Signaling Resilience: A Computational Assessment of Narratives in Local Government Budgets

Abstract

Abstract Local governments consider a wide range of policies to increase resilience in the face of myriad risks and employ a variety of tactics to communicate about these policies to external actors. An important platform to signal resilience as a policy priority is through the budget process wherein local communities decide “who gets what, when, and how.” Using computational text mining techniques, we assess how county governments in California signal efforts toward resilience in their budgets during the 2012–2017 fiscal years, as well as whether and how those signals are received by the capital market. Comparable budget documents are available for 38 counties across the state for a total of 161 county-year observations. To test the relationship between local government resilience signals and capital market outcomes, we focus on county underlying credit ratings issued by counties. Empirical results show that county underlying credit ratings are insensitive to resilience signals in local government budgets. By examining the efficacy of resilience signals and their effects on the capital market, we offer evidence on the link between policy signaling and financial outcomes at the local government level.