The vast array of policies, programmes, interventions and stated intentions described in Chapter Four is evidence of a growing acceptance across the political spectrum of a legitimate role for the state in family life. In this chapter, we assess how effective these policies have been in improving children's outcomes and life chances, especially for those who are in poverty or facing disadvantage. Policies over this period had differing goals: to reduce child poverty; to ameliorate the impact of poverty on a child's life chances; to improve social mobility; and, in some cases, to reduce inequality. We explore whether family policies were based on evidence or research, their effectiveness in improving both parent and child capabilities, and outcomes and whether they changed national patterns of opportunity and disadvantage over the 20-year period.

What role for evidence?

It is striking just how many of the policies discussed in the previous chapters drew on evidence and research from a range of different disciplines (for a discussion of different types of evidence and research, see Chapter Three). The use of data to identify, quantify and understand social problems has multiplied. However, evidence of a need is not the same as evidence of what works to address the need. There has been a search for more effective solutions to address policy challenges, especially in the context of public spending pressures. The growing interest in Whitehall and elsewhere in evidence-informed public policy is illustrated by the creation and diversity of the What Works Centres and their earlier incarnations. They collectively represent a major investment by Labour, the Liberal Democrats and the Conservatives in the evaluation of social interventions. A particular and sometimes controversial feature of the What Works approach has been the application of scientific methods to social and economic issues, with a strong emphasis on the evaluation of public policy. Randomised Control Trials (RCTs) and Quasi-Experimental Designs (QEDs) have been increasingly used to determine whether an intervention has impact. While not all policies and practices can or should be subject to testing through RCTs, this approach enables us to estimate the direct causal impact of a policy on outcomes and may offer insights into how to improve it. However, there are risks in an overreliance on RCTs in evaluating social interventions (described in greater detail later in this chapter).