A Discourse on Impact Evaluation

Abstract

Evaluation of the impact of a social intervention is complex. A social intervention cannot be designed as an invariant input for un-mixed and singular results. Agents and conditions of implementation confound outcomes. But evaluators of impact in search of certainties have often focused on direct and more immediate outputs. When intermediate and long-term outcomes are studied, the focus is on the concrete and the measurable. The discourse on impact evaluation presented here includes a model and methodology for the evaluation of impact. The model as elaborated posits three categories of impact: (1) impact by design– outputs resulting directly and immediately from the intervention; (2) impact by interaction– outcomes arising from interactivity with concurrent interventions by other agents; and (3) impact by emergence– unimagined outcomes emerging from the original intervention through its interactions with other interventions and its interfaces with historical and cultural processes in place but not easily discernable. The methodology proposed for the study of impact is pluralistic. The model and methodology are confirmed in the context of the Ghana Literacy and Functional Skills Programme (LFSP) at the end of its first phase, 1992–7, showing how the comprehensive conceptualization of impact, together with methodological pluralism, can deliver rich data on the impact of interventions which can serve the needs of both policymakers and the public.