Political institutions can influence the likelihood of banking crises through both direct and indirect causal pathways. They may influence domestic economic conditions, thereby indirectly impacting the likelihood of a banking crisis, or they may directly affect the likelihood of banking crises through confidence and expectations-related mechanisms. I apply econometric moderated multiple-mediation to estimate this combination of effects for veto player theory - a common framework for analysing political institutional constraints - using a dynamic panel approach and a dataset of 111 developing economies and emerging markets from 1990-2012. I find more veto players indirectly reduce the likelihood of banking crises by reducing inflation and increasing GDP growth in the pre-crisis period. However, they also increase the likelihood of banking crises by increasing credit growth. When global risk is high, more veto players impede policy responses to changing conditions. This directly increases the likelihood of crises. When global risk is low, more veto players reduce policy volatility. This directly reduces the likelihood of crises. Rising global volatility has larger effect on the likelihood of crises in relatively constrained political systems.
Banking Crises Political Institutions Econometric Moderated Mediation Veto Player Theory Empirical International Finance