This project has two interrelated goals: 1) provide the best possible estimates of the ROI (benefit-cost ratio) for the services and programs provided by BGCA to youth; 2) identify the strengths and limitations of the estimates that can be made with currently available data, and offer ideas for obtaining more precise and robust estimates in the future.
The ROI can be thought as the multiplicative product of two components: the short-run causal effects of the BGCA activities on youth outcomes, and the longer-term economic consequences of those short-run outcomes. Both types of parameters are challenging to estimate.
To estimate the short-run effects, one strategy will be to draw from previous evaluation studies of BGCA programs and services. Evaluations with rigorous research designs (ideally randomized experiments, or at least quasi-experiments with pre-post changes as compared to a reasonable counterfactual group) will be given more weight. We will also examine evaluation studies of youth programs and services that are similar to what BGCA offers. In addition, we will examine data from BGCA?s National Youth Outcomes Initiative (NYOI), and compare outcomes to comparison groups constructed from other data sets with similar measures (e.g., YRBS).
After generating a range of plausible short-run effects from the analysis above, we will translate those effects into longer-term economic costs and benefits. In the absence of longitudinal studies that follow BGCA participants through adulthood, this translation to longer-term outcomes will necessarily require drawing from previous research that estimates how outcomes during youth (e.g., high school graduation) translate to lifetime outcomes (e.g., employment and earnings, crime, and health). Depending on the available research, we will likely use some basic modeling (e.g., Markov models) to project outcomes beyond the time frames available in other data sets. A valuable resource will be Washington State?s detailed economic analyses of youth programs and services. More generally, we will review the best practices for projecting economic benefits from child and youth interventions, and will draw from methods and parameters in other studies as appropriate. Given the large amount of uncertainty and the many assumptions that will be needed to estimate the ROI, our results will provide not only our best estimates but also a careful discussion of the estimated uncertainty surrounding the estimates. This discussion of uncertainty will yield ideas for increasing the confidence around the estimates in the future.