Abstract Agricultural carbon projects are increasingly promoted as instruments to address climate change while supporting sustainable agriculture. These projects combine sustainable land management (SLM) practices with a carbon credit component, creating complex governance structures involving diverse actors with unequal power. Understanding the governance challenges that may impede their effective implementation is therefore essential. Despite growing evidence on the potential of digital tools in agriculture, their role in agricultural carbon projects remains underexplored. This study employs a qualitative case study of two pioneering carbon projects in Kenya, alongside a participatory and visual mapping tool (Process Net-Map), to engage with stakeholders. It combines concepts of principal-agent and bargaining power theory to analyse the governance challenges of agricultural carbon projects, and the potential role of digital tools in addressing them. The findings reveal layered principal-agent relationships in the carbon credit component, characterized by strict monitoring and external accountability, exacerbating information asymmetries and shifting performance risks to actors with limited bargaining power. While women play a pivotal role in implementation and monitoring, intra-household power relations constrain their control over assets and benefits, thereby reinforcing gender inequities. Digital tools currently support data collection and reporting, with potential to reduce transaction costs and improve accountability, but their use remains largely confined to the SLM component. Expanding digital tools beyond monitoring to support participation, transparency, and feedback could strengthen smallholder bargaining power. The current study contributes to literature by highlighting how the carbon credit component reshapes and intensifies existing SLM governance challenges and offers insights for project developers and policymakers seeking to promote more equitable and effective agricultural carbon initiatives.
Mantey et al. (Tue,) studied this question.