1 The Context
Like the Roman god Janus who is portrayed with two faces one looking to the past and the other to the future, the doctrinal methodology has strong roots in the past, but it is now transitioning towards an electronic globalised future. This discussion concentrates on the future of legal scholarship and the evolving taxonomy for incorporation of insights from other disciplines, particularly the social sciences, into reform-oriented legal research.
Even when a non-legal response might be just as appropriate to resolve a broader social problem, doctrinal researchers have tended to continue to work within the parameters of the discipline in order to make recommendations for reform. They have confined their research to a critical analysis and synthesis of the law. However, realising that the scope of the doctrinal method is too constricting, academic lawyers are becoming eclectic in their use of research method. Legal scholars may not often utilise non-doctrinal methods themselves, but they do include the results of the use of these methods in their research. In this transitional time, legal academics are increasingly infusing evidence (and methods) from other disciplines into their reasoning to bolster their reform recommendations. Current studies suggest that this is not occurring to the same extent within the law reform commissions.x K. Tranter, ‘Citation Patterns within the Australian Law Reform Commission Final Reports 1992-2012’, 38(1) University of New South Wales Law Review 318 (2015).
This article considers three examples of research and writings by lawyers which are directed to law reform. This analysis examines the extent of the interplay between doctrinal analysis and research from non-doctrinal research methodologies within these sets of materials. Firstly the article reviews studies on the extent of methodologies and reformist frameworks in PhD research in Australia. Secondly it analyses a ‘snapshot’ of recently published Australian journal articles on criminal law reform. Thirdly it focuses on the law reform commissions, those independent government committees that play such an important role in law reform in common law jurisdictions.
This examination demonstrates that while the doctrinal core of legal scholarship survives intact, legal scholars are, to some extent, endeavouring to accommodate statistics, comparative perspectives, social science evidence and methods, and theoretical analysis, within the legal research framework, in order to provide additional ballast to the recommendations for reform.
2 Clarifying the Basics: What Is Doctrinal Research?
Historically, doctrinal analysis has been the dominant legal method in the common law world, although other categories of research such as reform oriented, theoretical, and fundamental have been acknowledged as important and to this extent doctrinal research has always included an interdisciplinary aspect. Nevertheless, legal academic success has been measured within a doctrinal methodology framework, which includes the tracing of legal precedent and legislative interpretation. The essential features of doctrinal scholarship involve ‘a critical conceptual analysis of all relevant legislation and case law to reveal a statement of the law relevant to the matter under investigation’.x T. Hutchinson, ‘Valé Bunny Watson? Law Librarians, Law Libraries and Legal Research in the Post-Internet Era’, 106(4) Law Library Journal 579, at 584 (2014). There is general consensus on this type of broad description. This ‘conceptual analysis critique’ is based on an understanding of the rules of precedent between the court jurisdictions, the rules of statutory interpretation, the tacit discipline knowledge such as the difference between civil and criminal jurisdictions, and various tests of liability, along with the acknowledged reasoning methods, borrowed from philosophy and logic, such as induction and deduction.
How does the doctrinal method relate to law’s discipline paradigm? Thomas Kuhn viewed paradigms as a shared frame of reference among researchers, which could be upset by new revelations leading to generational struggles between newer and more established researchers.x T. Kuhn, The Structure of Scientific Revolutions (1996). Thus, paradigms are shared worldviews within a discipline, which determine what topics are ‘suitable’ to study, what methodologies are acceptable, and what criteria may be used to judge success. Other descriptions of paradigms include ‘taken-for-granted mind sets’, and according to this view, socialisation into the discipline is instrumental in ensuring that newcomers take on these ‘ways of knowing’.x J. Jones, ‘Undergraduate Students and Research’, in O. Zuber-Skerritt (ed.), Starting Research — Supervision and Training (1992), at 54. A discipline paradigm encompasses any underlying philosophies, which again, in the common law world, has been predominantly liberalism, with its ideas of rationalism, the importance of personal property and individual self-determination. There are other aspects to the paradigm – the once-prevalent view of law as being objective and neutral, and positivism, with its view of law as being ‘what is’ rather than what ‘could be’ or ‘should be’ also form part of the paradigm. These characteristics are particularly ubiquitous in the British common law legal tradition. The established paradigm within research in the discipline of law has involved the individual scholar’s legal voice.
So doctrinal research was the predominant category identified in all the discipline assessments for law that took place in the 1980s.x H. Arthurs, Law and Learning: Report to the Social Sciences and the Humanities Research Council of Canada by the Consultative Group on Research and Education in Law (1983), at 66; D. Pearce, E. Campbell & D. Harding, Australian Law Schools: A Discipline Assessment for the Commonwealth Tertiary Education Commission (1987). In 1987, the Australian Pearce Committee highlighted doctrinal as the main category in its research taxonomy, describing it as research which ‘provides a systematic exposition of the rules governing a particular legal category, analyses the relationship between rules, explains areas of difficulty and, perhaps, predicts future developments’.x Pearce, Campbell & Harding, above n. 5, 2, 312 [9.17]. The Council of Australian Law Deans subsequently expanded on this earlier definition – ‘Doctrinal research, at its best, involves rigorous analysis and creative synthesis, the making of connections between seemingly disparate doctrinal strands, and the challenge of extracting general principles from an inchoate mass of primary materials’.x Council of Australian Law Deans, Statement on the Nature of Legal Research (2005), at 3. In 2006, Martha Minow, Dean of Harvard Law School, identifies ‘doctrinal restatement’ as one of the main contributions legal scholars make within their research.x M. Minow, ‘Archetypal Legal Scholarship – A Field Guide’, 63(1) Journal of Legal Education 65-69, at 65 (2013). Susan Bartie identifies ‘doctrinalism’ as a ‘unifying element in legal scholarship in England and Australia’.x S. Bartie, ‘The Lingering Core of Legal Scholarship’, 30(3) Legal Studies 345, at 350 (2010). Writing from a European perspective in 2011, Rob van Gestel and H.-W. Micklitz, describe the process in similar terms stating that in doctrinal work, ‘arguments are derived from authoritative sources, such as existing rules, principles, precedents, and scholarly publications’.x R. Van Gestel and H-.W. Micklitz, ‘Revitalizing Doctrinal Legal Research in Europe: What About Methodology?’, European University Institute Working Papers Law (2011)/05, at 26. Accordingly, they continue, the law ‘somehow represents a system’ so that ‘through the production of general and defeasible theories, legal doctrine aims to present the law as a coherent net of principles, rules, meta-rules and exceptions, at different levels of abstraction’, and ‘decisions in individual cases are supposed to exceed arbitrariness because they have to fit into the system’ so that the system remains coherent.xIbid. Therefore, there is widespread agreement on the basic tenets of doctrinal research.
The doctrinal method has been widely criticised, largely because it has never been explicated sufficiently for non-lawyers – or for lawyers themselves!x W. Twining, Taylor Lectures 1975 Academic Law and Legal Development (1976) (Lagos: University of Lagos Faculty of Law). Legal researchers have not been in the practice of describing their methodologies even within their academic work. In the past, few PhD theses have provided a separate description detailing the extent of the method. The method is assumed knowledge within the discipline – part of the grab-bag of skills associated with ‘thinking like a lawyer’. The doctrinal method is qualitative and idiosyncratic and, especially in the courts and in practice, the outcomes are often limited to the specific facts of the case. As a way of combating criticism from the physical sciences, Christopher Langdell, in the early part of the nineteenth century, had tried to promote law as a ‘legal science’, and the law library as a ‘lawyer’s laboratory’. In the Preface to Contracts, he commented:x B. Kimball, The Inception of Modern Professional Education: C.C. Langdell, 1826-1906 (2009), at 349, app., 2.
Law, considered as a science, consists of certain principles or doctrines. … Each of these doctrines has arrived at its present state by slow degrees; in other words, it is a growth, extending in many cases through centuries. This growth is to be traced in the main through a series of cases. … Moreover, the number of fundamental legal doctrines is much less than is commonly supposed. … If these doctrines could be so classified and arranged that each should be found in its proper place, and nowhere else, they would cease to be formidable from their number. … It seemed to me, therefore, to be possible … to select, classify, and arrange all the cases which had contributed in any important degree to the growth, development, or establishment of any of its essential doctrines.x C.C. Langdell, A Selection of Cases on the Law of Contracts (1871) from Kimball, above n. 13 (emphasis added).
A few years later, in the Harvard Law School Annual Report, Langdell again noted:
‘The work done in the Library is what the scientific men call original investigation. The Library is to us what a laboratory is to the chemist or the physicist, and what a museum is to the naturalist’.x C.C. Langdell, Annual Report 1873-74 from Kimball, above n. 13, at 67, 349, app., 2. In this respect Langdell was suggesting that the law ‘ought to be studied from its own concrete phenomena, from law cases, in the same way that the laws of the physical sciences are derived from physical phenomena and experiments’.x Kimball, above n. 13, at 351, app., 2, n. 10; J. Redlich, The Common Law and the Case Method in American University Law Schools (1914), at 15.
Historically, the doctrinal process has been described within a problem framework with a number of linear steps including assembling the facts, identifying the legal issues, analysing the issues with a view to searching for the law, undertaking background reading and then locating primary material, synthesising all the issues in context, and coming to a tentative conclusion.x T. Hutchinson, Researching and Writing in Law (2010) 41, at 42. There is certainly a need for a more sophisticated approach to tease out the doctrinal method. Whether the doctrinal method can ever be stated in a formulaic way is problematic. At its heart it is fluid. It is difficult to reduce to an algorithm.
2.1 Additional Categories of Legal Research
Doctrinal research was not the only type of research categorised within the early discipline reviews. The reports categorised other methodologies such as law reform research, legal theory research, and fundamental research separately.x Pearce, Campbell & Harding, above n. 5, 2, at 310 [9.12]. The Pearce Committee acknowledged ‘reform-oriented’ research, research which ‘intensively evaluates the adequacy of existing rules and which recommends changes to any rules found wanting’, as a separate category.xIbid., 3, app. 3, at 17 . Arguably what was being delineated here was also a form of doctrinal research. Pure doctrinal research identifies and analyses the current law. Reform-oriented research recommends change. Most ‘good’ quality doctrinal research goes well beyond description, analysis, and critique, and invariably suggests ways the law could be amended or the philosophy, processes or administration of the law could be improved. In many common law jurisdictions, there are separate organisations working to develop a reform agenda, these being the law reform commissions. These are discussed at more length later in this article. The reform-oriented research taking place in the commissions was primarily doctrinal, but depending on resources, had a consultative aspect and the potential to be ‘interdisciplinary’ in its methods.x D. Weisbrot, ‘The Future for Institutional Law Reform’, in B. Opeskin and D. Weisbrot (eds.), The Promise of Law Reform (2005), at 31.
The third type of research identified in the Pearce Committee Report was theoretical research – ‘research which fosters a more complete understanding of the conceptual bases of legal principles and of the combined effects of a range of rules and procedures that touch on a particular area of activity’.x Pearce, Campbell & Harding, above n. 5, 3, app. 3, at 17 . Legal theory is a crucial tool to provide a critical perspective on the law. However, in the past, the utility of theoretical research may have been diminished because of the limited exposure of the profession to theory (and the language of theory) and also because of the seeming gap between legal theory and practice.x Arthurs, above n. 5, at 68. Lawyers, even academic lawyers, have been so steeped in positivism that they have not sufficiently fostered knowledge of legal theory and the skills of critique and applied this to the law.
Certainly this is not the case currently. Research activity at postgraduate level always includes a conceptual framework, a component of which is the theory underlying the law itself, and the philosophy that best encapsulates the researcher’s view of the law. In the post-modern world, legal researchers understand that nothing is objective. Even the choice of topic for examination depends on the researcher’s world view. Very few doctrinal researchers would not acknowledge that fact in the twenty-first century.
The Canadian Arthurs Report identified a further category – ‘Research designed to secure a deeper understanding of law as a social phenomenon, including research on the historical, philosophical, linguistic, economic, social or political implications of law’, or fundamental research.xIbid., at 66. This type of research treats law as a phenomenon, as a problem with cause and effect.xIbid., at 69. Fundamental research uses social science methodologies to examine the law through the prism of another discipline’s. view – the economist or linguist or criminologist. Can fundamental research include a doctrinal component? On the basis that all research on law necessarily acknowledges the law as its basis then the answer to that must be a resounding ‘yes’. Interdisciplinary legal articles, even those being written by non-lawyers from an ‘outsider’ view, frequently acknowledge the ‘black letter’ or ‘doctrinal core’ of law as the starting point, whatever methodology is being used to pursue the author’s agenda.x Bartie, above n. 9. Once again there must be an acknowledgement that the boundaries between the various categories are not closed in the present century. The methodology denotes the difference.
There are at least two other important categories of legal research which were not categorised separately in the reports in the 1980s. These are policy research and comparative research. Public policy research normally takes place within government departments rather than in academia but it too has a doctrinal component. There is certainly a cross-over with the methods used by the institutional law reform commissions because the process includes public consultation, discussion papers, public submissions, surveys, public meetings, and written reports. The difference lies in the degree of political interference in public policy outcomes. The policy inquiries are funded and driven by politicians, whereas law reform terms of reference emanate from government, and the research is conducted independently of departmental interference.
Martha Minow identifies ‘Comparative and Historical Inquiries’ as another typology or ‘intellectual contribution’ of legal scholarship which ‘Describe an earlier era or contrasting legal regime; Contextualize the selected era or regime utilizing social sciences such as anthropology or history; and Illuminate differences, choices, or continuities when compared with contemporary domestic practice’.x Minow, above n. 8, at 68. Despite not being placed in a separate category, comparative research was acknowledged in the earlier taxonomies which included statements about the need for lawyers to ‘keep up’ with the ‘legal and other relevant literature of all common law jurisdictions including England, New Zealand, Canada and the United States’.x Pearce, Campbell & Harding, above n. 5, 3, app. 3 at 17 .
From this discussion it is evident that there is a need for a new interdisciplinary taxonomy that recognises the interplay of the changing methods and purposes within the legal discipline paradigm. Kuhn suggests that paradigms can and do change and there is no doubt that this is occurring within the discipline of law. The examples examined in Section 3 of this article demonstrate that the paradigm of the sole researchers working at their computer and involved in qualitative doctrinal scholarship remains. Even so, although the scholars do not always apply the non-doctrinal methods themselves, there is an increasing application of the research results from the use of such methods by legal scholars. The door is definitely ajar to further change though the link between doctrinal and non-doctrinal needs to be better articulated and explored.
3 Examining Methodologies Informing Recommendations
If we accept the methodological ground rules for doctrinal work, with the simplistic problem based structure as a naïve framework, and also recognise that theory, fundamental research, comparative research, and law reform have always played a role within the discipline’s research landscape but often as separate genres, the next issue is to examine how non-doctrinal methodologies are being infused into legal research in the twenty-first century. Is this fundamentally changing the doctrinal method? Are lawyers using non-doctrinal methodologies and data as a matter of course? At what point in the analysis is this data being infused into the discussion and how is this evidence being synthesised with the law in coming to a conclusion or recommendation for reform? What is current practice and how valid are the outcomes or any recommendations for change based on the studies? Is there adequate internal cohesion in the analysis reinforcing the recommendations? These are difficult questions and this article can only hope to provide basic assessments of trends using examples of previous studies on PhD theses, a snapshot of recent articles written by lawyers for Australian law journals and the studies of outputs of the law reform commissions.
3.1 Interdisciplinarity Evidenced in PhD Theses and Law Journal Articles
There are few empirical studies examining the methodologies employed in legal scholarship.x While statistical studies on methodologies being used in legal research are rare, the topic of how lawyers research is not a new area of legal academic concern. See, for example, E. Jones, ‘Some Current Trends in Legal Research’, 15(2) Journal of Legal Education 121 (1962-1963). Any recent studies of the use of legal research methodologies focus on the postgraduate research arena. A survey of postgraduate research in Australian law schools undertaken in 2002 demonstrated that only 20% of all doctoral research projects could be described as purely ‘doctrinal’.x D. Manderson and R. Mohr, ‘From Oxymoron to Intersection: An Epidemiology of Legal Research’, 6 Law Text Culture 159, at 164 (2003); and see D. Manderson, ‘Law: The Search for Community’, in S. Marginson (ed.), Investing in Social Capital (2002), at 152 on breakdown of empirical and doctrinal PhDs in Australia. A more recent examination of HDR theses submitted to the Australasian Digital Thesis Program website in the 5-year period 2004–2009 reveals that most of the legal theses include a doctrinal component, even though only a few students overtly identified the study they were conducting as being to any extent ‘doctrinal’.x T. Hutchinson and N. Duncan, ‘Defining and Describing What We Do: Doctrinal Legal Research’, 17(1) Deakin Law Review 83, at 99 (2012); The study of the Australasian Digital Thesis Program website was undertaken by Felicity Deane and Terry Hutchinson and completed in October 2010. According to this study, 16 of the 60 theses examined include a chapter to describe the use of non-doctrinal methodologies, 21 theses discuss methodologies as part of another chapter, and one deals with the methodology in an appendix. Any overt description of method in the thesis invariably signalled an interdisciplinary perspective, so the results demonstrate a higher proportion of doctrinal papers than occurred in the 2002 survey with 37% (n = 22) in this group being purely doctrinal. This demonstrates that law is still essentially a scholarly endeavour.
There are differences between the research process and methods undertaken in a PhD program compared to a great deal of legal academic scholarship. PhD requirements influence the scope and the depth of analysis of any study as well as the choice of research methodology. The PhD students will usually attempt a triangulation of methods and may well base their conclusions on several different collections of research data. The categories of methods the PhD students choose are constrained by the supervisors’ levels of expertise and the student’s training. The non-doctrinal methodologies are invariably fully reported in the thesis, but the extent of doctrinal analysis is rarely acknowledged, described, or unpacked.
Following completion, the PhD candidates may either translate their work into a monograph or publish several separate journal articles dealing with sections of the thesis. This entails segmenting the work. Broadly speaking, different topics will be packaged for the various audiences to best disseminate the research amongst the assorted discipline audiences. The doctrinal legal analysis will be published in a university law review or topic specific law journal targeting academic and practising subject specialists. As a rule, lawyers do not ‘like’ detailed statistics because many have not been fully trained in statistical method, so the law journal article will not include extensive discussion of empirical work undertaken in the thesis and certainly not mathematical formulas. The theoretical framework of the project will be emphasised for a law and society or sociology journal. More practical policy and reform aspects combining the law and social science studies will be directed to subject-specific ‘current issues in the law’ titles. Studies including details of the empirical analysis are diverted to the criminology and social science journals.
Undoubtedly law academics use a similar approach for publishing outputs from their larger funded and team-based projects. Different aspects of the broader studies are highlighted according to the publishing profile of the target journals. For this reason it is difficult to validly determine the total extent of methodologies being implemented by legal scholars simply by examining law journal articles.
Despite the existence of these limitations on assessing legal research activity, an examination was conducted of a set of journal articles on the topic of ‘reform of the criminal law’ which were published in 2013. This ‘snapshot’ of publications encompassed articles published in Australian law journals – but only those held on the AustLII database.x Australasian Legal Information Institute <www.austlii.edu.au/>. A basic search string resulted in a retrieved list displaying 60 items. When book reviews, speeches, and government publications were omitted, then only thirty-two refereed journal articles remained as a relevant subset of the database.x These were located on the AustLII database using the search term ‘crim* w/10 (reform* OR recommend*)’ within the Australian journal titles published in 2013. The search was conducted on 16 February 2015. There are approximately 100 Australian journal titles on this database including most of the university law reviews. This subset included articles which were written by lawyers, included doctrinal analysis, and specifically dealt with ‘criminal law and reform’. This entire group of articles had been published in refereed (blind peer reviewed) journals. The examination of the articles in the retrieved list focused on the author’s profession (all had legal qualifications), whether the articles included a doctrinal analysis of legislation or case law, the extent of the description of additional methods, and the point at which it was introduced into the discussion. Were statistics included in the analysis and if so, where were these sourced? Where a comparison was included, the study considered whether this was contextual or a full comparison and whether there was also reference to public international law. Did the author mention law reform commission recommendations?
All the articles used a doctrinal research methodology to some extent. In two articles there was more emphasis on theory, criminology, and international law rather than an analysis of specific case law or legislation. The doctrinal methodology design was tacit; not so any non-doctrinal methods. Surveys, for example, were outlined and explained using appropriate tables. While only two of the articles were reporting that the authors had themselves undertaken surveys, interviews, or statistical projects,x M. Brown, G. Lansdell, B. Saunders & A. Eriksson, ‘‘I’m Sorry But You’re Just Not That Special …’ Reflecting on the ‘Special Circumstances’ Provisions of the Infringement Act 2006 (Vic)’, 24(3) Current Issues in Criminal Justice 375 (2013) (semi-structured interviewing and court observations and quantitative, descriptive data extracted from CLC databases); H. Douglas and R. Fitzgerald, ‘Legal Processes and Gendered Violence: Cross-Applications for Domestic Violence Protection Orders’, 36(1) University of New South Wales Law Journal 56 (2013) (examination of Magistrates Court files). eighteen used statistics published elsewhere in discussing the basis of concerns. This provided foundation for the discussion of the prevalence of offences and involved the use of news article reports, law reform submissions, and social science studies (predominantly reports of surveys) from medical and other interdisciplinary journals. To this extent the statistics provided context for the legal discussion.
All the articles analysed pertinent secondary literature with only two including an explicit acknowledgement of the literature review. The review of the literature is an implicit quality indicator in the doctrinal methodology paradigm. The extensive footnoting used as the preferred citation style for this group of articles provides an updated record of the secondary literature on the topic. The scope and currency of these references provide a strong warrant for the author’s credentials and knowledge of the subject area. Relevant texts, journal articles, and law reform publications are referenced where applicable to the discussion rather than brought together under a formal literature review heading.
Fourteen of the thirty-two articles were jointly authored, signalling a definite movement away from the lone scholar paradigm. Six of the articles disclosed their funding sources as being either from external publicly funded grants or university internal grants, and five of these were jointly authored. At least 6 of the 14 jointly authored articles emanate from subject-specific university or faculty research centres and working groups. Those articles that were jointly authored were likely to include interdisciplinary approaches, such as criminology and law, or emanate from the research centres.
The actual number of discrete comparative analyses in this retrieved group was low. Only two of the articles had as their main objective a comparison of the law between jurisdictions. However, twenty-three of the articles include a comparative review of the existing law particularly for the Australian state jurisdictions as well as pertinent examples from international jurisdictions. The Arthurs Report had recognised that comparative and historical research involving legal rules was likely to lead to ‘new’ solutions.x Arthurs, above n. 5, at 68. However, their comment in 1983 was that ‘experience, and our own investigation, shows that historical or comparative research is not undertaken routinely even by scholarly investigators, and rarely carried out by practitioners. Perhaps we stand on the threshold of change in this regard …’.xIbid. The change has occurred. It is widely recognised that comparative research approaches are becoming the norm within the current doctrinal method, and this small snapshot of articles reinforces this perspective. There are discrete comparisons of legal provisions in two or three jurisdictions, comparisons of the legislation between numbers of jurisdictions in order to provide context, and at the very least the use of comparative data and information on the law in other jurisdictions using secondary literature. Law is less parochial in the twenty-first century. Globalisation and technology mean that the wider legal sphere is more accessible and pertinent for the legal scholar. This in itself is intriguing and needs more examination to test and confirm this practice using a larger body of evidence. Certainly reports published by law reform commissions have consistently included a comparative approach. The current expectation in the literature is that there will be some statement of the legal jurisdictional status quo or an acknowledgement of obvious discrepancies in practice elsewhere. This forms part of the context of the doctrinal discussion.
Many of the articles (twenty-three) include recommendations for reforming the law informed by the evidence presented. Suggestions for reform included calls for social reform to engender change, improved data collection and review mechanisms, or specific non-legislative action. Five of the articles simply critiqued the existing laws.
Some of the limitations of this pilot study must be acknowledged openly. It might be expected that more non-doctrinal methods would be found in studies of criminal law and law reform. Studies of reform in other areas of law such as tort law or corporate law or equity might well be more legalistic in approach. There is space for more extensive studies of the published literature to gauge such differences. Overall, within this pilot study of thirty-two doctrinal articles on reform of the criminal law written by lawyers, there was evidence of an increasing reference to comparative law and to published statistics and social science evidence to contextualise the law and to reinforce the doctrinal analysis and conclusions. Studies utilising empirical methods or with a mainly theoretical focus were less common.
3.2 Interdisciplinarity and the Work of the Law Reform Commissions
The third study centres on the reports emanating from the law reform commissions. Before examining the research within the commissions, it is necessary to understand a little more about how the commissions function. Law reform bodies have an established role in common law history. The commissions are independent government agencies charged with the task of reforming the law. Evidence exists of law reform commissions in Scotland in 1425 and various ad hoc committees set up to reform and rationalise the law over the centuries in England,x W. Hurlburt, Law Reform Commissions in the United Kingdom, Australia and Canada (1986), at Chapter 2 Law Reform Commissions in the United Kingdom 15-99 generally. with law reform committees being formed in Australia from the 1870s.xIbid., at 100-68: Chapter 3 Law Reform Commissions in Australia; See also Australian Law Reform Commission, Annual Report 1975 (1975), at 5-19. The permanent English Law Commission was established in 1965, and statutory law reform commissions on a similar model are now established in most common law jurisdictions.x There are Law Reform Commissions in most of the Australian jurisdictions. Canada too has multiple reform commissions or institutes, for example, The Law Reform Commission of Nova Scotia and the non-statutory Alberta Law Reform Institute. In the United Kingdom, the Law Commission and the Scottish Law Commission are statutory commissions established by the Law Commissions Act 1965 (UK); The United States has multiple committees or commissions, for example, The California Law Revision Commission and the Michigan Law Revision Committee. The Australian Law Reform Commission (ALRC) was established as an independent statutory body in 1975. Under s21 of the Australian Law Reform Commission Act 1996 (Cth) the Commission’s brief is to systematically develop and reform the law by:
bringing the law into line with current conditions and ensuring that it meets current needs; and
removing defects in the law; and
simplifying the law; and
adopting new or more effective methods for administering the law and dispensing justice; and
providing improved access to justice.xAustralian Law Reform Commission Act 1996 (Cth), and see the Public Governance, Performance and Accountability Act 2013 (Cth).
The Australian state law reform commissions are, with a few exceptions, modelled on the national body, though they invariably have fewer resources.x Commonwealth of Australia, The Senate, Legal and Constitutional Affairs References Committee: Inquiry into the Australian Law Reform Commission (2011), at 9. The terms of reference for law reform enquiries are normally set by the Attorney General in consultation with the Commission. Consultation with the general public and stakeholders is always a key element of the inquiry process. Figure 1 demonstrates the typical law reform process.
The Law Reform Process <www.alrc.gov.au/law-reform-process> (last visited 21 October 2015). The Australian Law Reform, ‘Law Reform Process’ under a Creative Commons License 3.0: <http://creativecommons.org/licenses/by-nc-nd/3.0/>.
There have been trends in the popularity of law reform commissions resulting in the periodic closure and rebirth of agencies in common law jurisdictions depending on government finances and reform agendas.x N. Rees, ‘The Birth and Rebirth of Law Reform Agencies’, Australasian Law Reform Agencies Conference 2008 Vanuatu 10-12 September (2008). The ALRC for example has been reviewed several times since 1975.x 1977-1979 Senate Standing Committee on Legal and Constitutional Affairs inquiry; 1993-1994 House of Representatives Standing Committee on Legal and Constitutional Affairs; 1997-1998 Possible Improper Interference with a Potential Witness before the Parliamentary Joint Committee on Native Title and the Aboriginal and Torres Strait Islander Land Fund (73rd Report); 2010-2011 Senate Legal and Constitutional Affairs Committee Inquiry into the Australian Law Reform Commission. Possibly because of this high level of scrutiny, the Commission’s Annual Reports contain very detailed information on performance and outcomes.x Australian Law Reform Commission, Report 125: Annual Report 2013-13 (2014), at 26. The levels of implementation of all ALRC reports are reasonably high:
60% are substantially implemented;
28% are partially implemented;
2% are under consideration;
3% are awaiting response; and
7% have not been implemented.xIbid., at 27.
The law reform commissions are touted as one of the main conduits for law reform.x Commonwealth of Australia, above n. 40, at 9 [2.18]. However, there are many other channels available for advocating legal and regulatory change. These include ‘parliamentary committees and ad hoc commissions of inquiry’,x L. Barnett, ‘The Process of Law Reform: Conditions for Success’, 39 Federal Law Review 161 (2011). as well as periodic reports from internal government policy units, reports from government instrumentalities such as the sentencing councils, the children’s commissions, the human rights commissions, and even the courts’ annual reports. Tranter has identified two possible approaches to law reform – one being the ‘research institute’ approach where ‘recommendations are generated by experts analysing relevant data and academic literature’ and the other evidencing a ‘community engagement’ approach where ‘recommendations are located as having emerged from a process of community consultation’.x Tranter, above n. 1; Barnett, above n. 46. By and large the law reform commissions fall within the latter ‘community engagement’ category. Many of these other bodies would be categorised as ‘research institutes’.
Although comparative legal perspectives and references to published statistics feature in the reports and publications of the formally constituted law reform commissions, the principal method employed by the commissions is public consultation. Justice Kirby, the founding chair, commented that the ALRC ushered in a new era for law reform in Australia with the ‘value adding’ involved in public consultation.x Justice Kirby has concluded that ‘the most original “value added” of the ALRC – and its chief contribution to the law reform technique in the years after its establishment – was its emphasis on public consultation’. M. Kirby, ‘Are We There Yet?’, in B. Opeskin and D. Weisbrot (eds.), The Promise of Law Reform (2005), at 435. The current ALRC Chair, Rosalind Croucher, also champions the consultation process:
Through its widespread and thorough consultation strategies, the ALRC is able to build consensus and understanding of its proposals within the community and this assists the government in turn to implement various recommendations, even in a context where change may be challenging.x Australian Law Reform Commission, above n. 43, at 5.
Public consultation constitutes a non-doctrinal method and as such is interdisciplinary in its approach. However, consultation was probably not what the former Commissioner of the Law Commission of Canada, Roderick Macdonald, had in mind when in the mid-90s he was arguing for a ‘reimagining’ of law reform processes, including ‘co-opting non-academic and academic’, ‘non-legal and legal’, with the notion of reform being to ‘transcend doctrine’.x R. Macdonald, ‘Recommissioning Law Reform’, 35 Alberta Law Review 831, at 870 (1996-1997). Macdonald advocated that law reform should be focusing on alternative reform processes (and outcomes) rather than simply presenting a report together with redrafted legislation as a standard response, and that there should be a different choice of projects moving away from substantive law topics to broader areas of social justice and ‘the relation of law and society’.xIbid., at 875. W.H. Hurlburt challenged all of these ideas in a spirited reply in the same issue of the journal.x W.H. Hurlburt, ‘The Origins and Nature of Law Reform Commissions in the Canadian Provinces: A Reply to “Recommissioning Law Reform” by Professor R.A. Macdonald’, 35 Alberta Law Review 880 (1996-1997). However, these methodological constraints were perceived as a failing in the Canadian Arthurs Report which held that ‘the basic problem with much law reform research is rather that it is located toward the doctrinal end of the methodological spectrum, and consequently fails to confront most problematic issues’.x Arthurs, above n. 5, at 70. Greycar and Morgan perceived that the law reform bodies were using ‘limited consultation processes that often leave out of account the concerns of those most affected’ and disadvantaged groups.x R. Greycar and J. Morgan, ‘Law Reform: What’s in It for Women?’, 23 Windsor Yearbook on Access to Justice 393 (2005). Greycar and Morgan provide examples of the lack of empirical methods and data in the formation of law reform recommendations in favour of so-called “common sense” anecdotal information’.xIbid. More recently Angela Melville’s comparison between New Zealand and Canadian law reform commissions pointed out the difference in approaches.x A. Melville, ‘Conducting Law Reform Research: A Comparative Perspective’, 28(2) Zeitschrift für Rechtssoziologie 153 (2007), at 153. Melville noted the methodological insularity, ‘top down’ approach and even paternalistic attitude exhibited by some commissions who confined the stakeholder list and limited the questions addressed to them.xIbid., at 158, 159. Other commissions were more interested in broader questions – presenting papers in open academic conferences prior to the reports being published so as to engage with stakeholders, and commissioning private empirical studies prior to writing the reports.
However, this article is examining the extent of interdisciplinarity and the use of non-doctrinal methodologies by lawyers to reform the law. The question, therefore, must be whether the evidence, gathered from the consultations and submissions sent to the law commissions in their enquiries, is being infused into the recommendations? Are the commissions using the submissions in drafting their recommendations? Kieran Tranter’s study into the citation practices within the ALRC final reports from 1992 to 2012 found that ‘submissions were the most frequently cited source’ (46%), supporting an argument that ‘the best way to influence the executive is to locate recommendations within what can loosely be called the ‘community’.x Tranter, above n. 1. Community participation not only provides ‘responses and feedback’, it also promotes ‘a sense of public ‘ownership’ over the process of law reform’.x R. Atkinson, ‘Law Reform and Community Participation’, in B. Opeskin and D. Weisbrot (eds.), The Promise of Law Reform (2005), at 160. However, whether there is a statistical correlation between the number of those respondents supporting a course of action and the final recommendation is more difficult to determine. The submissions are often divided as to their arguments and proposals for the most advantageous way forward. Not all responses are helpful in terms of the information or views they proffer, but the ability to refer back to those providing submissions can assist deliberations. The likelihood is that the recommendations are being based primarily on case law and the arguments provided by the judges and the weight of commission opinion favouring a particular line of action, rather than emanating from the views of those providing submissions. This issue requires further research.
To what extent do the law reform commissions have recourse to academic material on the issues they are studying? Many of the recommendations extend more broadly than simply reforming the legislation – but this very much depends on the scope of the terms of reference handed to the agency. While the reports are referencing legal reforms implemented (and sometimes evaluated favourably) in other jurisdictions, recent examination of the reports produced by the commission has demonstrated a paucity of reference to published academic commentary on the issues they are examining, and also little empirical data to back up the recommendations.x Tranter, above n. 1, at 349. So in Tranter’s study, ‘citations to secondary academic material in the form of books, journal articles and conference papers were quite low at only 6% of the total citations’.xIbid.
There may be clear explanations for this. The reports have narrow terms of reference, and it may be the case that only a very few academic articles are ever directly on point. In addition, the articles and texts referenced in the final reports are in no way fully indicative of the background literature reviews, extensive annotated bibliographies, and working papers produced by the Commissions, and which are never made public; all that is published is the final reports. In the past, publishing costs may have limited the materials that could be released. However, this is certainly not the case at present when electronic files can be easily uploaded onto the websites. It could well be argued that when extensive background research has been undertaken at the public expense, then these preliminary working papers should be made available. Until this occurs, it should be queried whether the low level of citation of secondary material truly reflects the background research undertaken for the reports.
A more worrying aspect of the research is the apparent lack of reference in the reports to empirical studies. The literature reviews cover secondary legal literature. There is no evidence of the law reform commissions undertaking literature reviews of the wider social science and scientific evidence base on the issues being covered apart from clarification and definition of existing processes when required. The emphasis always lies on the primary materials – the legislation and case law. The secondary literature is only used to assist in interpreting the law. There is little or no attempt to deal with the wider context apart from what is provided by the submissions and consultations. As Barnett has commented, there is a real need in law reform to ‘uncover the facts upon which law reform proposals are based’, so that ‘they need to see the entire picture and identify the real problem(s) before launching into a search for policy solutions’.x Barnett, above n. 46, at 181. So reference is being made to the community stakeholders, and their views on the issues. The law is being considered closely. But very rarely is new data compiled from within the commissions. There are reasons for this. The use of additional interdisciplinary methods is dependent on government budgets. Even attempting to educate the public about an issue can be an expensive process. The commissions focus on the role of clarifying the law for other lawyers and the general public. As detailed previously, the law reform commissions have always performed this role well.
Further research is necessary to determine if there are regularly gaps between existing empirical data, the consultation responses, and the recommendations. However, the recommendations in a law reform report are not the final word. Any major recommendations for legislative change must be presented to cabinet and then parliament for debate. If there are gaps in the arguments presented, then those aspects should be addressed at that point.
It is clear from this discussion that academic lawyers are using non-doctrinal methods, but they are often infusing these methods within their doctrinal research framework. Definite inroads have been made in relation to the use of comparative approaches. Arguably comparative law, extending far beyond a simple backward glance to the genesis of a legal proposition, is now an intrinsic part of legal scholarship. Published government statistics and the published results of social science research are also being included in doctrinal academic writing to provide contextual framing or to highlight the disparity between the law, social policy, and the existing social evidence base. There are some examples of joint authorship and lawyers working in tandem with those from other disciplines to investigate all sides of the contextual prism in an effort to best achieve enlightened critique. Theoretical and philosophical discussions invariably include footnotes to the doctrinal stasis to provide factual legitimacy. Theory is also increasingly used as a vehicle for critiquing and analysing the basis of the ‘black letter’ law. Theory is part of the contextual framing. Researchers query – ‘What was the pre-eminent theory at the point in history when this law commenced? Are those theories and those economic and political views infused in the law still relevant and valid today?’ The doctrinal method remains true to its core, but it is certainly less constrained than in the past.
This article has used examples of existing studies on PhD students’ theses, a snapshot of recent articles written by lawyers for Australian law journals and the outputs of the Australian Law Reform Commission to provide some basic assessments of trends in the use of interdisciplinary and doctrinal methods especially focusing on reform agendas. These are Australian examples. Is there any great disparity between Australian legal scholarship and that being undertaken elsewhere? Further empirical study is required on this issue. More research needs to be carried out to determine at what point in the legal analysis the non-doctrinal data is being infused into the discussion and how exactly doctrinal lawyers are infusing this evidence in coming to a conclusion or making recommendations for reform. The discussion has by no means finished. There is evidence of a broadening of the method overall, but we need a more sophisticated study of larger amounts of data to verify the trends observed so far.
While empirical arguments are potential game changers in American law, they only slowly gain traction in Germany. In this article, I wish to show the role empirical methods (can) play in public law scholarship. Specifically, I will discuss an underexposed area in the field of empirical legal studies, namely the interplay between empirics and doctrine, and the power of such arguments in the public law discourse. In the second part of the article, I will exemplify the merits, and the caveats, of “empiricizing” public law in such way, in a specific case (gambling law), showing how it can be used to prove wrong basic assumptions that are at the heart of long-standing doctrine. I suggest that empirical scholarship geared towards public law doctrine (as opposed to public policy), as it is currently emerging in Germany, can be a potentially influential addition to the menu of legal arguments, even beyond the German legal discourse.
1. Prevailing arguments in public law
It is a well-established fact that German legal scholarship typically focuses on coherent meaning, and thus on the lex lata, while American legal scholarship tends to zero in on the good decision, thereby being drawn to the lex ferenda.1 This is even reflected in a divergent usage of the word “normative.” While German lawyers, in following the (somewhat descriptive) Latin undertone, simply use it to refer to prevailing legal norms, for Americans it evokes the notion of (somewhat prospective) value judgments, in line with the ubiquitous English meaning.2 The methods for the scholarly interest in the construction of a coherent, smooth-running system of legal norms are to be found in the hermeneutics toolbox,3 while the behavioral perspectives of economics and psychology are apt to study decisions and their impact. Attentiveness to doctrine guides the former, interest in policy the latter. This also explains why certain of the Grundlagenfächer (literally “basic disciplines”)—such as philosophy and history—feature more prominently and more permanently in the German legal discourse (and, for that matter, in law school curricula, perpetuating such disciplinary focus) than others—such as economics or sociology: Philosophy is kindred in method, and history can make for a powerful argument in the interpretation of norms. Moreover, in the German tradition, law is not a professional qualification, but a self-sufficient academic discipline; students, practitioners, and scholars rarely hold a degree in another field. Their exposure to and propensity for other disciplines is usually marginal.
From a public lawyer’s perspective, there is a certain irony in this allocation of arguments to legal systems, in that there is a mismatch between the scholarship produced and its reception by the highest courts.4 While US academia provides distinctive legal scholarship deeply informed by social sciences, the Supreme Court—being rather reserved vis-à-vis scientific evidence more generally since Daubert—holds such scholarship to be of limited value for its work, claiming it was too far away from practice. Moreover, public law and constitutional doctrine seems to be understood as more of a political exercise that should engage the people, rather than a legal exercise that lawyers should be entrusted with.5 It is also claimed that the “interpretative community” creating the disciplinary constraints, was partly destroyed particularly in constitutional law by legal realism, and that the Supreme Court lost interest in the scrutiny of its professional community.6 The German Federal Constitutional Court, on the contrary, has been open towards legal scholarship generally, and towards scholarship informed by social sciences more particularly. In fact, the court time and again talked about the necessity of such more interdisciplinary scholarship, while the German legal academy is rather reluctant to follow up on this lead.7 However, the legal scholarship produced can also be seen as a reaction to where it is well received. The US courts traditionally do not have a high propensity to adopt arguments from academia, so scholarship turns to the legislature to exert its influence there; thus its focus on policy. Moreover, the restrictions on the design of public policy seem fewer, and it allows for broader, more fundamental institutional changes: if you devise a system, you do not need to fit your argument into an existing order. In Germany, starting with legal education being oriented towards training judges, courts and academia are close neighbors: academics, in focusing on hermeneutics and doctrine, work in a similar fashion as do judges, and the arguments they produce are readily received by courts.8 Legal scholarship-induced developments in doctrine can have substantial impact even if the screws turned seem small; and it is easier to convince a “colleague” to sanction a change in the legal doctrine with the subset of legal arguments than to move scores of members of Parliament drawing on a broader range of reasons to adapt the law.
At any rate, it seems quite natural that, under such different circumstances on the two shores of the Atlantic Ocean, the legal scholarship that emerges is quite different. This, of course, impacts the type of arguments that are prevalent in the different legal discourses.
2. Empirical arguments and public law
Traditionally, insights from social sciences—and especially from economics—have first been absorbed in private law. Law and economics (to which empirical legal studies have strong ties) had its early great moments in contracts and torts, and in explaining and justifying common law concepts; public law regulation in proximity of private lawyers (antitrust, corporate finance) was soon to follow.9 As Cooter and Ulen so wittily put it: “Like the rabbit in Australia, economics found a vacant niche in the ‘intellectual ecology’ of the law and rapidly filled it.”10 But social science arguments since also gained traction in the public law discourse. Public law is, to a large extent, about regulating norm addressees’ behavior. Thus, knowledge about human behavior is key to proper regulation. Public lawyers need to be experts for behavioral regulation through law. This is true for theory—and accounts for the success of law and economics, because its concept of rational choice provides a theoretically conclusive model of human behavior, and of its derivations such as public choice theory which can provide important specific insights for public law. Primarily theoretic models of the social sciences (mostly economics) were introduced to expand the set of (policy) arguments in public law. The “law of democracy”11 and with it constitutional law12 and the political process13 were seen through this lens, as were environmental law14, administrative law,15 and public international law,16 to name just a few examples.
With theoretical models of the social sciences being used to make behaviorally sound public law arguments, the emergence of empirical arguments was just a matter of time. The predictions the theoretical models generate can be tested empirically, and with the law focusing on the decision of hard cases and real-life public policy have lead to behavioral extensions of the model, drawing on insights from cognitive and social psychology. At the same time, the social sciences saw a surge of (quantitative) empirical scholarship that ultimately led to extensions of the classical behavioral model of the homo oeconomicus. These extensions—summarized under the broad umbrella term of “behavioral law and economics”—also opened up space for new models of legal regulation, most commonly referred to as “nudges.”17
2.1. Empirical arguments for policy design
In the US, with doctrinal constructivism being a tool “that self-respecting professors [in elite American law schools] would not waste much time with,”18 empirical arguments focus on policy. The prevalent approach in empirical legal studies,19 emanating from the US, therefore employs empirical methods chiefly to identify (social, psychological) distortions that the law should act upon,20 prominently by “nudging”21 individuals; to use key findings from behavioral scholarship for institutional reform;22 to measure and evaluate the impact of legal norms, often with a view to alternatives;23 and to understand and adequately describe legal developments and law (or norm addressees, lawyers, judges24) as a social phenomenon.25 In fact, if you go through the schedule of the 2012 Conference on Empirical Legal Studies, the overwhelming majority of papers presented follow one of the aforementioned agendas. To synopsize, empirical legal studies today concentrates on how the law should be (given the effects current law or the lack of regulation has), and to a lesser extent on why it is the way it is. Empirical legal studies thus follow the general prevalence of policy arguments in the US legal discourse, mainly addressing policy arguments to the lawmaker.
2.2. Empirical arguments for doctrine
By contrast, Empirical legal studies are a relatively recent phenomenon in German legal scholarship.26 The reason why the legal academy here adopts empirical legal studies (and other social science methods) hesitantly is straightforward, given what we have seen regarding the focus of traditional German legal scholarship. It is a consequence of its emphasis on the coherence of the law, on doctrine, rather than on policy, while at the same time current empirical scholarship has a strong policy-orientation.27
To put it differently, policy arguments traditionally are not worth a whole lot in Germany (and, for that matter, in many Roman law oriented civil law jurisdictions).28 They are often suspected of arbitrariness and concealed subjective ideology—which is one of the reasons for the strong urge for empirical foundations of these arguments in the American discourse. The German discourse requires you to establish the relevance of your findings for mainstream legal arguments, i.e., primarily for doctrine. Doctrinal arguments are addressed to, and used by, the courts, administration, and attorneys.29 The question asked is: What is the law? Given this paramount question, why should the expertise of other disciplines matter? In the application of the law, for factual questions, we may require expert opinions. And while we can understand that an economist can devise a good regulatory cap-and-trade policy to reduce carbon dioxide emissions, or that psychological research may bring valuable insight on an appropriate consumer protection framework to the table, why should any of these have a say in the interpretation of the applicable law? What is more, would we not come close to falling victim to the naturalistic fallacy if we incorporated insights from social sciences, from the realm of reality, in the law, seated in the realm of ideas?30 Because of this understanding of the law, the framework for input from other disciplines is a matter of avid and serious academic attention, particularly in public law.31 A number of dissertations specifically discuss this question,32 and the German textbook on economic methods in law begins with a description of the place of social sciences in legal scholarship.33
The argument that doctrine is self-sufficient and cannot be open to empirical insight does not hold, for a number of reasons. Legal concepts are inherently ill-defined,34 which makes law, by definition, an exercise in hermeneutics (as every doctrinally working lawyer will agree). There is always an insurmountable ambiguity in the law. To disambiguate law, to choose a proper interpretation of the law, we have to employ reason guided by rigorous method, so that it is generally (i.e., inter-subjectively) acceptable. Of course, the boundaries of “acceptable” arguments and methods are set by the scholarly discourse in the epistemic community of lawyers. But if the produced arguments speak to the law, answer questions of legal scholarship, in a methodologically rigorous manner, then that discourse cannot shun an empirical argument. Moreover, the normative models used in law—unlike mathematical models—are not exclusively ideal constructs, they are, on the contrary, contingent on factual assumptions about reality. We accept this readily in the application of the law (e.g., when we hear expert witnesses about legal prerequisites), but even statutes are based on (often implicit) factual assumptions. This is especially true when the law has a clear regulatory goal, and its interpretation is supposed to help apply the law in such way that the regulatory goal is best achieved. In these cases, having an empirical understanding of human behavior can be of benefit for the impact of the law.
Against this backdrop, there appear to be three major ways for public law doctrine to receive and absorb injections of empirical knowledge. First, empirical insights can be used to (in)validate normative concepts based on (factual, often behavioral) assumptions. Second, when the hermeneutic exercise leaves us with an indeterminate concept, empirical knowledge may provide arguments to help offer a proper interpretation. Third, we may use empirical analyses of legal writing to actually determine what doctrine is. We will briefly go through these three modes of making use of empirical scholarship for public law doctrine.
(a) Empirically validating normative concepts
Empirics can be conceptualized as tests for the normative models employed by the law, much in the way the “behavioral” approach in economics scrutinizes economic models of behavior. An example may best illustrate this point, so let me briefly anticipate one aspect of the gambling study that will be presented in greater detail later on. Gambling is regulated (i.e., largely prohibited) by law because of certain dangers the legislature associates with this pastime. But what constitutes a gamble? Legal doctrine had to develop an interpretation of this concept. It came up with the distinction between “games of chance” and “games of skill” (that is common in US federal and state law as it is in German law and in many other jurisdictions), thus excluding skill games from the definition of gambling, effectively permitting such games. As gambling regulation aims at preventing certain kinds of “danger,” this distinction ultimately hinges on the factual assumption that games of skill are less dangerous (in this specific sense of danger) than games of chance. Now while defining “dangerous” includes a normative valuation, actually determining that something is dangerous (games of chance) or is not dangerous (games of skill) in the sense of the legal conceptualization is an empirical statement. But it is not an individual factual statement subsumed in a specific case in court; it is an abstract factual statement, generalized and “normatized” for a large number of cases: It is doctrine based on an empirical assumption. This assumption can be tested. The result of the test can suggest modifying doctrine.
(b) Determining indeterminate legal concepts
Moreover, there is ample space for empirics in doctrine when hermeneutic methods cannot sufficiently determine the meaning of the law.35 Of course, empirical insight will not be able to fully determine indeterminate concepts either, but on the “determinate⇔indeterminate” continuum it may help us move towards a better, more determinate understanding. It has been shown that instances of such situations can be observed when we engage in teleological interpretation, when balancing in constitutional law (and I would contend: not only in constitutional law), or when (judicially) concretizing margins of appreciation (for here the effects of the decision are of special importance).36 One may add resembling concepts such as comprehensive clauses, equitable discretion of courts or administration, and the principle of proportionality. You get the idea: whenever the law opens space for some sort of discretion or expediency, whenever it refers to knowledge it cannot itself provide, whenever hermeneutics cannot convincingly guide the law’s proper interpretation, a social science argument cognizant of the structure of the law can provide an intersubjectively acceptable judgment, i.e., give more sound reason than subjective ideology.
(c) Pinpointing doctrine
Finally, we may want to determine what actually is prevailing doctrine. Legal commentaries and textbooks may claim that a certain interpretation of the law is prevailing, or they may systematize court cases and come up with categories that explain the application of a norm in different situations. These assessments are usually based on the author’s experience and knowledge of the field, on a careful selection and reading of (the important) cases. In some cases, however, we may find that such approach is insufficient, for example if there is no established doctrine that has developed and solidified over an extended period of time. In such situations, an empirical evaluation of all court cases decided regarding a specific legal question can add insight and help clarify what at least the majority of courts and judges hold to be doctrine, as revealed by the decisions they hand down.37
2.3. Linking empirics and public law doctrine
How do we respond to the necessity of empirical knowledge in the aforementioned situations? Customarily, lawyers tend to make implicit or explicit assumptions about reality themselves, without recourse to scrutinized empirical methods.38 This is due to the fact that most lawyers do not have a formal statistical or econometrical training. Therefore, we contend ourselves with checking the “plausibility.” Sometimes we tend to refer to anecdotal evidence, failing to recognize, as the saying goes, that the plural of anecdote is not data. Such lowbrow clumsiness threatens the validity of our findings. It “sounds plausible” to assume that I have control in a game depending on my individual skill, while a mere gamble activates irrational behavioral patterns that can be “dangerous” in terms of addiction. But it also sounds plausible to assume that games of chance are not dangerous at all because, after all, I know it only depends on luck, a factor beyond my control. When you read about the empirical results of our study, you are likely to say: “I’ve known it all along, that’s obvious!”—an effect called “hindsight bias” in psychology and “common wisdom fallacy”39 in the legal literature, and brilliantly described in a recent book titled “Everything is obvious—once you know the answer.”40 But this does not discharge us from either importing from other disciplines or generating the empirical evidence that is needed to make our doctrine sound and solid. “Doctrine” is the expert knowledge lawyers generate. It may depend on empirical facts. To be able to integrate empirical knowledge into such legal expertise, forming a distinctively legal perspective,41 empirical evidence must satisfy three conditions: (1) the definition of the measured variables and their operationalization must follow the normative decision of the law;42 (2) the results have to be valid in the sense that they must be generalizable to the legal context. The definition and operationalization requirements affect the formulation of the research question; therefore (3) the specific research design, methods and statistics employed to generate the empirical results have to stand the tests of validity.
(a) Definition of legal concepts
First, we have to make sure that the adopted empirical results rely on the definition of the relevant concept as mandated by the law. This can be intricate, as oftentimes the social sciences and even the methods applied have implicit normative assumptions that need to be in line with the law’s normative assumptions. If “dangerous” in gambling law is meant to capture whether a game is addictive, an empirical study based on the risk of cardiac failure due to the stress induced by the game cannot be introduced into doctrine.43 Austrian legal scholarship speaks of “watergate terms” (Schleusenbegriffe44) that necessitate empirical imports but ultimately remain legal concepts; the “social science argument does not trump the legal norm.”45 In other words, the definitions of the concepts to be empirically measured must come from the law and follow normative value judgments. In this limited sense, and without devaluating the other disciplines,46 there is a hierarchy between the law (in the form it has found through doctrine) and social science methods.
(b) Operationalization of legal concepts
Second, we have to validate the operationalization of this concept. How do we measure addictiveness? Often, we cannot directly measure the concept, and the operationalization—both of the dependent and of independent variables—is contingent on judgments calls, too; these valuations again have to be guided by the law.47 The operationalization may, in turn, impact the methods we (can) use. Especially quantitative empirical methods (statistics) often have strict assumptions that need to be met in order to produce valid results. If the law conceptualizes a term in a particular way, this may requires a certain way of operationalizing the variable. This can make the empirical measurement of the variable more difficult, and it may prevent the application of certain statistical tests. However, it is precisely this “prerogative of the law” that renders empirical studies into empirical legal studies.
(c) Validity of empirical results for legal concepts
Third, the empirical results have to be valid for the specific legal and doctrinal context. This poses some caveats to the use of empirical scholarship in doctrine, both on the side of doctrine and on the side of empirics. Doctrine structures and prepares decisions, while the social sciences help us describe, understand, and explain effects. When deciding, at times you will have to deal with arguments even if you cannot fully determine their empirical validity because the underlying processes are not (yet) well understood. Empirical evidence may also be inconclusive, and still lawyers have to decide cases to the best of their knowledge.48 Moreover, once an empirical argument is introduced into doctrine, it tends to assume an independent existence and lose its connection to the methods that helped to make it in the first place; as mentioned earlier, it is important always to bear in mind the (again, often implicit) normative assumptions of the social sciences when dealing with empirical findings in the law. Lawyers must also ward the temptation to consider empirical “knowledge” as axiomatic (or incontrovertible).
On the empirical side of the aisle, we have to ascertain the validity of the results vis-à-vis our legal research question. Once we have measured our empirical effect, we can derive practical implications and draw conclusions from them; these conclusions, however, have limits—limits we refer to as “validity.” Our conclusions are valid if we can argue that the effect found in the collected data follows some sort of general regularity. Traditionally, validity is assessed in three categories: internally, statistically, and externally.
Internal validity refers to the coherence of our empirical result; it lacks if we have not applied strict methodological scrutiny. For example, this may be the case if there is noise in the data that is correlated with both the dependent and the independent variable, without further statistical control, or if the sample was not drawn randomly from the population.49 Problems of internal validity can occur due to a flawed empirical design or due to practical problems (e.g., it may be impossible to observe or measure certain factors of systematical noise). Basically, a lack of internal validity voids the empirical results. Statistical validity is established if we can assume with a sufficiently high probability that our results are not based on chance, but that they are rather driven by a general regularity in the data. If the observed effect is statistically significant, we also assume that the results are statistically valid. To ascertain the validity of our empirical results before implementing important changes in policy or doctrine, it seems advisable to replicate the data in another setup.50 Even if the statistical probability is low by definition, we do not want to risk implementing changes based on arbitrary empirical findings. Internal and statistical validity are common requirements in any empirical endeavor, and in my view there are no aspects specific to the law that would add weight to these types of validity.
This is different when we talk about external validity. External validity refers to the generalizability of the results beyond the situation in which the data was collected. Does the general regularity we have observed in the data extend to other contexts? It will often be limited to special contexts, situations or groups of people. For example, a sample can only render externally valid results for the population it was drawn from. An experiment with male German blue-collar workers can (only) be generalized to this group (if a sufficiently large sample was drawn randomly from that population), and probably not to female Chinese entrepreneurs.51 In law, many of the effects we see actually depend on context, for example, on a certain framing of the situation; stripping it off too radically to generate a testable hypothesis may actually suppress what really matters. In other words, when empirics and doctrine meet, external validity is even more intricate. In the legal context we have to ask ourselves whether our research design features all important characteristics of the situation relevant under the law. There is also a strong connection between the proper definition of concepts (and variables) and their operationalization within the boundaries set by the law and the external validity of our results. Obviously, external validity is the main argument to contest empirical data in the law, and so a focus has to be laid on assuring the generalizability of empirical findings.
2.4. Interdisciplinarity and the comparative advantage of lawyers in this exercise
This superstructure thus clears the view for a strand of empirical legal studies that has thus far been largely neglected: empirics geared at public law doctrine may form a promising additional pillar of empirical legal research.52 But an important question remains: Is this a disciplinary or an interdisciplinary endeavor, or to put it more bluntly—who should embark on such journey? Should we lawyers simply adopt findings from other disciplines (e.g., economics or social psychology), should we outsource the generation of replies to empirical questions in the law to social scientists, or should we rather engage in such cumbersome endeavors ourselves?53
It seems to me that generating empirical evidence for use in public law doctrine is both a disciplinary and an interdisciplinary exercise. It is disciplinary, for we ask questions that are relevant for the law, and we speak to (and wish to convince) lawyers—whether they reside in wood-paneled courtrooms, in stately old piles of administration, in the venerable halls of legislature or in awe-inspiring academic libraries. Only for them a doctrinal argument is meaningful. If we focus on the research question, we thus clearly find to be involved in a disciplinary endeavor. At the same time, it is an interdisciplinary exercise to the extent that the methods employed are (currently) relatively new to lawyers, and only few are capable of actually handling empirical methods on a level that is comparable to the general level achieved in the more advanced neighboring disciplines. With view to the methods that we import and use we will thus have to concede that they are interdisciplinary.54
This view gives us leeway to answer the “who?”-question pragmatically. Of course we can adopt non-law scholarship, provided that we can assure that the implicit normative assumptions and valuations that underlie the research are in line with the valuations of the law. There are three law specific elements (definition, operationalization, and validity) that the design of an empirical study must satisfy so that we can employ it in responding to questions of legal scholarship; and this applies to any empirical research, whether conducted by researchers of other disciplines or from the legal academy. If these are satisfied, then we can incorporate this research into a doctrinal argument. The thinner the normative valuations by the law, the greater the scope of possible empirical arguments.
There are, however, two reasons why (at least some) lawyers working on doctrine should equip themselves with the appropriate methods to be able to produce methodologically strictly scrutinized responses to fundamentally legal, empirical quests. First, even to import non-law empirical scholarship we need to understand the questions, designs, methods and results of empirical social sciences. We at least have to understand enough to be able to tell whether the three aforementioned criteria are satisfied so that we can utilize and benefit from empirical findings of other disciplines. Second, my observation is that other disciplines rarely provide the precise empirical information lawyers need in doctrine, and for good reasons: they follow a different set of research questions, and they lack the intimate knowledge of the law.55 Lawyers in this respect have a comparative advantage on their home turf. Some phenomena may even require an advancement of methods, a specific modification to suit legal scholarship’s needs. But this seems like a worthwhile endeavor, as doctrine is a distillate of abstractions from specific decisions, so we can assume that we are indeed examining a relevant phenomenon, that its parameters and robustness are established.
In the specific example that will follow now to illustrate the points I have made thus far, the “law experiment” was conducted collaboratively by a (German) public law lawyer and a cognitive psychologist. We wrote the results up in two distinct publications: one addressed at the legal community and going into the details of doctrine, focusing on a discussion of the consequences of the law;56 and a (more technical) companion paper mainly addressing the community of psychologists, focusing on establishing the cognitive effect in a special context.57
3. Empirically challenging gambling law doctrine
In this section, I will give an example for an empirical analysis of legal norms in such a way that long-standing legal doctrine can be evaluated (and ultimately rejected), with high external validity, while maintaining strict methodological scrutiny. I thus propose introducing a methodology that may be called “law experiment” in a narrow sense, and that can be distinguished from experiments in the fields of economics or psychology.58 At the same time, psychological benchmarks are used where the law makes factual assumptions that are beyond its own (normative) scope. I propose that this type of design may provide a good paradigm for experimental (and, more generally, empirical) research in law. 59
3.1. Reasons for regulation: teleology
Gambling law is currently a much-debated topic in Germany. New regulatory models are being discussed, not least because the Court of Justice of the European Union (CJEU) declared the sports betting monopoly of the federal Länder as unlawful with regard to EU law. The Federal Constitutional Court (BVerfG) has, time and again, voiced serious reservations, too.
German gambling law has two main purposes, among a few others:60
the fight against pathological gambling61 (i.e., protection against addiction),
and, in a somewhat paternalistic vein reminiscent of consumer protection,62 the protection from game organizers exploiting cognitive effects that may lead to irrational behavior.63
The severe restrictions of the freedom of occupation and free movement of services that are imposed by gambling law are only justified if such legitimate purposes are pursued. Therefore, the law has to exempt games from regulation that do not pose the threats the law seeks to control. To differentiate, thus, between “dangerous” and “harmless” games, gambling law doctrine resorts to a series of gambling terms, at the center of which lies the distinction between “games of chance” (ludus fortunæ) and “games of skill” (ludus artis), very similar to the “predominant factor test” applied in US law (for the ease of reference in this text, I will refer to the distinction of games of chance and skill games using this American terminology). Both games of chance and skill games are games in which an option to win must be bought with a substantial monetary investment that may be lost if the player loses; stated more obviously, the way a game of chance ends will depend (for the most part) on happenstance, while a skill game principally depends on the players’ talents. Legal doctrine subjects a game to severe restrictions if chance, rather than skill, is its predominant trait. For the implementation of gambling law, doctrine thus transforms the question, “is this a dangerous game?” to the question, “is this a game of chance?”
3.2. Research question: are games of chance more “dangerous” than games of skill?
Therefore the question emerges whether distinguishing between dangerous and harmless games by differentiating between games of chance and games of skill is empirically vindicated,64 or to use the previous section’s terminology: we want to empirically validate a normative concept.
Implicitly, this view is based on the assumption that a lack of influence on the game makes that game more dangerous (an argument that seems seductively reasonable but, as we shall show, is wrong nonetheless). The assumption that games of skill are typically less dangerous than games of chance can be empirically tested. The validity of this assumption, however, cannot ultimately be tested without normative valuations.65 If we are engaging with doctrine, we are not free in our research question. In defining “dangerous,” we are bound to the stated purposes of the law. Games are only considered “harmful” if they are addictive and evoke irrational behavior that can be exploited. In this narrow sense, we can empirically test doctrine.
Furthermore, we are not free in how we operationalize the concepts we test. How do we turn “dangerous,” in the sense of our definition, into a measurable variable?66 In the case at hand, in referring to “addictiveness” and “exploitable cognitive patterns,” the law uses psychological concepts, so it seems valid to use operationalizations that are acceptable in that discipline. Neither the addictive power of a behavior nor the exploitability of behavioral patterns by vendors can be measured directly. What can be measured are factors that increase the addictive power and exploitability. In clinical psychological research, one of the most prominent cognitive factors that mediate addiction is illusion of control;67 one of the cognitive biases that could easily be exploited by vendors is overconfidence.68 We chose to use these two established behavioral patterns as an operationalization of “dangerousness” in the sense of the law. In Section 3.3, I will explain how we empirically measured these concepts.
3.3. Research design and method
To answer this research question (among a few others), we conducted an online survey with 214 participants.69 The participants were questioned in three test groups (treatment #1: 95 participants; treatment #2: 74 participants; treatment #3: 45 participants).70 The survey asked the participants to predict the results of real soccer matches to be held in the near future. Those participants whose predictions turned out to be correct could win €5. For each bet they placed, we asked test persons to state on a scale to what degree they thought the correct prediction depended on chance or on skill, and how certain they felt in their prediction. If we compare these results with our insights on whether skill has influence on a bettor’s success, then we can draw conclusions on whether illusion of control and overconfidence are at play.
To ascertain how the dangerousness of sports betting acts in relation to games of chance, we also asked the subjects to place bets on the first or last two digits in a popular German lottery (Spiel 77), with participants stating here also to what degree they thought correctly predicting the lottery outcome depended on skill or chance, and how certain they felt in their prediction. The same experimental subjects were, as a further comparative measure, also asked to place bets predicting the development of blue chip stocks in a fashion similar to futures trading.
The questionnaire ended by testing the “skill” of participants in two ways: first, they were asked to self-assess their soccer-related knowledge, and then they had to answer a sports quiz consisting of 20 questions; self-assessment and the results from the sports quiz were highly correlated, i.e., participants properly assessed their skill level.71
(a) Illusion of control
To measure the degree of the illusion of control in test participants in this study, the participants were asked to what extent they thought the correct prediction of the particular soccer match depended on their personal skill. If the participants were realistic in their self-assessment, then their answers would closely reflect the actual results. That is, as the participants’ skill increased, the importance they attributed to their skill to predict the outcome accurately should also increase, albeit only to the level that skill actually does influence predicting accuracy. If we look at the individual level, this is probably an exaggerated aspiration, but the data of all test persons and all games should, on aggregate, give a balanced account of over- and underestimation of the influence of one’s own skill on the betting result, if we are really dealing with random rather than systematic effects.
A similar approach was chosen for measuring self-confidence in the case of the current bet. Self-confidence was tested simply by asking participants how certain they felt in their predictions. If participants are able to measure their skill accurately, their self-confidence should be strongly positively correlated to their ability to correctly predict soccer match results. That is, the dependent variable “self-confidence” should be explained by the independent variable “correct predictions.” The subjects of the third group were also asked to state how many of their predictions they thought would turn out to be true. This figure was compared with the actual number of correct predictions to determine the presence or absence of (excessive) overconfidence. These conclusions can be used to determine whether sports betting is dangerous.
(a) Illusion of control
Our data show that participants with higher skill assume that the result of the bets depends more on their skill: in other words, the more skill a participant displays, the stronger she estimates the influence of skill to be on the result. We can, however, only reckon with an illusion of control if the higher estimation of the influence on the result is not justified by betting results that are in fact better. The question can be answered statistically by controlling for the correctness of bets. In other words, we need to check whether the illusion of control effect stays, even when one includes in the statistics the increased number of correct betting results, where the increase can be explained with higher skill. Even with this statistical control, however, the effect remains.72 We can thus conclude that we observe an illusion of control in people who give themselves a high skill-related self-assessment. If we compare the illusion of control to which test persons are subjected in sports betting with that present in normal lotteries, we can furthermore see that, on aggregate and throughout all test groups, illusion of control is more prominent in the context of sports betting, though not reaching conventional significance levels.73
The ability to make correct predictions is only weakly correlated with self-confidence.74 Instead, we found that participants’ self-assessed skill level was strongly correlated to their self-confidence.75 In other words, perceived skill had a strong influence on self-confidence, while self-confidence was only weakly correlated to the ability to correctly predict results. Taken together, one can conclude that how excessive a participant’s overconfidence is correlates strongly with how skilled the participant thinks he is. That is, highly skilled people demonstrate excessive overconfidence when it comes to predicting results. The comparison with lottery bets also leads to the conclusion that excessive overconfidence levels are higher in sports betting.
To analyze more concretely whether excessive self-confidence is actually present, we directly asked the participants in treatment group #3 how many of their soccer bets they believed would win. This allows a direct comparison between bets that were perceived to be correct and bets that were actually correct. If we subtract the latter from the former, placing the result in relation to skill level, then we find that those players with little skill only slightly over-estimate themselves, while self-overestimation grows disproportionately with increased skill.76 If we divide people at the median according to their skill level, we get an over-estimation of only 0.33 of nine matches (3.7%) for test persons with the lowest levels of skill, but an over-estimation in people with high skill levels of one of nine matches (11.1%).
(c) Validity of the results
In terms of internal validity, we do not see any possible objections to our research design. From a statistical point of view, the sample size is large enough to draw a statistically valid conclusion.
External validity, too, seems to be high. Participants77 were recruited using the DecisionLab of the Max Planck Institute for Research on Collective Goods. Soccer expertise covered the entire spectrum. All test persons can therefore be seen as potential participants in sports betting, or as a target audience for such betting games.78 Roughly one in ten of the participants had gambled commercially; we have also checked whether there are significant statistical differences between experimental subjects who had previously participated in sports bets or who were supporters of a soccer club and the rest of the sample; this was not the case. Conducting the experiment as an online study also allowed the test persons to use all the possibilities of a search for information that would also be at their disposal if they placed bets in regular circumstances; there were no time restrictions either, apart from the closure of the study prior to the event.79 The study was hence conducted in a natural environment; doing so gives the study additional external validity, i.e., the results can be transferred and generalized beyond the test scenario. Also, as shown above, definitions and operationalizations of the tested concepts were derived from the law, considering its normative valuations. We are thus confident that the results are robust. However, to actually base doctrinal changes on this research, one may consider replicating the results in another lab, and maybe with more carefully selected samples of experi mental subjects.
3.5. Implications for legal doctrine
Empirical insights per se have no direct consequence on the (interpretation of) law as it is. But in legally evaluating concrete cases, the law has to assess facts empirically outside of the legal arena to determine the factual basis for the application of the law. The findings have to be translated (back) into normative statements within the framework of legal hermeneutics. Such statements are close to reality and can be infused as doctrinal or systematic arguments into the legal discourse. Empirical insights about whether games of chance can be considered more dangerous than skill games can be related to the principle of proportionality.80 Whether state intervention, as effected by gambling law, is justified, is not an entirely normative question; rather, it can only be determined in an encounter between legal constructs and social reality. Even if the lawmaker is entitled to an assessment prerogative when designing policy and hence is given leeway for its evaluation and prognosis,81 this does not discharge from the obligation to base decisions on empirical facts. If we cannot establish a difference in harmfulness between games of chance and skill games, using this very division to distinguish dangerous and harmless games is arbitrary. Therefore, our empirical insights will need to be considered in doctrine.
In this context, the results of the empirical analysis have consequences for gamb ling law in general. For they prove that differentiating between games of chance and skill games does not solve the problem of separating dangerous from harmless games. On the contrary, if no skill was necessary and the result of the game merely depended on chance, then there would be no space for an illusion of control. One could even go so far as to say that gambling becomes dangerous precisely because of the skill that is necessary for playing. Skill being an element of a given game may cause participants to fall victim to an illusion of control; furthermore, because it seems plausible for a certain measure of skill to have a positive effect on the gamble, more skilled players are over-confident. On the other hand, every lottery player knows that his or her success depends on chance alone—hence, apart from the skill to “pick the right numbers,” there is not much room for illusion of control or overconfidence. It follows that distinguishing between games of chance and skill games to justify a limitation of the freedom to choose an occupation, or fundamental freedoms of private gambling operators is not an appropriate means (and thus not justifiable under the principle of proportionality) for achieving the two regulatory purposes. We are dealing with an arbitrary demarcation of limits, because the empirical assumptions underlying the ‘predominant factor test’ are inapplicable.
This leaves us with two possible solutions. Either we liberalize gambling altogether (through a change of statutes—a policy change, if you will), or we modify legal doctrine. If the legislature does not change the law to liberalize the gambling market, it is necessary and imminent to adapt legal doctrine, taking the consequences drawn from our empirical results into account. The “predominant factor test” should therefore be done away with altogether. Ultimately, it does not seem possible to make a general distinction between hazardous and harmless games, using only a single criterion. This brings us back to square one. The administration will have to take the trouble of assessing the danger potential of each game and its variants, on the basis of independent empirical findings, deciding the question for each game on its own merits. These empirical findings need to be evaluated in light of the purpose of the law. In this contribution, we have not engaged in evaluating whether our findings suggest categorizing sports betting as a “dangerous game”; we have merely shown that they seem to be more dangerous than other games (lotteries, futures trading) which are subject to strict regulation.
The argument put forward here itself purports to be a single contribution to the eminently complex discussion on regulating gambling in Germany and elsewhere. We wish to place more emphasis on the legally relevant empirical aspects of gambling. However, we only concentrate on two of the regulatory goals in gambling legislation (albeit weighty ones)—curbing gambling addiction and preventing the exploitation of the human passion for games. Moreover, we focus on a specific kind of sports betting; there may be other types of games, other rules, that impact how harmful in the legal sense a game is. Also, we have a particular way of defining and operationalizing legal concepts such as “danger” and, more precarious, “skill.” One may always argue that there are better ways of staying true to the law when defining and operationalizing skill. Moreover, there may be further steps taken to ensure (external) validity and replicability of the results (e.g., different samples, different lab/online environment).
To conclude, I think I have shown how empirical studies can be a worthwhile exercise to enrich public law doctrine. There are a number of precautionary measures to be taken when designing empirical studies to suit doctrinal needs. Caution is also commendable when adopting empirical scholarship from neighboring disciplines, as their research questions are often not aligned to the normative judgments of the law. Empirical legal scholarship for doctrinal purposes is thus more restricted in its scope than its counterpart geared at generating policy suggestions. But it can have a great impact on the law by creating arguments administrations and courts have to submit to, even without reaching the ear of the lawmaker to adopt new statutory law reflective of empirically backed policy advice.
These thoughts, admittedly, are a somewhat rough sketch of empirical approaches to the law. The differences between the German and American legal tradition are probably not as stark as I have drawn them above, especially when it comes to empirical legal studies. Moreover, the difference between orientation towards “doctrine” and towards “policy” is less pronounced than implied. At the end of the day, every legal scholar wants to advance the law, and therefore ultimately aims at refinements or changes of policy—be it through an adaptation of statute, common law or interpretation.82
As legal scholarship is showing signs of convergence globally, so are the methods applied. Having a strong focus on doctrine has been, in the history of law, one of the major contributions of German jurisprudence to the international legal discourse; maybe it can also play a role in empirical legal studies. The connection of doctrine and empirics may serve as a bridge to a more explicit and method-driven policy orientation in German law.83 It may, moreover, provide another avenue of scrutinizing the normative models underlying doctrine, an important function of legal scholarship. And grounding empirics in the lex lata