Determinants of branching decisions of State-controlled commercial banks: evidence from China

PurposeBranching is one of the crucial strategic non-price actions for banks. Previous studies on the impact of state ownership upon banks focus on bank lending behavior. This paper aims to offer a novel investigation of how state ownership affects bank branching behavior by examining state-controlled commercial banks (SCCBs) in the context of the largest developing and transitional country China.Design/methodology/approachThe two-part model (TPM) is applied to analyze the branching decision process. In the first stage, the dependent variable is the choice of bank branching dynamics and in the second stage the dependent variable is the number of new branches or the number of closed branches. For robustness check, the ordered probit selection model allowing for interdependence of the two stage decisions is also employed.FindingsUsing a unique dataset of bank branches in China, this paper finds that the branching decisions of Chinese SCCBs are driven by both profit motivated factors including population size, population density, income level, financial development and banking competition and politically motivated factors as represented with the proportion of SOEs. As a comparison, branching decisions of joint-stock banks in China are fully determined by profit motivated factors.Originality/valueFirst, this study is the first to explore the effect of state ownership on bank branching decisions, providing a new insight on the literature regarding to the impact of state ownership on bank decisions. Second, this study explores the potential effect of politically motivated factors on bank branching decisions, filling the gap in bank branching literature. Third, this study can contribute to bank branching literature by enriching the limited understanding of how SCCBs make branching decisions. Lastly, this study applies novel empirical strategies to analyze bank branching decisions, including the TPM and the ordered probit selection model.