Introduction

The preceding quotes point to the complexities of homelessness legislation. This chapter looks to homelessness law and policy for the following interconnected reasons. Local authorities have a duty to house certain individuals, which is imposed by parliamentary legislation. So, homelessness law is an important form of coordinative discourse between state institutions. However, unlike immigration law, homelessness law establishes more duties than powers. It allows us to see how accountability for policy decisions is affected by indeterminate legislation. This chapter therefore looks to the second prong of policy feasibility. Where the previous chapter focused on delegation, this chapter looks to accountability. In addition to which, this chapter analyses coordination between different levels of government: national and local. Chapter Four concerned the powers of central government as delegated by the national Parliament. However, coordinating across different levels and among multiple local authorities adds another degree of complexity. Also, this chapter interrogates the effects of adjectives where Chapter Four demonstrated a more substantial effect from conditional conjunctions. Data come from local government housing decisions in England and Wales. Northern Ireland and Scotland are omitted from consideration as their local government laws are entirely separate and can less easily be compared. The aim of this chapter is to provide evidence taken from another level of government that language is a variable for explaining policy effectiveness.

The specific question for this chapter is as follows: what effect does legislative language have on the accountability of policy delivery? Presenting a difference-in-differences (DD) analysis of a natural experiment, it is demonstrated that the adjectives ‘vulnerable’ and ‘likely’ had a strong and statistically significant effect on the delivery of homelessness policy in England after 2002. With all 22 Welsh local authorities as the control group, the 22 most similar English local authorities were identified using propensity score matching.