Coalition Development for Drug Prevention in Mexico: Longitudinal Evaluation of Coalition Functioning

Introduction: Illegal drug use rose 87% from 2002 to 2011 in Mexico. To help address this problem, 19 substance use prevention coalitions were organized across Mexico. This study evaluates changes in coalition functioning through their first four years of functioning. Initial coalition funding was reduced after the second year and ended completely after the third year, along with intensive training and technical assistance. This study provides insight into how coalition functioning develops initially and how it evolves as external support is withdrawn.

Method: Coalition members completed coalition functioning self-assessment surveys at 6 months (253 members from 19 coalitions), 18 months (177 members from 19 coalitions), 3 years (79 members from 7 coalitions) and 4 years (81 members from 7 coalitions). Measures of coalition functioning under investigation were organized into three categories – member engagement (role involvement and time investment), process competence (participatory leadership style and coordinator skills), and outcomes (sustainability planning, community support, and community improvement). We tested for changes in coalition functioning constructs across time using paired t-tests and estimated Cohen’s d effect sizes for the changes.

Results: Coalition functioning improved significantly as the coalitions progressed from 6 to 18 months. In particular, role involvement increased from a mean of 1.49 to 1.73 (Cohen’s d = .75). Time investment increased from 29.55 to 45.19 hours per month (Cohen’s d = .74). Participatory leadership style and coordinator skill both increased significantly (Cohen’s d = .97 and .80 respectively). Coalitions did not have significant changes in coalition functioning from 18 months to 3 years but had several from year 3 to year 4. In particular, role involvement and time investment decreased (Cohen’s d = .93 and 1.72 respectively) while coordinator skill and community support increased (Cohen’s d = .89 and .92 respectively).

Conclusions: Results suggest that among the substance use prevention coalitions in Mexico, coalition functioning initially increased during their first 18 months of operation and then plateaued. In their fourth year of functioning, after startup funding and intensive technical assistance terminated, member engagement declined while coordinator skill and community support increased. Taken together, these results suggest the coalitions successfully developed high levels of coalition functioning but began to struggle with member engagement after facing funding declines. Future large-scale implementations of substance use prevention coalitions may want to consider more gradual reductions in financial support, training, and technical assistance.