among high sophisticates in the wake of the Mabo decision, which was observed in Table 4.1. The crystallization hypothesis also asserted that neither of the High Court’s subsequent decisions—Wik or the Trio cases—would alter the overall structure of opinion among high sophisticates. These expectations were born out in Tables 4.2 and 4.3. Neither the case dummy variables nor the likelihood ratio tests for Wik (p < 0.8070) or the Trio cases (p < 0.4952) were statistically significant, which was anticipated by the crystallization hypothesis. Overall, then, it appears there is modest support for the crystallization hypothesis, which asserted that the structure of opinion among high sophisticates would shift in response to the High Court’s landmark Mabo decision, but would not respond to subsequent native title decisions. Of course, without the Mabo case dummy variable being statistically significant it is much harder to claim that the overall change in the structure of opinion was due to the High Court’s decision. However, given the level of engagement that high sophisticates have with politics, generally, it is not surprising that, 117 as a group, attitudes would crystallize in the wake of a landmark decision and remain stable despite subsequent rulings. 4.5 Results: Volatility Hypothesis The volatility hypothesis conceives of the attitudes of low sophisticates as being much more pliable and responsive to judicial decisions even if the decision is not of the landmark variety. Low sophisticates are expected to be much more likely to respond to the heuristic cues associated with the High Court and its rulings regardless of a decision’s landmark status. In other words, the overall structure of opinion for low sophisticates is expected to be volatile, where opinion structure will change after each decision, as opposed to the crystallized structure of opinion exhibited by high sophisticates. Tables 4.4, 4.5, and 4.6 present the constrained and unconstrained models of public attitudes of low sophisticates before and after the High Court’s native title decisions. The models presented and interpretations are analogous to those discussed above for the crystallization hypothesis. [Insert Tables 4.4, 4.5, and 4.6 about here] The unconstrained models in Tables 4.4 (Mabo), 4.5 (Wik) and 4.6 (Trio) reveal a much more complicated set of reactions among groups of low sophisticates compared to those of the high sophisticates. In response to the Mabo decision, two of the low sophisticate groups, in particular, exhibited erratic behavior. Liberal-National coalition identifiers were 9.5% less likely to support the transfer of land rights, but unexpectedly 118 became 6.7% more likely to support land right transfers in response to the ruling. The switch is unexpected because the parties have an anti-Aboriginal history and the Mabo ruling was clearly meant to benefit Aboriginal people. One explanation for the sudden about-face lies with the heuristic cue of the High Court as a trustworthy and credible institution, which, theoretically, is expected to heavily influence the behavior of low sophisticates. Interestingly, the exact opposite behavior is observed among union members, which were 3.4% more likely to support land right transfer before Mabo, but became 6% less likely to do so after the decision. The negative reaction of union members to the Mabo decision is similar regardless of level of sophistication. Table 4.5 presents the results of the High Court’s Wik decision. Once again, low sophisticates that identify with the Liberal-National coalition parties were less likely to support the transfer of land rights, but after the decision became more likely to do so. Part of the explanation clearly lies with the heuristic cue of the High Court, but also with the nature of the Wik decision itself that while affirming the principle of native title nonetheless made it more difficult for Aboriginals to make land rights claims. Additionally, several groups of low sophisticates exhibited similar behavior prior to and after the Wik decision. Females and Labor Party identifiers were 4.7% and 7.7% more likely, respectively, to support land rights for Aborigines prior to the decision, but became more muddled in their respective group attitudes after the decision. Married identifiers were likewise 7.5% less likely to support the transfer of land rights before Wik and moderated after the decision. Each of these groups had consistent within-group attitudes prior to the High Court issuing a decision and afterwards within-group attitudes became much more heterogeneous or less consistent. The High Court’s power as a 119 heuristic cue does not explain why females and Labor Party supporters would backtrack in their support because, if anything, the decision should have lead to sustained support among those groups. A similar set of results is observed in low sophisticates response to the Trio cases with Liberal-National and married group identifiers being 12.4% and 4.2%, respectively, less likely to support land rights before the decision and again moderating after the decisions. Consistent with the respective models of high sophisticates, the models for low sophisticates did not reveal polarizing attitudes in response to the High Court’s native title decisions. Once again the decisions did not lead groups of low sophisticates to become systematically more or less supportive of the transfer of land rights. The expectations of the volatility hypothesis was that the post-decision dummy variables for each of the High Court’s native title decisions would have a statistically significant impact on the likelihood of low sophisticates to support the transfer of land rights to Aborigines. Unfortunately, the post-decision dummy variables for Mabo and Wik were not statistically significant, but the dummy for the Trio cases was statistically significant and positively signed, which indicates the High Court’s decision had the effect of increasing support for the transfer of land rights to Aborigines by 5.4%. The results of the likelihood ratio tests are much more consistent with the expectations of the volatility hypothesis, which suggests that the overall structure of opinion among low sophisticates will change after each High Court decision. As expected, the overall structure of opinion does change after Mabo (p < 0.0339) and after Wik (p < 0.0339), but not after the Trio of cases (p < 0.2597). On one hand, the likelihood ratio tests provide support for the volatility hypothesis in terms of the High 120 Court’s Mabo and Wik decisions, but not for the Trio cases. On the other hand, the case post-decision dummy variables only found support in the Trio cases model. The volatility hypothesis asserted that the overall structure of attitudes among low sophisticates would respond each time the High Court made a native title ruling. The evidence from the models is mixed. The likelihood ratio tests for the Mabo and Wik decisions indicate that the overall structure of attitudes did change, but because the postdecision dummy variables were not statistically significant the change cannot be directly attributed to the High Court’s decisions. Likewise, the post-decision dummy for the Trio cases was statistically significant, but the likelihood ratio test was not. Attitudes among low sophisticates was changing in regards to the transfer of land rights to Aborigines, but the results are unable to provide strong and consistent evidence that the decisions of the High Court were responsible. 4.6 Discussion: Crystallization and Volatility Hypotheses The results yield some degree of confirmatory evidence to support both the crystallization and volatility hypotheses, but certainly are not overwhelming. The overall structure of opinion of high sophisticates did change before and after the landmark Mabo decision, but because the case decision variable was not statistically significant it is not clear whether the change in opinion can be directly attributed to the decision. As expected though, the structure of opinion of high sophisticates did not change or respond to subsequent decisions made by the High Court. For the low sophisticates, the overall structure of opinion did change as a result of the High Court’s landmark Mabo decision as well as for the subsequent Wik decision. Once again though, the case decision variable 121 for both Mabo and Wik were not statistically significant, which does not allow for a direct link to be made between a change in the overall structure of opinion and the native title decisions. The Trio decisions did not elicit an overall change in opinion, but the case decision variable was statistically significant. Overall, neither of the crystallization or volatility hypotheses yielded clear and unambiguous results, but each hypothesis did provide some confirmatory support for the behavior of high and low sophisticates in response to the High Court’s native title decisions. Yet, it is clear that low and high sophisticates do react differently to judicial decisions, which was expected by the ELM. The question that now remains is whether the differences between low and high sophisticates leads to other differences in how each responds to judicial decisions. 4.7 Attitude Extremity Past research has consistently claimed that the increased cleavages that occur in betweengroup attitudes, as a result of court decisions, demonstrate attitude polarization (Franklin and Kosaki 1989; Johnson and Martin 1998; see also Persily et al 2008). This claim, however, is misleading as it confuses increasing homogeneity within group attitudes with attitude extremity. While it may be true that court decisions may elucidate different responses within and between groups, contending that opinion of Catholics and NonCatholics (Franklin and Kosaki 1989) or the higher educated and non-higher educated (Johnson and Martin 1998) have polarized merely means that different groups have become more or less likely to support a policy as the result of a court decision. In this sense, between-group attitudes may move away from each other (supporting a policy 122 versus not supporting a policy), but this does not mean that within-group attitudes are becoming more extreme as a result of a decision. In other words, within-group opinion may sort itself, as a result of a decision, and become more homogenous, but it is unclear whether this sorting leads to more extreme attitudes. After all, attitude extremity is the essence of attitude polarization; whether opinion becomes more (less) extreme from one time period to the next defines attitude polarization (Lord, Ross and Lepper 1979; see also Taber and Lodge 2006). Fortunately, insights from both social and political psychology provide theoretical guidance and expectations regarding who is likely to exhibit extreme attitudes and why. Individuals that seek out and think through information in order to evaluate people or situations in their world are considered to have a high need for cognition (Cacioppo and Petty 1982). The information seeking behavior exhibited by these individuals leads to greater knowledge about political events and political figures compared to less engaged individuals (Cacioppio, Petty, Kao and Rodriguez 1986; Tidwell, Sadowski and Pate 2000). Luskin (1990) argues that these individuals who have a high need for cognition and seek information about the political world are more properly defined as politically sophisticated. The attitudes of politically sophisticated individuals are markedly different from the less politically sophisticated in terms of their durability and strength (Lavine et al 1998; Petty and Krosnick 1995). The difference in attitudes between these so-called high and low sophisticates can be traced back to the amount of processing that goes into their formation. Increased thought about an issue leads to the repeated accessing of the attitude structure and greater integration of new and old attitude-relevant information, which fosters the accessibility and consistency of attitude-related beliefs (Eagly and 123 Chaiken 1995). In other words, the politically sophisticated or high sophisticates are more likely to gather political information and assess that information in light of previously held information or attitudes. Abraham Tesser and his colleagues (Tesser and Conlee 1975; Tesser and Cowan 1977; Tesser and Leone 1977), based on a series of timed experiments, demonstrate that increased thought about an issue may polarize one’s attitude toward it. The implication is that politically sophisticated individuals that spend time and effort thinking about an issue will increasingly exhibit polarized or extreme attitudes. Experiments by political psychologists exploring motivated or biased processing of political information further elaborates why the attitudes of politically sophisticated individuals become extreme. Taber and Lodge (2006; see also Taber et al 2009) find, as the result of a series of experiments, that attitude polarization is the result of biased processing. Essentially, Taber and Lodge argue that attitude polarization should be observed among the politically sophisticated because they are most able to assimilate congruent evidence and information while also being able to easily argue against incongruent information (see also Ditto and Lopez 1992; Rucker and Petty 2004). Low sophisticates, though, do not possess the ability or store of knowledge to be able to resist new information and attempts at persuasion, which leads to attitude moderation. High sophisticates, however, through the way they process and build stores of knowledge and information have built an asymmetrical bias towards their preferred attitude that is not easily overcome. In the language of the ELM model, high sophisticates have fully elaborated their attitudes to the point that they become crystallized making moderation unlikely. Repeatedly accessing an attitude that is built on an asymmetrical store of 124 knowledge leads high sophisticates toward attitude extremity. The attitudes of low sophisticates are expected to be moderate not extreme. This is because they are the least likely to repeatedly engage in the type of biased information processing that lead high sophisticates to develop extreme attitudes (Taber and Lodge 2006; Taber et al 2009). When a court hands down a landmark decision and then makes subsequent rulings in the same issue area the responses of high and low sophisticates are expected to be different. In the wake of a court’s landmark decision, the overall structure of attitudes of high sophisticates becomes set. The decision has, in effect, served to solidify their beliefs. Subsequent rulings will cause high sophisticates to repeatedly access their attitudes, which results in biased information processing that reinforces previously held positions and cause them to feel more strongly towards the issue. Each subsequent decision, then, is expected to lead to more extreme attitudes among high sophisticates. Extremity Hypothesis: The overall structure of opinion of high sophisticates will change, becoming more extreme, as a court issues subsequent decisions. Low sophisticates are not expected to respond to landmark or subsequent judicial decisions by becoming more extreme in the attitudes they hold. Each ruling issued after the landmark decision is expected to moderate the overall structure of opinion among low sophisticates. In other words, low sophisticates have not fully elaborated their attitudes to the point that subsequent information processing becomes biased towards a particular position leading towards extremity. 125 Moderation Hypothesis: The overall structure of opinion of low sophisticates will change, becoming less extreme or moderate, as a court issues subsequent decisions. 4.8 Data and Modeling Strategy: Extremity and Moderation Hypotheses In order to test the extremity and moderation hypotheses the dependent variable used in the preceding analysis the support for the transfer of land rights to Aborigines must be modified. This variable ranges on a 1 to 5 scale asking respondents whether the transfer of land rights had gone much too far, too far, about right, not far enough and not nearly far enough. The extremity and moderation hypotheses are interested in whether the overall structure of opinion has become more extreme or more moderate in response to judicial decisions. To capture extremity or moderation the 1 to 5 scale will be folded around the middle or moderate category, “about right”. This means that the most extreme categories, “gone much too far” and “not nearly far enough”, will be combined. The leaning categories, “too far” and “not far enough” will also be combined into a single category. In essence, a five-category variable has been transformed to a three-category variable ranging from the most moderate to the most extreme attitudes toward the transfer of land rights to Aborigines. Figure 5.3 reports the mean level of attitude extremity toward the transfer of land rights to Aborigines among both high and low sophisticates from 1990 through 2004. The overall mean level of attitude extremity among high sophisticates is .97 while the mean level for low sophisticates is 1.04. Interestingly, the figure indicates that it is in 126 fact the low sophisticates who, on average, held more extreme attitudes compared to the high sophisticates, which is contrary to the theoretical expectations of both the extremity and moderation hypotheses. The mean level of attitude extremity does increase over time for high sophisticates while the mean level decreases for low sophisticates. This does indicate that attitudes between the two groups are changing over time, which is theoretically expected. [Insert Figure 5.3 about here] Figure 5.4 reports the mean difference in attitude extremity amongst high and low sophisticates towards the transfer of land rights to Aborigines from 1990 through 2004. The figure indicates that high sophisticates were less extreme than low sophisticates, on average, until the Wik decision where the high sophisticates reversed course and became more extreme, on average, than the low sophisticates. Theoretically, the expectation is for the high sophisticates to be consistently more extreme in their attitudes than the low sophisticates, but the figure indicates this dynamic occurs after the Wik decision and not before. [Insert Figure 5.4 about here] The data comes from five Australian Election Studies conducted in 1990, 1996, 1998, 2001 and 2004. The same modeling strategy employed in the previous section will be employed here. Two surveys, one conducted before the decision and one conducted 127 after, are pooled together to empirically distinguish between group attitudes, predecision and postdecision, as a result of a judicial decision. A dummy variable indicating a decision is interacted with all of the independent variables and creates two sets of coefficients: predecision !1,k coefficients and postdecision !2,k coefficients. The interaction allows for a distinction between attitudes before and after a decision. The postdecision dummy variable equals 0 before the decision and 1 afterward and interacts with the !2,k coefficients (the postdecision individual characteristics). In the model, the coefficients are then able to vary between predecision and postdecision samples (Franklin and Kosaki 1989, 755-56; Johnson and Martin 1998, 302). When the postdecision dummy equals 1 the model is “unconstrained” as all of the model coefficients and interactions are present. The second estimated equation constrains the postdecision dummy to 0 and suppresses the interaction effect with the !2,k coefficients. This second equation is the constrained model. The results of the constrained model are only used and reported for use in a likelihood ratio test, which serves as the direct test of the extremity and moderation hypotheses. The results of the likelihood ratio tests indicate whether an overall change in attitudes occurs after a judicial decision. Given that the dependent variable consists of discrete ordinal numbers (0, 1 or 2), an ordered probit model is utilized because it takes into consideration that each observation is treated as a rank ordering. The models are estimated using maximum likelihood estimation. This modification of the dependent variable allows for statistical tests to capture the likelihood of moderate or extreme attitudes in response to judicial decisions. 128 Two models are reported for each hypothesis, one for each court decision (the Wik and Trio cases) addressing the extremity or moderation of opinion regarding the transfer of land rights or native title. The extremity hypothesis is assessed in terms of how high sophisticates respond to the High Court’s native title decisions. The expectation is for the likelihood-ratio test to be statistically significant for each model. Likewise, the moderation hypothesis is assessed in terms of how low sophisticates respond to the High Court’s decisions. The expectation is for the likelihood-ratio test to be statistically significant for each model. The principal independent variable in each model, for both high and low sophisticates, is the dichotomous case decision variable (the Wik and Trio cases Anderson, Ward and Yorta Yorta) and it is expected to be statistically significant. For high sophisticates, the case variable is expected to be positive indicating increased likelihood of attitude extremity. For low sophisticates, the case variable is expected to be negative, which indicates a decreased likelihood of attitude extremity (or an increased likelihood of moderation). As with the models in the previous section, a number of demographic and other control variables are included, such as gender, marital status and union membership. Measures of party identification are included as well as an indicator of whether a respondent works in the mining or farming industries. Previous studies of court influence suggest that polarization or differences between group attitudes should increase with each decision. Given the reactions of the major political parties to the native title decisions as well as historical support or opposition to Aboriginal policies there is an expectation that supporters of Labor and the Liberal-National coalition will be more likely to exhibit 129 extreme attitudes. Likewise, those individuals engaged in the mining and farming industries are likely to hold extreme attitudes towards the native title decisions as they have a real economic stake in the outcomes. Again though, there are no a priori assumptions about likelihood of extreme or moderate attitudes of females, married individuals or union members. 4.9 Results: Extremity and Moderation Hypotheses Both the extremity and moderation hypotheses expect that the overall structure of attitudes among both high and low sophisticate groups to change as a result of each of the High Court’s native title decisions. For high sophisticates, each decision is expected to further bolster their previously held attitudes toward the transfer of land rights and increase the likelihood of holding an extreme attitude. The expectation for low sophisticates is to be persuaded by each decision because their previously held attitudes are not firm enough to resist the influence of subsequent judicial decisions. On the individual level, there is an expectation that identifiers of the major political parties and those engaged in economic activity related to the decisions will likely become more extreme as a result of each decision. [Insert Tables 4.7 and 4.8 about here] Tables 4.7 and 4.8 report the results for high sophisticates and their likelihood of holding an extreme attitude with regard to the High Court’s native title rulings in the Wik and Trio cases. The expectation of the extremity hypothesis is for each decision to alter 130 the overall structure of opinion, which is assessed through a likelihood-ratio test as well as a case decision dummy variable. Confirming expectations, the likelihood-ratio tests are statistically significant for the Wik (p < 0.0043) and Trio (p < 0.0210) decisions. These indicate that the overall structure of attitude extremity did change before and after the latter two or subsequent High Court decisions. However, only the Wik decision variable was statistically significant. The Wik dummy was positively signed, which indicates that the decision contributed to a 28.3% increased likelihood of holding an extreme attitude toward the transfer of land rights to Aborigines. The Trio case dummy was not statistically significant, which does not allow for a causal attribution to be made regarding the overall change in opinion to the decision. While these results do not confirm the extremity hypothesis in each instance, the result of the Wik decision does confirm that the overall structure of opinion became more extreme as a result the High Court’s first subsequent ruling confirming the principle of native title. Contrary to general assertions made in the literature regarding between-group attitude polarization, these results indicate very little presence of such a phenomenon. Among high sophisticates, only did Liberal-National party identifiers in response to Wik and females in response to the Trio cases become more/less extreme compared to others. Liberal-National party identifiers were 16.6% more likely to hold extreme attitudes before the Wik decision, but, and contrary to expectations of polarization, they became 35.8% less likely than non-Liberal-National identifiers to hold extreme attitudes toward the transfer of land rights. Likewise, females before the Trio decisions were 10.7% less likely compared to non-females to hold extreme attitudes, but after the decisions females were 17.6% more likely to hold extreme attitudes compared to non-females. The general 131 lack of polarization among and between group attitudes stands contrary to the expectation of extremity that is the norm in the United States at least among high sophisticates. [Insert Tables 4.9 and 4.10 about here] Tables 4.9 and 4.10 report the results for low sophisticates and their likelihood of holding an extreme attitude with regard to the High Court’s native title rulings in the Wik and Trio cases. The expectation of the moderation hypothesis is for each decision to alter the overall structure of opinion, which is assessed through a likelihood-ratio test as well as a case decision dummy variable. Supportive of expectations, the likelihood-ratio tests are only statistically significant for the Wik (p < 0.0133) and Trio (p < 0.0768) decisions. The results of the likelihood ratio tests indicate that the overall structure of attitude extremity did change before and after the High Court’s subsequent native title decisions. Additionally, both the Wik and Trio case decision variables are statistically significant. However, the Wik dummy was positively signed, which indicates, contrary to the moderation hypothesis, that the decision increased the likelihood of extreme attitudes among low sophisticates by 7.3%. The Trio dummy was negatively signed and supports the moderation hypothesis of the decision decreasing attitude extremity or increasing the likelihood of attitude moderation by 10.7%. These results prove much more problematic for the moderation hypothesis as the Wik decision actually increased the likelihood of low sophisticates holding extreme attitudes. Consistent with the results from the high sophisticates, between-group attitude polarization among low sophisticates was not prominent. Among low sophisticates, only 132 Liberal-National party identifiers and married identifiers in response to the Wik decision were the only groups to exhibit between-group attitude changes. Both Liberal-National partisans and married individuals were 4.3% and 3.4% more likely to exhibit extreme attitudes before the decision, but both groups became less likely to exhibit extreme attitudes compared to non-identifiers after the decision, 8.6% and 5.4%, respectively. Again, the results point to an overall lack of polarization or attitude extremity in betweengroup attitudes among low sophisticates. 4.10 Discussion: Extremity and Moderation Hypotheses Both the extremity and moderation hypotheses marshaled a good deal of evidence to assess their validity, but neither can be fully accepted. The results from the empirical assessment of both hypotheses raises some interesting questions though. Both hypotheses asserted that the overall structure of attitudes would change in response to each subsequent High Court decision. Consistent with the extremity hypothesis, but counter to the moderation hypothesis, the High Court’s subsequent Wik decision resulted in an increased likelihood of attitude extremity among both high and low sophisticates. While the Trio decisions resulted in a decreased likelihood of attitude extremity among low sophisticates, but was inconclusive among high sophisticates. The results seem to suggest that the first subsequent decision may be the most important in terms of assessing how extreme overall attitudes within the public may become. The Wik decision affirmed the principle of native title, but limited its potential reach and impact by allowing pastoral leases to nullify or extinguish potential claims. Certainly, the decision left much to be desired by opponents and proponents of native 133 title, but it did give something to each side. Perhaps, the fact that the High Court’s decision split the difference led the public to become much further entrenched toward the side of the issue that they supported. It should also be noted that between-group attitudes, either among high or low sophisticates, did not uniformly result in more or less attitude extremity. The only group that seemed to consistently shift in response to the decisions, even across sophistication level, was identifiers of the Liberal-National coalition. This may be in response to those parties historical opposition to Aboriginal policies and a general hostility towards the High Court’s native title decisions. 4.11 Conclusion The Elaboration Likelihood Model has served as the cornerstone for scholars seeking to explain the process of judicial influence on public opinion. The ELM assumes that individuals are motivated to hold correct attitudes, but the ability and effort that goes into processing of attitude-relevant information varies by individual characteristics. It is the combination of several characteristics – ability, opportunity and motivation – that determine how well formed or elaborated attitudes will be. In other words, the strength and durability of attitudes are determined by how much effort goes into forming them. Political Scientists term this political sophistication and frequently observe differences between high and low sophistication levels. As a result, distinctions can be made between individuals based on their propensity to engage in this kind of attitude formation. Yet, studies of judicial influence of public opinion has not taken advantage of or explored 134 how the differences in high and low sophisticates may affect how court decisions are received by the public. The first section of this chapter assesses the extent to which high and low sophisticates may react differently in response to landmark and subsequent judicial decisions. The crystallization hypothesis theorizes that high sophisticates will behave exactly as Johnson and Martin’s (1998) conditional and subsequent hypotheses suggest: a landmark decision will affect the overall structure of opinion while subsequent decisions will have no affect. The volatility hypothesis suggested that the overall structure of opinion of low sophisticates would be more volatile and respond to each judicial decision regardless of its status. The results of both empirical investigations reveal differences in the behavior of high and low sophisticates, but not in the manner anticipated by the crystallization and volatility hypotheses. For both high and low sophisticates the likelihood-ratio tests revealed that the overall structure of opinion had changed after a decision, but the case decision variables were not statistically significant, preventing a link being made between the overall change and the High Court’s decisions. The second section of this chapter engaged with the question of polarization or attitude extremity. It has become commonplace in studies of court influence on the public to observe differences in group attitudes after decisions, which is often considered to be evidence of polarization. I suggest that attitude polarization may be more properly thought of as attitude extremity since it is the movement towards extreme positions that is of particular interest. The extremity and moderation hypotheses refer to the expected behavior of high and low sophisticates in response to subsequent judicial decisions. High sophisticates are expected to exhibit more extreme attitudes in the face of subsequent 135 decisions as their prior attitudes become reinforced as the result of repeated information processing. Low sophisticates, on the other hand, are expected to exhibit more moderate attitudes as a result of less rigid information processing. Though neither of the hypotheses could be fully accepted the behavior of both high and low sophisticates proved interesting. An overall change in the opinion structures of both high and low sophisticates occurred in response to the High Court’s Wik decision leading towards an increased likelihood of attitude extremity. The first subsequent case in which the High Court reexamined the question of native title clearly led both high and low sophisticates to revisit their attitudes on the transfer of land rights to Aborigines. Among both sophistication groups as well as Liberal-National coalition party identifiers, the decision did not result in a unified consensus or agreement about native title. Instead the decision led to an increased likelihood of and support for extreme positions on the transfer of land rights to Aborigines. Far from settling the issue of native title amongst the public, the High Court’s Wik decision led to increasingly divisive attitudes towards the issue. This is hardly the response the High Court expected would occur by bringing the issue of native title onto the national stage for public discourse. The High Court clearly thought that contemporary Australian values would support their effort to right an historical wrong (Patapan 2000, 117; Russell 2000, 250), but instead their decisions wrought a deeply divided public. Chapter 5 develops a theory based on geographic proximity (Hoekstra 2003; Hoekstra and Segal 1996) and the politicization of those environments (Hopkins 2010) that helps people connect their everyday experiences to judicial decisions. The 136 sociodemographic makeup of those environments will help or hinder judicial influence of opinion. 137 4.12 Tables and Figures TABLE 4.1 High Sophisticates: Likelihood of Support for the Transfer of Land Rights to Aborigines, Pre- and Post-Mabo v. Queensland Constrained Model Unconstrained Model Parameter Coeff. Prob. Coeff. Prob. Mabo -.18 (.12) .124 -.23 (.35) .508 Labor Party .26 (.15) .086 -.00 (.27) .990 Labor Party Postdecision --.44 (.33) .177 Liberal-National Party -.74 (.15) .000 -.89 (.27) .001 Lib-Nat Party Postdecision --.22 (.32) .507 Union .15 (.12) .226 .55 (.18) .003 Union Postdecision ---.67 (.25) .006 Female .16 (.11) .163 .37 (.18) .035 Female Postdecision ---.32 (.23) .170 Married -.10 (.12) .407 -.32 (.21) .124 Married Postdecision --.24 (.26) .363 Mine/Farm Worker -.85 (.48) .078 -1.29 (.72) .074 Mine/Farm Postdecision --.97 (1.03) .347 Cut 1 Cut 2 Cut 3 Cut 4 Pseudo R-square Log Likelihood Chi-Square Number of Observations Likelihood-Ratio Test -1.21 (.19) -.48 (.19) .51 (.19) 1.50 (.20) -1.32 (.30) -.57 (.30) .44 (.30) 1.42 (.31) 0.07 -517.43 77.42 368 0.08 -511.17 89.95 368 12.53 0.0511 Note: Coefficients are probit coefficients. Standard errors are in parenthesis. The dependent variable is coded 1 through 5 if respondents answered the transfer of land rights to Aborigines has gone much too far, too far, about right, not far enough, not nearly far enough. Source: Australian Election Studies 1990, 1996 138 TABLE 4.2 High Sophisticates: Likelihood of Support for the Transfer of Land Rights to Aborigines, Pre- and Post-Wik Peoples v. Queensland Constrained Model Unconstrained Model Parameter Coeff. Prob. Coeff. Prob. Wik .28 (.10) .005 .05 (.26) .865 Labor Party .41 (.13) .002 .40 (.19) .032 Labor Party Postdecision --.06 (.27) .832 Liberal-National Party -.45 (.13) .001 -.64 (.19) .001 Lib-Nat Party --.38 (.27) .160 Postdecision Union -.04 (.11) .743 -.13 (.16) .423 Union Postdecision --.20 (.23) .385 Female .05 (.10) .615 .05 (.15) .736 Female Postdecision ---.00 (.20) 1.000 Married -.07 (.11) .524 -.08 (.16) .607 Married Postdecision --.02 (.21) .925 Mine/Farm Worker -.41 (.37) .267 -.28 (.73) .700 Mine/Farm Postdecision ---.22 (.85) .798 Cut 1 Cut 2 Cut 3 Cut 4 Pseudo R-square Log Likelihood Chi-Square Number of Observations Likelihood-Ratio Test -.94 (.15) -.16 (.14) .72 (.15) 1.53 (.16) -1.05 (.20) -.27 (.19) .62 (.20) 1.42 (.21) 0.05 -675.68 65.44 459 0.05 -674.18 68.45 459 3.01 0.8070 Note: Coefficients are probit coefficients. Standard errors are in parenthesis. The dependent variable is coded 1 through 5 if respondents answered the transfer of land rights to Aborigines has gone much too far, too far, about right, not far enough, not nearly far enough. Source: Australian Election Studies 1996, 1998 139 TABLE 4.3 High Sophisticates: Likelihood of Support for the Transfer of Land Rights to Aborigines, Pre- and Post-Ward, Anderson and Yorta Yorta (Trio) Constrained Model Unconstrained Model Parameter Coeff. Prob. Coeff. Prob. Trio .16 (.09) .081 -.03 (.25) .894 Labor Party .07 (.12) .540 -.09 (.16) .588 Labor Party Postdecision --.36 (.24) .138 Liberal-National Party -.91 (.12) .000 -.94 (.17) .000 Lib-Nat Party Postdecision --.06 (.23) .783 Union .25 (.10) .015 .26 (.15) .076 Union Postdecision ---.03 (.21) .905 Female .01 (.09) .955 -.06 (.13) .661 Female Postdecision --.11 (.19) .550 Married -.14 (.10) .162 -.12 (.14) .412 Married Postdecision ---.01 (.20) .969 Mine/Farm Worker -.54 (.80) .495 -5.42 (188.68) .977 Mine/Farm Postdecision --5.79 (188.69) .976 Cut 1 Cut 2 Cut 3 Cut 4 Pseudo R-square Log Likelihood Chi-Square Number of Observations Likelihood-Ratio Test -1.67 (.15) -.83 (.14) .04 (.14) .88 (.14) -1.76 (.18) -.92 (.18) -.05 (.17) .80 (.18) 0.07 -770.33 118.86 533 0.08 -767.64 124.25 533 5.39 0.4952 Note: Coefficients are probit coefficients. Standard errors are in parenthesis. The dependent variable is coded 1 through 5 if respondents answered the transfer of land rights to Aborigines has gone much too far, too far, about right, not far enough, not nearly far enough. Source: Australian Election Studies 2001, 2004 140 TABLE 4.4 Low Sophisticates: Likelihood of Support for the Transfer of Land Rights to Aborigines, Pre- and Post-Mabo v. Queensland Constrained Model Unconstrained Model Parameter Coeff. Prob. Coeff. Prob. Mabo -.14 (.04) .000 -.19 (.12) .124 Labor Party .06 (.06) .317 -.11 (.08) .177 Labor Party Postdecision --.33 (.11) .003 Liberal-National Party -.44 (.06) .000 -.55 (.08) .000 Lib-Nat Party Postdecision --.19 (.11) .093 Union .03 (.04) .490 .10 (.06) .085 Union Postdecision ---.17 (.09) .052 Female .16 (.04) .000 .19 (.05) .001 Female Postdecision ---.05 (.08) .546 Married -.16 (.04) .000 -.11 (.06) .081 Married Postdecision ---.12 (.09) .178 Mine/Farm Worker -.06 (.11) .560 -.06 (.14) .673 Mine/Farm Postdecision ---.01 (.22) .960 Cut 1 Cut 2 Cut 3 Cut 4 Pseudo R-square Log Likelihood Chi-Square Number of Observations Likelihood-Ratio Test -.72 (.07) .01 (.07) .91 (.07) 1.75 (.08) -.76 (.09) -.03 (.09) .88 (.09) 1.72 (.10) 0.03 -4082.87 189.62 2964 0.02 -4076.05 203.26 2964 13.64 0.0339 Note: Coefficients are probit coefficients. Standard errors are in parenthesis. The dependent variable is coded 1 through 5 if respondents answered the transfer of land rights to Aborigines has gone much too far, too far, about right, not far enough, not nearly far enough. Source: Australian Election Studies 1990, 1996 141 TABLE 4.5 Low Sophisticates: Likelihood of Support for the Transfer of Land Rights to Aborigines, Pre- and Post-Wik Peoples v. Queensland Constrained Model Unconstrained Model Parameter Coeff. Prob. Coeff. Prob. Wik .20 (.04) .000 -.03 (.12) .786 Labor Party .31 (.05) .000 .22 (.08) .005 Labor Party Postdecision --.17 (.11) .128 Liberal-National Party -.19 (.05) .001 -.35 (.08) .000 Lib-Nat Party --.31 (.11) .004 Postdecision Union .04 (.05) .435 .07 (.07) .280 Union Postdecision --.21 (.10) .027 Female .10 (.04) .012 .13 (.06) .025 Female Postdecision ---.06 (.08) .466 Married -.19 (.05) .000 -.21 (.06) .001 Married Postdecision --.04 (.09) .624 Mine/Farm Worker -.20 (.04) .097 -.08 (.17) .624 Mine/Farm Postdecision ---.27 (.25) .281 Cut 1 Cut 2 Cut 3 Cut 4 Pseudo R-square Log Likelihood Chi-Square Number of Observations Likelihood-Ratio Test -.44 (.06) .35 (.06) 1.15 (.07) 1.88 (.07) -.57 (.09) .22 (.09) 1.02 (.09) 1.76 (.09) 0.02 -3846.11 172.56 2736 0.02 -3839.01 186.77 2736 13.64 0.0339 Note: Coefficients are probit coefficients. Standard errors are in parenthesis. The dependent variable is coded 1 through 5 if respondents answered the transfer of land rights to Aborigines has gone much too far, too far, about right, not far enough, not nearly far enough. Source: Australian Election Studies 1996, 1998 142 TABLE 4.6 Low Sophisticates: Likelihood of Support for the Transfer of Land Rights to Aborigines, Pre- and Post-Ward, Anderson and Yorta Yorta (Trio) Constrained Model Unconstrained Model Parameter Coeff. Prob. Coeff. Prob. Trio .12 (.04) .003 .19 (.11) .084 Labor Party .11 (.05) .038 .08 (.07) .253 Labor Party Postdecision --.07 (.11) .502 Liberal-National Party -.49 (.05) .000 -.42 (.07) .000 Lib-Nat Party Postdecision ---.14 (.10) .173 Union -.02 (.05) .667 .06 (.07) .334 Union Postdecision ---.19 (.10) .054 Female .19 (.04) .000 .19 (.06) .001 Female Postdecision ---.01 (.08) .903 Married -.13 (.04) .003 -.14 (.06) .018 Married Postdecision --.02 (.09) .788 Mine/Farm Worker -.26 (.12) .029 -.28 (.17) .091 Mine/Farm Postdecision --.07 (.24) .781 Cut 1 Cut 2 Cut 3 Cut 4 Pseudo R-square Log Likelihood Chi-Square Number of Observations Likelihood-Ratio Test -.91 (.06) -.12 (.06) .82 (.06) 1.59 (.07) -.88 (.08) -.09 (.08) .85 (.08) 1.62 (.08) 0.03 -3969.27 235.29 2763 0.03 3965.41 243.00 2763 7.72 0.2597 Note: Coefficients are probit coefficients. Standard errors are in parenthesis. The dependent variable is coded 1 through 5 if respondents answered the transfer of land rights to Aborigines has gone much too far, too far, about right, not far enough, not nearly far enough. Source: Australian Election Studies 2001, 2004 143 TABLE 4.7 High Sophisticates: Likelihood of an Extreme Attitude for the Transfer of Land Rights to Aborigines, Pre- and Post-Wik Peoples v. Queensland Parameter Wik Labor Party Labor Party Postdecision Liberal-National Party Lib-Nat Party Postdecision Union Union Postdecision Female Female Postdecision Married Married Postdecision Mine/Farm Worker Mine/Farm Postdecision Cut 1 Cut 2 Pseudo R-square Log Likelihood Chi-Square Number of Observations Likelihood-Ratio Test Constrained Model Coeff. Prob. .11 (.11) .313 -.13 (.14) .349 ---.04 (.14) .795 --.22 (.12) .063 --.03 (.11) .812 --.04 (.11) .731 ---.14 (.39) .726 --- Unconstrained Model Coeff. Prob. .81 (.28) .004 -.02 (.20) .919 -.29 (.29) .310 .48 (.20) .018 -1.03 (.29) .000 .27 (.17) .118 -.11 (.24) .654 -.00 (.16) .978 .09 (.21) .675 .20 (.17) .247 -.31 (.23) .167 .07 (.76) .926 -.26 (.88) .771 -.40 (.15) .73 (.15) -.08 (.21) 1.08 (.21) 0.01 -492.22 4.96 459 0.02 -482.77 23.87 459 18.91 0.0043 Note: Coefficients are probit coefficients. Standard errors are in parenthesis. The dependent variable folds a 5 category variable around the middle category to create 3 categories and is coded 1 through 3 if respondents answered the transfer of land rights to Aborigines has been about right (1), gone too far/not far enough (2) and gone much too far/not nearly far enough (3). Source: Australian Election Studies 1996, 1998 144 TABLE 4.8 High Sophisticates: Likelihood of an Extreme Attitude for the Transfer of Land Rights to Aborigines, Pre- and Post-Ward, Anderson and Yorta Yorta (Trio) Parameter Trio Labor Party Labor Party Postdecision Liberal-National Party Lib-Nat Party Postdecision Union Union Postdecision Female Female Postdecision Married Married Postdecision Mine/Farm Worker Mine/Farm Postdecision Cut 1 Cut 2 Pseudo R-square Log Likelihood Chi-Square Number of Observations Likelihood-Ratio Test Constrained Model Coeff. Prob. .03 (.10) .730 .13 (.13) .324 ---.08 (.12) .527 --.12 (.11) .281 ---.06 (.10) .526 ---.09 (.10) .408 --.15 (.82) .851 --- Unconstrained Model Coeff. Prob. -.19 (.26) .463 .04 (.18) .825 .23 (.26) .374 .04 (.18) .825 -.17 (.25) .491 .23 (.16) .141 -.21 (.22) .340 -.31 (.14) .024 .51 (.20) .009 -.09 (.15) .552 .04 (.21) .841 6.80 (9835.67) .999 -13.55 (13897.63) .999 -.57 (.14) .61 (.14) -.66 (.18) .54 (.18) 0.01 -568.65 6.39 533 0.02 -561.20 21.29 533 14.91 0.0210 Note: Coefficients are probit coefficients. Standard errors are in parenthesis. The dependent variable folds a 5 category variable around the middle category to create 3 categories and is coded 1 through 3 if respondents answered the transfer of land rights to Aborigines has been about right (1), gone too far/not far enough (2) and gone much too far/not nearly far enough (3). Source: Australian Election Studies 2001, 2004 145 TABLE 4.9 Low Sophisticates: Likelihood of an Extreme Attitude for the Transfer of Land Rights to Aborigines, Pre- and Post-Wik Peoples v. Queensland Constrained Model Unconstrained Model Parameter Coeff. Prob. Coeff. Prob. Wik -.04 (.04) .364 .24 (.12) .054 Labor Party -.20 (.06) .000 -.22 (.08) .007 Labor Party Postdecision --.03 (.11) .816 Liberal-National Party -.00 (.06) .947 .14 (.08) .087 Lib-Nat Party Postdecision ---.28 (.11) .014 Union .02 (.05) .681 .10 (.07) .155 Union Postdecision ---.14 (.10) .153 Female -.06 (.04) .172 -.05 (.06) .472 Female Postdecision ---.02 (.09) .782 Married .02 (.05) .657 .11 (.07) .095 Married Postdecision ---.17 (.09) .065 Mine/Farm Worker .28 (.13) .025 .35 (.17) .041 Mine/Farm Postdecision ---.15 (.25) .565 Cut 1 Cut 2 -.81 (.07) .28 (.07) -.66 (.09) .43 (.09) Pseudo R-square Log Likelihood Chi-Square Number of Observations 0.01 -2930.49 30.24 2736 0.01 -2922.45 46.33 2736 Likelihood-Ratio Test 16.09 0.0133 Note: Coefficients are probit coefficients. Standard errors are in parenthesis. The dependent variable folds a 5 category variable around the middle category to create 3 categories and is coded 1 through 3 if respondents answered the transfer of land rights to Aborigines has been about right (1), gone too far/not far enough (2) and gone much too far/not nearly far enough (3). Source: Australian Election Studies 1996, 1998 146 TABLE 4.10 Low Sophisticates: Likelihood of an Extreme Attitude for the Transfer of Land Rights to Aborigines, Pre- and Post-Ward, Anderson and Yorta Yorta (Trio) Constrained Model Unconstrained Model Parameter Coeff. Prob. Coeff. Prob. Trio -.06 (.04) .157 -.31 (.11) .008 Labor Party -.06 (.06) .271 -.13 (.08) .104 Labor Party Postdecision --.13 (.11) .253 Liberal-National Party .10 (.05) .070 -.04 (.07) .613 Lib-Nat Party Postdecision --.29 (.11) .007 Union .06 (.05) .271 -.01 (.07) .859 Union Postdecision --.15 (.10) .151 Female -.11 (.04) .012 -.10 (.06) .090 Female Postdecision ---.02 (.09) .808 Married .04 (.05) .374 .00 (.06) .952 Married Postdecision --.09 (.09) .350 Mine/Farm Worker .07 (.12) .548 .15 (.17) .393 Mine/Farm Postdecision ---.17 (.24) .488 Cut 1 Cut 2 -.51 (.06) .57 (.06) -.63 (.08) .46 (.08) Pseudo R-square Log Likelihood Chi-Square Number of Observations 0.00 -2988.13 22.14 2763 0.01 -2982.43 33.54 2763 Likelihood-Ratio Test 11.40 0.0768 Note: Coefficients are probit coefficients. Standard errors are in parenthesis. The dependent variable folds a 5 category variable around the middle category to create 3 categories and is coded 1 through 3 if respondents answered the transfer of land rights to Aborigines has been about right (1), gone too far/not far enough (2) and gone much too far/not nearly far enough (3). Source: Australian Election Studies 2001, 2004 147 FIGURE 4.1 Mean Support for the Transfer of Land Rights Between High and Low Sophisticates, 1990-2004 148 FIGURE 4.2 Mean Difference in Support for the Transfer of Land Rights Between High and Low Sophisticates, 1990-2004 149 FIGURE 4.3 Mean Level of Attitude Extremity in Support for the Transfer of Land Rights Between High and Low Sophisticates, 1990-2004 150 FIGURE 4.4 Mean Difference in Level of Attitude Extremity in Support for the Transfer of Land Rights Between High and Low Sophisticates, 19902004 151 Chapter 5 Context and Place 5.1 Introduction “Geoffrey Blainey, a notable [historian and] opponent of Mabo, warned that Western Australians ‘are unlikely to see Mabo in the same light as residents of Canberra and Sydney’ because they ‘have so much to lose’”(Goot 1993b, 195). The statement is provocative and attention grabbing, but is it true? Why would people living in Western Australia view the High Court’s ruling and the issue of native title any differently than those living in a population center like Sydney, New South Wales or the Australian capital, Canberra? Or more to the point, why would people living in Western Australia be so concerned about a judicial decision that the state of Queensland, not Western Australia, was involved in? Goot (1993b, 201) notes that both Western Australia and Queensland share some pertinent contextual characteristics related to the native title issue: they have the second and third largest Aboriginal populations, they have the two largest state-based mining industries and have had long, well publicized, histories of Aboriginal-European conflict. In other words, the issue of land rights for Aborigines is unlikely to garner support in areas where claims to native title appear most likely. Blainey’s assertion touches on a little explored aspect of court influence of public opinion: the role of context in shaping reactions to and acceptance of judicial decisions. Judicial decisions and the public reception of those decisions do not occur in a vacuum, but is, and understandably so, shaped by individual evaluations of those decisions in light of their own experiences. For many people, their context or the makeup of their daily environment, their town, community or even state, will come to shape how court 152 decisions are viewed. Research on the effect of Supreme Court decisions on local opinion suggests that support for a decision is moderated by how closely an individual lives to the people and the community involved in the case (Hoekstra 2003, 95-99; see also Hoekstra and Segal 1996). After all, salient landmark decisions, like those involving native title, may invite national attention (Fleming et al 1997) because of their statutory or constitutional importance, but how the case is decided and interpreted by the public has local and regional consequences. Whether or not is able to influence the public may depend on those local and regional conditions. Scholars of judicial compliance and acceptance of decisions have noted the success or failure of converting the public depends largely on the sociodemographic makeup of localities (Canon and Johnson 1999; McGuire 2008). In this sense, the context that individuals find themselves may prove to be a hurdle to judicial influence of public attitudes. Whether individual context operates independently or whether it must be activated by external factors such as judicial decisions or national media coverage is invested here. This chapter thus explores the extent to which context rivals judicial decisions to affect public attitudes on the issue of native title land rights. 5.2 Perceiving the Threat of Native Title Racial threat or power threat are considered a subset of theories of realistic group conflict whose intellectual roots can be firmly placed within the research of Key (1949) and Blalock (1967). The central claim of these theories is that the presence of an outgroup in sufficient numbers will generate competition for scarce resources and thus engender hostility. Threat might be especially severe in places of relative or increasing resource 153 deprivation (Branton and Jones 2005; Gay 2006) or of rising outgroup political power (Key 1949), but the mechanisms that engender or facilitate threat tend to be localized by geography. Recent research on racial threat has oscillated around four primary positions. The first suggests that geographic proximity acts primarily by triggering political competition (Glaser 1994). The second holds that geographic proximity triggers a more diffuse, undifferentiated prejudice (Taylor 1998). The third argues that ethnic and racial diversity can stifle cohesion among both outgroups and ingroups (Putnam 2007). The fourth challenges the relevance of racial threat altogether (Voss 1996) and focuses instead on the role of socioeconomic contexts in shaping racial attitudes (Oliver and Mendelberg 2000). For classical threat to operate, first, an outgroup must be identified by the majority or dominant ingroup; second, people must perceive an outgroup as a threat to political or economic resources; and third, people must be aware of their ethnic and racial contexts. Aboriginal Australians have unquestionably been marked as the outgroup in Australia, in historical or contemporary terms. There is, and has been, a significant level of prejudice across Australian society toward Aborigines, who have been described as “by far the most ‘Outsider’ group in Australian society” (Angelico 1995, 253; see also Hamilton 1990; Rizvi 1996). Research in Australia has repeatedly and over time uncovered a deep reservoir of hostility, resentment and racism directed toward Aboriginal Australians. Much of this negativity is sustained through negative stereotypes that suggest Aborigines are dependent on special treatment in the form of government handouts, are drunkards and 154 are unwilling to assimilate (Larsen 1981; Pedersen et al 2000; Pedersen et al 2006). Some research suggests that these negative attitudes may actually be the result of collective guilt about past and present wrongs to Aboriginal Australians (Pedersen et al 1 2004). This is especially noteworthy given the position of the High Court, which asserted that native title and recognition of land rights may be a way to atone for past injustice (Patapan 2000, 126). Yet, old-fashioned and modern racism persists throughout Australian society (Mellor 2003; Pedersen and Walker 1997; Pedersen et al 2000; Walker 1994; Western 1969). Individual acts of discrimination – such as avoiding sitting next to an Australian Aborigine on a bus or refusing to serve them in a restaurant (Mellor 2003), or wanting the government to provide less services to Australian Aborigines (Gibson et al 2002), or voting for a political party that is hostile to Australian Aborigines (Jackman 1998; Goot and Watson 2001) – are diffused throughout the public and over time result in the maintenance of clear differences between groups. The potential economic consequences of native title were laid bare by entities with mining interests. In the wake of the High Court’s Mabo (No.2) decision, the discussions “have almost exclusively been conducted with a view to establishing the effect of native title upon the mining industry” (Coombs 1994, 111), which was setting off the proverbial economic alarm bells about how native title threatened new mining projects (Russell 2005, 291). Several full-page advertisements were placed by the AMIC in major news papers asserting that Mabo had created a crisis for the mining industry, more native title claims would arise and investment would be frightened off (Mercer 1 Recent examples of historical injustices committed against Australian Aborigines include the child removal policy that operated from 1869 to 1969 as well as the more general White Australia Policy, which finally ended with the adoption of the Racial Discrimination Act of 1975. 155 1997, 14). When an industry that accounts for approximately 8.8 percent of Australia’s gross domestic product and employs 3.8 percent of the country’s workforce, asserts that 2 the country will suffer as a result of native title its protestations are not easily ignored. This is especially the case when a group mostly seen in a negative light is seen as the source or cause of likely economic calamity. Lastly, Australians have tended to exaggerate the relative population size of the 3 Aboriginal population by almost an order of magnitude. For example, Goot (1993a, 143) presents the results of a poll that indicated the mean estimate of the Aboriginal population in 1993 to be 13 percent when the actual population size was closer to 1.5 percent. 4 If the majority of Australians think there are significant numbers of Aborigines then granting native title land rights may be seen as a threat to property and resources. 5 Despite the negative perception and attention that Aborigines receive, it is not absolutely clear that their presence will generate opposition towards native title. According to contact theory, the more one gets to know personally individual members of a minority group, the less likely one is to be prejudiced against that minority group 2 Australian Bureau of Statistics. 2000. The Australian Mining Industry: From Settlement to 2000. Cat. No. 8414.0. 3 This phenomenon is a familiar one in the United States especially regarding the majority white population’s assessment of minority population size (see Wong 2007). 4 Population statistics were gathered from the Australian Bureau of Statistics (2008), which conducts a census every five years. Population data from 1986 to 2006 was used. Aboriginal Australians made up 1.46% of the population in 1986 or numbered 227,645 out of 15,602,156 and have steadily increased in population size so much so that in 2006 they made up 2.29% of the population or numbered 454,795 out of 19,853,069. 5 The argument advanced has its roots in theories of racial threat that claim the presence of an outgroup in sufficient numbers will generate competition for scarce resources and thus generate hostility, which in this context will result in less support for native title land rights (Blalock 1967; Key 1949). 156 (Allport 1954). To the extent that contact theory holds (Pettigrew 1998; Stein et al 2000), encounters with Aborigines might reduce negative outgroup attitudes and grow support for an issue like native title. This is an admittedly optimistic view. 5.3 Politicized Places The politicized places hypothesis asserts that contextual effects on individual opinion will vary with the national salience of an issue (Hopkins 2010, 43). More specifically, the theory asserts that when an issue becomes nationally politicized individuals will become more aware of the makeup of the environment in which they live and find it easier to draw conclusions about that issue from their everyday experiences. The politicized places hypothesis, to date, has only been applied to the issue of immigration and the adoption of anti-immigration ordinances, both in the United States (Hopkins 2010; Hopkins 2011b) and the United Kingdom (Hopkins 2011a). Given that the theory is partly grounded in realistic group conflict theories, it stands to reason that any context featuring both an ingroup and outgroup is susceptible to fluctuations in political hostility. What sets the politicized places approach apart from theories of realistic group conflict is that negative attitudes towards the outgroup are activated or made salient by the combination of the outgroup’s presence and national political attention to and discussion of an issue related to the outgroup. At other times, when an issue is not nationally salient then individuals do not necessarily connect the presence of the outgroup to the issue. This approach assumes that people are highly selective in incorporating environmental information and that information acquisition needs to be explained (Hopkins 2010, 42). In order for people to connect their daily environment to changes in 157 politics, they need the media to “define what the problem is and how to think about it” (Kinder 1998, 170). Past research has demonstrated that the media can play a key role in politicizing day-to-day events and experiences (Iyengar and Kinder 1987; Kinder 1998; Mutz 1994). In terms of the native title issue and debate, the High Court’s decisions along with help from the media put the aboriginal problem front and center in the country’s political life (Russell 2005, 277). The country’s major newspapers regularly featured hysterical reactions to hypothetical and outrageous Aboriginal land claims over much of Australia. Headlines from major newspapers and television included: ‘the decision has the potential to destroy our society’, ‘80% of Western Australia could be claimed’, and ‘Many mining projects are at risk’ (Russell 2005, 280). Former minister for indigenous affairs, Robert Tickner (2001, 94) commented, “The reporting of the native title debate was…abysmal. It reached its lowest point when the front page of the Sydney Sunday paper seriously reported a Mabo land claim over Sydney Opera House, which was without legal foundation of any kind”. The reporting may have been abysmal, but the media undoubtedly brought and maintained the spotlight of national attention to the issue of native title land rights for Aboriginal Australians. In effect, an individual becomes aware of the sociodemographic makeup of his or her environment as a result of national controversy made salient by media coverage and then uses that contextual information to form an opinion on the controversy. This does not mean that the emergence of a salient national issue will have the same affect on all contexts, but will instead vary with respect to the sociodemographic character of the place in question. In other words, the extent to which a national controversy becomes 158 salient for an individual depends on whether his or her context contains elements specific to the controversy. In the case of native title land rights for Aborigines, an individual that lives in a state or territory with a large number of Aboriginal Australians and is relatively unpopulated may exhibit more negative attitudes towards land rights than an individual who lives in a populace area with few Aborigines. 5.4 Contextual Proximity to a Case Hoekstra (2003, 95-99; see also Hoekstra and Segal 1996), in her study of Supreme Court decisions on local opinion, argues that support for a decision is moderated by the geographic proximity of respondents to the place of the case’s origin. In two of her studies, Center Moriches and Monroe, those respondents who lived in the immediate community where the case originated were less likely to be influenced by the Supreme Court’s decision than those respondents who lived in the surrounding and outlying communities. In Chapter 3, tests of the conditional and subsequent hypotheses included a dummy variable indicating the state where the native title case originated in order to assess Hoekstra’s geographic proximity hypothesis. For the Mabo and Wik cases, respondents from Queensland, where both cases originated, were expected to be less supportive of native title land rights before and after the High Court’s decisions than those respondents living in other states and territories. The predecision state dummy variable for both cases was consistent with expectations, but not after the decisions were handed down. For both cases, respondents from Queensland were not statistically distinguishable from respondents in other states postdecision. 159 Initially, these results might suggest that the geographic proximity to a case cannot be adequately captured using the state or territory as the relevant contextual unit. The fact that geographic proximity has not been incorporated into large cross-national 6 studies supports this observation . However, the reason for this is straightforward: in the United States, surveys commonly used to analyze these sorts of questions do not reveal 7 the respondent’s place of residence . As a result, past analyses have been unable to incorporate place as a means of or a factor relating to court influence of public opinion. Another explanation might be that a dummy variable indicating what state or territory a respondent lives in may not adequately represent the concept Hoekstra was seeking to capture. Since her study was focused on the immediate and surrounding communities, the populations she compared were small and relatively homogenous. By moving the unit of contextual analysis from a small town to a state and surrounding towns to surrounding states the populations under comparison become much more varied and complex. This shift suggests accounting for contextual or environmental characteristics associated with the geographic units. After all, the theory that geographic proximity moderates the influence of court decisions recognizes the importance of individual and situational factors in the attitude formation process. Those individual and situational factors are deeply connected to contextual makeup of his or her environment. Respondents living in communities or 6 All three major studies of court influence on public opinion, Franklin and Kosaki (1989), Johnson and Martin (1998) and Stoutenborough et al (2006) have used nationallevel cross-sectional surveys, which omit the respondent’s place of residence. 7 For example, the General Social Survey (GSS) does not reveal the state of residence of respondents only the geographic region. Recently, though, the GSS has changed this policy and will now release this information for a substantial additional fee and usage agreement. 160 areas similar to where a case originated may be more likely to identify with the people and places involved (Boninger et al 1995) and hence will be more likely to personalize the case as opposed to viewing the case as if it was happening in some other place to some other people. Those individuals may have a greater motivation to pay attention to and spend time thinking about the issue and its possible implications (Petty and Cacioppo 1986b; Fiske and Taylor 2008; Krosnick et al 1993) both for themselves and for their community. The implication for the High Court’s native title decisions is that context and the surrounding environment plays a mediating role in the persuasion process. If an individual thinks about the issues involved in a case and connects some aspect to his or her local environment then that person will be less likely to be influenced by a court decision. In other words, individuals will exhibit more negative attitudes toward native title land rights if they connect the decisions with their surroundings. 8 5.5 Data, Measurement and Model In order to assess the role of context for each of the above set of hypotheses data from both surveys and newspapers was required. The analysis makes use six nationally representative cross-sectional surveys that included the same question about land rights for Aborigines: the 1990, 1996, 1998, 2001, 2004 and 2007 Australian Election Studies. Having six surveys allows for comparison of contextual effects over time, which is critical given the hypotheses under investigation. These surveys also include information 8 Summary statistics for all variables used in the analysis can be found in Table 1B in Appendix B. 161 on the state or territory where a respondent resides, which will serve as the relevant contextual unit for the analysis. To measure the salience of native title land rights, an index of yearly mentions of native title by three of Australia’s largest daily newspapers: The Sydney Morning Herald based in New South Wales, The Courier Mail based in Queensland and The Advertiser based in South Australia. Specifically, the LexisNexis database was used to identify all stories mentioning “native title” or “land rights” for each year from January 1990 through December 2007. The average number of native title-related stories per year for The Sydney Morning Herald (SMH) during the 1990-2007 period is 237.1. For The Courier Mail (TCM), the figure is 272.8, and for The Advertiser (TA) it is 173.3. The analysis 9 constructs a salience measure for each year t as follows : (number of SMH stories)/237.1 + (number of TCM stories)/272.8 + (number of TA stories)/173.3. This index equally weights all three outlets. Its average is thus 3.0 by construction, with a standard deviation of 2.67 and a one-year maximum of 10.9. Figure 5.1 plots the index over the 1990 through 2007 period. Two large spikes in sustained coverage of native title appear in 1993 and 1997, which follow the High Court’s two most important decisions involving 10 native title land rights, Mabo (No.2) and Wik. During those two peak years, 1993 and 1997, SMH averaged .7 and 2.6 stories per day on native title or land rights issues, TCM 9 The national salience index is based on the national salience of immigration index created by Hopkins (2010, 44). 10 The national salience measure does not indicate a spike in coverage that corresponds with the High Court’s trio of decisions in late 2002. Given that the cases did not necessarily break new legal ground it is understandable that the cases did not garner the media attention that produces national salience. 162 averaged 1.5 and 3 stories per day and TA averaged 1.5 and 1.4 stories per day. Not only was native title of national salience, but also the issue was given sustained daily attention. [Insert Figure 5.1 about here] To measure the influence of context, it is important to note two key components that are required to make an individual become more aware of his or her environment in regards to native title. The first is Aboriginal Australians and the second is the availability of land. Claims of native title require Aboriginal groups to show they have maintained a traditional connection to the land and that the land itself has not been subject to development such that native title would be extinguished. Acts that would extinguish native title include any kind of freehold property contract including housing. In order capture both elements needed for a native title claim, population density for Aboriginal Australians and overall population density for each state was collected from the Australian census (Australian Bureau of Statistics 2008; see also Geoscience Australian 2010 for size of state and territories). The census in Australia is conducted every five years, which yields five points in time: 1986, 1991, 1996, 2001 and 2006. In the following analysis, linear interpolation was used to assign values between census 11 years. To create this measure, Aboriginal population density was constructed by taking the total population of Aborigines by census year and by state and territory and dividing that total by the corresponding size in square kilometers of the state or territory. The same procedure was also carried out to construct an overall population density measure. 11 Except for 2007, which is assigned the census value from 2006. 163 Next, the population density of Aboriginal Australians was divided by the overall population density, which created a measure of native title density. Native title density has a mean of .02, with a standard deviation of .02, a minimum of .005 and a maximum of .28. Lower values of native title density indicate weak conditions for native title claims, while higher values indicate strong conditions. This is born out as the lowest values of native title density correspond to Victoria and the Australian Capital Territory, 12 while the highest values correspond to the Northern Territory. Three dichotomous case dummy variables are included to account for the three sets of native title decisions made by the High Court. Given that none of the case dummies were statistically significant in the aggregate level analysis of opinion conducted in Chapter 3, it is unlikely that any of the variables will be statistically significant here. However, they are included in order to create the contextual proximity measure that suggests decisions may awaken individuals to their surrounding environments. This connection will most likely decrease support for native title since individuals will not want to expose themselves and their communities to the potential costs of native title. The dependent variable is 1 for the 20% of respondents who want to increase native title (believe native title has not gone far enough) and 0 for the others. This dependent variable is measured in all surveys with the question: The statements below indicate some of the changes that have been happening in Australia over the years. For each one, please say whether you think the change has gone too far, not far enough, or is 12 Victoria is the third smallest state or territory by size, but second in terms of overall population. The Australian Capital Territory is the smallest state or territory by far; the next smallest area by size is Tasmania, which is 29 times its size. The Northern Territory has the largest number of Aborigines in Australia and last in overall population. 164 it about right? Transfer of land rights to Aborigines. For the following analysis, if a respondent reports not gone far enough or not nearly far enough the dependent variable is 1. If the respondent reports gone much too far, gone too far, or about right, the dependent variable is 0. The approach used here to estimate support for native title relies on multilevel or mixed modeling. This approach is a useful means for analyzing observations that are likely to be correlated within units or levels of analysis (e.g. Gelman and Hill 2007). The mixed modeling strategy accounts for the potential correlation among observations within units without having to separate the data into subsets that are assumed to be independent. Pooled surveys where individuals are contained in both states or territories and years naturally fit into this structure. Opinions among respondents in a state or territory are likely correlated across geographic space and time. The analysis presented here includes random intercepts for each state-year, which allows for any omitted state-year specific covariates to be accounted for in the model. Given that the dependent variable is binary, a fixed-effects logistic regression with random-intercepts model is estimated. The model yields 10895 total observations, which are grouped into 48 state-year groups. 5.6 Results The results of the model are presented in Table 5.1. The random intercepts do not appear in the table, however the model does imply forty-eight separate intercepts (one for each state-year), but these are not directly estimated. Instead the standard deviation of the random intercepts (.35) along with a standard error (.05) is estimated. If the standard deviation appears significantly different from zero, then the intercepts can be interpreted 165 as varying from place to place. This is the case as the standard deviation is seven standard errors from zero. A likelihood-ratio test for the model indicates that a simple pooled logistic regression would not be sufficient; fitting random intercepts is thus a 13 useful approach. Although the model does not directly calculate random effects, predictions can be recovered from the random effects covariance matrix. Those predicted values are plotted using a horizontal bar chart in Figure 5.2. For readability, approximately one-quarter of the random-intercepts are identified on the y-axis. Figure 5.2 reveals that, at any given level of covariates in the model, support for native title averaged about .8 less for respondents living in Queensland in 1990, or about .6 more for respondents living in the Australian Capital Territory in 2007, compared with middle of the road respondents living in Tasmania in 1996. [Insert Figure 2 about here] Given that logistic regression coefficients are not easily interpreted, odds ratios are employed to assess the impact of each covariate in the model. Odds ratios report the factor change in the odds of supporting native title associated with a certain change in each variable (e.g. Long and Freese 2006). Thus, an odds ratio below 1.0 indicates a negative relationship, while any value above 1.0 indicates a positive relationship. Given that most of the covariates or variables in the model are continuous, the odds ratios that are displayed are associated with a deviation change. For the covariates in the model that 13 The L-R test statistic is 77.77 and is highly statistically significant (p < .000). 166 are dichotomous or binary, the odds ratios that are displayed are associated with a unit change. [Insert Table 5.1 about here] Contrary to the expectations from the threat hypothesis, the measure of native title density was not statistically significant. However, the odds ratio is below 1.0, indicating a negative relationship, which was expected. As native title density increases the likelihood of supporting native title was expected to decrease. The polarized places hypothesis asserted that the perception of contextual effects, native title density, would vary with the national salience of native title land rights coverage across media outlets. Contrary to expectations, neither the national salience measure nor the interaction between national salience and native title density were statistically significant. Again, however, the odds ratios for both covariates exhibit a negative relationship, which was in the expected direction. Based on a previous analysis in Chapter 3, the individual case dummy variables were not expected to be statistically significant and they were not here. All three of the case dummies did have odds ratios over 1.0, indicating a positive relationship between all three decisions and support for native title. However, the case dummies were included here as part of an interaction term with native title density. Each case dummy was interacted with native title density in order to capture the contextual proximity effect that was thought to take place after a decision. The decisions were thought to make people more aware of their surroundings, which would result in native title losing support among those individuals. Once again, none of the covariates was statistically significant. The 167 expectations were for each term to have a negative relationship with the dependent variable and this was true of for the Trio cases, while positive relationships were observed for the Mabo and Wik cases. Overall, then, none of the expected relationships were confirmed in this analysis. 5.7 Conclusion The sociodemographic makeup of individual environments or context was thought to act as a potential hurdle to judicial influence of public attitudes either by itself or in combination with externally driven factors like court decisions or national media coverage, which would serve to activate an awareness of an individual’s surroundings. These hypotheses were assessed in terms of individual attitudes toward land rights for Aborigines over a period from 1990 through 2007. Unfortunately, none of covariates in the multilevel model proved to be statistically significant, although almost all of the variables were in the expected direction. Does that mean that individual context does not act as a hurdle to court influence of public opinion? Despite the null results from the analysis, there is reason to be cautious about such a conclusion. The random-intercepts for each state-year, in the multilevel model, revealed quite a bit of variation across and between states and years. It is clear that opinion toward native title land rights does vary across the contextual units analyzed here, state-year, but the covariates in the model did not accurately identify the factors contributing to those differences. The use of the state or territory as the relevant contextual unit may also have contributed to the null results. The theoretical arguments made above have their roots in 168 theories where the relevant contextual unit is a much smaller geographic unit, such as counties, zip codes or towns and their surrounding areas. By expanding the contextual unit to the state level, localized variation may have been covered up through the aggregation of opinion. Scholars of court influence on public opinion are particularly aware of the perils of the aggregation process (e.g. Franklin and Kosaki 1989). Unfortunately, the data available does not allow for a geographically finer grained assessment of public attitudes towards land rights below the state level. Nevertheless, the validity of the arguments and theories advanced here cannot be definitively accepted or rejected until they are assessed at a lower geographic unit of analysis. The next and final chapter, Chapter 6, provides an overall assessment of the main findings of this dissertation. The results are discussed in terms of the development of the literature on court influence of public opinion as well as the larger implications that the native title issue had on Australian society and the High Court. 169 5.8 Tables and Figures TABLE 5.1 Fixed-Effects Logistic Regression with Random-Intercepts Parameter Fixed Effects Mabo Wik Trio Native Title Density Native Title Salience Native Title Salience x Native Title Density Mabo x Native Title Density Wik x Native Title Density Trio x Native Title Density Constant Coefficient Std. Error Odds Ratio .20 1.39 .23 -.00 -.24 1.50 2.71 .26 16.35 .45 1.23 4.00 1.26 .99 .79 -1.43 8.71 9.56 -5.32 -.98 11.70 39.25 70.47 6.39 .66 .24 6037.16 14119.29 .01 -- .35 .05 -- Random Effects State-Year, SD (SE) Number of Obs Number of Groups Log Likelihood L-R Test 10895 48 -5326.83 77.77* Dependent variable is (1) support for native title and (0) otherwise. Australian Election Studies 1990-2007. *Significant at p=.000, two-tailed. 170 FIGURE 5.1 National Salience of Native Title 171 FIGURE 5.2 Predicted Random-Intercepts by State-Year Note: Selected random-intercepts by state-year appear on the y-axis. ACT (Australian Capital Territory); NSW (New South Wales); NT (Northern Territory); QLD (Queensland); SA (South Australia); VIC (Victoria); WA (Western Australia). 172 Chapter 6 Conclusion This dissertation has been concerned with answering the following question: under what conditions can courts influence the public? In order to answer this question, it was necessary to generalize theories that have been developed in the United States and test their viability in a foreign legal and political context. Australia provided a fertile environment for exploring whether a court outside the United States can influence public opinion with their decisions. The High Court of Australia’s decisions in the area of native title land rights for Aborigines, beginning with its landmark decision in the Mabo (No. 2) case, immediately electrified the mass public and the political establishments. The decision finding that native title could survive annexation, colonization and federation appeared to usher in a “judicial revolution” (Russell 2006, 5). Though the decision and public reaction has been thoroughly documented, empirical evaluations of the public’s response to this decision and the High Court’s subsequent decisions have never previously been undertaken. Though this dissertation seeks to make a general contribution to general theories relating court influence of the public, it also seeks to make a specific contribution to and a greater understanding of the Australian public’s response to the High Court’s native title decisions. Chapter 3 began by recognizing that the High Court of Australia has never been viewed by the public or political elites as a political institution. Given that the High Court is historically viewed by Australians as not being ideologically driven, not engaged in making policy decisions and more importantly not the same type of institution as the United States Supreme Court, theories of court influence might not have as much 173 explanatory value in Australia. But the rise of the Mason Court’s implied rights jurisprudence created an opportunity for the High Court to interject itself into the popular political consciousness with its rulings in the controversial issue area of native title land rights for Aboriginal Australians. Would the Australian public respond in a similar manner as the American public? The first empirical test was of the positive response hypothesis, which suggested that court decisions should result in increased public support of the issue and the High Court’s position, in particular. The results, consistent with previous studies in the United States, did not find support for the positive response hypothesis. The initial indication being that if the High Court can influence the public with its decisions, then like the United States Supreme Court, it is more probably not as easy as initially thought. Next, the conditional and subsequent response hypotheses were tested. These suggest that court influence of the public is much more nuanced and less straight forward than simply issuing a decision. Consistent with Johnson and Martin’s (1998) conditional and subsequent response hypotheses, the empirical evidence demonstrated that the High Court’s decisions affected the overall structure of public opinion in the first landmark decision, Mabo, but not in either set of subsequent decisions, Wik or the Trio of cases. Overall, these findings indicate that the High Court of Australia can and did influence public attitudes toward the transfer of land rights to Aborigines. However, the effect was conditional in nature and only in the first instance of a landmark decision. More broadly, these results indicate that judicial influence of public opinion may not be restricted to a particular class of cases or a particular set of issues, but may be a general characteristic of public responsiveness to highly salient, landmark decisions and 174 to courts in particular. Given that judges and courts around the world are surrounded by symbols of fairness, objectivity and impartiality it is no wonder that the public responds to their most important decisions. Chapter 4 sought to assess whether political sophistication helped or hindered the court influence of public opinion and whether sophistication was responsible for extreme attitude positions toward native title land rights. The theoretical foundation of the process of judicial influence on public opinion is based on the Elaboration Likelihood Model. Yet, studies of judicial influence of public opinion has not taken advantage of or explored how the differences in high and low sophisticates may affect how court decisions are received by the public. Unfortunately, the results of both empirical investigations reveal differences in the behavior of high and low sophisticates, but not in the manner anticipated by the crystallization and volatility hypotheses. For both high and low sophisticates the likelihood-ratio tests revealed that the overall structure of opinion had changed after a decision, but the case decision variables were not statistically significant, preventing a link being made between the overall change and the High Court’s decisions. The second section of this chapter engaged with the question of polarization or attitude extremity, which suggests that the movement of opinion towards extreme positions may be the result of how information is processed between levels of sophistication. Though neither of the hypotheses could be fully accepted the behavior of both high and low sophisticates proved interesting. An overall change in the opinion structures of both high and low sophisticates occurred in response to the High Court’s Wik decision leading towards an increased likelihood of attitude extremity. The first 175 subsequent case in which the High Court reexamined the question of native title clearly led both high and low sophisticates to revisit their attitudes on the transfer of land rights to Aborigines. Far from settling the issue of native title amongst the public, the High Court’s Wik decision led to increasingly divisive attitudes towards the issue. Chapter 5 sought to determine if contextual factors linked to the sociodemographic makeup of individual environments could act as a potential hurdle to judicial influence of public attitudes either by itself or in combination with externally driven factors like court decisions or national media coverage, which would serve to activate an awareness of an individual’s surroundings. Surprisingly, none of covariates in the multilevel model proved to be statistically significant, although almost all of the variables were in the expected direction. Despite the null results from the analysis, there is reason to be cautious about such a conclusion. The analysis revealed that opinion toward native title land rights does vary across time and between the different states and territories, but the model did not accurately identify the factors contributing to those differences. The use of the state or territory as the relevant contextual unit may also have contributed to the null results. The theoretical arguments made above have their roots in theories where the relevant contextual unit is a much smaller geographic unit, such as counties, zip codes or towns and their surrounding areas. By expanding the contextual unit to the state level, localized variation may have been covered up through the aggregation of opinion. Scholars of court influence on public opinion are particularly aware of the perils of the aggregation process (e.g. Franklin and Kosaki 1989). Unfortunately, the data available does not allow for a geographically finer grained assessment of public attitudes towards land rights 176 below the state level. Nevertheless, the validity of the arguments and theories advanced here cannot be definitively accepted or rejected until they are assessed at a lower geographic unit of analysis. 6.1 A Final Word on Judicial Retreat Native title was a prime example of a controversial question that the political community happily defers to the judiciary (Patapan 2000 133). If the native title question had arrived at any other time than during the tenure of Chief Justice Anthony Mason it is fair to say that Australia would continue to accept the doctrine of terra nullius. The justices of the Mason Court introduced a new vision for the High Court’s role that shifted its institutional focus away from simply resolving disputes to making policy that addressed some of the most controversial issues in Australian politics and society (Pierce 2006, 4). The cases brought by Eddie Mabo stand as shining examples of a High Court willing to be “more activist, more controversial and much more politicized” (Pierce 2006, 4). The High Court’s newly found commitment to justice and fairness over legal stability led to the recognition of native title within the common law and a fundamentally different future for Aborigines Australians. Unfortunately, the political backlash that erupted in response to the High Court’s native title decisions revealed the limits of its influence. The criticism of the High Court directed at its institutional legitimacy had given “most members of the court a sense that their political mandate” had “run out” (Russell 2005, 381). It also doomed a sustained institutional transformation after Chief Justice Mason retired in 1995. The Mason Court’s challenge to the traditional legal orthodoxy was not supported in the broader 177 legal community, which combined with an inadequate constitutional structure to entrench and expand individual rights led to many of the Mason Court reforms being gradually rolled back (Pierce 2006, 25). The High Court’s subsequent native title decisions in Wik and the Trio cases highlight the distress that the justices that succeeded members of the Mason Court had about losing institutional legitimacy (Pierce 2006, 25). Those decisions, while affirming native title, failed to expand its reach or provide any meaningful protections for Aboriginal communities seeking to establish land claims in states hostile to their interests. The frozen rights approach enshrined in the Trio cases moved native title away from exclusive possession to a right to occupy. For a people that had been denied justice over hundreds of years of colonial settlement and occupation, the right to use and stand on land that is rightfully theirs provides cold comfort. Yet, it is not entirely surprising that the High Court, unable to sustain support for its transformation into a political institution backed away from making policy and returned to simply resolving legal disputes. While this makes good strategic sense for justices seeking to build and grow the legitimacy of an already weak institution, it was devastating the Aboriginal community. As Russell asserts, “in a constitutional democracy, we should not be shocked or surprised that judges’ assessment of the boundaries of their political legitimacy should be a factor in their decision-making” (Russell 2006, 381). However, as Russell continues, “a judicial retreat, however explicable in terms of the pressures of majoritarian democracy, can never be good news for the minority, whose rights judicial activism was protecting” (Russell 2006, 381). 178 APPENDICES 179 APPENDIX A Table 1A. Summary Statistics Variable Aggregate Level Support Paper12 Abperc Obs Mean Std. Dev. Min Max 18 18 18 18.53 12.69 2.06 4.39 16.58 0.23 12.9 0 1.58 24.5 67.17 2.29 Individual Level Mabo Support Native Title Queensland Queensland Post Education Education Post Media Media Post Labor Labor Post Liberal-National Liberal-National Post Union Union Post Female Female Post Married Married Post Mine/Farm Mine/Farm Post 3834 3735 3834 3834 3657 3657 3648 3648 3834 3834 3834 3834 3402 3402 3834 3834 3834 3834 3834 3834 0.47 0.15 0.16 0.08 0.31 0.13 5.71 2.65 0.41 0.17 0.40 0.19 0.36 0.14 0.51 0.24 0.69 0.32 0.03 0.02 0.50 0.36 0.37 0.28 0.46 0.34 1.63 2.99 0.49 0.37 0.49 0.39 0.49 0.35 0.50 0.43 0.46 0.47 0.46 0.28 0 0 0 0 0 0 2 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 1 8 8 1 1 1 1 1 1 1 1 1 1 1 1 Wik Support Native Title Queensland Queensland Post Education Education Post Media Media Post 3694 3554 3694 3694 3476 3476 3545 3545 0.51 0.17 0.17 0.09 0.29 0.15 5.68 2.95 0.50 0.38 0.38 0.28 0.45 0.36 1.54 3.08 0 0 0 0 0 0 2 2 1 1 1 1 1 1 8 8 180 TABLE 1A. Summary Statistics, Cont. Variable Labor Labor Post Liberal-National Liberal-National Post Union Union Post Female Female Post Married Married Post Mine/Farm Mine/Farm Post Obs 3694 3694 3694 3694 3289 3289 3694 3694 3694 3694 3694 3694 Mean 0.38 0.20 0.39 0.19 0.28 0.13 0.51 0.26 0.67 0.34 0.03 0.01 Std. Dev. 0.49 0.40 0.49 0.40 0.45 0.34 0.50 0.44 0.47 0.48 0.17 0.11 Min 0 0 0 0 0 0 0 0 0 0 0 0 Max 1 1 1 1 1 1 1 1 1 1 1 1 Trio Support Native Title W. Australia W. Australia Post Victoria Victoria Post NSW NSW Post Education Education Post Media Media Post Labor Labor Post Liberal-National Liberal-National Post Union Union Post Female Female Post Married Married Post Mine/Farm Mine/Farm Post 3779 3609 3779 3779 3779 3779 3779 3779 3503 3503 3524 3524 3779 3779 3779 3779 3410 3410 3779 3779 3779 3779 3779 3779 0.47 0.22 0.10 0.04 0.25 0.12 0.32 0.15 0.33 0.17 5.45 2.59 0.33 0.15 0.42 0.20 0.25 0.12 0.51 0.24 0.66 0.30 0.03 0.04 0.50 0.41 0.29 0.20 0.43 0.32 0.47 0.35 0.47 0.37 1.57 2.93 0.47 0.35 0.49 0.40 0.43 0.32 0.50 0.43 0.48 0.46 0.16 0.20 0 0 0 0 0 0 0 0 0 0 2 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1 1 1 8 8 1 1 1 1 1 1 1 1 1 1 1 1 181 TABLE 2A. Likelihood of Support for the Transfer of Land Rights to Aborigines, Pre- and Post-Mabo v. Queensland Parameter Mabo Queensland Queensland Postdecision Education Education Postdecision Media Media Postdecision Labor Party Labor Party Postdecision Liberal-National Party Lib-Nat Party Postdecision Union Union Postdecision Female Female Postdecision Married Married Postdecision Mine/Farm Worker Mine/Farm Postdecision Constant Pseudo R-square Log Likelihood Chi-Square Number of Observations Likelihood-Ratio Test Constrained Model Coeff. Prob. -.12 (.06) .050 -.40 (.09) .000 --.42 (.06) .000 --.03 (.02) .149 --.05 (.07) .539 ---.68 (.08) .000 --.05 (.06) .399 --.20 (.06) .001 ---.18 (.06) .003 --.21 (.17) .230 ---1.06 (.14) .000 0.09 -1194.64 234.44 3109 15.64 Unconstrained Model Coeff. Prob. -.61 (.26) .020 -.38 (.12) .002 -.03 (.19) .873 .37 (.08) .000 .11 (.12) .379 -.00 (.02) .895 .07 (.04) .075 -.15 (.10) .143 .40 (.15) .007 -.76 (.11) .000 .11 (.17) .517 .10 (.08) .191 -.15 (.13) .239 .19 (.08) .015 -.00 (.12) .980 -.11 (.09) .216 -.16 (.13) .192 .08 (.24) .745 .28 (.35) .423 -.81 (.18) .000 0.10 -1186.82 250.07 3109 0.075 Note: Coefficients are probit coefficients. Standard errors are in parenthesis. The dependent variable is coded 1 if respondents answered the transfer of land rights to Aborigines has not gone far enough/not nearly far enough and 0 otherwise. Source: Australian Election Studies 1990, 1996 182 TABLE 3A. Likelihood of Support for the Transfer of Land Rights to Aborigines, Pre- and Post-Wik Peoples v. Queensland Parameter Wik Queensland Queensland Postdecision Education Education Postdecision Media Media Postdecision Labor Party Labor Party Postdecision Liberal-National Party Lib-Nat Party Postdecision Union Union Postdecision Female Female Postdecision Married Married Postdecision Mine/Farm Worker Mine/Farm Postdecision Constant Pseudo R-square Log Likelihood Chi-Square Number of Observations Likelihood-Ratio Test Constrained Model Coeff. Prob. .29 (.06) .000 -.32 (.09) .000 --.51 (.06) .000 --.05 (.02) .008 --.29 (.07) .000 ---.48 (.08) .000 --.06 (.06) .343 --.08 (.06) .162 ---.23 (.06) .000 --.06 (.20) .782 ---1.44 (.14) .000 0.11 -1205.44 289.10 2982 11.84 Unconstrained Model Coeff. Prob. .25 (.27) .343 -.41 (.14) .004 .15 (.18) .396 .47 (.09) .000 .06 (.12) .629 .06 (.03) .029 -.02 (.04) .582 .25 (.11) .021 .09 (.15) .536 -.65 (.13) .000 .30 (.17) .075 -.04 (.10) .655 .19 (.13) .137 .19 (.09) .034 -.19 (.12) .102 -.27 (.09) .003 .07 (.12) .551 .36 (.26) .168 -.71 (.43) .097 -1.42 (.20) .000 0.11 -1199.52 300.94 2982 0.222 Note: Coefficients are probit coefficients. Standard errors are in parenthesis. The dependent variable is coded 1 if respondents answered the transfer of land rights to Aborigines has not gone far enough/not nearly far enough and 0 otherwise. Source: Australian Election Studies 1996, 1998 183 TABLE 4A. Likelihood of Support for the Transfer of Land Rights to Aborigines, Pre- and Post-Ward, Anderson and Yorta Yorta (Trio) Parameter Trio W. Australia W. Australia Postdecision Victoria Victoria Postdecision NSW NSW Postdecision Education Education Postdecision Media Media Postdecision Labor Party Labor Party Postdecision Liberal-National Party Lib-Nat Party Postdecision Union Union Postdecision Female Female Postdecision Married Married Postdecision Mine/Farm Worker Mine/Farm Postdecision Constant Pseudo R-square Log Likelihood Chi-Square Number of Observations Constrained Model Coeff. Prob. .15 (.06) .006 -.03 (.11) .789 --.37 (.07) .000 --.05 (.07) .500 --.56 (.06) .000 --.10 (.02) .000 --.16 (.07) .018 ---.71 (.07) .000 --.10 (.06) .118 --.13 (.06) .022 ---.20 (.06) .001 ---.36 (.24) .133 ---1.51 (.13) 0.14 -1334.82 423.38 2949 .000 Unconstrained Model Coeff. Prob. .33 (.26) .199 -.04 (.15) .773 .04 (.21) .836 .41 (.10) .000 -.07 (.14) .651 .14 (.10) .150 -.20 (.14) .157 .55 (.08) .000 .03 (.12) .785 .12 (.03) .000 -.04 (.03) .315 .07 (.10) .457 .20 (.14) .140 -.80 (.11) .000 .17 (.11) .257 .19 (.09) .034 -.19 (.13) .136 .12 (.08) .148 .03 (.11) .786 -.20 (.08) .016 .01 (.12) .912 -.21 (.49) .553 -.29 (.49) .547 -1.61 (.18) .000 0.14 -1330.75 431.51 2949 Likelihood-Ratio Test 8.13 0.701 Note: Coefficients are probit coefficients. Standard errors are in parenthesis. The dependent variable is coded 1 if respondents answered the transfer of land rights to Aborigines has not gone far enough/not nearly far enough and 0 otherwise. Source: Australian Election Studies 2001, 2004 184 APPENDIX B TABLE 1B. Summary Statistics Variable Mabo Wik Trio Obs 11383 11383 11383 Mean .16 .17 .16 Std Dev .37 .37 .36 Min 0 0 0 1 1 1 Aboriginal Population Density 11309 .13 .19 .02 1.64 Total Population Density 11309 11.39 19.00 .12 144.54 Native Title Density 11309 .02 .02 .01 .28 Native Title Salience 11383 2.92 2.34 1.17 7.4 185 Max REFERENCES 186 REFERENCES Adamany, David and Joel B. Grossman. 1983. "Support for the Supreme Court as a National Policymaker." Law and Policy Quarterly 5: 405-37. Allport, Gordon W. 1954. The Nature of Prejudice. Cambridge: Addison-Wesley. Angelico, T. 1995. “Wallerstein and Contemporary Australian Racisim.” In Contemporary Racism in Australia, Canada and New Zealand, ed. J. Collins. Sydney: University of Technology. Australian Bureau of Statistics. 2000. The Australian Mining Industry: From Settlement to 2000. Cat. No. 8414.0. Accessed from: http://www.abs.gov.au/ausstats/abs%40.nsf/94713ad445ff1425ca25682000192af2/93136 e734ff62aa2ca2569de00271b10!OpenDocument [Accessed: March 20, 2009] Australian Bureau of Statistics. 2008. Australian Historical Population Statistics. Cat. No. 3105.0.65.001. Accessed from: http://www.abs.gov.au/ausstats/abs@.nsf/mf/3105.0.65.001 [Accessed: March 17, 2009] Bailey, Michael, and Forrest Maltzman. 2011. The Constrained Court: Law, Politics and the Decisions Justices Make. Princeton, NJ: Princeton University Press. Bartles, Brandon L. and Diana C. Mutz. 2009. “Explaining Processes of Institutional Opinion Leadership.” Journal of Politics 71(1): 249-61. Bean, Clive, Ian McAllister and David Gow. 1999. Australian Election Study 1998. Canberra: Australian Social Science Data Archives, The Australian National University. Bean, Clive, Ian McAllister and David Gow. 2001. Australian Election Study 2001. Canberra: Australian Social Science Data Archives, The Australian National University. Bean, Clive, Ian McAllister and David Gow. 2008. Australian Election Study 2007. Canberra: Australian Social Science Data Archives, The Australian National University. Bean, Clive, Ian McAllister, Rachel Gibson and David Gow. 2005. Australian Election Study 2004. Canberra: Australian Social Science Data Archives, The Australian National University. Berkson, Larry C. 1978. The Supreme Court and Its Publics: The Communication of Policy Decisions. Lexington, MA: Lexington Books. Bickel, Alexander M. 1986. The least dangerous branch: The Supreme Court at the bar of politics. 2nd ed. New Haven, CT: Yale University Press. Bizer, George Y., Zakary L. Tormala, Derek D. Rucker, and Richard E. Petty. 2006. 187 “Memory-Based Versus On-Line Processing: Implications for Attitude Strength.” Journal of Experimental Social Psychology 42: 646-53. Blalock, Hubert M. 1967. Toward a Theory of Minority-Group Relations. New York: John Wiley and Sons. Bohner, Gerd, and Nina Dickel. 2011. "Attitudes and Attitude Change." Annual Review of Psychology 62: 391-417. Bohner, Gerd, H-P Erb, and F. Siebler. 2008. "Information Processing Approaches to Persuasion: Integrating Assumptions from Dual- and Single-Processing Perspectives." In Attitudes and Attitude Change, ed. W. D. Crano and R. Prislin. New York: Psychology Press. Boninger, David S., Matthew K. Berent, and Jon A. Krosnick. 1995. "Origins of Attitude Importance: Self-Interest, Social Identification, and Value Relevance." Journal of Personality and Social Psychology 68 (1): 61-80. Branton, Regina P. and Bradford S. Jones. 2005. “Re-examining Racial Attitudes: The Conditional Relationship Between Diversity and Socioeconomic Environment.” American Journal of Political Science 49(2): 349-72. Brennan, Sean. 2003. “Native Title in the High Court of Australia a Decade after Mabo.” Public Law Review 14: 209-18. Cacioppo, John T. and Gary G. Berntson. 1994. “Relationship Between Attitudes and Evaluative Space: A Critical Review, With Emphasis on the Separability of Positive and Negative Substrates.” Psychological Bulletin 115: 401-23. Cacioppo, John T., Wendi L. Gardner, and Gary G. Berntson. 1999. "The Affect System has Parallel and Integrative Processing Components: Form Follows Function." Journal of Personality and Social Psychology 76: 839-55. Cacioppo, John T. and Richard E. Petty. 1982. “The Need for Cognition.” Journal of Personality and Social Psychology 42: 116-131. Cacioppo, John T., Wendi L. Gardner, and Gary G. Berntson. 1999. “The Affect System has Parallel and Integrative Processing Components: Form Follows Function.” Journal of Personality and Social Psychology 76: 839-55. Cacioppo, John T., Richard E. Petty, Jeffrey Feinstein, and Blair Jarvis. 1996. "Dispositional Differences in Cognitive Motivation: The Life and Times of Individuals Varying in Need for Cognition." Psychological Bulletin 119: 197-253. Cacioppo, John T., Richard E. Petty, Chuan Feng Kao, and Regina Rodriguez. 1986. "Central and Peripheral Routes to Persuasion: An Individual Difference Perspective." 188 Journal of Personality and Social Psychology 51: 1032-43. Caldeira, Gregory A. 1986. “Neither the Purse nor the Sword: Dynamics of Public Confidence in the Supreme Court.” American Political Science Review 80: 1209-26. Caldeira, Gregory A. 1991. “Courts and Public Opinion.” In John B. Gates and Charles A. Johnson, The American Courts: A Critical Assessment. Washington, D.C.: CQ Press. Canon, Bradley C., and Charles A. Johnson. 1999. Judicial Policies: Implementation and Impact. 2nd ed. Washington, D.C.: CQ Press. Chaiken, Shelly. 1980. “Heuristic versus systematic information processing and the use of source message cues in persuasion.” Journal of Personality and Social Psychology 39(November): 752-66. Chaiken, Shelly, and Mark W. Baldwin. 1981. "Affective Cognitive Consistency and the Effect of Salient Behavioral Information on the Self-Perception of Attitudes." Journal of Personality and Social Psychology 41 (1): 1-12. Choper, Jesse H. 1980. Judicial review and the national political process: A functional reconsideration of the role of the Supreme Court. Chicago: University of Chicago Press. Clawson, Rosalee A., Elizabeth R. Kegler and Eric N. Waltenburg. 2001. “The Legitimacy-Conferring Authority of the U.S. Supreme Court: An Experimental Design.” American Politics Research 29(6): 566-591. Converse, Philip E. 1964. "The Nature of Belief Systems in Mass Publics." In Ideology and Discontent, ed. David E. Apter. New York: Free Press. Coombs, Herbert Cole. 1994. Aboriginal Autonomy: Issues and Strategies. Melbourne: Cambridge University Press. Cunneen, Chris. 1992. “Judicial Racism.” Aboriginal Law Bulletin 2(58): 9-11. Dahl, Robert. 1957. “Decision Making in a Democracy: The Supreme Court as a National Policy Maker.” Journal of Public Law 6(Fall): 279-95. Davis, Richard. 1994. Decisions and Images: The Supreme Court and the Press. Englewood Cliffs, NJ: Prentice Hall. Delli Carpini, Michael X. and Scott Keeter. 1996. What Americans know about politics and why it matters. New Haven: Yale University Press. Dicey, Albert V. 1982. An Introduction to the Study of the Law of the Constitution, 8th edition. Indianapolis: Liberty Fund. 189 Ditto, Peter H. and David F. Lopez. 1992. "Motivated Skepticism: Use of Differential Decision Criteria for Preferred and Nonpreferred Conclusions." Journal of Personality and Social Psychology 63 (4): 568-84. Eagly, A. H. and Shelly Chaiken. 1995. "Attitude Strength, Attitude Structure, and Resistance to Change." In Attitude Strength: Antecedents and Consequences, ed. R. E. Petty and J. A. Krosnick. Mahwah, NJ: Lawrence Erlbaum Associates, Inc. Epstein, Lee. 1999. "The Comparative Advantage." Law and Courts Newsletter 9(3): 16. Epstein, Lee, and Jeffrey A. Segal. 2000. "Measuring Issue Salience." American Journal of Political Science 44(1): 66-83. Evans, Simon and John Williams. 2008. “Appointing Australian Judges: A New Model.” Sydney Law Review 30(2): 295-327. Fiske, Susan T., Donald R. Kinder and W. Michael Larter. 1983. "The Novice and the Expert: Knowledge-Based Strategies in Political Cognition." Journal of Experimental Social Psychology 19: 381-400. Fiske, Susan T. and Shelley E. Taylor. 2008. Social cognition: from brains to culture. 1st ed. Boston: McGraw-Hill Higher Education. Fletcher, Joseph F. and Paul Howe. 2000. “Public Opinion and the Courts.” Choices 6(3): 1-56. Flemming, Roy B., John Bohte and B. Dan Wood. 1997. “One Voice Among Many: The Supreme Court’s Influence on Attentiveness to Issues in the United States, 1947-92.” American Journal of Political Science 41(4): 1224-50. Franklin, Charles and Liane Kosaki. 1989. “Republican Schoolmaster: The U.S. Supreme Court, Public Opinion and Abortion.” American Political Science Review 83(September): 751-71. Franklin, Charles and Liane Kosaki. 1995. “Media, Knowledge, and Public Evaluations of the Supreme Court.” In Lee Epstein, ed., Contemplating Courts. Washington, DC: CQ Press. Galligan, Brian and Peter Russell. 1995. “The Politicisation of the Judiciary in Australia and Canada.” The Australian Quarterly 67(2): 85-100. Gay, Claudine. 2006. “Seeing Difference: The Effect of Economic Disparity on Black Attitudes Toward Latinos.” American Journal of Political Science 50: 982-97. 190 Gelman, Andrew and Jennifer Hill. 2007. Data Analysis Using Regression and Multilevel/Hierarchical Models. New York: Cambridge University Press. Geoscience Australia. 2010. Area of Australia – States and Territories. Accessed from: http://www.ga.gov.au/education/geoscience-basics/dimensions/area-of-australia-statesand-territories.html [Accessed: January 12, 2011] Gibson, James L., and Gregory A. Caldeira. 2009. Citizens, courts, and confirmations : positivity theory and the judgments of the American people. Princeton, N.J.: Princeton University Press. Gibson, James L., Gregory A. Caldeira, and Vanessa A. Baird. 1998. “On the Legitimacy of National High Courts.” American Political Science Review 92(2): 343-58. Gibson, James L., Gregory A. Caldeira, and Lester Kenyatta Spence. 2003. "The Supreme Court and the US Presidential Election of 2000: Wounds, Self-Inflicted or Otherwise?" British Journal of Political Science 33:535-556. Gibson, Rachel, Ian McAllister and Tami Swenson. 2002. “The Politics of Race and Immigration in Australia: One Nation Voting in the 1998 Election.” Ethnic and Racial Studies 25(5): 823-44. Gilens, Martin and Naomi Marukawa. 2002. "Elite Cues and Political Attitude Formation." In Research in Micropolitics, ed. M. X. Delli Carpini, L. Huddy and R. Y. Shapiro. New York: Elsevier Press. Glaser, James M. 1994. “Back to the Black Belt: Racial Environment and White Racial Attitudes in the South.” Journal of Politics 56(1): 21-41. Goot, Murray. 1993a. “Polls as Science, Polls as Spin: Mabo and the Miners.” The Australian Quarterly 65(4): 133-156. Goot, Murray. 1993b. “The Wild West? Yes, no and maybe.” The Australian Quarterly 65(4): 194-202. Goot, Murray, and Tim Rowse. 2007. Divided nation?: Indigenous affairs and the imagined public. Carlton, Vic.: Melbourne University Press. Goot, Murray and Ian Watson. 2001. “One Nation’s Electoral Support: Where Does it Come From, What Makes it Different and How Does it Fit?” Australian Journal of Politics and History 47(2): 159-191. Grosskopf, Anke and Jeffery J. Mondak. 1998. “Do Attitudes Toward Specific Supreme Court decisions Matter? The Impact of Webster and Texas v. Johnson on Public Confidence in the Supreme Court.” Political Research Quarterly 51(3): 633-654. 191 Gujarati, Damodar N. 2003. Basic econometrics. 4th ed. New York: McGraw-Hill. Hamilton, Alexander. 1961. “Federalist No. 78,” in The Federalist Papers, Ed. Clinton Rossiter. New York: New American Library. Hamilton, A. 1990. “Fear and Desire: Aborigines, Asians and the National Imaginary.” Australian Cultural History 9: 14-35. Hanley, John, Michael Salamone, and Matthew Wright. 2012. "Reviving the Schoolmaster: Reevaluating Public Opinion in the Wake of Roe v. Wade." Political Research Quarterly 65(2): 408-21. Haynie, Stacia L., Reginald S. Sheehan, Donald R. Songer, and C. Neal Tate. 2007. "High Courts Judicial Database." University of South Carolina Judicial Research Initiative. Hibbing, John R. and Elizabeth Theiss-Morse. 1995. Congress as Public Enemy: Public Attitudes Toward American Political Institutions. New York: Cambridge University Press. Hoekstra, Valerie J. 1995. “The Supreme Court and Opinion Change: An Experimental Study of the Court’s ability to Change Opinion.” American Politics Quarterly 23: 109129. Hoekstra, Valerie J. 2000. “The Supreme Court and Local Public Opinion.” The American Political Science Review 94(1): 89-100. Hoekstra, Valerie J. 2003. Public Reaction to Supreme Court Decisions. Cambridge ; New York: Cambridge University Press. Hoekstra, Valerie and Jeffrey Segal. 1996. “The Shepherding of Local Public Opinion: The Supreme Court and Lamb’s Chapel.” Journal of Politics 58(November): 1079-102. Holmes, John. 1991. “Land Tenure in the Australian Pastoral Zone: A Critical Appraisal.” In North Australian Research: Some Past Themes and New Directions, eds. Ian Mofatt and Ann Webb. Darwin: North Australia Research Unit, Australian National University. Hopkins, Daniel J. 2010. “Politicized Places: Explaining Where and When Immigrants Provoke Local Opposition.” The American Political Science Review 104(1): 40-60. Hopkins, Daniel J. 2011a. "National Debates, Local Responses: The Origins of Local Concern about Immigration in Britain and the United States." British Journal of Political Science 41(3): 499-524. Hopkins, Daniel J. 2011b. "The Limited Local Impacts of Ethnic and Racial Diversity." 192 American Politics Research 39(2): 344-79. Horrigan, Bryan. 2003. Adventures in Law and Justice: Exploring Big Legal Questions in Everyday Life. Sydney: University of New South Wales Press. Huckfeldt, Robert, Jeffrey Levine, William Morgan, and John Sprague. 1999. “Accessibility and the Political Unity of Partisan and Ideological Orientations.” American Journal of Political Science 43(3): 888-911. Iyengar, Shanto. 1991. Is anyone responsible?: How television frames political issues. Chicago: University of Chicago Press. Iyengar, Shanto and Donald R. Kinder. 1987. News that matters: television and American opinion. Chicago: University of Chicago Press. Jackman, Simon. 1998. “Pauline Hanson, the Mainstream, and Political Elites: The Place of Race in Australian Political Ideology.” Australian Journal of Political Science 33(2): 167-86. Jackson, Donald Wilson, and C. Neal Tate. 1992. Comparative judicial review and public policy. Westport, Conn.: Greenwood Press. Jacob, Herbert. 1996. Courts, law, and politics in comparative perspective. New Haven: Yale University Press. Johnson, Timothy T. and Andrew D. Martin. 1998. “The Public’s Conditional Response to Supreme Court Decisions.” American Political Science Review 92(June): 299-309. Jones, Roger, Ian McAllister and David Gow. 1996. Australian Election Study 1996. Canberra: Australian Social Science Data Archives, The Australian National University. Judd, C. M., Jon A. Krosnick and M. Milburn. 1981. "Political Involvement, and Attitude Structure in the General Public." American Sociological Review 46: 660-9. Kerkhof, Peter. 1999. “Applying the Unimodel to Political Persuasion.” Psychological Inquiry 10(2): 137-140. Key, V. O. 1949. Southern Politics in State and Nation. New York: A. A. Knopf. Kinder, Donald R. 1998. “Communication and Opinion.” Annual Review of Political Science 1: 167-97. Krosnick, Jon A., David S. Boninger and Yao Chuang. 1993. "Attitude Strength: One Construct or Many Related Constructs?" Journal of Personality and Social Psychology 65: 1132-51. 193 Kruglanski, Arie W. and Erik P. Thompson. 1999. “Persuasion by a Single Route: A View from the Unimodel.” Psychological Inquiry 10(2): 83-109. Larsen, K. S. 1981. “White Attitudes in Townsville: Authoritarianism, Religiosity and Contact.” Australian Psychologist 16(1): 111-122. Lavine, Howard, J. W. Huff, S. H. Wagner and D. Sweeney. 1998. "The Moderating Influence of Attitude Strength on the Susceptibility to Context Effects in Attitude Surveys." Journal of Personality and Social Psychology 75: 359-73. Lerner, Ralph. 1967. “The Supreme Court as Republican Schoolmaster.” The Supreme Court Review 1967: 127-80. Long, J. Scott. 1997. Regression models for categorical and limited dependent variables. Thousand Oaks: Sage Publications. Long, J. Scott, and Jeremy Freese. 2006. Regression Models for Categorical Dependent Variables Using Stata. Second ed. College Station, TX: Stata Press. Lord, Charles, Marc Ross and Mark Lepper. 1979. "Biased Assimilation and Attitude Polarization: The Effects of Prior Theories on Subsequently Considered Evidence." Journal of Personality and Social Psychology 37(11): 2098-109. Luks, Samantha, and Michael Salamone. 2008. "Abortion." In Public opinion and constitutional controversy, ed. N. Persily, J. Citrin and P. J. Egan. Oxford ; New York: Oxford University Press. Lupia, Arthur and Mathew D. McCubbins. 1998. The Democratic Dilemma: Can Citizens Learn What They Need to Know? Cambridge: Cambridge University Press. Luskin, Robert C. 1990. "Explaining Political Sophistication." Political Behavior 12(4): 331-61. Marshall, Thomas R. 1989. Public Opinion and the Supreme Court. Boston, MA: Hyman. Manfredi, Christopher P. 2001. Judicial Power and the Charter: Canada and the Paradox of Liberal Constitutionalism (2 ed.). New York: Oxford University Press. Matthews, J. Scott. 2005. “The Political Foundations of Support for Same-Sex Marriage in Canada.” Canadian Journal of Political Science 38(4): 841-866. Mayhew, David R. 1974. Congress: The Electoral Connection. New Haven: Yale University Press. McAllister, Ian, Roger Jones, Elim Papadakis, Elim and David Gow. 1990. Australian 194 Election Study 1990. Canberra: Australian Social Science Data Archives, The Australian National University. McGraw, Kathleen M. and C. Hubbard. 1996. "Some of the People Some of the Time: Individual Differences in Acceptance of Political Accounts." In Political Persuasion and Attitude Change, ed. D. C. Mutz, P. M. Sniderman and R. A. Brody. Ann Arbor, MI: University of Michigan Press. McGuire, Kevin T. 2009. "Public Schools, Religious Establishments, and the U.S. Supreme Court: An Examination of Policy Compliance." American Politics Research 37(1): 50-74. McGuire, Kevin T. and James A. Stimson. 2004. “The Least Dangerous Branch Revisited: New Evidence on Supreme Court Responsiveness to Public Preferences.” Journal of Politics 66(4): 1018-35. McGuire, Kevin T., Georg Vanberg, Charles Jr. Smith, and Gregory A. Caldeira. 2009. "Measuring Policy Content on the U.S. Supreme Court." The Journal of Politics 71 (4): 1305-21. McKelvey, Richard and William Zavonia. 1975. "A Statistical Model for the Analysis of Ordinal Level Dependent Variables." Journal of Mathematical Sociology 4: 103-20. Mellor, David. 2003. “Contemporary Racism in Australia: The Experiences of Aborigines.” Personality and Social Psychology Bulletin 29: 474-486. Mishler, William and Reginald S. Sheehan. 1993. “The Supreme Court as a CounterMajoritarian Institution?” American Political Science Review 87(1): 87-100. Mondak, Jeffery J. 1990. “Perceived Legitimacy of Supreme Court Decisions: Three Functions of Source Credibility.” Political Behavior 12: 363-384. Mondak, Jeffery J. 1992. “Institutional Legitimacy, Policy Legitimacy, and the Supreme Court.” American Politics Quarterly 20(4): 457-77 Mondak, Jeffery J. 1994. "Policy Legitimacy and the Supreme Court: The Sources and Contexts of Legitimation." Political Research Quarterly 47 (3):675-692. Mondak, Jeffery J. and Shannon Ishiyama Smithey. 1997. “Dynamics of Public Support for the Supreme Court.” Journal of Politics 57(4): 1114-1142. Mooney, Christopher Z. 1996. “Bootstrap Statistical Inference: Examples and Evaluations for Political Science.” American Journal of Political Science 40(2): 570602. 195 Mooney, Christopher Z. and Robert Duval. 1993. Bootstrapping: A Nonparametric Approach to Statistical Inference. Newbury Park, CA: Sage Publications. Mutz, Diana C. 1994. “Contextualizing Personal Experience: The Role of Mass Media.” Journal of Politics 56(3): 689-714. Myers, William M. and Reginald S. Sheehan. 2012. “The Impact of the Australian High Court on Elite Opinion.” Commonwealth and Comparative Politics 50(2): 171-189. National Native Title Tribunal. 2012. 20 Years of Native Title. Accessed from: http://www.nntt.gov.au/News-and-Communications/MediaReleases/Pages/20yearssinceMabo.aspx [Accessed: June 6, 2012] Nettheim, Garth. 2007. “Wik”. The Oxford Companion to the High Court of Australia. Eds. Michael Coper, Tony Blackshield, George Williams. Oxford University Press, Oxford Reference Online. Oliver, J. Eric, and Tali Mendelberg. 2000. “Reconsidering the Environmental Determinants of White Racial Attitudes.” American Journal of Political Science 44(3): 574-89. Patapan, Haig. 2000. Judging Democracy: The New Politics of the High Court of Australia. Cambridge: Cambridge University Press. Pedersen, Anne and Ian Walker. 1997. “Prejudice Against Australian Aborigines: Oldfashioned and Modern Forms.” European Journal of Social Psychology 27: 561-587. Pedersen, Anne, Brian Griffiths, Natalie Contos, Brian Bishop and Ian Walker. 2000. “Attitudes Toward Aboriginal Australians in City and Country Settings.” Australian Psychologist 35(2): 109-117. Pedersen, Anne, Jaimie Beven, Ian Walker and Brian Griffiths. 2004. “Attitudes Toward Indigenous Australians: The Role of Empathy and Guilt.” Journal of Community and Applied Social Psychology 14: 233-249. Pedersen, Anne, Pat Dudgeon, Susan Watt and Brian Griffiths. 2006. “Attitudes Toward Indigenous Australians: The Issue of ‘Special Treatment.’” Australian Psychologist 41(2): 85-94. Persily, Nathaniel, Jack Citrin, and Patrick J. Egan. 2008. Public opinion and constitutional controversy. Oxford ; New York: Oxford University Press. Pettigrew, Thomas F. 1998. “Intergroup Contact Theory.” Annual Review of Psychology 49: 65-85. Petty, Richard E. 1994. "Two Routes to Persuasion: State of the Art." In International 196 Perspectives on Psychological Science, ed. G. d'Ydewalle, P. Eelen and P. Berteleson. Hillsdale, NJ: Lawrence Erlbaum Associates, Inc. Petty, Richard E. and John T. Cacioppo. 1979. "Issue Involvement Can Increase or Decrease Persuasion by Enhancing Message-Relevant Cognitive Responses." Journal of Personality and Social Psychology 37:1915-26. Petty, Richard E. and John T. Cacioppo. 1981. Attitudes and Persuasion: Classic and Contemporary Approaches. Dubuque, IA: Wm. C. Brown. Petty, Richard E. and John T. Cacioppo. 1984. "The Effects of Involvement on Responses to Argument Quantity and Quality: Central and Peripheral Routes to Persuasion." Journal of Personality and Social Psychology 46(1): 69-81. Petty, Richard E. and John T. Cacioppo. 1986a. Communication and persuasion: central and peripheral routes to attitude change. New York: Springer-Verlag. Petty, Richard E. and John T. Cacioppo. 1986b. "The Elaboration Likelihood Model of Persuasion." In Advances in Experimental Social Psychology, ed. L. Berkowitz. New York: Academic. Petty, Richard E., C. P. Haugtvedt and S. M. Smith. 1995. "Elaboration as a Determinant of Attitude Strength: Creating Attitudes that are Persistent, Resistant and Predictive of Behavior." In Attitude Strength: Antecedents and Consequences, ed. R. E. Petty and J. A. Krosnick. Mahwah, NJ: Lawrence Erlbaum Associates, Inc. Petty, Richard E. and Jon A. Krosnick. 1995. Attitude strength: antecedents and consequences. Hillsdale, N.J.: Lawrence Erlbaum Associates. Phelan, J., B. G. Link, A. Stueve and R. E. Moore. 1995. "Education, Social Liberalism, and Economic Conservatism: Attitudes Toward Homeless People." American Sociological Review 60: 126-40. Popkin, Samuel L. 1991. The Reasoning Voter: Communication and Persuasion in Presidential Campaigns. Chicago: University of Chicago Press. Popkin, Samuel, John W. Gorman, Charles Phillips, and Jeffrey A. Smith. 1976. "Comment: What have you done for me lately? Toward an investment theory of voting." The American Political Science Review 70: 779-805. Pierce, Jason L. 2006. Inside the Mason Court Revolution: The High Court of Australia Transformed. Durham: Carolina Academic Press. st Putnam, Robert D. 2007. “E Pluribus Unum: Diversity and Community in the 21 Century: The 2006 Johan Skytte Prize Lecture.” Scandanavian Political Studies 30(2): 137-74. 197 Rizvi, F. 1996. “Racism, Reorientation and the Cultural Politics of Asia-Australia Relations.” In The Teeth Are Smiling: The Persistence of Racism in Multicultural Australia, eds. E. Vasta and S. Castles. Sydney: Allen and Unwin. Rosenberg, Gerald N. 2008. The Hollow Hope: Can Courts Bring About Social Change 2nd ed. Chicago, IL: University of Chicago Press. Rucker, Derek D. and Richard E. Petty. 2004. "Emotion Specificity and Consumer Behavior: Anger, Sadness, and Preference for Activity." Motivation and Emotion 28(1): 3-21. Russell, Peter H. 2005. Recognizing Aboriginal Title: The Mabo Case and Indigenous Resistance to English-Settler Colonialism. Toronto: University of Toronto Press. Russell, Peter H., and David M. O'Brien. 2001. Judicial independence in the age of democracy : critical perspectives from around the world. Charlottesville: University Press of Virginia. Segal, Jeffrey A. and Harold J. Spaeth. 2002. The Supreme Court and the Attitudinal Model Revisited. Cambridge: Cambridge University Press. Short, Damien. 2007. “The Social Construction of Indigenous ‘Native Title’ Land Rights in Australia.” Current Sociology 55(6): 857-76. Smith, Leonard R. 1980. The Aboriginal Population of Australia. Canberra: Australian National University Press. Smyth, Russell. 2005. "The Role of Attitudinal, Institutional and Environmental Factors in Explaining Variations in the Dissent Rate on the High Court of Australia." Australian Journal of Political Science 40 (4): 519-40. Sniderman, Paul M., Richard A. Brody and Philip Tetlock. 1991. Reasoning and choice: explorations in political psychology. New York: Cambridge University Press. Staton, Jeffrey K. 2006. “Constitutional Review and the Selective Promotion of Case Results.” American Journal of Political Science 50(1): 98-112. Stein, Robert M., Stephanie Shirley Post and Allison Rinden. 2000. “Reconciling Context and Contact Effects on Racial Attitudes.” Political Research Quarterly 53(2): 285-303. Stoutenborough, James W., Donald P. Haider-Markel and Mahalley D. Allen. 2006. “Reassessing the Impact of Supreme Court Decisions on Public Opinion: Gay Civil Rights Cases.” Political Research Quarterly 59(3): 419-33. 198 Taber, Charles S., Damon Cann and Siimona Kucsova. 2009. "The Motivated Processing of Political Arguments." Political Behavior 31: 137-55. Taber, Charles S. and Milton Lodge. 2006. "Motivated Skepticism in the Evaluation of Political Beliefs." American Journal of Political Science 50(3): 755-69. Taylor, Marylee C. 1998. “How White Attitudes Vary with the Racial Composition of Local Populations: Numbers Count.” American Sociological Review 63: 512-35. Tesser, Abraham and M. C. Conlee. 1975. "Some Effects of Time and Thought on Attitude Polarization." Journal of Personality and Social Psychology 31: 126-31. Tesser, Abraham and C. Cowan. 1977. "Some Effects of Thought and Number of Cognitions on Attitude Change." Social Behavior and Personality 3: 165-73. Tesser, Abraham and C. Leone. 1977. "Cognitive Schemas and Thought as Determinants of Attitude Change." Journal of Experimental Social Psychology 13: 34056. Tate, C. Neal and Torbjorn Vallinder. 1995. The Global Expansion of Judicial Power. New York: New York University Press. Tickner, Robert. 2001. Taking a Stand: Land Rights to Reconciliation. Crows Nest, NSW: Allen and Unwin. Tidwell, Pamela S., Cyril J. Sadowski and Lia M. Pate. 2000. "Relationship Between Need for Cognition, Knowledge, and Verbal Ability." Journal of Psychology 134: 63445. Tocqueville, Alexis de, and J. P. Mayer. 2000. Democracy in America. 1st Perennial Classics ed. New York: Perennial Classics. Vanberg, Georg. 2000. "Establishing Judicial Independence in West Germany: The Impact of Opinion Leadership and the Separation of Powers." Comparative Politics 32: 333-53. Vanberg, Georg. 2001. “Legislative-Judicial Relations: A Game-Theoretic Approach to Constitutional Review.” American Journal of Political Science 45(2): 346-361. Voss, Stephen D. 1996. “Beyond Racial Threat: Failure of an Old Hypothesis in the New South.” Journal of Politics 58(4): 1156-70. Walker, Ian. 1994. “Attitudes to Minorities: Survey Evidence of Western Australians’ Attitudes to Aborigines, Asians, and Women.” Australian Journal of Psychology 46: 137-143. 199 Western, J. S. 1969. “What White Australians Think.” Race 10: 411-434. Williams, John M. 2001. “Judicial Independence in Australia”. Judicial Independence in the Age of Democracy: Critical Perspectives from around the World. Eds. Peter H. Russell and David M. O’Brien. Charlottesville, VA: University Press of Virginia. Wong, Cara. 2007. “”Little” and “Big” Pictures in our Heads: Race, Local Context, and Innumeracy about Racial Groups in the United States.” Public Opinion Quarterly 71(3): 392-412. Zaller, John and Stanley Feldman. 1992. "A Simple Theory of the Survey Response: Answering Questions versus Revealing Preferences." American Journal of Political Science 36 (3):579-616. Law Cases Australia Australian Capital Television Pty Ltd v. Commonwealth (1992) 177 C.L.R. 106 Cooper v Stuart (1889) 14 App Cas 286 Koowarta v Bjelke-Petersen (1982) HCA 27 Mabo v Queensland (No. 1) (1988) 166 C.L.R. 186 Mabo v. Queensland (No. 2) (1992) 175 C.L.R. 1 Members of the Yorta Yorta Aboriginal Community v The State of Victoria (1998) FCA 1606 Members of the Yorta Yorta Aboriginal Community v. Victoria (2002) HCA 58 Milirrpum v Nabalco (1971) 17 FLR 141 Nationwide News Pty Ltd v. Wills (1992) 177 C.L.R. 1 Street v. Queensland Bar Association (1989) 168 C.L.R. 461 Western Australia v. Ward (2002) HCA 28 Wik Peoples v. Queensland (1996) 187 C.L.R. 1 Wilson v. Anderson (2002) HCA 29 200 Canada Calder v Attorney-General of British Columbia (1973) SCR 313 Guerin v The Queen (1984) 2 SCR 335 New Zealand New Zealand Maori Council v Attorney-General (1987) 1 NZLR 641 United States Baker v. Carr (1962) 369 U.S. 186. Bowers v. Hardwick (1986) 478 U.S. 186 Lawrence v. Texas (2003) 539 U.S. 558 201