Measuring the deadweight loss from taxation in a small open economy☆ A general method with an application to Sweden

a r t i c l e i n f o Keywords: Deadweight loss Tax policy in a small open economy The paper develops a simple general equilibrium framework for calculating the marginal deadweight loss from taxation in a small open economy. The framework allows a decomposition of the deadweight loss from each tax instrument into the losses stemming from the contraction of the different tax bases. The paper describes a method of calibrating the model which exploits the links between the various factor supply elasticities implied by the standard life cycle model. It also presents a method of estimating effective tax rates that is consistent with optimising household and firm behaviour. To illustrate how the model works, it is calibrated to a data set for Sweden. The quantitative results indicate that more than half of the marginal deadweight loss from taxes on capital may stem from their negative impact on the tax bases for labour income and consumption. Which tax structure will minimise the efficiency cost of raising the necessary revenue and achieving the desired amount of redistribution? This is a fundamental question for tax economists. To answer it we need to estimate the deadweight loss from the various types of tax. Half a century ago Arnold Harberger (1964) developed a general equilibrium approach to the calculation of deadweight loss. According to his analysis the (approximate) deadweight loss from imposing a tax on a particular good can be decomposed into the well-known " Harberger triangle " measure of the tax-induced distortion to the market for the taxed good itself and a sum of " tax interaction terms " reflecting the reduction or increase in tax distortions to other markets, as the tax on the good considered induces consumers to substitute towards or away from other taxed goods. Yet over the years numerous studies of deadweight loss have taken a partial equilibrium approach, focusing only on the Harberger triangle, despite the general equilibrium methodology advocated by the master himself. 1 Sometimes the partial approach has been motivated by the scarcity of empirical knowledge of the relevant cross price effects or by the fact that the error caused by ignoring tax interaction effects is likely to be small when the tax rates on other goods are low. But as Goulder and Williams (2003) point out, the bias from ignoring the interaction between commodity markets and the labour market will almost surely …