Decisions Under Uncertainty

advertisement
Tax Psychology (3)
Decisions under uncertainty
(risk)
Erich Kirchler
University of Vienna, Austria
TARC Master Class
London - 2014
Economics
Ideas of humankind: Homo Oeconomicus
• Utility maximization
• Rationality (consistent, goal-oriented behaviour)
2
Economics
Utility maximization
Rationality assumptions: Axioms
Completeness
(a > b; b = c; c > d; ...)
Transitivity
(if a > b and b > c, then a > c)
Reflexivity
(a = a)
Non-satiation
(a + 1 > a)
Continuity
(a, b) = (a – x, b + y)
Convexity
(Law of Saturation)
3
Paying taxes as outcome of a
decision under risk
How to increase and guarantee tax compliance ?
Decision under risk
Sure option
Risky option
Allingham & Sandmo (1972); Srinivasan (1973)
Becker, G. S. (1968). Crime and punishment: An economic approach. Journal of Political Economy, 76, 169-217.
Average compliance rate by fine rate and audit probability
(standard deviations in parentheses; Alm et al., 1995, p. 11)
Audit
Fine rate
probability
1
2
4
----------------------------------------------------------------------------0.05
9.0 ( 4.0)
6.9 ( 3.2) 12.2 ( 4.2)
0.30
10.9 ( 6.0)
21.4 ( 4.4) 39.8 ( 7.4)
0.60
9.8 ( 8.0)
54.8 (10.6) 70.3 ( 7.5)
What does the theory say about
what motivates tax compliance ?
The standard theory of human behaviour is based
on several assumptions:
• Individuals are rational,
• Individuals have unlimited willpower,
• Individuals are self-interested.
From James Alm, 2014, Tulane University, USA.
What does the theory say about
what motivates tax compliance ?
The starting point: Economics-of-crime model
• A rational individual weights the expected benefits of
successful cheating against the risky prospect of
detection and punishment.
• The individual pays taxes because he or she is afraid of
getting caught and penalized.
• Indeed, the individual pays taxes because – and only
because – of the fear of detection and punishment.
Compliance depends upon enforcement.
From James Alm, 2014, Tulane University, USA.
What does the theory say about
what motivates tax compliance ?
However, …
•
•
•
•
•
•
Individuals face limits of their ability to compute (e.g., bounded rationality)
They systematically misperceive, or do not perceive at all, the true cost of
their actions (e.g., fiscal illusion, saliency, overweighting of (low)
probabilities)
They face limits of their self-control (e.g., hyperboic discounting; Christmas
club savings)
They are affected by the ways in which choices are framed (e.g., reference
points, gains versus losses, loss aversion, risk-seeing behavior)
They are affected by the social context in which, and the process by which,
decisions are made
They are motivated by notions of fairness, altruism, trust, guilt, shame,
morality, alienation, emotions, patriotism, social customs, social norms, tax
morale, …
From James Alm, 2014, Tulane University, USA.
What does the theory say about
what motivates tax compliance ?
Behavioral economics suggests several main
conclusions about what motivates tax compliance:
•
•
•
Enforcement matters – but many other factors matter in
the tax compliance decision beyond enforcement.
An individual does not always behave as assumed in
the standard economic appraoch.
Individuals are social creatures and are influenced by
group considerations.
From James Alm, 2014, Tulane University, USA.
16 min
Dan Arieli Cheating
Metaanalyses
Andreoni, J., Erard, B., & Feinstein, J. S. (1998). Tax compliance.
Journal of Economic Literature, 36(2), 818-860.
Kirchler, E., Muehlbacher, S., Kastlunger, B. & Wahl, I. (2010). Why pay
taxes? A review of tax compliance decisions. In J. Alm, J. MartinezVazques & B. Torgler (eds.). Developing Alternative Frameworks for
Explaining Tax Compliance (pp. 15-31). London: Routledge.
Most people are honest !
Correlations between tax evasion and measures of
psychological instigations and constraints (Elffers et al., 1987)
Behavioural outcome measures
Psychological variables
Dissatisfaction
Dissatisfaction with tax
authorities
Comprehensibility of rules
Personality
Competitiveness
Alienation
Tolerance of deviance
Fear of punishment
Social control
Personal control (attitudes)
2-year self- Documented
report
status
Documented amount of
tax evaded
0
+
+
0
+
+
+
-
0
0
0
0
0
0
Enforcing tax compliance
1. Deterrence
To protect honest taxpayers from free riders, controls are
The effect of deterrence measures
necessary.
weak and
oppostite
Negative is
sanctions
aresometimes
necessary at an
adequate level
and in thetoproper
form, depending
on the ability of the tax
the intended
effect…
offender to pay.
Tax authorities need to cooperate intensively with legislators,
judges, and international authorities.
Enforcing tax compliance
What effect have fines ?
What effect have repeated audits ?
A fine is a price !
Gneezy, U. & Rustichini, A. (2000). A fine is a price. Journal of Legal Studies 29(1) 1-18.
Experimental group: Managements of 6 day care centers introduced a fine for late pick up of children;
Control group: 4 day care centers did not introduce a fine.
% of smokers classified as heavy smokers (25+)
Reducing undesirable behavior by special taxes
Australia’s National Tobacco Campaign
30
25
20
Shift from
a weight
to
a stick
based
system
15
10
Harmonization of fees
in different states
(to end cross-border
Evasion of cigarette taxes)
5
Extra
tax on
tobacco
products
0
May 97
Nov 97
Nov 98
Scollo, Younie, Wakefield, Freeman, & Icasiano, 2003
Nov 99
Nov 00
Potential side effects of special taxes 1
% of smokers using Roll Your Own Tobacco
Australia’s National Tobacco Campaign
25
20
15
Shift towards unhealthy
form of tobacco
consumption
10
5
0
May 97
Nov 97
Nov 98
Scollo, Younie, Wakefield, Freeman, & Icasiano, 2003
Nov 99
Nov 00
Potential side effects of special taxes 2
Cigarette sales & consumption in Washington State (USA)
Tax rise
from 25c
to 75c per
pack
Stehr, 2005
Consumption
remains stable 
Presumably due to
smuggling
Tax compliance
(Kirchler, Maciejovsky & Schwarzenberger, 2005)
.6
Compliance
.5
Audit probability 30 %
.4
.3
.2
.1
0
0
Base-line
1
2
3
4
Periods following audit
20
Robust phenomena: the “bomb crater effect“ and the „echo effect“
(Guala & Mittone, 2005; Mittone 2006; Kastlunger et al. 2009;
Maciejovsky et al. 2007)
“Echo” effect in experiments with audits in the first and second half of 60
business periods, respectively Guala and Mittone (2002, p. 12 and 13)
Tax payments (averages, first group)
Value (Italian Liras)
500
400
300
200
100
Tax due
Average tax paid
Audit
0
1
7
4
13
10
19
16
25
22
31
28
37
34
43
40
49
46
Round
55
52
58
“Echo” effect in experiments with audits in the first and second half of 60
business periods, respectively Guala and Mittone (2002, p. 12 and 13)
Tax payments (averages, first group)
Value (Italian Liras)
500
400
300
200
100
Tax due
Average tax paid
Audit
0
1
7
4
13
10
19
16
25
22
31
28
37
34
43
40
49
46
55
52
Round
58
Impact of immediate versus delayed audits on tax
compliance and perceived fairness of authorities
Kogler, C., Mittone, L. & Kirchler, E. (2014)
Timing of feedback of tax audits matters
– Uncertainty resolution in tax experiments: waiting for an audit increases tax
compliance (Muehlbacher et al., 2012)
– Problem: audits directly after filing taxes in most tax experiments, but in reality
often years after filing a tax report (e.g., Austria 5-10 years)
Feedback related to trust, trust related to compliance
– In other contexts (e.g., organizational psychology; Sapienza & Korsgaard, 1996)
timeliness of feedback was identified to increase trust and acceptance of
decisions
– In tax literature trust in state/authorities was identified as an important
determinant of tax compliance and tax morale (e.g., Torgler & Schneider, 2004;
Braithwaite & Wenzel, 2007; Kirchler et al., 2008)
Experimental Design
• 22 rounds of taxpaying in a laboratory experiment
• Regular income
– 3000 ECU, tax rate: 900 ECU
– Pay-off determined at the end (one round randomly chosen)
• Audit probability: 15%
– Rounds 3, 12, 21 for all participants to keep effects of audits constant
• Fine in case of detection
– 2 x evaded amount (paying back + fine equal to evaded amount)
• Feedback
– After each round vs. only at the end (summary after the last round)
• Rounds of reduced earnings
– Rounds of reduced earnings: R8-10, R17-18
– Reduced income of 2500 ECU vs. additional tax of 500 ECU (compulsory)
• Questionnaire:
– Items: perceived fairness of tax authorities, voluntary compliance, subjective audit
probability, perceived severity of fine, perceived fairness of timing of feedback, general
tax morale, socio-demographic data, etc.
Results I – Feedback x Compulsory Tax
Covariate Gender:
Mean females = 69.49%
Mean males = 50.60%
F(1, 121) = 7.44, p < .01
Audit
Reduced Earnings
Mean immediate feedback = 50.75%,
Mean delated feedback
= 69.29%
F (1, 122) = 11.30, p < .01
Immediate vs delayed audits
In experiments, uncertainty if tax report is audited is resolved
immediately, whereas in real life it can take up to 7 years to know if
your audited or not (Zeigarnik-Effect; unfinished businesses)
Compliance decision Control group
(pay 0-6 € tax)
Compliance
Rates:
Experimental group
5 minutes
3 weeks
28%
59%
Muehlbacher, Mittone, Kastlunger, & Kirchler, 2012
There is more than audits and fines, …
situational and personal characteristics;
social norms, fairness, …
and perhaps the perception of taxpayers
as cheaters is short-sighted.
Diffusion of income tax evasion
10-15% of underreported income
Porcano (1988), the U.S. IRS estimated 10-15% of underreported income in 1983. Five years later the tax gap was about 17% of true liability.
25% underpaid
7.3% non-compliance, wide variation across types of income and deductions
Slemrod et al. (2001) explain that the detected rate of non-compliance is 7.3%, but varies widely across types of gross income and deductions. In 1988, voluntary
Andreoni and colleagues (1998) estimate that over 25% of all U.S. taxpayers underpaid their taxes in 1988. In developed countries, tax evasion is estimated to
reach 20% of the level of tax revenues, while in developing countries the percentages are even higher (Orviska & Hudson, 2002).
reporting was 99.5% for wages and salaries, but only 41.4% for self-employed income. In 2002, King & Sheffrin report that according to the U.S. IRS, 99% of
wage income is correctly reported, but less than 70% of income from unincorporated businesses is correctly reported.
Elffers (2000) writes, “[…] the
gloomy picture of massive tax
4.2% corrected their tax files to their disadvantage, 23.8% corrected them to their
advantage.
Assuming that
those
negatively
correcting their files made unintentional
evasion
is
a
phantom”.
make
reports. 13.5%
overstate
their taxes,
presumably
duedid
to
mistakes,¼and
an accurate
equal percentage
of those
who positively
corrected
their files
voluntary reporting 99.5% for wages and salaries, but only 41.4% for self-employed
There is little doubt that non-compliance should be contained and evasion, in particular, needs to be combated. It is, however, wrong to assume that the majority
of people try to evade or avoid paying taxes.
Long and Swingen (1991) write that some taxpayers are not predisposed to evade and do not search for ways to cheat. Survey studies and experiments on
income tax behaviour show that honesty characterises a majority of participants (e.g., V. Braithwaite, 2003d; James & Alley, 2002; Kirchler, Muehlbacher, Hoelzl,
& Webley, 2005).
Antonides and Robben found that 4.2% of participants in their study corrected their tax files to their disadvantage, whereas 23.8% corrected them to their
advantage (Antonides & Robben, 1995). Assuming that those negatively correcting their files made unintentional mistakes, and an equal percentage of those who
positively corrected their files did so also undeliberately, then less than 20% were intending to cheat.
On the basis of 1982 U.S. IRS audit data, Alexander and Feinstein (1987) report that approximately one quarter of all taxpayers make accurate tax reports.
According to their analysis, 13.5% overstate their taxes, presumably due to errors in completing tax returns. If the same percentage understates their taxes due to
errors, then more than half of taxpayers tend to be honest.
errors. If the same
understates
taxes due to errors, then
so also undeliberately,
then percentage
less than 20%
intend to their
cheat.
more than half of taxpayers tend to be honest.
Hessing, Elffers, and Weigel (1988) estimate that more than two thirds of taxpayers declare their income honestly.
2/3 declare their income honestly
The assumption that taxpayers are generally compliant is challenged by the wide use of tax preparers and studies contending that taxpayers generally demand
aggressive advice (Duncan, LaRue, & Reckers, 1989; Jackson, Milliron, & Toy, 1988; Milliron, 1988). These studies were conducted from tax preparers’ view, but
investigations from taxpayers’ perspectives reveal a different picture. The use of a tax practitioner does not seem primarily driven by the desire to avoid paying
taxes, but by the uncertainty about the tax law and the motivation to report correctly.
The use of a tax practitioner does not seem primarily driven by the desire to avoid
paying taxes, but by the uncertainty about the tax law and the motivation to report
correctly.
Economic Psychology Agenda
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
Self-Conception
Objectivity & Intentions
(Psycho-) Logics
Rituals of Adults
The Invisible Hand
Mental Accounting
Trick & Treat
Courage & Risk
20 US Cents
Invulnerability
Whatever you want!
In the Shadow
Nudges
75.000 US$
Ideas of Mankind
Methods
Economic Decisions
Lay Economic Theories
Markets
Consumption, Saving, Loans
Marketing Policy
Businessmen/women
Work & Satisfaction
Financial Markets
Currency & Inflation
Counter-Productivity
State: Power and Trust
Prosperity and Happiness
31
Decision Anomalies:
Perception of information
Cognitive scrooges:
Reality is subjectively construed and interpreted
• Estimation of the diameter of coins
• Optical illusions
32
Paying back loans
You have lended 3.000 € and need to pay it back with
12% annual interest rate. Each month you pay 30 €.
How long does it take to pay the credit back?
(a) less than 5 years
(b) 5 to 10 years
(c) 11 to 15 years
(d) 16 to 20 years
(e) for ever
15%
31%
18%
10%
26%
33
Risk and emotions
Emotions
Probabilities are ignored
or low probabilities are overestimated
Money
18
16
Prices paid to
avoid electric
shock and $20
penalty
14
12
Shock
10
8
6
4
2
0
1%
99 %
Probability
The willingness to pay to avoid electric shocks barely depends on the
probability of being exposed to electric shocks (Rottenstreich & Hsee: Money,
Kisses, and Electric Shock: On the Affective Psychology Risk. Psychological
34
Science, 2001)
Risk-aversion
Daniel Bernoulli (18th century): Human beings are
risk averse!
A sure gain of € 8.000 is
preferred to a possible gain (p = .85) of € 10.000.
35
Risk-aversion and -inclination
a) Sure win of € 240
or
25 % chance of winning € 1.000
and 75 % chance of winning 0
b) Sure loss of € 750
or
75 % chance of loosing € 1.000 &
25 % chance of loosing 0
36
Prospect Theory
Descriptive decision theory explaining a number
of decision anomalies in decision making under
risk.
Phase 1: Editing
Simplification of the decision problem
Phase 2: Evaluation
Assessment of the alternatives
37
Prospect Theory
Phase 1: Editing
Simplification of the decision problem
Coding: What is an event related to (reference point)?
Combination: Independent events are often seen combined.
Segregation: Safe events are often separated from risky
ones.
Cancellation: Identical aspects of two alternatives are
ignored whereas disparities are weighted heavily in
judgments.
38
Prospect Theory
Phase 1: Editing
Simplification of the decision problem
Simplification: e.g. odd numbers are rounded. In the case
of very low or very high probabilities this process is
fundamental.
Detection of dominance: Salient alternatives are discarded
in the beginning and are not taken into account any further.
39
Prospect Theory
Phase 2: Evaluation
Assessment of alternatives
Assessment is guided by two principles
• The value is estimated in relation to a reference point.
• Probabilities are taken into account.
40
Prospect-Theorie
Subjective value +
concave
Loss
Gain
-A -B
A
convex
B
41
Prospect Theory: Weighting function
Decision weight
1
0.5
0
42
0
0.5
Probability p
1
42
Neuropsychology /
Neuroeconomics
Support for the assumptions of the prospect theory comes from neuroeconomics,
examining the brain activity while decisions are made.
Neuroscientific methods allow the analysis of the activities of the human brain (e.g. Prince
& Pawelzik, 2008). In addition to imaging and psychophysiological methods, the
measurement of individual neurons, the electrical brain stimulation or the elimination of
brain regions can be counted among the methodological tools of neuroscience and, more
recently, of neuroeconomics. First and foremost imaging and psychophysiological
methods are used. Apart from the electroencephalogram (EEG), positron emission
tomography (PET) is used, which allows the creation of cross-sectional images of the
brains of humans and mapping biochemical and physiological processes.
Currently, functional magnetic resonance imaging (fMRI) is most popular. This method
allows the representation of metabolic activity in the brain by measuring the magnetic
properties of oxygenated and deoxygenated blood.
Using psychophysiological methods, different physiological responses to a stimulus, such
as blood pressure, heart rate, sweating, dilated pupils or muscle tone, can be measured
(Sanfey, 2007). Many studies have shown that economic decisions are correlated with
neurophysiological processes (e.g., Delgado, Locke, Stenger, & Fiez, 2003; Knutson,
Taylor, Kaufman, Peterson, & Glover, 2005).
Sanfey, Rilling, Aronson, Nystrom, and Cohen (2003) examined fMRI responses of the
players to fair and unfair offers in the ultimatum game and found out that in the case of
rejected, unfair offers other brain regions are activated, as in the case of accepted, unfair
offers.
43
Experiment
5 25
1. Choose 5 or 25
2. Outcome: gain or loss
3. Result:
4. Brain activity:
25 gain & correct
5 gain & incorrect
25 loss & incorrect
5 loss & correct
265 milliseconds after the information about a
win or loss => higher amplitude of a potential
springing from the frontal area of the
Cerebrum Medium in the event of a loss.
44
Libertarian
Paternalism
• Asymmetric (libertarian) paternalism:
– Does not affect rational decision makers;
– Works to the advantage of biased decision makers.
• Policy applications
–
–
–
–
–
Providing information;
Framing;
Changing status quo;
Format effects;
Motivational effects.
Tax compliance of people who are liable to wage tax and of self-employed persons as a function
of the refund amount or balance due (Cox & Plumley, 1988; as cited in Webley et al., 1991)
Compliance rate in percent
100
wages/salaries
95
90
business income
85
80
75
70
65
<-1.000
<- 500
<- 100
<- 0
>- 0
>- 100
refund
46
Size of refund or balance due (US $)
>- 500 >-1.000
balance due
Download