Evidence from a tax audit experiment in Denmark. Claus Thustrup Kreiner University of Copenhagen, CESifo, CEPR, Danish Economic Council.

Transcription

1 What makes tax payers comply? Evidence from a tax audit experiment in Denmark Claus Thustrup Kreiner University of Copenhagen, CESifo, CEPR, Danish Economic Council ECFIN Workshop November 2011

2 Overview of talk A. Background for tax compliance experiment B. Results on the anatomy of tax evasion C. Impact of tax enforcement on evasion behavior D. Cost-benefit analysis of the current Danish audit strategy E. Lessons for tax administration

3 A. Main questions? How big a problem is tax evasion? Why do people evade taxes? What is the optimal tax enforcement strategy to fight tax evasion? How many resources should be devoted to tax enforcement?

4 A. Empirical measurement is difficult Many non-experimental empirical studies of tax compliance and tax enforcement but serious measurement Not possible to measure evasion directly in register data People don t tell the truth, even in anonymous surveys (and large samples of individuals are too expensive) and identification problems What does it tell us, if we find high evasion rates in states/countries using many resources on tax enforcement?

5 A. Our tax compliance experiment Unique tax audit experiment carried out in Denmark in and including over 40,000 individual income tax filers Unwilling or Unable to Cheat? Evidence from a Tax Audit Experiment in Denmark. Kleven, Knudsen, Kreiner, Pedersen & Saez, Econometrica, may 2011 Tax Evasion and the administration of the Danish Tax System Chapter 4 in the Report of the Danish Economic Council, Spring 2011

6 B. Detectable tax evasion in Denmark Amounts in DKK Euro 1 = 7.5 DKK Total audit adjustment Underreporting Overreporting Net income Amount 2,2% 2,3% -0,1% Individuals 10,7% 8,6% 2,2% Total tax Amount 2,8% 3.0% -0,1% Individuals 10,6% 8,4% 2,2%

7 B. Income types, 3rd party information and tax evasion Share of total net income (%) Evasion rate(%) Total net income 100 2,3 Personal income 102 1,1 Deductions ,2 Capital income -5 2,6 Stock income 3 5,0 Self-employment income 5 15,7 Third-party reported income 95 0,3 Self-reported income 5 41,5

8 B. Probability of underreporting: Social, economic and information factors

9 B. Conclusions on the anatomy of tax evasion Size of tax gap? Reasonably low ( 2-3%) in relation to standard theory and e.g. USA Why low? Difficult to evade more than good moral (under reporting of 42% on self-reported income and 0,3% out of 3rd party reported income) Why Difficult to evade? Because of 3rd party information from employees, banks, trade unions etc. (95% of net income)

10 B. Conclusions on the anatomy of tax evasion Socio economic factors versus information about income Less under reporting of women and members of the church. Higher for persons working in small companies and in construction, fishery and agriculture BUT these socio economic factors have little explanatory power compared with variables reflecting existence and size of income, that is difficult to detect Attempts at declaring less income than 3rd party reported income is extremely rare

11 C. Impact of tax enforcement on evasion behavior Change in reported net income due to audit correction in 2007 Audit correction in 2007 Difference: 100% vs. 0% control group IV-effect of correction Net Net Self- Third-party income income reported reported Net income Amount (DKK) ,301

12 C. Impact of tax enforcement on evasion behavior Self-reported change in 2008 relative to the pre-populated tax return due to information about high audit probability No letter Letter No letter 50% letter No letter 100% letter 50% letter Adjustment Adjustment Up Down Up Down Probability of self reported adjustment (%) (0.35) (0.47) (0.28) (0.40) (0.33) (0.33) Amount (DKK) (31) (42) (22) (34) (26) (26)

13 C. Impact of tax enforcement on evasion behavior Tax audits have positive behavioral effects Detection of evasion/mistakes has large effects on tax compliance behavior the year after In the experiment, behavioral responses raise tax revenue the year after by 30-40% of the original audit adjustment Knowledge of a high audit probability has a positive, but small effect (obs: only done for wage earners)

14 D. Cost-benefit analysis of current audit strategy All Selfemployed Wage Earners Wage earners: Flag Wage earners No flag Population share Procent Revenue DKK Mechanical Behavior Audit cost Net effect

15 E. Lessons for tax administration Third-party information Effective instrument to reduce underreporting Recent expansions of third-party information in Denmark: Stock income Fringe benefits It seems difficult to expand third-party info much more in DK. Difficult to obtain third-party info on self-employment l income Third-party info is effective in modern economies with sophisticated production and large firms but not necessarily in underdeveloped countries (Kleven-Kreiner-Saez, 2009)

16 E. Lessons for tax administration Optimal audit strategy Audit selection criteria: Should focus on income information variables. Socio-economic factors do not improve selection significantly Important to take into account that audits also have a positive behavioral effect High evasion on self-employment income, but self-employed are also very expensive to audit Cost-benefit analyses are difficult to carry out and do not necessarily give the socially optimal level of resources spend on audits The current level of audit resources in Denmark does not seem far away from the revenue-maximizing level

17 Relationship between tax rate and evasion? xpayers mber of tax 200 Num 100 Before Audit After Audit A. Self-Employed Taxable Income