Road infrastructure development in Kenya

Road infrastructure is widely regarded as a cornerstone of rural development, with the potential to reduce transport costs, expand market access, and stimulate economic activity. However, its broader developmental effects remain contested and uneven, reflecting a fundamental tension between its role in enabling economic opportunities and its potential to intensify environmental pressures. Existing research has largely examined the political allocation of infrastructure, its economic impacts, and its environmental consequences in isolation, providing only a fragmented understanding of infrastructure-led development. This thesis addresses this gap by examining road infrastructure development in rural Kenya as a politically mediated and spatially embedded process through which transport investments reshape accessibility and generate differentiated economic and environmental outcomes. It advances three interrelated objectives: (1) to analyze how political incentives influence the spatial allocation of road infrastructure; (2) to assess how changes in accessibility affect local economic activity over time; and (3) to investigate how accessibility improvements are associated with land degradation dynamics and through which mechanisms. Conceptually, the thesis develops accessibility as the central mechanism linking infrastructure provision to downstream outcomes. Methodologically, it combines newly assembled spatial datasets, travel-time–based accessibility measures, and empirical strategies that address endogenous infrastructure placement and continuous exposure to changes in connectivity, enabling analysis in a context characterized by gradual network expansion rather than discrete interventions. The findings show that road development is not politically neutral: the allocation of paved roads reflects both developmental considerations and electoral incentives. Improvements in accessibility generate positive but modest, spatially differentiated economic gains, with stronger effects in areas with higher agricultural potential and pre-existing development conditions. At the same time, accessibility improvements are associated with increased land degradation in certain areas, primarily through land-use pressures linked to population concentration and, to a lesser extent, overgrazing. Overall, the results demonstrate that road infrastructure does not produce uniform development outcomes but instead generates uneven and sometimes competing effects across space. This thesis makes three key contributions. First, it integrates political economy, transport geography, development economics, and environmental analysis within a single empirical framework. Second, it advances accessibility as the central mechanism linking infrastructure allocation to downstream outcomes. Third, it demonstrates that infrastructure-led development is inherently uneven, generating both spatially differentiated economic gains and localized environmental trade-offs. Although the empirical analysis focuses on Kenya, the findings identify generalizable mechanisms that link political allocation, changes in accessibility, and differentiated outcomes. While the magnitude of these effects remains context-dependent, these mechanisms are likely to extend to other low- and middle-income countries undergoing rapid infrastructure expansion, with implications for infrastructure planning beyond the study context.