Abstract
High yield gaps persist in sub‐Saharan Africa and increased fertilizer use is considered among the crucial measures to increase productivity. Despite high government investments, particularly in fertilizer subsidy programs (FSPs), overwhelming evidence has revealed their inefficiency. This study employs a multidisciplinary approach to identify the determinants of low maize yields in the Guinea savanna zone of Ghana. We conducted a socioeconomic household survey and analyzed plot‐specific soil samples. Econometric models were estimated based on both socioeconomic and soil variables. The results show that a common parasitic weed, Striga, and labile soil structure have significant effects on yield in the study region. Plot sizes were recorded both from farmers’ direct elicitation and using GPS devices. Considerable discrepancies were detected between self‐reported and GPS‐measured plot sizes. Fertilizer samples from randomly selected agro‐input shops were analyzed to control for adulterated or fake inputs. The measured nutrient contents of the samples reflected the composition indicated on the package labels. Findings underline the need of site‐specific data collection, supported by laboratory‐based soil test results, to efficiently address low productivity. Although there are no signs of fertilizer adulteration, governance challenges persist in targeting, timing, and elite capture in the distribution system of the subsidy program. The study shows that the FSP has not been an effective standalone measure. Rather, the government needs to invest in capacity building and extension services to address the site‐specific problems through comprehensive soil fertility management techniques and weed control. Promoting soil carbon management, minimum mechanical stress, crop rotation, and permanent soil cover should be further investigated as options for the region.