Social policy evaluators who focus on TANF programs have long associated job accessibility with welfare usage within urban area (Ong, & Blumenberg, 1997) and welfare duration (Fitzgerald, 1995). They incorporated the notion of distance decay into their evaluation framework, but failed to take local labor market conditions into account. This failure comes from the constraint in the lack of sub-state-level geographic identifiers in these studies (Herbst & Stevens, 2010), because of confidentiality concerns. Consequently, local economic conditions, especially with regard to employment opportunities and wage level within neighborhood, cannot be fully incorporated. This study integrates spatial and non-spatial analysis to evaluate the reemployment stability among participants after exiting from workfare program, which is Multiple Employment Development Program (MEDP). We focus on geographic factor, which is job accessibility, and non-geographic factors, including welfare usage, age, gender, and human capital of participants. A micro-level dataset that contain program administrative and labor insurance information from 3,384 participants is used to identify the determinant of their reemployment stability: first, we apply the notion of distance decay (distance from program to reemployment location) and spatial competition (local unemployment status) to construct our indicator of job accessibility; secondly, we integrate this indicator with non-geographic elements to conduct factor analysis. Through this analysis, we expect, theoretically, to propose an integrated evaluation framework for workforce policy, and empirically, to understand the influential factors for reemployment stability among program participants. It is expected to shed light on how program administrator allocate resource within locality.