Fault-based and strict liability in the law of neighbours

Abstract

By the end of the twentieth century, and after a long line of conflicting case law, the question about the basis of liability in nuisance was settled: in Scotland, damages are awarded only upon proof of fault (RHM Bakeries (Scotland) Ltd v Strathclyde Regional Council 1985 SC (HL) 17). Fault, in turn, can adopt many forms: malice, intention, recklessness, negligence, and conduct causing a special risk of abnormal damage (Kennedy v Glenbelle Ltd 1995 SC 95).

Many aspects of this seemingly clear picture, however, remain problematic. On the one hand, the way in which this model is interpreted and applied gives place to particular forms of liability that can actually be characterised as strict. On the other hand, two other areas of the law of neighbours that overlap with the scope of nuisance do not fit entirely this model, namely the regulation of disputes over uses of water and of those arising from withdrawal of support.

The main argument of this thesis is that damages claims in the context of neighbourhood are governed by two distinct rules: a general fault-based liability rule for nuisance, and an exceptional strict liability rule for abnormally dangerous conduct. For the first of these rules, the thesis offers an evaluation of the fault model adopted in Kennedy v Glenbelle Ltd, explaining the interaction between its different elements and highlighting the developments that can result in forms of strict liability. For the second of these rules, the thesis develops an analysis of its elements and nature, as well as a proposal that delineates its scope of application.

This two-rule model offers a justification for the current structure of the law applicable to disputes over uses of water. The strict liability rule applicable to interferences with the natural flow of watercourses, traditionally explained as based upon the infringement of property rights, is better explained as danger-based. The regulation of disputes arising from withdrawal of support, however, is not consistent with this model, even though they have also been characterised as nuisances. It is argued that this framework entails unjustified inconsistencies, both internal and by reference to the model proposed, and that it should be adjusted accordingly.