Legal Network Analysis (LNA) studies predominantly focus on citation analysis. LNA has served various purposes, including determining the relevance of court decisions in terms of them being precedents, to explore how the relevance changed over time, and to examine which cases are similar based on their proximity in the citation network (community detection). LNA researchers have relied on various network analysis measures when answering their research questions. This raises the question which approaches can or should be used in order for LNA to produce meaningful results. Focusing on case law, this contribution discusses the purposes and challenges of LNA. More specifically, it will be shown that LNA lacks a proper reference point for evaluating the results and that, as a result, a methodology needs to be developed in order to produce results that are valid. Four specific aspects are subsequently explored more in-depth: (1) how to select sub-networks, (2) which community detection method to select, (3) estimating the probability that the network and its relationships as observed in the data did not occur by chance, and (4) which centrality measure to select to determine the extent to which a decision is a precedent. By examining these purposes and challenges, we aim to develop a research agenda for conducting LNA. Possible avenues for future research are discussed.
|Title of host publication||Computational legal studies: The promise and challenge of data-driven legal research|
|Publisher||Edward Elgar Publishing|
|Publication status||E-pub ahead of print - 26 Aug 2020|