Implementation and First-Year Impacts of the UK Employment Retention and Advancement (ERA) Demonstration

By Richard Dorsett, Verity Campbell-Barr, Gayle Hamilton, Lesley Hoggart, Alan Marsh, Cynthia Miller, Joan Phillips, Kathryn Ray, James A. Riccio, Sarah Rich, Sandra Vegeris

Published by the UK Department for Work and Pensions

This report presents findings on the implementation and early effects of Britain’s Employment Retention and Advancement (ERA) demonstration, which is being evaluated though a large-scale randomised control trial. Aimed at helping low-income individuals sustain employment and progress in work, ERA offers a combination of job coaching and financial incentives to participants once they are working. It is administered by Jobcentre Plus in six regions: East Midlands, London, North East England, North West England, Wales, and Scotland.[1]

The initial results, covering individuals’ first 12 months in ERA, are encouraging. Despite early operational difficulties, ERA has had a number of positive effects. Across various types of people and places, it increased the receipt of services and training for working customers, increased participants’ average earnings, and produced some reductions in their benefit receipt. Still, it is much too soon to draw firm conclusions about ERA’s effectiveness. Progression in work is a gradual process that can take several years to unfold. Furthermore, many ERA customers were just entering or preparing for work during the first follow-up year, and none could have yet taken full advantage of the programme’s offer of up to two years of in-work guidance and incentives.

What Is ERA?

Launched in 2003, ERA was envisioned as a ‘next step’ in British welfare-to-work policies. A centrepiece of those policies is the New Deal programme, which offers job placement help from a Personal Adviser and other pre-employment assistance to out-of-work recipients of benefits. To the existing pre-employment New Deal services, ERA adds a new set of financial incentives and job coaching following customers’ entry into work. It is aimed at three groups that have difficulty getting and keeping full-time work or advancing to more secure and better-paid positions: (1) lone parents (mostly women) who receive Income Support and volunteer for the New Deal for Lone Parents (NDLP) programme; (2) longer-term unemployed people over age 25 (mostly men) who receive Jobseeker’s Allowance and are mandated to enter the New Deal 25 Plus (ND25+) programme; and (3) lone parents who are already working part time (16 to 29 hours a week) and are receiving Working Tax Credit (WTC).

For the two New Deal customer groups, ERA begins with up to nine months of pre-employment assistance, largely following the same procedures as the regular New Deal programme. It then continues into a unique two-year post-employment or ‘in-work’ phase. During that phase, ERA’s job coaches, known as Advancement Support Advisers (ASAs), help customers find suitable work, remain employed, and advance in their jobs. ERA also offers special cash incentives and other resources, including: an employment retention bonus; training tuition assistance and a bonus for completing training while employed; and access to emergency payments to overcome short-term barriers to staying in work. Members of the WTC group immediately enter the post-employment phase.

Qualifying members of the target groups who volunteered for the ERA study participated in a lottery-like process of random assignment. Half the volunteers were assigned to the ERA programme group and the remainder to a control group. The control group got the provisions they were normally entitled to receive from Jobcentre Plus; for the two New Deal customer groups, these included regular New Deal pre-employment services. Controls in the WTC customer group, who would not normally enter the New Deal programme, were not offered any special services or incentives but could seek other services or training on their own. For all three groups, the success of ERA is determined by comparing the programme group’s outcomes (e.g. future average earnings) with those of the control group. Because of random assignment, any statistically significant difference in these outcomes can be attributed with confidence to ERA.

Intake into the study began in October 2003, continued through the end of 2004 for most customers, and was completed for all by early 2005. Over 16,000 people were randomly assigned, making this study one of the largest randomised social policy trials ever undertaken in Britain.

Key Findings

  • Because the organisational culture of Jobcentre Plus was focused on meeting job entry targets, it was a challenging to reorient ERA staff to focus on employment retention and advancement.

Implementing the ERA programme within Jobcentre Plus offices required the districts and the Department for Work and Pensions (DWP) to address a number of critical organisational issues: (1) training frontline staff for the newly created role of ASA, whose job coaching responsibilities continue for two years after customers become employed; (2) contending with a Jobcentre Plus job entry targeting system that made it difficult, early on, for ERA staff to earn recognition for post-employment job coaching; and (3) equipping ERA supervisors to manage against new objectives focused on retention and advancement. Initially, additional resources DWP provided for staff dedicated to ERA were not used exclusively for ERA. Attention remained focused on meeting job entry targets, and too little effort was expended on ERA’s goals of sustaining and progressing in work. This prompted DWP to initiate corrective actions. It instituted stricter guidelines and accountability mechanisms for ‘ring-fencing’ ERA funds, plus new approaches to staff and management training.

  • During the first year, ERA staff informed most customers about the employment retention bonus, but they were less successful in building awareness of the training incentive and in supporting job progression.

When interviewed 12 months after entering the study, most ERA customers said they knew about the employment retention bonus. Roughly 20 per cent had already received a bonus payment (which required finding and sustaining full-time work for at least four months). Awareness and use of the incentives for combining training with work were much lower, especially among the two New Deal groups. However, most customers in the WTC group knew about the training incentive, and a sizeable share of them received the in-work training bonus. Qualitative data suggest that for many WTC customers, the training incentive was an important part of ERA’s appeal. In contrast, among the New Deal customers, finding a job was their immediate focus; this may explain why they were less likely to recall the incentive for participating in training courses while in work.

  • Despite ERA’s early implementation difficulties, ERA customers overall were more likely than the control group to get help or advice from Jobcentre Plus staff on progressing in work.

As a result of its more deliberate focus on in-work assistance, ERA increased the rates of receiving in-work help or advice by 21 percentage points for NDLP customers, and by almost 11 percentage points for ND25+ customers. For East Midlands WTC customers, who were already employed when they came to ERA, the increase generated by ERA in the proportion who got in-work help or advice was much larger. Eighty-five per cent of the programme group received it, compared with only 22 per cent of the controls — an increase of 63 percentage points.[2]

  • ERA customers — especially Working Tax Credit lone parents — were more likely than the control group to combine education or training with employment.

Among the East Midlands WTC target group, 58 per cent of the programme group combined training or education with work during the first year, compared with about 45 per cent of the control group, for a statistically significant increase of almost 14 percentage points. The absolute effects were much smaller, though still statistically significant, for the two New Deal customer groups, among whom employment rates were lower during the first year after entering the study. For the NDLP customers (with all districts combined), ERA increased the likelihood of combining work and training to 21 per cent, an increase of 5 percentage points over the control group rate. Among ND25+ customers, 11 per cent of the ERA group combined work and training, 3 percentage points more than the control group rate.

  • By the second year of operations, the six districts substantially improved their delivery of advancement services.

The evaluation’s qualitative field research, which, for this report, covers programme implementation through Spring 2006, shows that ERA operations grew stronger over time. The turning point came in mid- to late 2005 (near the end of the period covered by this report’s impact analysis). The districts’ focus on in-work services steadily improved as more participants entered work, as frontline staff and managers were trained, and as ERA resources were more carefully ring-fenced. Since these improvements took hold after the period covered by the 12-month survey for most customers, that survey could not fully capture the effects of these improvements on customers’ receipt of ERA services and incentives, or on ERA’s labour market and benefit outcomes.

  • Within a year of beginning ERA, New Deal for Lone Parents ERA customers earned substantially more than the control group, largely because they were more likely to work full time.

ERA generated positive impacts on various employment and earnings outcomes in the first 12 months after individuals’ entry into the study, especially for the NDLP customer group. When the results from all districts are combined, the NDLP ERA group, on average, earned £811 more than the £2,783 earned by the control group — a statistically significant increase. Put differently, customers in the ERA group earned 29 per cent more than they would have earned without ERA. The ERA group’s earnings were more than 20 per cent higher than the control group’s in five of the six districts (where the differences ranged from £631 to £1,386), and were particularly large and statistically significant in North West England and Scotland. ERA increased NDLP customers’ earnings by increasing the proportion of the ERA group that worked full time to 22 per cent — a statistically significant increase of 7 percentage points.

  • ERA’s first-year impacts on earnings were smaller, more mixed, and less certain for the New Deal 25 Plus customer group.

Among the long-term unemployed ND 25+ group, ERA’s earnings effect during the first 12 months of follow-up is estimated to be £291 — a 12 per cent increase over the control group average of £2,419. However, this difference is not statistically significant, meaning that there is less assurance that it is not due to chance. At the same time, the pattern of results is positive across all districts except Wales, measuring £177 to £891 per person, with the difference in the East Midlands reaching statistical significance. Included in the all-district average for the ND25+ group is a large (and statistically significant) negative impact on earnings detected in Wales. The reason for this anomalous result cannot be established with certainty. However, particular administrative problems may have caused the ERA ND25+ group in Wales to receive a delayed and weaker set of New Deal pre-employment services than they normally would have received. The negative result does not appear to stem from ERA’s distinctive in-work strategies.

  • ERA did not substantially increase first-year earnings among the Working Tax Credit lone parent group, although it did increase their likelihood of working full time.

WTC customers were already working part time when they entered ERA. In the East Midlands, nearly 28 per cent of the ERA group were working full time at the time of the 12-month survey, for a statistically significant increase of 10 percentage points over the control group’s rate. However, this increase had not yet translated into a substantial earnings gain — for reasons that are not yet clear.

  • For the two New Deal target groups, ERA caused small reductions in receipt of benefits (Income Support and Jobseeker’s Allowance).

The proportion of NDLP customers in ERA who were receiving Income Support at the time of the 12-month survey fell below the 50 per cent rate for the control group by a statistically significant 4 percentage points. Similarly for the ND25+ group, ERA reduced the proportion receiving Jobseeker’s Allowance by about 5 percentage points (a statistically significant reduction). Among East Midlands WTC customers, ERA did not produce a statistically significant reduction in the likelihood of receiving WTC.



The results of the ERA demonstration so far show that building on the New Deal and Jobcentre Plus services by adding an in-work support and financial incentives component is a feasible goal. However, it has required substantial efforts to build staff capacity to take on the new and complex role of ASA, to engage busy working customers, and to contend with a Jobcentre Plus culture that has until recently rewarded only job placement. The districts were able to deliver a considerably stronger ERA intervention as time went on. Much of that improvement occurred after the period covered by the 12-month survey, the foundation for this report’s impact analysis. Whether the encouraging early effects of the programme continue to grow as a result of those improvements will be discussed in future reports, including a report on ERA’s two-year impacts scheduled for early 2008.


[1] The Department for Work and Pensions is managing the implementation of ERA and is overseeing the evaluation. The study is being conducted by a research consortium that includes the Policy Studies Institute, the Institute for Fiscal Studies, the Office for National Statistics, and MDRC (a US-based research organisation experienced in conducting large-scale random assignment tests of new social policies).

[2] Only the East Midlands district managed to enrol an adequate number of WTC customers to support a reliable district-specific analysis for this target group. In this report, the analysis of the WTC group focuses primary on the East Midlands.

Dorsett, Richard, Verity Campbell-Barr, Gayle Hamilton, Lesley Hoggart, Alan Marsh, Cynthia Miller, Joan Phillips, Kathryn Ray, James Riccio, Sarah Rich, and Sandra Vegeris. 2007. Implementation and First-Year Impacts of the UK Employment Retention and Advancement (ERA) Demonstration. New York: MDRC.