What is Wrong with Statistical Evidence? The Attempts to Establish an Epistemic Deficiency

Abstract

Many people have an intuition that something is fundamentally different between statistical evidence (e.g. per cent of the market-share) and non-statistical evidence (e.g. eyewitnesses). The intuitive difference between these two types of evidence is demonstrated by empirical research, according to which people tend to disregard the background statistical evidence when non-statistical evidence is available. A popular example is the red/blue bus scenario, in which a lady who was hit by an unidentified bus claims damages from the Red Bus Company on the basis that sixty per cent of the buses in that area are operated by that company. To many, it would seem intuitively wrong to impose liability based solely on this evidence. As a result, it is sometimes thought that statistical evidence is inferior to non-statistical evidence when it is used to prove particular facts in court.

The 'epistemic accounts' are attempts to provide a justification for restricting the use of statistical evidence in court by identifying a certain quality that non-statistical evidence has but statistical evidence lacks. The proponents of these accounts share the view that statistical evidence lacks a quality, the effect of which is to make the drawing of inference from the statistical evidence to the particular case epistemologically deficient. Epistemology is the branch of philosophy which addresses questions about the nature, possibility and structure of knowledge, justification, belief and other related conditions. Unlike psychology, which attempts to describe how human beings actually think, epistemology asks how a rational agent ought to think and form his rational beliefs. Epistemology is therefore normative in nature. These accounts are classified as 'epistemic' in this paper because their strategy is to identify a quality that is lacking from statistical evidence which makes it epistemologically unwarranted to establish the fact this evidence is adduced to prove.

The paper seeks to show that the existing epistemic accounts fail to provide a justification for restricting the use of statistical evidence in court. The paper presents the major epistemic accounts and argues that each suffers from substantial problems which make it unsuccessful. Some of these accounts are under-inclusive: the alleged epistemic deficiency is not found in some types of evidence which are clearly statistical. Other accounts are over-inclusive: their alleged epistemic deficiency applies to non-statistical evidence too. None of the existing accounts successfully establishes an epistemic deficiency from which (only) statistical evidence suffers.

This paper does not argue that statistical evidence should be used freely because there is no epistemic difference between statistical and non-statistical evidence. It may well be the case that there is a good justification for not using statistical evidence to determine our legal rights and liabilities, at least in some situations. However, if there is such a justification, it has not been provided by the epistemic accounts.

Keywords: Statistical evidence, generalisations, proof, inference, probability, epistemology, philosophy, Stein, Cohen, Thomson, Reference Class, Allan, Pardo, Dant