European Journal of Political Research 35: 341­370, 1999.
 1999 Kluwer Academic Publishers. Printed in the Netherlands.
341
Trust in government and welfare regimes: Attitudes to
redistribution and financial cheating in the USA and Norway
JONAS EDLUND
Department of Sociology, Ume University, Sweden
Abstract. Claims have been made that national institutions influence public preferences, as
well as structuring patterns of social division. This article analyses attitudes to redistribution
and financial cheating in Norway and the USA. On the aggregate level the results show that
there are striking differences between the two countries regarding attitudes to redistribution
and confidence in the state, while similar attitude patterns are found regarding cheating with
taxes and benefits. Results endorse arguments emphasising that the design and scope of welfare
state policies shape and determine their own legitimacy. There is less support for political
trust arguments, which emphasise that the efficacy of political decision-making institutions
promotes beliefs about trust in the state and views on government responsibilities. Similarly,
arguments proposing that advanced welfare statism has undesirable effects on civic morality,
such as cheating on taxes and benefits, are not supported empirically. Finally, while conflicts
over redistribution are similarly structured in the USA and Norway, divisions over financial
cheating are less clear-cut and vary cross-nationally.
Introduction
When it comes to explaining variations in welfare state developments, implementation
and effects of welfare policies institutional approaches have
proven to be fruitful. While the types or aspects of institutions that have
been subject to research differ, scholars commonly acknowledge that institutions
affect preferences among social actors (Steinmo 1993; Steinmo & Watts
1995; Evans et al. 1985; Skocpol 1992; Weir 1992; Esping-Andersen 1990;
Korpi 1983; cf. Carnoy 1984). However, research within this field has largely
neglected the empirical examination of relationships between institutional
characteristics and public preferences towards the welfare state.
Among arguments claiming that institutions matter for public preferences,
different themes can be distinguished. One theme, represented by
Steinmo (1994), focuses on the role and functioning of political decisionmaking
institutions for building political trust in government and welfare
policies. Explaining why Americans in general distrust government and oppose
governmental intervention in the field of welfare, it is argued that "the
fragmentation of power and authority has stripped our political system of
342 JONAS EDLUND
efficacy. When American governments do act, they too often act badly. In
short, Americans have come to distrust their government because it doesn't
work very well" (Steinmo 1994: 106).
A second theme emphasises that the organisation and scope of welfare
state policies influence and feed attitudes and beliefs concerning the role of
the state in society (Korpi 1980). In essence it is argued that welfare states
relying on encompassing welfare policies promote broad coalitions of welfare
state support among the electorate, while welfare states in which welfare
programmes targeted at those worst off dominate, encourage hostile attitudes
to welfare state policies among the public at large.
A third theme is the claim that extensive welfare policies are harmful to
economic competitiveness. While empirical evidence for this claim is limited
and to some extent contradictory (Pfaller et al. 1991a; Atkinson & Mogensen
1993), political strategies based on such assumptions have been common,
perhaps most notably in Great Britain and the USA during the 1980s (Weir
1992; Pierson 1994; Ginsburg 1992; Sears & Citrin 1985). The core of these
arguments is that extensive taxation and state benefits promote economically
and to some extent morally undesirable behaviour, which undermine
economic competitiveness and the foundations of the society itself.
Approaching the issues raised by these themes through a comparative analysis
of popular attitudes in the USA and Norway concerning some aspects
of the welfare state ­ namely to what extent citizens support redistribution,
and the public's financial morality ­ some relevant research questions can
be formulated. To what extent do Americans and Norwegians have different
views on the welfare state? Are the observed attitudinal patterns better explained
by political trust arguments or by arguments emphasising the scope
of welfare policies? More specifically, do citizens distrusting the state have
different views on redistribution and financial cheating from those having
confidence in the state? Are tolerant attitudes to tax cheating and claiming
state benefits illegally more prevalent in an advanced welfare state such as
Norway than in the United States? Is support for welfare state redistribution
stronger in a country relying on encompassing welfare state programmes as
in Norway, than in the more limited and `residuaľ US welfare state?1
Secondly, the analysis will explore how support for the welfare state is distributed
along social cleavages. This analysis takes its departure in the field of
`welfare-state regimes' (Esping-Andersen 1990). In short, it has been argued
that different types of welfare state regimes will generate different attitude
patterns. The aim of this analysis is to examine whether different welfare state
arrangements present in the two countries have contributed to weakening or
strengthening different kinds of social conflict lines among the Norwegian
and American publics. Here, we will pay attention to examining whether
TRUST IN GOVERNMENT AND WELFARE REGIMES 343
class-related factors, religion, sector, age, gender, and political choice have
significant relationships with attitudes concerning the welfare state.
Institutions and preferences
`Institution' as a concept is less than clear-cut (Hall & Taylor 1996). For analytical
purposes, a narrow definition is to be preferred, as a broad definition
undermines the analytical utility of the concept (Rothstein 1996; Thelen &
Steinmo 1992). Excluding factors such as class structures, `political culture',
and informal norms, institutions can be defined as "formal arrangements
for aggregating individuals and regulating their behavior through the use of
explicit rules and decision processes enforced by an actor or set of actors
formally recognized as possessing such power" (Levi 1990: 405). With this
definition tax systems and welfare state programmes, as well as the set of
rules regulating the political system and state bureaucracy, may be considered
as institutions.
Institutions may be viewed as intervening factors, and they matter for
attitudes in at least two important respects. From a state- or polity-centred
perspective, it has been emphasised that the state and its institutional setting
and policy legacies define the frames for negotiations between different social
groups and for what is desirable for groups to achieve (Weir & Skocpol 1985:
118; Skocpol 1992). As institutions in certain respects are conceived as normative
orders, they influence and structure world views and preferences among
the public (Rothstein 1994: 24; Thelen & Steinmo 1992: 27; Korpi & Palme
1995). Public policies are thus not only products of political forces. Policies in
themselves create interests and shape political conflicts both among political
elites and in the electorate. These feedback processes in turn constrain and
shape government behaviour (Pierson 1993).
On the other hand, institutions may be viewed as a product of ­ or reflecting
­ causal factors such as ideologically-based conflicts among different
social actors (Hanson 1994; Korpi 1996). Across nations, these conflicts may
give rise to quite different institutional designs, and these institutions are the
channels through which popular attitudes and the preferences of organised interests
such as political parties and interest groups are translated into political
action.
Acknowledging that interests and preferences are context-dependent and
will vary in different institutional settings, it should however be stated that
research studies have been less clear about institutional feedback processes,
and "particularly those directly affecting mass publics ( . . . ) have yet to receive
sufficient attention" (Pierson 1993: 597). To what extent institutions
344 JONAS EDLUND
determine public preferences should rather be considered as an empirical
question, not as an axiom (Putnam 1993: 17f; Rothstein 1996: 148).
Norway and the USA: Different welfare states
The classification schema of welfare states provided by Esping-Andersen
(1990) focuses principally on the arrangements between the state, the market,
and the family, concerning the organisation of social protection.2
Although
different or modified classification criteria have been suggested, there seem
to be little disagreement that Scandinavian countries fit the social democratic
welfare state regime-type, while the United States is considered to be perhaps
the prime example of a liberal welfare state regime (Esping-Andersen 1990;
Castles & Mitchell 1993; Korpi & Palme 1995; Stephens 1994; Olsen 1994;
Siaroff 1994).
The liberal welfare state regime is characterised by the circumscribed role
of the state and the high reliance on market solutions. Social spending and tax
revenue levels are relatively low. The state offers a basic security net for those
with low incomes, while others that are better off are supposed to protect
themselves via various private forms of insurance. State benefits are generally
means-tested or provided at a low flat rate. In the social democratic regime
type, social protection is linked to citizenship and, in some cases, long-term
residency. Most of the population are covered by the core programmes of the
welfare state. Benefits are either income-related, together with a rather high
basic security level for those with low or no income, or provided as citizens'
rights. Social spending levels are comparatively high and are mainly financed
by tax revenues.
Welfare outcomes produced by these institutional designs differ substantially.
Almost regardless of which measurements used for classification,
among major OECD countries, the limited role of the American welfare
state is evident, while Scandinavian countries are located in the top bracket
regarding `welfare statism' (Esping-Andersen 1990; Fritzell 1991; Kangas
1991; Palme 1990; Castles & Mitchell 1993; Korpi & Palme 1995; McFate
et al. 1995a). The selected socio-economic indicators in Appendix 1
suggest that while labour force participation is quite similar in Norway and
the United States, striking differences exist regarding taxation levels, social
spending levels, public service employment, income distribution, poverty
levels, de-commodification scores, as well as differences in the organisation
and political power of left movements.
The effects of the strategies dealing with income inequality and poverty
differ radically between the USA and Norway. While Scandinavian countries
have adopted encompassing welfare policies, the USA has been concerned
TRUST IN GOVERNMENT AND WELFARE REGIMES 345
with the very worst off in society, implementing targeting and means testing
programmes instead of universalistic policies in the war on poverty. However,
known as the `paradox of redistribution', it has been demonstrated that countries
fighting poverty via flat rate benefits along with targeting benefits to the
poor have been less successful in reducing poverty than countries relying on
encompassing welfare programmes (Korpi & Palme 1995: 21).3
Relying on
statistics from the 1980s, the USA seems rather exceptional. While tax and
transfer systems in a number of countries lift a fairly large share of working
age households out of poverty, the American tax system pushes about
the same number into poverty as the transfer system lifts up (McFate et al.
1995b).4
It has in this context been argued that the scope and organisation of the
welfare state to a significant degree structure interests and patterns of political
coalition formation (Esping-Andersen 1990; Korpi 1980; Korpi & Palme
1995). The characteristics of the liberal welfare state, such as the USA are
likely to discourage coalitions between poor and better off citizens. Targeting
programmes as well as benefits at a low flat rate tend to split the working
class and promote coalitions between better off workers and the middle class
against segments of the working class with less economic resources (cf. Edelmann
1994: 244f). As issues of social protection within this regime type tend
to be framed as political conflicts between `us' (the majority paying taxes)
and `them' (those, often equipped with less political resources, receiving
benefits without paying), support for welfare policies in the USA is limited.
Welfare states which have institutionalised encompassing programmes, as in
Scandinavia, are expected to encourage more solid support and demand for
continuing welfare policies since the working class as well as the middle class
is included and covered by the same insurance programmes.
From a quite different perspective, a variety of arguments concerning the
negative effects of welfare statism have been raised by economists, politicians
and business interests during the two last decades. In different contexts
and levels of abstraction it has been argued that income security and heavy
taxation distort labour market functions, weaken work motivation, promote
an entitlement ideology, foster a culture of dependence, and feed tax resistance
(Lindbeck 1988; cf. works quoted in Pfaller et al. 1991b; Gough 1991;
Atkinson & Mogensen 1993). If these kinds of argument are valid, it is likely
that Norwegians are more prone to justify tax cheating and claiming state
benefits illegally than Americans. It should, however, be emphasised that
measuring attitudes to financial cheating, as attempted in this article, is an
imperfect means to estimate actual behaviour of cheating. While it is true that
links between attitudes and behaviour may be less than straightforward, it has
346 JONAS EDLUND
nevertheless been shown that those believing that cheating is a trivial crime
are themselves more likely to be cheaters (Laurin 1986; Wahlund 1991).
While the above approaches suggest that the characteristics of welfare
policies define interests, affect behaviour, and promote certain coalitions
among social groups, others argue that the structure of the political decisionmaking
institutions, within which these political struggles take place, critically
determine policy outcomes, political trust, and by extension, views on
the role of the welfare state.
Developed in an American perspective, Steinmo (1994) is particularly
critical of the traditional liberal cultural explanation in the tradition of Hartz
(1955), which claims that "the state plays a more limited role in America
than elsewhere because Americans, more than other people, want it to play a
limited role" (King 1973 quoted in Steinmo 1994: 107).5
Contrary to the cultural
approach, it is argued that public preferences about the role of the state
in society are particularly influenced by the efficiency of decision-making
institutions.
Following Steinmo, the institutional foundations upon which countries'
democratic systems are built seriously affect the possibilities to govern efficiently.
In sharp contrast to the Scandinavian model of democracy, with
its relatively centralised government and `non-politicaľ state bureaucracy,
democracy in the USA can be characterised as a system with multiple checks
and balances with fragmentation of power and political authority, as well as
a weak national administration with virtually "no sphere of `administration'
apart from politics" (Banfield & Wilson 1963 in Lipset 1991: 11; Weir &
Skocpol 1985: 136). In absence of strong ideologically-coherent political
parties rooted in class structures ­ as American parties tend to be nonprogrammatic
and internally divided with historical roots in regions rather
than social classes ­ the American system "has come to represent the interests
of citizens through a nexus between interest group and bureaucracy rather
than between political party and elected official" (Jillson 1994: 57).
The fragmentation built into the system profoundly shapes both the political
strategies of the actors involved and policy outcomes (Steinmo 1993;
Weir 1992; cf. Rothstein 1994: 106). Two implications of this hyper-pluralism
is first that it "leaves members of Congress individually accountable for their
actions. They thus face powerful incentives to pay closer attention to the
short-term electoral consequences of their votes than to the long-term policy
effects of their actions" (Steinmo 1994: 125).
Second, it promotes special interest politics, as "elected officials must
cater to local or highly particularistic constituency interests to an extent that
is truly unique in the democratic world" (Steinmo 1994: 117). Since the
"political history [of US social policy making] tells a common story: Political
TRUST IN GOVERNMENT AND WELFARE REGIMES 347
and administrative reformers design a plan which is radically altered, watered
down, or rejected as it moves through the legislative branch", Americans have
realised that their government is not efficient nor trustworthy (Steinmo 1994:
125; cf. Steinmo 1993: 142ff).6
"[W]hat citizens believe about politics, and what they think is possible
and desirable is fundamentally shaped by what government does for them.
( . . . ) Thus if a government is disorganized, inefficient, and ineffective citizens
will want it to do different things, than they will if their experience
tells them that it is efficacious" (Steinmo 1994: 128). In sharp contrast to the
USA, Sweden (and probably other Scandinavian countries) with its efficient
decision-making institutional design, it is argued, promoted economic growth
while at the same time providing the revenue for the expansion of the welfare
state. "This substantive experience built a continued faith in the legitimacy of
the state's role in society" (Steinmo 1993: 207). Similarly, Rose (1991: 210f)
believes that "[o]nly if the state is trusted to be caring and effective does it
make sense for people to put their welfare in the hands of officials armed
with the power of law and the resources of the fisc". It is then stated that
public officials "appear more fully trusted in Northern Europe than in Southern
Europe." In the USA, however, the situation seems considerably worse as
governments are perceived as "dishonest, uncaring, and ineffective".7
In the same vein, Lewis (1982: 90ff) argues that tax evasion is related to
beliefs concerning government legitimacy. Citizens distrusting government
are supposed to view tax evasion as less wrong than those trusting government.
Similarly, Sears & Citrin (1985: 8) track one of the roots of the
tax revolt in California (proposition 13) to a generalised political cynicism
expressed as antigovernment sentiment. Listhaug & Miller (1985) examining
Norwegian data, found fairly strong relationships between confidence
in government institutions and attitudes to tax evasion. On the basis of this
finding they predict that "if the distribution of public values was to shift in
an anti-system direction, as has occurred in the USA in recent years, we
would expect increased support for tax cheating, especially among the less
politically involved" (Listhaug & Miller 1985: 279; cf. Peters 1991: 223).
As the above arguments claim that experiences of state actions affect attitudes
about what government should and should not do, we therefore need to
examine whether citizens with a great deal of confidence in political institutions
have different views on redistribution and financial cheating than those
distrusting the political system.
348 JONAS EDLUND
Data and methods
Data used in this article were collected within the International Social Survey
Program (ISSP) in 1991.8
One main reason behind forming ISSP was to
create truly comparable data for attitude studies. Identical surveys on various
topics have been conducted every year in a growing number of countries since
1985 (Davis & Jowell 1989; Svallfors 1996).
While a comparative research design can be fruitful when studying relationships
between public attitudes and institutional characteristics, it is
also afflicted with problems that suggest cautious interpretations of results
(Küchler 1987). Besides the fact that survey data are sensitive to question
wording, question context, and to question sequence, one major obstacle in
comparative research is that the meaning of a particular question may differ
among countries (Svallfors 1996; cf. Smith 1987). This suggests that we
should focus on broad attitudinal patterns (a `gestalť) and neither pay too
much attention to small differences in marginal percentages on single questions,
nor take them at face value (Scheuch 1989). Further, to use indexes
instead of single questions may also help to increase the validity of the results,
since the impact of wording effects and other sources of measurement error
are reduced.
The response rate for Norway is 62 percent of a total net sample of 2426
citizens. The corresponding figures of the United States net sample are 70
percent and 1950, respectively. Both samples are nationally-representative,
unweighted, and randomly drawn (ISSP 1991). An analysis of non-responses
shows that neither the Norwegian sample nor the American are particularly
biased regarding age, gender, region, education or labour force status. In Norway,
those with university education and those in the labour force are slightly
over-represented, while in the USA a small gender bias appears, as females
are over-represented in the sample.
In the following analysis frequency tables are used for descriptive purposes.
Then factor analysis is applied to distinguish attitudinal patterns
among the variables. Finally, cross-tabulation tables and Multiple Classification
Analysis (MCA) are used for group comparisons. MCA, using the
F-statistic for significance test, is a statistical method which suits multivariate
analysis when the independent variables are categorical and the dependent
variable is at least on the interval scale level (Andrews et al. 1973). The
MCA output is in this article restricted to coefficients controlled for other
independent variables.
TRUST IN GOVERNMENT AND WELFARE REGIMES 349
Attitudes to redistribution, financial cheating and decision-making
institutions
In the following empirical analysis, American and Norwegian public attitudes
concerning the state will be compared. The areas covered in the analysis are
(1), confidence in parliament and government; (2), government responsibility
regarding redistribution, and (3), tax evasion and claiming state benefits
illegally.
As pointed out above, it is expected that American and Norwegian beliefs
concerning government trust, redistribution, and cheating on government will
be different. First, if the ways governments structure access to state benefits
have implications for popular attitudes, we would expect stronger support
in Norway for redistribution in comparison with the USA. Second, whether
tolerance for financial cheating is related to heavy taxation or to changing
moral standards caused by advanced welfare statism, it is from the `negative
effects' perspective that we expect that Norwegians will express more tolerant
attitudes to tax evasion and cheating on benefits than the American citizenry.
Third, according to political trust arguments, the state is supposed to receive
less support in the USA than in Norway. It is further hypothesised that citizens
with a high degree of confidence in government tend to be more oriented to
collective action than those distrusting government. Similarly, those distrusting
government may tend to be more politically alienated and thus prefer
individual to collective actions, including illegal actions `outside' the system,
such as evading taxes. Finally it is hypothesised that attitudinal differences
regarding redistribution and financial cheating across nations will decrease
holding confidence in the state constant.
Results in Table 1 indicate that confidence in the state is more prevalent
in Norway than in the United States. Further, it is evident that differences
between Americans and Norwegians are even more pronounced when it
comes to government responsibilities for redistribution.9
In Norway the support
for government is widespread, while Americans only show moderate
support with those opposing outnumbering those favouring government involvement.
Examining attitudes to tax evasion and cheating on state benefits,
we get an entirely different picture. Observed differences between the two
countries are much smaller, and it seems that the claiming of benefits from
government by a person who is not entitled to do so, is regarded as more
incorrect morally than cheating on taxes.
Moving to the kernel of the analysis, relationships between confidence in
government on the one hand, and views on government responsibilities, tax
evasion and getting benefits illegally, on the other hand, will be examined.
Beginning with a dimensional analysis, displayed in Table 2, it is clear that
attitudes are similarly structured in both countries. The variables form three
350 JONAS EDLUND
Table 1. Attitudes towards the state in Norway and the USA: Confidence in parliament and
civil service; government redistribution; financial cheating, 1991 (%)
USA Norway USA Norway
[Parliament] [Civil service]
How much confidence do you US Congress Norwegian Government Government
have in . . . Parliament department
Complete or a great deal 25 40 17 30
of confidence
Some confidence 45 36 48 42
Very little or no confidence 27 21 33 25
at all
Can't choose 3 4 3 3
Balance -2 +19 -16 +5
(n) (1331) (1460) (1312) (1433)
Do you think it should or should Provide a job for everyone Reduce income differences
not be the governmenťs who wants one? between the rich and the poor?
responsibility to . . .
Definitely should be 18 56 20 50
[Yes, completely]
Probably should be 25 31 25 26
[Yes, partially]
Probably should not be 25 10 24 17
[NO]
Definitely should not be 27 ­ 24 Can't
choose 5 3 8 7
Balance -9 +77 -3 +59
(n) (1338) (1468) (1326) (1431)
Do you feel it is wrong or not A taxpayer does not report A person gives the government
wrong if . . . all of his income in order to incorrect information about
pay less income tax? himself to get government benefits
that he is not entitled to?
Seriously wrong 29 33 48 63
Wrong 53 43 44 33
A bit wrong 12 17 4 3
Not wrong 4 5 2 0
Can't choose 3 1 2 1
Balance +66 +54 +86 +93
(n) (1332) (1472) (1332) (1466)
Source: ISSP (1991).  Response categories for Norway in brackets.  Balance: percentage
points difference between `state friendly' and `state hostility' attitudes.
TRUST IN GOVERNMENT AND WELFARE REGIMES 351
distinct dimensions. The first dimension consists of trust-items (factor I).
The second and third dimensions capture questions concerning redistribution
(USA factor III, Norway II) and financial cheating (USA factor II, Norway
III), respectively. Based on these results three additive indexes with two items
in each were constructed.
These indexes are labelled `government redistribution', `financial cheating',
and `confidence in state'. According to cross-tabulation tables, the
indexes were constructed in the following way: those with `state friendly' responses
(agreeing that it is the responsibility of government to reduce income
differences and provide jobs; opposing financial cheating regarding taxes and
benefits; and having confidence in parliament and the civil service) were assigned
a value of 2 for each indicator. Those with `state hostile' attitudes
were assigned a zero value. Those answering can't choose, and the middle
category in the two confidence items, received the value of 1. The values of
the indicators within each index were then summed. Missing values on one indicator
within an index were assigned the same value as the other indicator.10
These indexes vary between zero and 4. Values below 2 are pooled into one
category (`state hostility'), and values larger than 2 are pooled into a `state
friendly' category. Hence, these three indexes contain three categories each.
Analysing group differences using MCA, which will appear later, the additive
indexes are constructed following the same procedure, except that categories
are not pooled. Indexes vary initially between zero and 8 (except for the Norwegian
redistribution index which varies between zero and 6, due to different
response categories), but are transformed to vary between zero and 1.
Table 3 shows the results of the analysis.11
Beginning with the USA,
relationships between confidence in state and financial cheating are nonsignificant.
The only significant but rather weak relationship found is that
Americans with low confidence in the state are less in favour of government
involvement than those with strong or moderate confidence in the state. In
Norway on the other hand, the expected patterns are found in both areas,
however only weakly with regards to redistribution. But it is clear that trust
in government is related to views on financial cheating.12
In sum, there is considerable agreement between institutional characteristics
of welfare policies and support for redistribution. Arguments stressing
the role of political trust for attitudes to government involvement receive less
support, as the striking attitudinal differences to redistribution in Norway and
the USA remain largely unaffected by control for trust in government. Having
a great deal of confidence in the state does not result in any particular special
requirements concerning the role of the state.
Considering the large differences in terms of taxation and social spending
levels, as well as the structure of welfare state policies, it could be expected
352JONASEDLUND
Table 2. Dimensional analysis of attitudes towards the state in Norway and the USA. 1991. Factor loadings × 100. Varimax rotated principal
components analysis
USA Norway
I II III I II III
Government redistribution Provide job 08 10 84 -07 85 04
Reduce income differences -02 02 86 05 85 -11
Financial cheating Income tax -03 90 04 -10 -16 81
Government benefits 08 89 08 00 08 85
Confidence in state Parliament 92 01 04 89 01 -09
Civil service 92 05 03 89 -02 -01
Eigenvalue 1.89 1.58 1.32 1.72 1.52 1.23
Factor criterion = Eigenvalue > 1. Source ISSP (1991).
TRUSTINGOVERNMENTANDWELFAREREGIMES353
Table 3. Attitudes to government redistribution and financial cheating by confidence in the state. USA and Norway. 1991. Percent state friendly
USA Norway
Confidence in state index Confidence in state index
Index Strong Moderate Weak (p) Strong Moderate Weak (p)
Government redistribution 31 35 31 0.02 82 74 74 0.01
Balance -3 -1 -11 +76 +67 +68
(n) (354) (483) (493) (617) (421) (408)
Financial cheating 84 83 84 0.81 82 78 68 0.00
Balance +78 +78 +79 +80 +75 +63
(n) (347) (479) (485) (624) (427) (412)
 (p) indicates significance level using the 2-statistic. Source: ISSP (1991).  Balance: percentage points difference between `state friendly' and
`state hostility' attitudes.
354 JONAS EDLUND
from the negative effects perspective that acceptance of financial cheating
would be more prominent in Norway than in the USA. Results do not support
these expectations. Citizens in Norway, living in a high taxing-and-spending
society, tend to oppose tax cheating and claiming state benefits illegally to
about the same extent as Americans. The slightly higher level of acceptance
of tax cheating in Norway meshes with previous research, which in addition
has also demonstrated weak and peculiar relationships between tax cheating
and overall levels of taxation (Peters 1991; Listhaug & Miller 1985).
Finally, confidence in the state does not seem to be strongly related to
attitudes concerning financial cheating, except for one distinct deviation:
Norwegian attitudes to tax evasion. The absence of significant relationships
among American citizens, contrary to the expectations by Listhaug & Miller
(1985), shows that it sometimes can be precarious to make generalisations
about some countries based on empirical evidence found in another country.
Social cleavages and attitudes to redistribution and financial cheating
There is no doubt that the organisation of the welfare state has had a significant
impact on living conditions and inequalities, but it has also been
increasingly recognised that "[t]he welfare state is ( . . . ), in its own right,
a system of stratification. It is an active force in the ordering of social relations"
(Esping-Andersen 1990: 23). From a power-resources perspective one
of the main systems of stratification is the labour market (Korpi 1983; EspingAndersen
1985). The unequal distribution of risks ­ such as sickness and
unemployment ­ and socio-economic resources, is intimately connected with
occupational positions in capitalist societies. Those with weaker positions on
the labour market are more dependent on the state for social protection. The
welfare state may therefore be seen as a major arena for class conflict, where
the strongest supporters of welfare state policies are expected to be found
among those with less salable market resources.
However, arguments have been raised to suggest that changing occupational
structures ­ which in many cases are related to welfare policies ­ in
the transition from industrial to post-industrial societies generate new social
cleavages which will replace or complement class divisions. Here, gender
emerges as one of the most important `new' axes of conflict (Orloff 1993).
Arguments are built either upon self-interest, as women are seen as more dependent
on the state both in terms of employment and in terms of benefits and
services (Hernes 1987; Borchorst & Siim 1987), and/or different experiences
of socialisation among men and women (Waerness 1987). Although theories
and studies concerning gender and the welfare state have often lacked a comparative
perspective, which feminist scholars admit (Daly 1994), the various
TRUST IN GOVERNMENT AND WELFARE REGIMES 355
contributions in Sainsbury (1994) show clearly that outcomes of social policy
differ between men and women, while the magnitude of these differences
varies between countries.
Another source of social conflict, it has been argued, is private versus
public employment. Arguments can be divided into two groups: socialisation
experiences and self interest (Hoel & Knutsen 1989; Sears & Citrin
1985: 238). In some cases, issues of gender are closely attached to sector
employment. From a comparative perspective, Esping-Andersen argues that
gender and sector conflicts will be especially salient in the social democratic
regime type. Due to the heavily gender segregated labour market and the
significant job expansion in the public sector among Scandinavian countries,
it is proposed that conflicts will arise along the highly gender-segmented
private-public sector axis. In the USA on the other hand, gender and ethnic
conflicts are supposed to be translated into class conflicts, due to decreasing
segregation effects in the labour market (Esping-Andersen 1990: chapter 9).13
Other factors that have received attention when describing contemporary
post-industrial societies are of a more general nature, such as the effects of
on-going processes of secularisation, and the `silent revolution' of slowly
emerging post-materialist value patterns. Beginning with secularisation, it is
apparent that this process has gone furthest in Scandinavia. Religion seems
to have little practical influence over the daily lives of most Scandinavians
(Riis 1994). By contrast, Greeley (1991) argues that secularisation processes
have been weak and almost non-existent in the USA over the past decades.
For most Americans religion still plays a significant role. The consequences
of secularisation have been discussed and it has been shown that religious
beliefs have a significant impact on values and attitudes concerning morality,
family, and child upbringing (Harding et al. 1986). It could thus be expected
that religious people are less tolerant to financial cheating than non-religious.
To what extent religious beliefs may determine attitudes concerning welfare
policies is less clear, but it is suggested that religious values may be viewed
as a conservative force in contemporary Western societies (Ester et al. 1994).
Conflicts concerning welfare policies could also be related to age, both in
terms of life cycle and cohort effects. In the former case, we would expect
that support for the welfare state is located furthermost among the youngest
and the oldest cohorts, due to their relatively weak market position and
dependency on state welfare. Regarding the latter case, if the generational
theory suggested by Inglehart (1990) has an impact on views concerning
the welfare state, we would expect that attitude patterns differ among age
cohorts. Due to their different formative experiences concerning material
well-being, younger people are expected to be less supportive of the welfare
state than older people. Consequently, as the welfare state has been rather
356 JONAS EDLUND
successful in its mission, questions of redistribution and inequality gradually
lose their political significance as older age cohorts are replaced by younger
age cohorts. According to Inglehart, contemporary societies are facing new
political issues, and these mainly reflect different value preferences between
materialists and postmaterialists. Traditional class-based political conflicts
tend to be forced out of the political agenda (Inglehart 1990: 248­334). The
diminishing relevance for the traditional left-right axis, and the `out-dateď
issues of economic inequality suggests that links between political choice
and welfare state issues are slowly but surely withering away.
The following variables will be used in the analysis: class; household
income; gender; age; religiosity; political choice; and sector.14
While some
of these variables do not need any further presentation, some certainly do.
While class theoretically is often assumed to be a key variable, it has
been less easy to assess the concept of class empirically and make it crossnationally
comparable. However, the works of Ganzeboom & Treiman (1994)
and Erikson (n.d.) have made it possible to translate different occupational
coding schemes into a Erikson/Goldthorpe classification of occupations
(Erikson & Goldthorpe 1992).15
The usefulness of the class schema has been
empirically validated (see works in Nieuwbeerta 1996), and permits that classifications
differ in detail. The classification used here distinguishes between
unskilled workers, skilled workers, routine non-manuals, service classes II
and I (lower and higher level controllers and administrators, respectively),
and finally self-employed.
The problems of standardisation and making income cross-nationally
comparable are, due to different currencies and coding schemes, obvious.
The USA classification contains 21 different income strata, compared to Norway's
nine. In the analysis, income will simply be divided into quartiles.
As income has been counted as household income, married and cohabiting
couples' income has been divided by two.
Religiosity is measured by an additive index consisting of three variables.
The first asks whether respondents consider themselves religious or not. The
second and third measure the frequency of, respectively, praying and taking
part in church activities. Results from factor and correlation analyses confirm
the conceptual validity among these variables. Hardly surprising, levels
of religiosity differ substantially between Norway and the USA. To create
an index which is both comparable and reliable, it is necessary to collapse
categories. The religiosity index therefore distinguishes only between three
different levels: weak, medium, and strong.16
Finally, categories for political choice differ between the USA and Norway.
Among Americans, political choice will be divided into three groups:
liberal, moderate, and conservative, using the subjective self-locating left-
TRUST IN GOVERNMENT AND WELFARE REGIMES 357
right scale. For Norway, the party classification schema is reclassified into left
(Red Electoral Alliance, Labour Party, Socialist Left Party), middle (Christian
Democratic Party, Centre Party, Liberal Party), and right (Conservative
Party, Progress Party) generally following the categorisation used by Ringdal
& Hines (1995).
Relationships between political party preference and political attitudes are
troublesome in at least two important respects. First, to determine causal relationships
between those two factors is complicated. Second, relationships
between structural factors ­ such as class, gender et cetera ­ and political
attitudes may be distorted if these particular political attitudes are to are large
extent channelled through political party preferences. The strategy chosen
here is to run separate analyses. In the first step, all independent variables,
except political choice, are entered. In the second step, variables proven to
have significant relationships with redistribution and financial cheating respectively
are reentered in a second analysis. Results from the second analysis
are displayed in model I. In the following model II, political choice is added.
Additional models including variables with non-significant relationships to
redistribution and financial cheating respectively are reported at the foot of
the tables.
In Table 4, some multivariate models are assessed to measure different
group patterns to redistribution. Although levels of support for government
redistribution differ between the nations, group differences are very similar.
Workers, women, and those with low incomes are more in favour of redistribution
(model I). In model II, in which the political preference variable is
introduced, these earlier patterns persist, and the impact of political choice
is significant, especially in Norway. It is principally among conservatives
that hostile attitudes to government redistribution occur, while differences
between left-wing and centre preferences are rather small. In model III sector
location is introduced. Among Norwegians, it is apparent that employees in
the private sector do not support redistribution to the same extent as those
in the public sector. Finally, despite that neither age nor religiosity having
any significant impact on structuring attitudes to redistribution, the explained
variance is higher for Norway, which appears largely to depend on the close
links between political choice and views on redistribution.
Examining group differences to financial cheating in Table 5, models I
and II, a somewhat different picture emerges. While levels of opposition to
financial cheating are similar in Norway and United States, group patterns differ
in some respects. In both countries, class, religiosity and political choice
show significant relationships with financial cheating. Again the impact of
class is demonstrated, but patterns are different to redistribution, as workers
and the self-employed show the most tolerant attitudes to financial cheating.
358 JONAS EDLUND
Table 4. Attitudes to government redistribution in Norway and the USA by various
structural determinants. Multiple classification analysis. Adjusted index means and beta
coefficients, 1991
USA Norway
Government redistribution index Government redistribution index
Model I Model II Model I Model II Model III
Beta (p) Beta (p) Beta (p) Beta (p) Beta (p)
Adjusted Adjusted Adjusted Adjusted Adjusted
means means means means means
Grand Mean 0.46 0.46 0.74 0.74 0.73
Class 0.18 (0.00) 0.17 (0.00) 0.17 (0.00) 0.15 (0.00) 0.17 (0.00)
Service I -0.07 -0.06 -0.11 -0.09 -0.11
Service II -0.03 -0.03 -0.01 -0.02 -0.03
Routine non-manuals -0.01 -0.01 0.00 0.00 0.00
Skilled manuals 0.05 0.07 0.05 0.05 0.05
Unskilled manuals 0.08 0.07 0.06 0.05 0.06
Self-employed -0.08 -0.07 -0.05 -0.04 -0.04
Gender 0.12 (0.00) 0.11 (0.00) 0.13 (0.00) 0.10 (0.00) 0.11 (0.00)
Male -0.04 -0.04 -0.04 -0.03 -0.03
Female 0.03 0.03 0.04 0.03 0.03
Income 0.11 (0.00) 0.12 (0.00) 0.18 (0.00) 0.15 (0.00) 0.15 (0.00)
1st quartile 0.06 0.06 0.06 0.04 0.04
2nd quartile 0.00 0.01 0.04 0.04 0.04
3rd quartile -0.05 -0.05 -0.03 -0.03 -0.02
4th quartile -0.02 -0.02 -0.07 -0.06 -0.06
Political choice ­ 0.17 (0.00) ­ 0.32 (0.00) Liberal
[Left] 0.06 0.07
Moderate [Middle] 0.02 0.05
Conservative [Right] -0.07 -0.17
[No preference] 0.00
Sector ­ ­ ­ ­ 0.14 (0.00)
Public 0.04
Private -0.05
Not in labour force 0.03
R2 7.1% 10.1% 9.8% 19.8% 11.8%
(n) (1177) (1175) (1158) (1136) (1115)
 (p) indicates significance level using the F-statistic.  Response categories for Norway
in brackets. Variables not included in models are marked with ­. Independent variables
with non-significant relationships (p > 0.05) are not included in the models above. For
both Norway and the USA, age and religiosity have non-significant relationships with
redistribution, which the following models (a­d) indicate. Independent variable(s) and
significance levels in different models. United States: Model (a) Age (0.07); model (b)
Religiosity (0.96). Norway: Model (c) Age (0.32), Income (0.00), Class (0.00); Model
(d) Religiosity (0.12), Gender (0.00). Source: ISSP (1991).
TRUST IN GOVERNMENT AND WELFARE REGIMES 359
Table 5. Attitudes to financial cheating in Norway and the USA by various structural
determinants. Multiple classification analysis. Adjusted index means and beta
coefficients, 1991
USA Norway
Financial cheating index Financial cheating index
Model I Model II Model III Model I Model II Model III
Beta (p) Beta (p) Beta (p) Beta (p) Beta (p) Beta (p)
Adjusted Adjusted Adjusted Adjusted Adjusted Adjusted
means means means means means means
Grand mean 0.78 0.78 0.78 0.80 0.80 0.80
Class 0.13 (0.00) 0.13 (0.00) 0.13 (0.00) 0.19 (0.00) 0.19 (0.00) 0.17 (0.00)
Service I 0.05 0.04 0.04 0.04 0.05 0.04
Service II 0.02 0.03 0.03 0.05 0.05 0.05
Routine non-manuals 0.00 0.01 0.01 0.00 0.00 0.00
Skilled manuals -0.01 -0.02 -0.02 -0.02 -0.02 -0.02
Unskilled manuals -0.04 -0.03 -0.04 -0.02 -0.03 -0.03
Self-employed -0.02 -0.02 -0.02 -0.06 -0.06 -0.06
Religiosity 0.10 (0.00) 0.09 (0.01) 0.09 (0.01) 0.10 (0.00) 0.10 (0.01) 0.09 (0.01)
Weak -0.05 -0.04 -0.04 -0.01 -0.01 0.00
Medium -0.01 0.00 0.00 -0.01 -0.01 -0.01
Strong 0.02 0.02 0.02 0.04 0.04 0.04
Gender 0.08 (0.01) 0.08 (0.01) 0.08 (0.01) ­ ­ 0.02 (0.52)
Male -0.02 -0.02 -0.02 0.00
Female 0.01 0.01 0.01 0.00
Income ­ ­ ­ 0.08 (0.04) 0.08 (0.05) ­
1st quartile 0.00 0.00
2nd quartile 0.00 0.00
3rd quartile 0.03 0.03
4th quartile -0.02 -0.02
Political choice ­ 0.09 (0.00) 0.09 (0.00) ­ 0.15 (0.00) 0.16 (0.00)
Liberal [Left] -0.03 -0.03 0.03 0.03
Moderate [Middle] 0.00 0.00 0.02 0.02
Conservative [Right] 0.03 0.03 -0.04 -0.04
[No preference] -0.02 -0.03
Age ­ ­ 0.05 (0.48) 0.15 (0.00) 0.14 (0.00) 0.13 (0.00)
25 years or less 0.01 -0.03 -0.03 -0.03
26­35 years 0.01 -0.02 -0.02 -0.01
36­45 years -0.01 -0.01 -0.01 -0.01
46­59 years -0.01 0.03 0.03 0.03
60 years or more 0.01 0.04 0.04 0.03
R2 3.7% 4.6% 4.9% 8.0% 10.0% 9.2%
(n) (1244) (1241) (1241) (1138) (1118) (1216)
 (p) indicates significance level using the F-statistic.  Response categories for
Norway in brackets. Variables not included in models are marked with ­. Independent
variables with non-significant relationships (p > 0.05) are not included in the
models above, except for model III. For the United States, age and income have nonsignificant
relationships with financial cheating. For Norway, gender and sector have
non-significant relationships with financial cheating. This is indicated in the following
models (a­c). Independent variable(s) and significance levels in different models.
United States: Model (a) Age (0.37); model (b) income (0.38). Norway: Model (c)
Gender (0.14), Sector (0.22), Age (0.00), Class (0.00). Data on sector are not available
for the United States. Source: ISSP (1991).
360 JONAS EDLUND
Further, religious people are less accepting than non-religious, and among
Americans, liberals are more prone to accept cheating than conservatives.
In Norway relationships are inverted as acceptance of cheating is associated
with the political right.17
In model III a comparison between the two countries using an equivalent
set of variables is made. Leaving group patterns common in both countries
aside, it may be concluded that men and women in the United States appear
to have slightly different views on cheating, while in Norway clear-cut age
cleavages exist. Younger cohorts are less like than older cohorts to view
cheating with taxes and state benefits as wrong. The data used in this article
do not give any answer to the question whether these age differences are
mainly a generational effect or a life cycle phenomenon. However, interpreting
it as a symptom of slowly changing value pattern between generations,
two things need to be considered. First, it has to be considered whether the
higher acceptance of financial cheating among the young is a part of a broader
pattern among Western countries, or if it is circumscribed to a smaller group
of countries. Data from Sweden and Denmark suggest that the age pattern
observed can be generalised at least to Scandinavia (SOU 1994: 73: chap. 13;
Gundelach & Riis 1992: 50).18
Second, it needs to be established to what extent
attitude patterns show systematic linkages to welfare state characteristics.
If the higher acceptance of financial cheating among the young turns out to
be a generational effect, and if it can be traced to be a consequence of a high
taxing and generous spending welfare state, the Norwegian welfare state may
meet financial as well as legitimacy problems in the future. Nevertheless, this
is mere speculation, further research is needed.
Conclusions
Claims have been made that institutional designs of welfare states tend to
determine public preferences and behaviour in certain ways, as well as structuring
patterns of social division. Comparing attitudes to redistribution and
financial cheating in Norway and the United States, data suggest that support
for government redistribution is considerably stronger in Norway in comparison
with the United States. Interpreting attitudinal differences from a
welfare state regime perspective, the results endorse arguments emphasising
that the design and scope of welfare policies shapes and determines its own
legitimacy. The political trust approach, which argues that the efficiency of
political decision-making institutions promote beliefs about trust in the state
and furthermore views on government responsibilities, receives less support.
Though Norwegians show higher levels of trust than Americans, relation-
TRUST IN GOVERNMENT AND WELFARE REGIMES 361
ships between confidence in government and attitudes to redistribution are
relatively weak in both countries.
For those believing in the trust argument, these results might be disappointing.
It is of course possible to question the results on methodological
grounds, as the concept of trust is difficult to measure adequately. Certainly
the indicators have their drawbacks as the questions used are very general
and may fail to distinguish between confidence in the democratic system and
other expressions of trust, i.e. political responsiveness to public demands and
moral standards of politicians (Craig et al. 1990; Niemi et al. 1991; Smith
1981). While this objection has its merits, most of the relationships which
have been shown are weak and it is doubtful whether better indicators would
result in strong associations between trust in government and attitudes to the
welfare state.19
Another problem, which is probably even more important, is that although
"mistrust of government in general results in great part from dissatisfaction
with the outcomes of ongoing public policies", reasons behind expressed mistrust
may differ as "[s]ome are dissatisfied because they want government to
do more and some because they want it to do less" (Sears & Citrin 1985: 177;
Borre 1995). It is difficult to overlook the simplicity of the trust argument,
as it does not take into account that social groups may differ in preferences
regarding government actions. Indeed, it is perhaps too naive to assume that
trust in government goes hand in hand with positive attitudes to redistribution.
I do not disagree that "what citizens believe about politics, and what they
think is possible and desirable is fundamentally shaped by what government
does for them" (Steinmo 1994: 128). However, even if the inefficiency of the
American political decision-making institutions, as it has been argued, have
contributed to the failures of American social policy making and promoted
distrust in the state, it is perhaps more plausible to suggest that support for
redistribution is rooted in experiences of the features and performance of ongoing
social policies, rather than in political decision-making processes. First,
as pointed out above, the American welfare programmes aimed at helping the
poor and reducing inequalities have not turned out to be effective. Second,
it is well-known that universal and encompassing welfare programmes (i.e.
health care and pensions) receive higher public support than targeted and
means-testing programmes (i.e. housing and social assistance). The latter
group of programmes are also more likely to feed attitudes of abuse and
cheating among the public (Coughlin 1980; Taylor-Gooby 1995; Svallfors
1991; Hadenius 1986).
While overall levels of support differ, conflicts over redistribution are similarly
structured. In both nations, the strongest supporters are to be found
among workers, those with low incomes, women, and, in Norway, public
362 JONAS EDLUND
sector employees. Arguments claiming that social conflict scenarios will vary
systematically among different welfare state regimes cannot be sustained (cf.
Svallfors 1997a). Admittedly, we are leaving the industrial society behind us,
but contrary to some expectations, attitudinal links between class, political
choice and attitudes to redistribution suggest that issues of inequality still are
of importance for citizens in Norway and the USA. In spite of the changing
occupational structures during the last decades, which have been intimately
connected with the entrance into the post-industrial society, identities and
interests are still significantly structured by class. However, appearances of
gender and sector differences point out that vertical cleavages are complemented
by horizontal divisions. Dissimilarities between men and women do
not diminish by controlling for class and income. Interpreting these results
suggests that differences between males and females are not solely grounded
in different positions in markets.
Arguments proposing that advanced welfare statism has undesirable effects
on civic morality, such as cheating on taxes and state benefits, are
generally not confirmed, in that general levels of opposition are quite similar
in both countries. however, in Norway younger cohorts are significantly more
tolerant of cheating than older cohorts. Whether these age differences indicate
generational patterns as well as consequences of welfare statism remain as
open questions.
It is clear that conflicts over financial cheating, compared to redistribution,
show more cross-country variation. The thesis that acceptance of financial
cheating is associated with government legitimacy is corroborated in Norway.
as citizens trusting government are less tolerant of tax cheating than those
with little confidence. Among Americans relationships between trust and financial
cheating are insignificant. While tolerance for cheating in the USA is
associated with the political left, the relationship is reversed in Norway. Still,
with regard to class and religiosity similar patterns occur. In both countries
workers, self-employed, and non-religious tend to view financial cheating as
less wrong than service classes and religious people.
The very similar patterns of class and gender conflicts over redistribution
in the USA and Norway were perhaps less expected when we take into account
different configurations of national institutions, occupational structures
and strength of labour movements. While the demonstrated importance of
class for political attitudes and voting in Scandinavian countries may be well
known (Svallfors 1995; Nieuwbeerta 1996; Ringdal & Hines 1995), common
explanations referring to strong centralised labour movements embodied in
class based political parties and unions fail to explain the observed pattern in
the United States (cf. Svallfors 1991: 628­630).
TRUST IN GOVERNMENT AND WELFARE REGIMES 363
The observed cleavage patterns do however suggest that determinants of
distributive conflicts should be sought among factors embodied in structures
of employment. In this respect two different interpretations of the observed
patterns seem plausible. One interpretation suggests that common forces
among Western industrialised nations are at work. Inequalities generated by
capitalistic means of production give rise to similar frames of interpretation,
world views and interests. A second interpretation fits the scenario envisaged
by Esping-Andersen as the patterns noted here may be indicators of emerging
class conflicts in the United States. In Norway on the other hand, appearances
of class, gender and sector differences could indicate that class may be on
decline and will in the future be superseded by gender and sector cleavages.
In absence of comparative data covering a larger time-span, suggestions about
which interpretation, if any, that is appropriate should be cautious.
In accounting for the attitudinal similarities as well as differences, comparative
research on attitudes should perhaps devote more attention to clarify
which societal forces within national contexts that can serve as explanations
for observed attitudinal patterns, rather than applying grand theories which
often are too general to acknowledge the dynamics of cross-country attitudinal
differences. In this perspective, the welfare state regime approach with
its emphasis on institutional designs of welfare policies, as well as structures
of employment as potential bases for social conflicts, seems fruitful.
Acknowledgements
I would like to thank Eero Carrol, Richard Coughlin, Walter Korpi, Chris
Mateiu, Gregg Olsen, Stefan Svallfors, and colleagues at the Department
of Sociology, Ume University, for valuable comments on previous drafts.
The comments made by the two anonymous referees are also appreciated.
The research has been financially supported by the Swedish Council for Social
Research (SFR) and by the Swedish Council for Research in the Social
Sciences and Humanities (HSFR).
Notes
1. From a methodological point of view, it would have been desirable to include other countries
that can be subsumed under the social democratic and the liberal welfare regime
concept, respectively. Unfortunately, Norway is the only Scandinavian country available
in the ISSP 1991 module. Among countries that can be labelled as liberal regime types,
no other country except the U.S. is available. Two other possible candidates ­ Britain and
New Zealand ­ appear to deviate from the liberal welfare regime concept in certain key
respects (see works cited in Svallfors 1997a; Ginsburg 1992).
364 JONAS EDLUND
2. Some scholars argue that Esping-Andersen largely neglects the family component and
almost exclusively focuses on the state-market relationship (Sainsbury 1994; Orloff
1993).
3. Their equation, however, can be criticised. See berg (1989) for an analysis of the
relationships between the size of the public sector and redistributive effects.
4. In the United States (1986), the share of poor working age households increased slightly
after tax and transfers (Appendix). In comparison, the percentages of poor households
lifted out of poverty were in Canada 20.1% (1987), United Kingdom 46.1% (1986), West
Germany 36.4% (1984), the Netherlands 61.8% (1987), France 51.5% (1984), Sweden
43.8% (1987) (McFate et al. 1995b: 53).
5. For a critique of the liberal values explanation see Weir (1992); Skocpol (1992); Hanson
(1994).
6. For empirical evidence claiming the American political system's inability to govern efficaciously
and its lack of capacity to administer social policy, see Steinmo & Watts (1995),
Weir (1992), and works summarised in Steinmo (1994: 131) and Rothstein (1994: 122).
Some explanations concerning the cross-time stability of American political institutions
are commented in Lipset (1991).
7. While the prevalence of individualistic and anti-statist values among the U.S. public is
confirmed (Steinmo 1994; Kluegel & Smith 1986; Sears & Citrin 1985; Taylor-Gooby
1993), perhaps Rose and Steinmo overestimate the exceptionalism of anti-trust attitudes in
the United States. In comparison with West Germany, Britain and Australia, U.S. citizens
show comparably higher levels of political trust (Hayes & Bean 1993).
8. Data is available from Zentralarchiv für Empirische Sozialforschnung, at the University
of Köln, Germany. See: http://www.za.uni-koeln.de/data/en/issp/.
9. Note that response categories differ. For reasons unknown to the author, the Norwegian
response categories deviate from the ISSP standard. However, the large cross-country
differences seem not to be an effect of different response categories (cf. ISSP 1990: 53).
10. Additional analysis excluding cases which contain missing values have been performed.
These results show only minor deviations from the results presented in this article.
11. Results from additional analysis using the non-pooled `redistribution index' and `financial
cheating index', are, by and large, similar to those presented in Table 3. All relationships
except one (political trust are significantly related to cheating in Norway), are
non-significant on the 0.05 level. Tables may be obtained from the author on request.
12. Evidence from additional `item one-by-one' analysis reveal that most of this relationship
concerns tax cheating. Similar analysis for the United States shows that trust is not
significantly related to tax cheating nor claiming benefits illegally.
13. Race is however not included in the ISSP data-file, but see Coughlin (1991) for the significant
impact of race, independent of education and income, on attitudes to welfare state
policies.
14. Sector location is only available for Norway.
15. In this paper, the algorithms provided by Svallfors (1997a) have been used.
16. The `frequency of praying' variable distinguishes between: (1) Once or twice a year
or less; (2) Several times a year to every week; (3) Several times a week or more. The
`frequency of taking part in church activities, other than attending services' variable distinguishes
between: (1) Never; (2) Less than once a year to several times a year; (3) About
once a month or more. The `subjective self-description' variable distinguishes between:
(1) Non-religious; (2) Neither religious nor non-religious; (3) Religious. These variables
are then summed into an additive index, which, in order to assure a reliable number of
TRUST IN GOVERNMENT AND WELFARE REGIMES 365
respondents in each category, classifies respondents into three levels of religiosity. Tables
concerning frequencies and correlations may be obtained from the author on request.
17. Additional `item one-by-one' analyses yield basically the same results as those presented
in Table 5. While it is shown that political choice and class are similarly related to tax
cheating and claiming benefits illegally, relationships are more pronounced when it comes
to tax cheating.
18. The financial cheating items are also present in some `covering a lot of countries' studies,
but treated differently, which make comparisons problematic (Harding et al. 1986; Ester
et al. 1994).
19. Examining Norwegian and Swedish data, Svallfors (1997b), utilising a broad set of indicators,
did not detect any significant relationships between political trust and redistribution.
366 JONAS EDLUND
Appendix: Selected socio-economic indicators for the USA and Norway
USA (rank) Norway (rank)
1. Distribution . . . a
a. factor income (Gini coefficients c:a 1985) 0.44 8/12 0.34 2/12
b. disposable income (Gini coefficients c:a 1985) 0 33 11/11 0.23 3/11
2. Poverty (c:a 1985)b 17.9 11/11 3.5 11/11
3. Transfer system effectiveness (1986)c -0.5 7/7 n.a. n.a.
4. Government outlays of GDP (1989) 36.5 14/18 52.9 4/18
5. Social benefit expenditure of GDP (1985) 12.0 16/18 28.0 3/18
6. Tax revenue of GDP (1991) 29.5 17/18 47.1 4/18
7. Left share of . . . d
a. votes (mean 1950­1990) 0 18/18 49.2 2/18
b. seats in parliament (mean 1950­1990) 0 18/18 50.3 2/18
c. seats in government (mean 1950­1990) 0 16/18h 68.0 2/18
8. Labour force participation (1992)e 77.0 3/18 76.9 4/18
9. Female labour force participation (1992) 69.0 5/18 70.9 3/18
10. Public service employment of work force (1985)f 5.3 7/9 14.7 3/9
11. De-commodification scores (1980)g 13.8 17/18 38.3 2/18
aThe Gini coefficient measures income inequality. The higher the coefficient the more
unequal income distribution. Factor income = income from capital and work. Disposable
income = factor income + transfers - taxes.
bPoverty = percentage of households below 50% of median disposable income.
cTransfer system effectiveness = percentage of households (heads aged 20­55 years) lifted
out of poverty after tax and transfers.
dPolitical left = social democratic, socialist and communist parties.
eLabour force participation = total labour force divided by total population aged 15­64
fEmployment in education, health care, and welfare services divided by total population
aged 15­64.
gThe de-commodification score is a summary measure of the degree of market
independence for an average worker.
hIn three countries left parties had no seats in government.
Sources: 1a LIS; 1b, 2, 5 Korpi & Palme (1995); 3 McFate et al. (1995b); 4 Oxley et al.
(1990); 6 OECD (1994); 7 Saunders et al. (1991); 8, 9 OECD (1995: 215); 10 Cusack et al.
(1987); 11 Esping-Andersen (1990: 52). Sources 1a, 4, 7 adopted from Svallfors (1997a),
10 adopted from Therborn (1991). Notes: Rank orders are among various countries
participating in the Luxembourg Income Study (1­3), among 18 major OECD countries
(4­9, 11) or among 9 major OECD countries (10).
References
Andrews, F., Morgan, J. & Sonquist, J. (1973). Multiple classification analysis. Ann Arbor:
University of Michigan.
TRUST IN GOVERNMENT AND WELFARE REGIMES 367
Atkinson, A.B. & Mogensen, G.V., eds. (1993). Welfare and work incentives. Oxford:
Clarendon Press.
Banfield, E.C. & Wilson, J.Q. (1963). City politics. Cambridge: Harvard University Press.
Borchorst, A. & Siim, B. (1987). Women and the advanced welfare state ­ A new kind
of patriarchal power?, in A. Showstack Sassoon (ed.), Women and the state. London:
Hutchinson.
Borre, O. (1995). Scope-of-government beliefs and political support, in O. Borre & F.
Scarbrough (eds.), The scope of government. Oxford: Oxford University Press.
Carnoy, M. (1984). The state & political theory. Princeton: Princeton University Press.
Castles, F. & Mitchell, D. (1993). Worlds of welfare and families of nations, in F. Castles
(ed.), Families of Nations. Patterns of public policy in Western democracies. Aldershot:
Dartmouth.
Coughlin, R. (1980). Ideology, public opinion & welfare policy. Berkeley: Institute of
International Studies, University of California.
Coughlin, R. (1991). The economic person in sociological context, in A. Etzioni & P.R.
Lawrence (eds.), Socioeconomics: Toward a new synthesis. New York: M.E. Sharpe.
Craig, S.C., Niemi, R.G. & Silver, G.E. (1990). Political efficacy and trust: A report on the
NES pilot study items, Political Behavior 12: 289­314.
Cusack, T., Notermans, T. & Rein, M. (1987). Political-economic aspects of public employment.
Berlin: WZB.
Daly, M. (1994). Comparing welfare states: Towards a gender friendly approach, in D.
Sainsbury (ed.), Gendering welfare states. London: Sage.
Davis, J. & Jowell, R. (1989). Measuring national differences: An introduction to the international
social survey program (ISSP), in R. Jowell, S. Witherspoon & L. Brook (eds.),
British social attitudes. Special international report. Aldershot: Gower.
Edelmann, M. (1994). The social psychology of politics, in L. Dodd & C. Jillson (eds.), The
dynamics of American politics. Boulder: Westview Press.
Erikson, R. (n.d.). Kodningar av SEI och EGP i LNU91. Stockholm: Swedish Institute for
Social Research (mimeo).
Erikson, R. & Goldthorpe, J.H. (1992). The constant flux. Oxford: Clarendon Press.
Esping-Andersen, G. (1985). Politics against markets: The social democratic road to power.
Princeton: Princeton University Press.
Esping-Andersen, G. (1990). The three worlds of welfare capitalism. Cambridge: Polity Press.
Ester, P., Halman, L. & de Moor, R., eds. (1994). The individualizing society. Tilburg: Tilburg
University Press.
Evans, P., Rueschmeyer, D. & Skocpol, T., eds. (1985). Bringing the state back in. Cambridge:
Cambridge University Press.
Fritzell, J. (1991). Icke av marknaden allena: Inkomstfördelningen i Sverige. Stockholm:
Almqvist & Wiksell International.
Ganzeboom, H.B.G. & Treiman, D.J. (1994). Internationally comparable measures of occupational
status for the 1988 international standard classification of occupations. Dept. of
Sociology, Utrecht University (mimeo).
Ginsburg, N. (1992). Divisions of welfare. London: Sage.
Gough, I. (1991). The United Kingdom, in A. Pfaller, I. Gough & G. Therborn (eds.), Can the
welfare state compete? London: Macmillan.
Greeley, A.M. (1991). American exceptionalism: The religious phenomenon, in B.E. Shafer
(ed.), Is America different? A new look at American exceptionalism. Oxford: Clarendon
Press.
Gundelach, P. & Riis, O. (1992). Danskernes vrdier. Copenhagen: Forlaget Sociologi.
368 JONAS EDLUND
Hadenius, A. (1986). A crisis of the welfare state? Stockholm: Almqvist & Wiksell Interna-
tional.
Hall, P. & Taylor, R. (1996). Political science and the three new institutionalisms, Political
Studies 44: 936­957.
Hanson, R.L. (1994). Liberalism and the course of American social welfare policy, in L. Dodd
& C. Jillson (eds.), The dynamics of American politics. Boulder: Westview Press.
Harding, S., Phillips, D. & Fogarty, M. (1986). Contrasting values in Western Europe.
Hampshire: Macmillan.
Hartz, L. (1955). The liberal tradition in America. New York: Harcourt, Brace.
Hayes, B.C. & Bean, C.S. (1993). Political efficacy: A comparative study of the United States,
West Germany, Great Britain and Australia, European Journal of Political Research 23:
261­280.
Hernes, H. (1987). Women and the welfare state: The transition from private to public
dependence, in A. Showstack Sassoon (ed.), Women and the state. London: Hutchinson.
Hoel, M. & Knutsen, O. (1989). Social class, gender and sector employment as political
cleavages in Scandinavia, Acta Sociologica 32: 181­201.
Inglehart, R. (1990). Culture shift in advanced industrial society. Princeton, NJ: Princeton
University Press.
ISSP. (1990). Role of government II. Codebook and datafile compiled at the ZA, University of
Köln.
ISSP. (1991). Religion. Codebook and datafile compiled at the ZA, University of Köln.
Jillson, C. (1994). Patterns and periodicity in American national politics, in L. Dodd & C.
Jillson (eds.), The dynamics of American politics. Boulder: Westview Press.
Kangas, O. (1991). The politics of social rights. Studies on the dimensions of sickness
insurance in OECD countries. Stockholm: Almqvist & Wiksell.
King, A. (1973). Ideas, institutions and the policies of governments: A comparative analysis,
Part III, British Journal of Political Science 3: 409.
Kluegel, J.R. & Smith, E.R. (1986). Beliefs about inequality. New York: Aldine de Gruyter.
Korpi, W. (1980). Social policy and distributional conflict in the capitalist democracies. A
preliminary comparative framework, West European Politics 3: 296­316.
Korpi, W. (1983). The democratic class struggle. London: Routledge and Kegan Paul.
Korpi, W. (1996). Social citizenship, class, and gender. Stockholm: Swedish Institute for
Social Research (mimeo).
Korpi, W. & Palme, J. (1995). The paradox of redistribution and the strategy of equality.
Stockholm: Swedish Institute for Social Research (mimeo).
Küchler, M. (1987). The utility of surveys for cross-national research, Social Science Research
16: 229­244.
Laurin, U. (1986). P heder och samvete. Stockholm: Norstedts.
Levi, M. (1990). A logic of institutional change, in K. Schweers Cook & M. Levi (eds.), The
limits of rationality. Chicago: The University of Chicago Press.
Lewis, A. (1982). The psychology of taxation. Oxford: Martin Robertson.
Lindbeck, A. (1988). Consequences of the advanced welfare state, The World Economy 11:
19­38.
Lipset, S.M. (1991). American exceptionalism reaffirmed, in B.E. Shafer (ed.), Is America
different? A new look at American exceptionalism. Oxford: Clarendon Press.
Listhaug, O. & Miller, A.H. (1985). Public support for tax evasion: Self-interest or symbolic
politics?, European Journal of Political Research 13: 265­282.
McFate, K., Lawson, R. & Wilson, W.J., eds. (1995a). Poverty, inequality and the future of
social policy. Western states in the new world order. New York: Russell Sage Foundation.
TRUST IN GOVERNMENT AND WELFARE REGIMES 369
McFate, K., Smeeding, T. & Rainwater, L. (1995b). Markets and states: Poverty trends and
transfer system effectiveness in the 1980s, in K. McFate, R. Lawson & W.J. Wilson (eds.),
Poverty, inequality and the future of social policy. Western states in the new world order.
New York: Russell Sage Foundation.
Niemi, R.G, Craig, S.C. & Mattei, F. (1991). Measuring internal political efficacy in the 1988
national election study, American Political Science Review 85: 1407­1413.
Nieuwbeerta, P. (1996). The democratic class struggle in postwar societies, Acta Sociologica
39: 345­383.
OECD. (1994). Revenue statistics. Paris: OECD.
OECD. (1995). Employment outlook. July 1995. Paris: OECD.
Olsen, G. (1994). Locating the Canadian welfare state: Family policy and health care in
Canada, Sweden, and the United States, Canadian Journal of Sociology 19: 1­20.
Orloff, A.S. (1993). Gender and the social rights of citizenship: The comparative analysis of
gender relations and welfare states, American Sociological Review 58: 303­328.
Oxley, H., Maher, M., Martin, J.P. & Nicoletti, G. (1990). The public sector: Issues for the
1990's. Working Paper 90, OECD Department of Economics and Statistics. Paris: OECD.
Palme, J. (1990). Pension rights in welfare capitalism. Stockholm: Almqvist & Wiksell.
Peters, G. (1991). The politics of taxation. Oxford: Blackwell.
Pfaller, A., Gough, I. & Therborn, G., eds. (1991a). Can the welfare state compete? London:
Macmillan.
Pfaller, A., Gough, I. & Therborn, G. (1991b). The issue, in A. Pfaller, I. Gough & G. Therborn
(eds.), Can the welfare state compete? London: Macmillan.
Pierson, P. (1993). When effect becomes cause: Policy feedback and political change, World
Politics 45: 595­628.
Pierson, P. (1994). Dismantling the welfare state. Cambridge: Cambridge University Press.
Putnam, R.D. (1993). Making democracy work. Princeton: Princeton University Press.
Riis, O. (1994). Patterns of secularization in Scandinavia, in T. Petterson & O. Riis (eds.),
Scandinavian values. Uppsala: Acta Universitatis Upsaliensis.
Ringdal, K. & Hines, K. (1995). Patterns in class voting in Norway 1957­1989: Decline or
`trendless fluctuations'?, Acta Sociologica 38: 33­51.
Rose, R. (1991). Is American public policy exceptional? in B.E. Shafer (ed.), Is America
different? A new look at American exceptionalism. Oxford: Clarendon Press.
Rothstein, B. (1994). Vad bör staten göra? Stockholm: SNS.
Rothstein, B. (1996). Political institutions: An overview, in R.E. Goodin & H.D. Klingemann
(eds.), A new handbook of political science. Oxford: Oxford University Press.
Sainsbury, D., ed. (1994). Gendering welfare states. London: Sage.
Saunders, P., Stott, H. & Hobbes, G. (1991). Income inequality in Australia and New Zealand:
International comparisons and recent trends, Review of Income and Wealth 37: 63­79.
Scheuch, E.K. (1989). Theoretical implications of comparative survey research: Why the
wheel of cross-cultural methodology keeps on being reinvented, International Sociology
4: 147­167.
Sears, D.O. & Citrin, J. (1985). Tax revolt. Something for nothing in California. Enlarged
edition. Cambridge: Harvard University Press.
Siaroff, A. (1994). Work, welfare and gender equality: A new typology, in D. Sainsbury (ed.),
Gendering welfare states. London: Sage.
Skocpol, T. (1992). Protecting soldiers and mothers. Cambridge: The Belknap Press of
Harvard University Press.
370 JONAS EDLUND
Smith, T.W. (1981). Can we have confidence in confidence? Revisited, in D.F. Johnston
(ed.), Measurement of subjective phenomena. Washington, DC: U.S. Government Printing
Office.
Smith, T.W. (1987). That which we call welfare by any other name would smell sweeter,
Public Opinion Quarterly 51: 75­83.
SOU. (1994). Ungdomars välfärd och värderingar. SOU 1994: 73. Stockholm: Fritzes.
Steinmo, S. (1993). Taxation and democracy. New Haven: Yale University Press.
Steinmo, S. (1994). Rethinking American exceptionalism: culture or institutions?, in L. Dodd
& C. Jillson (eds.), The dynamics of American politics. Boulder: Westview Press.
Steinmo, S. & Watts, J. (1995). Iťs the institutions, stupid! Why comprehensive national health
insurance always fails in America, Journal of Health Politics 20: 329­372.
Stephens, J.D. (1994). Welfare state and employment regimes, Acta Sociologica 37: 207­211.
Svallfors, S. (1991). The politics of welfare policy in Sweden: Structural determinants and
attitudinal cleavages, British Journal of Sociology 42: 609­634.
Svallfors, S. (1995). The end of class politics? Structural cleavages and attitudes to Swedish
welfare policies, Acta Sociologica 38: 53­74.
Svallfors, S. (1996). National differences in national identities? An introduction to the
international social survey programme, New Community 22: 127­134.
Svallfors, S. (1997a). Worlds of welfare and attitudes to redistribution: A comparison of eight
western nations, European Sociological Review 13: 283­304.
Svallfors, S. (1997b). Political trust and attitudes towards redistribution: A comparison of
Sweden and Norway. Dept. of Sociology, Ume University (mimeo).
Taylor-Gooby, P. (1993). What citizens want from the state, in R. Jowell, L. Brook & L. Dowds
(eds.), International social attitudes. The 10th BSA Report. Aldershot: Gower.
Taylor-Gooby, P. (1995). Who wants the welfare state? Support for state welfare in European
countries, in S. Svallfors (ed.), In the eye of the beholder. Ume: Impello.
Thelen, K. & S. Steinmo. (1992). Historical institutionalism in comparative politics, in S.
Steinmo, K. Thelen & F. Longstreth (eds.), Structuring politics. Historical institutionalism
in comparative analysis. Cambridge: Cambridge University Press.
Therborn, G. (1991). Sweden, in A. Pfaller, I. Gough & G. Therborn (eds.), Can the welfare
state compete? London: Macmillan.
Waerness, K. (1987). On the rationality of caring, in A. Showstack Sassoon (ed.), Women and
the state. London: Hutchinson.
Wahlund, R. (1991). Skatter och ekonomiska beteenden. Stockholm: EFI.
Weir, M. (1992). Politics and jobs. Princeton: Princeton University Press.
Weir, M. & Skocpol, T. (1985). State structures and possibilities for `Keynesian' responses to
the great depression in Sweden, Britain and the United States, in P. Evans, D. Rueschmeyer
& T. Skocpol (eds.), Bringing the state back in. Cambridge: Cambridge University Press.
berg, R. (1989). Distributive mechanisms of the welfare state ­ A formal analysis and an
empirical application, European Sociological Review 5: 167­182.
Address for correspondence: Dr. Jonas Edlund, Department of Sociology, Ume University,
SE-901 87 Ume, Sweden
Phone: +46 90 786 7822; Fax: +46 90 7866694; E-mail: jonas.edlund@soc.umu.se