A sector-specific training program called WorkAdvance US increased the average yearly earnings of underemployed workers by almost $2,000. The program is marked out by its communication with local industries to identify skills that are in high demand. Specific training is provided, with an emphasis on career building. The program was tested in a randomised control trial across four sites in the US. Increases were seen in training attendance and completion, as well as job earnings, hours and benefits—though there was substantial variation between sites.
Results & Impact
WorkAdvance was tested in a randomised control trial. Between 2011 and 2013, roughly 2,500 unemployed or low-wage workers were enrolled and randomly assigned to two groups—one with WorkAdvance training and one without it. After two years, participants with WorkAdvance training were on average earning $1,969 a year more. Specifically comparing workers who had been unemployed for a long time at the outset - a group that tends to struggle to find work - the earnings difference was $2,347. Notably, the gains were not simply from working more but also from higher earnings. Jobs held by participants also offered more hours, better benefits and greater prospects for advancement.
New York City Mayor's Office for Economic Opportunity, Corporation for National and Community Service’s Social Innovation Fund (SIF), MDRC, Madison Strategies Group, Per Scholas, St. Nicks Alliance, Towards Employment, private funders
The basic WorkAdvance program was designed by the NYC Centre for Economic Opportunity and MDRC. Implementation was outsourced to four non-profits: Per Scholas and St. Nicks Alliance in New York City, Towards Employment in Cleveland, Ohio, and Madison Strategies Group in Tulsa, Oklahoma. Each non-profit started by making contact with one or two local industry groups to find out where the skills gap was. They modified their training curricula according to what was needed. Then unemployed and low-wage workers were screened to ensure they had an interest in the sector, the capability to complete training and the potential to meet the employer’s needs. They then received an average of 3 to 4 months of sector-specific training as well as CV and interview preparation. Finally participants were helped with job placements, and they were offered support and advice on career planning for two years after enrolment.
Tulsa, Oklahoma, Cleveland, Ohio, two sites in New York City
Low-income people, unemployed people
Cost & Value
$5,200 to $6,700 per worker—much more than running a job placement operation. In total, $15 million for the four sites over five years.
Running since 2011
Even once the basic WorkAdvance program had been designed, it took time and a substantial amount of technical assistance for the non-profit providers to translate the WorkAdvance model into a set of concrete services appropriate for their local job market. But each site faced its own challenges. In Tulsa, for example, it required a substantial change in mindset, moving from case management - getting someone in and out the door - to career building. In some instances, the benefits of the program only appeared towards the end of the two-year trial, so the results may even underestimate the true benefits. That said, one of the four trials was significantly less successful and failed to increase earnings. This was due to the vagaries of the market: they trained people for an economic need that vanished when construction in New York dried up. This highlights the need for foresight when choosing a sector, but also points to a certain amount of risk in training for the skills gap of the moment—something which is liable to change. However, if the training providers are in continual communication with employers, they should often be nimble enough to anticipate and change according to trends in the labour market.
The pilot took place in four sites simultaneously. In each case the program was modified to suit the needs of local employers, and there was significant variation in outcomes between sites. The program has continued in some form at three of the four sites and is now in its seventh year.
A five-year study found that participants enrolled in WorkAdvance, a sector-specific training program for the underemployed, were after two years earning on average almost $2,000 a year more than those who were not.
This figure rose to $2,347 for those who had been unemployed for a long time at the outset—a group that tends to struggle to find work at all.
Aside from increasing earnings, WorkAdvance training increased employment in the targeted sector everywhere it was tested by between 12 and 40%. It varied from site to site, but jobs held by WorkAdvance participants often offered more hours, better benefits and made workers feel as if there were genuine prospects for advancement.
Altogether, the results, gathered from a randomised control trial of roughly 2,500 participants, made a strong case for investing in sector-specific training.
WorkAdvance US was designed by MDRC, a New York-based social policy research group, and the New York City Mayor’s Office for Economic Opportunity in a bid to improve on the usual work placement services for the unemployed. At its heart was the idea that if you start with the economic need – where are workers required – and train people accordingly, you will be providing better jobs with genuine opportunities for career building.
Target sectors were identified on a local basis and non-profit providers were used to implement WorkAdvance. These providers developed relationships with local employers within targeted industries to figure out what kind of skills were in demand in those occupations and might pave the way to a career path of rising wages. They then tailored their programs according to employer needs and leveraged the relationship in order to ‘open the door’ to program participants.
Although the program was site-specific, there were four general components. First, participants were selected on the basis of an intensive screening procedure to ensure they were interested, capable and committed. They then received appropriate pre-employment and career readiness services. This consisted of an orientation to the sector, individualised career coaching and learning how to search for work in the sector. After that they underwent occupational skills training geared towards job openings, and received certificates marking their qualifications. Crucially, the training did not stop once they found a job—providers maintained close contact with participants, assessing their performance, helping them move up the career ladder and find new work if they lost their job. This support continued for up to two years.
This kind of program is far from cheap. WorkAdvance cost between $5,200 to $6,700 per worker and in total $15 million to operate across all fours sites over five years.
However, this should be weighed against the average earnings gains made by the workers, which, if they stayed as they were, would in the space of a little over three years exceed than the cost of their training. Furthermore, over a third of those who entered the program were, for instance, receiving food stamps, which they would not need if they had jobs and/or earned more. And a quarter of the workers enrolled had a criminal record; given that having work with prospects reduces the likelihood of a return to prison, savings would be made on the tens of thousands of dollars a year it costs to keep someone in prison.
Nonetheless the program was still a sizeable investment, and the designers sought to test its effectiveness in a definitive way.
Four pilot studies took part in a two-year randomised control trial with over 2,500 participants. Roughly half were selected to receive WorkAdvance training, and the rest were left to their own devices, able to engage in any other work training on their own initiative. The average participant was 34 years old, and the majority were male (73%) and single (67%). Only one in five were working when they started, and they had to be earning less than $15 per hour. One third had been unemployed for at least seven months. They did not skim the best for their trial: this was a realistic, challenging cohort.
Four non-profits implemented the program in three cities. Per Scholas and St. Nicks Alliance, both in New York City, targeted information technology and environmental remediation respectively. Madison Strategies Group in Tulsa, Oklahoma, focused on transportation and, later, manufacturing. Towards Employment in Cleveland, Ohio, targeted healthcare and manufacturing.
The results were strikingly positive, with one exception. The program at St. Nicks Alliance in New York failed to increase earnings. They were training for asbestos removal and other environmental remediation tasks, and they were caught off guard by the abrupt slowdown in construction in New York.
This highlights two things. For one, the need for foresight at the stage of choosing the sector for training. But it also underlines the intrinsic risk of training according to the skills shortages of the moment—something which is liable to change over the duration of the program.
This volatility could also be seen in the variation of the positive results. While the average earnings increase was about $2,000 in year two of the program, the I.T. training provided by Per Scholas had bumped workers’ earnings up an average of $3,746 by then.
To some extent this volatility can be anticipated and dealt with if providers maintain continual contact with employers, getting intelligence on trends within the labour market that allow them alter their training in response.
Nonetheless, targeting sectors like I.T. that are, in the bigger picture, on the rise, will provide valuable training. This showcases what WorkAdvance and sector-specific training in general can do: lead to not just a job, but a career.
The WorkAdvance program has continued in some form at three of the four sites. The randomised control trial added strong evidence to the growing consensus about the effectiveness of more expensive, sector-specific training for the underemployed.