Professional Archetype Change: The Effects of Restricted Professional Autonomy

One of the points on which researchers agree is the centrality of autonomy to professionalism. Moreover, a common conclusion in the studies of professions is that the profound changes in society over the last fifty years have threatened the autonomy and changed the archetype of professionalism. This paper contributes to the research on changes and continuities, challenges and opportunities for professionalism by discussing advantages and disadvantages of restricted professional autonomy. By describing the historical development in the Swedish and Canadian healthcare context, two major findings are discussed. First, although medical professionals have been subjected to certain constraints, they still appear to maintain a relatively high level of autonomy concerning the technical content of the work. Second, restricting professional autonomy is not negative merely due to the preservation of the professional archetype; rather, a «reasonable» limitation can be positive if professional autonomy is understood as a contract based on public trust.

or proletar-ianisation (Turner, 1993). Scott and his colleagues (2000) described the changes in the American healthcare field primarily in terms of changes in governance structures to demonstrate how the developmentfrom professional dominance to federal involvementand current managerial control and market mechanisms have greatly undercut the autonomy of physicians and eroded much of their professional dominance. The sheer fact that a greater reliance is put on hierarchical and managerial controls today indicates that alternative preferences with respect to how professionals should be organised and controlled as well as how their performance should be evaluated and rewarded exists in today's society (Brock et al., 1999). As Evetts (2003, p. 396) puts it: «Professions […] have been perceived as under threats from organisational, economic and political changes, [and] are portrayed as experiencing a reduction in autonomy and dominance, [and] a decline in their abilities to exercise the occupational control of work».
Fundamental to the above description of the changes challenging pro-fessions is the focus on professions as passive victims and professional autonomy as something worth protecting and preserving (Evetts, 2009). However, from the view of professionalism described as an archetype, an occupational value (Freidson, 2001) or a discourse (Fournier, 1999), the challenges and the contrast between professional and other occupational principles might not always be polar opposites and mutually exclusive (Bourgeault, Hirschkorn and Sainsaulieu, 2011). Instead, there might be opportunities associated with these changes which might improve both the conduct and the practices of professional work and perhaps even strengthen the archetype of professionalism (Evetts, 2011). In this paper, I try to contribute to the current and requested research on changes and continuities, challenges and opportunities for professionalism as an archetype or an occupational value by presenting a discussion of the advantages and disadvantages of restricted professional autonomy. By describing and comparing the historical development of one of the traditional professionsnamely, the medical profession in Sweden and Canada, I argue that when discussing the effects of restricted professional autonomy we need to take into consideration that professional autonomy is a social contract between a profession and policy elites to act in society's best interests.

Theoretical issues
The focus in this paper is on the advantages and disadvantages of restricted professional autonomy. In order to be able to investigate this, we need to clarify the concept of professional autonomy and for whom a restriction of professional autonomy may be negative or positive.

Professional autonomy
One of the key features of the archetype of professionalism is that the specialised work performed by the members of a profession is grounded in a body of theoretically based, discretionary knowledge and skills (Abbott, 1988;Freidson, 2001;Parson, 1968). Discretion can be described as a form of practical reasoning where the aim is to reach conclusions about what ought to be done in particular cases, but where warrants or rules of inference are weak (Molander and Grimen, 2010). Thus, having discretion means possessing a restricted and protected space in which the liberty to solve complex problems in accordance with a normative frame, abstract concepts and theories is provided. Hence, standardisationor, as Abbott (1991) puts it, 'commodification'of professional tasks is problematic while selfcontrol or professional autonomy is simultaneously an essential characteristic of the professional archetype (Greenwood, 1957;Meiksins and Watson, 1989;Wilensky, 1964). However, the concept of professional autonomy has been used in a variety of different ways.
First, professional autonomy can be understood as the existence of two separate yet intertwined levels: autonomy with respect to the individual professional and autonomy with respect to the profession as a group (Engel, 1969). On the collective level, the basic argument for self-regulation is that the profession is the one and only entity that can define and judge the competence and actions of its members (Eraut, 1994). Consequently, autonomy can be defined as legitimate statesanctioned control over standards of entry as well as mores and habits that develop around the work. On the individual level, autonomy consists of the professional's opportunity to decide the technical content of workthat is to say, the ability to judge, control and plan one's own work as well as the freedom to deal with the clients (Engel, 1969;Freidson, 1988). Freidson (1988) refers to the collective and the individual level of autonomy as socio-economic and technical autonomy. Similar distinctions of autonomy can be seen in Derber's (1982, p. 169) discussion of technical proletarianisation (loss of control over the process of work) and ideological proletarianisation (loss of control over the product) and Bailyn's (1985) division between the freedom to set goals and define problems and the freedom to decide how to pursue these goals and control the resources necessary to do so.
Another way of understanding professional autonomy in the literature is to tease out some separate elements of the concept. For instance, Horner (2000) describes professional autonomy as consisting of three aspects: (1) the ability to make and exercise decisions, (2) the ability to determine and judge the quality of professional work, and (3) the ability to value one's own work and determine an appropriate level of remuneration. A similar example can be found in Schulz and Harrison's (1986) division of the content, terms and conditions of work into clinical, economic and political autonomy. Clinical autonomy is defined as the ability to control clinical decisions related to treatment and judge the quality of medical care without being restricted by outside agencies such as the state, insurance companies or patients' rights. Economic autonomy refers to the ability to value one's own work and therefore control fees, volume and the mix of services provided. Finally, political autonomy involves the ability to make policy decisions and control the conditions of professional practice, such as location and hours. Although these three dimensions can be described as interrelated, variations can emerge over time and among professions in different countries in terms of the extent to which autonomy is experienced along these three dimensions and the extent to which they are affected by different reforms (Horner, 2000;Randall and Williams, 2009;Schulz and Harrison, 1986).
Although Freidson (1988) distinguishes between socio-economic (collective) and technical (individual) autonomy, he argues that «autonomy of technique is at the core of what is unique about the profession, and that, in fact, when this core autonomy is gained, at least segments of autonomy in the other zones follow after» (p. 45). Hence, a professional does not need to be an entrepreneur in a free market to be autonomous. Instead, Freidson (1988) claims that as long as a profession is free to decide on the technical content of work, its lack over the socio-economic terms of work will not significantly change its status and essential characteristics as a profession. Evetts (2002) seems to go even further when it comes to stressing the importance of the individual level of professional autonomy as she claims that professional autonomy is relevant only to the professional's decisions with regard to the individual client, and not the professional's work situation more generally. However, viewing professional autonomy only in the clinical or technical sense would be to take a narrow view of the concept as different elements of the concept tend to overlap and influence each other (Farough, 1996). Thus, this study adapts Schulz and Harrison's typology as the framework for the analysis and discusses professional autonomy from the aspects of clinical, economic and political autonomy (cf. Randall and Williams, 2009).

Professional autonomy as a social contract
When discussing professional autonomy, it is important to remember that the ability to control the content, terms and conditions of one's work is an outcome of a trust relationship established between a profession and society (Freidson, 1986;Larson, 1977). Autonomy is thus simultaneously a privilege and a responsibility to act in the best interests of the society (Sandstrom, 2007). As advocates for patients or clients, professional autonomy influences the broader society. Hence, when a physician makes decisions in terms of what illness a specific patient has, to what service this patient is entitled, and how many days in a hospital this person should receive, the decisions will have consequences not only for the patient's health, but also on the responsibilities of the community and the distribution of economic resources. For these reasons, professional autonomy extends well beyond the technical content of the work and the relationship between the patient and professional.
Although the literature suggests that the attainment of increased professional autonomy is of importance, Freidson (1986Freidson ( , 1988 postulates that complete professional autonomy is unrealistic. Given that a professional is dependent upon the wishes and demands of society, professional judgement must be limited by the public interest. Hence, in this study, advantages and disadvantages of restricted professional autonomy will be analysed from the perspective of what would be considered best for society.

Research design
In order to provide a discussion of the advantages and disadvantages of restricted professional autonomy, an approach focusing on describing and comparing the historical development between 1950 and 2000 of one of the traditional professions namely, the medical professionin Sweden and Canada is advocated. The reason for focusing on this era is that the medical profession has undergone several profound changes during this period, which earlier studies suggested have influenced the professional autonomy (Evetts, 2003;Nordgren, 2000;Parkin, 1995;Scott et al., 2000).
The reason for comparing Sweden and Canada is that the healthcare systems in the two countries have similarities but also differences. Healthcare in Sweden and Canada is both publicly funded and managed. Nevertheless, the situation of medical professionals in the two countries differs. Using Collins' (1990) distinction between Anglo-American and European contexts for professions, Swedish physicians can be categorised as operating within the state-influenced Continental model of professions. In Sweden, physicians are salaried employees directly employed by the county councils, the main providers of healthcare. In contrast, Canadian physicians fall within the Anglo-Saxon model of professions, which emphasises the freedom of the self-employed practitioner. In Canada, physicians are affiliated with but generally not employed by the hospitals. Their financial compensation is instead linked to their performance and paid by the federal state. The question is whether these different ways of organising healthcare also imply differences in the medical professionals' autonomy. In addition, if there is a difference in the degree of autonomy between the medical professionals in the two countries, how has it affected the way in which physicians act and provide services? The material used in this study consists primarily of historical data from earlier studies and public prints.

The development of the Swedish medical profession
At the beginning of the twentieth century, Sweden was home to approximately 500 district medical officers and municipal physicians, most of whom were general practitioners in private practices (Swartling, 2006). During the first part of the century, the number of physicians increased faster than the number of salaried appointments at hospitals, which resulted in a gradual decrease in the number of full-time physicians at the hospitals. However, at mid-century, the number of physicians affiliated with hospitals started to increase. By 1954, approximately 40 per cent of all physicians worked in hospitals; by 1972 the number was approaching 50 per cent (Nordgren, 2000;Swartling, 2006). Nordgren (2000) explains that more physicians applied to public healthcare in hospitals because of the lack of customer potential in the private market. With Sweden's small population density, low urbanism, initially slow economic development, and late industrialisation, the economic foundation for the expansion of physicians' practice was weak. However, in contrast to the medical profession in many other countries, Sweden's healthcare system wasand still isalmost entirely public in all main features (Bergman, 1998;Harrison and Calltorp, 2000). In 2000, 92 per cent of Sweden's physicians were publicly employed (Nordgren, 2000).
The idea that physicians should be public servants can be traced to the Höjerreport of 1948. The report suggested that physicians should be employed and salaried by the public sector rather than by the prevailing fee-for-service system. Physicians' income in the fee-for-service system was divided in two parts: a basic salary from the principal organisation and fees from the patients. The report proposed that physicians should be employed as public servants, a suggestion that eliminated each physician's right to run his or her own practice, have his or her own patients, and charge patients his or her own fees. Although unpopular at first, the ideas of the Höjer-report were eventually adopted (Nordgren, 2000). Private hospital beds and private fees for hospital patients at public hospitals were abolished by the Medical Care Act of 1959.
In 1963 the Swedish government decided that the county councils should assume the responsibility of district medical officers. This decision aroused protests and resistance; many previously state-employed physicians feared that the change would worsen their working conditions (Ministry of Health and Social Affairs, 1978). However, seven years later, the discretionary space for physicians became even more restricted as private fees at public hospitals were abolished and a standard rate was introduced in all public non-institutional care through the sjukronorsreformen. Before this reform, physicians had earned a basic salary for being available and for treating in-patients. Physicians also had the possibility to run their own practice at their workplace and compensate the principle for the costs of the room area and the use of staff. The new reform put an end to this practice as physicians were required to become salaried employees at public hospitals with no right to run private practices (Brage Nordlander, 2006).
The criticism against the reform was harsh. Opponents predicted that the reform would result in the socialisation and totalisation of the medical profession (Einarsdottír, 1997). Sweden's healthcare system became a public question, and physicians were transformed from autonomous workers to salaried employees at the mercy of productivity demands. Still, the rapid expansion of the healthcare sector during the 1970s and 1980s implied a great freedom of actions as hospitals were paid mainly in accordance with their costs. Hence, physiciansalthough salaried employeescould focus sincerely on how to distribute the yearly financial increase between departments (Anell, 1994).
However, the general weakness of the economy in the 1980s put an end to this as governments imposed constraints on healthcare budgets through ceilings on expenditures and restrictions on taxation (Harrison and Calltorp, 2000). Defects in the traditional approach to distributing resources became evident at the same time as issues around how to finance, organise, and control the healthcare organisation were lifted. This became the starting point for an extensive reform work whereby purchaser-provider models, performance-based reimbursement (DRG system) and waiting-time guarantees among others were introduced (Anell, 1994;Hanning, 1996;Forsberg, Axelsson and Arnetz, 2001).
Although the Swedish medical profession has long been strongly tied to the public sector, studies have indicated that the professional autonomy has been reduced over the years as a result of the increased reform work (Forsberg et al., 2001;Nordgren, 2000). Historically, the proportion of physicians in administrative positions at hospitals has beenand still issmall. According to Riska (1989), this indicates that physicians have not gained control over managerial positions at hospitals and in public administration. In addition, they have not been able to prevent hospital administrators, nurses, and business controllers from taking control of management positions. Such statements can of course be questioned as studies have shown that there are cases when professionals and organisations have mutual interests (Bourgeault et al., 2011) and that medical professionals can maintain a high level of autonomy even though the number of management positions decreases (Spehar and Kjekshus, 2012). Even so, Nordgren (2000) concludes that the medical profession has lost its professional authority, dominance and control over time: The classical medical profession and its position to be one's own manager and possess a great clinical freedom is about to fade away at the same time as the claims from the environment get tougher. The influence over one's own work situation has been reduced through administrative control and financial contractions and the medical practices as medical research becomes more and more industrialised (Nordgren, 2000, p. 83)

The development of the Canadian medical profession
The first entirely publicly funded healthcare insurance programme in North America was in Saskatchewan in 1947. Saskatchewan's provincial Prime Minister, Tommy Douglas, who had survived a serious illness in childhood, was committed to the creation of a publicly financed healthcare programme. The new insurance programme was built upon the principles that everyone -regardless of their ability to paydeserved access to hospital care (Rachlis and Kushner, 1994). According to Rachlis (2004), the new healthcare insurance programme was a great success. People who had not been able to afford healthcare were suddenly able to obtain the care that they needed. At the same time, physicians, who were accustomed to at least ten per cent of their invoices never being paid, were guaranteed payment.
Other provinces quickly followed Saskatchewan's lead. By 1955, five more provinces had introduced publicly funded hospital care insurance programmes. However, since it was harder for provinces to finance these healthcare programmes entirely on their own, the federal governmentafter pressure from several directionsenacted The Hospital Insurance and Diagnostic Service Act in 1957. The legislation became the foundation of Canada's present public healthcare programme. Yet the federal government's aim was not to take jurisdiction over healthcare from the provinces. Instead, it promised to compensate the provinces for half of their healthcare costs if the provinces agreed to preserve the principles of universality, accessibility, portability, comprehensiveness and public administration (Rachlis and Kushner, 1994).
This legislation gave the provinces the financial space necessary for developing and expanding hospital care; between 1961 and 1971, the number of hospital beds grew twice as fast as the population. In Saskatchewan, Tommy Douglas proposed expanding the provincial healthcare programme to cover medical treatment outside of hospitals. However, this proposal met hard resistance from medical professionals. In July 1962, more than 90 per cent of the physicians in the province went on strike and even refused to treat emergency patients. After negotiations, the parties agreed that medical treatment should be covered by the public healthcare programme as long as physicians were allowed to impose a surcharge above the negotiated provincial rate (Rachlis and Kushner, 1994;Rachlis, 2004). Nevertheless, the medical dominance was severely shaken as the national health insurance and the form of state invention accompanying it were gradually established (Coburn et al., 1983).
Two years later, apparently reflecting fears of inflation and a rising deficit, Canada approved The Medical Care Act (Medicare) in 1968. This new legislation covered both hospital care and physician medical treatment. Once again, the federal government would compensate the provinces for fifty per cent of their costs in exchange for the preservation of the five basic principles (Rachlis and Kushner, 1994;Rachlis, 2004).
When the boom of the 1960s was replaced by stagflation in the 1970s, the problem of fifty-fifty funding became obvious. Until then, each province had reported at the end of each year how much it had spent on hospital and medical care; the federal government was expected to write a cheque for half of the amount. This procedure made it difficult for the federal government to anticipate its expenses and prepare its budget. The federal government demanded a reliable system that made it easier to calculate expected healthcare expenses. The fifty-fifty funding was changed to budget funding with the effect that physicians' incomes began a gradual decline as physicians became subject to wage and price controls (Coburn, 1988;Rachlis, 2004).
The introduction of Medicare was financially beneficial for physicians, who were guaranteed payment. However, the law that made it possible for physicians to demand a smaller patient fee above the negotiated rate was only rarely used. In 1978, only ten per cent of the physicians in the Ontario province charged their patients the extra fees. Nevertheless, after the wage and price controls were lifted, the incentives for charging these extra fees increased. The Ontario Medical Association raised its charges by thirty per cent above the negotiated federation rate (Globerman, 1990) and, in the ensuing years, the share of physicians who used this possibility increased drastically. Similar situations arose in other provinces, creating a fear that Medicare and its principle of universal access to medical care were being threatened.
A political debate started about the existence of the patient fees and their consistency with the principles of Medicare. The debate produced The Canada Health Act, enacted in 1984, which made the five principles more stringent and prohibited patient fees above the negotiated rate. Provinces whose hospitals or physicians charged their patients would be punished with a reduction of their federal financial compensation (Rachlis and Kushner, 1994;Rachlis, 2004). Moreover, physicians who opposed the new regulations by charging their patients would be fined $10 000 (Coburn et al., 1987;Globerman, 1990). These reforms evoked strong protests from the medical professional associations and a physicians' strike. This time, the reform remained unchanged: Medicare in its present form was born.
So what is the situation for Canadian physicians today? One of the greatest differences between Swedish and Canadian physicians is that Canadian physicians are generally not employed by hospitals in the sense that the hospitals pay their salary. Instead, each physician is considered to be self-employed and receives repayment through a system called fee-for-service. The physicians' wages and fees for medical services are negotiated between the medical professional associations and the provincial government, with the province paying the physicians' wages (Westander, 2003). According to Farough (1996) the system of payment is crucial to the physicians' vision of themselves as freestanding professionals, rather than government employees.
As physicians are paid for the services they perform, the medical profession has a very powerful position in the Canadian healthcare system and is relatively independent. Rachlis (2004, p. 202) cites Dr. Gabor Maté, a family physician and columnist: «Our fee-for-service medical system actively punishes doctors who spend time with patients and financially rewards those who practise superficial medicine in profit-motive walk-in clinics, who see minor problems and refer more challenging cases to the emergency ward.» Consequently, in the Canadian healthcare system, the more patients a physician can meet and treat, the more he or she earns. As each physician is guaranteed payment by the provincial government, physicians can recommend more and longer visits. This implies that hospitals might be charged with costs that they cannot control.

Restricted professional autonomy
From these historical descriptions, we can establish that several changes have restricted the professional autonomy over the years. However, when analysing the situation from the three dimensions of Schulz and Harrison (1986), differences appear between Canada and Sweden.

Dimension 1: Economic autonomy
The first dimension, economic autonomy, is the degree to which a profession has control over remuneration (Schulz and Harrison, 1986). Although reforms in the two countries have resulted in the abolition of private hospital beds and fees over negotiated rates, this change can also be described as a restriction of the medical profession's possibility to establish and control compensation and fees. In both countries, interference in issues regarding remuneration from the government has increased as healthcare costs in the two countries have risen. However, this development has not always come at the expense of physicians' economic autonomy. Instead, the first publicly funded healthcare insurance programmes in Canada can be described as a government initiative that improved economic stability and increased economic autonomy for physicians. In contrast, the reform initiatives in the 1970s and 1980s put an end to this.
However, the degree of economic autonomy has been even lower among Swedish physicians due to the different healthcare systems. Operating in a typical Continental model of professions (Collins, 1990), Swedish physicians are more closely tied to the government by their position as salaried employees. Their financial compensation is thus a responsibility of managers in the organisation, and the only way physicians can influence their compensation is by working additional overtime hours. At the same time, physicians, in their role as employees, are held responsible for the financial outcome of the hospital.
In contrast, Canadian physicians are linked tobut not tied tothe government through their fee-for-service system. Payment is based on the number of client visits made rather than an hourly wage; thus, the possibility of influencing remuneration is higher among Canadian physicians than among Swedish ones. The financial compensation system based on fee-for-service rather than a basic salary means that Canadian physicians are not fully under the regulations and supervision of the organisation. Instead, as freestanding professionals they can refer complicated cases to the hospitals and recommend expensive treatments without being held responsible for the extra costs incurred for the hospital. Hence, we can conclude that, although medical professionals in both countries have experienced a decline in economic autonomy, the degree differs as Canada and Sweden belong to different models of professions.

Dimension 2: Political autonomy
Political autonomy can be defined in terms of the ability to make policy decisions and control the broader context in which professional work takes place. In other words, political autonomy applies to the ability to decide when, were, and under what circumstances services are provided to clients (Schulz and Harrison, 1986). When analysing the degree of political autonomy for the medical profession in the two countries, once again the influence of the organisation and its ability to restrict the political autonomy of the professionals becomes obvious. As salaried employees, Swedish physicians' ability to influence and decide on schedule of work, work conditions and location is limited by the rules and regulations of the organisation and its managers. The development of Swedish physicians from autonomous workers to salaried employees thus indicates that the degree to which the profession has control over the conditions of the professional practice has decreased. Instead, this has become an issue for hospital managers. Meanwhile, the situation for physicians in Canada is closer to the situation of independent practitioners; thus, the influence of the organisation on political autonomy is less appreciable and the ability to set working hours and location is greater.
However, political autonomy can also be discussed from the perspective of the relationship between the medical professional associations and the state. From this aspect, the Swedish and Canadian situations can be described differently. In a study of the historical development of the British Columbian Medical Association, Farough (1996) concludes that the government intervention has become more frequent and invasive over time while the boundaries between what is portrayed as a technical or a political matter have constantly changed. In other words, Farough continues, the medical association in Canada has been the least successful in protecting the political autonomy. Hence, we can conclude that, for medical professionals considering themselves to be freestanding professionals rather than government employees, the intervention of politicians has been crucial for their political autonomy whereas the relationship between the profession and the state in Sweden can be described as pretty much consistent over time.

Dimension 3: Clinical autonomy
The third dimension of professional autonomy is the degree to which a professional controls the content of his or her work. Such clinical autonomy is associated with the ability to make clinical decisions about treatment and assess patients' needs (Schulz and Harrison, 1986).
The ability of medical professionals in Sweden to make treatment decisions can once again be linked to their situation as salaried employees. At the same time, the situation indicates how a weakness in the political and economic autonomy has ramifications for clinical autonomy. For example, the public funding constraints during the 1980s and the 1990s and the budgetary control of county councils limited the medical profession's influence in terms of the kind of technical equipment to be brought in and the allocation of resources. In Canada, patients follow the physician rather than the organisation, meaning that hospitals are the only place where physicians can use the technical equipment that they need. This implies that the clinical autonomy for the medical profession in Canada has become less restricted by the fee-for-service system as physicians are not subordinated to the management of the hospital to the extent that Swedish physicians are.
The situation of the Canadian physicians can be illustrated with the following example. During the autumn of 2007, a former member of the board of directors of the Guelph General Hospital in Ontario was interviewed, during which the member told a story about a board discussion as to whether or not the hospital should tie a certain physician to it. It was generally known that this specific physician recommended that most of his patients undergo a certain examination. If the board tied the physician to the hospital, its laboratory costs would certainly increase. After a long discussion and numerous calculations, the board decided to tie the physician to the hospital, but offer him an outside office. Examinations could be conducted at either Guelph General Hospital or a private clinic on the other side of Guelph. Renting an office for the physician in a building closer to the private clinic, rather than offering one of the vacant offices at the hospital, would be less expensive for the hospital as patients would go to the private clinic rather than to the hospital for the examination. As a result, the hospital's laboratory costs would not increase.
The example indicates that the fee-for-service system in which physicians are tied to but not employed by hospitals means that the single physician and the clinical autonomy are only marginally affected by decisions made by the hospital and its managers. This sometimes forces hospitals into prolonged negotiations with physicians when changes must be made. Rachlis and Kushner (1994) argue that physicians in Canada have virtually complete clinical autonomy, and Farough (1996) concludes that of all the aspects of professional autonomy for medical professionals in Canada have been most successful in controlling clinical autonomy.
Although the organisation has affected the clinical autonomy of the medical profession to different degrees, the increased administrative control derived from the New Public Management debate (Hood, 1995) has probably had the greatest influence on its clinical autonomy. With tighter management control, focus on performance, and a growing number of documents in the form of manuals, quality assurance methods and standardisation of the process of treatment, the work of medical professionals has suddenly become controllable by people outside the profession (cf. Agevall and Jonnergård, 2007). Through the development of registers and reports of medical care in both Sweden and Canada, each hospital's performance is being compared to that of other hospitals. Even if different control documents exist in both contexts, we can claim that the medical professional's right to judge the quality of care has become even more restricted in Canada than in Sweden as the monitoring of medical treatment at the hospitals and indirectly by the physicians is more noticeable (cf. Randall and Williams, 2009). The question here is whether medicine will continue to have the dominant voice in determining and analysing these reports or if medical treatment is an area where we can expect changing boundaries between what is portrayed as a technical or as political or economic matter.

The social contract
Based on the discussion thus far, we can conclude that changes during the last decades have restricted the professional autonomy of the medical profession in Sweden and Canada. However, when discussing professional autonomy, it is important to remember that autonomy is an outcome of a trusting relationship between a profession and society (Freidson, 1986;Larson, 1977). In order to examine the effects of these restrictions, we have to analyse advantages and disadvantages from the perspective of what would be considered the best for society and how this is associated with the professional archetype. Three predominant changes found in the empirical material that have influenced the socio-economic and technical aspects of professional autonomy will be examined.

Government intervention and democratic ideals
When looking at the historical development of the medical profession, it is striking that in both countries the government's intervention in healthcare has increased over the past decades. This intervention has limited professionals' ability to make autonomous decisions on both a collective and an individual level. On the collective level, the economic incentivesin the form of the possibility to control fees and possess private beds at the hospitalshave been abolished. On the individual level, the new control instrument derived from the idea of New Public Management means that hospitals' performance is now a question for politicians and the public. Although the intervention from the government has restricted the so highly valued professional autonomy, it has become legitimised in society. One reason for this legitimisation is that the reform initiatives in both countries have been built on democratic principles. The main argument for transforming healthcare into publicly financed programmes has always been that everyoneregardless of their ability to payshould have access to hospital care. For this reason, charging patients extra fees will never be compatible with the democratic rights of free access to healthcare. The same can be said about the increased performance control over the last decades. The democratic ideal and its principle of openness for guaranteeing fairness, good governance, and transparency presume that hospitals' performance will be open to public scrutiny.
However, even if professional autonomy has become restricted, it is hard to state that democratic ideals clash with the professional archetype. Charging patients extra fees is associated with selfishness rather than altruism and commitment to the client. Hence, a first conclusion is that restricting professional autonomy in the economic dimension might not always be negative. We might even claim that government intervention and its emphasis on democratic ideals have strengthened the professional archetype and its principle to act in the best interests of society (cf. Freidson, 2001).

The organisation as the arena
The second thing that strikes one is the important role the organisation (the hospital) has come to play as the arena for the medical professionals and their work. The influence of the organisation over the medical profession is stronger in Sweden than in Canada due to their different healthcare systems, but in both countries the organisation influences the professional autonomy.
A consequence of the transformation of Swedish physicians into salaried employees is that the profession has been forced to bow to rules and regulations of the bureaucratic organisations. The fact that there are few professionals in management positions further indicates the limits of professionals' ability to influence organisational decisions, although studies have demonstrated that a loss in dominance does not automatically result in a loss of influence (Spehar and Kjekshus, 2012). Hence, operating within the boundaries of the organisation has resulted in restrictions when it comes to both economic and the political autonomy restrictions that might in fact influence the clinical autonomy.
Although the historical development implies that the medical profession has lost influence over socio-economic decisions to the administrative staff, its professional identity and characteristics have not been threatened. As long as decisions regarding the technical content of the work belong to the profession, without too many restraints of management decisions, professionalism as an occupational value can be maintained (cf. Freidson 1988). A similar line of argument is evident in Dohler's (1989, p. 196) comparative study as she concludes that clinical autonomy «is preserved best when doctor's financial freedom is restricted most» because state attempts to regulate the income of physicians seem to protect clinical autonomy even if regulations have only a marginal impact on cost containment.
However, as the Canadian example demonstrates, excessive socio-economic autonomy without any responsibility for the economic consequences might exaggerate practitioners' own occupational self-interests in terms of their salary and 'craft of work' in front of what would be best for society. Such professional acting suggests that healthcare spending will increase and governmental resources will be wasted. Moreover, over-treatment demands more physician hours, which might result in longer waiting lists for hospital care. Overtreatment can also cause patients to become unnecessarily worried about their health. Hence, a second conclusion is that too much socio-economic autonomy may not always be desirable. Instead, a reasonable restriction of the professional autonomy by the organisation and its decision makers can prevent physicians from over-treating patients and wasting governmental resources.

Documents for control
The historical development of the medical profession also shows a pro-liferation of management control instruments in the form of manuals, standards and reports that have invaded the public sector in general and healthcare organisations in particular. The new management control instruments have contributed to making the outputs and results in healthcare more transparent, which implies that the professional work has become controllable by people outside the profession. Does this means that pro-fessional autonomy has been reduced?
On a collective level, we can claim that insights from outsiders have increased. Healthcare performance is being inspected, monitored and compared, and politicians, the media, and the public can and do question healthcare decisions and the quality of care. However, the insights and comparisons occur at an aggregated hospital level. Hence, it is the performance of the hospitalnot that of the physicianthat is being monitored.
The new management documents and manuals have affected the technical autonomy of the medical profession regarding the ability to judge the quality of care, but they have had less of an impact on daily treatment decisions. The impact on medical professionals at the individual level may instead be described as professionals now having to defend and justify their actions in ways that they did not have to in the past. Given that the legitimacy of a profession depends on the client's recognition of the professional's knowledge and expertise and that the new insights enable more people to discuss and question the quality of care, we can claim that the professional autonomy and the privileges of freedom have changed the way in which today's medical professionals have to communicate what they intend to do and why more clearly than before.

Conclusion
More research on changes and continuities, challenges and opportunities for professionalism as an archetype has been requested (Evetts, 2011). This article contributes to this call for research by presenting a historical description of the development of the Swedish and Canadian medical professions and discussing the effects of restricted professional autonomy. The analysis suggests some key conclusions.
First, the findings indicate that medical professionals in both Sweden and Canada have been subjected to constraints which have reduced their professional autonomy in all three dimensions. However, whereas medical professionals in Sweden have been more likely to experience restrictions of their professional autonomy from the organisation, Canadian physicians have primarily been subjected to increased government intervention.
Second, although professional autonomy has been restricted by government intervention, the organisation and new document control, the technical content of treatment still appears to be prescribed the professionals. Hence, medical professionals still have control over the core autonomy of the profession (Freidson, 1988). At the same time, medical treatment can be described as an area where the boundaries among what is portrayed as a technical, a political or an economical matter currently seem to be changing the most. With greater focus on performance and a growing number of manuals and standards, people outside the profession are able to judge and comment on the quality of care, which has resulted in today's physicians needing to communicate with politicians and patientstwo interest groups with increasing influenceto a greater extent in order to defend and legitimise their actions in society.
Finally, the study shows that restricting professional autonomy is not merely negative for the preservation of professionalism as an archetype. Instead, a 'reasonable' limitation of the socio-economic freedom may in fact preserve altruistic behaviour, prevent overtreatment and overexploitation of government resources and guard professional interests at the same time. What emerges from the study is that we ought to reconsider the importance of professional autonomy. Instead of focusing on the degree of autonomy, it might be better to focus on how the protection and restriction of the space in fact change, preserve, challenge and strengthen professionalism as an archetype.