Governance effects on deforestation in the tropics: A review of the evidence.

• High scores for forest governance were frequently related to reduced tropical deforestation. • Revision of governance frameworks would be beneficial as some indicators were hardly applied, others were interrelated. • The structure – agency concept can help to structure the numerous governance indicators. • Governance was moderated by context factors like corruption, population and technology. • Results are based on a content analysis of 28 most frequently cited papers. In view of a continuing global decrease of forest cover, many authors mention forest governance as a basic concept contributing to reduced deforestation, specifically in the tropics. There are numerous definitions for forest governance. The concept is commonly understood as a broad approach and assessment tools comprise dozens of indicators. However, there is no information about relations between single indicators, their individual importance and whether contextual factors modify their effects. This article aims to analyze if the hypothesized relation between governance leading to reduced tropical deforestation in general holds true, to identify the most decisive governance components and to explore if the wider socio-economic and political context influences potential governance effects. The structure - agency concept is used as theoretical basis to identify underlying mechanisms and as conceptual basis for discussing individual indicators. We employ a quantitative literature review based on scientific articles on governance and deforestation. From a total of 810 articles we select the most frequently cited publications related to the subject. From the resulting 198 papers only those are studied that contain empirical relations between governance and deforestation. The remaining 28 studies are analyzed by applying the governance indicators of the World Resource Institute as categories for content analysis. Likert scores are used to quantify governance effects as input for subsequent principal component analysis and multiple linear regressions. Results show that indeed high governance scores frequently relate to lower rates of deforestation and we thus recommend continued political support for the concept. But the reviewed studies mostly focus on a smaller number of classical governance indicators, which suggests that the governance concept might benefit from streamlining into a more targeted approach. In this respect, several indicators were related to underlying principal components reflecting the agency and structure concept. Even though we cannot claim a statistically significant stronger effect of one or the other component, single structural indicators were more strongly and more consistently linked to reduced deforestation in comparison to agency related indicators. The concept can thus be helpful to guide policy design and implementation. We show this by applying it to the ongoing discussion on results based payments under REDD+ as a prominent example for agency related measures. The reviewed literature in addition suggests that governance effects are moderated by deforestation drivers such as corruption, illegal logging and population growth and by interventions like technology transfer. [ABSTRACT FROM AUTHOR]