Judicial Ideology in the House of Lords: A Jurimetric Analysis Author(s): David Robertson Source: British Journal of Political Science, Vol. 12, No. 1 (Jan., 1982), pp. 1-25 Published by: Cambridge University Press Stable URL: http://www.jstor.org/stable/193719 . Accessed: 09/05/2014 10:06

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp

. JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.

.

Cambridge University Press is collaborating with JSTOR to digitize, preserve and extend access to British Journal of Political Science.

http://www.jstor.org

This content downloaded from 194.29.185.42 on Fri, 9 May 2014 10:06:58 AM All use subject to JSTOR Terms and Conditions B.J.Pol.S. 12, 1-25 I Printed in Great Britain

Judicial Ideology in the House of Lords: A Jurimetric Analysis

DAVID ROBERTSON*

For a variety of reasons political science in Britain has made no serious attempt to study courts and judges as political institutions and actors. Or at least this was true until recently. Several works in the last few years, especially Griffith's,1 Stevens's2 and a forthcoming book on the law lords by Alan Paterson,3 have pointed to a much needed change in this attitude. However, none of them have been works of political science, even though they have considered politics. By this I mean two things: they have not principally considered the judges' thoughts as political ideology; and they have not used the techniques and assumptions of rigorous, analytic political science. Indeed one of the few slightly earlier studies of the political role of the courts in Britain, by Morrison,4 specifically denies that such approaches, especially the statistical approach of jurimetrics, is possible in Britain. This article is an attempt to do the impossible, not so much because the author believes that Morrison's point is necessarily wrong, but because it is never sound methodology to abandon techniques that have been useful elsewhere without trying to make them work on different data sets. But first we must attempt to characterize the judicial role, before we try any study of the politics that may be attendant on judicial ideology. For political science there are perhaps three broad characterizations of the judicial role. I will identify these, for convenience, as the realist model, the class model, and the orthodox model. The realist model comes from the American tradition in the early decades of this century, best exemplified in works like Jerome Frank's Law and the Modern Mind and Llewellyn's The Bramble Bush.6 These assume, broadly, that legal decisions are never determined in any firm way by the rules, precedents and arguments in the court. Rather a judge does, and must, come to his decision intuitively, and then only rationalizes it by the legal material. In itself this says nothing about the types of decision a judge makes, and not much about how he comes to his intuitions. It was left to a school of political scientists in the I950S and I96os, associated with Glendon Schubert,7 to fill in these details.

St Hugh's College, Oxford. J John A. G. Griffith, The Politics of the Judiciary (London: Fontana, 1977). 2 Robert Stevens, Law and Politics. the House of Lords as a Judicial Body, I80oo-976 (London: Weidenfeld and Nicolson, 1979). 3 Alan Paterson, The Law Lords (forthcoming, I98I). 4 Fred L. Morrison, Courts and the Political Process in England (London: Sage, I973). 5 Jerome Frank, Law and the Modern Mind (New York: Tudor, 1936). 6 Karl Llewellyn, The Bramble Bush (New York: Columbia University Press, I930). 7 Glendon Schubert, The Judicial Mind (Evanston: Northwestern University Press, I965). P.p.

B.J.Pol.S. 12, 1-25 I Printed in Great Britain

Judicial Ideology in the House of Lords: A Jurimetric Analysis

DAVID ROBERTSON*

For a variety of reasons political science in Britain has made no serious attempt to study courts and judges as political institutions and actors. Or at least this was true until recently. Several works in the last few years, especially Griffith's,1 Stevens's2 and a forthcoming book on the law lords by Alan Paterson,3 have pointed to a much needed change in this attitude. However, none of them have been works of political science, even though they have considered politics. By this I mean two things: they have not principally considered the judges' thoughts as political ideology; and they have not used the techniques and assumptions of rigorous, analytic political science. Indeed one of the few slightly earlier studies of the political role of the courts in Britain, by Morrison,4 specifically denies that such approaches, especially the statistical approach of jurimetrics, is possible in Britain. This article is an attempt to do the impossible, not so much because the author believes that Morrison's point is necessarily wrong, but because it is never sound methodology to abandon techniques that have been useful elsewhere without trying to make them work on different data sets. But first we must attempt to characterize the judicial role, before we try any study of the politics that may be attendant on judicial ideology. For political science there are perhaps three broad characterizations of the judicial role. I will identify these, for convenience, as the realist model, the class model, and the orthodox model. The realist model comes from the American tradition in the early decades of this century, best exemplified in works like Jerome Frank's Law and the Modern Mind and Llewellyn's The Bramble Bush.6 These assume, broadly, that legal decisions are never determined in any firm way by the rules, precedents and arguments in the court. Rather a judge does, and must, come to his decision intuitively, and then only rationalizes it by the legal material. In itself this says nothing about the types of decision a judge makes, and not much about how he comes to his intuitions. It was left to a school of political scientists in the I950S and I96os, associated with Glendon Schubert,7 to fill in these details.

St Hugh's College, Oxford. J John A. G. Griffith, The Politics of the Judiciary (London: Fontana, 1977). 2 Robert Stevens, Law and Politics. the House of Lords as a Judicial Body, I80oo-976 (London: Weidenfeld and Nicolson, 1979). 3 Alan Paterson, The Law Lords (forthcoming, I98I). 4 Fred L. Morrison, Courts and the Political Process in England (London: Sage, I973). 5 Jerome Frank, Law and the Modern Mind (New York: Tudor, 1936). 6 Karl Llewellyn, The Bramble Bush (New York: Columbia University Press, I930). 7 Glendon Schubert, The Judicial Mind (Evanston: Northwestern University Press, I965). P.p.

B.J.Pol.S. 12, 1-25 I Printed in Great Britain

Judicial Ideology in the House of Lords: A Jurimetric Analysis

DAVID ROBERTSON*

For a variety of reasons political science in Britain has made no serious attempt to study courts and judges as political institutions and actors. Or at least this was true until recently. Several works in the last few years, especially Griffith's,1 Stevens's2 and a forthcoming book on the law lords by Alan Paterson,3 have pointed to a much needed change in this attitude. However, none of them have been works of political science, even though they have considered politics. By this I mean two things: they have not principally considered the judges' thoughts as political ideology; and they have not used the techniques and assumptions of rigorous, analytic political science. Indeed one of the few slightly earlier studies of the political role of the courts in Britain, by Morrison,4 specifically denies that such approaches, especially the statistical approach of jurimetrics, is possible in Britain. This article is an attempt to do the impossible, not so much because the author believes that Morrison's point is necessarily wrong, but because it is never sound methodology to abandon techniques that have been useful elsewhere without trying to make them work on different data sets. But first we must attempt to characterize the judicial role, before we try any study of the politics that may be attendant on judicial ideology. For political science there are perhaps three broad characterizations of the judicial role. I will identify these, for convenience, as the realist model, the class model, and the orthodox model. The realist model comes from the American tradition in the early decades of this century, best exemplified in works like Jerome Frank's Law and the Modern Mind and Llewellyn's The Bramble Bush.6 These assume, broadly, that legal decisions are never determined in any firm way by the rules, precedents and arguments in the court. Rather a judge does, and must, come to his decision intuitively, and then only rationalizes it by the legal material. In itself this says nothing about the types of decision a judge makes, and not much about how he comes to his intuitions. It was left to a school of political scientists in the I950S and I96os, associated with Glendon Schubert,7 to fill in these details.

St Hugh's College, Oxford. J John A. G. Griffith, The Politics of the Judiciary (London: Fontana, 1977). 2 Robert Stevens, Law and Politics. the House of Lords as a Judicial Body, I80oo-976 (London: Weidenfeld and Nicolson, 1979). 3 Alan Paterson, The Law Lords (forthcoming, I98I). 4 Fred L. Morrison, Courts and the Political Process in England (London: Sage, I973). 5 Jerome Frank, Law and the Modern Mind (New York: Tudor, 1936). 6 Karl Llewellyn, The Bramble Bush (New York: Columbia University Press, I930). 7 Glendon Schubert, The Judicial Mind (Evanston: Northwestern University Press, I965). P.p.

This content downloaded from 194.29.185.42 on Fri, 9 May 2014 10:06:58 AM All use subject to JSTOR Terms and Conditions 2 ROBERTSON

The class model is broader still as a label, but refers mainly to Marxist attacks on the law and the courts of 'bourgeois' states. The best recent example is J. A. G. Griffith's The Politics of the Judiciary.8 He makes two characteristic claims in this book, although he seldom disentangles them himself. Primarily he argues that discretionary problems in law are solved by judges consulting their own notion of 'public interest', which is made up of a belief that the State must be protected from danger, including 'moral' danger, that legislation should be limited in its effects on property and other 'bourgeois' freedoms, and that a general political philosophy associated with the Conservative party should prevail.9 The second assumption that he seems to make is that judges, in so acting, are acting out of a class-conditioned perspective, as agents of a dominant socio-economic group. Thus it is not so much the private and perhaps idiosyncratic intuitions of judges unprovided with determinate law, but the deliberate and systematic protection of a particular class that characterizes discretionary judgments.10 One of the class model's major drawbacks is that, by treating judges as largely interchangeable, it does not enable us to differentiate various profiles of judicial ideology. If judges are taken as ideologically homogenous, we can learn nothing in detail about how they do operate, and nothing about the predisposing - I am hesitant to say causal - elements. The third, orthodox, model is hard to describe but approximates to the public lawyer's interpretation of a broadly Austinian theory. It has little of a systematic nature to say about the uses of discretion. I mean by it the working assumptions of lawyers and political scientists interested in public law. It contains perhaps the three following assumptions:

(i) Discretion does exist, because statutes are sometimes vague, or precedents missing or conflicting. (2) Most judges most of the time will try very hard to stick within the clear meaning of the statute or the guidance of precedent. Some will from time to time try to develop the law to fit modern needs, but this 'public policy horse' is dangerous, to be ridden seldom and cautiously. (3) Where 'difficult cases' do crop up, some judges will act from personal idiosyncrasy, but these are mainly self-cancelling unsystematic quirks, not representing an intrusion of class ideology, and too limited to fit the realist model of permanent intuition. It is difficult, if not impossible, to test any of these models. Not only are their predictions entirely unclear, but, as a philosopher of science would say, they use different observation languages. However, they are basically empirical propositions, and raise questions that need to be answered empirically. The empirical work on judicial decision making known to political science

8 See fn.i. 9 Griffith, The Politics of the Judiciary, p. I95. 10 Griffith, The Politics of the Judiciary, p. 2o8f.

The class model is broader still as a label, but refers mainly to Marxist attacks on the law and the courts of 'bourgeois' states. The best recent example is J. A. G. Griffith's The Politics of the Judiciary.8 He makes two characteristic claims in this book, although he seldom disentangles them himself. Primarily he argues that discretionary problems in law are solved by judges consulting their own notion of 'public interest', which is made up of a belief that the State must be protected from danger, including 'moral' danger, that legislation should be limited in its effects on property and other 'bourgeois' freedoms, and that a general political philosophy associated with the Conservative party should prevail.9 The second assumption that he seems to make is that judges, in so acting, are acting out of a class-conditioned perspective, as agents of a dominant socio-economic group. Thus it is not so much the private and perhaps idiosyncratic intuitions of judges unprovided with determinate law, but the deliberate and systematic protection of a particular class that characterizes discretionary judgments.10 One of the class model's major drawbacks is that, by treating judges as largely interchangeable, it does not enable us to differentiate various profiles of judicial ideology. If judges are taken as ideologically homogenous, we can learn nothing in detail about how they do operate, and nothing about the predisposing - I am hesitant to say causal - elements. The third, orthodox, model is hard to describe but approximates to the public lawyer's interpretation of a broadly Austinian theory. It has little of a systematic nature to say about the uses of discretion. I mean by it the working assumptions of lawyers and political scientists interested in public law. It contains perhaps the three following assumptions:

(i) Discretion does exist, because statutes are sometimes vague, or precedents missing or conflicting. (2) Most judges most of the time will try very hard to stick within the clear meaning of the statute or the guidance of precedent. Some will from time to time try to develop the law to fit modern needs, but this 'public policy horse' is dangerous, to be ridden seldom and cautiously. (3) Where 'difficult cases' do crop up, some judges will act from personal idiosyncrasy, but these are mainly self-cancelling unsystematic quirks, not representing an intrusion of class ideology, and too limited to fit the realist model of permanent intuition. It is difficult, if not impossible, to test any of these models. Not only are their predictions entirely unclear, but, as a philosopher of science would say, they use different observation languages. However, they are basically empirical propositions, and raise questions that need to be answered empirically. The empirical work on judicial decision making known to political science

8 See fn.i. 9 Griffith, The Politics of the Judiciary, p. I95. 10 Griffith, The Politics of the Judiciary, p. 2o8f.

The class model is broader still as a label, but refers mainly to Marxist attacks on the law and the courts of 'bourgeois' states. The best recent example is J. A. G. Griffith's The Politics of the Judiciary.8 He makes two characteristic claims in this book, although he seldom disentangles them himself. Primarily he argues that discretionary problems in law are solved by judges consulting their own notion of 'public interest', which is made up of a belief that the State must be protected from danger, including 'moral' danger, that legislation should be limited in its effects on property and other 'bourgeois' freedoms, and that a general political philosophy associated with the Conservative party should prevail.9 The second assumption that he seems to make is that judges, in so acting, are acting out of a class-conditioned perspective, as agents of a dominant socio-economic group. Thus it is not so much the private and perhaps idiosyncratic intuitions of judges unprovided with determinate law, but the deliberate and systematic protection of a particular class that characterizes discretionary judgments.10 One of the class model's major drawbacks is that, by treating judges as largely interchangeable, it does not enable us to differentiate various profiles of judicial ideology. If judges are taken as ideologically homogenous, we can learn nothing in detail about how they do operate, and nothing about the predisposing - I am hesitant to say causal - elements. The third, orthodox, model is hard to describe but approximates to the public lawyer's interpretation of a broadly Austinian theory. It has little of a systematic nature to say about the uses of discretion. I mean by it the working assumptions of lawyers and political scientists interested in public law. It contains perhaps the three following assumptions:

(i) Discretion does exist, because statutes are sometimes vague, or precedents missing or conflicting. (2) Most judges most of the time will try very hard to stick within the clear meaning of the statute or the guidance of precedent. Some will from time to time try to develop the law to fit modern needs, but this 'public policy horse' is dangerous, to be ridden seldom and cautiously. (3) Where 'difficult cases' do crop up, some judges will act from personal idiosyncrasy, but these are mainly self-cancelling unsystematic quirks, not representing an intrusion of class ideology, and too limited to fit the realist model of permanent intuition. It is difficult, if not impossible, to test any of these models. Not only are their predictions entirely unclear, but, as a philosopher of science would say, they use different observation languages. However, they are basically empirical propositions, and raise questions that need to be answered empirically. The empirical work on judicial decision making known to political science

8 See fn.i. 9 Griffith, The Politics of the Judiciary, p. I95. 10 Griffith, The Politics of the Judiciary, p. 2o8f.

This content downloaded from 194.29.185.42 on Fri, 9 May 2014 10:06:58 AM All use subject to JSTOR Terms and Conditions Judicial Ideology Judicial Judicial Ideology Ideology in the House of Lords

comes from the American realist tradition, under the name of 'judicial behaviour', and has been carried out principally on the Supreme Courts of the United States, Canada and Australia.11 Its celebrated paradigm is Schubert's The Judicial Mind,12 a classic of what has been called 'jurimetrics'. This involves complex statistical analysis of the voting by Supreme Court judges in 'political' cases during the I950s and I96os. Its principal conclusion, that these judges can be seen as arrayed on two intercorrelated dimensions, 'political liberalism' and 'economic liberalism', with several minor dimensions, is well known. It is often overlooked that Schubert's work is not just descriptive data-analysis. The data-analysis is informed by, and partially intended as a test of, a highly specified, if obscure, psychological model of stimulus perception and choice-making that is influenced by the 'cognitive psychology' of Coombs.13 The most important criticism of Schubert and his associates is Karl Becker's Political Behaviouralism and Modern Jurisprudence.14 This stresses the central importance of the judge's own notion of his role, and especially of the degree of his commitment to principles like that of stare decisis.15 Becker argues that although judges may have, or be forced to consult, a private ideology, and may be influenced by emotional reactions to the facts and personalities of the case, they are highly trained, and therefore highly socialized, professionals.16 Come what may, the ideology and the emotions consulted are those of judges. Although research like Schubert's is very useful, the dimensions that emerge as characterizing judicial ideology ought, at least partially, to represent judicial differences on essentially 'professional judicial' questions.17 However, Schubert's approach is markedly superior to most class analysis, and certainly to that of Griffith, because it fixes firmly on the judicial decision. We therefore need to offer, with due caution, an individuating descriptive account of judicial ideology, on the basis of unambiguous and standardizable data. Even the orthodox view invites research of Schubert's type, although not with his theoretical presuppositions. It accepts that there is discretion on the margin, and that sometimes it is influenced by judicial prejudice. But it is silent on how much and how salient and systematic such 'prejudicial' decision

ll For example, the case studies contained in Glendon Schubert and David Danelski, eds, Comparative Judicial Behavior (New York: Oxford University Press, I969). 12 Schubert, The Judicial Mind, and also Glendon Schubert, The Judicial Mind Revisited (New York: Oxford University Press, I974). 13 Clyde H. Coombs, A Theory of Data (New York: Wiley, i964). 14 Karl Becker, Political Behaviouralism and Modern Jurisprudence (Chicago: Rand McNally, I965). 15 Becker, Political Behaviouralism and Modern Jurisprudence, p. 105. 16 Becker, Political Behaviouralism and Modern Jurisprudence, p. 243. 17 This position has become common now in American political science, where survey examinations of judges' role expectations have often been carried out. Some comparative work of this type has also been done.

comes from the American realist tradition, under the name of 'judicial behaviour', and has been carried out principally on the Supreme Courts of the United States, Canada and Australia.11 Its celebrated paradigm is Schubert's The Judicial Mind,12 a classic of what has been called 'jurimetrics'. This involves complex statistical analysis of the voting by Supreme Court judges in 'political' cases during the I950s and I96os. Its principal conclusion, that these judges can be seen as arrayed on two intercorrelated dimensions, 'political liberalism' and 'economic liberalism', with several minor dimensions, is well known. It is often overlooked that Schubert's work is not just descriptive data-analysis. The data-analysis is informed by, and partially intended as a test of, a highly specified, if obscure, psychological model of stimulus perception and choice-making that is influenced by the 'cognitive psychology' of Coombs.13 The most important criticism of Schubert and his associates is Karl Becker's Political Behaviouralism and Modern Jurisprudence.14 This stresses the central importance of the judge's own notion of his role, and especially of the degree of his commitment to principles like that of stare decisis.15 Becker argues that although judges may have, or be forced to consult, a private ideology, and may be influenced by emotional reactions to the facts and personalities of the case, they are highly trained, and therefore highly socialized, professionals.16 Come what may, the ideology and the emotions consulted are those of judges. Although research like Schubert's is very useful, the dimensions that emerge as characterizing judicial ideology ought, at least partially, to represent judicial differences on essentially 'professional judicial' questions.17 However, Schubert's approach is markedly superior to most class analysis, and certainly to that of Griffith, because it fixes firmly on the judicial decision. We therefore need to offer, with due caution, an individuating descriptive account of judicial ideology, on the basis of unambiguous and standardizable data. Even the orthodox view invites research of Schubert's type, although not with his theoretical presuppositions. It accepts that there is discretion on the margin, and that sometimes it is influenced by judicial prejudice. But it is silent on how much and how salient and systematic such 'prejudicial' decision

ll For example, the case studies contained in Glendon Schubert and David Danelski, eds, Comparative Judicial Behavior (New York: Oxford University Press, I969). 12 Schubert, The Judicial Mind, and also Glendon Schubert, The Judicial Mind Revisited (New York: Oxford University Press, I974). 13 Clyde H. Coombs, A Theory of Data (New York: Wiley, i964). 14 Karl Becker, Political Behaviouralism and Modern Jurisprudence (Chicago: Rand McNally, I965). 15 Becker, Political Behaviouralism and Modern Jurisprudence, p. 105. 16 Becker, Political Behaviouralism and Modern Jurisprudence, p. 243. 17 This position has become common now in American political science, where survey examinations of judges' role expectations have often been carried out. Some comparative work of this type has also been done.

comes from the American realist tradition, under the name of 'judicial behaviour', and has been carried out principally on the Supreme Courts of the United States, Canada and Australia.11 Its celebrated paradigm is Schubert's The Judicial Mind,12 a classic of what has been called 'jurimetrics'. This involves complex statistical analysis of the voting by Supreme Court judges in 'political' cases during the I950s and I96os. Its principal conclusion, that these judges can be seen as arrayed on two intercorrelated dimensions, 'political liberalism' and 'economic liberalism', with several minor dimensions, is well known. It is often overlooked that Schubert's work is not just descriptive data-analysis. The data-analysis is informed by, and partially intended as a test of, a highly specified, if obscure, psychological model of stimulus perception and choice-making that is influenced by the 'cognitive psychology' of Coombs.13 The most important criticism of Schubert and his associates is Karl Becker's Political Behaviouralism and Modern Jurisprudence.14 This stresses the central importance of the judge's own notion of his role, and especially of the degree of his commitment to principles like that of stare decisis.15 Becker argues that although judges may have, or be forced to consult, a private ideology, and may be influenced by emotional reactions to the facts and personalities of the case, they are highly trained, and therefore highly socialized, professionals.16 Come what may, the ideology and the emotions consulted are those of judges. Although research like Schubert's is very useful, the dimensions that emerge as characterizing judicial ideology ought, at least partially, to represent judicial differences on essentially 'professional judicial' questions.17 However, Schubert's approach is markedly superior to most class analysis, and certainly to that of Griffith, because it fixes firmly on the judicial decision. We therefore need to offer, with due caution, an individuating descriptive account of judicial ideology, on the basis of unambiguous and standardizable data. Even the orthodox view invites research of Schubert's type, although not with his theoretical presuppositions. It accepts that there is discretion on the margin, and that sometimes it is influenced by judicial prejudice. But it is silent on how much and how salient and systematic such 'prejudicial' decision

ll For example, the case studies contained in Glendon Schubert and David Danelski, eds, Comparative Judicial Behavior (New York: Oxford University Press, I969). 12 Schubert, The Judicial Mind, and also Glendon Schubert, The Judicial Mind Revisited (New York: Oxford University Press, I974). 13 Clyde H. Coombs, A Theory of Data (New York: Wiley, i964). 14 Karl Becker, Political Behaviouralism and Modern Jurisprudence (Chicago: Rand McNally, I965). 15 Becker, Political Behaviouralism and Modern Jurisprudence, p. 105. 16 Becker, Political Behaviouralism and Modern Jurisprudence, p. 243. 17 This position has become common now in American political science, where survey examinations of judges' role expectations have often been carried out. Some comparative work of this type has also been done.

3

This content downloaded from 194.29.185.42 on Fri, 9 May 2014 10:06:58 AM All use subject to JSTOR Terms and Conditions 4 ROBERTSON

making is. If one can demonstrate that there are fairly fixed and specifiable dimensions of judicial ideology, operating over a fairly wide field, the orthodox model may have a certain amount of explaining, and explaining away, to do. The data analysis in this article is only exploratory. The whole exercise is an attempt to see if such an approach may be useful and justified in general. I give rather more attention therefore to the methodology and its potential weaknesses than is usual. Indeed, while I hope that the analysis itself is interesting, the methodology is in some ways the core of this discussion. Very broadly, the article attempts to create a statistical model of the similarities and differences between judicial styles. The model expresses these similarities in terms of abstract numerical dimensions. These dimensions have, in themselves, no substantive meaning. Meaning can only be given by a second stage in the analysis - interpreting these dimensions - which I attempt by looking at how groups of judges at different ends of these dimensions react to particular sorts of cases. The third stage is to show how changing membership of the court has affected the salience of the two primary dimensions I analyse. At this stage I also look at the relationship between a judge's position on each of the separate dimensions. The final part of the article attempts to sketch a more general theory of the intellectual orientations that might lead judges to occupy varied 'ideological' positions. I offer some evidence to support the idea that these are orientations to the very business of being an appellate judge. It should be borne in mind throughout that my attitude is one of great respect to the judiciary, and that I end by making the notion of judicial ideology a matter of intellectual and professional preference, not a psychological drive.

RESEARCH OUTLINE AND METHODOLOGICAL PROBLEMS

Basic Premises

Judicial decision making in English Appellate Courts is partly discretionary. My basic premise is that, where discretion exists, an individual judge's voting will be influenced by his beliefs and attitudes inside a specifically legal and 'professional' ideology. This 'ideology' will vary from judge to judge, and whether it is conservative or not, will be complicated both by individual variation with respect to any one aspect of the ideology, and also by the ideology having several aspects. There is no reason to believe that judges' views on the separate aspects will be correlated, although they may be. These ideological variances should in part account for fairly stable groupings of judges, and should be discoverable, at least in crude form, by analysis of voting.

making is. If one can demonstrate that there are fairly fixed and specifiable dimensions of judicial ideology, operating over a fairly wide field, the orthodox model may have a certain amount of explaining, and explaining away, to do. The data analysis in this article is only exploratory. The whole exercise is an attempt to see if such an approach may be useful and justified in general. I give rather more attention therefore to the methodology and its potential weaknesses than is usual. Indeed, while I hope that the analysis itself is interesting, the methodology is in some ways the core of this discussion. Very broadly, the article attempts to create a statistical model of the similarities and differences between judicial styles. The model expresses these similarities in terms of abstract numerical dimensions. These dimensions have, in themselves, no substantive meaning. Meaning can only be given by a second stage in the analysis - interpreting these dimensions - which I attempt by looking at how groups of judges at different ends of these dimensions react to particular sorts of cases. The third stage is to show how changing membership of the court has affected the salience of the two primary dimensions I analyse. At this stage I also look at the relationship between a judge's position on each of the separate dimensions. The final part of the article attempts to sketch a more general theory of the intellectual orientations that might lead judges to occupy varied 'ideological' positions. I offer some evidence to support the idea that these are orientations to the very business of being an appellate judge. It should be borne in mind throughout that my attitude is one of great respect to the judiciary, and that I end by making the notion of judicial ideology a matter of intellectual and professional preference, not a psychological drive.

RESEARCH OUTLINE AND METHODOLOGICAL PROBLEMS

Basic Premises

Judicial decision making in English Appellate Courts is partly discretionary. My basic premise is that, where discretion exists, an individual judge's voting will be influenced by his beliefs and attitudes inside a specifically legal and 'professional' ideology. This 'ideology' will vary from judge to judge, and whether it is conservative or not, will be complicated both by individual variation with respect to any one aspect of the ideology, and also by the ideology having several aspects. There is no reason to believe that judges' views on the separate aspects will be correlated, although they may be. These ideological variances should in part account for fairly stable groupings of judges, and should be discoverable, at least in crude form, by analysis of voting.

making is. If one can demonstrate that there are fairly fixed and specifiable dimensions of judicial ideology, operating over a fairly wide field, the orthodox model may have a certain amount of explaining, and explaining away, to do. The data analysis in this article is only exploratory. The whole exercise is an attempt to see if such an approach may be useful and justified in general. I give rather more attention therefore to the methodology and its potential weaknesses than is usual. Indeed, while I hope that the analysis itself is interesting, the methodology is in some ways the core of this discussion. Very broadly, the article attempts to create a statistical model of the similarities and differences between judicial styles. The model expresses these similarities in terms of abstract numerical dimensions. These dimensions have, in themselves, no substantive meaning. Meaning can only be given by a second stage in the analysis - interpreting these dimensions - which I attempt by looking at how groups of judges at different ends of these dimensions react to particular sorts of cases. The third stage is to show how changing membership of the court has affected the salience of the two primary dimensions I analyse. At this stage I also look at the relationship between a judge's position on each of the separate dimensions. The final part of the article attempts to sketch a more general theory of the intellectual orientations that might lead judges to occupy varied 'ideological' positions. I offer some evidence to support the idea that these are orientations to the very business of being an appellate judge. It should be borne in mind throughout that my attitude is one of great respect to the judiciary, and that I end by making the notion of judicial ideology a matter of intellectual and professional preference, not a psychological drive.

RESEARCH OUTLINE AND METHODOLOGICAL PROBLEMS

Basic Premises

Judicial decision making in English Appellate Courts is partly discretionary. My basic premise is that, where discretion exists, an individual judge's voting will be influenced by his beliefs and attitudes inside a specifically legal and 'professional' ideology. This 'ideology' will vary from judge to judge, and whether it is conservative or not, will be complicated both by individual variation with respect to any one aspect of the ideology, and also by the ideology having several aspects. There is no reason to believe that judges' views on the separate aspects will be correlated, although they may be. These ideological variances should in part account for fairly stable groupings of judges, and should be discoverable, at least in crude form, by analysis of voting.

This content downloaded from 194.29.185.42 on Fri, 9 May 2014 10:06:58 AM All use subject to JSTOR Terms and Conditions Judicial Ideology in the House of Lords 5

Selection of Data

Decisions on cases decided by the House of Lords over a twelve-year period were studied, mainly through statistical analysis. The House of Lords Appellate Committee was chosen because it is the only English court not absolutely bound by precedent, and thus the possibility of discretion in judgment is at its highest. As it is also the final court of appeal this possibility is increased, both because it only gets difficult cases, and because it need not fear being overruled. In addition, the incidence of dissent, vital for this type of analysis, is higher than in any other multi-member court in the United Kingdom. The principal analysis is carried out only on cases where decisions were not unanimous.

The Basic Method and its Justification

It is assumed that a judicial ideology of the type hypothesized will show itself in two ways - by causing some pairs of judges to think more alike than others, and by causing judges to vary in the preponderance of their votes for particular sides in cases characterized by certain issues. Both of these assumptions are also made by judicial behaviouralists in the United States, although they conceptualize the problem differently, and use different techniques. Thus the basic method has two stages. The first extracts from data on judicial voting some description of the degree of affinity between judges. The second explores whether or not these affinity relationships are correlated with differences in support for certain types of litigant. In the first stage we take the cases for the period and construct a measure of agreement for each pair of judges where this is possible. The 'agreement index' is simply the proportion of all cases the pair of judges heard together on which they voted the same way. This matrix says nothing about the direction of influence in agreement, or the depth of agreement in detail, but reflects only the fact of concurring or dissenting judgments. The 'agreement index' matrix is decomposed by multi-dimensional analysis into a two dimensional representation of judicial agreement. This represents the agreement/disagreement ratios of judges in what is hoped is an intuitively comprehensible way. The more judges agree, the closer their positions are in the underlying assumed 'ideological space'. The idea of 'ideological space' is, of course, only a metaphor, but a useful one. The second stage tries to identify the as yet meaningless dimensions of this space. This is done by taking several basic categories of cases where different 'sides' can be unambiguously and meaningfully identified. The tendency of a judge to vote for a particular side is measured by the proportion of all cases in a particular category in which he gave judgment for that side. These 'category scores' are correlated in various ways with the positions of the judges on each dimension. The resulting pattern, in which the dimensions of the spatial representation are shown to be correlated with category voting, is

Judicial Ideology in the House of Lords 5

Selection of Data

Decisions on cases decided by the House of Lords over a twelve-year period were studied, mainly through statistical analysis. The House of Lords Appellate Committee was chosen because it is the only English court not absolutely bound by precedent, and thus the possibility of discretion in judgment is at its highest. As it is also the final court of appeal this possibility is increased, both because it only gets difficult cases, and because it need not fear being overruled. In addition, the incidence of dissent, vital for this type of analysis, is higher than in any other multi-member court in the United Kingdom. The principal analysis is carried out only on cases where decisions were not unanimous.

The Basic Method and its Justification

It is assumed that a judicial ideology of the type hypothesized will show itself in two ways - by causing some pairs of judges to think more alike than others, and by causing judges to vary in the preponderance of their votes for particular sides in cases characterized by certain issues. Both of these assumptions are also made by judicial behaviouralists in the United States, although they conceptualize the problem differently, and use different techniques. Thus the basic method has two stages. The first extracts from data on judicial voting some description of the degree of affinity between judges. The second explores whether or not these affinity relationships are correlated with differences in support for certain types of litigant. In the first stage we take the cases for the period and construct a measure of agreement for each pair of judges where this is possible. The 'agreement index' is simply the proportion of all cases the pair of judges heard together on which they voted the same way. This matrix says nothing about the direction of influence in agreement, or the depth of agreement in detail, but reflects only the fact of concurring or dissenting judgments. The 'agreement index' matrix is decomposed by multi-dimensional analysis into a two dimensional representation of judicial agreement. This represents the agreement/disagreement ratios of judges in what is hoped is an intuitively comprehensible way. The more judges agree, the closer their positions are in the underlying assumed 'ideological space'. The idea of 'ideological space' is, of course, only a metaphor, but a useful one. The second stage tries to identify the as yet meaningless dimensions of this space. This is done by taking several basic categories of cases where different 'sides' can be unambiguously and meaningfully identified. The tendency of a judge to vote for a particular side is measured by the proportion of all cases in a particular category in which he gave judgment for that side. These 'category scores' are correlated in various ways with the positions of the judges on each dimension. The resulting pattern, in which the dimensions of the spatial representation are shown to be correlated with category voting, is

Judicial Ideology in the House of Lords 5

Selection of Data

Decisions on cases decided by the House of Lords over a twelve-year period were studied, mainly through statistical analysis. The House of Lords Appellate Committee was chosen because it is the only English court not absolutely bound by precedent, and thus the possibility of discretion in judgment is at its highest. As it is also the final court of appeal this possibility is increased, both because it only gets difficult cases, and because it need not fear being overruled. In addition, the incidence of dissent, vital for this type of analysis, is higher than in any other multi-member court in the United Kingdom. The principal analysis is carried out only on cases where decisions were not unanimous.

The Basic Method and its Justification

It is assumed that a judicial ideology of the type hypothesized will show itself in two ways - by causing some pairs of judges to think more alike than others, and by causing judges to vary in the preponderance of their votes for particular sides in cases characterized by certain issues. Both of these assumptions are also made by judicial behaviouralists in the United States, although they conceptualize the problem differently, and use different techniques. Thus the basic method has two stages. The first extracts from data on judicial voting some description of the degree of affinity between judges. The second explores whether or not these affinity relationships are correlated with differences in support for certain types of litigant. In the first stage we take the cases for the period and construct a measure of agreement for each pair of judges where this is possible. The 'agreement index' is simply the proportion of all cases the pair of judges heard together on which they voted the same way. This matrix says nothing about the direction of influence in agreement, or the depth of agreement in detail, but reflects only the fact of concurring or dissenting judgments. The 'agreement index' matrix is decomposed by multi-dimensional analysis into a two dimensional representation of judicial agreement. This represents the agreement/disagreement ratios of judges in what is hoped is an intuitively comprehensible way. The more judges agree, the closer their positions are in the underlying assumed 'ideological space'. The idea of 'ideological space' is, of course, only a metaphor, but a useful one. The second stage tries to identify the as yet meaningless dimensions of this space. This is done by taking several basic categories of cases where different 'sides' can be unambiguously and meaningfully identified. The tendency of a judge to vote for a particular side is measured by the proportion of all cases in a particular category in which he gave judgment for that side. These 'category scores' are correlated in various ways with the positions of the judges on each dimension. The resulting pattern, in which the dimensions of the spatial representation are shown to be correlated with category voting, is

This content downloaded from 194.29.185.42 on Fri, 9 May 2014 10:06:58 AM All use subject to JSTOR Terms and Conditions 6 ROBERTSON

held to be a rough description of certain aspects of the hypothesized judicial ideology. In a third and final stage a different question is asked. Can these 'substantive' patterns be seen as representing a more fundamental single difference that makes general theoretical sense of English judges' behaviour? By examining the voting patterns of two active and intellectually dominant Law Lords of the period it is argued that this may be the case

Research Problems, Solutions, and Some Working Assumptions

Why has no jurimetrics been done on the English House of Lords, when so much has been done on the Supreme Courts of the United States and Canada and on the High Court of Australia?18 Because it is very much harder, technically, to do such research in Britain. As mentioned above, one American political scientist has argued that jurimetrics simply cannot be applied in Britain.19 No other example is known to me. The most obvious problems relate to the structures and practice of our Higher Appellate Courts. Compared with the United States Supreme Court, the House of Lords (a) hears far fewer cases in a year (perhaps forty to fifty compared with well over a hundred); (b) decides far more of them unanimously (70-80 per cent in Britain, rarely even 50 per cent in the United States, according to Schubert's research); (c) has more judges but hears cases with less. (In the House, a bench of five Law Lords, drawn from eleven or twelve more or less at random, hears each case, while in the United States the norm is for all nine Justices to vote on each case.) The immediate result is the paucity of data. Because of these differences, Schubert could analyse each year's voting separately, and still have plenty of cases to determine the agreement for each pair of judges as well as having a large number of cases for both his 'political liberalism' and 'economic liberalism' dimensions. There is another difference of practice that raises for my research a problem Schubert did not have to face. On the United States Bench majority and dissenting judges usually group together, offering joint opinions, although there may be more than one opinion for both majority and minority. In Britain it is quite usual to issue five separate, and often radically different, opinions even when all the judges vote the same way! This involves the following very strong assumption: when two judges vote the same way in a case, they can be taken to be substantially in agreement in preferring that one side win the case rather than the other.20 Although this may not seem an

18 See Schubert and Danelski, eds, Comparative Judicial Behavior. 19 Morrison, Courts and the Political Process in England. Others have felt, however, that it was both possible and desirable. See, for example, Burton Atkin, 'Judicial Behaviour Approaches in America and Britain', paper presented to the Law and Politics panel, Conference of the Political Studies Association, Hull, April I981. 20 I am here ignoring the potentially important difference between giving a formal concurrence with another judge's reasoned opinion, and issuing one's own opinion. In the latter case two

held to be a rough description of certain aspects of the hypothesized judicial ideology. In a third and final stage a different question is asked. Can these 'substantive' patterns be seen as representing a more fundamental single difference that makes general theoretical sense of English judges' behaviour? By examining the voting patterns of two active and intellectually dominant Law Lords of the period it is argued that this may be the case

Research Problems, Solutions, and Some Working Assumptions

Why has no jurimetrics been done on the English House of Lords, when so much has been done on the Supreme Courts of the United States and Canada and on the High Court of Australia?18 Because it is very much harder, technically, to do such research in Britain. As mentioned above, one American political scientist has argued that jurimetrics simply cannot be applied in Britain.19 No other example is known to me. The most obvious problems relate to the structures and practice of our Higher Appellate Courts. Compared with the United States Supreme Court, the House of Lords (a) hears far fewer cases in a year (perhaps forty to fifty compared with well over a hundred); (b) decides far more of them unanimously (70-80 per cent in Britain, rarely even 50 per cent in the United States, according to Schubert's research); (c) has more judges but hears cases with less. (In the House, a bench of five Law Lords, drawn from eleven or twelve more or less at random, hears each case, while in the United States the norm is for all nine Justices to vote on each case.) The immediate result is the paucity of data. Because of these differences, Schubert could analyse each year's voting separately, and still have plenty of cases to determine the agreement for each pair of judges as well as having a large number of cases for both his 'political liberalism' and 'economic liberalism' dimensions. There is another difference of practice that raises for my research a problem Schubert did not have to face. On the United States Bench majority and dissenting judges usually group together, offering joint opinions, although there may be more than one opinion for both majority and minority. In Britain it is quite usual to issue five separate, and often radically different, opinions even when all the judges vote the same way! This involves the following very strong assumption: when two judges vote the same way in a case, they can be taken to be substantially in agreement in preferring that one side win the case rather than the other.20 Although this may not seem an

18 See Schubert and Danelski, eds, Comparative Judicial Behavior. 19 Morrison, Courts and the Political Process in England. Others have felt, however, that it was both possible and desirable. See, for example, Burton Atkin, 'Judicial Behaviour Approaches in America and Britain', paper presented to the Law and Politics panel, Conference of the Political Studies Association, Hull, April I981. 20 I am here ignoring the potentially important difference between giving a formal concurrence with another judge's reasoned opinion, and issuing one's own opinion. In the latter case two

held to be a rough description of certain aspects of the hypothesized judicial ideology. In a third and final stage a different question is asked. Can these 'substantive' patterns be seen as representing a more fundamental single difference that makes general theoretical sense of English judges' behaviour? By examining the voting patterns of two active and intellectually dominant Law Lords of the period it is argued that this may be the case

Research Problems, Solutions, and Some Working Assumptions

Why has no jurimetrics been done on the English House of Lords, when so much has been done on the Supreme Courts of the United States and Canada and on the High Court of Australia?18 Because it is very much harder, technically, to do such research in Britain. As mentioned above, one American political scientist has argued that jurimetrics simply cannot be applied in Britain.19 No other example is known to me. The most obvious problems relate to the structures and practice of our Higher Appellate Courts. Compared with the United States Supreme Court, the House of Lords (a) hears far fewer cases in a year (perhaps forty to fifty compared with well over a hundred); (b) decides far more of them unanimously (70-80 per cent in Britain, rarely even 50 per cent in the United States, according to Schubert's research); (c) has more judges but hears cases with less. (In the House, a bench of five Law Lords, drawn from eleven or twelve more or less at random, hears each case, while in the United States the norm is for all nine Justices to vote on each case.) The immediate result is the paucity of data. Because of these differences, Schubert could analyse each year's voting separately, and still have plenty of cases to determine the agreement for each pair of judges as well as having a large number of cases for both his 'political liberalism' and 'economic liberalism' dimensions. There is another difference of practice that raises for my research a problem Schubert did not have to face. On the United States Bench majority and dissenting judges usually group together, offering joint opinions, although there may be more than one opinion for both majority and minority. In Britain it is quite usual to issue five separate, and often radically different, opinions even when all the judges vote the same way! This involves the following very strong assumption: when two judges vote the same way in a case, they can be taken to be substantially in agreement in preferring that one side win the case rather than the other.20 Although this may not seem an

18 See Schubert and Danelski, eds, Comparative Judicial Behavior. 19 Morrison, Courts and the Political Process in England. Others have felt, however, that it was both possible and desirable. See, for example, Burton Atkin, 'Judicial Behaviour Approaches in America and Britain', paper presented to the Law and Politics panel, Conference of the Political Studies Association, Hull, April I981. 20 I am here ignoring the potentially important difference between giving a formal concurrence with another judge's reasoned opinion, and issuing one's own opinion. In the latter case two

This content downloaded from 194.29.185.42 on Fri, 9 May 2014 10:06:58 AM All use subject to JSTOR Terms and Conditions Judicial Ideology Judicial in the House of Lords Ideology Judicial in the House of Lords Ideology in the House of Lords

over-demanding assumption, it could make certain conclusions false if drawn from data heavily characterized by a special sort of case. This is a case, quite frequent in English law, where one or more judges who think X ought to win in fact vote for Y because they feel bound by a precedent they dislike, and where some vote for Y feeling the precedent is itself desirable. Because the American court is much freer to overturn precedents than the House of Lords, Schubert and his associates have less need to justify this assumption. The justification is that we are not only interested in the inner meaning of a judge's decision. To a major extent the statement that judge X has a particular type of judicial ideology is an external judgement based on the consequence of his vote. For example, in the IT Case Lord Reid's decision was just as much a vote against individual freedom of expression, and had just as much effect on the freedom of press enjoyed by International Tribune as Lord Morris's, even though Reid felt the binding precedent to be wrong and had voted against it in the original case while Morris had supported the prosecution willingly on both occasions.21 There is a final difference between the American and the English experience that makes Schubert's kind of analysis much harder to perform on English data. It does not affect this article, because I take a totally different approach to the characterization (the 'coding') of decisions from Schubert's. The difference is that, because the United States has a written constitution and a Bill of Rights, the clauses and principles of which enter into real political conflict and political language, it is much easier to assert that case N is an example of category Z. For example, in the United States an obscene publications case will be fought expressly in terms of a constitutional 'positive law' prohibition on censorship in the first amendment. So for Schubert to 'code' a case as belonging to the 'political liberalism' category is fairly safe. Here a similar case will be fought in terms of varying interpretations of one or more statutes, using legal categories that may not translate at all into the language of political ideology. To claim that a case involving contractual infringement on freedom of trade fits into a category of economic conservatism would be much more subjective and difficult to defend than an equivalent in the United States. There it would probably arise under the Inter-State Commerce clause where decades of Congressional debate make clear what political lines are drawn up. An analyst must, in the end, impose his own categories. I have attempted to stick fairly close to legally existing categories, because I wish to interpret the resulting dimensions of judicial ideology as far as possible in terms of legal ideologies rather than, as with Schubert, asserting ideologies that have nothing directly to do with what judges may think they are doing. Yet again it

judges agreeing on how to dispose of a case may in fact have very different views. From my point of view this difference is irrelevant here, although it might have been preferable to take the distinction so as to record degrees of agreement. Unfortunately the available data are too sparse for this. 21 Knuller v. DPP (1973) A.C.435.

over-demanding assumption, it could make certain conclusions false if drawn from data heavily characterized by a special sort of case. This is a case, quite frequent in English law, where one or more judges who think X ought to win in fact vote for Y because they feel bound by a precedent they dislike, and where some vote for Y feeling the precedent is itself desirable. Because the American court is much freer to overturn precedents than the House of Lords, Schubert and his associates have less need to justify this assumption. The justification is that we are not only interested in the inner meaning of a judge's decision. To a major extent the statement that judge X has a particular type of judicial ideology is an external judgement based on the consequence of his vote. For example, in the IT Case Lord Reid's decision was just as much a vote against individual freedom of expression, and had just as much effect on the freedom of press enjoyed by International Tribune as Lord Morris's, even though Reid felt the binding precedent to be wrong and had voted against it in the original case while Morris had supported the prosecution willingly on both occasions.21 There is a final difference between the American and the English experience that makes Schubert's kind of analysis much harder to perform on English data. It does not affect this article, because I take a totally different approach to the characterization (the 'coding') of decisions from Schubert's. The difference is that, because the United States has a written constitution and a Bill of Rights, the clauses and principles of which enter into real political conflict and political language, it is much easier to assert that case N is an example of category Z. For example, in the United States an obscene publications case will be fought expressly in terms of a constitutional 'positive law' prohibition on censorship in the first amendment. So for Schubert to 'code' a case as belonging to the 'political liberalism' category is fairly safe. Here a similar case will be fought in terms of varying interpretations of one or more statutes, using legal categories that may not translate at all into the language of political ideology. To claim that a case involving contractual infringement on freedom of trade fits into a category of economic conservatism would be much more subjective and difficult to defend than an equivalent in the United States. There it would probably arise under the Inter-State Commerce clause where decades of Congressional debate make clear what political lines are drawn up. An analyst must, in the end, impose his own categories. I have attempted to stick fairly close to legally existing categories, because I wish to interpret the resulting dimensions of judicial ideology as far as possible in terms of legal ideologies rather than, as with Schubert, asserting ideologies that have nothing directly to do with what judges may think they are doing. Yet again it

judges agreeing on how to dispose of a case may in fact have very different views. From my point of view this difference is irrelevant here, although it might have been preferable to take the distinction so as to record degrees of agreement. Unfortunately the available data are too sparse for this. 21 Knuller v. DPP (1973) A.C.435.

over-demanding assumption, it could make certain conclusions false if drawn from data heavily characterized by a special sort of case. This is a case, quite frequent in English law, where one or more judges who think X ought to win in fact vote for Y because they feel bound by a precedent they dislike, and where some vote for Y feeling the precedent is itself desirable. Because the American court is much freer to overturn precedents than the House of Lords, Schubert and his associates have less need to justify this assumption. The justification is that we are not only interested in the inner meaning of a judge's decision. To a major extent the statement that judge X has a particular type of judicial ideology is an external judgement based on the consequence of his vote. For example, in the IT Case Lord Reid's decision was just as much a vote against individual freedom of expression, and had just as much effect on the freedom of press enjoyed by International Tribune as Lord Morris's, even though Reid felt the binding precedent to be wrong and had voted against it in the original case while Morris had supported the prosecution willingly on both occasions.21 There is a final difference between the American and the English experience that makes Schubert's kind of analysis much harder to perform on English data. It does not affect this article, because I take a totally different approach to the characterization (the 'coding') of decisions from Schubert's. The difference is that, because the United States has a written constitution and a Bill of Rights, the clauses and principles of which enter into real political conflict and political language, it is much easier to assert that case N is an example of category Z. For example, in the United States an obscene publications case will be fought expressly in terms of a constitutional 'positive law' prohibition on censorship in the first amendment. So for Schubert to 'code' a case as belonging to the 'political liberalism' category is fairly safe. Here a similar case will be fought in terms of varying interpretations of one or more statutes, using legal categories that may not translate at all into the language of political ideology. To claim that a case involving contractual infringement on freedom of trade fits into a category of economic conservatism would be much more subjective and difficult to defend than an equivalent in the United States. There it would probably arise under the Inter-State Commerce clause where decades of Congressional debate make clear what political lines are drawn up. An analyst must, in the end, impose his own categories. I have attempted to stick fairly close to legally existing categories, because I wish to interpret the resulting dimensions of judicial ideology as far as possible in terms of legal ideologies rather than, as with Schubert, asserting ideologies that have nothing directly to do with what judges may think they are doing. Yet again it

judges agreeing on how to dispose of a case may in fact have very different views. From my point of view this difference is irrelevant here, although it might have been preferable to take the distinction so as to record degrees of agreement. Unfortunately the available data are too sparse for this. 21 Knuller v. DPP (1973) A.C.435.

7

This content downloaded from 194.29.185.42 on Fri, 9 May 2014 10:06:58 AM All use subject to JSTOR Terms and Conditions Unanimous and Split Cases Heard by Members of the House of Lords Appellate Committee, I948-76 Unanimous and Split Cases Heard by Members of the House of Lords Appellate Committee, I948-76 Unanimous and Split Cases Heard by Members of the House of Lords Appellate Committee, I948-76

Dates of No. of cases heard by this Name membership judge in the data set Unanimous Split

For both 'courts' Reid I948-75 Io8 56 Morris 1960-75 102 44 Diplock I968-77 92 37 Dilhorne* 1969-77 I 2 31 Wilberforce 1964-77 92 50

Plus, for 'early court' Hodson I960-71 60 28 Guest 1961-71 51 32 Pearce I962-69 38 20 Donovan 1963-71 24 14 Upjohn I963-71 42 25 Pearson 1965-73 62 26

Plus, for 'late court' Simon 1971-77 50 29 Kilbrandon I971-77 40 20 Salmon 1972-77 80 20 Edmund-Davies 1974-77 39 27 Russell 1974-77 43 I6 Cross 1971-76 48 17