View
30
Download
0
Category
Preview:
Citation preview
Recep Tekeli Policy Research Institute 1
1 Determinants of Tax Morale in Japan and in Turkey
The Determinants of Tax Morale: the Effects of Cultural Differences and Politics
Recep Tekeli1
Abstract: In this paper we analyze the tax morale in two OECD countries; Turkey and Japan, using data from the fifth wave of World Values Surveys. Using Survey data for comparative analysis we can see the differences in several factors affecting Tax Morale between Japan and Turkey We empirically test what shapes tax morale by using Ordered Probit model. We have followed the literature but used additional variables to see what determines the notion “intrinsic motivation to pay taxes i.e tax morale”. Most of our findings are in line with the earlier works in tax morale literature. Our findings also indicate significantly higher tax morale in Japan than in Turkey, controlling in a multivariate analysis for additional variables.
Keywords: Tax morale; Tax compliance
JEL classification: H26; H30
1. Introduction
Why do people pay taxes? This question recently has started to be pronounced more often and
attracted increased attention in the tax (non)compliance literature over the last few years. It is
supposed that nobody likes to pay taxes but governments can still generate tax revenues. One
possible answer among the scholars is to force people to pay their taxes by establishing a
deterrence policy.
In line with the economics-of-crime approach, Allingham and Sandmo’s (1972) model shows
that the extent of tax evasion is negatively correlated with the probability of detection and the
degree of punishment. However, because of the empirical and experimental findings, their
seminal model has attracted many criticisms by researchers (see, e.g., Graetz and Wilde 1985;
Alm et al. 1992; Frey and Feld 2002). These deterrence models predict a comparatively high
incidence of tax evasion; even in many countries, the actual level of deterrence is found to be
too low to explain the high degree of tax compliance.(see Torgler, Schneider and Shaltegger,
2010:294; Torgler and Shaltegger, 2006:396)2
1 Faculty of Economics and Administrative Sciences, Department of Public Finance, University of Adnan Menderes, Nazilli, Turkey (rtekeli@adu.edu.tr). The Research Grant by Higher Education Council of Turkey (YÖK) is greatly appreciated. I would like to thank Policy Research Institute at Ministry of Finance, Japan for granting Visiting Scholar position and Policy Research Institute staff. Also I would like to thank to participants at the workshop where this paper was presented. 2 Torgler Schneider and Shaltegger (2010:294) and Torgler and Shaltegger (2006:396) quoting from several researchers report that there is a considerable gap between the amount of risk aversion that is required to guarantee such compliance and the effectively reported degree of risk aversion. For example, in the United States and Switzerland, the estimated Arrow–Pratt measure of risk aversion is between 1 and 2, but approximately a value of 30 would explain the observed compliance rate in these countries. Also, they argued that a higher level of income declaration is reported in tax compliance experiments than the expected utility model would predict.
Recep Tekeli Policy Research Institute 2
2 Determinants of Tax Morale in Japan and in Turkey
In the tax compliance literature this seemed to be a “puzzle”, or in other words, it is indeed an
avenue to study further for the scholars. To resolve this puzzle of tax compliance, many
researchers have tried to link tax morale to the high degree of tax compliance (e.g., Schwartz
and Orleans 1967; Lewis 1982; Roth et al. 1989; Alm et al. 1992, 1999; Pommerehne et al.
1994; Frey 1997; Frey and Feld 2002; Feld and Tyran 2002; Torgler 2001 and Torgler and
Schneider 2009). In this perspective, Torgler, (2003:290) argued that Erard and Feinstein’s
work (1994) “demonstrates the relevance of integrating moral sentiments into the models to
provide a reasonable explanation of actual compliance behavior”. Also while reviewing the
tax compliance, Andreoni et al. (1998:852) suggest that “adding moral and social dynamics to
models of tax compliance is as yet a largely undeveloped area of research” (see Torgler
Schneider and Shaltegger 2010:294)
Following Torgler’s researches into the field of tax morale, rather than treating tax morale as
a black box, or a residuum, many papers have analyzed which factors shape or maintain tax
morale (see Torgler and Schneider 2007; Torgler and Murphy 2004; and Torgler Schneider
and Shaltegger 2010). In addition, this puzzle concerns policymakers because they need to
know the driving forces of tax morale and the possibility that it influences willingness to pay
taxes so that they can design efficient tax system.
In the literature, tax morale is defined as “a moral obligation to pay taxes”, or “a belief in
contributing to society by paying taxes”. It is also defined as “the existence of an intrinsic
motivation to pay taxes” (Torgler 2003; Torgler 2004; Torgler and Schneider 2009;
Cummings et al 2009; Torgller 2005). This is not an output variable like the size of shadow
economy; it measures individual attitude not individual behavior (Torgler 2004; Torgler and
Schneider 2009). Tax morale is also closely linked to the term taxpayer ethics by Alm and
Torgler (2006:228) and Torgler and Schneider (2007a:10; 2007b:444) and used the definition
by Song and Yarbrough (1978:443) that, “the norms of behaviour governing citizens as
taxpayers in their relationship with the government”. Torgler and Murphy(2004:301) defines
the concept as “tax morale can generally be understood to describe the moral principles or
values individuals hold about paying their tax” 3
3 Tax morale is sometimes related to the term tax culture. Although it is not well conceptually organized yet, tax culture is defined as all residual factors that have not been considered to explain the behavior of tax compliance. In this aspect, Nerre (2008:155) defines tax culture as follows: “A country-specific tax culture is the entirety of all relevant formal and informal institutions connected with the national tax system and its practical execution, which are historically embedded within the country’s culture, including the dependencies and ties resulting from their ongoing interaction.”
Recep Tekeli Policy Research Institute 3
3 Determinants of Tax Morale in Japan and in Turkey
Why Tax Morale is Important?
German scholars in the 1960s (e.g. Schmölders, 1970 and Strümpel, 1969) “tried to bridge
economics and social psychology by emphasizing that economic phenomena should be
analyzed from a perspective larger than the traditional neoclassical point of view (e.g., Lewis,
1979, 1982)” (see Torgler and Schneider 2007b:444).
In this respect, Torgler and Schneider (2007a:10-11) argue that values and attitudes can affect
individual behavior: Apart from sanctions, Spicer and Lundstedt (1976) argued that a set of
attitudes and norms might have affect on the choice between tax compliance and evasion.
Lewis (1982) points out that “it could be that tax evasion is the only channel through which
taxpayers can express their antipathy … we can be confident in our general prediction that if
tax attitudes become worse, tax evasion will increase” (p. 165, 177).” (excerpted from Torgler
and Schneider 2007a:10-11). Therefore, we can state that values and attitudes can affect
individual’s behavior.
Tax noncompliance is actually inevitable fact in all societies (see Cowell, 1990; Schneider
and Enste (2000, 2002); Schneider 2003; Alm et al. 2004; Schneider 2005; Schneider et al
2010 ). Schneider (2005:598) argues that most societies attempt to control shadow economic
activities “through measures such as punishment and prosecution, or by relying on economic
growth or education. Gathering statistics about who is engaged in shadow economy activities
and the frequencies with which these activities occur and magnitudes, is important”.
Schneider (2005:600) defines the shadow economy as: “ The shadow economy includes all
market-based legal production of goods and services that are deliberately concealed from
public authorities”: to avoid payment of taxes and social security contributions, to avoid
having to meet certain legal labor market standards, such as minimum wages, and to avoid
complying with certain administrative procedures. While the shadow economy is estimated
generally it was not dealt with typical economic activities that are illegal and fit the
characteristics of traditional crimes such as gambling, burglary, robbery, drug dealing,
prostitution etc. The informal household economy, which consists of all household services
and production, is not included either (see Schneider 2005; Schneider et al 2010).
Schnieder (2005) and Schneider et al (2010) have provided estimates of the size of the
shadow economies for developing, transition and highly developed OECD countries over the
periods of time using the MIMIC/DYMIMIC and the currency demand approach. The first
conclusion from these results is that for all countries investigated the shadow economy has
Recep Tekeli Policy Research Institute 4
4 Determinants of Tax Morale in Japan and in Turkey
reached a remarkably large size and increasing over the years. Also, the empirical results
convincingly demonstrate that an increasing burden of taxation and social security payments,
combined with rising state regulatory activities, are the major driving forces underlying the
size and growth of the shadow economy4.
Table 1: The size of shadow economy in 26 EU countries and Turkey Size of the shadow economy (as % of GDP)
EU Countries 1989/1990 1994/1995 1999/2000 2007
Austria 6.9 8.6 9.8 10.1
Belgium 19.3 21.5 22.2 23.1
Bulgaria 29.4 33.2 36.9 41.2
Czech Republic 15.9 17.2 19.1 21.2
Denmark 10.8 17.8 18 19
Estonia N.A N.A N.A 42.3
Finland 13.4 18.2 18.1 19.2
France 9 14.5 15.2 15.7
Germany 11.8 13.5 16 16.7
Great Britain 9.6 12.5 12.7 13.2
Greece 22.6 28.6 28.7 31
Hungary 21.4 23.9 25.1 26.4
Ireland 11 15.4 15.9 16.4
Italy 22.8 26 27.1 27.4
Latvia 32.5 36.3 39.9 44.3
Lithuania 24.7 27.1 30.3 34
Luxembourg N.A N.A N.A 10.2
Malta N.A N.A N.A 27.7
Netherlands 11.9 13.7 13.1 13.2
Poland 21.3 24.3 27.6 29.1
Portugal 15.9 22.1 22.7 22.5
Romania 26.2 30.6 34.4 38.9
Slovakia 14.3 16.2 18.9 21.1
Slovenia 21.5 24.3 27.1 29.5
Spain 16.1 22.4 22.7 23.1
Sweden 15.8 19.5 19.2 20.4
Turkey 26.3 29.4 32.1 35.2
Average 17.9 21.5 23.0 24.725 Source: the data are obtained from Schneider (2005) and Schneider et al (2010) with the average value is our own estimate.
4 Although it is difficult to estimate the accurate size of a shadow economy, there are three methods can be found in the literature to estimate the size of the shadow economy (see Schneider 2005): 1. Direct approaches, 2. Indirect approaches [2.1. The discrepancy between national expenditure and income statistics, 2.2. The transactions approach, 2.3. The discrepancy between the official and actual labor force, 2.4. The currency demand approach, 2.5. The physical input (electricity consumption) method, 2.5.1. The Kaufmann-Kaliberda Method, 2.5.2. The Lacko´ method] and, 3. The model approach (DYMIMIC).
Recep Tekeli Policy Research Institute 5
5 Determinants of Tax Morale in Japan and in Turkey
The size and development of shadow economies of 26 EU countries is shown in Table 1. The
average shadow economy size was 17.9% in 1989/90 and increased to 24.75% in 2007. In
Turkey the size of shadow economy shows a significant increase. The lowest shadow
economies have Austria, Luxembourg and Britain with the size of shadow economy around
10 to 13 percent in 2007. The highest shadow economies among 26 EU countries have
Romania with 38.9, Bulgaria 41.2 Estonia 42.3 and Latvia 44.3 percent. These countries’
sizes of shadow economy are above Turkey’s size in 2007 i.e 35.2% (see Graph 1).
Source: see the fourth column in Table 1.
Note: Graph is prepared by using shadow economy estimates of Schneider et al (2010)
Recep Tekeli Policy Research Institute 6
6 Determinants of Tax Morale in Japan and in Turkey
In Graph 2 above and Table 2 below the size of the shadow economy of 27 OECD countries
are shown. The lowest shadow economies have the USA, Switzerland, and Austria with an
average size of shadow economy varies from 9 to 10%. Japan also shows a low level of
shadow economy with 12.1%. The countries with the highest shadow economies among these
27 OECD countries are Turkey with 35.2%, Mexico with 31.3 %, Greece with 31% and South
Korea with 29.4%.
Table 2: The size of shadow economy in OECD countries (including Japan)
OECD Countries 2007 USA 9 Switzerland 9.1 Austria 10.1 Luxembourg 10.2 Japan 12.1 Great Britain 13.2 Netherlands 13.2 New Zealand 13.6 Australia 15 France 15.7 Ireland 16.4 Canada 16.6 Germany 16.7 Iceland 16.8 Denmark 19 Finland 19.2 Norway 20.2 Sweden 20.4 Portugal 22.5 Belgium 23.1 Spain 23.1 Italy 27.4 South Korea 29.4 Greece 31 Mexico 31.3 Turkey 35.2 Mean 18.8
The comparison of shadow economies in EU and OECD economies shows an important
phenomenon for some countries to deal with. The main problem in the shadow economy, or
black economy, is the fact that individuals are behaving dishonestly by providing false
Recep Tekeli Policy Research Institute 7
7 Determinants of Tax Morale in Japan and in Turkey
information. If so, what would lead citizens to behave more honestly, provide correct
information and improve the tax compliance rate? Some believe that tax morale is an answer
to this question (see Feld and Frey 2002). Also, Torgler and Schneider (2007b) argue that a
reduction of tax morale reduces the moral costs of behaving illegally and increases the
incentives to work in the black economy.
According to Alm, et al (2004), a negative correlation between the size of the shadow
economy, which is a measure of the extent of tax evasion, and tax morale indicates the extent
to which individuals’ revealed actions are related to their attitudes about paying taxes (Torgler
and Schneider 2009:230; Torgler and Schneider (2007b)). In this extent, a number of previous
studies have investigated the simple correlation between tax morale and the size of shadow
economy. For example, Alm and Torgler (2006:242) focusing on Europe and the United
States find a strong negative correlation (r=-0.460). Alm, Martinez-Vazquez, and Torgler
(2006) focused on transition countries and their results indicate a strong negative correlation
between both variables (-0.657) (see Torgler and Schneider, 2009:230). Torgler (2005)
investigates the correlation between the size of shadow economy and tax morale in Latin
America, and a strong negative correlation between both variables (-0.511) has been found.
Torgler and Schneider (2009) find a significant correlation between tax morale and the size of
shadow economy based on data from more than fifty countries. Their empirical results
strongly suggest that tax morale plays a significant role in determining the level of shadow
economy: higher tax morale leads to a smaller shadow economy. Also, the beta coefficients
show that its quantitative impact is comparable to other determinants.
It has been argued that if tax authority places a taxman under every bed it is highly likely to
achieve a high level of tax compliance rate but with high costs to the tax authority. Torgler
and Shaltegger (2006:397) argues that “even though taxation is enforced by law, there is a
moral dimension in paying taxes for many people”. For example, Slemrod (1992:7) “states
that methods that reinforce and encourage taxpayers’ devotion to their responsibilities as
citizens play an important role in the tax collection process” (excerpted from Torgler and
Shaltegger 2006:397).
If tax morale is thought to be an explanation for why tax compliance rates are so high, it
would be interesting to analyze what may shape tax morale among taxpayers. Next section
examines the magnitude and determinants of Tax Morale. Model and variables which are
considered to be important explaining tax morale are presented as well. Section 3 gives the
empirical results in Turkish and Japanese experience. Last section concludes the paper.
Recep Tekeli Policy Research Institute 8
8 Determinants of Tax Morale in Japan and in Turkey
2. Magnitude and Determinants of Tax Morale: 2.1 Magnitude
If tax morale is seen as an important factor to understand the puzzle of tax compliance and
plays a significant role in determining the levels of shadow economy, then it is necessary to
investigate the determinants of tax morale. Thus, this section focuses on the magnitude and
determinants of tax morale in the study countries: Turkey and Japan.
We define tax morale as “the intrinsic motivation to pay taxes” as defined by Torgler in his
papers. This is not physical output variable such as tax evasion or the size of shadow
economy. It measures an individual’s willingness to pay taxes, in other words, “the moral
obligation to pay taxes” or “the belief that paying taxes contributes to society”. Data for the
tax morale variable are extracted from the World Values Survey (WVS) (see Inglehart et al.,
2000). The World Value Survey is a worldwide investigation of socio-cultural and political
change, which includes the case of Turkey and Japan. There are so many variables for this
dataset and it covers quite a huge number of countries. They are based on representative
national samples of more than 1000 individuals. In the 2005 wave of the survey, a total of
1346 Turkish citizens and 1096 Japanese citizens agreed to participate in the study.
Data from these surveys can be accessed by researchers interested in how views change with
time and across countries. The World Values Surveys have produced comparative data on
what people value, what their beliefs are, what they want out of life and the facts of life. In the
survey, our main concern is the question number 200 which is about the attitudes toward
cheating on taxation. This attitude was measured with the 10 different scales, where the value
of one (1) indicates that cheating on taxation is never justifiable and the value of ten (10)
means it is always justifiable. This dataset has been regularly surveyed. The first one was
carried out in 1981 for European Value Survey. The second wave was completed in 1990-
1991, the third one was carried out in 1995-1996, the fourth one took place in 1999-2001, and
the fifth and last one took place in 2005/2007. As the last one is the most recent one which we
can access, we use the last wave for the analysis.
The general question to assess the level of tax morale in World Values Survey is that:
“Please tell me for each of the following statements whether you think it can always be justified, never be justified, or something in between”: V200. Cheating on tax if you have the chance.
1 2 3 4 5 6 7 8 9 10
never justified always justified
Recep Tekeli Policy Research Institute 9
9 Determinants of Tax Morale in Japan and in Turkey
The tax morale variable is generated by recoding this ten-point scale into a four point scale
(0,1,2, and 3), with the value 3 standing for “never justifiable”. As usual, the value of 0 is an
aggregation of the last 7 scale points (4-10), which were rarely chosen (ie, 0 = “always
justifiable” to 3 = “never justifiable”; original scores of 4-10 were recoded into 0 = “always
justifiable”).
We should point Elffers et al. (1987) that their findings indicate the differences between
actual evasion assessed and evasion reported in survey responses. This result shows that
subjective surveys are always prone to significant reporting errors. As argued by some
researchers, tax morale is also measured with subjective survey responses in WWS, thus the
measurement of tax morale is not free of bias (e.g. Torgler 2004; Torgler and Schneider 2007;
Torgler et al 2010). Because the available data from surveys are based on self-reports in
which respondents may tend to overstate their degree of compliance (Andreoni et al. 1998),
no objective or directly observable measure of tax morale is available. However, the degree of
honesty is expected to be higher because in the WWS the tax morale is defined less sensitive
than directly asking whether a person has evaded taxes. Moreover, the dataset is based on
wide-ranging surveys (based on more than two hundred questions) it was assumed to reduce
the probability of respondent suspicion and the framing effects of other tax context questions.
However, it can still be argued that a taxpayer who has evaded in the past will tend to excuse
this kind of behavior and report a higher degree of tax morale in the survey. As no sanctions
are involved in the survey many respondents might overstate their willingness to pay taxes or
degree of honesty.( see Torgler and Schneider 2007a; Torgler and Schneider 2007b; Torgler
and Shaltegger 2006) 5
On the other hand, it was further argued that the use of such a single question (single item
value) has the advantage of reducing problems of index construction complexity, especially in
regard to measurement procedure or low correlation between items (see Torgler and
Schneider 2007a; Torgler 2004; Torgler 2003; Alm and Torgler2006). According to Torgler
and Schneider (2007a:450) “tax morale is a multi-dimensional concept that requires a multi-
item measurement tool and the likelihood of a multi-item index being adversely affected by
random errors will produce more reliable measures. However, several previous studies have
found consistent results using single-item survey measurements and laboratory experiments
5 We should also indicate another disadvantage of working with survey data as argued by Alm and Torgler (2006:237-238) that “we cannot control for such traditional factors as the audit probability (because this is not known for each individual) and the fine rate (because this is identical for all individuals in a country)”.
Recep Tekeli Policy Research Institute 10
10 Determinants of Tax Morale in Japan and in Turkey
(e.g., Cummings et al. 2005; Alm and Torgler 2006)”. Despite these criticisms our approach
to measuring tax morale is consistent with the previous studies in this area (see Torgler, 2007;
Torgler and Schneider 2007a; Torgler 2004; Torgler 2003; Alm and Torgler2006).
Graph 3 provides a comparison of tax morale levels of 12 EU countries with the same method.
It shows the frequency of response for “tax cheating is never justifiable” with respect to the
total respondents and the EU average value for scaled response6. The descriptive analysis
reveals the percentage of individuals in each EU country stating that ‘tax evasion is never
justifiable’ (i.e., those with the highest level of tax morale) and the mean level of tax morale
among EU countries. Turkey with 80.50% shows a higher proportion of never justifiable
response than that of EU average (57%). The average score obtained for the tax morale
question across all EU countries in 2005/2007 was 2.2 (out of 10). The lowest level of tax
morale is observed for France and Norway.
6 Selected countries in WVS 2005 are: Bulgaria [2006], Finland [2005], France [2006], Germany [2006], Great Britain [2006], Italy [2005], Netherlands [2006], Norway [2007], Poland [2005], Romania [2005], Spain [2007], Sweden [2006] and Turkey [2007].
Recep Tekeli Policy Research Institute 11
11 Determinants of Tax Morale in Japan and in Turkey
Graph 4 provides a comparison of Tax Morale levels of 15 Asia Pacific Economic
Cooperation (APEC) Countries. It also shows that in 2005 the average number of people in
Japan stating that tax cheating is never justified was 82.90%, which was far above APEC
average of 61.80%. The average score obtained for the tax morale question across all APEC
countries in 2005/2007 was 2.1 (out of 10). The lowest level of tax morale is observed for
Malaysia and Thailand. As we can see there is a strong difference among APEC countries
with the highest tax morale.7
Table 3 shows the range of responses varied from 1.5 to 3.5. The country most opposed to tax
cheating is Japan with a mean score of 1.5 followed by South Korea and Indonesia. Malaysia
has the highest score at 3.5 implying that Malaysian has the least opposition to tax cheating.
7 APEC currently has 21 members, including most countries with a coastline on the Pacific Ocean. As Taiwan joined with full approval of mainland China under the name Chinese Taipei, the APEC uses the term Member Economies rather than member countries to refer to its members. Therefore we also used Member Economies rather than member countries. Selected countries/samples are Australia [2005], Canada [2006], Chile [2006], China [2007], Hong Kong, China [2005], Indonesia [2006], Japan [2005], Malaysia [2006], Mexico [2005], New Zealand [2004], Russian Federation [2006], South Korea [2005], Taiwan [2006], Thailand [2007], United States [2006], Viet Nam [2006] http://www.wvsevsdb.com/wvs/WVSAnalizeQuestion.jsp(04-07-2011)
Recep Tekeli Policy Research Institute 12
12 Determinants of Tax Morale in Japan and in Turkey
Table 3: Tax Morale in APEC Countries
APEC Countries cheating never justified mean score
Japan 82.90% 1.5 South Korea 74.20% 1.6 Indonesia 79.30% 1.6 Viet Nam 76.90% 1.7 Canada 66.10% 1.9 Chile 66.20% 1.9 Hong Kong 61.40% 1.9 Australia 63.90% 2 China 63.40% 2 Taiwan 61.90% 2 United States 63.80% 2.1 New Zealand 60.20% 2.2 Mexico 60.10% 2.6 Thailand 29.00% 2.8 Russian Federation 51.30% 3 Malaysia 26.90% 3.5
Mean 61.80% 2.1
Note: Data on tax cheating was not available for all OECD countries.
Recep Tekeli Policy Research Institute 13
13 Determinants of Tax Morale in Japan and in Turkey
Table 4 presents a basic descriptive analysis showing the percentage of individuals in each
OECD country saying that ‘tax cheating is never justifiable’ (that is, those with the highest
level of tax morale). Table 4 also presents the mean level of tax morale in Turkey and in
Japan in relation to other OECD countries.
Table 4: Tax Morale in OECD Countries
OECD Countries Never justifiable Mean Scores
France 47.90% 2.8 Norway 49.70% 2.3 Sweden 53.50% 2.2 Poland 55.10% 2.4 Finland 56.20% 2.1 Germany 56.60% 2.2 Great Britain 57.80% 2.3 Mexico 60.10% 2.6 New Zealand 60.20% 2.2 Netherlands 61.10% 2.3 Italy 61.50% 2.2 Switzerland 61.80% 2 United States 63.80% 2.1 Australia 63.90% 2 Spain 64.80% 2.1 Canada 66.10% 1.9 Chile 66.20% 1.9 South Korea 74.20% 1.6
Turkey 80.50% 1.3
Japan 82.90% 1.5 Mean 62.60% 2.1
First of all, it can be seen that in 2005 wave of WVS, the average number of people among all
OECD countries saying that tax cheating was never justified was 62.6 per cent. Further the
average score obtained for the tax morale question across all OECD countries was 2.1. When
comparing the Japanese and Turkish figures to those of the rest of the OECD, it can be seen in
Table 4 that both countries’ level of tax morale in 2005 and in 2007 (82.90 and 80.50 percent
respectively) was substantially well above the OECD average (62.60%).
In sum, the results for the magnitude of tax morale show that Japan and Turkey rank in the
highest as compared to other countries within their regions. Thus, this gives a task to explain
why tax morale is very high in these two countries; what determines tax morale and are there
any similarities between these two countries in the determination of tax morale level.
Recep Tekeli Policy Research Institute 14
14 Determinants of Tax Morale in Japan and in Turkey
The tax systems of the study countries
Although two countries with very different cultures and a different history and politics, a
comparison of tax morale and compliance between Turkey and Japan as a member of OECD
constitutes a good experiment since the tax systems themselves are broadly similar.
The elements of the personal income tax (PIT) systems in both countries are summarized in
Table 5. For the PIT (and Corporate Income Tax, CIT), the self-assessment are similar in both
countries although there are varying degrees of aggressiveness in tax enforcement. Both
countries rely on some form of tax withholding at source, the stamp payment system, official
assessment system and individual self-assessment and reporting of final tax liabilities. Tax
evasion seemed to be treated as more serious crime in Turkey than in Japan even though
general tax amnesty was conducted three times in the last decade (2003, 2008 and 2011) in
Turkey.
Table 5: Features of the tax system in the study countries Tax feature Turkey Japan Self reporting/assessment Yes Yes Withholding Yes Yes Stamp payment Yes Yes Official assessment Yes Yes Tax rates (%) PIT: 15, 20, 27, 35
CIT: 20 PIT: 10, 20, 30, 37 CIT: 22-30
The respective computations of the PIT and CIT bases are similar. In Japan, PIT and CIT are
taxed separately as in Turkey. The PIT base consists of wages and salaries as well as earnings
from interest and dividends (but these are taxed at the source). The marginal rate for PIT is
capped at 35 percent in Turkey, lower than the top rate in Japan (37 percent). The corporate
tax rates in Japan varies from 34.5 %(30% effective) for ordinary companies with large scale
to 25 %(effective 22%) for small size corporate and cooperative associations and public
service corporation. However in Turkey there is a basic corporate tax rate of 20 % applied to
all size corporations.
Both countries have some form of administrative and judicial penalties imposed for failure to
file if taxes are owed. Administrative penalties consist of interest and a penalty up to the tax
owed/evaded. Judicial penalties also can be imposed for tax evasion/fraud. In Japan there is a
delinquent tax fee at the rate of 14.6% per annum and substantial additional tax imposed at the
rate of 35% of the tax amount as a penalty for evasion. In case of fraud additional tax is
imposed at the rate of 40% of the tax amount (see Table 6). There is a judicial penalty
imposed against tax crimes in connection with the assessment and collection of tax.
Recep Tekeli Policy Research Institute 15
15 Determinants of Tax Morale in Japan and in Turkey
According to the Criminal Procedures Law, tax evasion that constitutes a crime is investigated
by police or prosecutors and is examined and tried in court in Japan.8
The Turkish penalty structure seems more severe. In addition to a delinquent tax fee at the
rate of 1.40 percent per month, a tax payer or a person responsible for a tax liability required
to file a return who fails to do so within the period mentioned in the law is liable to a penalty
equal to the tax owed (see Table 6). In case of fraud, taxable income may be estimated and
three times the amount of tax charged thereon. A tax payer is also liable to imprisonment for a
period of 18 months to 3 years. However, a taxpayer who knowingly and willfully makes
false statements in the return or evades or attempts to evade taxation (and any person who
assists a taxpayer to do so) by providing false documents and bookkeeping is liable to a
penalty equal to three times the tax owed/evaded and to imprisonment for a period of 3 to 5
years. As displayed in Table 6 the tax enforcement means differs considerably between
Turkey and Japan.
Table 6: Penalty structures in Japan and Turkey Japan Turkey Evasion without
intention in case of fraud Evasion without
intention in case of fraud
Timely filing but underreporting
10-15% 35% 100% 300%
Non-filing 15% 40% 100% 300% Tax Bureau, MOF (2005:210)
2.2 Determinants
After having discussed the magnitude of tax morale in EU, OECD and APEC countries, the
question of the determinants of tax morale in Turkey and in Japan arises. German scholars led
by Günter Schmölders (1960, 1962, 1970) known as the ‘Cologne school of tax psychology’
have emphasized that economic phenomena should not only be analyzed from the traditional
point of view with a dominance of price- and incentive mechanisms. In their surveys, using
the subjective tax burden as an indicator for the level of tax morale, they found that self-
employed people reveal lower tax morale than employees. On the other hand, in the last 15
years, with the excellent works of Torgler, the main research interest has shifted to discover
the reasons why people pay taxes? At first, this question has seemed as an unorthodox
research question. However, taking in account that many people pay their taxes although fines
8 It should be mentioned that in Japan “Under the law, a considerable difference exists between the handling of the evasion of direct taxes and that of indirect taxes”. (MOF 2005:209)
Recep Tekeli Policy Research Institute 16
16 Determinants of Tax Morale in Japan and in Turkey
and audit probability are low, the question seemed not only reasonable but also crucial for the
design of a country’s efficient tax system (Erard and Feinstein, 1994). (see Torgler and
Shaltegger 2006)
The results indicate that the high level of tax compliance cannot be explained entirely by the
level of enforcement (Graetz and Wilde, 1985). It is a long way before a person becomes a tax
evader. For example, Elffers (2000) sees the process of tax evasion as a staircase with three
different steps. First, taxpayers have to be seized by a will ‘not to comply’. Some researchers
have argued that many taxpayers do not even think of tax evasion and hence do not even
search for ways to cheat at taxes. In Long and Swingen’s words (1991:130) some taxpayers
are “simply predisposed not to evade”. Some researchers findings indicate that there are
individuals who always comply (Alm 1999). In a second step, Elffers (2000:187) goes on to
argue that not everyone with “an inclination to dodge his taxes is able to translate his intention
into action”, because many individuals do not have the opportunity or the knowledge and
resources to evade. In the last step, you can find taxpayers that feel inclined ‘not to comply’
and check for the opportunity to evade taxes. Elffers’s argument is that this is the phase where
standard economic theory works, where taxpayers evaluate the expected value of tax evasion.
Thus, the focus has been on the lower level (willingness step) where attitudes regarding tax
evasion (i.e. tax morale) play an essential role. (see Torgler 2003:290)
2.2.1 Model and Variables
The empirical analysis in this paper focuses on tax morale which is influenced by a variety of
factors. In empirical analysis social, demographic and economic factors as well as national
pride or religiosity play a role in determining tax morale in a society. Also, “there is the
institutional arrangement in which the government works. Here, the extent of democratic
participation (possibilities) by taxpayers as well as the level of institutional and political trust
are decisive. Tax morale is driven by the acceptance of government decisions.” (see Torgler
and Shaltegger (2006:397). Following discussions explain the variables and indicate the
expectations of the coefficient signs.
Elffers (2000), as already discussed above, argues that it is important that people remain at the
lower level of the staircase to tax evasion. Thus, it might be interesting to focus on the
“willingness step”, which means that attitudes regarding tax evasion play an essential role.
Thus, the empirical analyses focus on the attitudes regarding tax evasion which is simply
defined as tax morale and data on Japan and Turkey from 2005 World Values Surveys are
Recep Tekeli Policy Research Institute 17
17 Determinants of Tax Morale in Japan and in Turkey
evaluated (see Inglehart et al., 2000). Our main model for predicting tax morale in the study
countries has the following structure:
where TM denotes the individual degree of tax morale in a respected country; Our key
independent variables TRUST denotes individual’s trust in authority; PRIDE denotes pride of
individual; POL is individual’s political and pro-democratic attitudes, HAPPY denotes
individual’s well being (happiness) and X is a vector of variables which denotes numerous
socioeconomics and control variables such as age, gender, marital status, education,
awareness, employment and occupational status. Our choice of independent variables is in
line with the literature. Torgler and Shaltegger (2006) discuss the studies using the factors to
explain the determinants of tax morale in various countries. In Tale 7 summary of survey
studies in different regions are presented with the variables and signs.
Table 7: Literature: Determinants of Tax Morale (excerpted from Torgler and Shaltegger 2006:415)
Dependent Variable Tax Morale
Regions Europe North and Latin America
Transition, Asian Countries and Australia
Independent Variables Age + + + Gender (female) + + + Education (-/+) (+) + Married + + (+) Self-employed - (+) - Income - - - Financial Satisfaction + + + Church Attendance + + +(Asia) Religious + + (-/+) Trust in Government + + + Trust in Legal System + + Trust in court/legal system
+
Trust in the president + Trust public officials + + Pride + + Direct democracy +(Switzerland) +USA Pro-democratic attitudes
+ + +
Local autonomy + Notes: Summary of survey studies in different regions. Tendencies: +: significant positive coefficient, (+), (-), positive, respectively negative coefficient sign without being (consistently) significant, (±) positive and negative sign of the coefficient without being (consistently) significant.
i
k
kiiiiXHAPPYPOLTRUSTTM
133i21
PRIDE
Recep Tekeli Policy Research Institute 18
18 Determinants of Tax Morale in Japan and in Turkey
Dependent Variable
The dependent variable in our study is Tax Morale (TMi). The general question to assess level
of tax morale from the World Values Survey 2005– and for which has been used in several
previous research studies is already described above.
The WVS has been broadly used by researchers and the applied econometric method is
Ordered Probit models. The tax morale variable is scaled dependent variable and hence the
ordered probit models help to analyze the ranking information of this variable (Torgler 2006).
Following Cummings et al (2009:451-452) answers of “don’t know” and missing values were
not coded and were dropped from the sample. In the model the estimated equation has a non-
linear form and only the sign of the coefficient can be directly interpreted, not its coefficient
as in ordinary regression equation. Calculating the marginal effects is therefore a method to
find the quantitative effect of independent variable has on tax morale9. The marginal effect
indicates the change in the share of taxpayers (or the probability of) belonging to a specific
tax morale level (i.e. 0,1,2,3) when the independent variable increases by exactly one unit. We
will present only the marginal effects for the highest value - i.e. tax evasion is never justified-
from the ordered probit estimation.10
Table 8: Tax Morale
Country Justifiable:
cheating on taxes Japan (2005)
Turkey (2007)
Never justifiable 82.90% 80.50% 2 6.00% 12.50% 3 4.50% 4.00% 4 2.60% 1.10% 5 2.20% 1.10% 6 0.10% 0.20% 7 0.30% 0.20% 8 0.40% 0.10% 9 0.10% 0.10%
Always justifiable 0.90% 0.30% Total 1075
(100%) 1341
(100%) Mean 1.5 1.3
Standard Deviation 1.32 0.97
9 Bruin (2006) offers the “margin” command to find the marginal effect of a specific variable in the regression. 10 The empirical results on the tax morale can be given on request from the author.
Recep Tekeli Policy Research Institute 19
19 Determinants of Tax Morale in Japan and in Turkey
Table 8 shows the frequency of response for “tax cheating is never justifiable” with respect to
the total respondents, the average value and standard deviation for scaled response. Japan
shows a higher proportion of never justifiable response than that of Turkey. Although mean
response in Japan is slightly higher than that in Turkey, the mean scores appeared to be
basically the same; 1.5 for Japan and 1.3 for Turkey. On a scale from 1 to 10, it does not seem
like much of a difference. The Japanese sample had the highest percentage (82.90%) response
for that position but both countries had responses for the “never justifiable” position above
eighty percent.
Independent Variables
Our independent variables are described below.
TRUST: There is no doubt that the socio-demographic and the socio-economic background
of individuals affect tax morale. Torgler (2001) argued that there were a number of key
factors that seemed to be important for understanding tax morale. Two of these factors
included the relationship taxpayers had with their Government (ie, whether they trusted them
or not), and taxpayers’ moral rules and sentiments. Torgler (2005) confirmed that these
concepts predicted levels of tax morale in Latin America as well as in other countries (see
also Torgler and Murphy 2004). In order to see whether these variables were important in
understanding tax morale in the study countries, three questions that most closely measured
these concepts were chosen from the World Values Survey and used as predictors of tax
morale: trust in the government, trust in the legal system and trust in the parliament.
Trust in the government has been measured as follows:
Could you tell me how much confidence you have in the government in your capital: is it a great deal of confidence, quite a lot of confidence, not very much confidence or none at all? (1= “a great deal” to 4= “none at all”).
The other variables used to explore the relationship taxpayers had with authority are trust in
the legal system and trust in the parliament. Trust in the legal system was measured in the
World Values Survey as follows:
Could you tell me how much confidence you have in the legal system: is it a great deal of confidence, quite a lot of confidence, not very much confidence or none at all? (1= “a great deal” to 4= “none at all”).
Trust in the parliament was measured in the World Values Survey as follows:
Could you tell me how much confidence you have in the Parliament: is it a great deal of confidence, quite a lot of confidence, not very much confidence or none
Recep Tekeli Policy Research Institute 20
20 Determinants of Tax Morale in Japan and in Turkey
at all? (1= “a great deal” to 4= “none at all”).
These three variables allow us to analyze trust at the constitutional level, thereby focusing on
how the relationship between the State and its citizens is established, and also allow us to
analyze trust more closely at the current politico-economic level (see Alm and Torgler
2006:236). Torgler and Murphy (2004:309) see the relationship between taxpayers and the
State as “a relational contract or psychological contract, which involves strong emotional ties
and loyalties”. If citizens/taxpayers trust their government, representatives and justice system,
because the State is seen to be acting in a trustworthy manner, taxpayers/citizens might be
more willing to comply with their tax obligations.
The tax authority cannot achieve compliance rate completely; they would have to place a
taxman under every bed. In this respect, trust especially plays an important role where
detection and punishment are achieved by high cost. This might indicate that trust is an
important institution, which influences citizens’ incentive to commit themselves to obedience.
The findings by Scholz and Lubell (1998) from America and by Murphy (2004) from
Australian experience have led Torgler and Murphy (2004:310) to indicate that trust is an
important institution, which influences citizens’ incentives to commit themselves to
obedience. If this is indeed the case, one would therefore expect that those who have more
trust in the legal system, government and/or in Parliament might also have higher levels of tax
morale.
Furthermore, Torgler and Shaltegger (2006:410) assert that “trust is not an attitude that can be
demanded by the government, taxpayers bestow trust on their government or they don’t”. In
similar arguments, Torgler (2004:243) points out that “trust can only be created if the
government acts in line with citizens’ needs and desires…”.
In the lights of above arguments, the following hypothesis is going to be tested:
Hypothesis 1. The more extensive the citizens’ trust in the government, in the parliament and the legal system (justice or the court), the higher the tax morale.
We expect that trust in the government, the legal system and the parliament has a strong
impact on tax morale. In other words the willingness to pay taxes is highly influenced by the
positive treatments of taxpayers by the authorities.
Table 9 shows the frequency of response for “confidence in Justice System” with respect to
the total respondents. In Japan participants have higher confidence in justice system than in
Turkey. In Turkey 24.3% of respondents are not confident with justice system while 18% of
respondents are not confident with their justice system in Japan.
Recep Tekeli Policy Research Institute 21
21 Determinants of Tax Morale in Japan and in Turkey
Table 9: Trust Justice System
Country Confidence in Justice System Japan Turkey
A great deal 17.20% 36.80% Quite a lot 64.80% 38.90%
Not very much 15.50% 16.30% None at all 2.50% 8.00%
Total 1031 (100%)
1313 (100%)
Table 10 shows the frequency of response for “confidence in parliament” with respect to the
total respondents. In Turkey higher percentage of participants responded for “a great deal”
and “quite a lot” category than that of Japan. Japanese participants are appeared to be
unconfident with the political system especially for the parliament.
Table 10: Trust Parliament
Country Confidence in Parliament Japan Turkey
A great deal 1.20% 22.70% Quite a lot 22.00% 38.80%
Not very much 55.40% 24.00% None at all 21.30% 14.60%
Total 1012 (100%)
1277 (100%)
Table 11 shows the frequency of response for “confidence in the government ” with respect to
the total respondents. In the table Japanese participants seem to be less confident with their
government than Turkish participants. It is also consistent with the result for the confidence in
Parliament. In Turkey it is quite likely that people who are confident with the parliament they
also indicate higher confidence in the government. We can say that both countries quite differ
in politics.
Table 11: Trust The Government
Country Confidence in The Government Japan Turkey
A great deal 1.60% 27.60% Quite a lot 29.50% 36.00%
Not very much 50.20% 21.80% None at all 18.80% 14.50%
Total 1028 (100%)
1304 (100%)
Recep Tekeli Policy Research Institute 22
22 Determinants of Tax Morale in Japan and in Turkey
PRIDE: Although pride is a widespread phenomenon, pride and its effect on tax cheating or
tax evasion have not been discussed very much in economic literature. However, many results
are provided in recent literature. Boulding (1992:93) states that “The dynamics which governs
the creation, destruction, and distribution of various forms of pride and shame in society are
very little understood, yet nothing perhaps is more crucial to the understanding of the overall
dynamics of a particular society than the marked differences which exist among societies in
this regard’’ (excerpted from Torgler 2004:243 also see Torgler and Shaltegger 2006:412).
Everybody could be proud of own country. Pride produces a sense of group identification
which can be found in international games such as the FIFA/FIBA World Cup. Tyler (2000)
“argues that pride influences people’s behaviour in groups, organizations and societies. It
gives a basis for encouraging cooperative behavior” (see Torgler 2004:243; Torgler and
Shaltegger 2006:412). To analyze the effects of pride on tax cheating (or tax morale) it can be
hypothesized that people who are proud to be citizens of their country are more loyal, better
identified with the state and have thus a higher tax morale.
The question that measures national pride is:
How proud are you to be . . .? (specific nationality, e.g., ‘Japanese, Turkish, Korean, Dutch etc.’) (1 = very proud, 4 =not at all proud).
Hypothesis 2. The greater the citizens’ national pride, the higher the tax morale.
Table 12 shows the frequency of response for “how proud of nationality ” with respect to the
total respondents. Turkey shows a higher proportion of “very proud” response than that of
Japan. Almost 40 percent of participants in the survey are responded “not proud” of Japan,
while only 3.5 percent of participants responded for “not proud” of own country in Turkey.
Table 12: Pride.
Country How proud of nationality Japan Turkey
Very proud 22.20% 81.30% Quite proud 38.80% 15.10%
Not very proud 35.10% 2.90% Not at all proud 3.80% 0.60%
Total 1030 (100%)
1264 (100%)
POLITICAL ATTITUDES: Because of political attitudes (both partisan and ideological
positions) and/ or a general distrust of political institutions and a lack of confidence in their
own country’s institutions, tax payers might less comply with their own country’s taxes. For
Recep Tekeli Policy Research Institute 23
23 Determinants of Tax Morale in Japan and in Turkey
example, Penas and Penas (2008:3) argued that “Individuals who defend the limitation of
government’s revenues and hence a protection of parts of their own income and wealth from
taxation are probably more likely to cheat on their taxes”. Although political disaffection and
lack of confidence in the political process, politicians, and democratic institutions (Montero
and Torcal 2006:6) might affect tax morale, there are other political attitudes which affect tax
morale i.e. ideological biases (being left or right in politics). (see Penas and Penas 2008)
We test political attitudes by a variable ideology, which is the self-placement in a left-right
ideological scale from 1 (left) to 10 (right). It was measured as follows:
How would you place yourself in a political scale? (1=extreme left, 10=extreme right)
We do not expect a priori sign on the coefficient of this variable.
Table 13 shows the frequency of response for “self positioning in political scale” with respect
to the total respondents. In Japan, participants’ answer is indicating the center in political
scale(52.7%) but in Turkey participants are appeared to be skewed to the centre right in
politics (see the mean score is 6.2) as many people are conservatives and traditional values are
still highly respected.
Table 13: Ideology
Country Self positioning in political scale Japan Turkey
Left 2.40% 7.90% 2 2.20% 3.90% 3 9.00% 7.00% 4 9.70% 6.00% 5 27.10% 12.90%6 25.60% 18.80%7 9.50% 7.70% 8 10.30% 13.50%9 2.10% 6.60%
Right 2.00% 15.70%Total 863
(100%)1143
(100%)Mean 5.5 6.2
It might also be interesting to analyze whether a pro democratic attitude has a positive effect
(if any) on tax morale. In general, democracy offers citizens the possibility to express their
preferences. In a democratic countries citizens are able to better monitor and control
politicians and thus to reduce the asymmetry of information between them and their
Recep Tekeli Policy Research Institute 24
24 Determinants of Tax Morale in Japan and in Turkey
representatives. Torgler and Shaltegger (2006:413) shows that a higher pro democratic
attitude leads to higher tax morale in Latin America and in transition countries as well. As
observed in transition countries, a move toward more democracy might help to enhance tax
morale and civic virtue over time. (see Torgler and Shaltegger 2006 :413)
The degree of democratization might influence citizens’ tax morale and thus we have built a
variable that measures individuals’ support for democratic government (Pro-Democracy). The
question is:
Would you say that our country is governed by a democratic political system? (1= none at all (not democratically governed), 10=country is completely governed by a democracy).
Hypothesis 3. A stronger pro democratic attitude leads to a higher tax morale.
Satisfaction with the way democracy works could also be a measure of political disaffection.
If tax payers are politically discontent then they might less comply with taxes. For example,
Feld and Frey (2007) argue that tax morale is boosted if political processes perceived as fair
and legitimate.
Table 14 shows the frequency of response for “democraticness in own country” with respect
to the total respondents. Majority of respondents in Japan believes that their county is being
governed democratically. Percentage of participants who believe the country is not
democratic at all is higher in Turkey (26.50%) than that of Japan (7%). So in Turkey the level
of democracy is believed to be lower than in Japan.
Table 14: Democraticness
Country Democraticness in own country Japan Turkey
Not at all democratic 0.80% 6.90% 2 0.60% 4.20% 3 3.00% 7.10% 4 2.60% 8.30% 5 8.00% 11.20% 6 23.00% 16.50% 7 24.90% 15.50% 8 24.50% 13.80% 9 9.30% 7.70%
Completely democratic 3.10% 8.80% Total 998 (100%) 1277 (100%)Mean 6.9 6
Recep Tekeli Policy Research Institute 25
25 Determinants of Tax Morale in Japan and in Turkey
HAPPINESS: Frey and Stutzer (2002) summarize that the extent of happiness may influence
many important economic decisions e.g. consumption activities, work behavior, investment
behavior and political behavior. In this respect, it might also be interesting to analyze whether
the citizens’ happiness (i.e. individual’s well being) affect the tax morale or not. Recent
literature on happiness is developing, but in tax morale research there are not many results
provided. Only one exception is the study by Torgler (2004) who has included the happiness
variable in order to consider the fact that in some Asian countries such as India and the
Philippines many citizens have a low living standard. Here we assume that one’s happiness
level affects the moral attitude towards tax compliance. Thus we would like to find an answer
to whether happier people are more likely to report honestly, or not?
In the WVS happiness was measured by this question:
Taking all things together, would you say you are: very happy (4), quite happy (3), not very happy (2), not at all happy (1)
Hypothesis 4. Tax morale increases with individual’s well being.
Table 15 shows the frequency of response for “ feeling of happiness” with respect to the total
respondents. In both countries 85-90% per cent of participants were happy with the current
life. However, in Turkey the proportion of “not all happy” responses are higher than in Japan.
Table 15: Happiness Country Feeling of
happiness Japan Turkey Very happy 29.20% 37.40% Quite happy 60.50% 48.90%
Not very happy 9.20% 9.40%
Not at all happy 1.10% 4.30% Total 1066 (100%) 1345 (100%)
CONTROL VARIABLES:
In addition to the main independent variables discussed above we also use additional
independent variables as controls to more fully explore what factors might determine tax
morale in the study countries. Further control variables are age, gender, education, awareness,
marital status, economic and occupation status and religiosity. Each of these variables and
expected coefficient are discussed below:
Recep Tekeli Policy Research Institute 26
26 Determinants of Tax Morale in Japan and in Turkey
1. Age. We will use age as a continuous variable and also treat age as a categorical variable
(three classes are formed in the survey: 18-29, 30-49, 50+ with 18-29 as the reference group)
Do older people have higher tax morale than younger ?
Tittle (1980) argues that older people (aged greater than 65) are more sensitive to the threat of
sanctions (both social and financial) from others in the society. Although the findings of the
tax compliance studies show that the impact of age on compliance is still uncertain, some
researchers have found that older taxpayers are more compliant (see Torgler 2006:88 and
Torgler and Murphy 2004). It has been suggested that this is because older people while aging
have acquired more material goods, have obtained greater status in their community, and have
a stronger dependency on the reactions from others (Torgler and Murphy 2004:312). Thus, the
potential costs of sanctions from others increase for this group (see Torgler 2006:88). Here we
are concerned whether this sensitivity to sanction among older taxpayers also affects level of
tax morale or not.
On the other hand, it can be argued that elderly people might be more experienced in tax
matters, which might reduce their tax compliance. But we do not expect that happens for two
reasons. First one is elder’s attachment to the community; If they have been living for a
certain period of time in the same place they are more attached to the community
(Pommerehne and Weck- Hannemann, 1996) (see Torgler and Schneider, 2005). Therefore,
attachment might have a positive effect on tax morale. Second one is as older people are
generally not subject to income taxes, they might have a different attitude towards tax
compliance (or higher tax morale). (Torgler 2004)
Table 16: Age.
Country Age Group Japan Turkey
15-24 7.00% 20.70% 25-34 16.50% 31.60% 35-44 17.40% 21.80% 45-54 20.80% 14.20% 55-64 21.50% 6.20%
65 and more years 16.70% 5.50% Total 1096
(100%) 1346
(100%) Mean 48.1 36.5
Table 16 shows the frequency of response for “age group” with respect to the total
respondents. While more than 50 percent of participants are in the first two categories of age
Recep Tekeli Policy Research Institute 27
27 Determinants of Tax Morale in Japan and in Turkey
(lower age groups ) in Turkey, in Japan 23.5% of participants are in lower age categories.
Also, percentage of elders in Japan is much higher than in Turkey.
2. Gender (categorical variable: 1= male (the reference group), 2=otherwise)
Do gender differences affect the level of tax morale? Social psychological research suggests
that women are more compliant and less self-reliant than men (e.g. Tittle 1980). Differences
in tax morale levels may be due to differences in gender values or to lower female labour
participation rates. Some researchers in the tax compliance literature found that women are
more compliant than counterpart (see Torgler 2004; Torgler 2006; Torgler and Schneider
2005, Torgler and Valev 2006)
Torgler and Valev (2006:5) concludes from Dollar et al (2001) paper that there is an evidence
of a higher presence of women parliamentarians had a statistically significant negative impact
on corruption. Swammy et al (2001) also find that if women participation in parliament, in
government ministers and in the labor force increases, then corruption will decrease.
Experimental research findings indicate that gender may influence various behaviors such as
charitable giving, bargaining, and household decision making. Strong differences between
men and women are also observed in accident involvements, alcohol and drug abuse. The
criminology literature actually gives a basis to explain possible gender differences. (Torgler
and Valev 2006:5)
Table 17 shows the frequency of response for “gender” with respect to the total respondents.
In Japan female respondents are higher than in Turkey. In Turkey almost equal percentage of
gender participated in the survey. Survey sample also represents the national distribution of
sex in Turkey. In 2007 census, female comprises 49.88% of total populations while male
comprises 50.12% of total populations.
Table 17: Gender
Country Sex Japan Turkey
Male 44.10% 50.20% Female 55.90% 49.80%
Total 1096 (100%)
1346 (100%)
Recep Tekeli Policy Research Institute 28
28 Determinants of Tax Morale in Japan and in Turkey
3. Marital status [married (in the reference group), single, living together, divorced,
separated, widowed)
Do married people have different levels of tax morale than others? An individual’s marital
status might influence his/her behavior. It was argued that married people are more compliant
than others, especially compared to singles because they are more constrained by their social
networks (Torgler and Murphy 2004; Torgler and Schneider, 2005). The argumentation for
this is stated by Tittle (1980:111) as: “A long tradition in sociology, extending back to
Durkheim, postulates that proneness toward rule breaking varies inversely with the extent to
which individuals are involved in social networks with constraining content”. Tittle (1980)
found significant differences among the different marital status, with the greatest evidence for
the singles, followed by the separated or divorced. (see Torgler 2004; Torgler 2006).
In the tax compliance literature Torgler (2006:89) stated that “some studies have found that
non-compliance is more common and of greater magnitude among married taxpayers (see
Clotfelter, 1983; Feinstein, 1991)”. This is due to the tax system. If dual incomes are treated
as one (as in the US), “being thus taxed in a higher bracket than two separate incomes,
compliance might be lower (Hays 2000)” (see Torgler 2006:89). However this is not the case
in the study countries and there is still further evidence needed to clearly understand the
correlation between tax morale and marital status.
Table 18 shows the frequency of response for “marital status” with respect to the total
respondents. In Japan higher percentage of participants are responded for “married” category
than in Turkey. In Turkey the proportion of single/never married category is much higher than
that of Japan. Divorce rate seems to be lower in Turkey than in Japan. Living together as a
married couple does not seem to be common life style in Turkish society.
Table 18: Marital status.
Country Marital status Japan Turkey
Married 72.80% 66.40% Living together as married 1.70% 0.10%
Divorced 3.00% 1.40% Separated 0.60% 0.20% Widowed 4.70% 2.30%
Single/Never married 17.20% 29.60% Total 1090
(100%) 1346
(100%)
Recep Tekeli Policy Research Institute 29
29 Determinants of Tax Morale in Japan and in Turkey
Table 19: Educational level
Country Highest educational level attained Japan Turkey No formal education - 8.30%
Inadequately completed elementary education - 3.80%
Completed (compulsory) elementary education 8.20% 36.60%
Incomplete secondary school: technical/vocational type/(Compulsory) elementary education and basic vocational qualification
1.30% 2.60%
Complete secondary school: technical/vocational type/Secondary, intermediate vocational qualification
11.50% 6.90%
Incomplete secondary: university-preparatory type/Secondary, intermediate general qualification
4.00% 3.30%
Complete secondary: university-preparatory type/Full secondary, maturity level certificate
48.00% 26.70%
Some university without degree/Higher education - lower-level tertiary certificate
1.90% 1.00%
University with degree/Higher education - upper-level tertiary certificate
25.10% 10.80%
Total 1072 (100%)
1346 (100%)
4. Education [continuous variable for higher educational level attained: 1= low (never gone
to the school), 10 = higher education)]
Do more educated people have higher tax morale level than less educated people? “Education
is related to taxpayer’s knowledge about the tax law. Better educated taxpayers are supposed
to know more about tax law and fiscal connections, and thus would be in a better position to
assess the degree of compliance (see Lewis, 1982)” (see Torgler, 2006:89). Therefore
educated taxpayers might be more aware of the public services and benefits. If this
assumption is correct, then, we might expect that better educated taxpayers would be more
compliant with their tax obligations. This assumption, however, cannot be automatically
accepted because the degree of knowledge involving evasion or avoidance opportunities
might increase with the education level. Hence we might expect that better educated taxpayers
may be less compliant because they better understand opportunities for tax evasion and
avoidance and might be better aware of how the State uses tax revenues (see Torgler 2006:89;
Recep Tekeli Policy Research Institute 30
30 Determinants of Tax Morale in Japan and in Turkey
Torgler and Murphy 2004; Torgler and Schneider 2005). Thus, as the effect of education on
tax compliance is not clear, further empirical results are needed to distinguish which effects
are stronger.
Table 19 shows the frequency of response for “highest educational level attained” with
respect to the total respondents. One interesting result from the table is that 8.3% of
participants have not attained any formal education in Turkey. Percentage of participants
attained higher education level is higher in Japan than in Turkey; 25.1% of participants in
Japan have graduated from the university, while 10.80% of participants hold a university
diploma in Turkey.
Current literature on tax morale has only used formal education as a proxy of such awareness.
Therefore, Torgler and Shaltegger (2006:409) suggested that future studies should take a look
at informal education such as time spent for reading a newspaper to investigate whether
awareness affects tax morale or not. Hence we also constructed a variable to measure such
awareness to see its effect on tax morale.
Awareness is a categorical variable. It was measured by following question:
People use different sources to learn what is going on in their country and the world.
For each of the following sources, please indicate whether you used it last week or did
not use it last week to obtain information: Daily newspaper? (categorical variable:
1=yes, 2=no)
Table 20 shows the frequency of response for “ information source ” with respect to the total
respondents. In Japan 90.1 percent of participants used a daily newspaper as information
source while in Turkey 58.60% of participants used a daily news paper which is quite lower
than Japanese rate.
Table 20: Awareness.
Country
Information source: Daily newspaper Japan Turkey Used last week 90.10% 58.60%
Not used last week 9.90% 41.40% Total 1080 (100%) 1343 (100%)
5. Economic class [we used scale of incomes: 1=low income, 10=high income].
Do the levels of tax morale vary with the level of tax payers’ income? Although some
research into tax morale has also shown that a person’s economic situation can affect their
level of tax morale, the effects of income on tax morale are difficult to assess theoretically
Recep Tekeli Policy Research Institute 31
31 Determinants of Tax Morale in Japan and in Turkey
(Torgler 2004; Torgler and Schneider 2005:237; Torgler and Shaltegger 2006:410). This is
because, as argued by Torgler and Murphy (2004:31), “the findings have been mixed. The
findings have also been found to depend on an individual’s risk preference and the
progression of the income tax schedule” 11.
If the country has a progressive income tax schedule, taxpayers with a higher income earn a
higher income return by evading [i.e. a higher marginal tax rate makes tax evasion more
profitable]. On the other hand, taxpayers with a lower income might have lower social
“stakes” or restrictions but they are less in the position to take these risks because of a high
marginal utility loss (wealth reduction) if they are caught and penalized (See Torgler 2006:90;
Torgler and Schneider 2005:232). Also, as argued by Alm and Torgler (2006:237) that “a
rational choice theory of crime predicts that individuals in lower income classes are more
likely to engage in criminal activities due to their lower opportunity costs”. As a result we do
not expect the sign of coefficient on income variable in a priori. Here we will give more
evidence which effects are stronger.
Table 21: Incomes.
Country Scale of incomes Japan Turkey
Lower step 14.80% 22.90% second step 13.80% 26.60%
Third step 13.50% 14.10% Fourth step 11.40% 14.80%
Fifth step 9.10% 3.20% Sixth step 9.20% 7.50%
Seventh step 7.80% 1.30% Eigth step 5.80% 5.80%
Nineth step 7.30% 1.90% Tenth step 7.30% 1.90%
Total 1000 (100%) 1321 (100%)Mean 4.7 3.3
Table 21 shows the frequency of response for “scale of incomes” with respect to the total
respondents. Almost half of the participants fall into the first two low income scales in Turkey,
while in Japan about one third of the participants fall into first two low-scales of income
categories. Also percentage of participants who fall into higher income scales is higher in
11 For example, Jackson and Milliron (1986) argue that if the relationship between income level and compliance is not linear but curvilinear, then using linear models might produce biased estimates, and hence causing the difference in the empirical findings.
Recep Tekeli Policy Research Institute 32
32 Determinants of Tax Morale in Japan and in Turkey
Japan than in Turkey. It also shows the consistent feature that the mean value score for
response in Japan is higher than in Turkey.
6. Occupation status (full time employed (in the reference group), part time employed, selfemployed, unemployed, housewife, student, retired, other).
Does the occupation status of individual influence tax morale? The standard argument is that
self-employed taxpayers would have more opportunity to evade their taxes than taxpayers
who have their taxes deducted each payday by their employers (Torgler and Murphy
2004:313) 12 . Torgler (2006:90) concluded from Vogel’s survey in Sweden that “self-
employed taxpayers are more likely to think that large parts of taxes were used for
meaningless purposes, that the government had made a great number of unnecessary social
reforms, that they have had less benefit from government programs than the average taxpayer,
and that the burden of taxes was too high”. Also Lewis (1982) argued that self-employed have
higher compliance costs and hence taxes become more visible for them. Torgler and
Shaltegger’s (2006:409-10) survey findings show that the coefficient of occupation status is
not significant but in the transition countries the coefficient is mostly significant with a
negative sign. They argue that in these countries self-employed individuals might feel the
financial restriction much more, as the compliance costs and taxes become more visible.
Although, it is not easy to make a clear prediction about the influence of occupation status on
tax morale, in line with the findings of Schmölders (1960) in Europe and Torgler and
Shaltegger (2006:409-10) survey findings, we would expect that self-employed Turks and
Japanese would have a lower level of tax morale than those employed by others.
Table 22: Employment status.
Country Employment status. Japan Turkey
Full time 38.40% 22.20% Part time 15.50% 2.70%
Self employed 11.60% 14.50% Retired 9.90% 9.10%
Housewife 17.90% 34.60% Students 2.80% 8.40%
Unemployed 1.60% 6.10% Other 2.40% 2.40% Total 1055 (100%) 1346 (100%)
12 We should emphasize the difference between tax morale and tax evasion. As in Torgler (2006), here the arguments are related to tax evasion and not tax morale.
Recep Tekeli Policy Research Institute 33
33 Determinants of Tax Morale in Japan and in Turkey
Table 22 shows the frequency of response for “employment status” with respect to the total
respondents. Percentage of unemployed participants is higher in Turkey (6.1%) than in Japan
(1.6%). Because of the substantial young generation in Turkey, students’ proportion is higher
in Turkey than in Japan. Also one conclusion can be drawn from the table is that one third of
participants in Turkey is housewife indicating that women’s participation in the workforce is
quite low.
7. Religiosity
Does the religion make any differences to the levels of tax morale? To answer this question
we used religiosity as an additional independent variable that was analyzed in this study. In
his earlier studies Torgler (2001) argued that moral sentiments were important for
understanding tax morale. In the same vein Torgler and Murphy (2004:310) stated that “There
are many behavioral norms and moral constraints that are strongly influenced by religious
motivations”.
It was argued that religion might influence people’s habits and might have the function to
economize and simplify people’s actions. According to Torgler (2003:295) religion forms
habits of thought common to all human beings. While analyzing the question why morality
and religion are tied together, Margolis (1997) suggested that religion includes the belief in
the right behavior which has two components: “the performance of rituals, which is important
in a society and serves to bind society together, and right behavior in the secular sense of what
is fair and just” (see Torgler 2003:298 and Torgler 2006: 85).
According to Torgler (2006:84) and Torgler and Schneider (2007:449) religious organizations
provide moral social constitutions for a society and, to a certain extent, act as ‘‘supernatural
police’’ that enforce accepted rules. This is because religion can act as a sanctioning system
that “legitimizes and reinforces social values and may also inhibit illegal behavior (Hirschi
and Stark, 1969).” (see Torgler and Schneider 2007:449)
A negative correlation between religious membership and crime (less violent and non violent)
has been reported in criminology literature (see, Torgler and Schneider 2007:449; Alm and
Torgler 2006:237). Thus, in many religions, as religiosity affects the degree of rule breaking,
we might assume that it can be a restriction on tax evasion. The degree of religiosity could be
measured directly by asking. However, we include religiosity proxied by frequency of
religious service attendance. This is because it approximately shows how much time
individuals devote to religion. The frequency of religious service attendance and the
Recep Tekeli Policy Research Institute 34
34 Determinants of Tax Morale in Japan and in Turkey
involvement in a Mosque or a church or any other religious organization indicate that people
spend time devoted to religion. Torgler (2006:86) points Iannaccone (2002) who emphasizes
the aspect of time people devote to religion: “Attendance takes time, time that has an
opportunity cost because it preempts other activities”. Attendance to religious organizations is
also important because it involves ties to others in the societies.
In order to measure religiosity in our study, the following question was taken from the World
Values Survey 2005:
Apart from funerals, weddings, and christenings, about how often do you attend religious services these days? More than once a week, once a week, once a month, only on special holy days, once a year, less often, never or practically never (1=more than once a week to 7=never or practically never).
Based on the above arguments, we assume that Turks (and/or Japanese) who think of
themselves as religious might be more likely to have higher levels of tax morale than those
who do not consider themselves to be religious. Thus, as argued by Torgler and Murphy
(2004:310) “it might also be the case that religious beliefs might influence people’s habits in
the area of white collar crime”.
Table 23 shows the frequency of response for “attending religious services” with respect to
the total respondents. In Turkey higher percentage of participants are attending religious
services than in Japan. In general, Japanese participants are going to attend only on special
holy days or on a special occasion. However in Turkey about one third of the participants are
regularly attending religious services (i.e 5 times a day and Friday prayers), while at the other
side of the scale one third of people never attends religious services.
Table 23: Religious services attendance.
Country How often do you attend religious services. Japan Turkey
More than once a week 1.70% 14.00% Once a week 1.60% 20.20%
Once a month 6.50% 1.80% Only on special holy
days/Christmas/Easter days 42.10% 23.80% Once a year 22.30% 3.30%
Less often 15.00% 4.00% Never practically never 10.80% 32.90%
Total 1090 (100%)
1320 (100%)
Recep Tekeli Policy Research Institute 35
35 Determinants of Tax Morale in Japan and in Turkey
In order to compare Turkey with Japan we used another religiosity variable measured as
religious person. The question in the survey is:
Independently of whether you go to religious service or not would you say
you are 1= Religious 2= not a religious person 3= a convinced atheist.
Table 24 shows the frequency of response for “religious person” with respect to the total
respondents. In Turkey quite higher percentage of participants (82.60%) responded for the “a
religious person” while in Japan only 24.20% responded for the same category. On the other
hand, in Japan 13.70% of participants responded for the “a convinced atheist” which is quite
higher than in Turkey (0.50%).
Table 24: Religiosity
Country
Religious Person Japan Turkey A religious person 24.20% 82.60%
Not a religious person 62.10% 16.90% A convinced atheist 13.70% 0.50%
Total 958 (100%) 1320 (100%)
36 Determinants of Tax Morale in Japan and in Turkey
2.2.2 Descriptive analysis
First we will examine, compare and contrast the responses that were given by the World Value Survey participants in Japan and Turkey. Then in
the next section, we use the World Values Survey data to conduct a multivariate analysis of tax morale in the two countries.
In Table 25, in Japan opposition to tax cheating decreases when the happiness deteriorates. However, it does not significantly differ from one
group to another. In Turkey the result is not following a pattern in order to make a conclusion. The strongest opposition to tax cheating occurs in
the “very happy” group. The least opposition is for the “quite happy” group in Turkey.
Table 25: Tax Morale and Happiness
Japan Turkey
Feeling of happiness Feeling of happiness Justifiable: cheating on taxes
Total (%)
Very happy
Quite happy
Not very happy
Not at all happy
Total (%)
Very happy
Quite happy
Not very happy
Not at all happy
Never justifiable 83.20% 84.90% 82.50% 82.30% 81.80% 80.40% 86.00% 75.90% 82.50% 79.90%2 6.00% 5.60% 6.30% 5.20% 9.10% 12.50% 9.30% 15.30% 13.60% 7.30% 3 4.50% 3.90% 4.70% 5.20% - 4.00% 2.10% 5.20% 2.70% 9.40% 4 2.40% 3.30% 1.90% 3.10% - 1.10% 1.00% 1.30% 0.70% - 5 2.10% 1.00% 2.50% 3.10% - 1.10% 0.50% 1.50% - 3.40% 6 0.10% - 0.20% - - 0.20% 0.20% 0.10% - - 7 0.30% 0.30% 0.30% - - 0.20% 0.20% 0.30% - - 8 0.40% 0.30% 0.50% - - 0.10% - 0.20% 0.60% - 9 0.10% - - - 9.10% 0.10% 0.20% - - -
Always justifiable 1.00% 0.70% 1.10% 1.00% - 0.30% 0.50% 0.30% - - Total 1046 304 635 96 11 1340 501 655 126 58 Mean 1.5 1.4 1.5 1.5 1.8 1.3 1.3 1.4 1.2 1.4
Recep Tekeli Policy Research Institute 36
37 Determinants of Tax Morale in Japan and in Turkey
In Table 26, married people are strongly opposed to tax cheating than single/never married people. The least opposition to tax cheating occurs in
the “single/never married” category in Japan. The strongest opposition is for the “widowed” and “separated” groups in Japan while for the
“living together as married” and “separated” groups in Turkey (however it is not meaningful to compare these two groups with the others
because number of respondents in these groups is very small). The “single/never married” group is significantly less opposed to tax cheating
than the other groups. However, since the sample means for all groups significantly less than 2.0 on a scale of 1.0 to 10.0, we can conclude that
all groups are strongly opposed to tax cheating.
Table 26: Tax Morale and Marital Status
Japan Turkey
Marital status Marital status
Justifiable: cheating on taxes Total Married
Living together
as married Divorced Separated Widowed
Single/Never married Total Married
Living together
as married Divorced Separated Widowed
Single/Never married
Never justifiable 83.00% 84.10% 78.90% 86.70% 83.30% 88.00% 76.80% 80.50% 81.80% 100.00% 83.20% 100.00% 78.40% 77.40%
26.10% 5.90% 5.30% 3.30% 16.70% 2.00% 8.10% 12.50% 12.90% - 4.50% - 11.80% 12.20%
34.40% 4.60% - 3.30% - 2.00% 4.90% 4.00% 3.00% - 6.00% - 6.60% 5.90%
42.60% 2.20% - 3.30% - 6.00% 3.80% 1.10% 0.70% - - - 3.20% 1.80%
52.20% 1.50% 15.80% - - 2.00% 3.80% 1.10% 0.90% - - - - 1.60%
60.10% - - - - - 0.50% 0.20% 0.10% - - - - 0.30%
70.30% 0.10% - - - - 1.10% 0.20% 0.20% - - - - 0.20%
80.40% 0.50% - - - - - 0.10% 0.10% - - - - 0.30%
90.10% - - - - - 0.50% 0.10% 0.10% - - - - -
Always justifiable 0.90% 1.00% - 3.30% - - 0.50% 0.30% 0.20% - 6.40% - - 0.40%
Total1069 779 19 30 6 50 185 1341 891 1 19 2 31 397
Mean1.5 1.4 1.7 1.5 1.2 1.3 1.6 1.3 1.3 1 1.7 1 1.3 1.4
Recep Tekeli Policy Research Institute 37
38 Determinants of Tax Morale in Japan and in Turkey
In Table 27 people who are defined to be extremist (if it can be) at both side of the spectrum are strongly opposed to tax cheating than those who put themselves in the
center left. This is even more remarkable in Turkey than in Japan. Those who put themselves in 2-3 and 4 on political scales they are least opposed to tax cheating
than other political categories in Turkey. The proportion of participants defining themselves extreme left and right are higher in Turkey than in Japan. Taking out the
middle two groups in Japan, the leftist and the rightist are almost equally distributed. In Turkey, 280 participants responded for leftists while 497 participants defined
themselves the rightists which is almost doubled the left in politics. In Japan, the opposition to tax cheating first decreases when on a political spectrum one moves
from the left to the centre and then increase from the centre to the right on a political scale. The least opposition to tax cheating occurs in the centre left categories in
Japan. Also in Turkey the least opposition to tax cheating occurs in the left (non extremist) political scale. The strongest opposition is for the left and the right groups
in Turkey and also at the centre of political scale there is a high opposition to tax cheating. However, since the sample means for all 10 groups are less than 2.0 on a
scale of 1.0 to 10.0, we can conclude that all groups are strongly opposed to tax cheating (but with some cautions on being leftist as its mean score is 1.8).
Table 27: Tax Morale and Ideology
Japan Turkey
Self positioning in political scale Self positioning in political scale Justifiable: cheating on taxes Total(%) Left 2 3 4 5 6 7 8 9 Right Total(%) Left 2 3 4 5 6 7 8 9 Right
Never justifiable 82.90 90.00 84.20 85.70 75.90 77.60 86.80 81.20 89.80 77.80 93.80 79.40 89.10 62.30 60.50 70.00 80.40 86.10 68.90 79.30 76.30 87.90
2 6.90 - 10.50 5.20 4.80 9.90 5.50 8.80 5.70 11.10 - 13.10 3.10 24.70 23.80 11.90 12.50 9.80 25.00 14.40 20.10 5.70
3 4.00 - - 5.20 9.60 3.90 3.20 3.80 2.30 5.60 - 4.00 2.00 7.80 6.40 6.80 3.30 3.60 1.40 5.70 1.30 4.20
4 2.50 - - 1.30 3.60 4.30 2.30 1.20 1.10 - - 1.30 1.80 - 5.90 5.00 0.60 - 2.40 - 2.30 -
5 1.90 10.00 - 1.30 3.60 2.60 0.50 2.50 1.10 - - 1.20 4.00 - - 5.00 2.80 - 1.30 0.60 - 0.60
6 0.10 - - - - - - - - - 6.20 0.10 - 2.50 - - - - - - - -
7 0.20 - - - 1.20 - - 1.20 - - - 0.20 - - 2.50 - - - - - - 0.50
8 0.50 - - - - 0.90 - 1.20 - 5.60 - 0.20 - - 0.90 - - - - - - 0.70
9 0.10 - - 1.30 - - - - - - - 0.10 - - - - - - 1.00 - - -
Always justifiable 0.90 - 5.30 - 1.20 0.90 1.80 - - - - 0.40 - 2.70 - 1.30 0.60 0.50 - - - 0.50
Total 852 20 19 77 83 232 219 80 88 18 16 1140 90 45 78 67 148 215 88 154 75 180
Mean 1.8 1.4 1.6 1.4 1.7 1.5 1.4 1.5 1.2 1.6 1.3 1.6 1.3 1.8 1.8 1.7 1.4 1.2 1.5 1.3 1.3 1.3
Recep Tekeli Policy Research Institute 38
39 Determinants of Tax Morale in Japan and in Turkey
In Table 28, people who are not confident with the justice system are not strongly opposed to tax cheating than those who put more trust in their
justice system. The least opposition to tax cheating occurs in the “not very much” category in Japan. The strongest opposition is for the “a great
deal” group in Japan and in Turkey. It can be seen that tax morale deteriorates with the confidence level in Justice in Japan if we consider first
three groups. The “not very much” group is significantly less opposed to tax cheating than the other groups in Turkey and there is quite
remarkable difference in opposition to tax cheating between groups in confidence for justice system. However, since the sample means for all 4
groups are significantly less than 2.0 on a scale of 1.0 to 10.0, we can conclude that all groups are strongly opposed to tax cheating.
Table 28: Tax Morale and Confidence in Justice System
Japan Turkey
Confidence: Justice System Confidence: Justice System Justifiable: cheating on taxes Total
A great deal
Quite a lot
Not very much
None at all Total
A great deal
Quite a lot
Not very much
None at all
Never justifiable 82.90% 85.90% 83.10% 79.50% 80.80% 80.40% 88.90% 76.20% 70.40% 81.60%
2 5.80% 3.50% 6.20% 5.80% 11.50% 12.80% 6.80% 16.70% 17.80% 11.30%3 4.50% 3.50% 4.80% 5.10% - 4.00% 2.70% 4.30% 6.80% 2.90% 4 2.60% 1.80% 2.40% 3.20% 7.70% 0.90% 0.80% 1.10% 0.40% 1.60% 5 2.30% 1.80% 2.10% 3.80% - 1.00% - 0.80% 3.30% 1.70% 6 0.10% - 0.20% - - 0.20% - 0.20% 0.50% - 7 0.30% 0.60% 0.20% 0.60% - 0.20% 0.40% - 0.40% - 8 0.40% - 0.30% 1.30% - 0.10% - 0.40% - - 9 0.10% 0.60% - - - 0.10% - 0.20% - -
Always justifiable 1.00% 2.40% 0.80% 0.60% - 0.40% 0.40% 0.20% 0.40% 0.90%
Total 1013 170 661 156 26 1308 481 509 212 105 Mean 1.5 1.5 1.4 1.6 1.3 1.4 1.2 1.4 1.5 1.4
Recep Tekeli Policy Research Institute 39
40 Determinants of Tax Morale in Japan and in Turkey
In Table 29, people who are confident with their parliament are strongly opposed to tax cheating than those who do not put more trust in their
parliament. The least opposition to tax cheating occurs in the “none at all” category in Japan as same as trust in government. The strongest
opposition is for the “a great deal” and “none at all” groups in Turkey while for the “quite a lot” group in Japan. It can be seen that tax morale
deteriorates with the confidence level in parliament in Turkey but for the last group this result is not verified. The “not very much group” and the
“quite a lot” group is significantly less opposed to tax cheating than the other groups in Turkey. The results from the table indicate that there is
not any pattern to make a statement about the trust in parliament and tax morale level. However, since the sample means for all 4 groups are
significantly less than 2.0 on a scale of 1.0 to 10.0 (except for the first group in Japan which is 2.2), we can conclude that all groups are strongly
opposed to tax cheating.
Table 29: Tax Morale and Confidence in Parliament Japan Turkey
Confidence: Parliament Confidence: Parliament
Total A great
deal Quite a
lot Not very
much None at
all Total A great
deal Quite a
lot Not very
much None at
all Never
justifiable 82.80% 81.80% 89.10% 82.90% 76.20% 80.00% 87.10% 76.00% 76.00% 86.30%2 6.10% - 5.00% 6.00% 7.90% 13.00% 8.90% 15.10% 16.40% 8.10% 3 4.40% - 1.80% 4.90% 6.10% 4.10% 3.60% 5.60% 3.70% 1.60% 4 2.50% - 1.80% 2.60% 3.30% 1.00% 0.40% 1.50% 0.30% 1.40% 5 2.20% 9.10% 0.50% 2.60% 2.80% 1.00% - 0.50% 2.80% 1.00% 6 0.10% - - - 0.50% 0.20% - - 0.40% 0.50% 7 0.30% - 0.50% 0.20% 0.50% 0.20% - 0.40% 0.30% - 8 0.40% - - 0.40% 0.90% 0.10% - 0.20% 0.20% - 9 0.10% - - - 0.50% 0.10% - 0.20% - -
Always justifiable 1.00% 9.10% 1.40% 0.50% 1.40% 0.30% - 0.40% - 1.00%
Total 995 11 221 549 214 1274 288 494 306 185 Mean 1.6 2.2 1.3 1.4 1.7 1.4 1.2 1.4 1.4 1.3
Recep Tekeli Policy Research Institute 40
41 Determinants of Tax Morale in Japan and in Turkey
In Table 30, people who are not confident with their government are not strongly opposed to tax cheating than those who put more trust in their
government. The least opposition to tax cheating occurs in the “none at all” category in Japan. The strongest opposition is for the “a great deal”
and “quite a lot” groups in Japan while for the “a great deal” group in Turkey. It can be seen that tax morale deteriorates with the confidence
level in Government in Japan. Also the same conclusion could be said for Turkey but the last group cannot verify the conclusion. The “not very
much” group is significantly less opposed to tax cheating than the other groups in Turkey. However, since the sample means for all 4 groups are
significantly less than 2.0 on a scale of 1.0 to 10.0 (except for the first group in Japan which is 1.8), we can conclude that all groups are strongly
opposed to tax cheating.
Table30: Tax Morale and Confidence in Government
Japan Turkey
Confidence: The Government Confidence: The Government
Total A great
deal Quite a
lot Not very
much None at
all Total A great
deal Quite a
lot Not very
much None at
all Never
justifiable 82.90% 87.50% 87.30% 82.30% 77.00% 80.50% 90.00% 78.90% 70.10% 81.90%2 6.00% - 5.70% 5.60% 8.40% 12.60% 7.60% 13.20% 19.40% 10.40%3 4.50% - 2.70% 5.40% 5.20% 4.10% 1.50% 5.50% 5.40% 3.40% 4 2.60% - 2.00% 2.60% 3.70% 0.90% - 1.50% 1.20% 0.90% 5 2.20% 6.20% 1.00% 2.60% 2.60% 1.00% 0.30% - 2.90% 2.00% 6 0.10% - - 0.20% - 0.20% - - 0.40% 0.50% 7 0.30% - 0.30% 0.20% 0.50% 0.20% 0.30% 0.20% 0.30% - 8 0.40% - - 0.40% 1.00% 0.10% - 0.30% 0.20% - 9 0.10% - - - 0.50% 0.10% - 0.20% - -
Always justifiable 1.00% 6.20% 1.00% 0.80% 1.00% 0.30% 0.20% 0.30% - 1.00%
Total 1010 16 299 504 191 1299 358 469 284 188 Mean 1.5 1.8 1.3 1.5 1.6 1.4 1.2 1.4 1.5 1.4
Recep Tekeli Policy Research Institute 41
42 Determinants of Tax Morale in Japan and in Turkey
In Table 31, in Japan people who are confident with democraticness in their country strongly opposed to tax cheating than those who believe own
country is not democratic. In Turkey the strongest opposition to tax cheating occurs in the “not at all democratic” category while in Japan the
strongest opposition is for the “completely democratic” group. It can be seen that tax morale does not deteriorate with the democraticness in
Japan. However we cannot make the same statement for Turkey. The 9th group is significantly less opposed to tax cheating than the other groups.
However, since the sample means for all 10 groups are significantly less than 2.0 on a scale of 1.0 to 10.0, we can conclude that all groups are
strongly opposed to tax cheating.
Table 31: Tax Morale and Democracy
Japan Turkey
Democraticness in own country Democraticness in own country
TotalNot at all
democratic 2 3 4 5 6 7 8 9 Completely democratic Total
Not at all democratic 2 3 4 5 6 7 8 9
Completely democratic
Never 83.00 85.70 83.30 80.00 80.80 83.50 78.50 82.50 85.10 88.00 90.30 80.20 91.00 74.50 77.30 79.20 78.70 80.30 82.80 82.00 65.10 85.00
2 6.30 - 16.70 6.70 - 7.60 7.50 6.10 7.00 4.30 - 12.50 3.00 17.90 11.70 13.70 14.20 11.60 11.90 11.00 22.00 11.50
3 4.50 - - 3.30 7.70 - 5.30 6.10 3.70 3.30 6.50 4.10 0.90 7.60 7.80 1.80 3.70 3.60 3.30 4.20 10.40 1.50
4 2.60 - - 3.30 7.70 3.80 3.10 2.40 2.10 2.20 - 1.10 1.90 - 1.30 2.10 0.80 1.50 0.40 0.40 2.40 0.90
5 2.00 14.30 - - - 3.80 3.10 1.20 1.20 2.20 3.20 1.10 0.90 - 1.10 3.20 1.40 1.60 0.50 1.50 - -
6 0.10 - - - - - 0.40 - - - - 0.20 - - - - - 1.00 - - - -
7 0.30 - - 3.30 - 1.30 - 0.40 - - - 0.20 - - - - - 0.40 0.60 0.50 - -
8 0.20 - - - - - 0.40 0.40 - - - 0.10 - - 0.80 - - - - - - 1.10
9 0.10 - - - 3.80 - - - - - - 0.10 - - - - - - 0.40 - - -
always - - 3.30 - - 1.80 0.80 0.80 - - 0.40 2.30 - - - 1.30 - - 0.50 - -
Total 987 7 6 30 26 79 228 246 242 92 31 1274 89 53 88 106 142 211 198 176 99 113
Mean 1.6 1.6 1.2 1.7 1.7 1.4 1.6 1.4 1.3 1.3 1.3 1.4 1.3 1.3 1.4 1.4 1.4 1.4 1.3 1.3 1.5 1.2
Recep Tekeli Policy Research Institute 42
43 Determinants of Tax Morale in Japan and in Turkey
In Table 32, in Japan people who occasionally attend religious services are strongly opposed to tax cheating than those who attend religious services more than once a
week. It can be seen from the table that Japanese people are going to attend religious services on special holy days and on a specific occasion (once a year). This may
be because of the Buddhism religion as Buddhists are not required to attend regular prayers. The least opposition to tax cheating occurs in the “never practically
never” category in Turkey. Less than half of the participants responded that they never or practically never attend the religious services in Turkey. Taking the first
three groups together less than half of the respondents are attending religious services regularly in Turkey. Less than one fourth of people attend religious services
only on special holy days. The strongest opposition is for the “only on special holy days” groups in Japan while for the “once a year” group in Turkey. “More than
once a week” and the “never practically never” groups are significantly less opposed to tax cheating than the other groups in Japan. However, since the sample means
for all 7 groups significantly less than 2.0 on a scale of 1.0 to 10.0 (except for last group in Japan which is 1.9 and significant), we can conclude that all groups are
strongly opposed to tax cheating.
Table 32: Tax Morale and Religiosity
Japan Turkey
How often do you attend religious services How often do you attend religious services
Justifiable: cheating on taxes Total
More than
once a week
Once a week
Once a month
Only on special holy days/Christmas/Easter
days Once a
year Less often
Never practically
never Total
More than
once a week
Once a week
Once a month
Only on special holy days/Christmas/Easter
days Once a
year Less often
Never practically
never Never
justifiable 83.00 73.70 82.40 81.20 85.90 81.90 83.50 75.70 81.10 84.80 85.90 77.90 83.00 92.10 78.20 74.60 2
6.10 5.30 5.90 5.80 6.20 6.70 5.70 5.20 12.30 10.90 8.60 16.00 10.30 7.90 7.40 17.40 3
4.50 15.80 5.90 4.30 3.50 5.00 4.40 5.20 3.80 2.80 4.30 6.10 3.90 - 8.00 3.50 4
2.40 - - 2.90 1.30 2.90 3.80 4.30 1.00 0.60 - - 0.90 - 4.80 1.70 5
2.20 5.30 5.90 2.90 2.00 1.70 1.30 4.30 1.00 0.50 - - 1.20 - - 1.90 6 0.10 - - - - - - 0.90 0.20 - - - 0.30 - - 0.30 7
0.30 - - - 0.40 - 0.60 - 0.10 - 0.30 - - - 1.60 - 8
0.40 - - 2.90 0.20 - - 0.90 0.10 - - - 0.40 - - 0.20 9 0.10 - - - - - - 0.90 0.10 - 0.30 - - - - -
Always justifiable 0.90 - - - 0.40 1.70 0.60 2.60 0.40 0.40 0.60 - - - - 0.50
Total 1070 19 17 69 454 238 158 115 1315 184 266 23 314 43 53 432
Mean 1.5 1.6 1.4 1.6 1.3 1.5 1.4 1.9 1.4 1.2 1.3 1.3 1.3 1.1 1.5 1.4
Recep Tekeli Policy Research Institute 43
44 Determinants of Tax Morale in Japan and in Turkey
In Table 33, more aware people are strongly opposed to tax cheating than less aware people in Japan. The least opposition to tax cheating occurs
in the “used last week” category in Turkey. The strongest opposition is for the “used last week group” in Japan while for the “not used last week”
group in Turkey. The “not used last week” group is significantly less opposed to tax cheating than the “used last week” group in Japan while in
Turkey counter argument holds. While majority of people in Japan reads daily newspaper and gets information from daily newspaper, in Turkey
less than half of the participants do not obtain information from the daily news paper. This also reflects the national characters in relation to
education: Japanese people like reading very much. Since the sample means for all groups significantly less than 2.0 on a scale of 1.0 to 10.0, we
can conclude that all groups are strongly opposed to tax cheating.
Table 33: Tax Morale and Awareness
Japan Turkey
Information source: Daily newspaper
Information source: Daily newspaper Justifiable:
cheating on taxes Total
Used last week
Not used last week Total
Used last week
Not used last week
Never justifiable 83.00% 83.60% 77.40% 80.50% 76.90% 85.50% 2 6.10% 5.90% 8.50% 12.60% 14.30% 10.10% 3 4.30% 4.40% 3.80% 3.90% 4.40% 3.20% 4 2.50% 2.30% 4.70% 1.10% 1.60% 0.30% 5 2.30% 2.20% 2.80% 1.10% 1.70% 0.20% 6 0.10% 0.10% - 0.20% 0.30% - 7 0.30% 0.30% - 0.20% 0.20% 0.20% 8 0.40% 0.40% - 0.10% 0.10% 0.20% 9 0.10% - 0.90% 0.10% 0.10% -
Always justifiable 0.90% 0.80% 1.90% 0.30% 0.40% 0.30%
Total 1062 956 106 1339 787 552 Mean 1.5 1.4 1.7 1.3 1.4 1.2
Recep Tekeli Policy Research Institute 44
45 Determinants of Tax Morale in Japan and in Turkey
In Table 34, pride and opposition to tax cheating are cross tabulated. People proud of own country are strongly opposed to tax cheating than
those who are not proud of own country. The least opposition to tax cheating occurs in the “not at all proud” category for Japan and Turkey (in
Turkey there is a significant difference for this category). The strongest opposition is for the “very proud” groups in Japan while for the “quite
proud” group in Turkey. We should point that majority of participants have shown the “very proud” response in Turkey in contrast to Japan in
which majority of participants responded the “quite proud” or the “not very proud” answers. However, since the sample mean for all 4 groups are
significantly less than 2.0 on a scale of 1.0 to 10.0, we can conclude that all groups are strongly opposed to tax cheating in Japan but in Turkey
the conclusion slightly differs. In Turkey, people who show no pride in their own country they are most in favor of tax cheating than other
categories.
Table 34: Tax Morale and Pride
Japan Turkey
How proud of nationality How proud of nationality Justifiable: cheating on taxes Total
Very proud
Quite proud
Not very proud
Not at all proud Total
Very proud
Quite proud
Not very proud
Not at all proud
Never justifiable 82.90% 87.00% 82.20% 82.10% 74.40% 81.80% 81.80% 83.30% 82.50% 39.90%
2 6.30% 4.50% 5.60% 8.40% 5.10% 11.90% 11.60% 13.40% 10.50% 16.70%
3 4.50% 3.10% 5.60% 3.90% 7.70% 3.50% 3.90% 1.50% - 10.40%
4 2.20% 1.30% 2.50% 2.00% 5.10% 1.00% 1.00% 0.40% 2.30% 10.40%
5 2.30% 2.20% 2.30% 2.20% 2.60% 1.00% 0.90% - 2.30% 22.50%
6 0.10% - 0.30% - - 0.20% 0.10% - 2.40% -
7 0.30% - 0.50% 0.30% - 0.20% 0.30% - - -
8 0.30% - 0.50% 0.30% - 0.20% 0.10% 0.40% - -
9 0.10% - - - 2.60% 0.10% - 0.40% - -
Always justifiable 1.00% 1.80% 0.50% 0.80% 2.60% 0.30% 0.30% 0.60% - -
Total 1014 223 394 358 39 1263 1027 191 37 8
Mean 1.5 1.4 1.5 1.4 1.9 1.4 1.3 1.3 1.4 2.6
Recep Tekeli Policy Research Institute 45
46 Determinants of Tax Morale in Japan and in Turkey
Table 35 compares the scores for two sample populations by gender. By far most frequent response for the two groups is the “never justifiable”
option. All mean scores are around 1.4. In table, females are more opposed to tax cheating than men. This is especially strong for Japanese case.
Table 35: Tax Morale and Gender Japan Turkey
Sex Sex Justifiable: cheating on taxes Total Male Female Total Male Female
Never justifiable 82.90% 78.80% 86.10% 80.50% 80.40% 80.50% 2 6.00% 7.40% 5.00% 12.50% 13.00% 12.10% 3 4.50% 4.80% 4.20% 4.00% 3.60% 4.30% 4 2.60% 2.90% 2.30% 1.10% 1.00% 1.20% 5 2.20% 2.70% 1.80% 1.10% 1.40% 0.80% 6 0.10% 0.20% - 0.20% 0.10% 0.20% 7 0.30% 0.40% 0.20% 0.20% 0.10% 0.30% 8 0.40% 0.60% 0.20% 0.10% 0.20% 0.10% 9 0.10% 0.20% - 0.10% - 0.10%
Always justifiable 0.90% 1.90% 0.20% 0.30% 0.20% 0.50% Total 1075 476 599 1341 674 666 Mean 1.5 1.6 1.3 1.3 1.3 1.4
Recep Tekeli Policy Research Institute 46
47 Determinants of Tax Morale in Japan and in Turkey
Table 36 compares the scores for both sample populations by age. By far the most frequent response for all age groups is the “never justifiable”
option. All mean scores are around 1.5. In the table, both countries show the similar tendency. Never justifiable response increases with age
groups. As can be seen from the table the mean scores decline with age which means that as one gets older, opposition to tax cheating increases
in both countries.
Table 36: Tax Morale and Age
Japan Turkey
Age Age Justifiable: cheating on taxes Total
15-29 years
30-49 years
50 and more years Total
15-29 years
30-49 years
50 and more years
Never justifiable 82.90% 77.20% 80.60% 86.50% 80.50% 79.20% 79.50% 85.60%
2 6.00% 6.70% 5.90% 5.90% 12.50% 12.00% 13.50% 11.40% 3 4.50% 4.40% 7.30% 2.50% 4.00% 4.70% 3.80% 2.70% 4 2.60% 5.00% 3.00% 1.50% 1.10% 1.10% 1.50% - 5 2.20% 3.90% 1.60% 2.10% 1.10% 1.70% 0.90% - 6 0.10% 0.60% - - 0.20% 0.20% - 0.40% 7 0.30% 1.10% - 0.20% 0.20% - 0.50% - 8 0.40% - 0.30% 0.60% 0.10% 0.20% 0.10% - 9 0.10% 0.60% - - 0.10% 0.20% - -
Always justifiable 0.90% 0.60% 1.30% 0.80% 0.30% 0.70% 0.10% -
Total 1075 180 371 524 1341 537 565 240 Mean 1.5 1.6 1.5 1.4 1.3 1.4 1.3 1.2
Recep Tekeli Policy Research Institute 47
48 Determinants of Tax Morale in Japan and in Turkey
Table 37 shows that full time employed people are strongly opposed to tax cheating than unemployed people. The strongest opposition is for the
“Housewife” and the “part time employed” groups in Japan while for the “Retired” and the “full time employed” groups in Turkey. The
“unemployed” group is significantly less opposed to tax cheating than the other groups in Turkey. Since the sample means for unemployed group
is 1.9 on a scale of 1.0 to 10.0, and it is higher than the other categories (in Japan Self employed people show the least opposition to tax cheating
than the other categories) we can conclude that Tax Morale deteriorates with whether one is employed or not.
Table 37: Tax Morale and Employment Status
Japan Turkey
Employment status Employment status Justifiable: cheating on taxes Total
Full time
Part time
Self emp Retired Housewife Student Unemp Other Total
Full time
Part time
Self emp Retired Housewife Student Unemp Other
Never 82.70% 82.00% 85.60% 68.10% 85.30% 91.80% 76.70% 70.60% 82.60% 80.50% 82.60% 72.20% 81.10% 87.00% 81.20% 78.10% 64.20% 80.40%
2 6.10% 6.70% 4.40% 11.80% 3.90% 2.70% 10.00% 11.80% 4.30% 12.50% 11.00% 15.70% 15.80% 8.90% 12.10% 10.50% 18.50% 16.00%
3 4.40% 5.00% 3.80% 7.60% 2.00% 3.80% - 5.90% 4.30% 4.00% 2.90% - 1.00% 3.40% 4.00% 7.60% 13.80% -
4 2.60% 3.20% 3.80% 4.20% 1.00% 0.50% 3.30% - - 1.10% 0.90% 3.10% 1.50% - 1.00% 1.70% - 3.60%
5 2.30% 1.50% 1.90% 4.20% 4.90% 1.10% 3.30% 5.90% 4.30% 1.10% 2.10% 5.80% 0.60% - 0.60% 1.00% 1.00% -
6 0.10% - - - - - 3.30% - - 0.20% - 3.10% - 0.70% - - - -
7 0.30% 0.50% - 0.80% - - - - - 0.20% 0.30% - - - 0.40% - - -
8 0.40% - 0.60% 1.70% - - - 5.90% - 0.10% - - - - 0.20% 1.10% - -
9 0.10% - - - - - 3.30% - - 0.10% - - - - 0.20% - - -
Always 1.00% 1.00% - 1.70% 2.90% - - - 4.30% 0.30% 0.30% - - - 0.40% - 2.50% -
Total 1035 401 160 119 102 183 30 17 23 1341 299 36 195 123 462 114 82 31
Mean 1.5 1.4 1.4 1.9 1.6 1.2 1.8 1.9 1.7 1.3 1.3 1.6 1.2 1.2 1.3 1.4 1.7 1.3
Recep Tekeli Policy Research Institute 48
49 Determinants of Tax Morale in Japan and in Turkey
In Table 38, higher income scaled people are strongly opposed to tax cheating than other categories. The least opposition to tax cheating occurs
in the fourth income scale in Turkey and eight income scale in Japan. The strongest opposition is for the 4th and the 9th group in Japan while for
the 7th and 9th group in Turkey. There is not any pattern to conclude that tax morale level increases or decreases with income level. In Turkish
case, opposition to tax cheating first increases with income level then decreases and then increases with income level and it goes on this way. In
Japanese case there seems to be a pattern in which opposition to tax cheating increases with income scale and then decreases, if we consider only
the first six income scale. However, since the sample mean for all income groups significantly less than 2.0 on a scale of 1.0 to 10.0, we can
conclude that all groups are strongly opposed to tax cheating.
Table 38: Tax Morale and Income
Japan Turkey
Scale of incomes Scale of incomes Justifiable: cheating on taxes Total 1 2 3 4 5 6 7 8 9 10 Total 1 2 3 4 5 6 7 8 9 10
Never justifiable 82.80 80.70 80.60 87.30 88.60 82.40 83.50 77.90 75.90 90.30 76.10 80.40 78.00 84.90 86.50 71.80 82.90 74.80 93.40 77.70 91.50 75.40
2 6.10 8.60 6.00 2.20 7.00 4.40 6.60 7.80 10.30 1.40 8.50 12.60 14.60 10.70 7.80 17.40 9.20 15.90 - 15.30 4.50 13.10
3 4.70 3.60 3.70 3.70 3.50 5.50 4.40 9.10 5.20 4.20 7.00 4.00 6.10 2.20 2.50 5.70 7.90 2.10 6.60 2.60 - 7.10
4 2.50 2.10 3.00 4.50 - 4.40 - 1.30 5.20 1.40 4.20 1.10 0.90 - 1.10 2.20 - 4.40 - 1.50 - -
5 2.10 2.90 4.50 2.20 - - 3.30 - 3.40 1.40 2.80 1.10 - 0.80 1.70 1.50 - 2.00 - 2.90 4.00 -
6 0.10 - - - - - - - - - 1.40 0.20 - - - 0.50 - - - - - 4.40
7 0.20 0.70 - - - - 1.10 - - - - 0.20 0.30 0.30 0.40 - - - - - - -
8 0.40 0.70 1.50 - - - - 1.30 - - - 0.10 - - - 1.00 - - - - - -
9 0.10 - - - - - - 1.30 - - - 0.10 - 0.20 - - - - - - - -
Always justifiable 0.90 0.70 0.70 - 0.90 3.30 1.10 1.30 - 1.40 - 0.30 - 0.80 - - - 0.90 - - - -
Total 982 140 134 134 114 91 91 77 58 72 71 1316 298 350 186 195 42 99 17 77 25 26
Mean 1.6 1.5 1.6 1.3 1.2 1.6 1.5 1.6 1.5 1.3 1.5 1.4 1.3 1.3 1.3 1.5 1.3 1.5 1.1 1.4 1.2 1.5
1: Lower step, 2:second step, 3:Third step, 4:Fourth step, 5:Fifth step, 6:Sixth step, 7:Seventh step, 8:Eigth step, 9:Nineth step, 10:Tenth step
Recep Tekeli Policy Research Institute 49
Recep Tekeli Policy Research Institute 50
50 Determinants of Tax Morale in Japan and in Turkey
In Table 39, the mean score for tax morale is increasing with the education level which means that opposition to tax cheating deteriorates as the level of
education increases in Turkey. The differences become more significant as one drops off high school in Japan and university in Turkey. In Turkey
justification for tax cheating decreases with the education level attained while the opposition to tax cheating is not appeared to deteriorate with
education level in Japan. However the mean score for the highest level of education attained is around 1.4 and on a scale of 1 to 10. At the highest level
of education the mean score is 1.3 for Turkey and 1.5 for Japan which means that there is a significant opposition to tax cheating in both countries. In
Japan the fifth category of education (Complete secondary school) the mean score is higher than overall average score while in Turkey the mean score
for the eighth category of education (Some university without degree ) is higher than overall average score.
Table 39: Tax Morale and Education
Japan Turkey
Highest educational level attained Highest educational level attained Justifiable: cheating on taxes Total 1 2 3 4 5 6 7 8 9 Total 1 2 3 4 5 6 7 8 9
Never justifiable 82.80 - - 78.80 92.90 73.60 69.00 86.80 75.00 82.70 80.50 79.90 78.30 82.90 92.40 81.90 78.50 77.30 61.20 79.60
2 6.10 - - 7.50 - 10.70 21.40 3.70 15.00 5.30 12.50 18.80 19.80 10.00 4.40 12.40 11.30 13.40 16.40 13.60
3 4.60 - - 2.50 7.10 5.00 4.80 3.90 10.00 5.60 4.00 1.30 1.90 4.80 - 1.20 3.50 5.20 14.20 2.70
4 2.60 - - 1.20 - 5.00 - 2.80 - 2.30 1.10 - - 0.80 3.20 0.90 - 1.50 8.20 1.40
5 2.20 - - 3.80 - 5.00 4.80 1.20 - 2.30 1.10 - - 0.70 - 1.20 5.10 1.40 - 1.50
6 0.10 - - - - - - 0.20 - - 0.20 - - - - - - 0.30 - 0.80
7 0.30 - - 1.20 - - - 0.40 - - 0.20 - - 0.30 - - - 0.30 - -
8 0.40 - - 2.50 - - - 0.20 - 0.40 0.10 - - - - - 1.60 0.30 - -
9 0.10 - - - - - - - - 0.40 0.10 - - 0.20 - - - - - -
Always justifiable 1.00 - - 2.50 - 0.80 - 0.80 - 1.10 0.30 - - 0.20 - 2.40 - 0.20 - 0.60
Total 1052 0 0 80 14 121 42 509 20 266 1341 108 52 492 35 92 44 358 14 145
Mean 1.5 - - 1.8 1.1 1.6 1.5 1.4 1.4 1.5 1.3 1.2 1.2 1.3 1.1 1.4 1.5 1.4 1.7 1.4
Notes: 1: No formal education, 2: Inadequately completed elementary education, 3: Completed (compulsory) elementary education, 4: Incomplete secondary school: technical/vocational type/(Compulsory) elementary education and basic vocational qualification, 5: Complete secondary school: technical/vocational type/Secondary, intermediate vocational qualification, 6: Incomplete secondary: university-preparatory type/Secondary, intermediate general qualification, 7: Complete secondary: university-preparatory type/Full secondary, maturity level certificate, 8: Some university without degree/Higher education - lower-level tertiary certificate, 9: University with degree/Higher education - upper-level tertiary certificate.
Recep Tekeli Policy Research Institute 51
51 Determinants of Tax Morale in Japan and in Turkey
3. Empirical results
We observe for WVS data that tax morale is higher in Japan than in Turkey. However this
purely descriptive analysis gives information about the raw effects and not the partial effects.
The observed differences between Japan and Turkey might be explained in terms of
differences in socio-economic, political and socio-demographic factors. Thus multiple
regressions in the next section help us to separate the effects of several socio-economic and
socio-demographic factors from a possible culture difference.
We estimate separately the determinants of Tax Morale at the individual level for Japan and
Turkey. Then the determinants of Tax Morale at the combined data set will be estimated.
3.1. Empirical results on tax morale in Turkey
Table 40 presents the estimation results for Turkey, focusing only on ordered probit models.13.
In a next step we test Hypotheses 1–4. We use three trust variables, (i.e Trust in Legal System,
Trust in Government and Trust in Parliament), to check the robustness of the trust variables.
These variables allow us to analyze trust at the constitutional level (e.g., trust in the legal
system/justice), thereby focusing on how the relationship between the state and its citizens is
established; they also allow us to analyze trust more closely at the current politico-economic
level (e.g., trust in the parliament and government) (see Alm and Torgler 2006). Trust in
Legal System/Justice and Trust in Government are found highly significant. However Trust in
Parliament in some estimations is not significant while in some estimations it is significant
but with a negative coefficient implying that for those who does not have confidence in
parliament they have higher tax morale.
In all estimations both trust variables have a significantly positive effect on tax morale with
marginal effects varying between 2.5 and 4.7 (Trust in Government) and between 2.6 and 3.6
(Trust in Justice/Legal System) percentage points. Therefore, our study finds support for the
notion that trust matters for tax morale in Turkey. An increase in the trust in government
(trust in legal system) scale by one unit increases the share of subjects indicating the highest
tax morale by 2.5 to 4.7 (2.6 to 3.6) percentage points. As trust is highly and positively
correlated with tax morale, Hypothesis 1 cannot be rejected.
Specifications (1-9) show the relative importance of Pride in tax morale determination. In
five estimations we have significant coefficient on this variable. As can be seen from the 13After the first estimation we have not included variables “trust in the government” and “trust in the Parliament” in the same model because there is a strong correlation between these two variables (r=0.60) that makes it impossible to clearly separate the effects of the two variables if they were included in one model.
Recep Tekeli Policy Research Institute 52
52 Determinants of Tax Morale in Japan and in Turkey
tables, there is a positive correlation between Pride and tax morale. The coefficient is always
positive and the marginal effects indicate that an increase in Pride by one point raises the
share of persons indicating the highest tax morale value by 3 to 4 percentage points. Thus, the
results show that we cannot reject our main hypothesis.
The results in Tables also indicate that Hypothesis 3 is rejected. We observe a significant
correlation between a prodemocracy attitude and tax morale. As Table 40 shows, a one-point
increase in the prodemocracy index reduces the share of persons with the highest tax morale
approximately 2 percentage points, meaning that a higher degree of democracy leads to lower
tax morale for the reference group14. However, calculating the marginal effects at the lowest
tax morale level (i.e, 0 scale) we observe that an increase in the democracy index by one unit
reduces the probability of stating that tax evasion is always justified to less than one
percentage point.
Political attitude was tested by the variable ideology, the self-placement in a left-right
ideological scale from 1 (left) to 10 (right). It has positive sign but insignificant effect on tax
morale. We did not find any evidence to support the view that individual’s political
alignments affect his/her behavior toward taxes.
To get a broader view and to consider the fact that a considerable number of citizens have a
low living standard (as argued by Torgler 2004), we also include the variable happiness. In all
specifications this variable does not significantly affect tax morale. The happiness variable
has interestingly, negative marginal effects but it is insignificant to derive any conclusion
from it. Thus our hypothesis 4 cannot be proved statistically.
As argued by Torgler (2004:249) it was expected that “if the financial situation of a household
is bad, the tax payments might be seen as a hard restriction of their possibility set, which
might reduce tax honesty”. In line with this argument we included income variable to test
whether individuals in lower income classes are more likely to engage in criminal activities
due to their lower opportunity costs, or not. However, in all estimations income variable did
not have significant coefficient.
We include different employment status such as part time employed, self employed, retired,
and unemployed. In the tax compliance literature, it was argued that self-employed persons
14 I would like to thank Mr Osamu Tanaka Vice President of Policy Research Institute for pointing at the western democracies that when a country is advancing in a democracy historically there has been a resistance to the taxation because of democraticness. Therefore we may not be surprised to see the negative relation between tax morale and the level of democraticness.
Recep Tekeli Policy Research Institute 53
53 Determinants of Tax Morale in Japan and in Turkey
have higher compliance costs than employed (e.g., Lewis, 1982). Therefore, Torgler and
Schneider (2007:452) argued that “taxes are more visible for the self-employed, who have a
higher opportunity to evade or avoid them. Moreover, pensions of retired individuals are
incomes provided or at least heavily regulated by the state, so transparency is higher and the
control better”. However, we did not find significant coefficient on self-employed. On the
other hand we have significant coefficients on unemployed and housewife. An increase in
Housewife by one point reduces the share of persons indicating the highest tax morale value
by 8 percentage points.
Our results indicate that the proportion of unemployed who report the highest tax morale is
approximately 10 percentage points lower than the one of full-time employees. An increase in
unemployment by one point reduces the share of persons indicating the highest tax morale
value by almost 10 percentage points. Unemployed individuals might have a higher incentive
to act in the shadow economy, which might influence their attitude regarding tax evasion.
Tables present evidence regarding the variable religiosity, measured as frequency of Mosque
attendance. We first included the variable measured as frequency of Mosque attendance. The
result in the first specification indicates that an increase in the religious attendance increases
the tax morale around 2 percentage points15 . To make a comparison between Japan and
Turkey we included the variable religiosity, measured as Religious person on a three point
scale (i.e 1 for religious 3 for atheist). Robust across all 8 specifications is the negative
correlation between Tax Morale and Religiosity variables. As argued above we would expect
that religiosity might influence people’s habits, and might make individuals reluctant to
engage in tax evasion. Our findings are consistent with those of Alm and Torgler (2006) and
Torgler and Murphy (2004), who find that a higher religiosity is correlated with a higher tax
morale. Our results indicate that an increase in religiosity increases the tax morale between 8
and 10 percentage points.
McGee (1998) reports that the coefficient on Muslim is mostly not significant and also argues
that Muslims are not always obligated to pay all taxes. Torgler (2006:94) points that “If the
government engages in activities that are not legitimated, tax evasion might not be immoral
(for a list of possible immoral state activities, see Yusuf, 1971). It would not be immoral for a
Muslim not to pay indirect taxes, to avoid paying tariffs, to evade income taxes, or not to
comply with a law that causes prices to rise”. Our results on religiosity in Turkey do not seem
15 The results can be given on request by the author.
Recep Tekeli Policy Research Institute 54
54 Determinants of Tax Morale in Japan and in Turkey
to support McGee’s argument that “the Muslim view toward tax evasion seemingly falls
under category three that evasion is sometimes ethical” (2006:23). The result rather seems to
fit under McGee’s first category that “tax evasion is never ethical”.
The positive coefficients on education and negative coefficients on gender (woman) deserve
special attention. In the models we have run, women do not seem to have stronger tax morale
than men but this argument is not proved statistically.
Our results show that education does not have any significant effects.16 However, we found
that if the reading daily newspaper rate decreases tax morale increases by 8 to 10 percentage
points. This result indicates that unawareness has significantly positive effect on tax morale.
This result does not change even if we drop education variable from the estimation. It seems
that reading newspaper makes individual’s behavior somehow deteriorated.
We cannot say that individuals’ constraints by their social networks affect tax morale. In
tables our results indicate that married people having higher tax morale than singles is not
supported by the evidence from Turkey. Actually marital status did not have any effect on tax
morale in all estimations. Also, older individuals (third age group in the estimation) tend to
exhibit higher tax morale, but the coefficients on age group are not statistically significant in
all estimations.
Conclusions: The effects of trust on tax morale were analyzed on two different levels: (i) at
the constitutional level (trust in the legal system, or in the court/justice system) and (ii) at the
current politico-economic level (trust in the government and trust in parliament). We agree
with the statement that not only trust in the government might have an effect on tax morale,
but also trust in the court, or the legal system and hence the way the relationship between the
state and its citizens is established.
The probit estimations in all specifications for Turkey indicate that an increase in the level of
direct democracy by one unit decreases the probability of stating that tax evasion is never
justified by 2 percentage points. Our results indicate that tax morale of unemployed and
housewife is lower than full-time employed. We found that while unawareness has significant
effects on tax morale level in Turkey, education does not affect tax morale. In line with the
previous findings in the literature pride has positive effect on tax morale level in Turkey, too.
The results on religion, indicates that tax cheating is immoral for a Religious Muslim.
16 The result on education did not change even if we we used education as a continuous variable.
Recep Tekeli Policy Research Institute 55
55 Determinants of Tax Morale in Japan and in Turkey
Table 40: Determinants of tax morale in Turkey (2007)
Ordered Probit
Model (1) Model (2) Model (3) Variables
Coefficient Z statistic
Marginal Effect
Coeff. Z Marg Coeff. Z Marg
HAPPİNESS -0.009 -0.15 -0.020 -0.30 -0.024 -0.37
Living Together 3.214 0.02 3.078 0.02 3.140 0.02
Divorced -0.046 -0.11 -0.034 -0.08 0.042 0.1
Separated 3.962 0.03 3.803 0.04 3.884 0.03
Widowed -0.434 -1.33 -0.470 -1.45 -0.429 -1.34
MA
RIT
AL
S
TA
TU
S
Single/Never -0.036 -0.25 -0.042 -0.29 -0.037 -0.25
IDEOLOGY 0.002 0.09 0.0002 0.01 0.002 0.1
JUSTICE SYSTEM
0.105 * 1.65 0.026 0.096 1.5 0.023 0.054 0.86
THE GOVERNMENT
0.173 *** 2.52 0. 043 0.193 *** 2.79 0.047 0.103* 1.7 0.025
TR
US
T
PARLIAMENT -0.180*** -2.64 -0.044 -0.204 *** -2.95 -0.050
DEMOCRACY -0.081*** -3.77 -0.020 -0.085 *** -3.92 -0.021 -0.091 *** -4.25 -0.022
RELIGIOSITY 0.067 *** 2.56 0.016 -0.354 *** -2.88 -0.086 -0.343 *** -2.82 -0.084
PRIDE 0.069 0.78 0.145* 1.67 0.035 0.121 1.41 0.029
AWARENESS 0.416 *** 3.55 0.098 0.404*** 3.43 0.094 0.386 *** 3.31 0.090
GENDER (WOMAN)
0.027 0.17 -0.109 -0.74 -0.093 -0.64
30 - 49 -0.111 -0.88 -0.147 -1.14 -0.150 -1.18 AGE
Age ≥ 50 0.139 0.69 0.121 0.60 0.074 0.38
EDUCATION 0.009 0.36 0.008 0.29 0.011 0.42
PARTTIME -0.176 -0.64 -0.104 -0.37 -0.080 -0.29
SELF EMPLOYED
-0.1301 -0.81 -0.172 -1.07 -0.122 -0.77
RETIRED -0.107 -0.48 -0.110 -0.49 -0.067 -0.31
HOUSEWIFE -0.164 -0.88 -0.247 -1.29 -0.219 -1.16
STUDENT -0.045 -0.23 -0.141 -0.7 -0.112 -0.56
UNEMPLOYED -0.359 -1.58 -.094 -0.426 -1.82 -0.109 -0.405* -1.76 -0.105
EM
PL
OY
ME
NT
ST
AT
US
OTHER -0.032 -0.1 -0.082 -0.25 -0.083 -0.25
INCOME 0.008 0.37 0.012 0.52 0.005 0.21
Number of Observations 960 960 972
LM(prob>chi2) 0.0001 0.0000 0.0001
Pseudo R2 0.0503 0.0568 0.0489
Notes: Dependent variable: tax morale on a four-point scale. In the reference group: full-time employed, married, read newspaper, male, lower age group (18-29), Marg.: marginal effect. The marginal effect is calculated at the highest TAX MORALE score. Significance levels are: *0.05 < p < 0.10, **0.01 < p < 0.05, ***p < 0.01 . dy/dx is for discrete change of dummy variable from 0 to 1 and dy/dx for factor levels is the discrete change from the base level. In this table and following tables, variables Happiness, Trust Justice, Trust Parliament, Trust Government, Religious Service Attendance, Pride are rescaled in order to interpret the result straightforward. For example, in the survey, pride is measured on a 4 point scale i.e 1 for very proud 4 for not all. This was recoded into 4 for very proud and 1 for not all proud.
Recep Tekeli Policy Research Institute 56
56 Determinants of Tax Morale in Japan and in Turkey
Table 40: Determinants of tax morale in Turkey (2007) (continued)
Ordered Probit
Model (4) Model (5) Model (6) Variables
Coeff. Z Marg Coeff. Z Marg Coeff. Z Marg
HAPPINESS -0.013 -0.20
Living Together 3.246 0.02
Divorced -0.147 -0.35
Separated 3.888 0.03
Widowed -0.419 -1.3
MA
RIT
AL
ST
AT
US
Single/Never -0.057 -0.39
IDEOLOGY 0.017 0.89
JUSTICE SYSTEM 0.145*** 2.35 0.035 0.150 *** 2.67 0.036 0.062 1.09 0.015
THE GOVERNMENT
0.109** 2.03 0.026
TR
UST
PARLIAMENT -0.120 ** -1.98 -0.029 -0.089 -1.63 -0.021
DEMOCRACY -0.085 *** -3.94 -0.021 -0.072 *** -3.7 -0.017 -0.08 *** -4.16 -0.019
RELIGIOSITY -0.356 *** -2.91 -0.088 -0.402 *** -3.62 -0.097 -0.359 *** -3.24 -0.085
PRIDE 0.144 * 1.67 0.035 0.138 * 1.73 0.033 0.106 1.34 0.025
AWARENESS 0.421 *** 3.6 0.098 0.408 *** 3.94 0.094 0.359 *** 3.46 0.082
GENDER (WOMAN)
-0.079 -0.54
30 - 49 -0.173 -1.36 AGE
Age ≥ 50 0.098 0.49
EDUCATION 0.002 0.08
PARTTIME -0.059 -0.21 -0.101 -0.37 -0.116 -0.42
SELF EMPLOYED -0.153 -0.96 -0.164 -1.09 -0.153 -1.02
RETIRED -0.109 -0.49 -0.023 -0.14 -0.034 -0.2
HOUSEWIFE -0.273 -1.44 -0.343*** -2.68 -0.083 -0.324*** -2.54 -0.077
STUDENT -0.147 -0.73 -0.112 -0.66 -0.084 -0.49
UNEMPLOYED -0.403* -1.73 -0.104 -0.421** -2.1 -0.105 -0.423** -2.12 -0.105
EM
PL
OY
ME
NT
ST
AT
US
OTHER -0.076 -0.23 -0.066 -0.21 -0.079 -0.25
INCOME 0.009 0.4
Number of Observations 967 1123 1137
LM(prob>chi2) 0.0001 0.0000 0.0000
Pseudo R2 0.0509 0.0427 0.0435
Notes: Dependent variable: tax morale on a four-point scale. In the reference group: full-time employed, married, read newspaper, male, lower age group (18-29), Marg.: marginal effect. The marginal effect is calculated at the highest TAX MORALE score. Significance levels are: *0.05 < p < 0.10, **0.01 < p < 0.05, ***p < 0.01 . dy/dx is for discrete change of dummy variable from 0 to 1 and dy/dx for factor levels is the discrete change from the base level.
Recep Tekeli Policy Research Institute 57
57 Determinants of Tax Morale in Japan and in Turkey
Table 40: Determinants of tax morale in Turkey (2007) (continued)
Ordered Probit
Model (7) Model (8) Model (9) Variables
Coeff. Z Marg. Coeff. Z Marg Coeff. Z Marg
HAPPİNESS
Living Together
Divorced
Separated
Widowed
MA
RIT
AL
S
TA
TU
S
Single/Never
IDEOLOGY
JUSTICE SYSTEM
0.117** 2.28 0.027
THE GOVERNMENT
0.134 *** 2.79 0.032
TR
US
T
PARLIAMENT -0.030 -0.61 -0.007
DEMOCRACY -0.077*** -4.01 -0.018 -0.082 *** -4.21 -0.019 -0.070 *** -3.63 -0.017
RELIGIOSITY -0.381*** -3.5 -0.091 -0.365 *** -3.3 -0.086 -0.415 *** -3.76 -0.101
PRIDE 0.108 1.37 0.026 0.120 1.52 0.028 0.166 ** 2.08 0.040
AWARENESS 0.390*** 3.82 0.089 0.366 *** 3.55 0.083 0.416 *** 4.04 0.096
GENDER (WOMAN)
30 - 49 AGE
Age ≥ 50
EDUCATION
PARTTIME -0.103 -0.37 -0.110 -0.4 -0.066 -0.24
SELF EMPLOYED
-0.149 -1.01 -0.145 -0.97 -0.139 -0.93
RETIRED -0.012 -0.08 -0.032 -0.19 -0.003 -0.02
HOUSEWIFE -0.316 *** -2.5 -0.075 -0.327 *** -2.57 -0.077 -0.320 *** -2.53 -0.078
STUDENT -0.095 -0.57 -0.082 -0.48 -0.089 -0.53
UNEMPLOYED -0.379 ** -1.92 -0.093 -0.421 ** -2.12 -0.104 -0.415 ** -2.08 -0.105
EM
PL
OY
ME
NT
ST
AT
US
OTHER -0.051 -0.16 -0.059 -0.19 -0.035 -0.11
INCOME
Number of Observations 1156 1150 1135
LM(prob>chi2) 0.0000 0.0000 0.0000
Pseudo R2 0.0403 0.0437 0.0386
Notes: Dependent variable: tax morale on a four-point scale. In the reference group: full-time employed, married, read newspaper, male, lower age group (18-29), Marg.: marginal effect. The marginal effect is calculated at the highest TAX MORALE score. Significance levels are: *0.05 < p < 0.10, **0.01 < p < 0.05, ***p < 0.01 . dy/dx is for discrete change of dummy variable from 0 to 1 and dy/dx for factor levels is the discrete change from the base level.
Recep Tekeli Policy Research Institute 58
58 Determinants of Tax Morale in Japan and in Turkey
3.2. Empirical results on tax morale in Japan
Next we present the results of ordered probit models in Japan 17 . Table 41 presents the
estimation results for Japan. We use three trust variables (i.e Trust in Legal System, Trust in
Government and Trust in Parliament), which allow us to analyze trust at the constitutional
level (e.g., trust in the legal system/justice), thereby focusing on how the relationship between
the state and its citizens is established; they also allow us to analyze trust more closely at the
current politico-economic level (e.g., trust in the parliament and government) (see Alm and
Torgler 2006). Trust in Legal System/Justice, Trust in Parliament and Trust in Government
are found insignificant. However Trust in Parliament was found significant in only one
estimation with positive coefficient implying that for those who does have confidence in
parliament they have higher tax morale. An increase in the trust in Parliament scale by one
unit increases the share of subjects indicating the highest tax morale by 3 percentage points.
However this result is not robust across all estimations. In all estimations Trust in Legal
System/Justice and Trust in Government did not have a significant effect on tax morale.
Therefore, our study could not find support for the notion that trust matters for tax morale in
Japan. As trust is not highly and positively correlated with tax morale, Hypothesis 1 can be
rejected.
Specification (4) shows the relative importance of Pride in tax morale determination. Only in
this specification we have significant coefficient on this variable and there is a positive
correlation between Pride and tax morale. The marginal effects indicate that an increase in
Pride by one point raises the share of persons indicating the highest tax morale value by
almost 3 percentage points. Thus, as it is not robust across all estimations the results show that
we can reject our main hypothesis.
The results in Tables also indicate that Hypothesis 3 is rejected. We do not observe a
significant correlation between a prodemocracy attitude and tax morale. Thus we dropped
this variable from the models for Japan. Political attitude was tested by the variable ideology,
the self-placement in a left-right ideological scale from 1 (left) to 10 (right). It has negative
sign but insignificant effect on tax morale. We did not find any evidence to support the view
that individual’s political alignments affect his/her behavior toward taxes in Japan.
17After the first estimation we have not included “trust in the government” and “trust in the Parliament” in the same model because there is a strong correlation between these two variables (r=0.80) that makes it impossible to clearly separate the effects of the two variables if they were included in one model.
Recep Tekeli Policy Research Institute 59
59 Determinants of Tax Morale in Japan and in Turkey
In all specifications in preliminary regressions the variable happiness does not significantly
affect tax morale. The happiness variable has positive marginal effects but it is insignificant to
derive any conclusion from it. Thus our hypothesis 4 cannot be proved statistically. We also
dropped this variable from various estimations.
In line with Torgler’s (2004:249) argument, as previously stated, we included income
variable to test whether individuals in lower income classes are more likely to engage in
criminal activities due to their lower opportunity costs, or not. Robust across all estimations is
that income variable did have significantly negative coefficient. It seems that as Japan has a
progressive income tax schedule, taxpayers with a higher income earn a higher income return
by evading [i.e. a higher marginal tax rate makes tax evasion more profitable]. The result
indicates that an increase in income scale by one point reduces the share of persons indicating
the highest tax morale value by less than 1 percentage point.
In the tax compliance literature, it was argued that self-employed persons have higher
compliance costs than employed (e.g., Lewis, 1982). Therefore, Torgler and Schneider
(2007:452) argued that “taxes are more visible for the self-employed, who have a higher
opportunity to evade or avoid them”. We include different employment status to see if it
affects tax morale. Robust to all estimations is significant coefficient on self-employed. Our
results indicate that the proportion of self-employed who report the highest tax morale is
approximately 18 percentage points lower than the one of full-time employees. The result
indicates that an increase in self-employee by one point reduces the share of persons
indicating the highest tax morale value by 16 to 18 percentage points. The results are in line
with the argumentation that higher compliance costs reduce tax morale.
Tables present evidence regarding the variable religiosity, measured as Religious person. We
first included the variable measured as frequency of Religious attendance but it was
statistically insignificant (see specification 1). As for a Buddhist it may not be an obligation to
attend Religious services and also to make a comparison between Japan and Turkey, we
included the variable religiosity, measured as Religious person on a 3 point scale (i.e. 1 for
religious 3 for atheist). Robust across all 6 specifications is the insignificant correlation
between Tax Morale and Religiosity variables. As argued above we would expect that
religiosity might influence people’s habits, and might make individuals reluctant to engage in
tax evasion. However our findings for Japan are not consistent with those of Alm and Torgler
(2006) and Torgler and Murphy (2004), who find that a higher religiosity is correlated with a
higher tax morale.
Recep Tekeli Policy Research Institute 60
60 Determinants of Tax Morale in Japan and in Turkey
The positive coefficients on education and gender (woman) deserve special attention. In the
models we have run, women seem to have stronger tax morale than men in Japan. Robust
across all 9 specifications is the significant correlation between Tax Morale and Education
variables. Our results show that education has significant effects. We found that if the
individual’s education level increases tax morale increases by 1.5 to 2 percentage points.
However, we found that awareness does not have significant effect on tax morale. This result
does not change even if we drop education variable from the preliminary regressions we run.
Also, older individuals (third age group in the estimations) tend to exhibit higher tax morale
than younger age groups; the coefficients on higher age group are statistically significant in
all estimations18. The proportion of people aged over 50 who report the highest tax morale is
approximately 19 percentage points higher than the 18-29 year old reference group (lower age
category). The result indicates that an increase in old age category by one point increase the
share of persons indicating the highest tax morale value by 16 to 19 percentage points. The
proportion of people of the age 30-49 who report the highest level of tax morale is around 10
percentage points higher than for the 18-29 reference group. In fact we can see from the tables
that the marginal effects increase with an increase in age. For example 50+ age group reports
the highest tax morale higher than the other age groups.
We cannot say that individuals’ constraints by their social networks affect tax morale. In
tables our results indicate that married people having higher tax morale than singles is not
supported by the evidence from Japan either. Actually, marital status did not have any effect
on tax morale in all estimations.
Conclusion: The effects of trust on tax morale were not verified in Japan. Trust in the
government, trust in the court, or the legal system did not have any effect on tax morale. The
probit estimations in all specifications indicate that an increase in the level of income by one
unit decreases the probability of stating that tax evasion is never justified by less than 1
percentage point. Our results indicate that tax morale of self-employee is lower than full-time
employed. We found that while education has significant effects on tax morale level in Japan,
awareness does not affect tax morale. In line with the previous findings in the literature age
has positive effect on tax morale level in Japan. The results on religiosity, however, indicate
that there is no relation between tax cheating and Religion in Japan.
18 The result on age did not change even if we used age as a continuous variable.
Recep Tekeli Policy Research Institute 61
61 Determinants of Tax Morale in Japan and in Turkey
Table 41: Determinants of tax morale in Japan (2005)
Ordered Probit
Model (1) Model (2) Model (3) Variables
Coefficient Z statistic Marginal Effect
Coeff. Z Marg Coeff. Z Marg
HAPPINESS 0.059 0.57 0.014
Living Together 4.077 0.04 0.162 0.008 0.02 0.001 0.049 0.12
Divorced -0.099 -0.27 -0.024 0.044 0.13 0.009 0.056 0.16
Separated -0.213 -0.3 -0.053 -0.026 -0.04 -0.005 -0.076 -0.12
Widowed -0.028 -0.09 -0.006 0.023 0.08 0.005 0.004 0.01
MA
RIT
AL
ST
AT
US
Single/Never -0.224 -1 -0.057 -0.032 -0.15 -0.007 0.041 0.19
IDEOLOGY -0.010 -0.28 -0.002
JUSTICE SYSTEM
0.097 0.95 0.022 -0.041 -0.44 -0.009 -0.031 -0.34
THE GOVERNMENT
0.011 0.08 0.002 0.060 0.72
TR
UST
PARLIAMENT 0.123 0.82 0.029 0.100 1.12 0.022
DEMOCRACY -0.006 -0.16 -0.001
RELIGIOSITY -0.027 -0.54 -0.006 0.031 0.32 0.007 0.024 0.25
PRIDE 0.067 0.81 0.015 0.102 1.38 0.023 0.094 1.26
AWARENESS 0.001 0.01 0.0004
GENDER (WOMAN) 0.224 1.51 0.052 0.321 ** 2.25 0.073 0.285 ** 2.03 0.065
30 - 49 0.179 0.74 0.050 0.267 1.2 0.077 0.345 1.58 AGE
Age ≥ 50 0.494 ** 1.84 0.124 0.656*** 2.7 0.163 0.732 *** 3.08 0.186
EDUCATION 0.065 ** 1.7 0.015 0.073 ** 2 0.016 0.069 ** 1.92 0.015
PARTTIME -.177 -0.83 -0.042 -0.060 -0.3 -0.012 -0.087 -0.44
SELF EMPLOYED
-0.593 *** -3.2 -0.165 -0.656 -3.76*** -0.179 -0.674 *** -3.91 -0.186
RETIRED -0.087 -0.36 -0.019 -0.288 -1.28 -0.068 -0.265 -1.19
HOUSEWIFE 0.212 0.83 0.041 0.168 0.72 0.031 0.203 0.88
STUDENT 0.155 0.32 0.031 0.148 0.32 0.028 0.123 0.27
UNEMPLOYED -0.147 -0.33 -0.034 -0.510 -1.32 -0.132 -0.433 -1.16
EM
PL
OY
ME
NT
ST
AT
US
OTHER -0.134 -0.33 -0.031 -0.519 -1.38 -0.135 -0.309 -0.74
INCOME -0.041* -1.8 -0.009 -0.042 -1.97** -0.009 -0.041 ** -1.94 -0.009
Number of Observations 628 724 736
LM(prob>chi2) 0.0036 0.0001 0.0002
Pseudo R2 0.0610 0.0605 0.0575
Notes: Dependent variable: tax morale on a four-point scale. In the reference group: full-time employed, married, read newspaper, male, lower age group (18-29), Marg.: marginal effect. The marginal effect is calculated at the highest TAX MORALE score. Significance levels are: *0.05 < p < 0.10, **0.01 < p < 0.05, ***p < 0.01 . dy/dx is for discrete change of dummy variable from 0 to 1 and dy/dx for factor levels is the discrete change from the base level.
Recep Tekeli Policy Research Institute 62
62 Determinants of Tax Morale in Japan and in Turkey
Table 41: Determinants of tax morale in Japan (2005) (continued)
Ordered Probit
Model (4) Model (5) Model (6) Variables
Coeff. Z Marg. Coeff. Z Marg Coeff. Z Marg
HAPPINESS
Living Together
Divorced
Separated
Widowed
MA
RIT
AL
ST
AT
US
Single/Never
IDEOLOGY
JUSTICE SYSTEM
-0.001 -0.01
THE GOVERNMENT
0.040 0.52
TR
UST
PARLIAMENT 0.080 0.97
DEMOCRACY
RELIGIOSITY -0.004 -0.05 -0.006 -0.07 -0.002 -0.02
PRIDE 0.124* 1.73 0.028 0.099 1.34 0.101 1.38
AWARENESS
GENDER (WOMAN)
0.267** 1.97 0.060 0.246* 1.83 0.056 0.299** 2.2 0.068
30 - 49 0.294* 1.71 0.087 0.294* 1.72 0.085 0.273 1.57 AGE
Age ≥ 50 0.720*** 3.94 0.181 0.693*** 3.81 0.173 0.669 *** 3.63 0.166
EDUCATION 0.062* 1.77 0.014 0.068* 1.93 0.015 0.073** 2.04 0.016
PARTTIME -0.072 -0.37 -0.075 -0.4 -0.056 -0.29
SELF EMPLOYED
-0.675 *** -3.94 -0.183 -0.670*** -3.92 -0.184 -0.641*** -3.73 -0.174
RETIRED -0.364* -1.68 -0.352 -1.62 -0.337 -1.54
HOUSEWIFE 0.151 0.7 0.215 0.97 0.178 0.8
STUDENT 0.107 0.24 0.104 0.23 0.101 0.22
UNEMPLOYED -0.447 -1.22 -0.432 -1.17 -0.529 -1.4
EM
PL
OY
ME
NT
ST
AT
US
OTHER -0.525 -1.41 -0.244 -0.6 -0.434 -1.18
INCOME -0.037* -1.81 -0.008 -0.039* -1.91 -0.009 -0.039* -1.86 -0.008
Number of Observations 757 757 747
LM (prob>chi2) 0.0000 0.0000 0.0000
Pseudo R2 0.0569 0.0552 0.0581
Notes: Dependent variable: tax morale on a four-point scale. In the reference group: full-time employed, married, read newspaper, male, lower age group (18-29), Marg.: marginal effect. The marginal effect is calculated at the highest TAX MORALE score. Significance levels are: *0.05 < p < 0.10, **0.01 < p < 0.05, ***p < 0.01 . dy/dx is for discrete change of dummy variable from 0 to 1 and dy/dx for factor levels is the discrete change from the base level.
Recep Tekeli Policy Research Institute 63
63 Determinants of Tax Morale in Japan and in Turkey
Table 41: Determinants of tax morale in Japan (2005) (continued)
Ordered Probit
Model (7) Model (8) Model (9) Variables
Coeff. Z Marg. Coeff. Z Marg Coeff. Z Marg
HAPPINESS
Living Together
Divorced
Separated
Widowed
MA
RIT
AL
S
TA
TU
S
Single/Never
IDEOLOGY
JUSTICE SYSTEM
0.067 0.85
THE GOVERNMENT
0.101 1.39
TR
US
T
PARLIAMENT 0.141* 1.81 0.033
DEMOCRACY
RELIGIOSITY
PRIDE 0.086 1.29 0.081 1.21 0.106 1.63 0.025
AWARENESS
GENDER (WOMAN)
0.291** 2.38 0.070 0.249** 2.05 0.059 0.261** 2.13 0.062
30 - 49 0.306** 1.95 0.091 0.326** 2.1 0.097 0.317** 2.04 0.096 AGE
Age ≥ 50 0.683*** 4.08 0.177 0.698 *** 4.22 0.182 0.728*** 4.39 0.190
EDUCATION 0.089*** 2.75 0.021 0.083 *** 2.59 0.019 0.075** 2.33 0.017
PARTTIME -0.047 -0.27 -0.063 -0.37 -0.033 -0.19
SELF EMPLOYED
-0.615*** -3.92 -0.176 -0.648 *** -4.16 -0.188 -0.642*** -4.13 -0.183
RETIRED -0.202 -0.95 -0.220 -1.05 -0.232 -1.11
HOUSEWIFE 0.193 0.96 0.214 1.07 0.153 0.78
STUDENT 0.408 0.96 0.406 0.96 0.382 0.91
UNEMPLOYED -0.356 -0.97 -0.314 -0.86 -0.306 -0.86
EM
PL
OY
ME
NT
ST
AT
US
OTHER -0.158 -0.46 0.025 0.07 -0.240 -0.68
INCOME -0.031 -1.63 -0.007 -0.033* -1.73 -0.007 -0.032* -1.71 -0.007
Number of Observations 830 841 846
LM (prob>chi2) 0.0000 0.0000 0.0000
Pseudo R2 0.0589 0.0565 0.0553
Notes: Dependent variable: tax morale on a four-point scale. In the reference group: full-time employed, married, read newspaper, male, lower age group (18-29), Marg.: marginal effect. The marginal effect is calculated at the highest TAX MORALE score. Significance levels are: *0.05 < p < 0.10, **0.01 < p < 0.05, ***p < 0.01 . dy/dx is for discrete change of dummy variable from 0 to 1 and dy/dx for factor levels is the discrete change from the base level.
Recep Tekeli Policy Research Institute 64
64 Determinants of Tax Morale in Japan and in Turkey
3.3. Empirical results on tax morale in Japan and Turkey
We present the results of ordered probit models in Japan and Turkey combined. To test our
main hypotheses the same political- socio-economic and demographic variables are used to
control for factors affecting tax morale and a country dummy variable (JAPAN) is used to
control for unobserved differences (cultural) across the two countries. As discussed earlier
given the scaled ranking information of the dependent variables, we use an ordered probit
estimation. As the estimating equation is nonlinear, only the sign of the estimated coefficients
is interpreted. Therefore, we compute the marginal effects to estimate the change in the share
of taxpayers (the probability of) belonging to the higher tax morale level when the
independent variable increases by one unit.
Table 42 present the estimation results for the study countries19. We use three trust variables,
(i.e Trust in Legal System, Trust in Government and Trust in Parliament), to check the
robustness of the trust variable. These trust variables allow us to analyze trust at the
constitutional level (e.g., trust in the legal system/justice), thereby focusing on how the
relationship between the state and its citizens is established; they also allow us to analyze trust
more closely at the current politico-economic level (e.g., trust in the parliament and
government) (see Alm and Torgler 2006). Trust in Legal System/Justice and Trust in
Government are found highly significant in various specifications. However Trust in
Parliament in almost all estimation was not significant. While it is significant only in the first
estimation, it has negative coefficient implying that for those who does not have confidence in
parliament they have higher tax morale.
In a separate estimations both trust variables (Trust in Legal System/Justice and Trust in
Government) have a significantly positive effect on tax morale with marginal effects around 3
(Trust in Government) and 2.6 (Trust in Justice/Legal System) percentage points. Therefore,
our study finds support for the notion that trust matters for tax morale. An increase in the trust
in government (trust in legal system) scale by one unit increases the share of subjects
indicating the highest tax morale by 3.1 (2.6) percentage points. As trust is highly and
positively correlated with tax morale, Hypothesis 1 cannot be rejected.
19 As noted above after the first estimation we have not included the variables “trust in the government” and “trust in the Parliament” in the same estimation.
Recep Tekeli Policy Research Institute 65
65 Determinants of Tax Morale in Japan and in Turkey
Table 42: Determinants of tax morale in Turkey& Japan
Ordered Probit
Model (1) Model (2) Model (3) Variables
Coefficient Z statistic
Marginal Effect
Coeff. Z Marg Coeff. Z Marg
HAPPINESS 0.018 0.34 0.010 0.19 0.030 0.55
Living Together 4.170 0.04 4.161 0.04 4.130 0.04
Divorced -0.134 -0.5 -0.084 -0.32 -0.173 -0.64
Separated -0.017 -0.03 0.027 0.05 0.041 0.07
Widowed -0.276 -1.28 -0.267 -1.25 -0.236 -1.1
MA
RIT
AL
S
TA
TU
S
Single/Never -0.035 -0.31 -0.012 -0.11 -0.018 -0.16
IDEOLOGY 0.003 0.17 0.003 0.17 0.0172 1
JUSTICE SYSTEM
0.083 1.56 0.063 1.21 0.1187 **
2.3 0.029
THE GOVERNMENT
0.168 *** 2.79 0.041 0.107 ** 2.13 0.026
TR
US
T
PARLIAMENT -0.115 ** -1.9 -0.028 -0.029 -0.58
DEMOCRACY -0.062 *** -3.27 -0.015 -0.065 ***
-3.47 -0.016 -0.060 ***
-3.22 -0.015
RELIGIOSITY -0.120 -1.54 -0.125 -1.61 -0.030 -0.124 -1.6 -0.030
PRIDE 0.116** 2 0.028 0.107* 1.87 0.026 0.128 ** 2.22 0.031
AWARENESS 0.238** 2.36 0.055 0.233 ** 2.33 0.054 0.255 ** 2.55 0.059
GENDER (WOMAN)
0.031 0.31 0.019 0.2 0.044 0.44
0.008 **
2.19
0.002 0.009** 2.41 0.002 0.008 ** 2.22 0.002 AGE
EDUCATION 0.035 1.63 0.008 0.036* 1.68 0.008 0.032 1.49
PARTTIME -0.112 -0.68 -0.137 -0.85 -0.089 -0.54
SELF EMPLOYED
-0.289 ** -2.42 -0.074 -0.288** -2.43 -0.073 -0.277 **
-2.32 -0.071
RETIRED -0.106 -0.64 -0.117 -0.71 -0.100 -0.61
HOUSEWIFE -0.128 -0.9 -0.106 -0.75 -0.131 -0.93
STUDENT -0.013 -0.08 0.002 0.01 -0.024 -0.14
UNEMPLOYED -0.256 -1.28 -0.229 -1.16 -0.253 -1.27
EM
PL
OY
ME
NT
ST
AT
US
OTHER -0.138 -0.55 -0.130 -0.53 -0.236 -0.98
INCOME -0.015 -0.94 -0.017 -1.05 -0.014 -0.9
JAPAN 0.124 0.95 0.155 1.21 0.118 0.92
Number of Observations 1545 1572 1555
LM(prob>chi2) 0.0001 0.0002 0.0004
Pseudo R2 0.0310 0.0292 0.0280
Notes: Dependent variable: tax morale on a four-point scale. In the reference group: full-time employed, married, read newspaper, male, lower age group (18-29) if included, Marg.: marginal effect. The marginal effect is calculated at the highest TAX MORALE score. Significance levels are: *0.05 < p < 0.10, **0.01 < p < 0.05, ***p < 0.01 . dy/dx is for discrete change of dummy variable from 0 to 1 and dy/dx for factor levels is the discrete change from the base level.
Recep Tekeli Policy Research Institute 66
66 Determinants of Tax Morale in Japan and in Turkey
Table 42: Determinants of tax morale in Turkey& Japan (continued)
Ordered Probit
Model (4) Model (5) Model (6) Variables
Coeff. Z Marg. Coeff. Z Marg Coeff. Z Marg
HAPPINESS
Living Together
Divorced
Separated
Widowed
MA
RIT
AL
S
TA
TU
S
Single/Never
IDEOLOGY
JUSTICE SYSTEM
0.106 ** 2.27 0.026 0.057 1.2
THE GOVERNMENT
0.109** 2.48 0.026
TR
US
T
PARLIAMENT -0.003 -0.08 0.034 0.82
DEMOCRACY -0.047 *** -2.76 -0.011 -0.054 *** -3.21 -0.013 -0.045 *** -2.68 -0.011
RELIGIOSITY -0.167 ** -2.4 -0.041 -0.159 ** -2.28 -0.038 -0.187 *** -2.70 -0.045
PRIDE 0.131 ** 2.56 0.032 0.107 ** 2.11 0.026 0.139 *** 2.73 0.034
AWARENESS 0.195 ** 2.25 0.045 0.172** 1.98 0.040 0.213 ** 2.47 0.049
GENDER (WOMAN)
0.006 ** 2.04 0.001 0.006 ** 2.09 0.001 0.005* 1.93 0.001 AGE
EDUCATION
PARTTIME 0.035 0.24 -0.001 -0.01 0.049 0.34
SELF EMPLOYED
-0.297 *** -2.69 -0.077 -0.305*** -2.79 -0.079 -0.286 *** -2.61 -0.075
RETIRED -0.083 -0.55 -0.120 -0.81 -0.088 -0.59
HOUSEWIFE -0.065 -0.66 -0.062 -0.63 -0.060 -0.61
STUDENT -0.017 -0.12 -0.006 -0.04 -0.008 -0.06
UNEMPLOYED -0.312* -1.81 -0.082 -0.292* -1.72 -0.075 -0.316* -1.85 -0.083
EM
PL
OY
ME
NT
ST
AT
US
OTHER -0.212 -0.91 -0.106 -0.44 -0.183 -0.79
INCOME
JAPAN 0.195* 1.74 0.047 0.211* 1.92 0.051 0.249** 2.26 0.061
Number of Observations 1876 1905 1907
LM(prob>chi2) 0.0000 0.0000 0.0000
Pseudo R2 0.0225 0.0245 0.0207
Notes: Dependent variable: tax morale on a four-point scale. In the reference group: full-time employed, married, read newspaper, male, lower age group (18-29), Marg.: marginal effect. The marginal effect is calculated at the highest TAX MORALE score. Significance levels are: *0.05 < p < 0.10, **0.01 < p < 0.05, ***p < 0.01 . dy/dx is for discrete change of dummy variable from 0 to 1 and dy/dx for factor levels is the discrete change from the base level.
Recep Tekeli Policy Research Institute 67
67 Determinants of Tax Morale in Japan and in Turkey
Table 42: Determinants of tax morale in Turkey& Japan (continued)
Ordered Probit
Model (7) Model (8) Variables
Coeff. Z Marg. Coeff. Z Marg
HAPPINESS
Living Together
Divorced
Separated
Widowed
MA
RIT
AL
S
TA
TU
S
Single/Never
IDEOLOGY
JUSTICE SYSTEM
0.106 ** 2.46 0.025
THE GOVERNMENT
0.129 *** 3.22 0.031
TR
US
T
PARLIAMENT
DEMOCRACY -0.048 *** -2.89 -0.011 -0.054 *** -3.22 -0.013
RELIGIOSITY -0.174 *** -2.52 -0.042 -0.171 ** -2.48 -0.041
PRIDE 0.126 *** 2.5 0.030 0.114 ** 2.25 0.027
AWARENESS 0.196 ** 2.27 0.045 0.185 ** 2.13 0.042
GENDER (WOMAN)
0.006 ** 2.18 0.002 0.006 ** 2.03 0.001 AGE
EDUCATION
PARTTIME 0.0133 0.09 0.007 0.05
SELF EMPLOYED
-0.299 *** -2.74 -0.0774 -0.302 *** -2.77 -0.078
RETIRED -0.105 -0.71 -0.119 -0.8
HOUSEWIFE -0.058 -0.59 -0.065 -0.66
STUDENT -0.018 -0.12 -0.005 -0.04
UNEMPLOYED -0.270 -1.6 -0.068 -0.294* -1.73 -0.075
EM
PL
OY
ME
NT
ST
AT
US
OTHER -0.184 -0.8 -0.076 -0.32
INCOME
JAPAN 0.181* 1.67 0.043 0.244** 2.25 0.059
Number of Observations 1935 1934
LM(prob>chi2) 0.0000 0.0000
Pseudo R2 0.0226 0.0244
Notes: Dependent variable: tax morale on a four-point scale. In the reference group: full-time employed, married, read newspaper, male, lower age group (18-29), Marg.: marginal effect. The marginal effect is calculated at the highest TAX MORALE score. Significance levels are: *0.05 < p < 0.10, **0.01 < p < 0.05, ***p < 0.01 . dy/dx is for discrete change of dummy variable from 0 to 1 and dy/dx for factor levels is the discrete change from the base level.
Recep Tekeli Policy Research Institute 68
68 Determinants of Tax Morale in Japan and in Turkey
Specifications (1-8) show the relative importance of Pride in tax morale determination. In all
pooled estimations we have significant coefficient on this variable. As can be seen from the
Tables, there is a positive correlation between Pride and tax morale. The coefficient is always
positive and the marginal effects indicate that an increase in Pride by one point raises the
share of persons indicating the highest tax morale value by 2.6 to 3.4 percentage points. Thus,
the results show that we cannot reject our second main hypothesis.
We observe a significant but a negative correlation between a prodemocracy attitude and tax
morale. As Table 42 shows, a one-point increase in the prodemocracy index reduces the
share of persons with the highest tax morale approximately 1 percentage point, meaning that a
higher degree of democracy leads to lower tax morale for the reference group. Thus, the
results in Tables indicate that Hypothesis 3 is rejected.
Political attitude was tested by the variable ideology, the self-placement in a left-right
ideological scale from 1 (left) to 10 (right). It has positive sign but insignificant effect on tax
morale. We did not find any evidence to support the view that individual’s political
alignments affect his/her behavior toward taxes. Thus, we dropped this variable from the
regression.
We also included the happiness variable to get a broader view and to consider the fact that a
considerable number of citizens have a low living standard (see Torgler 2004). In all
specifications happiness variable does not significantly affect tax morale. The happiness
variable has insignificant marginal effects to derive any conclusion from it. Thus our
hypothesis 4 cannot be proved statistically and in various estimations this variable is omitted
from the regression.
It was argued by Torgler (2004:249) that “if the financial situation of a household is bad, the
tax payments might be seen as a hard restriction of their possibility set, which might reduce
tax honesty”. In line with this argument we included income variable to test whether
individuals in lower income classes are more likely to engage in criminal activities due to
their lower opportunity costs, or not. However, in all pooled estimations income variable did
not have significant coefficient.
In the tax compliance literature, it was argued that self-employed persons have higher
compliance costs than employed (e.g., Lewis, 1982). Therefore, Torgler and Schneider
(2007:452) argued that “taxes are more visible for the self-employed, who have a higher
opportunity to evade or avoid them”. We include different employment status such as part
Recep Tekeli Policy Research Institute 69
69 Determinants of Tax Morale in Japan and in Turkey
time, self employed, retired, housewife, student and unemployed. In pooled estimation we
find significant coefficient on self-employed and unemployed. An increase in self-
employment by one point reduces the share of persons indicating the highest tax morale
value by almost 8 percentage points. Our results indicate that the proportion of self-employed
who report the highest tax morale is approximately 8 percentage points lower than the one of
full-time employees. The results are in line with the argumentation that higher compliance
costs reduce tax morale.
Our results on employment status also indicate that the proportion of unemployed who report
the highest tax morale is 8 percentage points lower than the one of full-time employees. An
increase in unemployment by one point reduces the share of persons indicating the highest tax
morale value by 8 percentage points. Furthermore, unemployed individuals might have a
higher incentive to act in the shadow economy, which might influence their attitude regarding
tax evasion.
Tables present evidence regarding the variable religiosity, measured as Religious person. We
did not include the variable religiosity measured as frequency of Mosque attendance. As for a
Buddhist it may not be an obligation to attend Religious services and in order to make a
comparison between Japan and Turkey, we included the variable religiosity, measured as
Religious person on a three point scale (i.e. 1 for religious 3 for atheist).
Robust across 5 specifications is the negative correlation between Tax Morale and Religiosity
variables. The negative sign indicates that an increase in the religiosity increases the tax
morale around 4 percentage points20. As argued above we would expect that religiosity might
influence people’s habits, and might make individuals reluctant to engage in tax evasion. Our
findings are consistent with those of Alm and Torgler (2006) and Torgler and Murphy (2004),
who find that a higher religiosity is correlated with a higher tax morale. Our results indicate
that an increase in religiosity increases the tax morale between 3.8 and 4.5 percentage points.
We have positive coefficients on gender (woman) but it is statistically insignificant. In the
models we have run, we do not observe a tendency for women report significantly higher tax
morale level than men.
Our results show that education does not have any significant effects but only in one case we
have a significantly positive coefficient on education (second model). However, we found
20 The results can be given on request by the author.
Recep Tekeli Policy Research Institute 70
70 Determinants of Tax Morale in Japan and in Turkey
that if the reading daily newspaper rate decreases tax morale increases by 4 to 5 percentage
points. This result interestingly indicates that unawareness has significantly positive effect on
tax morale. This result does not change even if we drop education variable from the
estimation.
We cannot say that individuals’ constraints by their social networks affect tax morale. In
tables our results indicate that married people having higher tax morale than singles is not
supported by the evidence from pooled estimation. Actually marital status did not have any
effect on tax morale in the first three estimations and hence we dropped this variable from the
regression model.
We found that older individuals tend to exhibit higher tax morale as the coefficients on age
are statistically significant in all estimations21. A one-point increase in the age increases the
share of persons with the highest tax morale approximately 1 percentage point, meaning that
getting old leads to higher tax morale.
We observe that individuals in Japan have higher tax morale than those in Turkey. The
coefficient on the country dummy JAPAN is statistically significant in models 4 to 8. The
marginal effects indicate that being a resident/or a citizen of Japan rather than of Turkey
increases the probability of reporting the highest tax morale by around 4.4 to 6.1 percentage
points, and this result is robust across various specifications. Perhaps these results support the
argument that tax morale will be higher for countries with “better” governance. For example,
connected to their findings on the determinants of tax morale, Torgler and Shaltegger
(2006:415-16) observe that “there has been a noticeable shift toward a more balanced set of
strategies in tax administrations. In many countries there is an increased emphasis on quality,
customer service and reasonable and fair treatment of taxpayers. …Japan, for example, offers
a good example for a consequent strategy with the aim to intensify an interaction between tax
administration and taxpayers based on trust. … Today, Japan is known for a high degree of
tax morale, which might be influenced by the intensive interaction between the tax
administration and the taxpayers based on trust.”
21 The result on age did not change even if we used age as a categorical variable. However it is meaningful to use education as a continuous variable in a combined data because age groups are weighted differently for both countries (see descriptive data).
Recep Tekeli Policy Research Institute 71
71 Determinants of Tax Morale in Japan and in Turkey
Conclusions: We agree with the statement that not only trust in the government might have
an effect on tax morale, but also trust in the court, or the legal system and hence the way the
relationship between the state and its citizens is established.
The probit estimations in all specifications indicate that an increase in the level of democracy
by one unit decreases the probability of stating that tax evasion is never justified by more than
1 percentage point. Our results indicate that tax morale of self-employed and unemployed is
lower than full-time employed in Turkey and in Japan. We found that while unawareness has
significant effects on tax morale level, education does not affect tax morale in pooled
estimation. Also our findings indicate that older individuals tend to exhibit higher tax morale.
In line with the previous findings in the literature pride has positive effect on tax morale level
in the study countries. The results on religion, indicates that tax cheating is immoral for a
Religious individuals in both countries.
4. Conclusions
It has been argued in the tax compliance literature that audit rate, tax fine and tax rates have
effects on tax payers’ decision to comply. However over last decades researchers have argued
that individual’s tax compliance behavior cannot be explained by traditional economic
analysis considering entirely deterrence components of tax compliance. Tax morale has been
used as a residual while searching the reasons why people pay taxes. Several studies argued
that tax morale, or ‘‘the intrinsic motivation to pay taxes’’, might help to explain the puzzle of
why so many individuals pay their taxes.
Instead of taking tax morale as a black box or a residuum, a number of studies have examined
the determinants of tax morale, and what shapes tax morale or has it being changed over the
time in a specific country and/or across the countries over the world. Many social,
institutional and economic factors have been found to explain the notion of tax morale, or
intrinsic motivation to pay taxes.
We attempt to bring together the numerous insights from the earlier works on tax morale by
examining many social, economical and institutional factors in Turkey and in Japan. In
particular, we use data sets from the fifth wave of WVS that contains information on 1346
individuals in Turkey and 1096 individuals in Japan. WW Surveys offer the best possibility to
investigate tax morale. By analyzing tax morale as a dependent variable, our findings give
further evidence in tax morale literature.
Recep Tekeli Policy Research Institute 72
72 Determinants of Tax Morale in Japan and in Turkey
We report a large number of alternative specifications in the tables, and all specifications
show the marginal effects of the explanatory variables on the highest value of tax morale (i.e
tax cheating is never justified). We found that there are not many similarities between Turkey
and Japan in determination of tax morale. While variables trust (trust in the government and
trust in legal system) (+), pride (+), religiosity (+), the prodemocracy (-), unemployment (-),
house wife (-) and reading daily newspaper rate (-) are found significant in Turkey, self-
employment (-), education (+), gender (women) (+), age (+) and income (-) are significant to
explain tax morale level in Japan. The variable of most interest is JAPAN. The estimated
coefficient on JAPAN is positive and significant across all specifications, and indicates that
tax morale is significantly higher in JAPAN than in Turkey. The marginal effect indicates that
being from Japan rather than from Turkey increases the probability of stating that tax cheating
is never justified by between 4.3 percentage and 6.1 points. Thus our findings show that tax
morale is unambiguously higher in Japan
Bibliography
Allingham, Michael G and Sandmo, Agnar (1972), ‘‘Income Tax Evasion: A Theoretical Analysis.’’ Journal of Public Economics 1:323–38.
Alm, J, G. H. McClelland and W. D. Schulze (1992), “Why Do People Pay Taxes?”, Journal of Public Economics. 48: 21-48.
Alm, J. (1999), “Tax Compliance and Administration”, in: W. B. Hildreth, J. A. Richardson (eds.), Handbook on Taxation. New York: Marcel Dekker, pp741-768.
Alm, J., and Torgler, B. (2006), “Culture Differences and Tax Morale in the United States and Europe”, Journal of Economic Psychology, 27:224-246.
Alm, J., J. Martinez-Vazquez and B. Torgler (2006) “Russian Attitudes Toward Paying Taxes - Before, During, and After the Transition”, International Journal of Social Economics, 33: 832-857.
Alm, J., Martinez-Vazquez, J. and Schneider F. (2004), “Sizing the Problem of the Hard-To- Tax”. Working Paper. Georgia State University: USA.
Alm, J., Martinez-Vazquez, J., and Schneider, F. (2004). “Sizing the Problem of the Hard-To- Tax”, in J. Alm, J.Martinez-Vazquez, & S. Wallace (Eds.), Taxing the hard-to-tax: Lessons from theory and practice (pp. 11–75), Amsterdam: Elsevier/North Holland.
Alm, J., McClelland, G. H., and Schulze, W. D (1999), “Changing The Social Norm of Tax Compliance by Voting”, KYKLOS, 48, 141–171.
Recep Tekeli Policy Research Institute 73
73 Determinants of Tax Morale in Japan and in Turkey
Andreoni, James, Erard, Brian and Feinstein, Jonathan (1998), “Tax Compliance”, Journal of Economic Literature 36:818–60.
Boulding, K. E. (1992), “Towards a New Economics”, Cheltenham, UK: Edward Elgar.
Bruin, J. (2006), Newtest: command to compute new test. UCLA: Academic Technology Services, Statistical Consulting Group. http://www.ats.ucla.edu/stat/stata/ado/analysis/
(21.07.2011).
Clotfelter, C.T. (1983), “Tax Evasion and Tax Rate: An Analysis of Individual Return”, The Review of Economics and Statistics 65, 363–373.
Cowell, F. A. (1990) “Cheating The Government: The Economics of Evasion”, Cambridge: MIT Press.
Cummings, R. G., Martinez-Vazquez, J., McKee, M., and Torgler, B. (2009), “Tax Morale Affects Tax Compliance: Evidence from Surveys and an Artefactual Field Experiment”, Journal of Economic Behavior and Organization, 70:447-457
Cummings, Ronald G., Martinez-Vazquez, J., McKee, M., and Torgler, B. (2005), “Effects of Tax Morale on Tax Compliance: Experimental and Survey Evidence”, Working Paper No 2005–29. Basel: Center for Research in Economics, Management and the Arts.
Dollar; D Fisman R and Gatti R (2001) Are Women really the fairer sex? Corruption and Women in Government, Journal of Economic Behavior and Organizations, 46: 423-429.
Elffers, H. (2000), “But taxpayers do cooperate!”, in M. Van Vugt, M. Snyder, T. R. Tyler, and A. Biel (Eds.), Cooperation in modern society. Promoting the welfare of communities, states and organizations (pp. 184– 194), Routledge, London.
Elffers, Henk, Russell H. Weigel, and Dick Hessing (1987), “The Consequences of Different Strategies for Measuring Tax Evasion Behavior”, Journal of Economic Psychology 8:311-37.
Erard, B. and Feinstein, J. S. (1994), “The Role of Moral Sentiments and Audit Perceptions in Tax Compliance”, Public Finance, 49: 70-89.
Feinstein, J.S. (1991), “An Econometric Analysis of Income Tax Evasion and Its Detection”, RAND Journal of Economics 22, 14–35.
Feld, L. P., and Frey, B. S. (2002), “Trust Breeds Trust: How Taxpayers Are Treated”, Working paper no. 98, Institute of Empirical Research in Economics, University of Zurich.
Feld, L. P., and Frey, B. S. (2007), “Tax Compliance as the Result of a Psychological Tax Contract: The Role of Incentives and Responsive Regulation”, Law & Policy, 29: 102-120.
Feld, L. P., and Tyran, J.-R. (2002), “Tax Evasion and Voting: An Experimental Analysis”, KYKLOS, 55:197-222.
Recep Tekeli Policy Research Institute 74
74 Determinants of Tax Morale in Japan and in Turkey
Frey, B. S. (1997), “Not just for the money. An economic theory of personal motivation”, Cheltenham: Edward Elgar Publishing.
Frey, B. S. and Feld, L. P. (2002), “Deterrence and Morale in Taxation: An Empirical Analysis”, CESifo Working Paper No. 760, August 2002.
Frey, B. S. and Stutzer, A. (2002), “What can economists learn from happiness research?”, Journal of Economic Literature 40(2):402-435.
Graetz, M. J. and Wilde, L. L. (1985), “The Economics of Tax Compliance: Fact and Fantasy”, National Tax Journal, 38:355-363.
Hays, S. (2000), “An Empirical Analysis of Taxpayers’ Attitudes amd Behavioral Intentions Regarding Compliance with Federal Income Tax Laws, Dissertation”, College of Adminstration and Business, Louisiana Tech University.
Hirschi, Travis, and Rodney Stark (1969), “Hellfire and Delinquency”, Social Problems 17:202–13.
Iannaccone, L.R. (2002), “A Marriage Made in Heaven? Economic Theory and Religious Studies”, in: Grossbard- Shechtman, S., Clague, C. (Eds.), The Expansion of Economics. M.E. Sharpe, New York, pp. 163–193.
Inglehart, R (2000), “Codebook for World Values Survey”, Ann Arbor: Institute for Social Research.
Jackson, B. R. and Milliron, V. C. (1986), “Tax Compliance Research: Findings, Problems, and Prospects”, Journal of Accounting Literature. 5: 125-166.
Lago-Peñas, Ignacio and Lago-Penas, Santiago (2008), “The Determinants of Tax Morale in Comparative Perspective: Evidence from a Multilevel Analysis” P. T. N.o 2/08
Lewis, A. (1979), “An Empirical Assessment of Tax Mentality”, Public Finance, 34:245-257.
Lewis, A. (1982), “The Psychology of Taxation”, Martin Robertson, Oxford.
Long, S., and Swingen, J. (1991), “The conduct of tax-evasion experiments: Validation, analytical methods, and experimental realism”, in P. Webley, H. Robben, H. Elffers, & D. Hessing (Eds.), Tax evasion: An experimental approach (pp. 128–138). Cambridge: Cambridge University Press.
Margolis, H. (1997), “Religion as paradigm”, Journal of Institutional and Theoretical Economics 153, 242–252.
McGee, R.W. (1998), “The ethics of tax evasion in Islam”, in: McGee, R.W. (Ed.), The Ethics of Tax Evasion. The Dumont Institute for Public Policy Research, South Orange, pp. 214-219.
Montero, J. R., and Torcal, M. (2006), “Political disaffection in comparative perspective”, in Torcal, M., and Montero, J. R. (eds.): Political Disaffection in Contemporary Democracies: Social Capital, Institutions, and Politics, Routledge. London.
Murphy, K. (2004), “The Role of Trust in Nurturing Compliance: A Study of Accused Tax Avoiders”, Law and Human Behavior 28:187-209.
Recep Tekeli Policy Research Institute 75
75 Determinants of Tax Morale in Japan and in Turkey
Nerre, B. (2008), “Tax Culture: A Basic Concept for Tax Politics”, Economic Analysis & Policy, 38 (No. 1): 153-167.
Pommerehne, W. W. and Weck-Hannemann, H. (1996), “Tax Rates, Tax Administration and Income Tax Evasion in Switzerland”, Public Choice 88: 161-170.
Pommerehne, W. W., Hart, A., and Frey, B. S. (1994), “Tax morale, tax evasion and the choice of policy instruments in different political systems”, Public Finance 49(Suppl.), 52–69.
Roth, J. A., Scholz, J. T., and Witte, A. D. (1989), “Taxpayer compliance”, Vols. 1 and 2. Philadelphia: University of Pennsylvania Press.
Schmölders, G. (1960), “Das Irrationale in der öffentlichen Finanzwissenschaft”, Hamburg: Rowolt.
Schmölders, G. (1962), “Volkswirtschaftslehre und Psychologie”, Berlin: Reinbek.
Schmölders, G. (1970), “Survey Research in Public Finance: A Behavioral Approach to Fiscal Theory”, Public Finance 25: 300-306.
Schneider, F. (2003), “The Shadow Economy”, in: Rowley, C.K., Schneider, F. (Eds.), Encyclopedia of Public Choice, Kluwer Academic Publishers, Dordrecht.
Schneider, F. (2005), “Shadow Economies Around the World: What do we Really Know?”, European Journal of Political Economy, 21(3): 598–642.
Schneider, F. and Enste, D. (2000), “Shadow Economies: Size, Causes, and Consequences”, The Journal of Economic Literature 38(1): 77–114.
Schneider, F. and Enste, D. H. (2002), “The Shadow Economy An International Survey”, Cambridge: Cambridge University Press.
Schneider, Friedrich, Buehn, Andreas and Montenegro, Claudio E. (2010), “Shadow Economies all over the World: New Estimates for 162 Countries from 1999 to 2007 Submitted to the EPCS 2010”, Izmir, Turkey
Scholz JT and Lubell, M (1998), “Trust and Taxpaying: Testing the Heuristic Approach to Collective Action”, 42 American Journal of Political Science 398.
Schwartz, R., and Orleans, S. (1967), “On legal sanctions”, University of Chicago Law Review 34, 282-300.
Slemrod (ed.) (1992), “Why People Pay Taxes”, Tax Compliance and Enforcement. Ann Arbor University of Michigan Press.
Song, Y. and Yarbrough, Y. E. (1978), “Tax Ethics and Taxpayer Attitudes: A Survey”, Public Administration Review 38: 442-457.
Spicer, M. W. and Lundstedt, S. B. (1976), “Understanding Tax Evasion”, Public Finance 31: 295-304.
Recep Tekeli Policy Research Institute 76
76 Determinants of Tax Morale in Japan and in Turkey
Strümpel, B. (1969), “The Contribution of Survey Research to Public Finance”, in: A. T. Peacock (ed.), Quantitative Analysis in Public Finance, New York: Praeger Publishers: 14-32
Swamy, Anand, Stephen Knack, Young Lee, and Omar Azfar (2001) “Gender and Corruption”, Journal of Development Economics 64, 25–55.
Tittle, C. (1980), Sanctions and Social Deviance: The Question of Deterrence. Praeger, New York.
Torgler, Benno (2001), “Is Tax Evasion Never Justifiable?”, Journal of Public Finance and Public Choice XIX: 143-168.
Torgler, Benno (2001), “What Do We Know about Tax Morale and Tax Compliance?”, RISEC: International Review of Economics and Business 48:395–419.
Torgler, Benno (2002), “Speaking to Theorists and Searching for Facts: Tax Morale and Tax Compliance in Experiments”, Journal of Economic Surveys, 16:657–684.
Torgler, Benno (2003), “To Evade Taxes or Not: That is the Question”, Journal of Socio-Economics, 32:283-302.
Torgler, Benno (2004), “Tax Morale in Asian Countries”, Journal of Asian Economics, 15: 237-266.
Torgler, Benno (2005), “Tax Morale and Direct Democracy”, European Journal of Political Economy, 21:525-531.
Torgler, Benno (2005), “Tax morale in Latin America”, Public Choice, 122: 133-157.
Torgler, Benno (2006), “The Importance of Faith: Tax Morale and Religiosity”, Journal of Economic Behavior & Organization, 61: 81-109.
Torgler, Benno (2007), “Tax Compliance and Tax Morale: A Theoretical and Empirical Analysis”. Cheltenham: Edward Elgar.
Torgler, Benno and Murphy, K (2004), “Tax Morale in Australia: What Factors Shape It and Has It Changed Over Time?”, Journal of Australian Taxation, 7: 298-335.
Torgler, Benno and Schaltegger, C. A (2006), “Tax Morale: A Survey with a Special Focus on Switzerland”, Schweizerische Zeitschrift für Volkswirtschaft und Statistik, 142 (3): 395-425.
Torgler, Benno and Schneider, Friedrich (2005), “Attitudes Towards Paying Taxes in Austria: an Empirical Analysis”, Empirica 32:231–250
Torgler, Benno and Schneider, Friedrich (2007a), “Shadow Economy, Tax Morale, Governance and Institutional Quality: A Panel Analysis”, QUT School of Economics and Finance Working/Discussion Paper 210, February 2007.
Torgler, Benno and Schneider, Friedrich (2007b), “What shapes attitudes toward paying taxes? Evidence from multicultural European countries”, Social Science Quarterly, 88:443–470.
Recep Tekeli Policy Research Institute 77
77 Determinants of Tax Morale in Japan and in Turkey
Torgler, Benno and Schneider, Friedrich (2009), “The Impact of Tax Morale and Institutional Quality on The Shadow Economy”, Journal of Economic Psychology 30 (2009) 228–245
Torgler, Benno, Schneider, Friedrich and Schaltegger, Christoph A. (2010), “Local Autonomy, Tax Morale, And The Shadow Economy”, Public Choice 144: 293-321.
Torgler, Benno and Valev, Neven T (2006), “Women and Illegal Activities: Gender Differences and Women’s Willingness to Comply over Time”, CREMA Working Paper Series 2006-15, Center for Research in Economics, Management and the Arts (CREMA).
Tyler, T. R. (2000), “Why do people cooperate in groups?”, in M. Van Vught, M. Snyder, T. R. Tyler, & Anders Biel (Eds.), Cooperation in modern society. Promoting the welfare of communities, states and organizations (pp. 65–82). London: Routledge.
Vogel, J. (1974), “Taxation and Public Opinion in Sweden: An Interpretation of Recent Survey Data”, National Tax Journal 27:499–513.
Yusuf, S.M., (1971), “Economic Justice in Islam”, Sh. Muhammad Ashraf, Lahore.
Recommended