Sustaining Collective Action in Urban Community Gardens

Abstract

This paper presents an agent-based model that explores the conditions for ongoing participation in community gardening projects. We test the e ects of Ostrom's well-known Design Principles for collective action and use an extensive database collected in cases in Germany and two case studies in the Netherlands to validate it. The model uses the Institutional Analysis and Development (IAD) framework and integrates decision mechanisms derived from the Theory of Reasoned Action (TRA). This allows the analysis of volunteer participation in urban community gardens over time, based on the garden's institutions (Design Principles) and the volunteer's intention to join gardening. This intention is influenced by the volunteer's expectations and past experiences in the garden (TRA). We find that not all Design Principles lead to higher levels of participation but rather, participation depends on specific combinations of the Design Principles. We highlight the need to update the assumption about sanctioning in such systems: sanctioning is not always beneficial, and may be counter-productive in certain contexts.

Structural equation modeling reveals decoupling of ecological and self-perceived outcomes in a garden box social-ecological system

“…much or little quality time with friends and family, mental relaxation, and fresh vegetables. It seems that the importance of the physical resource is lower than expected 43 and the immediate human needs in this case might be fulfilled even when the ecological outcome is low. The importance of the resource for livelihood has been found to increase its sustainability 36 , which can partly explain the loss of connection here.…”

“…We first examined whether the multiple observed variables included in the study adequately measured self-perceived and ecological outcomes in the urban garden box resource system. The SES case studies rarely include more than one measure at a time, even though research would benefit from including multiple references 36 , 43 , 51 . Several measures describe more genuinely the multidimensional and complex reality of natural resource systems.…”

“…Our model fitted the data satisfactorily and all our measured variables had significant loadings onto their latent variable. In urban gardening research the typical ecological performance measure is either the species diversity or the gardeners` evaluation of their harvest, which simplifies the numerous aspects related to the harvest or yield 32 , 37 , 43 . Socio-economic performance has been measured for example, by self-organization, livelihood, gender equity or different socioeconomic factors 18 , 34 , 36 .…”

“…We expect a positive association between outcomes and variables related to the resource due to higher resource amount (box number), higher motivation and knowledge (effort), higher productivity (shade), and lower risks outside urban areas (privacy) 36 , 38 , 40 – 42 . We expect the variables from the governance sub-system to have a positive influence on outcomes due to higher motivation and shared strategy (rules), trust in everyone`s participation and shared knowledge (involvement) 2 , 36 , 43 , 44 . Variables related to the actors are expected to have a positive effect on outcomes due to social benefits and aggregate contributions (group size, others number), reciprocity and trust (social capital), and a negative effect due to higher cost of self-organizing (group size), lack of knowledge and experience (starting year), negative experiences (damage) and lack of control on the resource state (worries) 2 , 3 , 41 , 43 , 44 (for more details consult Table 1 ).…”

“…We expect the variables from the governance sub-system to have a positive influence on outcomes due to higher motivation and shared strategy (rules), trust in everyone`s participation and shared knowledge (involvement) 2 , 36 , 43 , 44 . Variables related to the actors are expected to have a positive effect on outcomes due to social benefits and aggregate contributions (group size, others number), reciprocity and trust (social capital), and a negative effect due to higher cost of self-organizing (group size), lack of knowledge and experience (starting year), negative experiences (damage) and lack of control on the resource state (worries) 2 , 3 , 41 , 43 , 44 (for more details consult Table 1 ). The variables have not been studied jointly before in this context and therefore we have no clear prediction on how their interrelations will influence the results.…”