(to NETWORK VISUALIZATION)

CITATION ANALYSIS

We have now identified, acquired, and validated a corpus of 2,345 regulatory takings decisions (7.1% false-positives) that have been handed down by Federal courts since the 1978 Penn Central decision (438 US 104, 128). In addition, we have identified and acquired over 2,000 pre-1978 Federal regulatory takings decisions (dating back to 1900), and are in the process of cleaning and validating that dataset. Because our approach has been to minimize "missed cases" (i.e., false-negatives) prior to identifying false-positives, we are confident that the resulting corpus (1900-present) will represent the entirity of federal caselaw on the question of regulatory takings.

Our principal interest is in the "communication" between courts, understood as the citations to precedent within each decision. The post-1978 pool contains 14,388 unique citations to others within that pool - that is, judges writing those 2,345 decisions cite one another more than 14,000 times.

According to a purely binding precedent (i.e., hierarchical) model, courts should always cite higher courts (or themselves). We find that in the post-1978 pool, 89 percent of all citations conform to this conception. However, this cannot explain 1,224 citations - nearly 60 percent of which are found in Federal Circuit Court decisions.

As a preliminary assessment of the citation structure, we have calculated the circuit-to-circuit mean for each court and produced a mapping (scaled by standard deviations) for the 2383 citations within that stratum. If citations were to be randomly distributed (i.e., courts show no preference to cite any other), the topographical map would be flat.

When we look at the distribution of citations between Circuit Courts (see below), we again see that the dominant pattern is reflexive citation - for judges to cite earlier decisions handed down within their own jurisdiction.

However, when we control for reflexive citations (see below), we see several “archipelagos” emerge (i.e., Circuit courts that received a disproportionate number of citations from their sister courts). In particular, the 9th Circuit – and to a lesser degree, the 11th – are heavily referenced by other Circuit courts. This pattern does not fit easily into a hierarchical model. Considering the Supreme Court's ongoing involvement in the regulatory takings question after 1978, this suggests that persuasive legal authority may sometimes play a stronger role than structural hierarchy.