Evaluating cropping intensity in registered and unregistered farms: Evidence from Kenya

Abstract

Sustainable agricultural intensification strategies are crucial in increasing Africa's agricultural productivity. Globally, literature shows strong linkages between land tenure security and sustainable agriculture. However, in Sub-Saharan Africa (SSA), findings remain inconsistent. Studies attribute this inconsistency to the use of cost-benefit and yield methodologies that are limited in capturing the nuance of subsistence farming and spatial and temporal distribution. Thus, this study seeks to fill this gap using Tharaka Nithi County in Kenya by employing remote-sensing and spatial econometrics to evaluate this relationship. The study uses the Normalised Difference Vegetation Index (NDVI) as the dependent variable and a proxy for cropping intensity, which similar studies have not used. NDVI datasets covered seven years (2014–2020) acquired from Copernicus Global Land Service. Explanatory factors include land registration status, topography and climatic characteristics. The study utilises a three-step analysis, which includes exploratory regression for model selection, global regression using Ordinary Least Squares (OLS), and localised regression using Geographically Weighted Regression (GWR). Spatial autocorrelation tests on OLS residuals using Moran's Index indicated spatial heterogeneity; hence, the study ran GWR. GWR performed better with an adjusted R2 value of 0.941, compared to OLS with an R2 value of 0.807. Mapping the GWR model illustrated that NDVI is lowest in unregistered land in the semi-arid zone of the County (0.012608) and highest in registered areas (0.883172). A hot-spot analysis using the Gettis-Ord Gi* Statistic validated the findings. The test showed cold spots in unregistered land in the semi-arid zone and hot spots in registered land, which was significant at 99 % confidence. The findings shows that the relationship between land registration and cropping intensity is highly significant, particularly in arid and semi-arid lands (ASALs), an indicator that land tenure security promotes sustainable agriculture in Kenya. The results enrich policy discussions aimed at land reforms and sustainable agriculture. The study recommends fast-tracking land registration prioritising ASALs, implementing the national land information system, and modernisation of land registries in Kenya to improve public access to land information. The study also recommends similar studies on various agroecological zones to supplement the findings, mainly where subsistence farming is dominant. The findings will add a critical layer of information in decision-making for sustainable agriculture and land reform policies.