Modelling the economy-wide effects of unilateral CO2 pricing under different revenue recycling schemes in Austria - Identifying structural model uncertainties

Abstract

Macroeconomic modelling is essential for assessing the impacts of carbon pricing, but significant uncertainties remain between modelling approaches. This paper examines structural uncertainties by comparing a Neoclassical computable general equilibrium model (WEGDYN-AT) and a New Keynesian model (DYNK). We qualitatively and quantitatively analyze the effects of non-ETS carbon pricing under different revenue recycling options in Austria. We identify six causal impact chains (ICs) that shape model outcomes. The first two - carbon pricing leading to a loss in economic output (IC1) and a shift towards labor-intensive sectors (IC2) - are common to both models. However, differences in economic paradigms emerge in the markets for labor (IC3), capital (IC4), and goods and services (IC5). For example, DYNK shows stronger transmission of external price shocks but smoother labor market adjustments, while WEGDYN-AT shows the opposite pattern. Structural differences from models’ historical development, such as the modelling of private consumption (part of IC5) and government budget closure (IC6), also contribute to divergence. Quantitative simulations show that, due to these structural differences, results for key indicators, such as price indices, consumption, and welfare, can differ in both magnitude and sign. Our results highlight the importance of tailoring policy recommendations to the prevailing economic context, such as whether the economy is experiencing an output gap with idle resources (DYNK) or a boom phase with scarce resources (WEGDYN-AT). Transparent documentation of impact chains is crucial to understanding the range of macroeconomic effects of carbon pricing, identifying more robust policy outcomes, and strengthening policy support.