Conclusion
Our review illustrates that forestry decentralization has gained the analytical attention of a variety of academic disciplines, resulting in a complex and varied body of literature. It also shows a tendency of compartmentalization, whereby the findings from one strand do not necessarily inform or draw on those of other strands. For instance, while studies have illustrated that decreasing forest cover is not necessarily an indication of unsustainable management [22], other studies use such indicators to examine relative sustainability of different management regimes. And while studies have shown that areas designated as under decentralized forestry may have highly centralized government management in practice, other studies will take such labels at face value. This is largely a consequence of different disciplinary and individual assumptions and preferences for study design and approach to data analysis that, in turn, shape empirical enquiry and, ultimately, what are presented as ‘findings’. Thus, economists and political and environmental scientists favoring large-N studies and associated statistical analytical approaches face a trade-off between reducing complexity and, thereby, introducing ambiguity, and attaining the ostensibly needed number of observations. Although we believe that such endeavors may bring important and unexpected findings, we believe that they must be complemented by grounded case-based empirical studies. Likewise, we argue for continued attention among authors of such grounded empirical studies to how their idiosyncratic findings may or may not reflect more general phenomena across time and space. Thus, as this review has illustrated, the four strands present differing and partial representations of forestry decentralization policies. Accordingly, reviews that exclude large parts of the literature through restrictive inclusion criteria, such as recent systematic reviews of ‘evidence-based’ studies [20 ,76], run the risk of reproducing and entrenching such partial representations. Rather, we argue, a comprehensive understanding of forestry decentralization policies should be informed by all of them.
