Class 13 Question 2 Estimating Taxable Income Responses Using Danish Tax Reforms Kleven and Schultz (2014)

Transcription

1 Class 13 Question 2 Estimating Taxable Income Responses Using Danish Tax Reforms Kleven and Schultz (2014)

2 Outline: 1) Background Information 2) Advantages of Danish Data 3) Empirical Strategy 4) Key Findings

3 Background Information: Data: Full Danish Population since 1980 Detailed Tax Return Information Large set of socioeconomic variables Danish Tax System: Personal income tax treats different income forms in a partially separate fashion, in contrast to a standard tax system which a progressive rate structure on a single measure of taxable income Labour Income (LI) Capital Income (CI) Personal Income (PI) Different tax bases with different rates. Reforms: Change definition and the rates

4 Background Information: Danish Tax System: Summary of the changes: 1987 reform, 1994 reform, 1999 reform and 2004 reform Lowering of marginal tax rates in each bracket Substantial broadening of tax base Focus: 1987 reform Labour and Capital Income

5 Previous Literature: Focused on the US tax reforms in the 1980s and 1990s Tax changes at the top of the income distribution (Feldstein, 1995; Gruber and Saez 2002) Problem (Generic): Identifying variation: Tax cuts to the top which are correlated with non-tax factors driving top incomes and tax data alone does not offer direct controls for these nontax factors.

6 Previous Literature: Identification Problems: 1. Bias from non-tax changes in inequality Hard to disentangle tax driven changes in top incomes from changes driven by nontax factors. Saez (2016)

7 Previous Literature: Identification Problems: 1. Bias from non-tax changes in inequality Hard to disentangle tax driven changes in top incomes from changes driven by nontax factors. Implications: Upward bias of the elasticity estimates

8 Previous Literature: Identification Problems: 2. Mean Reversion Problem Define treatments and controls according to pre-reform income levels- taxpayer with a positive income shock in the pre-reform year will tend to have lower income in the following years. Implications: For tax cuts at the top, this will bias the elasticity estimate downwards.

9 Advantages of the Danish Data: 1. Stability of income distribution Distribution of top income share has been stable which eliminates the threat to identification that comes from nontax changes in inequality

10 Advantages of the Danish Data: 1. Large Variation in the Danish Tax Reforms Differential changes in the marginal tax rates across different tax brackets. Changes in brackets cutoffs that move taxpayers to different brackets Rich tax variation that is not correlated with the income level as compared to the US data. The tax reforms are not targeted at specific income groups and hence these reforms created a lot of up and down movement in tax rates throughout the income distribution.

11 Empirical Strategy Elasticity of Taxable Income (ETI): Relate changes in taxable income over time to changes in marginal tax rates over time. Approach: Difference in Difference Strategy Treatment: Face with tax cuts Control: Face with tax increase Panel Regression

12 Empirical Strategy Approach: Panel Regression: + controlling for pre-reform income (mean reversion), time fixed effects and individual fixed effects Results: Labour Income Elasticity Capital Income Elasticity Size of the tax reforms on estimates of elasticity

13 DID: Labour Elasticity

14 DID: Labour Elasticity

15 DID: Capital Elasticity

16 Key Findings (Panel Regression): Modest labour income elasticities ( ) Compared to labour, larger capital income elasticities ( ) Larger reforms, larger elasticities Evidence of frictions. Large tax variation to overcome frictions Evidence of modest income shifting between labour and capital income (modest cross tax elasticities)

17 Summary: Results are robust to different specifications of pre-reform income controls that are different from papers which used US tax reforms data where elasticity estimates were sensitive to the specifications used. Results of low taxable income elasticities despite the presence of high marginal tax rates suggest small opportunities for avoidance and evasion Broad tax base with limited chances of deductions Effective compliance enforcement Implication: Policy makers can consider increasing tax base and strict monitoring as conditions for imposing high marginal tax rate with limited behavioural responses.

18 Bibliography: Kleven, H. J., & Schultz, E. A. (2014). Estimating taxable income responses using Danish tax reforms. American Economic Journal: Economic Policy, 6(4), Saez, E. (2016). Striking it Richer: The Evolution of Top Incomes in the United States (updated with 2015 preliminary estimates). Berkeley: University of California, Department of Economics.