Positioning the ageing subject: articulations of choice in Swedish and UK health and social care

ABSTRACT What happens when similar measures are being introduced in different national contexts? This article studies the ways in which patient choice has been articulated in public and official reports on health care in the two contexts of Sweden and the UK, whose welfare systems are typically comprehended as different. Specific interest is directed towards the construction of patient positions, and policy documents are analyzed using discourse theory. The results show many similarities between the national contexts; choice is primarily articulated with individuality, autonomy, consumption, and responsibility, as well as with support from state agencies, and patient choice is relentlessly normalized as the way forward. But there are also important differences that reveal that the presuppositions differ, for example, when pinpointing the stakeholders of patient choice reforms and how the different policies work to take the well-known edges off of patient choice ideology.


Introduction
Swedish and British welfare systems have, together with representatives of so-called conservative systems, been used as symbols of opposing ways of organizing welfare services (cf. Esping Andersen 1990;Breen 2010). Whereas the Swedish welfare state has been strongly associated with the notion of the Scandinavian or Nordic model (Christiansen and Petersen 2001;Christiansen 2006) and described as the "archetypal 'social democratic welfare state'" (Ruhs 2010, 269; see also Breen 2010), the UK has represented a somewhat different welfare system. England was the first country in Europe to marketise social care (Brennan et al. 2012), and the UK has a political agenda clearly driven by neoliberal ideology (Ferragina and Arrigoni 2017;Baldwin 2000) and market orientation (Scott-Samuel et al. 2014). What do such differences mean for how policy changes are being introduced and "sold" to a general public?
In recent decades there has been a general tendency towards choice policies and privatization measures in Western health and social care (cf. Self 1990). Both in Sweden and in the UK, choice practices have become intertwined, warranted, and applied in combination with other modern-day care and management ideals (Bevan, Helderman, and Wilsford 2010), and they have been implemented as part of marketization, personalization, and management ideologies (Ferguson 2007;Hasselbladh, Bejerot, and Gustafsson 2008). 1 The aim of this article is to study the ways in which "choice" in health and social care is articulated in public and official reports in the two contexts of Sweden and the UK, whose welfare systems are typically comprehended as different, and who have undergone a "remarkable shift from planning to market mechanisms" (Fotaki and Boyd 2005). We argue that although there are similarities, there are also differences in how choice is comprehended and "sold", as it were, and what it is thought to deliver. Such differences are telling of the prevailing ideological fantasies about what it means to be/come a patient or care user that are deployed by Swedish and UK policy makers. They also help constitute the conditions for taking different positions as a patient or care user. Our focus is not on describing the organizational changes as such, but on studying how change towards choice policies is made culturally comprehensible on a policy level and what arguments are used when describing and legitimizing this move. This article thus adds to the literature on policies for choice in health and social care and on the roles that ideological fantasies play in such policies, including their effects on the possible positions that can be taken by patients and care users.
Because systems revolving around choice constitute ongoing processes rather than fully accomplished conditions, policy reports make an apt entrance for the study of the discursive struggles that shape and re-shape their meanings. Discourses never simply mirror social experience or represent some objective will emanating from a position of power. Rather, they take part in the constitution of experiences as well as of political agendas, and, in the words of Irvine (2002, 31), play a part in "constituting social subjects, and the relation between subjects" (see also Mol 1999). In the sections that follow we will develop our theoretical points of departure and connect them to our empirical material: Swedish and UK policy texts on the introduction of choice in health and social care.
Patient choice: discourse, articulation, and (old age) patient positions Much research has touched upon the complexities of choice discourses, theoretically inspired by Foucault (1982) and Rose (1999) who have pointed out that "choice" often builds on an idealized notion of a rational subject. Such critical accounts problematize prevailing liberal understandings of choice in health and social care and the accompanying notions of an autonomous person who makes unrestricted and well-informed choices (e.g. Mol 2008). They also shed light on how different subject positions enable and prioritize certain patient or care user practices while impeding others (Moser 2005;Pols 2005Pols , 2011. In this sense, although having emerged in the 1960s and 1970s as a resistance against public care institutions (Needham 2011), the notion of choice seems to have become an apt tool in the reorganization of the health care system along neoliberal lines (Irvine 2002).
We follow this line of argument, and we do so specifically in relation to the production of elderly subjects in healthcare policies. In our analysis, we use political discourse theory (Laclau and Mouffe 1985), focusing on how choice is ascribed meaning in and through health care policies and how that in turn contributes to the construction of different positions for patients and care users. A focal point of the analysis is the practice of articulation in policy textsthe construction of nodes in which meaning is partially fixated. The concept of articulation describes the construction of a totality by the linking together of a variety of signs, practices, or identities (Laclau and Mouffe 1985). For example, within some studied policy texts "choice" might be articulated with "autonomy", "freedom", and "minimized bureaucracy", while in others with expressions like "insecurity", "cutbacks", and "economic austerity". The concept of articulation points at the core of meaning-making, and we sometimes refer to this articulatory practice in terms of "discursive strategies" or "discursive struggles", whereby different ways to create meaning around patient choice struggle for the right to define the phenomenon. When a specific type of articulation becomes established, the discourse about that phenomenon is being stabilized which means that it will also start to become materialized into policies and everyday routines. Concurrently, the concept of articulation also constitutes the subject positionsor patient positions as we will call them from here on afterthat are made available for patients within the realm of certain discourses (Mol 1999;Bacchi 2016). 2 The studied policy documents often proclaim the failure of the social, primarily the organization of health and social care, and announce the potentials of choice reforms. A central question regards the hopes and pitfalls ascribed to choice policies, as well as the ideological fantasiesor fantasmatic policy narratives (Howarth, Glynos, and Griggs 2016) evoked to legitimise and support the move towards such policies. In accordance with Žižek (1998), we use the concept of ideological fantasy to point out the way the "order of things" is described and made comprehensible (see also Glynos 2001;Glynos and Howarth 2007;Stavrakakis 1999), thus offering explanations for things that happen by contextualizing them in specific ways. These fantasies thereby work to conceal the radical contingency of the social, but they also direct the attention in certain directions while precluding others (Glynos 2011). An example of how ideological fantasy works is provided by Gunder (2016, 34) who describes how notions of the effectiveness of neoliberal governance, the market and unfettered growth tend to be portrayed as what will "ensure policies for market-led success", e.g. in policies for affordable housing for all. The fantasy suggests that if only the market is allowed to work unconstrained, affordable housing will be produced. To explain why such policy-promises are often not fulfilled despite their strong support, Gunder points at how phenomena such as planning and bureaucracy are included into the fantasy and placed in the role of scapegoats for policy failure (see also Fotaki 2007Fotaki , 2010a. In this way, policies and their ideological underpinnings can remain despite repeated failures to deliver their promises (Glynos 2008).
We argue that fantasies about patient choice "drove" the policy arguments and made them come forth as reasonable (Glynos 2001). It is important to note that rather than being understood as "false consciousness", we understand ideological fantasy as the necessary fabric of meaning that urges people to act in certain ways and that make these ways appear natural and/or necessary.

Methods and material
In accordance with the above, ideological fantasies not only exist as immaterial (spoken or thought) discourses but are continuously stabilized into actions and materialities through practices of articulation. We view the studied policy documents as constituting such materialized and temporarily stabilized discourses; they constitute one of many sites where the struggle over meaning ascribed to welfare, patienthood, and ageing takes place. Hence, the reports are inherently a part of what they hope to govern, and thus never intervene "from the outside" (Shove, Pantzar, and Watson 2012, 145). Building on this notion of written policy as being involved in the always ongoing constitution of the social, being itself a cultural product (Shore and Wright 1997;Bacchi 2009), the selected reports are seen as political actors that define, organize, and transform care practices such as patient choice (Timmermans and Berg 2010).
The studied material consists of policy reports on the development of patient choice in Swedish and UK health and social care that were written between 2005 and 2015, a particularly expansive period for patient choice both in Sweden and the UK. The chosen corpus includes a selection of official state reports, white/green papers, and other authoritative texts (Howarth 2000) used to provide information to the public and/or for consultation in policy and law making. What unites the selected reports is that they are possible to view as "authoritative claims" (Hajer 2009) in the sense that they constitute authoritative efforts to communicate specificbut potentially conflictingcomprehensions of patient choice. Therefore, they all take part in the political struggle over the meanings of patient choice. Because changes in health and social care systems evolve over time, it is important that the motives behind their introduction be closely scrutinized (Fredriksson, Blomqvist, and Winblad 2012). We argue that policy documents constitute an important arena for articulations that are ascribed decisive significance, not least because they have the capability to reach a large audience, whether in their printed or digitalized forms or via references in other texts.
The selection criteria were that the reports 1) are written as reports to do with choice in health care, 2) are investigative and explanatory in character, 3) are directed at a public with a general form of address, and 4) are referred to by other texts, and thus are recognized within the studied discourse of choice. The UK policy texts comprise green and white papers and consultation papers published by the Department of Health (DH) and one green paper published by ResPublica, an independent think tank. The Swedish policy texts comprise three Swedish government official reports (SOU), one report from the Swedish Association of Local Authorities and Regions (Statens kommuner och landsting, SKL) 3 , and one report from Nutek. 4 The reason for the broad scope, including official state reports as well reports published by stakeholder organizations, is that we wanted to study the on-going construction of a common understanding of what patient choice entails. We argue that such understandings are formed not by the instances of greatest power but as an effect of discursive struggles.

Analytical strategies
Because our central proposition is that "governing takes place through the formation of 'problems'" (Bacchi 2016, 12), the first step of the analysis was to identify how the policy-writers identified the "problem". What, to use Bacchi's (2009) words, is "the problem represented to be"? (see also Vallgårda 2015). Upon reading the reports, we specifically looked for how they conceptualized the reasons to go about change; what was described as problematic with the present state? We also asked what ideological fantasies are evoked to make these problematisations come across as reasonable, which in practice meant that we noted how the authors of the reports made efforts to create contexts that would legitimise the positions they took. What notions of, for example, "healthcare", "patient" or "today's society" were evoked in such contextualisations? The contextualisations were then studied for what constituted their respective nodal points, or the privileged signs that structured the other signs in the articulation (Laclau and Mouffe 1985, 112), and they were named accordingly.
In a second step, we systematically looked for articulations of patient choice in the selected documents. How is choice being intertwined with and given meaning through references to practices such as "less management" or adjectives such as "active"?
The third step of the analysis was to chisel out how patient positions are given form through the different articulations of patient choice. What patient positions become reachable through the policy reports?.
In what follows, we will first account for how the problem in health and social care is accounted for. We then move on to describe the central articulations of patient choice. In the discussions section we will deepen the analysis of the studied articulations in terms of their effects on patient positions.

Resultsselling choice in health and social care
Representing the problem: three ideological fantasies In both national contexts, the problem of the present and/or future health and social care is represented either in terms of quantity (the present system does not cover the needs) or quality (the present system does not suit future patients). We have chosen to account for these "ways of representing the problem" (Bacchi 2009) in terms of ideological fantasies (Žižek 1998;Glynos and Howarth 2007) that offer narratives of the situation that legitimise the formulation of the problem. Three different fantasies surface in the policy documents. These are either explicitly put forth as arguments or simply constitute taken for granted points of departure. They provide the reasons for why a change in the organization of health and social care is necessary, and they legitimate increased patient choice as the preferred way forward. Further, they repeatedly suggest that time is a factor in the process and declare a state of urgency that helps in packaging the turn towards patient choice as a pressing issue.
First, there is pervasive talk about present or expected economic deficits that have to be dealt with. The way economy and costs are referred to only makes sense from within what we have called a neoliberal fantasy of market economy that is strongly articulated with the possibility to cut costs (cf. Brennan et al. 2012;Payne 2000;Forssell 1999). But more than that, the fantasy of market economy offers understandings according to which state monopoly is questioned and in which competition is unreflectedly portrayed as a guarantor for quality, both in terms of efficiency and in terms of individual autonomy and equity (Blomberg 2008; Andersson and Kvist 2015). 5 Highlighting the possibilities to simultaneously cut costs and increase quality, the fantasy of market economy presents costs and a general lack of autonomy on the part of patients as the major problems of present health and social care. The focus on costs is obvious both in the Swedish and UK reports; economic conditions are persistently used as vantage points in the debate about care, whether this is explicitly stated or just implied. In Equity and excellence: Liberating the NHS (2010, 4) from the UK, it is affirmed that "(O)ur massive deficit and growing debt means there are some difficult decisions to make." It then continues to articulate care practices of "putting patients first" with the goal to "drive up standards, deliver better value for money and create a healthier nation." The competition between care providers is itself unproblematically assumed to drive up standards.
Second, and closely connected to the first, demographic changes are repeatedly said to put a strain on welfare systems. What could be labelled a fantasy of demography works to position demographic prognoses of increased longevity and demographic ageing as major threats to the welfare systems, thus compelling them to change. Such a fantasy has frequently been described in terms of "apocalyptic demography" (cf. Gee and Gutman 2000;Lundgren and Ljuslinder 2011) and has been shown to be closely connected to a sense of urgency (West 2013;West and Lundgren 2015). Almost all of the policy texts studied here include references to population ageing, and they unproblematically make this demographic process the reason why patient choice should be chosen as the necessary way forward. It is obvious that patient choice is presented as a way to avoid the horrific scenario of an implosion of current welfare systems on a national level as populations age, both in Sweden and the UK. In the Swedish Framtidens utmaning (SKL 2010), it is admittedly said that it is indeed a big success for the welfare state that people are living longer. However, the report then states that "there is no long-term relief to be found" (2010,27), suggesting that this "success" also has significant problems.
Third, there is a fantasy of changing cultural identities that highlights people's increased desire for control and self-management, people's increased purchasing power, and how people are becoming more and more prone to take responsibility for their own health, thus demanding the right to choose their health and social care. "People want to have more control over their own health, as well as their care", the UK authors of Our health, our care, our say (DH 2006, 13) stipulate, and one Swedish official report (SOU 2008:105, 193) states: "There is a will among older people to take greater responsibility." Perhaps because the reports all talk about changes in the making or changes to come, they often argue that people will be even more prone to self-control in the future. In the pursuit to account for these changes, the fantasy of changing cultural identities is often exemplified with the figure of the ageing baby boomer, who is described as having specifically high demands on choice and service and who is also healthy, active, and able to gain an overview of the variety of services from which to choose (cf. Healy 2004;Freedman 1999;Sperazza and Banerjee 2010).
In most cases, the fantasies are intertwined, often occurring in the same sentences. In its very first paragraph, the green paper Independence, Well-being and Choice. Our Vision for the Future of Adult Social Care in England (DH 2005) states that it "addresses the challenges for social care of a changing and ageing population, higher expectations, and our desire to retain control over our own lives for as long as possible and over as much as possible." Swedish policy texts perform similar articulations when arguing that demographic ageing "will lead to an ever smaller proportion of the population working and paying taxes to finance services for an ever growing and more demanding group of elderly" (SOU 2008:15, 252). In both quotes, a fantasy of demography is articulated with a fantasy of changing cultural identities, and this makes clear that demographic ageing and "a desire for" autonomy and control is what urges patient choice reforms.
These fantasies work to point out the reasons why change is needed. Simultaneously, they incorporate the policy suggestions into taken for granted narratives of the world and come to constitute the naturalized context in which the articulations of choice are made.

Articulating choice
Choice, individuality and autonomy At first glance, there is a lot that unites the Swedish and UK policy texts. They all primarily articulate "choice" with an increased possibility to customize one's own care. In this sense, "choice" is tightly connected to notions about patients' greater knowledge of their own needs and preferences as well as to their abilities to study and choose from different care alternatives. Slight differences can be detected where the Swedish reports articulate choice with words like "opportunity" and "possibility", while the UK reports tend to use words and expressions like "control", "autonomy", and "less administration" more often. In this sense, the UK policy texts more clearly articulate choice as the medium for patients to protect themselves from state interventions and bureaucratic inconveniences; choice seems to become the tool that will set patients free from the shackles of the state, but also "free the frontline from bureaucratic constraints" (DH 2010c). The Swedish reports only seldom contain such insinuations; patient choice is not thought to set anyone free, but to increase certain values. However, in both national contexts, "choice" seems firmly connected to "freedom", whether it is freedom from something or freedom to do something.

Choice and consumption
Not surprisingly given the embeddedness within a fantasy of market economy, both national contexts articulate "choice" with a consumerist discourse where the possibilities to pick and choose from a greater variety are emphasized, and where a language of consumerist discourse is deployed.
In the UK policy texts, words like "purchase" and "buy" are frequent and are sometimes articulated with the patient position of the "consumer" in order to explicitly produce a sense of agency, which is thus connected to the practice of choice: People who are currently the passive recipients of services become consumers with the ability to shape and control the services they are willing to buy and thus shift the culture of care planning (DH 2005, 35).
Although more sparsely, the Swedish reports also sometimes use words like "customers" or "consumers" to describe patients and care users and "producers" to describe care providers (e.g. SOU 2008:15, 230;Blomqvist 2004;Szebehely 2000). The system of patient choice is sometimes referred to as a "customer choice system" (Nutek 2008:01, 5), and the reports refer generally to the gains of "competition" and write about the consumer's right to "switch suppliers" if dissatisfied. The reports thereby symbolically and problematically equate the position of a patient with the position of a customer (Granqvist 2000;Andersson and Kvist 2015).
In UK policies, the investment in consumerist discourse is further explicated with the use of what is called personal or individual "budgets" (e.g. DH 2005DH , 2006DH , 2010aDH , 2010bDH , 2010cFox 2013). Roughly, patients are given customized and personal budgets with which to spend on social or health services of their choice. The word "budget" under-communicates the relation that the patient has to state agencies; it emphasizes the connection to a consumerist discourse, and it is not to be found in the Swedish reports. In the Swedish reports, the practice of payment is more often articulated as an administrative measure, whereby patients' choices govern the compensation to the provider. However, in the Swedish report Framtidens utmaningar (2010, 64), the authors claim: "With rising prosperity the possibilities to pay more from one's own pocket increase." This emphasizes that the relations between patient and private health care providers facilitate consumption of additional services. The report thus associates patient choice with consumer practices that lie outside of the welfare model but still lend it added value.
The articulation of choice with consumption results in patienthood being produced in a way that grants the patient the possibility and the right to consume and make consumerist choices, rather than to grant the patient the right to care. It is thus through the exercise of consumption that the patient is connected to their rights. The articulation of patient choice with consumption unquestionably also detangles patients from health and social care professionalism.

Choice and responsibility
Not surprisingly, and as Nordgren (2010) has pointed out, patient choice is often paired up with notions of personal responsibility. In the reports, patients' attention to themselves and their care is an unmistakeable feature, and patient choice is used as a specific technology, transferring the responsibility from the welfare state or health and social care providers to the individual subject, which is in turn articulated as a simultaneously "responsible" and "competent" patient. Generally, the UK texts are more prone to explicate what increased responsibility means: The challenge is also for each of us as individuals to take responsibility for our own lifestyles and to aim for a healthy and fulfilling old age. (DH 2006, 17) Social care is not solely the responsibility of the state. Communities and wider civil society must be set free to run innovative local schemes and build local networks of support. (DH 2010b, 4) In the first quotation, the reader is taught that although it is a challenge, the taking of individual responsibility is needed. In the second quotation, it is emphasized how social care is not the sole responsibility of the state. Through choice reforms, other actors are pointed out to possibly be "set free" to build supportive networks. In this sense, choice is further articulated as the tool that will set not only patients free from the shackles of the state, but that will also open up for a creativity that is assumed to exist in the local communities, but that has up till now been restricted. 6 Also, the otherwise critical ResPublica report (Fox 2013) agrees with the significance of responsibility. Here, however, responsibility is less articulated with taking responsibility for one's choices as a patient, and more with maintaining relationships and continuing to be a caring citizen.
Interestingly, "responsibility" is much more ideologically explicit in the UK reports than in the Swedish reports, in which responsibility is much more often articulated as "possibility" than as "demand". However, this does not mean that the Swedish reports are consistently gentler in their neoliberal approach to responsibility. The Swedish reports sometimes go so far as to write not only about the responsibilities to choose between different care providers, but also about the responsibility not to become ill in the first place (e.g. SOU 2008:15, 194).

Choice and support from state agencies
The policy texts all deal with the supposed consequences, and reactions, that the practice of choice implies. One such reaction is the supposed critique and concern that everyone does not have the possibility (or wish) to make informed choicesthat the ideal of rational consumer behaviour does not always accord with "the reality of people's everyday lives" (Henderson and Petersen 2002, 3), specifically regarding older people (Lymbery 2010;Woolham and Benton 2013).
The Swedish policy texts put much effort on fleshing out the new relations and are specifically anxious to assure the readers how the authorities have just as an important role to play as the choosing individual: The authority is responsible for facilitating and explaining to the individual what the choice entails and which providers are available (SOU 2008:15, 19).
In stressing the information part of choice and submerging it in a consumerist model of care, patient choice is in part re-articulated. With patients being increasingly re-articulated as consumers (Henderson and Petersen 2002;Lai 1994), it could be said that the state has taken on the role as a protector of consumer rights rather than its previous role as a protector of welfare rights. In the Swedish policy texts, the anxiousness to stress the articulation of "choice" with "support from authorities" seems to rather have to do with an expected critique. There is a well-known confidence and trust in the Swedish welfare state (Svallfors 2016) that has to be handled. 7 Articulating the opposites of choice The opposite of choice is implicitly equalled with a care that is insensitive to citizens' individual needs. This comes forth most clearly in the UK policy texts where the "autonomy" of the choosing patient is repeatedly referred to when describing the benefits of patient choice. The Swedish reports touch upon the matter by noting that patient choice would improve the possibilities to customize health and social care, thus describing the "old" system as less efficient in this matter, but not, as in the UK texts, something that the patient has to gain autonomyor even be "freed"from.
But the potential criticism that the suggested practice of choice is thought to provoke is also successfully countered. This is accomplished through articulations of "choice" with "activity" as opposed to "docility" or "dependence" on medical professionals. This articulation denounces ways that care and health care have been organized historically (Foucault 1973). It also engages with the activity norm that has been specifically successful in late modern old-age identities, for which "successful" Kahn 1987, 1997), "active" (WHO 2002), and "productive" (Butler and Gleason 1985;Holstein 1999) ageing has been the ideal (Calasanti and Slevin 2006).

Ideological fantasies as motive and driving force
In both national contexts, articulations of choice are firmly invested in ideological fantasies that far exceed the context of health and social care. The evocation of fantasies of market economy, demographic processes, and changing cultural identities make the suggested changes towards patient choice come forth as necessary and urgent cost-cutting measures, but also as liberating possibilities invented partly to fulfil the wishes of modern older people. The connection to these fantasies helps all of the suggestions that are put forth in the policy texts to come across as reasonable. What is more, they come across as quite apoliticalas the "normal" and "natural" way to deal with the present situation (cf. Forssell 1999). Because ideological fantasies are seldom confined only to delimited social contexts, but constitute more general narratives of "how it is", the naturalization of patient choice seems comprehensible.
Without denying the seriousness of the situation, the policy texts sometimes seem to be engaged in "a process where a solution (i.e. organizational models) looks for problems to underpin it, rather than as in the traditional rational choice approach the opposite is believed to occur, i.e. problems seek solutions" (Blomberg 2008, 427). The fantasies of market economy, demographic processes, and changing cultural identities are evoked to legitimate a solution that seems, at least in part, to be already decided on.
The UK policy texts are formulated rhetorically in a way that quite explicitly points out responsibility; it is often repeated that "we want" this and that, where the "we" is the writer of the reports. In the Swedish policy texts, the sender, and thus the actors of the suggested changes, are much vaguer, and responsibility is shared between the writer of the texts and patients, who are assumed to have the desire to be able to exercise choice.
At the same time, following the theorisations of ideological fantasy, the main function of the texts is to cover over the contingencies of the social (Žižek 1998;Glynos and Howarth 2007), offering narratives that are possible to hold on to and that are desirable to identify with. In this sense, the ideological fantasies not only constitute the reasons to implement change or the legitimization of the kind of changes that are suggestedthey also produce desired goals and horrifying scenarios that are to follow if patient choice is not carried through. The horrifying scenarios are generally a crumbling welfare system (in both national contexts) or a continued state straitjacket (in the UK). The desired goals comprise, apart from a functioning welfare system and increased individual autonomy, a privileged notion of the ageing patient who lives up to normative third-age characteristics such as activity and independence by practicing choice.

Positioning the ageing patient
The "selling" of models for health and social care also imply the construction of normative as well as unwanted patient positions. The articulations of choice primarily produce a patient position that is heavily invested in neoliberal consumerist discourse. It assumes a patient that is informed and willing to actively take control over their own health through care practices such as information seeking, choosing between a range of both public and private care providers, and deselecting providers with which one is displeased.
Yet, such ideal images of the neoliberal self have proven hard for individuals to uphold (McGuigan 2014). Especially older people seem to regard the prospect of having to choose as unwelcome (Moffatt et al. 2012), patient choice-based care practices have not always worked out for this age group (Lymbery 2010), and the marketization of welfare has not increased the quality of care (Hartman 2011). Possibly being aware of such results, it is precisely the ageing patient that the policy texts argue might sometimes need help and support when carrying out the practice of choice. The ageing patient is thus both used as the reason to implement patient choice and the reason for why patient choice cannot be carried out completely. By emphasizing, specifically in the Swedish official reports, that the state will continue to take responsibilityin providing information and support, but also in quality assurancea second patient position is formed: the patient who does not live up to the standards of being an informed consumer, but rather displays unprivileged patient practices like independence and passivity. 8 This patient position is partly formed by being the denounced, but in the Swedish texts seldom explicated, opposite of the normative patient position. In UK policy texts, the image of the dependent patient is evoked only to be described as what choice initiatives will themselves transform into independence (e.g. DH 2005, 10). Not much is said about this patient position other than that state agencies will help patients to make proper decisions.
While the assumed and privileged patient positions are quite similar in the two national policy contexts, there are also slight differences. In the Swedish texts there is a tendency to identify that people are, but above all are expected to become, autonomous and choosing in the future and that this is an important reason for why the system has to change in the direction of patient choice. In the UK texts, the citizen is described as already autonomous, and that the change towards personalization will free the individual to exercise what they already are (without being hindered by the state). Related to this is the tendency in the UK policies to describe the reduction of state intervention as a good thing in itself, while in the Swedish reports there is rather the tendency to describe this reduction as necessary due to circumstance, or as a way to enhance quality. The carefulness not to portray state intervention as bad might depend on the meanings that are historically ascribed to the large Swedish public sector, where it has been intrinsic to the social democratic notion of freedom as freedom from reliance on the market (Tilton 1991;Blomqvist 2004). It might also reflect how the Swedish welfare model has been a central part of a national identity construction (Lawler 1997;Browning 2007) that Swedish policy makers take into consideration (or are themselves invested in).
A related difference regards the contextualization of the patient. In the UK reports, much more emphasis is put on the community, including voluntary organizations and family, while in the Swedish reports the patient, the state, and the outsourced care providers are the primary agents. The UK policies' confidence in ideological enterprises like "Big Society" or "time banks", which both aim to encourage involvement in voluntary work to "contribute to local capacity building" (DH 2005, 70), is interesting because the data imply that the third sector is "not well prepared for a shift in the funding environment" and cast doubt on the willingness, as well as the capability and capacity, of third sector organizations to engage with such agendas (Chapman, Brown, and Crow 2008, 13).
One further factor conditioning the patient subject that is seldom recognized in the policy texts is geography. Studies on the effects of the choice reform have noted that proximity to urban centres conditions the possibilities to choose (Haynes, Lovett, and Sünnenberg 2003;Stolt and Winblad 2009;Swedish Competition Authority 2012). Because the private care providers that are thought to bring about a variety of choices are reliant on the customer base, people living in sparsely populated areas have fewer opportunities to take on this new position of the choosing patient. There is thus not only a neoliberal normativity, but also an urban normativity present in the discourse of choice, privileging urban residents as its idealized patients. 9

Ensuring ideological perseverance
The selling of choice, and of the positions of autonomy and control that are articulated with it, carries with it a transfer of responsibility from authorities and professionals to the patient, which in practice contributes to making criticism of the system difficult. There is the obvious risk that deficiencies and failures in the choice model are reinterpreted as problems in the relation between the patient and the chosen care provider. Such problems are then solved by the patient exercising their right to dismiss the provider and choose someone else. In cases where the patient for different reasons cannot choose, it is emphasized that support will be provided from state agencies. This support might be viewed as a "countermeasure" taken by the state in order to minimize the unfair or even destructive effects of liberalizing measures such as choice reforms (cf. Polanyi 1944Polanyi /2001Streeck and Thelen 2005, 4). It can also be seen as a way to pre-empt potential contestation and criticism (Glynos, Speed, and West 2015). Both interpretations point at the offering of support being central to establishing patient choice as safe and sound, and to reinforce patient choice as the logical way forward. Both interpretations also safeguard the enjoyment of the fantasy of patient choice and the promises that it makes. The latter aspect may explain why so much effort is put on retaining the fantasy and may serve as an important answer to why it proves to be so persistent (cf. Glynos 2008).
The strong fantasy of changing cultural identitiesof which the ageing baby boomer is an important exampleworks in a similar way. Specifically in the Swedish policy texts, the incorporation of the notion of the boomer, who is supposedly demanding the opportunity of choice, symbolically lays the claim that policymakers are only being responsive to patient demands. In the UK material, this "demanding patient" is supposed to already exist and does not have to be envisioned as a figure of the near future. In both national contexts, the policy texts come across as speaking for the public will, thus not only having a populist appeal, but also partly placing responsibility on the patients. In a sense, by repeatedly articulating patient choice with older people's demands, the reports offer what West, following Boltanski, calls "ritualized confirmation" (West 2013, 647); they re-constitute the symbolic frame through which they want patient choice to be interpreted. Criticizing patient choice would thus mean criticizing not only officials, but also a generalized mass of "other patients". Because these "other patients" are ascribed normative characteristics such as being active and independent, the one criticizing patient choice risks coming across as a less active, less independent, and less competent older patient.
The studied policy texts thus entail the performance of an operation of ideological perseverance. Fotaki (2010a) has pointed out how policy failures are often rectified by the introduction of smaller changes and how the relentless work to improve policy might in fact contribute to the survival of ideology (see also Gunder 2016). The countermeasures of offering patients support are an apt example of this. Another example is when in Equity and Excellence (DH 2010a, 4) it is stated that the policy writers are aware of the challenges to come: "But far from that being reason to abandon reform, it demands that we accelerate it." In the case studied here, such improvements are also already sewn into the policy fabric as a way to anticipate eventual criticism. It is "sewn into" it by engendering the normative, intertwined, and mutually reinforcing positions of the consumer rights discourse, the market economy discourse, and the discourse of the strong-headed baby boomers.

Concluding remarks
In analyzing how patient choice is articulated in Swedish and UK policy documents, we see a wide range of similarities. In order to present patient choice as a desirable care practice, it is articulated with neoliberal concepts such as individuality, autonomy, consumption, and responsibility, but also with support from state agencies.
However, we also detect some differences. We argue that these differences partly have to do with the way that the relation between the state and the citizen is supposedly comprehended in the two national contexts. While UK policies emphasize how patient choice will free the citizen from being restricted by the state, Swedish policies do not presuppose that the citizen is feeling restricted, and they are much more prone to emphasize the possibilities that increased patient choice would entail and are also much more careful to reassure the reader that the patient will not stand alone in this but that the state will still offer support. Further, much more explanatory weight is laid on the ageing patient in the Swedish policies; although present in the UK policy texts, the Swedish policies write more explicitly about the demands from the ageing cohort of baby boomers, thus partly placing responsibility for policy change on the (future) patients themselves. This simultaneously takes an urban environment for granted, presupposing that ageing patients are in fact living in areas where choice is a viable option.
That it matters how the welfare systems have been chiselled out in the respective countries is visible when pinpointing the stakeholders of patient-choice reforms. In the UK policies, the voluntary sector, the community, and the family are often mentioned. These almost never appear in the Swedish policies. Apart from the patients themselves, the Swedish policy texts only mention either the state or private businesses as healthcare providers. Selling choice in the UK thus evokes notions of a lively community, while the same sale in Sweden focuses on the patient, possibly with support from the state.
We also find two contradictory ageing-patient positions. Older people are generally positioned as competent patients, willing to engage in the care practices that come with patient choice reforms. This positioning is consistent with the norm of active, healthy, and successful ageing that has been dominant in recent decades. This helps to fixate everyday relations between healthcare and idealized ageing patients, making it visible to people what it means to be a patient. However, another ageing position is also present in the reports. This is the older patient who cannot choose and is therefore in need of support in carrying out this privileged care practice. This position is either articulated as one that should be supported and helped or as one that will mysteriously disappear once patient choice reforms are implemented.
Even though arguments as well as the actual proposed systems differ, the policy narratives work to normalize patient choice as the way forward. The histories of the respective welfare systems are different, and citizens are assumed to think about the welfare state in different ways in the two national contexts. By taking this into consideration, and by meeting and pre-empting expected criticism before it can be delivered, the policy texts work in favour of ideological preservation; apart from highlighting the pros of patient choice, they offer services that will mitigate the well-known problems of, and concerns about, patient choice, but they still advocate this neoliberal way of organizing welfare and of positioning patients. Patient choice is thus made culturally comprehensible by being articulated with signs that are central in the respective national contexts. In the UK, it is "community" and the "voluntary sector", and in Sweden it is "state agencies" and "support" that are offered to take the edge off of patient choice ideology.
Notes 1. In the UK, care management was introduced already in the National Health Service and Community Care Act in 1990 (Järkestig Berggren 2006), but we focus here on the introduction of what is generally referred to with words such as "personalization" and "personcentred" (Needham 2011). Its principles, primarily highlighting "choice" and "control" within adult social care, were introduced in 2005 and were later elaborated upon (Lymbery 2012(Lymbery , 2014. According to Fredriksson, Blomqvist, and Winblad (2012), Sweden was also one of the first countries to embrace the principle of patient choice. Swedish social services remain mainly tax-funded, although user fees have become more significant. Private enterprises have come to "deliver services on a regular basis, and public providers have been re-organized so as to compete internally and externally through 'quasimarkets'" (Lind 2014, 9; see also Blomqvist and Rothstein 2000;Blomqvist 2004). 2. For reasons of simplification, we have chosen to use the word "patient", although we are fully aware that not all users of care consider themselves to be patients. 3. The Swedish Association of Local Authorities and Regions (Statens kommuner och landsting, SKL) has as its main objective "to safeguard the interests of Swedish local and regional authorities" (skl.se, 29 Nov 2013). The SKL is a politically run organization representing the governmental, professional, and employer-related interests of Sweden's 290 municipalities and 20 county councils. Its 451 congress delegates are appointed by the politically elected members of municipalities, county councils, and regions. 4. Nutek was an administrative authority responsible for strengthening private sector businesses and for promoting sustainable regional growth. It was phased out in 2009, and the activities conducted by Nutek were transferred to the Swedish Agency for Economic and Regional Growth (Tillväxtverket) and Growth Analysis (Tillväxtanalys). 5. Counter to this stance, research has warned that increased patient choice might in fact increase inequity, and it raises criticism that the consequences of choice reforms have not been sufficiently investigated (cf. Fotaki 2010b; Fredriksson, Blomqvist, and Winblad 2012). 6. Mentions of local communities are sparse in Swedish policy. This is not to say that local communities do not play a part in Swedish social care. Much Swedish research has emphasized that after the 1990s there was an increase in local responsibility for providing elderly care, which led to large variations between municipalities (Lindqvist 1998;Socialstyrelsen 1996). There has also been a development in the direction from a traditional welfare model to a model that utilizes resources such as the family, neighbours, and volunteers as well as private actors (Thorslund and Parker 1995;Blomberg, Edebalk, and Petersson 2000). Importantly, however, this is not explicated within the public policy included in this study and must thus be considered a safety net evoked to compensate for flaws in the welfare system since the 1990s. 7. Interestingly, most studies of trust in the welfare system focus on public institutions and not on private alternatives (Lind 2014). 8. Nordgren (2010) states that the Swedish discourse on patient choice creates weak rather than strong and empowered patients. Among other things, he builds this assertion on the fact that "health policies are being restricted by a lack of information and support and institutional regulations". Regardless of whether the demand for information is satisfyingly carried through in reality, it is obvious that the need for information and support is present in the policy texts. 9. Swedish Official Report 2008:15 (SOU 2008:15, 230) recognizes the risk that sparsely populated areas might not have big enough customer bases to provide service users a real choice. In Sweden, the Act on System of Choice in the Public Sector (SFS 2008:962) was voluntary for the municipalities, which meant that many municipalities in the north of Sweden with strong social democratic support chose not to implement it (Swedish Competition Authority 2012).

Disclosure statement
No potential conflict of interest was reported by the author.

Funding
This work was supported by Forte: Swedish Research Council for Health, Working Life and Welfare.

Notes on contributors
Jens Lindberg holds a PhD in Ethnology and is currently a researcher at the Department of Social Work, Umeå University, and affiliated with the Centre of Demography and Ageing Research (CEDAR). His research areas include the organization and logics of care, gender and masculinity studies, rape care and discourse theory.