Innovative research in behavioral science has demonstrated that human decision-making is frequently imperfect and imprecise. We procrastinate, get overwhelmed by choices, lose our self-control and permit small changes in our environment to influence our decisions. However, many programs are designed with the implicit assumption that clients and staff will carefully consider options, analyze details, and make decisions that maximize their well-being. As a result of imperfect decision-making, desired goals may not always be achieved. Emerging research in the behavioral sciences has found that minor details, such as the presence of a default condition, can have significant effects on altering behavior and individual outcomes. Yet despite the promise such approaches have for improving the effectiveness of public social programs, scant attention has been paid by program operators, social welfare academic researchers, and evaluators to behavioral improvements and the opportunities to increase the effectiveness of social programs through approaches grounded in behavioral theory. This panel presents evidence from three randomized controlled trials from the Behavioral Interventions to Advance Self-Sufficiency (BIAS) project, sponsored by the Administration for Children and Families (ACF). Each pilot employed behavioral diagnosis and design to understand the program’s process, identify possible bottlenecks, and develop behavioral interventions geared to tackle one or more of the identified bottlenecks. Using this approach, one pilot program developed a behaviorally-informed mailing to induce incarcerated noncustodial parents to file an application for a modification of their child support order. Another pilot altered invoices in order to overcome barriers among noncustodial parents that may result in their falling behind on their child support payments. The third pilot created mailers and other outreach strategies that employ behavioral concepts to encourage low-income participants to work in order to qualify for an earnings supplement that is modeled after the federal Earned Income Tax Credit (EITC). Each presentation will discuss the dynamic process at the core of behavioral diagnosis and design, which seeks to understand program processes from the client’s point of view. This can lead to new perspectives on reasons programs are not achieving their intended outcomes. Understanding these “bottleneck” points allows program designers to account for behavioral reasons clients may not respond to a program and may increase the likelihood that interventions are well designed. The presentations will show participants how to identify behavioral bottlenecks, connect those bottlenecks to their possible underlying psychological causes, and design low-cost behavioral interventions to address both. Findings will also be presented where available.