This research explores the welfare consequences of substituting carbon taxes for conventional distortionary taxes in a small open economy, using Pennsylvania as case study. A computable general equilibrium (CGE) model is developed for Pennsylvania to simulate possible outcomes when carbon taxes are substituted for conventional taxes. Consumer welfare impacts from environmental tax reform are conventionally decomposed into three effects: the Pigouvian effect, the tax revenue recycling effect, and the tax interaction effects. The Pigouvian effect is a welfare gain from reducing environmental externalities. The tax revenue effect is a welfare gain from reduced distortion in factor and commodity markets that results from the substitution of the environmental tax for conventional distortionary taxes. The tax interaction effect is a welfare loss due to increased distortion of factor and commodity markets induced by environmental taxes, given that distortionary taxes are not eliminated. The 'Double dividend' hypothesis argues that the sum of welfare gains is larger than the welfare loss. Most research on this debate has been done using closed-economy models. Small open economies, however, have different aspects from closed economies. These include endogenous factor mobility and trade. Factor mobility triggered by changes in real wages and environmental quality can affect distortions of regional labor market, leading to the different welfare outcomes.. When the carbon tax is imposed in open economies, inter-regional trade sectors are more responsive than foreign trades, since the elasticity of substitution for inter-regional trades is higher than that for the foreign trades. This different inter-regional trading effect leads to different welfare outcomes.