ABSTRACT
In this chapter, the authors illustrate the subtle ways in which a soft analysis fails to displace the casual analysis when evidence must be brought to bear on the current policy debate. The analysis assumes that the particular symptoms of concern - economic deprivation, female-headed households, joblessness, crime- can be dealt with by determining their particular causes and then developing policies that are directed at these factors. The authors examine Seebohm Rowntree's analysis of poverty in some detail, and then they consider the policy context with respect to poverty in the United States. The authors discuss six issues: overselling, fragile rationales, weak causal chains, over-generalisation, elimination of personal responsibility, and the confounding of issues. They conclude by discussing what they see as the implications of making policies with 'weak' as opposed to 'strong' causal theories. Social science has had little success in demonstrating strong causal relations among factors.
