Resumé

The "Scandinavian welfare model" is often considered vulnerable to mass unemployment. The Danish welfare state provides an opportunity to examine the capacities of the "Scandinavian model" to adapt to this situation. This article explores a number of alleged crisis problems of the welfare state, grouped into budgetary pressures, incentive problems and legitimacy problems. It is concluded that most of these problems have been exaggerated and that the real threats to the economic foundations of the welfare state should be found in political steering and incentive problems rather than in the exogenous pressures from the social and economic system, or in pressures from the unintended side effects of welfare arrangements. It is furthermore argued that one of the main achievements of the Danish welfare state has been to prevent unemployment and labor market marginalization from developing into a broad-ranged social marginalization and a political polarization which could undermine citizenship and solidarity in society.

Introduction

The Danish welfare state has received much less attention in the international literature than the Swedish. Not without justification, Sweden is presented as the archetypal representative of the Scandinavian welfare model. This implies, however, that the crisis of the Swedish economy in the 1990s has raised serious questions about the viability of the once so highly praised Scandinavian model. As Table 1 shows, total public expenditure in Sweden rose dramatically from 60.0 percent of GDP in 1989 to 74.1 percent in 1993, which lead to an unprecedented government net lending in 1993 of 13.4 percent of GDP.

---

*Jørgen Goul Andersen, Department of Economics, Politics and Public Administration, Aalborg University, Fibigerstræde 1, 9220 Aalborg 0, Denmark.

However, the Swedish economy is likely to recover, and the economic problems which suddenly struck Sweden in the 1990s have almost been part of "everyday life" in the Danish welfare system throughout the last 20 years. Even in 1993, the Danish unemployment rate was still above the Swedish. Denmark had to adjust its welfare arrangements in the 1980s; Sweden is going through the same process in the 19905. Still, Denmark has by no means abolished its welfare system. On the contrary, at most points, there have been uninterrupted welfare improvements throughout the last 15 years. Thus, the Danish system may give both a more balanced impression of the problems of the Scandinavian welfare model and a more balanced account of what the Scandinavian welfare model may realistically hope to achieve.

This article assesses some of the main crisis arguments from the international literature whereas some of the more "parochial" national concerns such as the (largely flawed) arguments about the need to "adapt" the Scandinavian welfare state to the European Union are omitted in this article. The main issue, illuminated by the Danish experience, is whether the Scandinavian welfare model is still viable in the face of changing economic and social surroundings. Apart from economic pressures, we also consider arguments about citizenship and legitimacy. Besides, a brief account of the achievements of the Danish welfare state in the 1980s and 1990s may contribute to the debate about whether it is worth saving at all.

What Sort of Welfare State? - What Sort of Crisis?

To illuminate some basic characteristics of the Danish welfare model, a somewhat imaginary "Continental European Model" is presented as contrast. The description is inspired by the well-known works of Titmuss (1968) and Esping-Andersen (1990) but is more concrete and less theoretically ambitious (for detailed descriptions, see e.g. Ploug & Kvist 1994a,b,c; Goul Andersen & Munk Christiansen 1991). Some main aspects are described in Fig. 1.

Most basically, the Danish model is based on a principle of universalism or citizenship rather than social insurance contributions. This partly derives from the fact that Denmark consciously chose a different path than Germany in 1891, when old-age pension was first introduced (Petersen 1990), partly from the later dialectic between Social Democrats introducing reforms mainly for the working class and the efforts of non-socialist parties to extend such rights to their own middle-class constituencies (Baldwin 1988). As a consequence, means-tested social rights have become increasingly universal. A case in point is the Danish old-age pensions which are universal, flat-rate pensions.¹ In practice, even the unemployment benefit system approached the principle of universalism until 1996, as it required only one year of membership of an unemployment insurance fund and 26 weeks of paid employment to obtain full membership rights. As all members were entitled to seven years of unemployment benefits, the risk of being involuntarily transferred to the social assistance system was small.²

In accordance with the citizenship principle, the basic unit of the Danish welfare system has increasingly become the individual rather than the family. Social assistance and rent subsidies for tenants are still means-tested at the family level but most social rights are linked to the individual. For instance, student allowances are obtained by all students from the age of 18, and the allowances become entirely independent of parents' income when the students turn 20. Alongside the high labor market participation of women, this means that only some 1-2 percent of the adult population are privately supported (Goul Andersen, 1995d). All other adult citizens are either gainfully employed or receive some sort of transfer income from the state.

Nevertheless, it is a myth that welfare expenditures are exceptionally high in Denmark. Most of the difference between North West European welfare states reported in international statistics is caused by formal differences in accounting practices, taxation systems - and/or in the level of interest payments on public debt ("sins of the past"). When we correct for such differences, it turns out that expenditure levels for (current) public provisions are only slightly higher in Denmark than in Germany (Table 2). This is confirmed if we aggregate the social expenses and correct for differences in

taxation of transfer incomes which are typically paid as gross benefits
subject to tax in Denmark.3

The Danish tax system corresponds with the welfare system, i.e. social security contributions are almost entirely replaced by income taxes. However, the assumption that taxes on labor are extraordinarily high in Denmark is also a myth: If we add income taxes and social security

contributions into a "composite income tax," this (composite) income tax as a percentage of GDP resembles most other North European countries (OECD 1995c). The relative weight of composite income taxes, as a proportion of the total tax revenue, equals the average for the European Union (55 percent) but is much lower than e.g. in Germany (66 percent). Also at the individual level, we find that composite marginal tax rates for small and medium incomes are slightly lower than in Germany and the Netherlands, and under the tax reform act of 1993, they are scheduled to become quite a bit lower in 1998 (Regeringen 1993, 57). However, the Danish tax system is more progressive, and marginal taxes for higher incomes are among the very highest in Europe.

The Danish welfare state is strongly oriented towards the goal of equality, also beyond the sheer principle of universalism. Thus unemployment benefits are 90 percent of previous earnings but with a quite low ceiling of some 132,000 kr (1994/95 level). This means that in practice, unemployment benefits come close to a flat-rate benefit as most unemployed receive the maximum. However, since 1975, the average compensation rate has declined from some 75 percent to about 65 percent (Pedersen, Pedersen & Smith 1995, 41), i.e. close to that of other Northern European welfare states. Another example is the Danish educational system which assigns unusually high priority to basic education and to training skilled workers (Goul Andersen & Munk Christiansen 1991, Ch. 4) and less to higher education.

Next, the Danish welfare state gives high priority to service provisions,
whereas expenses for income transfers were quite small until recently (Goul
Andersen & Munk Christiansen 1991, Ch. 6). This means that public

employees constitute some 30 percent of the labor force; from 1970 to the early 1980s, the number of public employees almost doubled, not least because of a rapid expansion of public care for pre-school children and for the old. This went hand in hand with an explosion in female labor market participation (Table 7): Public care facilities not only enabled more women to enter the labor market but also provided a large number of the new jobs for women.

Overall, women are among the winners in the Danish welfare system, not only as far as employment opportunities are concerned. Thus more women than men receive unemployment benefits, and at higher compensation rates. The same holds for public pensions. Other beneficiaries include students, families with small children (provision of day-care institutions is even significantly higher than in Sweden. See Togeby 1987; Goul Andersen & Munk Christiansen 1991, 132), low income groups, and people with unstable employment. Higher income groups, on the other hand, are disadvantaged not only on the tax side but also as far as unemployment benefits and public pensions are concerned.

Thus, it is more the structure of the welfare state and the distributive aspects than levels of public expenditure that distinguishes the Danish (Scandinavian) welfare system from other welfare systems in North Western Europe. In some respects, the Danish welfare state is more generous, in other respects, other European countries have higher standards. However, the Danish welfare state contributes very significantly to social equality and gender equality, i.e. the welfare system compensates to a very large degree for differences in labor market positions. It is also at this point that we encounter the most important achievements of the Danish welfare state in the 1980s and 1990s (see below).

But, as we hear from many sides, all European welfare systems have serious problems. The literature on welfare state problems abounds, and each academic discipline seems to have developed its own notions of a welfare state crisis. Apart from the problem of achieving efficiency in service production on non-market conditions (a problem which is not considered here), economists have become increasingly concerned with unintended micro-level effects of welfare arrangements in terms of economic disincentives; political scientists have been concerned with steering problems such as factors leading to excessive growth of public expenditure; sociologists have been concerned with problems of alienation and paternalism; political sociologists have analyzed legitimacy problems, just to mention a few topics. Finally, all disciplines are concerned with the challenges of changing demographic and social structures such as aging populations and labor market marginalization. Not least long-term unemployment is claimed to generate a "two thirds society" (Dahrendorf 1988; 1994) which may undermine citizenship and solidarity as well as the

economy of the welfare state. Some economists even contend that
marginalization is largely an unintended side effect of the welfare state, as
it disturbs the flexibility of the market.

The analysis below concentrates on major aspects that are either very fundamental or might seem especially threatening to a welfare model of the Scandinavian type. In the following section we address the problem of "uncontrollable budgets." Next, we examine some of the problems highlighted by economists in the 19905, i.e. economic disincentives and inflexibility of labor markets - a problem which should be very profound in Denmark because of its high emphasis on economic equality. The next section examines theories of legitimacy problems in the light of recent survey evidence, and the final section assesses some of the achievements of the Scandinavian model as it has been functioning in Denmark during two decades of mass unemployment and economic problems interrupted only by short periods of prosperity.

Budgetary Pressures: Uncontrollable Expenditures?

Demand Overload and Ungovernability

In the 1970s and 1980s, discussions about the crisis of the welfare state were frequently phrased in terms of "demand overload" or "ungovernability" (Birch 1984). It was argued that the state was unable to withstand the demands of strong pressure groups and that the spiral of increasing demands and expectations, and declining economic capacity, would lead either to economic or moral bankruptcy of the welfare state, or both.

Certainly these pressures were and remain very strong but as the welfare state survived and even managed to put inflation under control, we know by now that the claims were exaggerated. But it is illuminating to examine at which point they went wrong. In retrospect, it appears that they (1) overestimated the power of interest groups and underestimated the powers of the state to withstand the pressures; (2) underestimated the capacity of public officials to learn from previous errors; and (3) underestimated the crisis awareness, the willingness to make sacrifices, and the adjustment of expectations among the general public. As the two last-mentioned factors largely explain the first-mentioned, they are discussed below.

When it comes to the power of pressure groups, it is important to distinguish between public consumption and transfers. Public services are usually not very redistributive (Hansen 1990; Goodin & LeGrand 1987),⁴ user groups tend to be strong and are supported by strong groups of public employees who deliver the services. Income transfers, on the other hand, are highly redistributive (Hansen 1990; 1995) but are received by people with fewer resources who often have difficulties organizing and have little

support from public employees. Thus social transfers should be much easier
to control than services.

This was also the lesson from the situation in the 1970s (Kristensen 1987). However, the point is that the lesson was learned. The bourgeois government from 1982-1993 took a number of initiatives to reverse the trend: It introduced fixed budgets to control total expenditures, it gave top priority to the goal of reducing the number of public employees, and the government put strong pressures on local authorities to reach this goal.

These goals were not reached. But the government did succeed in controlling the growth in public consumption (see Table 3). From 1982 to 1992, public consumption increased by only about 5 percent in real terms, that is, far less than the general economic growth. On the other hand, income transfers to households increased markedly from the mid-1980s, whereas interest payments on public debt which had exploded in the early 1980s stabilized at a high level. We return to the debt problem below; the point here is that public consumption expenditures are controllable when governments are sufficiently determined to control them. With few exceptions, it also seems that the general public did in fact accept the retrenchment policies as a whole because of the economic situation (Goul Andersen 1988).

Indeed, the mentality among the general public was very different from what was implicitly assumed in the "overload" theories. Thus, voter support for the new bourgeois government in 1982 increased dramatically in the first few months when it announced its harsh policies to restore the economy, whereas the subsequent improvement of the economic situation had no measurable effect (Goul Andersen 1995a). More generally, Danish citizens possess a high degree of crisis awareness and willingness to make sacrifices

in recession periods, and they adjust their expectations so that general satisfaction does not deteriorate (Petersen 1996). According to the Eurobarometer surveys, both satisfaction with life and satisfaction with democracy are unusually high in Denmark, and unlike in most countries, satisfaction has increased significantly (Goul Andersen 1994a, 59-61). Besides, there has been little dissatisfaction with taxation. This issue has not really been on the voters' agenda since 1973 (in spite of government attempts to mobilize on the issue of tax relief), and most people have believed that the state could not afford lower taxation (ibid., 46-56).

Aging Populations and Increasing Transfer Expenditures

After the decline of demand overload theories, attention has been directed more towards the increasing transfers, not least the costs of increasing unemployment and of demographic change. Like other advanced nations, Denmark will experience a considerable growth in the proportion of older people in the first half of the 21st century due to low birth rates. Various attempts to calculate the costs conclude that, other things being equal, the increasing costs (pensions, care, health expenditures) will amount to 5-6 percent of GDP over the next 35 years (Arbejdsmarkedspensionsudvalget 1988; Ministry of Finance 1993; Petersen 1994).

However, the economic burden of aging populations will increase less in Denmark than in many other European countries as fertility rates have increased from 1.4 in 1983 to 1.8 in 1994. Here in the mid-19905, the Nordic countries have the highest fertility rates in Europe, which probably mainly reflects cultural change. To some degree, it may also reflect welfare policies which assign high priority to working families with small children. At any rate, increasing fertility contributes to a somewhat less unfavorable ratio of old people relative to the working population in the future. Less fortunately, high mortality pulls in the same direction. Life expectancy in Denmark, which used to be among the highest in the world, has improved very little since the 1970s and is now the second lowest in the EU (Goul Andersen & Munk Christiansen 1991, Ch. 8; European Council 1995).

Turning from pure demography to welfare arrangements, we find that aggregate spending for health and public pensions in Denmark is relatively low, partly because of a very high pension age of 67 years. Like in other countries, the typical retirement age is lower, and it has declined considerably since an early retirement allowance was introduced in 1979 enabling people to retire at the age of 60 at favorable conditions. Increasingly, even better-educated people prefer early retirement, and this contributes to an increased burden of support. Even if it has a positive shortterm effect on unemployment, the major long-term effect tends to be a

decline in aggregate employment and a permanent increase in transfer
payments.

However, this is not a matter of demographic pressures on a "Scandinavian model." As such, long-term demographic changes are not very threatening, and probably less so than in most other countries. Early retirement arrangements are basically a matter of welfare priorities. And although they enjoy extremely high popular support, the arrangements were mainly introduced "from above," more than as a response to irresistible demands "from below." This brings us to the political determinants of budgetary pressures.

Endogenous Pressures: Political Factors

The budget experience from the last two decades points at the importance of
endogenous political factors rather than exogenous pressures from interest

groups and demographic changes. Table 4 shows the functional distribution of (total) public expenditure 1972-1992. It turns out that even though total expenditures increased from 50 to 70 percent of GFI, the share of GFI allocated to the health sector remained almost constant whereas educational expenses declined. Other social service expenditures (mainly care for preschool children and for the old) increased quite a bit. Nevertheless, total expenses for basic social service provisions increased its share of GFI by only 1.3 percentage points in spite of the dramatic increase in the number of public employees.

Social transfers (excluding unemployment benefits and social assistance) have increased considerably but mainly due to improvements of existing arrangements and the introduction of new programs. In spite of higher pensions and an increase in the number of old-age pensioners by nearly one third, the proportion of GFI assigned to old-age pensions has not increased. And even if we include social transfers, basic welfare expenditures (except unemployment) account for only one fourth of the total increase in public expenditures. The rest is explained by interest payments, unemployment, administration, and transfers to abroad.

This shows the importance of political factors. The low increases in basic welfare spending (and a relative decline from 1981 to 1992) reflect a conscious political effort necessitated by the growth in interest payments and other transfers. It may be added that in 1993, when the goal of zero growth was abandoned, public expenditure increased by 3.0 percent of GFI in one single year, that is, by one half of the expected increase derived from demographic changes over the next 35 years.

Now, increasing unemployment is usually blamed for most of the increases in social transfers in the 1980s and 19905, and this is often explained as the consequence of "old systems in a new world," i.e. of generous welfare arrangements from the 1960s working in the environment of mass unemployment today. However, as pointed out by Munk Christiansen (1995), this is not a satisfactory interpretation. As mentioned above, the average compensation rate for the unemployed has declined. And if we compare 1983 and 1993 (the two years with the highest unemployment in the 1980s and 19905), it furthermore turns out that the increases in expenditure for the "uncontrollable" transfers relating to age and unemployment have been far less dramatic (26 percent) than the increase in "other social transfers" (70 percent), that is, the more "controllable" part (see Table 5). Unemployment benefits only increased at the same rate as general economic growth, and old-age pensions increased far less. Even early retirement allowances only increased moderately, whereas social assistance and disablement pensions have increased rapidly. But "other transfers" exhibit by far the largest increase and furthermore account for around half of the total increases in transfers in real terms 1983-1993. These

figures even refer to a period before new leave arrangements and the transitional allowance became widespread. In 1994 and 1995, the lastmentioned transfers exploded, and by 1996, the costs are stipulated to equal the costs of the early retirement allowance. Thus, the conclusion is that the "politically controllable" expenses are less controllable than the "uncontrollable" transfers derived from unemployment and demographic changes. Or in other words, self-control by parliament and government seems to be a more important factor than demography when it comes to the question of controlling public expenditures.

However, "self-control by parliament and government" is a catchword without much content. The question is why it is so difficult for parliaments and governments to control themselves. The explanation can probably be found in party interests, not least in the need of small parties and opposition parties to obtain concessions in political negotiations. However, part of the explanation of increasing transfer expenditures from 1983 to 1993 may simply be found in a lack of attention. Decision makers were aware that it was difficult to control public consumption, and it seems that they were so concerned about this goal that they paid less attention to the increasing transfers. If this is the case, we should expect the opposite to happen in the following years as "increasing transfer expenditures" have now been defined as a most serious problem by influential experts and policy elites.

When it comes to the problem of the public debt (see Table 6), lack of
awareness was also part of the explanation. It took quite a long time to
abandon the Keynesian notion that state debt was simply a matter of "one

hand borrowing from the other." Increasing awareness of the distributional consequences and of the "disturbing" effects of taxes on economic efficiency means that in the 19905, decision makers have become highly aware of the problem. However, an even more serious obstacle in the 1990s is the political incentive structure (which is common to most countries): If governments choose to combat the problem, the short-term positive effects are very small, whereas the negative effects (e.g. higher unemployment) are felt immediately. This means that it is difficult to obtain significant effects within a single election term, and this undoubtedly contributes to a postponement of the solution. It is far more tempting to promise tax relief or increasing welfare, as the short-term (net) costs are small, and this is to a large degree what has been happening since the mid-1980s.

To conclude our diagnosis, although the forces are not mutually exclusive, the budgetary pressures of the welfare state seem to derive more from political logics (and lack of awareness) than from external pressures in terms of interest groups, demographic change, or even unemployment. "Democracy" may sometimes be a more appropriate label than "steering problems," but in Denmark, new or improved programs only seldom seem to be responses to irresistible pressures "from below." Thus it gives little meaning to speak of a crisis of the welfare system. Furthermore, there does seem to be counteractive forces, both among politicians and in the electorate, when expenditures are about to get out of control. This contributes to keeping the economic problems within certain limits.

Incentive Problems

In the 1980s and 19905, interest in the economic problems of the welfare state has increasingly been directed from the macro to the micro level, i.e. to the functioning of markets and to the incentives and disincentives which welfare arrangements generate for economic actors (Sandmo 1991). For instance, the welfare state has been accused of limiting incentives to private savings, "crowding out" investments, etc. Two recent concerns are discussed below: Work incentives related to taxation and to the relatively generous welfare arrangements for the unemployed. These problems are discussed all over Europe, not least in light of Europe's inability to create jobs, as compared with the United States (Møller 1995). However, as marginal taxes for upper, medium and high income groups are unusually high in Denmark, and as welfare arrangements for the unemployed have been unusually generous, in particular for women and (other) low income groups, such problems should be particularly threatening in Denmark. Welfare arrangements give low income groups little incentive to work, and income taxes give medium and high income groups little incentive to work.

The Danes Work A Lot

Beginning at the macro level, we can observe that the Danes work a lot. This does not exclude the possibility that they might work more if taxes were lowered but it puts the alleged problems in proportion. As revealed by Table 7, labor market participation is higher in Denmark than in most other countries.⁵ From the mid-19605, married women entered the labor market in large numbers, and twenty years later, housewives had nearly disappeared. This process was facilitated by the provision of large numbers of part-time jobs, but since 1980, the frequency of part-time labor has declined rapidly. Many women still work at slightly reduced hours, but half-time labor has

almost disappeared (Goul Andersen 1991b). Probably the increasing fulltime labor is the effect of three interconnected factors: (1) The integration of women at the labor market; (2) compensation for low wage increases; and (3) reduction of the work week (to 37 hours) and longer vacations (5 weeks) enabling more families to work on a double full-time basis and still have some time left. Even if we take account of the relatively short work week, however, the aggregate working effort is still unusually high. Besides, overtime and extra jobs are widespread, especially among younger males (Goul Andersen 1991b).

Tax Incentives

Even though labor supply is high, there has been much discussion about the negative incentives of high marginal taxes. However, econometric studies as well as survey analyses have faced unexpected difficulties proving a relationship between income taxes and labor supply, as well as between income taxes and wages (Goul Andersen and Munk Christiansen 1991, Ch. 7). In a summary of evidence from econometric studies, Pedersen (1993, 280) concludes that "empirical evidence so far does not indicate any significant impact from the income tax on the pre-tax wage." As far as the relationship between taxes and labor supply is concerned, an econometric analysis has estimated elasticity to somewhat less than 0.10 for women and just about 0.10 for men (Smith 1991).⁶ Pedersen (1993, 271) concludes that "overall, the estimated elasticities are fairly small compared with many studies on labor supply." By the same token, survey evidence indicates no

relationship between taxes and willingness to perform extra labor (Goul
Andersen 1994a, 98-105).7

This does not mean that tax incentives are irrelevant, but there is little to indicate that it is a more pertinent problem in Denmark than in other welfare systems, in spite of unusually high marginal taxes for upper-medium and high incomes. Another side effect is the "shadow economy" which has increased, partly in response to increasing income taxes since the 19605. It is now estimated at approx. 5 percent of GDP (Mogensen et al., 1995). The problem is largely limited to household services, however, (because social security contributions are nearly absent, employers have few incentives to employ illegal labor), and it does not seem that the amount of non-taxed labor is larger than in other Northern European countries, not to mention Southern Europe (Pedersen, forthcoming). There are probably strong effects on do-it-yourself labor, and on this point, Denmark and Sweden have the lowest employment in household services among the OECD countries (Goul Andersen 1994b).

Incentives of the Unemployment Benefit System

As indicated above, Danish arrangements for the unemployed have been unusually generous, as judged from a comparative perspective. This should, however, be seen in the light of unusually poor employment security in Denmark. Thus, the system has also been beneficial to Danish employers who have more hiring and firing flexibility than employers in most other countries.

Although the system has been tightened in most respects, except benefit levels from 1996 (the reforms are scheduled to be fully implemented by 1998), many economists remain concerned about the small differences between wages and unemployment benefits for low income groups. If we judge the incentive problems on the basis of data collected under the previous system, however, we obtain a very conservative test of such problems.

It turns out that the question of incentives must be split into two separate questions: One question concerning the effect of basic social security, and the other question concerning the effect of the size of differences between unemployment benefits and wages. As far as the first question is concerned, there is little doubt that the unemployment benefit system does reinforce high de facto minimum wages. This can be seen from the so-called "reservation wages" among the unemployed. In a representative, nationwide survey of long-term unemployed (in the following labeled the "1994 Unemployment Survey"),⁸ the respondents were asked about the income they expected to receive if they became employed, and about the lowest wage they were willing to work for (reservation wage). Only one third

among unemployed who expected a wage below DKK 165,000 (maximum benefits were 132,000) would be willing to accept a wage below their expected wage (Goul Andersen 1996a). Some 13 percent did not even want to work at their expected wage. Calculations by Pedersen & Smith (1995) also reveal that on average the wage demands of the unemployed are too high compared to the estimated market value of their labor power. Apart from small incentives to work, trade union consciousness and the fear of returning to unemployment at a lower level of unemployment benefits than previously may also play a role here, however.

When it comes to the association between motivation to work and expected gains above unemployment benefits, the results are far more ambiguous. Thus Pedersen & Smith (1995) have calculated that about 25 percent of the employed receive a monthly net gain of less than DKK 500 if we take into account taxes, transportation expenses and possible savings on child care.⁹ Furthermore, about one-third of the unemployed did not expect any improvement of their personal economic conditions if they became employed.

However, this is just to repeat that the welfare system is very equality oriented and has moved far towards "decommodification." The important question is whether such conditions affect people's motivation to work, and here the evidence is surprisingly negative. In Table 8, the unemployed are divided into four groups (for detailed operationalizations, see Goul Andersen 1996a):

(1) The passive, who are not very interested in having a job and who do
very little to find one.

(2) The busy, who are interested in having a job but not for the time
being, i.e. they are not actively looking for one.

(3) The reluctant, who are not interested in having a job but have to get a
job and are actively looking for one.

(4) The active, who are interested in having a job and are actively looking
for a one.

The marginal distributions between the four groups confirm that the provision of a basic social security level may reduce the eagerness to find a job, but the question here is if the difference between wage income and unemployment benefits is also an important factor, and the answer is overwhelmingly negative: The small group who expects an annual income above DKK 250,000 (i.e. almost twice the level of maximum unemployment benefits) is much more actively job seeking than is typical among the unemployed. By the same token, the small group who expects to earn an income below the level of maximum unemployment benefits (DKK 132,000 or less) is not very active. But among the remaining 85 percent of the longterm unemployed, the level of expected income virtually has no effect at all.

To conclude, there may be a certain trade-off between equity and efficiency in the sense that the provision of a basic social security level reduces the incentives of the unemployed to look for a job. But the size of the difference between wages and unemployment benefits does not seem to play an important role. Except at the extremes, money is not the problem. Rather, the "problem" is that a basic social security level reduces the interest in having unattractive jobs, almost regardless of wages. Like other people, the unemployed want jobs that give them something more than just a wage, i.e. not "junk jobs." On the other hand, they are willing to take more attractive "normal" jobs, regardless of low wages (Goul Andersen 1995b, Ch. 6 and 7). This is also confirmed by aggregate data from the period of prosperity after the prolonged Danish recession from 1987-1993. In certain high growth regions in Jutland, long-term unemployment was reduced by nearly 40 percent within the first year of prosperity, and more generally, long-term unemployment declines at a much faster speed than general unemployment (Goul Andersen 1996b).

It may be possible to increase employment by lowering minimum wages and social standards accordingly, thus increasing demand and forcing people to take even the least attractive jobs. It is estimated that a 10 percent reduction would provide about 25,000 new jobs (Ministry of Finance 1995b; Det økonomiske Råd 1995), that is, about a one percent increase in total employment. The costs, however, would not only be increasing inequality in society but also the risk of increasing social marginalization among the unemployed if benefits were too small to allow the unemployed to participate in society's standards and way of living. As the final section will show, the Danish welfare state has managed to avoid this social marginalization among the unemployed to a surprisingly large degree.

Legitimation Problems

Most "classical" theories of legitimation problems have been disproved by historical developments. This holds for Wilensky's (1975) "welfare backlash" of the middle mass; it holds for rising expectations arguments of neo-Marxists (Habermas 1973; Offe 1984) and demand overload theories (Birch 1984); and it holds for Rose & Peters' (1978) "bankruptcy" argument that welfare support would erode when real take-home pay declined due to increasing taxes, just to mention a few. Today, the question is more about ambivalence, new cleavages, alienation and morality. We address some of these questions briefly below.

Solidarity and Individualism

There are few opportunities to compare welfare attitudes in Denmark and other countries. However, ISSP questions on responsibilities of government were included in the Danish 1994 election survey, and the data in Table 9 indicate that in most respects, the demand for government responsibilities is very large in Denmark. There is one exception, however: Along with the Germans, the Danes are the least inclined to believe that it is a government responsibility to reduce income differences. Widespread support for the welfare state is also indicated by public expenditure questions (Goul Andersen & Munk Christiansen 1991, Ch. 10; Goul Andersen 1995c and 1995d).

In short, we can safely conclude that the Danish welfare state enjoys
widespread public support. This does not mean that people are entirely
uncritical. Needless to say, we find more negative attitudes when asking

about abuse, about the size of the bureaucracy, or about efficiency in the public sector (see e.g. Goul Andersen 1982; Svallfors 1989; Hviid Nielsen 1994; Goul Andersen 1995c). Some scholars would conceive of such differences as "ambivalence" (Feldman & Zaller 1992); however, at this point it is more plausible to speak of various subdimensions of attitudes which may be perfectly consistent (Papadakis 1992). The task of political leaders is to strike a balance, for instance between equity and risks of abuse; if they fail, attitudes towards abuse may become more salient and perhaps even undermine general welfare state support. Indeed, this seems to be an important part of the explanation of the 1973 Danish "welfare backlash" when even hard-core socialists complained about social abuse (Goul Andersen 1982). More generally, abuse and inefficiency, or the fear of excessive costs which are impossible to finance, rather than lack of solidarity seem to be the Achilles' heel of welfare legitimacy (Svallfors 1989; Rothstein 1994).

The arguments against the "ambivalence" interpretation are also relevant when we compare welfare attitudes with other social attitudes. Table 10 shows that overall support for the maintenance of welfare has remained quite stable since the 1960s (except for a temporary backlash in 1973/74), despite almost uninterrupted improvements and expansions. However, at the same

time, support for state regulation of the market has declined significantly,
and support for equality has declined even more.

This disjunction between welfare, state and equality which has also been observed in Norway (Martinussen 1988, 1993) probably reflects increasing individualism (Gundelach & Riis 1992) as well as a "victory of the market." It does not mean that people would welcome increasing inequality (for instance, only 16 percent would accept a lowering of minimum wages; see Goul Andersen 1996b), but there are limits to the demand for equality, and probably increasingly so. However, more individualist values do not imply a breakdown of solidarity, nor even "ambivalence" in relation to welfare state support.

Even if aggregate measures reveal no signs of declining solidarity, one might imagine a decline of solidarity in terms of increasing polarization between people with stable employment and people who are dependent on welfare state support. This is what can be expected from self-interested actor models, and exactly such a polarization between the employed majority and the welfare dependent minority is at the core of many recent debates.

Now, in most welfare states except the residual ones, it is quite difficult to identify who is dependent on the welfare state and who is not. This is particularly true for the universalist Scandinavian model, and it is exactly one of the rationalizations of universalism that it seeks to avoid such divisions. Leaving aside the question of welfare services, we may begin by distinguishing between those who are dependent on transfer income and those who earn their own income. However, if we simply compare these two groups, we find no attitudinal differences at all (Goul Andersen 1995c; 1995d). This is partly explained by quite negative attitudes among old-age pensioners who were socialized before the expansion of the welfare state in the 1960s and have lower expectations and different values.

Acknowledging that such generational differences may also be relevant, we could then confine the analysis to the 18-59 year olds. However, even here, we find no clear divisions between the employed and the publicly supported (in all instances, explained variance is below one percent). Again, the self-interest model is too simple: Firstly, some 30 percent of the labor force are public employees who hold positive attitudes towards the welfare state - not only towards social services, but also towards social security transfers. Secondly, arguments about polarization tend to forget that people live in families. As most women belong to the labor force, there is typically only one unemployed per household - but many more households than individuals that are affected by unemployment. If we include close relatives such as parents and children, even more people are affected. In short, if welfare support declines, it is likely to follow a parallel trend in all groups in society - among "insiders" and "outsiders" alike - and will not be a response to unwillingness to pay.

Alienating Effects of the Welfare State

A long tradition of welfare state criticism contends that legitimacy is not a matter of attitudes but rather a question of "alienating" effects of the welfare state, in particular the relationship between the bureaucratic rationality of the state and the norms and values of civil society. Ever since Tocqueville's (1835/40) vision of a mass society with a paternalist, benevolent state, a long intellectual tradition (both left and right-wing) has warned against a deterioration of the virtues of civil society. Thus, Habermas (1981) has argued that the "life world" is being colonized by the "system world," echoing to some degree Weber's arguments about the "iron cage" of rationality in modern society (Bild 1989); Wolfe (1989) has argued that the solidarity of civil society declines alongside with the institutionalization of solidarity in the welfare state; and others have warned against "welfare dependency." Scandinavian critics of the "patronizing state" include e.g. Arvidsson, Berntsson and Dencik (1993).

However, Tocqueville himself pointed at citizen participation in public life as a major cure against such tendencies. Inspired by Tocqueville, J. S. Mills and later social liberal writers also emphasized participation, but added that the state should provide the necessary resources to enable people to participate. In the modern welfare state, it may furthermore be added that the question of paternalism vs. empowerment is also a question of development of institutions that are responsive and invites citizen participation (Goul Andersen & Hoff, forthcoming).

At this point, we shall examine only a few aspects. Among others, we have to leave aside the question of whether the welfare state develops cynical citizens concerned only with satisfying their own needs,¹⁰ but we may examine the feeling of paternalism.

In Denmark, the younger generations have moved significantly to the right in the 1980s and 19905. This is associated with their increasing individualism and declining belief in equality (Gundelach & Riis 1992; Goul Andersen 1996b), and political observers have sometimes suggested that this is an effect of dissatisfaction with the paternalism of the welfare state. However, the very opposite turns out to be the case. As shown in Table 11, older voters tend to agree that "the welfare state leads to paternalism" but among the young generation, only one fifth answers the question affirmatively. It furthermore turns out that there is a very large generation gap among the supporters of the right-wing parties on this issue. Young liberals and conservatives are even less inclined to complain about paternalism than older Social Democrats.

Achievements

A record of the Danish welfare state would be incomplete without an account of its achievements. Needless to say, it is generally recognized even among welfare state critics that the provision of care facilities for children and the elderly has contributed significantly to improving women's conditions in Danish society. "Objective" judgments, taking comparative standards as criteria, would probably conclude that the Danish welfare state provides unusually high care standards for pre-school children and the elderly (although many Danes disagree on the last point). The attainments of the educational system and the health sector are probably more problematic, but here it is more important to evaluate to which degree the welfare state has been able to meet the challenges of increasing inequality and labor market marginalization as this is frequently seen more or less as the end of Marshall's (1949) idea of social citizenship or "full citizenship" (van Steenbergen, ed. 1994).

Even though increasing equality has lost importance as a top priority of public policies, as well as among citizens, equality remains a basic premise of the Danish welfare state. It has also been very successful in creating equality. As shown in Table 12, Denmark has about the lowest inequality in the distribution of disposable incomes (measured by Gini coefficients),¹¹ and unlike many countries, Denmark has not experienced increasing inequality during the 1980s (Hansen 1990, 1995). Furthermore, inequality in political participation has declined in almost all fields during the 1980s (Andersen et al. 1993; Goul Andersen & Hoff forthcoming).

It is generally agreed that the most important challenges to the European
welfare states are constituted by changes on the labor market, in particular

the tendencies towards marginalization and loss of citizenship of a
considerable proportion of the labor force.12

As far as material well-being is concerned, the unemployed do not appear marginalized in Denmark. As Fig. 2 demonstrates, only 3 percent of the households where the head of household is unemployed live below the (relative) poverty line. This may be compared to 22 percent in the Netherlands and to an EU average of 38 percent. It also turns out that among the unemployed who are married/cohabiting, the proportion of home owners remain above 50 percent, regardless of duration of unemployment, i.e., even among people who have had no ordinary employment for 10 years or more (see Goul Andersen 1996a).

Now, material well-being may be a necessary but not a sufficient condition to avoid social marginalization among the unemployed. However, even data on subjective well-being point in the same direction. Only 34 percent of the long-term unemployed express a lower level of well-being compared to when they were employed, and a huge majority have, according to their own information, the same or even more contacts with friends and acquaintances than previously (Goul Andersen 1996a).

This does not speak against innumerable reports which indicate unanimously that unemployment is associated with nearly all conceivable social problems. But the results do indicate that the majority of the unemployed, even of the long-term unemployed, do not experience such problems to a considerable degree. This is also confirmed by the fact that when the unemployed were asked whether they "could accept a period without work if (they) could receive unemployment benefits or social assistance as long as (they) wished," 44 percent answered that they "would welcome it," whereas 29 percent "wouldn't like it but could make the best of it" (Goul Andersen 1996a). Finally, it turns out that there is no significant causal effect of unemployment upon political participation (Svensson & Togeby 1991; Goul Andersen & Hoff 1995).

Conclusions

The Scandinavian welfare model is still alive and in good shape. The unusually severe problems of the Swedish economy in the 1990s is not an indication of a crisis of the Scandinavian welfare system. The ability of the Scandinavian model to withstand long-term economic problems is demonstrated convincingly by the Danish welfare state in the 1980s and 19905, and Sweden is likely to recover with its welfare systems largely intact (albeit on a slightly less ambitious level).

On the other hand, the achievements of the Scandinavian welfare systems should not be exaggerated. Several welfare arrangements may be more generous, more efficient, or both, in other welfare systems. Besides, most welfare systems, including Scandinavia's, have suffered from political incentive problems and other steering problems which tend to generate a large public debt with interest payments that limit the long term economic capacities of the welfare state. There is also a permanent challenge, not least in service-oriented welfare states of the Scandinavian type, to ensure efficiency in public service "production," and to avoid paternalism. Finally, the problem of striking a balance between equity on the one hand and economic efficiency or avoidance of abuse at the other always remains.

The fact that this article has concentrated on positive aspects should not be misperceived as an apologetic defense. Rather, the point is that most of the problems highlighted in theoretical debates seem, by and large, to have been solved quite successfully in the Danish welfare state, and this confirms the vitality of this welfare model or welfare system. Certainly economic management may be far less than optimal. But there is nothing to indicate that the economic problems constitute a "crisis," and the lesson from the 1980s is that such problems may be effectively counteracted when they become sufficiently threatening. And certainly, paternalism may be a latent

threat, but it may be effectively counteracted by providing citizens with
resources to defend their interests and by providing opportunities of user
influence in public service.

Perhaps most importantly, the Scandinavian welfare systems have had to give in as far as the goal of full employment is concerned. In 1990, Esping predicted that mass unemployment was the Achilles' heel of the Scandinavian or "Social Democratic" model; the Swedish experience from 1990-1993 comes close to confirming this statement, but with the important qualification, illustrated by Denmark, that it is possible to adjust to this situation. And indeed, the most convincing attainment of the Scandinavian model, demonstrated by Denmark, is exactly how it manages to live with unemployment, i.e. how it has managed to avoid increasing inequality and to maintain citizenship among the unemployed in a situation of enduring mass unemployment. There are virtually no signs of a developing two thirds society, nor of a political conflict structure characterized by loss of solidarity between labor market "insiders" and "outsiders." Furthermore, the maintenance of equality and citizenship has seemingly been achieved without significant loss of economic efficiency. This is not an ideal situation compared with full employment, but it is certainly not a bad "second-best" solution.

---

NOTES

1. A pension supplement is means-tested (only on income) but is received by most pensioners. Besides, there is a small, employment-related (but not earnings-related), funds-based pension supplement (ATP). A universal flat-rate pension system may be interpreted either as a typical aspect of the Scandinavian model which was abandoned in Norway and Sweden in favor of a contribution-related supplementary system (Rold Andersen 1993) or as a residual, liberal deviance from the model as it has paved the way for an enormous expansion of private pension schemes (Esping-Andersen 1990; Hagen 1991). As a "residual" model, however, the Danish system is unusually generous, as compensation rates for many single, unskilled workers come close to 100 percent (Arbejdsmarkedspensionsudvalget 1988).

2. Since 1996, the system has been tightened. Now 52 weeks of normal employment are required to qualify for benefits; the maximum length of an unemployment period will be limited to five years; and the unemployed will be obliged to participate in an activation program after two years of unemployment until the end of the period (previously four years). Further, the system has been tightened for unemployed aged less than 25 years.

3. An equivalent account by the Danish Ministry of Finance (1994) shows the following figures for 1991; Denmark: 53.4 percent; Netherlands 45.6; Germany 50.3 percent. (Correcting further for double counting of pensions for public employees, and for business cycle variations, the Ministry found that the entire difference between Denmark and Germany disappeared).

4. At least they are not (vertically) redistributive if we look exclusively at the consumption side. In Denmark, services as a whole arealmost neutral as far as redistribution is concerned (Hansen 1990). Still, state provision of such services nevertheless imply a massive redistribution compared to a situation where such services were provided by the market: Services are consumed equally, but the rich pay the most, even when taxes are proportional (see also Rothstein 1994, 179). Much of the welfare state criticism at this point fails by implicitly assuming that redistribution demands that both the consumption side and financing should be to the advantage of the poor. However, it is sufficient that one of these conditions is fulfilled (Goul Andersen 1991a), and in the real world, total privatization of services rather than means testing is the most relevant counterfactual. In fact, the tax equivalent of total privatization is a poll tax.

---

4. At least they are not (vertically) redistributive if we look exclusively at the consumption side. In Denmark, services as a whole arealmost neutral as far as redistribution is concerned (Hansen 1990). Still, state provision of such services nevertheless imply a massive redistribution compared to a situation where such services were provided by the market: Services are consumed equally, but the rich pay the most, even when taxes are proportional (see also Rothstein 1994, 179). Much of the welfare state criticism at this point fails by implicitly assuming that redistribution demands that both the consumption side and financing should be to the advantage of the poor. However, it is sufficient that one of these conditions is fulfilled (Goul Andersen 1991a), and in the real world, total privatization of services rather than means testing is the most relevant counterfactual. In fact, the tax equivalent of total privatization is a poll tax.

5. The figures in Table VII are a bit inflated by high frequencies of part-time supplementary labor among school children and students.

6. These figures refer to overall elasticities which include both an incentive effect and an income effect (as taxes are lowered, consumption possibilities increase, and people have to work less). The so-called "compensated elasticity" where incentive effects are isolated is about 0.15 for men (Smith 1990), but for practical purposes, it is the overall elasticity that is important.

7. In 1990, two surveys were carried out almost simultaneously. In the first survey, the respondents were asked whether they would want to perform extra work if income taxes were lowered. In the second survey they were asked whether they had performed extra work within the last year. The figures turned out to be entirely identical. Besides, in the second survey respondents were also asked if they wanted to perform extra work if they had the opportunity, and the proportion answering "yes" was substantially higher than the proportion indicating that they would work more if taxes were lowered.

8. Unemployment survey carried out by the Danish National Institute of Social Research, November 1994. Nationwide representative survey of approx. 1,250 respondents who had been unemployed for at least six months at the time of the interview. The respondents had been unemployed for at least six months before the interview took place. For details about the survey, see Ministry of Finance (1995a) and Goul Andersen (1995b).

9. Subtracting savings on child care is problematic. As the huge majority of 3-5 years old children attend kindergartens, children staying at home may become "deviant," and most parents want their children to attend kindergartens even if they have the opportunity to keep their children at home.

10. See Martinussen (1994) and Koch-Nielsen (1996) for counter-evidence against this alienation hypothesis.

11. Data from the European Commission (1995) further reveal that Denmark has by far the lowest proportion of households living in poverty among the EU countries, if poverty is defined as having less than one half of the median household in disposable income (corrected for household size and composition).

12. By marginalization, we do not mean "exclusion" (which in the Danish case appears to be more of a voluntary than involuntary nature) but long-term or recurring unemployment among those who belong to the labor force (following Halvorsen 1995; Johannessen 1995).

REFERENCES

Andersen, J., et al. 1993. Medborgerskab. Demokrati og politisk deltagelse. Herning: Systime.

Baldwin, P. 1988. "How Socialist is Solidaristic Social Policy?", International Review of
Social History 33, 121-147.

Bild, T. 1989. "Tvivlen om det rationaliserede samfunds fornuft," in Klausen, K. K. & Hviid
Nilsen, T, eds., Stat og marked. Copenhagen: Jurist- og Økonomforbundets Forlag.

Birch, A. H. 1984. "Overload, ungovernability, and Deligitimation: The Theories and the
British Case," British Journal of Political Science 14(2), 135-160.

Dahrendorf, R. 1988. The Modern Social Conflict. London: Weidenfeld & Nicolson.

Dahrendorf, R. 1994. "The Changing Quality of Citizenship," in van Steenbergen, 8., ed., The
Condition of Citizenship. London: Sage.

Danmarks Nationalbank. 1996. Statens låntagning og gæld 1995. Copenhagen: Danmarks
Nationalbank.

Det økonomiske Råd. 1989. Tax Burden and Public Expenditure in Denmark, Germany and
the Netherlands. Working paper. Copenhagen: Det økonomiske Råd.

Det økonomiske Råd. 1995. Dansk økonomi, juni 1995. Copenhagen: Det økonomiske Råd.

Esping-Andersen, G. 1990. The Three Worlds of Welfare Capitalism. Princeton: Princeton
University Press.

European Commission. 1994. Employment in Europe, 1994. Danish ed. 1995.

European Council. 1995. Recent Demographic Developments in the Member States of the
Council of Europe. Strasbourg: European Council.

Feldman, S. & Zaller, J. 1992. "The Political Culture of Ambivalence: Ideological Responses
to the Welfare State," Americal Journal of Political Science 36(1), 268-307.

Goodin, R. E., Le Grand, J., et al. 1987. Not Only The Poor: The Middle Classes and the
Welfare State. London: Allen & Unwin. .

Goul Andersen, J. 1982. "Den folkelige tilslutning til socialpolitikken: En krise for
velfærdsstaten?," Anckar, D., Damgaard, E. & Valen, H., eds., Partier, ideologier, väljare.
Åbo: Åbo Academy Press.

Goul Andersen, J. 1988. "Vælgernes holdninger til den offentlige udgiftspolitik," in Bentzon,
K-H., ed., Fra vækst til omstilling. Copenhagen: Nyt fra Samfundsvidenskaberne.

Goul Andersen, J. 1991a. "Mod klassisk demokrati. Ansvarlighed og uansvarlighed i dansk
politik," GRUS 32.

Goul Andersen, J. 1991b. "De arbejdsomme danskere. Om skatteplagede danskeres arbejdsvillighed og fleksibilitet," in Petersen, E. et al., De trivsomme og arbejdsomme danskere. Krisen og den politisk-psykologiske udvikling 1982-90. Århus: Psykologisk Institut/Aarhus Universitetsforlag.

Goul Andersen, J. 1993. Politik og samfund i forandring. Copenhagen: Columbus.

Goul Andersen, J. 1994a. "Samfundsøkonomi, interesser og politisk adfærd," in Petersen, E. et
al.. Livskvalitet og holdninger i det variable nichesamfund. Psykologisk Institut, Aarhus
Universitet (in commission at Aarhus Universitetsforlag).

Goul Andersen, J. 1994b. "Skal ledigheden bekfmpes ved at satse på servicesektoren?,"
Årsberetning fra Finanssektorens Arbejdsgivere 1993/94.

Goul Andersen, J. 1995a. "(Hvorfor) vinder regeringen ikke på den økonomiske fremgang?
Policy vurderinger, policy effects og andre foklaringer." Aarhus: Department of Political
Science, University of Aarhus. Working Papers from the Danish Election project no. 2.

Goul Andersen, J. 1995b. De ledige ressourcer. En analyse af de langtidsledige - deres
baggrund, forventninger og holdninger. Copenhagen: Ugebrevet Mandag Morgen.

Goul Andersen, J. 1995c. "Velfærdsstatens folkelige opbakning." Social Forskning
temanummer aug.1995 om velfærdssamfundets fremtid.

Goul Andersen, J. 1995d. "Welfare State Legitimacy in a one-third Society. Attitudes towards the Welfare State among Public Employees, Publicly Supported, and the State non Minority." Article to be published in Albæk, E., et al. The Danish Welfare State (working title).

Goul Andersen, J. 1996a. "Marginalization, Citizenship and the Economy: The Capacities of
the Universalist Welfare State in Denmark" in Eriksen, E. O. and Loftager, J., eds., The
Rationality of the Welfare State. Oslo: Universitetsforlaget.

Goul Andersen, J. 1996b. "Socialdemokratiets vælgerudvikling 1971-1994" in Callesen, G. et
al.. Udfordring og omstilling. Bidrag til Socialdemokratiets historie, 1971-1996. Copenhagen:
Fremad.

Goul Andersen, J. & Munk Christiansen, P. 1991. Skatter uden velfærd. De offentlige udgifter i
international belysning. Copenhagen: Jurist- og Økonomfobundets Forlag.

Goul Andersen, J. & Hoff, J. 1995. "Lighed i den politiske deltagelse," in Madsen. M„
Nielsen, H.J. & Sjöblom, G., eds., Demokratiels mangfoldighed: Tendenser i dansk politik.
Copenhagen: Forlaget Politiske Studier.

Goul Andersen, J. & Hoff, J. Forthcoming. Democracy and Citizenship in the Scandinavian
Welfare States. London: Macmillan.

Gundelach, P. & Riis, O. 1992. Danskernes værdier. Copenhagen: Forlaget Sociologi.

Habermas, J. 1981. Theorie des Kommunikativen Handelns. Frankfurt: Suhrkamp.

Hagen, K. 1991. "Linjer i velferdsstatsforskningen," Tidsskrift for samfunnsforskning 32,
187-216.

Halvorsen, K. 1995. Virker velferdsstaten? Arbeidsmarkedsmarginalisering og levekår.
Ø.K.S.-rapport, Høgskolen i Oslo.

Hansen, F. K. 1990. "De offentlige udgifters fordelingsmæssige virkninger," Politica 22(4),
457^69.

Hansen, F. K. 1995. "Velfærd og fordeling," Samfundsøkonomen 1995: 6, 9-14.

Hviid Nielsen, T. 1994. "Velfærdsstatens attituder: Ambivalens og dilemmaer," in Andersen,
J. & Torpe, L., eds., Demokrati og Politisk Kultur. Rids af et demokratisk medborgerskab.
Herning: Systime.

Johannessen, A. 1995. Arbeidsmarginalisering og levekår. ØKS-rapport 1995: 4. Oslo:
Høgskolen i Oslo.

Koch-Nielsen, I. 1996. "Den personlige indsats lever og har det godt." Social Forskning.
Særnummer, aug. 1996, 42-47.

Kristensen, O. P. 1987. Væksten i den offentlige sektor. Copenhagen: Jurist- og
Økonomforbundets Forlag.

Lindbeck, Asser. 1994. Turning Sweden Around. Cambridge, Mass.: MIT Press.

Martinussen, W. 1988. Solidaritetens grenser. En undersøkelse av oppsluttningen om
velferdsstatens verdier og virkemidler. Oslo: Universitetsforlaget.

Martinussen, W. 1993. "Welfare-State Support in Achievement-Oriented Hearts: The Case of
Norway," in Hansen, E. J. et al., eds., Welfare Trends in the Scandinavian Countries. New
York: M .E. Sharpe.

Ministry of Finance. 1993. Finansredegørelse 1993. Copenhagen: Finansministeriet.

Munk Christiansen, P. 1995. "De nordiske velfærdsstaters tilpasningsbehov og tilpasningsevne,"
in Effektivisering av välfärdsstaten. Rapport från ett seminar 16-17. mars 1995,
Reykjavik, Island. TemaNord 1995: 578. Copenhagen: Nordisk Ministerråd.

Møller, I. H. 1995. Labour Market Marginaliz.ation in North-West Europe: Some Causes and
Consequences. Copenhagen Business School: CID Studies Report.

OECD. 1993. Employment Outlook. Paris: OECD.

OECD. 1995. Income Distribution in OECD Countries: The Evidence from the Luxembourg
Income Study. Paris: OECD.

OECD. 1995a. Historical Statistics 1960-1993. Paris: OECD.

OECD. 1995c. Revenue Statistics of the OECD Member Countries. Paris: OECD.

Papadakis, E. 1992. "Public Opinion, Public Policy and the Welfare State," Political Studies
XL, 21-37.

Pedersen, P. J. 1993. "Welfare State and Taxation in Denmark," in Atkinson, A. B. &
Mogensen, G. V., eds., Welfare and Work Incentives. A North European Perspective.
Oxford: Clarendon Press.

Pedersen, S. Forthcoming, [title unknown] Copenhagen: The Rockwool Foundation Research
Unit.

Pedersen, L., Pedersen, P. J. & Smith, N. 1995. "The Working and the Non-Working
Populations in the Welfare State," in Gunnar Viby Mogensen, ed., Work Incentives in the

Danish Welfare State. New Empirical Evidence. Århus: Aarhus University Press/The
Rockwool Foundation Research Unit.

Pedersen, P. J. & Smith, N. 1995. "Unemployment and Incentives," in Viby Mogensen, G.,
ed., Work Incentives in the Danish Welfare State. New Empirical Evidence. Arhus: The
Rockwool Foundation Research Unit/Aarhus University Press.

Petersen, J. H. 1990. "The Danish 1891 Act on Old Age Relief: A Response to Agrarian
Demand and Pressure," Journal of Social Policy 19, 69-91.

Petersen, J.H. 1994. "ldreforsørgelsesbyrden - et dansk perspektiv," Nationaløkonomisk
Tidsskrift 132, 286-298.

Petersen, E. 1996. "Den samfundspsykologiske udvikling 1982-1986-1988-1990-1994," in
Petersen, E. et al., Danskernes trivsel, holdninger og selvansvarlighed under ' 'opsvinget.' '
Aarhus: Department of Psychology/Aarhus University Press.

Ploug, N. & Kvist, J. 1994a. Social Tryghed i Europa: Udvikling eller afvikling? Copenhagen:
Socialforskningsinstituttet.

Ploug, N. & Kvist, J. 1994b. Overførselsindkomster i Europa. Systemerne i grundtræk.
Copenhagen: Socialforskningsinstituttet.

Ploug, N. & Kvist, J., eds. 1994c. Recent Trends in Cash Benefits in Europe. Copenhagen:
Socialforskningsinstituttet.

Regeringen. 1993. Ny kurs mod bedre tider. Copenhagen: The Danish Government.

Rold Andersen, B. 1993. "The Nordic Welfare State Under Pressure: The Danish Experience,"
Polity and Politics, 21(2), 109-120.

Rose, R. & Peters, B. G. 1978. Can Government go Bankrupt? New York: Basic Books.

Sandmo, A. 1991. "Economists and the Welfare State," European Economic Review 35,
213-239.

Smith, N. 1990. "Household labour supply and taxes in Denmark." Centre of Labour Market
Economics.

Smith, N. 1991. "Labour supply, unemployment and taxes in Denmark." Department of
Economics, Aarhus University.

Statistiske Tiårsoversigter. Various issues.

Svallfors, S. 1989. Vein älskar välfärdsstaten? Attityder, organiserede intresser och svensk
välfärdspolitik. Lund: Arkiv.

Svensson, P. & Togeby, L. 1991. Højrebølge? Arhus: Politica.

Taylor-Gooby, P. 1995. "Who wants the welfare state? Support for state welfare in European
countries," in Svallfors, S. ed., In the Eye of the Beholder. Umeå: Impello Sja'lsupport/The
Bank of Sweden Tercentenary Foundation.

Titmuss, R. M. 1968. Commitment to Welfare., London: Allen & Unwin.

Tocqueville, A. de. 1835/40. De la démocratie en Amerique. quoted from Danish (abridged)
ed., 1984.

Togeby, L. 1987. Notat om børnepasningsdækningen i de nordiske lande (mimeo). Aarhus:
Department of Political Science, Aarhus University.

van Steenbergen, B. ed. 1994. The Condition of Citizenship. London: Sage.

Wilensky, H. 1975. The Welfare State and Equality. Berkeley: University of Califonia Press.