Computational Social Science and its Potential Impact upon Law

Computational Social Science and its Potential Impact upon Law

Nicola Lettieri, [1] Sebastiano Faro [2]

Cite as: Lettieri, N., Faro, S., 'Computational Social Science and its Potential Impact upon Law', European Journal of Law and Technology, Vol. 3, No. 3, 2012

Abstract

Over the past decades, social sciences have gradually learned to use quantitative methods and computational tools. According to the emerging paradigm of so called 'computational social science' (CSS), the study of social phenomena is more and more often identified with the use of statistical and analytical tools, with data mining techniques and with the construction of simulation models, in other words, with computation.

Law has substantially fallen behind in exploiting the methodological and scientific outcomes of CSS research. The ongoing change though should involve legal culture for at least two reasons. Firstly, the making, interpretation and application of legal rules conceived to regulate social life cannot ignore the scientific knowledge and methodologies illuminating social dynamics at both individual and collective level. Secondly, CSS are going to provide lawyers with methods and tools offering new scientific basis to their findings and their concrete choices in applying and interpreting law.

The goal of the paper is twofold: on the one hand, to present CSS and discuss, in general terms, how it can positively influence law, offering to lawyers new tools and approaches for addressing legal problems; on the other hand, to focus on one of the most interesting research sectors that can take advantage from CSS approaches: the study of support methodologies and tools for policy making.

1. Introduction

Over the past decades, social sciences have gradually learned to use quantitative methods and computational tools. According to the emerging paradigm of so called 'Computational social science' (CSS), the study of social phenomena is more and more often identified with the use of statistical and analytical tools, with data mining techniques and with the construction of simulation models, in other words, with computation. The use of formal and mathematical methods, the gradual surmounting of disciplinary boundaries and tools provided by information and communication technologies are orienting social sciences, even though new interpretations of the experimental method, [3] towards falsifiability and cumulativeness that are specific features of natural sciences. [4] So far law has substantially fallen behind in exploiting the scientific and methodological outcomes of CSS research. Although there are some examples of the application of CSS methodologies to problem solving or the way lawyers perform their tasks, research experiences and studies have been sporadic and non-systematic.

The ongoing change though should involve legal culture for at least two reasons. Firstly, legal phenomena are at the same time the outcome and the ordering factor of social life. The making, interpretation and application of legal rules conceived to regulate social life cannot ignore the scientific knowledge and methodologies illuminating social dynamics at both individual and collective level. For example, knowledge about conditions under which social norms emerge and evolve, gained by means of applications of the so called distributed artificial intelligence (artificial societies and social simulations), seems suitable to be applied also in investigating legal issues. [5] Secondly, CSS are going to provide lawyers with methods and tools offering new scientific basis to their findings and their concrete choices in applying and interpreting law. An interesting example is represented by the use of data mining and information extraction techniques: the automated analysis of big amount of legal data, such as the full text corpus of decisions of the United States Supreme Court, can provide quick aggregate insights regarding the relative importance of particular legal questions during different time periods (Katz et al., 2011).

Our claim is that CSS can significantly impact law as issues like quality of regulation, good governance, access to public information can be addressed in an innovative and more effective way with effects on areas of fundamental social and legal relevance: from human rights to justice management, from policy design in law enforcement to formulation of foreign policies. The goal of the paper is twofold: on the one hand, to present CSS and discuss, in general terms, how it can positively influence law, offering to lawyers new tools and approaches for addressing legal problems; on the other hand, to focus on one of the most interesting research sectors that can take advantage from CSS approaches: the study of support methodologies and tools for policy making. Knowledge and techniques that can be ascribed to CSS are, in effect, demonstrating a significant capacity to contribute to public policy making: the results of the analyses performed with CSS methods can flow into the rule making process that plays a key role in policy making.

The paper is structured as follows. Section 2 presents, in general terms, the CSS scientific paradigm illustrating the way, from a methodological viewpoint, it is articulated. Section 3 refers to some legal issues that have found or may find answers thanks to CSS methodologies. Section 4 focuses attention, in particular, on the relationship between CSS and policy making, making reference to important research projects dedicated to this topic. Section 5 highlights how CSS tools can, in practice, enter into the rule making procedure, by referring to regulatory quality tools, like a bridge between the application of CSS methodologies to policy making and legal rule making. Finally, Section 6 sets out some conclusions.

2. A New Paradigm for Social Sciences

In February 2009, Science published an article by 15 leading scientists dedicated to the emergence of a new research paradigm of social sciences in which computational tools play a key role: Computational Social Science (Lazer et al., 2009). The authors begin with the observation of a phenomenon that is a fundamental characteristic of the knowledge society: 'we live life in the network'; each transaction that occurs in the network 'leaves digital traces that can be compiled into comprehensive pictures of both individual and group behaviour, with the potential to transform our understanding of our lives, organizations, and societies'. The capacity of new technologies to collect and analyse massive amounts of data is destined to significantly transform the sciences. This has already happened in such fields as biology and physics and it will also happen in the social sciences.

The position of these authors, coming from strongly diversified disciplinary spheres - from physics to economics - is a confirmation that social sciences are going through a phase of profound change due to two main factors. Firstly, the resort to computational tools and approaches; understanding social phenomena means increasingly using statistical and analytical tools, exploiting data mining techniques, running simulation models or, in other words, exploiting the power of computation (Casti, 1996; Simon, 1996). Secondly, the integration (made possible by the use of computational tools) among different disciplines and sciences. Different areas of social sciences have gradually become interested in this process: from the economy to political science, from sociology to anthropology, etc. (Cioffi-Revilla, 2010). The article in Science mainly focuses on the availability of an increasingly greater quantity of data and the surfacing of CSS 'that leverages the capacity to collect and analyze data with an unprecedented breadth and depth and scale'. [6]

In this paper, though, we refer to a wider notion of CSS, that proposed by Cioffi-Revilla (2010),,which presents CSS as 'a fledging interdisciplinary field at the intersection of the social sciences, computational science, and complexity science.' Within the Cioffi-Revilla framework, CSS include social disciplines that have in common the use of computational approaches and a set of different research methodologies: (a) social simulation, (b) complexity modelling, (c) social network analysis, (d) automated information extraction, and (e) social GIS.

  1. Social simulation. Simulation models of social phenomena appeared across the social sciences within a relatively short time after the initial utilization of computers for data analysis purposes, thereby establishing new foundations for contemporary computational social science research. The two principal types of computational simulation models involved in basic social research are systems dynamics models and agent-based models. System dynamics models are computational simulations that model a given target or reference system as a set of state variables (stocks) and their associated rates of change ( flows). Today, system dynamics models are implemented in very numerous industrial, managerial, and natural science applications. Agent-based models are computational simulations that model a given target system representing classes of actors and other social entities that interact through a variety of relations (associations) in a given environment. The basic ontology or landscape of an agent-based model typically includes a set of actors/agents, a set of interaction rules, and an environment with features that may be static or dynamic, organizational and/or spatial. The general class of problems addressed by agent-based models is that of explaining - through a simulation model - the emergence of collective or macroscopic behaviour based on the individual behaviour of agents or actors. A particularly valuable feature of computational simulation models for both basic social research and for policy analysis is their ability to run current and alternative policies to observe their effects (alternative scenarios), assuming a sufficiently well-developed base model of a given 'target system'.
  2. Complexity modelling. Complexity-theoretic models provide mathematical systems based on concepts and principles used for the analysis of non-equilibrium dynamics. Such dynamics, that are far from equilibrium (in general, non-Gaussian distributions), [7] are quite often found in challenging research problems across the social sciences. Patterns and regularities of social phenomena like terrorist attacks, wealth and poverty in developing societies, political instability, foreign aid distributions are instances of non-equilibrium dynamics. Complexity models aim at finding regularities of these dynamics in order to comprehend how they evolve and, possibly, predict them.
  3. Social network analysis (SNA) . Networks are ubiquitous in the social world: social relationships can be described and studied as networks consisting of a set of nodes and a set of relations, each defined by a set of attributes. For this reason, scientific investigations of social systems or processes often include networks. SNA aims at providing insightful information and inferences on the functionality of social groups (organizations, institutions, etc.) given their structural patterns of nodes and relations. It has many computational applications across social sciences and is supported by a large family of metrics and exact methods. SNA can provide interesting insights by analyzing properties such as resilience, [8] vulnerability, decomposability, functionality that would not be inferred through plain observation or through more traditional methods. In addition, SNA can be applied to the design of more robust networks relevant to public policy (e.g., transportation, homeland security, public health).
  4. Automated information extraction is the method of parsing and coding documents in order to extract information from them. This technique has recently evolved into the computational analysis of multiple media (text, audio, images, video). The efficiency of these methods is improving thanks to the introduction of computational methods from artificial intelligence and other computational algorithms, an effort that is likely to yield significant breakthroughs in the future. Automated information extraction and text mining are a promising computational strategy in areas of social science that are 'text-rich' and 'numbers-poor' (Cioffi-Revilla, 2010).
  5. Social Geographic Information Systems (GIS). Social GIS are tools for visualizing and analyzing spatially-referenced data concerning the social world. It has found many applications across the social sciences and related disciplines (e.g., criminology and regional economics). Social GIS have also been combined with other quantitative techniques to produce unique new insights about spatial patterns and are also closely related to the vast field of spatial statistical analysis, but with a greater emphasis on visualization of layers of social data.

In general terms, it is worth underlining that the different methodologies so far mentioned can be (and often are) combined in order to increase their explanatory and predictive power. From this point of view, CSS is in tune with the attitude towards the methodological eclecticism that characterizes scientific enquiry today.

3. Computational Social Science Methods in the Legal World: Some Experiences

The encounter between CSS and law has in some cases already taken place. We can cite interesting research experiences, in many ways pioneering experiences, in which the methodologies mentioned so far are used as much for studying law as for delving into knowledge about social phenomena relating to the law itself. The research defines an already broad application horizon spanning from the study of normative corpora through data mining and information extraction techniques, to the use of network analysis or complexity modelling techniques for studying jurisprudential trends and their evolution.

Table 1 offers, without any pretence of completeness, a first glance at the most significant research on the matter considering the legal tasks under examination, the objectives of the research, the methodologies used and recently published selected research works. On a general level, we can see how, from the analysis of these initial experiments, there are advantages that potentially derive from the dialogue between law and CSS. First of all, CSS paradigm can promote in law a scientific (in the Galilean sense) mind set promoting a quantitative approach to legal problems and offering scientifically grounded answers to legal issues. Quantitative research, the empirical investigation of phenomena via statistical, mathematical and computational techniques that is transforming social science, can indeed improve the way we understand legal matters and the evolution of legal and social systems. CSS suggests innovative ways to exploit, also in the legal field, the 'data deluge' that characterizes science, the 'Big Data Science' in the knowledge society (Anderson, 2008; Hey et al., 2009), [9] offering a chance to bring together both quantitative and qualitative research.

Moreover, especially by means of simulation models, CSS is providing legal science with a new experimental method, a new approach to social science research, according to which modelling the structural properties of social systems and exploring their spatio-temporal development via computer simulation are crucial steps not only to provide explanations of complex social outcomes, but also to predict the evolution of social dynamics. In this way, CSS is promising to fill the gap between natural and social sciences bringing them closer to the objectivity, falsifiability and cumulativeness of natural science.

Tasks

Research goals

Methodologies

Relevant works

Analysis of legal systems

  • Analysis of (hidden) structural and dynamic properties of legal systems in order to characterize and compare different legal corpora or juridical cultures
  • Study of the evolution of legal systems over time
  • Automated information extraction
  • Complexity modelling
  • Social network analysis

Pagallo (2010); Katz et al. (2011)

Analysis of normative texts

  • Analysis of legal concepts and of their relations
  • Study of the evolution of norms over time.
  • Automated information extraction
  • Social network analysis

Bommarito and Katz (2009, 2010); Boulet, Mazzega and Bourcier (2010)

Case law analysis

  • Analysis of the relations between precedents.
  • Study of the evolution of case law over time.
  • Analysis of the relevance of precedents in case law
  • Automated information extraction
  • Social network analysis

Chandler (2005); Fowler et al. (2007); Pagallo (2007); Katz and Stafford (2010); Vining and. Wilhelm (2011)

Analysis of international alliances structures

  • Study of legal relations among members of international alliances
  • Analysis of structural properties (resilience, vulnerability, decomposability, functionality etc.) of alliances
  • Automated information extraction
  • Social network analysis
  • Social simulation

Deffuant et al. (2002); King, and Lowe (2003); Raczynski (2004); Scott Bennett (2008); Geller and Alam (2010).

Analysis of procedures regulated by law
(bargains, voting procedures, judiciary procedures, divorce proceedings)

  • Study of structural and dynamic properties
  • What-if analysis
  • Design and re-engineering
  • Social simulation

Thoyer et al. (2001); Rouchier and Thoyer (2006); Mumcu and Saglam (2008); Bonaventura and Consoli (2009); Cerulli (2012)

Crime analysis

  • What-if analysis
  • Pattern detection in behaviours and social phenomena relevant to criminal law
  • Social simulation
  • Social GIS
  • Complexity modelling

Makowsky (2006); Liu and Eck (2008); Rauhut and Junker (2009); Bosse and Gerritsen (2010); Giura et al. (2010); Fonoberova et al. (2012).

Table 1: Some experiences of CSS methods relevant to the legal world.

4. Walking the Fine Lines between Computational Social Science and Law: Policy Making

Along with the tasks set out in Table 1, a very promising and stimulating field for the application of CSS is represented by policy making. The topic, recently discussed with regard to the aspect of the relationship between science and policy making, [10] is closely linked to the law. Policy making, indeed, not only involves choices of a policy nature but also rule making. There are many reasons why it is appropriate to investigate the relationship between CSS and policy making strictly related to the nature and complexity of the problems that policy making must face in modern societies. As a matter of fact, modern societies are characterized by an unprecedented degree of complexity originated by the interdependence among their technological, social, cultural and economic components. This complexity, emphasized by globalization and by communication dynamics of an interconnected society, often produces counter-intuitive effects driven by feedbacks that are difficult to manage. CSS can offer a vital contribution in gathering these aspects of complexity and interdependence and in contributing to public policy making at different levels of government.

The numerous research projects that, in recent years, have begun to explore the positive relationship that can exist between CSS methodologies and policy making are proof of this. Here we wish to mention the following four projects from among the important projects that have been developed in this direction: FuturICT, FOC, Eurace and Riftland.

  • FuturICT is a response to the European Commission's Flagship Call in the area of ICT aiming at delivering new science and technology to understand and manage the complexity of modern connected societies. The idea underlying the proposal is to combine ICT with knowledge from the social and complexity sciences; ICT will provide the data to boost the social sciences; complexity science will shed new light on emergent phenomena of socially interactive systems, and the social sciences will provide a better understanding of the opportunities and risks of strongly networked ICT systems. FuturICT aims to promote a scientific endeavour revealing the dynamics of socially interactive systems in order to inspire the creation of new tools (social simulations, massive data mining) to manage social complexity. [11]
  • FOC (Forecasting financial Crises) is a project financed by the EU Commission aiming at understanding and forecasting systemic risks and global financial instabilities by means of a novel integrated and network-oriented approach to the issue. On one hand, it aims at offering a theoretical framework to measure systemic risk in global financial market and financial networks. On the other hand, its goal is to deliver an ICT collaborative platform for monitoring systemic fragility and the propagation of financial distress across institutions and markets allowing experts to evaluate algorithms and models to forecast financial crises as well as visualize interactively possible future scenarios. [12]
  • Eurace is an interdisciplinary project involving computer scientists and economists in the development of a simulation model of the European economy using advanced software and high performance computing. The project proposes an innovative approach to macroeconomic modelling and economic policy design. The interplay between the different components of public policy (fiscal and monetary strategies, knowledge exchange, R&D incentives etc.) and their effects are difficult to be understood and predict by means of classical economic approaches. Eurace investigates the potentialities of agent-based computational models in this field. [13]
  • RiftLand is multi-disciplinary research project funded by the US Office of Naval Research aimed at creating a simulation model for both basic fundamental research and applied exploratory policy analysis in the form of scenarios. Its primary goal is to develop an agent-based model for analyzing disaster scenarios in East Africa, with emphasis on coupled social, natural, and artificial systems. RiftLand exploits GIS for many of its data layers, and covers an area representing approximately 120 million inhabitants. Climate, topography, land use, hydrology, roads, boundaries, other infrastructure, and human inhabitants are among the main entities represented. [14]

All these projects are characterized by a wide use of CSS methodologies and by the same goal of supporting that which the EU defines as 'scientific evidence-based policy making' as policymakers need reliable knowledge to enhance the quality of their decisions. [15]

5. Computational Social Science and Regulatory Quality Tools

The methodologies and tools attributable to the area of CSS, also in combination amongst themselves, can contribute to defining objectives and content of policies as demonstrated by the projects we have just mentioned. But, how do the results obtained by the application of these methodologies and tools to the definition of policies in practice flow into the process of rulemaking? The answer is in the procedure followed for the formulation, implementation and evaluation of rules, a set of procedures and rules (more precisely, meta-rules) falling within the concept of the 'quality of regulation'. For several decades, the question of the quality of regulation has been part of the agenda of national governments and parliaments and, at the supranational and international level, of the OECD and the EU. The OECD has dedicated attention to the theme of better-quality regulation since 1995; it recently adopted the 2012 Recommendation of the OECD Council on Regulatory Policy and Governance. [16] The EU has also been very active in promoting better regulation and regulatory quality tools; in the 1990s, it introduced a series of ideas, experimentation with new methodologies was undertaken and the first mechanism of impact assessment was tested. In the past decade, the theme has been discussing in a systematic way especially within the context of the White Paper on Governance [17] where the EU Commission mentioned 'better regulation' stating that it is necessary to 'pay attention to improving the quality, effectiveness and simplicity of regulatory acts'.

The notion of regulatory quality that emerges from the national and supranational framework covers both process and outcome. With reference to the rule making process, three different tools are highlighted:

  1. Regulation Impact assessment (RIA). It is a methodology to evaluate the impacts of proposed legislation before its adoption. It consists in a socio-economic evaluation, through the comparison of different hypotheses of normative action (that which is otherwise defined as a what-if analysis), of the effects of these actions on the activities of citizens and businesses, as well as on the organisation and functioning of public administrations. It aims at directing policy-makers in adopting the most efficient and effective regulatory options. Moreover, RIA increases the explication of the justification of the regulatory solutions adopted and is also a tool for policy coherence and policy integration because it requires policy makers to examine effects on other policy areas. Good RIA gathers qualitative and, where possible, quantitative evidence from all the stakeholders; it involves the final decision maker in the sense that it changes the way in which rule makers think about public policy. More specific analyses like the analysis of administrative costs or risk analysis are included within RIA.
  2. Citizen involvement in policy making. Typically this occurs with consultation, in particular, through the mechanism of notice and comments (the public are requested to give their comments on specific issues to be regulated). Collecting information on the impact of regulation on the public, including their perception of regulation, helps governments to structure their policies to address perceived issues and better prioritise reforms to focus on those areas that may warrant regulation, or where regulation may be unnecessarily burdensome.
  3. Ex-post evaluation. This gives a judgment of interventions according to their results and impacts in relation to the needs they aim to satisfy and the resources mobilized. Evaluation generates relevant information that is essential for planning, designing and implementing policies. Evaluation tries to answer questions on relevance (do the objectives correspond to the needs and problems?) effectiveness (did they achieve the objectives?) and efficiency/cost-effectiveness (were the objectives achieved at reasonable costs?) of rules.

The scientific and methodological acquisitions reached in the field of CSS lend themselves basically to support the effectiveness of these tools in three ways, through

  1. acquisition of new knowledge
  2. extraction of information from the huge quantities of information available today in digital form, and
  3. what-if analysis based on predictive techniques.

With regard to RIA and ex post evaluation, the simulation models may help us, for example, to better understand the studied phenomena; that is, thanks to their capacity to draw inferences from a complex system of human and social dynamics that can be specified through variables and relations in both quantitative and qualitative form. The simulation models, in effect, help to throw light on important properties such as short- and long-term effects, feedback and feed forward effects, stability properties (or lack thereof), fluctuations, and other interesting features for understanding social dynamics and designing better policies (Cioffi-Revilla, 2010).

Regarding the phase of gathering the opinions of citizens, normally, in relation to the contribution that new technologies can make, reference is made to e-consultation procedures. However, the consultations do not always, even when carried out with the aid of computer tools, obtain a large number of responses; also the knowledge produced through the explicit responses of the parties consulted often do not allow, due to reticence, misunderstanding, false answers, etc., all the important aspects of the analyzed phenomenon to be captured. Other means that can assist in understanding the position of citizens with regard to specific issues are, therefore, undoubtedly useful. Various other methodologies attributable to CSS may be of assistance permitting us to see the way in which citizens perceive policies and their results. Here, we are thinking, for example, of the implicit knowledge that may emerge from inferences based on the analysis of large quantities of data coming from the media or social networks or a so-called sentiment analysis (Liu, 2010). In this way, for example, the realisation of more complex and articulated indicators than those traditionally used for evaluating the perception of policies could be supported (such as the World Bank's worldwide governance indicators [18] that measure the perception of citizens and businesses and experts about the quality of governance in six distinct dimensions).

At this point, the reasons why lawyers are interested in CSS methods applied to policy making issues appear more obvious. Making regulatory quality tools more effective contributes to improving regulation, with positive effects on the application of basic legal principles of action by public authorities and on their relationship with citizens (transparency, accountability, legitimacy, proportionality). Moreover, it is worth considering that there are increasing numbers of administrative authorities involved in regulation which have increasingly wider discretionary powers and whose powers are not based on direct democratic investiture. With regard to this last matter, the better functioning of regulatory quality tools can contribute to making decisions more transparent, limiting the regulator's discretion and guaranteeing greater objectivity in the regulation.

6. Conclusions

The relationship between CSS and law is promising for various reasons.

From a theoretical viewpoint, the scientific paradigm underlying CSS research may encourage lawyers to pay closer attention to the empirical dimension of legal phenomena and to be more open in their dialogue with other disciplines. These are two fundamental requirements for proper legal decision making in relation to the complexity and dynamism of the knowledge society. Then, from a methodological viewpoint, the CSS paradigm may promote a scientific (in the Galilean sense) mind set in law, fostering a quantitative approach to legal problems and providing them with more scientifically grounded and evidence-based answers.

Also from the application viewpoint, the impact of CSS on law is important. It seems to offer not only the possibility to study and understand some legal issues better but also to provide significant innovation in policy and rule making processes. From this last point of view, a series of research perspectives are opening up as much for legal informatics as for IT law. On one hand, there is an emergent need to design CSS methods and tools specifically aimed at supporting quality regulatory tools. On the other hand, the use of these methods and tools, within the sphere of rulemaking processes requires specific rules to be made, addressing specific questions like can RIA be exclusively based on simulation techniques? How far can we substitute consultation with stakeholders with sentiment analysis? How must personal data be protected where massive operations of data mining are carried out?

References

Anderson (2008), The End Of Theory: The Data Deluge Makes the Scientific Method Obsolete, Wired Magazine, http://www.wired.com/science/discoveries/magazine/16-07/pb_theory

Banks (2009), Evidence-based policy-making: What is it? How do we get it?, Lecture, ANZSOG/ANU Public Lecture Series, Canberra, 4 February 2009. Available at SSRN: http://ssrn.com/abstract=1616460.

Bommarito II, Katz (2010), A Mathematical Approach to the Study of the United States Code. http://arxiv.org/abs/1003.4146.

Bommarito II, Katz (2009), Properties of the United States Code Citation Network. Available at SSRN: http://ssrn.com/abstract=1502927.

Bonaventura, Consoli (2009), Priorities criteria and criminal caseload management, MPRA Paper No. 14559.

Bonaventura, Consoli (2013), Priorities for Backlog of Criminal Cases Pending in Courts: A Computational Agent-based Model, forthcoming in Faro, Lettieri (eds.), Law and Computational Social Science, special issue of the journal Informatica e Diritto, 1, 2013.

Bosse, Gerritsen (2010), Social Simulation and Analysis of the Dynamics of Criminal Hot Spots, Journal of Artificial Societies and Social Simulation 13 (2) 5, http://jasss.soc.surrey.ac.uk/13/2/5.html.

Boulet, Mazzega, Bourcier (2010), Network Analysis of the French Environmental Code, in Casanovas, Pagallo, Sartor, Ajani (eds.), AI Approaches to the Complexity of Legal Systems, Heidelberg, pp. 39-53.

Casti (1996), Would-Be Worlds. How Simulation Is Changing the Frontiers of Science, New York, Wiley.

Cerulli (2012), Are R&D Subsidies Provided Optimally? Evidence from a Simulated Agency-Firm Stochastic Dynamic Game, Journal of Artificial Societies and Social Simulation 15 (1) 7, http://jasss.soc.surrey.ac.uk/15/1/7.html.

Chandler (2005), The Network Structure of Supreme Court Jurisprudence, Houston: University of Houston Working Papers Series No. 2005-W-01;

Cioffi-Revilla (2010), Computational social science, Wiley Interdisciplinary Reviews: Computational Statistics, 2, 3, pp. 259-271.

Cukier (2010), Data, data everywhere, The Economist, http://www.economist.com/specialreports/displaystory.cfm?story_id=15557443 .

Deffuant, Amblard, Weisbuch, Faure (2002), How Can Extremism Prevail? A Study Based on the Relative Agreement Interaction Model, Journal of Artificial Societies and Social Simulation, vol.5, no.4. http://jasss.soc.surrey.ac.uk/5/4/1.html

Duijn, Immers, Waaldijk, Stoelhorst (2003), Gaming Approach Route 26: a combination of computer simulation, design tools and social interaction, Journal of Artificial Societies and Social Simulation vol. 6, no. 3, http://jasss.soc.surrey.ac.uk/6/3/7.html

Epstein (2006), Generative Social Science. Studies in Agent-Based Computational Modeling, Princeton (NJ), Princeton University Press.

Epstein, Axtell (1996), Growing Artificial Societies. Social Science from the Bottom Up, Cambridge (MA), The MIT Press.

Fonoberova, Fonoberov, Mezic, Mezic, Brantingham (2012), Nonlinear Dynamics of Crime and Violence in Urban Settings, Journal of Artificial Societies and Social Simulation 15 (1) 2, http://jasss.soc.surrey.ac.uk/15/1/2.html.

Fowler, Johnson, Spriggs II, Jeon, Wahlbeck (2007), Network Analysis and the Law: Measuring the Legal Importance of Precedents at the U.S.Supreme Court, Political Analysis, vol. 15, no. 3, pp. 324-346. Available at SSRN: http://ssrn.com/abstract=906827.

Geller, Alam (2010), A Socio-Political and -Cultural Model of the War in Afghanistan, International Studies Review 12 (1), pp. 8-30.

Giura, Giuffrida, Pennisi, Zarba (2010), Socio-Legal Analysis of Criminal Sentences: A Preliminary Study, Proceedings of the Fourth International AAAI Conference on Weblogs and Social Media, North America: http://www.aaai.org/ocs/index.php/ICWSM/ICWSM10/paper/view/1472

Hey, Tansley, Tolle (2009), In The Fourth Paradigm: Data-Intensive Scientific Discovery, Journal of Artificial Societies and Social Simulation vol. 11, no. 4 7 http://jasss.soc.surrey.ac.uk/11/4/7.html.

Katz, Bommarito, Seaman, Candelub, Agichtein (2011), Legal N-Grams? A Simple Approach to Track the 'Evolution' of Legal Language, available at SSRN: http://ssrn.com/abstract=1971953 .

Katz, Stafford (2010), Hustle and Flow: A Social Network Analysis of the American Federal Judiciary. Ohio State Law Journal, vol. 71, no. 3, pp. 457-509.

King, Lowe (2003), An Automated Information Extraction Tool for International Conflict Data with Performance as Good as Human Coders: A Rare Events Evaluation Design, International Organization, 57, pp. 617-642.

Kirkpatrick, Parker (2007), Regulatory Impact Assessment Towards Better Regulation?, Cheltenham, Edward Elgar Publishing.

Lazer, Pentland, Adamic, Aral, Barabási, Brewer, Christakis, Contractor, Fowler, Gutmann, Jebara, King, Macy, Roy, Van Alstyne (2009), Computational Social Science, Science 6 February 2009: vol. 323 no. 5915, pp. 721-723.

Liu (2010), Sentiment Analysis and Subjectivity, in Indurkhya, Damerau (eds.), Handbook of Natural Language Processing, second edition, Chapman and Hall/CRC.

Liu, Eck (eds.) (2008), Artificial Crime Analysis Systems. Using Computer Simulations and Geographic Information Systems, Hershey, Information Science Reference.

Makowsky (2006), An Agent-Based Model of Mortality Shocks, Intergenerational Effects, and Urban Crime, Journal of Artificial Societies and Social Simulation 9(2) 7, http://jasss.soc.surrey.ac.uk/9/2/7.html.

Moss (2002), Policy Analysis from First Principles, PNAS, 99, Suppl. 3, May 14, 2002, pp. 7267-7274.

Mumcu, Saglam (2008), Marriage Formation/Dissolution and Marital Distribution in a Two-Period Economic Model of Matching with Cooperative Bargaining, Journal of Artificial Societies and Social Simulation, vol. 11, no. 4 3, http://jasss.soc.surrey.ac.uk/11/4/3.html

Pagallo (2007), Small World Paradigm and Empirical Research in Legal Ontologies: A Topological Approach, in Ajani, Peruginelli, Sartor, Tiscornia (eds.), The Multiple Complexity of European Law: Methodologies in Comparison, Florence, EPAP, pp. 195-210.

Pagallo (2010), Power Laws and Levels of Abstraction in Understanding Legal Systems, in Andrighetto, Boella, Pagallo, Villata (eds.), Combined Proceedings of the International Symposium on Social Network Analysis and Social Norms in MAS, Leicester, UK: AISB, P.C., De Montford University, 2010, pp. 41-45.

Raczynski (2004), Simulation of The Dynamic Interactions Between Terror and Anti-Terror Organizational Structures, Journal of Artificial Societies and Social Simulation vol. 7, no. 2, http://jasss.soc.surrey.ac.uk/7/2/8.html.

Rauhut, Junker (2009), Punishment Deters Crime Because Humans Are Bounded in Their Strategic Decision-Making. Journal of Artificial Societies and Social Simulation 12 (3) 1, http://jasss.soc.surrey.ac.uk/12/3/1.html.

Rouchier, Thoyer (2006), Votes and Lobbying in the European Decision-Making Process: Application to the European Regulation on GMO Release, Journal of Artificial Societies and Social Simulation vol. 9, no. 3, http://jasss.soc.surrey.ac.uk/9/3/1.html.

Scott Bennett (2008), Governments, Civilians, and the Evolution of Insurgency: Modeling the Early Dynamics of Insurgencies, Journal of Artificial Societies and Social Simulation vol. 11, no. 4 7, http://jasss.soc.surrey.ac.uk/11/4/7.html.

Simon (1996), The Sciences of Artificial, Cambridge (MA), The MIT Press.

Thoyer, Morardet, Rio, Simon, Goodhue, Rausser (2001) A Bargaining Model to Simulate Negotiations between Water Users, Journal of Artificial Societies and Social Simulation vol. 4, no. 2, http://jasss.soc.surrey.ac.uk/4/2/6.html.

Troitzsch (2013), Legislation, Regulatory Impact Assessment and Simulation, forthcoming in Faro, Lettieri (eds.), Law and Computational Social Science, special issue of the journal Informatica e Diritto, 1, 2013.

Vining Jr., Wilhelm (2011), Measuring Case Salience in State Courts of Last Resort, Political Research Quarterly, 64, 559-572

Yücel, Daalen (2009), An Objective-Based Perspective on Assessment of Model-Supported Policy Processes, Journal of Artificial Societies and Social Simulation 12 (4) 3 http://jasss.soc.surrey.ac.uk/12/4/3.html.



[1] Nicola Lettieri is Researcher at ISFOL. He is adjunct Professor of Legal informatics at the University of Sannio, Benevento, and of Computational Social Sciences at the University of Salerno (Italy).

[2] Sebastiano Faro is Senior Researcher at the Institute of Legal Information Theory and Techniques (ITTIG), National Research Council (CNR), Italy.

[3] Experimental method, the cornerstone of scientific enquiry in natural sciences, finds now place in social sciences by means of computer simulations, virtual laboratories in which social scientists can study social phenomena. Simulations were the starting point of a new scientific paradigm theorized in 2006, with reference to social sciences, by Joshua Epstein, a mathematician from the Santa Fe Institute. According to Epstein, simulations are giving birth to a new kind of science, 'generative social science' that can be considered the third paradigm of science, after induction and deduction. According to the generative approach, to explain social phenomena, social scientists have to reproduce, 'generate' them within a computer simulation.

[4] Within the philosophy of science, falsifiability is the quality belonging to scientific hypothesis and theories that are testable by empirical experiment and thus conform to the standards of scientific method. Cumulativeness is the quality of scientific knowledge that increases by successive additions.

[5] See, e.g. the seminal work of R. Axelrod, An Evolutionary Approach to Norms, The American Political Science Review, Vol. 80, No. 4 (Dec., 1986), pp. 1095-1111. This work, based on a computer simulation of a game inspired by the Prisoner's Dilemma, deals with the emergence and stability of behavioral norms, in order to show the conditions under which norms can evolve and prove stable. The relevance for lawyers of this approach emerges from what Axelrod observes about norms: 'Today, norms still govern much of our political and social lives. In politics, civil rights and civil liberties are as much protected by informal norms for what is acceptable as they are by the powers of the formal legal system.' (p. 1096). Moreover, regarding to the relationship between norms and law he notes that: 'Social norms and laws are often mutually supporting. This is true because social norms can become formalized into laws and because laws provide external validation of norms. They are also mutually supporting because they have complementary strengths and weaknesses. Social norms are often best at preventing numerous small defections where the cost of enforcement is low. Laws, on the other hand, often function best to prevent rare but large defections because substantial resources are available for enforcement.' (p. 1107).

[6] Science

[7] In probability theory, the Gaussian (or normal) distribution is a probability distribution that has a bell-shaped probability density function often used as a first approximation to describe real-valued random variables that cluster around a single mean value. (Source: Wikipedia).

[8] Resilience is defined as 'the positive ability of a system or company to adapt itself to the consequences of a catastrophic failure caused by power outage, a fire, a bomb or similar' event. In recent years the term has been used to describe a burgeoning movement among entities such as businesses, communities and governments to improve their ability to respond to and quickly recover from catastrophic events such as natural disasters and terrorist attacks (Source: Wikipedia)

[9] According to Hey and Trefethen, 'The Data Deluge: An e-Science Perspective'. In Berman, Fox, Hey (eds.), Grid Computing - Making the Global Infrastructure a Reality, Wiley and Sons, 2003, pp. 809-824, 'It is evident that e-Science data generated from sensors, satellites, high-performance computer simulations, high-throughput devices, scientific images and so on will soon dwarf all of the scientific data collected in the whole history of scientific exploration. Until very recently, commercial databases have been the largest data collections stored electronically for archiving and analysis. .... As of today, the largest commercial databases range from 10's of Tbytes up to 100 Tbytes. In the coming years, we expect that this situation will change dramatically in that the volume of data in scientific data archives will vastly exceed that of commercial systems. Inevitably this watershed will bring with it both challenges and opportunities'.

[10] The discussion about the topic has overcome the boundaries of scientific debate (e.g. see a recent article on the Economist presenting some applications of computer simulation techniques in crisis and civil conflict management http://www.economist.com/node/21553006).