Resumé

In most western countries, public services have been the target of cut-backs during the past few years. It is often argued that support for a strong welfare state has declined. In this article, attitudes towards public services in Finland are examined. Factors affecting the willingness to accept cuts were studied using a structural equation model and correspondence analysis. The analysis revealed that piecemeal deterioration in public services can form a vicious circle. The lower the quality of public services, the more citizens are willing to accept cuts in them. The results also lend some support to the argument that the new middle classes are most willing to see market-based options implemented in service production.

Introduction

The present-day transition of the traditional welfare state can be seen not only in the Anglo-American world and in continental Europe, but also in the Nordic countries, where the state has traditionally been a strong operating force in society (Lane et al. 1993). Demands for changes in the existing structure of welfare services are voiced by the political left as well as the political right. The arguments for change vary from left to right, but they all amount to the same thing: a reduction in the obligations of government. To the extent that such shifts in public responsibility are supported by the citizens and carried out democratically, objections to them would appear to be contrary to the will of the majority and might unnecessarily delay the implementation of improvements in social policy.

The aim of this article is to examine the opinions which citizens have about the existing public service structure in Finland and their willingness to accept changes in it. This leads to several questions: why are people in favour of cutting public services, which social groups are in favour of cutting them, how should these cuts be made and what will follow if public services are provided in other ways?

In Finland welfare and other public services are heavily subsidized by

the state, but it is the municipalities that carry the main responsibility for organizing and providing them. In 1992 it was estimated that the municipalities had accumulated a combined budget deficit of more than two billion US dollars as a result of the current recession. This has created considerable pressure to reduce public expenditure.

As a consequence of several administrative and legislative reforms during the past decade, the Finnish municipalities have been given more authority over the production of services. On the one hand, a greater plurality of services is the anticipated consequence of this decentralization, but on the other hand it is feared that this very development will widen the gap between different parts of the country and thereby generate new forms of inequality.

Crisis in Public Welfare Services?

Traditionally, there are two explanations for the crisis in the welfare state. The first, which can be called the perspective of social integration, emphasizes citizens' alienation from the public sector. This tendency is manifested in their "illogical" demands on the political system. On the one hand, citizens demand more and better public services, but on the other hand they are not willing to pay more taxes. The second explanation may be referred to as the perspective of internal discrepancies in the welfare state. This perspective emphasizes the underlying contradictions, dysfunctions and suboptimality problems between the welfare state and the market economy. These two perspectives are not independent of each other. In fact they should be seen as two different sides of the same coin (Donati 1987, 27-28).

Nowadays high unemployment in Finland (and also in many other countries) poses the greatest challenge to the legitimacy of the welfare state. Just as new production techniques are making it possible to increase productivity, the only way to reduce the present high level of unemployment would be by increasing consumption. The problem is that a high unemployment rate reduces national consumption, which in turn causes losses to the private sector and increases the financial burden on the public sector. This can lead to a situation in which the decreasing number of people still working must finance the whole public sector. This development is, of course, manifested in growing tax rates. Thus, unemployment causes both an increasing public sector deficit and an increasing tax burden among those who are still employed. In the long run, this situation can be remedied only by reducing public expenditures. This kind of development can clearly be seen in Finland and also in Sweden.

From the perspective of social integration, the alienation of the citizens'

from the public sector is manifested in the decreasing political support for the present form of the welfare state. According to Offe (1987), this reduction in support can be explained by the increasing differences in employment opportunities among the wage earners, by growing individualistic political attitudes, by decreasing acceptance of bureaucratic and professional intervention and by the quantitative growth of the middle class. Offe claims that these structural changes in society are leading to a new situation in which the breaking up of the traditional welfare state can be accomplished through democratic processes.

Disparity among wage earners does not mean simply that workers are divided into those who are successful in the labour markets and those who have to take any low-paid work or face unemployment. The growing disparities among wage earners depend on numerous factors such as gender, skill level, region and industrial sector. In this kind of situation, it is very difficult to create any strong political alliances to defend existing welfare structures. Public support for the welfare state tends to prosper in times of economic growth, and tend to decrease under zero-sum conditions. Thus, the economic crisis in the welfare state might create individualistic political attitudes and may eventually turn into a political crisis.

The means by which the goals of the welfare state have traditionally been achieved may also be in crisis. Bureaucratic and professional interventions are frequently called into question. They are not seen as rational means for the implementation of the public good but rather as increasingly ineffective or even counterproductive attempts to respond to the needs of citizens.

A particularly important factor in understanding the emerging antiwelfare-state reorientation is the growth of the middle class, particularly the "new" middle class. The greater the income of this group of people, the more they begin to look for private arrangements instead of traditional welfare-state services. Moreover, the less one uses public services, the more one is dissatisfied with one's tax burden. This kind of development might lead to "tax revolts" and may increase support for political parties that call for tax and budget cuts.

Nevertheless, the decrease in support for the welfare state is by no means a simple process. It is still too early to speak of a crisis. After all, it makes sense to refer to it as a crisis only if the dissatisfaction is channelled into action against the existing public welfare-state structures in parliamentary or non-parliamentary arenas.

Changes in the Provision of Welfare Services

It is possible to distinguish three sectors, which provide services in a modern
society. These are the public, market and voluntary non-profit sectors (Hill

& Bramley 1986, 105-106). All of them mobilize citizens and acquire and direct resources. Each one has its own style of producing services. Behind the demands for change are claims that the only way to survive the present financial crisis is to cut public expenditures. Actual savings can take place not only by modifying the balance between the state, market and nonprofit sectors, but also by reorganizing the public-sector welfare provision.

Public, market and voluntary non-profit sectors are three arenas, where services are provided. In these arenas different principles of providing the services may be employed. The most important principles are the existence of the pricing system and the competition between different providers of the service. In the public sector, most services are provided free of charge by public organizations which usually do not have to face any real competition. It is, however, possible to collect payments and/or provide competition even for public services. When privatization is used, two possibilities emerge. Services could be quasi-privatized to the voluntary non-profit sector, where services are mainly free of charge or privatized to markets, where monetary transactions are used.

Ordinary citizens do not think about the providing of services in terms of abstract concepts (Rose 1989, 132); rather, they are more interested in things like the price, quality or the ease of receiving the service. Since these attributes of services usually vary according to the provider of the service, it is reasonable to assume that even ordinary citizens have preferences as to who provides the service and as to how it is produced. With regard to these dimensions, four main alternatives in providing public services can be formulated: (1) internal reform of the public sector, (2) public markets alternative, (3) voluntary alternative, and (4) market alternative.

Cuts in the public sector that do not involve the use of market mechanisms (i.e. pricing, competition) could be called the alternative of internal reform. This non-market orientation represents the traditional response strategy in times of economic recession to make the public sector work more effectively. In practice, this often results in using across-the-board cuts affecting all bureaux equally.

Introducing the alternative of public markets means employing markettype monetary solutions within the public sector. The current discussion in Finland seems to emphasize the importance of increasing this sort of arrangement, which, if implemented, could loosen the control of political decision-making on the way in which the service is provided (Dunleavy 1986). Increasing the role of the market principles in providing welfare services would certainly create more vigorous competition between the providers of those services. Apart from more satisfactory financial results in the short term, competition could cause long-term suboptimal solutions, especially under circumstances in which competition takes place within the public sector. Rival organizations could block information channels and

stop assisting one another, which could eventually increase the overall cost
of public activity.

Privatization of public welfare services can be accomplished in one of two ways, by transferring services either to the voluntary non-profit sector or to the market sector. When limited use of the market mechanism is combined with cuts in public services, the voluntary alternative is adopted. In practice, this necessitates activating voluntary associations, popular movements, cooperatives and self-help groups. It also puts a heavier burden on households, e.g. by encouraging child care and the nursing of the elderly at home.

Last but not least, cuts in the public sector combined with an extensive use of the market mechanism constitute the market alternative. Private hospitals and schools do exist, but it is not clear how much can we depend on the privatization of services without endangering the values of social equity and responsibility.

The three different kinds of service-producing sectors should be considered as ideal types. In practice, the boundaries between the systems are not that stringent. On the contrary, different models of producing services could easily be mixed. For both practical and analytic purposes, the production and provision of services should be separated from one another. Thus, the public sector could, as a provider, buy all of its health-care services from private enterprises and still guarantee the planned level of health-care provision through other measures, such as vouchers or free services to those of limited means.

Research Design and Empirical Data

In the theoretical section of this article, modes of producing services were discussed in the context of the dynamics of the welfare state. It was assumed that current trends in modern societies bring about changes in their class structure. One such fundamental change is "embourgeoisement", the rise in the number of employees who work on non-manual tasks. Modern societies have experienced an "elevation effect" which has augmented material well-being across the social classes. These developments result in pressure to change service production.

Current services are "mass-produced" and provided universally. This approach does not necessarily serve the interests of the enlarged middle classes, which favour individual services for which they have the means to pay. Thus, as far as the middle classes are concerned social policy should include more income-related services, more free choice in choosing the service and more high-level individually "tailored" services (Cochrane 1991, 284-288).

The empirical section of this article examines attitudes in favour of cutting public-sector services. It is presumed that general attitudinal market orientation, discontent with current public services, opinions concerning the financial situation of the public sector and views about the future of the respondents' own financial situation are linked with the willingness to accept cuts in public services. It is hypothesized that the middle classes are disappointed with the current mode of producing public services and are therefore eager to cut them more vigorously than other social groups. In the following empirical analysis two socio-economic groups represent the middle classes: higher officials and entrepreneurs.

The empirical data used in this analysis were quite substantial. The questionnaire was posted to 11,000 randomly chosen respondents in the summer of 1992. It was available both in Finnish and in Swedish so that the Swedish-speaking minority would be correctly represented in the sample. Eventually, 5,302 completed questionnaires were returned. The percentage of returned questionnaires (48 percent) was acceptable, considering that the questionnaire consisted of 232 single items, which made it rather tedious and time-consuming to complete.

The fit between the sample and the population was assessed with respect to age, gender, marital status and socio-economic status. Women were slightly overrepresented in the sample (55.2 percent) in relation to the percentage of women in the population (52.2 percent). Minor overrepresentation of those not working (i.e. students, housewives and the unemployed) was also found in the sample. However, the overall fit of the sample and the population was considered to be acceptable.

Empirical Analysis

A structural equation model which is a combination of confirmatory factor analysis and regression analysis (Loehlin 1987, 76-129) was used in the initial stages of the analysis. The fundamental purpose of structural equation models is to examine causal relationships between latent variables (or factors), or here, more specifically, to estimate the quantity of causal links between respondents' perceptions and attitudes in forms of unstandardized or standardized regression coefficients. The EQS-program was used in the actual model building and parameter estimation.

The choice of variables for the model was based mainly on a study of the perceptions and attitudes towards municipal services provided in Sweden (Birgersson 1975). The principal idea of the model is that two kinds of factors can be expected to influence people's attitudes towards public services: factors relating to the individual and factors relating to the municipality. There are two components relating to the individual in the model:

the respondent's general acceptance of market-based solutions as a good alternative to traditional public service provision (market sympathy) and his/her financial situation. Market sympathy was measured using Likertscale questions about general possibilities of replacing public-sector services with services produced in markets. The financial situation was measured using questions about the respondent's expectations about his/her personal economic circumstances in the near future. There are also two factors relating to the respondent's own municipality. The first one is his/her satisfaction with existing public services. This latent variable consists of several questions in which respondents were asked to evaluate different types of public service. The second factor is the respondent's perception of the financial situation of his/her own municipality. This latent variable consists of two related questions.

The final independent variable in the model is the respondent's willingness to accept cuts in public services. This is a latent variable which was measured with the help of numerous concrete questions. Respondents were given a list of different public services and asked to mark the ones they were willing to see privatized. The model is based on the basic hypothesis that the actual desire for privatization is related to the more general attitudinal willingness to accept market mechanisms in the society (market sympathy). These two latent variables are affected by the respondent's satisfaction with existing public services, his/her expectations of the development of his/her own financial situation, and perceptions about the economic situation of his/her municipality.¹ It is also possible that these three factors may directly affect the willingness to accept the privatization of public services.

The results of the estimates provided by the model are shown in Fig. 2. The arrows indicate the direction of causal relationships between the latent variables, and the numbers next to the arrows show the strength of the association. These numbers are standardized regression coefficients (beta weights). The curved arrows indicate the correlation relationships between variables. The asterisks attached to the figures indicate the results of the ttests of the parameter estimates (** p < 0.01).

The model explains 26 percent of the variance in willingness to accept cuts in public services, but only 7 percent of the variance in market sympathy. This means that market sympathy can be considered as a fairly stable attitude, independent of the exogenous variables in the model. As expected, the strongest causal link is between market sympathy and willingness to accept cuts. Of the exogenous variables, the one with the most explanatory power is dissatisfaction with existing public services. The perceived unsatisfactory financial condition of the municipality also correlated positively with market sympathy, but not with willingness to accept cuts in public services. This is a very interesting finding because the

difficult financial situation in a municipality is the most commonly stated justification for cuts in services. However, the model as a whole clearly shows that only half of the attitudinal market sympathy seems to transfer to concrete willingness to accept cuts. It is also interesting to note that when respondents are afraid that their own economic situation will worsen in the future, they want fewer changes in the system of services provided at public expense.

The most important conclusion to be drawn from the model is that the level of satisfaction with existing public services is a major variable explaining the willingness to accept cuts in them. In the current economic crisis, Finnish local governments are being forced to reduce their expenses. If these cuts are allocated "fairly", the deepest ones should be in services that are used primarily by individuals who do not belong to the poorest group of citizens. In practice, this would mean services that are, on the whole, used more by well-off people (e.g. child care, cultural services). However, if these kinds of cuts were made, the people who contribute most to tax revenues would probably become more and more dissatisfied with the existing public services. At worst, this could lead to an increasing

willingness among well-off people to accept cuts, which would in turn increase dissatisfaction with the existing services. This kind of development could be called the vicious circle of cutback policies: the more cuts there are, the more citizens want cuts.

It is of course possible to prevent the formation of this vicious circle by making cuts which do not affect all voter groups, but instead only marginal groups of voters who do not have prominent political position, such as disabled people or people using social assistance. There is already some proof that cuts are being partly made in this way in Finland (Uusikylä 1993). This solution cannot be justified if the ideal of the Nordic welfare state in Finland is to be maintained. It does, however, put the political decision-makers in a very difficult position: they must decide how to make cuts so that they have as moderate an effect as possible on the legitimacy of the public sector and are also morally justifiable.

Market Orientation and Willingness to Accept Cuts in Public Services

Market sympathy was measured using Likert-scale questions, and the willingness to accept cuts in public services was tapped using a battery of items in which respondents could choose an appropriate producer for a specific service. A choice had to be made between public and private sources of the service. In the previously presented model of cutting public services it is demonstrated that these dimensions do correlate positively.

Without stretching the interpretation of the empirical data too far, it can be suggested that market sympathy describes an abstract orientation towards the provision of services on the basis of market mechanisms, whereas a wish to cut concretely stated public services depicts a more profound attachment to non-public alternatives. This interpretation is supported by the fact that not all market sympathy was channelled into willingness to accept cuts in public services.

Table 1 illustrates the results obtained from the analysis of variance of the factor score means of the two endogenous latent variables presented earlier in the EQS model (market sympathy and willingness to accept cuts in public services). According to these results, higher officials were most sympathetic to markets (0.30) and most willing to accept cuts in public services (0.33). The oldest age groups, those not interested in municipal politics and pensioners, were the only groups of less-than-average means which exhibited both less market sympathy and less willingness to accept cuts in public services. They could be classified as the most probable groups favouring an internal reform of the public sector. It was suggested earlier that individually produced services suit well-off citizens best. Thus it is

quite logical that entrepreneurs and higher officials emphasize the market
alternative.

According to the model, blue-collar workers and the unemployed seem to be generally in favour of the public markets alternative. These groups may well accept the principles of the marketplace, but they are not that ready to abandon the public sector as an arena for providing the services. At first, this seems somewhat contradictory, but it might originate from the fact that market sympathy does not vary between the upper and lower classes. This distinction has more to do with the position of the public sector in society (i.e. in willingness to accept cuts in public services). Respondents of a lower social status are less likely to accept the weakening of the public sector, but they seem to be more willing to accept the strengthening of market orientation within it.

When trying to mobilize voluntary organizations to supply public services, one should note that the main potential users of volunteered services, namely pensioners and the elderly, are inclined towards internal reform of the public sector. What is more, the possible providers of volunteered services, students, housewives and the unemployed, are inclined towards the market alternative. The strongest tendency towards the voluntary alternative is found among farmers, which suggests that this alternative is considered as an option mainly among those who live in the

countryside, where the sense of community and traditions of neigbourhood
assistance might still exist.

The current system of services provided at public expense is supported most strongly by the elderly and those who are not interested in politics. The growing proportion of old people and the increasing disinterest in politics seem to serve as a base for public sector support. However, it is a bleak outlook for the future of a large, publicly maintained welfare state if those who pay for its functioning are most willing to see cuts in it.

Patterns of Support for the Various Arenas of Welfare Support

This section refers to the trichotomy presented earlier (public-markets - voluntary sector) in discussing what kind of citizens' support, which service production principle and for what kind of service. This will be done using correspondence analysis, which is a way to present and analyse crosstabulations visually. With correspondence analysis it is possible to examine relations between both row and column variables. The closer two row variables are in the correspondence plot, the more their distributions are alike on the column variable. On the other hand, the closer a row variable is to a column variable, the more observations it has in this column (proportionally) (Weller & Romney 1990).³

Figure 2 is a correspondence plot which displays the kind of service
providers respondents prefer with respect to various groups of services.4 If

a service is located near a certain corner of the triangle, respondents are,
on average, more willing to let this service be provided by that producer.

The figure shows clearly that the public sector is most preferred in the area of education and health service production, and almost as popular in care for the elderly. Child care is located between the public sector and the voluntary sector, which means that respondents are somewhat willing to let voluntary organizations and individuals provide these services. Cultural and leisure services differ most strikingly from other areas of service production. Respondents are undoubtedly willing to let this area be governed by market principles or to be provided by voluntary organizations. All in all, it seems that citizens are able to rank existing public services, and the ranking appears quite unanimous. There are certain basic services, such as schools and health services, which are considered the responsibility of the public sector. Cuts should start with services of secondary importance. The problem, of course, is that the basic services constitute an overwhelming part of the public budget in Finland.

Figure 3 shows preferences for different service-providing arenas among respondents from various socio-economic groups. The popularity of the public sector is highest among workers, housewives and the unemployed. The proportion of students and higher officials increases along the lower side of the triangle towards the market corner. Entrepreneurs are located

the furthest away from the public sector corner, indicating that this group is the most sceptical towards public service production. Farmers prefer voluntary service production. The location of pensioners outside the triangle at the extreme upper left may be explained by their significant distrust of market mechanisms.

It is interesting to note that all groups of wage earners are located in Fig. 3 on the lower side of the triangle (that is, on the axis between markets and the public sector). The ranking among these groups is also evident: the nearer the group is to the markets corner, the better its socio-economic status. The only group that does not fit this pattern is the unemployed, which is located very near to the public sector corner. This pattern supports the hypothesis that citizens preferring privatization to public service production are more likely to be found among groups of people who have good positions in the labour markets and high social status.

Conclusions

In principle, the concept of "welfare" does not automatically involve the public sector, although these two issues are frequently connected. Welfare can also be produced by other bodies, within non-profit organizations or in the market sector. The basic idea behind this article has been to examine citizens' attitudes toward these options. The present-day welfare state, which is mainly based on universal and standardized benefits, was developed to correct problems faced by industrializing societies. This is why it has been referred to as a "Fordist welfare state" (Jessop 1988, 7-9). It has been argued that this model has outlived its usefulness, because it is no longer in harmony with economic growth. Furthermore, it has been claimed that the structure of the welfare state must change along with the changes in the structure of society.

The central element undermining the legitimacy of the existing welfare state is the growth of the (new) middle classes. It has been argued that their lifestyle includes a certain stress on free choice and individuality, which can be seen in demands for more plurality in service production. The results of this study have given some support to this argument. The provision of services at private expense was supported most by higher officials and entrepreneurs. However, it has to be borne in mind that, in general, no great dissatisfaction with publicly financed welfare services was revealed.

In the long run, there seem to be two problems with the above-mentioned development. The first one is connected with the legitimacy of the public sector. If market options become widely used by the wealthy section of the population, its members would undoubtedly be less willing to finance public

services which they do not use themselves. This could start a vicious circle. During the present economic recession, many public services are facing across-the-board cuts. The more dissatisfaction there is with the public services, the greater the desire to reduce them. At worst this could lead to even greater dissatisfaction which, in turn, could be channelled into gradually increasing demands to cut public services even more. This piecemeal deterioration could, in the long run, create more problems than well planned structural changes made by politicians.

Another problem related to the pluralistic approach to providing services is that this development may create and intensify existing social inequality. The increase in private medical services may lead to a deterioration in corresponding public services, which would in turn broaden the gap between the wealthy and the non-wealthy. These new cleavages do not necessarily follow traditional lines among the employed, but tend to differentiate the unemployed and the employed. Demographic changes caused by the growing proportion of non-working elderly persons may reinforce this development in the future.

---

NOTES

1. In principle, it is possible that market sympathy affects the level of dissatisfaction with existing public services. However, the data provided no support for this causal link.

2. The goodness-of-fit indices of the estimated model are: Bentler-Bonnet normed fit index: 0.985; Bentler-Bonnet non-normed fit index: 0.977; Comparative fit index: 0.987; chi:(2o) = 133. All indices, except the chi-square test, indicate acceptable fit. The problem with the chi-square test is that the result is dependent on the sample size, which in this case is very large. Because of this dependency, one should not rely only on the chi-square test (Bollen-Long 1993, 8). The McDonald centrality index (Gerbing & Anderson 1993, 57 and 63), which is not sample size-dependent, indicates a good fit (0.988). Hence the conclusion that the fit of the model is acceptable.

3. The original cross-tabulation must be tested using the chi-square test. If the test does not reject the independence model, it would not make sense to apply correspondence analysis (Weller & Romney 1990, 57).

4. The figures around the plot indicate only the coordinates on which the plot is based.

ACKNOWLEDGEMENTS

The authors express their thanks to Stephan Lengauer (University of Vienna) and two
anonymous referees for helpful comments on this article.

REFERENCES

Birgersson, B. O. 1975. "Kommunen som service producent", Stockholm Studies in Politics
7. Stockholm: Stockholms Universitet.

Bollen, K. A. & Long, J. 5.1993. "Introduction", in Bollen, K. A. & Long, J. S., eds., Testing
Structural Equation Models. Newbury Park: Sage Publications.

Cochrane, A. 1991. "The Changing State of Local Government: Restructuring for the 19905",
Public Administration 69, 281-302.

Donati, P. 1987. "Traditional Political Theories and New Social Options: Replies to the Crisis
of the Welfare State", in Evers, A., Nowotny, H. & Wintersberger, H., eds., The Changing
Face of Welfare. Aldershot: Gover.

Dunleavy, P. 1986. "Explaining the Privatization Boom: Public Choice Versus Radical
Approaches", Public Administration 64, 13-34.

Gerbing, D. W. & Anderson, J. C. 1993. "Monte Carlo Evaluations of Goodness-Of-Fit
Indices for Structural Equation Models", in Bollen, K. A. & Long, J. S., eds.. Testing
Structural Equation Models. Newbury Park: Sage publications.

Hardin, R. 1988. Collective Action. Baltimore: The Johns Hopkins University Press.

Jessop, B. 1988. "Conservative Regimes and the Transition to Post-Fordism: the Cases of
Britain and West German", Essex Papers in Politics and Government 47. Department of
Government, University of Essex, Wivenhoe Park.

Lane, J.-E., Martikainen, T., Svensson, P., Vogt, G. & Valen, H. 1993. "Scandinavian
Exceptionalism Reconsidered", Journal of Theoretical Politics 5, 195-230.

Loehlin, J. C. 1987. Latent Variable Models. An Introduction to Factor, Path, and Structural
Analysis. Hillsdale: Lawrence Erlbaum Associates Publishers.

Offe, C. 1987. "Democracy against the Welfare State? Structural Foundations of Neoconservative
Political Opportunities", Political Theory 15, 501-537.

Rose, R. 1989. Ordinary People in Public Policy: A Behavioural Analysis. Trowbridge: Sage.
Trowbridge.

Uusikylä, P. 1993. "The New Politics of Retrenchmen". Paper presented at the European
Consortium for Political Research (ECPR) Joint Session in Leiden 2-8 April.

Weller, S. C. & Romney, A. K. 1990. Metric Scaling: Correspondence Analysis. Newbury
Park: Sage Publications.