Employs a rigorous quantitative approach to estimate overall impacts across 235 causal studies and identify the most effective approaches to improve employment and earnings outcomes among low-income adults. 

The Office of Planning, Research & Evaluation in the Administration for Children & Families sponsored the Employment Strategies for Low-Income Adults Evidence Review (ESER)—a systematic review of the literature published from 1990 to mid-2014 on the effect of employment and training programs and strategies for low-income individuals. This brief is one in a “series of briefs [that] offers a synthesis of the findings of…ESER…for policymakers, practitioners, and officials who seek to improve the employment and earnings outcomes of low-income adults through research-based interventions ” (p.2).

“Trained reviewers examined the strength of the causal evidence for each study…then rated each study [as high, medium, or low] based on its rigor (not on the effectiveness of the intervention).…The ESER team identified a ‘primary strategy’ for each intervention[:]…the service most treatment group members received and most comparison group members did not.…The team determined the primary strategy for each intervention by having two reviewers independently read the description of each intervention, identify a primary strategy, compare their assessments, and discuss until they reached agreement” (p.2). The team only reviewed studies that used “randomized controlled trials or comparison group designs” (p.2).

“In this brief, [the authors] use a rigorous quantitative approach known as meta-regression to identify not only those interventions that seem successful on the whole, but also those that are effective for particular labor market outcomes and for particular types of low-income workers.

Because any given intervention typically comprises sev­eral employment strategies, [the authors] also examine the specific employment strategies that appear to be successful (1) overall, (2) for certain outcomes of interest, and (3) for certain types of low-income workers. Being able to iden­tify the context in which particular strategies work best can help practitioners and policymakers make targeted and informed decisions about the types of interventions that could be most effective in their contexts” (p.1).

The brief “relied on the detailed ESER database of 235 high- or moderate-rated studies of 93 interventions testing employment strategies for low-income adults” (p.2).

Full publication title: The Right Tool for the Job: A Meta-Regression of Employment Strategies’ Effects on Different Outcomes

(Abstractor: Author and Website Staff)

Major Findings & Recommendations

“Several interventions are effective at improving low-income adults’ labor market outcomes. • Although most of these effective interventions are associated with relatively small impacts, 10 are highly likely to improve outcomes by at least 5 percent. • Only one intervention causes significantly unfa¬vorable impacts. Most individual strategies, although effective, are associated with modest positive effects. No single strategy on its own is associated with substantial gains. • The individual strategies that appear most effective are financial incentives and sanctions, education, work experience, and training. Each has over a 90 percent chance of improving out¬comes across population and outcome types. • Interventions that combine several strategies to help low-income workers find and keep jobs appear more effective than any single strategy. It is easier to improve education and training outcomes than it is to increase employment or independence from public assistance. • Interventions’ impacts on education and training outcomes were larger than impacts on employment or independence from public assistance, suggesting that the latter outcomes are more difficult to improve with employment and train¬ing interventions over the time period that the original studies examined. The effect of an intervention is more than the sum of the effects of that intervention’s strate¬gies. In this context, implementation and other idiosyncratic factors become all the more crucial to our understanding of effectiveness” (p.1). (Abstractor: Author)