Personal Assistance as a Strategic Game between State and Municipalities in Norway

The paper is based on a study of the law-making process of personal assistance in Norway. It discusses some tensions between the state and the municipalities concerning the type and quantity of services with regard to personal assistance. Does introduction of personal assistance imply an extended quantity of services or is it just another way of organising existing services? The signals from the state authorities are ambiguous, and when implementing the reform the municipalities are left to find a solution to this question. This might create a strategic game between the state and the municipalities. It might become an unpredicted side effect of the legislation of personal assistance that the municipalities will become more restrictive in the provision of personal assistance than they were in the experimental period.


Introduction
Personal assistance is an organised effort for disabled people where selfdetermination and control by the user is the important issue. In principle, this means that the user may choose his/her own helpers (personal assistants), decide what tasks they shall carry out, and decide their working hours. The user should be the manager for his/her assistants. Ideologically the arrangement has its roots in the so-called Independent Living-Movement, which appeared in the United States at the end of 1960s (Askheim & Guldvik 1999). Within the framework of user-control as an ideological principle, personal assistance has been designed in various ways in different countries. Askheim (2001) gives a comparative analysis of the arrangement in the Nordic countries.
Personal assistance has been tried out as an experiment in Norway since 1994. The experiment was conducted by the municipalities applying to a Government fund to establish personal assistance for named users, and being granted a fixed amount per user. The state financial support was supposed to cover the municipalities' additional expenditure for establishing and administering the arrangement to fulfil the condition of user-control. For instance, it was necessary to teach the users how to fulfil their obligations as managers, cover extra expenditure to recruit personal assistants and to provide guidance for the users when the arrangement had been implemented. The experiment was extended to include an increasing number of municipalities and users.
Evaluations of the arrangement during the experimental period concluded that personal assistance was a success. Important objectives were obtained, such as increased user-control, more flexibility in the services and a better quality of life for the users (Askheim 1999, Askheim & Guldvik 1999. The participating municipalities also seemed to be positive (Ressurssenter for Omsorgstjenester 1997, Guldvik 1998. A recently published study of the personal assistants' experiences also concludes that most of them appear to be satisfied with the arrangement (Guldvik 2001).
The general satisfaction of the arrangement as it was practised in the experimental period was an important factor when the politicians decided to make personal assistance a service, which all municipalities were obliged to include in their catalogue of welfare services. The spokesman for this matter in Parliament underlined this very clearly by pointing out that it is rare to find such comprehensive satisfaction with arrangements as in this case (Interview 1).
In the year 2000, personal assistance was incorporated in the Social Services Act as an arrangement the municipalities are obliged to include among their welfare services. It has become an explicit option in the Social Services Act section 4-2, letter a (the amendment is in cursive): "Social services shall comprise a. practical assistance and training, including user-controlled personal assistance, for those who are in special need of assistance owing to illness, disability, or age or for other reasons".
This integration in the Social Services Act entails offering personal assistance in accordance with the ordinary rules for the provision of services in this Act. The provision of social services is based on a discretionary assessment of needs. The conditions are discretionary both with reference to obtaining a right to services, what kind of services one may receive, and the quantity and quality of the services (Kj0nstad & Syse 1997).
Even though personal assistance is integrated in the Social Services Act, one element of the experiment is maintained. Municipalities can still apply for financial support to establish the specific service of personal assistants. But, because of the legislation the municipalities are not allowed to refuse establishing personal assistance even if they do not receive state support. The plan is that the arrangement will be integrated in the framework contributions from the state to the municipalities in the future (Sosial-og helsedepartementet 2000).
Even though the experiment has been a success, there are some tensions and possible conflicts of interests related to personal assistance as a welfare service (Askheim & Guldvik 1999). These factors might be stronger and more visible when personal assistance becomes an every-day arrangement as part of the ordinary municipal services. One of these is related to types and quantity of services being provided. Does the introduction of personal assistance purport an increased quantity of services or is it just another way of organising existing services? The answer to this question will have different financial consequences, and thereby actualise a possible conflict between the state and the municipalities. If an implication of the reform is increased expenditure, who is going to pay? In the paper I will show how the reform of personal assistance has brought to the surface a gap between the municipalities' actual practise in providing traditional social services and the intentions of the Social Services Act. If seen in relation to the intentions of the Act the introduction of personal assistance should not necessarily lead to an increase in the quantity of help, but if seen in relation to the actual municipal practise it might do. Furthermore, I will outline how the state authorities have ignored this as a matter for the state by handing the responsibility for solving the problem over to the municipalities. This may create a process of strategic games between the state and the municipalities, and will probably lead to variations between the municipalities in the implementation of the reform. I will discuss how the municipalities may react to this complex situation in their adaptation of the services. Finally, I will outline briefly the possible consequences for the reform and for the users of personal assistance.
The paper is based on a study of the decision-making process related to the legislation of personal assistance in Norway. The data material has consisted of documents related to the development of the experiment from 1994 and onward, and the law-making process in the years 1999 and 2000. In addition, I have conducted interviews with eight important actors during the process. The informants represent the administrative and political leadership in the Ministry of Health and Social Affairs, and the spokesman for this matter in Parliament. Furthermore, I have interviewed a representative of the Association of Local Authorities, representatives of three different user organisations, and finally a representative of a trade union organising some of the personal assistants. The data material is presented more completely in the published report from the study (Andersen 2001). In this paper the empirical material being used is documents related to the law-making process, but I will also touch briefly on some of the interviews. However, the main part of the paper is based on generally available research on personal assistance and the relationship between state and municipality.
An underlying conflict of interests between the state and the municipalities The municipalities have become increasingly more important as institutions for providing welfare services to the citizens (Rose 1996). Thereby, the integration between the state and the municipalities has become closer, resulting in inter-dependency between the parties (Kjellberg 1988). The state has become more and more dependent on the municipalities as implementers of national goals. At the same time the municipalities are dependent on the state for financial support to be able to implement their statutory obligations (Montin 2000).
The interaction between the state and the municipalities under this condition of inter-dependency has an underlying conflict of interests related to the question of responsibility and financial capacity. A frequent criticism is that the municipalities do not fulfil their statutory obligations towards the users of services (e.g. Eskeland 1993). The services are not good enough and the municipalities are too restrictive in their provision of services. The municipal authorities' standard response to this criticism is that they do not get sufficient compensation for their increased expenditure in connection with welfare reforms that the state authorities impose on them (e.g. Kleven 1997). They also claim that the state authorities generally impose on the municipalities a greater responsibility and more tasks than they have financial capacity to carry out (e.g . Oppedal 1993;St. meld. nr. 23, 1992-93). When doing so they blame the state for deficient services and a restrictive practise. The state authorities' argument against this is that the municipalities have a statutory duty and that they are given financial compensation for their extended responsibility in connection with new welfare reforms. In addition, the state authorities argue that the municipalities must become better at making priorities between different tasks, they must be more efficient, and they have to develop their ability to reorganise the use of resources in accordance with changing needs (St. meld. nr. 23, 1992-93). The result is that when the users complain about the municipal services the local authorities blame the state, while the state authorities throw the responsibility back to the municipalities. Thereby, we can talk about "a game" between the state and the municipalities in the sense that both parties try to blame each other (Hagen & Sørensen 1997). The relationship between the levels of government may be seen more as dominated by strategic games and bargaining, than by mutual recognition and co-operation (Rattsø & Sørensen 1997). Bernt (1994) has stated that it is a game, which in a cynical way may be described as the art of reducing welfare services without being held responsible for the consequences.
One important factor creating this process of bargaining and the strategic games is the complexity in estimating the financial consequences of implementing new reforms. As a consequence of the municipalities' large responsibility in the welfare policy, new reforms will often appear within policy areas where they already have a certain responsibility. Such reforms can reveal a gap between the municipalities' statutory obligations and their actual practise in the policy area (Andersen 1997).
Thereby, the municipal expenditure is not only related to the extended formal responsibility but also to adjusting established practise to the intentions of the law.
The problems related to gaps between intentions and realities may be more visible when legislation has the character of a framework Act. In such types of legislation the users' substantial rights to services are not decided in the Act, but are developed through practise (Astrom 1998a). The substantial content of the politics will be decided in the local implementation. The provision of services according to the Social Services Act is an example of such an arrangement, because the provision of social services is based on a discretionary assessment of needs. Thus, there will be a discussion of what kind of services and which quantity of services is necessary to obtain the objectives in the Act.
Also, reforms targeting certain groups of users will often carry increased ambitions related to the services, either as an integrated element in the reform or as a consequence of increased demands and expectations from the users. These kinds of situations make it more complicated to estimate the actual financial burdens related to welfare reforms and may create conflicts between municipal and state authorities.
In what way have such mechanisms been in action in the reform of personal assistance? Before I discuss this question further, I have to present a distinction between different types of services for disabled persons that I am going to use in the analysis.

Typologies of services for disabled persons
In the discussion about types of social services provided for disabled persons I will use a distinction made by Astrom (1998b), between compensating and developing efforts.
"Compensating efforts" aim to compensate for something the persons cannot do. We are here referring to efforts that exclusively, often in a physical sense, compensate for something the individuals cannot do by themselves. These types of services will mainly be related to practical assistance in or in relation to the home of the users, for instance home based services like cleaning the house. The services will be of vital importance for the users' daily life, but do not promote participation in society. In short we can say that compensating efforts represents services taking care of the users' basic needs of practical assistance in their everyday life.
"Developing efforts" aim to compensate for an impairment but also to give disabled people the same opportunities as non-disabled to live their own life in the community with others, for instance by receiving assistance outside their own home with the purpose of social integration. These types of services will to a large extent encompass help to enable the users to lead active lives outside their own home in social relationship with others. The most common social service with this intention is 'personal support' (st0ttekontakt). Developing efforts try to support the individuals in realising their own interests and abilities and make it possible to take part in social life. An important objective with these kinds of services is to include disabled persons in society.
In an arrangement like personal assistance, the user as manager can decide what kind of tasks the assistants shall carry out. For users with physical impairments the possibility of having a real choice in this matter is dependent on the quantity of services being offered. The quantity of help will be of vital importance with regard to what extent developing aspects can be achieved. With only a few hours of assistance the services will be tied to compensating efforts to meet the users' basic needs of practical assistance. As the Norwegian Association of the Disabled stated in the hearing: "It is important to remember that the arrangement is meant for people with extensive needs for help. Because of that it is not possible for the users to choose to do practical things in the household themselves, and use the assistants exclusively to activities outside home" (Høringsuttalelser 1999). With an increase in hours the users will have a greater opportunity to redirect some of the assistance towards social activities, holidays and so on. The opportunities to give priority to such activities are greater when the hours increase. One user illustrates the point in this way: "Just to come out and do exactly what I want without having to ask other people like family or friends for help. I have the opportunity to be out among people instead of being isolated" (NHF Oslo 2001: 24).
Thus, the opportunity for the users to take active part in social life outside

SJDR -Volume 5, No 2 -2003
their own home is dependent on the quantity of assistance exceeding the basic need for practical assistance in their homes. This is a necessary precondition for the users' opportunities to take active part in society.
With personal assistance now being integrated in the Social Services Act, the rules for the provision of services in this Act will be of importance. How can the rules be interpreted with regard to these two main types of efforts?

The gap between the intentions of the Social Services Act and the actual municipal practise
When we try to uncover the intentions of the Social Services Act, we may look at the preamble in the Act: "The purpose of the present Act is to contribute to giving individuals opportunities to live and reside independently and to achieve an active and meaningful existence in community with others" (Section 1-1, letter b).
According to judicial literature, the preamble expresses an underlying opinion that there is not just only the most basic needs for care that should be met (Kj0nstad & Syse 1997: 139). In addition, people needing help should be secured a service encompassing other aspects of life. The Social Services Act has a broader social dimension, related to the basic welfare state idea of integration of the individual citizen in society. Social services have to be provided in a form and at a level, which gives the client an opportunity to function as an autonomous human being in a social community (Bernt & Kjellevold 2000). Thus, the formulations in the preamble indicate that the help offered should also include "developing efforts". Generally, such formulations in the preamble do not impose any legal duties on the authorities, and they do not give individuals any specific legal rights to services (Kj0nstad & Syse 1997:37). However, the preamble may still have a legal importance because it will function as guidance when using discretionary powers.

The Ministry of Health and Social
Affairs also refers to the preamble when arguing that "developing efforts" are in accordance with the intentions of the Act. Furthermore, The Ministry emphasises that there is no formulations in Section 4 -2a indicating that practical assistance should only be restricted to the user's home. In the explanatory comments to the Act it is stated that the services "mainly" consists of help related to the user's own home (Sosialdepartementet 1993: 107).
The interpretation of the intentions of the Social Services Act both in judicial literature and by the central authorities is that it includes what I have called "developing efforts". But, if we make a distinction between the intentions of the Act on the one hand, and the actual municipal practise on the other, the picture becomes more complicated.
In my opinion the experiment with personal assistance and the law-making process have brought to the surface a gap between the intentions of the law and the actual municipal practise with regard to providing traditional services based on the Social Services Act. This gap is caused by the traditional services mainly being concentrated on compensating efforts and therefore not been in accordance with the intentions, which also place emphasis on developing efforts. The argument for this conclusion is twofold, firstly based on statements from actors in the process and secondly on the experiences from the experimental period with extended help when receiving personal assistance.
There are statements indicating that traditional services offered by the municipalities in accordance with the Social Services Act, have consisted of practical assistance in the users' homes and have been limited mainly to "compensating efforts". At the start of the experiment with personal assistance the state authorities' argued that the ordinary social services were related to the users' homes. In the law-making process, several of the hearing institutions also claimed the same thing (Høringsuttalelser 1999).
The second argument is a bit more complicated, but the starting-point is the fact that many of the users have obtained an extended service when they received personal assistance. As many as 85 percent of the users said that they obtained more help when they received personal assistance (Askheim & Guldvik 1999). The same tendency is also found in a study of personal assistance in the city of Oslo (NHF Oslo 2001).
There is no empirical evidence to come to a definite conclusion concerning why so many municipalities have extended the services when introducing personal assistance. A hypothesis is that the municipalities in addition to necessary "compensating efforts" also have accepted "developing efforts" as reasonable when introducing personal assistance. This has led to an increase in the total quantity of services. A lot of municipalities have used the financial contribution from the state to finance the extension of the services, and this has made the extension easier (Ressurssenter for Omsorgstjenester 1997, Guldvik 1998). I will give some reasons for supporting this hypothesis.
From the very beginning of the experiment, personal assistance was linked to disabled persons' opportunity to achieve an active participation in society. It was pointed out that traditional municipal services were given in the users' home, and that personal assistance was especially suited for disabled persons with a need for practical assistance outside their own homes (Regjeringens handlings-plan for funksjonshemmede 1994-97: 50). Furthermore, the objectives were to ensure flexibility in the services, to make the users independent, and to fulfil the principle of user-control by making the users managers of the service. Personal assistance was especially related to assistance outside the users' homes as a means to obtain the goals of participation and equality for disabled people in society. Therefore we can say that the arrangement was marketed to include "developing efforts". This was also emphasised in the law-making process, both from the Ministry and from politicians in Parliament (Sosial-og helsedepartementet 1999: 11). For example the spokesman for the matter in Parliament underlined that disabled people might be able to create a social network and thereby prevent isolation and loneliness with the help of their personal assistants, and "the user may have help to tasks that the home helps and the personal supporters normally do not carry out" (Odelstinget 14th December 1999).
Several of the hearing institutions emphasised that if the purpose of promoting an active life outside the home should be achieved a provision for personal assistance would also involve an extension of the total services (Høringsuttalelser 1999). They underlined that the need for practical assistance in the home would not decrease by introducing personal assistance, and therefore the quantity of services would have to increase to secure "developing efforts" in addition to "compensating efforts".
This also corresponds with the experiences from Sweden when introducing personal assistance. In Sweden two particular Acts came into force in 1994, where personal assistance was established as an individual right for users who qualified for the service: "The Act concerning Support and Service to certain groups of disabled people" (LSS) and "The Act concerning assistance compensation" o (LASS). In a case study Astrom (1998b) compared the development both in the quantity and type of services provided to two different groups of users. Both groups received services in accordance with the Social Services Act in 1993. Two years later one group was provided services according to LSS, while the other got their services based on the Social Services Act. It was a significant higher number of the first group that had obtained an increase in the quantity of services than among the users who still received services in accordance with the Social Services Act. The same significant difference between the two groups was also seen in the types of services. A higher number of the LSS-group received developing efforts than the other group. Even if there are no legal differences between the two Acts with regard to the kind of help offered, more people obtained "developing efforts" when they received help according to LSS.
Through the implementation of personal assistance in the experimental period, a lot of municipalities have probably followed the signals from the state authorities about encompassing "developing efforts" in the arrangement. This is in accordance with the intentions of the Social Services Act as formulated in the preamble. At the same time, this has brought to the surface the fact that municipal practise in providing traditional social services has not been in accordance with the intentions of the law. The introduction of personal assistance has therefore revealed a gap between the intentions of the law and municipal practise. On the whole, the municipalities are probably more restrictive in their quantitative provision of traditional services to the users than the intentions of the Social Services Act indicates. This means that introducing personal assistance has revealed an under-consumption of social services in general.
We have also seen that the ambitions with regard to types and quantity of help have increased when introducing personal assistance. The reform has been marketed especially to meet the users' need for help to participate in society, in addition to the traditional services more directed to practical assistance in the users' homes. In this way we may claim that increased ambitions were an integrated element in the reform. But, this is related to actual municipal practise, and not to the intentions of the Social Services Act. Thereby, the state authorities' signals towards the municipalities are ambiguous with regard to the question of the quantity of services. Referring to the existing statutory responsibility, the level of ambition regarding the quantity of services is not intended to increase. However, when referring to traditional municipal practise an increase is seen as desirable. It is also a fact that the users have increased expectations to the reform, compared to traditional services. The user organisations have pointed out that obtaining an active life outside the home has to involve an extension of the total services.
The spokesman for the matter in Parliament also admitted that the reform might lead to more assistance being given because there had been an under-consumption of assistance to participate in social and leisure activities. But, she was more uncertain about the effects in the long run because in a perspective of "help to self-help" the need for assistance might decrease (Interview 1). In my opinion the experiences from Sweden indicate that the effect may be the opposite, that there might be an increasing demand for assistance. One explanation for this is that the users with the help of their personal assistants have improved their living conditions more than they expected in the beginning by a greater opportunity to live active lives. This has lead to a change in the attitudes and a confidence in the arrangement from the users, leading to a demand for an increase in assistance (Bengtsson & Gynnerstedt 2001).
This is a situation that may create conflicts between municipal and state authorities. How the state authorities handle this complex situation will be of importance for the development of the relationship between the levels of government. To what extent does it create a game-like relation between the municipalities and the state?
The state authorities have handed the responsibility for solving the problem over to the municipalities When a reform process reveals a gap between the formal responsibility the municipalities have according to the law and their existing practise in the policy area, this may be handled in different ways by the state authorities. They can choose a formal-judicial approach by arguing that the municipalities have the responsibility to ensure that their actual practise is adjusted to the intentions of the law. It is the municipalities' own business to solve the problem of a discrepancy between intentions and reality. Additional expenditure to make sure that actual practise corresponds with the formal-judicial responsibility must be met by the municipalities. Another approach will be just to recognise the fact that the municipalities' actual practise is below the acceptable level and then to stimulate to an increased level of services by improving their financial opportunity to reduce the gap between actual practise and intentions. For instance, the latter was done in the reform for intellectual disabled people in Norway in 1991. The municipalities were granted financial support from the state to improve their services to intellectual disabled people living at home, even though they already had the statutory responsibility to provide services to these people (Andersen 1995).
The choice of strategies in this matter will be of decisive importance because they will have different financial consequences for the municipalities. This again will influence the municipalities' adaptations and thereby influence the services provided for the users. Using the statutory responsibility as an argument to instruct the municipalities will have a limited influence when confronted with financial limitations, especially when it comes to the quantity of services (Andersen 1997;Bernt 1997). At the same time it is the arguing about the gap between tasks imposed on the municipalities and their actual financial situation that creates the strategic game between the levels of government.
In connection with the arrangement of personal assistance the state authorities have assumed a position based on a formal-judicial argumentation. The Ministry of Health and Social Affairs has stated clearly that the municipalities have the responsibility to adjust their practise in accordance with the intentions of the Social Services Act, and to cover the expenditure followed by that responsibility. The Ministry emphasised that the amendment to the Act, making personal assistance a statutory duty, did not give the users any legal right to extended services. The municipalities were already obliged to provide necessary practical assistance outside the users' homes before the amendment to the Act was approved. Thus, the inclusion of "developing efforts" when deciding the total quantity of services was from the outset seen as an integral part of the rights laid down in the Social Services Act. The Ministry also emphasised that the municipalities have a full right to provide other services or better services than what they are obliged to. If personal assistance leads to an increase in the quantity of help because of efforts outside the home, this will be something the municipalities are obliged to anyway, or something the municipalities can provide without being obliged to (Interview 2).
The Ministry stated that it is the municipalities' own responsibility to find a reasonable balance. If the municipalities choose to extend the quantity of help when providing personal assistance, it is their own business: "Such an increase in municipal expenditure is a consequence of the general legislation and do not imply any particular state efforts" (Ot. prp. nr. 8, 1999(Ot. prp. nr. 8, -2000. This was also the background for the flat financial contribution from the state to the municipalities for implementing the arrangement, because the intention of the financial contribution was only to compensate for the additional expenditure related to establishing and administrating the arrangement. Furthermore, the state authorities have also argued that the municipalities are free to expand the quantity of services beyond their statutory duty. This is done by referring to the fact that the functions of the municipalities are negatively defined, which means that the municipalities may take responsibility for any task they like, unless prohibited by law (Larsen & Offerdal 2000). The municipalities are likely to meet such arguments by stating that this is only a theoretical freedom without practical relevance today, because different kinds of statutory obligations and state regulations are completely tying up their financial resources.
The process of incorporating personal assistance in the legislation has not solved the question of the quantity of the services. The state authorities want the municipalities to increase the quantity of services to be in accordance with the intentions of the Social Services Act, but they have to cover the expenditure themselves. What will be the possible consequences of this strategy from the state in implementing the reform in the municipalities?

How the municipalities may react to the problem of adapting the services
The state strategy has created opportunities for different municipal adaptations, or ways of dealing with the gap between legal responsibility and local practise. Seen in this way there is room for local priorities, which will open up for considerable variations between the municipalities. At the same time this also gives an opening for a strategic game between the state and the municipalities. Here I will discuss three possible municipal adaptations following the amendment to the Social Services Act.
One possibility is that personal assistance will function as a catalyst for an expanded practise regarding the provision of social services in general. If the preamble in the Act is given greater importance when practising discretion, "developing efforts" may be included in the provision to a larger extent than today. This will mean that the practise of the municipalities will be closer to the intentions of the Act. This adaptation will be in line with the hopes of the Ministry of Health and Social Affairs. We have seen that the Ministry has tried to use the reform of personal assistance to "force" the municipalities to adjust their practise to be more in line with the legal intentions. Through this reform the Ministry can bring to the surface the fact that the municipalities do not fulfil their general obligations in providing social services. At the same time this is probably the least likely result because it will be the most expensive adaptation for the municipalities.
A second possibility is that the municipalities continue the practise followed by a majority during the experimental period, namely to limit the extension of the services to personal assistance. Through the provision of personal assistance they include "developing efforts" to a larger extent than in traditional social services. This solution leads to a discriminatory treatment between the recipients of personal assistance and recipients of traditional social services. It corresponds with the Swedish experiences referred to earlier, that a number of the users receiving services based on the particular Act (LSS) obtain extended help compared to users receiving help pursuant to the Social Services Act. However, an important difference between Sweden and Norway is that users in Sweden may receive personal assistance based on a more specialised Act, while in Norway users receive services in accordance with the more general Social Services Act. It is probably easier to "discriminate" against users receiving services based on different Acts, than to do it to users receiving different kind of services based on the same Act. Such "discrimination" may also create reactions among users of other social services, because they may feel that they receive unfair treatment compared to the recipients of personal assistance. At the same time this adaptation will, at least in the short run, be made easier by the fact that personal assistance still has a special position because of the possibility to obtain financial contribution from the state.
To consider personal assistance as a particular service with more liberal criteria for provision than other social services is probably not in accordance with the state's intentions. However, municipalities continuing along this line may still argue that this is a consequence of the state's reform strategy. The state authorities "force" the municipalities to discriminate between users of social services. One argument is that the state authorities already from the beginning of the experiment gave clear signals about including "developing efforts" in the services, and that the municipalities followed this up in relation to personal assistance. When personal assistance was included in the Social Services Act this alone should not get any consequences for the quantity of other social services, because the reform did not encompass any general changes of the criteria for providing other social services. Neither was any financial compensation given.
A third possibility is that the municipalities are being more restrictive when they consider the quantity of help to recipients of personal assistance, by adjusting the criteria used to assess the ordinary social services. This means that they mainly consider "compensating efforts" as sufficient also in the provision of personal assistance. It might become an unpredicted side effect of the legislation that the municipalities will be more restrictive in the provision of personal assistance than they were in the experimental period. Taking the municipalities' financial situation into consideration this adaptation will probably be the most likely. The municipal authorities may in this situation also argue that the state has the responsibility because the municipalities have not been granted financial support to be able to continue a high level of services. Furthermore, it might be argued that a consideration of equality and justice towards all recipients of social services in the municipality makes it impossible to provide the users of personal assistance with more extensive services than other users. The general situation of tight budgets forces the municipalities to assume a restrictive practise. From the recipients of personal assistance' point of view this will of course be the most negative solution, and may create dissatisfaction among these users. This is a fear also shared by the organisations for disabled, and it is a decisive element in their demand for a state co-financing of the arrangement (Høringsuttalelser 1999). However, this effect may be postponed because of the preliminary continuation of the state financial contribution to the arrangement.
Can the right of appeal against municipal decisions counteract a more restrictive local practise?
The Social Services Act includes a guaranty of rights by giving the user a right to appeal to the County Governor against decisions made by the municipal social services. The right of appeal includes both refused applications for personal assistance and complaints about the total quantity of services. Is it possible to imagine that this right of appeal may counteract a more restrictive municipal practise and also reduce the variations between different municipalities in their provision of social services?
With regards to the total quantity of social services the municipalities must ensure that needs are being met at an "acceptable level" (Sosialdepartementet 1993: 112). This is also the formulation that the County Governor must consider when considering a complaint. This is a very discretionary formulation and opens up for a variation in the management of the appeals among the different County Governors. There are not many studies of the County Governors management of appeals with regard to the Social Services Act. One study of financial support conducted by Andenaes (1992) concluded that there were differences among the County Governors in their management of the appeals that could not be explained by differences in the cases. The actual practise in the matter of personal assistance is not yet studied. But, the statements from the County Governors in the hearing-process indicate that they do not consider it to be a simple process to manage the appeals (H0ringsuttalelser 1999). The hypothesis is that the County Governors management of the appeals will not be effective in protecting the users right to have an extended service, and that there might be variations among the appeal institutions.
The Social Services Act can be characterised as a framework Act where the users' substantial rights to services are developed through practise. In the debate about personal assistance the user organisations have wanted a more comprehensive opportunity for the County Governor to overrule municipal decisions (Interview 3 and 4). But, for such a system to function satisfactory it is probably necessary to define the rights to services more concrete. This may be a problem regarding social services because there is a risk that defining a minimum level may give too little assistance for people with great needs, or that the right do not meet the needs in an accurate way (Bernt & Kjellevold 2000). Assistance must be provided on the basis of individual and concrete assessment of needs. However, there might be possible to create more specific rights, for instance by formulating standards for livingconditions which the user's situation may be evaluated against (Bernt & Kjellevold 2000).
However, the municipal authorities will not accept more detailed standards for services or a more comprehensive control function from the County Governor. The Association of Local Authorities argues that the County Governors in practise already exert too much control over the municipal authorities. The Association also points to the fact that there are great variations between the County Governors in their management of the appeals (Interview 5). The concern about local autonomy was also the argument used when it was decided on a limited opportunity for the County Governor to overrule municipal decisions in the Social Services Act (Ot. prp. nr. 29, 1990-91).

Conclusion
The arrangement of personal assistance has been a definite success in the experimental period in Norway. The evaluations are positive with regard to user-control, flexibility, and the satisfaction of the users. A lot of the users have got an increased quantity of services when receiving personal assistance compared to their traditional services. This fact has probably given the users greater opportunities to practise user-control, because their hours of help have not solely been occupied by basic compensating efforts.
In the process of including the arrangement in the Social Services Act, the state authorities have not taken this aspect into consideration in their reform strategy. We have seen that the signals from the state authorities have been ambiguous with regard to the question of quantity of services. Referring to the existing statutory responsibility, the level of ambition regarding the quantity of services is not intended to increase. However, when referring to traditional municipal practise an increase is seen as desirable. The law-making process has brought to the surface a gap between the intentions of the Social Services Act with regards to type of services being given and the municipalities' actual practise. The state authorities have chosen a formal-judicial approach to this problem; by making it clear that it is the municipalities' own responsibility to solve this. The consequences may be a game-like situation between the state and the municipalities in the implementation of the reform. The municipal authorities may respond to this by being more restrictive towards the users and place the responsibility on the state's limited financial contributions to the municipalities. From the users' point of view this game of responsibility may be experienced as an attempt from the authorities to renounce their obligations as providers of social services. Thereby, it might become an unpredicted side effect of the legislation of personal assistance that the municipalities will be more restrictive in their provision of personal assistance than they were in the experimental period. If the quantity of services is restricted to purely compensating efforts the consequences might be a considerable weakening of important positive elements in the reform, such as self-determination, flexibility, and the opportunity to get a more active life in different areas of society.
One can argue that the problems of possible gaps between intentions and reality are "in-built" in the framework Acts. Thus, the arrangement will have a tendency to create conflicts and strategic games between the state and the municipalities. The user organisations have primarily wanted a more right-orientated Act for personal assistance, for instance incorporated in the social security legislation (Interview 3). In this way the elements of strategic game might disappear with the responsibility being placed on the state authorities. However, relating to the level of services such change will carry no guaranties because the criterions for assessment of needs may change. We have seen tendencies of this in Sweden where financial arguments have been important for the state authorities in trying to reduce the level of services by adjusting the criterions for assessment of needs (Askheim 2001). Besides, such an arrangement will not remove the discretionary assessment of needs related to individual users. Therefore, a state responsibility represents no guaranty for a faire and acceptable level of services.