Assessing Collaborative Conservation: A Case Survey of Output, Outcome, and Impact Measures Used in the Empirical Literature

Abstract

Much existing research on collaborative conservation has focused on process, even as researchers have called for greater attention to explaining what results these processes yield. It is time to take stock of collaborative conservation research by mapping what kinds of variables researchers are including in analyses. Here we conduct a case survey from the SCAPE database of environmental decision-making cases. We include cases involving collaboration across government, environmental protection, and resource exploitation interests in western democratic countries. Results reveal patterns in what researchers include in their outputs, outcomes, and impacts measures of collaborative conservation. While there is little difference by publication type (peer-reviewed journals, scholarly book chapters, or gray literature) or over time, we find significant differences in explicit measures across variable types. In particular, variables more proximate to process in a logic chain are more often measured, as are social rather than ecological variables.

Keywords:

• Case survey
• collaborative conservation
• logic chain
• measures
• variables

Notes

Acknowledgments

The authors wish to thank members of the SCAPE database team including Edward Challies, Elisa Kochskämper, Ana Adzersen as well as 24 student assistants who helped code the case studies.

Notes

1 29 of 32 models did not have a proper model fit, where the null model could not be rejected with model Chi-square showing insignificant values.

2 Model fit (Chi-square) for output: process goals attained for environmental conservation was less than p=.05, while output: process goal attainment for natural resource use (p=.055) and output: higher order policy for environmental conservation (p=.051) are slightly above the .05 significance level.

3 By means of the Kaiser-Meyer-Olkin measure, we verified sampling adequacy as ‘middling’ (KMO = 0.74). Bartlett’s test of sphericity indicated that correlations between variables were sufficiently large (Chi-square (253) = 3588.085, p < .001).

Additional information

Funding

Funding for this research was provided by the German Research Foundation (DFG) under grant number NE 1207/2–1 “ECOPAG” and by the European Research Council (ERC) through Starting Grant number 263859 “EDGE” to JN.