Child Welfare in Sweden - an overview Sven Hessle & Bo Vinnerljung Introduction The universal social welfare state Sweden has been described as a social welfare state from various theoretical points of view (Titmuss, 1974; Furniss & Tilton, 1977; Eriksson & Aberg, 1986; Gould, 1988; Spicker, 1988; Esping-Andersen, 1990; Cochrane & Clarke, 1993; Korpi & Paime, 1993). In addition to being committed to full employment - at least up until the 1990ies - and having a generous social insurance system as in the "social security state model" (Fumiss & Tilton, 1977)1, the social welfare state emphasises high quality living conditions through tax financed public services and environmental planning. The Scandinavian welfare state model is also characterised by a "universal approach" approach (directed to all, not to stigmatise the few) in opposite to a "residual approach". Everyone share the vast majority of benefits created by tax transfers sometime during their life span, without means tests. Residual tendencies in the universal welfare system Most of the national social insurance system, eg unemployment benefits and pensions, are based on individuals’ previous employment. The main exception is means tested economic assistance from local social services, a residual system with fresh roots in poverty relief. In the Social Services Law of 1980, this form of economic support was intended only to be temporary, a "social emergency relief'. Prevalence was stable for several decades, with around 6% of the adult population receiving this form of "welfare" annually (Tham, 1994). But in the last 15 years, the number of recipients has grown considerably. Means tested economic assistance is now the main source of income for large groups, eg young people with no education and recent immigrants, many of whom are more or less defacto excluded from ordinary jobs (Salonen, 1993; Ekberg & Andersson, 1995; Kamali, 1997). In other words, due to the post-industrial economic process, dependency on the residual components of the welfare state has become more widespread than anyone could imagine 20 years ago. These developments - paradox to the universal approach - have resulted in political confusion. By historical tradition and organisational affiliation, Swedish Child Welfare is closely related to this residual section of the social welfare state, serving, supervising and disciplining the poor and the excluded in the community. Like in most other countries, Swedish child welfare has never been able to "shed its poverty relief shell" (Sunesson, 1990).' A brief description of the country2 Sweden is one of northern outposts of Europe, sharing land borders with Norway and Finland on the Scandinavian Peninsula at roughly the same latitude as Alaska. Journeying the entire length of the country, 1.600 kilometres, a traveller would experience great variations in climate, topography and vegetation, going from fertile plains in the south through large evergreen forests in mid-Sweden to barren mountains and marshlands in the arctic north. If this traveller bypassed the areas around the nations three major cities, he/she would rarely be crowded. The population density outside these metropolitan centres is among the lowest in Europe. Half of the nation's nine million inhabitants are concentrated in just 3 % of the country's area. Roughly 1 8% of these nine millions have family roots outside Sweden, either by being foreign-born (10%) or having at least one non-native parent. Sweden was one of the most ethnically homogenous nations in Europe after the Second World War, but large-scale immigration during the 1960ies- 1990ies changed this forever. During a short period Sweden has become a multi-ethnic European nation. In the first two decades immigrants came mainly from southern Europe to work in the expanding industry but in the 1980ies and 1990ies most arrived as refugees, from repressive societies, war zones or other catastrophes in different parts of the whole world. In Malmoe, the country's third largest city, local statistics confirm this profound transformation: according to media reports, today 40% of the children under 18 years have some kind of immigrant background. The industrial backbone of the Swedish post WWWII -wealth was eroded during the late 1980ies and the early 1990ies. Large sections of the industry were dismantled, and an estimated half a million jobs disappeared, counting all sectors of the economy. This process, in combination with a strong influx of refugee immigrants at the same time, resulted in dramatically changed unemployment figures. Whereas 3% unemployment was seen as highly alarming in the beginning of the 1980ies, official statistics ran as high as 8-12% in the first half of the 1990ies.3 ' Other local social services managed to do this in the 1970ies: services for the elderly and child day care, both historically closely associated with help to the poor (Sunesson, 1990). 2Facts and figures quoted in this section come from various fact sheets of The Swedish Institute, all available on the Web (http://www.si.se/sverige/sverige/htmll, if otherwise is not stated. 3The two disparate figures stem from different definitions. The 12% figure includes those jobless who are in "labor market programs", eg adult education, while being payed unemployment insurance. Family policy4 Even during the economic recession in the early 1990ies with high unemployment, few Swedes questioned the fundamental goals of family policy: 1) to establish good condiditons for raising children, 2) to provide social security for families, and 3) to uphold the principle of the equal right of men and women to participate in life and work, while good child care is provided (see eg Kindlund, 1986) Historical roots Sweden's family policy should be understood within the broad context of a social policy that has its roots in the 18th century (Skoglund, 1992. But its modern socio-political framework, with the universalistic approach, dates back to the beginning of the 1930ies. Modern family policy was introduced in the beginning of the 1930ies at a time of deep economic recession. Because birth rates were very low, it was deemed necessary to improve conditions for families. In addition to the introduction of monthly child allowances, generous loans were given to newly married couples. The ideal of modern housing was promoted as a civil right rather than a privilege, and government subsidies were used to hasten the restructuring of the housing market.5 The reforms of the 1930ies and 1940ies reflected not just ambitions of material improvement, but also far reaching social engineering visions. Leading contemporary social scientist, like Alva and Gunmar Myrdahl, advocated science based social planning to set the everyday lives of Swedish families right (Hirdmand, 1989). Considering his close connections to the Social Democratic government, it is not totally without a factual base that Gunmar Myrdahl referred to Sweden as is "population laboratory" (quoted in ibid., 1989). The vision of fostering modern, rational Swedes included the use of eugenics to control economically wasteful growth among mentally and socially "inferior elements" in the population (Hirdman, 1989; Broberg & Tyden, 1991; see also Kock, 1996 for the history of sterilisation in Denmark). Family policy today Around 75% of Swedish children live in traditional nuclear families with birth parents (SCB:1994:2). But 50% are actually born out of wedlock, since many couples get legally married only after they have had their first child. Due to a high divorce rate, the probability of finding a child with divorced parents 4 As before, facts and figures quoted come from The Swedish Institute if otherwise is not stated (see footnote 1) That some people wanted to live in old, cramped housing was a matter of serious concern for politicians and planners in the 1930ies and 1940ies. Some advocated that it must be a civic duty to move if state functionaries recommended so (Hirdman, 1989). 5 increases with age of the child. Among 17 year olds, one in every three has experienced divorce and is most likely living with his/her mother. Social and medical support systems for parents and children were expanded radically during the decades after WWII. Today they include preventive services with free (or very inexpensive) health controls during pregnancy, prenatal and postnatal education programs, and regular health controls for children during the school years. The Swedish parental insurance system has gained international recognition, aiming at enabling both men and women to combine parenthood with employment. Parental insurance allows the mother or father to stay at home up to 360 days after the birth of the child, with 80% compensation for loss of income. Furthermore, parents have the right to stay home from work to care for a sick child in sum 60 days annually (per child under twelve years of age), with a cash benefit corresponding to 80% of their income. Labour laws secure parents employment unconditionally during absence for care of children and while using their parental benefits. New reforms have been added more or less continually over the last decades, even during economic recession. An example: when the unemployment figure was an all time high in 1994, parliament passed a new law requiring every local municipality to provide family counselling. This suggests that family policy can make it to the top of the Swedish political agenda, even during "hard times". But in other areas, tax financed benefits have been reduced. The municipalities are responsible for guaranteeing subsidised childcare for every child between one and six if parents are employed or studying. Since the overwhelming majority of parents work through their children's pre-school years, having access to childcare has become a necessity in everyday life. The formerly heavily subsidised high quality day care system - often regarded as a marvel by overseas visitors - has seen relatively big personnel cuts 6 and drops in the subsidy rate during the last ten years (e.g. Bjdrnstrom, 1996). Gender notes Women's participation in the labour force is among the highest in the world, 80% compared to 85% for Swedish men in 1995. A tax reform from the middle of the 1970s has been helpful in promoting female employment, conferring on married women the status of independent taxpayers. But the Swedish labour market has the same gender related segregation patterns as in many other countries. Women form the vast majority of employees in all "care sectors", e g nursing, childcare and social work. 30% work part-time, compared to 5% of the men. In national and local politics, 42-45% of representatives are women (parliament, local city councils etc.)7 6 Child-staff ratio in day care centres for pre-school children has risen from 3 -4 children per full-time employee to 5-6. Day care centres are still very well staffed compared to many other countries (Bjornstrom, 1996) 7 But in the higher echelons of the universities they are far fewer. As recent as 1996 only 22% of lecturers with a doctoral degree and 8% of the head professors were women. Even though there is unanimous agreement in the public discourse that fathers should share child related responsibilities, males use parental benefits substantially less than women. Attitudes among employers and colleagues to fathers exercising their legal right to a longer leave of absence for child caring reasons are still ambivalent (Hwang et al, 1984; Haas & Hwang, 1994; Bjornstrom, 1996). Defining/describing Swedish Child Welfare The problem of defining practice Defining practice is a precarious matter. What is child welfare, child protection, abuse, neglect - and what is not? The concepts are so imbedded in the legal, political and professional discourse, in everyday language, in cultural and time-related contexts etc., that attempts to define them in a more rigorous way frequently fail. This is so regardless whether one tries to define child abuse in a medical or sociological model (see eg Graham et al 1985; Finkelhor, 1986; Hutchinson, 1994; Rose & Meezan, 1997; Lagerberg, 1998). The concepts reflect norms of parental responsibility for adverse child experiences and for providing "good enough" standards of upbringing (Gough, 1993). Child welfare and child protection etc. are institutional organisations for state interventions when parents are not "good enough" (Levin, 1998). Law in combination with the political/professional discourse define the threshold of state interventions in families lives (see eg Donzelot, 1979; Dingwall et al, 1983a, 1983b; Harding, 1997). In other words, the social worker is also a moral worker (Hyden, 1996). Not surprising, many studies have found wide cultural disagreement about what constitutes "abuse" and when/what interventions are acceptable (eg Giovanni & Beccera, 1979; Gray & Cosgrove, 1985; Noh Ann, 1994; Soydan, 1995; Maitra, 1996; Colton et al, 1997). Vignette-studies in different countries also suggest that there are considerable differences within cultures, even among child welfare professionals (Lindsey, 1992a; Christopherson, 1998; Ostberg et a, 1999). Likewise, perceptions of "children's needs" and what is "in the best interest of the child" varies between cultures and maybe most strikingly within western societies over relatively short periods of time. Philanthropic societies in Britain sent children considered at risk for asocial development to residential and foster care in Australia and Canada as late as 1967, a practice that few would endorse today (Bean & Melville, 1989). Australia's policy of removing aboriginal children with some white descent, wholesale from their parents for life, continued until the late sixties. This practice was branded as "genocide" in a recent official inquiry (Human Rights and Equal Opportunity Commission, 1997).8 8 There are similar examples closer to home. Norwegian child welfare history include systematic removal of Traveller ('Tinker') children from their parents. Proposals of replicating this practice were voiced in Sweden as late as 1948 (Pfannenstill, 1948). All these ambiguities tend to lead the discussion into tautologies, eg parental behaviour becomes abusive when practitioners describe it as such (Dept of Health, 1995; Christopherson, 1998). The same line of reasoning can be applied to practice concepts like "children at risk' (Hessle, 1985; Andersson, 1993a; Vinnerljung, 1996a, 1997). Though many rightly abhor circular definitions, they are not totally without merit in social research. By describing child welfare practices as contextual, we are encouraged to continually study and review the basic norms and notions whereupon the legal political and professional discourse rests. Child welfare world history teaches us with clarity that this is necessary in a democratic, pluralist society. Swedish child welfare according to the legislators In consequence of this discussion, we have chosen to give an outline of child welfare as it is defined in Swedish law. The foundation is the legislator's interpretation of children's basic needs. The Children and Parents Code establishes that children have a right to care, security9 and good upbringing. Furthermore, the law states that children should be treated with respect for their individuality. They may not be subjected to physical punishment or other degrading treatment. By legally giving children these "rights", the law also defines the responsibilities of parents (or other caretakers). At the same time, the State ascribes to itself the right to intervene if basic needs are not respected of fulfilled. Logically, it would follow that the state has an obligation to make sure that interventions lead to the needs being met, but this is less often discussed. The equivalent of a child welfare law is found in the Social Services Act of 1980 (Sol) and the Care of Young Persons Act (LVU) from the same year. The latter regulates taking children and youth into care without consent from the parents or from children themselves, when aged 15 or more. The Social Services Act is a "frame law', regulating several areas of social support and interventions: economic assistance, pre school child care, care for the elderly and handicapped, care for substance abusers etc. In reality, the organisation of child welfare is amorphous and differs from one local authority to another. In some towns there are specialised units within the Social Services, in others child welfare is part of the local school organisation or there are no specialised child welfare social workers at all. This is due to social services being the legal and financial responsibility of municipalities, local self governing entities with far reaching rights of taxing their citizens. As long as basic standards in the "frame law' are respected, local authorities have the right to organise their child welfare as it suits them. Swedish child welfare legislation makes no strict distinction between child protection and youth justice. Asocial behaviour of young people under 20 is a child welfare problem, outside the realms of criminal justice. Neither are there organisational dividing lines between different means of intervention. 9 The law addresses "security", not "safety". Local authorities shall mainly work with social support to and in partnership with families, regardless of the age o the children or the reason for intervention. If support is not enough, care shall be used. The normal procedure is care with the consent of the families (eg 80% of all placements of teenagers; Vinnerljung et al, forthcoming). The social authorities cannot legally refrain from providing support or care, eg by referring to financial limitations. Child welfare related duties of local social authorities are roughly, besides creating secure and good conditions for children in the community: In partnership with families support children's personal, physical and social development Monitor children who show signs of unfavourable development In partnership with families make sure that children at risk get the protection and support they need, and - if it is deemed to be in the best interest of the child - place children in care outside their families. Generally, Swedish child welfare has its main emphasis on social support and service, rather than on child protection (same as in Norway, see eg Clausen, 1998 ; Jonassen et al, 1997). Definitions of child abuse and neglect in Swedish law Swedish lawmakers have a stern attitude to physical abuse. Even minor physical punishment may be a reason for a court order. According to a handbook for application of the LVU-law, mental abuse is defined as cases where "the child is submitted to psychological suffering through systematically degrading behaviour, terror or depreciation" (Socialstyrelsen, 1997, p 26). Examples include chronic negative discrimination against one sibling. Other circumstances that may lead to a court order include children who are repeatedly told they are stupid, clumsy, ugly or worthless. Frequent threats of physical or other punishment may also constitute reasons for intervention (ibid.). Exploitation mainly refers to sexual abuse but include cases where children are forced to do heavy labour. This criteria also applies to children who have "un-normal responsibilities in the home", eg children who have to look after an alcoholic parent (ibid.). Inadequate care is an approximation of the international concept of neglect. It covers a lot of ground: children receiving insufficient physical care, stimulation and emotional nurturing, substance abusing or mentally ill parents, children who do not get proper medical care, who are not receiving adequate supervision or are repeatedly left in the care of others etc (ibid.). The Child Welfare System in Sweden No evidence of down-sizing The recession in Sweden in the late 1980ies and early 1990ies led to a fiscal crisis in the public sector, both nationally and locally. Down-sizing public service was also motivated by forced priorities, since the number of elderly requiring extensive tax subsidised in-home and hospital care had increased tremendously between 1970 and 1990 for demographic reasons (SOU 1996:169). But several studies has found that local child welfare became a "protected zone" in this turbulent process, with financial resources mostly intact or even growing (Socialstyrelsen, 1994a, b; 1998a, b; Bergmark, 1995). Great local variation in service availability and delivery As mentioned before, child welfare is the responsibility of local government, who decides how to organize its services. Local authorities also have considerable freedom in choosing what services to provide. The result is great variation between different municipalities. This is so even within cities and large towns, since many are divisioned into smaller districts with large discretionary powers over "their own" child welfare. In one town families-at-risk are placed in an assessment centre during investigation, in another they are provided different forms of social support during the process and in a third they are mainly summoned to the agency's office (Ostberg et al, in press). In our attempt to disentangle the decentralised Swedish child welfare, services and activities are sorted under four categories for descriptive purposes (compare Sanders et al, 1996): Prevention Investigation Social support and in-home treatment Care Prevention In this passage we will use the concept of prevention restricted to services available to families without turning them into individual "cases" for the authority, but where child welfare personnel are at least involved on a regular basis. 10 Services where children and families become formal child welfare 10 In the medical definition (eg Sameroff & Fiese, 1990; Upshur, 1990) primary prevention is practiced prior to the origin of a disease, secondary aims to prevent progression of a disease and tertiary prevention is not really prevention at all but strives to restore health. We elect to use the concept in a practice related way, even though it mainly corresponds with the idea of primary prevention "clients" are discussed under the headline of "social support and treatment". These services often have preventive ambitions, mainly to hinder problems from getting worse (secondary prevention, see note 10 below). Even foster care has historically been called "preventive action", aimed at preventing small children with "inadequate parents" from turning into asocial teenagers or adults (Vinnerljung, 1996a; see Bloom, 1998 for a modern version of this argument). The main bulk of general and group targeted prevention work is done outside child welfare organizations, eg pre and post natal maternal care, subsidised child care, social support in schools etc. In daily child welfare work there are hardly any standardised preventive services, available to families regardless of what "municipality" they live in. Instead there is an array of different programs, attempting to prevent eg negative child development, family breakdown, placement in care and asocial behaviour in youths. The following listing is by no means complete, and should be seen as an attempt to cover mainstream examples. In the last five years agency cooperation and intra-agency service delivery have become increasingly popular (Socialstyrelsen, 1998a). With financial aid from the National Board of Health and Welfare, scores of local authorities have constructed services where maternity care, child health, pre-schools and child welfare either operate in the same locality ("family centrals") or intensify cooperation. The aim is to give information and social advice, provide non-stigmatising support to families-at-risk and strengthen their social network by group activities. The Family Centre at Hagalund located in greater Stockholm is a good example of this interagency collaboration (Hessle (Ed), forthcoming). Evaluations of these programs are so far too local, small sized and process oriented for conclusions about outcome (Bergman, National board of Health and Welfare, personal communication). In 1997, 1/3 of all local authorities organized and/or supported self-help groups for children with alcoholic parents. Other child welfare initiatives supported self-help groups targeted for example at single mothers, refugee parents, and young victims of sexual abuse (Socialstyrelsen, 1998a). The extent of availability and service delivery on a national scale is not known. Evaluations of self-help groups are encouraging, especially so for children with alcoholic parents (Lindstein, 1996). It is questionable if self-help groups can be labelled prevention, although many such programs have clear secondary prevention goals. Parent training/education programs are rare, but exist in some "municipalities" (Socialstyrelsen 1998a, - b). Again, the extent is unknown. We do not know of any Swedish evaluation studies of these programs (see e g Gough, 1993; Kazdin, 1994 for reviews of evaluations). "Summer families" and summer camps used to be part of every larger local authority's service delivery. Children from towns and cities were offered to spend a month as guest in a rural family or at a summer camp, for a very low fee. The general idea is to give a compensatory service to low income urban families. But preventive aims are also prominent, eg reducing the risk of family breakdown by providing parents with temporary relief from stress. Services were reduced in the late 1970ies and early 1980ies in preference to supporting low-income families vacationing together with their children. "Summer families" and summer camps still exist - and are in demand - but we have no figures of service delivery rates. Closely related are agency financed and staffed programs aiming to give at-risksyouth meaningful after-school activities and nurture relations between youths and community workers. Examples include café-, boat- and theatre- projects (Socialstyrelsen, 1998b). Youth in around 85% of the municipalities have access to a Youth Advice Centre. These centres give advice and guidance to youth in sexual matters, eg sexually transmitted disease protection and birth control, and are often run jointly by local social and medical authorities (ibid.). A few small centres that offer advice on drugs to youth and parents are in operation. Social workers in most local authorities regularly give information and advice on drug related matters in schools. Overseas experiences from community work have had apparent difficulties being integrated in the Swedish model of social work, where "authorities" are the main actors. Especially models built on Community development or Community action (Denwall, 1994; Wahlberg, 1997). Community work in Sweden has a tendency to either drift toward social planning (Denwall, 1994) or toward individual case work. But it thrives in several towns or cities (Socialstyrelsen 1998b). Usually community work is restricted to a neighborhood. Social workers take an active part in matters of importance to the local community, give advice and support to families and community groups without turning them into "cases". Work is done in liaison with schools, landlords and local community organisations. Some of the self-help groups mentioned before, are organised by community social workers. Youth gangs have become an important target group. Some programmes could be categorised between prevention and investigation, eg different social network programs that has been developed on the neighbourhood level during the last decades (Dominelli, 1999; Hessle (Ed), forthcoming; Wallmark, J (Ed), 1998). Focus is on risk groups in the neighbourhood, with different social work models activating supportive relations for the vulnerable groups. Investigation It is mandatory for all professionals who come into contact with children and youth in their work, to report anything that may cause child welfare authorities to intervene for a child's protection. This applies to all professionals, regardless if they are employed by a public or a private agency.11 For other citizens, the law says they "ought to" notify child welfare authorities. There is a notable exception in the 11 This does not mean that all professionals comply with the law. Sundell (1997) found that personnel in child day care report only a minority of suspected maltreatment cases. same paragraph: professionals working with marriage counselling are only obliged to report suspected physical and sexual abuse. Child welfare social workers must start an investigation when they receive a report, without delay. Any information valid for the investigation is automatically exempted from ordinary rules of secrecy. The investigator can directly access information from other authorities or organisations that work with children, including private hospitals and day care facilities. If needed, a child can be taken into compulsory care immediately, for investigation purposes or in emergency situations. These kind of placements have increased considerably during the 1990ies (Socialstyrelsen, 1998a) Research suggest that there are great variations in the procedures of investigation, naturally between different kind of cases, but also between different local authorities (and individual social workers) (Andersson, 1991; Socialstyrelsen 1998a; Ostberg et al, in press). Sallnas (1995) has reported that increasingly more residential homes offer assessment, of either the whole family or only the children. The expansion of residential assessment centres in latter years is interesting, since this practice was largely abandoned in Britain during the same time (e g Berridge & Brodie, 1998; see Sinclair et al, 1995 for an example of evaluation). The national prevalence and incidence of reports is unknown, since there are no uniform routines for registering and counting reports. If local statistics exist, reports are usually indistinguishable from applications for social support. Estimates indicate that local authorities in 1995 received reports or applications concerning 2% of all children under 18 (Socialstyrelsen, 1998a). But a recently completed survey reveals immense variations between different municipalities (Stina Holmberg, personal communication). A two year trial project with Family Group Conferences (FGC), after the New Zealand Model,12 has recently been completed in ten municipalities. Evaluation with quasi-experimental design found that 90% of families came up with plans that were accepted by local child welfare authorities. Participating families were generally satisfied with procedures, even though a larger number than in Britain declined offers of a FGC during investigation. Follow-up results after one year showed no differences in the number of cases that had to be re-opened due to new reports between 108 FGC-participants and 167 children whose cases were processed by ordinary routines, after controlling for backgrounds. Also there were no significant differences in the number of children who remained in care (Sundell & Haeggman, 1999). 12 See eg Marsh & Crow, 1998 for description of the model. Social support and treatment Like in prevention, there are great variations between local child welfare authorities regarding what services of social support and treatment that is offered. As before, programs mentioned below are only examples of main stream phenomena in child welfare today. One of few standardised services is providing a mixture of respite care and social support from volunteers (or volunteer families), in Swedish legislation called contact person or contact families. These volunteers receive a small monthly fee and a minor expense allowance for supporting another family, eg by taking care of the children one weekend a month (Andersson, 1993c; Sundell et al, 1994). When this support was constructed, proponents likened the arrangement to providing substitute relatives for families with weak social networks. Parents apply for a "contact person", or can be offered one as the net result of a child protection investigation. Single mothers used to be the main target group, but now "contact persons" are used also for supporting teenagers and adults without children. Even if this form of support is under-researched and evaluation is lacking, "consumer reports" tell that this is a genuinely popular form of service. It also seems relatively free from social control elements, as perceived by both families and volunteers (ibid. Official statistics for 1997 show that 1 % of all children 0- 1 8 had a contact person/family (Socialstyrelsen, 1998c). Children in care are in other words only a minority among Swedish child welfare clients. 13 Current research shows that voluntary work has important functions in Swedish social welfare, even in the welfare state where most of the responsibilities for solving social problems is borne by the state (Jeppsson Grassman, 1995; Lundstrom, 1994). This result may come as a surprise to those who believe that a strong public sector kills reduces people's willingness to help others on a voluntary basis. Actually, nearly 50% of the adult population devote part of their free time to some form of voluntary work (Svedberg, 1994). There is justification for the hypothesis that a strong welfare society, with a high standard of living, creates the freedom to use one's free time for the benefit of others. Information, advice, personal support and practical help from a child welfare worker, often scheduled over a long time, is among the most common service delivered, even though it never enters national statistics. Sometime meetings between social worker and families include family therapy, depending on the social workers skills. Standardized in-home services are few. To our awareness there are no specialised family preservation programs like Family First in the US (Schuerman et al, 1994), which has been noted by some foreign observers (Barth, 1992). "Family treatment" and "family support" are other forms of service and support, frequently delivered by child welfare ( eg in Socialstyrelsen, 1998a, -b). 76% of all local authorities offered some version of this service in 1997 (ibid.). Definitions are often vague, but generally refer to regular meetings with families-at-risk with qualified personnel. "Family educators' in many local authorities provide advice, support, skills training of parents and give practical help, often for several years in the same family. Typical target groups are young single mothers and mentally disabled parents (Socialstyrelsen, 1998a, -b). These supportive services have a long tradition in Sweden, originating in the 1960ies (Jonsson, 1973). They have also been used extensively in social work with gypsy families (Arnstberg, 1998). There are no national figures at present showing extent of service delivery. Providing and financing clinical individual and family therapy for children and families is a standard tool in child welfare, although (again) no statistics are available about scale or profile of service receivers. Some local authorities have their own therapists. Financing therapy with private practitioners is a common alternative, as is referrals to treatment by public child and youth psychiatry units. Child welfare personnel often have trouble gaining access to child and youth psychiatry services for "their" families, especially youth with asocial behaviour (Vinnerljung, 1989, 1990; Andersson, 1993b). Child Welfare has become increasingly involved with basic education. Social workers are part of schools social support service in many places, child welfare authorities are involved in running special education units in some etc (Socialstyrelsen, 1998a, -b). Since youth unemployment is rampant, partly due to exclusion mechanisms of the Swedish labour market (Salonen, 1993), child welfare has in recent years directed more attention toward school leavers with only a basic education. Many local authorities run different programs for unemployed youth, often in combination with life skill training (Socialstyrelsen, 1998b). Care Care for children and youth in Sweden can roughly be divided in three categories: • Foster (family) • Residential care • Homes for special supervision Adoption without parent's consent or procedures to "free children for adoption" does not exist today. From WWII to the mid-1960ies, young "immature" mothers were as a matter of routine recommended/persuaded by local child welfare authorities to give up illegitimate children for adoption (Allmanna Barnhuset, 1955; Socialstyrelsen, 1959; Vinnerljung, 1992). Now, only a handful of children are left annually for adoption by Swedish parents. Childless couples have turned to international adoptions since the late 1960ies. An estimated 1,6 % of all children in current teenage cohorts were originally adopted from abroad (statistics quoted in Runquist, 1996). Richard Barth, a leading US child welfare researcher, noted with some surprise that Swedish child welfare workers did not promote 13 Reports from other countries usually show the same, e g Dept of Health, 1995 for Britain; Clausen, 1997 for Norway; adoption of long term placed foster children, even when birth parents suggested it. He regarded the absence of adoption in Swedish child welfare practice as a problem, considering every child's primary need of a family for life (Barth, 1992). Foster care Foster care is by law and long tradition preferred to residential treatment. Swedish children have been placed in foster homes since the 18th century, 100 years before the American "orphan trainee' brought thousands of poor children from cast coast cities to substitute parents in the rural west (Holt, 1992). Taking foster children is a tradition in some rural districts (Kalvesten, 1974). Research from 1974-1992 clearly shows that the century old practice of placing children from urban families in rural foster homes continues (Vinnerljung, 1996b). In cross section data 75% of all children placed in care are in foster homes (Socialstyrelsen, 1998c). But since cross section samples contain a large overrepresentation of long placements (e g Vinnerljung, 1996c; Usher et al, 1999), the picture changes if we look at longitudinal data. Among initiated placements in 1995 of children all ages, around 55% were made in foster care (Vinnerljung et al, forthcoming/b). Time series data reveals that the use of foster care has decreased continually the last 13 years, from 72% of all initiated placements in 1983 to 55% 1995, while placements in different forms of residential care have increased (from 28% to 45%). This trend, valid for all age groups, cannot be explained by changes in the care population. Age and gender patterns have been remarkably stable throughout this period. The development of the care system has taken a path in the opposite direction of eg in Britain, contrary to the intentions of national policy makers and cost wise from cheap to expensive care (ibid.). Legislation for protection of children in non-parental care is rather strict in some aspects, compared to other countries. Unregulated private foster care is illegal, even in a wide sense. An example: if a mother wants her child to live with granny during the mothers convalescence after hospital care, granny is required to notify local authorities, agree to be investigated (including control of criminal records) and submit to annual inspections. Failure to comply can lead to prosecution. Child welfare authorities actually have the power to legally ban adults from taking care of any children in their homes, eg foster or day care children.14 Swedish child welfare professionals have in the last 30 years tended to regard kinship care (fostering by relatives) with suspicion, probably due to a prominent influence in practice of theories on intergenerational transmission of psychosocial problems (Vinnerijung, 1993, 1998). A media supported political campaign by a group of relatives to children placed in care (called the "grandma revolt") 14 Stockholm's child welfare authorities are now trying to get a precedent for a wider use of this law, as a tool against adults who are suspected of sexual exploitation of neighborghood children . resulted in a change of law in 1997. When children must be place in care, primary consideration must now be given to relatives (or other close adults) as substitute caretakers. Many local authorities use contracted foster homes for short term or emergency placements. These families usually have a contract, agreeing to care for a set number of children. Contracted foster families are used as replacements or complements to residential care (Vinnerljung, 1996b; compare Cliffe & Berridge, 1991). Residential care Sweden's care system is a good illustration to Berridge's (1985) and Colton's (1988) assertion that foster and residential care are not totally separable entities, but rather different intervals on a continual scale. The law of 1980 brought a change in definitions. It stated that if foster homes have four children or more and if foster parents main income came from fostering, they should be defined as a residential unit. The rationale behind this legal change was to bring professional care under stricter control. It actually paved the way for a private expansion in care, that only partly can be explained by former foster homes being redefined (Socialstyrelsen, 1990; Sallnas, forthcoming; Vinnerljung et al, forthcoming). There are now more residential units "in the market" than at the heyday of residential care in the 1930ies (Sallnas, forthcoming). Half of them were established during the 1990ies (Sallnas, forthcoming). Sweden's care system has - if legal definitions are applied - slowly developed toward reinstitutionalisation and privatisation during the last 15 years, even if foster family care still is the dominating form of care (Vinnerljung et al, forthcoming/b). Some of the new private residential units would probably be defined as "specialist foster care" in other contexts (e g Shaw & Hipgrave, 1983). More than half of its owners/operators has experience from fostering (Sallnas, forthcoming). In 1995 they received almost as many children - 0- 1 8 years old - as publicly owned children's homes (Vinnerljung et al, forthcoming). Residential care is mainly used for teenagers (60% of all initiated placements in 1995). But the private children's homes have increased their 53 market share" substantially for younger children, mainly for intermediate and long term care, a new phenomena (ibid). Most residential care units are small. 73% were dimensioned for 9 children or less. Around half of the personnel has a college degree (Salinas, forthcoming). Sweden has been spared tragic scandals of sexual and physical abuse in residential care, that have plagued child welfare in some other countries (see e g Levy & Kahan, 1991; Kirk-wood, 1993; Colton & Vanstone, 1996). Placing young children in temporary residential care together with their parents is very common. Among children's homes for 0- 12 year olds, 90% state that they receive children and parents together (Sallnas, forthcoming). The legendary Children's Village Ska outside Stockholm with decades of experience in treating children and parents together in residential settings, has been the model (Hessle, 1997). The impact of Ska actually revolutionised residential care in a very short time. Already in 1985, more than half of all children in children's homes had at least one parent staying with them (Socialstyrelsen, 1990). The proportion is smaller now, since private small, home-like residential units are increasingly used for longer placements (Salinas, forthcoming). Homes for special supervision Since youth justice is included in Swedish child welfare, criminal and drug-using youth make up a substantial part of teenagers in care. Sweden has for decades had special residential care for the "worst" in this group. These homes for special supervision have facilities for locking up youth, and are legally authorized to place violent youngsters in temporary solitary confinement and to perform (bodily) searches. There are 30 residential units of this category with around 600 beds today, since 1994 run by a national government agency (SiS). Local authorities apply for placement and have to pay for care (£200-300/day). Most homes have been in operation for decades (Korpi, 1996; Salinas, forthcoming). Residential care in this form does not exist in Denmark or Norway, but there are calls to replicate the Swedish example in these countries (Levin, 1998). Even though follow-up studies usually is depressing literature (ibid.), placements have increased since since 1994 (Vinnerljung et al, forthcoming/b). The responsible government agency has strong ambitions of turning this form of care - traditionally punitive and confining - into an evidence-based treatment organisation (Armelius et al, 1996). Historical and legislative contect in brief Just after the turn of the century (1902), Sweden got the first child welfare legislation. As a consequence of the nation's rapid industrialisation and urbanisation at the end of the 19'h century, and influenced by trends of social control within the European social policy, the child welfare legislation aimed to save young people from assumed future criminality (Lundstrom, 1993). Like in Norway (who passed the world's first- child welfare legislation in 1900) responsibility for administration and enforcement was given to special child welfare committees in the local communities and not, as in other countries, to special family or youth courts. At the turn of the century and for decades to come, the state's gravest concern was what to do with all abandoned and out-of-wedlock-born children that followed in the wake of industrialisation (eg Ohrlander, 1992). The "fortunate" single mothers in distress, were those who managed to secure a place in a children's home or foster home for infants they could not mind while working (eg Hegeland, 1978, 1988). Children born out of wedlock dominated foster care until the late 1950ies. As late as the end of WWII, only around half of all Stockholm children born out of wedlock, were in a parent's care (Granath, 1958; Sjoberg, 1959). The Child Welfare Act of 1924 replaced the legislation of 1902. The indications leading to compulsory care were extended also to small children abused at home. The new step taken was that the local child welfare authorities (committees) under specified rules were obliged to intervene in families where children were being abused. Even if experts, for the most part medical practitioners, took their place side by side with laymen in the communities, the practice of child welfare was for many decades and in most municipalities carried out by laymen (Lundstrom, ibid). During the post-second-world-war period, when the next step was taken with the passage of the Child and Young Persons, Act of 1960, Sweden had undergone dramatic change and had become a successful industrial nation with a social reformist policy. Moreover, the municipal organisations administrating child welfare had developed into bureaucracies. And the fast growing field of child psychiatry with its arsenal of theories became the dominant incitement for taking children into public care. "The theories of child welfare... changed... from explanatory models based on moral precepts to models based on psychological is grounds." (Ibid., p 268). The Child and Young Persons Act of 1960 added nothing new to the Child Welfare Act of 1924. But it did emphasise the preventive part of child welfare work and it regulated the legal procedures and rules for case documentation in child welfare. Current legislation The present child welfare legislation, the Social Services Act, was passed in 1980. Child welfare was organized as a part of other social services, e.g. child day care, care for the elderly, social assistance, and treatment of alcoholics/drug addicts. The professionalisation of social work had taken a big leap in as much as the personnel now became what we may call "counsellors" for social welfare consumers. One of the main issues driving the change in legislation was a wish to extend the process of democratisation to the poorest and most excluded segments of the communities. The law also had farreaching compensatory ambitions, of stretching traditional goals of shared welfare and political participation to include also poor and troubled citizens (Sol): Public social services are to be established on a basis of democracy and solidarity, with a view to promoting economic and social security, equality of living conditions, and active participation in the life of the community. With due consideration to the responsibility of the individual for his or her own social situation and of others, social services are to be aimed at liberating and developing the innate resources of individuals and groups. Social service activities are to be based on respect for the selfdetermination of the individual. The following discussion concentrates on interventions that form the Swedish equivalent of "child protection". We will mainly look at criteria for intervention in "inadequate" families with younger children and focus on the prerequisites for "hard" interventions, that is the taking of children into care without parent's consent. Even though most placements legally have the form of social support, the criteria for care without consent mark the threshold for traditional child protection. "If you agree to our proposals, you will get social support, if not we will take coercive action". Legal criteria for "hard" interventions Conditions that can lead to placement in care without parental consent are (LVU): • Abuse • Inadequate care/neglect • Other conditions in the home, e g very disturbed relations between the child and his/her parents • Children's and youth's substance abuse, crime or other socially destructive behaviour Adverse home conditions are not by themselves sufficient ground for a court order, but their presumed short and long-term consequences. The law specifies that home conditions (or a child's behaviour) must constitute an "apparent risk to the child's development or (physical/mental) health" (LVU, 1). Thus is the theme of prediction firmly embedded in the fundaments of Swedish child welfare legislation, and has been so since the beginning of the century (Lundstrom, 1993; Vinnerljung, 1996a). An example: a parent's alcohol abuse falls under the requisite "inadequate care". Even if there are no apparent signs of damages to the child's development, practice wisdom and cultural values say that having parents who abuse alcohol is a risk of negative development in the future. If we look to research, this is surely so. Most studies show that the risk of a maladaptive development for children of alcoholic parents is three-fourfold higher than for children without that parental deficit. But the great majority do not show signs of social or mental impairment in later life, (eg EI-Guebaly & Offord, 1977; West & Prinz, 1987, see Rutter, 1989; Maugham & McCarthy 1997 for a general discussion). How substantial should this risk be to motivate a court order, expressed in quantitative terms - 10%, 30%, 50%, 70% or 90%? These questions are rarely put forth, and have never been answered by legislators or policy makers. They mark a problematic crossroad between moral issues and the actual ability to make predictions of children's development from parental characteristics (Lindsey, 1992a; Weightman & Weightman, 1995; Vinnerljung, 1996a; Backe- Hansen, 1997). Given these conceptual problems, Swedish criteria for child protection are diffuse compared to many other countries. Evidence of abuse or harm to children is not even a typical precondition in practice. Andersson (1984, 1991) found that most pre school children taken into care during the 1980ies did not exhibit observable damages, dysfunction or maladaptive behaviour. They were placed due to parental life styles assumed to be harmful to children's development, mainly mothers' alcohol or narcotic abuse. Several British writers have approvingly described the predictive elements in Swedish child welfare legislation. They have expressed regret about British law not permitting similar "theory based", "pro- active" interventions ( e g Weightman & Weightman, 1995; Nixon, 1998). This may be understandable from a practitioner's point of view, but it is more problematic for a researcher. From empirical evidence, it seems impossible to predict the development of children-at-risk or the future behaviour of families-atrisk on an individual level with reasonable accuracy, especially so for low frequent phenomenas like child abuse (e g Dingwall, 1989; Gough, 1993; Lagerberg et al, 1994; Browne & Herbert, 1997; Vinnerljung, 1998). The number of "false positives" in theory based screenings of families are simply so many that interventions labelling parents as potentially abusive become ethically impossible. To sum up, Swedish child welfare stresses a social service approach by an emphasis on providing support. Child protection legislation is diffuse and includes a high confidence in social workers ability to make predictions of children's development. To this can be added another essential element: a strong belief in the beneficial results of state interventions in families (Vinnerljung, 1996a). Foreign observers have noted that Swedish child welfare has high legitimacy and support in the public, in spite of far reaching and diffuse powers. They credit this to a cultural tradition of consensus in social control issues (Gould, 1988, Weightman & Weightman, 1995). Gould coined the phrase: Swedish social workers know they are doing nothing wrong (1 986, quoted in Weightman & Weightman, 1995, p 76). A profile of children in the care system Prevalence and incidence of care Statistics and estimates of prevalence and incidence of care are often based on cross section data, which make yearly figures confusing and often misleading (Goerge, 1990; Vinnerljung, 1996c; Usher et al, 1999). Cohort statistics give us good complementary data.15 Based on a study of cohorts born 19721974, the prevalence of care experience among today's living adult population can be (conservatively) estimated to 4-5%, roughly the same figures as in Britain (Vinnerljung, 1996c; Bamford & Wolkind, 1988). 3-4% has been in foster care and around 1% only in residential care. In later cohorts, fewer have care experiences when they reach maturity (3-4%), but local variations seem to be very big.16 The risk of being admitted to care for the first time is greatest in infancy (0-1) and in teens (14-17) (Vinnerljung, 1996c). If we look at cross section statistics, 15.500 children 0-20 years old, were in care sometime in 1997, that is less than 8 of 1.000 children. Almost three in four were placed with formal consent from parents, and one in four after a court order (compulsory care). Of the latter, about three fourths were in care due to adverse home conditions and one fourth due to asocial behaviour (Socialstyrelsen, 1998c). The 15 16 See Samet & Munoz, 1998 for a general discussion on cohort studies. In one Stockholm district, 7% (!) of all 15 - 17 year old were in care 1998 (Begler & Wanell, 1990) proportion of compulsory/voluntary care can vary a lot between countries. In a comparison, the figure for compulsory care in Sweden is 20 times that of Japan (Hessle, et al, 1996). On December 31 1997, 1 1.000 children were in care (5-6 of 1.000 children), 74% in foster homes (Socialstyrelsen 1998c). Cross section statistics have been stable for the last ten years. But the numbers have been much reduced since 1960, when there were three times as many foster children compared to now, on any given day of the year (Vinnerljung 1996c). Around 5.500 children started a new placement during 1997, indicating a large annual turnover in the care system. Of these around 3.000 came into care for the first time (Socialstyrelsen, 1998c). The epidemiology of care Bebbington & Miles' (1989) classical study of the social epidemiology of care has been called "the most obscene statistics ever to emerge from child welfare research in Britain" (Ryburn, 1993). There is a dearth of epidemiological research in this field, but the few examples that exist show overwhelming evidence of the connection between placements in care and poverty, especially in combination with single motherhood (see c g Lindsay, 1992b). In the table below, results from Bebbington & Miles article are shown together with a little known piece of Swedish epidemiological research, the latter focusing on the risk a child being taken into care during the first six months after birth (Hoglund Davila & Landgren-Moller, 1991). "Child A" and "child B" are fictitious constructions with all statistical traits listed in the table. Risk for placement in care (great Britain), Bebbington & Miles, 1989 Risk for placement in care (Sweden), Hoglund Davila & Landgren-Moller, 1991 Child A Age 5 - 9 Two-parent family No social assistance <3 children in family "white family" Child B Age 5 - 9 Single parent Social assistance >4 children in family Child A Age 0 - 6 months Two-parent family Mother good income Middle class family Mixed racial background Family rents house House has one room or less per family member Risk 1/10 Risk<1/10.000 Family owns house17 House has more room than family members Risk 1/7.000 Child B Age 0 - 6 months Single young mother Mother low income Mother unqualified, jobless Mother not convicted of Mother convicted of a a crime crime Risk 1/10 In a review of Swedish research on foster care, Vinnerljung (1996b) noted that in all studies the large majority of foster children have single mothers and absent fathers. Hessle (1988) found strong evidence of a long history of social exclusion and continued marginalisation during the life course, in parents of foster children. Groups with specifically high risk of having their children placed in care are substance 17 "House" also includes apartment, flat etc. abusing and mentally disabled mothers (Vinnerljung, 1996b). By tradition, child welfare workers focus their main interest on the mother's behaviour and capabilities (Kristinsdottir, 1991). Fathers' behaviour is much less frequently defined as a problem for the child, unless it is abusive (Hikansson & Stavne, 1983; Claezon, 1987;Andersson, 1995; Vinnerljung, 1996b). Age and gender of children placed in care Looking at annually initiated placements in care and correcting for sonic of the problems in national statistics, age and gender patterns seems to have been stable between 1983-1995 (Vinnerljung et al, forthcoming/b). Pre-school children (0-6) make up for 25-30% of annual placements, school children (7-12) for 20-25%, younger teenagers (1 3 -16) 3 5% and older teenagers (> 1 6) 1 5%. This means that in the last decade, every second child placed in care has been a teenager. This is rarely acknowledged, since child welfare professionals tend to focus on small children (Claezon, 1987; Freeman et al, 1996). Age compositions vary between different forms of care. Around 45% of children coming into foster care are teenagers, while practically all placements in homes for special supervision concerns children 13 or older. (Vinnerljung et al, forthcoming/b). Generally, there has been an even gender composition of children coming into care for many years. In residential care placements though, boys form a small majority (55-60%) and among placements in homes for special supervision, boys dominate (68-75%) (ibid.). Immigrant overrepresentation Children with an immigrant background 18 are almost twice as likely to be in foster care, compared to "Swedish" children (Socialstyrelsen, 1995a; 1998 b). This overrepresentation is mainly limited to children of Finnish decent, who are four times as likely to be in foster care on any given day, and have been so since the beginning of the 1980ies (ibid.). Among teenage placements, 13-16 year olds, several other ethnic groups are strongly overrepresented, eg children with South American decent (4.3), or with an African background (10.6) (Vinnerljung et al, forthcoming/a). The latter group is very small in absolute numbers, but is also 17.5 times more likely to be in homes for special supervision (Bergstrom & Sarnecki, 1996). The only examples of this kind of extreme overrepresentation that we have found in international literature, concern native American children in the 1960ies and 1970ies and aborigines in Australia (Johnson, 198 1; Mannes, 1995; Human Rights and Equal Opportunity Commission, 1997). Reasons for coming into care National statistics give no information about reasons for care, except on what legal grounds the placement is made. Since the same reason for intervention can result in a placement with parent's consent or without, these figures offer little advice. Instead, we have turned to a review of Swedish foster care research (Vinnerljung 1996b) and an ongoing study of over 1.000 teenage placements (13-16 year old) made in 1991 (Vinnerljung et al, forthcoming/a). There are problems, also with these data. The studies in the research review are mostly 10+ years old and the foster care studies lack standardised categories of reasons for admission to care. As before, we stay away from cross-section studies and use only data on initiated placements. Foster care populations All figures on the foster care population are rough estimates from several local, often small studies. In the following passage, total percentages exceed 1 00% since many children are placed for several reasons and since different studies have used different categories. • In all studies from the 1980ies and 1990ies physically and sexually abused children are few, 5-10%. • A large proportion of placements is made due to parent's substance abuse, 25-35%. • Children with "incapable" parents, many mentally disabled, make up for 20% of placements • Children with mentally ill parents for 15%. • Possibly the largest group consists of children who have left their home due to family breakdowns, 30- 40% in two local studies. Some have been deserted, most often due to incarceration of parents. Others come into care due to hard conflicts with parents, resulting in that some children become "runaways" and some "throwaways" (parental rejection). Teennage placements In the study of teenage placements made in 1991, 16% were directly caused by breakdown in a previous placement (Vinnerljung et al, forthcoming/a; compare Berridge & Cleaver, 1987). The composition of reasons for admission in care was as follows. Figures are approximations. The total is larger than 100%, since some children came into care for several reasons. Asocial behaviour 40% Conflicts in the family 40% Inadequate care 30% Physical or sexual abuse 10% "Run-aways" or "throwaways" 10% Special needs care 19 19 5% Refers to physically or mentally impaired children, who cannot or do not get required special needs care from parents. Problems among children admitted to care There are no national statistics about behaviour problems of children admitted into care or already in care, neither any studies of the prevalence of educational or health problems. Trials with the Looking After Children instruments for assessments of children in care (Parker et al, 1991) are on-going and will probably shed light on these and many other issues. For teenagers coming into care, we have more substantial information, albeit from 1991 (Vinnerljung et al, forthcoming/a). At the time of placement, information from case files indicated that the majority had some kind of behavioural problem (68%) 57% had school related problems like frequent absconding or very disturbing behavior at school (65% of boys, 50% of girls) 34% had a recent history of crime (5 1 % of boys, 17% of girls) 25% had a recent history of substance abuse (same for both genders) 18% had obvious psychiatric problems (22% of boys, 14% of girls) 16% had a recent history of violence (28% of the boys, 4% of girls Maybe a surprise to some, behavioural problems in this extreme population were not statistically related to a history of known parental substance abuse or mental illness or to experience of abuse before placement. Teenagers coming into care from more "normal backgrounds" were just as likely to exhibit behavioural problems as placed peers from troubled homes, at least according to information in case records (ibid.' lends some support to those who question practice notions of child maltreatment as a leading cause of delinquency and social maladjustment (McCord, 1983; Schwarz et al, 1994). Length of placements Information on lengths of placements, lengths of spells of care and placements patterns are scarce, particularly if one attempts to describe care on a national level. Yearly statistics give information on the average length of terminated placements. But a spell in care often consists of two planned placements, eg first a short time in residential care and then a longer placement in a regular foster home, which make yearly statistics misleading. Most information from research comes from small, local samples and worse - from studies done in the 1980ies (Vinnerljung, 1996b). There is impressionistic evidence that placement patterns have changed in the last five years, toward a more restrictive attitude to long term care. Children's "care careers" could tentatively be categorised in four groups: 1. Children who come into care for a short period and then return home for good. 2. Children who repeatedly moves from birth home to care and back, sometimes due to breakdown of placements 3. Children who come into care, and who remain in one foster home for a long time 4. Children who make repeated moves between placements within the care system, mostly due to breakdowns in placements. It is not possible to quantify these categories. All information from research, national and local statistics shows that the first group is dominant. Recidivism in the US is estimated to be around 30% (Fein et al, 1983; Fein & Staff, 1993). Reports on very small samples from the 1980ies point to rates between 25% and 50% in Sweden, but a larger cohort study is long overdue. The earlier quoted work on teenage placements in 1991, found that slightly less than 30% of teenagers under 17 who left care returned within 12 months (Vinnerljung et al, forthcoming/a). Research suggest that as many as 50% of placement breakdowns lead to family unification. Swedish social workers are often hesitant to press for a renewed placement when the care system has failed (Andersson, 1988, 1995; Socialstyrelsen, 1995b). Foster home placements are by legal definition temporary, although there has been a long ideological tradition of long term care being preferred by professionals (Vinnerljung, 1996a, -b). The same thing has been noted in many other countries (e g George, 1970; Colton & Williams, 1997). In practice, shortterm placements are the rule. Less than one in three foster children in cohorts born in the 1970ies - very likely fewer now - stayed in care for five years of more (Vinnerljung, 1996a,). Studies done on placements from 1983, indicate that a majority of infants coming into care remain for several years (Socialstyrelsen, 1990). In the study of teenage placements, median placement length was 11 months. Teenagers placed with relatives had longer, and more stable placements. Almost 70% stayed a year or more (Vinnerljung et al, forthcoming/a). Kinship care make up for 15-25% of all foster care placements, according to several local studies (Vinnerljung, 1996b). Current issues and trends in addressing the needs of children through the child welfare system To label current issues is to summarise what has been approached in Swedish child welfare during the second part of this century. We believe that the rapid change in child welfare during the last decades can be dated back to three important conditions. Firstly, the emergence of the welfare state after VMI reduced the number of children in public care from 20 per 1,000 to 10 per 1,000. Child neglect because of material poverty and misery eradicated as a reason for taking children from their families and placing them in institutions or foster homes. Secondly, the emerging family continuity principle in child care has dismantled the institutions for children - the whole family has entered care in Sweden. Extended family and social networks are of increased importance in care, meaning that the world of the extended family is looked upon as valuable when considering children in placement rather than people to blame. Humanistic values and democratic ambitions that entered child care has created a revolution in child welfare. Individualisation of care and treatment has resulted in affirmation of the individual child in care. The welfare legislation has successively adjusted to these tendencies, and is constructed to offer services according to people's rights rather than being part of charity. Current issues that point to future dilemmas to solve seems to be the importance of adjusting child welfare competence to the context of a multi-ethnic society. The overrepresentation in care of children with roots in other countries is a worrying signal of the state of condition of the migrants in our society. Especially the young offenders with a background in other countries are over-represented in correctional care. The Swedish integration policy during the last decades has created suburbs close to the big cities that are internationally unique in its composition. Usually in these neighbourhoods the Swedish citizens are in minority, and migrants from at least 60 countries (and language groups) are forced to live together. That planned composition of the neighbourhoods seems to endanger the natural development of neighbourhood network. Instead it becomes an important factor of an emerging social exclusion of migrants of the Swedish society. References Allmanna Barnhuset (1 95 5) Adoption. Stockholm: Allmanna Barnhuset/Medicinalstyrelsen/ Socialstyrelsen. Andersson G (1984) Sma barn pi barnhem [Young children in children's homes]. Diss. Malmo: Liber. Andersson G (1988) En uppfoljning av barn som skilts fran sina foraldrar [A follow-up of children who were separated front their parents]. Lunds Universitet: Meddelanden fran Socialhogskolan 1988: 1. Andersson G (1991) Socialt arbele med sma barn [Social work with young children]. Lund: Studentlitteratur. Andersson G (1993a) Socialtjansten och de utsatta barnen [Social Services and children-at-risk]. Socialmedicinsk Tidskrift, 1993/9-10, pp 465-470. Andersson G (1993b) Socialtjansten, barnpsykiatrin och de sma barnen [Social Services, Child Psychiatry Services and young children]. Psykisk Halsa, 93/2, pp 110-125. Andersson, G .(1993 c). Support and relief: the Swedish contact person and contact family program. Int J Social Welfare, Vol 2:54-62 Andersson G. (1995) Barn i samhallsvard [Children in public care]. Lund: Studentlitteratur Armelius B-A, Bengtzon S, Rydelius P-A, Sarnecki J & Soderholm Carpelan K (eds) (1996) Vard av ungdomar med sociala problem - enforskningsoversikt [Care for youth with social problems - a research review]. Stockholm: Liber Utbildning/SiS. Amsberg K-O (1998) Svenskar och zigenare [Swedes and gypsies]. Stockholm: Carlssons. Backe-Hansen E (1997) Nar barna er sma og beslutningen vanskelig [When children are young and decisions are difficult]. In Backe Hansen E & Havik T (eds) Barnevern pa barns premisser [Child welfare on childrens' terms]. Oslo: Ad Notam Gyldendal. Bamford F & Wolind S (1988) The pshysical and mental health of childen in care: research needs. London: NCB. Barth R (1992) Child welfare services in the United States and Sweden: different assumptions, laws and outcomes. Scandinavian Journal of Social Welfare, 1992: 1, pp 36-42. Bean P & Melville J (1989) Children of the empire. London: Unwin Hyman. Bebbington A & Miles J (1989) The background of children who enter local authority care. British Journal of Social Work, 1915, pp 349-3 68. Begler A-M & Wanell S E (1990) Omhandertagande av barn och ungdom [Children and youth in care]. Socialtjansten i Stockholm: F1/Internrevisionen, unpublished local report.. Bergmark A (1995) Prioriteringar i socialtjansten [Priorities of Social Services]. Stockholms Universitet: inst for socialt arbete, Rapport i socialt arbetc. Bergstrom B & Sarnecki J (1996) Invandrarungdomar pa sarskilda ungdomhem i Stockholms lan aren 1990 - 1 994 [Immigrant youth in homes for special supervision, in Stockholm county 1990-19941. In Armelius B-A, Bengtzon S, Rydelius P-A, Sarnecki J & S6derholm Carpelan K (eds) Vdrd av ungdomar med sociala problem - en forskningsoversikt [Care for youth with social problems - a research review]. Stockholm: Liber Utbildning/SiS. Berridge D (1985) Closing children's homes. Oxford: basil Blackwell, Berridge D & Brodie I (1998) Children's homes revisited. London: Jessica Kingsley Publ. Berridge D & Cleaver H (1987) Foster home breakdown. Oxford: Basil Blackwell. Bjornstrom U (1996) Children's rights in a dual-earner family context in Sweden. In wintersberger H (ed) Children on the wayfrom marginality towards citizenship. Childhood policies: conceptual and practical issues. International seminar in Montebello, Canada 16-20 Oct, 1995. Vienna: European Centre for Social Welfare Policy and Research, report 6111996. Bloom M (1 998) Primary prevention and foster care. Children and Youth Services Review, 2018, pp 667- 696. Broberg G & Tyden M (1991) Oonskade I folkhemmet. Rashygien och sterilisering i Sverige [Race hygiene and sterilization policy in Sweden]. Stockholm: Gidlunds. Browne K & Herbert M (1997) Preventing family violence. New York: Wiley & Sons. Christopherson J (1998) Social work students' perception of child abuse: an international comparison and postmodern interpretation of its findings. British Journal of Social work, 28, pp 5 7-72. Claezon I (1987) Battre beslut [Better decisions]. Diss. Umea Universitet: Studier i Socialt Arbete 5. Claussen S-E (1998) Klientekarrierer i barnavemet 1990-1993 [Client careers in child welfare 199019931. Tidskrift for Norsk Psykologforening, 3 5, pp 115-123. Cliffe D & Berridge D (1991) Closing children's homes. London: NCB. Cochrane A & Clarke J (Eds). (1 993). Comparing welfare states. Britain in international context. London: SAGE. Colton M (1 988) Dimensions of substitue child care. Avebury: Aldershot. Colton K, Drakeford M, Roberts S, Scholte E, Casas F & Williams M (1 997) Child welfare and stigma. Principles into practice. Childhood, 413, pp 265-283. Colton M & Vanstone M (1996) Betrayal of trust. London: Free Association Books. Colton M & Williams (eds) (1 997) 7he world of foster care. Aldershot: Arena. Denvall V (1994) For samhallets basta. Socialtjanstens medverkan i samhallsplaneringen [For the good of society. Social Services participation in social planning]. Diss. Floda: Zenon. Department of Health (1995) Child protection. Messages from research. London: HMSO. Dingwall R (1989) Some problems about predicting child abuse and neglect. In Stevenson 0 (ed Child abuse. Public policy and professional practice. London: Harvester & Meatsheaf. Dingwall R, Eekelaar J & Murray T (1983a) 7he protection of children. State intervention and family Life. Oxford: Blackwell. Dingwall R, Eckelaar J & Murray T (1983b) Times change and we change with them? Community Care, June 16, 1983, pp 18-19. Donzelot J (1 979) The policing of families. New York: Pantheon Books. Dominelli, L. (1999). El-Guebaly N & Offord D (1977) The offspring of alcoholics: a critical review. The American Journal of Psychiatry, 13414, pp 357-365. Ekberg J & Andersson L (1995) Invandring, sysselsaftning och ekonomiska effekter [Immigration, employments and economic effects]. Stockholm: Finansdepartementet, ESO, Ds 1995:68. Eriksson R & Aberg J (eds). (1 986) Welfare in transition. A survey of living conditions in Sweden 1968- 1981. Oxford: Clarendon Press. Farran D (1990) Effects of intervention with disadvantaged and disabled children: a decade review. In Meisels S & Shonkoff (eds) Handbook of early childhood intervention. New York: Cambridge University Press. Fein E, Maluccio A, Hamilton J & Ward D (1983) After foster care: outcomes of permanency planning for children. Child Welfare, LXII/6, pp 485-558. Fein E, Maluccio A & Kluger M (1990) No more partings. An examination of long-term foster care. Washington DC: Child Welfare League of America. Fein E & Staff 1 (1993) Last best chance: findings from a reunification services program. Child Welfare , LXXII/1, pp 25-40 Finkelhor D (1986) A source book on child sexual abuse. Beverly Hills, Ca: Sage. Furniss N & TiltonT (1977) 7he case for the welfare state. Bloomington, Ind: Indiana University Press. George V (1 970) Foster care. Theory and practice. London: Routledge & Kegan Paul. Giovanni J & Becerra R (1979) Defining child abuse. New York: The Free press. George R (1990) The reunification process in substitute care. Social Services Review, Sept 1990, pp 422 - 457 Gough D (1 993) Child abuse interventions. A review of the research literature. London: IIMSO. Gould A ( 988) Conflict and control in welfare policy. The Swedish experience. London: Longman House. Graham P, Dingwall R & Wolkind S (1985) Research issues in child abuse. Social Science & Medicine, 21/11, pp 1217-1228. Granath K-E (195 8) Foraildrar och fosterfolraldrar [Parents and foster parents]. Barnavard och Ungdomsskydd, vol 33, pp 88-96. Gray E & Cosgrove J (1985) Ethnocentric perception of childrearing practices in protective services. Child Abuse &Neglect, vol 9, pp 3 89-3 96. Harding L F (1 997) Perspectives in child care policy, 2'd edition. London: Longman. Haas L & Hwang P (1995) Company culture and men's usage of family leave benefits in Sweden. Family Relations, 44, pp 28-36. Hegeland H (1978) Barnhemsbarn. Min kollektiva uppvaxt [Child in a children's home. My communal childhood]. Stockholm: Natur & Kultur. Hegeland H (1988) SkyddsIingarna pa Ekedalen [The wards of Ekedalen. Goteborg: Tre Bocker. Hessle S (1985) Rikilinjer for psykosocialt arbete [Guidelines for psycho-social work]. Stockholm: AWE/Gebers. Hessle S (1988) Familjer i sonderfall [Disintegrating families]. Goteborg: Nordstedts. Hessle, S. (1 997). Child Welfare and Child Protection on the Eve of the 21" Century. Stockholm studies in social work, 21, Sth Univ, Dep of Social work Hessle S (Ed) forthcoming; The Multicultural neighbourhood as a village - How to build strong network with children and families Hessle S loka B & Yamano N (1996) Family policy and child welfare in Japan and Sweden. University of Stockholm: Department of social work, Studies in Social Work, 11. Hirdman Y (1989) Alt Iagga lives till ratta [To set life straight]. Stockholm: Carlssons. Holt M (1 992) The orphan trains. placing out in America. Lincoln, Nebr: University of Nebraska Press. Human Rights and Equal Opportunity Commission (1 997) Bringing them home. National inquiry into the separation of aboriginal and Torres Strait children from their families. Sydney: Human Rights and Equal Opportunity Commission. Hutchison E (1994) Child maltreatment: can it be defined? In Barth R, Berrick J D & Gilbert N (eds). Child welfare research review, volume 1. New York: Columbia University Press. Hwang P, Elden G & Fransson C (1984) Arbeisgivares och arbetskamraters attityder till pappaledighet [Attitudes toward men's parental leave among employers and colleagues]. G6teborgs Universitet: psykologiska inst, Rapport nr 1. Hyden L-C (1996) Applying for money. The encounter between social workers and clinents - a question of morality. British Journal of Social Work, 26, pp 843-860. Hakansson H & Stavne K (1 983) Jag kanner mig sa himla osaker. [I feel so dammed unsure…] Stockholm: Skeab Forlag. Hoglund Davila & Landgren Mo1ler (1991) Vilka barn omhamdertas av samhallet ? [Which children are taken into community care?] In Hoem et al (ed ) Livsfporloppsanalys [Life course analysis]. Orebro: SCB, Demografiska funktionen. Jeppsson Grassman E (1995) Third age volunteering in Sweden. Skriftserien, no 1, Stockholm: Skondalsinstitutet. Johnson B B (1981) The Indian Child Welfare Act of 1978: implications for practice. Child Welfare, LX/7, pp 435-446. Jonassen W, Clausen S-E & Kristofersen L (1997) En gang klient - alltid klient? [Once a client - always a client?]. In Backe Hansen E & Havik T (eds) Barnevern pa barns premisser [Child welfare on childrens 'terms]. Oslo: Ad Notam Gyldendal. Jonsson G (1973) Att bryta det sociala arvet [To break social heritage]. Stockholm: Tiden/Folksam. Kamali M (1997) Distored immigration. Clientization of immigrants in Sweden. Diss. Uppsala: Uppsala Multiethnic Papers, no 4 1. Kazdin A (1994) Psychotherapy for children and adolescents. In Bergin A & Garfield S (eds) Handbook of psychotherapy and behavior change. New York: John Wiley & Sons. Kindlund G (1986) Family policy in Sweden. Unpublished paper from the Ministry for Health and Social Affairs. Kirkwood A (1993) 7he Leicestershire inquiry 1992. Leicester:Leicester County Council. Koch L (1996) Racehygiejne i Danmark 1920-1956 [Race hygiene in Denmark 1920-19561. Kopenhamn: Gyldendal. Korpi S (1996) Samhallets insatser for ungdomar [State interventions for youth]. In Armelius B-A, Bengtzon S, Rydelius P-A, Sarnecki J & S6derholm Carpelan K (eds) Vard av ungdomar med sociala problem - enforsknings4Oversikt [Care for youth with social problems - a research review]. Stockholm: Liber Utbildning/SiS. Korpi V & Palme J (1 993) Socialpolitik, kris och reformer: Sverige i internatione belysning. Social policy, crises and reforms: Sweden in international comparisons]. Sartryck, no 389, Stockhoms universitet: Institutet for social forskning. Kristinsdottir G (1991) Child welfare and professionalization. Diss. Umea universitet: Social Work Studies 15. Kalvesten A-L (1974) 40 fosterfamiljer med Ska-barn [40 foster families with Ska-children. Stockholm: Almqvist & Wiksell. Lagerberg D (1998) Sexual victimization (invited commentary). Acta Paediatrica, 87, pp 130-131 Lagerberg D, Mellbin T, Sundelin C & Vuille J-C (1994) Growing up in Uppsala: the "new mobility" in the adolescent period. A longitudinal epidemiological study based on school data and some external sources. Acta Paediatrica, vol 83, supplement 3 89. Levin C (1 998) Uppfostringsanstalten. Om tvang i foraldrars stalle {The reformatory. Coercion in loco Parentis} Diss. Lund: Arkiv. Levy A & Kahan B (1991) The Pindown experience and the protection of children. Staffordshire County Council. Lindsay D (1992a) Reliability of foster care placement decision: a review. Research and Social Work Practice, 211, pp 65-80. Lindsey D (1992b) Adequacy of income and the foster care placement decision: using an odds-ratio approach to examine client variables. Social Work Research & Abstracts, 2813, pp 29-3 6. Lindstein T (1996) Working with children o alcoholics. Stockholm University: School of Social Work, Stockholm Studies in Social Work, no 10. Lundstrom T (1993) Tvangsomhandertagande av barn [Compulsory care of children]. Diss, Stockholms Universitet: Institutionen for socialt arbete. Lundstrom T (1994) Staten och det frivilliga sociala arbetet i Sverige [The State and voluntary social work in Sweden]. Stockholm: Skondalsinstitutet. Maitra B (1996) Child abuse: a universal "diagnostic category"? The implication of culture in definition and assessment. International Journal of Social Psychiatry, 42/4, pp 287-3 04. Mannes M (1995) Factors and events leading to the passage of the Indian Child Welfare Act. Child Welfare, LXXIV/1, pp 264-282. Marsh P & Crow G (1 998) Family Group Conferences in child welfare. Oxford: Blackwell. Maugham B & McCarthy G (1 997) Childhood adversitites and psychosocial disorders. British Medical Bullentin, 53/1, pp 156-169. McCord J (1983) A forty year perspective on effects of child abuse and neglect Child Abuse & Neglect, 7/1983, pp 284-289. Nixon P (1998) Exchanging practice. Some comparisons, contrasts and lessons from the practice of Family Group Conferences in Sweden and the United Kingdom. Protecting Children, 1414, pp 13-18. Noh Ahn H (1994) Cultural diversity and the definitions of child abuse. In Barth R, Berrick J D & Gilbert N (eds). Child welfare research review, volume 1. New York: Columbia University Press. Ohrlander K (1992) I barnens och nationens intresse [In the interest of the children and the nation]. Diss. Stockholm: Hogskolan for Lararutbildning, inst for pedagogik, Studies of Psychology and Education no 30. Parker R, Ward H, Jackson S, Aldgate J & Wedge P (eds) (1 99 1) Looking after children: assessing outcomes in child care. London: HMSO. Pfannenstill B (1948) Tattarna - en sociologisk grupp och ett socialt problem, II [The tinkers/travellers sociological group and a social problem]. Statsvetenskaplig Tidskrift, 1948, pp 316-335. Rose S & Meezan W (1997) Defining child neglect: evolutions, influences and issues. In Berrick J D, Barth R & Gilbert N (eds) Child welfare research review, volume 2. New York: Columbia University Press. Runquist W (1996) Adoptivbarn, ungdom och omhandertagen [Adopted, young and in care]. Stockholm: SiS, Rapport 1996:4. Rutter M (1989) Pathways from childhood to adult life. Journal of Child Psychology & Psychiatry, 30/1, pp 23-5 1. Ryburn M (1 993) Empowering practice in family placement. Paper presented at the International Foster Care Conference in Dublin July, 1993 Sallnas M (1995) Att utreda foraldraskap. Om modern institutionsvard for barn [To assess parenthood. On modern residential care.] Socialvetenskaplig Tidskrift, 2/4, pp 334-349. Salinas M (forthcoming) Barn och ungdomsinstitutioner [Residential care for children and youth]. Diss. Salonen T (1 993) Margins of welfare. Diss. Toma Hallestad: Hailestad Press. Sameroff A & Fiese B (1 990) Transactional regulation and early intervention. in Meisels S & Shonkoff (eds) Handbook of early childhood intervention. New York: Cambridge University Press. Samet J & Munoz A (1998) Evolution of the cohort study. Epidemiologic Reviews, 20/1, pp 1-14. Sanders R, Jackson S & Thomas N (1996) The balance of prevention, investigation and treatment in the management of child protection services. Child Abuse &Neglect, 20/10, pp 899-906. SCB: Demografiska rapporter 1990:1; Family formation and family dissolution in the 1980s SCB, 1994:2 Sven, det hair air Din ref. bar Du den? SCB, 1994: S 32 SM 9401 Schuerman J, Rzepnicki T & Littell J (1994) Putting families first. An experiment in family preservation. New York: Aldine de Gruyter. Schwarz I, Rendon J & Hsieh C-M (1994) Is child maltreatment a leading cause of delinquency? Child Welfare, LXXXIII/5, pp 639-655. Sinclair R, Garnett L & Berridge D (1 995) Social work and assessement with adolescents. London: NCB. Shaw M & Hipgrave T (1 983) Specialist fostering. London: Batsford/BAAF. Sjoberg B (1959) Fosterbam och fosterlegor [Foster children and foster-parents' fees]. Barnavard och Ungdomsskydd, vol 34, ppISI-159. Skoglund, A-M.(1992). Fattigvarden pa den svenska landsbygden ar 1829 [Poverty relief in rural Sweden 18291. Diss. Stockholms Universitet: Institutionen for socialt arbete. Socialstyrelsen (1959) Adoption. Stockholm: Socialstyrelsen (Allmanna Rad och Anvisningar 117/1959. Socialstyrelsen (1 990) Vard utom hemmet [Out-of-home care]. Stockholm: Socialstyrelsen (SoSrapport 1990:4). Socialstyrelsen (1994a) Barns vilikor I forandringstider. Slutrapport [Conditions for children in times of change. Final report.]. Stockholm: Socialstyrelsen (Socialstyrelsen foljer upp och utvarderar 1994:4). Socialstyrelsen (1994b) Barn och ungdomsinstitutioner. Hur paverkas virden av kommunernas ekonomiska situation? [Residential care for children and youth. How is care affected by the economic situation in the municipalities?]. Stockholm: Socialstyrelsen (Socialstyrelsen foljer upp och utvarderar 1994:10). Socialstyrelsen (1995a) Invandrarbarn i familjehem [Immigrant children in foster care]. Stockholm: Socialstyrelsen (SoS-rapport 1995:7). Socialstyrelsen (1995b) Sammanbrott I familjehem [Breakdowns in foster homes]. Stockholm: Socialstyrelsen (SoS-rapport 1995:9). Socialstyrelsen (1997) Tillampningen av lagen (1990:52) med sarskilda besttammelser om vard av unga [Application of the law on care of youth]. Stockholm: Socialstyrelsen (SOFTS 1997: 1 5, Allmanna rad). Socialstyrelsen (1998a) Anmalan, utrednning, insats [Report, investigation, intervention]. Stockholm: Socialstyrelsen (SoS-rapport 1998:4). Socialstyrelsen (1998b) Individ och familjeomsorgens arbete med ungdomar [Social Services' work with youth]. Stockholm: Socialstyrelsen (SoS-rapport 1998:5). Socialstyrelsen (1998c) Insatser for barn och unga 1997 [Children and youth subjected to measures 19971. Stockholm: Socialstyrelsen (Statistik, Socialtjanst 1998: 10, with English summary). Soydan H (1995) A cross-cultural comparison of how social workers in Sweden and England assess a migrant family. Scandinavian Journal of Social Welfare, 1995:4, pp 85-93. Spicker, P. (1988). Principles of social welfare. London:Routledge. Statistics Sweden, S 32 SM 9401 Sundell K (1997) Child-care personnel's failure to report child maltreatment: some Swedish evidence. Child Abuse & Neglect, 21/ll, pp 93-1 05. Sundell K & Haeggman U (1999) Familjeradslag i Sverige [Family Group Conferences in Sweden]. Stockholm: Stockholms stad, FoU-rapport 1999: 1. Sundell K, Humlesjo E & Carlsson M (1994) Att hjalpa sin nasta. En undersokningav kontaktfamiljer i Stockholm [To help your neigbour. A study on contact families in Stockholm]. Stockholm: Stockholms Stad, FoU-rapport 1994: 1 5. Sunesson S (1990) Familjevarden, en del av individ- och familjeomsorgen i socialtjansten [Foster family care, a part of the social services adminstration]. I Socialstyrelsen Sju perspektiv pi barns och ungdomars levnadsforhanllen [Seven perspectives on the living conditions of children and youth]. Stockholm: Socialstyrelsen (SoS-rapport 1990:5). Svedberg L. (1 994) En bok om marginalitet [A book about marginality]. Stockholms Universitet: Institutionen for socialt arbete. Tham H (1994) Trends among social assistance recipients in Sweden since 1945. Scandinavian Journal of Social Welfare, 3, No2, pp 75-84. Titmuss R M (1974) Social policy: an introduction. London: Allen & Urwin, UN (1989) Convention on the Rights of the Child. New York: United Nations. Upshur C (1990) Early intervention as preventive intervention. In Meisels S & Shonkoff (eds) Handbook of early childhood intervention. New York: Cambridge University Press. Usher C, Randolph K & Gogan H (1 999) Placement patterns in foster care. Social Services Review, March 1999; pp 22-36. Vinnerljung B (1989) Svarplacerade ungdomar [Hard-to-place youth]. Unpublished local inquiry, Municipality of Helsingborg. Vinnerljung B (1 990) Asociala ungdomar och socialtjansten - en lages bild [Asocial youth and Social Services- a report on current situation]. Unpublished local inquiry, Municipality of Malmo). Vinnerljung B (1 992) 235 syskon med olika uppvaxtoden - en retrospektiv aktstudie [235 siblings raised in non-shared environments - a retrospective case file study]. Lunds Universitet: Meddelanden fran Socialhogskolan 1992:5. Vinnerljung B (1993) Slaktingplaceringar i fosterbarnsvarden [Relatives as foster carers]. Socionomens Forskningssupplement, no 3, pp 3-10. Vinnerljung B (1996a) Fosterbarn som vuxna [Foster children as adults]. Diss. Lund: Arkiv. Vinnerljung B (1 996b) Svensk forskning om fosterbarnsvard En oversikt [Swedish research on foster care. A review.]. Stockholm: Liber Utbildning/CUS. Vinnerljung B (1996c) Hur vanligt ar det att ha varit fosterbarn? [A descriptive epidemiological study of foster care in Sweden]. Socialvelenskaplig Tidskrift, 313, pp 166-179. Vinnerljung B (1998) Socialt arv [Social heritage] In Denvall V & Jacobson T (eds) Vardagsbegrepp i socialt arbete [Common-sense-concepts in social work]. Stockholm: Norstedts Juridik. Vinnerljung B, Sallnas M & Kyhle-Westermark P (forthcoming/a) Sammanbrott vid placeringar av tonaringar i dygnsvard [Breakdown in placements of teenagers in care]. Vinnerljung B, Sallnas M & Oscarsson L (forthcoming/b) Dygnsvird for barn och ungdom 1983-1995 [Care for children and youth 1983-19851. Wahlberg, S. (1997). Samhallsarbete - Strategier for bevarande eller fornyelse? [Community work - a strategy for conservation or transformation], Diss, Stockholm University, Dep of Pedagogy Wallmark, J (Ed). (1998). Weightman K & Weightman A (1995) "Never right, never wrong": child welfare and social work in England and Sweden. Scandinavial Journal of Social Welfare, 1995:4, pp75-84. West M 0 & Prinz R (1 987) Parental alcoholism and childhood psychopathology. Psychological Bullentin 10212, pp 204-2 1 8. Ostberg F, Wahlander & Milton P (in press) Social utredningar om barn - en jamforande studie av sex kommuner [Child protection investigations - a comparative study of six municipalities].