hear how older adults experienced the social unrest and how they thought about improving their community. Ten focus groups were conducted with 73 participants. Focus group recordings were transcribed; and three coders completed a thematic analysis and member checking. Participants averaged 75 years of age, 73% were female, and 59% were African American with the remainder White. Eight themes were identified. Issues related to safety were most commonly discussed. Some focus group members

more »

... ted in the protests but left before nightfall because they felt vulnerable. Concerns about going out after dark remain strong. Some felt intimidated to attend community meetings because of the angry nature of the discussion. There was recognition of ongoing racism and long-standing problems in the community. Participants expressed concern and understanding toward the challenging situations faced by today's youth, including lack of opportunity in education and employment and lack of strong parental and community support. Participants reported a breakdown in intergenerational communications and expressed a desire for more exchange. Findings are being discussed with relevant organizations and public officials to increase the involvement of older adults in on-going community development efforts and to provide opportunities for intergenerational dialogue. Formal volunteering in later life has been shown to improve health. However, few studies have examined how volunteering can help to buffer health during life-transitions (e.g., relocating from one residence to another). Relocation in later life, at times, is unavoidable. Older adults may move by choice or due to changes in health, widowhood, financial circumstances, etc. Although health outcomes of relocation are mixed, we ask: can volunteering buffer one's health for those who relocate? We utilized 2008 and 2010 Health and Retirement Study and included individuals who were aged 50+, did not volunteered at baseline, and relocated between 2008 and 2010 (N=835). With volunteer engagement as the "treatment," propensity score weighting was performed stepby-step: 1) To estimate the propensity score, the Generalized Boosted Regression (GBR) model was run with a comprehensive set of covariates at the baseline (socio-demographic, health, social network, etc.); 2) Using the propensity score, two types of propensity score weights were calculated: the average treatment effect (ATE), and the treatment effect for the treated (ATT); and 3) To examine the effect of the "treatment," volunteering, on health outcomes in 2010, regression models were performed with ATE and ATT weights. Results suggested engaging in volunteering increased self-report health status; and reduced number of ADL and IADL difficulties. Volunteering did not influence depressive symptoms. Implications for policy and practices intersecting relocation and volunteering will be discussed. Future research should examine the role of specific mechanisms of volunteering (e.g. social engagement, psychological motivations, neighborhood friendliness) and their impacts on health.