Government-Sourced Categorical Economic Policy Uncertainty

Abstract

This study develops a novel framework for measuring Categorical Economic Policy Uncertainty (GovCEPU) by analysing textual data from official government communications rather than traditional news media. Using Vietnam as a theoretical test case, a transition economy where centralised policy signals are paramount, we construct distinct monthly indices of monetary, fiscal, and trade policy uncertainty (2004-2023). Comparative analysis reveals that GovCEPU measures capture higher volatility and provide signals of policy shifts different from news-based counterparts, reflecting ex-ante deliberation in state-led contexts. Employing a Diebold-Yılmaz (2012, 2014) connectedness framework to account for systemic endogeneity, we demonstrate that monetary and fiscal uncertainties function as primary net shock transmitters to the macroeconomic system, whereas trade policy uncertainty operates as a net receiver. At the micro-level, heightened monetary and fiscal policy uncertainty have significant impacts on banks’ loan loss provisions and risk-taking, while trade policy uncertainty impairs non-financial firms’ performance and capital expenditures. These findings suggest that government-sourced indices offer a direct, cost-effective, and theoretically grounded alternative to media-based metrics for evaluating policy-related risks in emerging markets.