All publications
-
Evaluating the Legacy of Raila Odinga’s Influence on Local Governance Structures in Kenya
Suleiman Ibrahim Roba, Maxwell Muthini Kyalo
Kenya has gone a long way in democratizing and decentralizing governance. Raila Odinga has played a key role in the reformist politics in Kenya. Nonetheless, there exist very little academic sources that evaluate the impact of his political leadership in influencing the governance systems in Kenya. The discrepancy exists on whether his influence has brought structural and behavioral change that has strengthened the local institutions or it is actually just symbolic in the wider politics of Kenya. In order to seal this knowledge gap, this paper assesses the legacy of the influence of Raila Odinga on local governance structures (LGS) in Kenya. To be precise, the current study is interested in fulfilling two particular goals: i) To determine the effects of constitutional reforms conducted by Raila Odinga on devolution in Kenya, and ii) To identify the sustainability and long-term effects of the Raila Odinga legacy on the devolved system of governance in Kenya. The study uses a mixed-methods design that integrates data from secondary materials with qualitative evidence from 48 key informant interviews (KIIs). Secondary materials comprised journals, books, official policy documents and legislation, and online sources and archival media sources. Purposive sampling was used in selecting participants in the semi-structured KIIs. The participants included national and county officials, party actors, civil-society leaders, policy experts, and community representatives. Data analysis for the study involved content analysis of the gathered data. The analysed data were presented in narrative form. The study found that Raila Odinga acted as an effective institutional entrepreneur (EI). His coalition-building and framing helped secure broad popular legitimacy for the 2010 Constitution, and his political sponsorship and advocacy contributed materially to the constitutional architecture that created 47 county governments, embedded a constitutionally guaranteed minimum county share (15%). Simultaneously, the research reveals that such gains in Kenya are conditional in terms of their durability. Favorable improvements include the broadened political representation, consistent fiscal rights, and the empowered political agency on the local level. Yet, some of the issues that have remained are fiscal dependence and limited own-source revenues, lack of capacity in planning and budget implementation, cases of elite capture, and constant intergovernmental friction that deters service delivery. The legacy of Odinga is both important and the prerequisite to the existence of devolution, yet it does not consistently precondition positive long-term results. The institutionalization of the reforms he aided produced the environment of long-term decentralization. However, it will be realised only to its full extent in case of buttressed administrative capacity, financial sustainability, well-established accountability procedures, and long-term cross-party political will.
PDF -
A Stacking-Based Heterogeneous Ensemble Model for Customer Churn Prediction: Synergistic Integration of LightGBM and AdaBoost
Ziyu Zeng, Maoxin Li, Yuyi Huang, Zhaohong Cao, Cunjue Li
Customer churn prediction, typically framed as an imbalanced binary classification problem, poses significant challenges to traditional machine learning models and single ensemble methods, which often suffer from limitations in both predictive accuracy and model interpretability. To address these issues, this paper proposes a heterogeneous ensemble learning framework based on Stacking, which integrates LightGBM and AdaBoost as base learners to leverage their complementary strengths in computational efficiency and classification performance. The proposed model employs a five-fold cross-validation strategy to generate meta-features, thereby enhancing generalization capability. Experimental results demonstrate that the Stacking model achieves an AUC of 0.9132, representing a substantial improvement of 11.45% over standalone LightGBM and 8.69% over AdaBoost. Moreover, the model attains a recall rate of 0.9388, effectively aligning with the business priority of minimizing customer churn through high sensitivity. The innovation of this study lies in three aspects: (1) the design of a heterogeneous ensemble architecture that facilitates performance synergy; (2) the use of cross-validation for robust meta-feature generation; and (3) the incorporation of feature importance analysis to enhance model interpretability. The findings validate the effectiveness of the Stacking ensemble in customer churn prediction and provide both theoretical insights and practical guidance for developing intelligent customer relationship management systems.
PDF