Policy to date has largely focused on reattaching people to the labour market, making less provision for assisting people once in work. The Employment Retention and Advancement (ERA) Demonstration is testing a new strategy for improving job retention and advancement of individuals who are either already employed in low paying jobs or who have newly entered work from benefits. The scheme is directed at those eligible for New Deal 25-Plus, those volunteering for New Deal for Lone Parents and those Lone Parents on Working Tax Credit working part-time. ERA offers both pre- and post-employment assistance to encourage full-time stable employment as well as participation in advancement-focussed training while in work. Specifically, ERA offers two years of work-related services (one-to-one support from a dedicated adviser) combined with financial incentives for participants who work full-time (Retention bonus) and for those who complete training while in work (Training bonus). The ERA demonstration also promises to make a significant contribution to the process of evidence-based policymaking in Britain by testing the effectiveness of the intervention as a large-scale, multi-site, random assignment social experiment.

The IFS is part of the consortium in charge of the evaluation of the demonstration, the other members being the US MDRC; the Policy Studies Institute (PSI) and the Social Survey Division of the Office for National Statistics (ONS).

The overall aim of the evaluation is to assess whether the new measures cause improvements in employment stability and advancement above and beyond what would have occurred in their absence, and how any economic benefits they generate compare to their costs.

The IFS will contribute to the cost-benefit study and to the impact study, in particular to the non-experimental component of the latter. In fact, even though the randomisation ensures that participants and controls have the same characteristics at baseline, if ERA has any effect on job-finding or job-retention probabilities, the sub-sample of employed participants will have different observed and unobserved characteristics from the employed control group members. The experimental design cannot thus be directly used to assess the impact of ERA on those outcomes which are only defined conditional on being in work or on having found a job (e.g. wages, non-pecuniary work-related benefits, duration in a job or wage growth); instead, suitable non-experimental methods need to be applied to remove the potential selection bias.