Elasticiteten af skattepligtig arbejdsindkomst

This paper examines how Danish income tax changes since 1994 have impacted on taxable earned income. Results are presented from two competing ap - proaches to measuring the elasticity of taxable income. Until recently, the differencesin- differences model was the most common approach, but the method is not precise about whether short run or long run responses are being measured. This is problema tic, as behavioural adjustments to tax changes may take several years to complete. The problem is overcome with an error correction model, which explicitly identifies separate short and long run responses. The main results from the differences-in-differences model suggest that taxable earned income increases by 0.06 per cent for females and 0.11 for males in response to a 1 per cent reduction in the marginal tax rate (the uncompensated elasticity). The income elasticity is -0.013 for females and -0.017 for males. These results are broadly consistent with other results for Denmark. How ever, the results from the error correction model indicate that the short run elasticities are somewhat smaller and the long run elasticities considerably larger than those ob tained by the differencesin- differences model. This lends support to recent criticism that methods with a short run focus underestimate tax elasticities by ignoring slug gishness in the behavioural adjustment process. The results have important implica tions for considerations regarding optimal taxation policy. The paper also points to inherent methodological problems with the differences-in-differences model with overlapping differences that, in all like - lihood, will lead to underestimated long run elasticities.