Policing the Chain Gang : Panel Cointegration Analysis of the Stability of the Suffolk System, 1825-1858

Conventional monetary theory suggests that a closed system banking regime may lead to in-concert overexpansions of circulation by its banks. However, Selgin (2001, 2010) argues that this is unlikely as long as there are enough banks to ensure (i) routine interbank settlement and (ii) no collusion amongst banks refraining from redeeming one another's notes. Banks effectively form a “chain gang” where in-concert expansion requires coordination that is prohibitively costly in a system with many banks. In order to test this conjecture, we examine state-level data on circulations and reserves from the Suffolk Banking System (1825 to 1858) in New England. In addition to narrative evidence on the stability of the Suffolk, panel cointegration tests provide evidence of a long-run relationship between state-level circulations and total reserves. The estimated error-correction mechanisms suggest a deviation half-life of about two years. We argue that a cointegrating relationship between circulations and reserves, along with rapid error-correction, supports the Selgin hypothesis