CHAPTER 12 Welfare State Restructuring and Child Care in Sweden Christina Bergqvist and Anita Njberg Introduction The 1990s was one of the most critical periods for the Swedish welfare model. The employment rate fell dramatically and unemployment soared to levels unthinkable since the 1930s.1 The situation began to improve only as the decade came to an end.The employment crisis, in turn, produced an accelerating public sector deficit, with revenues plummeting and public expenditures skyrocketing.2 In addition to the economic crisis, there were also other factors that constituted a challenge to the stability of the traditional Swedish welfare state. First, the Social Democratic Party lost its historically dominant position, which opened the way for neoliberal ideas of marketization and privatization. The internationalization of capital markets and financial transactions, plus Sweden's participation in the European integration project, also posed new challenges. Given the unemployment situation, the financial strains, globalization, and the spread of neoliberal ideas, it is reasonable to assume that serious attempts to transform the Swedish welfare state might have been undertaken. The aim of this chapter is to analyze the characteristics of this restructuring and to determine whether retrenchment did indeed take place in the Swedish welfare state in the 1990s. We will do this through the lens of public child care. The Swedish welfare state has as a political goal included the provision of high quality child care since the 1970s.3 It was intended as a vehicle to promote gender equality and full employment, as well as equality between different kinds of families and equal provision of care and education for preschool children (Leira 1993; Hirdman 1998; Bergqvist 1999). As Paul Pierson has argued, retrenchment involves more than cuts (1996, IS7). In addition to expenditure cuts, changes in policies and program structure must also be examined. Our investigation accordingly relies on a combination of data on expenditures and qualitative analysis of the child care sector. We examine Swedish child care in relation to five important characteristics associated with the Swedish welfare state: (1) generous public spending; (2) high quality in social services; (3) limited scope for the private sector in social services; (4) universalism; and (S) egalitarianism (Esping-Andersen and Korpi 1987; Pierson 1996; Esping-Andersen 1999, 78—81; Sainsbury 1999, 259). By analyzing whether and how these characteristics have changed in the 1990s, we assess where the Swedish welfare state is heading. 288 Welfare State Restructuring and Child Care in Sweden The Historical Legacy and Development of Child Care Policies The Social Democratic regime, for which Sweden is often the exemplar, is typified by universal benefits, a level of compensation and quality of service high enough to meet middle class expectations, and a strong commitment to equality and full employment (Esping-Andersen 1990). While Gosta Esping-Andersen has focused on the relationship between the state, market, and class, feminist researchers have emphasized the importance of including the family and gender relations (Sainsbury 1994). The move from the male-breadwinner to dual-breadwinner norm has been one of the biggest social changes of the last thirty years, and the social democratic regime has tended to be more supportive of this change than other models (Hemes 1987; Borchorst 1994; Sum 1997).This is especially visible in Sweden's child care policies. Although small public grants for child care institutions have been available since the 1940s, the demand for public child care has always been greater than supply. For example, in 196S, only 3 percent of all preschool children were in public child care. At the same time, nearly 36 percent of the mothers of preschool children were employed. The majority of parents arranged child care in the informal sector and this was common well into the 1980s (Nyberg 2000). It was, however, in the 1970s that public child care, together with the transformation of maternal leave into an income-related parental leave, became important characteristics of the "new" Swedish welfare and gender equality model. The introduction of a National Preschool Act in 197S imposed on local authorities the obligation to expand public child care. The municipalities were now required to provide all six-year olds with at least S2S hours of free preschooling—that is, at least part-time child care for six-year olds.4 For younger children, however, eligibility for a place in public child care was linked to parent's employment (or study) status. Exceptions were made primarily for children in need of special support. The demand for public child care was always greater than supply. In 198S, the Social Democratic government increased its efforts to make good on its commitment to working parents, aiming to ensure places in public child care for all children between one and a half and six years of age whose parents worked or studied (Bengtsson 199S; Ministry of Education and Science 1999). As a result, the number of children enrolled in different forms of publicly subsidized child care increased substantially. The most common form of service is the child care center, but family child care units and preschools remain important. In 197S, only 17 percent of all children between one and six were enrolled in one of these forms of child care. By 1990 the number had risen to 57 percent (see table 12.1). The use of public child care has to be seen in relation to the age of the child and the length of the paid parental leave. Today, hardly any children under the age of one are enrolled in public child care because of the long parental leave period, which increased from around six months to over a year in the mid-1970s. Between the 1970s and 1990s the expansion of public child care services was Christina Bergqvist and Anita Nyberg 289 generally an accepted policy objective (Gustafsson and Antman 1996). Two of the three bourgeois parties certainly remained in favor of a care allowance and were critical, in principle, of too much state involvement. Nevertheless, even when they formed the government in the latter part of the 1970s, they refrained from introducing any significant changes. During the 1980s, however, a debate, embedded in a discourse of "choice," arose over whether to increase for-profit alternatives. The social democratic government opened up the possibility for public support to centers not run by the municipalities but continued to exclude private for-profit child care, which prompted severe criticisms from the opposition, but also from some social democrats (Bengtsson 1995; Mahon 1997). Choices and Challenges in the 1990s Throughout the 1980s, the nonsocialist parties intensified their promotion of neoliberal alternatives that stressed choice, decentralization, markets, and privatization. In addition to questioning the rule prohibiting public subsidies for commercial child care, it was argued that parents did not have a real choice between staying at home with children or working. Introduction of a care allowance—long favored by the Conservative, Center (agrarian), and small Christian Democratic Parties—was again touted as a means for providing such choice. With the election in 1991 of a bourgeois coalition government, headed by Carl Bildt of the Conservative Party, a child care policy in line with neoliberal and conservative ideas was on the agenda. Interestingly enough, while the Bildt government introduced elements of a bourgeois policy, it also expanded the social democratic/liberal line of earlier child care policies. Thus the care allowance was introduced in July 1994, just as new elections were approaching. The aim of the allowance was to make it possible for one of the parents to stay at home for a longer period than parental leave, which lasts for about a year with income compensation plus three months with a flat-rate benefit. This version of care allowance was, however, designed more as a child care check than as a mother's wage. The parents could use the money to enable one parent to stay at home or for child care. Parents could also choose between public and private forms of child care, both of which were eligible for public subsidies.5 Table 12.1 Children Enrolled in Publicly Subsidized Child Care Centers, Preschools, and Family Day Care Units, 1975-1990 (IN percent of the population by age) 1-2 YEARS OLD 3-6 YEARS OLD 1-6 YEARS OLD IN TOTAL 197S 16 17 17 1980 31 38 36 198S 45 55 52 1990 44 64 57 SOURCE: Skolverket (1998). 290 Welfare State Restructuring and Child Gare in Sweden The government coalition, however, also included the Liberal Party, which continued to favor the right to institutionalized child care and a more equal division of labor between women and men in the family. In particular, the Liberal minister for social affairs and gender equality, BengtWesterberg, was a driving force behind two important developments—a new Act on Child Care, and what is popularly known as the "daddy month," a measure designed to get fathers to take a greater share of parental leave (Bergqvist, Kuusipalo, and Styrkarsdóttir 1999). The Act on Child Care was introduced as part of a move from detailed state regulation of the municipalities towards a general decentralization of authority and responsibilities to the municipalities. This process involved a tightening and clarification of municipal obligations to provide child care without unreasonable delay for children between the ages of one and twelve. The main criteria remained, however, that the parents were working or studying or that the child had special needs. The municipalities were also given more freedom. During the phase of expansion in the 1970s and 1980s, public child care had been part of centralized government funding. Gross costs of child care were divided in such a way that the state and the municipalities each covered about 45 percent of the costs and parents' fees the rest. In 1993, program-specific state grants to the municipalities for child care were replaced by block grants (Ministry of Education and Science 1999; SOU 2000: 3; Oberhuemer and Ulich 1997).The municipalities thus have a higher autonomy today as to how these funds should be expended and how child care should be run and organized. At the same time, the act required the municipalities to provide child care "without unreasonable delay." To summarize, the family policies of the bourgeois government were grounded on a contradictory mixture of traditional family values, neoliberal ideas of privatization and of gender equality. It would thus be erroneous to conclude that a radical shift towards a new model was taken. The basic components of the parental leave legislation were maintained and the right to child care was strengthened. The care allowance, moreover, became but an historical parenthesis. The Social Democratic government that took office in the fall of 1994 abolished it while retaining the "daddy month," the Act on Child Care, and the opening to public funding for commercial child care centers. Public responsibility for child care remained strong throughout the 1990s, even though the state had assumed a new role. Authority and responsibilities had been decentralized and, at the local level, the provision of child care became more diversified. Public Expenditure on Child Care6 One criterion to distinguish types of welfare states is the level of expenditure on social services. In Sweden, a large proportion of national income is devoted to the goal of provision of high quality services for all. In terms of child care, the provision of accessible, high-quality child care services is understood to contribute to a good start Christina Bergqvist and Anita Nyberg 291 in life for children from all social backgrounds, as well as supporting parents, especially mothers, in managing the tension between paid work and caregiving responsibilities. The proportion of public resources spent on child care can thus be seen as an indication of the extent of the state's commitment to the two-breadwinner models and to equal opportunities for parents and children from different backgrounds. The most common way of measuring the extensiveness of the public sector is the cost of social services as a share of gross domestic product (GDP). In this respect, public child care has become an increasingly important sector in the Swedish economy. The contribution of public child care to GDP increased from 0.2 percent around 1970 to 1.68 percent in 1980 and 2.4 percent in 1990. This contribution is larger than that of agriculture and comparable in size to the chemical industry (Kjulin 1995 III, 2; Edebalk, Stahlberg, andWadensjö 1998, 134; SCB 2000).This reflects the high priority attached to public support for child care, at least until the beginning of the 1990s when Sweden entered a severe economic crisis. Given the state of the economy, the unemployment level, the increasing social expenditure, and the worsening of the public finances one might have expected calls for cost cutting and retrenchments in public child care. The costs of child care as a Figure 12.1: Costs of Public Child Care as a Share of GDP, Gross Costs for Public Child Care, Number of Children in Public Child Care, and Cost per Child (Index 1991 = 100) *Costs per hour (1991-1995) and costs per child full-time (1996-1998) in public child care centers. Unfortunately, due to changes in the statistics in 1998 it is not possible to make comparisons with the years after 1997 in the other measurements. Until 1997 preschool covered one- to six-year olds, in 1998 one- to five-year olds. SOURCES: SOU (2000 3, 115); SCB (2000); Skolverket (1999a, diagram 5). 292 Welfare State Restructuring and Child Care in Sweden share of GDP did decrease somewhat between 1991 and 1997 (see fig. 12.1). This provides, however, a rather crude measure of public commitment to child care provision, for it is affected by changes in the GDP as well as resources devoted to child care. If instead we look at the development of total gross costs in fixed prices, we find that the amount was the same in 1997 as it was in 1991, with a dip in between. This suggests that public child care remained a high priority. When we take the number of children in child care into consideration, we get a somewhat different picture. Around 188,000 more children were in public child care in 1998 than in 1990. In other words, fewer resources were spent per child. In fixed prices the decrease was about 14 percent per child between 1991 and 1998 (SOU 2000: 3, 113ff). Moreover, parents have been contributing a rising share of the costs. In 1990 parents paid 10 percent of the total gross costs of child care in direct child care fees but by 1998 this proportion had increased to 17 percent. As the total gross costs for public child care remained the same, the total amount, as well as the amount per child, of "public" money spent on child care has diminished. This development should not, however, be seen as a result of changed political priorities and restructuring, but as a "temporary" retrenchment. Harsh economic circumstances combined with increased demand for child care as a result of a high rate of fertility increased the strains on the system. Toward the end of the decade, the economy improved while fertility rates fell.7 The resources used for child care—independently of how it is measured—increased again. Due to new reforms, described in the next section, more resources will in the future be dedicated to the child care sector. Public Child Care and Universalism Social rights and benefits are usually classified as universal, labor market related, or income/needs tested. Universalism in a strict sense of the term means that all citizens or individuals in a specific category (e.g., a particular age group) get the same bene- Table 12.2: Demand and Supply of Public Child Care 1980-1996. Numbers in Thousands and Percent of All Children. Preschool-age Children (3 Months to Six Years) YEAR DEMAND SUPPLY DIFFERENCE PERCENT OF PERCENT OF PERCENT OF NUMBER ALL CHILDREN NUMBER ALL CHILDREN NUMBER ALL CHILDREN 1980 348 49 211 30 137 19 1986 374 58 299 47 75 12 1990 400 57 337 48 63 9 1993 443 56 386 49 57 7 1996 494 62 444 55 50 7 SOURCES: SCB (1994 and 1997). Christina Bergqvist and Anita Nyberg 293 fits and same rights without any restrictions based on need or labor-market status. The flat-rate child allowance for all children up to fifteen years of age is an example of a genuinely universal benefit and primary and secondary education are provided as a universal social service. As a concept, universalism is not always used in this way. In the Scandinavian context it is often used for social insurance and services that are widely available and apply to large segments of the population or all wage earners (Anttonen and Sipilä 1996). From a comparative perspective, Swedish policies can perhaps be considered universal, but many programs have indeed been tied to employment status (Clayton and Pontusson 1998). More specifically, the main social programs usually consist of a flat-rate universal component, based on citizenship or residence, and a more generous component dependent on one's labor-market participation. In the case of parental leave insurance this means that citizens and permanent residents who become mothers or fathers are entitled to parental leave for the same length of time and everyone is entitled to a low flat-rate benefit. When one has been employed during the last eight months before the child is born, one receives the more generous income-related benefit. As long as almost everyone—including women—is employed, the vast majority of citizens are eligible for the richer benefits. When unemployment grows, however, as it did during the 1990s, a growing number of citizens are forced to rely on the flat-rate benefit (SOU 2000: 3). As we have seen, the right to child care has never been universal in the same sense as education. Except for some hours of preschool for six-year-olds, the right to public child care has in practice been restricted to children whose parents are in paid work, or to children with special needs. That is, labor-market participation or need have formed the basis for eligibility to publicly financed child care. Moreover, supply has never matched demand, so even employed parents have been unable to find the child care they need, as table 12.2 shows. In times of cuts in public spending, moreover, one might expect a tightening of eligibility rules as one way to save money. This has not, however, been the route taken in Sweden. In fact, more children than ever had access to child care institutions in the 1990s. As can be seen from table 12.2, both demand and supply increased substantially between 1980 and 1996. In 1980, the gap was 19 percentage Table 12.3: Children Enrolled in Publicly Subsidized Child Care Centers, PRESCHOOLS, AND FAMILY DAY CARE UNITS, 1990-1997 (PERCENT OF THE POPULATION BY AGE) 1-2 YEARS OLD 3-6 YEARS OLD 1-6 YEARS OLD IN TOTAL 44 64 57 46 65 59 48 70 63 57 76 70 59 78 73 1990 1992 1994 1996 1997 SOURCE: Skolverket (1998). 294 Welfare State Restructuring and Child Care in Sweden points but by 1995, when the new Child Care Act came into effect, it had shrunk to 7. In that same year, S 5 percent of all preschool children had a place in publicly financed child care. Of the remaining 45 percent, parents on parental leave took care of a large majority of the children. Overall, as table 12.3 indicates, the percentage of children from one to six increased considerably from 1990 to 1997. Contrary to what might have been expected, then, universalism has been strengthened. This process of universalization and integration of child care and education reflects, in part, the commitment to "life-long learning." The goal of public child care is not only to serve working parents but also to educate and support the development of preschool children. This is, of course, not possible if eligibility is tied to the parent's status on the labor market. Of particular concern here are the children of families where the parents are unemployed. These children do not get the same support and early childhood education as those whose parents work and who therefore are eligible for a place in public child care.8 As a step toward a more universal preschool system, responsibility for child care at the national level was moved from the Ministry of Health and Social Affairs to the Ministry of Education and Science in July 1996, with the aim of strengthening the pedagogical profile of child care. This inaugurated a process of integrating child care and school into the same legislation. In the public documents the term child care center (daghem) has been replaced by references to preschool and preschool activities.9 Legislation for the whole child care sector has been brought into the School Act and the National Agency for Education has the supervisory responsibility. Thus, contrary to what might have been expected in times of economic restraint, on the aggregate level universalism has been strengthened. Supply has increased, but so has demand. In 1998, 95 percent of the municipalities could offer a place within three or four months after application (Socialdepartmentet 1999: 53, 26). In addition, the social democratic government has put on the agenda new reforms that move child care further in the direction of universalism. These include child care for the children of the unemployed, a universal and free preschool for all four- and five-year-olds, and the imposition, by the national government, of a maximum parental fee (Socialdepartmentet 1999, 53; Proposition 1999/2000, 129). In November 2000 the Swedish Parliament decided to implement the suggested reforms during the coming three years. The Social Democratic Party got support from the Left Party and the Green Party for these reforms. The four bourgeois parties in opposition did not support the reforms and instead suggested a lump sum (child care account) to all parents with preschool children, and the right to tax deductions of child care costs. The Quality of Public Child Care Cuts in public costs in child care have not been translated into fewer children in child care, which might suggest that quality has deteriorated. Quality in child care is not easy to measure but changes in child/staff ratios and group size provide a rough Christina Bergqvist and Anita Nyberg 295 indication. As figure 12.1 shows, the number of children in child care has expanded while gross costs and costs per child initially decreased, then rose somewhat during the latter part of the 1990s. This is reflected in the development of the child/staff ratio, with a lag. Until 1991, the number of children and the number of staff increased at about the same pace. Thereafter the number of children rose, but not the number of staff. While staff-child ratios were the same in 1980 and 1990 (see table 12.4), after that the number of children per staff rose rapidly such that by 1998 each staff member was taking care of 1.5 children more than in 1990. Also, if quality is measured by group size, we find that quality has deteriorated. In 1990 the average group size was 13.8 children but by 1998 it was 16.5 children. There was, however, some improvement after 1997 (SOU 2000: 3, 116; Skolverket 2000a, 18).,0 More children per group and per adult can be interpreted as a deterioration of the quality in child care, but it can also be associated with an increase in productivity. That is, it is possible that the municipalities are producing as high quality child care with fewer resources. It is difficult to assess whether lower costs mean lower quality or higher productivity. The large variation in child care costs among municipalities can be seen as evidence that productivity gains are possible." It does not seem unreasonable to suggest that decentralization and increased coordination between child care and school have increased productivity. That the municipalities have increasingly substituted family day care for child care centers and, within the centers, child minders for preschool teachers, might have raised productivity.12 There are, however, no systematic studies of how quality is affected by changes in resources over time (Svenska Kommunförbundet 1998). It is hard to give a straightforward answer on the question of changes in the quality of Swedish public child care over the last decade. However, a recent investigation concerning the quality of nine different sectors in the economy found that child care was rated highest while local traffic companies and insurance services were rated lowest (Konsumentverket 2000). Table 12.4: The Development of the Average Number of Children per Annual Worker in Child Care Centers YEAR no. of chile )REN/ANN 1980 4.2 198S 4.3 1990 4.2 1992 4.9 1994 S.2 1996 5.5 1998 S.7 1999 S.4 SOURCES: Välfárds Bulletinen (1994); Skolverket (1999a); Skolverket (2000a). 296 Welfare State Restructuring and Child Gare in Sweden Also, an Organization for Economic Cooperation and Development (OECD) evaluation of public child care in various countries notes that Swedish staff have been under increased pressure to provide services to greater number of children (1999, 38). At the same time, the report found the preschool system outstanding and waxed eloquent about the Swedish child care system, noting that "it is said that the merit of any nation may be judged by how it treats its children—particularly the poor and needy. If that adage is true, then Sweden surely sits at an international pinnacle. Nothing honors Sweden more than the way it honors and respects its young" (OECD 1999,43). Privatization of Child Care One dimension of variation between different welfare-state regimes is the role of the market in relation to the public sector. The characteristic of the Swedish welfare state has been to offer "high-quality public services that obviated the need for supplementary private solutions" (Mahon 1997, 385). During the last decades of the twentieth century, a political trend towards privatization has swept over the world. This has not left Sweden untouched, and this is visible in the production of social services such as child care. Private solutions can be of different kinds. Child care can be purchased from profit-making enterprises or nonprofit organizations or it can by provided by the family. The implications of privatization will be different depending on whether privatization means expanding the role of commercial enterprises, nonprofit organizations, and/or the family. During the 1990s, the bourgeois government introduced a care allowance as a way of trying to privatize to the family.'3 Upon their return to office, the Social Democrats abolished the care allowance. Privatization to for-profit or nonprofit organizations was more successful. Such privatization can be divided into three aspects: provision, financing, and regulation. Swedish child care today is, even when it is "private," publicly regulated, and financed. Private—meaning not provided by the municipality—child care is nothing new in Sweden. Private child care centers have existed for a long time.14 The most common form of private child care in Sweden, as in other countries, until the 1990s was privately arranged (and paid for), unregistered family child care. In the 1980s, nonmunicipal child care centers became eligible for public subsidies, as long as they met certain requirements. The center had to be run by a nonprofit organization, such as a parents' cooperative, or offer a special form of pedagogy or other similar grounds. In 1991, when the bourgeois parties were in power, the law was changed to include child care centers run by personnel cooperatives, the Swedish church, and for-profit organizations. Parental fees had to be kept on a "reasonable" level, so as not to diverge too much from the parental fees in the municipally run child care (Nyberg2000). Christina Bergqvist and Anita Nyberg 297 In 1990, there were some privately run, but publicly regulated and financed, child care centers. The proportion of children in private nonprofit and commercial child care centers rose from S percent in 1990 to IS percent in 1999 (SOU 2000, 3; Skolverket 2000a). Most of the private child care centers are run by nonprofit organizations. The most common form is parental cooperatives. In 1998, parental cooperative child care centers provided places for 47 percent of all children in private child care centers. This form of child care offered in the 1970s and 1980s parents who could not get a place in a public child care center a chance to bypass the often long queue for public child care (Antman 1996, ISO). In this form of child care, parents employ the personnel and often themselves participate on a rotating basis in the work at the child care center. Cooperatives run by staff and voluntary associations each accounted for an additional 10 percent of children in privately run child care centers (Skolverket 1999a, 21, table 1.2). The ideological difference between social democrats and the bourgeois parties has not concerned child care centers run by nonprofit organizations such as parental cooperatives, but for-profit, company-run child care centers. The bourgeois parties opened the way for public subsidies to for-profit child care centers at the beginning of the 1990s. The Social Democrats did not reverse this decision when they came back into power. In 1998 commercial child care accounted for around a quarter of the privately run child care centers. Another way of privatizing child care is to increase and restructure the fee system. During the 1990s a majority of the municipalities raised the level of fees and, to a greater extent, relied on time- and income-related fees to control demand and keep costs down.The difference between different municipalities is substantial.15 The general impression is that the private sector in child care has grown. This can, however, be contested. The proportion of children in private child care of all children in child care is smaller today (16 percent) than in the 1980s (40 percent).16 Private child care is also less "private" today since it is now publicly regulated and financed. In other words, the big difference between the 1980s and today is that the need for unregulated child care, organized and financed privately by the parents, has been crowded out by publicly financed child care. The great majority of child care is today run by the municipalities. At the same time, the way has been opened for public subsidies to nonprofit and for-profit child care. More or Less EgalitarianismP According to Esping-Andersen, "The social democratic model and egalitarianism have become basically synonymous. To many, the egalitarian element is simply the practice of universalism: everybody enjoys the same rights and benefits, whether rich or poor. To others, it refers to the active promotion of well-being and life chances—perhaps no more evident than for women. Still others equate egalitarian- 298 Welfare State Restructuring and Child Care in Sweden ism with redistribution and the elimination of poverty" (1999, 80). Public child care has contained all three of these elements. Here we examine how the increase in number of places in publicly financed and regulated child care has affected the distribution of public child care among different groups in society. Lone and Cohabiting Mothers It could be argued that the aim of early child care centers in Sweden was to enhance women's "capacity to form and maintain an autonomous household," to use the phrase of Ann Orloff (1993). The concern that women would be able to support themselves and their children, had little, however, to do with gender equality at that time. Child care was seen as a way to alleviate poverty by making it possible for poor mothers to work for pay and to be self-sufficient. Lone mothers have long had access to public child care to a greater extent than cohabiting mothers have. In 1966, 46 percent of the children in child care centers were children of lone mothers, which is well above their proportion of the population. From the 1970s through the 1990s, this proportion decreased. In 1994 only 16 percent of the children in public child care were children of lone mothers. This reflects the very substantial increase in the number of children of cohabiting/married mothers in public child care rather than a decline in access for lone parents. Access has, in fact, improved for the children of lone parents, too. In 1975, 56 percent of the preschool children of lone mothers were in public child care, but by 1994, 74 percent had places (Antman 1996, 148). Lone parents working full-time normally have their children in publicly subsidized child care and they make use of it—as always—to a higher degree than cohabiting parents (Prop. 1999/2000, l).17 In 1996, 83 percent of cohabiting, employed parents had their children in public child care as compared to 91 percent of lone parents (SCB 1997, Översiktstabell 6). Figure 12.2: Proportion of Children Two to Six Years of Age in Public Child Care, Sweden 100 90 80 70 60 50 40 -I 30 20 10 0 79 59 m4 82/83 90/91 82 72 63......... 94/95 ■ LO B TCO D SACO SOURCE: LO (1996). Christina Bergqvist and Anita Njfberg 299 Workers and the Professional Glasses In the first half of the twentieth century, full-time child care was clearly linked to poverty relief, and child care centers did not enjoy a good reputation as long as their main role was to mind the children while their mothers, of necessity, worked. When their role was expanded to stimulate and educate the children too, middle-class parents became interested. This was reflected in a rapid increase in the share of middle-class children and a concomitant decrease in the share of working-class children among the children in public child care in the 1960s and 1970s. By the beginning of the 1980s, 18 percent of the children of parents organized by unions belonging to the blue-collar umbrella organization, LO (SwedishTrade Union Federation), had a place in a public child care center. The corresponding share was 26 percent among children of white-collar workers belonging toTCO (Swedish Confederation of Professional Employees), the largest umbrella organization for white-collar workers, and 43 percent of preschool children of professionals, organized by SACO (Swedish Confederation of Professional Associations).18 If public family child care is included, the proportion was 44 percent of LO children, 59 percent ofTCO children, and 79 percent of SACO children (see Figure 12.2). In the 1980s, part of the controversy surrounding child care policy stemmed from the fact that not all children of appropriate ages had placements in public child care, either because they did not demand it or because they were locked out due to rationing of scarce spaces. The heavy subsidization of some, but not all, families with young children can be regarded as unfair. To make matters worse for a Social Democratic government, the children of high-income parents were the main beneficiaries. Contrary to the original intention of Social Democrats and their blue-collar union partners, public child care was contributing to enhanced differences between parents and children from different classes. As public child care expanded in the 1990s, however, the proportion of LO children with access to public child care increased faster than the other two categories. By 1994/5, the distance between the groups had appreciably diminished, as figure 12.2 shows. The main reason for the difference between blue-collar and white-collar and professional workers' access to child care is the difference in mothers' working time. Public child care is primarily used by families where both members work for pay on a full-time basis. Today almost all children of parents working full-time have a place in child care and the difference between those with higher and lower education levels is small (98 percent for higher educated, full-time working parents versus 87 percent for those with the lowest levels of education). The same picture emerges when the socioeconomic status of parents is examined. Around 97 percent of all children of higher-paid employees have a place in public child care, compared to 91 percent of children of full-time blue-collar workers (Prop. 1999/2000, 1 Bilaga 1). In the 1990s, however, attitudes toward part-time child care began to change. Child care centers were originally thought of as full-time care. Families who did not need full-time child care were not given the same priority. The municipalities did not 300 Welfare State Restructuring and Child Care in Sweden want part-time children in child care since they reduced state grants and the families themselves often avoided child care since the fee was high relative to the time needed. The grant system was reorganized at the end of the 1980s, however, and in many municipalities, more flexible fee schedules were introduced (Antman 1996). The Unemployed and Immigrants In Sweden, egalitarianism has traditionally been discussed in terms of class, and later, gender. Today the concern to promote equality has widened to include employment status and ethnic background. The two are related. When parents are working outside the home, children of immigrants are as likely to have a place in public child care as other children (Ministry of Education and Science 1999, 43; So-cialdepartmentet 1999, S3, 29). Immigrants, however, today experience a higher rate of unemployment than Swedes and a greater share work part-time. A recent study shows that in 40 percent of the municipalities, children lose their place in child care if a parent loses her job. Another 49 percent offer child care for a limited number of hours to children with unemployed parents and a few municipalities offer separate, short-term child care to unemployed parents to assist them in their job-seeking efforts. This means that it is already vulnerable groups are shut out of public child care. It should be observed that the number of children with unemployed parents today is a bigger group than the number of children with an "at-home mother." The Social Democrats have, however, proposed that the children of the unemployed be given a right to a place in public preschool. Differences between Municipalities In the 1980s, under the mantle of decentralization and democratization, new arrangements began to be worked out between the state and the municipalities designed to allow the latter greater latitude in adapting national legislation to local conditions. Some regarded this as a way of introducing more of a market orientation into the public sector, while others spoke about user influence and power of the citizens. During the 1990s, decision-making power was increasingly transferred from the state to the municipalities. This has led to a situation where child care, which once operated according to the same rules across the country, can vary substantially across municipalities.19 The question is, does child care in municipalities run by bourgeois governments differ from that provided by social democratic municipalities? Is public child care more universal, egalitarian and less oriented towards private provisioning in social democratic municipalities than in those where the bourgeois parties are in majority? There are indications that this is the case. Universalism and egalitarianism seems to be a more common aim in Social Democratic municipalities than in bourgeois municipalities. An example of this is that, in 1998, the share of municipalities where the children of unemployed parents could keep a place in public child care, was considerably higher in municipalities with a Social Democratic/Left Party majority (58 Christina Bergqvist and Anita Nyberg 301 percent) than in municipalities with a bourgeois majority (39 percent) (Skolverket 2000b, 9). Private child care—and, it is argued, freedom of choice—is higher in bourgeois than in socialist municipalities. In 68 percent of the Social Democratic municipalities, as compared to 86 percent of the bourgeois, there was an alternative to municipal child care (estimated from data from Skolverket 2000a).20 The partisan cast of government is also connected to the size of parents' fees and their tax levels. Municipalities with a socialist majority on average have higher local taxes than municipalities with bourgeois majority and are more likely to have lower parental fees than municipalities with bourgeois majority. The connection is strongest with regard to single parents' families, which is the family type with the lowest incomes. Of the quarter of municipalities with the lowest fees on average for single parents, 64 percent had a socialist majority and only 7 percent bourgeois majority. Of the quarter with the highest fees, 56 percent had bourgeois majority and 17 percent socialist majority (Skolverket 1999b, 10). On average, staff ratios tend to be somewhat lower in bourgeois governed municipalities than socialist municipalities (Asker and Kehnberg 1999). A maximum fee in publicly financed child care can be seen as a strategy for the Social Democratic government to retain some of the power from the municipalities. Conclusions As the 1990s unfolded, some were prepared to declare the Swedish model of the welfare state dead. Comparative welfare research has however found that most social programs enjoy a surprising durability (Pierson 1996; Stephens 1996). Our analysis of Swedish child care suggests that Sweden is no exception in this respect. This does not mean that no restructuring has taken place. We found that restructuring has, on the one hand, served to strengthen several of the core features of the Swedish model. This becomes evident if one takes into consideration the decisions taken in 2000 about a maximum fee, the right of children of the unemployed and of four- and five-year-olds to preschool, proposed at the end of the 1990s. On the other hand, new features have also been introduced. Certainly there was a modest reduction in the resources devoted to child care on a per-child basis in the 1990s, yet this is partly the result of changes in the number of children demanding child care. A "mini baby boom" around 1990 temporarily increased the demand for spaces. This also happened at a time when municipalities were trying to cope with sharper fiscal constraints while meeting an obligation to provide spaces for all children who need it, within a reasonable amount of time. This development probably affected the quality of the care, as each child care worker has more children to care for. There is, however, also some evidence of gains in productivity. The decisions taken in 2000 also mean that more economic resources will be devoted to publicly financed child care in the future. 302 Welfare State Restructuring and Child Care in Sweden The 1990s have witnessed a general trend toward a decentralization of responsibilities from the state to the local level and this has affected child care. State grants for child care and other social services now come in the form of block grants. In addition, regulation by central state authorities has become less detailed. As the system becomes more decentralized, it has also become more differentiated between municipalities concerning things like rules of eligibility, parental fees, and the share of private child care. At the same time, however, the 1995 Act on Child Care clarified the municipalities' obligation to provide child care for all who need it. Another new element is a move toward greater privatization. The number of children in child care run by for-profit organizations remains small, yet the basic principle of private service delivery—within the framework of public financing and public regulations—has been accepted. In addition, the basic rationale of trying to improve performance by exposing public service production to market disciplines has become a key part of public management reforms, including the provision of child care. The idea is that by strengthening market elements, gains in economy and efficiency will be made at the same time as higher-quality service and greater choice can be offered. At the same time, on aggregate, inequalities between different social groups have not increased. Today public child care provides places for a larger share of preschool children than it did in 1990. Those who earlier found it harder to get access to child care—particularly blue-collar workers—have found it easier to do so, reducing the "class gap" highlighted in the 1980s debates. The norm today is that children are in publicly financed child care except when parents are on parental leave. Nevertheless, the right to a place in public child care does not apply to children of unemployed parents in about half the municipalities. This has become an important issue as unemployment affected more people and as it became apparent that immigrants constitute a disproportionate share of the unemployed. However, from the year 2001 on, children of unemployed parents will have a right to at least three hours per day or fifteen hours per week of preschool. In the 1990s we saw a higher degree of diversity at the local level, but during the second half of the decade there was also a trend toward a more universalistic system. This could be seen in the shift from a focus on child care as a service for working parents and children in need to a right for all preschool children to preschool education. Responsibility for child care has been transferred from the Ministry of Social Affairs to the Ministry of Education. In line with this, the Social Democratic government, with the help of the Green Party and the Left Party, have decided to create a universal (but not obligatory) preschool for all four- and five-year olds. In the debate over the fate of the traditional Swedish welfare state some would argue that globalization and neoliberalism has put Sweden under pressure to reduce social standards and that profound retrenchment and restructuring is taking place. Others counter that the social system shows a high degree of path dependency and that the traditional Swedish welfare state still is strong. Our conclusion is that it is not a question of either I or but one of continuity ani change—or, perhaps, continu- Christina Bergqvist and Anita Nyberg 303 ous change. Generous public spending, high quality in social services, universalism and egalitarianism are retained, while change is seen in increased marketization and privatization. Notes 1 .The employment rate was 85 percent for men and 81 percent for women in 1990 and 75 percent for men and 71 percent for women in 1999. The unemployment rate was 1.7 percent for men and 1.6 percent for women in 1990 and 5.9 percent for men and 5.2 percent for women in 1999 (SCB 1999). 2. In 1990 the central government budget showed a surplus of almost SEK19 billion. In 1993 the deficit amounted to almost SEK210 billion and in 1994 close to SEK200 billion. In 1998 there was again a surplus, this time slightly more than SEK20 billion (SOU 2000: 3, 40). 3. We use the term child care in a broad sense, including parental leave and care allowances as well as child care institutions and the like. The child care system can be seen as comprising a broad variety of different resources for caring for children. In this chapter, however, we mainly analyze child care institutions and preschools. 4. Children used to start regular school at age seven, but today it is also possible to start at six. 5. The care allowance was given to all children between ages one and three, and the amount was 2000 Swedish crowns a month, which is far from sufficient to be able to support oneself. In practice the reform meant that the ninety flat-rate days in the parental leave were replaced with a care allowance lasting until the child was three (Lag 1994: 553, om rätt till várdnadsbidrag [Law 1994: 553, concerning the right to a care allowance]). 6. Public child care included publicly regulated, financed, and provisioned child care, but also publicly regulated and financed, but privately provided, child care centers. 7.The total fertility rates were: 1988: 1.961; 1990: 2.137; 1992: 2.090; 1994: 1.890; 1996: 1.606, and 1998: 1.504 (Statistisk Ärsbok 2000, table 68). 8. There are no national rules for how the children of unemployed parents should be treated, and regional differences are substantial. Some municipalities offer child care for children of the unemployed but others do not. These differences will be analyzed later. 9. Preschool activities include what is now called preschool (fórskola), which is part-time, and child care centers, family day care homes tfamiljedaghem), and open preschools, where parents can go together with their preschool children if the children are not attending any of the other activities. 10. The staff in Swedish public preschool child care fall mainly into two categories: preschool teachers and qualified child minders. Almost 98 percent of personnel working in preschool child care are trained to work with children: 60 percent of staff are university-trained preschool teachers, while the remaining staff are qualified child minders (Oberhue-mer and Ulich 1997). In family day care services, over 70 percent of day care mothers have been trained to work with children, having either earned a children's nurse certificate or 304 Welfare State Restructuring and Child Care in Sweden participated in special training provided by the municipality (Ministry of Education and Science 1999, 33). Up until 1990, child minders predominated in public child care, but today university-trained preschool teachers form the biggest category (VälfärdsBulletinen 6, 1994; Skolverket 1998). 11. In eight of ten municipalities, the cost of care per full-time child varies between SEK84,400 and 129,900, a difference of SO percent (Skolverket 2000a, 18). Various studies report productivity increases in public child care centers (ESO 1988, 1994; Socialstyrelsen 1997; Svenska Kommunförbundet 1998). 12. The number of children in family day care decreased throughout the 1990s and continues to do so. Since 1990 the number of children has been halved. In 1999 11 percent of all children between one and five were in family day care (Skolverket 2000c, 12). 13. There has also been discussion about subsidized domestic services in the home similar to those in France, but no such policy has as yet been adopted (Nyberg 1999). 14. The highest proportion of private child care centers was found in the first half of the twentieth century, when it reached almost 100 percent. The highest numbers before the 1990s were found in the 1950s. In the 1960s and 1970s, however, private child care centers almost disappeared altogether. One important reason for this was changes in the rules for state grants. Of almost 9,000 child care centers in 1981 only sixty-four—less than 1 percent—were run by nonmunicipal organizations (Nyberg 2000). 15. A family with two children and an average income can pay as little as SEK 1,300 and as much as 3,400 per month in different municipalities (Socialdepartementet 1999: 53,4). These differences will diminish greatly with the introduction of a maximum fee to be implemented in 2002. 16. Calculated from SCB (1994 and 1997). 17. Traditionally the employment of single mothers has been higher than cohabiting mothers. Today, however, single mothers have a higher unemployment rate than married/cohabiting mothers. Single mothers have greater difficulties supporting themselves through paid work, and they have lower earnings than cohabiting mothers (Nyberg 1997). 18.The categorization is based on the union to which the interviewed parent belonged. It should be pointed out that a very high proportion of the employees in Sweden are members of unions. 19. In 114 municipalities, the Social Democrats/Left Party is in the majority; in 92 the bourgeois parties are in majority and in 83 there is no majority. 20. Sixty-four out of 289 municipalities only had municipally-run centers. References Antman, Peter. 1996. Barn och äldreomsorg í Tyskland och Sverige.Välfdrdsprojektet. Sverigedelen (Child Care and Elder Care in Germany and Sweden. The Welfare Project: Sweden). Stockholm: Socialdepartementet. Anttonen, Anneli and Sipilä, Jorma. 1996. "European Social Care Services: Is It Possible to Identify Models?" Jburnai of European Social Policy 6, no. 2: 87—100. Christina Bergqvist and Anita Nyberg 305 Asker, Anna, and Ola Kehnberg. 1999. "AUtfärre vuxna i skula och daghem" (Fewer and Fewer Adults in Schools and Child Care Centers). Svenska Dagbladet (Swedish Daily), January 14: 8. Bengtsson, Hans. 1995. Förskolereformen. En studie i implementering av svensk välfärdspolitik 1985—1991 (Preschool Reform: A Study of the Implementation of Swedish Welfare Policy 1985-1991). Lund: Political Studies. Bergqvist, Christina. 1999. "Child Care and Parental Leave Models." In Equal Democracies? Gender and Politics in the Nordic Countries, ed. Christina Bergqvist, Anette Borchorst, Ann-Dorte Christensen, Viveca Ramstedt-Silén, Nina C. Raaum, and Audur Styrkásdótti. Oslo: Scandinavian University Press. Bergqvist, Christina, Jaana Kuusipalo, and Audur Styrkarsdóttir. 1999. "The Debate on Child Care Policies." In Bergqvist et al., eds. Equal Democracies? Borchorst, Annette. 1994. "The Scandinavian Welfare States—Patriarchal, Gender Neutral or Woman-Friendly?" International Journal of Contemporary Sociology 31, no. 1: 45—67. Clayton, Richard, and Jonas Pontusson. 1998. "Welfare-State Retrenchment Revisited: Entitlement Cuts, Public Sector Restructuring and InegalitarianTrends in Advanced Capitalist Societies," World Politics 51: 67-98. Edebalk, Per Gunnar, Ann-Charlotte Stáhlberg, and Eskil Wadensjö. 1998. So-cialfórsäkringarna: ett samhälisekonomiskt perspektiv (Social Security: A Social Economic Perspective). Stockholm: SNS. ESO (Expert Group on Public Finance). 1988. Kvalitetsutvecklingen inom den kommunala barnomsorgen: rapport till ESO—Expertgruppen fór studier i offentlig ekonomi (The Development of Quality in Public Child Care). (Ds Fi 1988:1). Stockholm: Allmänna förlaget. __________. 1994. Den offentliga sektorns produktivitetsutveckling 1980—1992: rapport till Expertgruppen fór studier i offentlig ekonomi (The Development of Productivity in the Public Sector 1980-1992). (Ds 1994: 24). Stockholm: Fritzse. Esping-Andersen, Gosta. 1990. TheThree Worlds of Welfare Capitalism. Cambridge: Polity Press. __________. 1999. Social Foundations of Postindustrial Economies. Oxford: Oxford University Press. Esping-Andersen, Gosta, and Walter Korpi. 1987. "From Poor Relief to Institutional Welfare States." In The Scandinavian Modeh.Welfare States andWelfare Researcy, ed. Robert Erikson et al. New York: Sharpe. Gustafsson, Rolf A, and Peter Antman. 1996. "Välfärd pá entreprenad" (Tenders onWel-fare). In Generell välfdrd. Hot och möjligheter? (Universal Welfare: Threat or Possibilities?), ed. Joakim Palme and Irene Wennemo. Stockholm: Socialdepartementet. Hernes, Helga. 1987. Welfare State and Women Power. Olso: Norwegian University Press. Hirdman, Yvonne. 1998. Med kluven tunga (With Forked Tongue). Stockholm: Atlas. Johansson, Susanne. 2000. Genusstrukturer och lokala välfárdsmodeller. Fyra kommuner mater om-vandlingen av den offentliga sektorn (Gender Structures and Local Welfare Models: Four Municipalities and the Transition of the Public Sector). Regionstudier nr 40. Uppsala: Uppsala Universitet. Kjulin, Urban. 1995. Economic Perspectives on Child Care. Nr 56 avhandling (dissertation). 306 Welfare State Restructuring and Child Gare in Sweden Ekonomiska studier utgivna av Nationalekonomiska Institutionen (Economic Studies published by Economics Dept.). Göteborg: Handelshögskolan vid Göteborgs univer- sitet. Konsumentverket. 2000. Tjänste bra? Markriad, kvalitet och klagomal—konsumenter om nio tjän- stebranscher (Is It Good? The Market, Quality and Complaints—Consumers about Nine Service Sectors). Rapport 2000: 20. Stockholm: Konsumentverket. Leira, Arnlaug. 1993. "The Woman-Friendly Welfare State? The Case of Norway and Sweden." In Women and Social Policies in Europe, ed. Jane Lewis. Aldershot, Eng.: Edward Elgar. LO. 1996. Barnomsorg, fórvärvsarbete och jämställdhet. Ett Jaktamaterial omfórhůllandena vid 1990-talets mitt (Child Care, Employment and Gender Equality: A Fact Sheet about the Situation in the 1990s). Nr 29. Stockholm: LO. Mahon, Rianne. 1997. "Child Care in Canada and Sweden: Policy and Politics." Social Politics 4, no. 3:382^-18. Ministry of Education and Science. 1999. Early Childhood Education and Care Policy in Sweden. Background report prepared for the OECD Thematic Review of Early Childhood Education and Care Policy. Stockholm: Ministry of Education and Science. Nyberg, Anita. 1997. Women, Men and Incomes. Gender Equality and Economic Independence, SOU 1997:87. Stockholm: Fritzes. __________.1999. "Hemnära tjänster—kvinnornas befriare?" (Domestic Services—The Liberator ofWomen?) Kvinnovetenskaplig tidsknft (Women's Studies Journal) 3: 31-48. __________. 2000. "From Foster Mothers to Child Care Centers: A History of Working Mothers and Child Care in Sweden." Feminist Economics 1, no. 1: 5—20. Oberhuemer, Pamela, and Michaela Ulich. 1997. Working withYoung Children in Europe: Provision and Staff Training. London: Paul Chapman. OECD. 1999. OECD Country Note: Early Childhood Education and Care Policy in Sweden. Paris: OECD. Orloff, Ann. 1993. "Gender and the Social Rights of Citizenship." American Sociological Review 58: 303-28. Pateman, Carol. 1988. "The Patriarchal Welfare State." In Democracy and the Welfare State, ed. Amy Gutmann. Princeton, N.J.: University Press. Pierson, Paul. 1996. "The New Politics of the Welfare State." World Politics 48:143-179. Prop. 1999/2000: 1 Bilaga 1. Fördelningspolitisk redogörelse (A Report of the Distribution). Stockholm: Inansdepartementet. Prop. 1999/2000: 129. Maxtaxa och allmänfórskola m.m. (A Maximum Fee and Universal Preschool, etc.). Stockholm: Inansdepartementet. Sainsbury, Diane, ed. 1994. Gendering Welfare States. London: Sage. __________. 1999. Gender and Welfare State Regimes. Oxford: Oxford University Press. SCB (Statistiska Centralbyrán—Statistics Sweden). 1990, 1999. AKU Arbetskraftsunder- sökningarna (Labor Force Surveys). Stockholm: SCB. __________. 1994. Barnomsorgsundersökningen 1994 (The Child Care Investigation 1994). Stockholm. Christina Bergqvist and Anita Nyberg 307 __________. 1997. Barnomsorgsundersökningen, Hosten 1996 (The Child Care Investigation, Fall 1996). Stockholm: Heia Riket. __________. 2000. Kommunernas finanser, Nationalräkenskaper (The Finances of the Municipalities, National Accounts). Örebro: SCB. Siirn, Birte. 1997. "Dilemmas of Citizenship in Denmark: Lone Mothers between Work and Care." In Lone Mothers in European Welfare Regimes, ed. Jane Lewis. London: Jessica Kings-ley Publishers. Skolverket. 1998. Barnomsorgen i siffror (Child Care in Numbers). Rapport nr 152. Stockholm: Skolverket. __________. 1999a. Barnomsorg och škola i sffror 1999. Del 2—Barn personal, elever och ľdrare (Child Care and Schools in Numbers 1999. Part 2—Children, Personnel and Teachers). Rapport nr. 167. Stockholm: Skolverket. __________. 1999b. Avgifter i fórskola ochfritidshem 1999 (Parental Fees in Preschools and Leisure Homes). Rapport nr. 174. Stockholm: Skolverket. __________. 2000a. Barnomsorg och škola. Jämfórelsetalfór huvudmdn. Resurser (Child Care and School. Comparisons between Providers). Rapport nr. 172. Stockholm: Skolverket. __________. 2000b. Skolverkets ůrsredovisningfór budgetdret 1999 (Annual Report of the Board of Education 1999). Stockholm: Skolverket. __________. 2000c. Barnomsorg och škola, jämfórelsetal fór huvudmän. Del 1 Organization—Personal—Resultat (Child Care and School: Comparative Numbers for Providers. Part 1: Organization, Personnel—Results). Rapport nr 183. Stockholm: Skolverket. Socialdepartmentet. 1999. Ds 1999: 53. Yalfrihetsrevolutionen i praktiken (The Revolution of Choice in Practice). Stockholm: Socialdepartementet. Socialstyrelsen. 1997. Social service, vard och omsorg i Sverige 1996 (Social Service and Care in Sweden 1996). Stockholm: Socialstyrelsen. SOU (Government Official Reports) 1996:169. Förändrinsmodeller ochfórändringsprocesser i kommuner och landsting (Models and Processes of Change in Muncipalities and County Councils). Stockholm: Fritzes. __________. 2000:3. Väfärd vid vägskäl. Utvecklingen under 1990-talet (Welfare at the Crossroads. The Development during the 1990s). Stockholm: Fritzes. Statistisk Ärsbok (StatisticalYearbook). 2000. Stockholm: SBC. Stephens, John D. 1996. "The Scandinavian Welfare States: Achievements, Crisis, and Prospects." In Welfare States inTransition: National Adaptations in Global Economies, ed. Gosta Esping-Andersen. London: Sage. Svenska Kommunförbundet. 1998. Förändringar av kvalitet och produktivitet i barnomsorg, skola och äldreomsorg (Changes in Quality and Productivity in Child Care and Elder Care). Stockholm: Svenska Kommunförbundet. Uddhammar, Emil. 1993. Partierna och den stora staten: en analys av statsteorier och svensk politik under 1900-talet (The Parties and the Big State: An Analysis of Theories of State and Swedish Policy during the 1990s). Stockholm: City University Press. VälfárdsBulletinen (The Welfare Bulletin) 6. 1994. Stockholm: SCB.